3243 J&J Vulnerability
1/11/07
10:22 am
Page i
Climate Change and Vulnerability
3243 J&J Vulnerability
1/11/07
10:22 am
Page ii
3243 J&J Vulnerability
1/11/07
10:22 am
Page iii
Climate Change and Vulnerability Edited by Neil Leary, Cecilia Conde, Jyoti Kulkarni, Anthony Nyong and Juan Pulhin
London • Sterling, VA
3243 J&J Vulnerability
1/11/07
10:22 am
Page iv
First published by Earthscan in the UK and USA in 2008 Copyright © The International START Secretariat, 2008 All rights reserved Climate Change and Adaptation: ISBN 978-1-84407-470-9 Climate Change and Vulnerability: ISBN 978-1-84407-469-3 Two-volume set: ISBN 978-1-84407-480-8 Typeset by FiSH Books, Enfield Printed and bound in the UK by Antony Rowe, Chippenham Cover design by Susanne Harris For a full list of publications please contact: Earthscan 8–12 Camden High Street London, NW1 0JH, UK Tel: +44 (0)20 7387 8558 Fax: +44 (0)20 7387 8998 Email:
[email protected] Web: www.earthscan.co.uk 22883 Quicksilver Drive, Sterling, VA 20166-2012, USA Earthscan publishes in association with the International Institute for Environment and Development A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data Climate change and vulnerability / edited by Neil Leary ... [et al.]. p. cm. ISBN-13: 978-1-84407-469-3 (hardback) ISBN-10: 1-84407-469-2 (hardback) 1. Climatic changes. 2. Climatic changes—Developing countries. I. Leary, Neil. QC981.8.C5C5625 2007 304.2’5—dc22 2007034503
The paper used for this book is FSC-certified and totally chlorine-free. FSC (the Forest Stewardship Council) is an international network to promote responsible management of the world’s forests.
3243 J&J Vulnerability
1/11/07
10:22 am
Page v
Contents List of figures and tables
vii
Acknowledgements
xv
Foreword by R. K. Pachauri
xvii
Part I: Introduction
1
1
3
For whom the bell tolls: Vulnerabilities in a changing climate Neil Leary, James Adejuwon, Wilma Bailey, Vicente Barros, P. Batima, Rubén M. Caffera, Suppakorn Chinvanno, Cecilia Conde, Alain De Comarmond, Alex De Sherbinin, Tom Downing, Hallie Eakin, Anthony Nyong, Maggie Opondo, Balgis Osman-Elasha, Rolph Payet, Florencia Pulhin, Juan Pulhin, Janaka Ratnisiri, El-Amin Sanjak, Graham von Maltitz, Mónica Wehbe, Yongyuan Yin and Gina Ziervogel
Part II: Natural Resource Systems
31
2
Vulnerability of southern African biodiversity to climate change Graham P. von Maltitz and Robert J. Scholes
33
3
Forest responses to changing rainfall in the Philippines Rodel Lasco, Florencia Pulhin, Rex Victor O. Cruz, Juan Pulhin, Sheila Roy and Patricia Sanchez
49
4
Vulnerability of Mongolia’s pastoralists to climate extremes and changes Punsalmaa Batima, Luvsan Natsagdorj and Nyamsurengyn Batnasan
67
5
Resource system vulnerability to climate stresses in the Heihe river basin of western China Yongyuan Yin, Nicholas Clinton, Bin Luo and Liangchung Song
88
Part III: Coastal Areas 6
115
Storm surges, rising seas and flood risks in metropolitan Buenos Aires 117 Vicente Barros, Angel Menéndez, Claudia Natenzon, Roberto Kokot, Jorge Codignotto, Mariano Re, Pablo Bronstein, Inés Camilloni, Sebastián Ludueña, Diego Riós and Silvia González
3243 J&J Vulnerability
1/11/07
10:22 am
Page vi
vi Climate Change and Vulnerability
7
Climate and water quality in the estuarine and coastal fisheries 134 of the Río de la Plata Gustavo J. Nagy, Mario Bidegain, Rubén M. Caffera, Frederico Blixen, Graciela Ferrari, Juan J. Lagomarsino, Cesar H. López, Walter Norbis, Alvaro Ponce, Maria C. Presentado, Valentina Pshennikov, Karina Sans and Gustavo Sención
8
Climate change and the tourism dependent economy of the Seychelles Rolph Antoine Payet
Part IV: Rural Economy and Food Security
155
171
9
Household food security and climate change: Comparisons from 173 Nigeria, Sudan, South Africa and Mexico Gina Ziervogel, Anthony Nyong, Balgis Osman-Elasha, Cecilia Conde, Sergio Cortés and Tom Downing
10
Vulnerability in Nigeria: A national-level assessment James D. Adejuwon
11
Vulnerability in the Sahelian zone of northern Nigeria: 218 A household-level assessment Anthony Nyong, Daniel Dabi, Adebowale Adepetu, Abou Berthe and Vincent Ihemegbulem
12
Livelihoods and drought in Sudan Balgis Osman-Elasha and El-Amin Sanjak
13
Social vulnerability of farmers in Mexico and Argentina 257 Hallie Eakin, Mónica Wehbe, Cristian Ávila, Gerardo Sánchez Torres and Luis A. Bojórquez-Tapia
14
Climatic threat spaces in Mexico and Argentina Cecilia Conde, Marta Vinocur, Carlos Gay, Roberto Seiler and Francisco Estrada
279
15
Climate variability and extremes in the Pantabangan–Carranglan watershed of the Philippines: An assessment of vulnerability Juan M. Pulhin, Rose Jane J. Peras, Rex Victor O. Cruz, Rodel D. Lasco, Florencia B. Pulhin and Maricel A. Tapia
307
16
Climate risks and rice farming in the lower Mekong river basin 333 Suppakorn Chinvanno, Somkhith Boulidam, Thavone Inthavong, Soulideth Souvannalath, Boontium Lersupavithnapa, Vichien Kerdsuk and Nguyen Thi Hien Thuan
198
239
3243 J&J Vulnerability
1/11/07
10:22 am
Page vii
Contents vii
17
Vulnerability of Sri Lankan tea plantations to climate change Janaka Ratnasiri, Aruliah Anandacoomaraswamy, Madawala Wijeratne, Senaka Basnayake, Asoka Jayakody and Lalith Amarathunga
Part V: Human Health
351
373
18
Vulnerability to climate-induced highland malaria in East Africa 375 Shem O. Wandiga, Maggie Opondo, Daniel Olago, Andrew Githeko, Faith Githui, Michael Marshall, Tim Downs, Alfred Opere, Pius Z. Yanda, Richard Kangalawe, Robert Kabumbuli, Edward Kirumira, James Kathuri, Eugene Apindi, Lydia Olaka, Laban Ogallo, Paul Mugambi, Rehema Sigalla, Robinah Nanyunja, Timothy Baguma and Pius Achola
19
Vulnerability to dengue fever in Jamaica 398 Charmaine Heslop-Thomas, Wilma Bailey, Dharmaratne Amarakoon, Anthony Chen, Samuel Rawlins, Dave D. Chadee, Rainaldo Crosbourne, Albert Owino, Karen Polson, Cassandra Rhoden, Roxanne Stennett and Michael Taylor
Index
417
3243 J&J Vulnerability
1/11/07
10:22 am
Page viii
List of Figures and Tables
Figures 2.1 3.1 3.2 3.3 3.4 3.5 3.6 4.1 4.2 4.3 4.4 4.5 4.6 5.1 5.2 5.3
5.4 5.5 5.6 5.7
Current and future predicted distributions of the major South African biomes The Holdridge system of vegetative cover classification Potential Holdridge life zones in the Philippine forests without human influence Projected change in area of existing life zones in the Philippines under various climate change scenarios (25%, 50% and 100% increase in precipitation) Holdridge life zones in the Philippine forests under Scenario 1 (25% increase in rainfall) and with a 2°C temperature increase Holdridge life zones in the Philippine forests under Scenario 2 (50% increase in rainfall) and with a 2°C temperature increase Holdridge life zones in the Philippine forests under Scenario 3 (100% increase in rainfall) and with a 2°C temperature increase Summer drought index, 1940–2002 Winter severity index, 1940–2002 Relationship between drought indexes and summer/autumn live-weights of sheep and goats Winter severity and livestock mortality Areas vulnerable to black and white zud Areas vulnerable to drought, black zud and white zud Flow-chart of the general research approach Map of the Heihe river basin with approximate population distribution shown in shades of grey (black is highest population density) Growing season PDSI for lower reach of the Heihe river basin, 1961–2000 (a&b) Growing season PDSI for upper part of the middle reach of the Heihe river basin, 1961–2000 (c&d) Trend of water-use conflicts (number of violent events in competition for water) in the study basin Areas with high irrigation water demand (negative units are in millimetres of deficit) Per capita water resources (in cubic meters per capita per annum) Geographic distribution of vulnerability to adverse weather conditions in the Heihe river basin
34 52 55 58 59 60 70 71 75 75 78 79 92 93 102 103 104 105 106 109
3243 J&J Vulnerability
1/11/07
10:22 am
Page ix
List of Figures and Tables ix
5.8 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 7.1
7.2 7.3 7.4 7.5 8.1 8.2 8.3 9.1
9.2 9.3 10.1 10.2 10.3 10.4 10.5 11.1 13.1
Histogram of composite water-use vulnerability levels in the 110 Heihe river basin The Plata river estuary 118 Frequency of levels (m) above mean sea level at the Buenos Aires 121 port, 1990–1999 Mean Plata river level (m) calculated by the HIDROBID II at 121 the Buenos Aires port, 1990–1999 Maximum heights calculated for the storm surge tide with return 122 period of 100 years in the Buenos Aires port Damages in real estate per event in millions of US dollars as a 123 function of the water-level rise over the mean present level Flood return periods in years for present time 124 Changes in the return period between 2070–2080 and present 125 time in years Social risk index: Present conditions 127 Variation of the social risk map between 2070 and present 128 Río de la Plata estuarine front location under different ENSO 135 conditions: a) strong La Niña event (1999–2000), b) typical, c) moderate El Niño (winter 1987), d) strong El Niño 1997–1998/2002–2003 Climate baseline scenarios for the Río de la Plata for the period 137 1961–1990 River Uruguay discharge at Salto from 1921 to 2003 139 DPSIR framework of trophic state and symptoms of 141 eutrophication for the Río de la Plata estuary Long-term gross income of fishermen (local currency-1999) 144 Share of economic losses by sector from the 1997–1998 El Niño 156 and 1998–2000 La Niña events in the Seychelles Number of tourist nights per year in the Seychelles – Actual 158 number of tourist nights and predictions from the Vision 2 master plan and a statistical model are shown Primary reasons tourists visit the Seychelles 162 Assessment of sustainability of financial capital before and after 183 intervention of the Rangeland Rehabilitation Project, based on availability of information, effectiveness of credit repayment, suitability of local institutions, and support of credit systems and government policy to income-generating activities Determinants of vulnerability to food insecurity in study villages 190 Causal chain of drought risk 191 The ecological zones of Nigeria 199 Mean monthly maximum temperature projections, 1961–2099 204 Mean monthly precipitation projections, 1961–2099 204 Maize yields for projected climate change, 1961–2099 206 Vulnerability of peasant households to climate change by state 215 Bhalme and Mooley drought intensity (BMDI) index for selected 221 stations in northern Nigeria, 1930–1983 Vulnerability of farm groups in Laboulaye and the wider 265
3243 J&J Vulnerability
1/11/07
10:22 am
Page x
x Climate Change and Vulnerability
13.2 13.3 13.4 13.5 14.1 14.2 14.3 14.4 14.5 14.6 14.7 14.8 14.9 14.10 14.11 14.12 14.13 14.14 14.15 14.16 15.1 15.2 15.3
south-centre of Córdoba Province Structure of vulnerability for different farmer groups in 266 Laboulaye, Argentina Sources of income for communal and private tenure farmers in 270 González, Mexico Proportion of communal and private tenure farmers associated 271 with each vulnerability class in González, Mexico Structure of vulnerability for low, medium and high vulnerability 273 classes among communal farmers (top) and private tenure farmers (bottom) in González, Mexico The state of Veracruz and the region under study 283 Normal climatic conditions for Teocelo 284 Threat space for Atzalan, Veracruz, in spring (MAM) 285 Coffee production anomalies for Mexico and for the state of 286 Veracruz Climatic threat space for coffee during spring (MAM), considering the minimum temperature and precipitation requirements of the coffee plant 287 Threat space for Atzalan, Veracruz, in summer (JJA) 288 Study region: Location of the city and flood-prone area 290 Normal climatic conditions for Laboulaye, Argentina, in terms 291 of maximum temperatures (Tmax), minimum temperatures (Tmin) and precipitation (Pcp) Climatic threat space for Laboulaye, Córdoba, for summer season 292 (DJF) Maize yield deviations from linear trend of the series 1961–1999 293 for the Department of Roque Sáenz Peña (Córdoba, Argentina) Projected changes in a) temperature and b) precipitation for the 295 central region of Veracruz (2020 and 2050) Climatic threat space for coffee production in the central region 296 of Veracruz, considering climate change scenarios Frequency maximum temperatures (Teocelo, Veracruz, 297 1961–1998) Current mean and variability conditions and climate change 298 scenarios for 2020 Projected changes in a) temperature and b) precipitation for 300 January in the southern region of Cordoba, Argentina (2020 and 2050) Observed (1960–2002) and projected DJF maximum 301 temperature (2020) based on ECHAM4 under the A2 emissions scenarios for Laboulaye, Argentina Location of the Pantabangan–Carranglan watershed on Luzon 310 Island in the Philippines Land uses in the Pantabangan–Carranglan watershed, 1999 311 The location of vulnerable areas in the Pantabangan–Carranglan 322 watershed as identified by analysis of biophysical data and by participatory mapping with community members
3243 J&J Vulnerability
1/11/07
10:22 am
Page xi
List of Figures and Tables xi
16.1 16.2
Framework for climate risk and vulnerability assessment 334 Average temperature in the lower Mekong river basin: Baseline 338 and projected changes 16.3 Average rainfall in the lower Mekong river basin: Baseline and 339 projected changes 16.4 Climate risk levels of farm households under current climate 344 conditions 16.5 Climate risk profiles under current climate conditions 345 16.6 Changes in climate risk scores in response to climate change and 347 extremes 17.1 Baseline temperature distribution 353 17.2 Baseline rainfall distribution (mean of 1961–1990) 354 17.3 Increased mean temperature during mid-2100 corresponding to 355 HadCM3/A1FI scenario 17.4 Increased precipitation during mid-2100 corresponding to 356 HadCM3/A1FI scenario 17.5 Agro-ecological regions where tea is cultivated 357 17.6 Variation of monthly yield with ambient temperature with 360 optimum value at 22°C 17.7 Variation of yield with rainfall up to optimum rainfall for the 361 intermediate up-country region 17.8 Variation of annual yield in different agro-ecological regions in 362 the drought year (1992) and the previous year 17.9 Basic flow chart of the crop simulation model 363 17.10 Variation of tea yield in wet zone mid-country estimated by the 364 tea crop model with change of climatic conditions 17.11 Percentage yield change expected with respect to baseline values 366 for 2025, 2050 and 2100 and emission scenarios A1FI, A2 and B1 under the three GCMs in low-, mid- and up-country regions 18.1 Geographical information system maps of the three study sites in 379 Uganda, Kenya and Tanzania 18.2 Trends in malaria hospital admissions in Kenya, Tanzania and 383 Uganda 18.3 Trends in malaria in children <5 years in Muleba, Tanzania 384 18.4 Modelled climate and malaria data for Litein, Kenya 392 19.1 Time series graph of reported cases of dengue with rainfall and 400 temperature for Jamaica 19.2 Variation of annual reported cases and rate of change 401 19.3 Health regions in Jamaica 405 19.4 Health centres in Jamaica 406 19.5 Distribution of dengue cases in St James, Jamaica, 1998 407
Tables 1.1 1.2
Water resource vulnerabilities Land vulnerabilities
7 10
3243 J&J Vulnerability
1/11/07
10:22 am
Page xii
xii Climate Change and Vulnerability
1.3 1.4 1.5 1.6 3.1 3.2 3.3 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 5.1 5.2 6.1 6.2 7.1 7.2 7.3 7.4 8.1 8.2 8.3 9.1 9.2
Ecosystems and biodiversity vulnerabilities 12 Coastal area and small island vulnerabilities 14 Rural economy vulnerabilities 18 Human health vulnerabilities 26 Synthetic climate change scenarios used in the study 53 Comparison of potential and actual (1993) life zones in the 56 Philippines Adaptation options to climate variability and extremes for forest 64 lands in the Pantabangan–Carranglan watershed, Philippines Standing peak biomass by ecosystem type, 1960–2000 72 First- and second-order impacts of drought 74 Zud forms and their description 76 Frequency of extremes and levels of vulnerability 78 Projections of temperature and precipitation changes in 82 Mongolia relative to 1961–1990 baseline for SRES A2 and B2 emission scenarios Estimated changes in above-ground biomass and carbon:nitrogen 83 ratio for scenarios derived from the HadCM3 climate model Estimated changes in above-ground biomass and carbon:nitrogen 84 ratio for scenarios derived from the ECHAM4 climate model Estimated changes in above-ground biomass and carbon:nitrogen 84 ratio for scenarios derived from the CSIRO-Mk2b climate model Projected changes in ewe weight (%) with combined effects of 85 changes in pasture productivity, forage quality and grazing time for HadCM3 climate change scenarios Potential determinants (climate and other variables with forcing) 96 and resource indicators Water withdrawal ratio in the Heihe river basin, 1991–2000 100 Water heights (above mean sea level) at the Buenos Aires port 119 for return periods Present population living in areas that are, or will be, flooded 125 under different scenarios Aggregated impact matrix of HABs in the Uruguayan coastal 142 zone of the Río de la Plata for the period 1991–2000 Aggregated impact index of HABs for the Río de la Plata, 142 1991–2000 Weighted aggregated impact index of HABs (index of 143 persistence/extension/toxicity, IPET) for each species and sector on the northern coast of the Río de la Plata, 1991–2000 Assessment of the vulnerability of the coastal fishing community 145 Intense rainfall events of above 200mm in a 24-hour period over 157 the island of Mahe, Seychelles, 1972–1997 Results from estimation of the model of tourist demand 163 Social and economic impacts from increases in average number 165 of wet days and temperatures Summary of the importance of determinants 189 Determinants of vulnerability situated in a causal chain of 192
3243 J&J Vulnerability
1/11/07
10:22 am
Page xiii
List of Figures and Tables xiii
10.1 10.2 10.3 10.4 11.1 11.2 11.3 11.4 12.1 12.2 13.1 13.2 13.3 13.4 13.5 13.6 15.1 15.2 15.3 15.4 15.5 15.6 16.1 16.2 16.3 16.4 17.1 17.2 17.3
drought risk Mean monthly minimum and maximum temperatures by zone in Nigeria, 1971–2000 Variability of monthly maximum temperature, 1971–2000 Variability of monthly minimum temperature, 1971–2000 Variability of monthly rainfall, 1971–2000 Self-perceived risks Indicators and weights for vulnerability assessment in northern Nigeria Distribution of households by vulnerability class Ordinal logit results for determinants of drought vulnerability Word picture of households’ access to/use of livelihood capitals The main impacts of drought identified by the communities across the three case studies Main production systems of Laboulaye, Córdoba Province, Argentina Indicators of the adaptive capacity of farm households in Laboulaye, Argentina, by farm type Sensitivity of farm households in Laboulaye, Argentina, to climate impacts on crops by farm type and climate event Vulnerability of different farm production systems within the Laboulaye area Selected indicators of adaptive capacity of farm households in González, Mexico Selected indicators of sensitivity of farm households in González, Mexico, to climate hazards Multi-level indicator of vulnerability of households to climate variability and extremes using varying weights Major climate events identified by participants in key informant interviews and focus group discussions Vulnerability index values for farmers, non-farmers and the full sample, based on weights provided by researchers and local communities Groups vulnerable to climate variability and extremes Results of Spearman correlation analysis of vulnerability indices Results of stepwise regression analysis Villages by zone in Thailand Simulated rice yields under different climate scenarios Indicators used in evaluating farmers’ risk from climate impact Scenarios of changes in rice yields in response to changes in average climate and extreme climate Projected change in mean temperature under different emission scenarios at mid-2100 Projected change in rainfall under different emission scenarios for June, July and August 2100 Agro-ecological regions where tea is grown and their characteristics
200 201 202 203 223 225 227 230 243 248 261 262 264 264 269 269 316 320 323 324 327 329 336 340 343 346 355 356 357
3243 J&J Vulnerability
1/11/07
10:22 am
Page xiv
xiv Climate Change and Vulnerability
17.4
Climate and yield data at four sites representing four 359 agro-ecological regions 17.5 Annual (baseline) production of made-tea during 1996–2005 359 17.6 Effect of rainfall on productivity of tea lands 361 17.7 Projected yield changes under different climate change scenarios 365 17.8 National projected production changes under different models 367 and climate change scenarios 17.9 Matrix of indices of impact, adaptive capacity and vulnerability 369 vs. ownership 17.10 Vulnerability of tea plantations at regional/national level 369 18.1 Geographical positions of streamflow gauging stations, Kericho 379 area 18.2 The long-term context of temperature changes in the Lake 380 Victoria basin, showing results for highland sites based on linear regression 18.3 Ranked Tmax and Tmin with high Tmax and low Tmin for the period 380 1978 to 1999, compared with occurrence of El Niño and La Niña years 18.4 Ranked mean monthly cumulative precipitation with wet years 381 and dry years for the period 1978 to 1999, compared with occurrence of El Niño and La Niña years 18.5 Selected indicators of vulnerability to malaria epidemics 387 18.6 Type of health facility visited 388 18.7 Visits to hospitals in the last three months by household 388 members 19.1 Distribution of epidemic peaks among ENSO phases, 1980–2001 401 19.2 Sample size 407 19.3 Composite of ranking for communities in St James 410 19.4 Identification of vulnerable groups 411 19.5 Characteristics of the most and least vulnerable groups 411 19.6 Characteristics of Groups 4 and 2 412
3243 J&J Vulnerability
1/11/07
10:22 am
Page xv
Acknowledgements The two volumes Climate Change and Vulnerability and Climate Change and Adaptation are products of Assessments of Impacts and Adaptations to Climate Change (AIACC), a project that benefited from the support and participation of numerous persons and organizations. AIACC was funded by generous grants from the Global Environment Facility, the Canadian International Development Agency, the US Agency for International Development, the US Environmental Protection Agency and the Rockefeller Foundation. The initial concept for the project came from authors of the Third Assessment Report of the Intergovernmental Panel on Climate Change (IPCC) and was championed by Robert Watson, Osvaldo Canziani and James McCarthy, the IPCC chair and IPCC Working Group II co-chairs, respectively, during the Third Assessment Report. The productive relationship between AIACC and IPCC was continued and nurtured by Rajendra Pachauri, the current Chair of the IPCC, and Martin Parry, who joined Dr Canziani as co-chair of IPCC Working Group II for the Fourth Assessment Report. The project could not have succeeded without the very capable and dedicated work of the more than 250 investigators who undertook the AIACC case studies, many of whom are authors of chapters of the two books. The project also benefited from the valuable and enthusiastic assistance of the many committee members, advisers, resource persons and reviewers. These include Neil Adger, Ko Barrett, Bonizella Biagini, Ian Burton, Max Campos, Paul Desanker, Alex De Sherbinin, Tom Downing, Kris Ebi, Roland Fuchs, Habiba Gitay, Hideo Harasawa, Mohamed Hassan, Bruce Hewitson, Mike Hulme, Saleemul Huq, Jill Jaeger, Roger Jones, Richard Klein, Mahendra Kumar, Murari Lal, Liza Leclerc, Bo Lim, Xianfu Lu, Jose Marengo, Linda Mearns, Monirul Mirza, Isabelle Niang-Diop, Carlos Nobre, Jean Palutikof, Annand Patwardhan, Martha Perdomo, Roger Pulwarty, Avis Robinson, Cynthia Rosenzweig, Robert Scholes, Ravi Sharma, Hassan Virji, Penny Whetton, Tom Wilbanks and Gary Yohe. Patricia Presiren of the Academy of Sciences of the Developing World (TWAS) and Sara Beresford, Laisha Said-Moshiro, Jyoti Kulkarni and Kathy Landauer of START gave excellent support for the administration and execution of the project. Finally, thanks are owed to Alison Kuznets and Hamish Ironside of Earthscan and to Leona Kanaskie for assistance with copy-editing and production of the books.
3243 J&J Vulnerability
1/11/07
10:22 am
Page xvi
3243 J&J Vulnerability
1/11/07
10:22 am
Page xvii
Foreword Climate change is increasingly recognized as a critical challenge to ecological health, human well-being and future development, as underscored by the award of the Nobel Peace Prize for 2007 to the Intergovernmental Panel on Climate Change (IPCC). The award recognizes the substantial advances in our shared understanding of climate change, its causes, its consequences and its remedies, which have been achieved by more than 20 years of work by the thousands of contributors to the IPCC science assessments, and which draw from the research and analyses of an even larger number of scientists and experts. This work has culminated in the unprecedented impact of the Panel’s most recent report, the Fourth Assessment Report. The Fourth Assessment Report advances our understanding on various aspects of climate change based on new scientific evidence and research. A major contribution in this regard has come from the work promoted under the project Assessments of Impacts and Adaptation to Climate Change (AIACC). The AIACC project was sponsored by the IPCC to fill a major gap in the available knowledge about climate change risks and response options in developing countries that existed at the completion of the Third Assessment Report in 2001. Twenty-four national and regional assessments were executed under the AIACC project in Africa, Asia, Latin America and small island states in the Caribbean, Indian and Pacific Oceans. The two volumes Climate Change and Vulnerability and Climate Change and Adaptation present many of the findings from the AIACC assessments. The findings not only give us a fuller scientific understanding of the specific nature of impacts and viable adaptation strategies in different locations and countries, but have contributed to a much better appreciation of some of the equity dimensions of the problems as well. In simplified terms, the biggest challenge in confronting the negative impacts of climate change lies in the developing world, where people and systems are most vulnerable. Not only are these negative impacts likely to be most serious in the subtropics and tropics, where most developing societies reside, but the capacity to adapt to them is also limited in these regions. An important element in understanding vulnerabilities to climate change is in linking current and projected exposures to climate stresses with other existing stresses and conditions that are responsible for hardship and low levels of economic welfare. Climate change often adds to these existing stresses, increasing the vulnerability of such communities and ecosystems. Unfortunately, limited research is carried out in developing countries on likely impacts and appropriate responses related to climate change. This is where the knowledge provided by the assessments of the AIACC has been particularly valuable. There is considerable interest in interpretation of Article 2 of the United
3243 J&J Vulnerability
1/11/07
10:22 am
Page xviii
xviii Climate Change and Vulnerability
Nations Framework Convention on Climate Change (UNFCCC), the focus of which defines the ultimate objective of the Convention, namely that of preventing a dangerous level of anthropogenic interference with the climate system. Research on impacts of climate change focusing on specific parts of the world that are highly vulnerable enhances our understanding of what may constitute a dangerous level of anthropogenic interference with the world’s climate system. In the absence of such knowledge, any value judgement defining a dangerous level would apply to, and be determined by, knowledge only from particular regions of the world, primarily the developed nations. Understanding the critical nature of impacts in some of the most vulnerable parts of the world, which are largely in developing countries, will assist our determination of what might constitute a dangerous level of interference with the earth’s climate system. Such knowledge would help appropriately to include and consider those locations which are perhaps much closer to danger than was known earlier. The record and outputs of the AIACC are impressive. The project, funded by the Global Environment Facility and coordinated by the Global Change System for Analysis, Research and Training (START), the Academy of Sciences for the Developing World (TWAS) and the United Nations Environment Programme (UNEP), engaged investigators from more than 150 institutions and 60 countries to execute the assessments. The quality of the assessments is demonstrated by the more than 100 peer-reviewed publications produced, which benefited substantially the IPCC’s Fourth Assessment Report. In view of this success, it is imperative that we build on the experience and achievements of AIACC and develop the next phase of such work to help advance new knowledge for a possible Fifth Assessment Report of the IPCC. While the material contained in the two volumes from AIACC and the substantial amount of knowledge developed through the case studies presented in the following pages are valuable, the need for further work is enormous. There remain many countries in the developing world where very little is known about the nature and extent of the impacts of climate change, and these gaps would not permit the development of plans and programmes to address climate change risks or to put in place response measures that would help communities and ecosystems to adapt to the impacts of climate change. These clearly would get much more serious with time unless suitable mitigation measures are taken in hand with a sense of urgency. Yet, even with the most ambitious mitigation actions, the inertia of the system will ensure that the impacts of climate change will continue for centuries, if not beyond a millennium. Knowledge of impacts and the manner in which they would grow over time is therefore critical to the development of capacity and measures for adaptation to climate change. The work of the AIACC provides an extremely important platform to take such steps, but there is yet very far to go to meet the challenges ahead. It is hoped that the material contained in this volume is just the start of a process that must expand and continue in the future. R. K. Pachauri Director General, The Energy and Resources Institute (TERI) and Chairman, Intergovernmental Panel on Climate Change (IPCC)
3243 J&J Vulnerability
1/11/07
10:22 am
Page 1
Part I:
Introduction
3243 J&J Vulnerability
1/11/07
10:22 am
Page 2
3243 J&J Vulnerability
1/11/07
10:22 am
Page 3
1
For Whom the Bell Tolls: Vulnerabilities in a Changing Climate Neil Leary, James Adejuwon, Wilma Bailey, Vicente Barros, Punsalmaa Batima, Rubén M. Caffera, Suppakorn Chinvanno, Cecilia Conde, Alain De Comarmond, Alex De Sherbinin, Tom Downing, Hallie Eakin, Anthony Nyong, Maggie Opondo, Balgis Osman-Elasha, Rolph Payet, Florencia Pulhin, Juan Pulhin, Janaka Ratnisiri, El-Amin Sanjak, Graham von Maltitz, Mónica Wehbe, Yongyuan Yin and Gina Ziervogel
No man is an island, entire of itself; every man is a piece of the continent, a part of the main. If a clod be washed away by the sea, Europe is the less, as well as if a promontory were, as well as if a manor of thy friend’s or of thine own were: any man’s death diminishes me, because I am involved in mankind, and therefore never send to know for whom the bell tolls; it tolls for thee. JOHN DONNE, 1623
Introduction People have evolved ways of earning livelihoods and supplying their needs for food, water, shelter and other goods and services that are adapted to the climates of the areas in which they live. But the climate is ever variable and changeable, and deviations that are too far from the norm can be disruptive, even hazardous. Now the climate is changing due to human actions. Despite efforts to abate the human causes, human-driven climate change will continue for decades and longer (IPCC, 2001a). Who is vulnerable to the changes and their impacts? For whom does the bell toll? We ask, against the oft-quoted advice of the poet John Donne, because understanding who is vulnerable, and why, can help us to prevent our neighbour’s home from washing into the sea, a family from suffering
3243 J&J Vulnerability
1/11/07
10:22 am
Page 4
4 Climate Change and Vulnerability
hunger, a child from being exposed to disease and the natural world around us from being impoverished. All of us are vulnerable to climate change, though to varying degrees, directly and through our connections to each other. The propensity of people or systems to be harmed by hazards or stresses, referred to as vulnerability, is determined by their exposures to hazard, their sensitivity to the exposures, and their capacities to resist, cope with, exploit, recover from and adapt to the effects. Global climate change is bringing changes in exposures to climate hazards. The impacts will depend in part on the nature, rate and severity of the changes in climate. They will also depend to an important degree on social, economic, governance and other forces that determine who and what are exposed to climate hazards, their sensitivities and their capacities. For some, the impacts may be beneficial. But predominantly harmful impacts are expected, particularly in the developing world (IPCC, 2001b). To explore vulnerabilities to climate change and response options in developing country regions, 24 regional and national assessments were implemented under the international project ‘Assessments of Impacts and Adaptations to Climate Change’ (AIACC). The case studies, executed over the period 2002–2005, are varied in their objectives, geographic and social contexts, the systems and sectors that are investigated, and the methods that are applied. They are located in Africa, Asia, Latin America, and islands of the Caribbean, Indian and Pacific Oceans. The studies include investigations of crop agriculture, pastoral systems, water resources, terrestrial and estuarine ecosystems, biodiversity, urban flood risks, coastal settlements, food security, livelihoods and human health. One factor that is common to most of the studies is that they include investigation of the vulnerability of people, places or systems to climatic stresses. Vulnerability studies take a different approach from investigations of climate change impacts. The latter generally emphasize quantitative modelling to simulate the impacts of selected climate change scenarios on Earth systems and people. By contrast, vulnerability studies focus on the processes that shape the consequences of climate variations and changes to identify the conditions that amplify or dampen vulnerability to adverse outcomes. The climate drivers are treated as important in vulnerability studies, but drivers related to demographic, social, economic and governance processes are given equal attention. Understanding how these processes contribute to vulnerability and adaptive capacity in the context of current climate variations and extremes can yield insights regarding vulnerability to future climate change that can help to guide adaptive strategies (Leary, 2002). This volume presents a collection of papers from the AIACC case studies that address questions about the nature, causes and distribution of vulnerability to climate change. In this first chapter we introduce the case studies and present a synthesis of lessons from our comparison of the studies. A companion to this volume, Climate Change and Adaptation (Leary et al, 2008), explores options for adapting to climate change, capacities for implementing these and obstacles to be overcome. Our synthesis of lessons about vulnerability is a product of a week-long
3243 J&J Vulnerability
1/11/07
10:22 am
Page 5
For Whom the Bell Tolls 5
workshop held in March 2005. During the workshop we applied a three-step risk assessment protocol previously used by Downing (2002). In the first step we identified domains of vulnerability that correspond to resources or systems that are important to human well-being, are very likely to be affected by climate change, and are a focus of one or more of the case studies. Four major domains were selected, around which we organized our discussions in the workshop and which have been used to structure both this chapter and the book as a whole: 1) natural resources, 2) coastal areas and small islands, 3) rural economy and food systems, and 4) human health. In the second step, outcomes of concern within each domain were identified and ranked as low-, medium- or high-level concerns. In selecting and ranking outcomes, we attempted to take the perspective of stakeholders concerned about national-scale risks. Outcomes are included that our studies and our interpretation of related literature suggest are plausible, and that would be of national significance should they occur. Our rankings of low-, medium- and high-level concerns are based on the following criteria: potential to exceed coping capacities of affected systems, the geographic extent of damages, the severity of damages relative to national resources, and the persistence versus reversibility of the impacts. The rankings do not take into account the likelihood that an outcome would be realized. They represent the degree of concern that would result if the hypothesized outcomes do materialize. While we have not formally assessed the likelihood of the different outcomes, each is a potential result under plausible scenarios and circumstances. In step three we identified the climatic and non-climatic factors that create conditions of vulnerability to the outcomes of concern within each domain. Where climatic and non-climatic drivers combine to strongly amplify vulnerability, the potential for high-level concern outcomes being realized is greatest. Conversely, where some of the drivers interact to dampen vulnerability, outcomes of lower-level concern are likely to result. The lessons produced from this protocol are presented below in this chapter. The case studies from which they are derived are elaborated on in the chapters that follow.
Natural Resources Natural resources, under pressure from human uses, have undergone rapid and extensive changes over the past 50 years that have resulted in many of them being degraded (MEA, 2005). Population and economic growth are likely to intensify uses of and pressures on natural resource systems. Global climate change, which has already impacted natural resource systems across the Earth, is adding to the pressures and is expected to substantially disrupt many of these systems and the goods and services that they provide (IPCC, 2001b; IPCC, 2007; MEA, 2005). Our case studies investigated vulnerabilities to climate hazards for a variety of natural resources, which are grouped into the contexts of water, land, and ecosystems and biodiversity.
3243 J&J Vulnerability
1/11/07
10:22 am
Page 6
6 Climate Change and Vulnerability
Water Population and economic growth are increasing water demands, and many parts of the world are expected to face increased water stress as a result (Arnell, 2004). Water resources are highly sensitive to variations in climate and, consequently, climate change will pose serious challenges to water users and managers (IPCC, 2001b). Climate change may exacerbate the stress in some places but ameliorate it in others, depending on the changes at regional and local levels. Vulnerabilities from water resource impacts of climate change are addressed by several of the case studies. Outcomes of concern for water resources from these studies and the climatic and non-climatic drivers of the outcomes are identified in Table 1.1. Scenarios of future climate change indicate that many of the study regions, including parts of Africa and Asia, face risks of greater aridity, more variable water supply and periods of water scarcity from drought. In contrast, scenarios suggest that the climate may become wetter and water supply greater in southeastern South America and southeastern Asia. Changes in water balances will impact land, ecosystems, biodiversity, rural economies, food security and human health; vulnerabilities to these impacts are discussed in later sections of this chapter. The outcomes are strongly dependent on factors such as the level and rate of growth of water demands relative to reliable supplies; water and land-use policies, planning and management; water infrastructure; and the distribution and security of water rights. Where water becomes less plentiful and climates drier, the changes have the potential to retard progress towards the Millennium Development Goals. The impacts that can result from persistent and geographically widespread declines in water balances have been demonstrated all too frequently. OsmanElasha and Sanjak (Chapter 12) and Nyong et al (Chapter 11) examine the impacts of decades of below average rainfall and recurrent drought in two parts of the Sudano-Sahel zone with case studies in Sudan and Nigeria respectively. The reduced availability of water in these arid and semi-arid areas has resulted in decreased food production, loss of livestock, land degradation, migrations from neighbouring countries and internal displacements of people. The effects of water scarcity have contributed to food insecurity and the destitution of large numbers of people; they are also implicated as a source of conflict that underlies the violence in Darfur. Non-climate factors that have contributed to the severity of impacts of past climatic events in Sudan and Nigeria create conditions of high vulnerability to continued drying of the climate and future drought. Both studies find that large and growing populations in dry climates that are highly dependent on farming and grazing for livelihoods, lack of off-farm livelihood opportunities, reliance of many households on marginal, degraded lands, high poverty levels, insecure water rights, inability to economically and socially absorb displaced people, and dysfunctional governance institutions create conditions of high vulnerability to changes in water balances. While projected water balance changes for the Sahel and Sudano-Sahel zones are mixed (Hoerling et al, 2006), they include worrisome scenarios of a drier, more drought-prone climate for these regions.
Low • Seasonal droughts
• Increase in heavy precipitation events
• More frequent flood events that increase loss of life, damage to infrastructure, loss of crops and disruption of economic activities
• Losses to water users from localized, temporary and manageable fluctuations in water availability
• Persistent and moderate decrease in rainfall, increase in aridity • More variable rainfall and runoff • More frequent severe drought events • Changes in timing of runoff and water availability
More severe effects kept in check by: • Effective management, planning and policies for water demand and supply
• Growth in populations and infrastructure in flood-prone locations • Poorly managed land-use change, including clearing of vegetation and filling of wetlands that can provide flood protection • Ineffective disaster prevention, preparedness, warning and response systems
• High and growing water demand relative to supply • Extensive land use changes • Pollution from industrial, agricultural and domestic sources • Undefined or insecure water rights • Poor performance of institutions for water planning, allocation and management
• High dependence on subsistence or smallscale rain-fed crop farming and herding • Land degradation • High poverty rate • Insufficient investment in rural development • Inequitable access to water • Lack of social safety nets • Governance failures • High and growing water demand relative to reliable supply
• Philippines (Pulhin et al, Chapter 15) • Western China (Yin et al, Chapter 5) • Thailand and Lao PDR (Chinvanno et al, Chapter 16) • South Africa (Callaway et al, 2006)
• Argentina (Eakin et al, Chapter 13) • Argentina (Barros et al, Chapter 6) • Thailand and Lao PDR (Chinvanno et al, Chapter 16) • Philippines (Pulhin et al, Chapter 15)
• Western China (Yin et al, Chapter 5) • Philippines (Pulhin et al, Chapter 15) • South Africa (Callaway et al, 2006)
• Sudan (Osman-Elasha and Sanjak, Chapter 12) • Northern Nigeria (Nyong et al, Chapter 11) • Mongolia (Batima et al, Chapter 4) • Mexico (Eakin et al, Chapter 13)
AIACC Studies
10:22 am
• Losses from reallocations of water among competing users • Non-violent but costly conflict among competing water users
• Persistent, regional decrease in rainfall, increase in aridity • More variable rainfall and runoff • More frequent severe drought events
• Water scarcity that retards progress on Millennium Development Goals and threatens food security
• Lack of alternative water sources • High and growing water demand relative to reliable supply • Failure of water and land-use policy, planning and management • High dependence on single vulnerable water source
Other Drivers
1/11/07
Medium
• Persistent and severe decline in water balance due to reduced rainfall and/or higher temperatures • Sea level rise causing salt-water intrusion into shallow aquifer of small island • Disappearance of glaciers
• Collapse of water system leading to severe and long-term water shortage
High
Climate Drivers
Outcomes of Concern
Level of Concern
Table 1.1 Water resource vulnerabilities
3243 J&J Vulnerability Page 7
For Whom the Bell Tolls 7
3243 J&J Vulnerability
1/11/07
10:22 am
Page 8
8 Climate Change and Vulnerability
The Heihe river basin of northwestern China has experienced more modest drying over the past decade (Yin et al, Chapter 5). But with increasing development in the basin, water demands have been rising and intensifying competition for the increasingly scarce water. As a result, water users in the basin have become more vulnerable to water shortage, reduced land productivity and non-violent conflict over water allocations. These effects illustrate outcomes of medium- and low-level concern. A drier climate, as some scenarios project for the region, would exacerbate these conditions and could result in outcomes of higher-level concern if future development in the basin raises water demand beyond what can be supplied reliably and sustainably. For the case study regions in the eastern part of the southern cone of South America (Conde et al, Chapter 14; Eakin et al, Chapter 13; Camilloni and Barros, 2003), the Philippines (Pulhin et al, Chapter 15), and the Lower Mekong river basin (Chinvanno et al, Chapter 16), climate change projections suggest a wetter climate and increases in water availability. In the southern cone of South America, increased precipitation over the past two decades has contributed to the expansion of commercially profitable rain-fed crop farming, particularly of soybeans, into cattle ranching areas that were previously too dry for cropping. While this has generated significant economic benefits, the increased rainfall has also brought losses from increases in heavy rainfall and flood events. In the future, a wetter climate in these regions would also bring benefits from increased water availability, but may cause damages from flooding and water-logging of soils (Eakin et al, Chapter 13). Furthermore, farmers may face greater risks from greater rainfall variability in South America that could include both heavier rainfall events and more frequent droughts (IPCC, 2001a). In the Pantabangan–Carranglan watershed of the Philippines, increases in annual rainfall and water runoff would benefit rain-fed crop farmers, irrigators, hydropower generators and other water users. But changes in rainfall variability, including those related to changes in ENSO variability, could intensify competition for water among upland rain-fed crop farmers, lowland irrigated crop farmers, the National Power Corporation and the National Irrigation Administration (Pulhin et al, Chapter 15). Changes in flood risks are also of concern in the watershed. In the Lower Mekong, while increases in annual rainfall may bring increases in average rice yields, shifts in the timing of rainy seasons and the potential for more frequent flooding are found to pose risks for rice farmers (Chinvanno et al, Chapter 16). Those most vulnerable to changes in variability in the Lower Mekong and in Pantabangan–Carranglan are small-scale farmers with little or no land holdings, lack of secure water rights, limited access to capital and other resources, and limited access to decision-making processes.
Land The quality and productivity of land is strongly influenced by climate and can be degraded by the combined effects of climate variations and human activities. Land degradation has become one of the most serious environmental
3243 J&J Vulnerability
1/11/07
10:22 am
Page 9
For Whom the Bell Tolls 9
problems, reducing the resilience of land to climate variability, degrading soil fertility, undermining food production and contributing to famine (UNCCD, 2005a). Seventy per cent of the world’s drylands, including arid, semi-arid and dry sub-humid areas, are degraded, directly affecting more than 250 million people and placing 1 billion people at risk (UNCCD, 2005b). Human-caused climate change is likely to affect land degradation processes by altering rainfall averages, variability and extremes, and by increasing evaporation and transpiration of water from soils, vegetation and surface waters. The effects on land will depend in part on how the climate and water balances change. But they will also depend strongly on non-climate factors that shape human pressures on land. The human consequences will, in turn, be shaped by the ability of people to cope and respond to the effects and to reduce the human pressures that drive land degradation. Two of our case studies, one in Sudan (Osman-Elasha and Sanjak, Chapter 12) and the other in Mongolia (Batima et al, Chapter 4), have land degradation as a central focus. Other studies also examine land degradation as both a potential outcome as well as an amplifier of climate change vulnerability in the Philippines (Pulhin et al, Chapter 15), Tlaxcala, Mexico (Ziervogel et al, Chapter 9) and Tamaulipas, Mexico, and the Argentine Pampas (Eakin et al, Chapter 13). Table 1.2 lists some of the outcomes of concern from the studies that are related to land degradation. The ranking of outcomes is based on the spatial extent, severity of impacts, and the reversibility or irreversibility of land degradation. The climate drivers of land degradation outcomes are increases in aridity and increases in the frequency, severity and duration of droughts. Nonclimate drivers include population growth and economic incentives that create pressures to intensify land uses, expand farming and grazing activities into marginal lands, and clear vegetation. Contributing to this are land tenure systems, land policies and market failures that limit incentives for good land and water management. Widespread poverty, breakdown of local support systems and ineffective governance institutions heighten the vulnerability of populations to income and livelihood losses, food insecurity and displacement from their homes as a result of land degradation. In northern and central states of Sudan, the dry climate, sandy soils and heavy human pressures on the land create conditions of high vulnerability to desertification. Below average rainfall over the past 20 years and growing landuse pressures have degraded grazing and crop lands in North Darfur and reduced food and fodder production and the availability of water (OsmanElasha and Sanjak, 2005). The scarcity of these lifelines has triggered southward migrations of people and their livestock within North Darfur. In addition, people fleeing civil war in neighbouring Chad also migrated into western Sudan. The resulting rapid increases in human population and the number of livestock have intensified pressures on the already fragile environment, including over-grazing and excessive cutting of gum arabic (Acacia senegal) trees to clear land for cultivation and provide fodder and firewood. The reduction in vegetation cover has increased vulnerability to loss of soil and soil fertility by
• Localized but reversible land degradation
Low
• Moderate, temporary drying of localized extent
• Increased aridity of limited geographic extent • Increase in climate variability, including more frequent extreme droughts
More severe effects kept in check by: • Tenure systems and land policies that promote good land management • Households that have sufficient resources with which to cope with reduced food and fodder production • Social systems that function to absorb shocks
• Locally severe overuse of land • Population pressures • Poverty
• Intensive use of land that degrades land productivity during dry periods but does not irreversibly alter soils • Population pressures • Poverty • Inability of land management systems to adapt to climate variations
• Mexico (Eakin et al, Chapter 13) • Philippines (Pulhin et al, Chapter 15)
• Mongolia (Batima et al, Chapter 4) • Mexico (Eakin et al, Chapter 13) • Philippines (Pulhin et al, Chapter 15)
• Sudan (Osman-Elasha and Sanjak, Chapter 12) • Northern Nigeria (Nyong et al, Chapter 11) • Mexico (Eakin et al, Chapter 13)
• Sudan (Osman-Elasha and Sanjak, Chapter 12) • Northern Nigeria (Nyong et al, Chapter 11)
AIACC Studies
10:22 am
• Land degradation of limited geographic extent that is irreversible
• Arid, semi-arid or sub-humid climate • Increase in climate variability, including more frequent extreme droughts
• Widespread but reversible desertification of lands
• Severe overuse of land, including overly intense cropping with poor soil management, poor irrigation practices, extension of cropping into marginal lands, overgrazing of rangelands, removal of vegetation and deforestation • Land tenure systems, land-use policies, market failures and globalization forces that create pressures for overuse and limit incentives for good land management • Population pressure • Breakdown of support systems • Poverty • Poor, erodable soils
Other Drivers
1/11/07
Medium
• Arid, semi-arid or sub-humid climate • Persistent decrease in rainfall, increased aridity • Increase in climate variability, including more frequent extreme droughts
• Widespread desertification of lands with irreversible changes to soil structure or nutrient status
High
Climate Drivers
Outcomes of Concern
Level of Concern
Table 1.2 Land vulnerabilities
3243 J&J Vulnerability Page 10
10 Climate Change and Vulnerability
3243 J&J Vulnerability
1/11/07
10:22 am
Page 11
For Whom the Bell Tolls 11
exposing soils to wind erosion and encroachment of desert sands. Similar processes are degrading lands in Sudan’s North Kordofan state (Ziervogel et al, Chapter 9). The human consequences of drought and land degradation in Sudan are explored later in this chapter. If, as some climate projections suggest, the future climate of the region becomes drier and the frequency and severity of droughts increase, desertification processes would be exacerbated. Mongolia, a nation for which livestock herding is the dominant livelihood activity, is also experiencing serious land degradation (Batima et al, 2005). Over the past 40 years, although rainfall has stayed relatively constant, increases in mean temperatures ranging from near 1°C in the low mountains and on the plains of the Gobi Desert to more than 2°C in the high mountains have resulted in drying of the climate and soils. With the dryer climate, pasture production has declined by 20 to 30 per cent over the same period. Overstocking and overgrazing of pastures in the drier conditions, driven in part by institutional changes that have turned Mongolia’s pastures into an open access commons, has led to degradation of lands in parts of Mongolia. Climate projections indicate that temperatures will continue to rise and suggest that the region may continue to become drier. Such scenarios would likely worsen problems of land degradation in Mongolia.
Ecosystems and biodiversity Habitat change, overexploitation, invasive alien species, pollution and climate change are identified by the Millennium Ecosystem Assessment as presently the most important drivers of ecosystem change and biodiversity loss. By the end of the 21st century, it is possible that climate change may become the dominant driver (MEA, 2005). Case studies in South Africa (von Maltitz and Scholes, Chapter 2) and the Philippines (Lasco et al, Chapter 3) investigate the potential changes in the spatial extent of ecosystem types and biodiversity loss for scenarios of climate change. The findings of these studies are summarized here. Other studies examine the impacts of climate change on the productivity of ecosystems and the consequences for human livelihoods; these are examined in sections of this chapter on water, land, coastal systems and the rural economy. Outcomes of high, medium and low levels of concern from the South African and Philippine studies are presented in Table 1.3. At the high end of the scale, the two studies find that loss of some entire ecosystems, along with extinction of many of their species, is probable for changes in climate that are projected for a doubling of atmospheric concentration of carbon dioxide. In the South African example, projected increases in aridity in the western half of the country would cause current biomes to contract and shift towards the eastern half of the country. A large proportion of South Africa would be left with a habitat type that is not currently found in the country. The impacts vary by location and biome type, and for individual species. The savanna systems of South Africa and their species are found to have relatively low vulnerability to climatically driven extinctions. By comparison, species of the fynbos biome are potentially more vulnerable to climate change
• Slow changes in climate • Small absolute changes in temperature and precipitation that do not fundamentally alter water balances
More severe effects kept in check by: • Managing pressures on ecosystems to a low level • Connections of suitable habitat enable species to migrate
• Sufficient connections of suitable habitat persist across the landscape to enable species to migrate
• South Africa (von Maltitz and Scholes, Chapter 2
• South Africa (von Maltitz and Scholes, Chapter 2)
• South Africa (von Maltitz and Scholes, Chapter 2)
• South Africa (von Maltitz and Scholes, Chapter 2) • Philippines (Lasco et al, Chapter 3)
AIACC Studies
10:22 am
• Genetic loss • Loss of genetic variability, loss of sub-species and varieties
• Slow changes in climate that allow most species to migrate
• Species loss and change in habitat compositional structure
• Moderate pressure on ecosystems due to habitat loss and fragmentation, overexploitation, competition from invasive species and pollution • Changing fire regimes • Changes in grass–tree interactions due to increased CO2 in atmosphere
• Narrow climate tolerances of dominant species of an ecosystem • Extensive habitat loss and fragmentation due to land-use change • Severe pressure from overgrazing, over-harvesting, over-fishing, etc • Severe competition from invasive species • Severe pressure from pollution • Changing fire regimes • Physical barriers to species migration (e.g. islands, mountain tops, isolated valleys) • Changes in grass–tree interactions due to increased CO2 in atmosphere
Other Drivers
1/11/07
Low
• Greater water stress from higher temperatures and lower precipitation
• Species loss and retrogressive succession
Medium
• Rapid rate of change in mean temperature • Changes in water balance across an ecosystem’s geographic distribution that are beyond tolerance limits of dominant species • Changes in seasonal climate extremes, variability and means
• Collapse or loss of entire ecosystem and extinction of many of the system’s species
High
Climate Drivers
Outcomes of Concern
Level of Concern
Table 1.3 Ecosystems and biodiversity vulnerabilities
3243 J&J Vulnerability Page 12
12 Climate Change and Vulnerability
3243 J&J Vulnerability
1/11/07
10:22 am
Page 13
For Whom the Bell Tolls 13
than are those of the savannas. The fynbos is the major vegetation type of the Cape Floral Kingdom, which is the smallest of the world’s six floral kingdoms. It is located entirely in South Africa, has the highest concentration of species of any of the floral kingdoms and has a species endemism rate of 70 per cent. While the fynbos biome is projected to have relatively little loss in spatial extent, climatic habitats would move for many individual species and some climatic habitats would disappear completely. Model simulations suggest that many species of the fynbos will be able to migrate with their moving habitats, but some would not and would be lost. The situation for the succulent karoo biome is more dire. The succulent karoo is an arid ecosystem of southwestern South Africa and southern Namibia that is also rich in biodiversity and high in species endemism. Model simulations for climate change scenarios corresponding to a doubling of carbon dioxide project that almost the entire extent of the succulent karoo would be lost to a new climatically-defined habitat type. Extinction of many of the species endemic to the biome would probably result. In the Philippines, increasing temperature and rainfall are projected by Lasco et al (Chapter 3) to result in the dry forest zone being completely replaced by wet forests and rainforests. They estimate that a 50 per cent increase in precipitation would cause dry forests, which occupy approximately 1 million hectares, to disappear completely from the Philippines and moist forests, which occupy 3.5 million hectares, to decline in area by two thirds. Most of these forest areas would become wet forests, which would more than double their area from their present size. If precipitation were to increase by 150 per cent, which is within the range of climate model projections for the end of the century, all dry and moist forests would disappear, wet forests would decline by half and rain forests, a forest type not currently present in the Philippines, would grow to 5 million hectares. The warmer, wetter climate that is projected for the Philippines would increase the primary productivity of the forests and produce associated benefits. But the disappearance of dry and possibly moist forest types would result in loss of species.
Coastal Areas and Small Islands Coasts and small islands are highly exposed to a variety of climate hazards that may be affected by global climate change. The climatic hazards converge with local and regional human pressures in coastal zones to create conditions of high vulnerability, particularly in areas with high concentrations of people and infrastructure along low-lying coasts. Barros et al (Chapter 6) investigate flood risks from storm surges along the Argentine coast of the Río de la Plata. Nagy et al (Chapter 7), also working in the Río de la Plata basin, examine changing dynamics of the estuarine ecosystem and their implications for fisheries on the Uruguayan side. Payet (Chapter 8) explores problems of coastal erosion and also risks to tourism in the Seychelles, while Mataki et al (2005 and 2006) assess the vulnerability of coastal towns of Fiji to flooding. Outcomes of concern from these studies are summarized in Table 1.4.
• Changes in number of wet days and frequency of storms
• Increase in frequency and intensity of extratropical and tropical storms • Sea level rise • Sea level rise • Changes in winds, water temperatures, and freshwater inflow to estuaries and coastal waters
• Loss of tourism-related income, export earnings and jobs
• Severe coastal erosion
• Damage to coastal ecosystems and their services and resulting impacts on fishing livelihoods
• Pollution discharges into waters • Nutrients carried into coastal waters by runoff • Use of fertilizers that runoff into coastal waters • Removal of vegetation that increases erosion • Hardening of shoreline to protect against storm surges • Over-harvesting of fish and shellfish
• Uruguay (Nagy et al, Chapter 7)
• Seychelles (Sheppard et al, 2005; Payet, Chapter 8) • Fiji (Mataki et al, 2005 and 2006)
• Seychelles (Payet, Chapter 8)
• Central America (Project LA06, www.aiaccproject.org) • Argentina (Barros et al, Chapter 6)
AIACC Studies
10:22 am
• Intensive land uses in the coastal zone • Loss of coastal wetlands and bleaching of coral reefs
• Damages to infrastructure, beaches, water supply and ecosystems that provide tourismrelated services • High dependence on tourism for income and employment
• Large and growing population and infrastructure in exposed coastal areas • Lack of land-use policies to avoid/reduce exposures • Lack of maintenance of flood control infrastructure • Loss of wetlands and reefs • Ineffective disaster prevention, preparedness, warning and response systems
Other Drivers
1/11/07
Medium
• Increase in frequency and intensity of extratropical and tropical storms • Sea level rise
• More frequent and greater loss of life, infrastructure damage, displacement of population and disruption of economic activities
High
Climate Drivers
Outcomes of Concern
Level of Concern
Table 1.4 Coastal area and small island vulnerabilities
3243 J&J Vulnerability Page 14
14 Climate Change and Vulnerability
Low
Level of Concern
• Increase in frequency and intensity of extratropical and tropical storms • Sea level rise
• Sea level rise • Changes in water balance and ENSO and monsoon variability
• Diminishing and less reliable water supply
• Modest acceleration of coastal erosion and modest infrastructure damage
Climate Drivers
Outcomes of Concern
More severe effects kept in check by: • Low concentrations of population and infrastructure in areas exposed to erosion • Intact coastal wetlands and inland vegetation • Good coastal policies and management practices
• Increasing water demand from growing population and economic activity • Increasing extraction of groundwater
Other Drivers
Table 1.4 (continued)
• Argentina (Barros et al, Chapter 6)
• Fiji (Mataki et al, 2005 and 2006)
AIACC Studies
3243 J&J Vulnerability 1/11/07 10:22 am Page 15
For Whom the Bell Tolls 15
3243 J&J Vulnerability
1/11/07
10:22 am
Page 16
16 Climate Change and Vulnerability
Barros and colleagues find that sea level rise would permanently inundate only small and relatively unimportant areas along the southern coast of the Río de la Plata during this century. However, the area and population that would be affected by recurrent flooding from storm surges would increase considerably. They estimate that sea level rise and changes in wind fields would increase the population affected by storm surges with a five-year return period from 80,000 persons at present to nearly 350,000 in 2070. For storm surges with a 100-year return period, the population affected would rise from 550,000 at present to nearly 900,000 by 2070. Economic costs resulting from real estate damage and increased operational costs of coastal public facilities are estimated to range between 5 and 15 billion US dollars for the period 2050–2100, depending on the rate of sea level rise. These estimates are based on the current population and development in the basin. Continuation of trends that have been concentrating both people and infrastructure on the coast would increase the number of people exposed and the potential economic damage. Coastal erosion is common to all coasts, but the level of concern that it engenders ranges from low to high depending on local circumstances. The study by Barros and colleagues finds that coastal erosion is presently of little concern in the Río de la Plata basin, though concern could rise if newly accreted lands in the Parana delta are allowed to be settled and developed. In contrast, Payet finds that concern about coastal erosion is high in the Seychelles and Mataki, and colleagues make the same conclusion for Fiji and the Cook Islands. In each of these cases, infrastructure and resources are more exposed to the impacts of erosion than is the case in the Río de la Plata. Another recent study of the Seychelles (Sheppard et al, 2005) found that coastal erosion is significantly heightened as a result of coral bleaching events that reduce the ability of reefs to dissipate wave energy. The authors conclude that areas that have experienced mass bleaching are at a higher risk from coastal erosion under accelerated sea level rise. In the Seychelles, as in many island states, tourism is a major contributor to incomes. The attributes that make the Seychelles and other islands attractive tourist destinations can be highly sensitive to climate stresses. The high economic dependence on tourism and the sensitivity of tourist resources to climate create a situation of high socioeconomic vulnerability to climate change (Payet, Chapter 8). Climate change can impact tourism by accelerating beach erosion, inundating and degrading coral reefs, damaging hotels and other tourism-related infrastructure, and discouraging tourists from visiting because changes in climate reduce an area’s appeal. In a scenario that assumes a substantial increase in the number of wet days per month, Payet estimates that tourist visits would be reduced by 40 per cent. They also estimate that the decrease in tourist visits would reduce tourism expenditures by 40 million US dollars and cause over 5000 jobs to be lost, representing 15 per cent of the national labour force. The effects would be felt in all areas of the economy. The trophic state of the estuary of the Río de la Plata has degraded since the mid-1940s. The eutrophication of the estuary is due primarily to nutrients introduced by increased fertilizer use and changes in human land uses, but climatic factors such as changes in river flows and wind patterns have also contributed (Nagy et al, Chapter 7). A consequence of the eutrophication is an
3243 J&J Vulnerability
1/11/07
10:22 am
Page 17
For Whom the Bell Tolls 17
increase in the frequency of harmful algae blooms in the last decade, resulting in considerable economic harm to commercial fisheries and tourism as well as negative impacts on public health in Uruguay. Climate change would impact the estuary through changes in freshwater input from tributaries and changes in winds that would modify the circulation, salinity front location, stratification and mixing patterns. These changes would, in turn, alter oxygen content, nutrients and primary production in the estuary. Many of the estuary’s services would be altered. But the specific changes are difficult to predict as they depend on the balance of multiple and complex interactions. One of the concerns is the sustainability of fisheries in the Río de la Plata. A case study of an artisanal fishery located on the northern coast of the estuary finds the fishermen and the fishing settlement to be vulnerable to climate driven shifts in the salinity front location and other changes in the estuary that would alter fish catch or the effort and cost required to catch fish.
Rural Economies and Food Security Several of the case studies investigate the vulnerability of rural livelihoods to climate variability and change, and the implications for rural economies and food security. Rural livelihoods and economies are based on and dominated by agricultural, pastoral and forest production systems that are highly sensitive to climate variations. Climate change can and will have both positive and negative impacts on the productivity of these systems, which will, in turn, impact incomes, costs of production, supplies of food and other commodities, stores of food, livestock and financial savings, and food security. Table 1.5 highlights some of the potential negative outcomes identified as concerns by the studies. The focus is on negative outcomes because our interest is in understanding who is vulnerable in rural economies, how they are vulnerable and why. The productivity of farm fields, pastures and forests will be impacted by exposures to changes in the averages, ranges and variability of temperatures and precipitation, water balances and frequencies, severities of droughts, floods and other climate extremes, and the ameliorating effects of higher carbon dioxide concentrations on plant processes. One of the common findings of the studies is that systems with similar exposures to climate stimuli can vary considerably in their vulnerability to damage from such exposures. The particular factors that determine vulnerability are context-specific and vary from place to place. But some commonalities can be identified. Rural households’ sensitivity to climate shocks and capacity to respond vary according to their access to water, land and other resources and the condition and quality of these resources. Large and growing populations, a high proportion of households engaged in subsistence or small-scale farming and herding, land degradation, high poverty rates and governance failures create conditions of vulnerability for rural economies and households. Declining local authority, lack of social safety nets, violent conflict, gender inequality and competition from market liberalization are also factors that add to vulnerability in the different case study areas. These issues are developed in the sections below.
• Persistent below average rainfall, increased aridity • Severe, multi-year, geographically widespread drought events
• More frequent climate extremes over large portion of growing area of key export crops • Changes in average climate or shifts in rainy season that stress export crops
• Multi-year collapse of rural production systems • Widespread and persistent loss of livelihoods and impoverishment • Chronic hunger and malnutrition for large percentage of population • Long-term or permanent out-migration on large scale
• Loss of export earnings • Loss of national income • Loss of jobs
• Sri Lanka (Ratnasiri et al, Chapter 17)
• Sudan (Osman-Elasha and Sanjak, Chapter 12) • Northern Nigeria (Nyong et al, Chapter 11) • Mongolia (Batima et al, Chapter 4)
• Sudan (Osman-Elasha and Sanjak, Chapter 12) • Northern Nigeria (Nyong et al, Chapter 11)
AIACC Studies
10:22 am
• High dependence on small number of agricultural commodities for export earnings, national income and employment • Declining or volatile export crop prices • Insufficient investment in research, development and diffusion of agricultural technology
• Large and growing population in dryland areas • High percentage of households engaged in subsistence or small-scale farming and herding on lands with poor soils and no irrigation • Overuse or clearing of lands leading to land degradation • Lack of or insecure water rights • High poverty rate • Lack of off-farm livelihood opportunities • Lack of social safety nets • Governance failures
• Tensions among rival groups • Migrations of herders into lands of sedentary farmers • Collapse of local authorities • Governance failures • Scarcity of food, water and other resources
Other Drivers
1/11/07
Medium
• Persistent below average rainfall, increased aridity • Severe, multi-year, geographically widespread drought events
• Violent conflict • Famine
High
Climate Drivers
Outcomes of Concern
Level of Concern
Table 1.5 Rural economy vulnerabilities
3243 J&J Vulnerability Page 18
18 Climate Change and Vulnerability
Climate Drivers • Region-wide increase in frequency of climate extremes that cause losses of crops, livestock and income • Changes in average climate or significant shifts in rainy season that stress traditionally grown crops and available substitutes
• Increase in frequency of climate extremes that cause losses of crops, livestock and income • Changes in average climate or shifts in rainy season that are less optimal for traditionally grown crops
Outcomes of Concern
• Increased rural poverty rates • Declining and more variable net farm incomes for many rural households • Failures of small farms • Accelerated rural-to-urban migration
• Declining and more variable net farm incomes for some rural households • Decreased and more variable quality of crop and livestock output • Temporary migrations as strategy to obtain off-farm incomes
More severe effects kept in check by: • Robust and diversified rural development • Equitable access to resources (e.g. improved seed varieties) • Adequate household savings • Maintenance of social safety nets • Political stability • Well maintained rural infrastructure and services • Access to credit and insurance
• Declining output prices (e.g. due to trade liberalization) • Rising input prices (e.g. due to removal of subsidies) • Lack of income diversification of rural households • Lack of access to credit by small farmers • Stagnant rural development • Poor rural infrastructure (e.g. roads, water storage) • Lack of social safety nets
Other Drivers
• Argentina and Mexico (Eakin et al, Chapter 13) • Thailand and Lao PDR (Chinvanno et al, Chapter 16) • Philippines (Pulhin et al, Chapter 15) • Sri Lanka (Ratnasiri et al, Chapter 17)
• Argentina and Mexico (Eakin et al, Chapter 13) • South Africa, Nigeria, Sudan and Mexico (Ziervogel et al, Chapter 9) • Thailand and Lao PDR (Chinvanno et al, Chapter 16) • Philippines (Pulhin et al, Chapter 15)
AIACC Studies
1/11/07
Low
Level of Concern
Table 1.5 (continued)
3243 J&J Vulnerability 10:22 am Page 19
For Whom the Bell Tolls 19
3243 J&J Vulnerability
1/11/07
10:22 am
Page 20
20 Climate Change and Vulnerability
Household access to resources Access or entitlements to land, water, labour and other inputs to rural production processes are important determinants of the vulnerability of rural households. They shape the sensitivity of households’ livelihoods and food security to variations in climate and land productivity. They also underpin the capacity of households to withstand and respond to the impacts. Ziervogel et al (Chapter 9) compare the determinants of food insecurity from four case studies: Mangondi village in Limpopo Province, South Africa; Gireigikh rural council in North Kordofan, Sudan; Chingowa village in Borno State, Nigeria; and Tlaxcala State, Mexico. Each of the study sites has a dry, drought-prone climate and exposure to declining average precipitation and frequent drought are sources of risk to household food security. Comparison of the cases demonstrates that household characteristics related to resource access play a dominant role in determining household vulnerability. These include household income, income diversification, area of land cultivated, soil quality, household labour per hectare cultivated and the health status of household members. Factors external to the household also control access to resources needed to cope with and recover from climate shocks. These include the existence of formal and informal social networks, availability and quality of health services, and prices of farm inputs and outputs. In each of the case studies, labour available to the farm household is adversely affected by rural–urban migration and infectious disease such as HIV/AIDS and malaria. Adejuwon (Chapter 10) compares the vulnerability of peasant households to climate shocks in different states of Nigeria using household census data. He finds that the percentage of households employed in agriculture, poverty rate, dependency ratio, access to potable water, health status and educational attainment are important determinants of vulnerability. Also important is the aridity of the climate and quality of soils. The comparison identifies households in the northern states of Nigeria as the most vulnerable in the country. Nyong et al (Chapter 11) conduct detailed surveys of households in these northern states to identify household characteristics that determine vulnerability. Key characteristics include ownership of land and livestock, area and quality of land cultivated, sufficiency of annual harvest relative to household food needs, dependency ratio, cash income, livelihood diversification, gender of household head, and connections to family and social networks. Women in this patrilineal society can be particularly vulnerable. In the Pantabangan–Carranglan watershed of the northern Philippines, households are exposed to variability in rainfall and water supply as well as to flood events (Pulhin et al, Chapter 15). The vulnerability of households to these exposures is found to be correlated with variables that determine access to and control of resources: ownership of land, farm size, farm income, gender, and status as a native or migrant to the basin. Large landowners are less vulnerable to variable incomes and other impacts of climate events than are smallholder farmers due to their greater resources for coping and recovery, their ability to live in locations that are less exposed to flooding and erosion, and their ability to capture more of the benefits from development projects due
3243 J&J Vulnerability
1/11/07
10:22 am
Page 21
For Whom the Bell Tolls 21
to their ties to the institutions that implement these projects. Projections of future climate change suggest the potential for greater precipitation in the Philippines, which would ease water scarcity in most years. However, flooding may become a more frequent stress and would likely impact poor, smallholder farmers of the basin the hardest. Chinvanno et al (Chapter 16) similarly find that land ownership and other indicators of economic vitality are important determinants of the vulnerability of rice farmers in the Lower Mekong basin. Farmers of rain-fed rice in Thailand and Lao People’s Democratic Republic (PDR) are exposed to variations in rice harvests and other impacts from seasonal flooding, shifts in the dates of onset and cessation of the rainy season and variations in rainfall amounts. Farm households with small land holdings produce low volumes of rice and incomes from rice, which are often only enough to sustain the household on a year-to-year basis. As a result, smallholder farm households have very limited buffering capacity to deal with losses or to cope with anomalies during the crop season. Small land holdings also limit the ability of the farmer to implement other activities to diversify their income sources. Comparing farm households from the Thai study sites with those at the Lao PDR sites, Chinvanno and colleagues find that a larger proportion of the Thai farmers are at high risk from climate shocks despite their higher monetary incomes. They attribute this to higher food costs relative to farm income, lack of income diversification, little savings in the form of financial assets, livestock or food stores, and high debt relative to income among farmers at the Thai sites. Farmers at the Lao PDR study sites also have the advantage of being able to supplement their food supplies with harvests of various products from relatively healthy natural ecosystems adjacent to their farms, an option that is not available at the Thai sites.
Land degradation Land degradation is both an outcome of climate stress and a source of additional stress that can amplify the vulnerability to climate impacts of people making a living from the land. It is found to be an important factor in several of the case study areas, including Mongolia (Batima et al, Chapter 4), Sudan (Osman-Elasha and Sanjak, Chapter 12), northern Nigeria (Nyong et al, Chapter 11), the Philippines (Pulhin et al, Chapter 15), and Argentina and Mexico (Eakin et al, Chapter 13). In the grazing lands of Mongolia, land degradation has been severe due to the harsh and variable climate, drying of the climate over a 40-year period, and heavy grazing pressures. As noted earlier, these conditions have depressed pasture productivity and livestock production. Batima et al (Chapter 4) find that these stresses, combined with the effects of profound economic and institutional changes that accompanied Mongolia’s transition from a socialist to a market system, have increased vulnerability among herders to climate extremes, as was demonstrated by events in 1999–2003. Several years of summer droughts and severe winter conditions (called zud) combined to drastically reduce pasture production, animal weights at the start of winters and stores of
3243 J&J Vulnerability
1/11/07
10:22 am
Page 22
22 Climate Change and Vulnerability
fodder for winter months. Approximately 12 million head of livestock died as a result, roughly a quarter of Mongolia’s herds. Thousands of families lost animals, an important source of savings, which increased poverty and reduced further the capacity of livestock-dependent households to cope with shocks. Many lost their livelihoods with their animals and migrated to urban centres where unemployment is high and few opportunities awaited them. Climate scenarios suggest the potential for further drying of the climate and Mongolia’s herders continuing to be in a state of high vulnerability to the effects of land degradation, drought and zud. In the Philippines’ Pantabangan–Carranglan watershed, reforestation and community development projects were implemented to reverse land degradation problems and provide other benefits. However, the projects developed a dependency on external assistance for livelihoods. Many of the jobs associated with the development projects ended when the projects were terminated. Affected households have resorted to charcoal making and kaingin (slash and burn) farming, which are damaging the fragile environment of the watershed, including the reforested areas, and increasing vulnerability to flooding (Pulhin et al, Chapter 15). Food insecurity is increasing in Tlaxcala, Mexico, for a variety of causes, including a shortage of farm labour due to out-migration of young males, declining maize prices and severe soil erosion problems (Ziervogel et al, Chapter 9). The shortage of farm labour constrains the use of soil conservation practices, which are labour-intensive, and leads to the expansion of monocropping of maize, a system that increases soil erosion. Eakin et al (Chapter 13) find that mono-cropping has contributed to land degradation and heightened vulnerability in their study areas in Mexico and Argentina. In Tamaulipas, Mexico, mono-cropping of sorghum, which is resilient to water stress, was promoted by the national government as a strategy for managing drought risks. However, farming of sorghum under persistent drought conditions in the 1990s may have resulted in degradation of soils that is adding to farmers’ risks from drought. Now the government is using incentive payments to farmers to encourage them to switch to other alternatives. In the Argentine Pampas, the dramatic expansion of soybean mono-cropping is also observed to be associated with land degradation. This contributes to flood problems and raises concern about the long-term sustainability of soybean farming in the region.
Conflict Persistent low rainfall, recurrent drought, land degradation, high population growth, governance failures and other factors have deepened poverty and resulted in food and resource scarcity in the Sahel. The scarcities have contributed to tensions between competing groups and tribes. Cereal production in the region has grown at a meagre 1 per cent per annum over the past decade, while the population has grown at an estimated 2.7 per cent. Against this backdrop of generally tightening food scarcity, climatic and other events have created conditions of crisis. Responses can and have inflamed tensions that
3243 J&J Vulnerability
1/11/07
10:22 am
Page 23
For Whom the Bell Tolls 23
contribute to violent conflict, compounding the vulnerability of populations to climatic and other stresses. Events in Sudan’s North Darfur State illustrate a case of extremely high vulnerability of the population to loss of livelihoods, livestock, lands and personal security leading to destitution, hunger, famine and violent death (Osman-Elasha and Sanjak, Chapter 12). The drivers of this human misery are multiple and complex. Among them are 20 years of below average rainfall that has severely reduced the availability of water, food and fodder in this dry region of infertile soils. The drying climate and human pressures on the land, exacerbated by migrations of people and their livestock into the area, are degrading the land. Traditional land management systems and practices have been disrupted, bringing nomadic and sedentary tribes into more frequent contact and conflict over land and other scarce resources. These resource conflicts are an important factor behind the widespread violence that has taken tens of thousands of lives in Darfur and forced many more to flee their homes. The lack of physical security and access to resources have devastated livelihoods, eroded capacities to cope with climate and other stresses, and threaten people of the region with famine. Farmers and herders of northern Nigeria face similar pressures. In their case study of northern Nigeria, Nyong et al (Chapter 11) find that food scarcity and rising food prices have led to intensification of farming and grazing and expansion of these activities into more marginal lands. The greater land-use pressures, combined with the persistent decline in average rainfall, have added to land degradation problems. The productivity of grazing lands has declined in the north. In response, herders have migrated south into sedentary farmers’ lands. The resulting conflicts have led to the loss of lives, the destruction of crops, livestock and farmlands, and food insecurity for those affected.
Commodity export-oriented economies The sensitivity of cash crop yields to climate variability and change is of considerable importance to countries that depend heavily on the contribution of cash crops to national income and foreign exchange earnings. In Sri Lanka, coconut and tea production are the largest sources of export earnings, major contributors to national income and significant employers of labour. Ratnasiri et al (Chapter 17) investigate the effects of past climate variations on the plantation tea sector of Sri Lanka and develop crop models to simulate yield responses to future climate change. Tea yields are affected by variations in both precipitation and temperature. The 1992 drought in Sri Lanka caused a 25 per cent decline in tea production and a corresponding 22 per cent decline in foreign exchange earnings from tea. Projections of future climate change indicate warmer temperatures and both increases and decreases in precipitation. In the lowlands, where temperatures are near the optimum for tea yields, warming would decrease yields. In the cooler uplands, tea yields would increase with warming. Hence the lowland plantations, owned largely by smallholders with low adaptation capacity, are more vulnerable than the upland plantations, which are owned by large companies. An important factor for vulnerability in
3243 J&J Vulnerability
1/11/07
10:22 am
Page 24
24 Climate Change and Vulnerability
these cash crop sectors will be the effect of climate change on climate variability, particularly the frequency of drought, which, as shown by past events, is a significant source of risk for these sectors.
Market forces and social safety nets The case studies by Eakin et al (Chapter 13) of crop and livestock farms in Cordoba, Argentina, and Tamaulipas, Mexico, demonstrate the influences of international market integration and government social programmes on the vulnerability of farmers. Both countries have pursued policies of trade liberalization, privatization and deregulation. The policies have opened access to international markets and foreign investments, allowing, for example, the profitable expansion of soybean farming in Argentina. But competition from overseas producers and removal of price supports and input subsidies have created a ‘price squeeze’ for farmers, particularly for maize farmers in Mexico. In this highly competitive environment, farm households have less margin for absorbing shocks, including crop and livestock losses from climate extremes, and so are more vulnerable. The pressures are leading to greater concentration of farms into larger-scale commercial operations as smaller family farms face a number of disadvantages, including higher cost of credit, lack of access to technical skills, high dependence on crop income, greater problems with pests and lack of economies of scale. The problems for small farmers are compounded by cutbacks in state-supported social security mechanisms, resulting in declining rural incomes and increasing inequality between small and large landholders. In Tamaulipas, communal and private tenure small farmers are responding to declining and uncertain farm incomes by diversifying into off-farm sources of income, a trend that is reducing their vulnerability to direct climate impacts. Mongolia’s transition from a socialist to market system in the early 1990s brought dissolution of the collectives through which the livestock sector was managed and subsidized services delivered to herders and their families. As described in Batima et al (Chapter 4), herders received livestock for private ownership but found themselves lacking access to health, education, veterinary, water and marketing services, facing higher prices for transport, fodder and fencing materials, and lacking institutions to regulate access to pastures (which remain state owned), enforce seasonal migration of herds or reserve pastures for emergency use. The dissolution of the social safety nets heightened the vulnerability of pastoralists to climate and other hazards. Traditional customary institutions of the pre-socialist era are beginning to re-emerge and to fill some of these needs.
Human Health The paths by which climate can affect human health are diverse and involve both direct and indirect mechanisms. The most direct mechanisms operate through human exposures to climatic extremes that can result in injury, illness
3243 J&J Vulnerability
1/11/07
10:22 am
Page 25
For Whom the Bell Tolls 25
and death. Climate and climate change also affect human health by influencing human exposure to infectious disease through effects on the biology, habitats and behaviours of disease pathogens, hosts and vectors. Even less directly, climate and climate change can affect human health through impacts on the resources that individuals and communities need to maintain good health. Health outcomes of concern highlighted in the synthesis workshop are summarized in Table 1.6. Many vector-borne infectious diseases are climate-sensitive and epidemics of these diseases can occur when their natural ecology is disturbed by environmental changes, including changes in climate (McMichael et al, 2001). For example, observations of numbers of malaria and dengue cases vary with interannual variations in climate (Wandiga et al, Chapter 18; Heslop-Thomas et al, Chapter 19; Kilian et al, 1999; Lindblade et al, 1999). In the Lake Victoria region of East Africa, significant anomalies in temperature and rainfall were recorded during the El Niño period of 1997–1998 and these were followed by severe malaria outbreaks. A similar association of dengue fever occurrences with ENSO variability has been observed in Jamaica. Other infectious diseases that have been observed to be sensitive to climate variability and change include other insect-borne diseases such as encephalitis, yellow fever and Leishmaniasis and water-borne diseases such as cholera, typhoid and diarrhoea (Aron and Patz, 2001; McMichael et al, 2001). Projected changes in rainfall and temperature have the potential to expose more people to vector-borne diseases by expanding the geographic range of vectors and pathogens into new areas, increasing the area of suitable habitats and numbers of disease vectors in already endemic areas, and extending transmission seasons. For example, average temperature and precipitation in the East African highlands are projected to rise above the minimum temperature and precipitation thresholds for malaria transmission and extend malaria into areas from which it has been largely absent in the past (Githeko et al, 2000). Other studies suggest that if El Niño events continue to increase in frequency, the elevated temperatures and precipitation would increase malaria transmission (Kilian at al, 1999; Lindblade et al, 1999). In rural communities of the highlands studied by Wandiga et al (Chapter 18), risks for developing malaria and complications from the disease are amplified by low utilization of hospitals and clinics because of distance, cost and low incomes. The health outcome identified as the highest-level concern is sustained or often repeated in geographically widespread epidemics with high mortality rates. At medium and low levels of concern are more frequent epidemics or outbreaks of infectious disease that may be associated with mortality but which are geographically and temporally limited. Another concern is that changes in climate may allow more virulent strains of disease or more efficient vectors to emerge or be introduced to new areas. Whether changes in climate result in greater infectious disease incidence or epidemics, the geographic extent and severity of epidemics or outbreaks that might result depend on complex interactions that include not just the effect of climate stresses on the ecology of infectious disease, but also on demographic, social, economic and other factors that determine exposures, transmission,
Outcomes of Concern
• More frequent geographically widespread and sustained epidemics of infectious and waterborne disease with high human mortality
• Emergence of new or more virulent strains of infectious disease and more efficient disease vectors • More frequent but geographically and temporally limited epidemics with high or moderate mortality • Increase in number of infectious disease cases and mortality in endemic areas and seasons
• More frequent but geographically and temporally limited epidemics with no mortality • Increase in number of isolated infectious disease cases that are not life-threatening
Level of Concern
High
Medium
Low
• Changes in climate that alter disease and vector ecology and transmission pathways • Changes in climate that moderately increase exposures by expanding endemic areas and seasons
More severe effects kept in check by: • Access to healthcare • Effective disease surveillance, vector control and disease prevention • Good nutritional and health status of population • Access to potable water and sanitation
• East Africa (Wandiga et al, Chapter 18) • Caribbean (Heslop-Thomas et al, Chapter 19)
• East Africa (Wandiga et al, Chapter 18) • Caribbean (Heslop-Thomas et al, Chapter 19)
• East Africa (Wandiga et al, Chapter 18) • Caribbean (Heslop-Thomas et al, Chapter 19)
AIACC Studies
10:22 am
• Land-use changes that increase habitat for disease vectors and reservoirs for zoonotic diseases • Crowding • Drug resistance • International migration, travel and trade • Water storage and sanitation practices • Poor programmes for disease surveillance, vector control and disease prevention • Declining quality and increasing cost of healthcare
• Severely degraded or collapsed healthcare system • Poor and declining immunity, nutritional and health status of large portion of population • High poverty rates that limit access to healthcare • Poor or non-existent programmes for disease surveillance, vector control and disease prevention • Large portion of population lack reliable access to potable water and sanitation • Land-use changes that increase habitat for disease vectors and reservoirs for zoonotic diseases
Other Drivers
1/11/07
• Changes in climate that alter disease and vector ecology and transmission pathways • Changes in climate that moderately increase exposures by expanding endemic areas and seasons
• Geographically widespread changes in climate that increase the geographic area and number of disease vectors • More frequent heavy rainfall and drought events that disrupt water supply and sanitation and expose people to waterborne pathogens
Climate Drivers
Table 1.6 Human health vulnerabilities
3243 J&J Vulnerability Page 26
26 Climate Change and Vulnerability
3243 J&J Vulnerability
1/11/07
10:22 am
Page 27
For Whom the Bell Tolls 27
results of infection, treatment and prognosis. Vulnerability to severe health outcomes are greatest where the healthcare system is severely degraded, large numbers of people lack access to healthcare, the immunity, nutritional and health status of the population is low, and effective programmes for disease surveillance, vector control and disease prevention are lacking (see Table 1.6). Where the converse of these conditions holds, the likelihood that the most severe health outcomes would be realized is much diminished. Many of the climate change impacts described in previous sections can also have health impacts by reducing individuals’ resilience to disease, the resources available to maintain and protect their health and obtain access to healthcare, and the ability of their communities to deliver quality healthcare services. Examples of these indirect effects include households placed at greater risk of illness as a result of loss of livelihood, assets and support networks from severe and persistent drought, health risks associated with displacement and crowding of populations that migrate in response to climate impacts, healthcare systems being overburdened by increases in case loads as a result of direct health effects of climate change, and impacts of climate extremes on healthcare infrastructure and personnel. The severity of the indirect health outcomes that are realized will depend on the geographic extent, persistence and return period of the triggering climatic event, the severity of impact on resource productivity, livelihoods and healthcare infrastructure, and the resilience of the affected area as indicated by the diversity of economic opportunities, poverty rate, health status and capacity of the healthcare system relative to the population.
Conclusion In all of the case studies, climate hazards are a significant danger now, not just in the distant future. Potential outcomes from exposure to climate hazards and climate change identified as high-level concerns include water scarcity that retards progress towards development goals, land degradation, losses of entire ecosystems and their species, more frequent and greater loss of life in coastal zones, food insecurity and famine, loss of livelihoods, and increases in infectious disease epidemics. All of these are plausible outcomes of exposure to climate hazards. Whether they are likely outcomes will vary and will depend on the degree and nature of vulnerability of the exposed systems. Vulnerability to impacts from climate variation and change is shown by our case studies to have multiple causes. The causes include not only exposure to the climatic stressors, but also to stressors that derive from interactions among environmental, demographic, social, economic, institutional, political, cultural and technological processes. The state and dynamics of these processes differ from place to place and generate conditions of sensitivity, adaptive capacity and vulnerability that differ in character and degree. Consequently, populations that are exposed to similar climatic phenomenon are not necessarily impacted the same. Differences in vulnerability are also apparent for different sub-populations or groups inhabiting a region, and even from household to household within a
3243 J&J Vulnerability
1/11/07
10:22 am
Page 28
28 Climate Change and Vulnerability
group. Factors such as sources and diversity of a household’s livelihood, experience and skills, level of wealth, ownership and access to land, water and other resources, support from social networks, and access to technical assistance and knowledge give rise to differences in vulnerability between households. Our synthesis focuses on four domains of vulnerability: natural resources; coasts and small islands; rural economies and food security; and human health. A common finding across the domains of vulnerability is that impacts ranked as high-level concerns generally are not likely to result from climate stress alone. They are most likely to be realized when multiple stresses act synergistically to create conditions of high vulnerability. A climate shock or stress has the potential to do the most damage in a context in which natural systems are being severely stressed and degraded by overuse and in which social, economic or governance systems are in or near a state of failure and thus not capable of effective responses. Unfortunately, such conditions exist in many parts of the world, particularly the developing world. Places where this is true are consequently vulnerable to some of the high-level concern outcomes from exposure to climate stresses, both now, from current climate variations and extremes, and increasingly in the future as the climate changes. An exception is the potential loss of some ecosystems or their biodiversity, which might, in some instances, be triggered by climate change alone. For example, the rate of climate change is a key factor that threatens the succulent karoo biome of South Africa, and a rapid rate of change could, by itself, be sufficient to cause its demise. But for many other ecosystems it will be the interaction of a changing climate with pressures from human uses, and management of land and other resources that will probably determine their fate. More optimistically, our studies suggest that the potential severity and risk of many of the outcomes are less where social, economic and governance systems function in ways that enable effective responses to prevent, cope with, recover from and adapt to adverse impacts. For example, a healthcare system that is effective in delivering services to a population, combined with public health programmes that promote preventive behaviours, disease monitoring and disease vector control, can substantially limit the risk that climate change would cause widespread and persistent epidemics. Disaster prevention, preparedness, early warning and response systems can similarly help to limit the extent of harm from changes in the frequency or severity of extreme climate events. Poverty reduction can provide households with access to all manner of resources that can help them to cope with and overcome climate-related impacts. These and other examples indicate that improving the performance of human systems can reduce vulnerability. Doing so can yield near-term payoffs, as we improve our management of existing climate risks, as well as the longerterm benefits associated with building resilience to a changing climate. But optimism should be tempered by the reality of how challenging it has been to achieve even minimal progress where key human systems have been dysfunctional.
3243 J&J Vulnerability
1/11/07
10:22 am
Page 29
For Whom the Bell Tolls 29
References Arnell, N. W. (2004) ‘Climate change and global water resources: SERES emissions and socioeconomic scenarios’, Global Environmental Change, vol 14, pp31–52 Aron, J. L. and J. A. Patz (2001) Ecosystem Change and Public Health: A Global Perspective, John Hopkins University Press, Baltimore, MD Callaway, J., D. Louw, J. Nkomo, M. Hellmuth and D. Sparks (2006) ‘The Berg river dynamic spatial equilibrium model: A new tool for assessing the benefits and costs of alternatives for coping with water demand growth, climate variability and climate change in the Western Cape’, AIACC Working Paper No 31, International START Secretariat, Washington, DC, www.aiaccproject.org Camilloni, I. A. and V. R. Barros (2003) ‘Extreme discharge events in the Parana river and their climate forcing’, Journal of Hydrology, vol 278, pp94–106 Donne, J. (1623) Devotions Upon Emergent Occasions, Meditation No 17, available at www.incompetech.com/authors/donne/bell.html Downing, T. E. (2002) ‘Linking sustainable livelihoods and global climate change in vulnerable food systems’, Die Erde, vol 133, pp363–378 Githeko, A. K., S. W. Lindsay, U. E. Confaloniero and J. A. Patz (2000) ‘Climate change and vector-borne disease: A regional analysis’, Bulletin of the World Health Organization, vol 78, no 9, pp1136–1147 Hoerling, M., J. Hurrell, J. Eischeid and A. Phillips (2006) ‘Detection and attribution of 20th century northern and southern African monsoon change’, Journal of Climate, vol 19, pp3989–4008 IPCC (2001a) Climate Change 2001: The Scientific Basis, J. T. Houghton, Y. Ding, D. J. Griggs, M. Noguer, P. J. van der Linden, X. Dai, K. Maskell and C. A. Johnson (eds), contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York IPCC (2001b) Climate Change 2001: Impacts, Adaptation and Vulnerability, J. J. McCarthy, O. F. Canziani, N. A. Leary, D. J. Dokken and K. S. White (eds), contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York IPCC (2007) ‘Summary for policymakers’, in M. Parry, O. Canziani, J. Palutikof and P. van der Linden (eds) Climate Change 2007: Impacts, Adaptation and Vulnerability, contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York Kilian, A., P. Langi, A. Talisuna and G. Kabagambe (1999) ‘Rainfall pattern, El Niño and malaria in Uganda’, Transactions of the Royal Society of Tropical Medicine and Hygiene, vol 93, pp22–23 Leary, N. (2002) ‘AIACC, contributing to a second generation of climate change assessments’, START Network News, no 7, May Leary, N., J. Adejuwon, V. Barros, I. Burton, J. Kulkarni and J. Pulhin (eds) (2008) Climate Change and Adaptation, Earthscan, London Lindblade, K., E. Walker, A. Onapa, J. Katunge and M. Wilson (1999) ‘Highland malaria in Uganda: Prospective analysis of an epidemic associated with El Niño’, Transactions of the Royal Society of Tropical Medicine and Hygiene, vol 93, pp480–487 Mataki, M., K. Koshy and V. Nair (2006) ‘Implementing climate change adaptation in the Pacific islands: Adapting to present variability and extreme weather events in Navua, Fiji’, AIACC Working Paper No 34, International START Secretariat, Washington, DC, www.aiaccproject.org
3243 J&J Vulnerability
1/11/07
10:22 am
Page 30
30 Climate Change and Vulnerability Mataki, M., K. Koshy, R. Lata and L. Ralogaivau (2005) ‘Vulnerability of a coastal township to flooding associated with extreme rainfall events in Fiji’, unpublished working paper, University of the South Pacific, Suva, Fiji McMichael, A., A. Githeko, R. Akhtar, R. Carcavallo, D. Gubler, A. Haines, R. S. Kovats, P. Martens, J. Patz and A. Sasaki (2001) ‘Human health’, in J. J. McCarthy, O. F. Canziani, N. A. Leary, D. J. Dokken and K. S. White (eds) Climate Change 2001: Impacts, Adaptation and Vulnerability, Cambridge University Press, Cambridge, UK, and New York MEA (2005) Ecosystems and Human Well-being: Synthesis, Millennium Ecosystem Assessment, Island Press, Washington, DC Sheppard, C., D. Dixon, M. Gourlay, A. Sheppard and R. Payet (2005) ‘Coral mortality increases wave energy reaching shores protected by reef flats: Examples from the Seychelles’, Estuarine, Coastal and Shelf Science, vol 64, pp223–234 UNCCD (2005a) ‘Fact Sheet 3: The consequences of desertification’, www.unccd.int/publicinfo/factsheets UNCCD (2005b) ‘Fact Sheet 1: An introduction to the United Nations Convention to Combat Desertification’, www.unccd.int/publicinfo/factsheets
3243 J&J Vulnerability
1/11/07
10:22 am
Page 31
Part II:
Natural Resource Systems
3243 J&J Vulnerability
1/11/07
10:22 am
Page 32
3243 J&J Vulnerability
1/11/07
10:22 am
Page 33
2
Vulnerability of Southern African Biodiversity to Climate Change Graham P. von Maltitz and Robert J. Scholes
Introduction Southern Africa, and in particular South Africa, is home to a vast variety of endemic flora and fauna which, besides its intrinsic value, is vital for sustaining human livelihoods as well as many formal and informal economic sectors. Seven broad terrestrial ecological zones (biomes) occur in this region: the savanna, grassland, nama karoo, succulent karoo, forest, desert and fynbos biomes (Rutherford and Westfall, 1994). Some authorities also recognize an eighth biome, the thicket biome (Low and Rebelo, 1996). Of these, the fynbos biome in the Cape region and succulent karoo biome in the succulent karoo region possess particularly high levels of plant biodiversity and are two of the eight centres of plant endemism identified within southern Africa (Cowling and Hilton-Taylor 1994; Cowling and Hilton-Taylor, 1997). The biomes of southern Africa are characterized by unique climatic parameters, though edaphic (soil related) factors are also important in defining the habitats of individual species (Rutherford and Westfall, 1994). Early assessments using simple biome-level climatic envelope models have confirmed that global climatic change is likely to severely impact future biome distribution patterns (Scholes, 1990; Rutherford et al, 1999), with existing biomes likely shifted eastwards and compacted into the eastern half of the country (Figure 2.1). Based on projections of the HadCM2 (including and excluding sulphates) and CSM scenarios, by 2050, 38–55 per cent of the current area of South Africa would have a climatic envelope that did not match any current biome. Consequently, future climate conditions will possibly produce species assemblages with no counterpart in present communities (Hannah et al, 2002) and wide-scale species loss is also probable (see, for example, Thomas et al, 2004). The succulent karoo biome is of particular concern since it would likely be almost entirely replaced by an area with hotter and dryer conditions (Rutherford et al, 1999). In addition, the main area of its future distribution is predicted to be disjunct from its area of current distribution, being separated
3243 J&J Vulnerability
1/11/07
10:22 am
Page 34
34 Climate Change and Vulnerability
Figure 2.1 Current and future predicted distributions of the major South African biomes (fynbos, succulent karoo, nama karoo, savanna and grassland), generated using simple bioclimatic modelling techniques based on five ecologically important bioclimatic parameters Source: Rutherford et al (1999).
by a range of mountains. The extents of other biomes – savanna, grasslands and nama karoo – are also reduced, with both the savanna and nama karoo invading into some grassland areas. The fynbos is the only biome to show limited changes in total extent, but there is evidence that many individual species may be impacted. The forest biome is very small and dispersed, and the most likely scenario is that only its distribution in the north-eastern part of the country would be impacted. Very little desert is presently found in South Africa, though its extent may increase in future due to desertification caused by climate change. In order to better understand the potential impacts of climate change on biodiversity in southern Africa, we selected three of these biomes – fynbos,
3243 J&J Vulnerability
1/11/07
10:22 am
Page 35
Vulnerability of Southern African Biodiversity to Climate Change 35
succulent karoo and savanna – as pilot study sites primarily on the basis of the following important characteristics: •
• •
Though the spatial extent of the fynbos is less likely to be reduced as much as other biomes, it is very important in terms of its biodiversity and there is significant concern about individual species response to climate change within this biome. The succulent karoo, which has a unique biodiversity, is likely to experience some of the most severe impacts from climate change based on the results of previous modelling studies. The savanna is a widespread biome in Africa and is very important from a livelihood perspective. It possesses a unique co-dominance of tree and grass lifeforms, which makes it perfect for studying systems-level responses of functional groups of plants to climate change.
Furthermore, data on the three biomes are also readily available from previous and ongoing work in these areas, which facilitates the analysis of their vulnerability to climate change.
Description of the Biomes Studied Fynbos Fynbos, literally meaning fine-leaved bush, is a local term for the heath-like vegetation found in areas of South Africa that have a Mediterranean-type climate with winter rainfall (Cowling et al, 1997). Evergreen, fire-prone vegetation on a low-nutrient substrate is a key feature of the fynbos. It is largely confined to the rugged and steep quartzitic Folded Mountains in the Cape Floristic region1 located in the south-western tip of South Africa and occupies an area of about 71,000km2. This biome has exceptionally high species richness, with an estimated 7300 species, of which 80 per cent are endemic (Cowling and Hilton-Taylor, 1994 and 1997). Altitude, rainfall, aspect and soil are important determinants of vegetation structure, while stochastic factors such as fire frequency and intensity play an important role in determining species composition (Cowling et al, 1997). There is a strong dependency on insects as pollinators and some species depend on ants as seed dispersers (myrmecochory). Others depend on the wind or are passively dispersed. Most species have specific strategies to ensure regeneration from seeds after fires, although a few species can also resprout following fire. Species that do not resprout after fire but are limited by seeds for recruitment face a high risk of extinction if their pollinator or disperser mutualisms collapse (Bond, 1984). With respect to species composition, approximately 40 per cent of the fynbos plant species are comprised of the highly speciated ericoid family. It is, however, the Proteaceae family that makes up the bulk of the shrub biomass and holds ecological and economic importance. The standing biomass of the Proteaceae is a major driver of the fire regime and though a few protioids can resprout after fire, the majority are obligatory reseeders. This is an important
3243 J&J Vulnerability
1/11/07
10:22 am
Page 36
36 Climate Change and Vulnerability
factor when considering climate change impacts on the fynbos (Bond and Breytenbach, 1985).
Succulent karoo The succulent karoo biome is an area of about 100,000km2 that stretches along the west coast of southern Namibia and South Africa and as a narrow band across the northern edge of the fynbos. It is found predominantly on coastal plains and intermountain valleys, mostly at an altitude of less than 1000m. The biome receives only 20–290mm of annual rain, more than 40 per cent of which falls in winter. It is the high summer aridity and the finer grained and more nutrient-rich soils that differentiate this area from the fynbos. The biome has more than 5000 plant species, of which about 50 per cent are endemic (Milton et al, 1997). The unique feature of the karoo vegetation is the high concentration of leaf-succulents of the two dominant families, Mesembryanthemaceae and Crassulaceae. Most species are insect pollinated and seed dispersal occurs mostly by wind or water, with long-lived perennials and short-lived annuals having different seeding strategies. The annuals use a number of mechanisms to delay germination to periods of suitable rainfall and different species respond to different timing of rain. Generally, annuals have a high proportion of seedlings reaching reproductive maturity, whereas long-lived species tend to have a very low annual recruitment rate (Milton, 1995; Rosch, 1977; van Rooyen et al, 1979). Disturbance of the landscape also promotes the establishment of annuals. The re-establishment of long-lived plants onto disturbed areas is, in contrast, very slow and may take up to 80 years (Beukes et al, 1994; Dean and Milton, 1995). Fire is not a feature in the succulent karoo but the vegetation is susceptible to drought, which can kill off many of the perennials. Once again, annuals tend to recover rapidly from drought while perennials recover very slowly (Milton 1995).
Savanna The co-dominance of trees and grass is what distinguishes the savanna from other biomes. Savannas cover 54 per cent of the land area of southern Africa and 60 per cent of sub-Saharan Africa. They are found in predominantly frost-free regions with a moist hot season (Scholes and Walker, 1993; Scholes, 1997). The species richness of the Savanna is second only to the fynbos in southern Africa, with about 5780 species in an area of about 600,000km2. Although plant diversity at the small plot scale (‘alpha diversity’) is high, the species turnover (‘beta diversity’) and landscape (‘gamma’) diversity are generally low (Scholes, 1997). In other words, adjacent patches tend to have very similar species. This is in sharp contrast to the fynbos, which has high beta and gamma diversity. The savanna grasses largely depend on animals or wind for seed dispersal, whereas trees have ballistic dispersal, wind dispersal and animal dispersal. The herbaceous layer is a mix of annual and perennial forbs2 and grass species, with annuals being more dominant in arid areas. Trees can be exceptionally long
3243 J&J Vulnerability
1/11/07
10:22 am
Page 37
Vulnerability of Southern African Biodiversity to Climate Change 37
lived (a few centuries in some species) and many species are able to resprout if damaged by animals or fire. Fire is a common occurrence, and its frequency and intensity are largely governed by rainfall and herbivory. Fire and herbivory play an important role in determining the proportional representation of trees and grasses in the savanna (Scholes, 1997). The savannas are home to most of Africa’s large mammals and support an important tourism industry. They also house a large livestock ranching industry. Historically, the savannas have supported most of the indigenous human populations of Africa and the area still plays an important role in supporting subsistence economies through the supply of grazing, fuelwood and other natural resources (Scholes, 1997).
Research Framework and Methodology Three different approaches were developed to assess the vulnerability of biodiversity to climate change in the succulent karoo, fynbos and savanna. We also attempted to determine the potential impacts of this vulnerability on biodiversity-dependent human communities in these biomes. For the fynbos biome a ‘time-slice’ modelling approach was used to investigate the migratory corridors required for individual species to track climate change (for details see Midgley et al, 2006; Williams et al, 2004). The Proteaceae were used as representative species because of the ready availability of detailed distribution data and their economic importance and because their distributions have been shown to be closely correlated with the distributions of other fynbos species. Species distribution data were obtained from the Protea Atlas Project database (http://protea.worldonline.co.za/default.htm). Climate data were interpolated for the one-minute grid (Schulze et al, 1997) and future projections were based on the HadCM2 general circulation model, using IS92a emissions scenarios (Schulze and Perks, 1999). Soil categorization, species nomenclature and species dispersal modes were some of the other data obtained. In the karoo case study, a simpler approach was taken to examine the link between a single animal species and its habitat and food resources. Two species, the highly endangered riverine rabbit (Bunolagus monticularis) from the nama karoo and the padloper tortoise (Homopus singnatus) from the succulent karoo (for details see Hughes et al, 2005a and b) were used for this purpose. The savanna study, on the other hand, investigates key functional properties – tree cover, fire frequency, grass and browse production, and carrying capacity – for major guilds of herbivores and carnivores, rather than individual species interaction. A model based on the Lotka–Volterra approach (Lotka, 1925; Volterra, 1926) was developed for this purpose (Scholes, 2005) and attempts were made to predict the functional responses of grasses and trees to changes in temperature, rainfall and CO2 levels.
3243 J&J Vulnerability
1/11/07
10:22 am
Page 38
38 Climate Change and Vulnerability
Vulnerability of Different Biomes and Species to the Impacts of Climate Change The primary factors affecting the vulnerability of individual species are the magnitude of climate impacts and the adaptive capacity of species. Different global circulation models (GCMs) and emission scenarios result in a range of projections for future climate in southern Africa. Common to all is an increase in temperature, but the magnitude of this change varies from under 1°C to up to 3.5°C for 2080 (Scholes and Biggs, 2004). There is less agreement between GCMs regarding the impacts of climate change on rainfall. Most models project reduced rainfall in the western half of southern Africa, though there may be increased rainfall in the eastern subcontinent (Scholes and Biggs, 2004). From the perspective of plants, an increase in temperature combined with reduced rainfall greatly decreases the available moisture for growth. In other words the ratio of rainfall to potential evaporation tends to decrease, resulting in more desert-like conditions. The fertilization effect of increased CO2 may, to some extent, offset the impacts of this increased aridity. The fynbos as a biome appears less affected by climate change than many of the other southern African biomes (Rutherford et al, 1999). In a study of climate impacts using the Proteaceae as an indicator group, Midgley et al (2002a) reported that 17 of the 28 Protea species examined would experience potential range contraction, while 5 species would likely experience range elimination. Changes in climatic factors were predicted to have a greater impact than land transformation as the Proteas tend to move to areas of higher altitude in the mountains that are not likely to be impacted by land transformation. This study was repeated for 330 Protea species (Midgley et al, 2002b), with similar results. A loss of 51 per cent to 65 per cent of the area extent of the fynbos biome was predicted, depending on the climate scenario used. One third of the Protea species were found to have complete range dislocation (in other words there is no overlap between their future range and their current range). Only 5 per cent would retain more than two thirds of their current range, while 10 per cent of the species would have no predicted range in the future scenario. To further examine Proteaceae responses to climate change we developed a dynamic movement model to determine whether the Proteas could disperse from their current distribution to predicted 2050 habitats (Williams et al, 2004). The model assumed a shorter migration range for Protea species with ant-dispersed seeds rather than those whose seeds are wind dispersed. A time-sliced approach was used to see whether suitable ‘stepping stone’ habitats existed to allow the Protea to move from their current distribution to their future distribution. The study found that, of the 282 species investigated, 262 maintained overlapping habitats and did not need to disperse (though their range may be reduced), 18 species were obligatory dispersers (those that would have to move to new habitats) and could reach their new habitats, while 2 species were unable to disperse because of land transformation. Thirty-four species were removed from the analysis, as they had no future habitat and would therefore likely become extinct unless ex situ conservation was initiated.
3243 J&J Vulnerability
1/11/07
10:22 am
Page 39
Vulnerability of Southern African Biodiversity to Climate Change 39
These detailed studies on the Proteas are unique and other genera of the fynbos have not been subjected to similar scrutiny. Although Proteas may serve as an example to represent general trends, they only account for a small proportion of the actual diversity, and the rest still need more detailed analysis. Additionally, the many assumptions and uncertainties associated with climatic envelope modelling in general and the dynamic movement models in particular have not been taken into account in the current results. The succulent karoo biome is expected to suffer the greatest impact from climate change, losing almost all of its current range according to studies by Rutherford et al (1999). The small area of future range is projected to be disjunct from the current range, separated by mountain escarpments (Rutherford et al, 1999). In the Namibian portion of the succulent karoo, many species are also highly localized endemics that probably lack the ability for long-distance migration, even though their projected future habitats may bear similarities to their present habitats. So far a decline has already been observed in the population of Aloe dicotima, one of the most conspicuous species of the succulent karoo (Foden, 2002). Among animals in the succulent karoo, the padloper tortoise will likely adapt to the impacts of climate change because it has a broad range of plant food sources growing over a widespread area. In contrast, the riverine rabbit from the nama karoo is predicted to become extinct because its specialized habitat and food requirements are unlikely to be met under future climate conditions (Hughes et al, 2005a and b). The core of the savanna case study was an investigation of the dynamics of trees and grasses, given climate change. Trees and grasses have different responses to both changes in soil moisture and temperature, and increased atmospheric CO2 concentrations, which tend to have a growth promoting effect. The CO2 effect is more pronounced in trees than in tropical grasses, but in both cases begins to saturate at around 500ppm under natural conditions (Scholes et al, 1999). According to preliminary model runs for the northeastern lowveld savanna, the negative impacts of the decrease in soil moisture and increase in temperature would more than compensate for the small advantage trees have from elevated CO2 levels. A slight increase in woodiness is, however, predicted. It is also predicted that habitat suitability for browsers and grazers is likely to remain relatively constant in the 50-year timeframe, provided appropriate management of fire and elephant population, the key controls of future habitat structure. Overall, the carrying capacity for large herbivores is projected to decrease by about 10 per cent.
Other Factors Likely to Influence Ecosystem and Species Vulnerability Flatness of topography Altitudinal gradients and aspect both affect local climate conditions. In hilly topography with steep environmental gradients, plants and animals would only
3243 J&J Vulnerability
1/11/07
10:22 am
Page 40
40 Climate Change and Vulnerability
have to move short distances to find new habitats (Cowling et al, 1999). For example, Protea species are expected to move up altitudinal environmental gradients in response to climate change (Williams et al, 2004). A potentially negative consequence of such higher altitude ‘escapes’ is that species may be trapped on these ‘islands’ and their dispersal into new areas might be prevented. By contrast, on a flat topography like that of the succulent karoo, the horizontal rate of climatic envelope shift would be more rapid, making it extremely difficult for plants to disperse fast enough to track a changing climate. While this may not be a problem for widespread species adapted to long-distance seed dispersal and rapid establishment, it is likely to be devastating for species with localized distributions, short-range seed dispersal or slow rates of establishment (in other words many of the long-lived perennial species).
Availability of refugia Small climatically suitable refugia could potentially support remnant populations of a species in an otherwise changed climate; for example, a cooler and moister south slope of a mountain (in comparison to a hot north slope) or a deep river gorge that provides moisture and temperature regulation and protection from fire. The fynbos biome and the savanna biome are more likely to have a number of refugia options than the karoo biome.
Edaphic barriers Most plant species are adapted to specific soil (edaphic) conditions and are unlikely to be able to move over barriers of unsuitable soils. Similarly, the large mammals and birds of the savanna biome tend to be either ‘fertile soil’ specialists or ‘infertile soil’ specialists. Factors such as texture, water holding capacity, nutrient status and acidity are likely to be the most important in terms of soil suitability. An example is Colophospernum mopane, a common species of the low-lying and hot savanna areas which is predicted to expand extensively with global warming on the basis of its climate envelopes (Rutherford et al, 1999) but which may, in practice, show little or no response as the species appears to be limited to the heavy soils of river valleys (Scholes, 1997). Similarly, mountains are likely to form barriers, especially if the substrate on the mountains greatly differs from that of the surrounding plains, as is the case of the Cape Folded Mountains, which separate the existing succulent karoo from areas of future suitable habitat.
Land transformation and habitat fragmentation The combined impacts of habitat loss due to land transformation (into cropland) and climate change will make some species exceptionally vulnerable. Land transformation is typically most severe in the relatively flat areas suitable for crop production (Haplin, 1997). The west coast renosterveld3 of the fynbos, for instance, has already lost an estimated 97 per cent of its habitat (Low and
3243 J&J Vulnerability
1/11/07
10:22 am
Page 41
Vulnerability of Southern African Biodiversity to Climate Change 41
Rebelo, 1996) and just 100m of cropland may be sufficient to prevent dispersal of some Protea species (Williams et al, 2004).
Life history characteristics A number of life history characteristics of species are likely to influence their vulnerability to climate change, more so in the case of plants than animals. Some of the key characteristics are discussed below: •
•
•
•
•
•
Niche specialists versus generalists: Species with broad climate tolerances are less likely to be impacted than specialist species with narrow habitat niches. For example, a large number of Proteas as well as the padloper tortoise are likely to maintain habitats into the future while a few Proteas and the riverine rabbit would likely lose their habitat niche (see Williams et al, 2004; Midgley et al, 2002a and b; Hughes et al, 2005a and b). Species movement: In plants, the mechanism of seed dispersal will determine the distance that species can migrate per generation. This can be hundreds of metres or even kilometres where wind, animals, birds or water are the dispersal agents. Many seeds lack long-distance dispersal mechanisms and tend to concentrate very close to the parent plant. For instance, many of the Proteas have ant-dispersed (myrmecohory) seeds (Bond and Slingsbey, 1983), which limits the distance they can move to about a few metres per generation (Bond, 1984). In the case of animals and insects, some birds and large mammals may travel hundreds of kilometres, while others may be limited because of mobility or reluctance to cross changed habitats. Samango monkeys, for example, though highly mobile, will not cross extensive open ground to move from one forest patch to another. Species interactions – competition and facilitation: Species do not move as total habitats (Hannah et al, 2002); rather, individual species move at different rates. In a new area, a species is likely to experience new competition from those species already present. The impacts of this are exceptionally difficult to understand or model. Dependency on pollinators and dispersers: Where species have a mutual dependency on a pollinator or seed disperser, both organisms would have to adapt simultaneously to climate change for their long-term sustainability. Obligatory pollinators and seed dispersers are common in the fynbos and succulent karoo, but relatively rare in the savanna (see, for example, Milton et al, 1997; Cowling et al, 1997). The risk is far less where there are generic pollinators such as the honeybee. Species establishment: The establishment niche (Grubb, 1977) will be a key limiting process for many plant species because, once established, many species can cope with slightly modified climatic niches (for example, the numerous garden trees that grow way outside of their natural climatic niche). Additionally, plants that require many years to mature and fruit may be more vulnerable to climate impacts than annual species. Species longevity: It is probable that many established plant species might be able to persist in a changed climate, especially long-lived species that
3243 J&J Vulnerability
1/11/07
10:22 am
Page 42
42 Climate Change and Vulnerability
•
might persist for centuries and possibly regenerate if climate mitigation is successful and optimal conditions return. Short-lived species are likely to be more severely impacted and could become extinct within a few generations. Seed ecology: Some plant species require a cool period to stimulate germination and climatic warming could negatively affect their ability to germinate. In the succulent karoo, plant species appear to be dependant on seasonal rainfall for germination.
Impacts of alien invasive species Alien invasive species pose a strong threat to some biomes, especially the fynbos. Any indigenous species attempting to establish itself in a new habitat may have to compete with alien vegetation simultaneously moving into the new niche. It has been suggested that invasive alien species may invade more rapidly under conditions of climatic change, as they are frequently opportunistic species adapted to a wide range of habitats (Macdonald, 1994).
Impacts of fire Both the fynbos and savanna are fire-dependent ecosystems (Cowling et al, 1997; Rutherford and Westfall, 1994; Scholes, 1997), and any changes in fire frequency or intensity would change the dynamics of these systems. On the other hand, the succulent karoo is not adapted to fire and its spread into habitats where fire occurs would therefore be limited.
Impacts of misaligned conservation strategies Strategic conservation planning typically attempts to ensure that all current habitat types are conserved in at least one location. It does not take into consideration possible impacts of climatic change that will result in new species assemblages and changes in the spatial location of at least some species. The current conservation network may thus prove inadequate for protecting biodiversity in the future. It may be poorly aligned to facilitate the movement of biodiversity and may not provide sufficient protection of important areas that could serve as biodiversity refugia. Therefore, a better solution would be to protect gradients and transitional areas in addition to core areas (Peters and Darling, 1985; Haplin, 1997; Hannah et al, 2002; Williams et al, 2004). This would require a strategic expansion of the conservation network, including enhancing conservation outside of protected areas (also referred to as managing the matrix). This has been studied in detail only in the case of the fynbos biome (Williams et al, 2004), where it was determined that for effective conservation, a doubling of the current conservation network is needed. The conservation of ‘persistence areas’, or areas that remain suitable over time despite climate change, is also recommended. Of the three biomes considered, species in the succulent karoo are likely to be more vulnerable to climate change than those of the savanna or fynbos,
3243 J&J Vulnerability
1/11/07
10:22 am
Page 43
Vulnerability of Southern African Biodiversity to Climate Change 43
primarily because of the potential extent of climate impacts on this habitat, but also because of other factors that compound climate change impacts such as the relatively flat topography, edaphic conditions, specialist pollinators and the slow rate of re-establishment of long-lived perennial species. Impacts on the fynbos will, to some extent, be mitigated by the mountainous terrain, though species loss is still probable. The high level of land transformation in the lowland fynbos habitats makes species in these habitats especially vulnerable. In contrast, the savannas, as a system, are considered less vulnerable to climate change, with most functional aspects likely to be maintained. Direct impacts on individual species are less certain and still unexamined. According to our study on the vulnerability of biodiversity in the savanna, the widespread and generalist nature of most savanna species is likely to result in less severe impacts when compared to the fynbos and karoo biomes.
Human Dependencies on Biodiversity In addition to the intrinsic value of biodiversity, there are also direct and indirect benefits that contribute to human livelihoods. So far the value of biodiversity to human well-being has been poorly researched and therefore we have made an initial attempt to predict the likely impacts of climate-induced biodiversity change on livelihoods in our three case study biomes.
Succulent karoo This near-desert area has low plant production potential and a very low human population density. It has historically been used prominently for livestock grazing, though even with borehole water it is still an inhospitable environment. Livestock stocking rates have decreased over the last couple of decades and this is attributed to species changes in response to herbivory (Dean and McDonald, 1994; Milton et al, 1997). It can be assumed that with increasing aridity, as predicted by most climate models, this area will become less suited to livestock production. Changes in the global economics of livestock production are likely to exacerbate this impact. The succulent karoo is also an increasingly popular ecotourism spot during its annual periods of wild flower blooms. These blooms are all from annual species, mostly of the Asteraceae, Liliaceae and Mesembryanthemaceae families, and are dependent on seasonal rainfall. A decline in rainfall or shift in seasonality of rainfall may therefore have a direct impact on the tourism industry.
Fynbos In contrast to the succulent karoo, the fynbos is an area of high human population in the plains, while the mountains have a relatively low population density. Land transformation for agriculture and settlement has had strong impacts on biodiversity, but natural biodiversity per se is not the basis of the
3243 J&J Vulnerability
1/11/07
10:22 am
Page 44
44 Climate Change and Vulnerability
main economic activity in the biome. There are, however, individual species of the fynbos that are extensively used for the cut flower and dry flower industry, for flavouring brandy (buchu) and for herbal tea (rooibos tea). The vegetation is also used for fuelwood, though much of this is from alien invasive species. Any changes in the availability and location of such commercial species would directly impact these industries. In addition, the fynbos is also a major tourist attraction, though the relationship between tourism and species biodiversity is unclear and is unlikely to be linear.
Savanna The savanna biome is very important in terms of livelihood benefits, especially for the rural poor (see, for example, Campbell, 1996; Shackleton et al, 2004) and is used predominantly for livestock grazing and wildlife management, including nature-based tourism. Within South Africa, most of the communally managed areas occur in the savannas and these house extensive rural communities that depend on the natural biodiversity of the savanna for a large proportion of their livelihoods. This trend is even more prominent in other southern African countries where there is a smaller cash economy and a poorer state social support network. By far the most important resources from the savannas to these subsistence communities are fuelwood and grazing, although many other resources, including construction timber, edible plants, medicinal plants and craft material, are also collected. In the future it is likely that the range of services derived from savannas would still be maintained (possibly at a somewhat lower level), though there may be a change in specific species. However, the change in climate may cause these areas to become less suitable for crop production and hence reliance on natural products may increase. The wildlife-based tourism industry in the savannas is expected to survive the impacts of climate change, as is the livestock industry, although climate change may accelerate a switch from livestock to wildlife management. This is due to the greater resilience of wildlife to a hot environment and also the possibility of increased pathogen infestations in livestock due to climate change.
Conclusion The vulnerability of individual species to climate change depends on the interaction between the magnitude of the impact and the adaptive capacity of the species. Of the three South African biomes considered, the fynbos and succulent karoo are predicted to be the most vulnerable to projected climate changes in the 21st century, while the savanna is predicted to be more resilient. The plant species-rich succulent karoo is at risk partly because of the magnitude and nature of the projected climate change (in the most extreme scenarios, its current climate niche is not represented in the future) but also because of the relatively flat topography, which does not provide altitudinal or aspect-based refugia. This means that species need to disperse very rapidly to keep track with a changing environment. Their capacity to do so is constrain-
3243 J&J Vulnerability
1/11/07
10:22 am
Page 45
Vulnerability of Southern African Biodiversity to Climate Change 45
ed by their strong soil specificity in this geologically complex location. Additionally, the slow establishment rates of many long-lived species places them at greater risk than annual species with more rapid establishment and greater seed dispersal. Not only is there a large probability of extensive species loss in the succulent karoo, but the biome’s ability to support livelihoods is also expected to decline. Within the fynbos, the extent of species loss may exceed that predicted from habitat loss alone but it is expected to be less severe than in the succulent karoo. The complex reproduction ecology of many of the species adds to their vulnerability and this is not taken into consideration in simple climatic envelope-based modelling. Species loss is an important concern given that this biome is an international biodiversity hotspot. The mountainous terrain will, to some extent, assist in adaptation to climate change. Economic sectors dependent on the biodiversity in this biome, such as the wild flower industry, may suffer severe impacts from climate change. A key consideration in both the fynbos and succulent karoo biomes is the close and obligatory association between species; for instance, between plant species and their pollination and seed dispersal agents. The entire suite of organisms would have to move jointly in response to a changing environment for the long-term integrity and viability of the community and its constituents. The current knowledge base and models are unfortunately inadequate to analyse future climate impacts on these relationships. The savanna biome is predicted to have relatively limited functional or species change. Though the biome overall is species-rich, the mean range size of the organisms is large (in other words the same species are found over extensive areas and across a wide climate niche). The diverse mammals and birds for which this biome is world-renowned are relatively mobile and not highly dependent on particular plant species, but on broad functional categories of plants, which are projected to persist. There is some evidence that enhanced CO2 may promote C3 trees over C4 grasses4, but our models suggest that this effect may be overwhelmed by differences in temperature and water response functions between trees and grasses. Coupled with changing fire regimes and elephant densities, these impacts could result in a structurally changed savanna, which would, in turn, have consequences on habitat structure and species proportions in the community, though not necessarily their persistence. The ability of the savanna biodiversity to support rural subsistence livelihoods or the wildlife and tourism industry is not expected to change very substantially, although climate change may accelerate the current shift from cattle to wildlife ranching. From a human vulnerability perspective, it is the savannas that have the highest density of human settlements and it is in these areas where there is a high dependency on the use of savanna resources, particularly fuelwood, but also numerous other livelihood-enhancing products. These areas are also exceptionally important for nature-based tourism and wildlife-based production. In contrast, the fynbos and succulent karroo have a lower level of subsistence livelihood dependency, though a number of enterprises have developed around the commercial harvesting of particular plant species such as Proteaceae for the flower market and herbal teas and brandies.
3243 J&J Vulnerability
1/11/07
10:22 am
Page 46
46 Climate Change and Vulnerability
Notes 1 2 3
4
The Cape Floristic region is a biodiversity hotspot with Mediterranean-type climate and is one of the only two hotspots that encompass an entire floral kingdom (for details see www.biodiversityhotspots.org/xp/Hotspots/cape_floristic/). Forbs are non-grass, broad-leaved plants with little or no woody material and are usually found on fields, prairies or meadows. This group includes most herbs, vegetables, and wild and garden flowering plants. Renosterveld vegetation is a type of fynbos that dominates on the flatlands of the western Cape. It is dominated by the daisy family (Asteraceae) and has low levels of typical fynbos genera such as Proteas, Ericas and Restios, It has been extensively cleared to make way for agricultural crops and housing development (Low and Rebelo, 1996). C3 plants produce a three-carbon compound during photosynthesis and include most trees and common crops like rice, wheat, barley, soybeans, potatoes and vegetables. The C4 plants produce a four carbon compound during photosynthesis and include grasses and crops like maize, sugar cane, sorghum and millet. Under increased atmospheric concentrations of CO2, C3 plants have been shown to be more responsive than C4 plants (see IPCC, 2001).
References Beukes, P. C., R. M. Cowling and F. Ellis (1994) ‘Vegetation and soil changes across a succulent karoo grazing gradient’, Arid Zone Ecology Forum Abstracts, Foundation for Research Development, Pretoria, South Africa, p23 Bond, W. (1984) ‘Fire survival of Cape Proteaceae: Influence of season and seed predation’, Vegitatio, vol 56, pp65–74 Bond, W. and G. J. Breytenbach (1985) ‘Ants, rodents and seed predation in Proteaceae’, South African Journal of Zoology, vol 20, pp150–155 Bond, W. and P. Slingsbey (1983) ‘Seed dispersal by ants in shrublands of the Cape Province and its evolutionally implications’, South African Journal of Science, vol 79, pp213–233 Campbell, B. (ed) (1996) The Miombo in Transition: Woodlands and Welfare in Africa, CIFOR, Indonesia Conservation International (no date) ‘Cape Floristic region, Biodiversity hotspots’, www.biodiversityhotspots.org/xp/Hotspots/cape_floristic/ Cowling, R. M. and C. Hilton-Taylor (1994) ‘Patterns of plant biodiversity and endemism in southern Africa: An overview’, in B. Huntley (ed) Botanical Diversity in Southern Africa, National Botanical Institute, Pretoria, South Africa Cowling, R. M. and C. Hilton-Taylor (1997) Phytogeography, Flora and Endemism, in R. M. Cowling, D. M. Richardson and S. M. Pierce (ed) Vegetation of Southern Africa, Cambridge University Press, Cambridge, UK Cowling, R. M., R. L. Pressey, A. T. Lombard, P. G. Desmet and A. G. Ellis (1999) ‘From representation to persistence: Requirement for a sustainable reserve system in the species-rich Mediterranean-climate deserts of southern Africa’, Diversity and Distributions, vol 5, pp1–21 Cowling, R. M., D. M. Richardson and P. J. Mustart (1997) ‘Fynbos’ in R. M. Cowling, D. M. Richardson, and S. M. Pierce (eds) Vegetation of Southern Africa, Cambridge University Press, Cambridge, UK Dean, W. R. J. and I. A. W. McDonald (1994) ‘Historic changes in stocking rates of domestic livestock as a measure of semi-arid and arid rangeland degradation in the Cape Province South Africa’, Journal of Arid Environments, vol 26, pp281–298
3243 J&J Vulnerability
1/11/07
10:22 am
Page 47
Vulnerability of Southern African Biodiversity to Climate Change 47 Dean, R. and S. J. Milton (1995) ‘Plant and invertebrate assemblages on old fields in the arid southern karoo, South Africa’, African Journal of Ecology, vol 33, pp1–13 Foden, W. (2002) ‘A demographic study of Aloe dichotoma in the Succulent Karoo: Are the effects of climate change already apparent?’, Unpublished MSc thesis, Percy Fitzpatrick Institute of African Ornithology, University of Cape Town, Cape Town, South Africa Grubb, P. J. (1977) ‘The maintenance of species richness in plant communities: The importance of the regeneration niche’, Biological Reviews, vol 52, pp107–145 Hannah, L., G. F. Midgley and D. Millar (2002) ‘Climate change-integrated conservation strategies’, Global Ecology and Biogeography, vol 11, pp485–495 Haplin, P. N. (1997) ‘Global climate change and natural-area protection: Management responses and research directions’, Ecological Applications, vol 7, pp828–843 Hughes, G. O., W. Thuiller, G. F. Midgley and K. Collins (2005a) ‘A fait accompli? Environmental change hastens the demise of the critically endangered riverine rabbit (Bunolagus monticularis)’, Unpublished report, South African Botanical Research Institute, Cape Town, South Africa Hughes, G. O., V. J. T. Loehr, W. Thuiller, G. F. Midgley and T. E. J. Leuteritz (2005b) ‘Global change and an arid zone chelonian: The case of Homopus signatus’, Unpublished report, South African Botanical Research Institute, Cape Town, South Africa Lotka, A. J. (1925) Elements of Physical Biology, Williams and Wilkins, Baltimore, MD, US Low, B. and A.G. Rebelo (eds) (1996) Vegetation of South Africa, Lesotho, and Swaziland, Department of Environmental Affairs and Tourism, Pretoria, South Africa Macdonald, I. A. W. (1994) ‘Global change and alien invasion: Implications for biodiversity and protected area management’, in O. T. Solbrig, P. G. van Emden and W. J. van Oordt (eds) Biodiversity and global change, CAB International, Wallingford, UK IPCC (2001) ‘Glossary of terms’ in J. McCarthy, O. Canziani, N. Leary, D. Dokken and K. White (eds) Climate Change 2001: Impacts, Adaptation and Vulnerability, Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US Midgley, G. F., L. Hannah, D. Millar, W. Thuiller and A. Boot (2002a) ‘Developing regional species-level assessments of climate change impacts on biodiversity in the Cape Floristic region’, Biological Conservation, vol 112, pp87–97 Midgley, G., F. L. Hannah, D. Millar, M. C. Rutherford and L. W. Powrie (2002b) ‘Assessing the vulnerability of species richness to anthropogenic climate change in a biodiversity hotspot’, Global Ecology and Biogeography, vol 11, pp445–451 Milton, S. J. (1995) ‘Spatial and temporal patterns in the emergence and survival of seedlings in the arid karoo shrubland’, Journal of Applied Ecology, vol 32, pp145–156 Milton, S. J., R. I. Yeaton, W. R. J. Dean and J. H. J. Vlok (1997) ‘Succulent karoo’, in R. M. Cowling, D. M. Richardson and S. M. Pierce (eds) Vegetation of Southern Africa, Cambridge University Press, Cambridge, UK Peters, R. L. and J. D. Darling (1985) ‘The greenhouse effect and nature reserves’, Bioscience, vol 35, pp707–717 Rosch, M. W. (1977) ‘Enkele plantekologiese aspekte van die Haster Malan-natuurreservaat’, MSc thesis, University of Pretoria, Pretoria, South Africa Rutherford, M. C. and R. H. Westfall (1994) ‘Biomes of Southern Africa: An objective characterisation’, Memoirs of the Botanical Survey of South Africa, vol 63, pp1–94 Rutherford, M.C., G. F. Midgeley, W. J. Bond, L. W. Powrie, R. Roberts and L. Allsopp (1999) Plant biodiversity: Vulnerability and Adaptation Assessment, South African Country Study on Climate Change, National Botanical Institute, Cape Town Scholes, R. J. (1990) ‘Change in nature and the nature of change: Interactions between
3243 J&J Vulnerability
1/11/07
10:22 am
Page 48
48 Climate Change and Vulnerability terrestrial ecosystems and the atmosphere’, South African Journal of Science, vol 86, pp350–354 Scholes, R. J. (1997) ‘Savanna’ in R. M. Cowling, D. M. Richardson and S. M. Pierce (eds) Vegetation of Southern Africa, Cambridge University Press, Cambridge, UK Scholes, R. J. and R. Biggs (2004) ‘The regional scale component of the Southern African Millennium Ecosystem Assessment’, Millennium Ecosystem Assessment, CSIR, Pretoria, South Africa Scholes, R. J. and B. H. Walker (1993) An African Savanna: Synthesis of the Nylsvley Study, Cambridge University Press, Cambridge, UK Scholes, R. J., E. D. Schulze, L. F. Pitelka and D. O. Hall (1999) ‘The biogeochemistry of terrestrial ecosystems’, in B. H. Walker, W. L. Steffen, J. Canadell and J. S. I. Ingram (eds) The Terrestrial Biosphere and Global Change, Cambridge University Press, Cambridge, UK, pp88–105 Schulze, R. E., M. Maharaj, S. D. Lynch, B. J. Howe and B. Melvil-Thompson (1997) South African Atlas of Agrohydrology and Climatology, Water Research Commission, Pretoria, South Africa Schulze, R. E. and L. A. Perks (1999) Assessment of the Impact of Climate, School of Bioresources Engineering and Environmental Hydrology, University of Natal, Pietermaritzburg, South Africa Shackleton, C. M. and S. E. Shackleton (2004) ‘Use of woodlands resources for direct household provision’ in M. J. Lawes, H. A. C. Ealey, C. M. Shackleton and B. G. S. Geach (eds) Indigenous Forests and Woodlands in South Africa: Policies, People and Practice, University of Kwazulu-Natal Press, Scottsville, South Africa Thomas, C. D., A. Cameron, R. E. Green, M. Bakkenes, L. J. Beaumont, Y. C. Collingham, B. F. N. Erasmus, M. F. de Siqueira, A. Grainger, L. Hannah, L. Hughes, B. Huntley, A. S. van Jaarsveld, G. F. Midgley, L. Miles, M. A. OrtegaHuerta, A. T. Peterson, O. L. Phillips and S. E. Williams (2004) ‘Extinction risk from climate change’, Nature, vol 427, pp145–148 Van Rooyen, M. W., G. K. Theron and N. Grobbelaar (1979) ‘Phenology of the vegetation in Hester Malan Nature Reserve in Namaqualand Broken Veld: The Therophyte population’, Journal of South African Botany, vol 45, pp433–452 Volterra, V. (1926) ‘Variations and fluctuations of the number of individuals in animal species living together’, reprinted (in 1931) in R. N. Chapman (ed) Animal Ecology, McGraw-Hill, New York, US Williams, P., L. Hannah, S. Andelman, G. Midgley, M. Araujo, G. Hughes, L. Manne, E. Marinez-Meyer and R. Pearson (2004) ‘Planning for climate change: Identifying minimum-dispersal corridors from the Cape Proteaceae’, Conservation Biology, vol 19, no 4, pp1063-1074 Yee, T. W. and N. D. Mitchell (1991) ‘Generalized additive models in plant ecology’, Journal of Vegetation Science, vol 2, pp587–602
3243 J&J Vulnerability
1/11/07
10:22 am
Page 49
3
Forest Responses to Changing Rainfall in the Philippines Rodel Lasco, Florencia Pulhin, Rex Victor O. Cruz, Juan Pulhin, Sheila Roy and Patricia Sanchez
Introduction Among the world’s forests, tropical forests have critical importance in terms of their natural resources, the enormous variety of biodiversity they house and their vast potential to conserve existing carbon pools and serve as carbon sinks (Brown et al, 1996). Forests are highly dependent on climate since they are limited by water availability and temperature. The IPCC Fourth Assessment report concludes that the resilience of natural ecosystems could be exceeded due to projected climate change in the next century coupled with multiple stresses (Fischlin et al, 2007). In addition, a global warming of more than 2–3°C above pre-industrial levels could lead to substantial changes in the structure and functioning of terrestrial ecosystems. Philippine forests have extremely high floral and faunal diversity, being one of the biodiversity ‘hot spots’ of the world (McNeely et al, 1990). They harbour about 13,000 species of plants, comprising 5 per cent of the world’s total plant species (DENR/UNEP, 1997). The main strategy in biodiversity conservation is through the implementation of the National Integrated Protected Area System (NIPAS) Law of 1992, which has provided a stronger legal basis for the establishment and management of protected areas. To date, 18 terrestrial and marine reserves have been proclaimed as initial components of NIPAS. However, many of these areas are protected merely on paper because of a lack of resources. The major land-cover types in the Philippines in terms of area coverage are classified into six categories (Lasco et al, 2001): upland farms, secondary forests, protected forests, brush lands, grasslands and tree plantations. About 70 per cent (21Mha) of the Philippines’ total land area was covered with lush forests at the end of the 19th century (Garrity et al, 1993) but only about 5.789Mha (20 per cent) of forest remains (Earth Trends, 2003), with less than 1Mha of old-
3243 J&J Vulnerability
1/11/07
10:22 am
Page 50
50 Climate Change and Vulnerability
growth forests (FMB, 1998; Lasco et al, 2001). The current status of primary land-cover types in the Philippines is described as follows (Lasco et al, 2001): •
•
•
•
•
•
Old growth and other protected forests: Since 1992, the Government of the Philippines has banned logging on all old-growth forests, mossy forests and forests above 100m asl (above sea level) and with slope greater than 50 per cent as part of NIPAS. In 1995, the area of forest under protection was estimated at 2.7Mha, consisting of mossy forests (1.1Mha), old-growth Dipterocarp forests (0.8Mha), pine forests (in high elevation areas, mainly Pinus kesiya), mangroves and sub-marginal forests. Post extraction secondary forests: Logging is permitted only in the post extraction secondary Dipterocarp forests since these forests are the main source of wood in the country. There were about 2.8Mha of post extraction secondary forests in the country as of 1997 (FMB, 1998). Upland Farms: It is estimated that there are 5.7Mha of upland farms in the Philippines, consisting of forest tree-based farms, swidden fallow secondary forests, coconut plantations (typically intercropped) and fruit orchards. However, large portions of the swidden fallow areas are probably devoted to annual crops and are therefore not true swidden fallow systems. There is therefore uncertainty regarding the exact distribution of upland farms and the related swidden fallow secondary forest since these systems are highly dynamic. In response to the problem of shifting cultivation in the uplands, the government is promoting agroforestry as the main alternative production system (Agroforestry Communications, 1986; Lasco and Malinao, 1993; Nera, 1997). Tree Plantations: The rehabilitation of vast denuded areas through reforestation activities and private commercial tree plantations is a primary objective of the Philippines Government. Typically, fast growing species such as Gmelina arborea, Acacia spp. and Eucalyptus spp. are used. In government reforestation activities, trees planted are intended primarily for establishing a permanent forest cover and are not to be harvested. On the other hand, the commercial plantations established by private developers on farms are usually harvested after about 10–15 years. The rate of establishment of tree plantations was estimated at 65,233ha in 1995 (FMB, 1998; Lasco et al, 2001). There is no accurate estimate of the total area actually planted. Official records show that from 1976 to 1995, 1.3Mha were planted, but of this area only 0.6Mha can be assumed to exist (Lasco and Pulhin, 1998). Grassland: There are no natural grasslands in the Philippines except in very small patches located in high-altitude areas and human-induced and maintained grassland ecosystems (about 2Mha) that are usually the result of severe land degradation (Earthtrends, 2003). These areas are expected to regenerate back to tropical forests if protected, but regular burning (for example, the slash-and-burn method commonly practised in the Philippines) prevents plant succession. Brushlands: As of 1997, there were 2.4Mha of brushland areas in the Philippines, consisting essentially of remnants of tropical forests progressively degraded by excessive cutting. Forest cover in these areas is less than
3243 J&J Vulnerability
1/11/07
10:22 am
Page 51
Forest Responses to Changing Rainfall in the Philippines 51
20 per cent and the vegetation consists of relic trees, shrubs and grasses. Like grasslands, brushland areas are expected to regenerate back to mature forests if adequately protected. Not all global vegetation models agree on whether tropical forests will increase or decrease in extent as a result of climate change. But any major shift in rainfall pattern will affect distribution of vegetation types. Shifts in rainfall patterns could increase conversion of forests to agricultural land by increasing migration from areas affected by drought, erosion and so on. Productivity will increase or decrease depending on rainfall. Temperature change could affect the climate of a certain area drastically, leading to a loss of a few species of plants and animals that may significantly drain the biodiversity resources of these forests. A 2–3°C increase in temperatures will have marginal effects in the tropics but extended exposure to temperatures of 35–40°C combined with water shortage may damage plant tissue (Hudson and Brown, 2006). To date, there has been no study that quantifies the effects of climate change on Philippine forests. It was earlier hypothesized that under various GCM scenarios, tropical forest areas in the Philippines will likely expand as temperature and precipitation increase in many parts of the country (Cruz, 1997). At the same time a change in temperature could potentially result in significant biodiversity losses. It is also possible that species may adapt to stresses in the environment over the period of time during which climate change occurs, resulting in no significant changes in biodiversity. In response to this lack of scientific information on the potential implications of climate change for Philippine forests, a quantitative assessment of the impact of climate change scenarios on Philippine forest types was undertaken with the following objectives: • •
to determine the potential vegetative cover of the Philippines without human intervention using the Holdridge life zones; and to simulate changes in present vegetative cover as a result of climate change using GIS tools and the Holdridge life zones.
Methodology Simulating potential forest types using the Holdridge life zones The Holdridge life zones are an ecological classification system based on the three climatic factors: precipitation, heat (biotemperature) and humidity (potential evapotranspiration ratio) (Holdridge, 1967). Holdridge (1967) defined a life zone as a group of associations related through the effects of these three major climatic factors. Figure 3.1 shows the classification of the most common life zones on the Earth based on these parameters. All Philippine forests can be classified under the tropical belt because the biotemperature here is always greater than 24°C. Thus, the main determinant of life zone classification in the Philippines would be precipitation (expressed as mean annual rainfall).
3243 J&J Vulnerability
1/11/07
10:22 am
Page 52
52 Climate Change and Vulnerability
Figure 3.1 The Holdridge system of vegetative cover classification Source: Holdridge (1967).
The mean annual biotemperature is the measure of heat that is utilized in the life zone chart. The biotemperature mean is the average of the temperatures in Celsius at which vegetative growth takes place relative to the annual period. The range of temperatures within which vegetative growth occurs is estimated to lie between 0°C as a minimum and 30°C as a maximum. The positive temperatures within this range must be averaged out over the whole year in order to make it possible to effectively compare a given site with any other on Earth. The biotemperature is thus calculated as: Mean annual biotemperature (MAB) = (0 < T < 30)/365 days (or divide by 12 rather than 365 to use months rather than days as units). The third climatic factor that determines the boundaries of life zones is humidity, best described by the potential evapotranspiration (PET) ratio. PET is the theoretical quantity of water given up to the atmosphere within a zonal climate and on a zonal soil by the natural vegetation of the area throughout the growing season. Both evaporation and transpiration are directly correlated with temperature, and other factors being equal, the mean annual PET in millimetres at any site may be determined by multiplying the mean annual biotemperature by the factor 58.93.
3243 J&J Vulnerability
1/11/07
10:22 am
Page 53
Forest Responses to Changing Rainfall in the Philippines 53
The PET ratio is determined by dividing the value of the mean annual PET in millimetres by the value of the mean annual precipitation in millimetres. The PET ratio is thus a measure of the humidity that can be utilized for comparing distinct sites. It can be expressed as: PET ratio = mean annual PET/mean annual precipitation, where mean annual PET = MAB 58.93. The Holdridge life zones can also be graphically represented by means of the Holdridge life zone chart, created on the basis of annual temperature and precipitation values, to demonstrate relationships between different vegetation types. The classification of forest types in the Holdridge model for a given set of temperature and precipitation conditions can be considered as rough estimates of the potential forest types that would thrive under those conditions in the future.
Climate change impacts on Philippine forest cover Based on the three parameters, Holdridge life zones for the Philippines were identified using the ArcView 3.2 GIS program. Changes in temperature (1°C, 1.5°C and 2°C increases from current values) and precipitation (25 per cent, 50 per cent and 100 per cent increases from current values) based on the projections of climate scenarios were next used to determine the future distribution of forest types in the Philippines (Table 3.1). These precipitation and temperature scenarios are within the limits of GCM projections for the country (Government of the Philippines, 1999). Table 3.1 Synthetic climate change scenarios used in the study Increase in Rainfall (% relative to present) 25 50 100
Increase in Temperature (°C) 1
1.5
2.0
Scenario 1a Scenario 2a Scenario 3a
Scenario 1b Scenario 2b Scenario 3b
Scenario 1c Scenario 2c Scenario 3c
The various maps generated by ArcGIS 8.1 for the classification of Holdridge life zones for the Philippines include a rainfall map, a Thiessen map generated using temperature data and a land-use map. The rainfall map was based on the data collected by the Philippine Atmospheric, Geophysical and Astronomical Services Administration (PAGASA). Average annual rainfall in the Philippines ranged from 1000mm to 4000mm in 1961–1990. Temperature data were also obtained from the PAGASA and a Thiessen map was created from the 55 weather stations throughout the Philippines. The average diurnal temperature in the Philippines ranged from 19.3°C to 28.2°C in 1949–2002.
3243 J&J Vulnerability
1/11/07
10:22 am
Page 54
54 Climate Change and Vulnerability
The land-use map is based on the 1993 map prepared by the Presidential Task Force on Water Resources Development and Management. Presently there are only about 6 million hectares of forests left (excluding brushland and man-made forest plantations), a mere 20 per cent of the country’s total land area. Of these, 1.6 million hectares are non-production and less than 1 million hectares are old growth forest. This land-use map served to delineate the boundaries of Philippine forests and was used as an overlay on the calculated Holdridge life zone in the Thiessen map.
Results and Discussion Potential vs actual life zones In the absence of any anthropogenic influence, simulation of potential forest types for current temperature and precipitation showed that the Philippines would be dominated by the dry tropical, moist tropical and wet tropical forest life zones (Figure 3.2). Such a condition probably existed when the Spanish colonizers first set foot in the Philippines in 1521. At that time it is estimated that 90 per cent of the country was covered with lush tropical rainforest. By 1900, there was still 70 per cent or 21Mha of forest cover (Garrity et al, 1993). However, by 1996 there was only 6.1Mha (20 per cent) of forest cover remaining (FMB, 1998). The average deforestation rate from 1969 to 1973 was 170,000 hectares per year (Forest Development Center, 1987) while over the past 20 years it has been about 190,000 to 200,000 hectares per year (Revilla, 1997). However, the average rate of deforestation in the more recent past years has declined somewhat (largely due to reduced logging and stronger reforestation and forest protection efforts) and is in the vicinity of 100,000 hectares per year (Lasco and Pulhin, 1998; Pulhin et al, 2006). In the Philippines, the direct and indirect causes of deforestation include shifting cultivation, permanent agriculture, ranching, logging, fuel wood gathering and charcoal making (Kummer, 1992). Using ArcView, we overlaid the actual forest cover of the Philippines in 1993 over the potential life zones predicted by the Holdridge system. As might be expected, all the forest types showed a decline, with the highest decline in dry forests and the least decline in wet forests (Table 3.2).
Effects of the synthetic climate change scenarios on vegetative cover When the temperature and precipitation projections generated by climate change scenarios were considered, the change in forest cover was found to vary depending on the magnitude of change in climate parameters. For a 25 per cent increase in precipitation for the entire range of temperature increases (1°C, 1.5°C and 2°C), the following changes in forest cover were observed (Figures 3.3 and 3.4):
3243 J&J Vulnerability
1/11/07
10:22 am
Page 55
Forest Responses to Changing Rainfall in the Philippines 55
Figure 3.2 Potential Holdridge life zones in the Philippine forests without human influence
3243 J&J Vulnerability
1/11/07
10:22 am
Page 56
56 Climate Change and Vulnerability
Table 3.2 Comparison of potential and actual (1993) life zones in the Philippines Life Zone Type
Area Distribution (ha)
Percentage Distribution
Potential
1993 Life Zone
Potential
1993 Life Zone
Dry Forest
8,763,696.10
1,082,197.20
29.65
3.66
Moist Forest
15,149,315.26
3,534,636.30
51.25
11.96
Wet Forest
5,646,414.43
2,266,455.20
19.10
7.67
TOTAL
29,559,425.79
6,883,288.71
100.00
23.29
• • •
a total loss of all dry forests even at the lowest temperature change of 1°C this was expected considering the increase in available water; a 30.5 per cent increase in moist forest across all temperatures; and a negligible (0.1 per cent) increase in wet forest.
It is noteworthy that the temperature increase had negligible drying effects on the life zones in the Philippines. Normally, even slight increases in temperature (even by 0.5°C) can give rise to the El Niño phenomenon in some areas of the country. This is probably because all parts of the country already fall within the tropical belt under the Holdridge system (>24°C) and the precipitation estimates for the different scenarios are quite large, negating any effect on the increase in temperature. A 50 per cent increase in precipitation results in the following changes in the Holdridge life zones (Figures 3.3 and 3.5): • • • •
total loss of all dry forests even at the lowest temperature change – this is again expected considering the increase in available water; a 47 per cent decline in moist forests across all temperatures; an increase in rainforests from non-existent under current conditions to 365,000ha under a 1°C increase in temperature – as temperature increases, there is a slight decline in rainforest area; and a 106 per cent increase in wet forest cover as a result of greater precipitation.
It is assumed in this study that the total forested area remains fixed in the analysis. In other words, we assumed that non-forested areas could never revert back to forests because of human influence (for example, agriculture and settlements). Finally, a 100 per cent increase in precipitation will result in the following changes in the life zone pattern in the Philippines (Figures 3.3 and 3.6):
3243 J&J Vulnerability
1/11/07
10:22 am
Page 57
Forest Responses to Changing Rainfall in the Philippines 57
Figure 3.3 Projected change in area of existing life zones in the Philippines under various climate change scenarios (25%, 50% and 100% increase in precipitation)
• • • •
total loss of all dry forests (>1Mha); a 50 per cent decline in moist forests area; a significant rise in the area of rainforests from zero under current conditions to more than 2Mha; and a 32 per cent increase in wet forests, about one-third of the increase under Scenario 2.
As in the previous scenarios, the impact of temperature change is observed to be minimal due to the very high precipitation increases that mask any temperature effects. The impacts of the various climate change scenarios on Philippine forest ecosystems are presented in Figure 3.3. Overall, the simulation study showed that increases (of 25 per cent, 50 per cent and 100 per cent) in precipitation and temperature would result in a redistribution of forest types as classified by the Holdridge scenarios. The dry forests are the most vulnerable. They could be totally wiped out even under a 25 per cent increase in precipitation. Moist forests are also vulnerable, especially under a higher precipitation increase. On the positive side, there will be a significant increase in rainforest types as precipitation levels increase.
3243 J&J Vulnerability
1/11/07
10:22 am
Page 58
58 Climate Change and Vulnerability
Figure 3.4 Holdridge life zones in the Philippine forests under Scenario 1 (25% increase in rainfall) and with a 2°C temperature increase
3243 J&J Vulnerability
1/11/07
10:22 am
Page 59
Forest Responses to Changing Rainfall in the Philippines 59
Figure 3.5 Holdridge life zones in the Philippine forests under Scenario 2 (50% increase in rainfall) and with a 2°C temperature increase
3243 J&J Vulnerability
1/11/07
10:22 am
Page 60
60 Climate Change and Vulnerability
Figure 3.6 Holdridge life zones in the Philippine forests under Scenario 3 (100% increase in rainfall) and with a 2°C temperature increase
3243 J&J Vulnerability
1/11/07
10:22 am
Page 61
Forest Responses to Changing Rainfall in the Philippines 61
The results of the study are generally consistent with the findings of the IPCC’s Third Assessment Reports (McCarthy et al, 2001). At the eco-physiological level, while the net effect of climate change on the net primary productivity (NPP) of the Philippine forest system is not yet clear, there is generally a positive correlation between NPP and temperature. However, if rainfall is not sufficient, water stress could be a problem. Global vegetation models (for example, BIOME, MAPSS and IMAGE) do not agree on whether tropical forests will increase or decrease (McCarthy et al, 2001), but any major shift in rainfall pattern will affect distribution of vegetation types. Under enhanced CO2 conditions, tropical evergreen broadleaf forests could readily establish after deforestation. On the other hand, shifts in rainfall patterns could increase conversion of forests to agricultural land by increasing migration from areas affected by drought, erosion and so on. Productivity will increase or decrease depending on the amount of rainfall.
Adaptation Strategies and Options for Philippine Forests National policy framework and potential adaptation strategies To date, there has been little consideration of an overall climate change adaptation strategy and options for the Philippine forest ecosystems. The 1999 Philippines Initial National Communication contains adaptation options for watershed management that partly apply to forest ecosystems. These are mainly contained in the laws and policies governing the use and conservation of forest resources in the Philippines, which include the following: • • •
•
•
Presidential Decree 705 of 1975 (Revised Forestry Code of the Philippines) – embodies the general mandate of the Constitution in managing and conserving forest resources. DENR Administrative Order No 24, Series of 1991– promulgates the shift of logging from old-growth forests to secondary (residual) forests, effective as of 1992. Prior to this, logging was confined to old-growth forests. Republic Act No 7586 – ‘National Integrated Protected Areas Systems Act of 1992’ – stipulates that the management, protection, sustainable development and rehabilitation of protected areas shall be undertaken primarily to ensure the conservation of biological diversity. However, not all of the remaining natural forests are covered by NIPAS. All remaining old-growth forests are protected but logging is still allowed in secondary forests. Republic Act No 8371 – ‘Indigenous People’s Rights Act of 1997’ – recognizes the vested rights of indigenous peoples over their ancestral lands within forestlands, including secondary forest. The implementing guidelines of this law are still being finalized. Executive Order 363 of 1995 – adopts community-based forest management (CBFM) as a national strategy to ensure the sustainable development of the country’s forests and promote social justice.
3243 J&J Vulnerability
1/11/07
10:22 am
Page 62
62 Climate Change and Vulnerability
•
Executive Order 318 of 2004 – ‘Promoting Sustainable Forest Management in the Philippines’ – is an attempt to revise Presidential Decree 705 and aims to attain sustainable forest management in the country’s production forests.
All of the above provide the overall framework for climate change adaptation in the Philippines forest sector. Watershed management, forest conservation and greater local community participation are strategies that could also contribute towards climate change adaptation. For example, protecting existing forests allows for natural adjustment to a new climate regime. Greater local community involvement could minimize the financial cost to state agencies of adaptation. In terms of actual activities on the ground, the government has been actively pursuing several initiatives in spite of its limited resources. These include: • • •
conservation of remaining forests in NIPAS sites and watershed areas; reforestation and rehabilitation of barren upland areas through tree planting and agroforestry; and community-based forestry activities such as community organizing and development.
The private sector is less involved today compared to their involvement during the height of logging activities in the 1950s and 1960s. However, civil society is more involved as community-based programmes increase. The incorporation of climate change concerns in planning for forest resources is unfortunately not yet a priority for the government. The more urgent concern is the protection of remaining forests from human exploitation, which is viewed as the more imminent threat. However, the incorporation of climate change concerns early on in the planning process might help to avert some of the negative impacts and improve the coping capacity of forest ecosystems. As we have shown earlier in the results of our analysis based on the Holdridge life zone system, certain forest types in the Philippines, especially the dry forest types, are highly vulnerable and could, in future, be entirely replaced by other types of forests. The laws and regulations mentioned above may therefore need to be reassessed and updated to focus more on how forest management can be improved to mitigate climate change impacts, with special focus on areas classified as dry forests, such as areas in Northern Luzon, Negros, Cebu, Palawan, Basilan and General Santos. In this light, a national adaptation strategy should probably focus on identifying forested areas that are more at risk and the unique species they harbour. Specific adaptation options could include conservation and management of vulnerable species and assisting local communities that are highly dependent on forests at risk, among others.
3243 J&J Vulnerability
1/11/07
10:22 am
Page 63
Forest Responses to Changing Rainfall in the Philippines 63
Adaptation Strategies Identified by Stakeholders Local stakeholders in the Philippines that depend to various extents on forest resources generally have little awareness about climate change issues. However, they have abundant experience in coping with climate variability and extremes such as tropical storms and high rainfall and drought brought about by the ENSO phenomenon. Their current adaptation strategies to such extreme events might serve as a guide to their possible responses to the impacts of future climate change, under the implicit assumption that adaptation options to the impacts of climate variability could be applicable to the impacts of climate change. Stakeholder responses to current climate events were gauged on the basis of stakeholder interactions during a workshop held in the Pantabangan– Carranglan watershed of the Philippines in 2004 aimed at identifying the impacts of climate variability and extremes and determining local coping strategies. Thirty participants from different organizations within the Pantabangan–Carranglan watershed – the National Power Corporation, the National Irrigation Authority, local government units, non-government organizations and people’s organizations – attended the workshop. A wide variety of adaptation options were identified by stakeholders. These centred on the use of appropriate species/crops, scheduling, technical innovations (for example, water conservation), capacity building and law enforcement (see Table 3.3). The high degree of knowledge on adaptation options exhibited by participants indicates a relatively strong level of awareness on adjusting to climate variability and extremes. This existing knowledge base could have the potential to serve as a building block for the determination of future climate change adaptation strategies. In general, the adaptation options identified are consistent with those recommended in the Philippines Initial National Communication (1999) and the IPCC Third Assessment Report (McCarthy et al, 2001). Since these adaptation options were identified in response to climatic variability at a particular site, their applicability to other areas would have to be determined.
Conclusions Under potential climate change scenarios, forest types in the Philippines could change dramatically, especially with increasing precipitation levels. The most vulnerable are the dry forests and to a lesser degree the moist forests, which could be completely eliminated with the rise in precipitation. They would most likely be replaced by the rainforest type. At the national and local levels, climate change issues are not yet considered in the planning and implementation of forestry activities. However, there is a set of laws and regulations at the national level that could provide a framework for the implementation of adaptation options. At the local level, stakeholders have been adapting to climate variability and extremes for a long time and their strategies could also serve to inform future adaptive responses to climate change.
3243 J&J Vulnerability
1/11/07
10:22 am
Page 64
64 Climate Change and Vulnerability
Table 3.3 Adaptation options to climate variability and extremes for forest lands in the Pantabangan–Carranglan watershed, Philippines Land Use
Adaptation Options
Tree plantation
Adjusted silvicultural treatment schedules and proper silvicultural practices Plant species that can adjust to variable climate situations Proper timing of tree planting projects or activities Construction of fire lines Controlled burning Supplemental watering
Natural forest
Safety net measures for farmers by local and national government Cancellation of logging permits (total logging ban)
Grasslands
Reforestation – adaptation of contour farming in combination with organic farming Promote community-based forest management Increased funds for forest protection and regeneration from national government Increase linkages among local government, national government and non-governmental organizations Introduction of drainage measures Controlled burning Introduction of drought resistant species Intensive information dissemination campaign among stakeholders
It is recommended that future studies examine the impacts of climate change and the accompanying changes in forest types on biodiversity at the species level, with a special emphasis on rare, threatened and endangered species. It is also recommended that a thorough review of existing policies on managing forest ecosystems in the country be conducted in order to be able to assess their effectiveness in terms of addressing the projected future impacts of climate change and safeguarding the country’s forest resources from irreparable damage.
References Agroforestry Communications (1986) The Philippine Recommendations for Agroforestry, PCARRD, Los Baños, Laguna Brown, S., J. Sathaye, M. Cannel and P. Kauppi (1996) ‘Management of forests for mitigation of greenhouse gas emissions’, in R. T. Watson, M. C. Zinyowera and R. H. Moss (eds) Climate Change 1995: Impacts, Adaptations and Mitigation of Climate Change: Scientific-Technical Analyses, Contribution of Working Group II to the Second Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge and New York, pp775–797 Cruz, R.V.O. (1997) ‘Adaptation and mitigation measures for climate change: Impacts on the forestry sector’ in Proceedings of the Consultation Meeting for the International Conference on Tropical Forests and Climate Change, Environmental Forestry Programme (ENFOR), CFNR, UPLB, College, Laguna, Philippines
3243 J&J Vulnerability
1/11/07
10:22 am
Page 65
Forest Responses to Changing Rainfall in the Philippines 65 DENR/UNEP (Department of Environment and Natural Resources/United Nations Environment Program) (1997) Philippine Biodiversity: An Assessment and Action Plan, Bookmark Inc, Makati City, Philippines Earthtrends (2003) ‘Earth trends country profiles: Forests, grasslands, and drylands: Philippines’, World Resources Institute, http://earthtrends.wri.org Fischlin, A., G. F. Midgley, J. T. Price, R. Leemans, B. Gopal, C. Turley, M. D. A. Rounsevell, O. P. Dube, J. Tarazona and A. A. Velichko (2007) ‘Ecosystems, their properties, goods, and services’, in M. L. Parry, O. F. Canziani, J. P. Palutikof, P. J. van der Linden and C. E. Hanson (eds) Climate Change 2007: Impacts, Adaptation and Vulnerability, Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, pp211–272 FMB (1998) Forestry statistics (1997), Forest Management Bureau, Quezon City, Philippines Forest Development Center (1987) Towards a Successful National Reforestation Program, Policy paper no 24, UPLB College of Forestry and Natural Resources, Laguna, Philippines Garrity, D. P., D. M. Kummer and E. S. Guiang (1993) The Upland Ecosystem in the Philippines: Alternatives for Sustainable Farming and Forestry, National Academy Press, Washington, DC Government of Philippines (1999) Philippines Initial National Communication to the UN Framework Convention on Climate Change, Government of Philippines, Manila Holdridge, L. R. (1967) Life Zone Ecology (Revised Edition), Tropical Science Center, San Jose, Costa Rica Hudson, J. C. and D. Brown (2006) ‘Rethinking Grassland Regionalism’, University of Minnesota website, www.geog.umn.edu/Faculty/brown/grasslands/RGR1.htm Kummer, D. M. (1992) Deforestation in Post-war Philippines, Ateneo de Manila University Press, Philippines Lasco, R. D. and E. P. Malinao (1993) ‘Height growth and herbage production of seven MPTS uses as hedgerows in alley cropping system: An on-farm experiment’, Sylavatrop: The Technical Journal for Philippine Ecosystems and Natural Resources, vol 3, no 1, pp97–107 Lasco, R. D. and F. P. Pulhin (1998) Philippine Forestry and CO2 Sequestration: Opportunities for Mitigating Climate Change, Environmental Forestry Programme (ENFOR), CFNR, UPLB, College, Laguna, Philippines Lasco, R. D., R. G. Visco and J. M. Pulhin (2001) ‘Secondary forests in the Philippines: Formation and transformation in the 20th century’, Journal of Tropical Forest Science, vol 13, pp652–670 Nera, B. S. (1997) ‘Agroforestry of DENR’s social forestry program’, in Developments in Agroforestry Research, Book Series Number 160/1997, Philippine Council for Agriculture, Forestry and Natural Resources Research and Development, Los Baños, Laguna, Philippines, pp35–44 McCarthy, J., O. Canziani, N. Leary, D. Dokken and K. White (eds) (2001) Climate Change 2001: Impacts, Adaptation and Vulnerability, Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US McNeely, J. A., K. R. Miller, W. V. Reed, R. A. Mitternmeier and T. B. Werner (1990) Conserving the World’s Biological Diversity, IUCN, Gland, Switzerland and Washington, DC Pulhin, J. M., U. Chokkalingam, R. J. J. Peras, R. T. Acosta, A. P. Carandang, M. Q. Natividad, R. D. Lasco and R. A. Razal (2006) ‘Historical overview’, Chapter 2 in U. Chokkalingam, A. P. Carandang, J. M. Pulhin, R. D. Lasco, R. J. J. Peras and T. Toma (eds) One Century of Forest Rehabilitation in the Philippines: Approaches,
3243 J&J Vulnerability
1/11/07
10:22 am
Page 66
66 Climate Change and Vulnerability Outcomes and Lessons, Center for International Forestry Research (CIFOR), Bogor Barat, Indonesia Revilla, A. V. (1997) ‘Working paper for the forestry policy agenda for the incoming administration’, UPLB College of Forestry and Natural Resources, Laguna, Philippines
3243 J&J Vulnerability
1/11/07
10:22 am
Page 67
4
Vulnerability of Mongolia’s Pastoralists to Climate Extremes and Changes Punsalmaa Batima, Luvsan Natsagdorj and Nyamsurengyn Batnasan
Introduction Over a million of Mongolia’s roughly 2.7 million inhabitants live in rural areas, most herding livestock for their livelihood. More than three quarters of the country’s land area is used for grazing livestock, making pasture by far the largest category of land use. Herding on the open pastures of Mongolia is very risky, however. The climate is one of harsh extremes such as summer drought, severe winter conditions called zud, spring and autumn frost, dust storms, blizzards, heavy snowfall and cold rain. These and other climate hazards negatively impact herders’ livelihoods by reducing access to forage, fodder and water, weakening livestock and limiting their weight gain, and even killing large numbers of animals. Mongolian pastoralists adopted nomadic practices long ago as an adaptation that provides resilience to the highly variable climate and productivity of Mongolia’s grasslands. But, as our study shows, Mongolia’s pastoralists have become more vulnerable to climatic hazards. This condition of high and increasing vulnerability is cause for serious concern given the potential consequences of future climate change. Mongolia is heavily reliant on the pastoral system and impacts on it will be felt throughout Mongolian society and the Mongolian economy. The livestock sector, which includes not only herders but also processors and distributors of livestock products, employs almost 50 per cent of the population, produces roughly 35 per cent of agricultural gross production and accounts for 30 per cent of the country’s exports (National Statistical Office, 2001). Recent experiences have demonstrated that severe and widespread drought and zud that impact livestock herds can trigger large internal migrations of people, cause widespread unemployment, deepen poverty, sharply reduce export earnings and impact incomes throughout the economy. The growing vulnerability to climate hazards is driven by institutional and government policy changes and by recent climate changes. The transition to a
3243 J&J Vulnerability
1/11/07
10:22 am
Page 68
68 Climate Change and Vulnerability
market economy in the early 1990s brought important changes to the institutions of Mongolia’s pastoral system and to government policies that have contributed to growing vulnerability. Changes have also been observed in the climate of the region which have had negative impacts on Mongolia’s grasslands, livestock and herders. In our study of the vulnerability of Mongolia’s pastoralists to climate change we constructed indices of recent trends in drought and severe winters from observational data, examined the impacts of a drying climate and severe weather events on grassland productivity, livestock and livelihoods, and developed maps of current vulnerability to climate extremes. The effects of changing institutions and policies on vulnerability to climate hazards were also explored. The potential impacts of future climate change were investigated using model simulations of changes in grassland productivity, grazing behaviour of animals and animal weight gain for different scenarios of future climate. Our research also included assessment of options for adapting to changing climate stresses. We present here in this chapter our findings on the changing vulnerability of Mongolia’s pastoralists; findings about adaptation options are reported in Batima et al (2008).1
Changing Climate: Observed Trends Mongolia’s climate is semi-arid to arid and is characterized by a long-lasting cold winter, a dry and hot summer, low precipitation, large temperature fluctuations between day and night and between summer and winter, and a relatively high number of sunny days (an average of 260 days per year). There are four sharply distinct seasons and the months in each season are quite different. January, the coldest month, averages -30 to -34°C in the high mountain areas of Altai, Khangai, Khuvsgul and Khentii, -20 to -25°C in the steppe, and -15 to -20°C in the Gobi Desert. July, the warmest month, averages 12–15°C in the mountain areas and 20–25°C in the southern part of the Eastern steppe and the Gobi Desert. Annual mean precipitation is 300–400mm in the mountain regions, 150–250mm in the steppe, 100–150mm in the steppe desert, and 50–100mm in the Gobi Desert. About 85 per cent of total precipitation falls from April to September, of which 50–60 per cent falls in July and August. The clear skies in winter due to high anticyclone dominance over Mongolia results in low snowfall. Snow contributes less than 20 per cent of the annual precipitation. The first snowfall typically occurs from the middle of October to the beginning of November. It is usually short-lived and disappears because of late-autumn warming and wind, but sometimes a late-autumn first snowfall persists through the winter as snow cover in mountainous regions. Climate change studies conducted in Mongolia show a gradually increasing trend in air temperature and a slightly decreasing trend in precipitation. During the past 60 years, the annual mean air temperature has increased by 1.80°C. Clear warming started in the 1970s and intensified from the end of 1980s. Thirty of the years during the period 1940–2003 had positive air temperature anomalies, with 23 of these occurring after 1970. Similarly, all eight
3243 J&J Vulnerability
1/11/07
10:22 am
Page 69
Vulnerability of Mongolia’s Pastoralists to Climate Extremes and Changes 69
years that exceeded a 1°C anomaly were observed after 1990, including three consecutive years in 1997, 1998 and 1999. The year 1998 was the warmest year ever measured instrumentally in Mongolia. The warming has been most pronounced in winter, with a mean temperature increase of 3.61°C. The spring and autumn temperature has also increased from 1.4 to 1.5°C. Geographically, more intensive spring and autumn warming of >2°C was observed at high altitudes in mountainous regions while warming of <1°C was observed at lower altitudes, as well as in the steppe and the Gobi Desert. With the warming there has been a corresponding increase in the number and duration of hot days. For example, heat wave duration has increased by 8 to 18 days, depending on location. The greatest increase, of 15–18 days, was recorded in the mountainous regions, while in the Gobi Desert, heat-wave duration increased by only 6–8 days. In 1998, heat-wave duration reached 70 days in the high mountains and 30 days in the Gobi Desert, which was the most anomalous event that has occurred in the past 40 years. Changes in annual precipitation averaged over the entire country are small and not statistically significant, but some trends have been observed at regional scales. In the central region, annual precipitation decreased by 30–90mm, but it increased by 2–60mm in the far western area and by 30–70mm in the extreme southeastern part of the country. The magnitude of these changes ranges from 5 to 25 per cent of normal precipitation. Seasonally, autumn and winter precipitation increased by 4–9 per cent while spring and summer precipitation has decreased by 7–10 per cent (Batima, 2006). With the higher temperatures and mixed trends for precipitation, evaporative demand has increased and the climate of Mongolia has become slightly drier. To examine trends in drought conditions, an index of summer drought was constructed using data from 64 meteorological stations distributed evenly across the country (Batima, 2006). The index is calculated as a function of anomalies of mean monthly temperature and precipitation for summer months from long-term averages: n
Ssummer =
4
i =1 j =1
– T – T –––––– – T ij
n
4
i =1 j =1
– P – P –––––– P
ij
where T and P are the j monthly mean temperature and precipitation at i meteorological station, T and P are the long-term average temperature and precipitation at the same meteorological station, T and P are standard deviations of temperature and precipitation. Analysis of the drought index shows a clear increasing trend over the period 1940 through 2002 that is significant at the 95 per cent confidence level (see Figure 4.1). The increase in drought conditions is most pronounced over the last decade of the available time series, with four consecutive years of summer drought from 1999 through 2002 that affected 50–70 per cent of the territory. Such long-lasting and widespread drought had not been observed in Mongolia in the previous 60 years. During these 4 years, about 3000 water sources,
3243 J&J Vulnerability
1/11/07
10:22 am
Page 70
70 Climate Change and Vulnerability
Drought Index
including 680 rivers and 760 lakes, dried up (Ministry of Nature and Environment, 2001 and 2003).
Figure 4.1 Summer drought index, 1940–2002 Harsh winter conditions called zud pose the greatest climate hazard for herders in Mongolia. The Mongolian Language Vocabulary Dictionary defines zud as ‘a very severe situation of food insufficiency for both people and animals caused by natural factors in winter’ (Tsevel, 1966). Zud has been described as ‘livestock famine’ and can result in mass death of animals from hunger, freezing and exhaustion. It comes in a variety of forms that include heavy snowfall or ice cover that prevents animals from grazing, lack of snow cover and water, and extreme cold. An index of the severity of winter is constructed as a function of monthly mean temperature and precipitation anomalies for winter months: n
Ssummer =
4
i =1 j =1
– T – T –––––– – T ij
n
4
i =1 j =1
– P – P –––––– P
ij
where T and P are winter j monthly mean temperature and precipitation at i meteorological station, T and P are long-term average temperature and precipitation at the same meteorological station, T and P are standard deviation of temperature and precipitation. The winter severity index is calculated using data from the same 64 stations as for the summer drought index. Analysis of the winter index shows slightly decreasing occurrence of severe winter conditions. The normalized values of the winter indices for the country are shown in Figure 4.2.
3243 J&J Vulnerability
1/11/07
10:22 am
Page 71
Winter Indexes
Vulnerability of Mongolia’s Pastoralists to Climate Extremes and Changes 71
Figure 4.2 Winter severity index, 1940–2002 As can be seen from Figure 4.2, climatic conditions for winters in Mongolia grew milder as measured by the index. This is consistent with the observed winter warming of 3.61°C since 1940, which is twice the rate of warming of any other season (Natsagdorj et al, 2005). One might argue that herders should benefit from warmer winters, but unfortunately, this is not what has been observed. While the nationally averaged index for winter severity shows a trend toward milder conditions, the frequency of zud covering more than 25 per cent of the country’s territory doubled in 1980–1990 and tripled in 1990–2000 compared to the 1960s. This important spatial pattern is hidden by the nationally averaged data. In many cases, the warmer winter temperatures were accompanied by a number of unusual or unseasonal weather phenomena. For example, usually there are no wind storms during winter in Mongolia. But in recent years, wind storms have become more common in December and January. Wind storms during the coldest period of the year cause drifting of snow over a large area, causing animals to run long distances to avoid wind-drifted snow. Many animals die of exhaustion as a result. Air temperatures that rise rapidly to anomalous levels for short (3–7 day) periods and then fall to levels that are normal for the winter months cause rapid melting of snow cover and refreezing into an ice sheet that may cover a large area and prevent grazing. Animals also show temperature stress from short-period temperature fluctuation. Thus, despite the trend toward milder winter temperatures, herders’ exposure to zud conditions has been increasing.
3243 J&J Vulnerability
1/11/07
10:22 am
Page 72
72 Climate Change and Vulnerability
Climate, Environment and Livelihood In normal years, pasture growth begins in late April; standing biomass reaches its peak in August and decreases to 70–80 per cent of the peak by late autumn, 50–60 per cent in winter and 30–40 per cent in spring. Between August and spring, the quality of the vegetation, measured by the carbon:nitrogen ratio, decreases by a factor of 2–3 and protein content falls by a factor of 3–4. The productivity of Mongolia’s grasslands, measured as standing peak biomass, varies from 100 to 1000kg per hectare (ha), decreasing from the north to the south of the country. Grassland productivity declined by 20 to 30 per cent over the latter half of the 20th century as temperatures and a tendency towards drought conditions increased. Standing plant biomass, monitored since 1964 in a national network of 25m 25m fenced stations, declined at rates of 2 to 7kg/ha/yr in the different biomes of Mongolia (Table 4.1). Standing biomass has also decreased for the critical months of April and May, when animals weakened by the long winter need to rebuild their strength. In the steppe and forest steppe, April peak standing biomass has decreased by 10–20 per cent and the May peak has decreased by 30 per cent (Tserendash et al, 2005). Because observations of pasture biomass were carried out in fenced fields, the reduction is considered to be the result of climate change. The decline in pasture productivity observed in the monitoring stations is consistent with wider trends as shown by analysis of normalized difference vegetation index (NDVI) changes for the last 10 days of each July for the period 1982–2002, which shows a clear decline of NDVI in 69 per cent of the country’s territory over the past 20 years. Table 4.1 Standing peak biomass by ecosystem type, 1960–2000 Ecosystem
No of stations
Length of time series (years)
Average peak biomass (kg/ha)
Change in peak biomass (kg/ha/yr)
Forest-steppe Steppe Desert-steppe Desert Altai Mountains
18 13 18 2 5
16–31 19–34 17–29 18–21 17–24
590 300 220 170 170
-6 -5 -3 -7 -2
Animal breeds native to Mongolia have small body sizes and low productivity relative to exotic breeds. But they are well adapted to the severe climatic conditions of Mongolia’s winter and spring, as well as to various diseases. Even so, an average of 2.4 per cent of the total livestock population dies each year. When the climate is extreme, the results can be disastrous. Climate hazards such as summer drought, zud, spring and autumn frost, strong dust storms, blizzards, heavy snowfall and cold rain can cause mass death of animals and weaken the livelihood of herding families.
3243 J&J Vulnerability
1/11/07
10:22 am
Page 73
Vulnerability of Mongolia’s Pastoralists to Climate Extremes and Changes 73
Mongolian livestock obtains over 90 per cent of its annual feed intake by grazing year-round on open pastures. The animals start to gain weight in early summer, when high quality grass is available, and attain their maximum weight by the end of autumn. By winter, animals can take in only 40–60 per cent of their daily feed requirements by grazing because of the decreased quantity and quality of forage. They lose weight during this time and reach their minimum weight in the spring. In the case of ewes, the weight loss from autumn to spring is approximately 34 per cent. March and April are usually the most difficult months for livestock and are when livestock losses are greatest because animals are weakened and hay and fodder are in short supply. It is still cold, dry and extremely windy, with frequent dust storms and occasional snowfall, and new grass has not yet grown. Spring is also the time when animals bear young, and pregnant animals that are weakened are at particular risk. Spring plant growth is critical for animals to rebuild strength lost during winter and to produce milk for newborns. The observed decreases in the productivity of Mongolia’s grasslands have negatively impacted livestock, as indicated by decreases in animal weights. Observations of animal weights made at three sites in three different biomes over the period 1980–2001 indicate that average weights of ewes, goats and cattle have fallen. Ewe weight has declined by 4kg on an annual basis, or roughly 8 per cent, since 1980. An even greater decline, of 8kg, has been observed for springtime. Animal weight is an indicator of biocapacity for surviving harsh winter and spring conditions. The observed weight declines are evidence of decreased biocapacity and increased vulnerability to climate variations and extremes. Another factor contributing to declining animal weights may be the impact of high summer temperatures on animal grazing behaviour. When midday summer temperatures exceed certain thresholds, animals cease grazing, reducing the amount of time spent grazing per day and thus food intake. Over the period 1980–2001, the time spent grazing decreased by an average 0.8 hours per day for the months of June and July (Bayarbaatar et al, 2005).
Drought impacts The growing season in Mongolia is very short, lasting from May to September, and pasture productivity can vary widely depending on summer rainfall and temperatures. Dry summers and droughts have been observed to decrease rangeland productivity by 12–48 per cent in the high mountains and by 28–60 per cent in the desert steppe (Tserendash et al, 2005. But drought is not commonly perceived as a natural hazard in Mongolia because it is not the proximate cause of significant livestock mortality during or immediately following a summer of below normal rainfall. Nevertheless, drought does result in decreased pasture vegetation, decreased palatable plant species, reduced water availability, an absence of grass on pasture and increased prevalence of pests. Furthermore, drought prevents herders from preparing hay and other supplementary feed for animals and dairy products for themselves. These effects of drought reduce the avail-
3243 J&J Vulnerability
1/11/07
10:22 am
Page 74
74 Climate Change and Vulnerability
ability of forage and fodder for livestock. Most importantly, in drought years animals are unable to build up the necessary strength or biocapacity during the summer and autumn that is necessary to cope with the harsh, cold winter and spring, making them more vulnerable to mortality in those months. Some of the first- and second-order impacts of drought are described in Table 4.2. Table 4.2 First- and second-order impacts of drought Affected activity
First-order impacts
Second-order impacts
Animal husbandry
Insufficient vegetation on pasture Drying of water sources Increased number of hot days and heat-wave duration
Animals cannot take the necessary energy and nutrients Animals cannot graze on pasture because of hot weather and decrease daily intake Decreased animal weight Decreased biocapacity of animals to survive winter and spring
Grazing and supplemental feed production
Reduced hay making area Grazing areas normally preserved for winter/spring used during the summer
Limited forage produced and reserved for winter Reduced food reserves for livestock Reduced winter pasture
Livelihood
Reduced livestock productivity; Reduced production of milk, dried curd, clotted cream, butter
Reduced food reserves Reduced cash income Food insecurity and malnutrition
Conservation of pasture
No grass on affected area Increase in pests (e.g. Brandt’s No hay on affected area vole and grasshoppers) Degradation of pasture land Increase frequency of fire in forest and steppe Soil erosion and loss of nutrients
Animal weight is an important indicator of animal biocapacity and changes in weight integrate the effects of multiple environmental changes on biocapacity. There is a direct correlation between our drought index and sheep and goat weights, which demonstrates the negative impact of drought on animals’ biocapacity (Figure 4.3).
Zud impacts Winter zud represents a high risk both to livestock and to humans who depend on livestock for income, food, transportation, fuel, and materials for clothing and shelter. The impacts of zud are illustrated in Figure 4.4, which shows a positive correlation between our index of winter severity and livestock mortality.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 75
Drought Indexes
Drought Indexes
Vulnerability of Mongolia’s Pastoralists to Climate Extremes and Changes 75
Percentage of animal loss to total number
Figure 4.3 Relationship between drought indexes and summer/autumn live-weights of sheep and goats
Figure 4.4 Winter severity and livestock mortality
3243 J&J Vulnerability
1/11/07
10:23 am
Page 76
76 Climate Change and Vulnerability
There are several forms of zud, depending on their characteristics, contributing factors, and causes: tsagaan (white), khar (black), tumer (iron), khuiten (cold), khavsarsan and tuuvaryin zud. These are described in Table 4.3. Zud can last from one to several months. Mongolian herders are practised at coping with zud when it occurs in a small area and is of short duration. But zud that covers a large area and is long in duration strains the coping capacity of herders and their livestock and can cause high animal mortality. Very severe zud occurred in the years 1944–1945, 1967–1968, 1978–1979 and 1999–2002, during which abnormally high numbers of animals were killed. Nearly one third of livestock in Mongolia, 8.8 million animals, died during the zud of 1944 and 12 percent, or 2.6 million animals, died in the 1967–1968 zud. Table 4.3 Zud forms and their description Zud Form
Description
Climatic Criteria
Tsagaan (white) Results from high snowfall that prevents livestock from reaching the grass. Herders zud used to leave the zud area if the area was small. Can cause a very serious disaster if it covers a large area. Tsagaan is the most common and disastrous form of zud.
Long lasting: large amount of snowfall in the beginning of winter. Short lasting: large amount of snowfall at the end of winter.
Khar (black) zud Occurs when lack of snow in grazing areas leaves livestock without any unfrozen water supplies where wells are not accessible. Both human and animals suffer from lack of water to drink. This form usually happens in the Gobi Desert region.
Very little or no snowfall in winter. No winter forage on pasture because of drought in summer. No winter forage on pasture due to overgrowth in number of voles (Microtus brandtii) and grasshoppers or increased incidence of forest and steppe fire.
Tumer (iron) dzud
Occurs when snow cover melts and refreezes Short rapid warming in wintertime to create an impenetrable ice-cover that (3–7°C higher than monthly mean prevents livestock from grazing. temperature) followed by return to sub-freezing temperatures.
Khuiten (cold) zud
Occurs when air temperature drops to very low levels for several consecutive days. Extreme cold temperatures and strong freezing wind prevent animals from grazing; the animals expend most of their energy in maintaining their body heat.
Air temperature falls by 5–10°C lower than the monthly mean.
Khavsarsan A combination of at least two of the above (combined) zud phenomena occurring at the same time. Tuuvaryin zud
Geographically widespread white, black, iron or cold zud combined with overcrowding of livestock and migration of livestock over certain territory that results in overgrazing and depletion of pasture land resources.
Geographically widespread zud.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 77
Vulnerability of Mongolia’s Pastoralists to Climate Extremes and Changes 77
Mongolian herders experienced the worst zud of the last 30 years in 1999–2000 and then faced severe zud in 2000–2001 and 2001–2002 as well. This period also coincided with severe and extensive summer droughts. More than 50 per cent of the territory was affected and over 12 million livestock animals died, representing 25 per cent of the total (National Statistical Office, 2001). More than 12 thousand herders’ families lost their entire herds, while thousands more were pushed to subsistence levels below the poverty line by their loss of animals. Such long-lasting winter zud for three consecutive years combined with summer droughts had not happened in Mongolia for 60 years. The government’s disaster relief funding for a disaster of this magnitude was inadequate to meet the urgent demands of the affected population and Mongolia requested international relief assistance in February 2000. The mass death of livestock significantly affected the herders’ daily lives. Many people lost their means of transportation, which prevented them from moving away from the affected areas. Those who could move from their normal grazing areas placed an additional burden on areas not directly affected by the zud. Overgrazing in these areas increased the geographical extent of the impacts of the zud. Herders also faced an insufficient supply of heating materials (manure), food and cash income, which negatively affected their health. The elderly and newborn are the most vulnerable during a zud and mortality rates rose for these groups. Massive death of livestock not only affects the herders’ livelihood but also causes severe socioeconomic damage to the whole country. Many herders migrated from rural to urban areas, reversing the trend of net urban to rural migration of the 1990s and raising the urban population of Mongolia from just under 50 per cent in 1999 to over 57 per cent. Unemployment and poverty rose in the urban areas. Gross agricultural output in 2003 was 40 per cent below the 1999 level and its contribution to national gross domestic product decreased from 38 to 20 per cent (National Statistical Office, 2003).
Regions vulnerable to zud and drought Zud represents the greatest single climatic hazard for Mongolia’s herders. But, as demonstrated by the 1999–2002 events, the joint occurrence of summer drought and winter zud can have even greater impacts. Data from the Institute of Meteorology and Hydrology are used to calculate the frequency of white zud, black zud and drought conditions for the Mongolian landscape and classify the degree of vulnerability of each area as shown in Table 4.4. The classifications of frequencies of different extremes have been mapped and the results of an overlay of black and white zud are shown in Figure 4.5. As can be seen from this map, almost 60 per cent of the country’s territory is classified as having high, very high or severe vulnerability to zud. When drought is included (Figure 4.6), 90 per cent of the county is revealed to be vulnerable to natural disasters. Especially vulnerable are the western edge of the Khangai Mountains and the Altai Mountains, including the Great Lakes basin and the Gobi Desert. The Altai and Khangai mountain regions, including the Great Lakes basin, fall within the Altai-Sayan ecoregion, which is designated
3243 J&J Vulnerability
1/11/07
10:23 am
Page 78
78 Climate Change and Vulnerability
Table 4.4 Frequency of extremes and levels of vulnerability Vulnerability Level
Drought Frequency
White Zud Frequency
Black Zud Frequency
Slight Moderate High Very high Severe
No drought 1–20% 21–40% 41–60% 61–70%
No zud 1–10% 11–20% 21–40% 41–70%
No zud 1–10% 11–20% 21–40% 41–60%
as a Global 200 Ecoregion by the World Wildlife Fund for its outstanding biological diversity. Thus an increased incidence of drought and zud under climate change would affect not only the livestock sector but also some of the unique ecosystems and biodiversity of Mongolia (Batima et al, 2004). During the disastrous period of 1999–2002, the regions that lost more than half of their animals (from 800,000 to 1,400,000 animals) correspond to the areas classified as severely vulnerable to drought and zud.
Figure 4.5 Areas vulnerable to black and white zud
Changing Institutions, Policies and Resources Thousands of years of traditional pastoral practices were fundamentally changed with collectivization of the livestock sector in the early 1960s. Prior to this, allocation of pasture land, sharing of labour and resources, and the settle-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 79
Vulnerability of Mongolia’s Pastoralists to Climate Extremes and Changes 79
Figure 4.6 Areas vulnerable to drought, black zud and white zud
ment of disputes were organized through a hierarchy of customary institutions such as the khot ail (a group of 2 to 12 households that camp together), neg nutgiinkhan or ‘people of one place’ (a neighbourhood-level group of 20–50 households) and bag (group of 50 to 100 households) (Chuluun et al, 2003). Through these institutions a nomadic system of seasonal migration of households and livestock herds operated such that access to forage was provided year round, grazing pressures were regulated and drought and zud risks managed. In 1960 the livestock sector was collectivized and pasture land, animals and tools of production became the property of collectives called negdels. The negdels took over the regulatory functions of the traditional customary institutions for allocating pasture and resources and enforcing seasonal movement of herds. Within the negdel, each herding family specialized in the production of a single animal species and sometimes age class of animals (Chuluun et al, 2003). The negdel was the management and economic unit responsible for supplying inputs, consumer goods, fodder, transport, health services, education and veterinary services to its members. The government, through the negdel, provided inputs for fencing, veterinary services, cross-breeding to improve livestock breeds and irrigated forage production, all at greatly subsidized prices. A State Emergency Fodder Fund was established to provide supplementary feed during hazardous weather events. During the collectivized period of 1960–1990, income inequality was low, poverty was virtually unknown, access to primary healthcare was nearly universal and adult literacy reached 97 per cent (Mearns and Dulamdary, 2000). The collectively supplied and subsidized services and resources reduced the exposure and sensitivity of herders’ livelihoods to climate hazards and provided the
3243 J&J Vulnerability
1/11/07
10:23 am
Page 80
80 Climate Change and Vulnerability
means for coping with climate-driven variability in their production system and recovering from shocks to their production. However, herders became dependent on centralized and heavily subsidized services and inputs. In the beginning of the 1990s, the country was transformed from a socialist system to a market economy. With the transition came dissolution of the negdels, an end to state subsidies to the livestock sector, and privatization of livestock, shelter, wells and other tools of production, while pasture land remained the property of the state. The transition was accompanied by a loss of external subsidies from the Soviet Union, collapse of the state budget, retrenchment of state employment, drastically reduced provision of health and education services, and deep cuts in investment in basic infrastructure. The end of the 1990s also saw deterioration in terms of trade and loss of export earnings as prices fell for Mongolia’s three main exports (copper, gold and cashmere) and oil prices rose. As a result, Mongolia experienced a widening of income inequality, unemployment, a doubling of maternal mortality, declining literacy and a rapid rise in poverty that has put more than one third of the population below the poverty line (Mearns and Dulamdary, 2000). Privatization of livestock created a large number of herders with small animal herds who constitute the bulk of the poor. During the socialist period, there was some private ownership of animals, which provided a supplement to herders’ incomes and diet, but almost no herders owned more than 100 animals. With the end of the socialist system, privately owned animals provide most of a herder’s income and diet, and poverty in the rural areas is tightly linked to the number of animals owned by a household. In the market economy era, households with fewer than 50 animals are considered by the government to be poor and those with 51–100 animals are considered to be vulnerable to poverty. Households in these two categories, about 60 per cent of herders, are highly vulnerable to climate hazards because severe events such as zud can eliminate their herds or reduce herd sizes below levels needed for a subsistence income. Households with 101–200 animals, approximately 25 per cent of herders, are usually able to generate sufficient income to support a comfortable life, yet they are still vulnerable to events that cause a large loss of animals and reduce their incomes to subsistence levels or below. Thus nearly 85 per cent of herder households are vulnerable to climate extremes such as the series of severe droughts and zud from 1999 through 2002 that devastated livestock herds. Many of the poorer herder households were newcomers who migrated in the early 1990s from urban to rural areas in response to reduced state employment, economic contraction and the opportunity to receive private livestock (Mearns and Dulamdary, 2000). The migration was large enough to reverse the earlier rural-to-urban migration trend and more than double the number of herders and their households during the 1990s (Chuluun et al, 2003). The newcomers typically have small herds with low diversity of herd structure, are young, have high dependency ratios, have little herding experience and lack knowledge of local customary arrangements. This latter characteristic impeded them from assimilating and gaining access to social networks, informal credit sources, wells, pastures and winter shelters, creating a condition of high
3243 J&J Vulnerability
1/11/07
10:23 am
Page 81
Vulnerability of Mongolia’s Pastoralists to Climate Extremes and Changes 81
vulnerability to climate and other shocks (Mearns and Dulamdary, 2000). Many of these herders were forced to migrate back to urban areas when they lost their herds during the 1999–2002 period of drought and zud. The dissolution of the negdels left a vacuum in which the use of pasture is largely unregulated and customary rights weakened. Pasture effectively became an open access commons in which herders lack incentives to conserve pasture or make investments in land, water and other infrastructure. Without the negdels or appropriate incentives, investment has been lacking and infrastructure such as winter shelters, wells, marketing systems, veterinary services and production of feed supplements has deteriorated. Herders responded to the open access nature of pastures by increasing the numbers of animals stocked and grazed. Many reduced the frequency and distance of movement of herds, trespassed into pastures traditionally reserved for use only during climate calamities, moved herds closer to urban centres with services that were increasingly unavailable in rural areas and even grazed herds on the same pasture year round. As a result, pastures in some areas have been overstocked and degraded and have reduced productivity (Chuluun et al, 2003). The changes since 1990 outlined above have diminished the capacity of Mongolia’s pastoralists to cope with climatic, economic and other shocks. The elimination of subsidies and services to the livestock sector, the loss of the safety nets of the socialist system, and the contraction of opportunities due to the economic downturn and deteriorated terms of trade eroded the livelihood resources available to pastoralists. A large class of poor households with small herds, insufficient human capital and lack of access to social and financial capital has been created in the rural areas. The dismantling of socialist collective institutions has resulted in a loss of critical infrastructure and transformed the landscape into an open access commons that is susceptible to misuse and degradation. As a result of these changing conditions and pressures, the vulnerability of pastoralists has been exacerbated. On the positive side, some customary institutions such as the khot ail are re-emerging to fill the vacuum left by the dissolution of the socialist collective system. Herders are organizing themselves into traditional communities to share common seasonal camps and water resources, share labour, regulate access to pastures, and support traditional practices such as seasonal migration of herds and maintenance of reserve pastures. Herders are also diversifying their herds to include multiple species and age classes, which helps them to hedge against climate and market risk. These customary institutions offer a potential model for reducing vulnerability to climate and other risks, and enabling adaptation (Batima et al, 2008).
Climate Change: Future Projections and Impacts The most recent assessment of the Intergovernmental Panel on Climate Change (IPCC) concludes that increases in annual mean temperature will be well above the global mean in the Tibetan Plateau and above the mean in East Asia (Christensen et al, 2007). The 25th to 75th percentiles of projections of
3243 J&J Vulnerability
1/11/07
10:23 am
Page 82
82 Climate Change and Vulnerability
annual mean temperature change between 1980–1999 and 2080–2099 from 21 models reviewed by the IPCC for the A1B emission scenario range between 3.2 and 4.5°C, with a median of 3.8°C for the Tibetan Plateau. In East Asia, the 25th–75th percentile range is 2.8–4.1°C and the median is 3.3°C. Greater warming is projected for winter than summer and for higher altitudes. Longer, more intense and more frequent heat-waves as well as fewer very cold days are very likely. Precipitation increases are considered likely for winter and summer in East Asia and very likely for winter in the Tibetan Plateau. The 25th–75th percentile range of precipitation changes for the 21 models is +6 to +17 per cent in winter, +5 to +11 per cent in summer and +4 to +14 per cent annual average in East Asia over the 100 year period. In the Tibetan Plateau, the projected changes are +12 to +26 per cent in winter, 0 to +10 per cent in summer and +2 to +13 per cent annual average. More frequent intense precipitation events are very likely in the region. In our analyses of the possible impacts of climate change on pasture productivity and livestock, we used climate projections from three general circulation models (GCMs), HadCM3, ECHAM4 and CSIRO-Mk2b, for two scenarios of future emissions (A2 and B2) to construct scenarios for Mongolia (Natsagdorj et al, 2005). Our scenarios of temperature and precipitation changes derived from these model experiments are presented in Table 4.5. Table 4.5 Projections of temperature and precipitation changes in Mongolia relative to 1961–1990 baseline for SRES A2 and B2 emission scenarios Model
A2 – medium-high emissions B2 – medium-low emissions 2020
2050
2080
2020
2050
2080
T,°C P,% T,°C P,% T,°C P,% T,°C P,% T,°C P,% T,°C P,% Winter HadCM3 ECHAM4 CSIRO-Mk2b
0.9 23.6 2.4 38.7 3.9 67.0 1.0 16.5 1.7 34.4 2.5 54.7 3.6 59.0 5.7 80.9 8.7 119.4 3.7 11.5 6.0 82.0 6.6 90.3 1.7 12.6 2.9 27.2 5.2 49.0 1.7 14.2 2.7 24.9 3.7 36.8 Summer
HadCM3 ECHAM4 CSIRO-Mk2b
2.0 –2.5 3.5 1.9 7.2 3.7 1.3 –2.1 2.9
7.1 6.5 0.5
6.4 6.4 2.2 6.6 11.3 2.1 5.5 –2.3 1.9
3.1 7.6 0.4
3.3 8.7 4.7 4.5 3.8 5.7 4.9 8.6 3.0 –1.4 4.1 –1.3
The scenarios constructed from all three GCM projections yield greater warming in both winter and summer and much wetter winters than is suggested by the IPCC’s 2007 assessment. The ECHAM4-derived scenarios are the warmest and wettest while the CSIRO-Mk2b-derived scenarios are less warm and less wet. According to all the scenarios, summers in Mongolia will be hotter and drier. The projected evapotranspiration is higher than the projected increase in
3243 J&J Vulnerability
1/11/07
10:23 am
Page 83
Vulnerability of Mongolia’s Pastoralists to Climate Extremes and Changes 83
precipitation, which will dry the soil. Future winters are projected to have milder temperatures but greater snowfall. The impacts of the climate change scenarios on rangelands are simulated using the CENTURY 4.0 model (Metherell et al, 1993). The model yields estimates of above ground biomass and carbon:nitrogen ratio in the vegetation. Results are presented in Tables 4.6, 4.7 and 4.8 for the time periods 2020, 2050 and 2080. Above-ground biomass in 2080 is negatively impacted in the steppe and forest steppe biomes for all scenarios, as is forage quality, represented by the carbon:nitrogen ratio. The reductions in biomass range from -5 to -22 per cent in the steppe and -14 to -37 per cent in the forest steppe. The most severe reductions are for the HadCM3- and ECHAM4-derived scenarios because of their drier summers. In contrast to the steppe and forest steppe, increases in biomass are projected for the high mountains and desert steppe. In most cases, decreases in forage quality are projected for the high mountains and desert steppe. Given the relative importance of the steppe and forest steppe biomass, the overall impact of climate change on Mongolia’s pastures is expected to be negative. Table 4.6 Estimated changes in above-ground biomass and carbon:nitrogen ratio for scenarios derived from the HadCM3 climate model Ecosystem zones
A2 2020
2050
B2 2080
2020
2050
2080
Above-ground biomass Forest steppe Steppe High mountains Desert steppe
–3.8 –2.9 20.2 16.9
–6.0 –5.0 24.9 40.6
–37.2 –19.9 25.3 46.8
–10.4 –0.6 2.5 18.9
–11.9 –0.6 26.2 43.7
–27.8 –7.1 27.9 52.1
–3.8 –1.7 –1.7 –0.4
–7.1 –3.9 –2.2 –0.7
Carbon:nitrogen ratio Forest steppe Steppe High mountains Desert steppe
–1.2 –1.7 –1.2 –1.0
–1.4 –0.7 –1.4 –0.7
–10.4 –7.2 –2.8 –2.0
–2.8 –1.9 –2.1 –0.9
Reduced pasture productivity and forage quality would negatively impact livestock. Livestock would also be negatively impacted by the effects of warmer temperatures on the amount of time they can graze. Models of animal grazing behaviour project decreased grazing time in summer and increased time in winter over much of the land area. Estimates of combined effects of changes in pasture yield, forage quality and grazing time on summer–autumn ewe weight for the HadCM3 scenarios are presented in Table 4.9. Substantial reductions in animal weight are projected for the steppe and forest steppe zones, with modest reductions in the high mountains and nearly no change in the desert steppe.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 84
84 Climate Change and Vulnerability
Table 4.7 Estimated changes in above-ground biomass and carbon:nitrogen ratio for scenarios derived from the ECHAM4 climate model Ecosystem zones
A2 2020
2050
B2 2080
2020
2050
2080
Above-ground biomass Forest steppe Steppe High mountains Desert steppe
8.5 11.8 90.8 50.6
–7.2 3.8 77.6 57.0
–22.3 –11.0 74.0 52.1
7.9 14.0 104.4 60.8
–10.1 –0.4 71.4 51.1
–14.0 –5.5 74.9 31.7
–4.5 –2.6 1.5 –1.0
–5.0 –3.1 1.1 –1.5
Carbon:nitrogen ratio Forest steppe Steppe High mountains Desert steppe
0.6 0.8 4.0 0.5
–3.1 –1.4 2.4 –0.4
–9.1 –5.7 –0.5 –2.2
1.5 1.4 5.6 1.1
Table 4.8 Estimated changes in above-ground biomass and carbon:nitrogen ratio for scenarios derived from the CSIRO-Mk2b climate model Ecosystem zones
A2 2020
2050
B2 2080
2020
2050
2080
Above-ground biomass Forest steppe Steppe High mountains Desert steppe
–4.6 –4.2 11.3 11.1
–6.8 –1.6 34.0 36.9
–30.8 –21.9 34.8 53.0
–7.6 –7.3 11.0 27.9
–12.7 –11.4 8.5 25.9
–16.9 –15.2 62.7 57.6
–4.5 –2.6 1.5 –1.0
–5.0 –3.1 1.1 –1.5
Carbon:nitrogen ratio Forest steppe Steppe High mountains Desert steppe
0.6 0.8 4.0 0.5
–3.1 –1.4 2.4 –0.4
–9.1 –5.7 –0.5 –2.2
1.5 1.4 5.6 1.1
Conclusions Mongolia’s pastoralists historically have been strongly impacted by the extremes of the region’s climate. They are highly vulnerable to climate hazards and their vulnerability has grown in recent decades. Warming and drying of the climate has reduced the quantity and quality of forage on Mongolia’s pasture lands and this has negatively affected livestock and herders’ livelihoods. Despite the warmer temperatures, spatially extensive winter zud events increased significantly in frequency in the 1980s and 1990s relative to the
3243 J&J Vulnerability
1/11/07
10:23 am
Page 85
Vulnerability of Mongolia’s Pastoralists to Climate Extremes and Changes 85
1960s. More than 60 per cent of the country’s land has been shown to be vulnerable to climate-driven extremes.
Table 4.9 Projected changes in ewe weight (%) with combined effects of changes in pasture productivity, forage quality and grazing time for HadCM3 climate change scenarios Ecosystem zones
Forest steppe Steppe High mountain Desert steppe
A2
B2
2020
2050
2080
2020
2050
2080
–10.68 –12.85 –2.92 2.02
–34.40 –31.67 –3.05 3.87
–57.75 –39.50 –9.03 –0.18
–26.97 –15.86 –2.76 2.18
–38.33 –24.10 –3.64 0.96
–53.99 –34.37 –3.77 –0.36
These climatic and environmental changes have happened against a backdrop of tremendous social and economic change as Mongolia transformed itself from a socialist to a market-oriented system. The social and economic changes created greater vulnerability to climate and other shocks within the pastoral communities. Economic opportunities and income contracted, subsidized inputs and services to the livestock sector were eliminated, socialist institutions for the organization and management of livestock production were dissolved, overgrazing degraded pastures in some areas, investment was insufficient to maintain critical infrastructure in the rural areas, and a new class of poor rural households was created. These changes reduced the ability of herders’ households to cope with zud, drought, market variations and other stresses, and increased their vulnerability. New opportunities and new institutions are emerging, and these are helping to reduce vulnerability. Yet vulnerability remains high. This is a worrying situation as Mongolia faces a climate that is changing because of human actions. Our analyses of the future impacts of climate change suggest that pasture yields and forage quality will continue to decline and that these will negatively impact livestock. Our analyses have not taken into account possible changes in zud in the future. As temperatures warm, winters will become milder, offering the potential for less frequent and severe zud events. However, zud is a complex phenomenon that is not simply a function of mean temperature. Higher projected winter precipitation could result in more frequent and greater snowfall. Changes in patterns of daily temperature variability, frequencies of wind storms and other climate parameters will affect zud occurrence. Recent decades have been a time of warming, yet as noted previously, the frequency of zud events that impact more than 25 per cent of Mongolia has increased substantially. How human-driven climate change will affect the frequency, intensity or spatial extent of zud is highly uncertain, but
3243 J&J Vulnerability
1/11/07
10:23 am
Page 86
86 Climate Change and Vulnerability
the outcome will have important consequences for all Mongolia. It will be critical for Mongolia to take up this challenge and explore options for adapting to a changing climate, a challenge that we examine elsewhere (see Batima et al, 2008).
Note 1
Details of this research are reported in Batima (2006).
References Batima, P. (2006) ‘Climate change vulnerability and adaptation in the livestock sector of Mongolia’, Final Report, Project AS06, Assessments of Impacts and Adaptations to Climate Change, International START Secretariat, Washington, DC, US Batima, P., N. Batnasan and B. Lehner (2004) The Freshwater Systems of Western Mongolia’s Great Lakes Basin: Opportunities and Challenges in the Face of Climate Change, ADMON, Ulaanbaatar, Mongolia Batima, P., B. Bat, S. Tserendash and B. Myagmarjav (2008) ‘Adapting to drought, zud and climate change in Mongolia’s grasslands’, in N. Leary, J. Adejuwon, V. Barros, I. Burton, J. Kulkarni and R. Lasco (eds) Climate Change and Adaptation, Earthscan, London, UK Bayarbaatar, L., G. Tuvaansuren and D. Tserendorj. 2005. ‘Climate Change and Livestock’. In P. Batima and B. Bayasgalan, eds, Vulnerability and Adaptation Assessment for Grassland Ecosystem and Livestock Sector in Mongolia: Impacts of Climate Change. ADMON. Ulaanbaatar (in Mongolian) Christensen, J. H., B. Hewitson, A. Busuioc, A. Chen, X. Gao, I. Held, R. Jones, R. Koli, W. Kwon, R. Laprise, V. Rueda, L. Mearns, C. Menendez, J. Raisanen, A. Rinke, A. Sarr and P. Whetton (2007) ‘Regional Climate Projections’, in S. Solomon, D. Qin, M. Manning, Z. Chen, M. C. Marquis, K. Averyt, M. Tignor and H. L. Miller (eds) Climate Change 2007: The Physical Science Basis, Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US Chuluun, T., A. Enkh-Amgalan and D. Ojima (2003) ‘Tragedy of the commons during transition to market economy and alternative future for the Mongolian rangelands’, African Journal of Range and Forage Science, vol 20, no 2, p115 Mearns, R. and E. Dulamdary (2000) ‘Sustaining livelihoods on Mongolia’s pastoral commons’, Presented at the 8th Biennial Conference of the International Association for the Study of Common Property, Bloomington, Indiana, May 31 – June 4, Digital Library of the Commons, http://dlc.dlib.indiana.edu/archive/ 00000574/ Metherell, A. K., L. A. Harding, C. V. Cole and W. J. Parton (1993) CENTURY Soil Organic Matter Model Environment, Technical Documentation Agroecosystem, version 4.0., Colorado State University, Fort Collins, CO, US Ministry of Nature and Environment (2001) Annual Report on Environmental Resources, Ministry of Nature and Environment, Ulaanbaatar, Mongolia Ministry of Nature and Environment (2003) Report on Water Resources Inventory, Ministry of Nature and Environment, Ulaanbaatar, Mongolia National Statistical Office (2001) Mongolian Statistical Yearbook 2001, National Statistical Office, Ulaanbaatar, Mongolia
3243 J&J Vulnerability
1/11/07
10:23 am
Page 87
Vulnerability of Mongolia’s Pastoralists to Climate Extremes and Changes 87 National Statistical Office (2003) Mongolian Statistical Yearbook 2003, National Statistical Office, Ulaanbaatar, Mongolia Natsagdorj, L., P. Gomboluudev and P. Batima (2005) ‘Climate change in Mongolia’, in P. Batima and B.Myagmarjav (ed) Vulnerability and Adaptation Assessment for Grassland Ecosystem and Livestock Sector in Mongolia: Current and Future Climate Change, ADMON, Ulaanbaatar (in Mongolian) Tserendash, S., B. Bolortsetseg, P. Batima, G. Sanjid, M. Erdenetuya, T. Ganbaatar and N. Manibazar (2005) ‘Climate change and pasture’, in P. Batima and B. Bayasgalan (eds) Vulnerability and Adaptation Assessment for Grassland Ecosystem and Livestock Sector in Mongolia: Impacts of Climate Change, ADMON, Ulaanbaatar (in Mongolian) Tsevel (1966) The Mongolian Language Vocabulary Dictionary, Academy of Sciences, Ulaanbaatar, Mongolia
3243 J&J Vulnerability
1/11/07
10:23 am
Page 88
5
Resource System Vulnerability to Climate Stresses in the Heihe River Basin of Western China Yongyuan Yin, Nicholas Clinton, Bin Luo and Liangchung Song
Introduction China’s Western Region Development Strategy, announced in 1999, aims to accelerate economic growth in the western provinces and close disparities between the eastern and western regions. The vast region of 6.85 million square kilometres encompasses 71 per cent of the land area of China and is home to 355 million people, or 28 per cent of the total population (Liu and Nielson, 2004). The region is rich in many natural resources, but is poor in fertile land and water, is subject to highly variable precipitation, and its soils suffer constant wind erosion. Western lands account for 80 per cent of China’s total eroded lands (Glantz et al, 2001). These features have been and will continue to be important limiting constraints on development of the region. Human-caused climate change is likely to affect the availability of water, agricultural productivity and soil fertility, and could make resource and environmental constraints more biting in the future. Successful development of the west that attains economic and other goals in a sustainable manner will need to be cognizant of the current constraints and how they may evolve with a changing climate. This requires information about vulnerability to current and changing climate hazards, including its nature and causes, its distribution among the population, and its spatial distribution across the landscape, that can be used by decision makers to reduce vulnerability and improve the adaptive capacity of resource systems to cope with climate change. In this chapter we present an approach that we have developed for the assessment of water and land resource system vulnerability along with preliminary results from its application to the Heihe river basin of northwestern China. The basin is a region of predominantly arid and semi-arid climate with highly variable rainfall, scarcity of water and fragile ecological systems. The basin is poor in financial resources, infrastructure, education levels, and access
3243 J&J Vulnerability
1/11/07
10:23 am
Page 89
Resource System Vulnerability in the Heihe River Basin of Western China 89
to technology and markets, creating a situation of low adaptive capacity relative to the coastal region of China. People in the region are facing substantial and multiple stresses, including rapidly growing demands for food and water, poverty, land degradation, water pollution and conflicts over resources. Our analysis shows that resource-related stresses and conflicts result in a high degree of vulnerability in the basin to climatic risks and that these risks may be amplified by climate change. The growing resources vulnerability poses a big challenge for government agencies to manage natural resources more effectively while facilitating development in the basin.
Assessing Resource System Vulnerability We define the vulnerability of a system to climate variation and change as its propensity to undergo impacts from climate perturbations that disrupt the nominal functionality of the system. Our focus is on the vulnerability of natural resource systems, specifically water and land, which provide the physical basis for human life and society. Our conception of natural resource systems encompasses not only the physical resources themselves, but also human interactions with and use of these resources. Natural resources provide three groups of functions: 1) raw materials and energy as inputs of economic production; 2) services that support recreation, amenity wilderness and non-use values and 3) air, water, food and waste assimilation required as life-support for human society. Vulnerability of natural resource systems is concerned with shocks or stresses that disrupt any of these functions. Vulnerability to climate hazards has been described by different authors as having three dimensions: exposure, sensitivity and adaptive capacity (see, for example, Kasperson et al, 2002). Exposure relates to the frequency, intensity and nature of climate events or phenomena with which elements of the system of study come into contact. Sensitivity describes the degree to which a resource system is affected by exposure to climate stimuli. It is obvious that the sensitivity of a system to environmental stresses is influenced by the properties of the system, including its state or condition. A system that is in a degraded, weakened or fragile state from previous exposure to climate or other stresses is often highly sensitive to new exposures. Environmental risk is mainly determined by exposure and sensitivity. Adaptive capacity is the ability of a system to respond to exposures and their effects so as to limit harm, or even to benefit. It is related to resilience, which is the ability of the exposure unit to resist or recover from the damage associated with the convergence of multiple stresses. Smit et al (1999) and Downing et al (1997) use ‘coping range’ to characterize a system’s adaptive capacity. It is suggested that a system will not be seriously damaged as long as the environmental change or variation falls within a system’s coping range, while changes beyond the boundaries of the coping range can result in severe or even disastrous consequences. Smit et al (2000) suggest that defining a coping range can help to measure the adaptive capacity of the system.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 90
90 Climate Change and Vulnerability
Resource vulnerability and environmental risk Resource system vulnerability is closely linked to environmental risk, which can be expressed as: R = P(e) D where R is environmental risk, e is exposure to an environmental perturbation or hazard, P(e) is the probability or likelihood of exposure to the hazard, and D is the damage or consequence of exposure, which is a function of the intensity of the exposure, the sensitivity of the system and its adaptive capacity. In similar fashion we define vulnerability in probabilistic terms: V = P(s) P(e) [1–P(a)] where V is the vulnerability of the system, P(s) is the probability that the system is sensitive to the hazard, P(e) is the probability of exposure to the hazard and [1-P(a)] is the probability that the system lacks capacity to cope with or adapt to the hazard. In this construction, vulnerability is the joint probability that the system is sensitive, that it will be exposed and that it lacks the capacity to adapt, assuming independence among exposure, sensitivity and adaptive capacity. Values of V would lie in the interval [0, 1].With zero probability of any one of the component factors, the system is not considered to be vulnerable. An alternative conception would be to interpret each of the components as an index number, normalized to lie between 0 and 1, whose value is the proportion of the maximal degree of exposure, sensitivity or capacity.
Statistical measurement methods Climate risk also has been expressed as a statistical measure of the extent or duration of a resource system failure under climate stresses, should a failure occur. The extent of a system failure is the amount an observed statistical value exceeds or falls short of a critical threshold. For example, water system vulnerability can be measured as the volume of water by which supply falls short of meeting demand for a certain amount of water for a municipality, or to continually release water above a minimum flow rate from a reservoir. Water system vulnerability can also be measured as the average deficit occurring during failures to meet a target, as well as the severity of failures. For example, if we use river flow, F, as an indicator to measure vulnerability, the water system vulnerability can be calculated by: EVf = Max [0, LFt - Ft, Ft - UFt] where EVf is the water system’s maximum-extent vulnerability based on the river flow indicator; LFt and UFt are the lower and upper critical thresholds of the coping range respectively; and Ft is the observed river flow (ASCE and UNESCO/IHP, 1997). If the observed data for the system performance indi-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 91
Resource System Vulnerability in the Heihe River Basin of Western China 91
cator values lie within the upper and lower thresholds (within the coping range), we assume that the range of values is satisfactory, acceptable or not vulnerable. Statistics or observed data above the upper threshold or below the lower limit are considered as unsatisfactory or vulnerable. It should be noted that these coping ranges may change over time.
Using indicators to measure resource vulnerability In resource vulnerability assessment, indicators are used as decision criteria or standards by which the degree or the group of resource vulnerability class can be identified. ASCE and UNESCO/IHP (1997) suggest a list of vulnerability indicators that can be used for water resource vulnerability assessment. Each indicator is composed of a number of attributes or variables which are measurable by using existing sources of information in most cases. For example, Lane et al (1999) define reservoir system vulnerability as the magnitude of a water supply failure relative to annual yield. The South Pacific Applied Geoscience Commission (SOPAC) developed an environmental vulnerability indicator (EVI) that uses a hierarchical structure to aggregate 47 individual indicators (Kaly et al, 1999). The purpose of the EVI is to evaluate the significance of environmental vulnerability of a nation facing alternative stresses or hazards. The EVI includes three levels. The top level, the EVI, consists of three composite sub-indices: a risk exposure subindex (REI), an intrinsic resilience sub-index (IRI) and an extrinsic resilience index (ERI). Each sub-index is constructed from a number of indicators used to measure different determinants of the environmental vulnerability of a nation. The Famine Early Warning System (FEWS) of the US Agency for International Development (USAID) adopted various methods to conduct comprehensive assessments of vulnerability in poor southern African countries (USAID, 1997). FEWS uses an indicator method to identify geographic locations which are vulnerable to climate variables and stresses. The results are used to classify areas as slightly, moderately, highly or extremely vulnerable to famine. A vulnerability analysis mapping (VAM) project was also used by the same study in Mozambique for vulnerability assessment. The VAM applied classification methods to generate flood risk maps, drought risk maps, food systems maps, land-use maps, market access maps, and health and nutritional profile maps.
Vulnerability assessment approach developed for this study The climate vulnerability assessment in this study, outlined in Figure 5.1, has four components: 1) climate scenarios, 2) sensitivity analysis, 3) vulnerability indicators and 4) system vulnerability to climate stimuli. Given the large amount of research and literature on climate change scenario development and climate sensitivity analysis, we focus in this chapter on the selection of vulnerability indicators, their integration into an assessment of vulnerability and mapping of vulnerability.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 92
92 Climate Change and Vulnerability
Sensitivity analysis
Assessing present-day vulnerabilities of water and land resources to climate stresses, and mapping climate vulnerabilities in the Heihe River Basin
Climate vulnerability indicators
Multiple stakeholders, planners, analysts, and public
Current climate variability and climate change scenarios
Figure 5.1 Flow-chart of the general research approach Several key vulnerability indicators are selected to measure resource vulnerability under current climate conditions in the Heihe river basin. GIS (geographic information system) is used to identify the spatial distribution of water resource vulnerability by combining the indicators of domestic water deficit (Srdjevic et al, 2003) and irrigation deficit (Qi and Cheng, 1998) into a composite indicator. Maps, tables and figures provide visual displays of resource system vulnerability, which can help policymakers identify the most vulnerable subunits. It should be pointed out that there have been few practical applications of such an approach, particularly in climate vulnerability research. Results of our assessment of vulnerability to current climate variation establishes a baseline set of measurements and observations that could be used to measure progress toward reducing vulnerability to future climate change. The various vulnerability indicators can be applied to project potential vulnerabilities of the resource systems in the future using climate change scenarios. In this way, the research on present vulnerabilities of resource systems can provide insights into potential impacts and vulnerabilities associated with future climate change. The methodology developed and applied by our study provides, we hope, a useful approach that could be replicated and extended in other studies.
Case Study of the Heihe River Basin The Heihe river basin is located in a region bounded by latitudes 35.4 and 43.5°N and longitudes 96.45 and 102.8°E. A map of the study region is shown in Figure 5.2. The study area is the second largest inland river basin in the arid region of northwestern China. The basin includes parts of two provinces, Qinghai and Gansu, and the Inner Mongolia Autonomous Region. With an area of 128,000 square kilometres, the basin accommodates a population of 1.8 million living in 11 counties, three small cities and five prefectures. The region
3243 J&J Vulnerability
1/11/07
10:23 am
Page 93
Resource System Vulnerability in the Heihe River Basin of Western China 93
is composed of diverse ecosystems, including mountain, oasis, forest, grassland and desert. The Heihe river flows from a headwater in the Qilian Mountain area to an alluvial plain with oasis agriculture, and then enters deserts in Inner Mongolia, representing the upper, middle and lower reaches of the basin respectively. The total length of the Heihe river is 821 kilometres.
Figure 5.2 Map of the Heihe river basin with approximate population distribution shown in shades of grey (black is highest population density)
3243 J&J Vulnerability
1/11/07
10:23 am
Page 94
94 Climate Change and Vulnerability
The Heihe river basin has a typical arid and semi-arid continental climate, characterized by low and irregular rainfall, high evaporation and recurrent drought. The basin can be divided into three typical climate zones following the altitudinal gradient. On the southern slope of the Qilian Mountain, the climate is wet and cold with a mean annual precipitation ranging from 300 to 500mm. In the middle reach, the climate becomes much dryer and the mean annual precipitation is only 100–200mm. In the lower reach, the average annual precipitation is less than 60mm, making it one of the driest areas at the same latitude on Earth (Digital River Basin, undated). Great temporal variations in temperature and precipitation also exist over the Heihe River Basin, with about 50–70 per cent of the precipitation recorded during the summer. Mean daily temperature ranges from –14°C to 3°C in January and from 11°C to 27°C in July (Gansu Meteorological Bureau, 2000). The Heihe river basin is a poor region in China with a harsh environment and fragile ecological systems. The region is critically short of water and arable land, deficient in educated, technical and scientific personnel, and far from even domestic markets. The major economic sector in the region is agriculture, and irrigation is crucial for crop production. The leading crops are wheat, potatoes and corn. The oasis agriculture relies on irrigation from the Heihe river and its tributaries. While the basin has fostered the development of much oasis agriculture in the middle reach, rangeland farming in the upper reach and herdsmen in the lower reach, towns, small hydropower plants, a large number of rural communities, and government agencies, climate stresses have imposed considerable economic, social and environmental impacts. With a resource-based economy, the study region is very sensitive to climate. People in the basin face substantial and multiple stresses, including rapidly growing demands for food and water, large populations at risk of poverty, degradation of land and water quality, and other issues that may be amplified by climate change. Drought is one of the main climate hazards in the basin, with characteristics of high frequency and significant damage. For example, droughts occurred in the middle reach in about 50 per cent of the years since 1951. During drought years, while the precipitation volume of May and June is remarkably lower than the mean, the annual evaporation demand remains at 2000 to 2650mm (Chen and Qu, 1992). Under climate change conditions, periods of drought are likely to become more frequent and severe, and water shortages may increase water-use conflicts. Land degradation problems and limited water supplies restrict present agricultural production and threaten the food security of the region. Climate change may cause negative impacts on food and fibre production in the region (Shi, 1995). In addition, decreases in water availability and food production would lead to indirect impacts on human health. Kang et al (1999) suggested that spring outflow at the mountain outlet would increase while summer flow would decline by 2030 under climate change scenarios. Irrigation demand in the summer accounts for more than 70 per cent of the total agricultural water consumption in the region. This seasonal shift of water supply will affect agricultural production considerably.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 95
Resource System Vulnerability in the Heihe River Basin of Western China 95
There is already some evidence of an observed positive trend in temperatures over the past 50 years, with a more significant rise in the Qilian Mountain area. During this period, annual average temperature increased more than 1°C in Sunan County and 0.9°C in upper mountain areas. The Qilian Mountain glaciers are already undergoing a rapid retreat at a rate of about one metre annually. The region depends on spring melt from the glaciers for the main water supply (Cheng, 1997). A declining water supply has already affected land resources, with large areas of farmland undergoing desertification (Digital River Basin, undated).
Climate exposure In identifying present-day climate risks, existing climate variation patterns need to be specified. The climate change trend in northwest China for the past 50 years was investigated by analysing temperature and rainfall data from 1951 to 2004 (Wang, 2005). Climate change scenario specification for this study represents the possible future climate conditions under various assumptions. Based on eight coupled global atmospheric and oceanic circulation models (AOGCMs), the climate change projection over west China for the 21st century was calculated by the NCC/IAP (National Climate Center/Institute of Atmospheric Physics) T63 simulation model (Xu et al, 2003). Ding et al (2005) applied a regional climate model (Ncc/RegCM2) nested with a coupled GCM (NCCT63L16/T63L30) and Hadley Center model (HadCM2) for climate change studies. Outputs of the Chinese regional-scale climate model were used to design scenarios (Li and Ding, 2004). Results for western China from the regional climate experiment for the IPCC SRES AS emission scenario project an increase of 0.4ºC in mean annual temperature for the period 2020–2030, with greater warming in summer, and a slight decrease in precipitation (Yin, 2006).
Sensitivity analysis The purpose of sensitivity analysis is to identify those climate variables possessing relative importance in determining resource system vulnerability. In addition, sensitivity analysis can indicate those key aspects of resource systems which are sensitive to certain climate variables. Since the relations between climate variables and various system aspects are based on historical statistics or experience, this kind of information can be derived from experts or stakeholder consultation. In this connection, stakeholder workshops and surveys were carried out to provide sensitivity information. The consultation with stakeholders on climate sensitivity is also part of the capacity-building process of the study. In the consultation process, stakeholders identified key climate stresses, vulnerability indicators and critical thresholds they use in resource management. Climate sensitivity analysis through stakeholder consultation in this study followed an approach used in the Hunter Valley case study (Hennessy and Jones, 1999). A potential sensitivity matrix was generated during a stakehold-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 96
96 Climate Change and Vulnerability
er workshop to identify climate variables with the greatest forcing and activities with the broadest sensitivity to climate in the study region. The sensitivity results indicate that water shortage is the main concern of the study region. Rainfall variability and soil moisture levels have the greatest impact, while temperature has only a moderate effect. Obviously, rural resource-use activities are very sensitive to climate events.
Identification of critical vulnerability indicators The research procedure follows with an identification of indicators to measure resource vulnerability in the study. To select critical indicators for vulnerability assessment, the first major source of information used for the study was government reports, documents and other published materials on resource issues. Based on existing key policy concerns in the region, indicators for measuring resource vulnerability under climate stimuli were identified. Some operationally useful key indicators in vulnerability and adaptive capacity assessment are listed in Table 5.1. Table 5.1 Potential determinants (climate and other variables with forcing) and resource indicators Climate and Other Related Factors
Resource Vulnerability Indicators
Rainfall variability Maximum temperature Soil moisture Minimum temperature Wind Cold snap Heat stress days Accumulated degree days Cropping area Population growth Economic growth Technology Consumption Urban expansion Resource management Government policies
Food security Farm income Water scarcity (withdrawal ratio) Drought hazards Palmer drought severity index (PDSI) Water use conflicts Arable land loss Groundwater stress Salinity Soil erosion Grassland deterioration
The preceding discussion indicates that climate risks and vulnerabilities are determined by many factors, including climate stimuli, system sensitivity to climate, adaptive capacity and other response options to deal with risks. The factors that influence the system exposure risk (Table 5.1, left column) can mainly be divided into climate stimuli (rainfall variability, temperature extremes and so on), properties of the resource use systems (resource management), and economic and social forces (economic and population growth). These factors
3243 J&J Vulnerability
1/11/07
10:23 am
Page 97
Resource System Vulnerability in the Heihe River Basin of Western China 97
affect the spatial distribution of climate impacts and adaptive capacity which could result in significant differences in climate risks and vulnerabilities geographically. Since resource system vulnerability is related to failures of the resource system to provide economic, social and ecological functions to meet societal demands, indicators listed in the right column of Table 5.1 reflect some aspects of these functions. It is obvious that farm income is one of the most important indicators for measuring vulnerability. Improvements in economic return will also reduce system vulnerability (Yohe and Tol, 2002). In China, providing enough food for the country’s 1.3 billion people is always a big challenge, and there is increasing concern about China’s food security or its ability to feed itself. The provision of adequate food on a continual basis is a major indicator of regional sustainability. The food security indicator reflects the ability to achieve higher levels of self-sufficiency, and it can be used to check whether the resource base can provide enough food supply. Early in 2000, the Chinese central government launched a major new initiative to develop China’s poor, underdeveloped western regions. China’s Western Region Development Strategy has opened a new chapter of economic growth and expansion in China’s western provinces. The motivations behind the Western Region Development Strategy are aimed at rapid changes in western China over the next few decades and easing the income disparities between coastal and interior China. Stimulated by this new development strategy, many industrial and housing developments have been sited in productive farmland, forestry and wilderness areas. How to slow down the conversion of farmland to urban and industrial uses is critical for regional sustainability in western China. Thus a further indicator, to protect and conserve arable land, reflects this concern. It is now generally realized that environmental concerns should be incorporated in decision making in an effort to achieve sustainable development (World Commission on Environment and Development, 1987). There are a large number of parameters that can be used as indicators of ecological vulnerability. In Table 5.1, environmental concerns are reflected by the indicators for salinity, soil erosion and grassland deterioration. There is an increasing concern about the implications of climate change for water management (Gleick, 1990) and water-use conflicts in the semi-arid region of western China (Kang et al, 1999). Dealing with potential water-use conflicts under changing climate is therefore considered as an important indicator. The competition over access to water resources in the Heihe river basin has led to disputes, confrontation and many cases of violent clashes. Changing water supply induced by climate warming may increase water-use conflicts in the region. In order to improve the reliability of the information on indicators derived from existing sources, workshops and surveys with stakeholders and decision makers in the region were conducted to discuss major policy issues related to climate change and environmental risks. Various government officials and experts were invited to workshops in the Heihe river basin. Stakeholder rep-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 98
98 Climate Change and Vulnerability
resentatives were consulted to identify key concerns related to resource use in the region and to prioritize these indicators. These representatives included officials from various ministries of Gansu Province, bureaus of municipal governments, research institutes, women’s groups and universities. It was indicated during these workshops that water shortage was the key problem relating to sustainable development in the region. Almost all the problems of ecological unsustainability are caused by water shortage in the region.
Mapping vulnerability Mapping the spatial distribution of system vulnerability is useful in helping policymakers identify the most vulnerable subunits. To show the geographical distribution of the vulnerability levels using indicators, several spatial scales have been considered, ranging from square kilometres, the county level and the sub-basin level to the whole basin, based on data availability and other logistical reasons. For test purpose, this study applied a GIS mapping technique to illustrate resource system vulnerabilities using existing and modelled data at different scales. The probabilistic concept of vulnerability is directly applicable to the geographic assessment of vulnerability and facilitates interdisciplinary synthesis of geographic information. For example, outputs from climate models in the form of projected mean temperatures and precipitation provide an indication of future exposure to potentially deleterious environmental conditions. These outputs are commonly available as geographically referenced grids, which can be used to calculate probability of exposure based on the distributional assumptions of climate model accuracy. Similarly, social information, by county, jurisdiction or other geographic entity, can be collected and assembled into indicators of adaptive capacity (or lack thereof). The probability of adaptive capacity could be inferred from [0,1] scaled indices or percentage of a threshold level for the index (percentage of threshold revenue, for example). Adopting this approach, we have undertaken to map sensitivity as a geographic data set (layer) to be integrated as above with layers that represent exposure probability and lack of adaptive capacity probability. As discussed earlier, the requisite ingredients for vulnerability to a particular event include exposure to the event, sensitivity to events of that magnitude and duration, and lack of adaptive capacity to handle events of that type. Any lack of these key ingredients means the system is not vulnerable. For example, if P (exposure to event) = 0, then vulnerability = 0, and the same applies for the other terms. A map of vulnerability is created by multiplying (using map algebra, a form of overlay analysis for rasters in which the maps are multiplied cell by cell) the input maps of probability of exposure, sensitivity and lack of adaptive capacity. Thus teams with different emphases can produce maps that are easily integrated under this conceptual framework. While probability maps may be readily created (of frequency of observable or projectable events, for example), indicators may also be used as a proxy. The following sections illustrate the use of geographic data to create maps of sensitivity to be integrated with the results of climate simulations and social studies of adaptive capacity.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 99
Resource System Vulnerability in the Heihe River Basin of Western China 99
Preliminary Results of Resource Vulnerability in the Heihe River Basin For illustration purposes, measurements of some key vulnerability indicators were carried out in the case study. While all indicators listed in Table 5.1 were investigated by the research team, this study only calculated resource vulnerability indicators of water withdrawal ratio, water-use conflicts and Palmer Drought Severity Index (PDSI), as well as conducting vulnerability mapping. Based on data availability, PDSI was calculated at river reach level; water withdrawal ratio and water conflict (events of disputes, confrontation and violent clashes for accessing water resources) were measured at the basin scale. Results presented here are for water system vulnerability under current climate conditions. Vulnerability assessments for other indicators and under climate change scenarios are presented in the final report for the research project (Yin et al, 2006).
Water withdrawal ratio One important water vulnerability indicator is the water withdrawal ratio, defined as the ratio of average annual water withdrawal to water availability. Critical thresholds for indicators were set to enable the comparison of indicator values for different areas and to identify their vulnerability levels. If indicator values do not exceed the threshold level, it is assumed that the system will have a relatively benign experience under climate stresses, but beyond the threshold level, the system will suffer significant stress under climate variation and/or change. For example, a critical threshold level for the drought indicator can be determined by the amount of rainfall required in a specific region. It also can be set using more complex methods such as the accumulated deficit in irrigation allocations over a number of seasons (Jones and Page, 2001). For the annual water withdrawal ratio indicator, the World Meteorological Organization (WMO) suggests that values exceeding 20 and 40 per cent of annual water availability be considered as medium and high water stress respectively (WMO, 1997). In northern China, however, 60 per cent is considered by the government to be the threshold for high water stress (Xie, 2000; Gansu Meteorological Bureau, 2000). Annual water availability, water withdrawal and water withdrawal ratios in the Heihe river basin for the years 1991 to 2000 are presented in Table 5.2. The current water withdrawal ratios are extremely high, ranging between 80 and 120 per cent, far exceeding the critical threshold levels set by both the WMO and the Chinese Government. These values suggest very high levels of water stress in the basin. Water stress in the region might intensify in the future because of growing water withdrawals (driven by population and economic growth) and/or decreasing water availability as a result of climate change.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 100
100 Climate Change and Vulnerability
Table 5.2 Water withdrawal ratio in the Heihe river basin, 1991–2000 Year
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 8
3
Water Availability (10 m )
35.90 34.33 35.46 35.57 34.04 34.64 34.63 34.33 34.70 34.84 8
Total water withdrawal (10 m ) Water withdrawal ratio (%)
3
29.02 27.38 35.40 28.81 29.55 35.76 28.01 41.35 35.45 32.33 124
125
100
124
115
96.9
124
83
97.9
108
The Palmer Drought Severity Index The Palmer Drought Severity Index (PDSI) is used as an indicator of the frequency of drought hazards over time. The PDSI was introduced by Palmer (1965) for measurement of meteorological drought. It has been widely used in different regions of the world to study severity of drought hazards (Briffa et al, 1994; Kothavala, 1999; Ntale and Gan, 2003). Because the PDSI can simulate monthly soil moisture content, it is suitable for comparing the severity of drought events among regions with different climate zones and seasons (Makra et al, 2002). The computation of the PDSI begins with a climatic water balance using historic records of monthly precipitation and temperature. Soil moisture storage is considered by dividing the soil profile into two layers. The indicator operates on a monthly time series of precipitation and temperature to produce a single numerical value. Negative PDSI values indicate drought conditions and positive values indicate wet conditions relative to normal conditions for an area. Index values in the range –0.5 to 0.5 represent near normal conditions, –0.5 to –1.0 an incipient dry spell, –1 to –2 a mild drought, –2 to –3 a moderate drought, –3 to –4 a severe drought and beyond –4 an extreme drought. We calculate the PDSI using time series data for monthly temperature and monthly precipitation from 15 meteorological stations in the basin obtained from the meteorological service of Gansu Province. In order to identify spatial variations of drought conditions within the Heihe river basin, the basin is subdivided into four areas based on land use, population distribution and climatic conditions. The low reach is covered by desert and wetland; the lower part of the middle reach includes degraded land; the upper part of the middle reach has most of the agricultural land and includes the majority of the population in the basin; and the upper reach is mountains covered by snow and glacier. The calculated point PDSI values are interpolated into these four areas. Growing season temperature has been increasing gradually in recent decades in the low reach of the Heihe river basin, while growing season precipitation has trended downwards. This implies an increasing tendency for drought conditions. Figure 5.3a shows the growing season PDSI values in the low part of the basin. The averaged PDSI value is –0.546, which confirms a tendency towards drought conditions during the past decades. In the lower and upper parts of the middle reach of the Heihe river basin, both temperature and precipitation have trended upwards. The growing sea-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 101
Resource System Vulnerability in the Heihe River Basin of Western China 101
son PDSI in the lower part of the middle reach was negative in most years of the study period, as shown in Figure 5.3b, which indicates that that part of the basin has had a preponderance of drought conditions. In contrast, drought conditions have not dominated in the upper part of the middle reach, though calculated PDSI values are highly variable for the period 1961–2000, as shown in Figure 5.3c. Both temperature and precipitation were increasing during the growing season in the upper reach during the study period. The calculated growing season PDSI values show that wet conditions have prevailed over the region, as shown in Figure 5.3d. This is partly due to high annual precipitation in that mountainous area.
Water-use conflict Water use conflicts are disputes, confrontations and violent clashes about accessing water resources. The number of these events can be traced to show a system’s failure in supplying a certain amount of water for multiple users. In the Heihe river region, various water-use policies and plans have been implemented or designed to limit or prohibit the utilization of water by sectors or regions. Controversies have occurred, of course, as a result of such policies. For the water diversion policy in the Heihe river basin, farmers in the upper and middle reaches of the river already argue that less water for irrigation has led to detrimental consequences in the agricultural sector, while others have indicated that the new policy has been able to revive a dried lake located in the downstream region. Obviously, water policies or regulations may make some sectors or regions worse off and others better off because of their redistributive nature. It is this redistributive nature of policies that often aggravates water-use conflicts. Figure 5.4 illustrates the trend of water-use conflict in the study basin. It shows that the trend of water-use conflict has been increasing in the past decade. The trend of this social indicator suggests that water shortage in the growing season is becoming more and more serious because of decreased water supply and increasing population and per capita water use.
Mapping vulnerability in the Heihe river basin The following indicators were computed from a wide variety of ancillary data sets. These data will be described for each indicator. The indicators focus on weather, agriculture and water resources. Although numerous assumptions are made in the analysis, it is important to keep in mind that the objective is to elucidate geographic patterns. Thus, relative differences between areas are the important trend under investigation. The agricultural sector relies heavily on irrigation water, mainly from river flow and groundwater sources, due to the aridity of the region. For this reason, vulnerability of this sector is largely a function of water resources. To evaluate agricultural vulnerability, both crop water demand and water resource availability were assessed.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 102
102 Climate Change and Vulnerability
a
b Figure 5.3 Growing season PDSI for the Heihe river basin, 1961–2000: a) lower reach; b) lower part of middle reach
3243 J&J Vulnerability
1/11/07
10:23 am
Page 103
Resource System Vulnerability in the Heihe River Basin of Western China 103
c
d Figure 5.3 Growing season PDSI for the Heihe river basin, 1961–2000: c) upper part of middle reach; d) upper reach
3243 J&J Vulnerability
1/11/07
10:23 am
Page 104
104 Climate Change and Vulnerability
Figure 5.4 Trend of water-use conflicts (number of violent events in competition for water) in the study basin
Precipitation and temperature data were collected from meteorological stations distributed through the Heihe river basin. The data were averaged, on a monthly basis, over the period from 1995 to 2000 in order to represent current conditions. Using a method similar to that of Tan et al (2002), these data were interpolated to one-kilometre grid cells and partitioned into infiltration and runoff fractions (in millimetres) using the ‘rational’ method in which runoff is estimated from rainfall, catchment size and infiltration. Rational runoff coefficients were created from inputs of a digital elevation model (DEM), soil type and land cover using the method described as follows. Unique combinations of soil type and land cover were determined from spatial data sets over the project area. Each soil/land cover combination was assigned a United States Department of Agriculture Natural Resources Conservation Service (USDA/NRCS) ‘curve number’ for characterizing soil infiltration and runoff (USDA/NRCS, 1999). The curve numbers were converted to rational runoff coefficients using Tan et al’s (2002) Equation 3 and the DEM on the project area. Runoff was computed based on the rational runoff equation Q = cIA, where Q is discharge, I is rainfall (assumed at constant intensity) and A is catchment area. The runoff was routed to a theoretical channel network (derived from the DEM) to assess monthly values of discharge, geo-located at one-kilometre resolution. The difference between estimated runoff and monthly rainfall was assumed to be infiltration. The spatial distribution of the cropland map of the Heihe river basin was created from data of the International Geosphere-Biosphere Program (IGBP). Crop types were also based on the data in the land cover data set from the IGBP. Estimated crop evapotranspiration was computed using the FAO method described in FAO Drainage and Irrigation Paper 56 (Allen et al, 1999).
3243 J&J Vulnerability
1/11/07
10:23 am
Page 105
Resource System Vulnerability in the Heihe River Basin of Western China 105
The FAO Crop Water Requirement (or CROPWAT) model can be used to estimate some critical values of crop growth and water requirements. The computation of indicators of crop stress or yield index can be achieved using the relationship Yield Index = ETc-stressed/ETc-max (Allen et al, 1999). Weather conditions were based on a combination of the measured data and the CLIMWAT database, which is a set of weather records from observation stations distributed globally. The simulated evapotranspiration data were compared to the estimated rainfall on a monthly basis. It is assumed that all infiltrated rainfall is available for crop growth, while runoff is not. This analysis indicates areas of crop water deficit, meaning that the infiltrated rainfall is insufficient to meet crop demand. Thus, the deficits indicate the amount of irrigation water needed to maximize crop growth. Areas of high deficit will place more pressure on irrigation infrastructure and neighbouring areas of surplus. Figure 5.5 shows the distribution of these high-demand areas.
Figure 5.5 Areas with high irrigation water demand (negative units are in millimetres of deficit) The map indicates that there are geographic differences in demand for irrigation. This is due to the combination of crop type under cultivation (some crops require more water) and local variation in climate conditions, particularly rainfall, temperature and humidity. The areas where high crop water demand and low rainfall converge are areas with high irrigation demand. In the likely event of climate change, areas with high water deficits, as shown in Figure 5.5, will be more affected by fluctuations in the supply of irrigation water. Thus, this deficit can serve as an indicator of vulnerability.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 106
106 Climate Change and Vulnerability
However, irrigation will compete with humans for available water. The Landscan data set was utilized in the analysis of per capita water supply in the Heihe river region (Dobson et al, 2000). It was assumed that none of the infiltrated water was available for humans and that human consumption would rely entirely on runoff. Thus, the per capita water resources index was computed as the annual runoff (supply) divided by the population (demand). Where this index is very low, it indicates that either there are a lot of people or there is a low supply, or a combination of the two. Similar to the analysis of crop water requirements, the areas of low per capita water yield indicate areas of high demand from external sources. These areas may exert more pressure on neighbouring regions with a surplus of water resources. However, there may be cumulative impacts associated with high demand for water for both domestic and agricultural use. Figure 5.6 indicates the areas of high demand for domestic water supplies.
Figure 5.6 Per capita water resources (in cubic meters per capita per annum) Note: Areas of low value (dark colours) indicate a high demand for water resources not available through local supply. Source: The break points are taken from values in Feitelson and Chenoweth (2002).
A simple comparison of available runoff and population – what Feitelson and Chenoweth (2002) call ‘annual per capita internal renewable water resources’ – indicates the geographic areas of high demand. Figure 5.6 shows areas where
3243 J&J Vulnerability
1/11/07
10:23 am
Page 107
Resource System Vulnerability in the Heihe River Basin of Western China 107
the local demand for water is not met by a ‘readily available’ local supply (from local runoff). The ‘dependence ratio’ indicator suggested by Lane et al (1999) describes this deficit as the fraction of the local demand that must be met through water transfers. Areas of supply deficit are likely to be of varying degrees of vulnerability based on the extent of local resources available for remedying this supply/demand imbalance (in other words, the adaptive capacity). For example, the income per capita and the existing level of water supply infrastructure development will determine how well these areas can import water from elsewhere or otherwise provide local people with a safe source of water. This map should be interpreted as the sensitivity of the system, rather than the vulnerability, which is a function of sensitivity, exposure and lack of adaptive capacity. It is notable that some areas are barely at the subsistence level in terms of access to water. Similar to the indicator for the agricultural sector, the areas of high demand for water will be more affected by any fluctuations in supply that result from changes in climate. The above metrics are fairly indirect indicators of sensitivity of the system. Its potential exposure to climate change to future climate change can be illustrated by current trends in weather and climate. The following indicators of exposure – the number of months during which rainfall was 20 per cent lower than the long-term average for that month and the number of days during which the maximum temperature was more than 5°C higher than the mean monthly maximum – are based on the analysis of weather at seven observation stations between 1999 and 2003. These indicators are computed based on the methods proposed in Kaly and Pratt (2000). It is notable that the periods of dry weather are in the upper indices of vulnerability, while the ‘heat-wave’ indicator is much higher than anything discussed in Kaly and Pratt (2000). While Kaly and Pratt (2000) describe the indicators for ‘vulnerability’, here we are using them for exposure. We also created a map to show the number of months over the five-year period during which rainfall was 20 per cent lower than the long-term average for that month. This is an indicator of drought stress and, as such, should be considered in conjunction with the per capita domestic water indicator and the crop water deficit indicators as shown in Figures 5.5 and 5.6. Similarly, another map was created to show the number of days over the five-year period during which the maximum temperature was more than 5°C higher than the mean monthly maximum. This is a ‘heat-wave’ map that should also be considered in conjunction with the other indicators, in terms of illustrating the areas of likely weather extremes. The exposure indices we have used are for historical and illustrative purposes. Ideally, gridded GCM output regarding future exposure to climate extremes would be used to predict areas of likely future vulnerability. These indicators can be scaled and multiplied to obtain a geographic amalgam of vulnerability using the probabilistic framework. To do this, the weather indicators (dry periods and hot periods) were divided by their relative maxima and multiplied to obtain a composite indicator of weather extremes on a scale of [0,1], with 1 representing areas with more frequently observed extreme weather. This composite was used in the construction of two additional indicators, one representing the vulnerability of agriculture to weather extremes (the
3243 J&J Vulnerability
1/11/07
10:23 am
Page 108
108 Climate Change and Vulnerability
agricultural vulnerability indicator) and the other representing the vulnerability of domestic water availability to weather extremes (the domestic vulnerability indicator). The per capita water indicator and crop water deficit indicators were normalized to [0,1], with 1 (most vulnerable) corresponding to the lowest per capita water availability and the highest irrigation deficit respectively. Each normalized indicator was then multiplied by the composite weather indicator in the manner described by the probabilistic equation of vulnerability (and assuming each area to have equivalent adaptive capacity) to obtain two indices of vulnerability for the respective water uses. Here we assume adaptive capacity is constant across the project area; it is therefore not included in the analysis. However, this additional characteristic could be readily incorporated into the analysis given suitable geographic data. By using the scaled indicators as proxies for probability estimates, the vulnerability indices represent the likelihood of there being a confluence of extreme weather and ‘marginal’ conditions in the form of very low water availability or a high irrigation deficit on local cropland. This interpretation is justified by the fact that these marginal conditions will be aggravated by hot weather increasing evapotranspiration of crops and evaporation of water from channels or other impoundments, and by a lack of rainfall that restricts replenishment of supply. As agricultural and domestic water are competing uses, it is logical to combine these two vulnerability indictors into a composite indicator that reveals in what areas there are likely to be shortages and conflicts when there are adverse weather conditions. The composite indicator was created by adding the agricultural and domestic vulnerability indicators, summing over a rectangular 3-pixel neighbourhood and scaling to [0,1]. The rationale behind this manipulation is based on the assumption that in a three-kilometre neighbourhood, competing water uses will compound each other and result in a localized area of higher vulnerability (of both uses) to adverse weather conditions. This composite indicator is shown in Figure 5.7. In keeping with our probabilistic conceptual framework, Figure 5.7 can be interpreted as the likelihood of water resource system vulnerability, a critical system in this arid region. The high vulnerability areas (values close to one) are determined by the confluence of water resource system sensitivity and exposure to environmental extremes. Since domestic and agricultural uses compound each other (are additive effects) in this analysis, having a low vulnerability in either of these sectors does not render the area not vulnerable, but vulnerability is reduced by the absence of competing uses. Of course, the vulnerability is relative to the scale of analysis, since the indicators were normalized based on regional extremes. By expanding the scope of the analysis, it is likely that new extremes would be introduced and the indices would automatically adjust themselves accordingly. The obvious area of high vulnerability in the southeast portion of the region consists of a population centre, with high agricultural production, in an area that has historically experienced deleterious climate conditions (not enough rain and too hot). Areas of very low vulnerability may be due to either the absence of extreme weather conditions or the absence of an irrigation deficit and a high per capita water supply. This is consistent with what we
3243 J&J Vulnerability
1/11/07
10:23 am
Page 109
Resource System Vulnerability in the Heihe River Basin of Western China 109
would expect for this region, but with one caveat. As noted earlier, we did not incorporate a lack of adaptive capacity indicator in this analysis. Therefore, the vulnerability of rural areas with no means of coping with climatic extremes and high sensitivity may be underestimated. Similarly, the vulnerability of relatively urban areas with more extensive infrastructure, monetary resources and/or political clout may be over estimated. It is notable that there is a large portion of Figure 5.5 (irrigation deficit) that shows the ‘No Data’ value but that this value has not made its way into Figure 5.7, the map for vulnerability. This is explained by the fact that the map sources we used do not show agriculture in the ‘No Data’ areas (at the one kilometre resolution scale). We therefore assume that there is no demand for irrigation water in those areas. The infiltrated water would probably be used by local vegetation or enter the groundwater, but is, nevertheless, assumed to be unavailable for humans.
Figure 5.7 Geographic distribution of vulnerability to adverse weather conditions in the Heihe river basin Reduced water availability resulting from low rainfall is compounded by the decreased quality of the diminished supply (Qi and Cheng, 1998). While the demands for water resources increase as populations and economies grow, the availability of water is being reduced by climate variation (See Figure 5.3). The competition over access to water resources in the Heihe river basin has led to increasing disputes, confrontation and many cases of violent clashes (See Figure 5.4). The growing water-use conflicts have posed a big challenge for local government agencies to implement effective water allocation policies.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 110
110 Climate Change and Vulnerability
The composite indicator represents the vulnerability of both agricultural and domestic water users to unfavourable weather conditions in the form of long hot and dry spells. It should not be interpreted as an absolute measure of vulnerability, rather as a way of identifying areas of high relative vulnerability within the region. The frequency distribution of different vulnerability levels is shown in Figure 5.8. The histogram illustrates a bimodal distribution with values concentrated in the lower range and a fairly small number in the upper range. This is appropriate given that the analysis was designed to identify areas of high vulnerability for the purpose of building the adaptive capacity in those areas.
Figure 5.8 Histogram of composite water-use vulnerability levels in the Heihe river basin Note: Vertical axis is frequency of grid cell values.
Conclusions The chapter seeks to provide answers to some important questions in relation to climate vulnerability assessment. It provides information on the geographical distribution of current climate vulnerability levels in the Heihe river basin region. The results indicate the relative vulnerability levels of water and land resources in different areas exposed to current climate stimuli. The vulnerability indicator measurements for resource systems can be applied to project potential vulnerabilities of the resource systems to future climate change scenarios. In addition, the chapter contributes to methodological development in vulnerability assessment and mapping.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 111
Resource System Vulnerability in the Heihe River Basin of Western China 111
By using vulnerability indicators, the climate vulnerability of the study region under current climate conditions has been investigated. The methods for the compilation of indicators, geographic allocation and synthesis are valid for other regions as well. By taking a probabilistic approach, the framework automatically scales up due to the consistency of the [0,1] scale. If true probability measures are unavailable (though they frequently are available), the analysis is automatically normalized to regional extremes. Thus, the method should be viewed as portable, but intra-region comparisons will not, in general, be valid. The applications presented here are intended to benefit future studies that aim to assess resource system vulnerability. The consideration of scale will be important in the determination of what indicators are necessary and feasible for inclusion in any potential climate vulnerability assessment. In vulnerability and adaptive capacity measurement, many of the indicators can be expressed in numerical terms, particularly for climate and physical variables. It is also recognized in the case study, however, that many indicators cannot be quantified and that many of the threshold levels can only be qualitatively described. As a result, some data used in the case study are fairly abstract and not particularly meaningful out of context. It is also notable that the indicator is only mapped over areas of agriculture, as the crop indicator was included in the composite, and is undefined over areas without agriculture. However, assuming food production to be an important element of society and a logical starting point for vulnerability reduction, this composite indicator is informative. Specifically, the areas of highest vulnerability, as evidenced by the indicators, have been narrowed down to several square kilometres. Assuming constraints to the adaptive capacity of the entire region, these places could be designated as high priority in terms of implementing effective adaptation strategies to prevent long-term damage from climate change. For examining system vulnerability to climate change, a natural resource system representing particular future conditions needs to be proposed. For example, water resource system design, operation and management policy can be specified over time. The specification will include assumptions regarding system design, operation, and hydrologic and other inputs and demands that are all key aspects of a water system scenario representative of what could occur in the future. Incorporated into that scenario are key indicators of resource vulnerability. The uncertainties arising in estimating future demand or operational changes can be comparable to those associated with projecting climate change, and can be equally complicated for vulnerability assessments. As a pilot study, the methods here can be critiqued. When applying vulnerability assessment methods in the study, vulnerability indicator selection and vulnerability measurement were not carried out in a comprehensive and systematic way. This is in large part a result of spatial data availability over the study area. However, these methods are effective in vulnerability assessment and mapping spatial distributions of resource system vulnerability. When future climate change and socioeconomic scenarios are available, these methods can also be applied to estimate indicator values in the future. This will produce future vulnerability data for each indicator.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 112
112 Climate Change and Vulnerability
References Allen, R. G., L. S. Pereira, D. Raes and M. Smith (1999) Crop Evapotranspiration: Guidelines for Computing Crop Water Requirements, FAO Irrigation and Drainage Paper No. 56, Food and Agriculure Organization of the United Nations ASCE and UNESCO/IHP (1997) Sustainability Criteria for Water Resource Systems, Task Committee on Sustainability Criteria, Water Resources Planning and Management Division, American Society of Civil Engineers and Working Group of UNESCO/IHP IV Project M-4.3 Briffa, K., P. Jones and M. Hulme (1994) ‘Summer moisture availability across Europe, 1892–1991: An analysis based on the Palmer drought severity index’, International Journal of Climatology, vol 14, pp475–506 Chen, L. H. and T. G. Qu (1992). ‘Rational development and utilization on water and soil resources in the Heihe region’ (in Chinese with English abstract), Science Press, Beijing, pp143–176 Cheng, G. (ed) (1997) Assessing Climate Change Impacts on Snow Pack, Glaciers and Permafrost in China, Gansu Culture Press, Lanzhou, China (in Chinese) Digital River Basin (undated) ‘Digital River Basin’, Heihe River Basin website, http://heihe.westgis.ac.cn (accessed July 2007) Ding, Y. H., Q. P. Li and W. J. Dong (2005) ‘A numerical simulation study of the impacts of vegetation changes on regional climate in China’ Acta Meteorological Sinica, vol 63, no 5, pp604–621 (in Chinese) Dobson, J. E., E. A. Bright, P. R. Coleman, R. C. Durfee and B. A. Worley (2000) ‘LandScan: A global population database for estimating populations at risk’, Photogrammatic Engineering and Remote Sensing, vol 66, pp849–857 Downing, T. E., L. Ringius, M. Hulme and D. Waughray (1997) ‘Adapting to climate change in Africa’, Mitigation and Adaptation Strategies for Global Change, vol 2, pp19–44 Feitelson, E. and J. Chenoweth (2002) ‘Water poverty towards a meaningful indicator’, Water Policy, vol 4, pp263–281 Gansu Meteorological Bureau (2000) Agricultural Ecosystem Database, report, Gansu Meteorological Bureau, Lanzhou, China Glantz, M. H., Q. Ye and Q. Ge (2001) ‘China’s western region development strategy and urgent need to address creeping environmental problems’, Arid Lands Newsletter, vol 49 (http://ag.arizona.edu/OALS/ALN/ALNHome.html; accessed July 2007) Gleick, P. H. (1990) ‘Vulnerabilities of water system’, in P. Wagonner (ed) Climate Change and US Water Resources, John Wiley and Sons, New York, pp223–240 Hennessy, K. J. and R. N. Jones (1999) Climate Change Impacts in the Hunter Valley: Stakeholder Workshop Report, CSIRO Atmospheric Research, Melbourne Jones, R. N. and C. M. Page (2001) ‘Assessing the risk of climate change on the water resources of the Macquarie river catchment’, in F. Ghassemi, P. Whetton, R. Little and M. Littleboy (eds) Integrating Models for Natural Resources Management Across Disciplines, Issues and Scales, vol 2, Modelling and Simulation Society of Australia and New Zealand, Canberra, Australia, pp673–678 Kaly, U. L., L. Briguglio, H. McLeod, S. Schmall, C. Pratt and R. Pal (1999) Environmental Vulnerability Index (EVI) to Summarise National Environmental Vulnerability Profiles, Report to NZODA, SOPAC, Suva, Fiji Kaly, U. and C. Pratt (2000) Environmental Vulnerability Index: Development and Provisional Indices and Profiles for Fiji, Samoa, Tuvalu, and Vanuatu, Phase II Report for NZODA, SOPAC Technical Report 306, SOPAC, Suva, Fiji Kasperson, J. X., R. E. Kasperson, B. L. Turner, W. Hsieh and A. Schiller (2002) ‘Vulnerability to global environmental change’, in A. Diekmann, T. Dietz, C. Jaeger
3243 J&J Vulnerability
1/11/07
10:23 am
Page 113
Resource System Vulnerability in the Heihe River Basin of Western China 113 and E. Rosa (eds) The Human Dimensions of Global Environmental Change, MIT Press, Cambridge, MA, US Kothavala, Z. (1999) ‘The duration and severity of drought over eastern Australia simulated by a couple ocean-atmosphere GCM with a transient increase in CO2’, Environmental Modelling and Software, vol 14, pp243–252 Lane, M. E., P. H. Kirshen and R. M. Vogel (1999) ‘Indicators of impacts of global climate change on US water resources’, Journal of Water Resources, Planning and Management, vol 125, pp194–204 Li, Q. P. and Y. H. Ding (2004) ‘Multi-year simulation of the East Asian monsoon and precipitation in China using a regional climate model and evaluation’, Acta Meteorologica Sinica, vol 62, no 2, pp140–153 (in Chinese) Liu, D. and W. Neilson (eds) (2004) China’s West Region Development: Domestic Strategies and Global Implications, World Scientific Publishing, Singapore Makra, L., Sz. Horváth, P. Pongrácz and J. Mike (2002) ‘Long-term climate deviations: An alternative approach and application on the Palmer drought severity index in Hungary’, Physics and Chemistry of the Earth, vol 27, pp1063–1071 Ntale, H. K. and T. Y. Han (2003) ‘Drought indices and their application to east Africa’, International Journal of Climatology, vol 23, pp1335–1357 Palmer, W. C. (1965) ‘Meteorological drought‘, Research Paper, vol 45, Weather Bureau, US Department of Commerce Qi, F. and G. Cheng (1998) ‘Current situation, problems and rational utilization of water resources in arid North-Western China’, Journal of Arid Environments, vol 40, pp373–382 Shi, Y. (1995) Impacts of Climate Change on Water Resources in North-western and Northern China, Shandong Science and Technology Press, Jinan, China Smit, B., I. Burton, R. J. T. Klein and R. Street (1999) ‘The science of adaptation: A framework for assessment’, Mitigation and Adaptation Strategies for Global Change, vol 4, pp199–213 Smit, B., I. Burton, R. J. T. Klein and J. Wandel (2000) ‘An anatomy of adaptation to climate change and variability’, Climatic Change, vol 45, pp223–251 Srdjevic, B., Y. D. P. Medeiros and A. S. Faria (2003) ‘An objective multi-criteria evaluation of water management scenarios’, Water Resources Management, vol 18, pp35–54 Tan, C. H., A. M. Melesse and S. S. Yeh (2002) ‘Remote sensing and GIS in runoff coefficient estimation in China, Taipei’, Proceedings of the 23rd Asian Conference on Remote Sensing, Kathmandu, Nepal, November, www.gisdevelopment.net/ aars/acrs/2002/pos3/217.pdf USAID (1997) FEWS Project: Vulnerability Assessment, published for USAID, Bureau for Africa, Disaster Response Co-ordination, Arlington, VA USDA/NRCS (1999) ‘SCS Runoff Equation’, Employee Training Module 205, www.wcc.nrcs.usda,gov/hydro/hydro-training-course.html Wang, Z. Y. (2005) ‘Climate change analysis for Western China: 1951–2004’, PhD thesis, China Meteorological Administration, Beijing, China WMO (1997) Comprehensive Assessment of the Freshwater Resources of the World, overview document, World Meteorological Organization, Geneva World Commission on Environment and Development (1987) Our Common Future, Report of the World Commission on Environment and Development, Oxford University Press, Oxford, UK Xie, J. (ed) (2000) Northwestern China Arid Climate Change Research and Projection, Volume II: Drought and Flooding Indicators, China Meteorological Press, Beijing (in Chinese) Xu Y., Y. H. Ding, Z. C. Zhao and J. Zhang (2003) ‘A scenario of seasonal climate change of the 21st century in Northwest China’, Climatic and Environmental Research, vol 8, no 1, pp19–25 (in Chinese)
3243 J&J Vulnerability
1/11/07
10:23 am
Page 114
114 Climate Change and Vulnerability Yin, Y. Y. (2006) ‘Vulnerability and adaptation to climate variability and change in Western China’, Final Report, AIACC Project No AS25, International START Secretariat, Washington, DC, US, www.aiaccproject.org Yohe, G. and R. S. J. Tol (2002) ‘Indicators for social and economic coping capacity: Moving toward a working definition of adaptive capacity’, Global Environmental Change, vol 12, pp25–40
3243 J&J Vulnerability
1/11/07
10:23 am
Page 115
Part III:
Coastal Areas
3243 J&J Vulnerability
1/11/07
10:23 am
Page 116
3243 J&J Vulnerability
1/11/07
10:23 am
Page 117
6
Storm Surges, Rising Seas and Flood Risks in Metropolitan Buenos Aires Vicente Barros, Angel Menéndez, Claudia Natenzon, Roberto Kokot, Jorge Codignotto, Mariano Re, Pablo Bronstein, Inés Camilloni, Sebastián Ludueña, Diego Rios and Silvia G. González
Introduction The metropolitan region of Buenos Aires, which is the political, financial and cultural centre of Argentina, is located on the Argentine coast of the Plata river and is home to nearly a third of the country’s population. The total number of people living along the Plata river coast, including the Buenos Aires region, is almost 14 million. A considerable portion of this coastal area is low-lying land, between 2.8 and 5m above mean sea level, and is often subject to recurrent storm surge floods, which are common to this region due to the unique features of the Plata river estuary. Such storm surge floods, locally known as sudestadas, are expected to become more frequent as the mean sea level rises due to global climate change. The very low-lying areas will, in fact, probably be permanently flooded by the end of this century. These areas are, however, largely uninhabited due to their frequent exposure to storm surges and, as a result, the social impact of future permanent flooding is expected to be small. Climate change vulnerability in this coastal zone would therefore be mostly conditioned by its future exposure to extreme storm surges, especially in the densely populated areas of metropolitan Buenos Aires, where this phenomenon is presently not as common. This potential future vulnerability of the socioeconomically important Argentine coast therefore raises several important questions regarding the exact nature of the climate change impacts and their specific implications for population, infrastructure and the economy, namely: •
How many people and how much infrastructure and real estate are presently affected by storm surge floods with different return periods?
3243 J&J Vulnerability
1/11/07
10:23 am
Page 118
118 Climate Change and Vulnerability
• •
What are the present conditions of this population in terms of social vulnerability and exposure to storm surge floods? Under climate change scenarios, how will the return period of these floods change during the present century and, consequently, how much additional population will be affected at each return period? What will be the extent of damage to real estate and infrastructure?
Responses to these questions are critical for the assessment of the strengths and weaknesses of current socioeconomic and governance systems in this region and for the evaluation of their ability to manage the present and future impacts of extreme storm surges and flooding events due to climate change. We have therefore attempted to fill this important knowledge gap using various investigative techniques with the objective of developing an informative resource that can effectively advise the planning process in the region and help shape appropriate responses to the threat of climate change impacts on the physical, social and economic spheres.
The Physical System The Plata river is a freshwater estuary with unique features. It begins with a width of 50km and widens to 90km at the Montevideo–Cape Piedras section (Figure 6.1), also known as the inner Plata river. The salinity front between fresh- and saltwater is a little downstream of a line connecting Montevideo and Cape Piedras. From here the salinity gradually increases towards the boundary line between Punta del Este and Cape Rasa, where it reaches the ocean’s level of salinity. This 200km-wide boundary line is considered the outer border of the Plata estuary.
Figure 6.1 The Plata river estuary
3243 J&J Vulnerability
1/11/07
10:23 am
Page 119
Storm Surges, Rising Seas and Flood Risks in Buenos Aires 119
The dimensions and shape of the Plata river estuary, together with its very small slope (of the order of 0.01m/km), create quasi-maritime dynamic conditions due to the effects of wind as well as astronomical tides from the sea, which tend to grow in size as they propagate towards the shallower and narrower, inner part of the estuary. Wind storms generated by southeasterly winds are associated with some of the highest wind speeds and when combined with astronomical tides result in events locally known as sudestadas that flood lowlying areas along the coast (Escobar et al, 2004). The floods are more severe on the Argentinean coast than on the Uruguan side because of the Coriolis effect1 and also because the Argentine coast is lower and therefore subject to frequent flooding, especially at Samborombóm Bay. Other low-lying coastal areas are to the south of Greater Buenos Aires, in the floodplains of the MatanzasRiachuelo and Reconquista rivers and at the tip of the Paraná delta. In the city of Buenos Aires, the alert level due to flooding by sudestadas is not raised until water reaches a height of 2.30m above mean sea level because much of the area near the shore is located at an elevation of 1m or more above the mean high water line, which, in turn, is located at 0.99m above the mean sea-level at the Buenos Aires port. Floods caused by sudestadas typically last from a few hours to two or three days. Table 6.1 shows details of flood events, including their return periods and water heights, calculated from a 50-year tide record at the Buenos Aires port. These values are only representative for this particular location, since the coastal storm surge varies at different locations along the coast, intensifying as it progresses towards the interior of the estuary. Table 6.1 Water heights (above mean sea level) at the Buenos Aires port for return periods Return Period (years) 2.5 5 11 27.5 79 366
Height Above Mean Sea Levels (m) 2.50 2.80 3.10 3.40 3.70 4.00
Source: Adapted from D’Onofrio et al (1999).
Three primary factors that can influence water levels in the Plata river estuary include sea-level forcing, wind forcing and tributary forcing. A hydrodynamic model of the estuary was therefore used to assess the response of the river water level to changes in the mean sea level, the direction and intensity of surface winds, and tributary contributions (Barros et al, 2003). The physical influences of these variables on the Plata river estuary are outlined below: •
Sea-level forcing: Because of the small slope and the exceptionally high ratio between the width and length of the estuary, model simulations
3243 J&J Vulnerability
1/11/07
10:23 am
Page 120
120 Climate Change and Vulnerability
•
•
suggest that sea-level increases will propagate towards the inner part of the estuary without great alteration. A decrease is expected only in the inner portions of the Plata river, and at the Paraná delta front the increase is likely to be only 10 per cent of the initial height (Re and Menéndez, 2003). Wind forcing: Wind tension over the surface of the Plata river drags water into or out of the estuary, depending on wind direction, and plays a significant role in modifying water levels. Hydrodynamic model simulations forced by wind data from the National Center for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) (Kalnay et al, 1996) indicate that changes in wind direction and intensity during 1951–2000 due to a southward displacement of the South Atlantic high pressure belt (Escobar et al, 2003), explain about 5cm of the 13cm rise observed at the Buenos Aires port during the 20th century. Model simulations also indicate that the approximately 15cm difference between the summer and winter water levels is caused by the directional shift of wind between these seasons (Re and Menéndez, 2003). Tributary forcing: The mean discharge from the Plata river is about 25,000m3/s, which includes 20,000m3/s from the Paraná River and 5000m3/s from the Uruguay river, its two main tributaries. However, the impact of changes in tributary stream flows was not observed to be proportional. The most extraordinary flooding from both tributaries totalled 80,000m3/s in 1983 (Re and Menéndez, 2003), but the effect of this massive stream flow on the Plata river level was almost insignificant in the outer part of the estuary and was very small in its inner part, except for the Paraná delta front. At Buenos Aires, streamflows greater than 75,000m3/s cause only a 5cm rise in the water level (Barros et al, 2003).
Assessing Present and Future Vulnerability to Recurrent Floods A multidisciplinary approach that accounted for both physical and socioeconomic factors was employed for assessing vulnerability along the Plata river coast in Argentina. Statistical data on physical and social aspects was integrated in a geographical information system (GIS), which was then used to estimate, for different return periods of flood, the population and the public infrastructure (schools, hospitals and so forth) affected and the damage to real estate. A two-dimensional hydrodynamic model with high spatial (2.5km) and temporal (1 minute) resolution and based on the HIDROBID II software (Menéndez, 1990) was used to represent mean and storm surge levels. The model was forced with the astronomical tide at the southern border, the river discharges at the upstream border and the wind field over the whole domain. The model’s capacity to reproduce water-level distribution at the Buenos Aires port for the 1990–1999 period was first verified (Figures 6.2 and 6.3), followed by the estimation of extreme tide values along the coast of the Plata river, thus overcoming the lack of basic data (Figure 6.4). Future sea-level data and future
3243 J&J Vulnerability
1/11/07
10:23 am
Page 121
Storm Surges, Rising Seas and Flood Risks in Buenos Aires 121
wind data from the Intergovernmental Panel on Climate Change’s (IPCC’s) HADCM3 A2 scenario were then used to determine future scenarios of mean water level in the Plata river estuary. For sea-level data other IPCC model outputs were also considered (Church et al, 2001).
Figure 6.2 Frequency of levels (m) above mean sea level at the Buenos Aires port, 1990–1999 Note: Observed frequency is shown by heavy line; simulated frequency by the HIDROBID II model is shown by light line.
Figure 6.3 Mean Plata river level (m) calculated by the HIDROBID II at the Buenos Aires port, 1990–1999 Note: Modelled level is shown in black and observed level is shown in grey.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 122
122 Climate Change and Vulnerability
Figure 6.4 Maximum heights calculated for the storm surge tide with return period of 100 years in the Buenos Aires port For calculating the return periods of floods over land, a surface-level model was constructed with adequate vertical and horizontal resolution. The topography was put together with data from different sources: topographic maps of the Military Geographical Institute, altitude measurements taken by the Buenos Aires city authorities at certain points, field measurements with a differential GPS, and altitude maps constructed with satellite radar. The susceptibility of public buildings was evaluated by means of surveys in each of the 27 administrative districts of the Plata river basin that have territory under the 5m above sea level boundary. Present and future risk to public service infrastructure (including water supply, the sewage network, power facilities, highways and railroads) was assessed, as was the risk to real estate. Costs were estimated as a function of the Plata river level rise over its mean present level and its implications for infrastructure and real estate. Figure 6.5 shows the function of the cost of each flood event, including its depreciation effect on real estate as a function of the rise of the Plata river over its present level. Similar functions were calculated for each component of infrastructure. The sum of costs per event for each of these components was then used to determine the mean annual costs of flooding and, in combination with modelled flood recurrence and duration data, enabled the assessment of damages. Social vulnerability was estimated at the district scale due to the lack of data at smaller sub-levels, recognizing that this option leads to only a firstorder approximation of the geographical distribution of this vulnerability. Social data was obtained from the census and a modest one per cent annual population growth rate was assumed for future scenarios. A social vulnerability index (SVI) was calculated using indicators related to demography, living conditions of the population, and structural production and consumption processes. Values ranging from 1 (lower vulnerability) to 5 (higher vulnerability) were assigned to each indicator. Next an index of social risk (SRI) was developed by multiplying the SVI by an index of exposure to floods (EFI). The latter was calculated from the recurrence of floods at each 1km2 cell and is approximately the inverse of the return period of flooding (RPF).
3243 J&J Vulnerability
1/11/07
10:23 am
Page 123
Storm Surges, Rising Seas and Flood Risks in Buenos Aires 123
Figure 6.5 Damages in real estate per event in millions of US dollars as a function of the water-level rise over the mean present level
Present and Future Vulnerability Flood risk Our findings suggest that the areas at permanent flood risk during the 21st century along the Greater Buenos Aires coast due to climate change-related sea-level rise are very small. The only major concern with regard to permanent flooding could be for the few and sparsely populated very low lying areas at the tip of the Paraná delta and for the new lands that could be added to it in the next few years.2 Therefore, climate change impacts in this region are expected to be mostly due to the increasing inland reach of storm surges or sudestadas in the heavily populated areas. Figure 6.6 shows the flood return periods for the present time. The areas more affected by sudestadas are the southeastern coast of Buenos Aires and the district of Tigre. In both areas, there is a mix of low income communities and upper middle class gated communities. The differences in the storm surge return periods for the 2070/2080 A2 scenario in comparison to present conditions are shown in Figure 6.7. No significant change is observed in the Plata river coast of Buenos Aires and in the districts located immediately to the north, but towards the south of the city and in the Tigre district a significant increase in exposure to recurrent floods can be seen. The A2 scenario predicts a considerable reduction in flood return periods in the valleys of the
3243 J&J Vulnerability
1/11/07
10:23 am
Page 124
124 Climate Change and Vulnerability
Reconquista and Matanzas-Riachuelo rivers, where a socially vulnerable population predominates.
Figure 6.6 Flood return periods in years for present time Note: The number 1 denotes an area below the Reconquista valley and the number 2 denotes the Matanzas-Riachuelo valley; lines show the boundaries of the districts of the metropolitan area of Buenos Aires.
The survey of public properties in the area that could be affected by floods at least once in 100 years in the 2070/2080 A2 scenarios indicates that there are 125 administrative public offices, 17 social security offices, 205 health centres, 928 educational buildings, 92 security buildings and 306 recreational areas, including parks and squares, at risk. In addition, there are also 1046 private industries potentially at risk. Assuming little change in population density and distribution, under the scenario of maximum sea-level rise during the 2070 decade (2070max scenario), the number of people living in areas at flood risk with a return period of 100 years is expected to be about 900,000, almost double the present at-risk population. The relative increment of affected population is even larger for a flood recurrence time of 1 to 5 years, which more than triples for the 2070max scenario in comparison to the present period (Table 6.2). It must be noted that these figures were calculated without considering any growth in population. When a modest 1 per cent annual population growth over the next 70 years is assumed, maintaining the present geographical distribution, the number of people affected for each return period in the 2070s would be double the values in Table 6.2. This means that the population exposed to some risk of flood recurrence every 100 years would be about 1,700,000.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 125
Storm Surges, Rising Seas and Flood Risks in Buenos Aires 125
Figure 6.7 Changes in the return period between 2070–2080 and present time in years
Table 6.2 Present population living in areas that are, or will be, flooded under different scenarios Return Period (years) 1
5
10
20
50
100
1990–2000
33,000
83,000
139,000
190,000
350,000
549,000
2030–2040
102,000
297,000
390,000
500,000
643,000
771,000
2070–2080
113,000
344,000
463,000
563,000
671,000
866,000
Cost of damages Current real estate damage was estimated at US$30 million/year; assuming no socioeconomic changes or no construction of defences, it would reach US$80 and US$300 million/year by 2030–2040 and 2070–2080 respectively under an anticipated 1.5 per cent annual growth rate in the infrastructure. These figures do not include losses to the upper middle class gated communities that are increasingly being built in the coastal areas and neither do they include working hours lost, which could be significant given the size of the population likely to be affected. The figures therefore should be considered as very conservative estimates.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 126
126 Climate Change and Vulnerability
Social vulnerability and risk In tandem with increasing storm-surge risk, the communities that display the highest social vulnerability are those that currently experience a relatively low recurrence of floods and are therefore densely occupied. They are populated by varying social classes, ranging from the upper middle class to the socially vulnerable. The increased inland reach of floods due to the impact of an increased recurrence of sudestadas or storm surges is expected to result in significant economic impacts for these neighbourhoods, including important real estate losses. As previously mentioned, the two most socially vulnerable areas in this region that are highly susceptible to floods are not located directly on the coast of the Plata river, but in the flood valleys of two of its tributaries, the Reconquista and the Matanzas-Riachuelo. The low-lying areas along the coast of the Plata river, which are more exposed to frequent floods and storm surges and are at risk of permanent flooding, are, on the other hand, relatively less vulnerable because they are quite sparsely occupied. Vulnerability in these areas is high only in the few places that are occupied by poor squatter settlements that lack many basic needs as well as any access to social security, and have a child mortality rate higher than the national average. Many households in these settlements are also headed by women. Crime insecurity, unemployment and violence are among the common social issues faced by these communities. The existing social vulnerability of this small group of people is further worsened by the impacts of floods, which are expected to become more severe as the mean sea level rises due to climate change. In contrast to the poor squatter settlements in the flood risk zones of the Plata river and its tributaries, a more recent trend is the occupation of these suburban coastal areas by upper middle class gated communities. These include the southern coast of the Plata river, 20 to 50km to the southeast of Buenos Aires, and the county of Tigre, immediately to the south of the Paraná delta. Beginning in the 1980s, and especially since the 1990s, these zones have been increasingly occupied by upper middle class private gated communities built at an artificially elevated level, above the impact of storm surges. These developments have been largely driven by security concerns about living in the city; the attractions of nature, countryside and greenery (Ríos, 2002); and the enticing view offered by the shorelines. The increased accessibility to these areas, aided by new highways, further fuelled the growth of these private communities. In the early 1990s, private neighbourhoods occupied an area of about 34.4km2; by 2000, this area had grown nearly 10 times to about 305.7km2. In the district of Tigre, 90 private residential developments were authorized in 1998, of which 50 have already been constructed. About 40 per cent of the district’s population now lives in these gated communities. New projects are continually springing up all along the coast, both in the southeastern and in the northern extremes of the Buenos Aires metropolitan area and even in the Paraná delta front (Ríos, 2002). The construction of these gated communities in low-lying lands, historically frequently flooded, requires a massive transformation of the terrain and of the surface drainage, with the associated destruction and replacement of the
3243 J&J Vulnerability
1/11/07
10:23 am
Page 127
Storm Surges, Rising Seas and Flood Risks in Buenos Aires 127
original ecosystems, in order to obtain an elevation which is assumed to be secure from flood waters. Most of these private neighbourhoods are constructed at an altitude of 4.4m above sea level, which is considered to be a level safe from the impacts of floods. However, this assumption may not hold true in the future, when the impacts of climate change may result in more intense storm surges and increase the inland reach of floods. Therefore, though these communities are secure in terms of crime and other social hazards, they now face a new risk from climate change. Outside these gated communities are some of the original squatter settlements, which now face an increased susceptibility to flooding due to the massive land transformation and changes to surface drainage resulting from the construction of these elevated communities, which has effectively destroyed the natural drainage systems in the region. Figure 6.8 shows the current social risk index derived from social vulnerability and flood exposure. This figure is similar to the flood return period figure because of the coincidence of the areas of maximum exposure and social vulnerability. However, there is an important difference in the district of Tigre, in the north, where there is little social vulnerability despite it being a high flood risk zone. This is because these areas are less populated and the occupants are presently the well-to-do private gated communities that are situated at an elevated level, along with perhaps a few poor squatter communities. Therefore, the area of maximum social risk is to the south of this district and in the Reconquista river valley, which is home to the socially vulnerable communities.
Figure 6.8 Social risk index: Present conditions Note: The varying ranges of the social risk index are represented by the various shades of grey (e.g. the darkest shade stands for the range of 25 to 75).
3243 J&J Vulnerability
1/11/07
10:23 am
Page 128
128 Climate Change and Vulnerability
In comparison, in the 2070–80 A2 scenario (Figure 6.9), there is a change in the social risk index due to an increase in flood risk and a worsening of the situation of the more socially vulnerable areas along the Reconquista and Matanza-Riachuelo valleys and to the south of Greater Buenos Aires in a zone relatively far from the coast where few gated communities are expected. There is also a slight worsening of the social risk index of the areas of gated communities in the Tigre district and to the southeast of Buenos Aires due to the increased reach of extreme storm surges, possibly above the secure height of these communities.
Figure 6.9 Variation of the social risk map between 2070 and present Note: Values as in Figure 6.8.
An interesting observation made during the course of this investigation was that, in some cases, future social vulnerability to increased flooding may also be exacerbated by the status of current adaptive measures employed by some communities living in flood-prone areas. This could be due either to the changing dynamics of the communities or to the inadequacy of existing strategies to manage floods. Examples of such communities include the La Boca neighbourhood in the Buenos Aires district and the Avellaneda District, both of which are situated close to downtown Buenos Aires. These two communities have existed since the late 19th century and have historically lived with frequent flood impacts. Of these, the La Boca neighbourhood is an old harbour town built on a marshy area to the left of the mouth of the Riachuelo river. It has historically been one of the poorest areas of the city of Buenos Aires and
3243 J&J Vulnerability
1/11/07
10:23 am
Page 129
Storm Surges, Rising Seas and Flood Risks in Buenos Aires 129
the main climatic hazards it faces are floods due to sudestadas and intense rainfall. The Avellaneda District is an old urban industrial town situated on a flood plain to the south of La Boca on the other side of the Riachuelo river. It is also greatly impacted by floods as well as water and air pollution. In these neighbourhoods, a critical element of adaptation to storm surge floods has traditionally been informal networks among neighbours that support local practices and strategies that aid in anticipating the arrival of floods (including an early warning system and self-help and evacuation strategies) and tend to diminish local vulnerability. However, of late there has been an increasing influx of newcomers into these areas, which is gradually eroding the collective memory of these cultural practices of adaptation and could increase the vulnerability of this population to the more severe floods in the future. Moreover, the construction of the coastal defence structure for the city of Buenos Aires, including the La Boca neighbourhood, during the last decade has successfully provided protection against recent floods, but this has also unfortunately created a sense of complacency among the local population regarding the threat. This could prove dangerous in the long run as the coastal defences, built with current flooding levels in mind, may be insufficient protection against the higher future flood levels resulting from climate change (Gentile, 2002). By this time the collective knowledge of flood coping strategies would also have been lost and institutional responses would likely have been significantly weakened.
Other Factors that Enhance Social Vulnerability to Climate Change Other factors that tend to increase vulnerability to climate change impacts or hamper the processes of adaptation are those that are common to many developing countries and include issues such as lack of basic information; physical, social or institutional weaknesses; and a lack of public awareness about climate change and its consequences. The specific manner in which these factors play out in the Argentine context is briefly outlined below.
Lack of basic information Determination of vulnerability to climate change impacts in developing countries can often be hampered by the lack of data necessary for such efforts. Long-term data on many variables critical to the determination of vulnerability is often found to be insufficient or absent. This is especially true for data on variables that change rapidly. In this study two critical variables for which useful and sufficient data were lacking were land altitude maps or records for some locations and tide data for the Plata river coast. As a result a new digitized model of land altitude had to be specially developed to generate proxy data on land altitude that would otherwise either have been impossible or taken a very long time to obtain. Similarly for tide data a hydrodynamic model had to be developed to fill the information gaps.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 130
130 Climate Change and Vulnerability
Institutional weakness The vulnerability of coastal areas in Argentina is also impacted by the weak institutional structures, which have so far been unable to effectively put in place policies and programmes to reduce the risk of current and future flood impacts. Frequent changes in the national government tend to make the various institutions responsible for aspects of flood management very unstable in terms of their persistence, and even when they do persist, their policies change often. In the case of flood management, whatever little planning has been done or the few successful programmes and projects that have been implemented have unfortunately not always been continued with each change of government. One example of this is the Emergency Federal System created in the 1950s. Since then, it has been moved several times to and from different ministries and has had its plans and policies variously altered. In general, the current institutional management style, typical of the national culture, is not adequate for long-term planning. There is also a lack of coordination and poor communication between the different institutions that hold responsibility for flood management. As a result their policies and measures are often fragmented and lack coherence in totality, thus generating a high degree of uncertainty. Communication between institutions and the general public is also poor and civil participation in the determination of flood defence strategies tends to be nominal, with decision making being largely top–down.
Lack of public awareness Awareness among the general public about climate change and its potential risks is also found to be very low, largely due to the absence of any instituted programmes and measures to communicate this kind of information publicly. In fact, government officials, investors and engineers who make decisions that affect the coastal area also possess no knowledge of climate change impacts and of the dynamics of the system they are altering. As a result, the development of the entire coastline has been undertaken without accounting for the rising water level, which rose 17cm during the last century and is now rising even faster. There is also a high level of interest in the development of the new land environments that have been created over the past 300 years at the tip of the Paraná delta, which is now advancing over the Plata river at a pace of 70m per year. Plans are being made to use these new lands for real estate development, and disputes over the nature of their use, jurisdiction and property have already begun among various sectors of civil society. Despite this significant business interest there is little awareness that these are also one of the most vulnerable areas in the region to rising water levels. A continuation of this sort of poor coastal management would, in future, necessitate new investments in expensive remedial structures.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 131
Storm Surges, Rising Seas and Flood Risks in Buenos Aires 131
Conclusion Future climate change risk to coastal areas along the Greater Buenos Aires coast is expected primarily from the greater inland reach of recurrent storm surge floods or sudestadas, rather than due to the impact of permanent flooding. This is because the area likely to be permanently flooded in the future due to an increase in the level of the Plata river is very small and is sparsely populated. Therefore, unless there is a change in the population distribution resulting in a much greater inhabitation of this low-lying area, the risk due to permanent flooding in this century is expected to be relatively minor. The greatest impact from increased storm surge recurrence will be felt in those neighbourhoods that presently have a relatively low recurrence of floods and are therefore densely occupied. As a result, large social damages and substantial real estate losses can be expected. These neighbourhoods are currently occupied by a wide social spectrum, ranging from the upper middle class to a socially vulnerable population. In the absence of any adaptation measures, the lower limit of the total infrastructure losses, including real estate, for the Buenos Aires region is calculated to range between 5 and 15 billion US dollars for the period 2050–2100, depending on the speed of the sea-level rise and the increase in the infrastructure value. The increasing popularity of gated communities in low-lying, flood-prone coastal areas is an additional factor that will add to the future real estate risks due to more frequent and intense storm surges if climate change impacts continue to be ignored in the planning of these neighbourhoods. The massive land transformation brought on by these construction activities is destroying the critical natural hydrological drainage capacity of the area and endangering not only the rich occupying the gated communities but also the poor squatters outside. Some non-governmental organizations have now begun calling for an urgent regulation of the Plata river coastal zones in the Greater Buenos Aires region. In order to be effective, such regulation must include consideration of future climate change scenarios and the associated Plata river levels. At the moment, unfortunately, institutional preparation to deal with either current or future climate impacts is almost non-existent, and the ability to develop an effective flood management strategy is quite weak. Moreover, many of the historical collective adaptation strategies to floods among the local population are also gradually being eroded due to a sense of complacency arising out of the construction of flood control structures around Buenos Aires, which unfortunately may not afford adequate protection in the future. This cultural loss can also be attributed to the increasing influx of newcomers into these areas. Given these risks, the dissemination of the findings of this research (Barros et al, 2005) in collaboration with a local non-governmental organization, Fundacion Ciudad, can help to fill the knowledge gap about the impacts of and vulnerability to climate change in the region and generate awareness about the associated issues, especially among decision makers. These findings can serve as effective tools to guide institutions in the determination of strategies and implementation of programmes that can help in adapting to the impacts. They
3243 J&J Vulnerability
1/11/07
10:23 am
Page 132
132 Climate Change and Vulnerability
can also assist individuals and private developers in making appropriate decisions regarding future property development. These findings therefore also underline the importance of scientific research in helping develop and maintain the collective awareness of both present and future climate hazards. The preliminary results of this study have so far been presented at a forum for stakeholders and at numerous other institutional meetings. These activities are successfully contributing to a gradual awareness building among key individuals and have already resulted in frequent consultations. It is hoped that the continuation of these efforts will eventually build an increased understanding about the implications of climate change for this region at the public and institutional levels, and will lead to the development of effective adaptation strategies for the Plata river coast.
Notes 1
2
The Coriolis effect is an impact of the Earth’s rotation which causes the deflection of winds to the right of their direction of travel in the northern hemisphere and to the left of their direction of travel in the southern hemisphere. This effect was first described by the French scientist Gaspard-Gustave Coriolis in 1935. The Paraná delta is growing over the Plata estuary by the addition of sediments brought by the Paraná river. This is a natural process that is probably being enhanced by deforestation processes in part of the basin.
References Barros, V., I. Camilloni and A. Menéndez (2003) ‘Impact of global change in the coastal areas of the Río de la Plata’, AIACC Notes, vol 2, International START Secretariat, Washington, US, pp9–11, www.aiaccproject.org Barros, V., A. Menéndez, and G. Naggy (eds) (2005) Climate Change in the Plata River, Center for Atmospheric and Ocean Research, Buenos Aires, Argentina (in Spanish) Church, J. A., J. M. Gregory, P. Huybrechts, M. Kuhn, K. Lambeck, M. T. Nhuan, D. Qin, and J. L. Woodworth (2001) ‘Changes in sea level’, in J. T. Houghton, Y. Ding, D. J. Griggs, M. Noguer, P. J. van der Linden and D. Xiaosu (eds) Climate Change 2001: The Scientific Basis, Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US D’Onofrio, E., M. Fiore and S. Romero (1999) ‘Return periods of extreme water levels estimated for some vulnerable areas of Buenos Aires’, Continental Shelf Research, vol 4, pp341–366 Escobar, G., V. Barros and I. Camilloni (2003) ‘Desplazamiento del anticiclón subtropical del Atlántico Sur’ [‘The shift of the South Atlantic subtropical high’], X Congreso Latinoamericano e Ibérico de Meteorología, Proceedings of Tenth Latin American and Iberic Congress of Meteorology, Latin American and Iberic Federation of Meteorological Societies, La Habana, Cuba Escobar, G., W. Vargas and S. Bischoff (2004) ‘Wind tides in the Río de la Plata estuary: Meteorological conditions’, International Journal of Climatology, vol 24, pp1159–1169 Gentile, E. (2002) ‘La incorporación de la gestión del riesgo por inundaciones en la
3243 J&J Vulnerability
1/11/07
10:23 am
Page 133
Storm Surges, Rising Seas and Flood Risks in Buenos Aires 133 gestión urbana pública: El caso del barrio de La Boca’ [‘The inclusion of flood risk in the urban public management: The neighbourhood of La Boca case’], Instituto Gino Germani, Buenos Aires, Argentina Kalnay, E., M. Kanamitsu, R. Kistler, W. Collins, D. Deaven, L. Gandin, M. Iredell, S. Sha, G. White, J. Woollen, Y. Zhu, M. Chelliah, W. Ebisuzaki, W. Higgins, J. Janowiak, K. C. Mo, C. Ropelewski, J. Wang, A. Leetmaa, R. Reynolds, R. Jenne and D. Joseph (1996) ‘The NCEP/NCAR 40-year reanalysis project’, Bulletin of the American Meteorological Society, vol 77, pp437–471 Menéndez, A. N. (1990) ‘Sistema HIDROBID II para simular corrientes en cuencos’ [‘HIDROBID II System to simulate currents in basins’], Revista Internacional de Métodos Numéricos para Cálculo y Diseño en Ingeniería (International Magazine of Numeric Methods for Calculation and Engineering Design), vol 6, pp1–6 Re, M. and A. Menéndez (2003) ‘Modelo numérico del Río de la Plata y su frente marítimo para la predicción de los efectos del cambio climático’ [‘Numeric model of the Plata River and its marine front for the prediction of the effects of the climate change’], Mecánica Computacional, vol 22 (M. Rosales, V. Cortínez and D. Bambill, eds), Bahía Blanca, Argentina Ríos, D. (2002) ‘Vulnerabilidad, urbanizaciones cerradas e inundaciones en el partido de Tigre durante el período 1990–2001’ [‘Vulnerability, closed urbanizations, and floods in the Tigre District during the period 1990–2001’], thesis, Facultad de Filosofía y Letras, University of Buenos Aires, Buenos Aires
3243 J&J Vulnerability
1/11/07
10:23 am
Page 134
7
Climate and Water Quality in the Estuarine and Coastal Fisheries of the Río de la Plata Gustavo J. Nagy, Mario Bidegain, Rubén M. Caffera, Frederico Blixen, Graciela Ferrari, Juan J. Lagomarsino, Cesar H. López, Walter Norbis, Alvaro Ponce, Maria C. Presentado, Valentina Pshennikov, Karina Sans and Gustavo Sención
Introduction The Third Assessment Report of the Intergovernmental Panel on Climate Change (McCarthy et al, 2001) identified two main environmental problems in South America: land-use changes and El Niño Southern Oscillation (ENSO) variability. A good example is the Río de la Plata basin and estuary, an area that has been substantially influenced by human activities in recent decades and is highly sensitive to both climate extremes and changing precipitation patterns. The Río de la Plata basin, which measures 3,100,000km2, includes the Uruguay river basin, which in turn covers an area of 297,000km2 in Brazil, Argentina and Uruguay, and has a human population density of 25 persons/km2. The population increased by almost 90 per cent between 1961 and 1994 (Baethgen et al, 2001), which led to increasing pressure on watersheds due to their extensive use for agriculture and water storage, fertilizer application, and the heavy discharge of wastewater from point sources (wastewater from domestic use is about 20 per cent) and of nutrients from non-point sources (Pizarro and Orlando, 1985; Tucci and Clarke, 1998; Nagy, 2000; Nagy et al, 2002a). The main climatic changes reported in the Río de la Plata basin include an increase in inter-annual variability, especially of ENSO variability and precipitation (≥20 per cent change in recent decades); the southward displacement of the quasi-permanent Atlantic subtropical high-pressure circulation; and asso-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 135
Climate and Water Quality in Río de la Plata 135
ciated changes in frequency of the prevailing winds, increased surface air and water temperatures (≥0.8ºC), runoff, soil moisture and the Pantanal’s extent (Díaz et al, 1998; Camilloni and Barros, 2000; Escobar et al, 2004; Bidegain and Camilloni, 2004; Liebmann et al, 2004). The Río de la Plata river and estuary is a large (38,000km2) and wide funnel shaped (30–240 km width) coastal plain system. In this study, we focus on the estuarine frontal zone related to the salt intrusion limit (Figure 7.1). This productive ecosystem sustains the ecological and biogeochemical processes (nutrient assimilation, denitrification and production of organic matter) that determine the goods (fisheries) and services (fish reproduction, CO2 fixation and denitrification) obtained from the estuary.
Figure 7.1 Río de la Plata estuarine front location under different ENSO conditions: a) strong La Niña event (1999–2000), b) typical, c) moderate El Niño (winter 1987), d) strong El Niño 1997–1998/2002–2003 Note: RA – República Argentina; ROU – República Oriental del Uruguay. Source: Modified from Nagy et al (2002b).
Goal In this chapter the current biophysical and human vulnerability and the adaptive capacity to deal with impacts of climate variability and change in the estuarine waters of the Uruguayan coast of the Río de la Plata are synthesized. The main questions addressed are as follows: • • •
How sensitive is the system to climate variability and change? Is eutrophication related to climate change and variability? Is the coastal fishery system sustainable under increased river flow variability?
3243 J&J Vulnerability
1/11/07
10:23 am
Page 136
136 Climate Change and Vulnerability
ENSO-related inter-annual rainfall variability and associated river discharge fluctuations and changes in wind patterns are examined using observational records of recent decades. Our attempt is to understand the ecosystem response to climatic and anthropogenic influences and to estimate potential future impacts and vulnerability due to trophic state changes (increase in symptoms of eutrophication) and coastal fisheries activity. The associated livelihood potential from the present to the next several decades is also assessed. Specifically, we summarize the overall climatic background, develop future climate scenarios, and reference projections for a range of climate and non-climate factors and related vulnerability scenarios for future decades based on analysis of indicators of impacts and vulnerability. The basic data are the observational record from the past 30 years; some climatic and environmental trends (precipitation and drivers of eutrophication) are based on longer-term records from 1940. Our analytic approach includes: 1
2
3
multi-level indicators of vulnerability to climate change and a driver-pressure-state-impact-response (DPSIR) index for water resources, ecosystem and coastal fisheries, and settlement adapted from Moss (1999) and the Stockholm Environment Institute (SEI) (2001); calculations of vulnerability indicators, indices and vulnerability matrices, related regression models, and economic analysis of fishing activity, a combination of objective values (such as net income and education, wind speed, and ENSO-related sea-surface temperature (ENSO 3.4 index)) and the use of expert judgement in order to assess social, economic, environmental and legal indicators of sensitivity of fishermen and to assess the impacts of harmful algal blooms (HABs); use of the IPCC’s Special Report on Emissions Scenarios (SRES) A2/B2 climatic scenarios and Hadley CM3/ECHAM-4 global climate models (GCMs) (Bidegain and Camilloni, 2004).
Río de la Plata Basin Climate Baseline (1961–1990) The mean annual temperature in the Río de la Plata basin ranges from around 15°C in the south to more than 25°C in the mid-western Chaco region (Figure 7.2, left panel). In the austral (southern hemisphere) winter, monthly mean temperatures have a clear north–south gradient. In July, for example, the mean temperature over the northwest part of the basin is more than 20°C, while that in the province of Buenos Aires is around 10°C less. In the austral summer the gradient is more zonal, reacting to the land–ocean distribution. In January, maximum mean temperatures reach over 27.5°C in western Argentina, while they are less than 22.5°C along the coastlines of southern Brazil, Uruguay and Buenos Aires (Hoffmann, 1975). The eastern zone of the Río de la Plata basin is dominated by the influence of the South Atlantic high pressure (Figure 7.2, middle panel), whereas the northern portion shows a typical low pressure system (the Chaco low) that is
3243 J&J Vulnerability
1/11/07
10:23 am
Page 137
Climate and Water Quality in Río de la Plata 137
Figure 7.2 Climate baseline scenarios for the Río de la Plata for the period 1961–1990 Note: Sea-level pressure in hPa (above right), temperature in degrees Celsius (above left) and precipitation per year in mm (below).
more intense in the austral summer. Surface winds are northerly over most of the region, with a maximum in the northeastern sector; to the south, westerlies are present throughout the year but are strongest during the austral winter (Kalnay et al, 1996). The annual average precipitation in the region is about 1100mm (Figure 7.2, right panel). Annual mean rainfall tends to decrease both from north to south and from east to west. Amounts range from 1800mm in the maritime uplands along the Brazilian–Paraguayan border to 400mm along the western boundary of the region. The amplitude of the annual cycle in rainfall decreas-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 138
138 Climate Change and Vulnerability
es from north to south. The northern part of the region has a well-defined annual cycle with maximum precipitation during summer (December to February). The central region (northeast Argentina/southern Brazil) has a more uniform seasonal distribution, with maximum precipitation occurring during the southern spring and autumn.
The Río de la Plata River Estuary: Setting, Subsystems and Sectors The funnel-shaped Río de la Plata is a coastal plain estuary with a river paleovalley (called Canal Oriental) along the northern coast (López Laborde and Nagy, 1999) that behaves as a conduit channel for water and particles to the coastal ocean (Nagy et al, 2002b). Microtidal systems like the Río de la Plata estuary characteristically have low mixing capacity, primarily due to the prevailing wind; hence river inflow to the coastal zone is the ‘master variable’ that controls stability (defined as a function of the vertical difference in salinity as a function of depth, nutrient excess, flushing time of water and particles, stratification and gravitational circulation, salinity, and bottom-water hypoxia) (EPA, 2001; Nagy et al, 2002b; Nagy, 2003 and 2005). Total freshwater inflow (QV) to the Río de la Plata, estimated as the sum of the discharges of the Paraná (QP) and Uruguay (QU) rivers, typically varies from 1500 to greater than 3000m3/s for strong La Niña and El Niño years respectively. A strong relationship has been reported between the ENSO SST 3.4 Index (defined as the temperature anomaly in ENSO Regions 3 and 4 in the Pacific Ocean) and freshwater inflow, especially the Uruguay river flow during 1998–2002. Both the location and structure of the estuarine front and salinity off Montevideo closely follow river flow on monthly to inter-annual timescales (Severov et al, 2004) and have important effects on most ecological and biogeochemical processes in the estuarine waters and resources (Nagy et al, 2002a and b; Nagy et al, 2003). However, total freshwater inflow extremes also depend on other climatic and human-driven factors which are not considered in the analysis presented here. Freshwater inflow has increased by about 35 per cent over the past 50 years because of increased precipitation and runoff and land-use changes (García and Vargas, 1998; Tucci and Clarke, 1998; Kane, 2002; Nagy et al, 2002a; Menéndez and Re, 2005) and closely follows inter-annual fluctuations of ENSO events (Figure 7.3) as shown for the long-term series for the Uruguay river (Nagy et al, 2002a and b). Small changes in precipitation are reflected as doubled streamflow, though this amplification of signal varies on inter-annual to inter-decadal time scales (Berbery and Barros, 2002), revealing a high vulnerability of the region to increased precipitation (Figure 7.3). These physical factors dictate the nutrient condition of the Río de la Plata estuary. Hence consideration of the related trophic state change is a valuable exercise. Eutrophication is the process by which a body of water is enriched with organic material (Nixon, 1995). The change in trophic state associated with
3243 J&J Vulnerability
1/11/07
10:23 am
Page 139
m3/s
Climate and Water Quality in Río de la Plata 139
Figure 7.3 River Uruguay discharge at Salto from 1921 to 2003 Source: Bidegain et al (2005).
nutrient excess is driven by physical driving forces such as increase in human population density and related economic activities such as the use of synthetic fertilizers (Seitzinger et al, 2002). The balance between production and respiration of organic matter dictates the magnitude and direction of air–sea flux of biologically active elements (Gordon et al, 1996). The occurrence of eutrophication includes effects or symptoms such as excess algal biomass, hypoxia and harmful algal blooms (HABs) that are indications that the system cannot cope with the available nutrient inputs (NRC, 2000; De Jonge et al, 2002). The main factors in the expression of symptoms of eutrophication are 1) flushing time (ft), 2) turbidity, 3) nutrient inputs (NI) and 4) mixing state (stratification–destratification cycle) (De Jonge et al, 2002; Nagy et al, 2002b). The expression of these symptoms depends on both nutrient inputs (NCR, 2000; De Jonge et al, 2002) and the balance between river flow (QV) and wind stress (W), which determine mixing and transport processes (EPA, 2001; Nagy, 2003; Nagy et al, 2004). According to Seitzinger et al (2002), by 2050, export of nitrogen over South America will have increased nearly two to three fold from current values. Generically, the temperate regions export less than half the nitrogen from anthropogenic sources compared to tropical regions, which is partially explained by lower precipitation. However, ENSO variability and other hydroclimatic drivers also play a major role in controlling the trophic state changes in the Río de la Plata basin (Nagy et al, 2003b, 2004 and 2005). Since 1940, the trophic state changes in the Río de la Plata region have been due to increasing trends in pressure indicators and a shift of state indicators during the 1980s (Nagy et al, 2003b; López and Nagy, 2005; Calliari, Gomez and Gomez, 2005). For example, eutrophic concentrations of nitrate
3243 J&J Vulnerability
1/11/07
10:23 am
Page 140
140 Climate Change and Vulnerability
and chlorophyll were observed during extreme floods (1983 El Niño and 1999 La Niña respectively) (Nagy et al, 2002a and b). Human-driven land-use change and changing climatic drivers have indirectly altered trophic state. Continued trends of these parameters suggest that vulnerability and impacts may increase over the next few decades. As an indicator of this symptom, HABs have occurred since at least 1980, becoming more frequent and causing great economic damage due to their impacts on molluscs and fish, on the tourism industry, and on public health in Uruguay (Méndez et al, 1996; Méndez et al, 1997; Ferrari and Nagy, 2003). Freshwater cyanobacterial blooms have also multiplied in recent years, becoming persistent during the summer (López and Nagy, 2005). The occurrence of HABs is closely connected to changes in freshwater inflow or QV (Méndez et al, 1996; Nagy et al, 2002b; López and Nagy, 2005). Not far from the Plata river estuary is the mouth of the Santa Lucia river estuary, which lies within the same estuarine frontal zone, located a few miles to the west of Pajas Blancas (see Figure 7.1), close to a major fishing area. This system is eutrophic because of nutrient excess derived mostly from fertilizer application. The inter-annual variability of water height and river flow (QSL) here are related to ENSO effects that predominate during August to February. Since 1979, yearly means of QSL and persistence of floods have increased by about 25 per cent because of the increase in precipitation in the basin (Caffera et al, 2005). Such river discharge fluctuations can greatly influence local variability of salinity, sea level, turbidity, nutrient content and trophic state during the peak of primary productivity and fishing activity. The evolution of the trophic state in the Santa Lucia river estuary was related to the river flows of both the Santa Lucia and the Uruguay rivers (QSL and QU) during the studied period (October 2002 to May 2004). We selected one state/impact variable of trophic state (chlorophyll-a) to show the response and coping capacity of the ecosystem to ENSO events. The increase in both QSL and QU decreased salinity from about 5 per cent to 1 per cent, blocking the expression of trophic state symptoms (chlorophyll-a <5µg/l), whereas during normal flows, salinity increased to about 10 per cent and chlorophyll-a increased to eutrophic levels (>20µg/l) (Nagy et al, 2004; Caffera et al, 2005). To summarize: • • •
within normal river flow and salinity ranges the system expresses symptoms of trophic state change; when river flow is high (for example, 2002–2003 El Niño years) the river does not develop symptoms of trophic state change because both water and nutrients are exported to the Plata river estuary; and hydrological fluctuations associated with ENSO events seem to exert some control on the coping capacity of the ecosystem.
These results are in close agreement with previous observations for the adjacent waters of the Plata river (Nagy et al, 2002a and b), which showed that during La Niña years and low river flow (1999–2000), the system would be prone to increased trophic state changes.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 141
Climate and Water Quality in Río de la Plata 141
Indicators of Susceptibility, Impact and Vulnerability Indicators of the susceptibility of the trophic state for both cross-system and long-term comparisons as well as for the assessment of the impact of nutrient excess in estuaries (NRC, 2000; De Jonge, Elliott and Orive, 2002; EPA, 2001) are an important component of this kind of analysis. Nagy et al, (2002b and 2004), Nagy (2003) and Ferrari and Nagy (2003) have developed such indicators for fresh and estuarine waters of the Uruguayan coastal zone. Their analysis suggests that some expected impacts and responses under current and future scenarios would be an increase in hypoxia, HAB events and changes in biodiversity. In the present analysis, the following variables are considered: 1 2 3 4 5
susceptibility (for example, flushing and residence times of water, buoyancy and mixing state); drivers (population density, fertilizer use, point sources and QV); pressure (nitrogen (N), phosphorus (P) and silicon (Si) load); state variables (N, P); and response/impacts or symptoms (such as changes in algal biomass (Chlorophyll-a), dissolved oxygen (O2), occurrence of HABs, nutrient ratios (N/P and N/Si) and dominant species).
Some variables can be both state and impact (Figure 7.4).
Figure 7.4 DPSIR framework of trophic state and symptoms of eutrophication for the Río de la Plata estuary Note: N = nitrogen; P = phosphorus; HABs = harmful algal blooms; Chl-a = chlorophyll-a; Si = silicon.
Table 7.1 shows an example of an impact matrix of an important symptom of trophic state changes (HAB occurrence) developed by Ferrari and Nagy (2003) for the Uruguayan coastal zone of the Plata river for the decade 1991–2000. Four indicators were estimated and aggregated: 1) intensity of HABs (cells/l), 2) persistence (months), 3) extension along the coast (percentage of coverage) and 4) toxicity (toxin concentration from low to very high). Both freshwater (cyanobacteria HABs, for example, Mycrocystis aeruginosa) and estuarine/ marine species as well as toxic and noxious species were considered.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 142
142 Climate Change and Vulnerability
Table 7.1 Aggregated impact matrix of HABs in the Uruguayan coastal zone of the Río de la Plata for the period 1991–2000 Prevailing Indicators
0
1
Extreme 2
2
3 3
4
10
4
10
105
0.75
1–6
>6
Intensity (cells/l)
0
10
Persistence (months)
0
0.25
Extension (% of coverage)
0
10
25
50–75
>75
Toxicity (concentration)
0
Low
moderate
high
very high
A second aggregated impact matrix was then built in order to take into account the occurrence of the four main HAB species weighted from absence (0) to very high (4) according to the criteria defined in Table 7.1; weights of the HAB species were summed and ranked from 0 to 100 (very low to very high). Weighting criteria were based on both the literature (EPA, 2001) and local occurrence range according to the expert judgement of the authors (Table 7.2). The ENSO SST 3.4 index, which is well correlated with the Uruguay river flow QU, is also reported on an annual basis (from March to February) (see Table 7.2). Table 7.2 Aggregated impact index of HABs for the Río de la Plata, 1991–2000 Year
1990–1991 1991–1992 1992–1993 1993–1994 1994–1995 1995–1996 1996–1997 1997–1998 1998–1999 1999–2000 Average
SST 3.4
0.3 1.0 0.4 0.4 0.6 -0.4 -0.3 1.8 -0.7 -1.3 0.2
1 Alexandrum tamarense
2 Gymnodium catenatum
3 Dinophysis acuminata
4 Mycrocistis aeruginosa
Index 0–100
Impact
3 2 2.75 1.25 2 3 1.25 0 1.25 1.0 1.75
0 3 2.75 2.25 2.25 2.25 1.5 2 1.25 1 1.83
0 2 0 1.5 0 1.5 0 0 0 0 0.5
0 2 0 2.75 0 0 3 0 2 0 0.98
3 9 5.5 7.75 4.25 7 6 2 4.5 2 4.7
21 64 39 55 30 50 43 14 32 14 36
low high low medium low medium medium very low Low Very low Low to medium
Finally, these indicators were weighted taking into account their impacts on three sectors (human health (-H); economic activity (-Ec), specifically molluscs’ consumption; and environmental health (-Ev)) according to the expert judgement of the authors. Coefficient weights were 0.75, 0.50 and 0.25 for H, Ec and Ev respectively, which were multiplied by the values reported in Table 7.2. Thus a weighted aggregated impact index of persistence/extension/toxicity (IPET) was built (Table 7.3). An impact coefficient was assigned for each of the four species. Relative impact of each indicator (IPET from 0 to 1) is shown for each species and sector. Extreme values are toxicity of Gymnodinium for human health and intensity of Dinophysis for the environment (Table 7.3).
3243 J&J Vulnerability
1/11/07
10:23 am
Page 143
Climate and Water Quality in Río de la Plata 143
Table 7.3 Weighted aggregated impact index of HABs (index of persistence/extension/toxicity, IPET) for each species and sector on the northern coast of the Río de la Plata, 1991–2000 Weighted Index for each sector (0–1) Species
Health: 0.75
Economic activity: 0.50
Environment: 0.25
Gymnodinium 0.50 Alexandrum 0.42 Microcystis 0.33 Dinophysis 0.22
T: 0.37 E: 0.32 P: 0.25 I: 0.19
T: 0.25 P: 0.21 E: 0.17 I: 0.13
P: 0.13 E: 0.11 T: 0.08 I: 0.06
The overall impact of HAB occurrence was moderate, with only one highimpact year (1992) and two low-impact years (1998 and 2000), which coincide with strong ENSO events (1997–1998 and 1999–2000 respectively). The only species that reached high values in 1991 and 1993 (Alexandrium tamarense) was due to both the northward displacement of the Malvinas current, where Alexandium is present, and a low freshwater inflow or QV (Méndez et al, 1996), especially because of a low QU (Nagy et al, 2002b). Usually, both drivers are associated with El Niño and La Niña years respectively, which reduces the vulnerability of the Plata river to the presence of these blooms. To date HABs have occurred during spring time, but they could be potentially dangerous to human health if they were to occur during summer. Such examples – the blooms of Alexandrium, the relatively high index of El Niño multi-year events (1991–1994), and the extreme events of 1998 and 2000 – suggest some hydroclimatic control on HAB occurrence. Recent years (2001 and 2003, not shown here) have shown a marked increase in cyanoHABs (López and Nagy, 2005) at intensities higher (>3) than those found during the past decade. El Niño and La Niña events could stimulate cyanoblooms, the former by advecting freshwater and the latter by decreasing it.
Coastal Fishery System Vulnerability in the Pajas Blancas Fishing Community An artisanal fleet exploits fisheries 5–7 miles off the Uruguayan coast in the estuarine zone of the Río de la Plata close to the Santa Lucia river mouth. The main community is based at Pajas Blancas (see Figure 7.1) within the estuarine front (EF). The peak of the fishing period, October to December (Figure 7.5), is controlled by QV (especially by QU) and wind rotation (Norbis, 1995; Nagy et al, 2003a; Norbis et al, 2004). The location of the estuarine front displaces as a function of: 1
total river flow QV (which, in turn, is strongly associated with both seasonal and ENSO-related inter-annual variability in precipitation); and
3243 J&J Vulnerability
1/11/07
10:23 am
Page 144
Gross income
144 Climate Change and Vulnerability
Figure 7.5 Long-term gross income of fishermen (local currency-1999) Note: Average – black; strong ENSO years – light grey; maximum – dark grey. Months: 1 (October)–12 (September). Source: Nagy et al (2003a).
2
offshore and onshore winds on a weather development timescale (1–10 days) (Norbis et al, 2004).
Climatic stimuli (river flow and wind) thus displace the estuarine front upward and downward (Framiñán and Brown, 1996; Severov et al, 2003 and 2004), out of the fishing area. Displacement of the estuarine front induces changes in the spatial distribution of fish and their recruitment. Frequency patterns of winds have also changed over the past few decades (Pshennikov et al, 2003), with an increase in onshore east–southeast winds (Escobar et al, 2004). During strong ENSO events, the estuarine front tends to be located far from the Pajas Blancas community of fishermen (see Figure 7.1). In spite of the increase in the variability of the Plata and Uruguay river flows (QP and QU) and related extreme locations of the estuarine front, as well as the increase in onshore winds, fishermen have shown good adaptive capacity since 1988. Many of them have migrated seasonally or permanently away from the estuarine front along the coast following resources (Hernández and Rossi, 2003; Norbis et al, 2003) in order to reduce their long-term vulnerability to the fluctuations of QV and avoid bad years (Norbis et al, 2004).
Sensitivity and vulnerability of coastal fisheries We estimated proxy variables, classified and valued respectively as low (1), moderate (2) and high (3), in order to assess social, economic, environmental and legal indicators of sensitivity of the fishermen’s community (Table 7.4). The sum of all indicators (non-weighted index of vulnerability) suggests that the community is subject to moderate to high vulnerabilities yet seems to be resilient. Only those strong ENSO events whose effects are noticeable during the peak of the fishing period (El Niño 1992, 1997 and 2002 and La Niña 1989
3243 J&J Vulnerability
1/11/07
10:23 am
Page 145
Climate and Water Quality in Río de la Plata 145
Table 7.4 Assessment of the vulnerability of the coastal fishing community Proxy variable
Vulnerability High (3)
Social Family Education Housing Employment Health Social organization Sub-total: 2.2/3 Economic Boats Engines Fishing gear Communications Refrigeration Catch Prices Net income Subtotal: 2.0/3 Environmental Climate – ENSO Winds Storm surges/Flooding Eutrophication Habitat loss Subtotal: 2.2/3 Legal/Institutional Laws Territorial planning Coast Guard controls Conflicts with industrial fleet Conflicts with neighbours Legal organization X Subtotal: 2.5/3 Total: 2.2/3
Moderate (2)
Low (1)
X X X X X X
X X X X X X X X
X X X X X
X X X X X
Note: Unweighted total index of vulnerability (IV) = 2.2 (scale 1–3). Source: Modified from Nagy et al (2003a) and Norbis et al (2004).
and 1999) seem to impact fishing activity and net income. Therefore, less than one third of the peak fishing periods are considered to be very bad in economic terms (when the net income of fishermen is estimated to be reduced by ~60 per cent with regard to normal years as shown in Figure 7.5) (Nagy et al, 2003a; Norbis et al, 2004). The question regarding the sustainability of coastal fisheries was thus empirically answered because the fishing activity remains sustainable regardless of their (estimated) high vulnerability. Our analysis suggest that their resiliency is (or was) due to:
3243 J&J Vulnerability
1/11/07
10:23 am
Page 146
146 Climate Change and Vulnerability
• •
•
the combination of planned and reactive adaptation measures to hydroclimatic variability (for example, migration to escape from the high variability of the estuarine front); the good fishing performance of most fishermen – many fishermen have acquired a high level of appropriate skills and developed a high capacity to understand weather and environmental conditions, for instance, they usually conduct exploratory samplings of bottom waters with hand-made domestically produced hand-made equipment (Hernández and Rossi, 2003); and their (dominant) cautious behaviour (to avoid weather-related risks in spite of economic losses), which reduced their vulnerability. Fishermen do not risk fishing for at least one day after unfavourable wind conditions occur, even if fish are often then available (Norbis, 1995). Even if this behaviour should be considered a bad practice in terms of cost–benefit analysis, it has not significantly affected long-term income. However, real-time weather forecasting applied to fisheries would be a good adaptation practice, provided fishermen can trust the information, which is not always up-to-date.
Thus, the following question arises: Will fishermen have the adaptive capacity to continue to be successful under increasing climatic and economic pressures such as those that occurred in 2002?: which consisted of: • • • •
severe economic crisis; increase in fuel prices; a moderate El Niño year that decreased surface salinity at Pajas Blancas to close to zero because of the seaward displacement of the estuarine front; and increase in the occurrence of unfavourable wind conditions for fishing activity (>8m/s) (Nagy et al, 2003a; Norbis et al, 2004).
According to Norbis (1995) and Norbis et al (2003) southern winds (SW to SE) of >8m/s are not favourable for fishing and, as mentioned above, most fishermen prefer not to risk fishing on the first day of favourable conditions after bad weather, usually losing one favourable day. In fact, fishing activity for the 1998–1999 period shows that the number of fishing boats and fish haul increased during the favourable days, suggesting that there is no resource limitation and that the community tends to follow the advice of the fishermen who are the leaders (i.e. the expert fishermen or fishermen who perform well) (Norbis et al, 2004 and 2005). Generally, the average wind speed in the region is 5–6m/s, but it increases to >6m/s during spring and summer because of the prevailing SE winds (Nagy et al, 1997; Escobar et al, 2004). This means that fishermen are both highly exposed and resilient to developing their activity within a narrow wind range close to the limit of 8m/s. However, the overall conditions during 2002 forced fishermen to change their no-risk behaviour (Nagy et al, 2003a; Norbis et al, 2004 and 2005).
3243 J&J Vulnerability
1/11/07
10:23 am
Page 147
Climate and Water Quality in Río de la Plata 147
Scenarios of coastal fisheries activity Norbis et al (2005) developed an empirical scenario of fishing activity and productivity based on both the long-term yearly fishing activity gross income (see Figure 7.5) and the fishing period 1998–1999. This economic scenario was built for: • • • •
30 boats; fishing period (2, 3 and 4 months); monthly fishing days (8, 14 and 17 days); and efficiency of fishing units (26, 38 and 46 boxes).
Thus, nine different empirical scenarios were developed using various combinations of the first three sets of assumptions in the list above (i.e. number of boats, fishing period and monthly fishing days). These scenarios represent maximum (1) and minimum (9) fishing activity within the range of observed conditions during 1998/1999, for each variable. Of these, six scenarios yield captures greater than the maximum observed in 1998/1999 and the most successful are those based on four- and three-month seasons of fishing activity. It must be noted that these scenarios do not include captures during the low fishing activity period (February to September, see Figure 7.5), when many fishermen migrate seaward of the estuarine front to San Luis (see Figure 7.1). Both the observations and the economic scenario suggest that: • • • • •
the longer the peak of the fishing period the greater the capture, which depends mainly on hydroclimatic conditions (the ENSO-induced estuarine front location); the number of fishing days per month is crucial and this is highly dependent on the occurrence of southern winds (wind induced estuarine front location and mixing state); the sustainability of the fishing activity depends on several factors for which thresholds have been estimated: fishing activity must be ≥15 days/month and the fishing period must be >2 months; in the event of an increase in climatic constraints only the most able fishermen would be able to maintain present net income; and it will be necessary to continue this analysis for at least two more years with different climatic, environmental and socioeconomic conditions, as well as to incorporate anthropological research.
Future Climate Scenarios and Some Environmental Impacts Future climate changes for the Río de la Plata were extracted from global circulation model (GCM) runs (Bidegain and Camilloni, 2004). Estimates of future changes in mean temperature and precipitation over the region, for the baseline period 1961–1990, are based on two recent GCMs: the HADCM3 model from the Hadley Centre (UK) and the ECHAM4 model from the Max
3243 J&J Vulnerability
1/11/07
10:23 am
Page 148
148 Climate Change and Vulnerability
Planck Institute (Germany) run with the IPCC SRES emission scenarios A2 (medium high) and B2 (medium low). Based on the model results, changes in annual precipitation across the Río de la Plata are expected to vary between +0.1 and +0.2mm/day by the 2050s according to HADCM3 and between +0.0 and +0.6 mm/day by the 2050s according to ECHAM4 for the high emissions scenario (A2). In the case of the low emissions scenario (B2) it should vary between +0.0 and +0.3 mm/day by the 2050s according to HADCM3 and by +0.0 to +0.5mm/day by the 2050s according to ECHAM4. Annual temperature across the region would rise between +1.5°C and +3.0°C by the 2050s according to HADCM3, or by +0.5°C to +2.0°C by the 2050s according to ECHAM4, for the high emissions scenario (A2). For the low emissions scenario (B2) it should rise by +1.3°C to +2.5°C by the 2050s according to HADCM3, or +0.4°C to +2.0°C by the 2050s according to ECHAM4. Annual sea-level pressure across the Plata river basin by the 2050s, according to ECHAM4, indicates a southern displacement of the Atlantic subtropical high pressure. Under this scenario, and projecting the trends observed over the past three decades (changes in South Atlantic high pressure and sealevel pressure) and the increase in east and southeast winds reported by Escobar et al (2004), we can assume an increase in the frequency of onshore winds.
Some Conjectures about Climate Scenarios, Environmental Vulnerability and Impacts Current environmental scenarios (1971–2003) in the Plata river basin and estuary discussed in this chapter are dominated by the following main stresses: • • • •
increase in temperature, precipitation, streamflow, sea level and onshore winds; increase in population, use of natural resources and export of nutrients; increase in economic activities, land-use changes, soil erosion and runoff/infiltration ratio; and increase in symptoms of eutrophication.
Current climate and future scenarios (time horizon 2020–2050) presented in the previous sections for the Plata river basin and estuary suggest a potential change (since the 1961–1990 baseline conditions) in precipitation within the range +5 per cent to +20 per cent and in temperature from +1°C to +2°C. During the last few decades these changes have been observed to be +20 to +25 per cent for precipitation and +0.5°C to +0.8°C for temperature, along with a +25 to +40 per cent change for total river flows (QV). Trends for QV are very difficult to estimate because of both the uncertainty of regional human drivers (for example, land-use change) and the varied regional scenarios from different GCMs. Both the GCM scenarios used here i.e. the HADCM3 and ECHAM-4, have systematically underestimated the baseline precipitation over the region (i.e. estimated precipitation values lower
3243 J&J Vulnerability
1/11/07
10:23 am
Page 149
Climate and Water Quality in Río de la Plata 149
than the observed baseline precipitation for 1961–1990), but not necessarily the increase in it. Indeed, during the past few decades the observed increase in precipitation (Liebmann, 2004) was associated with the observed increase in river flow. Moreover, small changes in precipitation are doubled in the streamflow signal (Berbery and Barros, 2002). Tucci and Clarke (1998) have suggested that one third of observed increases in streamflows were attributable to land-use changes. From an environmental point of view, under one future scenario (for 2020–2050) in which streamflow remains similar or slightly lower (0 per cent to -10 per cent) with regard to present values, we do not expect a significant increase in present environmental stresses on the estuarine system (which are already moderately high). Our concern is about a future scenario where total river flow QV increases within the range 10–25 per cent, together with projected temperature increases, for which significant impacts are expected in the estuarine and coastal systems (increase in the vulnerability to trophic state changes and air–water and sediment–water gas and nutrient exchanges), besides changes in the shelf and contour current circulation (water temperature changes and ocean-driven HABs). Thus, the following question arises: Is such a scenario (increase in temperature and river flow) plausible under projected changes in temperature (of +1–2°C) and precipitation (~5–20 per cent), taking into account the consequent increase in evapotranspiration? Some assumptions are as follows: 1) about a third of QV changes may be due to land-use changes (Tucci and Clarke, 1998); 2) observed vs projected changes are of the same sign in all variables; and 3) the expected value for temperature change is 2 to 4 times greater than that formerly observed. A key question relates to the relative amounts of potential and actual evapotranspiration rates in the future. This uncertainty makes it very difficult to establish any coherent scenario about future streamflow, especially if current land-use changes continue increasing the runoff/infiltration ratio (which could reduce the impact of temperature rise on evapotranspiration). Considering the fact that seasonal temperature, precipitation and streamflow cycles are not superposed, any changes should modify seasonal circulation, stratification and mixing patterns, inducing further environmental shifts (such as changes in gas and nutrient exchanges), with a probable increase in both the degree and occurrence of hypoxic events (estimated to be about 20 per cent) in deep bottom waters and denitrification (emission of N2 to the atmosphere) (Nagy et al, 2002a and b; Nagy et al, 2003b), as well as an increase in the vulnerability of fishermen and low-lying areas. The long-term evolution of salinity off Montevideo (Nagy et al, 2002a; Bidegain et al, 2005) and the monthly evolution during the period 1998–2002 (Severov et al, 2004), as well as the evolution of the estuarine fronts location within the Río de la Plata estuary and adjacent shelf since 1998 (Severov et al, 2003 and 2004) allow the development of a conceptual model (not detailed here) for annual as well as monthly time periods: when the Uruguay river flow or QU is 4000m3/s, typical yearly salinities are >10–12 per cent, when QU is
3243 J&J Vulnerability
1/11/07
10:23 am
Page 150
150 Climate Change and Vulnerability
>5500m3/s they are ~7–9 per cent (present average: ~8 per cent) and when QU is >7000m3/s salinity is <5 per cent on timescales greater than weather development, whereas when QU is >10,000m3/s freshwater prevails in most of the Uruguayan coast and the estuarine front is displaced tens of miles to the mouth. Under a hypothetical environmental scenario for 2020–2050, based on climate models outputs, past trends, reference projections and expert judgement, some predictions can be made. Both long-term and monthly analysis of recent years suggests that a hypothetical increase in the QU of ~20 per cent should reduce average salinity at Montevideo by 2–3 ppt or parts per thousand (reaching 5 or 6 ppt) and displace riverward the estuarine front. If these changes were coupled with a plausible increase in onshore winds, especially during spring months, when ENSO is active, and biological processes and goods are in their activity peak period, significant changes in both the structure and location of the several fronts of the Río de la Plata as well as of the biological and biogeochemical functions (including fish reproduction, fishing activity and biogeochemical dynamics) can be expected (Severov et al, 2004; Nagy et al, 2004; López and Nagy, 2005; Lappo et al, 2005). If projected scenarios of human drivers (for example, +70–100 per cent population and ~150–200 per cent N input) (Nagy et al, 2003b) as well as landuse changes for the Uruguay river basin are to be accounted for, significant changes in symptoms of eutrophication and coastal fisheries livelihood can be expected. We need to assume a plausible pessimistic point of view. Taking into account the evolution, variability and extremes of temperature, precipitations, river flow and trophic state variables change during the past 10 to 30 years, and projecting them through the next 10–30 years, significant environmental impacts and changes can be expected soon.
Conclusion •
•
•
•
The overall vulnerability of the region and the impacts on the fresh and estuarine waters, coastal zone, ecosystem goods, processes and services on the Uruguayan coast of the Río de la Plata are primarily associated with ENSO-related climate variability. In addition, human and climatic shifts are drivers of environmental pressure and state variables within the Río de la Plata basin and sub-basins at different spatial scales and timescales, which produce, stimulate or trigger impacts, ecosystem responses and shifts within the estuarine waters of the Río de la Plata and its coastal zone. Increase in precipitation plays a major role in controlling buoyancy, eutrophication and production and respiration (P-R) processes of organic matter, since sensitivity to trophic state changes depends on the balance between river flow and winds. Projected scenarios for 2050 will increase vulnerabilities of the exposed sectors, and some of them will likely be heavily impacted and/or will
3243 J&J Vulnerability
1/11/07
10:23 am
Page 151
Climate and Water Quality in Río de la Plata 151
•
become unsustainable. While we have limited our discussion to the impacts on coastal fisheries in this chapter, other sectors affected include low-lying wetlands and beaches, tourism and supply of drinking water. The impacts on these sectors could be the subject of future studies. Changes in climate and the resulting changes in river flow will modify the circulation, stratification and mixing patterns of water in the Río de la Plata estuary, inducing further environmental impacts and increasing the vulnerability of fishermen.
References Baethgen, W., V. Barros, E. Berbery, R. Clarke, H. Cullen, C. Ereño, B. Grassi, D. P. Lettenmeier, R. Mechoso, and P. Silva Dias (2001) ‘Climatology and hydrology of the Plata Basin’, WRCP/VAMOS, www.clivar.org (also available from www.atmos.umd.edu/~berbery) Berbery, E. and V. R. Barros (2002) ‘The hydrologic cycle of the La Plata basin in South America’, Journal of Hydrometeorology, vol 3, pp630–645 Bidegain, M. and I. Camillloni (2004) ‘Performance of GCMs and Climate Baselines Scenarios for Southeastern South America’, paper presented at Latin America and Caribbean Assessment of Impacts and Adaptation to Climate Change (AIACC) Workshop, Buenos Aires, Argentina Bidegain, M., R. M. Caffera, V. Pshennikov, J. J. Lagomarsino, G. J. Nagy and E. A. Forbes (2005) ‘Tendencias climáticas, hidrológicas y oceanográficas en el Río de la Plata y costa Uruguaya’, in V. Barros, A. Menéndez and G. J. Nagy (eds) El Cambio Climático en el Río de la Plata, vol 14, Centro de Investigaciones del Mar y la Atmósfera (CIMA) – Universidad de Buenos Aires (UBA), Buenos Aires, pp137–143 Caffera, R. M., M. Bidegain, F. Blixen, J. J. Lagomarsino, G. J. Nagy, K. Sans, G. Dupuy, and R. Torres (2005) ‘Vulnerabilidad presente de los recursos hídricos y estado trófico del Río Santa Lucía a la variabilidad climática’, in V. Barros, A. Menéndez and G. J. Nagy (eds) El Cambio Climático en el Río de la Plata, vol 14, Centro de Investigaciones del Mar y la Atmósfera (CIMA) – Universidad de Buenos Aires (UBA), Buenos Aires, pp33–39 Calliari, D., M. Gómez and N. Gómez (2005) ‘Phytoplankton biomass and composition in the Río de la Plata estuary: Large-scale distribution and relationship with environmental variables’, Continental Shelf Research, vol 25, pp197–210 Camilloni, I., and V. Barros (2000) ‘The Paraná river response to El Niño 1982–1983 and 1997–1998 events’, Journal of Hydrometeorology, vol 1, pp412–430 De Jonge, V., M. Elliott and E. Orive (2002) ‘Causes, historical development, effects and future challenges of a common environmental problem: Eutrophication’, Hydrobiologia, vol 475/476, pp1–19 Díaz, A. F., C. D. Studzinski and C. R. Mechoso (1998) ‘Relationship between precipitation anomalies in Uruguay and Southern Brazil and sea surface temperature in the Pacific and Atlantic Oceans’, Journal of Climatology, vol 11, pp251–271 EPA (US Environmental Protection Agency) (2001) Nutrient Criteria Technical Guidance Manual: Estuarine and Coastal Marine Waters, Government Printing Office, Washington, DC Escobar, G., W. Vargas and S. Bischoff (2004) ‘Wind tides in the Río de la Plata estuary: Meteorological conditions’, International Journal of Climatology, vol 24, pp1159–1169 Ferrari, G. and G. J. Nagy (2003) ‘Assessment of vulnerability of trophic state in estu-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 152
152 Climate Change and Vulnerability arine systems: Development of a highly aggregated impact index’, paper presented at the First AIACC regional workshop for the Latin America and the Caribbean, San Jose, Costa Rica, Assessments of Impacts and Adaptations to Climate Change, International START Secretariat, Washington, DC, US, www.aiaccproject.org Framiñán, M. B. and O. B. Brown (1996) ‘Study of the Río de la Plata turbidity front: Part I: Spatial and temporal distribution’, Continental Shelf Research, vol 16, pp1259–1282 Garcia, N. O. and W. M. Vargas (1998) ‘The temporal climatic variability in the Río de la Plata basin displayed by the river discharges’, Climatic Change, vol 38, pp359–379 Gordon, D. C. Jr, P. R. Boudreau, K. H. Mann, and T. Yanagi (1996) ‘LOICZ Biogeochemical Modelling Guidelines’, LOICZ Reports and Studies, vol 5, LOICZ, Texel, The Netherlands Hernández, J. and P. Rossi (2003) ‘Characterization of the artisanal fishermen settlements of the frontal zone of the Río de la Plata’, in Research for Environmental Management, Fisheries Resources and Fisheries in the Saline Front, vol 16, EcoPlata, Montevideo, Uruguay, pp217–234 Hoffman, J. A. J. (1975) Atlas Climático de América del Sur, vol I, WMO-UNESCOCartographia, Geneva Kalnay, E., M. Kanamitsu, R. Kistler, W. Collins, D. Deaven, L. Gandin, M. Iredell, S. Saha, G. White, J. Woolen, Y. Zhu, M. Chelliah, W. Ebisuzaki, W. Higgins, J. Janowiak, K. C. Mo, C. Ropelewski, J. Wang, A. Leetma, R. Reynolds, R. Jenne and D. Joseph (1996) ‘The NCEP/NCAR 40-Year reanalysis project’, Bulletin of the American Meteorological Society, vol 77, pp437–471 Kane, R. P. (2002) ‘Precipitation anomalies in southern America associated with a finer classification of El Niño and La Niña events’, International Journal of Climatology, vol 22, pp357–373 Lappo, S. S., E. Morozov, D. N. Severov, A. V. Sokov, A. A. Kluivitkin and G. Nagy (2005) ‘Frontal Mixing of River and Sea Waters in Río de la Plata’, Transactions of the Russian Academy of Sciences (Earth Science Section), vol 401, pp267–269, (in Russian) Liebmann, B. C., L. Vera, L. Carvalho, I. Camilloni, M. Hoerling, D. Allured, V. Barros, J. Baez and M. Bidegain (2004) ‘An observed trend in Central South American precipitation’, Journal of Climatology, vol 17, pp4357–4367 López Laborde, J. and G. J. Nagy (1999) ‘Hydrography and sediment transport’, in G. M. Perillo, E. M. Pino and M. C. Piccolo (eds) Characteristics of the Río de la Plata: Estuaries of South America: Their Geomorphology and Dynamics, Springer-Verlag, Berlin, pp137–159 López, C. H. and G. J. Nagy (2005) ‘Cambio global, evolución del estado trófico y floraciones de cianobacterias en el Río de la Plata’, in V. Barros, A. Menéndez and G. J. Nagy (eds) El Cambio Climático en el Río de la Plata, vol 16, CIMA-UBA, Buenos Aires, Argentina, pp157–166 McCarthy, J. J., O. F. Canziani, N. A. Leary, D. J. Dokken and K. S. White (eds) (2001) Climate Change 2001: Impacts, Adaptation, and Vulnerability, Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK Méndez, S., D. N. Severov, G. Ferrari and C. Mesones (1996) ‘Early spring Alexandrium tamarense toxic blooms in Uruguay waters’, in Y. T. Oshima and Y. Fukuyo (eds) Harmful and Toxic Blooms, Intergovernmental Oceanographic Commission of UNESCO, Paris, pp113–117 Méndez, S., M. Gómez and G. Ferrari (1997) ‘Planktonic studies of the Río de la Plata and its oceanic front’, Chapter 4 in P. Wells and G. Daborn (eds) The Río de la Plata: An Environmental Overview, EcoPlata Project Background Report, Dalhousie University, Halifax, Nova Scotia
3243 J&J Vulnerability
1/11/07
10:23 am
Page 153
Climate and Water Quality in Río de la Plata 153 Menéndez, A. and M. Re (2005) ‘Hidrología del Río de la Plata’, in V. Barros, A. Menéndez and G. J. Nagy (eds) El Cambio Climático en el Río de la Plata, vol 7, CIMA-UBA, Buenos Aires, Argentina, pp69–83 Moss, R. (1999) Vulnerability to Climate Variability and Change: Framework for Synthesis and Modeling: Project Description, Battelle Pacific Northwest National Laboratory, Richland, WA, US Nagy, G. J. (2000) ‘Dissolved inorganic NP budget for the frontal zone of the Río de la Plata system: Estuarine Systems of South American Region: Carbon, nitrogen, and phosphorus fluxes’, in V. Dupra, S. Smith, L. I. Marshall-Crossland and C. J. Crossland (eds) LOICZ Reports and Studies 15, LOICZ, Texel, The Netherlands, pp40–43 Nagy, G. J. (2003) ‘Assessment of vulnerability and impacts to global change: Trophic state of estuarine systems’, paper presented at the First AIACC regional workshop for the Latin America and the Caribbean, San Jose, Costa Rica, Assessments of Impacts and Adaptations to Climate Change, International START Secretariat, Washington, DC, US, www.aiaccproject.org Nagy, G. J. (2005) ‘Vulnerabilidad de las aguas del Río de la Plata: Cambio de estado trófico y factores físicos’, in V. Barros, A. Menéndez and G. J. Nagy (eds) El Cambio Climático en el Río de la Plata, vol 15, CIMA-UBA, Buenos Aires, Argentina, pp145–155 Nagy, G. J., M. Gómez-Erache and A. C. Perdomo (2002a) ‘Río de la Plata’, in T. Munn (ed) The Encyclopedia of Global Environmental Change, vol 3 (Water Resources), John Wiley and Sons, New York, US Nagy, G. J., M. Gómez-Erache, C. H. López and A. C. Perdomo (2002b) ‘Distribution patterns of nutrients and symptoms of eutrophication in the Río de la Plata estuarine system’, Hydrobiologia, vol 475/476, pp125–139 Nagy, G. J., G. Sención, W. Norbis, A. Ponce, G. Saona, R. Silva, M. Bidegain and V. Pshennikov (2003a) ‘Overall vulnerability of the Uruguayan coastal fishery system to global change in the estuarine front of the Río de la Plata’, Poster Session, The Open Meeting Human Dimensions of Global Environmental Change, Montreal, Canada, October Nagy, G. J., C. H. López, F. Blixen, G. Ferrari, M. C. Presentado, A. Ponce, J. J. Lagomarsino, K. Sans, and M. Bidegain (2003b) El Cambio Global en el Río de la Plata: Escenarios, Vulnerabilidad e Indicadores de Cambio de Estado Trófico en la Costa Norte, V Jornadas Nacionales de Ciencias del Mar, Mar del Plata, Argentina Nagy, G. J., J. J. Lagomarsino, K. Sans and E. Andrés (2004) ‘Evaluación de las Cargas Contaminantes a la Costa Uruguaya del Río de la Plata y Atlántico. Evaluación del Estado Trófico (Nutrientes, Clorofila, Oxígeno disuelto e Indicadores Tróficos)’, final report, FREPLATA Project, Montevideo, Uruguay National Resource Council (2000) Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution, Ocean Studies Board and Water Science and Technology Board, Commission on Geosciences, Environment, and Resources, National Research Council, National Academy of Sciences Press, Washington, DC, Nixon, S. W. (1995) ‘Coastal marine eutrophication: A definition, social causes, and future concerns’ Ophelia, vol 41, pp199–219 Norbis, W. (1995) ‘Influence of wind, behaviour and characteristics of the croaker (Micropogonias furnieri) artisanal fishery in the Río de la Plata’, Fisheries Research, vol 22, pp43–58 Norbis, W., J. Verocai and V. Pshennikov-Severova (2003) ‘Artisanal fleet activity in relation to meteorological conditions (synoptic scale) during the fishing period October 1998–March 1999’, in Research for Environmental Management, Fisheries Resources and Fisheries in the Saline Front, vol 14, EcoPlata, Montevideo, Uruguay, pp191–197 Norbis, W., G. Saona, V. Pshennikov, G. J. Nagy, and G. Sención (2004) ‘Vulnerability
3243 J&J Vulnerability
1/11/07
10:23 am
Page 154
154 Climate Change and Vulnerability of Uruguayan artisanal fisheries of the frontal zone of the Río de la Plata to River Flow and wind variability’, paper presented at the Second AIACC regional workshop for the Latin America and the Caribbean, Buenos Aires, Argentina, Assessments of Impacts and Adaptations to Climate Change, International START Secretariat, Washington, DC, US, www.aiaccproject.org/meetings/meetings.html Norbis W., A. Ponce, D. N. Severov, G. Saona, J. Verocai, V. Pshennikov, R. Silva, G. Sención and G. J. Nagy (2005) ‘Vulnerabilidad y capacidad de adaptación de la pesca artesanal del Río de la Plata a la variabilidad climática’, in V. Barros, A. Menéndez and G. J. Nagy (eds) El Cambio Climático en el Río de la Plata, vol 3, CIMA, BsAs, Argentina, pp181–187 Pizarro, M. and A. Orlando (1985) ‘Distribución de fósforo, nitrógeno y silicio disuelto en el Río de la Plata’, Serv. Hidr. Naval. Secr. Marina, Publ. H-625, pp1–57 Pshennikov, V., M. Bidegain, F. Blixen, E. A. Forbes, J. J. Lagomarsino and G. J. Nagy (2003) ‘Climate extremes and changes in precipitation and wind patterns in the vicinities of Montevideo, Uruguay’, paper presented at the First AIACC regional workshop for the Latin America and the Caribbean, San Jose, Costa Rica, Assessments of Impacts and Adaptations to Climate Change, International START Secretariat, Washington, DC, US, www.aiaccproject.org Seitzinger, S., C. Kroeze, A. F. Bouwman, N. Caraco, F. Dentner and R. V. Styles (2002) ‘Global patterns of dissolved inorganic and particulate nitrogen inputs to coastal systems: Recent conditions and future projections’, Estuaries, vol 25, pp640–655 Severov, D., G. J. Nagy, V. Pshennikov and E. Morozov (2003) ‘SeaWifs fronts of the Río de la Plata estuarine system’, Geophysical Research Letters, vol, 5, p01914 Severov, D. N., G. J. Nagy, V. Pshennikov, M. De los Santos and E. Morozov (2004) ‘Río de la Plata estuarine system: Relationship between river flow and frontal variability’, paper presented at 35th COSPAR Scientific Assembly, Paris, France, July Stockholm Environment Institute (2001) ‘Framework for vulnerability’, The George Perkins Marsh Institute at Clark University and The Stockholm Environment Institute, Research and Assessment Systems, Stockholm, Sweden Tucci, C. E. M. and R. T. Clarke (1998) ‘Environmental issues in the La Plata basin’, Water Research Development, vol 14, pp157–174
3243 J&J Vulnerability
1/11/07
10:23 am
Page 155
8
Climate Change and the TourismDependent Economy of the Seychelles Rolph Antoine Payet
Introduction The Seychelles is a small island state of 115 islands spread over an exclusive economic zone covering an area of 1.37 million square kilometres in the Indian Ocean. They share characteristics with other small island states that, in the judgement of the Intergovernmental Panel on Climate Change (IPCC), make small islands especially vulnerable to the effects of climate change, sea-level rise and extreme events (Mimura et al, 2007). With a total land area of just 455.3 square kilometres, most of the population, infrastructure and economic activities of the Seychelles are concentrated on the coasts of three large islands. These islands are among the granitic group of 43 islands with mountainous peaks that rise steeply and constrain development to the very narrow 1 to 2kmwide coastal plains. The people and infrastructure that are concentrated in these coastal plains are highly vulnerable to storm surges and coastal erosion. The economy is heavily dependent on tourism and, to a lesser extent, on fisheries, both of which have their main infrastructure located in the exposed coastal plains and are dependent on natural resources such as coral reefs and freshwater supplies that are sensitive to climate variations, sea-level rise and other stressors. Tourism plays a role of unmatched importance in the economy of the Seychelles and any shocks that negatively impact the tourism industry are felt throughout the islands. This situation is not unique to the Seychelles and is a feature of the economies of many other small island states. Yet the vulnerability of tourism-dependent island states to climate change has been largely unstudied. Because of the importance of this issue, we undertook an analysis of the potential impacts of climate change on the tourism industry of the Seychelles. Results of the analysis and their implications for vulnerability for the wider economy and society are presented in this chapter. Also addressed are adaptation strategies for managing climate risks to the tourism industry.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 156
156 Climate Change and Vulnerability
Past Climate Events and their Impacts The impacts of the 1997–1998 El Niño and the 1998–2000 La Niña events demonstrate the sensitivity of the economy of the Seychelles to climate events (Payet, 2005). Combined losses in monetary terms caused by these events amounted to an estimated US$22 million, or 4 per cent of GDP. Figure 8.1 shows the distribution of losses by sector of the economy. Fisheries suffered the greatest loss, accounting for 45 per cent of the total, followed by agriculture, with 28 per cent of the losses. The tourism sector suffered an estimated US$2.6 million loss, or 12 per cent of the total.
Agriculture 28%
Fisheries 45% Forestry 3% Industry 7% Tourism 12%
Construction 5%
Figure 8.1 Share of economic losses by sector from the 1997–1998 El Niño and 1998–2000 La Niña events in the Seychelles Source: Payet (2005).
Other impacts included damage to coastal infrastructure such as roads and airstrips, as well as to the natural environment, including coral reefs, beaches and forests. Increased recession of the coastline has been observed since that period, primarily because of the destabilizing nature of higher tide levels and resulting damage to dune vegetation. Critical infrastructure necessary to support the economy and human welfare, such as houses, hospitals, power stations, electric supply, communication networks, emergency response centres, water supplies and schools, all located in the low-lying areas, were also affected by the torrential floods of August 1998. Intense rainfall events, nearly all of which are directly or indirectly associated with the Intertropical Convergence Zone and tropical cyclones in the region, are a significant climate hazard in the Seychelles. The Indian Ocean is the most prolific of all oceans in generating tropical cyclones. Tropical cyclones
3243 J&J Vulnerability
1/11/07
10:23 am
Page 157
Climate change and Tourism in the Seychelles 157
do not directly impact the granitic islands of the Seychelles as they lie too close to the equator. But the passage of cyclones to the south can result in the deepening of the intertropical convergence, culminating in feeder bands which can bring in heavy rainfall to the island of Mahe. Heavy rainfall events in Mahe of more than 200mm within 24 hours are shown in Table 8.1. Such heavy rainfall can lead to flooding and landslides resulting in loss of property and significant damage to agricultural crops on the granitic islands. Table 8.1 Intense rainfall events of above 200mm in a 24-hour period over the island of Mahe, Seychelles, 1972–1997 Region
District
West Mahe
Barbarons Anse Boileau Tracking Station Tea Factory Grand Anse East Mahe La Misere Cascade Seychelles Airport North Mahe La Gogue Belombre North East Point Le Niol South Mahe Val D’Endore La Plaine South Mahe (cont.) Anse Forbans Anse Royale Central Mahe New Port St. Louis Rochon Hermitage
Amount (mm)
Date Occurred
Weather Event
210.0 266.5 284.8 291.5 480.0 232.4 242.0 245.1 240.0 247.5 279.2 282.0 202.6 215.6 216.5 220.5 240.0 277.4 281.1 288.2
4 November 1996 15 August 1997 21 December 1986 4 November 1986 15 August 1997 3 February 1984 21 May 1990 2 February 1981 30 January 1991 30 January 1991 29 January 1991 15 August 1997 15 August 1997 15 August 1997 15 August 1997 15 August 1997 28 January 1978 4 November 1996 4 November 1996 27 January 1987
ITCZ Storm (El Niño) Storm (El Niño) Storm (El Niño) Storm (El Niño) Tropical cyclone TC ‘Ikonjo’ ITCZ Storm (El Niño) Storm (El Niño) Storm (El Niño) Storm (El Niño) Storm (El Niño) Storm (El Niño) Storm (El Niño) Storm (El Niño) Tropical cyclone ITCZ ITCZ El Niño/ITCZ
Note: ITCZ – Intertropical Convergence Zone; TC – tropical convergence. Source: Seychelles Meteorological Services.
Analysis of the frequency of cyclones and tropical depressions for the period 1959–1980 shows that there are three peak periods of activity: the last 10 days of December, the 10–20th of January and almost the whole month of February. No correlation is found between the activity in any one year and occurrence of El Niño.
Tourism in the Seychelles The tourism industry in the Seychelles is based on the natural beauty of its islands, the uncrowded beaches, clear coastal waters and year-round sunshine.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 158
158 Climate Change and Vulnerability
A study undertaken in 2004 found that more than 65 per cent of tourists choose to visit the Seychelles for the predominantly pristine nature of their coastal resources and the natural beauty in general (Cesar et al, 2004). Tourism has grown from a mere 3000 annual visitor arrivals in the early 1970s to over 132,000 in 2002. With an average length of stay of 10 days, the number of visitor nights spent by tourists in the Seychelles is roughly 1.3 million per year (Figure 8.2). This growth in tourism has been a principal driver of economic growth in the Seychelles. Tourism now accounts for about 29 per cent of foreign exchange earnings, 20 per cent of GDP and one third of employment.
Figure 8.2 Number of tourist nights per year in the Seychelles – Actual number of tourist nights and predictions from the Vision 2 master plan and a statistical model are shown Source: Cesar et al (2004).
However, tourist travel to the Seychelles is very sensitive to external market forces and regional political stability. For example, tourism numbers declined sharply from 1980 to 1983 as a result of economic recession in European markets (Gabbay and Gosh, 1997) and in 1991–1992 and 2003 during the two wars in the Persian Gulf region. The terrorist attacks of 11 September 2001 and the recent SARS outbreak (2004) also caused a drop in tourist arrivals. In the late 1990s, competition from other destinations also affected the performance of tourism in the Seychelles. The industry is also sensitive to changes in climate, as demonstrated by our study and described later in this chapter. The industry faces a number of challenges, including the need to: 1
compete more effectively with other destinations by providing greater value for money and enhanced quality levels;
3243 J&J Vulnerability
1/11/07
10:23 am
Page 159
Climate change and Tourism in the Seychelles 159
2 3 4 5
improve air access and airport infrastructure; diversify attractions and activities; address shortages of trained and qualified personnel; and protect the environment (MTT, 2001).
The Seychelles have traditionally demonstrated a strong development policy aimed at ensuring that tourism is environmentally sustainable. But with growing development pressure, many of these policies are being reviewed. The benefits of a policy to promote environmentally sustainable tourism have been significant in terms of allowing many areas of the country to remain undisturbed and retain most of the original landscape. With its pure white beaches that hug the granitic coastline, the Seychelles have been described by many tourist magazines as among the most beautiful paradise islands in the world, coining the tagline ‘unique by a thousand miles’. Consequently, the policy of the government has been to exploit the comparative advantages of the Seychelles as an ecotourism destination by promoting a wide range of naturebased attractions and activities related to the Seychelles’ unique natural environment. However, climate change can impact on the very resources that today are the basis for tourism.
Tourism and Climate Climate can and does impact tourism through its effects on the resources and infrastructure that are critical to tourism services and on the climate-related amenities that tourists seek when visiting destinations such as the Seychelles. The effects of climate change on tourism in small islands are expected to be largely negative (Mimura et al, 2007). The projected changes will expose the Seychelles to rising sea level, warmer temperatures and, probably, increasing rainfall (Christensen et al, 2007). The intensity of precipitation events is projected to increase generally, and, in particular, in tropical areas that experience increases in mean precipitation, where extremes are projected to increase more than the mean. Projections also suggest that tropical cyclones are likely to become more severe, with greater wind speeds and higher mean and peak intensity of precipitation (Meehl et al, 2007). Sea-level rise and greater intensities of rainfall and tropical cyclones would accelerate coastal erosion and beach erosion, threatening one of the primary tourist attractions of the Seychelles. Sea-level rise, rising sea surface temperatures, increased tropical cyclone intensity and changes in ocean chemistry from higher carbon dioxide concentrations are likely to negatively impact the health of coral reef systems, another major tourist attraction of the Seychelles and also important to the islands’ fisheries and conservation of biodiversity. Increased coral mortality would also accelerate coastal erosion, as demonstrated by the effects of coral mortality over the past decade in the Seychelles (Sheppard et al, 2005). Damage to or disruption of infrastructure located in the highly exposed coastal plains, such as the airport, hotels, roads, communication and electrical networks, water works and health facilities, would strongly and negatively
3243 J&J Vulnerability
1/11/07
10:23 am
Page 160
160 Climate Change and Vulnerability
impact tourism. Tourism places heavy demands on the scarce water supplies of the Seychelles and the industry is highly vulnerable to climate events that disrupt the supply of water (Payet and Agricole, 2006). Traditional economic tools have been used to estimate in monetary terms the direct and indirect impacts of climate on welfare, but most of this work is confined to temperate regions. Exceptions to this are estimates of losses from coral bleaching. Westmacott et al (2000) estimate that the tourism industry of the Maldives suffered a financial loss of more than US$3 million from severe coral bleaching in the Indian Ocean as a result of the 1997/1998 El Niño event. Aish and Cesar (2002), using the contingent valuation method, estimate an economic welfare loss of US$5 million from the mass coral bleaching of 1996. Cesar et al (2004) go on to estimate the welfare changes for scenarios of future coral reef bleaching. In their baseline scenario, there is no further bleaching and the coral reef environment is assumed to recover. Recovery of the reef is estimated to yield benefits of US$5 million to US$10 million over the next 25 years. An alternate scenario assumes that bleaching events continue and that there is no adaptation response. The result is catastrophic, with a potential loss of more than US$500 million over the next 25 years. Viner and Agnew (1999) examined the potential impacts of climate change on tourism in the Maldives and concluded that the main impacts would come from sea-level rise, elevated sea-surface temperatures and seawater intrusion on coastal lands. These changes in the coastal environment are expected to have impacts on tourism demand and also tourism resources such as beaches, coral reefs and water. Changes in the natural capital of tourism, as Viner and Agnew project for the Maldives as a result of climate change and which are likely in other small island destinations, would cause changes in the tourism attractiveness of a destination and tourist revenues. A number of studies have investigated the effects of climate on tourism demand. Research in the UK by Agnew (1995) and Benson (1996) shows that the main driver for sun-sea-sand tourism was primarily linked to climatic conditions in the UK, with people taking more tropical holidays to escape the cold European winters. This is confirmed by the UK Department of Environment and Transport study that found temperature to be the most important climate variable influencing annual domestic holiday tourism (Agnew, 1998). Similarly, work by Giles and Perry (1998) links increases in domestic tourism spending in the UK, and corresponding decreases in international travel, to warmer than usual summers at home. These examples indicate that both local and destination climate factors play a role in tourism demand in long-haul destinations such as the Seychelles. In a study of the German tourist market for travel to both local and international destinations, Hamilton (2003) found temperature, humidity and rainfall at the destination site influence tourism demand for those destinations. A positive relationship is found between demand and average temperature during the period of travel until the optimum mean temperature is reached, estimated to be 24°C, after which demand decreases for temperatures in excess of the optimum. Hamilton also estimates that 11.5 wet days per month is optimal, with demand falling when the number of wet days exceeds that amount.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 161
Climate change and Tourism in the Seychelles 161
Studies for the tourist markets in the UK (Maddison, 2001) and The Netherlands (Lise and Tol, 2002) also find that temperature influences tourist travel and estimate optimal temperatures of 29°C and 21°C for British and Dutch travellers respectively. Neither of these studies found precipitation to have a significant effect.
A Survey of Tourists to the Seychelles We conducted a survey of 400 tourists in the Seychelles to investigate their sensitivity to various climate variables such as precipitation, sea temperature, sunshine and cloud cover, state of the sea and wind conditions. The first part of the survey asked questions about the different activities that respondents might undertake during their stay in the Seychelles and how changes in climatic conditions might affect their overall satisfaction with the destination. The second part documented basic demographic variables of the interviewees. Written questionnaires were prepared in English, French, German and Italian; other tourists were all interviewed in English. The surveys were undertaken at the airport in the departure lounge (thus the tourists would also be able to comment on their level of satisfaction). Among the factors cited by respondents as their primary reasons for visiting the Seychelles include sunshine, tranquility, culture, clear waters, nature and food. Approximately 20 per cent of the tourists surveyed came to the Seychelles to enjoy the sun, while 18 per cent did so to enjoy peace and tranquility (see Figure 8.3). Because sunshine is closely linked to the number of wet days and cloudiness, the responses suggest that an increase in rainy days will affect the satisfaction of tourists in the Seychelles. Roughly 35 per cent of the respondents did state that their enjoyment of their stay would have been affected by rainfall, whereas only 10 per cent said that cloudiness would affect their enjoyment. The Seychelles experience an average of 12.6 wet days per month, or 42 per cent of the year, and considerably higher than the optimum estimated by Hamilton (2003), suggesting that increases in numbers of wet days would have a negative affect on tourist visits to the Seychelles. Tourists were asked about their attitude to rain days, including the number of rainy days they would tolerate during the period of their stay and the point at which rain days would cause them to change their travel plans. Almost 50 per cent of tourists reported concern about the number of rainy days during their stay, while the other half were not bothered. Very few tourists would change their plans following one day of rainfall, while 18 per cent of those surveyed indicated that they would change their travel plans after three consecutive days of rainfall. Temperature is an important factor for tourists visiting the Seychelles, but not as critical as rainfall. Tourist activities are primarily focused on the beach during the day and in the hotels in the evenings. The survey results indicate that, on average, tourists spend approximately 4.2 hours a day on the beach and about another 3 hours hiking, diving or sightseeing. Average daily temperature in the Seychelles is 27°C, which exceeds the estimated optimal
3243 J&J Vulnerability
1/11/07
10:23 am
Page 162
162 Climate Change and Vulnerability Don’t Know 2% Fresh Air 7%
Culture 16%
Clear Waters 15%
Tranquility 18% Food 9%
Nature 13%
Sun 20%
Figure 8.3 Primary reasons tourists visit the Seychelles
temperatures sought by travellers from Germany and The Netherlands. Increases in daytime temperatures during the hours when tourists are active would be likely to adversely affect comfort levels. Warmer evening temperatures would also probably reduce comfort levels at night because of the high humidity, but this could be ameliorated by cooling of indoor spaces. When asked about preferences for cooling during the night-time, such as fans and air-conditioning, 41 per cent stated that they would prefer to use airconditioning while only 28 per cent would prefer the ambient air temperature. One reason stated for the preference for the use of air-conditioning in the evening is the relatively high humidity levels, which may make it uncomfortable to sleep. Data collected in this survey show that tourists increasingly verify meteorological information for their travel destinations, either with their travel agent before planning a holiday or using the internet. Of the 400 tourists surveyed, 37 per cent undertake their own research into the climatic conditions prevailing in their destination of choice, while 30 per cent depend on the advice of their travel agent. Information presented and available on climate conditions such as rainfall or warm sunny periods can therefore influence the length and quality of the tourist visits to the Seychelles.
A Statistical Model of Tourist Visits to the Seychelles Results of the survey, as well as an examination of the literature on tourism demand, were used to specify a statistical model of climate effects on tourist visits to the Seychelles. Tourist arrivals are modelled as a logarithmic function
3243 J&J Vulnerability
1/11/07
10:23 am
Page 163
Climate change and Tourism in the Seychelles 163
of annual average air temperature, hours of sunshine, precipitation and relative humidity. Many other variables that are likely to influence demand for travel to the Seychelles are omitted due to lack of data. The model was estimated using data for the period 1972 to 2000. The results of the multivariate analysis are presented in Table 8.2. The partial coefficient for temperature shows a positive relationship with tourist arrivals, while sunshine hours and relative humidity show negative correlations. However, none of the estimated coefficients are statistically different from zero. Table 8.2 Results from estimation of the model of tourist demand Unstandardized Coefficients Predictors
B
Std. Error
Constant Temperature Sunshine Precipitation Relative humidity R Square Adjusted R Square Standard error of the estimate
4.880 0.024 –0.051 0.000 –0.015 0.281 –0.13
2.763 0.036 0.063 0.001 0.031
T
1.766 0.660 –0.806 –0.204 –0.466
Sig.
0.121 0.530 0.447 0.844 – 0.655
95% Confidence Interval for B Lower Bound
Upper Bound
–1.653 –0.061 –0.200 0.002 –0.088 0
11.414 0.108 0.098 0.001 .059
0.054
The lack of a statistically significant effect of climate variables on tourist arrivals in the Seychelles contrasts with the results of Hamilton (2003), Maddison (2001), Lise and Tol (2002) and others, as well as with the findings of our own survey of tourists. A key difference between our analysis and that of the other cited works is that these other studies estimated demands for travel to multiple destinations using cross-sectional data. In multiple destination, cross-section models, differences in the average climate conditions of the different destinations result in different demands for travel to the different destinations. In that context, climate does influence tourists’ choice of travel destinations. Our analysis sought to use time-series data of inter-annual climate variations to explain annual variations in travel to a single destination. The results suggest that inter-annual variations in climate do not strongly influence tourist travel to a destination.
Implications for the Seychelles Economy Hypothetical scenarios of increases in the number of wet days and in average temperatures are constructed to illustrate the sensitivity of the Seychelles
3243 J&J Vulnerability
1/11/07
10:23 am
Page 164
164 Climate Change and Vulnerability
economy to possible impacts of climate change on tourism. In constructing the scenarios it is assumed that the optimal number of wet days is 11.5 days per month, as estimated by Hamilton (2003), and that the optimal temperature is 29°C, as estimated by Maddison (2001). The average number of wet days per month is assumed to increase from the present average of 12.5 to 14.3 days by 2020. The increase is based on extrapolation of recent trends in the number of wet days. Such a scenario is plausible given climate model projections of rainfall increases in the Indian Ocean. We assume, as a worst case scenario, and based on responses to our survey, that the increase in wet days would decrease tourist arrivals by 39 per cent. Average temperature is assumed to increase from the present average of 27°C to 31°C in 2020, an increase that is extreme in comparison to recent estimates from the IPCC (Christensen et al, 2007). Estimates from Maddison imply that average temperatures at a destination site 2°C in excess of the optimum would cause 10 per cent of tourists to opt for local tourism, and that of those who do go, 24 per cent would prefer to remain inactive during the day due to discomfort. Additional effects would be an increase in the use of airconditioning, estimated at 28 per cent, and an increase in water demand, estimated at 2 per cent. The economic impacts are measured as estimated changes in tourism earnings and the social impacts are measured as effective job losses, which includes losses in tourism and its multiplier effect in the economy. The results are shown in Table 8.3. Economic costs would amount to approximately US$72 million in net present-value terms. This works out at approximately US$743 per capita. The impact would be felt in all areas of the economy. Hotel developers would face reduced occupancy rates, revenues and profits. Government receipts from tourism would be reduced by US$13 million in net present-value terms, which would, in turn, affect social programmes and infrastructure investments. Recreational revenue losses could undermine the management and protection of many protected areas, including marine parks, which would result in additional welfare losses. Increases in the consumption of water and electricity would cost over US$3 million. The estimated economic and social impacts are based on hypothetical scenarios that make some rather extreme assumptions. However, they are useful for understanding the sensitivity of the Seychelles economy to climate impacts on tourism and also as a benchmark of the scale of potential losses. Omitted from these estimates are potential losses from impacts on coral reefs, beaches, water supplies and other natural resources, as well as impacts to infrastructure.
Adaptation Strategies Tourism in the Seychelles and the wider economy of the islands are vulnerable to impacts of climate change on tourism demand and on the natural resources that support the tourism industry. While further research is needed to address many unanswered questions about how global climate change will affect the climate of the Seychelles, and how resources and tourism will be impacted,
3243 J&J Vulnerability
1/11/07
10:23 am
Page 165
Climate change and Tourism in the Seychelles 165
Table 8.3 Social and economic impacts from increases in average number of wet days and temperatures Scenario assumptions
Tourist response
Annual Economic Cost (US$ million)
Annual Net Present Effective Value* Job Losses (US$ million) (US$ million)
Increase in number of wet days (>11.5 days/month) Increase in temperature during the active hours of the day (>29°C)
39% of tourists may cancel their trip
–52.3
–21.0
–46.9
Local tourism favoured over long-haul, 10% tourists may opt for a local holiday Increased periods of inactivity, est. 24% loss by recreational sector Increased use of A/C – power demand increased by 28% Increase in water demand by 2% above normal demand rate
–13.4
–5.4
–12.1
–12.5
–2.7
–9.8
+0.4
–3.0
–0.4
–0.3
*Discounted over 15 years at 3 per cent net present value.
action is needed to prepare for a changing climate. Four adaptation strategies for tourism in small island states are proposed. Each is needed and would yield significant benefits whether or not climate change brings new or greater hazards to the Seychelles. But the threat of climate change gives greater impetus and urgency to their consideration and implementation, which would build necessary resilience for coping with climate change.
Long-term sustainable planning and management of tourism infrastructure The importance and long-term benefits of planning for tourism development have been demonstrated in many countries, including small island states. Planning for sustainable tourism development involves the harnessing of the powerful earning power of tourism to promote sound environmental management, economic development and social progress, without compromising the integrity of existing support systems. In the Seychelles, several poorly planned tourism developments are now being used as lessons for the future. One example includes a hotel that was built too close to the beach – with increasing erosion, the costly undertaking of maintaining the coastal infrastructure is decreasing the overall profits of the company. As a result, the hotel is finding
3243 J&J Vulnerability
1/11/07
10:23 am
Page 166
166 Climate Change and Vulnerability
it difficult to obtain further financing for hotel improvements and enhancements for its staff. Poor practices such as this reduce the ability of tourism infrastructure to face the emerging hazards in the changing global climate. While tour operators and other support services can easily exit tourism markets, hotels cannot do so, because of their huge investments in immovable property. Hence the need for proper policies that ensure those properties retain their maximum ecological and tourism value over time. Several hotels in the Seychelles have dedicated conservation programmes in which they allocate a certain amount of their profits towards habitat restoration and protection on or adjacent to their property. A number of new hotels also provide examples of good practice, having been built with good set-back distances from the beach with adoption of proper beach management plans, better architectural design and landscaping to promote cooler garden spots and living environment for tourists, as well as energy conservation through better building design. Modification of coastlines for private hotel services such as marinas and slipways, including removal of coral rock and discharge of pollution, all contribute to stress on vital coral reef ecosystems. In addition to the biodiversity and fisheries supported by coral reefs, they also play an effective role in protecting coastlines against extreme events. Protection of the reefs will be important for preventing coastal erosion in the long term. In the Seychelles, the effective implementation of environmental impact assessment regulations has been crucial in redressing some of the past mistakes (Payet, 2003). Stresses on reef systems from sea-level rise and warmer sea surface temperatures can be addressed through improved building and landscape design as well as maintenance of green areas and beach vegetation.
Diversify recreational tourism resources Many of the resources in small island states are undervalued in terms of their ecotourism services. The recent mass coral bleaching event which threatened to collapse the diving industry in the Maldives highlighted the fact that many small island states have too many of their eggs in one basket. Although diversification can be another challenge in itself, many Caribbean island states have become financial and banking centres, which provide additional sources of revenue. Other island states like the Seychelles have diversified their economies into fisheries and offshore services with the aim of reducing dependency on tourism. However, in many cases, and certainly in the Seychelles, tourism will remain the mainstay of the economy in the next 25 to 50 years. Consequently, it is important that all tourism activities cause minimum impacts on the environment and ecotourism activities are as non-intrusive as possible. Involvement of the local communities in local tourism development and diversification is a proven method to develop adequate resilience when external shocks in the tourism industry propagate to a small island state. Such efforts will optimize revenue from recreational activities and also ensure that all members of local communities benefit. In turn, they can better commit to the protection of the coastal resources.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 167
Climate change and Tourism in the Seychelles 167
Opportunities for diversification in other types of marine ecotourism are available but, as with any ecotourism activities, proper guidelines and policies will need to be agreed on and proper training given. In 2001 the Seychelles Government adopted an ecotourism strategy to improve and diversify tourism experiences as part of a policy of encouraging ecotourism (MTT, 2001). Opportunities such as whale, dolphin, whale shark and turtle watching provide tremendous opportunities for small island states to recover from losses as a result of coral bleaching. Creating artificial reefs, which later can become areas for diving, is still an expensive option for many small island states but worth consideration. On the other hand, making substantial investments in the protection and management of important coral reef areas is critical for future re-growth of the coral reefs (Engelhardt, 2002). The Seychelles are also studying opportunities for developing low-scale eco-lodges on some of the inhabited outer islands in an effort to ensure proper management and conservation of those islands.
Offer unique experiences and refrain from tourism intensification As discussed earlier, with a gradual warming in the European region, many potential long-haul tourists would be induced to stay in Europe and take local holidays. However, beaches in Europe would be likely to become so crowded that making a long-haul trip to a non-crowded, quiet and unique beach destination would become an increasingly attractive option. This was especially emphasized by tourists interviewed in our survey, who ranked tranquility and cultural experiences as major motivations for visiting the Seychelles. The uncrowded beaches and unspoiled environment of the Seychelles may command an increasing premium in the tourist market, provided that development is properly managed to preserve the attributes that give the Seychelles an advantage. The survey also indicates that some tourists are insensitive to climatic variables such as rainfall and temperature, as long as the destination promises to offer a unique and thoroughly holistic rejuvenation for the city dweller. The current proliferation of spas and other holistic treatment centres are too focused on infrastructure; nothing beats a spa in a natural environment. While city-based spas are also successful, island destinations can offer such unique experience in a non-aggressive environment and, in many cases, enable the development of community services in terms of providing such goods as herbal extracts and other possibilities.
Capacity building and coordination The cornerstone of any adaptation strategy is the ability to build resilience across the entire population strata. Capacity building for sustainable management needs to penetrate government agencies, the private sector, non-governmental organizations and civil society. Achievements in the first three adaptation strategies will not be sustainable if these are not driven by sus-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 168
168 Climate Change and Vulnerability
tainability principles, efficiency gains across the various sectors of the economy, improved coordination of activities aimed at restoration of particular degraded habitats and human values. Implementing the different strategies will require significant financial and technical resources. Implementing sustainable planning and management of the tourism sector is estimated to cost US$600,000 annually. Diversification of tourism resources and offering unique tourist experiences would cost US$300,000 and 500,000 per year respectively, while capacity building costs are estimated at roughly US$1 million per year. The total net present value of costs to implement these strategies over the next 15 years is US$21 million. It is expected that through such comparatively small investments, tourism losses as a result of the two climate variables considered in this study may be greatly minimized. Although it is not possible to avoid all potential losses due to limitation in adaptation, such strategies will yield a net benefit of more than the adaptation investment made. Because tourism is a very dynamic industry, with fierce competition and high risks, achieving common ground and translating these plans into action is a challenge, and one by which varied political, business and community interests will be affected. The integrated coastal zone management framework provides a good platform for such approaches to be effectively implemented, and many successful coastal zone management plans are now in place in various parts of the world, including many small island states.
References Agnew, M. (1995) ‘Tourism’, in J. Palutikof, S. Subak and M. Agnew (eds) Economic Impacts of the Hot Summer and Unusually Warm Year of 1995, Department of the Environment, Norwich, UK, pp139–147 Agnew, M. D. (1998) ‘Domestic holiday tourism’, in Indicators of Climate Change in the UK: A Pilot Study for the Global Atmosphere Division, Department of Environment and Transport, ITE, Penicuik, Scotland, pp49–51 Aish, A. and H. Cesar (2002) Economic Analysis of Coral Bleaching in the Indian Ocean Phase II, report no 0-02/08, prepared for the World Bank in support of the coral reef degradation in the Indian Ocean (CORDIO) Programme, Institute for Environmental Studies, Amsterdam, The Netherlands Benson, K. (1996) ‘Focus on weather economics’, in Window on the Economy, Kleinwort Benson Research, London, UK, pp4–22 Cesar, H., P. Van Buekering, R. A. Payet and E. Grandcourt (2004) Evaluation of the Socio-Economic Impacts of Marine Ecosystem Degradation in the Seychelles, World Bank/GEF/Government of Seychelles, Seychelles Christensen, J. H., B. Hewitson, A. Busuioc, A. Chen, X. Gao, I. Held, R. Jones, R. Koli, W. Kwon, R. Laprise, V. Rueda, L. Mearns, C. Menendez, J. Raisanen, A. Rinke, A. Sarr and P. Whetton (2007) ‘Regional climate projections’, in S. Solomon, D. Qin, M. Manning, Z. Chen, M. C. Marquis, K. Averyt, M. Tignor and H. L. Miller (eds) Climate Change 2007: The Physical Science Basis, Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US Engelhardt, U., M. Russel and B. Wendling (2002) ‘Coral communities around the
3243 J&J Vulnerability
1/11/07
10:23 am
Page 169
Climate change and Tourism in the Seychelles 169 Seychelles Islands 1998–2002’, in O. Linden, D. Souter, D. Wilhelmsson and D. Obura (eds) Coral Reef Degradation in the Indian Ocean: Status Report 2002, CORDIO, Univeristy of Kalmar, Sweden, pp212–231 Gabbay, R. and R. Ghosh (1997) ‘Tourism in the Seychelles’, Discussion Paper 97.08, Centre for Migration and Development Studies, Department of Economics, University of Western Australia, Perth, Australia Giles, A. and A. Perry (1998) ‘The use of a temporal analogue to investigate the possible impact of projected global warming on the UK tourist industry’, Tourism Management, vol 19, pp75–80 Hamilton, J. M. (2003) Climate and the Destination Choice of German Tourists, Working Paper FNU-15 (revised), Research Unit Sustainability and Global Change, Centre for Marine and Climate Research, University of Hamburg, Hamburg, Germany Lise, W. and R. Tol (2002) ‘Impact of climate on tourist demand’, Climatic Change, vol 55, pp429–449 Maddison, D. (2001) ‘In search of warmer climates? The impact of climate change on flows of British tourists’, Climatic Change, vol 49, pp193–208 Meehl, G., T. Stocker, W. Collins, P. Friedlingstein, A. Gaye, J. Gregory, A. Kitoh, R. Knutti, J. Murphy, A. Noda, S. Raper, I. Watterson, A. Weaver and Z. Zhao (2007) ‘Global climate projections’, in S. Solomon, D. Qin, M. Manning, Z. Chen, M. C. Marquis, K. Averyt, M. Tignor and H. L. Miller (eds) Climate Change 2007: The Physical Science Basis, Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US Mimura, N., L. Nurse, R. McLean, J. Agard, L. Briguglio, P. Lefale, R. Payet and G. Sem (2007) ‘Small islands’, in M. Parry, O. Canziani, J. Palutikof and P. van der Linden (eds) Climate Change 2007: Impacts, Adaptation and Vulnerability, Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US MTT (Ministry of Tourism and Transport) (2001) Vision 21: Tourism development in Seychelles 2001–2010, Government of Seychelles, Seychelles Payet, R. A. (2003) ‘Effectiveness of the environment impact assessment process in managing tourism development in Seychelles’, in B. Chaytor and K. R. Gray (eds) International Environmental Law and Policy in Africa, Environment and Policy Series vol 36, Kluwer Academic Publishers, Amsterdam, The Netherlands, pp327–349 Payet, R. A. and W. Agricole (2006) ‘Climate change in the Seychelles: Implications for water and coral reefs’, AMBIO, vol 35, no 4 pp182–189 Sheppard, C., D. Dixon, M. Gourlay, A. Sheppard and R. Payet (2005) ‘Coral mortality increases wave energy reaching shores protected by reef flats: Examples from the Seychelles’, Estuarine, Coastal and Shelf Science, vol 64, pp223–234 Viner, D. and M. Agnew (1999) Climate Change and Its Impact on Tourism, Report Prepared for WWF-UK, Climatic Research Unit, University of East Anglia, Norwich, UK Westmacott, S., H. Cesar, and L. Pet-Soede (2000) ‘Socio-economic assessment of the impacts of the 1998 coral reef bleaching in the Indian Ocean: A summary’, in D. Souter, D. Obura, and O. Lindén (eds) Coral Reef Degradation in the Indian Ocean, status report, CORDIO, Stockholm, Sweden, pp143–159
3243 J&J Vulnerability
1/11/07
10:23 am
Page 170
3243 J&J Vulnerability
1/11/07
10:23 am
Page 171
Part IV:
Rural Economy and Food Security
3243 J&J Vulnerability
1/11/07
10:23 am
Page 172
3243 J&J Vulnerability
1/11/07
10:23 am
Page 173
9
Household Food Security and Climate Change: Comparisons from Nigeria, Sudan, South Africa and Mexico Gina Ziervogel, Anthony Nyong, Balgis Osman-Elasha, Cecilia Conde, Sergio Cortés and Tom Downing
Introduction Food security, which became a catch phrase in the mid-1990s, can be defined as the success of local livelihoods in guaranteeing access to sufficient food at the household level (Devereaux and Maxwell, 2001). The failure of early solutions to the problem of food insecurity in the 1970s and 1980s was largely attributed to their technological bias, stressing production rather than equitable distribution, access, affordability and utilization. Since then, it has become clear that food security revolves around complex issues that encompass a wide range of interrelated environmental (and climatological), economic, social and political factors. Addressing food security, therefore, requires an integrated approach (as highlighted in Chapter 1 of this volume) that can enable many regions to find adequate and effective solutions to food access and availability issues. Early models projecting world food demand and supplies into the 21st century generally showed that global food supplies would match or exceed global food demand at least within the next two to three decades (Devereux and Edwards, 2004). One shortcoming of these models, however, is that their scales are very coarse and conceal regional disparities that are a major concern for already food-insecure regions (Stephen and Downing, 2001). Another shortcoming is that the models paid little or no attention to climate variability and change. Climate variability and change are major threats to food security in many regions of the developing world, which are largely dependent on rainfed and labour-intensive agricultural production (Parry et al, 1999 and 2004; Döös and Shaw, 1999; McCarthy et al, 2001; Gregory et al, 2005). Although there is research on the impact of climate on food production,
3243 J&J Vulnerability
1/11/07
10:23 am
Page 174
174 Climate Change and Vulnerability
there is limited understanding of how climate variability currently impacts food systems and associated livelihoods (Downing, 2002; Ziervogel and Calder, 2003; Gregory et al, 2005). This needs to be better understood before assessing the impact of climate change on food security. Variability is a measure of the frequency distribution of the value of climate variables and their range over a given time period. Temperature and precipitation are the most critical climate variables to measure with regard to food systems. Not only does the range between high and low values matter, but also the frequency at which these extremes occur and the intensity of the events. The focus in this chapter is on the impact of below-normal rainfall and drought on food security. The Third Assessment Report of the Intergovernmental Panel on Climate Change (Watson et al, 2001) projects that areas that are currently dry might experience an average increased dryness with global warming. It is important to note that with a climate that is warmer on average, even if there is no change in the amplitude of El Niño, the risk of droughts and floods that occur with El Niño will increase. In southern Africa, there is evidence that the drought experienced during the second half of the 20th century has been influenced by greenhouse gases, and drying trends are projected to continue (Hoerling et al, 2006). Variability is also expected to increase with more rain falling in intenserainfall events, larger year-to-year variations in precipitation in areas where increased mean precipitation is projected and increased variability of Asian summer monsoon precipitation (Watson et al, 2001). Although the issue of food security is directly linked to climate variability and change (Winters et al, 1999; Reilly, 1995), it must be noted that climate is not the single determinant of yield, nor is the physical environment the only decisive factor in shaping food security (Parry et al, 2004). Despite understanding the multidimensional nature of food insecurity, it remains a key concern affecting the livelihoods of marginal groups. Therefore, understanding the impacts of climate variability, as well as the possible changes in this variability on food security, is critical to making improvements in food security. Food insecurity at the household level often results in resources being diverted. For example, resources that might have been used to support the development of livelihoods (for example, education, healthcare and employment) get reallocated to ensure that basic food needs are met. The acquisition of food for marginal groups often entails a delicate balance of producing food for the household under stressed conditions at the same time as drawing on social and economic resources to access available food. When conditions in the environment vary (for example, climate, soil and water characteristics and land-use changes), this can place an additional stress on food production (McConnell and Moran, 2000). There are many levels at which a food system can be examined (Stephen and Downing, 2001). Food policy, trade and resource use are governed by decisions at national, regional and global levels. Global climate is part of a global system but influenced by the actions of individual large countries such as the US, China and India. The boundaries between these systems are not clear. And this is true for the impact of these global, regional and national systems on the local level of food systems. Yet it is at the local, individual and
3243 J&J Vulnerability
1/11/07
10:23 am
Page 175
Household Food Security and Climate Change 175
household levels that food is used, and it is people who must ensure their access to food; otherwise, they can become food insecure. Although individual access depends on factors at numerous scales, the first level of analysis that determines the nature of food insecurity is the local level of household livelihoods. In this chapter, we examine household food security from a livelihood perspective for a number of regions around the world where food insecurity is a stress to rural livelihoods. We draw on case studies from semi-arid regions in Nigeria, Sudan, South Africa and Mexico,1 and tease out the commonalities and differences from the case studies to learn some lessons about climate variability and food security on a livelihood scale. Although vulnerability to climate change, including vulnerability to food insecurity, is highly differentiated across continents, countries and livelihood systems, it is important to explore common strands across regions regarding food security and its determinants. Although each of the regions selected for this study is a drought-prone region, they are not all equally vulnerable to drought. Some, through various policies and adaptation strategies, have reduced their food insecurity resulting from droughts. Sharing their experiences could help other vulnerable regions deal with their food insecurity. Furthermore, in a globalizing world economy, regional integration is commonly being adopted to solve environmental problems. Identifying regional drought-related food insecurities could also lead to devising regional solutions to tackle such problems. The four case studies use different research approaches, yet all of the projects focus on food security, climate variability and climate change. Data for the studies were collected from households using ethnographic and interview research techniques. The household was adopted as the unit of analysis, as the household level tends to be where decisions about household production, investment and consumption are made in most agrarian societies, particularly under long-lasting drought conditions. Questions that this paper seeks to answer include: • • •
What factors determine a household’s vulnerability to food insecurity? What are the differences and commonalities with respect to these factors across the study regions? and What are the implications of these in addressing climate policy as related to food insecurity at regional or country-wide scales?
Climate Variability and Change and Food Security Food security depends on availability of food, access to food and utilization of food (FAO, 2000). Food availability refers to the existence of food stocks for consumption. Household food access is the ability to acquire sufficient quality and quantities of food to meet all household members’ nutritional requirements. Access to food is determined by physical and financial resources, as well as by social and political factors. Utilization of food depends on how food is used, on whether food has sufficient nutrients and on whether a balanced diet
3243 J&J Vulnerability
1/11/07
10:23 am
Page 176
176 Climate Change and Vulnerability
can be maintained. It is these three facets of the food system that all need to be met in order for food security to be realized. Each of these facets can be impacted by climate variability (Gregory et al, 2005); these impacts are discussed below.
Impact of climate variability and change on food availability The consensus of scientific opinion is that countries in the temperate, high-latitude and mid-latitude regions are generally likely to enjoy increased agricultural production, whereas countries in tropical and subtropical regions are likely to suffer agricultural losses as a result of climate change in coming decades (Arnell et al, 2002; Devereux and Edwards, 2004). It should be noted that the favourable assessment for temperate and high-latitude regions is based primarily on analyses of changes in mean temperature and rainfall; relatively little analysis done to date takes account of changes in variability and extremes. Impact of climate variability on crop production should be a priority given that analyses of agricultural vulnerability indicate that the key attributes of climate change are those related to climatic variability, including the frequency of nonnormal conditions (Bryant et al, 2000; Smit et al, 2000). Climate variability directly affects agricultural production, as agriculture is inherently sensitive to climate conditions and is one of the most vulnerable sectors to the risks and impacts of global climate change (Parry et al, 1999). Climate change could impact on growing season and plant productivity (Gregory et al, 2005). Many factors impact the type of policies implemented at a national level (such as domestic politics, redistribution of land/wealth, exchange rates and trade issues). Climate variability should be factored into these policies, as these policies can impact the availability of staple foods, for example, by providing incentives to grow crops appropriate for the climate conditions. In the case study sites, the two major forms of agricultural production are arable farming and pastoralism. Because of the limited amount and uneven distribution of rainfall over both time and geographic scope at the study sites, rainfall represents the most limiting factor for agricultural and livestock production. Its consequences are well known to local populations: the drying out of water sources, scarcity of grazing land, shortage of dairy products, loss of wild plants for gathering, migration of grazers, bad harvests and livestock losses, among others. For instance, it has been estimated by the World Bank that around 10 per cent of the population of Sub-Saharan Africa is primarily dependent on their animals, while another 58 per cent depend to varying degrees on their livestock. Increasing population pressures, interacting with declining rainfall and reduced pasture, have already begun to impact the livestock sector negatively. Rangeland condition is directly affected by the climate (as highlighted in the section on desertification in Chapter 1 of this volume) and, in turn, directly affects the quality and quantity of small and large livestock and associated livelihood activities.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 177
Household Food Security and Climate Change 177
Impact of climate variability and change on food access Individuals have sufficient access to food when they have ‘adequate incomes or other resources to purchase or barter to obtain levels of appropriate foods needed to maintain consumption of an adequate diet/nutrition level’ (USAID, 1992). Food access depends on the ability of households to obtain food from purchases, gathering, current production or stocks, or through food transfers from relatives, other members of the community, the government or donors. Intra-household distribution of these resources is an important determinant of food security for all household members. Food access is also influenced by the aggregate availability of food in the market, market prices, productive inputs and credit (USAID, 1992). Poor market infrastructure and an unfavourable policy environment may lead to high and variable prices for food and inputs, further undermining agricultural productivity, food supplies and derived incomes. Access depends on both physical factors and social and economic factors. After food is produced, it needs to be moved from the point of production to the point of consumption. This often depends on transport systems. In many developing countries, inefficient and ineffective transport systems retard delivery and increase the price of food. Climate change is expected to place a strain on transport systems (McCarthy et al, 2001). For example, increased heat stress may reduce the life of roads, and windstorms can impact transit at air and sea terminals as well as damaging infrastructure which may create delays (Perry and Symons, 1994). During droughts, people are known to move into marginal lands. Most of these marginal lands may not have good road access, and transporting food from such marginal farms poses a huge challenge.
Impact of climate variability on food utilization Adequate food utilization is realized when ‘food is properly used, proper food processing and storage techniques are employed, adequate knowledge of nutrition and child care techniques exists and is applied, and adequate health and sanitation services exist’ (USAID, 1992). Food utility involves how food is used. This can include how often meals are eaten and of what they consist. Constraints on food utilization include loss of nutrients during food processing, inadequate sanitation, improper care and storage, and cultural practices that negatively impact consumption of nutritious foods for certain family members. In many areas where food is produced and consumed locally, food utility changes with seasonal variation and food availability changes throughout the year. The hungry season is the time before the planted crops are ready to be eaten. Similarly, at harvest time, there might be festivals and a lot of food consumed. If there has been a drought and food availability is low, the range of food available often decreases, and so the meal frequency can decrease and the balance of nutrients can be inadequate. This can lead to malnutrition in children. It is also important to note that climate can have an impact on food utility indirectly. For example, if there are hot dry days, crops and vegetables may be
3243 J&J Vulnerability
1/11/07
10:23 am
Page 178
178 Climate Change and Vulnerability
dried so that they can be used later in the year. At the same time as seasonal crop production, many households face fluctuations in cash and in-kind income, both within a single year and from year to year. Agricultural households may face seasonal fluctuations in income related to crop cycles. Year-to-year fluctuations in income can result from varying agroclimatic conditions and climate variability.
Household food security and livelihoods Livelihoods can be considered as the combined activities and available social and physical assets that contribute to the household’s existence (Carney, 1998). Each individual has his or her own means of securing a livelihood, and the individuals together make up the household’s packages of livelihood assets and strategies. These strategies are pursued within a larger context that often determines whether these strategies will succeed or fail. The livelihoods approach is useful for understanding food insecurity as it emphasizes the importance of looking at an individual’s capacity for managing risks, as well as the external threats to livelihood security, such as drought (Chambers, 1989; Scoones, 1998; Carney, 1998; Moser, 1998). It enables the agency of individuals to be captured in their decision-making process (Ziervogel, 2004). For example, if one household has a member who works in the city and remits money and the household has a productive field, access to food sources may be spread through own production and food purchasing. If there is a drought and crops fail, the household may still have access to food if money continues to be remitted. If a household absorbs more children through the death of family members, then utilization of the existing food sources may be stressed and the number of meals reduced. This may result in a family being forced to remove children from school so that they can work to try and increase access to food that will result in improved utilization. A government grant may ameliorate this impact. It is clear that food insecurity depends on the agency of individuals and the components of household livelihoods that are interlinked with the three facets of food security, as explained above. Livelihoods of households can be compared if similar characteristics and activities in household livelihoods are grouped together to cluster livelihood typologies. Examples of typologies might include small-scale farming livelihoods or informal trade-based livelihoods. These help focus on an intermediate system level that draws from the local but has a unit of analysis which is greater. Recognizing livelihood typologies is a useful construct for comparing livelihood systems between regions (Dixon et al, 2001). A number of livelihood typologies can coexist and can vary in their geographical extent. In some instances, a livelihood typology may draw on certain environmental resources, such as coastal resources for fishing; in other instances, they may cross national boundaries, such as livestock-based livelihoods. The predominant livelihood typology for each case study is expanded in detail in the following case studies section.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 179
Household Food Security and Climate Change 179
Case Studies Case studies examining food security in Nigeria, Sudan, South Africa and Mexico are compared to identify commonalities and differences in how local food systems are impacted by global environmental change.1
Case Study 1: Mangondi village, Limpopo Province, South Africa Mangondi village is situated within the Vhembe district, Limpopo Province, in the northeast region of South Africa. Many of the previously disadvantaged farmers in the area have begun increasing the size of their production to try and enter into the fruit and vegetables market. Although there has been an increase in productivity among some previously disadvantaged farmers, there are still many constraints that are faced, and many of the poorer households remain food-insecure and do not have access to production or employment opportunities. A key constraint on farming in this area is high climate variability, as numerous droughts and floods have occurred in recent decades (for example, the 2000 floods or the 2002/2003 drought). Managing this climate variability as well as possible is of paramount importance when many other stressors (such as land access, political instability, market fluctuations, globalization and HIV/AIDS) enter into the equation (Ziervogel and Calder, 2003). Marketing is also a key concern. Although former subsistence farmers have started growing products for sale, particularly when they have access to irrigation, it is challenging to find markets that will buy products consistently, due to variable demand, prices and quality of produce. It is also hard for the producers to ensure quality and quantity of supply, as credit is limited and input and environmental conditions vary (Ziervogel et al, 2006). A communal farming project was initiated in Mangondi village in 1993 that aimed to support women in the production of vegetables to combat malnutrition among children (Archer, 2003). In the first year, subsistence crops were planted; in later years, vegetables were planted with the intention of selling them. Through the years, the success of the project has fluctuated. In some years, the farmers have had a functional irrigation scheme and money for inputs and have made profits. In other years, the irrigation pump has failed, people have not planted, and harvest and marketing has been poor. Research in this village has been undertaken since 1999 to examine the role of seasonal forecasts and agricultural support among smallholder farmers. Surveys, participatory approaches and computer-based knowledge elicitation tools have been used to determine the types of adaptive strategies followed under certain conditions and why, with a focus on the role of climate, market and livelihood needs (Bharwani et al, 2005). This is supported by the knowledge of household livelihood profiles, local responses to climate, availability of forecast information over time and access to information sources such as market demand and technical knowledge (Ziervogel et al, 2006).
3243 J&J Vulnerability
1/11/07
10:23 am
Page 180
180 Climate Change and Vulnerability
Food security This research has focused on participants involved in the communal garden scheme. The available data, therefore, do not represent the entire village. However, extensive work has been done in the area that enables a picture of food security to be painted. The members involved in the scheme have identified an increase in food security since the project started. Participants have stated that they now have vegetables to eat, which they did not have before there was an irrigation project, and that, as a result, health has improved and vegetables can be sold and the money used to send children to school. When there are surplus vegetables, they dry them so they can be used in months when vegetables are not readily available. The availability of vegetables has also impacted on the quality of livelihoods, as participants now spend less time travelling to nearby markets to obtain them. The disadvantage has been that the irrigation project relies on a pump that is often broken, primarily because it is too small for the garden area (Archer 2003). This has meant that time and resources have been invested in the garden area, and when the pump is broken these resources are wasted. It is also uncertain when the pump will be fixed, which makes it hard to plan and hard to market crops. The limitations of the pump can therefore make people more vulnerable when they are expecting and investing in a harvest that does not materialize. The food security of members not involved in the project depends on individual situations. The land is relatively fertile and there are numerous fruit trees. Many households have some livestock (such as cattle or chickens). The field crops can provide food when rains are good, but it is not uncommon for whole crops to fail when the rain is insufficient and irregular. Farmers who produce surplus are able to sell some of their vegetables, but marketing is a key constraint, and there are not large consistent markets. There are many households in the area in which members are sick, and households struggling to survive as they do not have access to labour for production or access to alternative employment. In South Africa, there is a grant system that supports many households and enables them to buy food. Grants are available in the form of pensions, child grants (for each child up to eight years old if household income is below a certain amount) and disability grants. Disability grants are available for the physically disabled, which includes people who are unable to work because of AIDS. The increased occurrence of HIV/AIDS has an impact on food security, as increased amounts of food and appropriate nutrition are needed, while labour decreases and resources are spent on health rather than agricultural production.
Case Study 2: Gireigikh Rural Council of Bara Province, North Kordofan State, Sudan The Gireigikh area lies in North Kordofan State in drought-prone western Sudan, typified by the semi-arid and desert scrub of the African Sahel region. The area is characterized by harsh climatic conditions and erratic seasonal rainfall. The predominant socioeconomic grouping consists of a mix of
3243 J&J Vulnerability
1/11/07
10:23 am
Page 181
Household Food Security and Climate Change 181
agro-pastoralists and transhumants, who are extremely vulnerable to drought. The key ongoing pressures are the degraded rangelands and strong sand dust storms in the region. The current vulnerabilities result from changes in climate variability, particularly aggravated by the long-term and intense droughts. Hoelzmann et al (1998) found that much of what is desert in Africa today used to be covered in steppe vegetation, and that many small lakes and streams existed above a latitude of 23°N, where they currently do not exist. This desertification is combined with problems of soil erosion and failing livestock and crop production. These factors directly affect the food security situation, which leads to loss of rural livelihoods and displacement of the rural people. Food security The Community-Based Rangeland Rehabilitation Project was implemented in Bara Province of North Kordofan State, a semi-arid land that receives a longterm average rainfall of 275mm annually (Dougherty et al, 2001). This amount used to be sufficient to maintain people’s livelihood and establish their staple food crops, which are mainly millet and sorghum, in addition to raising animals. However, the area was affected by drought episodes three times between 1976 and 1992, with the most severe drought occurring in 1984. The drought of 1980–1984 highlighted the basic problems that have been ignored for too long: family and tribal structures and their autonomous traditional practices of resource management and land tenure had broken down (Dougherty et al, 2001). Two tribes inhabit the study area: the Gawama’a and the Kawahla. The Kawahla are a nomadic tribe that settled in Gireigikh after they lost their herds due to a drought that hit their previous areas of settlement in Eastern Kordofan during the period 1967–1973. The Gawama’a were originally farmers and herders of cattle, sheep, camels and goats. After continuous drought cycles hit the area, the Gawama’a lost all their cattle and most of their sheep, camels and goats and were forced to shift from keeping livestock to farming crops. In the semi-arid Gireigikh area the majority of the interviewed farmers indicated that production declined and vast tracts of land became completely barren with no trees and all the below-canopy herbaceous species removed. The amount that was produced was not enough to sustain their food requirement for the whole year, and they often ended up with severe food shortages. To make up for the meagre income made from this kind of agricultural practice, young men had to travel, leaving the women behind, to seek jobs in nearby towns or in the capital city of Khartoum. With the droughts, water quantity became a limiting factor for the people to practise any alternative livelihood activities, for example, growing vegetables or fruits or raising poultry. Women were known to walk long distances to fetch meagre quantities of water from a hand-dug well, which was both time and health consuming. To address these problems, a project titled ‘CommunityBased Rangeland Rehabilitation for Carbon Sequestration and Biodiversity Conservation in North Kordofan State’ was initiated by the United Nations Development Programme (UNDP) Global Environment Facility (GEF) during the period 1994–2000. The project objectives were 1) to sequester carbon through the implementation of a sustainable, local-level natural resources
3243 J&J Vulnerability
1/11/07
10:23 am
Page 182
182 Climate Change and Vulnerability
management system that prevents degradation of, rehabilitates or improves rangelands, and 2) to reduce the risks of production failure in the droughtprone area by providing alternatives for sustainable production, increasing the number of livelihood alternatives so that out-migration would decrease and population would stabilize. The reported outcomes from the Bara case study (Osman-Elasha, 2006) include: • • • • • • • •
the establishment of local institutions, such as the village community development committees, to coordinate community natural resource management and community development activities; the development of land-use master plans to guide future resource use and implementation of sustainable rotational grazing systems and establishment of community mobilization teams to conduct outreach and training; the reforestation and stabilization of 5km of sand dunes to halt desert encroachment and soil erosion; the restocking of livestock by replacing goat herds with more resilient and less resource-damaging sheep; the creation of a water management subcommittee to better manage wells; the establishment of 17 women’s gardens to produce vegetables for household consumption, with surplus sold at local markets; the establishment of five pastoral women’s groups to support supplemental income-generating activities (including sheep fattening, handcrafts, milk marketing); and the planting of 195km shelterbelts around 130 farms to act as windbreaks, to improve soil moisture and to increase fertility (Dougherty et al, 2001).
Figure 9.1 shows the profound change from the poor financial conditions that used to prevail before the project to the much improved situation after the intervention (up to 80 per cent level of financial sustainability was observed as compared to the less than 10 per cent level of financial sustainability before). Information on indicators such as effectiveness of credit repayment revealed that the money from the revolving fund was primarily used for buying animals for fattening and marketing them in large state markets, hence fetching higher prices. Moreover, it offers necessary funds for the Natural Resources Committee to support other activities such as seed distribution for women’s irrigated gardens and the purchase of improved stoves. This has led to the conclusion that the improvement of financial capital could contribute to the improvement of other livelihood capitals, such as natural, human and physical. Moreover, the presence of reliable local-level institutional structures (for example, the Sudanese Environment Conservation Society (SECS) Bara Branch, along with the Community Credit Committees and Coordination Committee) has provided a guarantee to the local community that has enabled their access to credits.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 183
Household Food Security and Climate Change 183
Figure 9.1 Assessment of sustainability of financial capital before and after intervention of the Rangeland Rehabilitation Project, based on availability of information, effectiveness of credit repayment, suitability of local institutions, and support of credit systems and government policy to income-generating activities
Case Study 3: Chingowa village, Magumeri Local Government Area, Borno State, Nigeria Chingowa village is located in the Magumeri Local Government Area of Borno State in the Sahel zone of northeastern Nigeria. It has a mean annual rainfall of 600mm, which falls in the four months of the rainy season, which lasts between June and September. Agriculture, which is rain-fed, is primarily supported during the short rainy season, except around the oasis, which supports perennial vegetable farming. The vegetation consists mainly of shrub grassland, which is favourable for extensive grazing. Drought has been a recurrent feature in the Sahel, with records of drought dating back to the 1680s. The magnitude and intensity of these droughts have been on the increase over the last 100 years, however, and consequently so has the resulting destruction (Hulme et al, 2001). The lack of water, in association with high temperatures (up to 45°C at certain periods of the year), is the most limiting factor for agricultural productivity in the village. The main crops cultivated in the village include maize, millet, sorghum and beans. Farmers are predominantly smallholders using traditional farming systems, which mix food crops and cash crops on the same farming unit. Because crop farming is largely rain-fed in the village, the increase in the magnitude and intensity of droughts will result in a shift from the production of certain traditional crops, with all the possible negative consequences that this may bring to the local people. The rearing of livestock is a very important aspect of life in the village, as it represents livelihood, income and employment. Recurrent droughts have forced some pastoralists to dispose of their cattle and lose their livelihood sys-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 184
184 Climate Change and Vulnerability
tems, which ultimately increases their vulnerability. The southward movement of the isohyets has also resulted in the southward migration of pastoralists into lands formerly occupied by sedentary farmers. This has been a major source of conflict in the village, leading to widespread destruction of farmlands and cattle, with adverse implications for food security. The methodology used for this case study included data collection through the administration of questionnaires, focused group discussions, stakeholder analyses and field sampling. The case study looks at vulnerability to food insecurity in Chingowa village in Borno State of Nigeria, where questionnaires were administered to 30 farm households in the village. The research shows that one of the main concerns among the respondents is the fear of famine, which highlights the problem of food insecurity. Food security In assessing food insecurity in the village, various factors that predispose households to being vulnerable were examined. These factors include agricultural productivity and production, labour availability and land tenure, food storage and processing, transportation and distribution, population factors, income and conflicts. The size of the landholding that a household cultivates directly affects their production and hence food security. Population growth has led to a high level of fragmentation of land in the village. Hence, acquiring a relatively large tract or tracts of land for farming is a difficult task. A majority of the farmers in the village (61.2 per cent) are smallholders who cultivate less than 5 hectares of land. Only about 19 per cent of the respondents used irrigation, with just less than a third of them irrigating more than half of their total farm lands. The existing tenurial arrangements in the village also affect agricultural production in the village and pose a constraint on sustainable food security. There are sociocultural factors that prevent women from having title to land in many parts of the country, including Chingowa village. In Chingowa, ownership of farmland is predominantly acquired through paternal inheritance, which, to a large extent, excludes the women. However, some respondents cultivate lands that are communally owned, purchased outright, rented or borrowed. Other factors that affect agricultural productivity in the village include the unavailability of inputs, particularly fertilizer and improved seeds, and poor transportation infrastructure, which has adversely affected market access. Chingowa village is situated about 1.5km from a good road and 6km from a daily market. With low agricultural productivity in the village, the population has had to continue to rely on other sources of income in order to meet household demand for food and other needs. In addition to arable farming, pastoralism is a major economic activity in the village. The most common domesticated animals in the village are cattle, poultry, goats and sheep. Labour is a critical input in the traditional, subsistence farming system practised in the village. The farmers plant very small areas at a time, using crude implements and labour-intensive practices. As a result, the demand for labour is generally very high at the time of planting, weeding and harvesting.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 185
Household Food Security and Climate Change 185
Two main factors that affect the availability of labour in the household are rural–urban migration and the quality of the household, conceptualized as the ratio of healthy working members of the household over the sick members, as captured in the dependency ratio. With the spread of killer diseases such as malaria, HIV/AIDS and cerebrospinal meningitis, the quality of available household labour is seriously compromised. The problem of inadequate storage facilities has compounded the problem of food security in the village. It is estimated that about 15–20 per cent of cereals and up to 40 per cent of perishable crops produced are lost before they can be consumed. This situation is made worse by the dearth of any agro-processing industry close to the village. It also has a discouraging effect on the farmers as the struggle to sell most of their crops immediately after harvest results in very unprofitable competition and lower prices. The village has witnessed several communal crises, largely between the pastoralists and the sedentary farmers. These conflicts have largely arisen through the struggle for resources, which has been exacerbated by the frequent droughts and the downward shifts of the isohyets, and are a major constraint to food security in the village. The crises usually occur during the planting, weeding or harvesting periods, and with the flight of farmers from the area, irrespective of the stage of farming, food security is threatened as most, if not all, the crops are lost. The pastoralists also suffer significant losses of livestock.
Case Study 4: Rain-fed maize production in Tlaxcala, Mexico Rain-fed maize production is the most important agricultural activity for the majority of subsistence farmers in Mexico. The maize is traditionally cultivated on a surface called milpa, which includes other cultivars (such as beans and chillies) and plants used for medical and food preparation purposes. This activity is strongly affected by climate variability, particularly drought events, which have forced farmers to apply different coping strategies (Florescano, 1995). In Mexican history, hunger and famine have been common and are related to severe drought events, in which great losses in maize production have affected both rural and urban populations. The impacts of these events have been exacerbated during the most important civil wars in the country (the War of Independence and the Mexican Revolution). Nowadays, farmers who rely on rain-fed crops apply different strategies to cope with drought, including switching to more pest-resistant maize varieties, changing cultivars, seeking temporary jobs in urban areas, renting their fields, or even emigrating to the capital city of the state, cities in other states or the US. In 2004, 10.23 million Mexican migrants lived in the US, a population that has grown at an annual rate of 4.2 per cent since 1994 (CONAPO, 2004). In the same year, 16,613 million US dollars were sent by migrants to Mexico (Banco de Mexico, 2004). This flux of income has represented a basic support to preserve or enhance the levels of nutrition, health and education of the rural population, particularly for rural villages with less than 2500 inhabitants (Lozano, 2005). This situation was mainly forced by the aggressive changes in governmental policies related to the economic liberalization of agriculture, particularly since
3243 J&J Vulnerability
1/11/07
10:23 am
Page 186
186 Climate Change and Vulnerability
the North American Free Trade Agreement (NAFTA) came into effect in 1994 (Nadal, 2000). In the last decade, basic grain imports have increased by almost 40 per cent and maize imports have doubled, even though in NAFTA it was established that the total liberation of maize (and beans) imports will occur by 2008 (Bartra 2003). In general, the structural reforms in the agricultural sector have implied the removal of subsidies for seeds, agrochemicals, energy and water, the reduction of credits, and the elimination of the governmental control of prices. All of these measures have caused an increase in production costs and a reduction of profits for maize producers. The case of maize production in Tlaxcala, Mexico, illustrates how climate impacts on food security. The study builds on climate change and climate variability studies that have been carried out in this state since 1997 (Ferrer, 1999; Conde and Eakin, 2003). The state of Tlaxcala is located in the centre of the country, in the Mexican high plateau. It is the smallest state in the country, and 98 per cent of agriculture is developed under rain-fed conditions (INEGI, 1996) There is only one harvest a year (Trautmann 1991), and maize is the most important crop (71 per cent of the total planted area). During spring, farmers in Tlaxcala wait for the onset of the rainy season by April; should there be a delay in this, they start considering changing varieties of maize or even changing to another crop. This situation could be clearly observed during the strong 1997–1998 El Niño event, when oats (for fodder) were planted. Oats are locally known as the ‘hopeless crop’ and were planted as a desperate measure just to prevent cattle losses, with the expectation of maize planting being abandoned (Aviles, 2005). When normal climatic conditions returned, farmers reverted to planting maize (Conde and Eakin, 2003), because alternative crops are planted only during extreme climatic events, and not as a rule to adapt to adverse market conditions. Since maize prices have significantly decreased, the cost of fertilizers and other inputs have increased, and there is a lack of labour force in terms of younger farmers, a drought event can severely affect the capacity of the farming community to cope with its impacts. Furthermore, agricultural policies and supports are more centred towards the production of fruits and vegetables for export, not taking into account the high consumption of water of those products in a context of climate variability and change, associated with past and future droughts in the country. Other environmental factors raise the risk of increasing losses in maize production. Tlaxcala is the state with the worst soil erosion conditions in the country (SEMARNAT, 1996). Even when farmers are aware of soil conservation techniques (Conde and Eakin, 2003), they cannot practise them, because those require strong collective work, and families are being reduced in number because of migration or changed labour, so they are forced to develop maize monoculture production in extended areas, which reduces soil productivity and increases the soil erosion processes. Furthermore, the reduction of crop diversity (Altieri and Trujillo, 1987) increases climatic risks and reduces the farmers’ nutritional opportunities.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 187
Household Food Security and Climate Change 187
Food security The agricultural policies developed over the last 20 years in Mexico related to food security have shifted from a view of self-sufficiency agricultural production to a policy that seeks to secure access to food resources, following the globalization of economic processes. Government support for the rural population now focuses on programmes that deliver economic help for those needing to acquire basic goods (for example, tortillas or milk) at reduced prices, not on a policy to sustain the traditional maize production (milpas). The current government has declared that subsistence farmers have five years to be ‘efficient and competitive’ in the international markets, or to ‘look for another activity’ (Bartra, 2003). Besides the impacts of changes in Government policy, the NAFTA provisions have also established that until 2008 the Mexican market will be totally open to corn importations. Maize and tortilla prices have therefore been adjusted to the requirements of international markets, which has resulted in a decrease in national maize prices and an increase in tortilla prices. Before 1998, tortilla prices were subsidized by the government in a policy that focused on guaranteeing food security for the increasing urban population (Appendini, 2001). The costs of production were mainly transferred to consumers, but the benefits of the increase in prices of tortilla did not reach the maize producers. The massive imports of corn from the US at lower prices than maize produced in the country (Bartra, 2003) reduced the possible profits for Mexican maize producers. Traditional production of tortillas has also been mostly abandoned, and the corn flour production has been controlled since the 1990s by huge companies such as Grupo Maseca (Rosas-Peña, 2005). These companies are free to import forage corn from the US and use it for processing tortillas for Mexico City, for example. Food insecurity for maize farmers in Mexico, particularly in Tlaxcala, is therefore increasing, due to the described policies and other factors such as the decrease in human resources (migration and farmers’ aging) and in environmental resources. The financial support given by the governmental programmes to maize farmers that have suffered the impacts of climatological contingencies, which are delivered late and in reduced amounts (Cortés, 2004), does not solve the decreasing productivity and the lack of a market for their products. These governmental and macroeconomic policies and trends tend to reduce farmers’ ability to manage with adverse climatic events. In this context, new threats, such as transgenic maize or climate change, will be difficult to cope with. However, farmers with rain-fed land continue to plant maize as a means to subsist, making total corn production in the country more or less stable, but increasing planting areas and impacting soil and forest ecosystems. Several authors have considered that a national food security policy should be linked to environmental sustainability and social equity (Appendini, 2001), particularly related to the social right of rural communities to work (Bartra, 2003) and of consumers to chose the quality of the tortillas they eat.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 188
188 Climate Change and Vulnerability
Analysis and Discussion These four case studies both exhibit a number of similar trends and demonstrate distinct differences. In the South African case study, it is evident that there is a clear tension between natural environmental stress of precipitation variability and the social system that determines the welfare options and employment options available. If access to money through employment or grants is available, then food can be acquired through exchange, but those depending on household production are directly exposed to climatic variability, as well as market fluctuations. It is important to recognize that those households relying on grants or income might still be impacted through secondary-order sensitivities, where jobs may decline due to climate variability, for example, decreased employment on commercial farms when harvests have failed. Research has shown that the poorer households tend to respond to climate variability in multiple, low-input strategies, whereas the better-off households focus on a few strategies, often with higher risk (Bharwani et al, 2005). Fieldwork has supported the idea that food security interventions need to be sensitive to user characteristics and strategies at the same time as supporting institutional developments that enable vulnerable households to cope with loss of livelihood options through a range of stresses, including climate variability, unemployment or ill health. Although drought is often regarded as the major cause of food insecurity in semi-arid northern Nigeria, the study found that household factors contributed more to food insecurity than climate factors. These household factors include the size of landholding available to the household for cultivation and the labour available in the household to cultivate the land. The farmers are largely subsistence farmers using crude farming implements with little farm inputs such as fertilizers, irrigation, insecticides, pesticides or improved crop varieties. Where crop production is largely dependent on the size of a household’s landholding, a common strategy for increasing production is through land intensification. This usually results in the cultivation of marginal lands that have hitherto been occupied by pastoralists. This encroachment often results in conflicts and destruction of cattle and crops, further reinforcing food insecurity in the region. Belonging to a community organization was another major factor in food security in the region. There is no organized welfare system in Nigeria, and these local community organizations provide safety nets to their members in times of crisis. The study estimated that about 15–20 per cent of cereals and up to 40 per cent of perishable crops produced are lost before they can be consumed. Food storage was thus an important issue in food security. Minimizing this level of waste would make more food available for consumption. Food insecurity in northern Nigeria is also a ‘food access problem’. This could be linked to poor governance, whereby infrastructure such as roads are concentrated in the urban centres with very little being provided in the rural areas. This makes it difficult for farmers to transport their goods from the farms to the markets. Besides the general lack of roads and transportation services, the high cost of fuel adds to the cost of transportation, which makes the final prices
3243 J&J Vulnerability
1/11/07
10:23 am
Page 189
Household Food Security and Climate Change 189
of the goods very expensive and beyond the reach of the majority of the rural poor. It could also be linked to poverty and the inability of poor people to access food and other resources. Over the longer term, poverty contributes further to food insecurity as it restrains households’ potential for accumulation and growth. In view of this, it is important that government seeks to provide physical infrastructure in the rural areas, as well as provide seasonal input credit and long-term financing of farm investments. In Sudan and Mexico, it was also clear that climate variability alone does not determine vulnerability to climate change. Rather, it is the livelihood characteristics overlain with social and economic environments and climate variability that determine the vulnerability of households to food insecurity and climate change. In order to assess the similarities across the cases more systematically, the key vulnerabilities in each case study have been compared by assessing the role of determinants in five groups: climate, environment, food economy, household factors, and social and human environment. The determinants were scored on a scale ranging from 0 to 2 (where 0 indicates that the factor does not appear to be a key determinant of vulnerability, 1 suggests it is an important determinant and 2 that it is very important). These scores were derived from the expert judgement of the authors, based on their interpretation of stakeholder perspectives. This method has its limitations, as it does not incorporate stakeholder feedback, but it provides an initial analysis of where similarities and differences between the cases lie. It is important to note that this comparison is based on the four case study sites – Mangondi village in South Africa, Gireigikh Rural Council in Sudan, Chingowa village in Nigeria and Tlaxcala in Mexico – rather than on the countries as a whole. Table 9.1 Summary of the importance of determinants Ranked as very important by 3 or more case studies
Ranked as very important by 2 case studies
Ranked as a combination of very important and important by all case studies
Ranked as important by all case studies
Trends in precipitation Input price Income diversification
Recent drought Area cultivated Labour available per hectare Off-farm employment
2-year/seasonal drought 3–5 year ENSO 5–10 year drought
Ability to subsist Household size Health
Land degradation
Poor training and education Poor nutrition and human health
Income Local community institutions Disintegration of social fabric
Poor health services
Note: Very important = score of 2; and Important = score of 1
A summary of the key determinants is presented in Table 9.1. It is clear from the comparison that there are many common factors that influence household
3243 J&J Vulnerability
1/11/07
10:23 am
Page 190
190 Climate Change and Vulnerability
food security. The most important determinants of household vulnerability to food insecurity across all four case studies are trends in precipitation, input price, household income, income diversification, belonging to local community institutions and disintegration of social fabric. The next most important determinants include the occurrence of recent droughts, the size of landholding cultivated, the labour per hectare of farm land, poor health services and participation in off-farm employment. It is clear that there is a mix of physical, social and economic factors that determine vulnerability to food insecurity in all of the case studies. Given that the common vulnerabilities have been established, an objective way of comparing the strengths of the determinants across the various study sites is needed. The key determinants for five categories (climate, environmental, food economy, household factors, and social and human environment), as mentioned above, were ranked as to their role in determining vulnerability in each case study. The mean score for each group of determinants is plotted on a radar graph for each case study (Figure 9.2). This illustrates that climate factors played a similar role in determining vulnerability to food insecurity in Mangondi village, Gireigikh Rural Council and Chingowa village. Climate did not appear to be a major factor in Tlaxcala. Generally, household factors played a more dominant role among all four countries and appeared to have the largest influence on vulnerability to food insecurity. This is a significant finding that should be explored further.
Figure 9.2 Determinants of vulnerability to food insecurity in study villages Another way to compare the relative importance of the factors affecting food security is to place each factor in a conceptual framework. For the purposes of this synthesis, we chose a straightforward set of factors that link the underlying use of resources, exposure to drought and the consequences of food shortage
3243 J&J Vulnerability
1/11/07
10:23 am
Page 191
Household Food Security and Climate Change 191
(Figure 9.3). The focus on drought is justified, as this is the major climatic factor affecting food security in these four case study areas. Of course, drought alone is not a sufficient cause of food insecurity, and the framework attempts to place drought in context.
Human Needs: Nutrition
Human Wants:
Choice of Means:
Initiating Events:
Intermediate Events:
Outcome:
Exposure:
Consequence:
Consequence:
Household food scarcity
Hunger of household members
Morbidity, Loss of livelihood
Death
Dietary preference
Cropping system
Drought
Crop failure
Modify Wants:
Modify Means:
Cope with Event:
Cope with Event:
Cope with Outcome:
Block Exposure:
Mitigate Consequence:
Mitigate Consequence:
Alter choice of foods
Choose drought crops
Irrigate
Replant
Sell assets, buy food
Migrate to find food
Reduce activity
Emergency relief, recovery, rehabilitation
Figure 9.3 Causal chain of drought risk The upstream context begins with the identification of human needs (nutrition) and wants (the choice of diet to fulfil nutritional needs), along with the choice of cropping systems (or food procurement systems, more generally) to fulfil the dietary preferences. The hazard-sensitivity elements include the initiating events (drought, or a combination of drought and other factors) and the first-order impacts (such as crop failure), leading to initial outcomes, including household food scarcity. Differential vulnerability is apparent in the range of exposures to the first-order impacts and sensitivity to the consequences (from increased disease burden to death, plus environmental, social, economic and political consequences). At each stage, a range of actions can intervene to disrupt the causal chain (in other words to prevent further impacts and consequences) or to shift the chain of events to other pathways (for instance, to shift household food scarcity to regional markets, leading to increased food prices and imports to the region). This framework is used to map the determinants from the case studies against the causal chain as shown in Table 9.2. The most striking similarity, based on the interpretations of the expert teams, is in the initiating events – the importance of trends in decreasing precipitation, generally accompanied by drought. Health status and health services are seen as the major factors influencing the outcomes of climatic stress. The range of factors under the categories of structural vulnerability and impact sensitivity are similar, but often with different degrees of importance. For example, off-farm employment is a major factor for the wealthier two countries, South Africa and Mexico. These two countries show subsidies, pensions and welfare systems as the most important structural factors, but these are not as prominent in Sudan and Nigeria. This could be due to the fact that a healthier economy leads to more job opportunities.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 192
192 Climate Change and Vulnerability
Table 9.2 Determinants of vulnerability situated in a causal chain of drought risk Structural vulnerability: Needs Wants Means
Initiating events: Climatic Economic Environmental
1st order sensitivity: Impacts Exposure
Consequences: Nutrition Health Livelihood Death
Sudan
Water harvesting Land degradation Deforestation Land pressure Pests & disease Market access Storage Market prices Welfare
Precipitation Drought ENSO Early warning
Area cultivated Income/ diversification Training/ education Community institutions Off-farm employment Size of holding Household size
Health/health services
Nigeria
Land degradation Deforestation Land pressure Pests & disease Storage Market access Market prices
Precipitation Drought ENSO Floods Early warning
Area cultivated Labour per ha Income/ diversification Off-farm employment Community institutions Training/education
Health/health services
South Africa
Land degradation Water harvesting Storage Welfare Market access Market prices
Precipitation Drought ENSO Heat waves Floods Early warning
Area cultivated Off-farm employment Income/ diversification Community institutions
Health/health services
Mexico
Land degradation Deforestation Pests Market access Market prices Welfare
Precipitation Drought ENSO Heat waves Floods Early warning
Area cultivated Labour per ha Off-farm employment Income/ diversification Training/education Community institutions
Health/health services
Key: Bold = very important (2); regular font = important (1) from Table 9.1.
Conclusion Vulnerability to food insecurity is common across the world in semi-arid areas where marginal groups rely on rain-fed agriculture. This chapter has started to compare some of the dynamics associated with these commonalities. This is particularly important because it is well established that food insecurity is not solely about climate, but about a range of social, economic and political factors that are linked to physical factors. At the same time, the shift in climate patterns associated with climate change requires an understanding of how climate
3243 J&J Vulnerability
1/11/07
10:23 am
Page 193
Household Food Security and Climate Change 193
variability has an impact on food security in conjunction with other determinants. The causal chain of drought risk helps to highlight the process of becoming vulnerable to drought. In most of the case studies, the determinants of vulnerability are spread throughout the continuum, indicating that there are multiple ways to modify and change the risks. This highlights the need to understand the problem so that interventions can be appropriate in nature and timing. If one looks at the four case studies, it can be seen that household factors played a dominant role in determining vulnerability. Although this is not unexpected, it suggests that there needs to be a continued emphasis on the multidimensional integrated approach to assessing vulnerability to climate variability. This needs to be followed through when responding to climate variability, whether through climate adaptation options or through development policies to support drought-affected households. Another determinant that cut across all cases was the health status and health services that households have access to. Health stress is related to climate variability but can also be seen as a basic service and need that should be addressed to reduce vulnerability at multiple stages in the chain of drought risk. Off-farm income is important for case studies located in the two countries that are relatively wealthy, Mexico and South Africa. This highlights the differences that national-level policies might have on local impacts of drought on food security. In many places, the term ‘food security’ is still equated with ‘food availability’. The result is that government strategies to address food security, such as strategic grain reserve programmes and various agricultural development strategies, end up addressing only availability. These do not achieve the desired goal of improving food security, however, other key determinants that impact directly on that goal have not been integrated by government into their policies. There is the need to develop effective long-term agricultural policies that are situated within a wider development framework. For example: 1 2 3 4
productive commercially-oriented smallholder farming systems that employ cheaper means of enhancing farm productivity could be promoted; irrigation development for drought-mitigation strategies and sustainable food production could be encouraged; barriers to land ownership and secure tenure could be addressed; and the capacity of farmers and rural institutions to continue to provide safety nets in times of food crisis could be better supported.
The implications of this study for climate policy as related to food security are clear. The impacts of climate change on food security cannot be seen solely as food production issues. Food security depends on livelihood security that, in turn, depends on many factors, including social, economic and environmental determinants. The second key policy issue is that understanding the context is of paramount importance. Depending on the local and national situation, certain institutions support access to, availability and utilization of food. It is difficult to generalize about coping strategies in response to stress. Support for adaptation
3243 J&J Vulnerability
1/11/07
10:23 am
Page 194
194 Climate Change and Vulnerability
measures therefore needs to be grounded in the local context. What might be effective and contribute to improving food security in South Africa might be ineffective in Sudan and end up increasing the vulnerability of marginal groups. It is therefore critical to verify and screen adaptation options and support. This chapter has highlighted that there are commonalities and differences in understanding food security in the light of climate extremes such as drought. In cases in which there are commonalities, more could be done to look at how other countries have managed both the response to drought and the efforts to reduce the impact of drought. A potential increase in drought frequency and increased temperatures requires that understanding these processes of risk is a priority in order to respond appropriately with support for the most vulnerable groups.
Notes 1
For further details on the four case studies in this chapter, see Chapters 10, 11, 12, 13 and 14 in this volume.
References Altieri, M. A. and J. Trujillo (1987) ‘The agroecology of corn production in Tlaxcala, México’, Human Ecology, vol 15, pp189–220 Appendini, K. 2001. De la Milpa a los Tortibonos: La restructuración de la política alimentaria en México, 2nd edn, Instituto de Investigaciones de las Naciones Unidas para el Desarrollo Social, El Colegio de México, Mexico Archer, E. M. (2003) ‘Identifying underserved end-user groups in the provision of climate information’, Bulletin of American Meteorological Society, vol 84, pp1525–1532 Arnell, N. W., M. G. R. Cannell, M. Hulme, R. S. Kovats, J. F. B. Michell, R. J. Nicholls, M. Parry, M. J. Livermore and A. White (2002) ‘The consequences of CO2 stabilisation for the impacts of climate change’, Climatic Change, vol 53, pp413–446 Aviles, K. (2005) ‘Ofensiva neoliberal crea pueblos fantasmas en zonas agrícolas’, La Jornada, vol 7 BancodeMexico (2004) ‘Informe Anual 2004’, www.banxico.org.mx Bartra, A. (2003) Del Teocintle a los Corn Pops: Sin maíz no hay país, Conaculta, Museo Nacional de Culturas Populares, Mexico, pp219–250 Bharwani, S., M. Bithell, T. E. Downing, M. New, R. Washington and G. Ziervogel (2005) ‘Multi-agent modelling of climate outlooks and food security on a community garden scheme in Limpopo, South Africa’, Philosophical Transactions of the Royal Society: B, vol 360, pp2183–2194 Bryant, C. R., B. Smit, M. Brklacich, T. Johnston, J. Smithers, Q. Chiotti and B. Singh (2000) ‘Adaptation in Canadian agriculture to climatic variability and change’, Climatic Change, vol 45, pp181–201 Carney, D. (ed) (1998) Sustainable Rural Livelihoods: What Contributions Can We Make? Department for International Development (DFID) Natural Resources Advisers’ Conference (July 1998), DFID, London Chambers, R. (1989) ‘Vulnerability, coping and policy’, IDS Bulletin, vol 20, pp1–7 Clover, J. (2003) ‘Food security in sub-Saharan Africa’, African Security Review, vol 12, pp1–11
3243 J&J Vulnerability
1/11/07
10:23 am
Page 195
Household Food Security and Climate Change 195 CONAPO (2004) ‘Población nacida en México que reside en Estados Unidos: 1990 a 2004’, www.conapo.org.mx Conde, C. and H. Eakin (2003) ‘Adaptation to climatic variability and change in Tlaxcala, Mexico’, in R. K. J. Smith and S. Huq (eds) Climate Change, Adaptive Capacity and Development, Imperial College Press, London Cortés, S. (2004) ‘Criterio 2: Beneficios del Programa’, Evaluación Externa 2003 al Fondo para Atender a la Población Rural Afectada por Contingencias Climatológicas (FAPRACC), Centro de Ciencias de la Atmósfera, UNAM, Mexico Devereux, S. and J. Edwards (2004) ‘Climate change and food security’, IDS Bulletin, vol 35, no 3, pp22–30 Devereux, S. and S. Maxwell (eds) (2001) Food Security in Sub-Saharan Africa, Intermediate Technology Development Group Publishing, London, UK Dixon, J., A. Gulliver and D. Gibbon (2001) Farming Systems and Poverty: Improving Farmers’ Livelihoods in a Changing World, Food and Agriculture Organization and World Bank, Washington, DC, US Döös, B. R. and R. Shaw (1999) ‘Can we predict the future food production? A sensitivity analysis’, Global Environmental Change, vol 9, pp261–283 Dougherty, B., A. Abusuwar and K. A. Razig (2001) Sudan Community-Based Rangeland Rehabilitation for Carbon Sequestration and Biodiversity, Terminal Evaluation Report, SUD/93/G31.UNDP GEF Downing, T. E. (2002) ‘Linking sustainable livelihoods and global climate change in vulnerable food systems’, Die Erde, vol 133, pp363–378 FAO (2000) Guidelines for National FIVIMS: Background and Principles, Food and Agriculture Organization of the United Nations, Rome Ferrer, R. M. (1999) ‘Impactos del cambio climático en la agricultura tradicional de Apizaco, Tlaxcala’, thesis, Facultad de Ciencias, Universidad Nacional Autónoma de Mexico, Mexico City, Mexico Florescano, E. and S. Swan (1995) Breve Historia de la Sequía en México, Biblioteca Universidad Veracruzana, Veracruz, Mexico Gregory, P. J., J. S. I. Ingram and M. Brklacich (2005) ‘Climate change and food security’, Philosophical Transactions of the Royal Society: B, vol 360, pp2139–2148 Hoelzmann, P., D. Jolly, S. P. Harrison, F. Laarif, R. Bonnefille and H. J. Pachur (1998) ‘Mid-Holocene land-surface conditions in northern Africa and the Arabian peninsula: A data set for the analysis of biogeophysical feedbacks in the climate system’, Global Biogeochemical Cycles, vol 12, pp35–51, www.ncdc.noaa.gov/paleo/ abrupt/references.html#hoelzmann1998 Hoerling, M. P., J. W. Hurrell, J. Eischeid and A. Phillips (2006) ‘Detection and attribution of 20th century northern and southern African monsoon change’, Journal of Climate, vol 19, no 16, pp3989–4008 Hulme, M. (2001) ‘Climatic perspectives on Sahelian desiccation: 1973–1998’, Global Environmental Change, vol 11, pp19–29 INEGI (Instituto Nacional de Estadistica, Geografia e Informatica) (1996) Sonora. Datos por Ejido y Comunidad Agraria. XI Censo General de Poblacion y Vivienda. Aguascalientes, Ags. INEGI, Hermosillo Lozano, F. and F. Olivera (2005) ‘Impacto económico de las remesas en México: un balance necesario’, presented at Problemas y Desafíos de la Migración y el Desarrollo en América seminar, Cuernavaca, Mexico, April 7–9, 2005, available at www.migracionydesarrollo.org.mx McCarthy, J. J., O. F. Canziani, N. A. Leary, D. J. Dokken, K. S. White (eds) (2001) Climate Change 2001: Impacts, Adaptation, and Vulnerability, Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK McConnell, W. J. and E. F. Moran (eds) (2000) Meeting in the Middle: The Challenge of Meso-Level Integration, LUCC Report Series no 5, International Workshop on
3243 J&J Vulnerability
1/11/07
10:23 am
Page 196
196 Climate Change and Vulnerability the Harmonization of Land Use and Land Cover Classification, Ispra, Italy, 17–20 October, Indiana University Press, Bloomington, IN, US Moser, C. O. N. (1998) ‘Reassessing urban poverty reduction strategies: the asset vulnerability framework 1998’, World Development, vol 26, pp1–19 Nadal, A. (2000) ‘The environmental and social impacts of economic liberalization on corn production in Mexico’, study commissioned by Oxfam, GB and WWF International Osman-Elasha, B. (2006) ‘Environmental strategies to increase human resilience to climate change: Lessons for eastern and northern Africa’, Final Report, Project AF14, Assessments of Impacts and Adaptations to Climate Change, International START Secretariat, Washington, DC, US, www.aiaccproject.org Parry, M., C. Rosenzweig, A. Iglesias, G. Fisher and M. Livermore (1999) ‘Climate change and world food security: A new assessment’, Global Environmental Change, vol 9, ppS51–S67 Parry, M. L., C. Rosenzweig, A. Iglesias, M. Livermore and G. Fischer (2004) ‘Effects of climate change on global food production under SRES emissions and socio-economic scenarios’, Global Environmental Change, vol 14, pp53–67 Perry, A. H. and L. J. Symons (1994) ‘The wind hazard in the British Isles and its effects on transportation’, Journal of Transport Geography, vol 2, pp122–130 Reilly, J. (1995) ‘Climate change and global agriculture: Recent findings and issues’, American Journal of Agricultural Economics, vol 77, pp727–733 Rosas-Peña, A. M. (2005) ‘Un mercado hecho bolas’, La Jornada en la Economia, vol 5 Scoones, I. (1998) ‘Sustainable rural livelihoods: A framework for analysis’, Working Paper no 72, Institute of Development Studies, University of Sussex, Brighton, UK: SEMARNAT (1996) ‘Estadísticas Selectas: Agua: Balance de agua superficial y subterránea’, Secretaría de Medio Ambiente y Recursos Naturales, www.semarnap. gob.mx/naturaleza/estadística-am/ Smit, B., E. Harvey and C. Smithers (2000) ‘How is climate change relevant to farmers?’ in D. Scott, B. Jones, J. Audrey, R. Gibson, P. Key, L. Mortsch and K. Warriner (eds) Climate Change Communication: Proceedings of an International Conference, Environment Canada, Kitchener-Waterloo, Canada, ppF3.18–F3.25 Stephen, L. and T. E. Downing (2001) ‘Getting the scale right: A comparison of analytical methods for vulnerability assessment and household-level targeting’, Disasters, vol 25, no 2, pp113–135 Trautmann, W. (1991) ‘Los cultivos indígenas de Tlaxcala y la Mesa Central: Tipología y problemas de su datación’ in Historia y sociedad en Tlaxcala (ed) Memorias del 4º y 5º Simposios Internacionales de Investigaciones Socio-Históricas sobre Tlaxcala, Gobierno del Estado de Tlaxcala, Tlaxcala, Mexico, pp62–65 USAID (United States Agency for International Development) (1992) Policy Determination 19: Definition of Food Security, United States Agency for International Development, Washington, DC, US Vogel, C. and J. Smith (2002) ‘The politics of scarcity: conceptualizing the current food security crisis in southern Africa’, South African Journal of Science, vol 98, pp315–317 Watson, R. T. and the Core Writing Team (eds) (2001) Climate Change 2001: Synthesis Report, Contribution of Working Groups I, II and III to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US Winters, P., R. Murgai, A. de Janvry, E. Sadoulet and G. Frisvold (1999) ‘Climate change and agriculture: effects on developing countries’, in G. Frisvold and B. Kuhn (eds) Global Environmental Change and Agriculture, Edward Elgar Publishers, Cheltenham, UK Ziervogel, G. (2004) ‘Targeting seasonal climate forecasts for integration into household level decisions: The case of smallholder farmers in Lesotho’, The Geographical Journal, vol 170, pp6–21
3243 J&J Vulnerability
1/11/07
10:23 am
Page 197
Household Food Security and Climate Change 197 Ziervogel, G. and R. Calder (2003) ‘Climate variability and rural livelihoods: Assessing the impact of seasonal climate forecasts’, Area, vol 35, pp403–417 Ziervogel, G., S. Bharwani and T. E. Downing (2006) ‘Adapting to climate variability: Pumpkins, people and policy’, Natural Resource Forum, vol 30, pp294–305
3243 J&J Vulnerability
1/11/07
10:23 am
Page 198
10
Vulnerability in Nigeria: A National-level Assessment James O. Adejuwon
Introduction Vulnerability to climate change is the extent to which climate change may damage or harm a system and is a function of a system’s exposure, sensitivity and adaptive capacity (McCarthy et al, 2001). Applying this conception of vulnerability, an index of vulnerability of peasant households is constructed and used to rank the 30 states of the Nigerian Federation from least to most vulnerable. The analysis produces a depiction of the spatial distribution of vulnerability that is a synthesis of several investigations of climate variability, climate change and food security in Nigeria, the details of which are reported in Adejuwon 2005a, 2006a and 2006b. The components of the vulnerability index correspond to the three main aspects of vulnerability: exposure, sensitivity and adaptive capacity. Exposure is the nature and degree to which a system is physically exposed to climatic variations and changes. Sensitivity is the degree to which a system is affected, either adversely or beneficially, directly or indirectly, by climate-related stimuli. Adaptive capacity refers to the ability of a system to adjust to climate change, including variability and extremes, to moderate potential damages, to take advantage of opportunities, and to cope with the consequences. Indicators of each of these three components of vulnerability are selected, secondary statelevel data for the indicators are collected from various sources, and the indicators are combined into an aggregate measure of vulnerability and mapped. In addition, analyses have been made of the contemporary pattern of climate variability in Nigeria, projected future climate change and the impacts of climate change on future crop yields in major ecological zones of Nigeria.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 199
Vulnerability in Nigeria: A National-level Assessment 199
Climate Variability and Climate Change The contemporary regional climate The ecological zones of Nigeria, depicted in Figure 10.1, are determined largely by climatic conditions. The lowest and highest mean monthly minimum and maximum temperatures are presented in Table 10.1 for the major ecological zones. Maximum temperatures during the warmer months generally increase as you travel from south to north, from the forest zone, through the Guinean savannahs and to the savannahs of the Sudan and Sahel zones. But in the coolest month, January, temperatures are higher as you travel south. The observations confirm the well-known fact that diurnal variations in temperature are more pronounced than intra-annual variations and the common conceptualization that night is the winter of the tropics.
Figure 10.1 The ecological zones of Nigeria
3243 J&J Vulnerability
1/11/07
10:23 am
Page 200
200 Climate Change and Vulnerability
Table 10.1 Mean monthly minimum and maximum temperatures by zone in Nigeria, 1971–2000 Zone
Mean monthly minimum temperature (°C)
Mean monthly maximum temperature (°C)
Lowest (month)
Highest (month)
Lowest (month)
Highest (month)
Forest
21 (Jan)
24 (Apr) 2
8 (Aug)
34 (Mar/Apr)
Guinea savannah
17 (Jan)
24 (Apr)
30 (Aug)
38 (Mar/Apr)
Sudan/ Sahel savannah
12 (Jan)
26 (May)
28 (Dec/Jan)
40 (May)
Within the forest zone, mean annual rainfall varied between 1250 and 3000mm. Although rain can be expected during each month, there is usually a dry period of two to four months with significantly lower rainfall. During the dry period, the air remains humid. In the Guinea savannah zones constituting the Middle Belt of the country, the year is sharply divided into rainy season and dry season. During part of the dry season, a dry air mass, which comes in from the Sahara Desert, overlies the area. Dryness is expressed in terms of both low rainfall and low humidity of the air. There is little difference between the northern, drier boundary and the southern wet boundary in terms of total annual rainfall. However, the dry season is about seven months long in the northernmost areas, whereas it is only five months long in the south. The boundaries between the Northern Guinea Zone and the Sudan correspond to a sharp drop in mean annual rainfall from 1200mm to about 900mm, while the boundary between the Sudan and the Sahel corresponds to a mean annual rainfall of 600mm. In the Sahel, the rainy season is barely three months long, whereas in the Sudan, the rainy season extends over a period of four months. Inter-annual variability of maximum temperature (Table 10.2) is spectacularly low, averaging less than 5 per cent across the ecological zones and from January to December. The implication of this is that each year is very much like another with respect to daytime temperatures. The stability of high daytime temperatures from month to month and from year to year is a well-known marker of typical tropical climates. This notwithstanding, one can discern temporal and spatial patterns in the variability of mean monthly maximum temperatures. In forest zone locations, variability is uniformly low compared with the other zones. In the Sudan and Sahel zones, variability appears to be higher in the months of December, January and February compared with the other months of the year. Variability of minimum temperature (Table 10.3) demonstrates an unmistakable contrast between December–February on the one hand and the rest of the year on the other. In almost all of the stations, the highest coefficients were for January, followed by December. This pattern is much more pronounced in the drier northern areas than in the south. Although the average for December is more than 10 per cent, the averages for April, May, June, July, August, September and October are less than 5 per cent.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 201
Vulnerability in Nigeria: A National-level Assessment 201
Table 10.2 Variability of monthly maximum temperature, 1971–2000 Ecological Zone
Climate Station
Jan
Sahel Sudan N. Guinea N. Guinea S. Guinea S. Guinea S. Guinea Dry Savanna Forest Forest Forest Forest Forest
Maiduguri 5 Kano 9 Bauchi 10 Jos 6 Minna 3 Ilorin 6 Lokoja 4 Enugu 3 Ibadan 2 Benin 1 Lagos 3 Calabar 2 Port Harc 3
Feb
Mar Apr
May Jun
Jul
Aug Sep Oct
Nov Dec
6 7 6 5 3 2 2 3 3 2 2 2 2
3 4 3 1 3 2 3 4 3 5 2 4 4
3 5 5 5 5 4 3 2 2 2 1 1 1
4 4 10 6 2 2 1 2 4 3 2 4 11
4 5 3 3 3 3 1 2 3 2 2 3 3
3 3 4 1 3 3 2 2 3 2 1 1 2
2 3 2 2 3 3 4 4 3 3 2 3 1
10 2 3 3 3 2 2 2 2 3 2 2 3
4 3 3 2 2 2 1 1 3 7 2 2 1
5 2 2 2 1 2 2 2 2 2 1 1 1
5 9 5 2 2 2 2 3 2 2 2 2 2
Note: Numbers in the table are the coefficients of variation of monthly maximum temperature, or the standard deviation of monthly maximum temperature as a percentage of the mean.
The generally low variability of temperature depicted in Tables 10.2 and 10.3 explains why these parameters are usually relegated to a minor role in the literature on climate variability compared to the emphasis placed on rainfall. Although the coefficient of variation of monthly temperature normally falls between 1 and 5 per cent, that of rainfall can be as high as 600 per cent and hardly ever falls below 20 per cent. Table 10.4 presents the monthly and total annual coefficient of variation of rainfall. The low coefficients of variability of
Table 10.3 Variability of monthly minimum temperature, 1971–2000 Ecological Zone
Climate Station
Jan
Feb
Mar Apr
May Jun
Jul
Aug Sep Oct
Nov Dec
Sahel Sudan N. Guinea N. Guinea S. Guinea S. Guinea S. Guinea Dry Savanna Forest Forest Forest Forest Forest
Maiduguri Kano Bauchi Jos Minna Ilorin Lokoja Enugu Ibadan Benin Lagos Calabar Port Harc
13 13 12 13 4 10 18 8 8 10 6 5 10
9 7 8 18 3 6 17 6 3 4 4 3 3
9 10 9 7 7 2 12 2 2 3 4 4 3
4 3 12 3 4 4 4 2 2 3 4 4 2
3 4 3 2 1 2 2 2 2 6 4 3 2
4 3 12 3 2 2 1 2 3 2 4 4 1
9 5 6 8 2 5 5 5 3 4 5 2 2
4 4 3 5 2 3 6 3 2 3 4 3 2
4 3 2 3 2 3 4 2 2 8 4 2 2
4 3 2 4 2 2 1 2 1 1 2 3 1
6 5 3 6 2 2 1 3 2 2 3 3 1
12 8 8 12 5 11 16 8 3 3 4 4 4
Note: Numbers in the table are the coefficients of variation of monthly minimum temperature, or the standard deviation of monthly minimum temperature as a percentage of the mean.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 202
202 Climate Change and Vulnerability
the annual totals compared with the variability of monthly totals are noteworthy: while the coefficients for annual totals vary from 9 to 26 per cent, those of monthly totals vary from 0 to 600 per cent. There is no significant spatial pattern in the variability of annual total rainfall. However, one can discern a tendency for variability of annual totals to increase as the total rainfall decreases. The station with the least variable annual total, Jos, is distinguished by being the only high-altitude location among the list selected for this analysis. Table 10.4 Variability of monthly rainfall, 1971–2000 Ecological Climate Jan Feb Mar Zone Station
Apr May Jun
Jul Aug Sep Oct
Nov Dec Year
Sahel Sudan Sudan N. Guinea N. Guinea N. Guinea N. Guinea S. Guinea S. Guinea S. Guinea Dry Savanna Forest Forest Forest Forest Forest
183 148 176 102 78 82 68 102 52 62 46 41 38 53 37 49
41 37 39 29 31 29 23 27 50 48 38 39 36 73 35 35
0 600 0 50 271 462 291 244 187 225 137 109 82 75 60 68
Maiduguri 0 Sokoto 0 Kano 0 Bauchi 0 Yola 0 Kaduna 0 Jos 379 Minna 582 Ilorin 308 Lokoja 296 Enugu 188 Ibadan 223 Benin 170 Lagos 143 Calabar 115 Port Harc 104
0 500 415 500 0 321 326 372 162 204 143 136 175 100 116 76
493 292 477 240 228 211 128 138 75 89 75 71 51 63 52 62
74 76 72 50 76 43 34 49 40 37 32 44 28 41 27 31
59 65 51 45 41 31 18 33 31 37 35 38 29 38 33 31
39 32 46 23 28 30 24 29 66 57 41 105 53 115 32 39
59 174 44 145 63 160 51 72 40 76 33 83 18 80 34 61 35 67 31 43 32 45 49 32 28 42 49 63 35 26 23 36
500 0 0 0 0 0 0 0 384 489 226 165 96 136 152 102
22 21 26 18 16 14 9 26 24 16 18 25 11 20 12 14
Note: Numbers in the table are the coefficients of variation of monthly rainfall, or the standard deviation of monthly rainfall as a percentage of the mean.
December and January are perennially dry in the Northern Guinea, Sudan and Sahel zones. Therefore the coefficients of variability for these months and for these stations are zero. Apart from these, dry season months throughout the country are characterized by very high coefficients of variability. The dry season months with coefficients greater than 100 per cent in the Sudan and Sahel zones are February, March, April, October and November. In the Northern Guinea zone, the affected months are February, March, April and November. In the areas extending from the coast through the forest zone to Southern Guinea, coefficients greater than 100 per cent are recorded for December, January and February. Within the rainy season, very high coefficients are recorded for the months representing onset and cessation of the rains throughout the country. Such months include June and September in the arid zones and March, April, October and November in the forest zone. In summary, the percentage of periodic change is greatest in months with the smallest average precipitation and decreases as rainfall increases.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 203
Vulnerability in Nigeria: A National-level Assessment 203
Climate change The most recent assessment of climate change projections concludes that Africa is very likely to warm more than the global average. The median projection of 21 global climate models for annual mean temperature increase is 3.3°C for West Africa from 1980–1999 to 2080–2099, with a range of 2.7 to 3.6°C for 25–75 per cent of the projections, compared to a global mean increase of 2.8°C (Christensen et al, 2007). Rainfall could either decrease or increase. Projected rainfall changes for West Africa range from decreases (9 per cent annually and 18 per cent for the wet months of June–August) to increases (13 per cent annually and 16 per cent for June–August). These projections correspond to the A1B emission scenario of the Intergovernmental Panel on Climate Change (IPCC, 2000), which assumes rapid economic and population growth and a balanced mix of energy technologies and yields emissions in the upper-mid range relative to other scenarios. Our analysis for Nigeria is based on earlier projections of climate change. Specifically, we used experiments of the Hadley Centre’s HadCM2 model for emissions of greenhouse gases that increase at annual rates of 0.5 and 1.0 per cent. The climatic parameters included in the analysis are precipitation, minimum temperature, maximum temperature, cloud cover and vapour pressure. The patterns of change with respect to maximum temperature and precipitation for the higher emission scenario, applied to selected stations in Nigeria, are shown in Figures 10.2 and 10.3. This scenario projects increases in mean temperatures of the order of 5°C or more by 2099, with minimum temperatures increasing more than maximum temperatures. On average, mean vapour pressure may rise by as much as 5–8hPa (hectopascals, equal to 100 pascals) with the potential for a significant increase in atmospheric energy. One would expect from this scenario an increase in the frequency and intensity of stormy weather (IPCC, 2001). A general decrease in mean cloudiness is projected. This could improve the availability of sunlight for primary biological productivity. There has been an observed trend toward aridity in sub-Saharan West Africa (Adejuwon et al, 1990; Nicholson, 2001; Hulme et al, 2001). The HadCM2 model projects increasing rainfall for the region, suggesting that this trend could be put on hold or reversed as the century progresses. Note, however, that some other models project decreases in rainfall for West Africa and that there is not sufficient agreement across models to conclude whether rainfall will increase or decrease (Christensen et al, 2007). One aspect of the current climate pattern that seems likely to be carried forward into the climate of the future is more pronounced zonation. Rainfall, cloudiness and humidity are projected to decrease with distance from the ocean, whereas temperature and incident solar energy are projected to increase.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 204
204 Climate Change and Vulnerability
Figure 10.2 Mean monthly maximum temperature projections, 1961–2099
Figure 10.3 Mean monthly precipitation projections, 1961–2099
3243 J&J Vulnerability
1/11/07
10:23 am
Page 205
Vulnerability in Nigeria: A National-level Assessment 205
Sensitivity of Crop Yield to Climate Variability and Change Sensitivity of yields to contemporary climate variability Adejuwon (2005a) investigated the sensitivity of crop yield to climate variability in a case study of sites in the arid region states of Bornu and Yobe. The study area covers 70,000km2 and lies between latitudes 10°N and 13°N and between longitudes 10°E and 15°E in the northeastern corner of the country. The climatic data used for the analyses are from the records of the two meteorological stations within the study area, namely Maiduguri (11.53°N, 13.16°E) and Potiskum (11.40°N, 11.03°E). The crop yield and climate data cover the 17-year period from 1983 to 1999. The major food crops included in the analyses are maize, sorghum, rice, millet, cowpeas and groundnut. Bivariate correlation, multiple correlation and regression analyses were employed to demonstrate the relationship between crop yield and climate variability. The predictive models generated for cowpeas, groundnut, millet and sorghum are statistically significant with < 0.05. Among the more powerful determinants of crop yield were rainfalls in the onset and cessation months of the growing season. Inter-annual changes in the yields of maize and rice are less sensitive to rainfall variability. In general, the predictive models failed to incorporate the separation of crop yield variability from rainfall at higher levels of precipitation. During the long periods with normal and above normal rainfall, crop yield sensitivity to rainfall tended to be weak. However, during the years with unusually low precipitation, crop yield sensitivity became more pronounced. Thus, for purposes of developing the appropriate adaptation strategies, a distinction should be made between drought, with its associated disasters, and the less hazardous but more frequent inter-annual climate or rainfall variability.
Sensitivity of yields to projected climate change Our analyses of sensitivity of crop yield to projected climate change are based on simulations using the crop model Erosion Productivity Impact Calculator (EPIC) (Williams et al, 1989; Adejuwon, 2005b) and the HadCM2 climate projections. The simulations indicate both benefits and risks. The case of maize is depicted in Figure 10.4 as an example. Port Harcourt (forest zone), Ilorin (S. Guinea), Kaduna (N. Guinea), Kano (Sudan) and Maiduguri (Sahel) represent different zones along the climatic profile from the humid to the semi-arid areas. Jos is presented as an example of a zonal high-altitude ecology. Outputs of the EPIC runs include biomass production, economic yield, water stress, temperature stress, nitrogen stress, phosphorus stress and aeration stress. These are useful in identifying the limiting environmental factors. The general pattern of projected changes in the yield of maize, as depicted in Figure 10.4, is an initial increase from 1961–1990 to 2010–2039, followed by a decline towards the end of the 21st century. The exception is at Jos, representing a high-altitude ecosystem, where yield increases from less than 3 metric tons per hectare in Jos for the 1961–1990 period to more than 6 metric tons per hectare during 2070–2099. For Kano, representing the Sudan savannah
3243 J&J Vulnerability
1/11/07
10:23 am
Page 206
206 Climate Change and Vulnerability
Figure 10.4 Maize yields for projected climate change, 1961–2099
ecology, Maiduguri, representing the Sahel ecology, Kaduna, representing northern Guinean ecology and Port Harcourt, representing the forest zone ecology, peak yields are projected for 2040–2069. For Ilorin, representing the southern Guinean ecology, peak yields are projected for 2010–2039. The yield predicted for the end of the 21st century is lower than the yield for 1960–1990. This means that over the 140-year period, a net decrease in the maize yield is projected for Ilorin.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 207
Vulnerability in Nigeria: A National-level Assessment 207
The general increase in yield is explained in part by the effects of increases in the atmospheric concentration of carbon dioxide, the principal greenhouse gas, on plant water use efficiency and photosynthesis. The steady increase in yield at the high-altitude location of Jos corresponds to a steady decline in water stress from 47 days during 1961–1990 to 15 days during 2070–2099. Temperature stress is also projected to decline from 2.6 days to 0.5 days. Because the high-altitude site also experiences the projected general warming trend, it might be concluded that the current temperature levels are suboptimal in the region represented by Jos and that the warmer days ahead will provide environmental conditions for enhanced yield of maize in that region. However, it seems clear that the steady and significant increases in the yield of maize projected for Jos are related to increased rainfall and not increased temperature. In Maiduguri, which falls in the Sahel ecological zone, the increases in yield from 1961–1990 to 2040–2069 correspond to a decline in the level of water stress from 46 to 28 days. However, the further decline to 25 days during 2070–2099 seems not to be reflected in continued increase in crop yield. It appears that by the time of the 4th time slice, the rising temperature would have taken over as the limiting factor. In other words, the decline in crop yield from 2.75 metric tons per hectare to 2.31 metric tons per hectare could be ascribed to the increase in temperature stress from 8 to 11 days. It is important to bear in mind that the above analysis is based on a single climate change scenario that projects increases in rainfall for the region, while other scenarios of climate change indicate the potential for rainfall to decrease. Should rainfall fail to increase, or if it should become more variable, rainfall would not offset the temperature stress effects of warming and the potential for yield losses would be greater. One should also note here that across the country and across the four time ranges, the experiments did not show any stress related to nitrogen, phosphorus or aeration. In selecting the soil with which to create the soil file used in the EPIC model simulations, we adopted the more commonly cultivated soil series, which, in most cases, are of moderate productivity. In addition, the simulations assume that 300kg/ha of NPK is applied as part of the operations schedule.
Climate-driven risks and benefits Risks are the expected losses (losses of life, persons injured, property damaged and economic activity disrupted) because of a particular hazard for a given area and reference period (Downing et al, 2001). From the foregoing analysis, the risks to the livelihood of peasant households are related strongly to the potential for reductions in the yields of their crops. The risks vary from widespread crop failure affecting all crops in the arid zone or some of the crops throughout the country, to regional, local and individual farm-level crop failures affecting the more sensitive crops. Depending on its severity, crop failure could result in food inadequacy and famine, loss of livelihood, and long-distance emigration from the arid zone to the rainforest belt. Usually it is the able-bodied young men that emigrate. A large proportion of those left behind, especially
3243 J&J Vulnerability
1/11/07
10:23 am
Page 208
208 Climate Change and Vulnerability
the women and the children, pay with their lives. Moreover, the crop production systems are constrained by poor soils, disease, pest infestation and chronic peasant production diseconomies. These serve as amplifiers of the climate factor as the chief driver of the risks. The drivers of crop failure are droughts; in other words, inadequate growing season rainfall. Droughts have always been a common, though irregular feature of the arid region of Nigeria. The areas most exposed to the incidence of disastrous droughts are the Sudan and Sahel ecological zones, which cover about 240,900km2 and 20,700km2 respectively and thus constitute nearly 27 per cent of the country’s land area. These areas are characterized by a mean annual rainfall averaging 600 to 800mm and a short rainy period of 100 to 120 days. From the records, five major drought periods resulting in human deprivation were documented in the Nigerian arid zone during the 20th century. These were the 1913–1914, 1931–1932, 1942–1943, 1972–1973 and 1983–1984 droughts (Mijindadi and Adegbesin, 1991). Recurring drought periods were not limited to the 20th century. Oladipo (1988) traced the occurrence of periodic droughts in Africa from before the birth of Christ up to the 20th century. The farmers in the arid zones are quite conscious of the weather factors in their lives. According to the farmers interviewed during field work (Adejuwon, 2005a and 2006a), when the seasonal rains come early (in May), crop performance is high. When seasonal rain onset is delayed until July, however, crop yield is low. This is because unless cessation of rain is also delayed, the growing season is shortened and there is insufficient time for crops to mature. Rain coming in June is judged to be favourable and is associated with good yields of crops. Most crop failures are, however, associated with premature cessation of the rainy season. There can also be a low yield of crops when there are prolonged dry spells within the growing season. However, the most significant negative impacts of climate on crop production are delivered by extreme events such as season-long droughts with decadal frequencies of occurrence. Disaster usually comes in the form of a late arrival and/or a premature termination of the growing season. In the broad band of years that could be described as normal, noticeable, rather than significant, responses of crops to changes in climate can be observed. These responses reflect changes in the environmental drivers from one locality to another. For the rest of the current century, the risks outlined above will most likely remain for the Nigerian peasant household. However, the severity, spatial resolution and temporal resolution could differ as the century progresses. Instead of oscillating or periodic timing, the risks could gather momentum as the century progresses. Instead of cellular occurrence, the risks could affect the entire country with comparable severity. For the scenario of increasing rainfall that was analysed using the EPIC model, risks would most likely be least during the first half of the century, during which crop yields increase in response to higher levels of solar radiation, atmospheric humidity, rainfall and carbon dioxide. As the century progresses, however, the risks will be driven by a new set of climatic factors. The favourable moisture-based drivers projected for the first half of the century will be overshadowed by temperature-based factors in the second half. In the context of ecological factor interaction, the negative
3243 J&J Vulnerability
1/11/07
10:23 am
Page 209
Vulnerability in Nigeria: A National-level Assessment 209
impacts of supra-optimum temperatures will tend to mask the positive impacts of increases in solar radiation, moisture and carbon dioxide during the second half of the century.
Indirect Sensitivity of Peasant Households to Climate By definition, a peasant household depends on agriculture and related activities for whatever livelihood its members are able to eke out of their environment. It is the view in certain quarters that ‘one major cause of vulnerability to climate change is dependence of the exposure unit on sectors such as agriculture, forestry and fishery that are sensitive to changes in climate’ (Sperling, 2003). The logic of this viewpoint is quite easy to appreciate. Crop production, on which the peasant householder depends for his livelihood, is sensitive to climate variability and climate change. Whatever effects climate has on crop production affect the peasant household. In essence, sensitivity of crop production to climate is thus a good measure of the sensitivity of the peasant household to climate variability and climate change. The National Agricultural Sample Survey (FOS, 1983–97) indicates that 94 per cent of agricultural holdings are involved in crop farming. A good measure of the extent of dependence on agriculture is the percentage of employed persons in the sector. For the country as a whole, 65 per cent of employed persons worked in the agricultural sector in 1993 (FOS, 1996b). The data for 1993 also indicate that the percentage was above 50 in 20 out of the 30 states in the country. In general, the proportion of employed persons in agriculture tends to be higher in the northern, drier parts of the country than in the wetter south. The major crops grown by the households include maize, guinea corn and cassava. During the 1993/1994 crop season, maize was the most widely cultivated crop and was grown on 54 per cent of the household land holdings, while guinea corn (sorghum) was cultivated on 48 per cent and cassava on 47 per cent of the holdings (FOS, 1983–1997). Most of the states of the Nigerian Federation recorded high percentages for maize. The exceptions were Jigawa, Sokoto and Yobe, located in the Sudan and Sahel ecological zones, where less than 10 per cent of land is planted in maize. Sorghum cultivation was concentrated in the same Sudan and Sahel zones, where more than 90 per cent of the households cultivated the crop. 60 to 70 per cent of the households in the Guinean (middle belt) zones cultivated guinea corn, whereas in the southern forest zone less than 10 per cent of the households were engaged in cultivating the crop. The core of high-intensity cultivation of cassava is in the southeastern states of Anambra, Imo, Enugu, Akwa Ibom and Abia and extends westward into Delta, Edo and Oyo states, all lying outside the cocoa belt. The intensity of cultivation of cassava decreases northward. The Sudan and Sahel zones could be considered as lying outside the proper cassava growing areas. What all this implies is that any disaster overtaking maize, sorghum or cassava cultivation as a result of climate variability or climate change would also be a disaster for the Nigerian peasant household. Beans, millet, yams and ground-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 210
210 Climate Change and Vulnerability
nuts are also regionally important crops, cultivated by between 30 and 40 per cent of the households during the 1993/1994 growing season. Yam cultivation is intensive in the southeast and the middle belt, while beans and millet are important crops in the far north.
Constraints to Adaptive Capacity A low capacity to adapt to climate change automatically implies vulnerability. Among the factors imposing limitations on adaptive capacity, the most significant is persistent poverty, which signifies absence of the resources necessary for adapting to climate change (Sperling, 2003). In addition, there are disabilities such as poor health that could undermine labour availability for the farming activities both in quality and in quantity. Relatively low levels of educational attainment in parts of the country could also constrain the ability to acquire the technological capacity for combating the negative consequences of climate change. The rate of population increase, which at present stands at 28 per 1000, could also increase the rates of child dependency, increase pressure on social infrastructure and impose limitations on the capacity to cope with the negative impacts of climate change.
Poverty Widespread poverty has been cited as the main cause of a low capacity to adapt to climate change in Africa (Desanker et al, 2001). Resources, including social, financial, natural, physical and human capital, are required for planning, preparing for, facilitating and implementing adaptation measures. Poor persons, poor communities and poor nations do not have enough of these resources, hence their low adaptive capacity. Existent or pre-impact poverty connotes vulnerability. A summary of the latest results from the National Integrated Survey of Households of 1995 (FOS, 1996b) showed that practices that could have boosted the adaptive capacity of the peasant households were still being constrained by lack of funds at the individual household level. With respect to crop production, the use of pesticides and insecticides was limited to 4 per cent; the use of improved seeds was limited to 11 per cent and the use of chemical fertilizers was limited to 32 per cent of the peasant holdings. Among those who did not use fertilizer, 51 per cent considered the cost too high, 8 per cent found the distance to the source to be too far, 23 per cent did not know where to obtain fertilizer and 12 per cent felt they did not need fertilizers. Of those who were not using pesticides and insecticides, 36 per cent felt the cost was too high, 24 per cent felt no need for them and 22 per cent did not know where to obtain them. All of these reasons are due to inadequate financial resources and ignorance, which are the hallmarks of the poor. Ninety-four per cent of the holders had no credit for their farm work. Only 1 per cent had credit through formal banking and a cooperative society system. Informal credit systems accounted for only 2.6 per cent. Friends and relatives were the source of credit for 2 per
3243 J&J Vulnerability
1/11/07
10:23 am
Page 211
Vulnerability in Nigeria: A National-level Assessment 211
cent of households, while 0.4 per cent obtained credit from money lenders (FOS, 1996b). There are currently two official poverty lines that are set relative to the standard of living in the country (FOS, 1999). There is a moderate poverty line, which is equivalent to two thirds of the mean per capita expenditure, and a core poverty line, which is equivalent to one third of the mean per capita expenditure. Using the two lines, households are classified into three mutually exclusive groups: core poor, moderately poor and non-poor. There has been a change for the worse in the poverty structure of the country in recent years. The incidence of poverty increased from 27.2 per cent in 1980 to 46.3 per cent in 1985 and 65.6 per cent in 1996. Over the same period, the incidence of core poverty increased from 6.2 per cent in 1980 to 12.1 per cent in 1985 and 29.3 per cent in 1996 (FOS, 1999). Whether poor or non-poor, most of the income is derived from basic cash (i.e. from wages, salaries, rents, interests on investments, etc., as opposed to COP (consumption of own production) income such as gifts in kind, loans of all types etc.). However, basic cash is 10 per cent higher in the income of the non-poor compared with the income of the poor. On the other hand, the consumption from the own production component in the income of the poor is as high as 20 per cent compared with 12 per cent in the income of the non-poor. The core poor and the moderately poor spent 76.2 and 72.9 per cent, respectively, of their total income on food compared with 58 per cent spent for the same item by the non-poor in 1996 (FOS, 1999). There are significant differences between the poor and the non-poor with regard to the proportion spent on non-food items. The respective percentages for the core poor, moderately poor and non-poor were 24, 27, and 41 in 1996 (FOS, 1999). The incidence of poverty is considerably higher among households engaged in agriculture compared with households employed in the other sectors. In 1996, 77 per cent of farming households were classified as poor compared with 65.6 per cent for the total population. Among the farming households, 48 per cent fell into the core poor category compared with 29.3 per cent for the country as a whole (FOS, 1999). There are also regional disparities in the incidence of poverty. In 1996, householders in the southern forest-based states were less poor compared with those in the middle belt (Guinean) states, while those in the Sudan and the Sahel states were the poorest.
Demography-induced constraints Nigeria has a youthful population (FOS, 1996c). Going by the estimates for 1995, children under the age of 15 years constituted 44.0 per cent of the population. The youthful age structure creates a built-in momentum for future population growth. Even if it were possible to reduce the growth rate to replacement level, births would outstrip deaths and the population would continue to increase until the very large number of young females had passed through their reproductive years. The percentage of the female population that is in the reproductive age bracket, after declining between 1980 and 1990, increased from 1990 to 1995, giving indications that the growth rate might be accelerating. In 1993, women in the reproductive age brackets constituted 46.0
3243 J&J Vulnerability
1/11/07
10:23 am
Page 212
212 Climate Change and Vulnerability
per cent of the female population (FOS, 1996c). With such high percentages within the reproductive age bracket, coupled with the youthful nature of the population, policies aimed at reducing fertility may not produce the desired results within a short time. In the short run, the population of children would remain high relative to total population even if the fertility rate declined. The youthfulness of the population is directly responsible for the high rate of child dependency burden in the country. Child dependency burden is calculated as the population of children below the age of 15 divided by the population of working adults aged from 15 to 59. (Going on the data for 1993, the southwest has the least burden, followed by the southeast and the northwest.) The main consequence of this at the household level is that each adult household member must strive to provide for many more persons than can be conveniently accommodated. In countries with such high proportions of children relative to the proportions of the working age population, a high percentage of national income is expended on consumable goods for children. The higher the percentage of income expended on these consumables, the lower the percentage of income left for savings and investments. Thus a high dependency burden has an effect similar to poverty. The capacity to cope with any additional stress in the form of negative consequences of climate change will thus be lower in situations with high dependency burden because the first inclination would be to care for children rather than to prepare for a future under a changed climate. The rapidly expanding population is exerting increasing pressure on the social and economic infrastructure of the country. Schools, hospitals and houses become inadequate almost as soon as they are completed. Similarly, electricity, water and waste disposal facilities designed for a given population are being made to serve much more than the population on the day they are commissioned. In other words existing facilities are being used to a higher capacity than that for which they were designed. The result is a high rate of infrastructure deterioration. There is the possibility that these inadequacies will tend to command greater attention from policymakers and draw away funds from proactively responding to the consequences of a potential climate change.
Educational status Education definitely enhances personal, community and national capacity to respond to external stresses placed on human livelihood and well-being. Therefore inadequate or substandard education is a measure of the vulnerability of human exposure units to expected negative impacts of climate change. It is easy to appreciate the fact that education is one of the means for achieving the goals of better health, higher labour productivity and more rapid GDP growth, all of which are required as the need arises to anticipate, manage or adapt to a worsening climatic factor (Basher et al, 2000). The higher levels of education are especially necessary to enable individuals and countries to understand and participate more fully in the technological and administrative processes of the modern global economy. Since achieving independence from
3243 J&J Vulnerability
1/11/07
10:23 am
Page 213
Vulnerability in Nigeria: A National-level Assessment 213
colonial rule in 1960, a considerable proportion of the national income has been invested in raising the standard of education. Because of this, enrolment ratios have been trending upwards (FOS, 1996c). The national adult literacy rate averaged 25 per cent in 1970. By 1995, it had climbed to 49 per cent. However, regional disparities are well marked. In 1995, literacy rates among male adults varied from a low of 19 per cent in Jigawa state in the far north, to a high of more than 93 per cent in Lagos state, located in the humid forest zone. In general, the northern parts of the country are chronically handicapped in terms of the necessary literacy level for adapting to climate change.
Health and adaptive capacity The view has been expressed that climate change’s impact on human health will increase vulnerability and reduce opportunities by interfering with education and the ability to work (Sperling, 2003). In the absence of mechanization and other forms of modernization, the main input to crop and animal production in Nigeria is labour. Most of the labour used on peasant farms is supplied by members of the household. In households headed by women, hired labour could be employed for the more strenuous activities such as tilling in preparation for planting. Households engaged in cash crop production use more hired labour than households engaged in food crop production. During the harvest season, households cooperate to ensure that farm output is brought in as soon as possible. There is always limited time available for harvests as delay may expose the harvested crops to pests, diseases and destruction by the weather. Thus sufficient and timely availability of labour is crucial to the level of yield realized. The effect of the HIV/AIDS epidemic in limiting farm productivity is common knowledge. Hands that could have been employed in production are either lost through death or immobilized by sickness. Statistics on morbidity and mortality due to HIV/AIDS are not yet in the public domain and estimates supplied through the news media are largely unreliable. The rates of losses in farm worker days due to the disease are probably now as high as those reported for more traditional human ailments. However, most of the reported cases of death and sickness are children under the age of 5. Sick children are likely to reduce the normally substantial contributions of their mothers to farm labour.
Spatial Pattern of Vulnerability An aggregate index of vulnerability is constructed from three groups of attributes: exposure to climate variability and projected climate change, the sensitivity of households to climate, and the inherent capacity of the households to adapt to climate-related risks. Exposure to climate variability is ranked according to the variability of the onset of rainfall. The variability indices of rainfall in March, April/May and June are used as indicators of exposure to climate variability for the forest zone, Guinean zones (north and
3243 J&J Vulnerability
1/11/07
10:23 am
Page 214
214 Climate Change and Vulnerability
south) and semi arid zones respectively. The variability index of the onset of the rainy season can be as low as 34 per cent and as high as 65 per cent. The index is lowest at Jos and highest at Maiduguri. Hence Plateau state, where Jos is located, is ranked first, or least exposed, while Bornu state, in which Maiduguri is located, is ranked 30th, or most exposed. Our simulations of crop yields found that temperature effects have a negative effect on yields in most zones of Nigeria and that temperature becomes the dominate driver of yield changes towards the latter half of the 21st century. Consequently, exposure to climate change is measured by the projected average temperature for the growing season for the 2070–2099 time range. Growing season mean temperature for 2070–2099 is projected to be lowest along the coast and highest at the boundary with the Sahara Desert in the north. Hence, coastal states are ranked highest, or least exposed to climate change stresses, followed by the middle belt states, while the northernmost states are ranked lowest, or most exposed. As noted previously, high proportions of land holdings devoted to agricultural activities and of dependence of households on agriculture for livelihoods connote a condition of high sensitivity to climate stresses. We measure the sensitivity of households to climate by their dependence on crop production as indicated by the percentage of the households employed in agriculture. Hence Lagos state, where the percentage of households employed in agricultural production is lowest, is ranked first, or least sensitive, while Jigawa state, where the percentage is highest, is ranked 30th, or most sensitive. The attributes of adaptive capacity include economic, health, education and demographic conditions of the households. The indicators used to measure adaptive capacity are poverty head count for the economic attribute of adaptive capacity, the under-five mortality rate for the health attribute, adult literacy rate for education and the child dependency ratio for the demographic attribute. The state with the lowest percentage classified as poor, the lowest under-five mortality, the highest adult literacy rate and the lowest child dependency burden was ranked first in adaptive capacity. Other states were ranked lower in succession with respect to each attribute. To construct the index of vulnerability, the ranks for exposure, sensitivity and adaptive capacity for each state were averaged; the results of the ranking are summarized in Figure 10.5. The mean rankings show clearly that the peasant households in the arid zone states of the Sudan and Sahel are potentially the most vulnerable to climate change, followed by those in the middle belt or Guinea savannah, while households in the forest-based states are potentially the least vulnerable. The spatial pattern of vulnerability depicted in Figure 10.5 can be explained in terms of differences in cultural, historical and environmental attributes. The north is predominantly Moslem while the south is predominantly Christian. Because of this, European influence in the form of development has been consistently stronger in the south than in the north. Thus attributes favouring higher levels of adaptive capacity such as the provision of health and education facilities and availability of non agricultural employment tend to favour the south. Also the south, with its longer rainy season, is characterized by less
3243 J&J Vulnerability
1/11/07
10:23 am
Page 215
Vulnerability in Nigeria: A National-level Assessment 215
Figure 10.5 Vulnerability of peasant households to climate change by state
exposure to water stress, less sensitivity to climate due to greater diversity of agricultural products and a higher degree of resilience against the hazards and risks of a changing climate. This is based not only on a multiplicity of crop products – including cocoa, oil palm, coffee, cassava, yam, plantain and banana – but also on the plant and animal natural resources of the rainforest zone. Contemporary average temperature in the north is higher than the average temperature in the south; thus a general increase in temperature of the same magnitude will attain levels that will breach the upper limits of tolerance for crops earlier in the north than in the south. For this reason, the risks confronting peasant householders in the event of a changed climate are greater in the north than in the south.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 216
216 Climate Change and Vulnerability
Summary and Conclusions Exposure of crop production to contemporary climate variability, especially to droughts of varying severity, is a major source of existing vulnerability for the Nigerian peasant household. Climate change during the 21st century in Nigeria will be manifested by a statistically significant rise in temperature. Precipitation changes are projected with less confidence and either increases or decreases are possible. For the scenario of increased precipitation and atmospheric humidity analysed in our study, crop yield is projected to increase substantially during the first half of the century. Such increases, if they occur, would tend to relieve contemporary vulnerability. However, during the second half of the century, benefits from increases in precipitation and atmospheric humidity would tend to be overshadowed by the negative consequences of the higher temperatures. The net result of these higher temperatures could be that climate change will pose considerable risks to peasant household livelihood, health, crop production and food security. The risks will most likely be intensified towards the end of the century and will tend to be greater in the north than in the southern parts of the country. Peasant households of Nigeria are vulnerable to climate change because they are dependent on agricultural livelihoods that are highly sensitive to climate, because exposures to climate stresses are likely to become increasingly adverse for crop production as temperatures rise and because their capacity to adapt is low. The low adaptive capacity derives from poverty, food insecurity, poor health, low educational attainment, inadequate social and economic infrastructure, and explosive population growth. Vulnerability seems to be highest in the northern areas of Nigeria, where current and projected climate are less favourable for crop production and adaptive capacity is lowest. Vulnerability is lower in the coastal and high-altitude regions, where there are indications that the projected climate change may present opportunities for improving the quality of human life.
References Adejuwon, J. O. (2005a) ‘Food crop production in Nigeria: I – Present effects of climate variability’, Climate Research, vol 30, pp53–60 Adejuwon, J. O. (2005b) ‘Assessing the suitability of EPIC Crop Model for use in the study of impacts of climate variability and climate change in West Africa’, Singapore Journal of Tropical Geography, vol 24, pp44–60 Adejuwon, J. O. (2006a) Food Security, Climate Variability and Climate Change in SubSaharan West Africa, Final report of AIACC Project Number AF 23, The International START Secretariat, Washington, DC, US Adejuwon, J. O. (2006b) ‘Food crop production in Nigeria: II – Potential effects of climate change’, Climate Research, vol 32, pp229–245 Adejuwon, J. O., E. E. Balogun and S. A. Adejuwon (1990) ‘On the annual and seasonal patterns of rainfall fluctuations in Sub-Sahara West Africa’, International Journal of Climatology, vol 10, pp839–848 Basher, R., C. Clark, M. Dilley and M. Harrison (2000) Coping with the Climate: A Way Forward, International Research Institute for Climate and Society, Palisades, NY, US
3243 J&J Vulnerability
1/11/07
10:23 am
Page 217
Vulnerability in Nigeria: A National-level Assessment 217 Christensen, J. H., B. Hewitson, A. Busuioc, A. Chen, X. Gao, I. Held, R. Jones, R. Koli, W. Kwon, R. Laprise, V. Rueda, L. Mearns, C. Menendez, J. Raisanen, A. Rinke, A. Sarr and P. Whetton (2007) ‘Regional Climate Projections’, in S. Solomon, D. Qin, M. Manning, Z. Chen, M. C. Marquis, K. Averyt, M. Tignor and H. L. Miller (eds), Climate Change 2007: The Physical Science Basis, Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US Desanker, P., C. Magadza, A. Allali, C. Basalirwa, M. Boko, G. Dieudonne, T. E. Downing, P. O. Dube, A. Githeko, M. Githendu, P. Gonzalez, D. Gwary, B. Jallow, J. Nwafor and R. Scholes (2001) ‘Africa’, in J. McCarthy, O. Canziani, N. Leary, D. Dokken, and K. White (eds) Climate Change 2001: Impacts, Adaptation and Vulnerability, Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US Downing, T., R. Butterfield, S. Cohen, S. Huq, R. Moss, A. Rahman, Y. Sokona and L. Stephen (2001) Vulnerability Indices: Climate Change Impacts and Adaptations, Policy Series No 3, United Nations Environment Programme, Nairobi, Kenya FOS (1983–1997) National Agricultural Sample Survey, Annual Series, Federal Office of Statistics, Lagos, Nigeria FOS (1996a) General Household Survey, Federal Office of Statistics, Lagos, Nigeria FOS (1996b) The Nigerian Household 1995: Summary of latest results from the National Integrated Survey of Households, Federal Office of Statistics, Lagos, Nigeria FOS (1996c) Socio-Economic Profile of Nigeria, Federal Office of Statistics, Lagos, Nigeria FOS (1999) Poverty Profile for Nigeria 1980–1996, Federal Office of Statistics, Lagos, Nigeria Hulme, M., R. Doherty, T. Ngara, M. New and D. Lister (2001) ‘African climate change: 1900–2100’, Climate Research, vol 17, pp145–168 IPCC (2000) ‘Summary for policymakers’, in N. Nakicencovic and R. Swart (eds) Emission Scenarios, Special report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK IPCC (2001) ‘Summary for policymakers’, in J. Houghton, Y. Ding, D. Griggs, M. Noguer, P. van der Linden and D. Xiaosu (eds) Climate Change 2001: The Scientific Basis, Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US McCarthy, J., O. Canziani, N. Leary, D. Dokken and K. White (eds) (2001) Climate Change 2001: Impacts, Adaptation and Vulnerability, Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US Mijindadi, N. B. and J. O. Adegbesin (1991) ‘Drought, desertification and food production in Nigeria’, Savannah, vol 12, pp25–40 Nicholson, S. E. (2001) ‘Climatic and environmental change in Africa during the last two centuries’, Climate Research, vol 17, pp124–144 Oladipo, E. O. (1988) ‘Drought in Africa: A synthesis of current scientific knowledge’, Savannah, vol 9, pp64–82 Sperling, F. (ed) (2003) Poverty and Climate Change: Reducing the Vulnerability of the Poor, World Bank, Washington, DC, US Williams, J. R., C. A. Jones, J. R. Kiniry and D. A. Spaniel (1989) ‘The EPIC growth model’, Transactions of the American Society of Agricultural Engineers, vol 32, no 2, pp497–511
3243 J&J Vulnerability
1/11/07
10:23 am
Page 218
11
Vulnerability in the Sahelian Zone of Northern Nigeria: A Household-level Assessment Anthony Nyong, Daniel Dabi, Adebowale Adepetu, Abou Berthe and Vincent Ihemegbulem
Introduction The arid and semi-arid region of West Africa commonly known as the Sahel is characterized by a strong gradient of decreasing annual rainfall from south to north. Rains fall during a single wet season consisting of short intense storms over a 3- to 5-month period, with about 90 per cent of the rains falling during the months of July, August and September. Total seasonal rainfall ranges from 100 to 650mm in the northern region and from 650mm to over 1000mm in the semi-humid Sudan climate of the south (Ingram et al, 2002). Drought has been a recurrent feature in this region, with early records dating back to the 1680s. Annual rainfall levels decreased in the region over the course of the 20th century, with an increase in inter-annual and spatial variability (Dai et al, 2003; Brooks, 2004) and southward shifts of isohyets by 200km (L’Hôte et al, 2002). Whether this trend of increasing dryness in the western Sahel will continue or be exacerbated as a result of climate change is very unclear, as there is presently no convergence in predictions of climate change for the region. A comparison of climate change projections for Sub-Saharan West Africa, encompassing the Sahel, finds quite different outcomes (Dietz et al, 2001). Recent studies have suggested that increases in atmospheric carbon dioxide will lead to an enhanced West African Monsoon, wetter conditions in parts of the Sahel and an expansion of vegetation into the Sahara (Claussen et al, 2003; Maynard et al, 2002; Haarsma et al, 2005; Hoerling et al; 2006). Other models, however, project a significant drying (Hulme et al, 2001; Jenkins et al, 2005). Other studies have noted that the models that project a trend to wetter conditions do not take into account landuse changes and degradation, which are capable of inducing drier conditions (Huntingford et al, 2005; Kamga et al, 2005).
3243 J&J Vulnerability
1/11/07
10:23 am
Page 219
Vulnerability in the Sahelian Zone of Northern Nigeria 219
One area of agreement between most models, however, is that the response of the African Monsoon to increased carbon dioxide concentrations is likely to be highly non-linear, and the Sahel is likely to continue to experience a high degree of climatic variability on a range of timescales for the foreseeable future (see Wang and Eltahir, 2002; Mitchell et al, 2000). Rainfall will likely remain highly variable, and drought and associated conflicts could still pose a major problem in many areas in the Sahel (FAO, 2005). This could negatively affect the major livelihood systems of the Sahelians and increase their vulnerability, particularly to food insecurity. In climate change research, two distinct notions of vulnerability have been recognised – biophysical vulnerability and social vulnerability. Biophysical vulnerability is concerned with the ultimate impacts of a hazard event, and is often viewed in terms of the amount of damage experienced by a system as a result of an encounter with a hazard. Social vulnerability, on the other hand, is viewed more as a potential state of human societies that can affect the way they experience natural hazards (Vincent, 2004; Adger, 1999; Adger and Kelly, 1999). Social vulnerability depends on a range of factors including composition of resource endowments of the household, social relations in the community and the institutional capacity of local organizations. The most vulnerable are considered those who are most exposed to shocks, who possess a limited coping capacity and who are least resilient to recovery (Adger and Kelly, 1999). The nature of social vulnerability will depend on the nature of the hazard to which the human system in question is exposed, as certain properties of a system will make it more vulnerable to certain types of hazard than to others. In this chapter, we focus on the social vulnerability of rural farm households to drought in the Sahel region of West Africa, using a case study of selected villages in northern Nigeria. While the effects of future climate change on the frequency and intensity of drought in the Sahel is uncertain, drought is and very likely will remain the greatest climatic hazard faced by people of the region. Adapting to reduce climate risks in the future will necessarily require improved understanding of present day vulnerability to drought. Developing meaningful adaptation initiatives at the local level to reduce vulnerability to drought should begin by assessing household vulnerability among local populations. To be able to assess the vulnerability and adaptive capacity of rural households, we adopt a livelihood systems framework. This framework recognizes that a household’s vulnerability to drought is affected by its exposure to drought and its first order effects on the biophysical environment and on the livelihood assets at its disposal to ameliorate the impacts of the drought or to adapt to the situation. The main objective of this chapter is to identify the determinants of the vulnerability of farm households to drought in northern Nigeria and the implications for policy. The study uses a participatory process to identify indicators of vulnerability and rank households into various classes of vulnerability. The remainder of the chapter is organized as follows. The next section describes the study region and the study villages; the physical and socioeconomic characteristics are presented in order to better situate drought impacts in the region. This is followed by a description of the study methodology and
3243 J&J Vulnerability
1/11/07
10:23 am
Page 220
220 Climate Change and Vulnerability
then the results of our household vulnerability assessments. The chapter ends with a summary of findings and recommendations.
The Study Area Semi-arid northern Nigeria is characterized by low and variable rainfall. The Sahel is intermediate between the desert and the sub-humid zones of Africa, with average departures from the annual mean rainfall of 20 to 40 per cent over the period 1901 to 1973 (Janowiak, 1988). The key source of variation – and of drought – is seasonal rainfall. Two distinct seasons are observed, dry and wet. The dry season lasts about seven to eight months, from October to April or May, while the wet season lasts about four to five months, from May or June through September. Since the major droughts of the early 1970s, droughts have become persistent in the region (Lamb, 1982), and in 1983–1984 some stations recorded even lower rainfall than in the early 1970s (Mortimore, 1998). From the beginning of the 20th century, drought events have been recorded in Northern Nigeria in 1904–1912, 1914–1930, 1942, 1950–1952, 1966–1968, 1969–1974, 1983–1984 and 1987 (Apeldoorn, 1981, Okechukwu, 1997). During the 1968–1974 droughts, the region lost about 300,000 animals (13 per cent of the livestock population in northeastern Nigeria), agricultural yields fell to about 40 per cent of normal yields and the population at risk was about 14 million persons. Between 1983 and 1984, there was localized drought in Borno, Jigawa and Yobe resulting from deficient rainfall. In parts of Borno, the impact was as severe as the drought of 1971–73. During this drought, about 5 million metric tons of grains was lost, accompanied by constraints on biological productivity and forced migrations (Tarhule and Lamb, 2003; Ojo and Oyebande, 1985). The patterns of drought, as computed using the Bhalme and Mooley Drought Intensity Index (BMDI) for selected stations in the study area between 1930 and 1983 (Bhalme and Mooley, 1980), are presented in Figure 11.1. For the purposes of this chapter, the BMDI nine-class schema was reduced to five: BMDI values >3.00 = extreme wet, >2 = wet, >1 = Normal, –1 to –3 = mild to moderate drought and <–3 = severe drought. Observed data show that the region has experienced a trend towards increased aridity, as can be seen in the figure. Based on the 1991 census, the region has an estimated population of 47.3 million. Infrastructure to serve this population is extremely weak in terms of the quality and distribution of roads, schools and health facilities. Major indigenous ethnic groups in this region include the Hausa, Fulani, Kanuri, Shuwa, Burbur, Gerewa and Ningawa. Members of other ethnic groups have also migrated into the region from within and beyond Nigeria in recent decades. The different groups have different interests in the resource base, possess different skills with which to use it, and claim rights over different resources and areas. The region is rich in agricultural production, but the large inter-annual variability of rainfall subjects it to frequent dry spells, which sometimes result
3243 J&J Vulnerability
1/11/07
10:23 am
Page 221
Vulnerability in the Sahelian Zone of Northern Nigeria 221
Figure 11.1 Bhalme and Mooley drought intensity (BMDI) index for selected stations in northern Nigeria, 1930–1983
in severe and widespread drought that can impose serious socioeconomic constraints. Irrigated agriculture is widely practised and agricultural crops include rice, wheat, soybeans, beans, maize, millet, cotton, sorghum and groundnut. Besides arable farming, pastoralism is the main economic activity in the region. A particularly large stock of cattle is found in this zone, originating mainly from the neighbouring countries of Chad, Niger and Cameroon. These countries are currently experiencing particularly dry periods, and so livestock has been transferred from them to the transitional zone in Nigeria, where fodder is still available around the patches of wetland areas, fadamas (flood plains of rivers) and those river valleys that still contain water. Two major pastoral corridors exist in the region and 3 million hectares of wetlands dot these corridors, with an average livestock density of 13 animals per hectare, well above the carrying capacity. Herders traditionally move along these corridors on a seasonal basis, following the rains as they move from north to south and back. The land for these corridors was acquired by the federal government, which has created a number of consequent problems. The original owners of the land did not believe they were adequately compensated and have since attempted to repossess their lands. The movements of pastoralists are not monitored, and in many instances pastoralists have allowed their herds to stray from the corridor to graze on farmland, provoking conflicts. This continues because the density of herds using the corridors is greater than the forage available along them, and vegetation is consequently unable to regenerate sufficiently.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 222
222 Climate Change and Vulnerability
Data Collection Data were collected to examine vulnerability to droughts in 27 communities in northern Nigeria. The data collection started with a reconnaissance survey that took the research team from the east to the west of northern Nigeria, covering all the states that fall within the study area. The states that we visited include Bauchi, Gombe, Borno, Yobe, Kano, Jigawa, Katsina, Sokoto, Zamfara, Kebbi and Adamawa. The essence of the reconnaissance survey was to familiarize ourselves with the study area, identify the major livelihood systems and identify the major drought and other issues faced by households. We attempted to learn from members of the surveyed communities the nature and magnitude of the problem of drought and climate change. A major problem noted in the visit was the high water deficit in the area. Hand dug wells exploiting ground water are common features and it is currently estimated that the rate of abstraction of ground water in this zone is highly unsustainable, with a continuous decline in the water table. It is also common to find once perennial rivers having turned to dry valleys. We identified four major livelihood structures in the study: fishing, pastoralism, sedentary farming, and other informal livelihood systems such as mat making. We found that livelihood systems are differentiated spatially across the study region. Villages that are farther north are home predominantly to pastoralists, while those in the south are largely home to sedentary farmers. At the interface of these two livelihood systems lies a zone of mixed livelihood systems comprising both pastoralists and arable farming. Using information from the reconnaissance survey, we selected a number of communities for more comprehensive study using a three-tier sampling strategy. First, we selected all the states that are vulnerable to drought resulting from climate change. We then listed all the local government areas (numbering 250) in all the affected states. From the list, we aggregated these into larger units, numbering 27, and selected 1 community each from the larger units. The factors that we considered in selecting the study villages included the occurrence of past and/or repeated drought disasters, willingness of the women to participate in the survey, the main livelihood system in the village and the size of the village. We grouped the villages into three classes based on their population and physical size – large, medium and small. We ensured that the villages selected represented a mix of these classes. Data collection in the selected villages was done through the administration of questionnaires, focus group discussions and stakeholder analysis. The questionnaire comprises 17 sections that solicit information about household processes and relations, socioeconomic and drought-related variables, and livelihood systems and strategies. The questionnaire was developed in conjunction with relevant stakeholders identified during the reconnaissance survey and pre-tested in three communities to assess its validity and reliability in collecting pertinent information for the study. After the pre-test, some questions were modified to suit local norms and customs. 910 questionnaires were administered to household heads in 27 communities in Nigeria, of which 828 were completed.1 We define a household as a social and economic unit consisting of one or more individuals, whether they
3243 J&J Vulnerability
1/11/07
10:23 am
Page 223
Vulnerability in the Sahelian Zone of Northern Nigeria 223
are relatives or not, who live together and share both ‘the pot and the roof’ (in other words dwelling and food). The questionnaire administration was supplemented by focus group discussions. The questionnaires could only reveal past and present vulnerabilities; they do not explain the processes that are at the root of the households’ present situation of vulnerability. Focus-group discussions served to achieve this latter goal.
Self-Assessed Perceptions of Risk Although the focus of the study is to assess drought vulnerability, it is important to situate this within the overall risks faced by communities and groups in the study region. The inhabitants’ perception of risk may be based not only on the objective risk they face, but also on their subjective assessment of risk. The vulnerability of households and groups to these risks may differ and also affect the way they respond to them. In this work we asked broader questions about what worries the people face in their lives overall in order to look at how their perceptions of these worries can inform our understanding of their vulnerability. These worries were translated into risks. The risks that respondents identified are presented in Table 11.1. It should be noted that households listed more than one risk or worry and respondents were not restricted in the number of risks they could list. The risk incidence index, Ij, presented in Table 11.1, is a measure of the proportion of respondents that mentioned a particular risk. Table 11.1 Self-perceived risks Serial Perceived Risk no
Incidence Index (Ij)
Severity Index (Sj)
Risk Index (Rj)
1 2 3 4 5 6 7 8 9 10 11 12
0.55 0.55 0.52 0.50 0.40 0.37 0.37 0.34 0.31 0.28 0.26 0.25
1.03 1.14 1.24 1.25 1.47 1.55 1.62 1.63 1.74 1.76 1.89 1.92
0.93 0.91 0.86 0.77 0.60 0.47 0.44 0.33 0.30 0.18 0.15 0.14
Insufficient food for people Shortage of water for domestic use Shortage of water for animals Shortage of crops for cultivation Insufficient pasture for animals Animal diseases Limited land for cultivation Crop failure Human diseases Conflicts/insecurity Low prices for animals Lack of employment
Note: The incidence index (Ij), is a measure of the proportion of respondents that mentioned a particular risk; the severity index (Sj), measures the severity of each risk on a scale from 1 (most severe) to 2 (least severe); and the risk index (Rj), is the ratio of the incidence index (Ij) to the severity index (Sj).
Severity and risk indices also were computed using a methodology adopted from Quinn et al (2003). The severity index measures the severity of each risk
3243 J&J Vulnerability
1/11/07
10:23 am
Page 224
224 Climate Change and Vulnerability
on a scale from 1 (most severe) to 2 (least severe). It was calculated for each risk identified by each respondent as: (r – 1) Sj = 1 + ––––––– (n – 1) where Sj is the severity index value for a particular risk, r is its rank, based on the order in which it was mentioned by the respondent, and n is the total number of problems identified by that respondent. The mean value of Sj was then calculated for the subset of respondents who identified the particular problem. Finally, for each risk, a risk index was calculated as: Ij Rj = –––– Sj Since higher values for Ij indicate higher incidence and lower values for Sj indicate more severity, Rj increases with the overall risk associated with each type of problem. The risk index ranges between 0 (no incidence of risk) and 1 (severe risk). The results in Table 11.1 show that the greatest concern of the respondents was the risk of insufficient food, followed by shortage of water for domestic use. Without the respondents specifically mentioning drought, the table shows that most of the concerns are problems related to drought, an indication that drought is a major problem in the study area. This justifies the focus of the study on drought vulnerability.
Household Drought Vulnerability Assessment An index of drought vulnerability Vulnerability is a relative term differing between socioeconomic groups or regions, rather than an absolute measurement of deprivation. The analyst or decision maker must assign the thresholds of vulnerability that warrant specific responses. For our study, we developed a drought vulnerability index (DVI), which is constructed from indicators developed through a combination of top–down and bottom–up approaches. In conjunction with stakeholders, 14 indicators of vulnerability were identified along with their weights (Table 11.2). Indicators that directly impact on vulnerability are given a weight of 1, while those that affect vulnerability indirectly have a weight of 0.5. The original scores for the indices are transformed linearly so that the scores for each indicator range from 0 to 1. Adding up the weighted household scores on the 14 indices resulted in an overall vulnerability score for each household in the sample, with the scores ranging from 0.89 to 10 (see Table 11.2). Indicators that respondents felt directly explain vulnerability include crop yield, dependency ratio, livestock ownership, livelihood diversification and drought preparedness.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 225
Vulnerability in the Sahelian Zone of Northern Nigeria 225
Table 11.2 Indicators and weights for vulnerability assessment in northern Nigeria Indicator
Weight
Derived from:
Mean
Min
Max
Acreage under cultivation
0.5
Hectares/consumer units
0.21
0.04
1
Crop yield
1
Total bags of cereals harvested/consumer unit
0.50
0.06
1
Dependency ratio
1
Labour units/consumer units (inverted)
0.63
0.38
1
Livestock ownership
1
Tropical livestock units/ consumer units
0.45
Gender of household head Dummy: 1 if the household head is male, 0 otherwise
0.5
Value given to sex of household head
0.91
0
1
Livelihood diversification
1
Weighted number of nonagricultural income generating activities/ consumer units
0.29
0
1
Annual cash income
1
In 1000 naira/consumer units
0.43
0.25
1
Drought preparedness 1 Dummy: 1 if the household uses drought resistant crops or received drought-related information, 0 otherwise
Value given to use of drought resistant crops and livestock and received drought-related information and advice
0.34
0
1
Educational background of 0.5 the household head Dummy: 1 if household had completed at least primary education, 0 otherwise
Value given to highest school level attained by the head of the household
0.28
0
1
Land tenure situation 0.5 Dummy: 1 if household owns more than 50% of the land it cultivates, 0 otherwise
Value given to land tenure situation
0.57
0
1
Type of house Dummy: 1 if house is a modern house, 0 otherwise
0.5
Value given to type of house lived in
0.13
0
1
Self-sufficiency in food production
0.5
Logged number of years surplus foodstuffs were sold/number of years foodstuffs were bought in the past 10 years
0.56
0
1
Family and social networks
0.5
Number of organizations the 0.33 household is involved in
0
1
Quality of household
0.5
Number of able persons/ 0.14 number of disabled and/or sick persons in the household (inverted)
0.02
1
Note: Mean, minimum and maximum values from household surveys are normalized to [0,1] interval, with 1 representing lowest vulnerability.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 226
226 Climate Change and Vulnerability
The next step in constructing the DVI was to identify different thresholds within which households can be grouped into various levels of vulnerability. We could have adopted the Delphi method using ‘expert judgement’ or a top–down approach based only on the earlier computations of vulnerability. However, we considered neither of these methods satisfactory, as we believe that the respondents should contribute to defining their vulnerability. We therefore adopted a methodology that combined a top–down and bottom–up approach to delineating vulnerability thresholds. During the questionnaire administration, respondents participated in a vulnerability ranking exercise in which they placed themselves in any of three classes of vulnerability: very vulnerable, vulnerable and least vulnerable. These allowed us to factor in their perceptions and self-reported assessments of vulnerability. We took all those who put themselves in each group, found the average scores for each group based on our earlier computation and used those as the midpoints of the various vulnerability classes and then built class intervals about them. We arrived at the following ranges: those that scored less than 4 were categorized as highly vulnerable, scores between 4 and 7 were categorized as vulnerable and those above 7 as least vulnerable. The distribution of households by vulnerability class for each of the 27 communities is presented in Table 11.3.
A model of household vulnerability to drought As described in the previous section, households’ self-assessments of their level of vulnerability to drought and contributing factors were established through household surveys and participatory exercises. We then performed a multivariate analysis with the aim of identifying how these and other factors influence the probability of a household being classified as less vulnerable, vulnerable or highly vulnerable. Our model of vulnerability to drought, which is based on the sustainable livelihoods framework, postulates that a household’s vulnerability is a function of the household’s exposure to droughts and the livelihood assets available to the household, which provide capacity to cope with, recover from and adapt to drought and its impacts. Four classes of assets were used in the analysis – natural, financial/economic, human and social. The statistical analysis was performed using ordinal logistic regression analysis. The ordinal logit model is used when the outcome variable is categorized on an ordinal scale, as in this case where vulnerability is ordered as (1) least vulnerable, (2) vulnerable and (3) highly vulnerable (McCullagh and Nelder, 1989). This model is particularly useful in that it can show movement between vulnerability classes, explaining who moves in and out of vulnerability. Following Green (1993), the reduced form of the ordinal logit model is given as: y* = ’Z + ε where y* is the given state of vulnerability, Z is the set of explanatory variables,
3243 J&J Vulnerability
1/11/07
10:23 am
Page 227
Vulnerability in the Sahelian Zone of Northern Nigeria 227
Table 11.3 Distribution of households by vulnerability class Serial Community no.
Latitude Major (°N) occupation
Vulnerability class of Total household Less Vulnerable Highly vulnerable vulnerable
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
11.578 12.612 13.210 11.425 11.908 11.214 12.466 12.501 12.118 13.212 11.098 13.090 11.480 12.185 11.690 11.663 12.429 12.524 11.349 12.290 13.098 11.715 11.300 12.045 11.899 12.537 11.018
8 10 6 8 7 7 8 7 6 9 11 6 7 7 7 8 9 11 9 8 7 7 6 7 5 9 8
10 13 9 11 13 7 9 8 10 9 11 10 9 13 10 9 8 10 10 11 11 8 9 9 10 12 13
13 7 15 12 11 17 15 16 16 12 8 14 14 10 15 14 13 11 11 12 12 16 15 15 16 9 10
31 30 30 31 31 31 32 31 32 30 30 30 30 30 32 31 30 32 30 30 30 31 30 31 31 30 31
208
272
349
828
Buni Yadi Dan Matamachi Kalalawa Maimallamri Daki Takwas Shanga Maguru Kofin Soli Guruma Marnona Tabanni Garin Ahmadu Dabai Auki Madara Zangon Buhari Kwanar Gaki Chanchanda Sara Jinjimawa Damasak Badrama Zandam Chingowa Andarai Kajiji Kubani Total
farming pastoralism pastoralism farming farming farming pastoralism pastoralism mixed pastoralism farming pastoralism farming mixed farming farming mixed pastoralism farming pastoralism pastoralism farming farming mixed farming pastoralism farming
’ is the vector of coefficients to be determined and is a random error with zero mean and unit variance. y* is unobserved; what we do observe is: y = 1, if y* ≤ 2 y = 2, if 2 < y* ≤ 3 y = 3, if 3 ≤ y* The -values (2, 3), referred to as cut-off points, are unknown parameters to be estimated along with . A positive coefficient indicates an increased chance that a respondent with a higher score on the independent variable will be observed in a higher class of vulnerability. A negative coefficient indicates the chances that a respondent with a higher score on the independent variable will be observed in a lower class of vulnerability.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 228
228 Climate Change and Vulnerability
The explanatory variables are grouped as exposure, natural assets, economic assets, human assets and social assets. Following Ziervogel and Calder (2003), we emphasize livelihood assets rather than capitals, which tend to have economic associations. The explanatory variables used in the model are described in the Appendix to this chapter. Variables used to characterize household exposure to drought include the number of drought episodes previously experienced by the household, receipt of drought information, use of drought-resistant practices and existence of a household drought contingency plan. Previous exposure to drought could either increase vulnerability by depleting household assets or decrease vulnerability by increasing a household’s knowledge about and adoption of drought-resilient practices. However, we hypothesize that the greater the number of droughts experienced by a household, the more vulnerable it is. Receipt of drought-related information, either in the form of seasonal forecasts or information about drought-resilient crop varieties or management practices, and current use of drought-resistant crop varieties or livestock breeds are expected to reduce vulnerability. Some households have drought contingency plans and those that do are expected to be less vulnerable than those that do not. Most rural livelihoods in the West African Sahel are substantially reliant on the natural resource base. The natural asset variables used in the logit analysis capture the ability of the natural production system to maintain productivity when exposed to drought. The explanatory variables include self-assessed fertility of the soil, size of household landholdings and the proportion of household landholdings under irrigation. We hypothesize that households with soils that they consider fertile, larger landholdings and more than 50 per cent of their land under irrigation are less vulnerable than those that have infertile soils, small landholdings and less than 50 per cent of their lands irrigated. Another group of variables represents the economic capital base of the household, which is essential for the pursuit of any livelihood strategy (Scoones, 1998). Each household’s main occupation is identified as farming, pastoralism, agro-pastoralism or other. Households that engage in agro-pastoralism are expected to be less vulnerable than the other occupations. Financial asset related variables used in the analysis include identification of the main occupations of the household, number of income sources, level of cash income per capita, access to credit, existence of a savings account and percentage of income spent on household upkeep. Variables were also included for non-financial economic assets such as per capita livestock ownership, per capita cereal harvest, number of years in last ten that the household was selfsufficient in food, use of modern farm equipment, distance of the household to the nearest road suitable for motor vehicles and distance of the village to the nearest market. All of the financial and non-financial asset variables are hypothesized to confer lesser vulnerability to drought. Household composition and structure – gender relations, household cycle, the number of members and the number of potential contributors to the household economy – are crucial to understand households’ and household members’ vulnerability. Variables related to human assets that are included in the analysis include the age and gender of the household head, the highest level
3243 J&J Vulnerability
1/11/07
10:23 am
Page 229
Vulnerability in the Sahelian Zone of Northern Nigeria 229
of education attained by any member of the household, the ratio of household labour units to consumer units and household size. Studies elsewhere have shown that age is a major factor that affects poverty (UNDP, 1998). For example, data from Lebanon and Palestine suggest that households headed by young and old workers – in the early stage and the dispersion years of the domestic cycle – are much more vulnerable to poverty than households whose providers are between 40 and 54 years old. There exist divergent views among researchers regarding the vulnerability of female- versus male-headed households, with some suggesting that female-headed households are more vulnerable (Buvinic and Gupta 1994) and others arguing that female-headed households are as economically and socially viable as male-headed ones (Chant, 1997). Households with more education and more workers per dependent are expected to be less vulnerable, while larger households could be associated with either greater or lesser vulnerability. Social assets also play a vital role in sustaining livelihoods, particularly through benefits coming from networks of community based organizations and relatives. Membership of community organizations provides safety nets at times of crisis and the number of organizations a household belongs to reduces its vulnerability to droughts. Assistance that households receive from community organizations and from family members and relatives outside the community in the form of remittances can also reduce vulnerability.
Results from the Statistical Model of Vulnerability The results of our empirical estimation of the model are presented in Table 11.4. All the exposure variables are significant either at the 95 or 90 per cent confidence levels and have the expected signs. The most significant exposure variable is the use of drought-resistant crop and livestock varieties. Using drought-resistant varieties reduces the likelihood of being in a higher vulnerability class. While many of the households are using drought-resistant varieties, a significant proportion do not. Some of the reasons given by households who do not use them ranged from not knowing where to buy them to not having the services of extension workers to explain how to use them. Another exposure variable that is very significant is the number of drought episodes experienced. The higher the number of drought episodes a household has experienced, the higher the likelihood of belonging to a higher vulnerability class. The odds of a household’s vulnerability being above any level (j) are estimated to be 1.3 times higher for every additional exposure to drought. About 57 per cent (476) of the households reported experiencing droughts, 11 per cent of these in a previous settlement, 57 per cent in their present location and about 33 per cent in both previous and present locations. These included households who had migrated because of droughts in their previous communities only to experience them again in their destination communities. This category of people are very likely to be highly vulnerable as many lost almost all their assets in their previous communities to warrant migration, which is often seen as a last resort.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 230
230 Climate Change and Vulnerability
Table 11.4 Ordinal logit results for determinants of drought vulnerability Variable Exposure Drought episodes experienced Drought information received Use of drought resistant varieties Drought preparedness Natural Asset Soil fertility Cultivated land size Land under irrigation Economic/Financial asset Main occupation of household head Farming Pastoralism Agro-pastoralism Income diversification activities Total annual income Tropical livestock units (TLUs) per household Total crop harvest Use of modern farm equipment Access to credit Existence of savings account Household expenditure Distance of household from major access road Distance of settlement from market Household type i.e. traditional or modern Food self sufficiency Human Asset Age of household head Sex of household head Highest level of education in household Primary Secondary Tertiary Household dependency ratio (labour/consumers) Household size Social Asset Membership of community organizations Assistance from organizations Assistance from family outside the community cut1 cut2 Sample size Log likelihood statistic Pseudo R2
Coefficient
p>|z|
Odds ratio
0.2596443 –0.1936687 –0.3356348 –0.2849131
0.004 0.052 0.002 0.051
1.3 0.8 0.7 0.8
–0.1370299 –0.0421065 –0.5540021
0.034 0.121 0.027
0.9 1.0 0.6
0.4037844 0.0295531 –0.3861327 –0.0318834 –0.5844522 –0.5100647 –0.5330721 –0.3383219 –0.1439984 –0.1275478 0.5833401 0.4973665 0.6683497 –0.0025662 –0.492291
0.039 0.048 0.024 0.055 0.002 0.063 0.046 0.175 0.074 0.051 0.023 0.218 0.049 0.258 0.085
0.0161207 0.3218215
0.188 0.230
1.5 1.0 0.7 1.0 0.6 0.6 0.6 0.7 0.9 0.9 1.8 1.6 2.0 1.0 0.6 1.0 1.0 1.4
–0.1021783 –0.2410995 –0.4805231 0.7230225 0.0026388
0.195 0.047 0.203 0.000 0.093
0.9 0.8 0.6 2.1 1.0
–0.1921738 –0.4529962 –0.5562897 –0.5684523 0.8932861 828 –14925.141 0.401
0.381 0.047 0.033
0.8 0.6 0.6
Note: For a detailed description of the variables in Table 11.4, please refer to the Appendix at the end of the chapter.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 231
Vulnerability in the Sahelian Zone of Northern Nigeria 231
Availability of drought information, including forecasts and information about drought-resilient practices, can be used by households to reduce their exposure to drought impacts. But, as the study shows, about 58 per cent of the population is not aware of the existence of drought and rainfall forecast information. Only 27 per cent regularly receive forecasts, while 15 per cent are aware of but do not receive them. Lack of information, therefore, reduces the coping capacity to drought, hence increasing vulnerability to drought hazards. The population that receives forecast information gets the information from different sources, including radio, television, agricultural extension officers, village heads and farmers’ associations. Those who have access to the forecasts generally consider them satisfactory in terms of reliability and presentation. But often the forecast information is not used. Reasons given for non-use of the forecasts include language constraint, poor user friendliness and lack of timeliness. Greater efforts are needed to make forecasts available to a larger proportion of households, to provide forecasts in local languages and in forms that are more easily understood and used, and to issue forecasts farther in advance of forecasted events.2 While one cannot stop the onset of droughts, it is assumed that the extent of one’s preparedness affects the impacts of the droughts when they do occur. The statistical analysis confirms that those with contingency plans for drought are less vulnerable. The three natural asset variables had the expected signs, although only soil fertility and land irrigation were significant at the 95 per cent confidence limit. Most of the farmers practise rain-fed farming and harvests are strongly affected by rainfall variability. But those that do use irrigation on more than 50 per cent of their land are 40 per cent more likely to fall within a lower vulnerability class compared to households that do not irrigate or irrigate less than 50 per cent of their land. The data collected show that only 13 per cent of the farmlands are irrigated. There are five major sources of water for irrigation in northern Nigeria. These include rain harvesting, ponds and dams, streams and rivers, and wells and boreholes. The few households that have land under irrigation were those that live close to streams and rivers or are engaged in rainwater harvesting. Digging of boreholes and wells is an expensive venture and the technology to pump water from the ground is also expensive for most of these farmers. Efforts to make rainwater harvesting more widely available and the cost of pumping water more affordable could help to reduce vulnerability to drought. With respect to the economic and financial assets, the variables for main livelihood occupation are all significant. Being a farmer increases by 50 per cent the likelihood of belonging to a higher vulnerability class compared to other occupations. On the other hand, households whose main occupation was agro-pastoralism are 30 per cent more likely to belong to lower classes of vulnerability than those with ‘other’ occupations. Pastoralism is positively associated with vulnerability, but the odds ratio is statistically insignificant. Farming is the most common livelihood in the study area, followed by pastoralism. Pastoralism is the main economic activity in 33 per cent of the sampled villages and is the main livelihood for about 37 per cent of the sam-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 232
232 Climate Change and Vulnerability
pled households. The most common domesticated animals in the research area are cattle, poultry, goats and sheep. Cattle are generally kept at the commercial level, while the smaller animals – goats, sheep and poultry – are kept at the subsistence level. When expressed in tropical livestock units (TLUs), cattle are the most important type of animal among households, followed by goats, poultry and sheep. The poultry include chickens, ducks, turkeys and guinea fowl. Virtually every household keeps poultry and some goats and sheep. However, the distribution of livestock in the sample was very unequal in the communities and among the various vulnerable groups, with the less vulnerable households owning more livestock per household than the other groups. Livelihood diversification has become a well-accepted adaptation strategy in the Sahel. Households are less vulnerable to climatic stress if they have multiple sources of livelihood to fall back on in times of scarcity. The survey obtained information from households regarding all their non-agricultural income-generating activities and assigned weights to them based on the reliability and the income generated by each activity as reported by the respondents. The assumption is that the greater the number of non-agricultural income-generating activities, and the more income generated from them, the less vulnerable a household was. Income diversification activities identified in this study are separated into four main types: crafts, trade, labouring and hunting/gathering. Labouring on other people’s farms is a widespread phenomenon and to many a primary method of supplementing their own farm income. Labouring is the largest income diversification activity in the region, followed by trading. Migration on its own was not considered as an income diversification activity, rather what the migrant does to generate income. Many of the migrants go outside their immediate communities to work on other people’s farms or work in non-farm income generating activities. Hunting/gathering is the least practised income diversification activity in the region. Income and per capita crop harvest were all significant at the 95 per cent confidence level, and households with higher values of these were less likely to belong to higher vulnerability classes. Crops cultivated in the area include maize, millet, sorghum, beans and rice. About 28 per cent of households harvested less than 20 bags of cereals, 38 per cent harvested between 20 and 60 bags, 23 per cent harvested between 60 and 150 bags, while 11 per cent harvested more than 150 bags of cereals. Households that spent more than 70 per cent of their income in running the household were 1.8 times more likely to belong to a higher class of vulnerability. Also, longer distances from the house to a major road (inaccessibility) increased by 80 per cent the likelihood of belonging to higher classes of vulnerability. The same holds true for market distance, where the odds of belonging to a higher vulnerability class is estimated to be 2.0 times for every additional kilometre increase in distance. It is instructive to note the importance that distance to market plays in farm household vulnerability. Many of the farmers harvest their crops and cannot transport them to the markets because of distance. This inaccessibility to markets makes them receive suboptimal prices for their crops, reducing their income and further exacerbating
3243 J&J Vulnerability
1/11/07
10:23 am
Page 233
Vulnerability in the Sahelian Zone of Northern Nigeria 233
their vulnerability. Investment in rural infrastructure is needed to improve access to markets, which can improve farmers’ livelihoods and lessen their vulnerability to drought. With regards to human assets, the household dependency ratio is the most significant variable. A household that has more mouths to feed than pairs of hands to cultivate the land is very vulnerable as the cycle of subsistence is reinforced. Regarding education, only secondary education is significant, with those with secondary education more likely to be less vulnerable than those without any education. Household size and age and gender of household head are not statistically significant determinants of vulnerability. Of the social asset variables, assistance from community organizations and from family are significant at the 95 per cent confidence level. Those who receive assistance are less vulnerable than those who do not. This is where migration plays an important role in reducing vulnerability. The data show that remittances decline the farther relatives are from home. For example, there is more assistance from those within the villages than from those outside the village but within the country, and there is more assistance from those outside the community but within the country than from those outside the country. Membership of community organizations is not statistically significant. About 60 per cent of the population in the study area do not belong to any group or organization. Those that do are members of organizations such as farmers’ associations, trade unions, aid groups and religious associations. The most common type of organization among the rural communities under study is the farmers’ association. Those that belong to some organization have varied reasons for being members, such as for economic progress and membership benefits, for assistance from government, to praise and worship God, and to assist one another in the community.
Conclusion Drought is a persistent problem in the West African Sahel and has contributed to the under-development of the region. Many of the sampled households have experienced and been impacted by multiple droughts. But vulnerability is not a physical entity that can be seen or measured directly. It is a relative term differentiating between socioeconomic groups or regions, rather than an absolute measure of a physical characteristic. In our study villages, there was no agreement on what is meant by vulnerability to drought. Rather, households experience and perceive risks in terms of the consequences of the occurrence of events such as drought. For example, most households are more concerned about the risk of hunger or shortage of water for domestic use, which can be triggered by drought, than they are about the risk of drought itself. The perceived risks, filtered through various household assets and conditions, form the basis for households’ understanding of vulnerability. These fears or vulnerabilities vary among households and are shaped by economic, social and environmental processes. Because of the lack of uniformity in meas-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 234
234 Climate Change and Vulnerability
uring vulnerability, the usual practice in assessing vulnerability is that the analyst or decision maker assigns thresholds of vulnerability that warrant specific responses. We took a different approach, combining top–down and bottom–up methods to construct an index of vulnerability from indicators and weights derived from the literature and through stakeholder analysis. Using a simple self-assessment procedure in combination with statistical classification, households in the study were classified as highly vulnerable, vulnerable or less vulnerable. About 25 per cent of the households classified themselves as less vulnerable, 33 per cent as vulnerable and about 42 per cent as highly vulnerable. Numerous potential explanatory variables of vulnerability levels were identified with respondents. Some of these variables have direct impacts on vulnerability (for example, size of crop harvest, dependency ratio, livestock ownership, livelihood diversification and drought preparedness), while others act indirectly on household vulnerability. We consider this method of vulnerability assessment to be more robust as it captures both objective and subjective vulnerabilities and incorporates the perceptions of the local people. A multivariate analysis was conducted to statistically test which factors explain the vulnerability of rural farm households to droughts. This was based on a conceptual framework that postulates that vulnerability is a function of exposure and resilience, where resilience is determined by the availability to the household of natural, economic, human and social assets. Exposure variables found to be significant determinants of household vulnerability include the number of droughts a household was exposed to, the availability of drought information for planning and the use of drought-resistant varieties. Of the household assets, household dependency ratio and income are the most significant, but many others contribute to shaping the household’s vulnerability. These include soil fertility, use of irrigation, main source of livelihood, diversification of income, distance from markets, and assistance from community organizations and family. The needs of rural households in the Sahelian zone of Nigeria are still very basic, and efforts to build resilience to drought in these communities should first address issues of access to food and water, and to help households to build the assets that are needed to sustain their livelihoods. Concerns about food, water and livelihood assets cut across pastoralist groups and sedentary farmers. Issues such as human diseases and conflicts, though often considered important in the literature, did not feature very prominently among the risks perceived by respondents.
Notes 1
The study in Nigeria was conducted as part of a larger project that included research on vulnerability to drought in Mali. Six hundred questionnaires similar to the ones used in the Nigerian analysis were administered in 16 communities in Mali between April 2003 and March 2004. Information about the full project can be found in the AF92 Project ‘Final Report to Assessments of Impacts and Adaptation to Climate Change’ (forthcoming; www.aiaccproject.org).
3243 J&J Vulnerability
1/11/07
10:23 am
Page 235
Vulnerability in the Sahelian Zone of Northern Nigeria 235 2
In Chapter 10 of this volume, James Adejuwon presents a more complete analysis of the provision and use of seasonal weather forecasts and provides further recommendations for their improvement.
References Adger, W. N. (1999) ‘Social vulnerability to climate change and extremes in coastal Vietnam’, World Development, vol 27, pp249–269 Adger, N. and M. Kelly (1999) ’Social Vulnerability to Climate Change and the Architecture of Entitlements’, Mitigation and Adaptation Strategies for Global Change, vol 4, pp253–266 Apeldoorn, G. J. van (1981) Perspectives on Drought and Famine in Nigeria, George Allen and Urwin, London Bhalme, H. N. and D. A. Mooley (1980) ‘Large-scale drought/floods and monsoon circulation’, Monthly Weather Review, vol 108, pp1197–1211 Brooks, N. (2004) ‘Drought in the African Sahel: long-term perspectives and future prospects’, Tyndall Centre Working Paper no 61, Tyndall Centre for Climate Change Research, Norwich, UK Buvinic, M. and G. R. Gupta (1994) Poor Woman-Headed Households and WomanMaintained Families in Developing Countries: Views on a Policy Dilemma, International Center for Research on Women, Washington, DC, US Chant, S. (1997) Women Headed Households: Diversity and Dynamics in the Developing World, Macmillan, London Claussen, M., V. Brovkin, A. Ganopolski, C. Kubatzki and V. Petoukhov (2003) ‘Climate change in northern Africa: The past is not the future’, Climate Change, vol 57, pp99–118 Dai, A., P. J. Lamb, K. E. Trenberth, M. Hulme, P. D. Jones and P. Xie (2004) ‘The recent Sahel drought is real’, submitted to International Journal of Climate Change, available online from www.cgd.ucar.edu/cas/adai/publication-dai.html Dietz, A. J., R. Ruben and A. Verhagen (2001) ‘Impacts of climate change on drylands with a focus on West Africa’, Report No. 410 200 076, Dutch National Research Programme on Global Air Pollution and Climate Change, ICCD, Wageningen FAO (2005) ‘West African food crisis looming: Millions at risk of food shortages’, news release, 20 June, Food and Agriculture Organization of the United Nations, Rome Greene, W. H. (1993) Econometric Analysis, Second Edition, Macmillan, New York Haarsma, R., F. Selten, N. Weber and M. Kliphuis (2005) ‘Sahel rainfall variability and response to greenhouse warming’, Geophysical Research Letters, vol 32, pL17702 Hoerling, M., J. Hurrell, J. Eischeid and A. Phillips (2006) ‘Detection and attribution of 20th century northern and southern African monsoon change’, Journal of Climate, vol 19, pp3989–4008 Hulme, M., R. Doherty, T. Ngara, M. New, and D. Lister (2001) ‘African climate change: 1900–2100’, Climate Research, vol 17, pp145–168 Huntingford, C., F. Lamber, J. Gash, C. Taylor and A. Challinor (2005) ‘Aspects of climate change prediction relevant to crop productivity’, Philosophical Transactions of the Royal Society: B, vol 360, no 1463, pp1999-2009 Ingram, K. T., M. C. Roncoli and P. H. Kirshen (2002) ‘Opportunities and constraints for farmers of West Africa to use seasonal precipitation forecasts with Burkina Faso as a case study’, Agricultural Systems, vol 74, pp331–349 Janowiak, J. E. (1988) ‘An investigation of interannual rainfall variability in Africa’, Journal of Climate, vol 1, pp241–255 Jenkins, G., A. Gaye and B. Sylla (2005) ‘Late 20th century attribution of drying trends in the Sahel from the Regional Climate Model (RegCM3)’, Geophysical Research
3243 J&J Vulnerability
1/11/07
10:23 am
Page 236
236 Climate Change and Vulnerability Letters, vol 32, pL22705 Kamga, A., G. Jenkins, A. Gaye, A. Garba, A. Sarr and A. Adedoyin (2005) ‘Evaluating the National Center for Atmospheric Research Climate System model over West Africa: Present day and the 21st century A1 scenario’, Journal of Geophysical Research-Atmospheres, 110(D3), D03106. Lamb, P. J. (1982) ‘Persistence of sub-Saharan drought’, Nature, vol 299, pp46–47 Lebel, T., F. Delclaux, F. Le Barbe and J. Polcher (2000) ‘From GCM scales to hydrological scales: Rainfall variability in West Africa’, Stochastic Environmental Research and Risk Assessment, vol 14, pp275–295 L’Hôte, Y., G. Mahé, B. Somé and J. P. Triboulet (2002) ‘Analysis of a Sahelian annual rainfall index from 1896 to 2000: The drought continues’, Hydrological Sciences Journal, vol 47, pp563–572 Maynard, K., J. F. Royer and F. Chauvin (2002) ‘Impact of greenhouse warming on the West African summer monsoon’, Climate Dynamics, vol 19, pp499–514 McCullagh, P. and J. Nelder (1989) Generalized Linear Models, second edition, Chapman and Hall, London Mitchell, J. F. B., T. C. Johns, W. Ingram and J. A. Lowe (2000) ‘The effect of stabilising the atmospheric carbon dioxide concentrations on global and regional climate change’, Geophysical Research Letters, vol 27, no 18, pp2977–2980 Mortimore, M. (1998) ‘Roots in the African dust: Sustaining the sub-Saharan drylands’, Cambridge University Press, Cambridge, UK Mortimore, M. J. and W. M. Adams (2001) ‘Farmer adaptation, change and “crisis” in the Sahel’, Global Environmental Change, vol 11, pp49–57 Ojo, O. and L. Oyebande (1985) ‘Trends in occurrences and severity of droughts in Nigeria’, in Ecological Disasters in Nigeria: Drought and Desertification, Proceedings of the National Workshop on Ecological Disasters, Kano, pp19–53 Okechukwu, G. C. (1997) ‘Survey of drought history in northern Nigeria‘, sub-project report no 8, Jos–McMaster Drought and Rural Water Use Project, University of Jos, Nigeria and McMaster University, Canada Quinn, C. H., M. Huby, H. Kiwasila and J. C. Lovett (2003) ‘Local perceptions of risk to livelihood in semi-arid Tanzania’, Journal of Environmental Management, vol 68, pp111–119 Scoones, I. (1998) ‘Sustainable rural livelihoods: A framework for analysis’, IDS Working Paper 72 Tarhule, A. and P. J. Lamb (2003) ‘Climate research and seasonal forecasting for West Africans: Perceptions, dissemination, and use’, Bulletin of the American Meteorological Society, vol 84, pp1741–1759 UNDP (1998) ‘Mapping of living conditions in Lebanon: An analysis of the housing and population database, 1998’, United Nations Development Programme/ Ministry of Social Affairs Vincent, K. (2004) ‘Creating an index of social vulnerability to climate change for Africa’, Tyndall Centre Working Paper 56, Tyndall Centre for Climate Change Research, Norwich, UK Wang, G. L. and E. A. B. Eltahir (2002) ‘Impact of CO2 concentration changes on the biosphere-atmosphere system of West Africa’, Global Change Biology, vol 8, pp1169–1182 Ziervogel, G. and R. Calder (2003) ‘Climate variability and rural livelihoods: Assessing the impact of seasonal climate forecasts in Lesotho’, Area, vol 35, no 4, pp403–417
3243 J&J Vulnerability
1/11/07
10:23 am
Page 237
Vulnerability in the Sahelian Zone of Northern Nigeria 237
Appendix Variables in the ordinal logit model of determinants of drought vulnerability in northern Nigeria Variable
Variable description
Min
Max
Mean Std Dev
Exposure Droutepi
Number of drought episodes experienced
0
4
2.18
1.19
Droutinfo
Dummy variable: 1 if household has ever received formal drought-related information, 0 otherwise
0
1
0.15
0.34
Droutvar
Dummy variable: 1 if household uses drought-resistant varieties, 0 otherwise
0
1
0.57
0.41
Droutpln
Dummy variable: 1 if household has any drought preparedness plan, 0 otherwise
0
1
0.49
0.36
0
1
0.42
0.56
0.1
3.5
0.84
1.22
0
1
0.34
0.46
Categorical variable representing the main occupation of the household head: 1 if farmer, 2 if pastoralist, 3 agro-pastoral, 4 other occupations
1
4
1.98
1.72
Weighted number of income diversification activities in the household/consumer units
0
3.1
1.05
0.95
2.3
10.6
4.7
2.41
Natural Assets soilfert
Dummy variable for self-assessed fertility of farmlands: 1 if fertile, 0 otherwise
landsize
Total size of land cultivated/number of consumer units in the household (ha)
irrigland
Dummy variable for proportion of land under irrigation: 1 if more than 50% of the land is under irrigation, 0 otherwise.
Economic/Financial Assets mainocc
incomdiv income
Total annual income to the household from all sources/consumer units (1000 naira)
TLU
Tropical livestock unit in the household/number of consumer units.
crophvst
Crop total bags of cereals harvested/consumer unit
farmquip
0
11.2
5.3
4.33
1.2
24.6
11.3
5.44
Dummy variable for use of modern farm equipment: 1 if household uses modern farm equipment, 0 otherwise.
0
1
0.32
0.51
credit
Dummy variable: 1 if household has access to credit, 0 otherwise
0
1
0.21
0.65
banksav
Dummy variable: 1if any household member has a savings account, 0 otherwise
0
1
0.27
0.44
3243 J&J Vulnerability
1/11/07
10:23 am
Page 238
238 Climate Change and Vulnerability Variable
Variable description
Min
Max
hhexpen
Mean Std Dev
Dummy variable for proportion of household expenditure spent in running the household: 1 if more than 70%, 0 otherwise
0
1
0.64
0.44
roadist
Distance of home to major access road (km)
0.01
3.8
1.02
1.27
Markdist
Distance of settlement to major market (km)
0.5
12.8
6.3
2.41
foodsuff
Food self sufficiency: number of years in the past 10 years that household produced enough food to feed itself
0
10
4.23
5.1
Human Assets Age
Age of household head in completed years
15
83
49.26
22.4
Sex
Dummy variable for sex of household head: 1 if male, 0 otherwise
0
1
0.96
0.12
educat
Categorical variable for highest level of education completed by any member of the household: 0 for none, 1 for primary, 2 for secondary and 3 for tertiary 0
3
1.02
1.79
depend
Household dependency ratio (labour units/consumer units) 0.24
0.88
0.61
0.37
hhsize
Number of persons in the household
1
31
12.9
9.3
Number of community organizations household belongs to Receives assistance from organizations
0 0
4 1
1.5 0.45
1.3 0.24
Has family member outside the community that sends money and other forms of assistance
0
1
0.23
0.38
Social Assets comorg orgasist famasist
3243 J&J Vulnerability
1/11/07
10:23 am
Page 239
12
Livelihoods and Drought in Sudan Balgis Osman-Elasha and El-Amin Sanjak
Introduction Drought is a primary agent of famine in Africa today, where the agricultural production environment is under increased stress from many factors. Studies of past environmental change and the northern African archaeological record indicate just how variable climatic and environmental conditions are in the Sahel-Sahara zone on timescales of centuries to millennia. Such studies also illustrate the sensitivity of rainfall in this region to hemispheric and global-scale changes in climate (Brooks, 2006). In Africa, where around half of cultivable land is arid and semi-arid, about 65 per cent of the croplands and 30 per cent of pastureland have been affected by degradation, with resultant declines in crop yields and food insecurity. High seasonal rainfall variability is endemic to the arid regions of Sudan and the resulting droughts have affected many inhabitants, who live with constant vulnerability and possess weak or poor coping ability to deal with hunger, famine, dislocation and material loss (OsmanElasha, 2006). However, such vulnerability is not caused by climate variability or climate change alone. The highly variable climate is an underlying characteristic, a chronic state of the region’s fragile livelihood systems. Given the constant exposure to this state of fragility under frequent harsh conditions, most human inhabitants have developed traditional knowledge regarding occurrences and likely consequences of extreme climatic events. However, while they possess the know-how to adapt and modify their livelihood systems to buffer against potential disasters and prepare themselves with whatever means at their disposal for these anticipated threats, more recently, particularly since the 1980s, some human livelihood systems seem to have lost their ability to adapt to or recover from sustained droughts (Teklu et al, 1991). Hence, there is an urgent need to examine the underlying causes and the factors that determine enhanced exposure and declining ability to recover from extreme climate events. Consequently, there is also a need to find answers to questions such as ‘How do people manage their livelihoods under current vulnerability?’ and
3243 J&J Vulnerability
1/11/07
10:23 am
Page 240
240 Climate Change and Vulnerability
‘How could they avoid potential vulnerability in the face of anticipated enhanced future climate variability and change?’ This chapter attempts to address the above questions by exploring the relationship between climate variability and the livelihoods of rural agricultural and pastoral populations in this semi-arid Sahelian region of Africa. While focusing on climate-related vulnerabilities, it also considers the socioeconomic context of the inhabitants by examining the many other stresses they face and how these contribute to increasing their vulnerability. Climatic events in these parts of Africa, particularly droughts, trigger frequent subsistence crises, sharply increasing crop failures. An example is the drought that occurred in the whole of Sahelian Africa in 1984, resulting in widespread hunger and famine that has been ascribed to a combination of climate variability and socioeconomic factors. According to Braun et al (1998), the genesis of food crises in this region is a result of the interaction between environmental and socioeconomic factors, in both the short and the long terms, and a failure of policy to deal with them. Most survivors of the 1984 drought were left with fewer assets and with an increasingly risky agricultural income base that offers little buffer against future crises (Braun et al, 1991; Teklu et al, 1991). This situation was also a striking example of the collective impact of multiple factors (poor infrastructure, lack of capacity, illiteracy and under-development) undermining the coping abilities and resilience of entire populations, rendering them helpless, limiting their coping abilities, and resulting in extreme suffering and large scale migration away from the region.
Vulnerability to climate variability and change Vulnerability to climate change is generally understood to be a function of a range of biophysical and socioeconomic factors. According to McCarthy et al (2001), vulnerability may be characterized as a function of three components: adaptive capacity, sensitivity and exposure. Household vulnerability is defined as the capacity to manage shocks. Adaptive capacity describes the ability of a system to adjust to actual or expected climate stresses, or to cope with the consequences. It is considered ‘a function of wealth, technology, education, information, skills, infrastructure, access to resources, and stability and management capabilities’ (McCarthy et al, 2001, p8). Sensitivity refers to the degree to which a system will respond to a change in climate, while exposure relates to the degree of climate stress on a particular system, represented as a function of either long-term change in climate conditions or changes in climate variability, including the magnitude and frequency of extreme events. In the most general sense, the term ‘climate variability’ is often used to denote deviations from the mean of climate statistics over a given period of time. In this analysis, the following scales of variability are recognized: • • •
micrometeorological variability: from fractions of a second to several minutes; mesometeorological variability: from several minutes to several hours; synoptic variability: from several hours to two–three weeks; and
3243 J&J Vulnerability
1/11/07
10:23 am
Page 241
Livelihoods and Drought in Sudan 241
•
climate variability: from three weeks to several decades.
Variability on such a scale characterizes internal climatic oscillations, climate variability or climate fluctuations (Gruza and Rankova, 2004). Variability could also cover isolated extreme events or catastrophic weather conditions, such as floods, droughts or storms. Severe drought has occurred throughout the recorded history of Sudan. However, the Sudano-Sahelian zone has experienced a general decline in rainfall since the late 1960s (Brooks, 2006). In western Sudan, annual rainfall variability increased from 16 per cent in the 1960s to 21 per cent in the 1970s and 32 per cent in the 1980s. Drought in many cases has been followed by famine. The two greatest famines since 1684, when the historical record begins, are those of 1888–89 and 1984–85, both triggered by consecutive years of poor rain and resulting massive crop failures (Teklu et al, 1991). This study also revealed that drought-associated famine in Sudan tended to occur in the arid and semi-arid zones of the west and northeast, where the resource-base is poor, rainfall generally low and erratic, the income and asset base of the population thin and variable, and the agricultural environment marginal. In this chapter we focus on the exposure of people to climate variability as documented in the historical records of the meteorological department and as observed by the local communities. Indicators of sensitivity and adaptive capacity are used to represent vulnerability to climate variability, given that local inhabitants have historically been coping with periodic droughts and rainfall variability. Based on this vulnerability profile, in this study we assumed that current exposure to climate variability affects the community’s livelihoods such that communities respond to climate changes by trying to employ measures that enhance their adaptive capacity. However, vulnerability is not a passive state, but a dynamic process. Consequently, vulnerable people are caught up in a state of continuous struggle, attempting to lessen their vulnerability and gaining more resilience. Since the future state of the regional climate is uncertain, there is no guarantee that a specific coping mechanism that is appropriate today will still be appropriate in 20 or 30 years. Nevertheless, we assume that the social and individual capacity for community action, flexible management of natural resources, and diversification of livelihoods are valuable strategies for years to come (Osman-Elasha, 2006).
Impacts of Climate Variability The severe drought of 1984 is considered one of the most devastating in the history of Sudan. Poor rainfall contributed to low and variable food production. For instance, sorghum and millet production in Northern Kordofan declined by 92 and 86 per cent respectively compared with the average of 1974–81 (Osman-Elasha, 2006). The decline in cereal production led to rapid price increases that spread throughout the country. Large losses of cattle and camels among the pastoralists also resulted in great hardships for nomads. At the household level, the large drop in production (crops and rangeland) trans-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 242
242 Climate Change and Vulnerability
lated into large reductions in farm employment and income. The drop in income, coupled with rapidly increasing food prices, resulted in a severe and widespread decline in purchasing power and contributed to extreme displacement of the rural population as hundreds of thousands became destitute and moved out of their villages. Farmers and livestock herders sought employment away from their traditional lands and sold their assets in depressed markets to maintain their food entitlements. The dramatic deterioration in purchasing power and in the level and quality of food consumption, combined with a high incidence of disease, translated into an increased incidence of severe child malnutrition (Osman-Elasha, 2006; Teklu et al, 1991). The largest groups of affected households were the poor and those families consisting of the elderly or missing able adult males. By mid-1985 more than 700,000 people in Kordofan and about 800,000 in Darfur were at high nutritional risk. These two regions experienced significant demographic change resulting from high death rates and large-scale emigration (Osman-Elasha, 2006; Teklu et al, 1991). Climate variability had a similar impact in eastern Sudan, where frequent occurrences of drought and consequent famine in the Red Sea hills were largely the norm during the 20th century. The long-term drought and famine of the 1980s brought devastating effects, shattering the traditional pattern of natural short-term recovery and causing a major depopulation of the herds of the Beja tribe of Sudan, with losses estimated at 80 per cent of their animal wealth. A complex of human and other factors combined to produce a situation wherein the area available for the Beja’s livestock rearing was rapidly diminishing, coupled with a complete failure of all their livelihood systems (Osman-Elasha, 2006).
Vulnerability Assessment Most of the information contained in this chapter was drawn from three case studies conducted plus a number of studies and papers on the subject of impacts and adaptation to climate variability and change in Africa. This project attempted to address the local vulnerability of agro-pastoral people to global environmental changes in the context of sustainable livelihoods. The case studies examined the past experiences and coping capacities of the rural population in the face of drought, in order to assess the likely impacts of future climate change. A system called the ‘livelihood asset status tracking system’ was used to measure the changing asset base in five livelihood capitals, which could, in turn, serve as a proxy for determining the impacts on households. Participatory interviews conducted by a team of researchers is the main method used in this assessment. The process involves group discussions, brainstorming and clustering of criteria, field testing and validation. The main objective is to evolve ‘word pictures’ for livelihood assets tracking. The word picture is a method for constructing verbal descriptions of asset status. Such word pictures can depict ‘worst off’ and ‘better off’ households and also intermediate positions. To measure asset status for each of the five capitals available to a household, a locally meaningful scale of stages is considered from the
3243 J&J Vulnerability
1/11/07
10:23 am
Page 243
Livelihoods and Drought in Sudan 243
worst known situation to the best, while maintaining a balance between aspects related to production, equity and sustainability. An illustration of the word picture obtained from the case studies is given in Table 12.1. The illustration covers two word pictures of worst off and better off households. Table 12.1 Word picture of households’ access to/use of livelihood capitals Livelihood capital
Worst Picture under drought condition without interventions
Best picture under drought condition with intervention
Natural capital
Degraded land with v. low fertility and low productivity of grains Little water pools insufficient for animals No fodder for livestock Weak and fewer number of animals No access to forest produce
More fertile soil; abundant amount of fodder More moisture retention power; more produce from land; fertile soil that sustains growth of diverse crops, vegetables and fodder; access to forest produce and fuel wood from shelterbelts (some have government permit to grow opium); has many fruit trees
Physical capital
No water pumps or irrigation facilities Poor human health services (few clinics) Poor extension services Poor vet services Poor roads and marketing places Grain stores with small capacity
Irrigation facilities available round the year Availability of extension services Improved health and vet services Large-capacity stores for excess grains Improved marketing Availability of improved inputs and spare parts
Financial capital
Lack of options for income generation Unstable income condition No credit system granted to individuals and no savings
Availability of home grown food throughout the year; more livestock, high returns from livestock Better income levels and stability Introduction of revolving funds and better savings
Human capital
Poor skills (no training or education opportunities) Poor housing type
Improved skills and better access to extension, health, education, training and veterinary services
Social capital
Poor managerial skills Poor organizational set-up Poor participation in the decision making process
Better ability to manage natural resources (pasture, land, water, livestock etc.) Better organizational set-up (local village committees) Improved participation in the decision making process More membership in civil society organizations
3243 J&J Vulnerability
1/11/07
10:23 am
Page 244
244 Climate Change and Vulnerability
The study involved data collection and analysis of ‘resiliance indicator’ data to show how certain sustainable livelihood measures were effectively implemented and supported in order to ensure their lasting impact. In spite of the variation across the five capitals assessed across the three case studies, our study succeeded in reflecting the state of vulnerability of the communities before the employment of interventions as well as in illustrating how some environmental management/sustainable livelihood measures have succeeded in increasing the community’s overall resilience.
Coping deficit Our analysis has shown that, in their attempts to cope with climatic and other related stresses, people may adopt specific measures that further aggravate their vulnerabilities and undermine their productive assets, for example, through overexploiting their over-stressed natural resource base. This is illustrated by the farmers in Kordofan region, who faced decreased land capacity during the 1984 drought. When the per-hectare yields dropped so low that overall production could no longer sustain the households, people tried to compensate for the lost productivity by increasing the agricultural areas and expanding into marginal rangelands. The less productive the land, the more the farmers expanded. This situation eventually led to a situation of high tension and conflict with other tribes and land-users, especially the pastoralists. At the national level, government policies that encourage expansion of agricultural land through mechanized farming to improve the food security situation and safeguard against food shortages under climate variability, often negatively impact the pastoralists by forcing them onto marginal lands and exacerbating the problems. According to Tekulu et al (1991) and Braun et al (1998), the over-expansion of agriculture and consequent encroachment of pastoralists into historically marginal areas reflect the failure to appreciate the nature of long-term (in other words multi-decadal scale) climatic variability in the region. Well-meant policies have had inadvertent and catastrophic impacts on livelihoods in the arid parts of Sudan, including massive loss of life and livestock, destruction of communities and livelihood systems, and massive societal disruption on a regional scale. This is a typical condition in Sahelian Africa highlighted by Thébaud and Batterby (2001), who mention that the expansion of agriculture during the wet 1950s and 1960s and a shift to agro-pastoralism pushed pastoralists into more marginal areas and led to a breakdown in the networks connecting herders and farmers, contributing to conflict between these groups. Brooks (2006) writes that ‘Pushed into more marginal areas, and with their access to pasture regulated and restricted by both colonial and postcolonial government, pastoral communities became more vulnerable to drought’. Yet this is not always the case, and some local and national strategies have proven viable in producing some positive tangible impacts on the communities’ livelihoods (Osman-Elasha, 2006). The results from the three case studies emphasized this and suggest that planners adopt and build on environmental management/sustainable livelihoods, emphasizing the need for the develop-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 245
Livelihoods and Drought in Sudan 245
ment planning and adaptation to proceed by progressively improving on those proven coping strategies.
The sustainable livelihood approach Here we define livelihood as the capabilities, assets and activities required for a means of living. Livelihood is sustainable when one can cope with and recover from stresses and shocks and maintain or enhance one’s capabilities and assets both now and in the future, without undermining the natural resource base (Carney, 1999). Vulnerability in this context could be taken as the risk that the household’s capitals will fail to buffer against impacts of drought. According to Goldman (2000), the sustainable livelihoods approach sees poverty as vulnerability to shocks and seeks to reduce vulnerability by building on the livelihood assets (natural, physical, financial, human and social) of households. The sustainable livelihood approach provides the ‘bottom–up tool’ for assessing vulnerability as perceived by the local people themselves. Using this approach we were able to account for climate variability events and the social and economic conditions to model the processes that shape the negative consequences on the livelihoods of the studied communities. In this exercise, we got answers to questions such as ‘Who are the most vulnerable groups?’, ‘To which stresses are they exposed?’, ‘What sustainable livelihoods/environmental management strategies have they used to improve their coping capacity?’ and ‘What other factors contributed to improving their coping capacity?’ We also gained an insight into their views regarding future adaptation options. Some of the benefits of using a sustainable livelihoods framework are that it ‘provides a checklist of important issues and sketches out the way these link to each other, draws attention to core influences and processes, and emphasizes the multiple interactions between the various factors which affect livelihoods’ (DFID,2004). In the context of this analysis, the sustainable livelihood analysis has been used to include considerations of issues related to impacts of policies and institutions on the coping capacity of the rural communities in an attempt to show how macro-, meso- and micro-level policies – ranging from land-tenure and marketing policies to taxes – have played major roles in both security and vulnerability through their effects on different aspects of livelihood capitals. The sustainable livelihoods approach might assist in solving some of the problems inherent to vulnerable regions of Sudan. Our analysis showed that in some cases, vulnerability has been lessened through access to alternative income generating opportunities and productive resources such as land, irrigation, credit, fertilizers and improved seed. Our study also showed that vulnerability to climate variability and change can be reduced through certain carefully selected policies and institutional set ups.
The Case Studies Our research was conducted in Sudan. Three case studies representing different community settings in eastern, west-central and western parts of Sudan
3243 J&J Vulnerability
1/11/07
10:23 am
Page 246
246 Climate Change and Vulnerability
were undertaken. These case studies were selected partly on the basis of advance knowledge that the communities represent successful examples of sustainable livelihood measures for reducing vulnerability to drought. The first case study focused on a UNDP-funded project, ‘CommunityBased Rangeland Rehabilitation for Carbon Sequestration and Biodiversity’, in a rural area of western Sudan. This UNDP project was introduced into the area after the severe drought period that affected the Sahel zone during the 1980s. The study area lies within Kordofan State in west-central Sudan and is a part of Sahelian Africa that has undergone a general decline in rainfall since the late 1960s. Between 1961 and 1998, episodes of drought have inflicted the region with varying degrees of severity. This period witnessed two widespread droughts, during 1967–1973 and 1980–1984, the latter being the more severe. Available records show that drought episodes were intense and long lasting, resulting in enhanced vulnerability of the local population, particularly during and after the droughts mentioned above (Osman-Elasha, 2006). The droughts resulted in severe impacts in the region, including chronic poverty, socioeconomic marginalization and food insecurity, leading to a rural development crisis which required integrated and cross-sectoral responses (Warren and Khogali, 1991). In response to this, many viable interventions were identified, providing a starting point for addressing adaptation needs through issues of rural development and poverty reduction. The second case study is an SOS Sahel intervention called the ‘Khor Arba’at Rehabilitation Programme’. The study area is located in the Red Sea State in northeastern Sudan, about 50km north of Port Sudan, the state capital. The Red Sea State, one of Sudan’s 26 states, falls between latitudes 17°00 and 23°01N and longitudes 33°14 and 38°32E in the extreme northeastern part of Sudan. Administratively, Arba’at is part of the Red Sea locality, one of the four localities comprising the Red Sea State. The area lies in the catchment of the Khor (small stream) Arba’at, after which it is named. The Khor Arba’at drains a catchement of 4750km2 (Bashir, 1991) and flows in a west–east direction from the Red Sea Hills, where it originates, to the Red Sea. The region is generally characterized by relative isolation and harsh terrain, highly variable rainfall with recurrent spells of drought, small area of cultivable land and low population density. The people who live here are mainly the Beja pastoralist and agro-pastoralist tribal groups, some of whom are transboundary tribes moving between Sudan, Ethiopia and Eritrea. Rainfall is highly variable, but averages recorded between 1900 and 1980 range between 26mm and 64mm per annum. Both rainfall amounts and geographical distribution show a high degree of variability that generally increases from south to north. Available records indicate that over the last four decades, the general trend has been negative, with frequent and successive droughts. The rainwater received in Arba’at is too small to support cultivation, except in exceptional cases. For that reason, since the 1920s agriculture has depended on the Khor Arba’at. Although rainfall in the area has rarely been of significant benefit to agricultural production, it does support the natural growth for livestock to graze on, though even that resource has declined since the 1970s, reaching a minimum in 1984. The volume of run-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 247
Livelihoods and Drought in Sudan 247
off (Khor water) varies considerably between a minimum of 168mm3 (90 per cent probability of occurrence) and a maximum of 1662mm3 (10 per cent probability of occurrence), with high evaporation losses (Osman-Elasha, 2006). The hilly nature of the topography and the absence of aquifers due to the basement complex formation of the base rock make surface run-off the only source of fresh water in the Red Sea area. Of the total area of the Arba’at deltaic fan, the arable lands of Arba’at are estimated at 23,215 feddans (9750ha). Of these, about 9285 feddans (3900ha) can readily sustain irrigation agriculture. The third case study focused on water harvesting techniques as a coping mechanism in the face of climate variability and change in the North Darfur State. North Darfur is situated in western Sudan, on the northern transitional margin of the Intertropical Convergence Zone. Consequently, most of the area is deficient in water even in the wettest months of July to September (80 per cent of the rainfall). During June, the hottest month, temperatures regularly reach over 45°C and in January, the coldest month, temperatures reach 18°C. North Darfur is one of the most drought-affected regions of the Sudan. The drought years of 1983–85 greatly affected the demographic and socioeconomic conditions of the area, leaving large numbers of people homeless and facing the increasing impacts of poverty, famine and social dislocation. This was accompanied by tribal conflicts, particularly between livestock herders and subsistence farmers. During this time, most people lost the majority of their cattle (the animals most vulnerable to droughts), as well as considerable numbers of sheep, goats and camels. Our study attempted to locate the causes of rural vulnerability and related land degradation under climate fluctuations, taking into consideration other factors such as policies and institutions. In the three case studies, we assessed vulnerability and adaptation across the five livelihood capitals in terms of their productivity, equity and sustainability as well as risk factors. The target groups in the three case studies were the vulnerable households practicing subsistence farming and raising livestock in the most drought-prone areas of Sudan. Table 12.2 below summarizes the main impacts of drought as identified by the local communities in the three case study areas.
Autonomous Response Strategies Autonomous response strategies are traditional strategies employed by the local communities in response to recurrent climate variability features such as erratic rainfall and severe droughts. Although the details of these response strategies may vary from one region in Sudan to another, broad commonalities appeared in the type and sequence of responses adopted between the three study areas. Great similarities were seen in the two case studies of Kordofan and Darfur, both situated within the same ecological region of western Sudan. The pattern of household response generally involves a succession of stages. Often the pattern starts by the family moving towards more dependence on markets for the acquisition of food in response to crop failure, as well as look-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 248
248 Climate Change and Vulnerability
Table 12.2 The main impacts of drought identified by the communities across the three case studies Stress
Impact
Drought (declining rainfall)
Leading to
Resulting in
Contributing factors
Water scarcity and Poor hygiene deteriorating quality
Spread of waterborne diseases Poor health
Poor health services High illiteracy rate Poor water harvest and storage
Loss of soil moisture Erosion of topsoil
Low production Crop failure
Poor nutrition Poor health
Low storage capacity Low inputs Lack of technology Poor extension services
Reduced vegetation cover, less fodder, reduced carrying capacity and reduced number of animals
Exposure of topsoil Loss of soil fertility Loss of animals Over logging of trees (selling charcoal)
Loss of animal products Protein deficiency Poor health Increasing dust storms (burying of properties)
Lack of skills Poor rural extension services Poor health services Lack of alternative income sources
Loss of soil moisture Erosion of topsoil
Low production Crop failure
Poor nutrition Poor health
Low storage capacity Low inputs Lack of technology Poor extension services
ing for employment to provide the necessary cash. If conditions get worse and people are unable to pay for their necessities; then they start exchanging their properties and personal possessions for food and water. Some tend to reduce their livestock numbers or change the composition of herds (replacing cattle with smaller animals like sheep and goats). The frequency and size of meals is also reduced. Ultimately, under chronic conditions, they increasingly liquidate their assets and sell their livestock. Those with little or no assets wait either for external assistance to arrive or hope that their livelihoods will improve due to a change in climatic or non-climatic conditions. If nothing happens, then the deteriorating pattern eventually leads to mass migration, including that of whole families. These local measures for coping with climatic shocks do not necessarily represent a standard procedure followed by each and every household in the impacted community, however. Furthermore, the responses may not follow the above-mentioned sequence. Depending on the household’s decision, some families may move immediately on the onset of the crisis, while
3243 J&J Vulnerability
1/11/07
10:23 am
Page 249
Livelihoods and Drought in Sudan 249
others may stay behind and never move off their lands. During the period when the households follow the response sequence described above, it becomes increasingly difficult for them to decide between the few difficult alternatives. Many sad stories from the famous 1984 drought describe the situation of livestock herders who lost all their animals because they refused to sell them at a depressed price and kept moving around in search of forage and water, mostly in vain. In the case of Arbaat, aware of their environment’s vulnerability to drought and famine outbreaks, the Beja agro-pastoralists have developed over time various mechanisms that worked well for some time, contributing to the preservation of their livelihood system and the post-drought recovery. Unlike the responses identified in the Kordofan and Darfur case studies, household responses in Arbaat are not short-term, immediate responses to climatic events/shocks but a series of gradual changes that have been adopted by the communities exposed to the high variability of the climate system over the long term. The adopted survival patterns therefore involve strategic adjustments of livelihood systems, such as preparing in advance for expected climatic shocks. This early strategic action by the community and local leaders reflects the community’s awareness of the fragility of their natural resource base, particularly in view of the frequent droughts and famines that hit the region and the location of the area close to the most populated city in the region, Port Sudan. The ownership of land by the community contributes to the conservation of an important livelihood capital (natural capital). Other community adjustment measures include the adoption of a dispersed pattern of settlements that maintain land carrying capacity and reduce competition and conflict over resources. Another practice is the spatial and temporal migration up and down the Red Sea Hills in pursuit of water, pasture and cultivable lands. This is in addition to the temporary migration for work outside home areas, primarily to Port Sudan. Another significant measure is the strong social sanctioning system that dictates the use of resources, especially pasture and tree conservation, imposed by tribal leaders (the ‘native administration’) and adhered to by all the community members (Osman-Elasha, 2006). These measures have worked well for a period of time, enabling the Beja to survive their harsh, changing climatic conditions. Unfortunately, this traditional pattern of recovery was shattered with the system failure to re-configure after the long and severe drought of the 1980s (Osman-Elasha, 2006).
Planned Interventions In the case studies, planned interventions represent specific types of strategies that have been induced by means of external assistance, government agencies or NGOs to assist the local communities to cope with a number of stresses that they face under their harsh environmental conditions, especially the extreme droughts that are the dominant characteristic of arid and semi-arid Africa. Given enhanced climate variability, in the future more robust and sustainable solutions are needed to protect livelihoods and enable the people of the region
3243 J&J Vulnerability
1/11/07
10:23 am
Page 250
250 Climate Change and Vulnerability
to better utilize their meagre resources. We sought answers to the following questions: • • • • •
What conditions are necessary to cultivate and raise livestock in these dry, arid, and hence drought-prone, areas with the least risk? What types of interventions are necessary? What kinds of livelihood systems are needed? What skills and capacities need to be acquired? Who are the targeted stakeholders?
It is generally argued that enhancing the capacity of Africans to handle climate variability is an appropriate means to increase resilience and reduce the vulnerability of the continent to climate variability and change on all time scales (DFID, 2004). In the three case studies, the assessment of livelihood assets involved analysis of data collected around a set of indicators spanning four areas: productivity, sustainability, equity and risks. It was found that Sahelian Sudan is highly vulnerable to climate variability and change. Many factors were found to contribute to this vulnerability, including the general low level of economic development, the lack of alternative income-generating opportunities and the lack of adaptive technologies. These factors result in heavy dependence on and consequent over-exploitation of natural resources, leading to the tragic conditions such as those faced by the majority of the rural population during the 1980s drought. The identification of these factors guided the efforts of development agencies to reduce the local vulnerability. Major interventions were aimed at increasing ecological sustainability, including water harvesting and conservation, rehabilitation of range lands, creating shelterbelts, employment of a sustainable management system, and the application of strict regulations and sanctions against over-utilization of the meagre resources. The analysis of conditions following the interventions revealed that productivity and ecological management practices have evolved favourably and are, to a great extent, sustainable. Some important interventions aimed at reducing subsistence vulnerability include diversification of income sources, increasing savings and purchasing power, providing better storage facilities, and improving access to markets. Moreover, the formation of locally based organizations (community development committees) has enabled the participation of community members in activities and decision making processes. This has helped in the effective implementation of training and skill development programmes, helped to target local priorities, identify measures acceptable to the community, mobilize community resources and provided a sense of ownership and empowerment that can help sustain efforts. The failures of the local livelihood system in many parts of the Red Sea State in eastern Sudan have rendered it completely dependent on central government support and foreign aid organizations, and made long-term planning, including that of combating desertification, a low priority. In response to this, the Khor Arba’at Rehabilitation Project (KARP) was launched by the UKbased organization SOS Sahel. Its main objective was to contribute to the
3243 J&J Vulnerability
1/11/07
10:23 am
Page 251
Livelihoods and Drought in Sudan 251
development of the Khor Arba’at area through the application of a wide-ranging programme designed to improve community livelihoods and including: the sustainable management of natural resources in order to meet local community needs, provide for food security and enhance grassroots participation in overall development. The main areas of intervention included water management and harvesting, soil reclamation and protection, provision of extension services, community organization, support of education and health services, the empowerment of women and literacy classes. The application of these interventions led to positive impacts on the rural populations through the overall improvement of their livelihood capitals and increased diversity of crop and economic activities through provision of access to resources and livelihood options. This contributed to increased savings and purchasing power for the rural population. We conclude that the control of the Khor waters and the registration and management of land have been the key factors that underpin and shape the resilience pattern in the area. This was substantially helped by the homogeneity and prevailing spirit of cooperation among the community. In the case of Kordofan, a UNDP-funded project, ‘Community-Based Rangeland Rehabilitation (CBRR) for Carbon Sequestration’, has been implemented in response to the disastrous conditions that prevailed in the region after the 1984 drought. The project had two main developmental objectives. The first was to sequester carbon through the implementation of a sustainable, local-level natural resources management system that prevents degradation of and rehabilitates or improves rangelands; the second was to reduce the risks of production failure in a drought-prone area by providing alternatives for sustainable production and increasing the number of livelihood alternatives so that out-migration may decrease and population might stabilize. In essence, the project included both mitigation and adaptation outcomes. This was made possible through the implementation of a simple model of community-based natural resource management aimed at preventing the over-exploitation of marginal lands and rehabilitating rangelands for the purpose of biodiversity conservation and carbon sequestration. These activities represented an integrated package of sustainable livelihood measures with tangible impacts in relation to improving the coping capacity of rural communities in the study area. Most important was the development of a new type of social organization, community development committees, that participated in the process of rangeland rehabilitation, land management, livestock improvement, agro-forestry and sand dune fixation to prevent overexploitation and restore the productivity of rangelands. The availability of essential infrastructure, ranging from irrigation, health, veterinary and extension services to marketing facilities, contributed to the overall development. Social security and credit systems are also essential determinants of whether these interventions can be sustained over time. Given that sheep and goats can survive harsher environmental conditions than cattle, as conditions became drier sheep and goats feature more prominently in the Sahara resource base (di Lernia and Palombini, 2002). However, in the case of Kordofan it was found that goats can contribute to environmental degradation since they graze and browse anything and everything, including
3243 J&J Vulnerability
1/11/07
10:23 am
Page 252
252 Climate Change and Vulnerability
young seedlings. This led to the introduction of sheep to replace the goats, which served two purposes: first, they are more readily marketable and could contribute more to income generation, and second, they are less harmful to the environment, being less aggressive grazers and more manageable. In the case of Darfur, despite the fact that people in this region are able to survive extreme climate conditions and shocks, since the 1984 drought their coping capacity has weakened (Osman-Elasha, 2006). The current situation in Darfur presents a striking example of drought coupled with a number of other factors and stresses creating a devastating cycle of environmental collapse, conflict and displacement. According to Keen (1998), it is often the lack of viable economic alternatives that drives poor people to engage in violence. One of the important interventions that came in response to the 1980s drought is promoted by the Intermediate Technology Development Group (ITDG). This group started its work in Darfur in 1988 on a livelihood support programme, focusing on increasing poor people’s ability to improve their livelihoods through improved food production methods and processing, rural transport and building materials. A key element of ITDG’s food security programme has focused on building on the indigenous knowledge regarding water-harvesting techniques with the involvement of local communities. The aim of the water-harvesting component was to harvest as much of the rain that does fall in parched North Darfur as possible and store it for as long as possible in order to provide enough water for irrigation and domestic use. Various methods are used for water harvesting, including the use of earth dams to capture increased amounts of rainy season floodwater from streams. Other response strategies include construction of a central grain store for surplus production to be used in times of scarcity, diversifying income sources (for example, selling fruits and vegetables), establishing a gum garden and taking advantage of other employment opportunities. Training of farmers to manage their resources and diversify production, involvement of women in productive activities and public life, social networks and cooperatives, and the provision of a credit system are additional important factors.
Policy Assessment We also conducted assessments of policy and institutional aspects by identifying the levels at which policies are developed and implemented, assessing their impacts on the people and identifying the resources that could be influenced by them. We found that it is important in policy analysis to understand contextual factors that shape the policy with respect to the social, political and economic environment. Policy analysis conducted in this context highlighted the possible means by which policies could impact different aspects of people’s livelihoods, including their livelihood assets, the vulnerability context within which they operate and their capacity to choose effective livelihood strategies. Since some government policies on land use in arid and semi-arid areas have been considered among the causes of vulnerability (Brook, 2006), it is important to put in place the types of policies that assist in reducing vulnerability,
3243 J&J Vulnerability
1/11/07
10:23 am
Page 253
Livelihoods and Drought in Sudan 253
enhance the knowledge base and capabilities, and facilitate adaptation and innovation. Our studies have shown that development/adaptation measures will not be sustainable until they address issues related to land use and resource management, and contribute to the maintenance and development of ecologically sound agricultural practices. While short-term relief programmes curbed the negative impacts of the drought and famine conditions in the Red Sea State, they rendered the population heavily dependent on central government support and foreign aid organizations and made long-term policies, including that of combating drought, a low priority. This approach is unsustainable as it may undermine traditional coping capacity and contributes to the creation of relief-dependent societies. Instead, it is preferable to support local capacities and help sustain a local food production system. A similar situation was faced by the people in the Bara area of western Sudan. In this case, the appropriate intervention was to avoid the creation of a situation of relief dependency by selling the relief food to the villages participating in the project activities at nominal prices instead of distributing it for free. The revenue generated was used to supplement the revolving funds of the villages. Initially the idea was not favoured by the communities involved, but later on they concluded that it had replenished their financial capital and contributed to their sustainability (Osman-Elasha, 2006). Moreover, the creation of favourable marketing policies and stable infrastructure in the three study areas led to increased inter-state trading opportunities, social networks, and greater freedom and capacity to participate in decision making. The involvement of women in public life enhanced their participation in economic activities and related events, and ultimately reduced household vulnerability.
Conclusions and Lessons Learned 1
2
Increasing vulnerability of arid and semi-arid lands: Arid and semi-arid lands in Africa are experiencing a state of chronic high climate variability and extremes. Disasters associated with these are expected to become more frequent and more intense in the future (McCarthy et al, 2001). To address this issue, there is a need to deal with many problems faced by the population of the region, including underdevelopment, poverty and other aspects of vulnerability. Use of accumulated knowledge generated by other initiatives such as disaster mitigation, poverty reduction strategies and development programmes, and developing longer-term strategies are both important. Climate change is an additional source of uncertainty and risk: Vulnerability to hunger, famine, dislocation or material loss in arid and semi-arid areas results from the interaction of multiple stresses shaping the socioeconomic system of rural households and impacting their livelihood capitals. Hunger, famine, dislocation and material loss can best be understood and redressed by learning from the past. Analysis of current and historical impacts of climate variability reveals the underlying causes of vul-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 254
254 Climate Change and Vulnerability
3
4
5
6
nerability and assists in the identification of responses necessary to address these causes. Sudan has experienced more than 20 years of below average rainfall, during which there have been many localized droughts as well as a severe and widespread drought from 1980 to 1984. These events may be unrelated to the human contribution to global warming. Nevertheless, it is accepted wisdom that people are changing the Earth’s climate and that changes will be manifested in Sudan. Projections suggest that the climate of Sudan may become drier in the future. With this may come greater risk of drought and its impacts. Drought, population pressures and conflict are degrading lands and undermining resilience: Population and economic pressures have driven people to intensify cultivation of drylands, extend cultivation into more marginal areas, overgraze rangelands and over-harvest vegetation. Recurrent and severe drought overlaid on these activities has degraded lands, reduced availability of water, depressed production of food, fodder and livestock, and eroded livelihoods. Competition for resources has been a source of conflict and has contributed to the tragic violence that engulfs parts of Sudan. Policies and institutions can contribute to increasing vulnerability or improving adaptation: Vulnerability is shaped by the ongoing processes of institutional set-ups, such as different roles played by formal and informal organizations, macro-, meso- and micro-level policies and conflicting interests, and also by the distribution of and access to livelihood assistance, including training opportunities, services and social networks. It is not possible to consider the sustainability of the livelihood in a specific area separately from the broader geographical, social and political-economic context in which the area is located. Some of the causes of and responsibility for environmental degradation and vulnerability of livelihoods at the local levels could be attributed to national-level policies and irrational plans, for example, the expansion of mechanized farming at the expense of rangelands and forests. Communities are adapting: Sudan’s rural communities are adapting to reduce risks in a harsh, variable and changing environment. The adaptation strategies are not necessarily driven by climate change. Nonetheless, they do help enhance resilience to climate change. The measures being adopted include water harvesting and trus cultivation, expanding food storage facilities, managing rangelands to prevent overgrazing, replacing goat herds with sheep, planting and maintaining shelterbelts, planting of backyard farms or jubraka to supplement family food supply and income, supplying micro-credit and educating people about the use of credit, and forming and training of community groups to implement and maintain the various measures. Adaptation measures will not be sustainable until we address issues related to the productivity, equity and sustainability of livelihood assets, and contribute to the maintenance and development of ecologically sound agricultural practices. Adaptation requires local involvement to be effective: In some instances the adaptation strategies have been initiated within communities and
3243 J&J Vulnerability
1/11/07
10:23 am
Page 255
Livelihoods and Drought in Sudan 255
7
8 9
implemented largely with local resources. In other instances the adaptation strategies derive from externally funded and implemented projects. In both cases, the involvement of local institutions and community leaders has been critical for targeting local priorities, identifying measures that can be accepted by the community, mobilizing community resources and developing ownership that can sustain the effort. Adaptation falls short of what is needed: Existing efforts to cope and adapt are too little to manage present day risks. Drought and other climate disturbances impose an unacceptably high burden on the Sudanese, and this burden is likely to grow with global warming and associated climate change. The adaptive responses that have been applied and shown to be successful in building resilience need to be replicated and expanded, and innovative approaches need to be explored. Use the sustainable livelihoods approach for planning: This approach could enable effective national and international responses to manage reduction of risk and enhance adaptive capacity. Current coping capacities could contribute to future adaptation: The future climate is uncertain. However, by enabling current populations to buffer themselves against today’s climatic variations, they will be better able to cope with future contingencies. The social and individual capacity that is being built for community action, flexible management of natural resources and diversification of livelihoods will be valuable for years to come, whatever the climate may be.
References Bashir, S. (1991) ‘Surface Runoff in the Red Sea Province’, RESAP Technical Papers, no 5, May Braun, J. von, T. Teklu and P. Webb (1998) Famine in Africa: Causes, Responses and Prevention, The Johns Hopkins University Press, Baltimore, MD, US Brooks, N. (2006) ‘Climate change, drought and pastoralism in the Sahel’, discussion note for the World Initiative on Sustainable Patoralism, November Carney, D. (1999) Approaches to Sustainable Livelihoods for the Rural Poor, Overseas Development Institute, London, UK DFID (2004) African Climate Report, report commissioned by the UK Government to review African climate science, policy and options for action, Department for International Development, London Di Lernia, S. and A. Palombini (2002) ‘Desertification, sustainability, and archaeology: indications from the past for an African future’, Origini, vol 24, pp303–334 Goldman, I. (2000) Micro to Macro: Policies and Institutions for Empowering the Rural Poor, Department for International Development, London Gruza, G. and E. Rankova (2004) Detection of Changes in Climate State, Climate Variability and Climate Extremity, Institute for Global Climate and Ecology (IGCE), Russia Keen, D. (1998) ‘The economic functions of violence in civil wars’, Adelphi Paper, International Institute of Strategic Studies, London McCarthy, J. J., O. F. Canziani, N. A. Leary, D. J. Dokken and K. S. White (eds) (2001) Climate Change 2001: Impacts, Adaptation, and Vulnerability, Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel
3243 J&J Vulnerability
1/11/07
10:23 am
Page 256
256 Climate Change and Vulnerability on Climate Change, Cambridge University Press, Cambridge, UK Osman-Elasha, B. (2006) ‘Environmental strategies to increase human resilience to climate change: Lessons for eastern and northern Africa’, Final report, Project AF14, Assessments of Impacts and Adaptations to Climate Change, International START Secretariat, Washington, DC, US, www.aiaccproject.org Teklu, T., J. von Braun and E. Zaki (1988) ‘Drought and famine relationships in Sudan: Policy implications’, research report no 88, IFPRI, Washington DC, US Thébaud, B. and S. Batterby (2001) ‘Sahel pastoralists: Opportunism, struggle, conflict and negotiation: A case study from eastern Niger’, Global Environmental Change, vol 11, pp69–78 Warren, A. and M. Khogali (1992) ‘Desertification and drought in the Sudano-Sahelian region 1985–1991’, United Nations Sudano-Sahelian Office (UNSO), New York, US
3243 J&J Vulnerability
1/11/07
10:23 am
Page 257
13
Social Vulnerability of Farmers in Mexico and Argentina Hallie Eakin, Mónica Wehbe, Cristian Ávila, Gerardo Sánchez Torres and Luis A. Bojórquez-Tapia
Introduction Within regions with similar exposure to climate hazards, the sensitivity of particular farm units to climate impacts will vary considerably, as will the capacity of agricultural producers to adapt in relation to a wide variety of socioeconomic, institutional and psychological variables (Easterling, 1996; Brklachich et al, 1997; Eakin, 2002). These variables are not always easily observed or measured at the household level, posing considerable challenges to assessing the vulnerability of specific farm populations. In response to this challenge, we focus on the analysis of the variety of factors that differentiate farm enterprises and farm households in terms of both their sensitivity to climate events and their capacity to adjust to changing climatic and market circumstances. For this analysis, two case studies are presented in two different Latin American socioeconomic and climatic contexts: the community of Laboulaye, in Córdoba Province, Argentina, and the county (municipio) of González, in the state of Tamaulipas, Mexico. Although the cases are quite distinct, their production systems have similar exposure to political and economic uncertainty originating from intensified processes of economic liberalization and market integration in both countries. In very broad terms, the focus of production is also similar: grains and livestock in different combinations for commercial markets. The comparison of the cases, however, also reveals important differences in the distribution of livelihood resources, the relationship between farmers and the public sector, and thus the flexibility of agriculture in the face of both economic and environmental challenges.
Agricultural vulnerability In this study, the social vulnerability of farm households is considered to be a
3243 J&J Vulnerability
1/11/07
10:23 am
Page 258
258 Climate Change and Vulnerability
function of their exposure to climate shocks and extreme events, the sensitivity of the farm to such events in terms of both direct crop impacts and indirect livelihood impacts, and the capacity of households to adapt and adjust to protect themselves from future harm. In the analysis presented below, we explore in depth two of these three attributes of vulnerability: sensitivity and adaptive capacity. Although exposure to climate hazards is in part a product of the social construction of risk through, for example, the historical political and economic factors that have affected the geographic distribution of landholdings, physical infrastructure and populations (see Liverman, 1990), we consider that at the household level these differences in exposure are captured in differential sensitivities to climate impacts.1 The sensitivity to climatic hazards in agriculture is often measured in terms of the degree of decline in yields, losses in agricultural profits or farm value, increased costs of production or falls in production quality (Easterling, 1996; Reilly and Schilmmelpfenning, 1999). Thus sensitivity is also a product of the organization of a farm system, the technology and information employed by the system, and its exposure to other socioeconomic and biological factors as mentioned above (Anderson and Dillon, 1992; Chiotti et al, 1997; Smithers and Smit, 1997). If one uses farmers’ own assessments of climatic impacts on their production (as we have in this study), then sensitivity becomes a function of risk perception and risk tolerance – variables that are rarely captured in impact studies at sector scales (Risbey et al, 1999; Dessai, et al, 2003). In general terms, adaptive capacity can be viewed as a function of a system’s flexibility, stability and access to key resources, attributes that overlap and interact. Farmers’ capacities to respond to stress and uncertainty depend on several factors such as landholding size and soil quality, availability of machinery and equipment, access to credit and insurance, availability of technical assistance and information, social networking, the existence of public support programmes and farmer education and age (Blaikie et al, 1994; Scoones, 1998; Ellis, 2000). Both the degree of diversification within the agricultural production system and the economic diversification of the farm household have also been posited as important factors in determining the sustainability of farm households over time, particularly in peasant farm systems (Ellis, 2000). For example, the importance of non-farm income in a household’s income portfolio, or, conversely, the dependence of the household on agricultural income, can be a measure of household sensitivity to climate impacts (Adger, 1999). As described in the following section, the expectation that diversification will enhance adaptive capacity runs counter to current policy trends, which favour specialization.
The political-economic context of vulnerability The rapid rate of agricultural change that has occurred in Latin America over the last several decades has profoundly altered farmers’ relations with markets, their use of technology and their management of resources (de Janvry and Sadoulet, 1993; Loker, 1996; Bebbington, 2000; Berdegué et al, 2001). Neoliberalism, characterized by a suite of policies such as privatization,
3243 J&J Vulnerability
1/11/07
10:23 am
Page 259
Social Vulnerability of Farmers in Mexico and Argentina 259
decentralization, liberalization and deregulation that are designed to open up economies to foreign investment and international trade, has been the driving force behind the sweeping reforms and has become the dominant paradigm of economic development in the region. In both Mexico and Argentina, agricultural policy reforms have been implemented in concert with substantial changes in macroeconomic policy. Over the course of the 1980s and 1990s, in both countries protectionist policies, price supports and input subsidies for agricultural products were largely withdrawn, farm service agencies privatized and agricultural markets deregulated (Appendini, 2001; Obschatko, 1993; Ghezan et al, 2001). Smaller scale farmers – principally the ejidatarios or communal farmers of Mexico and the small family farmers of Argentina – have been particularly sensitive to these changes. In Argentina, since the beginning of the 1990s, the loss of agricultural income purchasing power has resulted in the concentration of land in larger production units, while those smaller farm units that have remained in production have been forced to restructure and have faced an increasing burden of debt (Wehbe, 1997; Peretti, 1999; Latuada, 2000). The economic crisis of 2002 was followed by an agricultural boom led by growth in soybean production, which helped some farmers get out of debt. The soybean boom, however, is considered to be augmenting soil stress and degradation, particularly under extreme climatic conditions (Cisneros et al, 2004). Beef production and pork and poultry farming have become less profitable in comparison with export crops such as soybeans. Many agricultural analysts in Argentina increasingly fear that these changes are resulting in the exchange of more sustainable agricultural practices for practices that are highly dependent on external inputs (Pengue, 2001; Solbrig and Viglizzo, 1999; Solbrig, 1996). In Mexico, the ejidatarios have been offered title to their land through a federal process initiated in 1992 in the hope that this would encourage the more efficient and entrepreneurial farmers to expand their production, while enabling others to leave agriculture (Ibarra Mendívil, 1996; Cornelius and Myhre, 1998). Today access to and use of technology and farm services (such as credit and insurance) has polarized the sector, dividing those who have ‘commercial potential’ from those who are considered ‘unviable’ (Myhre, 1998). While a relatively small number of agribusinesses have enjoyed rapid growth in productivity and exports over the 1990s, rural incomes have generally stagnated or declined in real terms (Kelly, 2001; Hernández Laos and Velásquez Roa, 2003).
Methods Our analysis employed both quantitative and qualitative data. We undertook an analysis of trends in agricultural and economic policy both at the national scale and the farm level for the two regions of study in order to evaluate some of the non-climatic stress factors that were hypothesized to have affected agricultural sensitivity and capacities in each case. At the farm level, we implemented a household survey using a similar survey instrument in each
3243 J&J Vulnerability
1/11/07
10:23 am
Page 260
260 Climate Change and Vulnerability
region to evaluate vulnerability attributes and outcomes. The samples were designed to be representative of the estimated number and diversity of production units in each study site (see Gay, 2006, for more details on the sample design and other methodological details). In each case study, a selection of the survey variables are grouped according to the particular attributes of adaptive capacity or sensitivity they are intended to represent. In Argentina, adaptive capacity is measured by four attributes: material resources, human resources, management capacity and adaptations. Sensitivity is calculated by the main climatic events affecting each main crop, frequency of adverse events, percentage of area usually affected and type of damage. Crop loss is taken into account as the difference between planted and harvested area within each group and for each of the main crops for the surveyed year. Impacts on livestock production and on infrastructure are also considered. In Mexico, adaptive capacity is measured by five attributes: human resources, material resources, financial resources, information access and use, and economic and agricultural diversity. Sensitivity is defined by variables measuring direct climate impacts on crops and by variables which are hypothesized to indicate greater sensitivity of the farm livelihood to climate shocks. Using a slightly different method in each of the two case studies, aggregate scores for each attribute are calculated and the attribute scores are combined to create the values of a single multivariate indicator of adaptive capacity and a single multivariate indicator of sensitivity. These two indicators are then combined (qualitatively in the Argentinean case, quantitatively in the Mexican case) to create an overall measure of vulnerability. In the Mexican case study, each household is categorized according to its values for sensitivity and adaptive capacity in one of three vulnerability categories (low, moderate and high). In Argentina, the production units are first grouped into different production systems and then for each system group indices of sensitivity and adaptive capacity are obtained, assigning each group to a particular vulnerability level (low, moderate and high) in relation to both their sensitivity and adaptive capacity. The particular variables that appear to contribute most to the vulnerability of farmers in each case are identified through radar diagrams. The results, described below, illustrate the characteristics of vulnerability in each case and the particular resources that currently differentiate the sensitivity and capacities of farmers.
Case Study 1: Laboulaye, Argentina Laboulaye City and its surrounding area belong to the Presidente Roque Saenz Peña Department in the southeast of Córdoba Province. It is a region in which agriculture has been a primary activity since the ‘Desert Conquest’ of the late 19th century, when white settlers were encouraged to expand the agricultural frontier into what was then indigenous territory. Today, agriculture and services for farmers continue to drive the local economy, although the circumstances of production have become increasingly difficult for the area’s family farms.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 261
Social Vulnerability of Farmers in Mexico and Argentina 261
These producers have traditionally pursued a variety of farm strategies based on diverse crops and livestock (Table 13.1). Table 13.1 Main production systems of Laboulaye, Córdoba Province, Argentina Farms
System Bovine livestock Mixed crop/livestock Small landholdings Dairy livestock Cash crop Dairy – mixed Bovine and ovine livestock (mixed) TOTAL
Worked area
Number
%*
Hectares
%*
641 509 222 186 169 94 83 1904
25.4 20.2 8.8 7.4 6.7 3.7 3.3 75.4
284,962 425,165 21,260 79,140 63,538 85,834 78,126 1,038,025
20.4 30.4 1.5 5.7 4.5 6.1 5.6 74.3
Average hectares per farm 444.6 835.3 95.8 425.5 376.0 913.1 941.3 545.2
Note: *Farms included in this table are a subset of the ‘most representative’ of those in the studied area, which is why the percentage columns do not add up to 100. Source: INTA (2002).
Reflecting the same trends that have been noted at the national scale, these family farmers have been negatively affected by the declining prices for livestock and rising living costs. Official statistics show an expansion of cash-cropping area (by 50 per cent) and a decline in livestock area (by 13 per cent) and in livestock numbers (by 37 per cent) in the region since the late 1980s (INDEC, 2004; INTA, 2002). Although soybean expansion was already apparent before the 2002 crisis, the recent rapid land-use changes have now raised concerns about possible environmental impacts (Moscatelli and Pazos, 2002; Pengue, 2001). Increasingly these changes in production practices – the lack of crop rotation, mono cropping and the absence of complementary practices to no-tillage systems – are being associated with the increased impacts from flood events (Cisneros et al, 2004). The agro-ecologic zone to which this area belongs to is characterized as a semi-arid to sub-humid region (INTA, 1987). Annual rainfall average (1961–1990) is 842mm, concentrated in spring, summer and autumn, predominating in summer and autumn (67 per cent). As in much of the Pampas, the region is relatively flat, with slight undulations. While the area is exposed to a variety of climatic hazards, in recent decades floods have raised the most concern among farmers. Excessive rainfall can cause the main rivers of the area, Río Cuarto and Río Quinto, and small streams to overflow (Seiler et al, 2002). The floods have incurred a high social cost locally, causing losses in harvests, problems with livestock mobilization, the spread of diseases, and property damage in both rural and urban areas. Flood management thus requires soil and crop management techniques and high investment in infrastructure and sanitation plans (SAGyP/CFA, 1995).
3243 J&J Vulnerability
1/11/07
10:23 am
Page 262
262 Climate Change and Vulnerability
Adaptive capacity and sensitivity of farm systems in Laboulaye The 47 farm units that were surveyed in Laboulaye are classified according to land use, resulting in 4 groups: cash-cropping farms (6 cases), large-scale mixed cash-crop/livestock units (8 cases, representing those cases with more than 890ha), small-scale mixed cash-crop/livestock ranches (20 cases, representing those cases with less than 890ha) and livestock-specializing farms (13 cases). Measures of sensitivity to climate hazards and adaptive capacity are constructed for each category of farmers. Variables selected as indicators of adaptive capacity are presented in Table 13.2. Their selection is based on previous knowledge about different production systems in the region and according to hypotheses about what types of resources might enhance the farmers’ flexibility to adjust to or cope with climatic variability. Weights for aggregating these variables and variable groupings into a measure of adaptive capacity are decided through a process of consultation with farmers to establish their relative importance for adapting to climate risk. Table 13.2 Indicators of the adaptive capacity of farm households in Laboulaye, Argentina, by farm type Capacity Attribute
Variable
Cash Crop
Mixed Large
Mixed Small
Livestock
32 37 40 47 Potential experience (yrs)a Education (yrs)a 11.7 12.25 8.9 7.3 Landholding size (ha)a 506 2,030 435 426 (Min/max) (120/1200) (900/3600) (172/800) (50/1270) Machinery indexa 1.3 2.3 1.3 0.6 Gross margin (Arg $) (Income)a 207,000 807,000 66,000 0 Good soil quality (% of cases) 66.66 62.50 35.00 30.77 Management Rented land (as % of worked 89.5 46.4 33.8 40.9 Capacity area )b Financial Other sources of income (% of cases)c 17 12 25 54 Resources Hail insurance (% of cases) 33 70 40 7 Information No of sources of technical assistancea 1 2.25 1.5 1.93 Consult any type of climate 83 50 100 84 information (% of cases) Diversity No of cropsa 1.83 3.0 2.05 0.15 % of hectares dedicated to cash-cropsa 98.7 61.7 36.7 0 Social/Human Resources Material Resources
Notes: a Average data; b Average data, weighted by landholding size; c Other sources of income refers only to the same or greater amount of money from activities other than agriculture.
Cash-crop farmers have fewer years of experience than other types of farmers, but more years of education for all groups except the mixed large farm producers. They work farms of moderate size with generally good soil quality and earn an average gross margin greater than do the mixed small and livestock producers. However, a large percentage of the area they cultivate is on rented
3243 J&J Vulnerability
1/11/07
10:23 am
Page 263
Social Vulnerability of Farmers in Mexico and Argentina 263
land and their incomes are not well diversified. Mixed small farm producers and livestock producers have similar human capital, land resources and access to technical assistance but differ in terms of income level and diversity. The mixed large farm producers have the greatest adaptive capacity, with the most years of education, larger farms, higher incomes, greater use of insurance and greater access to technical assistance. While sensitivity to climate in agriculture is often interpreted as a function of crop physiology, soils and management, farmers’ perception of their risk can also contribute to their sensitivity. The farmers are well aware of the implications of recent changes in public policy and land use in the region for their sensitivity to climate. Frustration at the lack of planning and water resource management, poor infrastructure development, and the volatility of the market were also expressed as elements that enhanced their sensitivity to climate impacts. We develop a sensitivity matrix to differentiate farmers’ sensitivity not only by the four farm types of Laboulaye, but also by the type of climate events and the nature of their impacts on crops, livestock and infrastructure. Information to construct the matrix is taken from the survey, and thus is based on farmers’ perceptions of the impacts of climate on their enterprises and livelihoods. By summing the sensitivity scores of each type of climate event for those farm groups reporting climate impacts on crops, livestock and infrastructure, each considered separately, weighted aggregate scores are calculated for each group, climate event and impact category. Results for sensitivity to crop impacts from flood, drought and hail are reported in Table 13.3. Similar analyses were performed for impacts on livestock and infrastructure. The analysis reveals that for cash-crop producers, climate is of relatively small concern. Flooding and drought are the most worrisome climate events for the mixed large farm group, while flooding is more important for the mixed small group of farmers. Surprisingly, neither mixed large nor mixed small groups reported impacts on livestock, despite the fact that these groups were both grain and livestock producers, which may be explained by the relatively higher participation of grain production relative to livestock in the total income of these farm units. The scores for impacts on crops are combined with indicator scores for impacts on livestock and infrastructure reported by farmers in all of the groups. The resulting value is the aggregate sensitivity index.
Vulnerability of farm systems in Laboulaye Aggregate indicators for sensitivity, adaptive capacity and vulnerability are reported in Table 13.4 for the different categories of farmers in Laboulaye. The results suggest that the mixed small and mixed large farmers are the most sensitive of the farm groups to climate hazards. The types of farmers who are most numerous in Laboulaye – the mixed small and livestock farmers – are found to have less adaptive capacity than the mixed large farm systems. According to the survey data, the mixed small and livestock producers tend to have the smallest landholdings and report problems with soil quality. In the case of livestock farmers, their choice of production strategy may be a result of the
3243 J&J Vulnerability
1/11/07
10:23 am
Page 264
264 Climate Change and Vulnerability
Table 13.3 Sensitivity of farm households in Laboulaye, Argentina, to climate impacts on crops by farm type and climate event Flood Drought Hail All
Cash crop
Mixed small
Mixed large
All
0.8 0.3 0.15 1.27
3.52 0.58 0.25 4.36
2.07 2.57 0.28 4.92
6.39 3.45 0.68
Note: Figures in this table are the result of linearly adding weighted values of number, frequency, percentage of area usually affected and type of damage, calculated for the main climatic events affecting each main crop. Weights are determined on the basis of the proportion of agriculture producers concerned with each particular event within their group and by the area dedicated to that particular crop in proportion to the total worked area by each farmer. Crop loss is also taken into account as the difference between planted and harvested area within each group and individually for each of the main crops for the surveyed year.
limitations of the soils they have available to them, which prohibit intensive crop production. Although both these groups report less total income than the other two groups, they tend to be more diversified economically. Cattle constitute a capital asset and source of income that is typically far less sensitive to climate impacts than crop income. The mixed small and livestock farmers also tend to rely more on technical assistance, primarily related to veterinary services, than the other farm groups. Table 13.4 Vulnerability of different farm production systems within the Laboulaye area Farmer Group
Vulnerability
Mixed Small Livestock Cash-Crop Mixed Large
High Moderate Moderate Low
Sensitivity
Adaptive Capacity
5.21 1.89 1.27 5.67
6.82 6.27 6.09 11.95
Note: Sensitivity indices in this table represent those in Table 13.3 plus impacts on livestock production (such as pasture damages) and on infrastructure (damages to roads, water mills, etc.). Adaptive capacity indices were obtained after weighing indicators through consultation with farmers. The indicators were grouped into four categories: material resources; human resources; management capacity and adaptations. All farmer groups were then assigned to vulnerability classes (High, Medium and Low) according to dispersion criteria (defining three ranges from average values of both indices). This vulnerability ranking is only relative to the considered groups within the studied area.
Cash-crop producers have adaptive capacity that is similar to that of mixed small and livestock farmers, although they report farming at a range of scales (120ha to 1200ha), and the interviewed farm managers have higher average incomes and high education levels. However, unlike the mixed small and livestock farmers, these farmers tend not to own their land or machinery, and do not rely on family labour for production. The highest level of adaptive capacity is associated with the mixed large farm systems, which are able to enjoy relatively high incomes from farming increasing amounts of soybeans on large farms of 900 to 3600ha with high quality soils. These farmers also report high education levels, high crop diversity (although low economic diversity) and high use of machinery and other inputs.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 265
Social Vulnerability of Farmers in Mexico and Argentina 265
Adaptive capacity
It is the combination of the farm’s sensitivity to climate and its capacity to manage its impact that determines its vulnerability, which is assessed qualitatively by comparing the aggregate scores for the sensitivity and adaptive capacity indices (Table 13.4 and Figure 13.1). The four Laboulaye farm groups’ sensitivity and adaptive capacity scores are plotted in Figure 13.1. Situations of higher sensitivity (i.e. higher values on the x-axis) and lower adaptive capacity (lower values on the y-axis) correspond to higher vulnerability, as denoted by the arrow in Figure 13.1. For simplicity, vulnerability is assumed to be a linear function of sensitivity and adaptive capacity, implying that points along the diagonals represent equal vulnerability. From the figure it can be inferred that the mixed large group of farmers in the Laboulaye area are less vulnerable than the cash-crop and livestock groups and these three groups are all less vulnerable than the mixed small group. In general, the four groups from Laboulaye are relatively more vulnerable than the majority of other groups of farmers evaluated in the wider south-centre of Córdoba Province, represented by the black dots. This is explained partly because the geomorphology of the Laboulaye area makes it prone to floods in addition to the droughts and hail that occur in the rest of the south-centre of Córdoba.
Figure 13.1 Vulnerability of farm groups in Laboulaye and the wider south-center of Córdoba Province Note: Higher sensitivity and lower adaptive capacity represent conditions of higher vulnerability.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 266
266 Climate Change and Vulnerability
The analysis illustrates that while the exposure to climate variability is similar across the Laboulaye area, the sensitivity and adaptive capacity of each group differed according to the nature of their production activities, their soil conditions and use, and their material assets, landholding size and income (Figure 13.2). The livestock and cash-crop groups are attributed with moderate levels of vulnerability, reflecting similar overall scores for sensitivity and adaptive capacities. However, the factors contributing to their vulnerability are quite different. Livestock production is an activity that is less affected by climate, but presently is less profitable and tends to take place in marginal cropping areas more susceptible to floods. Cash-crop farmers (generally soybean farmers) are far more sensitive to the direct impacts of climate events, but tend to rent land for this activity with higher quality soils that are less prone to flooding and offer better probabilities for high-income generation (90 per cent of worked area is devoted to soybeans). While cash-crop farmers are not economically diversified, livestock producers are highly diversified in their sources of income (58 per cent of cases have other income similar or greater than that from agriculture).
Figure 13.2 Structure of vulnerability for different farmer groups in Laboulaye, Argentina Note: normalized scores for total sensitivity and nine indicators of adaptive capacity show their relative contributions to the level of vulnerability of each group.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 267
Social Vulnerability of Farmers in Mexico and Argentina 267
The largest differences in vulnerability can be seen between mixed small and mixed large producers in the Laboulaye area (high and low vulnerability respectively). While both of these groups show a high sensitivity to a variety of climate events, the very high adaptive capacity of the mixed large producers outweighs this sensitivity and differentiates the two groups. The landholdings of the mixed large group are, on average, five times those of the mixed small, and, on average, the former group devotes more than 60 per cent of their land area to cash crops compared with the 37 per cent for small landholdings. This translates into 12 times greater income for mixed large producers than that of mixed small farmers (net of direct production and land renting costs). Farmers’ participation in organizations is not a distinguishing factor between the groups, a result that was expected given the general perception among all farmers that farmers’ organizations are not particularly helpful for the less favoured rural sectors. In summary, we find that the social vulnerability of agricultural producers is highly related to access to physical and material resources that allow producers greater flexibility in a changing economic and institutional environment.
Case Study 2: González, Mexico The municipio of González is located in the southern extension of the northeastern state of Tamaulipas, Mexico. Unlike much of the state of Tamaulipas, which tends to have a relatively arid climate, González is characterized by subhumid conditions with an average temperature of 24ºC and average annual rainfall totals of 850mm. Historical precipitation records illustrate a decadal pattern rather than any defining trends (Conde, 2005). Some analysts have also observed a correlation between winter precipitation, the Pacific North American Oscillation and El Niño Southern Oscillation events in this region (Cavazos 1999). In addition to periods of drought and flooding, the southern part of the state is particularly susceptible to the impact of the hurricanes that occasionally climb the Gulf of Mexico, as occurred in 1955, 1966, 1988, 1995 and 2000. Frost is relatively infrequent, although hailstorms cause occasional crop losses. Other than climate variability, pests have been a consistent problem in the state. Recently, locusts have been causing significant damage to annual grain crops. In contrast to the northern municipios of Tamaulipas, González has few factories or assembly plants and is primarily agricultural, with 28 per cent of land under crops and 24 per cent under pasture. The municipio’s 3491km2 area is also relatively flat, averaging 56 metres above sea level, which facilitates mechanized agriculture and contributes to the relatively uniform climatic conditions. In 2000, 51 per cent of the population was rural, living in localities of less than 2500 people, and 44 per cent of the economically active population was dedicated to agriculture. It is a relatively poor municipio, with 47 per cent of its economically active population earning less than 2 minimum salaries (INEGI, 2000). Although 87 per cent of adults are literate, over one third have not completed primary school.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 268
268 Climate Change and Vulnerability
As in other municipios in Mexico, the cultivated area in González is divided between private farmers (pequeños propietarios), representing 30 per cent of landholders and farming 70 per cent of agricultural land, and smaller-scale communal farmers (ejidatarios), who represent 70 per cent of landholders but farm only approximately 30 per cent of the municipio’s land. Several of the municipio’s ejidos (communal farms) are incorporated into irrigation districts along the Tamesi and Guayalejo rivers, and this provides farmers with the opportunity to plant irrigated vegetables, grains and fruit trees. The remainder of the municipio specializes in the crops for which Tamaulipas is best known: sorghum, maize, safflower and soybeans. Sorghum was introduced to the region in the 1950s and 1960s to supply the US’ and Mexico’s growing livestock industries and to address what was perceived as Tamaulipas’ vulnerability to drought (Barkin and DeWalt, 1988). Ironically, given the initial marketing of sorghum as a drought-tolerant crop, sorghum is now being actively discouraged in the more arid northern part of Tamaulipas, in response to the government’s observation of a progressive desertification of soils that they believe is associated with sorghum farming under persistent drought conditions in the 1990s (ASERCA, 1997).
Adaptive capacity and sensitivity of farm systems in González As described in the previous section, the agricultural population of González consists of both communal and private farmers (ejidatarios and pequeños propietarios). Of the sample of 181 farm households used to analyse vulnerability in González, 34 cases are private farmers and 147 are communal farmers. Data was collected from the households for variables related to their human capital, material and financial resources, information access, and diversification of livelihood, which we consider to be indicators of their adaptive capacity. As Table 13.5 shows, the two groups are distinguished not only in terms of landholding size, but also in terms of education, age and access to key resources such as credit and insurance. The survey data reveal that the average values for attributes of adaptive capacity are generally higher for private farmers than those of communal farmers. The private farmers are more educated, younger (and thus hypothetically more likely to be receptive to new technologies and ideas), and have far more land with which to experiment with alternative crops. The pequeños propietarios reported a higher average number of crops planted, but tend to devote more of their total landholding to crop cultivation than to livestock. Farmers with private tenure are thus, on average, less diverse in terms of land use, while only slightly more diverse in terms of crop choice. A higher per cent of private farmers report having received credit and insurance and are far more likely to have the mechanical equipment necessary for farmland production. These physical and financial resources could give these farmers more flexibility to respond to unexpected challenges in the future – whether from market shocks or climatic events. The sensitivity of farmer households in González to climate hazards is measured in terms of perceptions and impacts of past climate events and the
3243 J&J Vulnerability
1/11/07
10:23 am
Page 269
Social Vulnerability of Farmers in Mexico and Argentina 269
Table 13.5 Selected indicators of adaptive capacity of farm households in González, Mexico Capacity Attribute
Variable
Human capital
Age of farmer (yrs) Education (yrs) Adults with primary school education Landholding size (ha) Animal units Tractor ownership (%) Irrigation (%) Credit access (%) Insurance access (%) Technical assistance (%) Consult climate information (%) Number of crops Land allocated to crops (%)
Material resources
Financial resources Information Diversity
Private
Communal
46 4.6 1.5 332 30 91 15 44 21 29 53 2.2 90
52 3.2 2.15 23 10 31 37 15 7 29 70 1.5 69
Note: Mean values are given for private and communal farmers.
degree of dependency on crop income (Table 13.6). Results from the survey indicate that sensitivity, both indirect (for example, impact on livelihoods) and direct (for example, impact on crop yields), are similar for the ejidatarios and the private tenure farmers. For example, despite the apparent drought-tolerant nature of sorghum, all farmers reported equally variable yields for sorghum as for maize and similar spatial extent of impacts on land planted with sorghum and maize. Farmers report that yields decline by an average of 73 per cent for all crops in bad years relative to good years. In the 2002–2003 season, the percentage of planted area adversely affected by climate ranged between 40 and 43 per cent for sorghum and 47 to 48 per cent for maize. Table 13.6 Selected indicators of sensitivity of farm households in González, Mexico, to climate hazards Sensitivity Attributes Average number of past climate events remembered Average number of pests and diseases that frequently affect crops and livestock Average % area affected by hazards, summer 2002 Average decline in yields of summer crops between good and bad years (%) Farmers who think climate is changing (%) Farmers reporting loss in income 1998–2003 (%) Dependency of household on crop income (%) Note: Mean values are given for private and communal farmers.
Private
Communal
2.0
1.6
2.4 35
1.7 45
72 71 32.4 60
74 92 35.4 37
3243 J&J Vulnerability
1/11/07
10:23 am
Page 270
270 Climate Change and Vulnerability
The two groups do differ in some respects. Although private farmers tend to recall more damaging climate events in the past than communal farmers, the communal farmers are more inclined to believe that the climate is changing. The ejidatarios also report less frequent problems with pests and crop diseases compared to the pequeños propietarios. Communal farmers in general also rely more heavily on non-farm income sources, either as the primary income source or in combination with crop income, while crop income was the primary income source for 47 per cent of private farmers (see Figure 13.3).
Figure 13.3 Sources of income for communal and private tenure farmers in González, Mexico Note: Data represent percentage of households receiving 66 per cent or more of income from different sources (farm or non-farm).
Vulnerability of farm systems in González Each of the variables associated with adaptive capacity and sensitivity were transformed into a 0 to 1 scale and weighted using the analytical hierarchy process (Saaty, 1980). This process produced two indicators for each household, with values between 0 and 1, representing ‘absence of adaptive capacity’ and ‘degree of sensitivity’. The average score for ‘absence of adaptive capacity’ is predictably higher for the ejidatarios than for the private farmers (0.70 vs 0.59), reflecting the long history of unequal access to services and resources between the two groups (see Yates, 1981; Sanderson, 1986). However, the opposite is true for sensitivity. Average sensitivity scores are 0.38 for ejidatarios and 0.51 for private farmers. The higher sensitivity scores for private farmers
3243 J&J Vulnerability
1/11/07
10:23 am
Page 271
Social Vulnerability of Farmers in Mexico and Argentina 271
% of households
can be attributed to higher reported sensitivity to crop pests and diseases, as well as their greater dependence on crop income. To analyse the overall vulnerability of the households, the values for the sensitivity and adaptive capacity indicators were combined using fuzzy logic (see Bojórquez et al, 2002, for a description of this method). The resulting values are used to assign each household to one of three vulnerability classes: low, moderate and high. In the overall sample, 57 per cent of households are classified as moderately vulnerable, 39 per cent as highly vulnerable and only 4 per cent in the low vulnerability class. In comparison with the ejidatarios, a higher percentage of private farmers are, as expected, associated with the low vulnerability category. However, these farmers are also proportionally better represented in the high vulnerability class (Figure 13.4), suggesting that the land tenure classes alone are not good predictors of vulnerability.
Figure 13.4 Proportion of communal and private tenure farmers associated with each vulnerability class in González, Mexico By plotting the transformed values of the variables that were used to construct the indices for adaptive capacity and sensitivity on radar diagrams, one can see that for both pequeños propietarios and ejidatarios, access to financial resources (credit and insurance) and technical assistance, together with crop-income dependence and problems with crop pests, are what primarily distinguishes the households in the high and low vulnerability classes (Figure 13.5). It is interesting to note that in both of the tenure groups, income diversity and crop diversity are associated with both low and high vulnerability classes. This may reflect the fact that income diversity – particularly diversification into temporary low-skilled positions – is a coping strategy for income insecure
3243 J&J Vulnerability
1/11/07
10:23 am
Page 272
272 Climate Change and Vulnerability
households, and thus could equally be an indicator of poverty and marginalization as an indicator of flexibility in the face of risk, depending on the type of non-farm activity and the other endowments of the household. While crop diversity (in this case, the number of crops planted by the household in 2002/2003) theoretically provides households with alternatives should a climatic hazard affect one particular crop, greater diversity can also mean increasing one’s exposure to a broader variety of climatic hazards and thus increasing the probability that a household will experience crop loss. The greater sensitivity of diversified households was also seen with the mixed small and mixed large farm systems of the Laboulaye case study.
Discussion and Conclusions Despite the differences in the agricultural histories and structure of farming in the two countries of Argentina and Mexico, the case studies reveal important similarities. First, the drivers of vulnerability are similar. In the context of neoliberalism, farmers in both regions are feeling renewed pressure to specialize in one or two commercially viable commodities, and the bias in policy is in favour of larger-scale more entrepreneurial farm units, putting the smallholder farm system at a disadvantage. In Argentina, the importance of agricultural diversification in climate risk mitigation may also be diminishing in the face of the changing technologies and markets, which encourage farmers to accept greater climate risks whenever these risks are coupled with higher economic returns. However, if crop specialization is pursued without the support of financial mechanisms such as insurance to assist in coping with loss, this strategy can be associated with high vulnerability. Pursuing soy monocultures entails higher production costs and, as a result, some households have been forced by debt and economic hardship to rent out their land or abandon agriculture altogether. In González, it is apparent that income diversification still remains the primary risk reduction strategy for farmers operating on the economic margins. These farmers – almost exclusively ejidatarios – are spreading the impact of their losses through access to a variety of income sources. In contrast, for private farmers pursuing a crop specialization strategy, the determining factor in their vulnerability is their access to financial and material resources that can buffer a large-scale producer against climatic risk. Our study provides a snapshot of vulnerability in a particular tumultuous period in the histories of both countries. While our selection of indicators reflects the vulnerability of farm units to very dynamic social and climatic processes, we cannot argue that our assessment captures the changing nature of vulnerability in either location. We can, however, consider our assessment in the context of plausible scenarios for each region. One such scenario would be the continuation of current policy trends, with the likely result – particularly in Laboulaye – of further land concentration, the continued expansion of monocropping, and the continued economic marginalization of the small family farm. The particular resources identified in our case studies as important in
3243 J&J Vulnerability
1/11/07
10:23 am
Page 273
Social Vulnerability of Farmers in Mexico and Argentina 273
Figure 13.5 Structure of vulnerability for low, medium and high vulnerability classes among communal farmers (top) and private tenure farmers (bottom) in González, Mexico
3243 J&J Vulnerability
1/11/07
10:23 am
Page 274
274 Climate Change and Vulnerability
adaptive capacity (credit, insurance, landholding size and farm profit) would continue to play an important role in determining future vulnerability, by facilitating adaptations, reducing sensitivity or improving coping capacities. Under this scenario, the farmers who would be most likely to exit the agriculture sector are those who have been unable to engage fully and profitably in commercial markets and those who have assumed too much economic risk in the face of both volatile markets and climatic conditions. This scenario has environmental implications at the local and regional levels that could – and likely would – feed back into increased vulnerability at the scale of the farm enterprise. For several decades researchers have cautioned that very capital intensive models of agricultural development may, in some cases, make production systems less resilient by creating an unsustainable dependency on exogenous inputs and increasing the sensitivity of production to ecological and economic disturbances such as salinity, water scarcity and pests (Conway, 1987; Buttel and Gertler, 1982; Marsden, 1997). In one sense, the González case offers a good example of this process. According to the state’s agricultural ministry, the problem of erosion and soil degradation from sorghum mono-cropping are beginning to be evident in northern Tamaulipas, illustrating the environmental consequence of what was imagined in the 1960s to be a perfect adaptation to both water scarcity and market opportunity. A similar future could await Laboulaye, with disturbing implications for the regions’ susceptibility to floods and droughts. Another scenario is possible, although less probable. In Argentina, greater concern over the environmental impacts of expansive agriculture might encourage the development of new regulations to conserve fragile lands and enable more diversified land use once again. A drop in soy prices and increased support for Argentina’s livestock industry, perhaps in response to renewed consumer interest in locally-produced organic beef, would help revive the opportunities for small-scale family farms. In Mexico, although the promotion of pasture as an alternative to sorghum may, in the long run, also produce unconsidered environmental consequences (particularly because the promoted pasture is buffle grass, an invasive plant that has become very controversial in the Sonoran Desert, see Tobin, 2004, and Tucson Weekly, 1996), such a policy might provide those farmers practicing mixed grain/livestock farming with the resources they need to adjust to new opportunities. One of the benefits of globalization is that it can also facilitate the growth of new approaches that can improve the resilience of production – such as low-tillage farming, rainwater harvesting for irrigation and improved management of organic manures – by spreading information about these techniques and linking producers to consumers who are increasingly interested in the production process. Tamaulipas is already home to a large number of Mennonite farmers, who have practised low-input, high-yielding agriculture for decades. Public support for the formation of farm associations and producer groups would be a key element in such a scenario. Regardless of the future scenario, it is clear that vulnerability in both regions will continue to be a product of the incidence and biophysical impact of climate events, the structure and resources of the affected farm units, and
3243 J&J Vulnerability
1/11/07
10:23 am
Page 275
Social Vulnerability of Farmers in Mexico and Argentina 275
the institutional and policy environment in which farmers are operating. The complex and multivariate nature of vulnerability challenges any simple interpretation of current sensitivities and makes evaluating future risk problematic. The methodologies we used in this study we hope provide useful insights into how the evolving strategies of farm households are changing the landscape of vulnerability at the local level. Our analysis identifies not only processes presently occurring which may well have important implications for future risk, but also some areas of possible intervention, both at macro and micro levels, that could enhance farmers’ coping capacities to climate risks today and in the future. In our analysis we have raised important questions about the sustainability of current agricultural development pathways and the implications of these trajectories for future climate risk. The particular variables that make a farm system adaptive are not absolute or invariable but rather products of the ambitions and visions of progress held by the broader society and articulated through policy. Who will be adapting and by what means is thus ultimately a normative question inseparable from the ideology and outcomes of present development processes.
Note 1
Exposures to climate hazards in the two study areas in Argentina and Mexico are examined in Chapter 14 of this volume.
References Adger, N. (1999) ‘Social vulnerability to climate change and extremes in coastal Vietnam’, World Development, vol 27, no 2, pp249–269 Anderson, J. and J. Dillon (1992) Risk Analysis in Dryland Farming Systems, FAO, Rome ASERCA (1997) ‘El sorgo Mexicano: Entre la autosuficiencia y la dependencia externa’, Claridades Agropecuarias, vol 46, pp3–13 Appendini, K. (2001) De la Milpa a los Tortibanos: La reestructuracion de la politica alimentaria en Mexico, Colegio de Mexico City, Mexico City, Mexico Barkin, D. and B. R. DeWalt (1988) ‘Sorghum and the Mexican food crisis’, Latin American Research Review, vol 23, no 3, pp30–59 Bebbington, A. (2000) ‘Re-encountering development: Livelihood transitions and place transformations in the Andes’, Annals of the Association of American Geographers, vol 90, no 3, pp495–520 Berdegué, J., T. Reardon and G. Escobar (2001) ‘Rural non-farm employment and incomes in Chile’, World Development, vol 29, no 3, pp411–425 Blaikie, P., T. Cannon, I. Davis and B. Wisner (1994) At Risk: Natural Hazards, People’s Vulnerability and Disasters, Routledge, London Bojórquez-Tapia, L. A., L. Juárez and G. Cruz-Bello (2002) ‘Integrating fuzzy logic, optimization and GIS for ecological impact assessments’, Environmental Management, vol 30, no 3, pp418–433 Brklachich, M., D. McNabb, C. Bryant and J. Dumanski (1997) ‘Adaptability of agriculture systems to global climate change: A Renfrew County, Ontario, Canada pilot
3243 J&J Vulnerability
1/11/07
10:23 am
Page 276
276 Climate Change and Vulnerability study’, in B. Ilbery, Q. Chiotti and T. Rickard (eds) Agricultural Restructuring and Sustainability: A Geographical Perspective, CAB International, Oxon and New York, pp185–200 Buttel, F. H. and M. E. Gertler (1982) ‘Agricultural structure, agricultural policy and environmental quality’, Agriculture and Environment, vol 7, pp101–119 Cavazos, T. (1999) ‘Large-scale circulation anomalies conducive to extreme precipitation events and derivation of daily rainfall in northeastern Mexico and southeastern Texas’, Journal of Climate, vol 12, pp1506–1522 Chiotti, Q., T. Johnston, B. Smit and B. Ebel (1997) ‘Agricultural response to climate change: A preliminary investigation of farm-level adaptation in southern Alberta’, in B. Ilbery, Q. Chiotti and T. Richard (eds) Agricultural Restructuring and Sustainability, CAB International, Oxon and New York, pp201–218 Cisneros, J. M., J. de Prada, A. Degioanni, A. Cantero Gutiérrez, H. Gil, M. Reynero, F. Shah and B. Bravo Ureta (2004) Erosión Hídrica y Cambio de Uso de los Suelos en Córdoba. Evaluación Mediante el Modelo Rusle 2, unpublished manuscript Universidad Nacional de Río Cuarto, (In Spanish) Conway, G. (1987) ‘Properties of agroecosystems’, Agricultural Systems, vol 24, pp95–117 Cornelius, W. and D. Myhre (eds) (1998) The Transformation of Rural Mexico: Reforming the Ejido Sector, Center for US–Mexico Studies, San Diego, CA, US Conde, C. (2005) ‘El Clima de Tamaulipas’, powerpoint presentation to stakeholders, Gonzalez, Tamaulipas, 6–7 May de Janvry, A. and E. Sadoulet (1993) ‘Market, state, and civil organizations in Latin America beyond the debt crisis: The context for rural development’, World Development, vol 21, no 4, pp659–674 Dessai, S., W. N. Adger, M. Hulme, J. Köhler, J. Turnpenny and R. Warren (2003) Defining and Experiencing Dangerous Climate Change, Tyndall Center for Climate Change Research, University of East Anglia, Norwich Eakin, H. (2002) ‘Rural households’ vulnerability and adaptation to climatic variability and institutional change’, unpublished doctoral dissertation, Department of Geography and Regional Development, University of Arizona, Tucson, AZ Easterling, W. (1996) ‘Adapting North American agriculture to climate change in review’, Agricultural and Forest Meteorology, vol 80, pp1–53 Ellis, F. (2000) Rural Livelihoods and Diversity in Developing Countries, Oxford University Press, Oxford Gay, C. (2006) ‘Vulnerability and adaptation to climate change: The case of farmers in Mexico and Argentina’, Final Report AIACC Project No. LA29, International START Secretariat, Washington, DC, US, www.aiaccproject.org Ghezan, G., M. Mateos and J. Elverdin (2001) Impacto de las Políticas de Ajuste Estructural en el Sector Agropecuario y Agroindustrial: El Caso de Argentina, Serie Desarrollo Productivo no 90, Red de Desarrollo Agropecuario, ECLAC (UN), Santiago de Chile Hernández Laos, E. and J. Velásquez Roa (2003) Globalización, Desigualdad y Pobreza: Lecciones de la Experiencia Mexicana, Universidad Autónoma Metropolitana, Mexico City, Mexico Ibarra Mendívil, J. L. (1996) ‘Recent changes in the Mexican constitution and their impact on agrarian reform’, in L. Randall (ed) Reforming Mexico’s Agrarian Reform, M. E. Sharpe, Armonk, NY, pp49–60 INDEC (2004) www.indec.mecon.ar/ INEGI (2000) XII Censo General de Población y Vivienda, 2000. INEGI, Aguascalientes INTA (1987) Análisis de la Evolución, Situación Actual y Problemática del Sector Agropecuario del Centro Regional Córdoba, INTA Centro Regional Córdoba INTA (2002) Plan de Tecnología Regional (2001–2004) Centro Regional Córdoba,
3243 J&J Vulnerability
1/11/07
10:23 am
Page 277
Social Vulnerability of Farmers in Mexico and Argentina 277 Ediciones INTA, Buenos Aires, Argentina Kelly, T. (2001) ‘Neoliberal reforms and rural poverty’, Latin American Perspectives, vol 28, no 3, pp84–103 Latuada, M. (2000) ‘El crecimiento económico y el desarrollo sustentable en los pequeños productores agropecuarios Argentinos de fines de siglo XX’, Conferencia Electrónica sobre Políticas Públicas, Institucionalidad y Desarrollo Rural en América Latina, www.rlc.fao.org/foro/institucionalidad Liverman, D. (1990) ‘Drought impacts in Mexico: Climate, agriculture, technology and land tenure in Sonora and Puebla’, Annals of the Association of American Geographers, vol 80, pp49–72 Loker, W. M. (1996) ‘“Campesinos” and the crisis of modernization in Latin America’, Journal of Political Ecology, vol 3, pp69–88 Marsden, T. (1997) ‘Reshaping environments: Agriculture and water interactions and the creation of vulnerability’, Transactions of the Institute of British Geographers, vol 22, no 3, pp321–337 Myhre, D. (1998) ‘The Achilles’ heel of the reforms: The rural finance system’, in W. Cornelius and D. Myhre (eds) The Transformation of Rural Mexico: Reforming the Ejido Sector, The Center for US–Mexico Studies, San Diego, CA, US Moscatelli, G. and M. S. Pazos (2002) ‘Soils of Argentina: Nature and use’, www.elsitioagricola.com/articulos/moscatelli/soils%20of%20argentina%20-%20nature% 20and%20use.asp Obschatko, E. (1993) Efectos de la Desregulación Sobre la Competitividad de la Producción Argentina. Grupo Editor Latinoamericano, Buenos Aires, Argentina Pengue, W. (2001) ‘The impact of soybean expansion in Argentina’, www.grain.org/seedling/?id=116 Peretti, M. (1999) ‘Competitividad de la empresa agropecuaria Argentina en la década de los ’90’, Revista Argentina de Economía Agraria, vol 11, pp27–42 Reilly, J. M. and D. Schilmmelpfenning (1999) ‘Agricultural impact assessment, vulnerability and the scope for adaptation’, Climatic Change, vol 43, pp745–788 Risbey, J., M. Kandlikar and H. Dowlatabadi (1999) ‘Scale, context and decision making in agricultural adaptation to climate variability and change’, Mitigation and Adaptation Strategies for Global Change, vol 4, pp137–165 SAGyP/CFA (1995) El Deterioro de las Tierras en la República Argentina, SAGyP Ed. Saaty, T. (1980) The Analytic Hierarchy Process, McGraw-Hill, New York and London Sanderson, S. (1986) The Transformation of Mexican Agriculture: International Structure and the Politics of Rural Change. Princeton University Press, Princeton, NJ, US Scoones, I. (1998) ‘Sustainable rural livelihoods: A framework for analysis’, IDS Working Paper, Institute for Development Studies, Brighton, UK Seiler, R., M. Hayes and L. Bressan (2002) ‘Using the standardized precipitation index for flood risk monitoring’, International Journal of Climatology, vol 22, pp1365–1376 Smit, B., D. McNabb and J. Smither (1996) ‘Agricultural adaptation to climatic variation’, Climatic Change, vol 33, pp7–29 Smithers, J. and B. Smit (1997) ‘Agricultural system response to environmental stress’, in B. Ilbery, Q. Chiotti and T. Richard (eds) Agricultural Restructuring and Sustainability: A Geographical Perspective, CAB International, Oxon and New York Solbrig, O. (1996) ‘Towards a sustainable Pampa agriculture: Past performance and prospective analysis’, Working Paper No.96/97-6, The David Rockefeller Center for Latin American Studies (DRCLAS), Harvard University, Cambridge, MA, US Solbrig, O. and E. Viglizzo (1999) ‘Sustainable farming in the Argentine Pampas: History, society, economy, and ecology’, Working Paper No.99/00-1, The David Rockefeller Center for Latin American Studies (DRCLAS), Harvard University, Cambridge, MA, US
3243 J&J Vulnerability
1/11/07
10:23 am
Page 278
278 Climate Change and Vulnerability Tobin, M. (2004) ‘Exotic grass vs. palo verde’, The Arizona Daily Star, 8 August, Accessed online: http://www.azstarnet.com/dailystar/printDS/33361.php Tucson Weekly (1996) ‘The grass that ate Sonora’, cover story, The Tucson Weekly, 18–24 April, www.tucsonweekly.com/tw/04-18-96/cover.htm Wehbe, M. B. (1997) ‘Regional consequences of the agro-food system: Rural changes in the south of Córdoba (Argentina)’, unpublished thesis, Institute of Social Studies, The Hague, The Netherlands Yates, P. L. (1981) Mexico’s Agricultural Dilemma, The University of Arizona Press, Tucson, AZ
3243 J&J Vulnerability
1/11/07
10:23 am
Page 279
14
Climatic Threat Spaces in Mexico and Argentina Cecilia Conde, Marta Vinocur, Carlos Gay, Roberto Seiler and Francisco Estrada
Introduction Extreme climatic events associated with climate variability have exposed the vulnerability of human systems to such events (Le Roy, 1991; Jáuregui, 1995; Florescano and Swan, 1995; Stern and Easterling, 1999). In response, human systems have generated various adaptation strategies and measures according to their differing socioeconomic capacities to cope. Economic globalization processes have, on the one hand, extended, in principle, the access to knowledge and technology that can support a wide range of coping capabilities. On the other hand, in the developing countries, they have also influenced an accelerated loss of resources for many social groups and contributed to a deterioration of the social organizations that have applied and supported these capabilities. These kinds of globalization impacts are especially true for the climate sensitive agricultural sector and as O’Brian and Leichencko (2000) have stated, climate change and economic globalization are two ‘external’ processes that affect agricultural systems in the developing world. Mexico and Argentina are two examples of developing countries that are currently under the pressure of economic globalization. Current social conditions here are such that a relatively small change in normal climatic conditions might trigger important impacts on agricultural activities and generate varying responses based on their specific circumstances. In order to better understand the vulnerability of these two countries to present and future climate risks we analysed coffee and maize cultivation in two regions in Mexico and Argentina respectively (Gay et al, 2002): the central region of Veracruz, Mexico, and Roque Sáenz Peña Department in Córdoba Province, Argentina. Farmers in both countries aim to export most of their agricultural production and are, therefore, highly susceptible to both climatic and market variations. We have used the term ‘climatic threat spaces’ to
3243 J&J Vulnerability
1/11/07
10:23 am
Page 280
280 Climate Change and Vulnerability
describe cases in which climatic variables have played a major role in agricultural losses. In cases where no climatic explanation is found for agricultural losses, other non-climatic stressors that could be responsible are examined. The detection of other stressors – different from climate – that affect agricultural production is particularly important since events such as changes in agricultural policies or prices have in the past, affected crop production even when climate conditions have been favourable. The concept of threat spaces developed in this analysis is therefore intended to aid in the description of climatic variability in the regions under study. The concept could help in attributing losses in agricultural production during a particular year to specific climate anomalies and, in the absence of any climate-related explanation for the losses, would point to the need to examine other stressors in the socioeconomic and sociopolitical spheres in the region during the period. The establishment of threat spaces for current climate variability could additionally help to determine future vulnerability to climate change using outputs from climate scenario projections. At the same time, the assessment of current adaptation strategies to these climate threats can serve to inform future options for coping with climate impacts.
Methods Climatic threat spaces (Conde, 2003) used in this analysis were constructed by means of seasonal or monthly scatterplots of precipitation and temperature, similar to those constructed for climate change scenarios (see, for example, Hulme and Brown, 1998; Parry, 2002). For current climate anomalies, we used the interquartile range of the two variables to determine the threat space for crop cultivation. This can also be done using the one standard deviation criterion, which is currently used (Gay et al, 2004b) by the Mexican Ministry of Agriculture to determine economic support for farmers affected by extreme climatic events (contingencias climatológicas). However, this approach has certain limitations and therefore the interquartile range method is proposed, because it is more robust. According to Wilks (1995, p22) the interquartile range is ‘generally not sensitive to particular assumptions about the overall nature of the data’. Also, the interquartile range ‘is a resistant method that is not unduly influenced by a small number of outliers’ (Wilks, 1995, p22) because the 50 per cent of the distribution (25 per cent in each tail, independent of the shape of the distribution) in which extreme values occur is excluded in this process. To get a more precise description about the tails of the data, a crossed schematic plot can be used to classify extreme values with respect to their degree of unusualness. Climatic data included in the plots cover a period of more than 30 years, with the years 1961–1990 as the reference period. Years of extreme conditions in temperature and/or precipitation are selected as potential circumstances under which crop production might have been affected. Particularly, El Niño or La Niña years are specified in the plot to visualize the possible effects of strong El Niño Southern Oscillation (ENSO) episodes, which have been doc-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 281
Climatic Threat Spaces in Mexico and Argentina 281
umented in previous studies (Podesta et al, 2000; Magaña, 1999; Seiler and Vinocur, 2004) in these regions. For the purposes of this study, El Niño and La Niña years in the threat spaces are considered when the bimonthly values of the multivariate ENSO index (see Wolter) are above (or below) 1 (or -1). Particular attention is given to the strongest ENSO events since 1950. The quartile range for climatic variables is considered as a first order approximation of the limits of the coping range for agricultural activities. The hypothesis is that normal climatic conditions (with respect to 1961–1990) should be within the optimal or near-optimal conditions for crop production in the regions under study and that small variations around that average are tolerable for the system. Most of the crop models used to analyse the crops’ sensitivity to climate are based on this hypothesis (for example, Baier, 1977; Jones and Kiniry, 1986). Optimal conditions for a specific crop can be described within the threat space to help determine whether the region provides optimal or near-optimal conditions, or whether the climatic circumstances represent an important threat for that particular crop production. Thus the coping range’s boundaries, which were initially defined in terms of the interquartile range, could then be redefined in terms of the climatic requirements for the specific crop. In our research, we studied the impacts of climate on two specific crop production systems: coffee production in Mexico and maize production in Argentina. We analysed a series of crop production years and those showing relevant decreases or increases were searched within the threat space for a potentially responsible climatic factor. Such climatic anomalies are, however, only one of the possible stressors that affect production. Other stressors that can impact production in a given critical year include changes in markets and pricing, changes in agricultural policies, and environmental factors such as soils and pests. Documentation of such non-climatic factors that cannot be addressed within the threat spaces helps to determine the relative weight of climatological factors on agricultural production and therefore aids in assessing farmers’ vulnerability. Importantly, such factors can help to explain important crop losses in years when climatic conditions were favourable. In this analysis, we relied on the historical records of prices and agricultural policies; rural studies (Eakin, 2002); in-depth interviews and focus group discussions with producers (Gay et al, 2002; Maurutto et al, 2004; Maurutto, unpublished); and newspaper articles (Martínez, 2002; La Red, 2004, Diario La Communa, 1979a, b and c) and other secondary sources of information as tools to account for the influence of non-climatic stressors. In summary, threat spaces seen under the scope of the specific crop requirements can be used as a tool to assess the current vulnerability of agricultural production to extreme climatic extreme events (drought, floods, frosts, heat waves) in a particular region. Once the role of climatic factors in current vulnerability of crop production is established, future vulnerability under a changed climate can be determined using climate change scenarios. We obtained future climate scenarios for 2020 and 2050 from three general circulation models (GCMs), EH4TR98, GFDLTR90 and HAD3TR00, using the Model for the Assessment
3243 J&J Vulnerability
1/11/07
10:23 am
Page 282
282 Climate Change and Vulnerability
of Greenhouse Gas-Induced Climate Change and a Scenario Generator (Magicc/Scengen Model; version 4.1) (Wigley, 2003; Hulme et al, 2000), and considering the two emission scenarios A2 and B2 (Nakicenovic et al, 2000; IPCC, 2001). Simple interpolation methods have been applied to obtain the possible changes in mean temperature and precipitation values for specific locations (Sánchez et al, 2004; Palma, 2004). The changes in these variables are then introduced into the threat spaces constructed to assess current vulnerability, to visualize possible future climatic threat conditions and to assess future vulnerability to climate change. When the anomalies for both variables are outside the limits of the coping range defined above, the climate threat is considered to increase considerably under the particular climate scenario. Finally, besides the analysis of anomalies in the means of temperature and precipitation, the distribution or variability of the mean is also important. This can be determined by a simple approach that tracks changes in the variability of the mean, assuming no change in the other parameters of the distribution of the data. The distribution then gets transposed to the new mean without any alteration in its shape. Changes in the frequency of extreme events can be used to describe the possible increase in climatic threat. Another approach to include variability in climate change threat spaces is to draw schematic plots constructed with observed data around the future mean value, providing a plot of future minimum, lower quartile, median, upper quartile and maximum values. Thus the probability of having climatic events outside the coping range of a given crop (or activity) can be estimated. Although these methods are based on the assumption that current and future variability are the same, they can provide a rough scenario of future variability that can guide stakeholders and decision makers.
Case Studies Case Study 1: Coffee production in Veracruz, Mexico Veracruz is one of Mexico’s largest states (Figure 14.1), located in the Gulf of Mexico between Tamaulipas and Tabasco. Agriculture is an important economic activity here, generating 7.9 per cent of the state’s gross domestic product (GDP) and providing jobs for 31.7 per cent of the state’s labour force (Gay et al, 2004a). Veracruz is the second largest coffee producer in the country, with 153,000 hectares devoted to coffee production and coffee production involving 67,000 producers in 2000. Ninety-five per cent of this coffee was exported, with a production value of 151.1 million dollars (Gay et al, 2004a). Coffee plantations in Veracruz State are a relatively recent development, with coffee production becoming an important agricultural activity in the 1940s and 1950s, particularly because of the good prices after World War II (Bartra, 1999). Until the 1980s, favourable governmental policies led to an increase in production of nearly 75 per cent on that of the 1940s and a doubling of the number of coffee producers in the country, most of them with plantations of less than 10ha. However, since the late 1980s and early 1990s
3243 J&J Vulnerability
1/11/07
10:23 am
Page 283
Climatic Threat Spaces in Mexico and Argentina 283
Figure 14.1 The state of Veracruz and the region under study Note: Locations described here are situated between about 1000m and 1500m above sea level.
there has been a ‘megacrisis’ (Bartra, 1999) in the coffee sector, brought on by a saturation of production in the coffee market and decreasing international prices for coffee. The current trend of importing Vietnamese coffee into Mexico has also affected the producers’ competitiveness within the national coffee markets. At the same time, government support for coffee producers to cope with adverse market or environmental conditions has declined (Castellanos el al, 2003, Eakin et al, 2005; Eakin et al, 2006). These conditions have contributed to the exacerbation of poverty in the state, with about half the number of municipalities classified under very high and high poverty levels in 2000 (Consejo Estatal de Población, undated). Besides the economic and policy-oriented concerns (Bartra, 1999), historical records of crop losses as well as results from a recent study (Gay et al, 2004b) show that changes in temperature and precipitation have severely affected coffee production in Mexico. However, awareness regarding such current climatic threats is very low, and this lack of knowledge, in combination with declining government support, serves to increase the vulnerability of coffee production and reduces the response capacity of the producers to adverse climatic events. For the purpose of our analysis we selected a region in the central part of the state of Veracruz (Figure 14.1), between latitudes 18°30’ and 20°15’N and longitudes 95°30’ and 97°30’W. It occupies an area of about 183,600 km2 (Palma, 2004), with high altitudes providing almost optimal conditions for cof-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 284
284 Climate Change and Vulnerability
fee production. It currently contributes almost 90 per cent of the total coffee production in Veracruz (Araujo and Martínez, 2002). The region was analysed as a single unit, given that precipitation and temperature regimes (depending on altitude) are similar in the majority of the meteorological stations or municipios (Palma, 2004). The optimum average annual temperature range for coffee production is from 17 to 24°C, and the optimal annual precipitation is between 1500 and 2500mm (Nolasco, 1985). These climatic conditions are observed in Teocelo, in central Veracruz, which has a mean annual temperature of 19.5ºC and an annual precipitation of 2046.9mm (Nolasco, 1985) (Figure 14.2).
Figure 14.2 Normal climatic conditions for Teocelo Note: Tmax = Maximum temperature; Tmin = minimum temperature; Pcp = precipitation
Specific seasonal requirements for coffee include dry weather just before flowering: a small decrease in precipitation or ‘relative drought’ during one or two months in spring (March, April, May: MAM) is considered optimal (Nolasco, 1985; Castillo et al, 1997). Excess rain during this period could damage the flowering process. Figure 14.2 shows this condition for the month of April in Teocelo. This decrease in precipitation must, however, not be confused with an actual drought, since a drought can severely impact production, as has been noted during strong El Niño events (Martínez, 2002, La Red, 2004), particularly during May 1970, May 1983 and May 1998. In Figure 14.3, spring anomalies for minimum temperature and precipitation are shown for Atzalan, Veracruz (19º80’N, 97º22’W, 1842m above sea level), also located in the same region, but with a more reliable and complete data series (Bravo et al, 2006).
3243 J&J Vulnerability
1/11/07
10:23 am
Page 285
Climatic Threat Spaces in Mexico and Argentina 285
Figure 14.3 Threat space for Atzalan, Veracruz, in spring (MAM) Note: Anomalies for minimum temperature (TminºC) and for precipitation (%) and the year they occurred are indicated by the dots. Years are represented by their last two digits (97 equals 1997). The rectangle represents the quartile range (1961–1990) for this season. N represents strong El Niño years. Years with greater anomalies lie outside the rectangle.
In winter and spring, minimum temperature is another critical factor that can affect coffee production. In summer, critical variables include maximum temperature and precipitation since heat waves, drought or floods can affect the development and maturity of the coffee cherry. In autumn, the coffee fruits mature and therefore damaging factors such as climatic extreme events and pests are important. Factors such as high relative humidity, high temperatures and drought can increase the danger of crop pests (Castillo et al, 1997). Harvesting occurs during the end of autumn, winter and the beginning of spring. Adverse climatic conditions, particularly colder and wetter winters during strong El Niño events, can affect labour activities by preventing access to plantations located far into the forests or at higher altitudes. The effects of climatic events on coffee agrosystems in Mexico are unfortunately difficult to detect, since these systems are typically classified as rustic, immersed in a complex forest ecosystem that makes coffee plants quite resilient to climatic variations (Nolasco, 1985). For this reason, the anomalies reported for the extreme climatic events that have severely affected coffee production in the past will be taken as the thresholds to which the system is more vulnerable to determine current and future climatic threats (see Figure 14.2).
3243 J&J Vulnerability
1/11/07
10:23 am
Page 286
286 Climate Change and Vulnerability
Given the vulnerability of coffee plantations to spring drought (Castellanos, 2003), this factor was used in the determination of a coping range for the crop. The average reduction in accumulated (total) spring precipitation during the El Niño years of 1970, 1983 and 1998 was around 60 per cent in the region under study, severely affecting several plantations in the study area. Losses of 25 per cent and 17 per cent in coffee production occurred in Veracruz in 1970 and 1998 respectively (Figure 14.4).Given this damaging impact of a 60 per cent reduction in average rainfall, the lower boundary of total precipitation should be established at -50 per cent of the conditions (approximately 300 mm). At the higher boundary, precipitation should not exceed +50 per cent, as this could diminish the relative drought necessary for coffee. Excess rain can also affect the plant and soil conditions by favouring the spread of pests.
Figure 14.4 Coffee production anomalies for Mexico and for the state of Veracruz Note: Arrows show critical years in coffee production, either for both the country and the state (1989) or for only the state (1971, 1982 and 1998)
The temperature anomaly for spring of 1970 was a ‘mixed’ signal, since during April the temperature rose to more than 40ºC, but during May, after the intense heat-wave, one of the worst frosts to affect the region occurred. This combination of high temperatures and very low precipitation during April, followed by the very intense frost with a continued drought during May, caused great damage to the coffee crop (see data for 1970 in Figures 14.3 and 14.4) that was to be exported (El Universal, in La Red database, 2004). This frost event occurred on several days, but the seasonal averages used in developing
3243 J&J Vulnerability
1/11/07
10:23 am
Page 287
Climatic Threat Spaces in Mexico and Argentina 287
the threat spaces might obscure such climatic events (for example, frost or heavy rains). This highlights a shortcoming of the threat spaces, which cannot substitute for the analysis of daily data to study the behaviour of extreme climatic events. Information on such events must therefore be obtained from other sources. Considering the requirements for coffee with respect to precipitation and minimum temperature, the threat space for spring should include those cases in which minimum temperature could be below 10ºC, which is damaging to the coffee plant (Castillo et al, 1997). Thus minimum temperature anomalies of ≤–1ºC during spring would lie in the threat space for coffee in Veracruz (Figure 14.5). No limitation is depicted for an increase in minimum temperature, because even +3ºC is within the range of optimal requirements.
Figure 14.5 Climatic threat space for coffee during spring (MAM), considering the minimum temperature and precipitation requirements of the coffee plant Note: The square box represents the initial coping range proposed. Climatic anomalies outside the rectangle are considered to be risky for coffee. N represents strong El Niño years.
A threat space that considers the anomalies for maximum temperature and precipitation indicates possible combinations of climatic threats that could occur when anomalies for maximum temperature are greater than +1.5ºC, combined with a 20 per cent reduction in precipitation, a situation that occurred in 1975 (when almost half the coffee production was lost in the central region of Veracruz (see Figure 14.4). The greater losses in other years were caused by other stressors, in other words, different from climatic extreme events, such as changes in the political or economic conditions (Figure 14.4).
3243 J&J Vulnerability
1/11/07
10:23 am
Page 288
288 Climate Change and Vulnerability
It must be noted that adverse climatic conditions during spring could increase the sensitivity of the coffee plants and their environment such that in the subsequent seasons changes in climatic conditions even inside the coping range could cause crop damages. This issue is not discussed here, but it is certainly an important element in the ecological study of this agrosystem. In summer (June, July, August: JJA), total precipitation is quite high in the central region of Veracruz. Nevertheless, exceeding the optimal limit for coffee can cause important damages through flooding or severe storms (Figure 14.6). Such floods were recorded in 1970, 1973, 1996, 1997 and 1981 (La Red, 2004; Martínez, 2002). Generally, for a good coffee crop, rainfall should be neither more that 30 per cent above nor more that 40 per cent below normal. Strong ENSO years might be associated with important drought periods, for example, the episodes of 1982, 1989 and 1997. They may also be associated with an increase in maximum temperature, such as the episodes of 1982 and 1991 (see Figure 14.6). Significant losses in production can be attributed to these events, such as the 30 per cent loss in production in 1982 and the 36 per cent loss recorded in 1989.
Figure 14.6 Threat space for Atzalan, Veracruz, in summer (JJA) Note: Anomalies for maximum temperature (TmaxºC) and for precipitation (%). The rectangle represents the limits of the proposed coping range. The square box represents the quartile range calculated for the 1961–1990 period. N: El Niño year. Na: La Niña year. The scale is adjusted to the minimum and maximum values in the series; for visual purposes, it is not the same as in Figures 14.3 and 14.5
Another interesting observation about the climatic threat space for Veracruz is that there were years when climatic conditions inside the box characterized normal conditions and yet resulted in important losses in coffee production in
3243 J&J Vulnerability
1/11/07
10:23 am
Page 289
Climatic Threat Spaces in Mexico and Argentina 289
Veracruz and indeed throughout Mexico (for example, 1992 and 1993, see Figure 14.4). In such cases, precise studies must be developed to analyse in greater detail monthly and daily data distributions, identifying possible extreme events that were lost in the seasonal averages. If no important climatic signal is detected, then it is highly probable that other stressors impacted the coffee production system, which must be investigated. An important example is the ENSO year 1989, which not only had a strong climatic influence but also coincided with the economic ‘megacrisis’ that additionally affected coffee production. In this year the central region of Veracruz was affected by a severe frost, and during this period the governmental institute (Instituto Mexicano del Café – INMECAFE) that regulated coffee prices and the coffee market was also abolished. The effects of the unfavourable climatic conditions combined with the very low coffee prices due to market restructuring resulted in a significant reduction in coffee production (Figure 14.4). Even before 1989, coffee prices were already on the decline and coffee producers had begun to substitute coffee plantations with sugar plantations, clearing the shade trees on the plantation in this process and causing severe ecological damage (Martínez, 2002). The tendency to switch from coffee to sugar cane has persisted and has intensified in Veracruz, despite an overall increase in total coffee production in the region. The abandonment of the agro-ecosystem caused by low coffee prices and the occurrences of extreme climatic events described above could also be responsible for the spread of pests (such as broca, Hypothenemus hampei), which is an important environmental stressor in plantations in the lower altitudes. According to farmer surveys, climatic events such as drought, heat-waves, strong winds and frosts were the most worrisome climatic events and have been responsible for most of their losses (Gay et al, 2004a; Eakin and Martinez, 2003; Castellanos et al, 2003). However, coffee farmers did not find climate as relevant a variable to their activity as the economic factors. Although such survey responses must be interpreted with care, since respondents could be biased by recent climatic events, it is important to note that a low perception of climatic threats might serve to increase the vulnerability of producers to extreme climatic events in the future. This analysis of the central region of Veracruz in Mexico thus highlights the importance of threat spaces in assessing the significance of climatic factors in coffee cultivation. Coffee production was found to be highly correlated with climate variables, with spring precipitation and summer and winter temperatures being the most relevant climatic variables (Gay et al, 2004a). These findings are also supported by data from newspaper sources and other literature. A separate regression analysis and fieldwork were also undertaken, which arrived at the same conclusions noted above.
Case Study 2: Maize production in Roque Sáenz Peña County, Argentina Argentina is an agro-exporter country with most of the agricultural production based on the pampas, one of the world’s major agricultural regions. Córdoba
3243 J&J Vulnerability
1/11/07
10:23 am
Page 290
290 Climate Change and Vulnerability
province, in the centre of the country, has 83 per cent of its area dedicated to different agricultural activities developed under variable edaphic (soil related) and climatic conditions. The province is ranked fifth in size among all the Argentine provinces and its agricultural outputs represent 25 per cent of the state gross geographical product1 (INTA, 2002). It contributes approximately 14 per cent of national agricultural GDP, with 14 per cent of the national livestock, 17 per cent of the cereal and 25 per cent of the oilseed production. The province is the second largest maize producer in the country, accounting for about 32 per cent of the total national production (SAGPyA, 2004). For our case study, we selected the city of Laboulaye (34°08’S, 63°14’W), situated in Presidente Roque Saénz Peña County in the southern half of Córdoba Province (Figure 14.7). This is a flood-prone area and is typical of the poorly drained plains in the south of Córdoba Province, characterized by semiarid/subhumid temperate conditions (INTA, 1987). Annual mean precipitation here is 845mm (1961–1990) with most of the precipitation concentrated during the warm period (October to March). Seasonal distribution shows that 28.5 per cent of the rain occurs in the autumn, 38.4 per cent in the summer, 26.4 per cent in spring and only 6.7 per cent in winter. Mean annual temperature is 16.3°C, with the month of July being the coldest (8.8°C) and January the warmest (23.7°C). Mean annual precipitation and maximum and minimum temperatures are shown in Figure 14.8.
Figure 14.7 Study region: Location of the city and flood-prone area Climate and climatic variability are major factors driving the dynamics of agricultural production in the area. Inter-annual and inter-seasonal climatic fluctuations result in a high variability in crop production, thus negatively affecting the local and regional economy. Floods and droughts of varying fre-
3243 J&J Vulnerability
1/11/07
10:23 am
Page 291
Climatic Threat Spaces in Mexico and Argentina 291
Figure 14.8 Normal climatic conditions for Laboulaye, Argentina, in terms of maximum temperatures (Tmax), minimum temperatures (Tmin) and precipitation (Pcp)
quencies, intensities and extents alternate in occurrence. Floods are caused mainly by excess rainfall in the flood-prone basin and by the overflowing of rivers and streams. Additional factors (some of them human-induced) such as soil saturation, volume of runoff, physical characteristics of the area, control structures and management characteristics also play a significant role in their occurrence (Seiler et al, 2002). Besides the climatic and soil characteristics, ethnographic research (including surveys, in-depth interviews and focus groups meetings) involving key regional and local actors such as farmers, government officials and cooperative managers also informs us of the perception of climate threats and the importance of climate in agricultural decision making. Farmers here were found to have a ‘naturalized’ perception of climate, since it is a part of their daily life and relates to their living reality as well as to their historical past (Maurutto et al, 2003). Droughts, floods and hailstorms were identified as the most important events affecting farming activities, with flood being the most damaging threat (Rivarola et al, 2002; Vinocur et al, 2004). In particular for the maize crop, air temperatures between 10°C and 34°C are necessary during the growing period, with different thresholds depending on the stage of the crop in the growth cycle (Andrade et al, 1996, Andrade and Sadras, 2002). High temperatures during plant germination to flowering could shorten the development period, as the thermal time required by the crop for flowering is completed earlier. This will result in a shorter time for the interception of solar radiation, leading to a yield reduction (Andrade, 1992). Water
3243 J&J Vulnerability
1/11/07
10:23 am
Page 292
292 Climate Change and Vulnerability
requirements for maize are around 450–550mm during the crop cycle. Water shortage is very critical during the period from 15 days before flowering to 21 days after flowering. During this period, the number of grains per square meter, the principal component of maize yield, is determined (Uhart et al, 1996). The magnitude of yield losses depends on the time of occurrence and the intensity and the duration of any water stress. Because the probability of water stress is higher in January, the time of planting becomes important to ensure that flowering occurs before 15 December of the previous year. Precipitation and maximum temperature anomalies for summer (December, January and February: DJF) for the city of Laboulaye are shown in Figure 14.9. The summer season is critical for maize since the principal components of the crop’s yield are determined during this season. The importance of temperature and precipitation values during this period has been explained above. The box in Figure 14.9 represents values of the interquartile range for maximum temperature and rainfall in Laboulaye for the 1961–1990 period. The anomalies of rainfall and maximum temperature can be related to maize yield deviations from the linear trend (Figure 14.10) to identify events that may cause yield reduction or surplus.
Figure 14.9 Climatic threat space for Laboulaye, Córdoba, for summer season (DJF) Note: Quartile ranges for the anomalies of the climatic variables are shown as a box. Years are represented by the last two digits (97 equals 1997). N and Na represent an El Niño and a La Niña year respectively.
3243 J&J Vulnerability
1/11/07
10:23 am
Page 293
Climatic Threat Spaces in Mexico and Argentina 293
Figure 14.10 Maize yield deviations from linear trend of the series 1961–1999 for the Department of Roque Sáenz Peña (Córdoba, Argentina) Note: 1961 = 1960–1961 cropping season. The dotted lines above and below zero represent plus and minus one standard deviation.
Yields above the trend of the crop yield series 1961–1990 (linear regression, R2 = 0.68, P < 0.00001) were observed for different crop seasons (for example, 1961–1967, 1973–1974 and 1992–1995). If a coping range of one standard deviation from the trend is established (SD = ±26.2 per cent), maize yields exceeded this threshold in the 1962, 1963, 1964, 1978, 1982 and 1998 crop seasons. Despite the above average yields, three of these seasons (1998, 1982 and 1978) were also identified in the risk space (see 97, 81 and 77 in Figure 14.9) due to high rainfall anomalies during the summer, indicating that exceeding this threshold does not imply yield losses. Crop yield during the crop seasons of 1969, 1972, 1976, 1980, 1990 and 1997 was below the one standard deviation threshold (Figure 14.10). Some of these seasons can be identified in the risk space due to low rainfall and high maximum temperatures (for example, 71 and 89 in Figure 14.9). Although a discussion about the effects of La Niña/El Niño events is outside the scope of this chapter, it is interesting to point out that some of the years characterized by high/low rainfall anomalies during summer and high/low yields coincide with El Niño/La Niña events (for example, 1971–1972, 1988–1989 and 1997–1998). In this region, precipitation tends to be low from October to December in cold events (La Niña) and high from November to January during warm events (El Niño) (see Magrin et al, 1998), periods coinciding with the critical period of the maize crop when the grain number is determined. Extreme events in the area, such as floods and droughts, which can be obscured by the seasonal average used to construct threat spaces were documented from newspaper and other literature records (Diario (newspaper) La
3243 J&J Vulnerability
1/11/07
10:23 am
Page 294
294 Climate Change and Vulnerability
Comuna, Laboulaye; Holguín de Roza, 1986). Severe droughts were recorded in February 1967 and June 1988 and during 1989 (La Niña events) causing decreases in maize yield (Figure 14.10). Flood events were documented in 1979 (from January to August), in 1986–1987 (from May 1986 to June 1987) and in 1998–1999 (from March 1998 to May 1999) (Holguín de Roza, 1986). The 1986–1987 flood event (also an El Niño event) resulted in significant yield reductions in the affected local area. The impact of this flood event, however, shows up only as a small reduction in maize yield in Figure 14.10 since maize yields in this case have been determined for the entire study area, including areas not affected by floods. Heat-waves or unusual frost events were not documented for the area during the period of analysis. Based on anomalous values for summer precipitation and maximum temperatures, we identified events that exceed normal values, depicting a risk space for maize for Laboulaye. When rainfall anomalies are below 30 per cent and maximum temperature anomalies are above 1.5°C, important yield decreases are found (for example, in the 1971–1972, 1967–1968 and 1988–1989 crop seasons; see Figures 14.9 and 14.10). In contrast, when the summer precipitation anomaly is above the quartile threshold and the maximum temperature is below that threshold – in other words, events are located in the upper left portion of Figure 14.9 – maize yields are not affected, indicating that scenarios tending toward these conditions, at least for the summer, will be harmless for maize. Finally, there are very few examples of years with conditions representing those in the upper right corner of Figure 14.9 (summer precipitation and temperatures above the quartile threshold), so we were unable to assess their effects on maize yield and production.
Climate Change Scenarios for Mexico As stated in the introduction, climate change scenarios were constructed for 2020 and 2050 using the MAGICC/SCENGEN model version 4.1 (Wigley, 2003; Hulme et al, 2000). The outputs of the EH4TR98, GFDLTR90 and HAD3TR00 models were used under A2 and B2 emission scenarios (Nakicenovic et al, 2000). Simple interpolation methods (Sánchez et al, 2004) and a downscaling technique (Palma, 2004) were applied to obtain the possible changes for specific locations. The results obtained were also compared with those from the IPCC Data Distribution Center (http://ipcc-ddc.cru.uea. ac.uk/) and the Canadian Institute for Climate Studies (www.cics.uvic.ca/scenarios/data/ select.cgi). The EH4TR98 model (German Climate Research Centre/Hamburg Model) model and the HAD3TR00 model from the Hadley Centre provided the best approximations to the observed climate for Mexico (Morales and Magaña, 2003; Conde, 2003). The GFDL (US Geophysical Fluid Dynamics Laboratory) model was also used to generate climate change scenarios, since it has been used in previous research (Gay, 2000) in Mexico. Using the month of July to illustrate climate changes for summer, Figures 14.11a and 14.11b show the possible changes in temperature and precipitation
3243 J&J Vulnerability
1/11/07
10:23 am
Page 295
Climatic Threat Spaces in Mexico and Argentina 295
for the central region of Veracruz. Considering all the GCM outputs, the projected temperature changes range from an increase of 0.9ºC in 2020 to an increase of 2.7ºC in 2050. The projected changes in precipitation range from a decrease of 29 per cent to an increase of 42 per cent, depending on the year and the model used.
Figure 14.11 Projected changes in a) temperature and b) precipitation for the central region of Veracruz (2020 and 2050) Note: Projections made using 3 GCM (G: GFDL; H: Hadley; E: ECHAM) outputs and two SRES scenarios (A2 and B2).
3243 J&J Vulnerability
1/11/07
10:24 am
Page 296
296 Climate Change and Vulnerability
In order to determine the combinations of temperature and precipitation that could represent a climatic threat in the future, changes in temperature and precipitation for each scenario could be introduced in the threat spaces described in the previous sections. When the anomalies for both variables are outside the limits of the coping range (Figure 14.12), the climate scenario is considered to significantly increase the climate threat in the future, and therefore special attention must be paid to it in terms of assessing its potential future impacts on agricultural activities in the regions.
Figure 14.12 Climatic threat space (outside the rectangle) for coffee production in the central region of Veracruz, considering climate change scenarios Note: The projected changes are based on the models ECHAM4 (E_A2 and E_B2), GFDL (G_A2 and G_B2) and Hadley (H_A2 and H_B2).
Figure 14.12 shows the limits of the threat space for coffee production for July (similar to that for summer discussed previously) that is related to precipitation and to maximum temperature. In this case, the proposed coping range is such that precipitation must not exceed an increase of 30 per cent or a decrease of 40 per cent, and maximum temperature must not exceed an increase of 1.5ºC. Using the results of future climate scenarios for 2020 described in this section, it is observed that the projected changes from ECHAM4 (A2 and B2) and GFDL (B2) models are within the coping range for coffee. On the other hand, the projected changes for the Hadley model, in the emission scenario A2 (H_A2, Figure 14.12) lie within the threat space, implying possible important decreases in production, considering the historical impacts and the current climatic threat spaces.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 297
Climatic Threat Spaces in Mexico and Argentina 297
However, if climate variability is considered, then even the scenarios that are within the coping range could represent a climatic threat. As an example, if the ECHAM4 (E_A2 and E_B2) scenarios are considered, which are within the coping range (Figure 14.12), it can be observed in Figure 14.13 that instead of 2 years (5 per cent) with temperatures equal to or greater than 30ºC, there could be 4 years (11 per cent) with maximum temperatures equal to or exceeding 30ºC. This means that while anomalies in the mean temperature for the period continue to fall within the coping range, there is a possibility of an increase in the occurrence of extreme temperature events, which could have negative impacts on the coffee crop. Thus, areas that are not threatened now could be threatened in the future due to the impacts of climate variability, and these scenarios should be further explored and accounted for in future vulnerability studies.
Figure 14.13 Frequency maximum temperatures (Teocelo, Veracruz, 1961–1998) Note: Frequency maximum temperatures considering the observed data (July-base) and with the increase in temperature as proposed by the ECHAM4 (A2) model (July, Clim. Ch.) for 2020.
According to the models’ projections for the future climatic mean values, a relocation of the observed minimum, lower quartile, median, upper quartile and maximum values can be performed, providing a scenario of possible changes in climate variability. Each marker in Figure 14.14 represents different means and variability in temperature (horizontal lines) and precipitation (ver-
3243 J&J Vulnerability
1/11/07
10:24 am
Page 298
298 Climate Change and Vulnerability
tical lines). Current mean and variability are represented by the markers TO and PO, and all of the other markers are future scenarios for the emission scenario A2. The box represents the coping range for July (see Figure 14.12) and it is used to illustrate how, once a variability scenario is provided, relatively small and moderate changes in mean can imply important changes in the probability of adverse conditions for a specific crop. These changes in probability could be interpreted as changes in the viability of a certain crop (or activity) given climate change conditions. It also reveals the possible increase in future vulnerability of the coffee producers to climatic hazards.
Figure 14.14 Current mean and variability conditions and climate change scenarios for 2020 Note: The crosses show a mean value and variability for current and for each future scenario of temperature and precipitation. T0 and P0 are current temperature and precipitation conditions and the black box represents the coping range (Figure 14.12). The scenarios were constructed using mean temperature (T) and precipitation (Pcp) for HadCM3 (A2 scenario) and GFDL (B2 scenario), which project the highest and lowest changes in temperature respectively (see Figure 14.11a).
Using regression equations constructed in previous work (Gay et al, 2004a), which relate climate and economic variables with coffee production, we observed that the most important decreases are expected to occur when the projected changes in spring, summer and winter precipitation are considered. The regression parameters derived are presented in Equation 1. Pcoffee = –35965262 + 2296270(Tsumm)–46298.67(Tsumm)2 + 658.01618(Pspr) +813976.3(Twin)–20318.27(Twin)2 – 3549.71(MINWAGE)
3243 J&J Vulnerability
1/11/07
10:24 am
Page 299
Climatic Threat Spaces in Mexico and Argentina 299
where: Pcoffee = projected changes in coffee production, Tsumm = mean summer temperature, Twin = mean winter temperature, Pspr = mean spring precipitation, and MINWAGE = the real minimum wage. Considering that optimal temperatures for coffee in the summer and winter are 24.8ºC and 20.0ºC respectively, and that the mean precipitation for spring is ~81mm, the expected production of coffee in the central region of Veracruz is 549,158.4 tons. Changes in these variables based on scenario results for April, July and January indicate a decrease of 9 per cent to 13 per cent in coffee production by 2020.
Climate Change Scenarios for Argentina The same models and methodology used above were applied to develop climate change scenarios for Argentina. The results show minor increases in temperature, compared to the Mexican case study. The likely changes in temperature and precipitation for 2020 and 2050 for Laboulaye in January are shown in Figures 14.15a and 14.15b. The projected changes in precipitation range from decreases of 0.7 per cent to 4 per cent, while the projected changes in temperature range from increases of 0.2°C to 1.5°C, depending on the model used. These projections are consistent with the findings of other authors (Ruosteenoja et al, 2003) for southern South America (2010–2039), which have ranged from -3 per cent to +5 per cent changes in precipitation during summer (December, January, February: DJF) and an increase of about 1ºC in temperature for the same season. In this case, even though the increase in temperature is minimal (+0.33ºC as projected by the ECHAM4 scenario for 2020 for January), changes in extreme events should be considered. If the distribution of values of maximum temperature prevails under climate change conditions, the frequency of extreme values for that variable might increase. The situation in Laboulaye is depicted in Figure 14.16, with two events with maximum temperature above 34.0ºC (4.3 per cent) in the baseline years and a projected increase to 4 (10.3 per cent) events with maximum temperature crossing 34.0ºC in 2020. With reference to the threat space constructed for the region (see Figure 14.9), it can be stated that the climate change scenarios that might represent future threats are the ones that project increases in temperature and decreases in precipitation, similar to the climatic conditions during strong La Niña years. However, all the models project a decrease in precipitation for January in 2020 and 2050 (Figure 14.15b), indicating risky conditions for maize. When the summer season (DJF) is considered (data not shown), all the climate change scenarios projected increases in both temperature and rainfall, although the
3243 J&J Vulnerability
1/11/07
10:24 am
Page 300
300 Climate Change and Vulnerability
Figure 14.15 Projected changes in a) temperature and b) precipitation for January in the southern region of Cordoba, Argentina (2020 and 2050) Note: Projected changes using 3 GCM (G: GFDL; H: Hadley; E: ECHAM) outputs and two SRES scenarios (A2 and B2)
3243 J&J Vulnerability
1/11/07
10:24 am
Page 301
Climatic Threat Spaces in Mexico and Argentina 301
Figure 14.16 Observed (1960–2002) and projected DJF maximum temperature (2020) based on ECHAM4 under the A2 emissions scenarios for Laboulaye, Argentina
increases varied between the different scenarios. This latter situation, similar to the conditions in 1975, could lead to decreases in maize yields, but as there are few previous events that could illustrate the farmers’ responses to these conditions, it is difficult to determine the ultimate effects of these climatic changes on crop yield. Sensitivity analyses using crop models, which also include changes in variability besides changes in the mean values of rainfall and temperature, will help to better assess the conditions that could affect crop yield (Vinocur et al, 2000).
Conclusions Analysis of regional climatic variability can help in defining the current climatic threat via the construction of climatic threat spaces, which can serve as a useful tool for defining ‘threat’ for stakeholders and decision makers and for communicating risk. The dispersion diagrams for temperature and precipitation used in this study help to define the coping range and also illustrate the climatic threat for specific agricultural crops. The magnitude of the hazard and of the crop losses incurred due to unfavourable changes in climatic variables can be used to characterize the existing vulnerability of agricultural producers. The outputs of regional climate change scenarios can then be introduced into
3243 J&J Vulnerability
1/11/07
10:24 am
Page 302
302 Climate Change and Vulnerability
these climatic threat spaces to determine any future threats or opportunities that can once again serve to inform key stakeholders in the region. It is important to note that even though the changes in the mean values of climatic variables projected by the outputs of several GCMs may appear to be within the coping range for a particular crop, the pattern of extreme events in the future must also be considered. Even though climate change variability under future conditions is similar to the current observed variability, there could be an increase in the future frequency of extreme climatic events that would contribute to increased future vulnerability of the system under study. Climate threat spaces are limited tools in this regard: they are unable to analyse the frequency of extreme events since they are constructed using seasonal means, which can hide the effects of daily extreme values. However, other data sources such as newspaper articles, interviews and surveys, along with specific daily climatic studies, can help to overcome this limitation. Climate threat spaces can thus serve as a valuable tool to provide an initial assessment of potential risks to crop production or other climate-dependent activities under a changing climate. In combination with other analytical tools such as frequency analysis and sensitivity analysis, a more detailed understanding of present and future climate-related risks can be determined.
Notes 1
The gross geographic product (GGP) of a particular area amounts to the total income or payment received by the production factors (land, labour, capital and entrepreneurship) for their participation in the production within that area. It is a regional GDP.
References Andrade, F. A. (1992) Radiación y Temperatura Determinan los Rendimientos Máximos de Maíz [Radiation and temperature determine maximum yields of maize], Boletín Técnico no 106, Estación Experimental Agropecuaria Balcarce, Instituto Nacional de Tecnologia Agropecuaria, Buenos Aires, Argentina Andrade, F. A., A. Cirilo, S. Uhart and M. Otegui (1996) Ecofisiología del Cultivo de Maíz [Ecophysiology of maize], Dekalb Press, Buenos Aires, Argentina Andrade, F. and V. Sadras (eds) (2002) Bases Para el Manejo del Maíz, el Girasol y la Soja [Basis for the management of maize, sunflower and soybean], Estación Experimental Agropecuaria, Instituto Nacional de Tecnologia Agropecuaria Balcarce – Facultad de Ciencias Agrarias, Universidad Nacional de Mar del Plata Buenos Aires, Argentina Araujo, R. and J. Martínez (2001) AIACC workshop presentation, Vulnerabilidad Social de la Producción de Café en el Estado de Veracruz, Mexico Araujo, R. and J. Martínez (2002) ‘Vulnerabilidad Social de la Producción de Café en el Estado de Veracruz’ [‘Social Vulnerability of Coffee Production in the State of Veracruz’], AIACC workshop presentation, Centro de Ciencias de la Atmósfera, UNAM, Mexico, D.F. May 17– 21 Baier, W. (1977) Crop-Weather Models and their Use in Yields Assessment, WMO no 458, technical note no 151, World Meteorological Organization, Geneva
3243 J&J Vulnerability
1/11/07
10:24 am
Page 303
Climatic Threat Spaces in Mexico and Argentina 303 Bartra, A. (1999) ‘El aroma de la historia social del café’ [The smell of the coffee social history’], La Jornada Delcampo, 28 July, pp1–4 Bravo, J. L., C. Gay, C. Conde, F. Estrada (2006) ‘Probabilistic description of rains and ENSO phenomenon in a coffee farm area in Veracruz, Mexico’, Atmosfera, vol 19, no 2, pp49–74 Castellanos, E., C. Conde, H. Eakin and C. Tucker (2003) Adapting to Market Shocks and Climate Variability in Mesoamerica: The Coffee Crisis in Mexico, Guatemala and Honduras, Final Report IAI SGP 1-015 Castillo, G., A. Contreras, A. Zamarripa, I. Méndez. M- Vázquez, F. Holguín and A. Fernández (1997) Tecnología para la Producción del Café en México [Technology of Coffee Production in Mexico], Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias, Folleto Técnico no 8. Div. Agrícola Veracruz, Mexico CONAPO (1995) Índices de Marginación por Entidad Federativa: Resultados Principales [Marginality Indices for Federal Entities: Principal Results]. Consejo Nacional de Población. CONAPO – Secretaria de Gobernación. México Conde, C. (2003) ‘Cambio y variabilidad climáticos: Dos estudios de caso en México’ [‘Climate variability and change. Two case studies in Mexico’], postgraduate thesis, Universidad Nacional Autónoma de México, Distrito Federal, Mexico Diario La Comuna (1979a) Inundaciones: No hay Solución a Nivel Oficial. Delicada Situación por el Avance de las Aguas [Floods: There is No Solution at Official Level. Delicate Situation Due to the Advance of the Water] 17/05/1979 – Año LIV, N° 2628. Laboulaye, Córdoba Diario La Comuna (1979b) Inundaciones: Gestiones de la Sociedad Rural. [Floods: Management of the Rural Society] 07/06/1979 – Año LIV, N° 2630 Diario La Comuna (1979c) Inundaciones: Las Aguas Llegaron a la Ciudad. [Floods: Water Reached the City] 19/07/1979 – Año LIV, N° 2686 Eakin, H. (2002) ‘Rural households’ vulnerability and adaptation to climate variability and institutional change: Three cases from central Mexico’, PhD dissertation, Department of Geography and Regional Development, The University of Arizona Eakin, H. and J. Martínez (2003) Presentation to coffee farmers and experts in Coatepc, Veracruz, 6–8 December, unpublished Eakin, H., C. M. Tucker, E. Castellanos (2005) Market shocks and climate variability: The coffee crisis in Mexico, Guatemala, and Honduras’, Mountain Research and Development, vol 25, no 4, pp304–309 Florescano, E. and S. Swan (1995) Breve Historia de la Sequía en México [Brief History of Drought in Mexico], Biblioteca Universidad Veracruzana, Xalapa, Veracruz, Mexico Gay, C. (2000) México: Una Visión hacia el siglo XXI: El Cambio Climático en México [Mexico: Climate Change in Mexico: A Vision Towards XXI Century], Instituto Nacional de Ecología (INE), US Country Studies Program (USCSP), SEMARNAP, Universidad Nacional Autónoma de México Gay, C., C. Conde, H. Eakin, M. Vinocur, R. Seiler and M. Wehbe (2002) ‘Integrated assessment of social vulnerability and adaptation to climate variability and change among farmers in Mexico and Argentina’, Final report, Project LA29, Assessments of Impacts and Adaptations to Climate Change, International START Secretariat, Washington, DC, US, www.aiaccproject.org Gay, C., F. Estrada, C. Conde and H. Eakin (2004a) ‘Impactos potenciales del cambio climático en la agricultura: Escenarios de producción de café para el 2050 en Veracruz (México)’ [‘Potential impacts of climate change in agriculture: Scenarios of coffee production for 2050 in Veracruz (Mexico)’], in J. C. García, C. Diego, P. Fernández, C. Garmendía and D. Rasilla (eds) El Clima, Entre el Mar y la Montaña, Asociación Española de Climatología, Santander, Spain, pp651–660 Gay. C, C. Conde, N. Monterroso, H. Eakin, F. Echanove, B. Larqué, A. Cos, H. Celis, S. Cortés, R. Ávila, J. Gómez, H. García, A. Monterroso-Rivas, M. P. Medina, G.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 304
304 Climate Change and Vulnerability Rosales and R. Araujo (2004b) ‘Evaluación externa 2003 del Fondo para Atender a la Población Afectada por Contingencias Climatológicas (FAPRACC)’ [‘External evaluation of the Fund to Support Rural Population Affected by Climatological Contingencies’], Project supported by the Mexican Minister of Agriculture Holguín de Roza, M. C. (1986) Laboulaye, Cien Años de Vida [Laboulaye, One Hundred Years of Life], Dirección General de Publicaciones, Universidad Nacional de Córdoba, Córdoba, Argentina Hulme, M., T. M. L. Wigley, E. M. Brown, S. C. B. Raper, A. Centella, S. Smith and A. C. Chipanshi (2000) Using Climate Scenario Generator for Vulnerability and Adaptation Assessment: MAGICC and SCENGEN, Version 2.4 Workbook, Climate Research Unit, Norwich, UK Hulme, M. and O. Brown (1998) ‘Portraying climate scenario uncertainties in relation to tolerable regional climate change’, Climate Research, vol 10, pp1–14 INTA (Instituto Nacional de Tecnologia Agropecuaria) (1987) Análisis de la Evolución, Situación Actual y Problemática del Sector Agropecuario del Centro Regional Córdoba [Analysis of the Evolution of the Current Situation and Problems in the Agricultural Sector of the Central Region of Cordoba], Ediciones INTA, Centro Regional Córdoba, Buenos Aires, Argentina INTA (2002) Plan de Tecnología Regional (2001–2004) [Regional Technological Plan (2001–2004)], Ediciones INTA, Centro Regional Córdoba, Buenos Aires, Argentina IPCC (2001) ‘Summary for policymakers’, in B. Metz, O. Davidson, R. Swart and J. Pan (eds) Climate Change 2001: Mitigation, Contribution of WGIII to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US Jáuregui, E. (1995) ‘Rainfall fluctuations and tropical storm activity in Mexico’, Erkunde: Archiv Für Wiseenschagtliche Geographie, vol 49, pp39–48 Jones, C. A. and J. R. Kiniry (1986) CERES–Maize: A Simulation Model of Maize Growth and Development, Texas A&M Press, College Station, TX La Red (2004) Social Studies Network for Disaster Prevention in Latin America website, www.desinventar.org/desinventar.html Le Roy, E. (1991) Historia del Clima Desde el Año Mil [History of Climate since the Year One Thousand], Fondo de Cultura Económica, Mexico City, Mexico Magaña, V. (ed) (1999) Los Impactos de El Niño en México [Impacts of El Niño in México], Centro de Ciencias de la Atmósfera, Universidad Nacional Autónoma de México, Mexico, with Dirección General de Protección Civil, Secretaría de Gobernación, Mexico, http://ccaunam.atmosfcu.unam.mx/cambio/nino.htm Magrin, G., M. Grondona, M. Travasso, D. Boillón, Rodríguez and C. R Messina. (1998) Impacto del Fenómeno ENSO Sobre la Producción de Cultivos en la Región Pampeana Argentina [Impacts of ENSO Phenomenon on the Crop Production in the Argentinean ‘Pampas’ Region]. Instituto Nacional de Tecnología Agropecuaria, Instituto de Clima y Agua, Cautelar, Argentina. Martínez, J. (2002) ‘Work Report: Newspaper analysis 1980–2002’, unpublished report, El Diario de Xalapa, Veracruz Maurutto, M. C., M. Vinocur, C. Quiroga and R. Seiler (2003) ‘Assessment of social vulnerability and change among farmers in central Argentina: Importance of the subjective dimension’, Open Meeting of the Global Environmental Change Research Community, Montreal, Canada, 16–18 October, http://sedac.ciesin. columbia.edu/openmtg/absSrch results.jsp Maurutto, M. C. (2004) ‘Assessment of social vulnerability and adaptation to climate variability and change among farmers in central Argentina: importance of the subjective dimension’, Informe final Beca de Investigación, unpublished, Universidad Nacional de Río Cuarto, Argentina Morales R., V. Magaña, C. Millán and J. L. Pérez. (2003) Efectos del Calentamiento
3243 J&J Vulnerability
1/11/07
10:24 am
Page 305
Climatic Threat Spaces in Mexico and Argentina 305 Global en la Disponibilidad de los Recursos Hidráulicos de México [Effects of Global Warming on the Availability of Hydrological Resources in Mexico], Proyecto HC0112, Instituto Mexicano de Tecnología del Agua (IMTA), Centro de Ciencias de la Atmósfera, Universidad Nacional Autónoma de México (CCA–UNAM) Nakicenovic, N., J. Alcamo, G. Davis, B. de Vries, J. Fenhann, S. Gaffin, K. Gregory, A. Grübler, T. Y. Jung, T. Kram, E. L. La Rovere, L. Michaelis, S. Mori, T. Morita, W. Pepper, H. Pitcher, L. Price, K. Riahi, A. Roehrl, H. H. Rogner, A. Sankovski, M. Schlesinger, P. Shukla, S. Smith, R. Swart, S. van Rooijen, N. Victor and Z. Dadi (2000) Special Report on Emissions Scenarios, Special Report of Working Group III of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK Nolasco, M. (1985) Café y Sociedad en México [Coffee and Society in Mexico], Centro de Ecodesarrollo, Mexico City, Mexico O’Brian, K. L. and R. M. Leichenko (2000) ‘Double exposure: Assessing the impacts of climate change within the context of economic globalization’, Global Environmental Change, vol 10, pp221–232 Palma, B. (2004) ‘Generación de escenarios de cambio climático para la Zona Centro del Estado de Veracruz, México’ [‘Generation of climate change scenarios for the central region of Veracruz State, Mexico’], postgraduate thesis, Universidad Nacional Autónoma de México, Mexico City, Mexico Parry, M. (2002) ‘Scenarios for climate impacts and adaptation assessment’, Global Environmental Change, vol 12, pp149–153 Podesta, G., D. Letson, J. Jones, C. Messina, F. Royce, A. Ferreyra, J. O’Brien, D. Legler and J. Hansen (2000) ‘Experiences in application of ENSO-related climate information in the agricultural sector of Argentina’, in Proceedings from the International Forum on Climate Prediction, Agriculture and Development, IRICW/00/1, International Research Institute for Climate Prediction (IRI) Publications, Palisades, NY Rivarola, A. del V., M. G. Vinocur and R. A. Seiler (2002) ‘Uso y demanda de información agrometeorológica en el sector agropecuario del centro de Argentina’ [‘Use and demand of agrometeorological information in the agricultural sector of the centre of Argentina’], Revista Argentina de Agrometeorología, vol 2, pp143–149 Ruosteenoja, K., T. Carter, Jylhä, K. and H. Tuomenvitra (2003) Future Climate in World Regions: An Intercomparison of Model-Based Projections for New IPCC Emission Scenarios, Finnish Environmental Institute, Helsinki, Finland SAGPyA (Secretaría de Agricultura, Ganadería, Pesca y Alimentos de la República Argentina) (2004) ‘Estimaciones agrícolas para maíz’, www.sagpya.mecon.gov.ar/ Sánchez, O., C. Conde and R. Araujo (2004) ‘Climate change scenarios for México and Argentina’, AIACC internal report, supported also by CONACYT, SEMARNAT project SEMARNAT-2002-CO1-0615 Seiler, R. A., M. Hayes and L. Bressán (2002) ‘Using the standardized precipitation index for flood risk monitoring’, International Journal of Climatology, vol 22, pp1365–1376 Seiler, R. and M. G. Vinocur (2004) ‘ENSO events, rainfall variability and the potential of SOI for the seasonal precipitation predictions in the south of Cordoba-Argentina’, in Proceedings of the 14th Conference on Applied Climatology, CD. JP1.10., http://ams.confex.com/ams/pdfpapers/71002.pdf Stern, P. C. and W. E. Easterling (eds) (1999) Making Climate Forecasts Matter, National Academy Press, Washington, DC, US Uhart, S., F. Andrade and M. Frugone (1996) Producción Potencial de Maíz. Impacto de la Sequía Sobre el Crecimiento y Rendimiento [Potential Production of Maize: Impacts of Drought on Maize Development and Yield], Material Didáctico no 12, Estación Experimental Agropecuaria, Instituto Nacional de Tecnologia Agropecuaria, Buenos Aires, Argentina
3243 J&J Vulnerability
1/11/07
10:24 am
Page 306
306 Climate Change and Vulnerability Vinocur, M. G., R. A. Seiler and L. O. Mearns (2000) ‘Predicting maize yield responses to climate variability in Córdoba, Argentina (Abstract)’, in Proceedings of the International Scientific Meeting on Detection and Modelling of Recent Climate Change and its Effects on a Regional Scale, Tarragona, Spain, May 29–31, 2000 Vinocur, M., A. Rivarola and R. Seiler (2004) ‘Use of climate information in agriculture decision making: Experience from farmers in central Argentina’, in Proceedings of the Second International Conference on Climate Impacts Assessment (SICCIA), Grainau, Germany, 28 June–2 July, www.cses.washington.edu/cig/outreach/workshopfiles/SICCIA/program.shtml Wigley, T. (2003) MAGICC/SCENGEN 4.1, technical manual and user manual, Boulder, CO, US Wilks, D. (1995) Statistical Methods in Atmospheric Sciences: An Introduction, International Geophysics Series, vol 59, Academic Press, San Diego, CA, US, pp21–27 Wolter, K. (no date) ‘Multivariate ENSO Index’, Earth System Research Laboratory, National Oceanic and Atmospheric Administration, US, www.cdc.noaa.gov/ ENSO/enso.mei_index.html
3243 J&J Vulnerability
1/11/07
10:24 am
Page 307
15
Climate Variability and Extremes in the Pantabangan–Carranglan Watershed of the Philippines: An Assessment of Vulnerability Juan M. Pulhin, Rose Jane J. Peras, Rex Victor O. Cruz, Rodel D. Lasco, Florencia B. Pulhin and Maricel A. Tapia
Introduction In the Philippines, watershed areas are among those likely to be adversely affected by climate change. Watersheds are critical to sustainable economic development and environmental protection. More than 70 per cent of the country’s total land area lies within watersheds, including much of the remaining natural forests, which provide a host of environmental services. An estimated 1.5 million hectares or more of agricultural lands presently derive irrigation water from these watersheds. Moreover, around 20 to 24 million people – close to a third of the country’s total population – inhabit the uplands of the many watersheds, the majority depending on their resources for survival. Previous studies relevant to climate change in watershed areas have focused on the biophysical aspects (see, for instance, Jose et al, 1996). Completely lacking are studies that delve into the human dimension of climate change in these areas. In particular, there is little information on the impacts of climate change on local communities inhabiting watersheds. Even more limited is knowledge about the vulnerability of these communities to climate variability and extremes, and their capacity and mechanisms for coping with and responding to climate stresses. This chapter tries to fill this gap. It synthesizes the results of pioneering research on the vulnerability of local communities to climate variability and extremes within the Pantabangan–Carranglan watershed located in northern Philippines. The focus is on the local scale, that is to say on households and communities living within the watershed. Specifically, the study sought to
3243 J&J Vulnerability
1/11/07
10:24 am
Page 308
308 Climate Change and Vulnerability
answer questions about exposures of communities in the watershed to stresses from climate variability and extremes in recent decades; the nature and degree of vulnerability to negative impacts from these exposures; the distribution of vulnerability demographically and geographically within the watershed; the social and economic factors that determine vulnerability; and the implications for coping with and responding to human-driven climate change. In the following sections, we outline the key concepts and analytical framework that guided our assessment of the present vulnerability of watershed households and communities to climate variability and extremes, describe the study area and research methodology, present key findings, and conclude by pointing out key research and policy measures that could help advance the body of knowledge and improve policy responses to managing risks associated with climate variability and extremes.
Analytical Framework Vulnerability has many different definitions and is subject to various interpretations and usage. A number of authors have reviewed the various definitions and approaches to vulnerability in relation to climate change (see, for instance, Cutter, 1996; Adger, 1999; UNEP, 2001; Brooks, 2003; Leary and Beresford, 2007). Despite this, confusion appears to continue, and the term seems to defy consensus usage (Few, 2003). In our assessment of vulnerability in the Pantabangan–Carranglan watershed, we define vulnerability as the likelihood of households and communities suffering harm and their ability to respond to stresses resulting from climate variability and extremes. This conceptualization is consistent with Moss (1999), who views vulnerability as a function of at least two major variables: sensitivity of the system to climate-related events and its coping capacity. Climate variability refers to the variations in the mean state and other statistics of the climate (such as standard deviations and the frequency of extremes) on all temporal and spatial scales beyond that of individual weather events (IPCC, 2001). It may be due to natural internal processes within the climate system (internal variability) or to variation in natural or anthropogenic external forcing (external variability). This definition of climate variability encompasses human-driven climate change. A climate extreme is an event that is rare within its statistical reference distribution at a particular place. For the purposes of our study and in consideration of the climatic type in the study area, the occurrence of the following forms of climate variability and extremes are assessed: El Niño, La Niña, early onset or delay of the rainy season, prolonged rains and the occurrence of typhoons. At the operational level, the nature and degree of people’s vulnerability to the above-mentioned climate-related events are examined at two levels: the community and the household. At the community level, the degree of vulnerability of various socioeconomic groups is assessed by looking at the extent of impacts (positive or negative) of climate variability and extremes on four major areas of concern to local communities, namely food availability, water supply,
3243 J&J Vulnerability
1/11/07
10:24 am
Page 309
Climate Variability and Extremes in the Philippines 309
livelihoods and health. In addition, the communities’ adaptation strategies are identified and their effectiveness determined as a measure of their degree of vulnerability.1 To better understand the nature of the household’s vulnerability to climate variability and extremes, an index of vulnerability is constructed from selected indicators that are related to the above-mentioned four major areas of concern. This approach is based on the framework of Moss (1999) using multiple indicators of vulnerability to climate variability and climate change, from which a vulnerability index was developed on the basis of the system’s sensitivity and coping capacity. In the present study, a number of factors are hypothesized to influence vulnerability: demographic factors (age, gender, ethnic affiliation, educational attainment, household size and migration); socioeconomic factors (income, household assets, expenditures, land ownership, farm size, farm practices, number of organizations, and access to transportation, credit and information); geographic factors (distance to market); and a number of coping mechanisms.
The Pantabangan–Carranglan Watershed Physical characteristics The Pantabangan–Carranglan watershed lies between 15°44’ and 16°88’N and 120°36’ and 122°00’E, roughly 176km north of Manila on the island of Luzon (Figure 15.1). Located within the watershed are the municipalities of Pantabangan and Carranglan in the province of Nueva Ecija, the municipalities of Alfonso Castañeda and Dupax del Sur in the province of Nueva Vizcaya, and the municipality of Maria Aurora in the province of Aurora. The watershed has a total area of 97,318ha, of which 4023ha comprise the water reservoir (Saplaco et al, 2001). It is considered to be a critical watershed under the government’s classification as it supports a multipurpose dam for irrigation and hydroelectric generation. The Pantabangan reservoir provides water for domestic and industrial uses and serves to tame the flood waters, which for years damaged farm crops in Central Luzon. At present, it supplies the irrigation requirements of 24 municipalities in the provinces of Nueva Ecija, Bulacan and Pampanga. The reservoir serves an area of 102,532ha, which is divided into 4 districts. A total of 369 irrigators’ associations, consisting of 62,039 farmers, depend on the watershed and reservoir for their farm irrigation needs. The dam also generates 100,000 kilowatts of hydroelectric power, which supplies electricity to the adjacent region of Central Luzon (NPC, 1997). The Pantabangan–Carranglan watershed area largely falls under the Philippine Climatic Type I, with a pronounced dry season from December to April and a wet season the rest of the year. A small portion of the watershed, especially at the boundary of the province of Aurora, falls under Climatic Type II, characterized by no dry season and very pronounced maximum rainfall from November to January. Annual rainfall, based on measurements from 1960
3243 J&J Vulnerability
1/11/07
10:24 am
Page 310
310 Climate Change and Vulnerability
Figure 15.1 Location of the Pantabangan–Carranglan watershed on Luzon Island in the Philippines
to 1999 in four gauging stations within and adjacent to the watershed area, ranges from 1800 to 2300mm (Saplaco et al, 2001). Minimum and maximum monthly temperatures are recorded as 23°C and 34°C respectively, while the average annual relative humidity is 83 per cent (NPC, 1995 and 1997). The topography of the Pantabangan–Carranglan watershed is characterized by complex land configuration and mountainous, rugged terrain. It ranges from nearly level, through undulating and sloping, to steep hilly landscapes. Its soils originated mostly from weathered products of meta-volcanic activities and diorite. Surface soil textures are silty clay loam and clay loam to clay. There are four types of soils in the watershed, known locally as Annam, Bunga, Guimbaloan and Mahipon (Saplaco et al, 2001). The major land-use types found in the watershed are forestlands, open grasslands and reforestation sites (Figure 15.2). Vegetation in the watershed is predominantly second growth. Since the logging boom of the 1960s, primary forest in the watershed has greatly declined, though remnants of dipterocarp forest can still be found (Saplaco et al, 2001). Nevertheless, there has been a significant increase in the area of reforested sites, although these sites are now under intense pressure from increasing population. Residential and barangay (the smallest unit of local government) sites, as well as cultivated areas, are included in the alienable and disposable areas.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 311
Climate Variability and Extremes in the Philippines 311
Figure 15.2 Land uses in the Pantabangan–Carranglan watershed, 1999
Administrative, demographic and socioeconomic characteristics Spearheading the management of the Pantabangan–Carranglan watershed are three national government agencies, the Department of Environment and Natural Resources (DENR), the National Irrigation Administration (NIA) and the National Power Corporation (NPC). Each institution has specific areas within the watershed under its jurisdiction. This institutional arrangement comes from the need to sustainably manage the watershed so that there will be sufficient water in the reservoir for irrigation and hydroelectric power generation (Cruz, 2003). Supporting these institutions in the performance of their functions are the local government units, or barangays, which through the process of devolution instituted under the 1991 Local Government Code were given the mandate to conserve, manage and protect natural resources. There are a total of 36 barangays found in the Pantabangan–Carranglan watershed – 17 in Carranglan, 14 in Pantabangan, 3 in Alfonso Castañeda and 2 in Maria Aurora. As of 2000, about 61,000 people resided in the watershed, which comprises around 12,400 households (National Statistics Office, 2000). Three ethnic groups inhabited the watershed long before the Spanish occupation: the Aetas, Irol-les and the Italengs. They were later joined by several groups of migrants, among them the Pangasinensis, Ibaloi, Ifugao, Waray, Bicolano, Pampango, Kalinga, Kankanai, Ibanag, Cebuano and Ilongot.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 312
312 Climate Change and Vulnerability
However, the construction of the Pantabangan Dam in 1971 has led to relocation of the residents of the town and caused waves of out-migration through the 1970s and 1980s. Today, residents in the Pantabangan–Carranglan watershed are predominantly Tagalog and Ilocano. Other groups present in the area are Pangasinensis, Pampango, Waray, Bicol, Ifugao and Ibaloi (Saplaco et al, 2001). The largest portion of the watershed is located in the municipalities of Pantabangan and Carranglan in the province of Nueva Ecija. More than half of the productive population of Pantabangan and Carranglan are in the labour force. However, unemployment is a problem due to limited employment opportunities in these areas (Municipality of Pantabangan, undated; Municipality of Carranglan, undated), hence many residents depend on the goods and services provided by the watershed for their livelihood. The major source of livelihood in these municipalities comes from agricultural activities. In Pantabangan, about 5400ha (12 per cent of the total land area) is devoted to agriculture, while the corresponding figure for Carranglan is about 19,700ha (28 per cent). Among the major crops produced are rice, corn, onions and vegetables. Although the Pantabangan reservoir is located within the area, it only stores irrigation water for the Central Luzon area. Farmlands in the watershed are unirrigated and dependent on rain because of the topography, and slash-and-burn farming (kaingin) is commonly practised. Fishing is the second largest industry in these areas, much of it located in Pantabangan. This is because the area houses the dam reservoir, which is one of the biggest fishing reservoirs in Asia. The municipality of Carranglan, on the other hand, depends on large fishponds for their fish production. Other sources of income of the residents are cottage and business activities, which include wood and rattan craft, animal grazing and small stores (Municipality of Pantabangan, undated; Municipality of Carranglan, undated). Charcoal-making is also common.
History of development intervention in the area The human-made lake that forms part of the Pantabangan Dam reservoir submerged the old Pantabangan town and seven outlying barangays (Saplaco et al, 2001). Residents of the old town were resettled in the upper portion of Pantabangan. This resettlement process, which was a joint responsibility of the NIA and the Department of Agrarian Reform (DAR), started in May 1973 and was completed in August 1974 (Toquero, 2003). Because of the displacement of people caused by the construction of the dam, the area has continually received support from various agencies and institutions in the form of projects or programmes. Raising the economic conditions of the relocated settlers was a prime concern of the government, and the DAR was the leading agency that took care of this mission. One of the most prominent projects implemented in the watershed was the RP-Japan reforestation project, which was launched in partnership with the DENR. The project commenced in 1976 and ended in 1992. It aimed to reforest the open and denuded areas of the watershed and provide technical
3243 J&J Vulnerability
1/11/07
10:24 am
Page 313
Climate Variability and Extremes in the Philippines 313
support through the establishment of the Afforestation Technical Cooperation Center and the Training Center for Forest Conservation. The project has not only rehabilitated the denuded parts of the watershed but also created jobs for the local residents. Moreover, more than 600 Filipino forestry personnel were trained through this project and are now actively working in environment departments (Yoshida, 2000). Aside from the joint project with the Japanese government, the DENR launched several reforestation programmes, particularly in the municipality of Pantabangan. These are the Regular Reforestation Program, covering a total area of 823ha, and the Integrated Social Forestry Program, which reforested about 856ha. The department also engaged in the Contract Reforestation Program with the NIA from 1989 to 1990. In this programme, the DENR contracted the NIA to reforest a total of 900ha in the Pantabangan–Carranglan watershed (Municipality of Pantabangan, undated). The NPC and the NIA also have their share of projects implemented in the watershed area. Aside from training and extension services, the NPC conducts yearly reforestation and extension projects in the three sectors under its jurisdiction. The reforestation projects cover an average of 30–40ha a year. The biggest project implemented by the NIA in the Pantabangan–Carranglan watershed has been the Watershed Management and Erosion Control Project, which lasted from 1980 to 1988. This project was funded by the World Bank and aimed to control soil erosion and minimize sedimentation and siltation in the reservoir. It had the following components: reforestation; a feasibility study of an integrated development, waste management and smallholder agroforestry pilot project; and an integrated forest protection pilot programme. Aside from opening employment opportunities for 3800 residents in Pantabangan in 1982, the project also provided revenue and profit sharing to the communities in the watershed in the form of facilities such as domestic water supply, school building and road improvements The Casecnan Multipurpose and Irrigation Project, which began operations in 2001, was designed to collect some of the water from the Casecnan and Taan rivers in Nueva Vizcaya and transport it to the Pantabangan reservoir. It was designed to irrigate 35,000 new hectares of agricultural lands and stabilize the water supply of the current areas serviced by the Pantabangan–Carranglan watershed. Moreover, it will generate approximately 150MW of hydroelectric capacity for the important Luzon grid (Calenergy Company, 2004). As already mentioned, the above projects have significantly helped the residents in the watershed through the provision of jobs, livelihood programmes and various forms of assistance. But despite the three-decade-long development effort of the government, amounting to PHP1.5 billion or US$30million, there is still widespread poverty in the resettlement, as shown by a high percentage of families with income below the poverty threshold of PHP7377. The residents also perceived the services provided by the government organizations to be unsatisfactory. This implies the failure of the government in providing an economically viable resettlement area for the residents. A point of contention which could have contributed to this failure is the lack of participation of the residents in the planning and monitoring of the
3243 J&J Vulnerability
1/11/07
10:24 am
Page 314
314 Climate Change and Vulnerability
development projects or programmes. Consequently, some residents were unfamiliar with the livelihood activities that were introduced by the projects. Moreover, these development projects and programmes, which were temporary, may have resulted in the dependency of some people on these forms of assistance. With the recent completion of these development projects and programmes, the local people resort to charcoal-making and open kaingin or slash-and-burn farming practices, livelihoods that are practised by more than 50 per cent of the residents in the watershed (Toquero, personal communication, 2005). These practices are negatively impacting the watershed and areas reforested by the various projects.
Research Methodology Household survey A household survey was conducted to collect information about the vulnerability of households to climate variability and extremes and the socioeconomic factors influencing their vulnerability. The survey included question about 1) the socioeconomic profile of the respondent; 2) the household’s use of and benefits from the Pantabangan–Carranglan watershed; 3) climate variability and extremes experienced in the last few decades and their impacts; 4) food availability, water supply, livelihood and heath; and 5) adaptation strategies. The survey was administered in the four municipalities of the three different provinces encompassing the watershed. These are Pantabangan and Carranglan in Nueva Ecija, Alfonso–Castañeda in Nueva Vizcaya, and Maria Aurora in Aurora. Twenty-six of the 36 barangays within the watershed area were represented. The other 10 barangays were excluded because only small portions of their respective areas are within the watershed boundary and very few of their inhabitants live within the watershed. A total of 375 respondents were randomly selected using the barangay records. This sampling technique employed was adopted from Chua (1999), which allows a 0.05 permissible error and 95 per cent confidence interval level. Focus group discussions also were conducted in 21 barangays to complement the household survey. Of the 26 barangays included in the household survey, invited representatives from four barangays did not show up for the scheduled discussions. A minimum age of 40, equal distribution of males and females, and representation of different socioeconomic groups were considered in the selection of participants. The focus group discussions employed a combination of participatory techniques such as time-line analysis, stakeholder analysis, participatory vulnerability assessment and community mapping of vulnerable areas (see Pulhin, 2002, for discussions of these techniques). Time-line analysis was used to determine the local communities’ exposures to climate variability and extremes from the 1960s to the present. A combination of stakeholder analysis, participatory vulnerability assessment and community mapping techniques were used to identify which socioeconomic groups in the communities are more vulnera-
3243 J&J Vulnerability
1/11/07
10:24 am
Page 315
Climate Variability and Extremes in the Philippines 315
ble, their location in the watershed, and the extent and nature of their vulnerability. A scenario-building activity was conducted to explore the potential impacts of more extreme climate conditions that might be experienced in the future as a consequence of human-driven climate change and the vulnerability of the different socioeconomic groups to the impacts of climate change. The focus group discussions were also used to incorporate the communities’ perspectives in determining the weights of the different indicators for construction of a vulnerability index. Direct field observations were also conducted to validate information gathered through the household survey and focus group discussions. Vulnerable areas identified by the participants during the community mapping were verified on the ground and documented through photographs. In addition, GPS readings of vulnerable areas were taken for purposes of mapping these areas.
Index of vulnerability A multi-level index of the vulnerability of households to climate variability and extremes was constructed from information gathered in the household surveys. The index consists of four major component sub-indices: food, water, livelihoods and health. These indices are further divided into subcategories. Drawing on the framework of Moss (1999), the subcategories comprised relevant variables that involved certain characteristics of the component indicators in relation to climate variability and extremes (representing the household’s sensitivity in relation to these components) and the presence or absence of adaptation strategies (representing the household’s coping capacity). The subindices and their component indicator variables are shown in Table 15.1. The vulnerability index is constructed as a weighted aggregation of the four major sub-indices, which, in turn, are weighted aggregations of their component indicators. The weights are selected such that the maximum possible weighted vulnerability score is 100. Two different schemes are used to determine the weights for the sub-indices and indicators. One scheme applies the researchers’ judgements while the second scheme uses results of focus group discussions to develop weights that represent the perspectives of local stakeholders. The vulnerability index computation based on the researchers’ judgements assigns equal weights of 25 to all of the 4 sub-indices (food, water, livelihood and health). The subcategories under each major index were also given equal weights (see Table 15.1). For instance, the food index, assigned a total possible weighted score of 25, has two sub-components – availability of seeds and crop yield – each with a total possible weighted score of 12.5. The sub-component for availability of seeds is further divided into three separate indicators: seasonality of seed availability, sensitivity of seed availability to climate variations and availability of strategies to adapt to constraints on seed availability. Each of these indicators is assigned a total possible weighted score of 4.17 (a third of the 12.5 weight for seed availability). Thus, if seeds are available to a household only seasonally, a score of 4.17 is given, whereas year-round availability would result in a score of zero for the indicator for seasonality of seed availability.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 316
316 Climate Change and Vulnerability
Table 15.1 Multi-level indicator of vulnerability of households to climate variability and extremes (CV and E) using varying weights Weighted Scores Sub-Indices and Indicators A. Food a.1. Seed availability a.1.1. Availability of planting materials i. Available any time of the year ii. Seasonal or hard to find a.1.2. Is it affected by CV and E? i. Yes ii. No a.1.3. Adaptation strategies i. With adaptation ii. Without adaptation a.2. Crop yield a.2.1. Per cent (%) lost in rice production a.2.2. Is it affected by CV and E? i. Yes ii. No a.2.3. Adaptation strategies i. With adaptation ii. Without adaptation B. Water b.1. Domestic water b.1.1. Sources of domestic water i. Natural sources ii. Through agencies 1 b.2.1. Distance of house to sources of water i. 0–250m ii. 251–500m iii. 501–1000m iv. >1000m b.1.3. Observation for the supply of domestic water i. Declining supply ii. Increasing supply iii. No change b.1.4. Is domestic water supply affected by CV and E? i. Yes ii. No b.1.5. Adaptation strategies i. With adaptation ii. Without adaptation
Weights Weights Provided by Local Communities Provided by Pantabangan Carranglan P&C Researchers Combined 25 12.5 4.17 0
25 20 8
40 15 7
32.5 17.5 7.5
3
2
2.5
4.17 4.17 4.17 0 4.17 0 4.17 12.5 4.17
5 9 9 0 3 2 1 5 1.5
5 5 4 1 3 1 2 25 10
5 7 6.5 0.5 3 1.5 1.5 15 5.75
4.17 4.17 0 4.17 0 4.17 25 12.5 2.5 2.5 .25 2.5
2 2 0 1.5 0.5 1 40 33 11 8 3 5
10 7 3 5 2 3 40 15 7 6 1 2
6 4.5 1.5 3.25 1.25 2 40 24 9 7 2 3.5
0.62 1.25 1.88 2.5 2.5
0.4 1 1.5 2.1 7
0.2 0.3 0.5 1 2
0.3 0.65 1 1.55 4.5
2.5 0 1.25 2.5
3 2 1 5
1 0.5 0.5 2
2 1.25 0.75 3.5
2.5 0 2.5 0 2.5
3 2 5 1 4
1.5 0.5 2 0.5 1.5
2.25 1.25 3.5 1.25 2.75
3243 J&J Vulnerability
1/11/07
10:24 am
Page 317
Climate Variability and Extremes in the Philippines 317
Table 15.1 (continued) Weighted Scores Sub-Indices and Indicators b.2. Irrigation water b.2.1. Regularity/problem with supply? i. Problem with supply ii. No problem with supply b.2.2. Effects of scarcity i. Decrease in production/ income ii. No (zero) production/ income iii. Delayed harvest b.2.3. Adaptation strategies i. With adaptation ii. Without adaptation C. Livelihood c.1. Seek sources of income in cases of CV and E? i. Yes ii. No c.2. Is income from other sources sufficient? i. Sufficient ii. Not sufficient c.3. Adaptation strategies i. With adaptation ii. Without adaptation D. Health d.1. Experienced health problems during CV and E? i. Yes, experience health problems ii. No d.2. Kinds of health problems experienced during CV and E i. Diarrhoea, amoebiasis, dehydration, dysentery ii. Dengue, typhoid, malaria iii. Others: hepatitis, bronchitis, sore eyes, etc. d.3. Access to medical services i. Sufficient ii. Not sufficient d.4. Adaptation strategies i. With adaptation ii. Without adaptation
Weights Weights Provided by Local Communities Provided by Pantabangan Carranglan P&C Researchers Combined 12.5 4.17
7 3
25 10
16 6.5
0 4.17 4.17 2.78
1 2 2 1
3 7 10 7
2 4.5 6 4
4.17
0.5
1
0.75
1.39 4.17 0 4.17 25 8.33
0.5 2 0.56 1.44 15 6
2 5 2 3 10 2
1.25 3.5 1.28 2.22 12.5 4
0 8.33 8.33
4 2 6
0.5 1.5 6
2.25 1.75 6
0 8.33 8.33 0 8.33 25 6.25
2 4 3 2 1 20 6
2 4 2 0.5 1.5 10 2
2 4 2.5 1.25 1.25 15 4
6.25
4
1.5
2.75
0 6.25
2 7
0.5 4
1.25 5.5
4.17
3
2
2.5
6.25 2.09
2 2
1 1
1.5 1.5
6.25 0 6.25 6.25 0 6.25
3 1 2 4 1.8 2.2
2 0.5 1.5 2 0.5 1.5
2.5 0.75 1.75 3 1.15 1.85
3243 J&J Vulnerability
1/11/07
10:24 am
Page 318
318 Climate Change and Vulnerability
Similarly, the indicator for sensitivity of seed availability to climate is scored 4.17 if availability to the household is sensitive and zero if it is not, and the lack of adaptation strategies for the household would result in a score of 4.17 for the adaptation indicator, while existence of adaptation strategies would give a score of zero. Similar logic is followed in giving weights to the other subcategories under the other three component indicators of vulnerability, water, livelihood and health. As an alternative to the researchers’ equal weighting of indicators, we used the local communities’ perspectives to determine the weights. In two separate focus group discussions conducted with two different clusters of barangays in the municipalities of Pantabangan and Carranglan, participants were asked to provide their own weights for the indicators. The objective was to determine whether there would be significant variation in the weights provided by the two groups and to determine the likely implications of this in the use of multilevel indicators of vulnerability. Consensus was sought from the participants during the focus group discussions on the specific weights that they should assign for each component indicator at various levels. The computed vulnerability indices, which are the weighted summations of the food, water, livelihood and health sub-indices are correlated with factors hypothesized to influence vulnerability using Spearman correlation. These include a combination of demographic, socioeconomic and geographic factors, as well as the number of coping mechanisms practised by each household. Regression analysis is used to determine the combined effects of the hypothesized determinants of household vulnerability.
Mapping of vulnerable areas Vulnerable areas within the watershed are mapped using two different approaches. One uses spatially referenced data for biophysical variables considered to be associated with the vulnerability of different land-use categories to map vulnerability. The other uses a participatory mapping approach in which stakeholders identify vulnerable areas. In the first approach, areas of grassland/ brushland, cultivated land and forests are classified as having low, moderate and high vulnerability using five parameters: slope, elevation, distance from the nearest road, distance from the nearest river and distance from the nearest community centre. A single vulnerability map is developed by overlaying all of the individual maps produced for each of the five parameters. The vulnerability of grassland and brushland in the watershed is assessed in terms of their susceptibility to human-induced fire, which is common in the watershed and which has been observed to increase during a prolonged dry season. Proximity to human activities is a key determinant of vulnerability to fire. The closer an area is to a road, river or community, the greater is the chance that the area may suffer from human-induced fire. On the other hand, vulnerability to fire is low when the slope gradient and elevation of an area are low, as it is easier to control fire in favourable terrain. In agricultural areas, vulnerability is assessed in terms of their susceptibility to soil erosion from large rain events, a commonly observed problem
3243 J&J Vulnerability
1/11/07
10:24 am
Page 319
Climate Variability and Extremes in the Philippines 319
associated with climate variability and extremes in the watershed, and in terms of their accessibility to farmers. Higher elevation and more steeply sloped farmlands are more vulnerable to erosion than are lands at lower elevations or with less slope, while farms that are far from a road, river, or community centre are more difficult to manage and therefore more vulnerable. For forests, the vulnerability level is assessed in terms of ease of management and protection, associated with terrain conditions. Areas with favourable gradient and elevation are rated with low vulnerability since management and protection are less challenging here than in areas with adverse terrain. Forests that are more accessible (closer to roads, rivers and community centres) are rated as more vulnerable than forests that are difficult to access due to the greater chances of their being encroached on or cleared and converted to cultivated areas. The alternate approach had local communities identify vulnerable places during the focus group discussions using a participatory vulnerability mapping technique. Unlike the GIS-generated vulnerability map, which categorized the physical vulnerability of the watershed into low, moderate and high levels, the places identified by the communities as vulnerable do not have these categories. Instead, the participants of focus group discussions were just asked to identify locations of vulnerable areas on the barangay map, which they themselves had drawn, and to explain the reasons why they think these areas are vulnerable. GPS readings were made of the vulnerable places and plotted on the vulnerability map of the watershed developed using GIS. The idea is to examine the congruences and divergences between vulnerable areas identified using biophysical parameters and the areas that stakeholders judge to be vulnerable.
Results and Discussion Past climate variability and extremes in the Pantabangan–Carranglan watershed Considering the watershed’s geographic location, it can be said that all the communities living there are significantly exposed to natural hazards such as climate variability, climate extremes and other natural calamities like earthquakes. Data available from the Philippine Atmospheric, Geophysical and Astronomical Services Administration (PAGASA) indicate that from 1980 to 1995 a total of 58 strong typhoons – an average of about 4 typhoons per year – inflicted major damage in the area. In addition, three major drought episodes occurred during the same period, with an average interval of only about four years between episodes. These drought episodes occurred in 1983, 1987 and 1991, the three years with the lowest total annual rainfall and water inflow in the period 1980–2001. Participants in our focus group discussions were asked to recall past climate variations and extremes; the major events they identified are listed in Table 15.2. The respondents recalled two particularly strong typhoons that
3243 J&J Vulnerability
1/11/07
10:24 am
Page 320
320 Climate Change and Vulnerability
occurred in the 1970s, the local names of which, Kading and Didang, left an indelible mark in the minds of the respondents because of the great destruction that they brought to the communities within the watershed. Respondents also noted several symptoms of El Niño episodes in 1979–1980, 1982–1983 and 1997–1999, which correspond with El Niño events as recorded by PAGASA. Prolonged rains were also observed by the respondents in 1984, which marked the occurrence of a weak La Niña event. Table 15.2 Major climate events identified by participants in key informant interviews and focus group discussions Year
Climate Variability and Extremes
1974 1978 1979–1980 1982–1983 1984 1989 1997–1999 2000 2001 2002 2003
Typhoon Didang Destructive typhoon Kading Drought/El Niño El Niño Prolonged rains Delay in the onset of rainy season El Niño Delay in the onset of rainy season Early onset of rainy season Delay in the onset of rainy season Early onset of rainy season
Greater variability in the onset of the rainy season has been observed since 2000, making the onset of rains less predictable. Forest fires have been more frequent in the area since the 1980s. Between 1980 and 1988, the DENR recorded an average of 43 forest fires annually in the Pantabangan–Carranglan watershed, damaging a total area of 25,783ha over the 9 years. Although the high frequency of forest fires coincided with the almost cyclic occurrences of climate variability and extremes, such as El Niño and delays in the onset of the rainy season, greater fire frequency cannot be attributed to climate factors with high confidence. This is because most forest fires, according to respondents, are set off intentionally by people practising kaingin (slash-and-burn farming) and charcoal-making, practices that have also increased since the 1980s. Although the exact value of damages inflicted by past climate-related events in the watershed is not available, anecdotal evidence gathered during the survey and focus group discussions affirm that significant losses have been incurred. These losses include human lives, property and infrastructure, and sources of livelihood, especially farmlands and fishing areas. Pronounced decreases in crop yields have been observed in years with climate patterns that are unfavourable for the region’s agriculture. For instance, records from the NIA indicate that rice yield fell by more than two cavans (1 cavan = 50kg) per hectare below average in both the wet and dry season cropping periods of 1990 as a result of drought and typhoons. Individual yield changes, however, can
3243 J&J Vulnerability
1/11/07
10:24 am
Page 321
Climate Variability and Extremes in the Philippines 321
vary substantially from the average, and locals reported that crop loss from droughts and floods can be as much as 100 per cent for individual farms. Indeed, some community members are so vulnerable that even before they can fully recover from the adverse impacts of previous events, another calamity will strike again and force them back to their original desperate state.
Mapping vulnerable areas Our assessment of the spatial distribution of biophysical vulnerability of different land-use types using the five parameters discussed earlier (slope, elevation, distance from the nearest road, distance from the nearest river and distance from the nearest community centre) results in more than 65 per cent of the entire watershed being classified as moderately vulnerable and more than 25 per cent as highly vulnerable (Figure 15.3). Most of the forest and brushland areas are classified as highly vulnerable, with forests comprising most of the highly vulnerable land area, mainly due to their location in steep and highly elevated areas. Some grassland areas are also classified as highly vulnerable, but most grasslands are classified as moderately vulnerable, as are most of the forest plantations. The alienable and disposable lands, which consist of residential and cultivated areas, are mostly classified as moderately vulnerable, with some patches of low and high vulnerability. Also shown in Figure 15.3 are the places identified as vulnerable by the local communities themselves during focus group discussions. The participants tended to emphasize risks to people and their livelihoods over biophysical parameters in their identifications of vulnerable places. Places identified as vulnerable include low-lying flood-prone settlement areas, agricultural areas prone to floods and drought impacts, intermittent streams/rivers, farmlands at the tail end of irrigation canals, highly erodible areas on steep slopes along riverbanks, unstable areas with steep slopes that support infrastructure, and grasslands, forests and forest plantations near roads and settlements that are susceptible to fire. In total, 86 places were identified as vulnerable by participants in the participatory mapping exercises. These tend to cluster in areas of human settlement and cultivated farmlands, reflecting the concern with potential impacts on people and their livelihoods. Most of these (74 per cent) lie in areas classified as moderately vulnerable based on biophysical parameters, while 15 per cent and 11 per cent are in areas classified as having high and low vulnerability respectively. None are located in the forest areas along the northeastern border or in the southeastern corner of the watershed, which are classified as highly vulnerable in biophysical terms but which are relatively distant from the main population centres. While the vulnerable places identified by the local communities do not distinguish between areas of moderate and high vulnerability, they are more specific in their location. They also reflect the concerns of local communities with respect to risks associated with climate variability and extremes such as flood damages, soil erosion, water shortage and forest fires. On the other hand, the vulnerability map based on biophysical parameters and land-use categories
3243 J&J Vulnerability
1/11/07
10:24 am
Page 322
322 Climate Change and Vulnerability
Figure 15.3 The location of vulnerable areas in the Pantabangan–Carranglan watershed as identified by analysis of biophysical data and by participatory mapping with community members
can have advantages for macro-level planning to reduce vulnerability in the entire watershed. An approach that combines the two methods of identifying vulnerable areas could thus produce a more comprehensive assessment of vulnerable areas to be used to better address vulnerability in the watershed.
Groups vulnerable to climate variability and extremes Vulnerability indices were calculated for 108 households using data collected from the household survey and the 2 different schemes for weighting the component indicators. Mean, minimum and maximum values of the indices for the entire sample and for farmers and non-farmers are presented in Table 15.3.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 323
Climate Variability and Extremes in the Philippines 323
Farmers, in general, are more vulnerable to climate variability and extremes compared to non-farmers. This finding is true regardless of the source of the weights used in the index, in other words whether determined by the researchers or the local communities themselves. However, the index developed using the researchers’ weights produced both the highest (67) and lowest (4) index values (the index developed with weights provided by the local communities had 59 as the maximum and 12 as the minimum). Table 15.3 Vulnerability index values for farmers, non-farmers and the full sample, based on weights provided by researchers and local communities Source of Index Weights Researchers Farmers Non-farmers Combined Local Communities Farmers Non-farmers Combined
No of Vulnerability Index (Possible Value from 0 to 100) Respondents Mean Minimum Maximum 70 38 108
38 25 33
7 4 4
67 43 67
70 38 108
43 26 37
19 12 12
59 55 59
The values of the index are sensitive to the perceptions or experiences of whoever is giving the weights. For example, under the seed availability sub-index, households with an adaptation strategy get a weighted score of 2, but those with no strategy are scored 1. This suggests that those with an adaptation strategy are more vulnerable, which appears counter-intuitive. However, when asked in the focus group discussions if this was a mistake, participants maintained that the weights are logical. They noted that some adaptation strategies used in response to the lack of seeds, for example, shifting to other crops or buying hybrid varieties of seeds, can result in higher risk because of the high cost and the possibility that financial debt would be incurred. If the higher cost seeds fail to produce an adequate crop, for example, due to flooding or dryness, the indebted farmers are exposed to even greater financial losses and even the risk of losing their farms. Sometimes, in the experience of the farmers, doing nothing but waiting until seeds become available (no adaptation) can be a better option. Farmers are not a homogeneous group and are not all equally vulnerable. During the focus group discussions, the local community members identified at least three categories of farmers, as well as other socioeconomic groupings in the Pantabangan–Carranglan watershed, that have varying degrees of vulnerability to climate variability and extremes. These are small, average and rich farmers, fishermen, employees and small-business entrepreneurs (see Table 15.4). The focus group discussions identified small farmers as the most vulnerable group. These farmers are characterized as having very low educational
3243 J&J Vulnerability
1/11/07
10:24 am
Page 324
324 Climate Change and Vulnerability
attainment, not owning land, having meagre income, and lacking capital and access to other productive resources. Often they live in places considered to be vulnerable and have adaptation strategies for managing risks caused by variable and extreme climate that are ineffective. Table 15.4 Groups vulnerable to climate variability and extremes Description
Socioeconomic Groups Small Farmers
General socioeconomic characteristics
Average Farmers Employees/Small Rich Farmers and and Fishermen Entrepreneurs Overseas Workers
Landless; low Elementary and educational some secondary attainment; no education; some capital; very low access to productive income; almost no resources such as access to other land, capital and productive resources. ltechnology Nature of impacts Decline in crop Decline in crop/fish of climate production, food, harvest and income, variability and livelihood, health food availability, extremes condition; more livelihood sources; debt incurred health condition may or may not be affected Degree of negative High Moderate impacts Examples of High-interest loans Plant vegetables adaptation or borrowing from along rivers, plant strategies relatives; plant other crops; engage vegetables along in other sources rivers; plant other of livelihood crops; work in nearby towns; engage in other jobs Effectiveness of Some effective, Effective adaptation others not strategies Location of Some are located in Some are located in settlement/ vulnerable areas vulnerable areas properties relevant to vulnerable areas Degree of High Moderate vulnerability
College or highschool graduates; some access to productive resources such as land, capital and technology Increase in prices of commodities; reduced sales
Moderate
College or highschool graduates; more access and control over productive resources
Decline in production and income
Low
Government loans; Store food and farm engage in backyard inputs projects; store food supply and other farm inputs for sale
Effective
Effective
Some are located in Generally located in vulnerable areas secure areas
Moderate
Low
The groups considered to be moderately vulnerable include fishermen, average farmers, owners of small enterprises, sawali makers (sawali is a walling material made from bamboo), and employees of various agencies and businesses. These groups generally are better educated than the small farmers and have
3243 J&J Vulnerability
1/11/07
10:24 am
Page 325
Climate Variability and Extremes in the Philippines 325
greater access to productive resources such as land, capital and technology, although they don’t always have control over them. Incomes are modest but usually sufficient to provide for basic necessities, though some have incomes below the annual per capita poverty threshold (13,843 pesos per year in the Central Luzon region). Some live in vulnerable places such as low-lying floodprone areas and those with limited water sources. Compared to the most vulnerable group, they are relatively less sensitive to climate-related losses because of their greater access to resources and have more effective options for adaptation. The least vulnerable groups are the rich farmers and households with family members working overseas. Affluent farmers in general are the best educated among the three groups of farmers. They usually own large tracts of farmland, possess investment capital, own farm machinery and tools, and have control over other factors of production. They live in areas that are less susceptible to flooding and have effective adaptation strategies. Overseas workers are also among the better educated, have access to financial resources and have linkages with other institutions outside the community. Their families are considered among the least vulnerable group because the financial support they provide is stable and not affected by variable and extreme climate events in the local area. Similar to the well-off farmers, their families also live in safe places and have effective options for adaptation.
Vulnerability to future climate change Projected annual mean temperature increases over Southeast Asia from 1980–1999 to 2080–2099, from climate model simulations reviewed in the recent report of the Intergovernmental Panel on Climate Change (IPCC), have a median of 2.5°C and a 25th–75th percentile range of 2.2 to 3.0°C (Christensen et al, 2007). Most of the models project increases in the average annual precipitation for Southeast Asia, with a median of 7 per cent and a 25th–75th percentile range of 3 to 8 per cent. Seasonal changes vary strongly, with areas away from the Intertropical Convergence Zone showing a tendency for precipitation decreases. However, results from different models are as yet too varied to generalize about the effects of global warming on precipitation in the region (Boer and Faqih, 2004). With respect to extremes, Southeast Asia is likely to share the global tendency for daily extreme precipitation to become more intense, particularly in areas where mean precipitation increases, and tropical cyclones in the Pacific are likely to see an increase in the intensity of rain and wind (Christensen et al, 2007). Changes in the climate are likely to have far reaching consequences for the Pantabangan–Carranglan watershed. Participants in the focus group discussions were engaged in a scenario building exercise to explore vulnerabilities to a future with higher temperatures, greater variability of rainfall and more intense precipitation events. Through this exercise, potential outcomes are identified that are of concern to the participants and to which they consider themselves to be vulnerable. The outcomes identified as concerns are not predictions or projections of future impacts, but they do provide information
3243 J&J Vulnerability
1/11/07
10:24 am
Page 326
326 Climate Change and Vulnerability
about where potential vulnerabilities lie and indications of priorities for directing efforts to prevent or adapt to adverse impacts. Declines in crop production, greater risk of poverty, hunger and malnutrition, greater indebtedness, loss of access to farms due to unpaid debts, and shortages of water for farm and domestic use were identified as concerns by small farmers. Strategies that small farmers say they could apply to reduce risks of climate impacts include labour on other people’s farms and non-farm labour both locally and distant from their homes, adoption of drought tolerant crops, and sawali production. Access to high-interest loans may no longer be an option since they may not have sufficient collateral to guarantee them. Temporary relocation in times of extreme weather events such as typhoons can be used to reduce risks, but permanent relocation to safer areas is not an option. Average farmers, fishermen, employees and small entrepreneurs are concerned about and would be moderately vulnerable to potential climate impacts on the productivity of farmland, fisheries and other livelihood resources, incomes, prices of commodities, and water scarcity. Future climate impacts are not seen as major threats to the livelihoods, food security or health of these groups as a few adjustments in their expenditures, investments and other activities would enable them to cope with potential negative impacts. Available adaptation options include use of short-cycle crops, food storage, changes in household expenditure patterns, investment in other businesses to diversify risks and employment in other areas. Should the need arise, they also have the capacity to transfer to less vulnerable places. Rich farmers generally have low vulnerability to future climate impacts. Adverse impacts are expected to be readily managed and would not pose important threats to their health or food security. Participants in the scenario building exercise think that rich farmers could even benefit from the situation by gaining farmlands and other possessions from poor farmers who default on their debts.
Factors that influence vulnerability While communities and households in the Pantabangan–Carranglan watershed are generally vulnerable to climate variability and extremes by virtue of their geographic location, their degree of vulnerability varies based on a combination of factors. These factors include the demographic and socioeconomic characteristics of the households, as well as the broader sociopolitical and institutional contexts of the communities. Spearman correlation analysis was conducted to identify household characteristics that are correlated with the two vulnerability indices; results are presented in Table 15.5. For the vulnerability index constructed using the researchers’ weights, we found three factors that have significant correlation with vulnerability: farm income, monthly food expenditures and farm distance to market. Farm income and distance to market are positively correlated with vulnerability, while monthly food expenditures are negatively correlated. In the case of farm income, those with high farm income tend to be more dependent
3243 J&J Vulnerability
1/11/07
10:24 am
Page 327
Climate Variability and Extremes in the Philippines 327
on a source of income that is highly sensitive to climate. Farmers who are more distant from markets can be cut off from markets and other services during the rainy season and floods, which can amplify their vulnerability. Monthly food expenditures are postulated to be negatively correlated with vulnerability because larger expenditures suggest greater financial capacity to acquire food and other necessities. Table 15.5 Results of Spearman correlation analysis of vulnerability indices Postulated Factors
Weights by Researchers Vulnerability Coefficients
1. Demographic age gender ethnic affiliation educational attainment household size 2. Socioeconomic total income household assets number of organizations joined farm size farm income number of transportation vehicles monthly food expenditures number of loans applied for number of information sources 3. Geographic farm distance to market 4. Overall coping mechanisms number of coping mechanisms
Level of Significance
Weights by Communities Vulnerability Coefficients
–0.079
–0.12
–0.06 0.014
–0.04 0.001
0.03 –0.18 0.18 –0.12 0.26* –0.07 –0.30* 0.07 0.01
0.01
0.03 –0.08 0.22* 0.32* 0.44 –0.017 –0.30* 0.13 0.11
0.24
0.05
0.01
–0.09
Level of Significance
0.05 0.01
0.01
0.21 0.03
*Indicates correlation coefficient estimates that are significant at 0.05 or 0.01 confidence levels.
The correlation analysis for the vulnerability index constructed using weights from the local communities also identified monthly food expenditure as significant and negatively correlated with vulnerability. Two other factors were also found to have a significant correlation with vulnerability: the number of organizations joined and farm size. A positive relationship is found between the number of organizations joined by farmers and their vulnerability. This positive correlation might result from more vulnerable households having greater motivation to join organizations, or it might be that the number of organizations joined fails to capture the nature and quality of organizations that can help households to cope with climate hazards. A positive correlation is also found between farm size and vulnerability, which runs counter to our finding that rich farmers, who have the largest farms, are the least vulnerable households in the watershed. Perhaps this can be explained by the fact that
3243 J&J Vulnerability
1/11/07
10:24 am
Page 328
328 Climate Change and Vulnerability
most farmers in the watershed usually devote their farms to a single commodity, rice, which could enhance their vulnerability. To further evaluate the influences of the different household characteristics on the households’ vulnerability, we conducted stepwise multiple regression analyses for both vulnerability indices. In the regressions for the vulnerability index based on weights provided by researchers, 5 of the 17 postulated predictor variables are found to be statistically significant determinants of household vulnerability (Table 15.6). The significant predictors are sex of household head, ethnic affiliation, number of organizations joined, land ownership and distance of farm to market. Households headed by women are found to be more vulnerable compared with male-headed households. The vulnerability of female-headed households may be attributed to overwhelming family burdens, lack of adult male labour, limited livelihood opportunities and higher poverty rates. Ethnic affiliation serves as a proxy to identify households that are migrants to the region. Ethnic groups that are migrants are found to be more vulnerable, which may reflect their difficulty in gaining access to land for cultivation since the watershed area is mostly classified as government land and therefore legally protected from further encroachment and cultivation by new settlers. Migrants are also unfamiliar with the area and may be unable to develop appropriate adaptation strategies or draw on social networks to cushion the adverse impacts of variable and extreme climate conditions. The number of organizations joined by the farmers is estimated to increase vulnerability, consistent with results of the correlation analysis. Land ownership is a significant factor, with households that don’t own land found to be more vulnerable than households that do. The distance of farms to market is positively and significantly related to vulnerability, affirming the relationship found in the correlation analysis. For the vulnerability index based on weights provided by the communities, we found four variables to be significantly related to households’ vulnerability. Two of the significant variables (ethnic affiliation and distance to market) that were significant predictors of vulnerability as measured using weights from the researchers are also found to be significant predictors of household vulnerability using the weights provided by the communities. Larger households are found to be more vulnerable compared with smaller households, probably because the former have more mouths to feed compared with the latter. Monthly food consumption is also found to be negatively and significantly related to vulnerability, consistent with the correlation analysis results. On the basis of the computed coefficient of determination, 46 and 44 per cent of the total variation in vulnerability using the weights provided by the researchers and the local communities respectively is explained by the above-mentioned significant variables (Table 15.6). This means that roughly 55 per cent of the vulnerability variance of the 2 indices is still unaccounted for at an aggregate level. There is therefore a need to look for other factors that may help explain household vulnerability apart from those identified in the regression model. In addition to the above-mentioned factors, the broader sociopolitical context in which communities interact influences their level of vulnerability. As mentioned earlier, the chain of development projects implemented in the area
3243 J&J Vulnerability
1/11/07
10:24 am
Page 329
Climate Variability and Extremes in the Philippines 329
Table 15.6 Results of stepwise regression analysis Predictors
Weights by Researchers Regression Coefficients
1. Demographic age gender –9.66 ethnic affiliation –10.11 educational attainment household size 2. Socioeconomic total income household assets number of organizations joined 9.74 farm size farm income number of transportation vehicles monthly food consumption number of loans applied for number of information sources land ownership –8.3 3. Geographic farm distance to market 0.0006 4. Overall coping mechanisms number of coping mechanisms Intercept 46.25 Coefficient of determination 0.46
Weights by Communities
Level of Significance
Regression Coefficients
Level of Significance
0.01 0.01
–0.29
0.01
0.28
0.05
–0.39
0.01
0.40
0.01
0.01
0.05 0.01
43.73 0.43
Note: Variables without corresponding coefficient values do not meet the 0.05 level of significance.
from 1971 to the present has in some ways created a sense of dependency on the part of the local communities on external assistance. This is because these projects, especially the resettlement scheme, were oriented to providing handouts with very little attempt towards building local capacities. Consequently, a culture of self-reliance was not fully developed, contributing to the vulnerability of some members of the local community, especially with the termination of these projects. Instead of perpetuating external dependency through such projects, it would be possible to create more positive impacts by implementing national policies and crafting a more responsive institutional support system to improve livelihoods and the capacities of households to cope with and manage risks. For instance, forest policy could be reformed to provide communities with greater participation in and control of forest projects in their localities so that they might reap a greater portion of the benefits. At present, the national forest policy prohibits timber harvesting in all watershed areas that support big infrastructure projects (like the Pantabangan–Carranglan watershed, which supports the Pantabangan dam), even if the communities themselves are involved in plantation establishment. This has discouraged their active partic-
3243 J&J Vulnerability
1/11/07
10:24 am
Page 330
330 Climate Change and Vulnerability
ipation in reforestation and forest protection activities and has led in many cases to deliberate burning of established plantations. In the absence of direct benefits from established plantations and because of limited sources of livelihood opportunities in the area, community members are compelled to engage in illegal cutting and charcoal-making to augment their meagre income. This has contributed to the degradation of some parts of the watershed and increases in biophysical vulnerability. The presence of the different institutions in the area, such as the NIA, NPC and DENR, could be a catalyst for increasing opportunities and resources for building local capacity. However, the main focus of these institutions is on protecting their investments. The interests of the local communities are only a secondary priority. Previous institutional efforts have not given attention to the provision of more sustainable sources of livelihood. Moreover, institutional support to anticipate and adequately plan for the occurrence of variable and extreme climate conditions has yet to be developed. Similarly, there are as yet no initiatives directed at enhancing current adaptation strategies and building capacity at the local level. Finally, the prevailing inequity that characterizes the Philippine social structure is evident in the Pantabangan–Carranglan watershed and contributes to the vulnerability of the poor community members. The communities’ own typology of small, average and rich farmers is a concrete reflection of the inequitable social structure that prevails in the area. The well-off members of the community have better access to and control over productive resources and have the option to live in safer places, putting them in a less vulnerable situation. They are also better able to capture most of the benefits from the different development projects due to stronger associations and linkages with the institutions that implement them.
Conclusions Given the same climate stressors, vulnerability varies among different socioeconomic groups depending on their access to production resources and other assets, options to live in or have their assets in less vulnerable areas, and the effectiveness of coping mechanisms or adaptation strategies. In addition, components of broader societal, policy and institutional contexts can exacerbate the adverse impacts of climate change, compounding the vulnerability of certain groups. Looking at the multiple stressors that contribute to people’s vulnerability – which include a combination of climate and non-climatic factors both at the micro and macro levels – is a useful way of understanding this complex concept. There is a need for bottom–up assessment and planning to address vulnerability and enhance adaptive livelihoods at the local level. Participatory action and research that engages the different stakeholders, particularly the local communities, should be pursued to minimize the vulnerability of the poor and enhance adaptive capacity at the local level. To reduce vulnerability, policies and development programmes should aim
3243 J&J Vulnerability
1/11/07
10:24 am
Page 331
Climate Variability and Extremes in the Philippines 331
to empower the local communities to broaden their range of choices of appropriate adaptation strategies, rather than making them dependent on external support. This should not, however, preclude questioning the large-scale structural causes of vulnerability such as poverty, inequity, and institutional and economic barriers to development.
Note 1
Adaptation strategies for the Pantabangan–Carranglan watershed are explored in Lasco et al (2008).
References Adger, W. N. (1999) ‘Social vulnerability to climate change and extremes in coastal Vietnam’, World Development, vol 27, pp249–269 Boer, R., and A. Faqih (2004) ‘Current and future rainfall variability in Indonesia’, Technical Report, AIACC Project no AS21, International START Secretariat, Washington, DC, US Brooks, N. (2003) Vulnerability, Risk and Adaptation: A Conceptual Framework, working paper no 38, Tyndall Centre for Climate Change Research, University of East Anglia, Norwich, UK Calenergy Company (2004) ‘Worldwide projects: Casecnan (Philippines)’, www.calenergy.com/html/projects5b.asp Christensen, J. H., B. Hewitson, A. Busuioc, A. Chen, X. Gao, I. Held, R. Jones, R. Koli, W. Kwon, R. Laprise, V. Rueda, L. Mearns, C. Menendez, J. Raisanen, A. Rinke, A. Sarr and P. Whetton (2007) ‘Regional climate projections’, in S. Solomon, D. Qin, M. Manning, Z. Chen, M. C. Marquis, K. Averyt, M. Tignor and H. L. Miller (eds) Climate Change 2007: The Physical Science Basis, Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US Chua, L. A. (1999) Understanding the Research Process, Department of Agricultural Education and Rural Studies, College of Agriculture, UP Los Baños, College, Laguna Cruz, R. V. O. (2003) ‘Watershed level SDI for global change’, in Sustainable Development Indicators for Global Change: A Multi-Scale Approach, Completion Report, Environmental Forestry Programme, College of Forestry and Natural Resources, University of the Philippines, Los Baños, Philippines Cutter, S. L. (1996) ‘Vulnerability to environmental hazards’, Progress in Human Geography, vol 20, pp 529–539 Few, N. (2003) ‘Flooding, vulnerability and coping strategies: Local responses to global threat’, Progress in Development Studies, vol 3, pp43–58 IPCC (Intergovernmental Panel on Climate Change) (2001) Climate Change 2001: Impacts, Adaptation and Vulnerability, edited by J. McCarthy, O. Canziani, N. Leary, D. Dokken and K. White, Contribution of the Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK Jose, A. M., L. M. Sosa and N. A. Cruz (1996) ‘Vulnerability assessment of Angat Watershed Reservoir to climate change’, Water, Air, Soil Pollution, vol 92, pp191–201
3243 J&J Vulnerability
1/11/07
10:24 am
Page 332
332 Climate Change and Vulnerability Lasco, R., R. Cruz, J. Pulhin and F. Pulhin (2008) ‘Spillovers and tradeoffs of adaptation: Examples from a Philippine watershed’, in N. Leary, J. Adejuwon, V. Barros, I. Burton, J. Kulkarni and R. Lasco (eds) Climate Change and Adaptation, Earthscan, London, UK Leary, N. and S. Beresford (2007) ‘Vulnerability of people, places and systems to environmental change’, in G. Knight and J. Jaeger (eds) Integrated Regional Assessment, Cambridge University Press, Cambridge, UK Moss, R. (1999) Vulnerability to Climate Variability and Change: Framework for Synthesis and Modeling: Project Description, Battelle Pacific Northwest National Laboratory, Richland, WA, US Municipality of Carranglan (undated) Development Master Plan of the Municipality of Carranglan, 2003–2007, Nueva Ecija Municipality of Pantabangan (undated) Master Plan of the Municipality of Pantabangan, 1998–2000, Nueva Ecija NPC (National Power Corporation) (1995) Pantabangan Watershed Rehabilitation Project, Watershed Management Department, Quezon City, The Philippines NPC (1997) Pantabangan–Carranglan Watershed Management Plan, Watershed Management Department, Quezon City, The Philippines National Statistics Office (2000) Census 2000: Philippines Population, CD-ROM, Barangay Pulhin, J. M. (2002) ‘Climate change and watershed communities: Methodology for assessing social impacts, vulnerability and adaptation’, paper discussed during the AIACC–AS 21 Regional Capability-Building Training Workshop on Climate Change Impacts, Adaptation and Vulnerability, College of Forestry and Natural Resources, 25 November to 8 December, University of the Philippines, Los Baños, College, Laguna, Philippines Saplaco, S. R., N. C. Bantayan and R. V. O. Cruz (2001) GIS-based ATLAS of Selected Watersheds in the Philippines, Department of Science and Technology, Philippine Council for Agriculture, Forestry and Natural Resources Research and Development (DOST-PCARRD) and Environmental Remote Sensing and GeoInformation (UPLB-CFNR-ERSG)University of the Philippines Los Baños, College of Forestry and Natural Resources Toquero, F. D. (2003) ‘Impact of involuntary resettlement: The case of Pantabangan resettlement in the province of Nueva Ecija’, PhD thesis, Central Luzon State University, Science City of Muñoz, Nueva Ecija UNEP (United Nations Environment Programme) (2001) Vulnerability Indices: Climate Change Impacts and Adaptations, edited by T. E. Downing, R. Butterfield, S. Cohen, S. Huq, R. Moss, A. Rhaman, Y. Sokona and L. Stephen, Policy Series, United Nations Environment Programme, Nairobi, Kenya Yoshida, S. (2000) ‘Pantabangan forestry development assistance project’, www.jica.go.jp/english/news/2000/16.html
3243 J&J Vulnerability
1/11/07
10:24 am
Page 333
16
Climate Risks and Rice Farming in the Lower Mekong River Basin Suppakorn Chinvanno, Somkhith Boulidam, Thavone Inthavong, Soulideth Souvannalath, Boontium Lersupavithnapa, Vichien Kerdsuk and Nguyen Thi Hien Thuan
Introduction Agriculture is one of the most important activities in the lower Mekong river basin. It is a source of livelihood for a large portion of the population and a significant contributor to national incomes. For example, in Lao PDR, agriculture employs 76.3 per cent of the country’s 5.7 million people (UNESCAP, undated) and agricultural products contributed 44.8 per cent of Lao PDR’s gross domestic product (GDP) of US$2.9 billion in 2005 (World Bank, 2007). In Thailand, agriculture contributed a much smaller 9.9 per cent of total GDP (US$176.6 billion) (World Bank, 2007), yet the sector employs 44.9 per cent of Thailand’s 63.1 million people (UNESCAP, undated). Rice is the most important agricultural product of the region in terms of the proportion of land area used, the quantity and value of output, and contribution to diet. In Thailand, rice is cultivated on 88 per cent of land used for cereal production and represents 43 per cent of the per capita daily caloric intake (FAO, 2004a). Rice is even more predominant in Lao PDR, where 94 per cent of cereal lands is planted in rice and 64 per cent of daily caloric intake is provided by it (FAO, 2004b). Most rice and other cereals are grown under rain-fed conditions as the irrigated land area is limited, accounting for 19 and 30 per cent of total harvested area in Lao PDR and Thailand respectively (Barker and Molle, 2004). Because of the high dependence on rain-fed rice cultivation, and the sensitivity of rain-fed rice yields to rainfall amounts and other climate conditions, the region is strongly affected by variations or changes in climate that adversely affect rice cultivation. Farmers of rain-fed rice are among the most vulnerable groups in the lower Mekong basin as their livelihood depends heav-
3243 J&J Vulnerability
1/11/07
10:24 am
Page 334
334 Climate Change and Vulnerability
ily on their annual production of rice, which is directly exposed to climate risk. In addition, most of these farmers are poor and have limited resources and other capacity with which to cope with the impacts of climate variability and change. The risk profile and vulnerability of rice farmers of the Mekong basin vary from place to place due to differences in the changes in climate to which they will be exposed, the sensitivity of the production systems to climate change, the socioeconomic condition and lifestyle of each community, and the condition of the surrounding natural environment (IPCC, 2001a). As part of a larger study of climate change in the lower Mekong basin, we investigated the existing climate risks faced by rice farmers in selected villages in Lao PDR and Thailand and how their risks may change with climate change (see Snidvongs, 2006). This chapter presents the results of our investigations. Our approach, which follows the Adaptation Policy Framework of the United Nations Development Programme (Lim et al, 2004), is outlined in Figure 16.1. The analysis includes development of climate change scenarios for the region, estimation of climate change impacts on rice yields, and assessment of the vulnerability of farm households to climate variations and change as a function of their sensitivity to climate risk, exposure and coping capacity. Strategies for adapting to climate change were also examined and are evaluated in Chinvanno et al (2008).
Figure 16.1 Framework for climate risk and vulnerability assessment
3243 J&J Vulnerability
1/11/07
10:24 am
Page 335
Climate Risks and Rice Farming in the Lower Mekong River Basin 335
The Study Sites Our study encompassed three countries of the lower Mekong basin: Lao PDR, Thailand and Vietnam. In this chapter, we focus on study sites in Savannakhet Province in Lao PDR and Ubon Ratchathani Province in Thailand. These two countries of Southeast Asia represent opposite ends of the scale of socioeconomic development, resulting in very different conditions that lead to differences in their farmers’ vulnerability to climate hazards and climate change. Thailand is far more economically developed than Lao PDR, as reflected by the per capita gross national income levels in 2005 of US$2720 in Thailand and US$430 in Lao PDR (The World Bank, 2007), and has a higher population and population density. The different level of development and socioeconomic conditions are reflected in different livelihoods (commercial farming vs subsistence), structure of household expenses, resources for coping and adapting to stresses, institutional support, and agricultural practices. But despite the differences, farmers of rain-fed rice in the two countries share the same cultural roots and are among the poorest members of their respective societies; in both countries their well-being is highly dependent on climatic conditions. We selected four villages for study in Lao PDR: Seboungnuantay, Lahakhoke, Khouthee and Dongkhamphou. The villages, all located within the Songkhone district of Savannakhet Province, have a total land area of 1851ha and a population of 2490 living in 434 households. Savannakhet Province is in the central to southern part of Lao PDR, has a land area of 21,774km2 and consists of 15 districts. The topography of the province is lowland with a slight slope from east to west towards the Mekong river. Savannakhet Province has the largest area of rice fields in the country, nearly 140,000ha or 19 per cent of all rice fields in Lao PDR (Committee for Planning and Cooperation, 2003). It is also the most populated province of the country, with a total population of 811,400, or approximately 15 per cent of the population of Lao PDR. Songkhone district, where the study villages are located, is in the southwest of Savannakhet Province. It is the largest district of the province, with a total area of 1406km2. The district consists of 142 villages with 13,919 households and a total population of 86,855. Most of the inhabitants are subsistence farmers who grow rice mainly for their own consumption and sell only a small amount of their farm output in markets. Rice farming is rain-fed and a single crop is grown each year. Households supplement their food supply and livelihoods by harvesting natural products from surrounding natural ecosystems, which are relatively intact. Eighteen villages were selected for study in Thailand, all located in Ubon Ratchathani Province in the lower northeastern region of Thailand. The province covers an area of 16,112km2. Most of the land area consists of highlands, averaging 68 metres above sea level, with mixed sandy soils of low fertility. The Mekong river and mountains form the border between the province and Lao PDR to the east and high mountains form the border between the province and the Democratic Republic of Cambodia to the south. Major rivers include the Chi river, which merges with the Mun river and flows
3243 J&J Vulnerability
1/11/07
10:24 am
Page 336
336 Climate Change and Vulnerability
through Ubon Ratchathani Province from west to east before joining the Mekong in Khong Chiam district. In 2005, Ubon Ratchathani maintained a total population of 1,774,808 in 432,923 households, which are mostly in the agricultural sector (Department of Provincial Administration, undated). The study area is part of the Ubon Ratchathani Land Reform Area (ULRA), which covers 55,000ha on the east bank of the Dome Yai river. This area has three slope classes: level to gently sloping, sloping to undulating, and undulating to rolling. Soils are generally sandy and of low fertility. Korat series is the major soil type in this area; these soils are fairly well drained and strongly acidic. Most of the area is cultivated for paddy rice, with some areas cultivated for upland crops. There are small patches of degraded forests. Water is plentiful in the wet season, but severe shortage occurs in the dry season. Average rainfall is about 1600mm, 90 per cent of which falls in the period May to October. Average monthly temperature ranges from a minimum of 17.0°C in December and January to a maximum of 35.9°C in March and April. There is very limited irrigation and cropping is mainly a wet season activity (Ubon Ratchathani Province Administration, undated). Farmers are mostly commercial farmers who grow a single rice crop each year on farms of moderate size and using mechanized farming methods. The study area is divided into five zones, which are characterized in Table 16.1. Table 16.1 Villages by zone in Thailand Zone
Characteristics of zone
Villages studied
#1
Deep sandy soils. Cropping patterns are rice plus plantation and forest. The forest trees are eucalyptus and cashew nut.
#2
This area lies along the Lam Dom Yai river. Soil has high fertility. It is a wet area. The dominant cropping system is rice and upland crops such as vegetables, cassava or kenaf.
#3
The area is partly upland rice. The cropping system is an encroached forest area.
#4
This area has an intensive rice system. Mostly commercial farming practice. There is low tree density.
#5
This area is similar to zone # 3 but has more lowland characteristics. Rice cultivation encroaches into forest areas.
1. Ban Mak Mai 2. Ban Mek Yai 3. Ban Khok Pattana 1. Ban Fung Pa 2. Ban Muang 3. Ban Bung Kham 4. Ban Bua Thaim 1. Ban Nong Sanom 2. Ban Udom Chart 3. Ban Pa Rai 4. Ban Non Sawang 1. Ban Bua Ngam 2. Ban Nong Waeng 3. Ban Rat Samakee 4. Ban Non Yai 1. Ban Pa Pok 2. Ban Sok Seang 3. Ban Non Deang
3243 J&J Vulnerability
1/11/07
10:24 am
Page 337
Climate Risks and Rice Farming in the Lower Mekong River Basin 337
Projected Climate Change in the Lower Mekong Basin Our analyses of potential impacts of climate change on rice production are based on climate change scenarios constructed for our study areas from projections of the conformal cubic atmospheric model (CCAM), a high- resolution regional climate model. The CCAM is a second-generation regional climate model developed for the Australasian region by the Commonwealth Science and Industrial Research Organization (McGregor and Dix, 2001). Evaluations of the model in several international model inter-comparison exercises have shown it to be among the best climate models for reproducing key features of the climate of the Southeast Asian region (Wang et al, 2004). The baseline climate for the analysis is developed using a steady state simulation of the CCAM with an atmospheric concentration of carbon dioxide (CO2) of 360ppm, which corresponds to the CO2 concentration during the 1980s. Scenarios of future climate are developed using steady state simulations for CO2 concentrations of 540ppm and 720ppm, which correspond to 1.5 times and double the baseline level. These concentrations would be reached by roughly the 2040s and 2070s, respectively, for the IPCC’s A1FI scenario of greenhouse emissions, the highest of the IPCC emission scenarios (IPCC, 2001b). Figures 16.2 and 16.3 display baseline temperatures and precipitation for the region and the changes projected by the CCAM for CO2 concentrations of 540ppm and 720ppm. The CCAM simulations have a spatial resolution of 0.1 degree, or approximately 10km2. No results are shown for Cambodia due to insufficient observational data. For the 540ppm scenario, the CCAM projects that the region would get slightly cooler. For the 720ppm scenario, warming of less than 1°C is projected over most of Thailand and Lao PDR. Annual precipitation is projected to increase throughout the region for both climate change scenarios, with greater precipitation projected for the 720ppm CO2 concentration scenario than for the 540ppm scenario. The increases are greatest in the eastern and southern part of Lao PDR.
Climate change in the study areas To create climate scenarios for our study sites, the outputs of the CCAM model must be adjusted to match local climate conditions. The adjustment focused on precipitation and used observed data from weather stations throughout the region. The statistical procedure used to adjust the model output is based on cumulative rainfall using a non-linear log–log function to exponentially increase the daily variability. An arbitrary rainfall threshold of 3mm/day was applied to reduce the number of rainy days. In Savannakhet Province in Lao PDR, the rainy season is extended slightly for the 540ppm CO2 scenario as the onset of the rainy season is projected to shift approximately 10 days earlier. Total annual rainfall increases roughly 10 per cent from the baseline average of 1624mm to 1780mm. In comparison, the rainy season length would settle back to the same condition as the baseline when the CO2 concentration rises to 720ppm. However, total rainfall increas-
3243 J&J Vulnerability
1/11/07
10:24 am
Page 338
338 Climate Change and Vulnerability
Figure 16.2 Average temperature in the lower Mekong river basin: Baseline and projected changes
es by a larger amount, about 20 per cent above the projection for 540ppm, to 2120mm. Projected temperatures for Savannakhet only change within the range of +/-1 degree C as more cloud cover locally dampens the global warming trend. In Ubon Ratchathani Province of Thailand, the onset of the rainy season is projected to start much earlier, by about 20 days, for both the 540 and 720ppm CO2 scenarios. The simulated 10-year average annual rainfall is 1688mm during the baseline period and it rises to 1734mm and 1901mm for the 540 and 720ppm scenarios, respectively. However, despite the increased rainfall, the mid-season dry spell becomes more prominent for the 540ppm simulation. The temperature in the area would change within a narrow range of +/-1°C, which is also projected for the study site in Lao PDR, again because more cloud cover in the region dampens the warming trend.
Comparison to other projections of climate change for Southeast Asia To put the CCAM-derived scenarios in context, it is useful to compare them to the range of climate projections from other models. The projections of future temperature increases in Southeast Asia that are assessed in the IPCC’s most recent report range from 1.5 to 3.7°C average annual warming over the 100-
3243 J&J Vulnerability
1/11/07
10:24 am
Page 339
Climate Risks and Rice Farming in the Lower Mekong River Basin 339
Figure 16.3 Average rainfall in the lower Mekong river basin: Baseline and projected changes
year period from 1980–1999 to 2080–2099 (Christensen et al, 2007). Seasonal warming is roughly the same for each season as the projected change in average annual temperature. The median projected warming for the region is 2.5°C, similar to the global average, while the 25th and 75th percentile projections span a range of 2.2 to 3.0°C. Somewhat greater warming is projected over Indochina and the larger land masses of the archipelago. Note that none of the projections assessed in the new IPCC report indicate cooling for the region and that the CCAM projection of temperature changes for 720ppm is below the range projected by other models. So, our analyses are based on scenarios that are significantly cooler than other models have projected. The projected increase in precipitation from the CCAM is consistent with other model projections for the region. Most of the models reviewed by the IPCC project increases in precipitation averaged over all Southeast Asia, with a median increase of about 7 per cent in all seasons (Christensen et al, 2007). But there is potential for substantial local variations in precipitation changes, as demonstrated by McGregor and Dix (2001). For example, precipitation decreases are often projected in areas away from the Intertropical Convergence Zone (ITCZ) (Christensen et al, 2007). In areas where mean precipitation is projected to increase there is also the potential for more intense daily extreme precipitation.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 340
340 Climate Change and Vulnerability
Impact of Climate Change on Rice Yields The Decision Support System for Agrotechnology Transfers (DSSAT) version 4.0 crop modelling software (Hoogenboom et al, 1998) and the climate scenarios generated from the CCAM climate model are used to simulate the impacts of climate change on rain-fed rice yields at the study sites. The crop modelling software uses daily climate data, including maximum and minimum temperature, precipitation and solar radiation, coupled with the crop management scheme and soil property of the study sites, to calculate the rice yields. By using daily climate data for the simulation process, our study is able to capture the impact of climate change on rain-fed rice productivity not only with respect to changes in average climate parameters such as rainfall and temperature, but also with respect to changes in temporal aspects of climate such as shifts in the onset of rains, changes in the length of the rainy season or changes in the pattern of the mid-season dry spell. The DSSAT simulations also incorporate the direct effects of higher CO2 concentrations, which can increase yields by increasing photosynthesis and plant water-use efficiency. The crop management scheme used in the simulations assumes homogeneous practice in each site. The crop management scheme is comprised of choice of rice cultivar, planting date, initial condition of the field before planting, planting method and density, water management, and application of organic and inorganic fertilizers. Results of the simulations are shown in Table 16.2. Simulation results for the baseline case differ somewhat from actual yields as recorded from field interviews. Differences in yields are due, in part, to differences between modelled and actual farm management practices and differences between the dataset used for the simulations and actual field conditions, particularly for soil properties. However, the simulations provide useful indicators of the future trend and potential impacts of climate change on rice productivity in the study areas. Table 16.2 Simulated rice yields under different climate scenarios Rice Yields (kg/ha) Climate Scenario
Lao PDR Savannakhet Province Songkhone District Thailand Ubon Ratchathani Province Zone 1 Zone 2 Zone 3 Zone 4 Zone 5
Change from Baseline
360ppm CO2 Average Average (Baseline) Climate for Climate for 540ppm 720ppm CO2 CO2
540ppm
720ppm
CO2
CO2
2535
2303
2470
–9.1%
–2.6%
1154 1920 2364 2542 3024
1235 2002 2408 2575 3051
1331 2072 2439 2592 3069
7.0% 4.3% 1.9% 1.3% 0.9%
15.3% 7.9% 3.2% 2.0% 1.5%
3243 J&J Vulnerability
1/11/07
10:24 am
Page 341
Climate Risks and Rice Farming in the Lower Mekong River Basin 341
According to the climate change scenarios simulated by the CCAM, climate change has a slight negative impact on rain-fed rice production in Savannakhet Province in Lao PDR. The simulated rice yield is reduced by nearly 10 per cent under climate conditions corresponding to a CO2 concentration of 540ppm, but for the 720ppm CO2 scenario, yields rise back to almost the same level as the baseline scenario. The simulated rice yields at the study sites in Ubon Ratchathani Province in Thailand show positive impacts from climate change. The increase in rice yield varies from zone to zone and is greater for the 720ppm CO2 scenario than for the 540ppm scenario. The increases range from roughly 1 to 7 per cent for the CO2 concentration of 540ppm and 1.5 to 15 per cent for the 720ppm climate. The mild impact of climate change on rice yields in the Lao PDR sites and the positive impacts at the Thai sites are due primarily to three factors: the beneficial effects of carbon dioxide and increased rainfall for rice cultivation and the relatively modest temperature changes of the climate scenarios used in the analysis. Scenarios with greater warming would likely result in less beneficial outcomes or even negative outcomes. It is also worth noting that the simulations do not take account of the potential effects of more intense rainfall, flooding and changes in the timing of rainfall, which are discussed in the following section.
Farmers’ Concerns and Extreme Events Interviews with farmers in the study areas revealed that farmers are already threatened by climate variability. Farmers are highly concerned about extreme climate events that can cause substantial losses of farm output and threaten their livelihoods. Extreme events identified by farmers as threats to rice cultivation in the study areas include prolonged mid-season dry spells, floods and late-ending rainy seasons. Farmers of rain-fed rice sow their rice at the start of the rainy season, typically in May, or transplant seedlings into their fields in mid-June to mid-July, and harvest their crop in October or November after the end of the rainy season. A mid-season dry spell after sowing or transplanting rice is common to the region. The dry spell can damage young rice plants or impose additional costs on farmers for water procurement to sustain the rice plants while waiting for the rains to resume. If plants are lost but the resumption of rains does not come too late, the farmer can replant to salvage his harvest, but again incurring additional expenses. In the worst case, the midseason dry spell is prolonged and rains resume too late for replanted rice to mature before the rainy season ends. When very prolonged dry spells occur, farmers are at risk of losing a substantial portion of their crop and income. Floods are also a significant threat to rice cultivation in the lower Mekong basin. Floods commonly occur near the end of the rainy season, around the months of October and November, when water flow is at its highest in the Mekong river and its tributaries. This period of frequent flooding coincides with the middle to end of the crop season. Late season floods have caused severe damage to rice production, and recovery is difficult as it is too late in the
3243 J&J Vulnerability
1/11/07
10:24 am
Page 342
342 Climate Change and Vulnerability
rainy season to replant. Another source of risk is a late end to the rainy season. Rains during and after harvest can damage the harvest or result in higher costs for drying the rice. Our simulations of the impacts of climate change on rice productivity do not take into account potential changes in the timing, duration or severity of events such as dry spells, heavy rains and floods. But the greatest climate risks to farmers are currently from extreme events and it is changes in extremes that are of greatest concern to farmers. Thus, a complete assessment of climate change risks and vulnerability needs to consider potential changes in the distribution frequencies of extreme events. However, this requires climate scenario simulations for longer time periods than the 10-year time slices constructed for our study. In order to gauge the sensitivity of farmers to the occurrence of extreme climate events, we examine the impacts of a hypothetical extreme event on farm household risk profiles. Group discussions with farmers and community leaders in the study sites indicate that an event causing a loss of approximately one third of rice production or higher would be a severe situation that would significantly affect a farmer’s livelihood. Therefore, a loss of 30 per cent of rice production is used as a proxy for an extreme climate event in our analysis.
Baseline Climate Risk The level of climate risk faced by farm households is a function of three broad determinants: the sensitivity of the household to stresses in climate variations and changes, the exposure of the household to climate stresses, and the capacity of the household to cope with climate impacts. A variety of indicators are used to measure these three determinants of risk (see Table 16.3). Indicators of household economic condition are used to measure the sensitivity of the farmer household to climate stresses. Households with current consumption that is sustainable within the limits of household income, land ownership and farm size, allowing self-sufficient food production, have low sensitivity to climate stresses. The degree of dependency on farm production and rice production are used to measure the exposure of the farmer household, with low levels of dependency indicating low exposure. Coping capacity is measured by the diversity and amount of resources available to the farmer household for responding to and recovering from climate impacts. Within this conceptual framework, farmer households are at high risk if they have an unstable or unsustainable household economic condition, are highly reliant on rice production for their livelihood, and have few resources for coping with climate impacts. Data on the indicators was collected through field interviews of 560 farmer households in Thailand and 160 farmer households in Lao PDR. The field assessment activity in Thailand was conducted by researchers from the Faculty of Agriculture of Ubon Ratchathani University during May–July 2004. The assessment in Lao PDR was conducted by researchers from the National University of Laos during September 2004.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 343
Climate Risks and Rice Farming in the Lower Mekong River Basin 343
Table 16.3 Indicators used in evaluating farmers’ risk from climate impact Criteria
Indicator
Measurement
Scoring
Min score
Max score
Household Economic Condition
Sustainability of household consumption
Total household production (or income)/total household consumption (or expenditure)
>1 = 0; 1–0.7
0
2
Stability of household production
Farmland: own or rent
Own = 0, Rent = 1
0
1
Self-sufficiency of household food production
Farmland/capita 0.8ha/capita for Lao PDR and 0.65 for Thailand are thresholds to produce annual food consumption for one family member
≥0.8 = 0; <0.8 = 1 (Thailand ≥0.65 = 0; <0.65 = 1)
0
1
0
4
= 1; <0.7 = 2
Sub-total Household Dependency on On-Farm Production
Availability of income from nonclimate sensitive sources Dependency on rice production to sustain basic needs
Total household consumption/ income from livestock + Fixed off-farm income
>1 = 0; 1–0.7 = 1; <0.7 = 2 0 2
0
2
Total rice production/total food expenditure (or Total household fixed expenditure)
>1 = 0; 1–0.7=1; <0.7 = 2
0
2
0
4
≤1 = 0; 1–1.3 = 1; >1.3 = 2
0
2
≤1 = 0; –1.3 = 1; 1>1.3 = 2
0
2
0
4
0
12
Sub-total Coping Capacity
Ability to use non- Total household consumption + farming income to Total cost of production/total maintain livelihood household saving + Total off-farm income + Income from livestock + Extra income Ability to use non- Total food expenditure (or farming income to Total fixed expenditure)/total maintain household household saving + Total offbasic needs farm income + Income from livestock + Extra income
Sub-total Total
The collected indicator data were combined into an index of climate risk using the scoring system outlined in Table 16.3. Farm households are grouped into three risk categories according to their scores as follows: • • •
Low risk: households with risk scores in the range 0–4; Moderate risk: households with risk scores in the range 5–8; and High risk: households with risk scores in the range 9–12.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 344
344 Climate Change and Vulnerability
The proportions of farm households classified as having low, moderate and high risk in the current climate for each of the study sites are shown in Figure 16.4. Farm communities in Savannakhet Province in Lao PDR are found to be highly resilient to climate stresses relative to the farm communities of Ubon Ratchathani Province in Thailand. More than 80 per cent of the households in the Laotian villages are classified as low risk and less than 5 per cent are classified as high risk. In comparison, farmers at the study sites in Thailand are at greater risk from climate impacts. Only about a third of the surveyed population are classified as low risk, while approximately 15–25 per cent are in the high risk category. The moderate risk group is the largest group of the population, in some study sites accounting for as many as half of the total surveyed population.
Figure 16.4 Climate risk levels of farm households under current climate conditions The contributions of exposure, sensitivity and coping capacity to household risk scores are displayed in the risk profiles in Figure 16.5. The low risk groups in every location have risk profiles that differ substantially from the moderate and high risk groups. Their risk scores are low in every criterion. In most cases the biggest difference between the low risk and higher risk groups is that the higher risk groups have much less coping capacity. Greater exposure to climate stresses is also a significant contributor to the greater risks faced by households classified as moderate and high risk. The total risk scores of the low risk groups in Lao PDR and Thailand range roughly from 1 to 2 points, while the total risk scores of the moderate and high risk groups average close to 7 and 10 points respectively.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 345
Climate Risks and Rice Farming in the Lower Mekong River Basin 345
Figure 16.5 Climate risk profiles under current climate conditions The large proportion of rain-fed rice farmers in Lao PDR that are at low risk from climate stresses have high coping capacity relative to other farmers in the study. This is partly because their household production is diversified over various activities, including both on-farm and off-farm sources. Consequently they can accumulate and draw on a wide range of resources with which to cope with climate and other stresses. Rice production for these farmers does not dominate household production and accounts for less than a third of total household output. Another advantage of farmers living in rural areas of Lao PDR is that, due to the low population, natural systems are still able to provide a significant alternate food source and forest products that can be converted or exchanged for other products required for daily use or sold for cash. In addition to relying on natural ecosystems as a coping mechanism, farmers in Lao PDR also have savings in the form of stored rice and cash-convertible livestock to help them cope with impacts from climate stresses, even though cash saving is almost non-existent. In addition, the debt level of farmers in Lao PDR is virtually zero, partly due to the limited availability of loans or other institutional lending mechanisms, but also to social norms that are against indebtedness (Boulidam, 2005). The majority of surveyed farmers in Thailand are categorized as moderate or high risk. The most important factor contributing to their risk level is very limited coping capacity due to their having few savings and high debts. In addition, the surveyed farmers in Thailand have little diversification in their production and income sources and are highly dependent on income from rice production. Their dependency on rice production creates conditions of high exposure and sensitivity to climate impacts.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 346
346 Climate Change and Vulnerability
Vulnerability to Climate Change Our analysis thus far has addressed current or baseline climate risks to farmers. Climate change will change the stresses to which farm households are exposed and result in a variety of impacts. Potential impacts include changes in yields of rice and other crops, the availability of water, costs of planting, replanting and water procurement, and the frequency and severity of crop losses to floods and dry spells. Here we examine the potential impacts of climate change on rice production and how these impacts would affect the risk scores and risk profiles of farm households. Changes in rice yields for four scenarios are presented in Table 16.4. These include scenarios of average climate conditions corresponding to the steadystate CCAM projections for CO2 concentrations of 540 and 720ppm. The changes in rice yields are those derived from the DSSAT simulations, which indicate potential yield reductions in Savannakhet Province and yield increases in Ubon Ratchathani Province for the average projected climates. Climate change may also bring changes in extremes, such as late season flooding. To investigate how future climate extremes might affect farm households, we construct two scenarios of rice yields that assume that extremes reduce yield by 30 per cent relative to the simulated yields for average climate conditions. Table 16.4 Scenarios of changes in rice yields in response to changes in average climate and extreme climate 540ppm CO2
Lao PDR: Savannakhet Province Songkhone district Thailand: Ubon Ratchathani Province Zone 1 Zone 2 Zone 3 Zone 4 Zone 5
720ppm CO2
Average climate
Extreme climate
Average climate
Extreme climate
–9.1%
–39.1%
–2.6%
–32.6%
7.0% 4.3% 1.9% 1.3% 0.9%
–23.0% –25.7% –28.1% –28.7% –29.1%
15.3% 7.9% 3.2% 2.0% 1.5%
–14.7% –22.1% –26.8% –28.1% –28.5%
We use these yield changes to recalculate our measures of household economic condition, dependency on rice and coping capacity. New risk scores and risk profiles are then constructed for the climate change scenarios and compared to baseline risks to determine the proportion of households that are vulnerable to climate change. We define households to be vulnerable if the change in climate increases their risk score. Figure 16.6 shows the percentage of households whose risk scores increase or decrease for each scenario.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 347
Climate Risks and Rice Farming in the Lower Mekong River Basin 347
Figure 16.6 Changes in climate risk scores in response to climate change and extremes In Lao PDR, there are no substantial changes in the proportion of households classified as low, moderate and high risk for any of the climate change scenarios compared to the baseline case. More than 80 per cent of households are still classified as low risk for each scenario. While the majority of households would still be in the low risk category, however, some households would experience an increase in their risk score. Under average climate conditions, 17.0 and 4.4 per cent of households would face increased risk for the 540ppm and 720ppm CO2 scenarios, respectively, and are therefore defined as vulnerable. For the scenarios of extreme climate, more than 50 per cent would experience an increase in risk and are thus vulnerable. In Thailand, there is also no substantial change in the risk groups for the scenarios of changes in average climate, with the moderate risk group still the largest. Because rice yields are projected to increase at the Thai sites for the CCAM-projected changes in average climate, risk scores decrease by 3 to 6 per cent for households in zones 2, 3 and 4 and by 12 to 20 per cent in zone 1. The decrease in climate risk is more pronounced for the 720ppm CO2 case than for the 540ppm case. In the extreme climate scenarios, there are noticeable changes in the moderate and high risk group, with some households moving from the moderate to the high risk group. In zone 3, the number of households classified as high risk increases for the extreme climate scenarios and accounts for more than onethird of households. Approximately 18 to 50 per cent of households have higher risk scores for the climate extreme scenarios compared to the baseline case.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 348
348 Climate Change and Vulnerability
Conclusion Our analysis shows that the relation between the level of development and climate risk is not a simple one and that development level is not a major determinant of risk profiles. Vulnerability to climate impacts is a place-based condition that depends on the socioeconomic, environmental and physical conditions of each location that shape the exposure, sensitivity and coping capacity of households. The profile of climate risk differs from community to community. Households with low climate risk tend to have high coping capacity and low exposure and sensitivity. Comparing the communities in Lao PDR and Thailand, coping capacity emerges as the most important factor contributing to the low risk of the majority of households in the Laotian communities. Households in the Thai communities have lower coping capacity than their counterparts in Lao PDR. They also tend to have higher exposures and sensitivity to climate stresses due to high dependence on rice production for their livelihoods. Consequently, the majority of Thai farm households face moderate to high climate risks. Changes in average climate conditions are found to have relatively small effects on the degree of climate risk faced by rice farmers in the lower Mekong. For the specific climate change scenarios analysed, average yields would increase in Ubon Ratchathani Province in Thailand, resulting in reductions in climate risks. Vulnerability to changes in climate extremes is potentially greater, as suggested by the increases in climate risk scores for our hypothetical scenario of extreme climate. Our study is an attempt to develop a quantitative assessment of vulnerability that captures the influences of local context. However, it should be viewed only as a pilot study on the subject in the Southeast Asia region, and there are many gaps in the approach that need to be improved. First of all, we did not cover other non-climate stresses, particularly changes in socioeconomic conditions, which are impacting and changing farmers’ livelihoods. Future socioeconomic conditions such as the cost of living, market structure and condition, and national and regional development policy could differ greatly from the current situation, especially in the timescales relevant to the study of climate change. These non-climate factors are important drivers that are likely to have a significant influence on the future vulnerability and risk of any social group. Appropriate scenarios of socioeconomic change should therefore be developed and used in future risk analyses. Impact on rice production was used as the single proxy of climate stress in the analysis of risk and vulnerability. While changes in rice production are critically important for farmers in the lower Mekong, climate change will have other impacts that also need to be considered. Our categorization of households into risk groups is based on our own judgements; future research should attempt to establish empirical thresholds of farmers’ tolerance to climate stresses for delineating low, moderate and high risk households. In addition, the cumulative impact on the household of multi-year or consecutive occurrences of extreme climate event should also be taken into consideration.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 349
Climate Risks and Rice Farming in the Lower Mekong River Basin 349
The issue of accumulated risk and vulnerability condition may be a serious one, especially in the case of farmers in Thailand, whose coping capacity is low. Most of the Thai farmers have limited resources to buffer climate impacts on their on-farm output and sustain themselves until the next cropping season. In addition, most of the households have debt, which, in many cases, is higher than their annual income. The impact from multi-year climate stresses, especially consecutive years of extreme climate events, may drive them into a very difficult economic state. Such cumulative effects can drain away a household’s resources for coping and recovery, and surpass thresholds for the sustainability of their livelihood. For example, they may not be able to repay their debts and end up losing their farmland, which is their most important resource, and be forced to change their way of life or social status from that of an independent farmer to that of a hired farm labourer or leave farming permanently to work in another economic sector. Future study might include annual household cash flow analysis over periods of time under different scenarios in order to understand the effects of multi-year climate stresses on household financial conditions. Only two projections of climate change were used in this analysis, both from the same climate model and both representing cooler climates for Southeast Asia than are projected by most other models. Future analyses should examine a broader range of future climate projections. Our assessment focused on impacts in a single year for the average climate projected for the future and for artificially constructed extreme climates. But in order to understand climate change vulnerability and adaptation of farmers in the lower Mekong region, it may be necessary to consider the impacts of climate variability over a number of years. Analyses are needed of potential changes in the frequency and magnitude of extreme climate events and their cumulative impacts over multiple-year time horizons.
References Barker, R. and F. Molle (2004) Evolution of Irrigation in South and Southeast Asia, Comprehensive Assessment Secretariat, Colombo, Sri Lanka Boulidam, S. (2005) ‘Vulnerability and adaptation of rainfed rice farmer to impact of climate variability in Lahakhok, Sebangnuane Tai, Dong Khamphou and Koudhi Villages of Songkhone District, Savannakhet Province, Lao PDR’, Mahidol University, Nakhon Pathom, Thailand Chinvanno, S., S. Souvannalath, B. Lersupavithnapa, V. Kerdsuk and N. Thuan (2008) ‘Strategies for managing climate risks in the Lower Mekong River Basin: A placebased approach’, in N. Leary, J. Adejuwon, V. Barros, I. Burton, J. Kulkarni and R. Lasco (eds), Climate Change and Adaptation, Earthscan, London, UK Committee for Planning and Cooperation (2003) Statistical Yearbook 2002, National Statistical Center, Vientiane, Lao PDR Christensen, J. H., B. Hewitson, A. Busuioc, A. Chen, X. Gao, I. Held, R. Jones, R. Koli, W. Kwon, R. Laprise, V. Rueda, L. Mearns, C. Menendez, J. Raisanen, A. Rinke, A. Sarr and P. Whetton (2007) ‘Regional Climate Projections’, in S. Solomon, D. Qin, M. Manning, Z. Chen, M. C. Marquis, K. Averyt, M. Tignor and H. L. Miller (eds) Climate Change 2007: The Physical Science Basis, Contribution
3243 J&J Vulnerability
1/11/07
10:24 am
Page 350
350 Climate Change and Vulnerability of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US Department of Provincial Administration (undated) website fo Department of Provincial Administration, Ministry of Internal Affairs, Thailand: www.dopa.go.th/xstat/p4834_01.html (in Thai) FAO (2004a) ‘FAO statistical yearbook country profile: Thailand’, available at www.fao.org/countryprofiles FAO (2004b) ‘FAO statistical yearbook country profile: Lao PDR’, available at www.fao.org/countryprofiles Hoogenboom, G., P. W. Wilkens and G. Y. Tsuji (eds) (1999) DSSAT version 3 v4, University of Hawaii, Honolulu, HI IPCC (2001a) ‘Summary for policymakers’, in J. McCarthy, O. Canziani, N. Leary, D. Dokken and K. White (eds) Climate Change 2001: Impacts, Adaptation and Vulnerability, Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US IPCC (2001b) ‘Summary for policymakers’, in J. Houghton, Y. Ding, D. Griggs, M. Noguer, P. van der Linden, X. Dai, K. Maskell and C. Johnson (eds) Climate Change 2001: The Scientific Basis, Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change (IPCC), Cambridge University Press, Cambridge, UK and New York, US Lim, B., E. Spanger-Siegfried, I. Burton, E. Malone and S. Huq (eds) (2005) Adaptation Policy Frameworks for Climate Change: Developing Strategies, Policies and Measures, Cambridge University Press, Cambridge, UK McGregor, J. L. and M. R. Dix (2001) ‘The CSIRO conformal-cubic atmospheric GCM’, in P. F. Hodnett (ed) IUTAM Symposium on Advances in Mathematical Modelling of Atmosphere and Ocean Dynamics, Kluwer, Dordrecht Shukla, P. R., S. K. Sharma, N. H. Ravindranath, A. Garg and S. Battacharya (2003) Climate Change and India: Vulnerability Assessment and Adaptation, University Press, India Snidvongs, A. (2006) Vulnerability to Climate Change Related Water Resource Changes and Extreme Hydrological Events in Southeast Asia, Final Report of AIACC Project no AS07, International START Secretariat, Washington, DC, US, www.aiaccproject.org Ubon Ratchathani Province Administration (undated) Ubon Ratchathani Province Administration website: www.ubonratchathani.go.th/ UNESCAP (undated) ‘Annual core indicators’, United Nations Economic and Social Commission for Asia and the Pacific (UNESCAP), available at http://unescap.org/stat/data/main/datatable.aspx Wang, Y., L. R. Leung, J. L. McGregor, D. Lee, W. Wang, Y. Ding and F. Kimura (2004) ‘Regional climate modeling: Progress, challenges, prospects’, Journal of Meteorological Society of Japan, volume 82, pp1599–1628 World Bank (2007) ‘Country data profiles’, available at www.worldbank.org/
3243 J&J Vulnerability
1/11/07
10:24 am
Page 351
17
Vulnerability of Sri Lankan Tea Plantations to Climate Change Janaka Ratnasiri, Aruliah Anandacoomaraswamy, Madawala Wijeratne, Senaka Basnayake, Asoka Jayakody and Lalith Amarathunga
Introduction Sri Lanka is an island state located to the southeast of the Indian sub-continent. It is the third largest tea producer in the world, after India and China, and ranks first as an exporter. Tea plantations in Sri Lanka currently cover a land area of approximately 210,000ha, which is about 3.1 per cent of the total land area or 12 per cent of the cultivable land of the country, and stretch from the lowlands below 50m in elevation to the highlands above 2000m in elevation. Tea was initially introduced to Sri Lanka in 1867 by British companies who established large plantations by clearing virgin forest in the central hill country. By the turn of the century, more than 120,000ha had been brought under tea cultivation, which doubled to 240,000ha by 1965. In the 1960s, the Government of Sri Lanka nationalized all foreign-owned plantations, and in the early 1990s, the management of plantations was handed back to the private sector. These changes in ownership and management over the years have also significantly impacted the productivity of the industry. Despite these changing circumstances within the industry, export of tea remains one of the key foreign exchange sources of the country. It is the second highest net foreign exchange earning industry on the island, second only to the clothing industry. During 2000–2004, the total foreign exchange earning of the country was in the range 420–584 billion rupees (US$5.25–5.58 billion), of which the tea industry alone contributed 53–75 billion rupees (US$664–716 million) or about 13 per cent (CBSL, 2005). Tea plantations also provide livelihoods for over 1 million people in Sri Lanka. The tea crop is very sensitive to climatic and other environmental factors (Devanathan, 1975; Wijeratne and Fordham, 1996) and has specific require-
3243 J&J Vulnerability
1/11/07
10:24 am
Page 352
352 Climate Change and Vulnerability
ments of rainfall, humidity and soil conditions. Any sustained changes in any of these conditions could adversely affect the yield and thus cause a significant drop in national revenue. This would also affect the livelihoods of the many involved in various aspects of tea growing, processing and other allied operations. Therefore an understanding of the vulnerability of this economically important crop to environmental stressors would serve as an important tool for stakeholders and decision makers in Sri Lanka. One of the biggest concerns for crop production presently is the environmental threat of future climate change whose specific impacts on individual crops are still not exactly clear. Climate change could have important implications for the tea producing sector in Sri Lanka, depending on the nature of the impacts and the sector’s capacity to adapt. An assessment of the vulnerability of tea production in Sri Lanka to the impacts of a changing climate was therefore undertaken in order to generate a knowledge base that can inform current understanding of the strengths and weaknesses of this sector with respect to its ability to cope. A crop simulation model was developed to study the specific impacts of climate change on tea production in the different tea growing regions of Sri Lanka, and a vulnerability matrix was developed for the assessment of regional and national vulnerability. The findings of this study are discussed in the sections that follow.
Current Climate and Future Climate Scenarios for Sri Lanka Current climate Sri Lanka has a hot and humid tropical climate except in the central uplands. Based on data on daily temperature observations made at 19 meteorological stations for the 30-year period 1961–1990, the mean annual temperature in the lowlands is 27.5°C, while that in the highlands is 18.0°C. A significant feature of Sri Lanka’s mean monthly temperature is its small seasonal variation, which rarely exceeds 3°C at any location, while the diurnal temperature range may vary between 5 and 10°C depending on the location and the season (DCS, 2003). According to past studies on long-term temperature variations carried out using linear regression analyses, the mean surface temperature in Sri Lanka increased during the 30-year period 1961–1990 at a rate of 0.016°C per year (Chandrapala, 1996). The annual mean maximum surface temperatures have shown increasing trends for almost all stations examined, with the maximum rate of increase 0.021°C per year at Puttalam, a station midway along the west coast. The night-time annual mean minimum surface temperatures have also shown increasing trends, with a maximum rate of increase reported as 0.02°C per year at Nuwara-Eliya, the highest central hill station. When 30-year (1961–1990) averaged values of temperature were interpolated and mapped into island-wide patterns, incorporating variation in elevation (using a software package, ANUSPLINE, developed by the Australian National University (Hutchinson, 1989)), the baseline annual mean temperature distribution was obtained as shown in Figure 17.1.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 353
Vulnerability of Sri Lankan Tea Plantations to Climate Change 353
Figure 17.1 Baseline temperature distribution
The average annual rainfall in Sri Lanka is 1861mm (1961–1990 data), predominantly from the two monsoonal wind patterns: the southwest monsoon (SWM) period from May to September, experienced mostly in the southwestern part of the country; and the northeast monsoon (NEM) period from December to February, experienced mostly in the northeastern and eastern parts of the country. There is a wide variation in rainfall across the country, with mean annual rainfall received in certain parts of the western slopes of the hill country exceeding 5000mm, while that in the northeastern coastal area is less than 100mm. Based on the rainfall patterns, Sri Lanka has traditionally been divided into 3 climatic zones: the wet zone, covering roughly the southwest quadrant, receives an annual rainfall more than 2500mm, mostly from the SWM; the dry zone, covering the other three quadrants, receives less than 1750mm, mostly from the NEM; and the intermediate zone receives rainfall in the range 1750–2500mm. Figure 17.2 shows the baseline annual rainfall distribution determined for the 30 year period 1961–1990 derived using the ANUSPLINE software. When rainfall variations are analysed, it is found that the average annual rainfall over Sri Lanka for the 30-year period 1961–1990, 1861mm, was about 7 per cent less than that during 1931–1960. Further, the rates of decrease are observed to be higher in the wet regions than in the intermediate regions. The NEM rainfall over Sri Lanka decreased between the periods 1931–1960 and 1961–1990, also with increasing variability, while the SWM rainfall did not show any significant change between the two periods.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 354
354 Climate Change and Vulnerability
Figure 17.2 Baseline rainfall distribution (mean of 1961–1990)
Future climate scenarios Future climate scenarios for Sri Lanka were determined on the basis of regional climate projections developed using several global circulation models (GCMs) (Cubasch et al, 2001) and with reference to the emission scenarios developed by the Intergovernmental Panel on Climate Change (IPCC) in its Special Report on Emission Scenarios (SRES) (Nakicenovic and Swart, 2000). Software developed by the International Global Change Institute (IGCI) at the University of Waikato, New Zealand, was used to downscale the GCM results to locations within the country using an interpolation technique (Warrick et al, 1996). The GCM results considered for our analysis were the outputs of HadCM3 (UK), CSIRO (Australia) and CGCM (Canada) (McAvaney, 2001). The HadCM3 gives the worst scenario for temperature rise, while CGCM gives the worst scenario for rainfall. These GCMs results were found to be closely comparable to the observed baseline data for the 1961–90 period, especially with regard to temperature. The SRES emissions scenarios used were those representing the highest emissions (A1FI), median emissions (A2) and the lowest emissions (B1). For the mid-2100 year, the highest projected mean temperature rise, of about 2.7°C, was obtained with the HadCM3/A1FI scenario, while the lowest,
3243 J&J Vulnerability
1/11/07
10:24 am
Page 355
Vulnerability of Sri Lankan Tea Plantations to Climate Change 355
about 1.0°C, was obtained with the CGCM/B1 scenario. Projections made with all other scenarios show increases of intermediate magnitude, as shown in Table 17.1. Table 17.1 Projected change in mean temperature under different emission scenarios at mid-2100 Range in Temperature rise (°C) at mid-2100 within Sri Lanka Emission Scenario A1FI A2 B1
HadCM3
CSIRO
CGCM
Mean
2.5–3.0 2.1–2.5 1.1–1.4
2.2–2.4 1.9–2.0 1.0–1.1
2.0–2.2 1.7–1.8 0.9–1.0
2.38 2.00 1.08
The range of values indicated for a given scenario is the variation within the country as interpolated by IGCI software. These projections are slightly lower than those in the IPCC Data Distribution Centre maps for the zone covering Sri Lanka (see IPPC-DDC, undated). The future projected temperature distributions applicable under any emission scenario are obtained by adding the corresponding changes to the baseline data. The distribution of the increased temperature corresponding to the HadCM3/A1FI scenario for mid-2100 is shown in Figure 17.3.
Figure 17.3 Increased mean temperature during mid-2100 corresponding to HadCM3/A1FI scenario
3243 J&J Vulnerability
1/11/07
10:24 am
Page 356
356 Climate Change and Vulnerability
Rainfall change projections show both increases and decreases with different GCMs relative to baseline values. Both the HadCM3 and CSIRO models project increasing rainfall while the CGCM projects decreasing rainfall. The HadCM3 projects an increase in the range 0–476mm above the baseline rainfall for the A1FI scenario for June–August 2100. The CGCM projects a decrease in rainfall in the range 190–6mm for the A1FI scenario for the same period. The projected changes in rainfall for the three GCMs and the three emission scenarios are given in Table 17.2. The future rainfall in absolute terms for a given scenario is obtained by adding the corresponding changes to the baseline values. The rainfall scenario projected by the HadCM3/A2 emission scenario for June–August 2100 is shown in Figure 17.4. Table 17.2 Projected change in rainfall under different emission scenarios for June, July and August 2100 Range in Temperature rise (°C) at mid-2100 within Sri Lanka Emission Scenario
HadCM3
CSIRO
CGCM
A1FI A2 B1
0 to 476 0 to 403 0 to 215
2 to 157 2 to 133 1 to 71
–190 to –6 –161 to –5 –86 to 3
Figure 17.4 Increased precipitation during mid-2100 corresponding to HadCM3/A1FI scenario
3243 J&J Vulnerability
1/11/07
10:24 am
Page 357
Vulnerability of Sri Lankan Tea Plantations to Climate Change 357
Climate and Tea Production Sri Lanka is divided into 24 agro-ecological regions (AERs), based on rainfall, annual rainfall distribution, elevation, soils and landforms (Panabokke, 1997), as shown in Table 17.3. The elevation component is, in turn, divided into three ranges – low-country (LC) (<300m), mid-country (MC) (300–900m) and upcountry (UC) (>900m). Of the 24 agro-ecological regions, only 11 are suitable for tea cultivation (Watson, 1986). Their distribution within the country is shown in Figure 17.5. Of the three climate zones in Sri Lanka (the wet zone, dry zone and intermediate zone described in the previous section), the wet zone and a substantial portion of the intermediate zone are suitable for tea cultivation. No tea is grown in the dry zone or low-country intermediate zone. Table 17.3 Agro-ecological regions where tea is grown and their characteristics Elevation Range (amsl) Description
Mean Temp (°C) Wet (1961–1990) Zone
0–300m
Low-country
27.3
300–900m
Mid-country
24.6
>900m
Up-country
18.1
Climate Zone
Wet zone lowcountry (WL) Wet zone mid-country (WM) Wet zone upcountry (WU)
Intermediate Zone No tea grown Intermediate zone midcountry (IM) Intermediate zone upcountry (IU)
Figure 17.5 Agro-ecological regions where tea is cultivated Note: See Table 17.3 for explanation of codes.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 358
358 Climate Change and Vulnerability
The tea industry uses its own classification of elevation ranges, different from those used in the classification of agro-ecological regions. The industry classifications for tea produced are low-grown (<600m), mid-grown (600–1200m) and high-grown (>1200m). Generally, the traditional tea plantations owned by large companies with resident labour produce mid- and high-grown tea. Lowgrown tea is mostly produced by smallholders who have recently taken up tea cultivation. Mid-grown tea is produced by both estates and smallholders. Here we use the term ‘region’ to express division of elevations according to agroecological regions and the term ‘category’ to express division of elevations as per industry classification. Since the early 1990s, the extent of high- and mid-grown tea estates has declined while the extent of low-grown estates has increased. This can be attributed to the government’s decision in 1992 to hand over the management of government-owned plantation estates to the private sector while at the same time increasing financial assistance to smallholders for investing in new plantations. The tea-land survey of 1994 conducted by the Tea Small Holdings Development Authority shows that there were 206,652 smallholdings (below 20ha), covering a total area of 82,918ha. The average size of the smallholdings is thus 0.40ha. A survey carried out by the Census and Statistics Department in 2002 has shown that the area under estate plantations decreased from 168,627ha in 1982 to 118,754ha in 2002 – a decrease of 49,673ha or 29 per cent (DCS, 2004). This decrease is partly due to fragmentation of larger estates into smaller units and partly due to abandonment of unproductive plantations. Concurrently there has also been a significant increase in low-grown tea production since the early 1990s, while the production of the high-grown and mid-grown categories has remained the same.
Annual production and yield In order to evaluate yield on a regional basis and determine the effect of the most influential climatic factors (rainfall and temperature), yield data and meteorological data were collected for selected tea estates. Only estates with time series monthly yield and weather data for a period of more than 20 years and tea plants of an age below 20 years were selected. The average climatic factors and yield data for the selected stations are shown in Table 17.4. For production data,1 the period 1996–2005 was selected; production in each of the three elevation categories is shown in Table 17.5. The mean productions of 161.06kt, 53.78kt and 80.40kt for the low-, mid- and high-grown categories respectively were used as the baseline data for determining future projections under various climate change scenarios.
Effect of temperature on yield The correlation between temperature and productivity in each agro-ecological region was less evident than that between rainfall and productivity. Hence all data were pooled and the relationship between temperature and productivity was established. Dry months were excluded in this analysis in order to elimi-
3243 J&J Vulnerability
1/11/07
10:24 am
Page 359
Vulnerability of Sri Lankan Tea Plantations to Climate Change 359
Table 17.4 Climate and yield data at four sites representing four agro-ecological regions Location
Maximum (°C)
Minimum (°C)
Rainfall (mm)
Yield kg/ha/yr
Ratnapura (WL) Kandy (WM) NuwalaEliya(WU) Badulla (IU)
32.0 29.0 20.5 28.7
22.9 20.2 11.5 18.5
3617 1863 1907 1777
2489 2217 2454 2651
Table 17.5 Annual (baseline) production of made tea during 1996–2005 Annual Production (tonnes) Year 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Mean Percentage
High-grown
Mid-grown
Low-grown
72,447 83,999 77,638 81,471 83,867 83,982 75,342 87,632 82,137 75,525 80,404 27.2
48,211 57,255 52,542 53,660 56,492 56,578 53,943 54,363 54,371 50,391 53,781 18.2
138,312 136,174 150,494 149,060 166,430 166,571 169,334 168,149 184,270 181,826 161,062 54.6
Total 258,970 277,428 280,674 284,191 306,789 307,131 298,619 310,144 320,778 307,742 295,247 100.0
Source: SLTB (2005).
nate the effect of moisture stress on productivity. Analysis of pooled data for all regions shows an increasing trend in productivity with rising temperature up to about 22°C, above which productivity dropped with increasing temperature as shown in Figure 17.6. Similar results have been observed in controlled experiments carried out under laboratory conditions (Wijeratne and Fordhum, 1996). It can therefore be expected that in the low-country, where the ambient temperature is above 22°C a drop in productivity is likely, while in the up-country, where the temperature is below this value, an increase is likely with rising temperatures. The equation below, which explains this relationship, was established taking temperature and yield during wet weather when there is no moisture stress. Thus the rainfall effect was removed to get the temperature response. Y = –508+63.7 T – 1.46T2,
3243 J&J Vulnerability
1/11/07
10:24 am
Page 360
Monthly yield (kg/ha)
360 Climate Change and Vulnerability
Figure 17.6 Variation of monthly yield with ambient temperature with optimum value at 22°C Source: STLB (2005).
where, Y = yield in kg of made tea/ha/month and T = temperature in °C. (‘Made tea’ refers to processed tea available in the market as loose tea; it is also referred to as ‘black tea’. To produce 1 kilogram of made tea, around 4 kilograms of green leaves are generally needed. Statistics usually give made tea values rather than green leaves’.)
Effect of rainfall on yield The rainfall effect was determined by considering only the relatively dry months. The effect of rainfall on the productivity of tea lands was studied by scatter diagrams and regression analyses with a one-month lag period, thus comparing yield with the rainfall of the previous month. The optimum rainfall for each region was first established using curvilinear relationships. The rate of reduction in yield expressed in kg/ha per millimetre loss of rainfall below optimum was next estimated by linear regression analysis. Analysis of productivity with rainfall showed that a 1 millimetre change in rainfall could change productivity by 0.3–0.8kg of made tea per ha during a dry month. The highest positive response to rainfall was seen in the mid-country intermediate zone, where soil fertility and the amount of solar radiation were high compared to other regions. The lowest response, again positive, was seen in the low-country wet zone with less fertile soil. The optimum rainfall and the gradient values of the variation of yield with rainfall for each of five climate zones are given in Table 17.6. The variation observed in one zone – intermediate up-country – is shown in Figure 17.7.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 361
Vulnerability of Sri Lankan Tea Plantations to Climate Change 361
Table 17.6 Effect of rainfall on productivity of tea lands Agro-ecological zone
Productivity change (kg (made tea)/mm rain)
223±38 303±34 417±49 227±10 350±20
0.55±0.07 0.39±0.03 0.36±0.06 0.81±0.11 0.29±0.03
Yield (kg/ha/month)
Up-country wet zone Up-country intermediate zone Mid-country wet zone Mid-country intermediate zone Low-country wet zone
Optimum rainfall (mm)
Figure 17.7 Variation of yield with rainfall up to optimum rainfall for the intermediate up-country region
Effect of drought on yield In 1992 a severe drought was experienced in many parts of the country and data from this year were used to assess the effect of drought on tea production. Data collected from different tea growing regions for this year show that the yield loss due to drought in 1992 was in the range of 14–28 per cent compared to the previous year. The total loss was estimated to be worth about US$70 million. The drought was particularly severe in the intermediate up-country region and relatively mild in the intermediate mid-country region in comparison to other regions (Figure 17.8). Generally, drought effects are more pronounced in the low- and mid-country tea growing regions due to the poor soil conditions resulting from the absence of proper conservation measures and the adoption of ecologically unsound practices in tea lands. The degree of drought impact thus varies between different tea growing regions.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 362
362 Climate Change and Vulnerability
Figure 17.8 Variation of annual yield in different agro-ecological regions in the drought year (1992) and the previous year
Yield response to CO2 elevation The growth rate of mature tea bushes planted in a CO2-enriched enclosure with CO2 concentration maintained at 550±50ppm was compared with that of bushes planted in a control enclosure maintained at ambient CO2 level. This control enclosure had four sides closed and the top open. These trials were conducted at both up-country and low-country locations. It was observed that the tea yield increased with the enhancement of CO2 levels at both elevations, by 33 per cent at the low elevation and 37 per cent at the high elevation. The increase in yield with raised CO2 was attributed to increases in shoot density, growth rate, shoot weight, net photosynthesis rate and enhanced water-use efficiency, factors that are also dependent on the ambient temperature (Wijeratne, 2001).
Climate Change and Tea Production A simple empirical simulation model incorporating the effects of temperature, rainfall, radiation, CO2 and soil organic carbon was developed to quantitatively describe climate impacts on tea yield. The model was based on similar work carried out at the Indian Agricultural Research Institute (Kalra and Aggarwal, 1996). The total biomass production was considered to be a function of leaf area index (LAI) and radiation use efficiency (RUE) of canopy, as well as of radiation level (Monteith, 1977). Solar radiation was determined based on maximum and minimum temperatures using an empirical relationship (Kalra,
3243 J&J Vulnerability
1/11/07
10:24 am
Page 363
Vulnerability of Sri Lankan Tea Plantations to Climate Change 363
personal communication, 2004). Nutrient levels were not considered limiting as the growers practise recommended fertilizer applications which ensure that nutrients will be supplied adequately. Since there is no data on how soil organic carbon or plant population are affected by climate change, a crop model incorporating only rainfall, temperature and CO2 was used to predict tea yield under varying climatic conditions for a given set of soil conditions and cultivars. In developing the model, it was assumed that there are 12,000 tea bushes per hectare and that a mature tea bush at its maximum productivity has about 2kg of above ground biomass. Therefore, the initial biomass (above ground) was taken as 24000kg/ha. The radiation use efficiency value for tea was taken to be 0.3gMJ-1 of total radiation (Carr and Stephens, 1992) and the leaf area index for a healthy tea bush at its maximum productivity was taken to be 5. Of the total carbohydrates produced, it is assumed that 60 per cent is lost through respiration, about 20 per cent is lost through harvesting and only the remaining 20 per cent accumulates as biomass in the bush (Barbora and Barua, 1988; Tanton, 1992; Magambo et al, 1988). The basic flow chart of the model is shown in Figure 17.9. The crop model was linked to the IGCI software, enabling the determination of changes in crop yield for different elevations corresponding to future climate scenarios.
Figure 17.9 Basic flow chart of the crop simulation model
The model runs on monthly data, generating accumulated yield per month (kg/ha/month), which can be summed up to obtain total annual yield in kg/ha/yr. The crop model was validated under varying field conditions in the different agro-ecological regions. It was found that the model can account for 70–96 per cent of yield variations, pointing to the fact that further improve-
3243 J&J Vulnerability
1/11/07
10:24 am
Page 364
364 Climate Change and Vulnerability
Yield (kg/ha)
ments are possible by including other ecological (for example, soil) and plant (variety) factors influencing tea yield. The variation of tea yield under conditions of increased temperature and higher CO2 levels at one location in the wet zone mid-country region, as estimated by the crop model, is shown in Figure 17.10. At this location, a temperature rise beyond the optimum temperature of 22°C was found to reduce the tea yield, while a CO2 rise enhanced it. Under the combined effect yield increased, but by less than that due to the CO2 effect alone due to the negative impact of temperature rise.
Figure 17.10 Variation of tea yield in wet zone mid-country estimated by the tea crop model with change of climatic conditions Note: Run 0: present climate; Run 1: temperature rise by 1°C only; Run 2: 400ppm CO2 only; Run 3: temperature rise by 1°C and 400ppm CO2. 1–12 denote months from January to December.
The projected yield changes for the three GCMs and emission scenarios considered are shown in Table 17.7. According to this analysis, yields are likely to fall in the low-country, where the present ambient temperature is higher than the critical value of 22°C. In the mid- and up-country regions, the yield is also expected to reduce by 2100 since by then the temperature in these regions is also projected to exceed this critical value (Figure 17.11). The reductions expected by 2100 are in the ranges of 19–35 per cent, 9–17 per cent and 0.9–6 per cent, for the low-, mid- and up-country regions respectively, with the high emission scenario HadCM3/A1FI projecting the high end of the ranges and
3243 J&J Vulnerability
1/11/07
10:24 am
Page 365
Vulnerability of Sri Lankan Tea Plantations to Climate Change 365
Table 17.7 Projected yield changes under different climate change scenarios Projected Yield Change (%) Low-Country
Mid-Country
Up-Country
Model/Emission Scenario
2025 2050
2100
2025
HadCM3+A1FI HadCM3+A2 HadCM3+B1 CGCM+A1FI CGCM+A2 CGCM+B1 CSIRO +A1FI CSIRO +A2 CSIRO +B1
–0.32 0.04 –0.24 –2.09 –1.73 –2.05 1.85 2.25 1.97
–34.67 –29.33 –19.37 –31.14 –30.57 –18.80 –33.95 –30.57 –20.41
–0.63 –1.94 –16.78 1.92 3.75 –0.68 –1.44 –15.02 1.69 3.51 –0.77 –1.26 –11.05 1.89 3.25 0.32 0.45 –13.17 0.23 3.02 0.27 0.50 –11.41 –0.07 2.25 0.32 0.50 –8.84 –0.07 1.82 0.90 1.31 –15.47 2.39 4.97 0.81 1.35 –13.04 2.09 4.38 0.90 1.26 –9.70 2.35 3.98
–5.66 –3.66 –2.81 –7.03 –5.14 –4.38 –3.54 –1.49 –0.68
2050
2100
2025
2050
2100 –6.13 –4.18 –2.35 –3.17 –2.29 –0.83 –6.13 –4.18 –0.93
the low emission scenario CGCM/B1 producing the low end. In all three regions, yield changes are observed to be relatively small up to 2050 and could be either positive or negative depending on the time frame and scenario adopted. By 2050, the highest increase of 5 per cent is predicted for the up-country using the CSIRO/A1FI scenario while the highest reduction of 7 per cent is predicted for the low-country under the CGCM/A1FI scenario. The majority of tea plantations appear to be adversely affected by rising temperatures. This is mainly because the present mean temperatures in tea growing regions in the wet zone low-country, wet zone mid-country, intermediate zone mid-country and intermediate zone up-country are above the optimum temperature for high productivity (22°C). Hence beneficial effects of temperature rise can be expected only in the wet-zone up-country region. Given that more than 50 per cent of national production comes from the wet zone low-country and wet zone mid-country regions and this tea also fetches a higher price on the world market, the implications for revenue could be significant.
Impact on national production and revenue The projected yield obtained from the crop model as biomass accumulated per ha per year for each elevation region can be used to determine total production by multiplying this quantity by the corresponding area of plantation for a particular region. However, given that accurate records of productive plantation areas are not available, future projections were instead made using production values. Plantation areas are not expected to change significantly due to a government freeze on the further expansion of tea cultivation, largely because of declining market prices, and also due to a control on encroachment of forested land by smallholder tea cultivators. The production values in the low-, mid- and high-country regions for 2025, 2050 and 2100 were obtained using the baseline production values computed from 1996–2005 data ( Table 17.5) and the percentage regional yield changes corresponding to different
3243 J&J Vulnerability
1/11/07
10:24 am
Page 366
366 Climate Change and Vulnerability
Figure 17.11 Percentage yield change expected with respect to baseline values for 2025, 2050 and 2100 and emission scenarios A1FI, A2 and B1 under the three GCMs in low-, mid- and up-country regions
3243 J&J Vulnerability
1/11/07
10:24 am
Page 367
Vulnerability of Sri Lankan Tea Plantations to Climate Change 367
emission scenarios and climate models (Table 17.7). It is assumed that the yield changes based on elevation regions would also apply to the elevation categories for which production data are available. The new regional production values corresponding to a given time and climate scenario were then summed up to obtain the national production values under similar conditions, with their deviations from the baseline values expressed as percentages (Table 17.8). Table 17.8 National projected production changes under different models and climate change scenarios GCM/Emission Scenario
National production of made tea (kt per annum) 2025
2050
2100
Baseline
295.24
% change
295.24
% change
295.24
% change
HadCM3+A1FI HadCM3+A2 HadCM3+B1 CGCM+A1FI CGCM+A2 CGCM+B1 CSIRO +A1FI CSIRO +A2 CSIRO +B1
295.93 296.30 295.96 292.23 292.54 292.05 300.63 300.98 300.79
0.23 0.36 0.24 –1.02 –0.91 –1.08 1.82 1.94 1.88
288.10 291.39 292.65 286.59 289.04 289.92 294.24 297.09 298.02
–2.42 –1.30 –0.88 –2.93 –2.10 –1.80 –0.34 0.63 0.94
225.45 236.56 256.21 235.45 238.03 259.54 227.31 235.63 256.40
–23.64 –19.87 –13.22 –20.25 –19.38 –12.09 –23.01 –20.19 –13.15
A very significant finding is that for both 2025 and 2050, the expected changes in national production are relatively small for all scenarios and models, at less than ±3 per cent, while the changes expected for 2100 would be in the range –12 per cent to –23 per cent, depending on the scenario and the model. A 3 per cent variation is of the same order of magnitude as that due natural climatic variability, land fragmentation, loss of fertility and various socioeconomic factors. When the expected temperature rise is small and rainfall increases (under HadCM3 and CSIRO models), the overall production change is positive, while with the CGCM model, in which rainfall shows a decrease, there is a negative trend. The positive effects of increasing CO2 concentration and the increasing rainfall more than compensate for the negative effects of increasing temperature. For 2100, however, when the temperature rise exceeds the critical temperature of 22°C for all three regions, the overall effect overrides the positive effects of the CO2 increase and rainfall increase and causes a significant reduction in production under all scenarios and models. Assuming that all other factors remain unchanged, the impacts of climate change on tea production could bring about a reduction in national revenue by approximately 2–4 per cent by 2100. This amount represents the direct loss of revenue from exports, but there could be additional economic losses due to loss of employment for estate workers and the loss of auxiliary industries and services that are associated with various aspects of tea production and processing prior to export.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 368
368 Climate Change and Vulnerability
Vulnerability Assessment Vulnerability in this case is considered to be the difference between the degree of impact and the degree of adaptive capacity; it will thus vary between different elevation regions (where the impacts differ) and for different owners (whose adaptive capacities differ). Based on the results obtained for changes in yield for the different regions using different models and scenarios (Table 17.7), it can be determined that yield reductions of 20–35 per cent, 9–17 per cent and 1–6 per cent are likely for the low-, mid- and up-country regions respectively by 2100, the ranges indicating the variability due to different models and scenarios. The mean impacts would then be –28 per cent, –13 per cent and –3.5 per cent for low-, mid- and up-country respectively, and the relative impacts would thus be 8, 4 and 1. In order to assess the coping capacity of planters, which depends on the ownership of land, information on tea land ownership was extracted from the 2002 survey by the Census and Statistics Department (DCS, 2004). Landholdings of over 20 acres (8 hectares) fall into the category of estates. Most owners of land have holdings of under about 2 acres and cultivate tea in their home gardens, with very little capital and depending mostly on government subsidies. They also usually lack resources to implement adaptation measures in response to adverse climate conditions and hence their adaptive capacity is considered to be zero. The smallholders with over 2 acre landholdings have relatively higher incomes and a better capacity to adapt in comparison to those with less than 2 acres, and hence they are assigned a capacity index of 1. Among the estate holdings, there are wide disparities in size and ownership. There are the relatively recent estate holdings of less than 100 acres, developed using capital raised from banks and depending on government subsidies for their running expenses. These are located mostly in the low-country and are assigned an index of 2. Then there are the estates, several hundred acres in extent, which are the successors to the British-owned companies and have adequate financial resources. These exist in the mid- and up-country and are assigned the highest adaptive capacity index of 3. Using the above information on yield impacts and adaptive capacity indices, a vulnerability matrix was constructed to determine vulnerability of the various classes of landholders. In order to place the impact index in the same range as the adaptive capacity index, the relative impacts derived earlier are normalized to be in the range 0.5–4, with the index 4 assigned to low-country, 2 to mid-country and 0.5 to up-country. The vulnerability index, computed as the difference between the impact index and the adaptive capacity index, shows that the vulnerability of low-country holdings, both smallholdings and estates, is high, with the index lying between –3 and –4. The vulnerability of mid-country smallholdings for both categories (less than and greater than 2 acres), on the other hand, is low, and mid-country estates show no vulnerability at all. In the up-country scenario, the estates and larger smallholdings show a positive response, which indicates a higher than necessary adaptive capacity, while the smallholdings below 2 acres show low vulnerability (Table 17.9).
3243 J&J Vulnerability
1/11/07
10:24 am
Page 369
Vulnerability of Sri Lankan Tea Plantations to Climate Change 369
Table 17.9 Matrix of indices of impact, adaptive capacity and vulnerability vs. ownership Elevation Region Low-country
Mid-country
Up-country
Ownership
Impact Index
Adaptive Capacity Index
EH SH<2A SH>2A EH SH<2A SH>2A EH SH<2A SH>2A
–4 –4 –4 –2 –2 –2 –0.5 –0.5 –0.5
2 0 1 3 0 1 3 0 1
Vulnerability Vulnerability Index Rank –2 –4 –3 1 –2 –1 2.5 –0.5 0.5
Low (L) High (H) High (H) Nil Low (L) Low (L) Positive Low (L) Positive
Note: EH: estate holding; SH: smallholding; 2A: 2 acres
The vulnerability of individual holdings would, in turn, impact the national economy depending on their individual contribution to national production. Since production is generally dependent on the size of the landholding, this was used as the contributory factor. Using the landholding sizes for each of the ownership categories described above (Table 17.9), the size index for each category was obtained by dividing the landholding sizes by 20,000 and rounding off to the first decimal (Table 17.10). The vulnerability of the entire area under each ownership category is then obtained by multiplying the individual vulnerability by the size index and this gives the regional vulnerability. The mean for the three regions is then taken as the national vulnerability (Table 17.10). Table 17.10 Vulnerability of tea plantations at regional/national level Elevation Region Low-country
Mid-country
Up-country
National
Ownership EH SH<2A SH>2A EH SH<2A SH>2A EH SH<2A SH>2A
Size (Acres)
Size Index
Individual Holding Vulnerability
30,147 65,668 40,085 85,905 39,230 23,947 135,191 17,440 10,646 448,259
1.5 3.3 2.0 4.3 2.0 1.2 6.7 0.9 0.5
–2 –4 –3 1 –2 –1 2.5 –0.5 0.5
Regional/ National Vulnerability Vulnerability Rank –3 –13 –6 4 –4 –1 16 –0.5 0.3 –0.8
High Ex-High V-High Positive High Low Positive Low Positive Low
The contribution to the regional vulnerability is extremely high from lowcountry smallholders with landholdings below 2 acres, while that from larger
3243 J&J Vulnerability
1/11/07
10:24 am
Page 370
370 Climate Change and Vulnerability
smallholders (above 2 acres) is very high. The contributions from low-country estates and mid-country smallholders with landholdings of below 2 acres are also both high. The mid-country larger smallholders (above 2 acres) as well as the up-country smallholders with landholdings below 2 acres show low vulnerability. The up-country larger smallholders show a marginal adaptation capacity to counteract the adverse impacts from climate change. Not surprisingly, the mid-country and up-country estates have more than the required capacity to mitigate any adverse impacts. The country’s tea sector as a whole shows low vulnerability to climate change because of the high positive contribution from the large up-country estates.
Discussion and Conclusion In terms of future climate change in Sri Lanka, the projected outputs of three GCMs for three emissions scenarios for the period up to 2100 indicate an increase in temperature with mean increases in the range 1.0–2.4°C for all cases considered. Rainfall, on the other hand, shows different trends under different GCMs. The HadCM3 and CSIRO models project increases in rainfall while the CGCM model projects decreases. The extent of change varies with the emission scenario, with A1FI producing the highest changes and B1 producing the lowest changes. Temperature and precipitation are critical factors in tea production and our vulnerability analysis suggests a varying range of vulnerability depending on the size and location of the tea estate. Tea production in the low-country displays the highest vulnerability and could suffer a decline in productivity of up to 35 per cent by 2100 according to the HadCM3/A1FI emission scenario or somewhat lower declines of up to 19 per cent according to other scenarios. In contrast, the corresponding decreases for up-country are in the range 6–1 per cent. When impact indices for the different regions are compared it can be observed that for shorter time frames, up-country plantations in fact show increases in yields because the ambient temperature there is still less than the critical temperature of 22°C, unlike in the low-country, where it has already been exceeded. Similarly, when adaptive capacity and vulnerability indices are considered, it is the smallholders with landholdings below 2 acres in the low-country that would display the highest vulnerability, in contrast to the estate owners in the up-country, who display greater than the required adaptive capacity to deal with any adverse impacts of climate change. Climate change impacts would thus have important socioeconomic implications, especially in the low-country, by affecting the livelihoods of smallholders, who have few or no capital assets. In turn, the viability of industries and businesses associated with various aspects of tea production and processing could also be impacted. This could also have socioeconomic implications, especially for the low income groups associated directly or indirectly with this sector. In terms of climate change impacts on the national economy, a reduction in foreign exchange revenue by about 2–4 per cent by 2100 is likely due to the direct loss of revenue from tea exports. However, the actual loss of revenue
3243 J&J Vulnerability
1/11/07
10:24 am
Page 371
Vulnerability of Sri Lankan Tea Plantations to Climate Change 371
would probably be higher due to the losses incurred by allied industries and businesses that support the tea production sector. It would therefore be in the national interest for the government to intervene by providing financial assistance to planters with low adaptive capacity in highly vulnerable areas so that they are able to implement the necessary adaptation measures. Several options, such as irrigation, fertilization, application of potassium sulphate and management of shade, have been suggested. The evaluation of potential adaptation options in order to determine the most appropriate strategies for the management of the tea production sector in Sri Lanka under a changed climate is therefore an important subject for future studies.
Note 1
Land production data are more accurate than land extent data. This is because the tea produced across the country is traded through the weekly auctions and statistics of quantities traded are maintained both by elevation category and product type (bulk tea, tea packets, teabags, green tea and instant tea) by the Sri Lanka Tea Board (SLTB, 2005).
References Barbora A. C. and D. N. Barua (1988) ‘Respiratory losses and partitioning of assimilates in tea bushes (Camellia sinensis L.) under plucking’, Two and a Bud, vol 35, pp75–82 Carr, M. K. V. and W. Stephens (1992) ‘Climate, weather and the yield of tea’, in K. C. Wilson and M. N. Clifford (eds) Tea: Cultivation to Consumption, Chapman and Hall, London, pp87–135 CBSL (2005) Annual Report 2004, Central Bank of Sri Lanka, Colombo, Sri Lanka Chandrapala, L. (1996) ‘Long Term Trends of Rainfall and Temperature in Sri Lanka’, in Y. P. Abrol, S. Gadgil and G. B. Pant (eds) Climate Variability and Agriculture, Narosa Publishing House, New Delhi, India Cubasch, U., G. A. Meehl, G. J. Boer, R. J. Stouffer, M. Dix, A. Noda, C. A. Senior, S. Raper and K. S. Yap (2001) ‘Projections of future climate change’, in J. T. Houghton, Y. Ding, D. J. Grggs, M. Noguer, P. J. van der Linden, X. Dai, K. Maskell and C. A. Johnson (eds) Climate Change 2001: The Scientific Basis, Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US DCS (2003) Statistical Abstract: 2003, Department of Census and Statistics, Colombo, Sri Lanka DCS (2004) Sri Lanka: Census of Agriculture – 2002, Department of Census and Statistics, Colombo, Sri Lanka Devanathan, M. A. V. (1975) ‘The quantification of the climatic constraints on plant growth’, Tea Quarterly, vol 45, pp43–72 Hutchinson, M. F. (1989) ‘A new method for gridding elevation and stream line data with automatic removal of pits’, Journal of Hydrology, vol 106, pp211–232 IPCC–DDC (undated) website for the Data Distribution Centre of the Intergovernmental Panel on Climate Change, www.ipcc-data.org/sres/gcm_data.html Kalra, N. and P. K. Aggarwal (1996) ‘Evaluating the growth response for wheat under varying inputs and changing climate options using Wheat Growth Simulator
3243 J&J Vulnerability
1/11/07
10:24 am
Page 372
372 Climate Change and Vulnerability WTGROWS’, in Y. P. Abrol, S. Gadgil and G. B. Pant (eds) Climate Variability and Agriculture, Narosha Publishing House, New Delhi, India, pp320–338 McAvaney, B. J., C. Covey, S. Joussaume, V. Kattsov, A. Kitoh, W. Ogama, A. J. Pitman, A. J. Weaver and R. A. Wood (2001) ‘Model evaluation’, in J. T. Houghton, Y. Ding, D. J. Grggs, M. Noguer, P. J. van der Linden, X. Dai, K. Maskell and C. A. Johnson (eds) Climate Change 2001: The Scientific Basis, Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US Magambo, M. J. S., J. G. Omolo and C. O. Othieno (1988) ‘The effect of plant density on yield, harvest index and dry matter production’, Bulletin of United Planters Association of South India, no 40, pp7–14 Monteith, J. L. (1977) ‘Climate and efficiency of crop production in Britain’, Philosophical Transactions of the Royal Society of London, Series B, vol 281, pp277–294 Nakicenovic, N. and R. Swart (eds) (2000) Special Report on Emissions Scenarios, Special Report of Working Group III of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US Panabokke, C. R. (1997) ‘Agro ecological regions’, in T. Somasekaram, M. P. Perera, M. B. G. de Silva and H. Godellawatta (eds) Arjuna’s Atlas of Sri Lanka, Arjuna Consulting Co. Ltd., Dehiwala, Sri Lanka, pp79–80 STLB (2005) Statistical Bulletin 2005, Sri Lanka Tea Board, Colombo, Sri Lanka Tanton, T. W. (1992) ‘Tea crop physiology’, in K. C. Wilsion and M. N. Clifford (eds) Tea: Cultivation to Consumption, Chapman and Hall, London, pp173–199 Warrick R. A., G. J. Kenny, G. C. Sims, N. J. Ericksen, Q. K. Ahmad and M. Q. Mirza (1996) ‘Integrated model system for national assessment of climate change applications in New Zealand and Bangladesh’, Journal of Water, Air and Soil Pollution, vol 92, no 1–2, pp221–227 Watson, M. (1986) ‘Soil and climatic requirements’, in P. Sivapalan, S. Kulasegaram and A. Kathiravetpillai (eds) Handbook on Tea, Tea Research Institute, Talawakelle, Sri Lanka, pp3–5 Wijeratne, M. A. (2001) Shoot Growth and Harvesting of Tea, The Tea Research Institute of Sri Lanka, Talawakelle, Sri Lanka Wijeratne, M. A. and R. Fordham (1996) ‘Effect of environmental factors on growth and yield of tea (Camellia sinensis (L) Kuntze) in the low-country wet zone of Sri Lanka’, Sri Lanka Journal of Tea Science, vol 64, pp21–34
3243 J&J Vulnerability
1/11/07
10:24 am
Page 373
Part V:
Human Health
3243 J&J Vulnerability
1/11/07
10:24 am
Page 374
3243 J&J Vulnerability
1/11/07
10:24 am
Page 375
18
Vulnerability to Climate-Induced Highland Malaria in East Africa Shem O. Wandiga, Maggie Opondo, Daniel Olago, Andrew Githeko, Faith Githui, Michael Marshall, Tim Downs, Alfred Opere, Pius Z. Yanda, Richard Kangalawe, Robert Kabumbuli, Edward Kirumira, James Kathuri, Eugene Apindi, Lydia Olaka, Laban Ogallo, Paul Mugambi, Rehema Sigalla, Robinah Nanyunja, Timothy Baguma and Pius Achola
Introduction Malaria is a mosquito-borne (Anopheles gambiae species) viral illness, which causes the greatest morbidity and mortality in tropical and subtropical countries, with an especially high prevalence in Africa. Approximately 90 per cent of the 1 million global annual deaths due to malaria occur in Africa and nearly three-quarters of these are children under the age of five (WHO, 1996; McMichael et al, 1996;). It is also associated with several complications such as severe anaemia (especially in children and pregnant women) and cerebral malaria. Low birth weight caused by malaria is responsible for about 6 per cent of the infant mortality. In Kenya, Uganda and Tanzania malaria is endemic in most regions, accounting for a third or more of the outpatient morbidity in the population. In Kenya, malaria accounts for 40,000 infant deaths annually. In Uganda, there were 5.7 million and 7.1 million cases of malaria in 2002 and 2003 respectively, resulting in 6,735 and 8,500 deaths. In Tanzania, malaria causes between 70,000 and 125,000 deaths annually, accounting for 19 per cent of the health expenditure (De Savigny et al, 2004). Thus malaria is also an economic burden, and deprives Africa of US$12 billion every year in lost gross domestic product (GDP) (Greenwood, 2004). Because of climatic and ecological diversity, there is regional variation in the epidemiology of malaria transmission in Africa: from negligible to high risk in high-altitude areas, low but stable transmission along the Indian Ocean, and
3243 J&J Vulnerability
1/11/07
10:24 am
Page 376
376 Climate Change and Vulnerability
intense, high transmission around the Lake Victoria basin. A recently observed phenomenon of an increased frequency of malaria in the highlands of Africa has become a matter of serious concern, given that its prevalence here is typically low. Fifteen per cent of the African population live in the highlands and are therefore at high risk from the impacts of epidemic malaria, particularly in the eastern and southern African regions (Worrall et al, 2004). The concern is relatively small for the lowland areas, where the disease is endemic and the population has developed immunity to it. In East Africa, highland malaria has been recorded since the 1920s and 1950s, when it was first reported (Garnham, 1945; Fontaine et al, 1960; Roberts, 1964; Githeko and Clive, 2005), but the early epidemics were not as severe or as frequent as they have been within the last two decades, with virtually no recorded epidemics between 1960s to the early 1980s. The recent resurgence of malaria epidemics in this region in the last couple of decades has been closely associated with climate variability and change by several scientists (Matola et al, 1987; Lepers et al, 1988; Fowler et al, 1993; Khaemba et al, 1994; Loevinsohn, 1994; Some, 1994; Lindsay and Martens, 1998; Malakooti, et al, 1998; Mouchet et al, 1998; Githeko and Ndegwa 2001; Zhou et al, 2004).1 In fact such zones of unstable malaria, are generally observed to be more sensitive to climate variability and environmental changes (Mouchet et al, 1998). There is concern therefore that future climate change could result in minimum temperature and precipitation thresholds for malaria transmission being surpassed in various parts of the region (Githeko et al, 2000), thus favoring further disease outbreaks. Short-term climate extremes, such as El Niño that similarly affect temperature and precipitation have also been implicated in increased malaria transmission (Kilian et al, 1999; Lindblade et al, 1999). The sensitivity of malaria transmission to climate variability and change is influenced by several factors pertaining to the development and propagation of the vector and the virus, the vector’s preference for human blood feeding, and suitability and availability of disease habitat. In the East African highlands, increases in human population density have led to deforestation and swamp reclamation (Mouchet et al, 1998, Afrane et al, 2005; Minakawa et al, 2005) leading to the creation of puddles and providing ideal breeding sites for mosquitoes. The removal of vegetation, especially Papyrus, in swampy areas, also results in relatively higher temperatures, further aiding disease transmission (Walsh et al, 1993; Minakawa et al, 2005; Munga et al, 2006; Mouchet et al, 1998; Lindblade et al, 2000)2. Other non-climatic factors implicated in disease transmission include environmental and socio-economic change, deterioration of healthcare and food production systems, and the modification of microbial/vector adaptation (Epstein, 1992, 1995; Morse, 1995; McMichael et al, 1996). Given this outlook for a potentially increased incidence of highland malaria in the Lake Victoria region of East Africa, we undertook a study to better examine the relationship between climatic factors and disease outbreaks in three high altitude communities in Kenya, Uganda and Tanzania. Available climate, health and hydrological data from 1960 to 2001 were utilized for this purpose. An integrated vulnerability assessment mechanism for malaria in the
3243 J&J Vulnerability
1/11/07
10:24 am
Page 377
Vulnerability to Climate-Induced Highland Malaria in East Africa 377
affected communities was developed and the manner in which sources of vulnerability are differentiated within the population of this region was analysed. The coping and adaptive capacities of those affected were also identified in this process. The key questions addressed by this research include: 1 2
Which target groups are the most vulnerable, i.e. how are sources of vulnerability differentiated within the population of the Lake Victoria region? What excess risk (the added risk above the normal malaria incidence in a community/household) could be attributable to climate variability?
A more comprehensive description about this study and the important findings that emerged is outlined in the sections that follow.
Study Sites One characteristic of highland malaria epidemics is that affected communities have yet to develop resistance/immunity to the disease, due to the fact that it has not been historically endemic to this area, which was formerly a high-altitude, colder region. As a result we selected communities that were located at altitudes higher than 1100 meters above sea level, where the existence of malaria vectors is limited by the cool temperatures. Households from various elevations above the 1100m level were included because previous studies (for example, Githeko et al, 2006) have shown that prevalence of highland malaria is affected by elevation. Other factors considered include proximity to a hospital and a meteorological station with reliable data. Kabale (Uganda), Kericho (Kenya) and Muleba (Tanzania) were selected as study sites (Figure 18.1) since these sites not only have a recorded history of malaria epidemics in the last two decades but also have been experiencing climate variability and change since the turn of the 20th century. Climate data (temperature and rainfall) were obtained from the nearest meteorological station for Kabale, Kericho and Bukoba (for Muleba, Tanzania) for the period 1961 to 2001 (for Kericho the temperature data were from 1978 to 2001). Streamflow data were obtained from National Water Ministries or Meteorological Agencies for rivers passing or close to the study sites. Only data for the Sondu-Miriu and Yurith rivers from the Kericho site (Kenya), covering the period 1961–1991, could be used because data from the Ugandan site were found to be of poor quality and the Tanzanian site lacked sufficient streamflow data (Table 18.1). An integrated approach using both quantitative and qualitative techniques was employed in assessing the vulnerability and adaptability of highland communities to malaria epidemics. A household survey using 150 semi-structured interviews3 was conducted in each study site at Kabale, Kericho and Muleba to establish the health, demographic and socioeconomic characteristics of the affected communities. The key variables collected include location, sociodemographic data, incomes, household food security, wealth indicators, health issues, knowledge of disease and coping mechanisms.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 378
378 Climate Change and Vulnerability
Figure 18.1 Geographical information system maps of the three study sites in Uganda, Kenya and Tanzania
3243 J&J Vulnerability
1/11/07
10:24 am
Page 379
Vulnerability to Climate-Induced Highland Malaria in East Africa 379
Table 18.1 Geographical positions of streamflow gauging stations, Kericho area ID
Longitude
Latitude
Altitude
Name
IJG01 1JD03
35°00’30”E 35°04’45”E
0°23’35”S 0°28’35”S
1500m >1500m
Sondu Yurith
Hospital records of the number of monthly malaria cases (both in- and outpatients) were collected for Kabale, Kericho and Muleba over a 30-year period (1971–2001), though only 6 years of data (1996–2001) could be analysed due to the poor quality of data from other years. A trial check on some of the earlier data indicated little variability of malaria cases with seasons. An alarming 750 per cent increase in malaria cases in Kericho was observed over a 13-year period from 1986 to 1998 (Shanks et al, 2000). Similarly, in Kabale, malaria cases increased from 17 to 24 cases per 1000 individuals per month during 1992–1996 and 1997–1998 (Kilian et al, 1999; Lindblade et al, 1999), yet this is not reflected in the health data collected, highlighting the poor quality of records, largely resulting from the fact that highland malaria epidemics were not recognized as a national health concern by the East African governments until recently. Health data were collected from Kabale Regional Referral Hospital, Litein Mission Hospital (Kericho) and Rubya District hospital (Muleba). Additionally, qualitative data derived from focus group discussions and participatory stakeholder meetings with community, health and local administrative officials were also used. A total of 12 focus group discussions (4 in each study site) were conducted in communities where households had been previously interviewed; issues discussed include indicators of wealth, knowledge of disease, attitude towards the disease, disease prevention and management practices and impact of disease, coping mechanisms, and interventions. Two participatory stakeholder meetings were conducted in Kabale, Kericho and Muleba, where stakeholders were invited to actively articulate their knowledge, values and preferences regarding vulnerability and adaptation to malaria epidemics.
Climate and Hydrology of the Study Sites Temperature Analysis of the temperature data sets indicates increases in maximum (Tmax) and minimum (Tmin) temperatures at both lowland and highland sites during various periods (Table 18.2). The temperature change (increases in both Tmax and Tmin) has generally been greater in the highlands than in the lowlands and of note is the marked increase in Tmax (3.6ºC) in Kericho (a highland site). An increasing trend in Tmax and Tmin was also noted for the lowland sites of Kisumu and Kampala, but there was a declining trend in Mwanza. The trend of temperatures increasing from the lowlands to the highlands is probably a factor that has enabled the malaria transmitting mosquitoes to find new habitats in the highlands; hence the creeping altitudinal ascent of unstable malaria epidemics.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 380
380 Climate Change and Vulnerability
Table 18.2 The long-term context of temperature changes in the Lake Victoria basin, showing results for highland sites based on linear regression Station
Period of analysis
Temperature change (ºC)
Kericho
1978–2001
Kabale
1960–2003
Bukoba
1960–2002
Max Min Max Min Max Min
3.6 0.5 1.1 1.6 0.7 1.1
The ranked Tmax and Tmin values in Table 18.3 indicate that the high Tmax years within the Lake Victoria region as a whole are associated with El Niño occurrences and concomitant high streamflow and flooding in the lake basin area. Table 18.3 Ranked Tmax and Tmin with high Tmax and low Tmin for the period 1978 to 1999, compared with occurrence of El Niño and La Niña years Site
High Tmax years
Kericho
1981, 1978, 1991, 1985 1994–1995, 1997, 1999
1987, 1989 1981, 1991
1982–1983, 1995, 1997
1983, 1997
Kabale
Bukoba
1983, 1987, 1997, 1999
Low Tmax years
1985
High Tmin years
Low Tmin years
El Niño years
La Niña years
1977–1978, 1982–1983, 1986–1987
1988–1989. 1991–1992, 1978, 1985, 1993
1996, 1987, 1997–1998 1993
1992–1993, 1994–1995,
1995–1996. 1997-1998
1998–1999. 1999–2000
Note: Bold – El Niño years; bold/italic – La Niña years; normal font – non-El Niño/La Niña years.
Precipitation Rainfall analyses for the period 1961–2002 show that Kericho (annual rainfall range from 897 to 2420mm) and Bukoba (884–2736mm) received relatively high amounts of rainfall with relatively high coefficients of variation, while Kabale (755–1282mm) received the least. Generally, annual time series analysis for the period 1961–2001/2 shows a decreasing trend in rainfall at all the stations except Kabale. In all the stations, March to May (MAM) received more rainfall than September to December (SOND). On a seasonal basis only Bukoba showed a statistically significant downward trend for the JF, MAM and JJA seasons.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 381
Vulnerability to Climate-Induced Highland Malaria in East Africa 381
Table 18.4 Ranked mean monthly cumulative precipitation with wet years (≥1 standard deviation from long-term mean) and dry years (≤1 standard deviation from long-term mean) for the period 1978 to 1999, compared with occurrence of El Niño and La Niña years Site
Wet years
Dry years
El Niño years
Kericho
1982, 1988–1989, 1992, 1994, 1996,
1978, 1980, 1984, 1986, 1993,
1977–1978, 1982–1983, 1986–1987,
1991–1992,
1999,
1992–1993,
1978, 1998, 1996, 1998,
1979, 1982, 1993, 1999,
1994–1995,
1985, 1986, 1994
1980, 1981, 1982–1983
Kabale
Bukoba
La Niña years
1988–1996,
1995–1996, 1997–1998
1998–1999, 1999–2000
Note: Bold – El Niño years; bold/italic – La Niña years; normal font – non-El Niño/La Niña years.
The ranked mean monthly cumulative precipitation data (1978–1999) show that in Kericho, wet years occur either during El Niño or La Niña years (Table 18.4), with the exception being the El Niño period 1997–98, which was not significantly wetter. In Kabale, wet years also appear to be associated with La Niña and El Niño, more consistently with La Niña. In Bukoba, wet years are associated with El Niño, though not as strongly, with one high rainfall episode during a non-El Niño/La Niña year (1985). Dry years in Kericho occur during both El Niño and non-El Niño/La Niña years. In Kabale, dry years occur during El Niño, with single occurrences of dry years during a La Niña and non-El Niño/La Niña year. In Bukoba, dry years are associated with non-El Niño/La Niña years, but it is of significance that during the strong El Niño of 1982–83, Bukoba was dry but experienced a ‘normal’ rainfall season in SOND. The observed heterogeneity in the rainfall patterns around Lake Victoria may partly be accounted for, to varying degrees, by a combination of factors such as differences in topography and aspect, changes in land use, the influence of Lake Victoria and land–ocean interaction (see Ogallo et al, 1989; Ropelewski and Halpert, 1987).
Hydrology No significant trends in the annual flows for the Sondu-Miriu and Yurith rivers were noted during the period of analysis (1961–1990), although moving averages showed some fluctuations in river flows. An association of high flows with El Niño years (1968, 1970, 1978–1979, 1982, 1988 and 1990) was observed.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 382
382 Climate Change and Vulnerability
The mean and median flows for the Sondu-Miriu and Yurith rivers show that highest flow occurs in the MAM ‘long rains’ season, with a subdued peak in August and a high flow in the SOND ‘short rains” season with a peak in November. The peak river flow lags behind two of the three observed rainfall peaks (April and August) by one month, but is coincident with the rainfall peak in November. Flood frequency analysis on the Sondu-Miriu river indicates that the return period for maximum flow is between two and five years. This suggests that besides the influence of El Niño (occurring approximately every two to seven years) in generating abnormally high flows in either or both MAM and SOND, other simultaneous mesoscale climate and weather systems can also account for the flooding events. The results of streamflow analyses demonstrate that precipitation and resulting flow in rivers within the area are a tightly coupled system, which means that the streamflow is directly dependent on the precipitation.
Possible links between climate, hydrology and malaria outbreaks The climate and hydrological data provide some important insights into the links between vulnerability to malaria outbreaks and climate variability and change. Positive trends in maximum temperature are significantly linked to the El Niño Southern Oscillation (ENSO) (Table 18.3), which, in turn, has been associated with serious malaria epidemics in the lake basin. These results indicate that the malaria exposure risk of the highland lake communities is dramatically increased during ENSO periods when anomalously high temperatures and widespread flooding favour the proliferation of the mosquito vector. Secondly, anomalously wet years are not always necessarily accompanied by anomalously high temperatures (see Tables 18.3 and 18.4), which indicates that other mesoscale climate or weather patterns, such as the Indian Ocean dipole reversal (Conway, 2002), that can generate heavy precipitation events equalling or even exceeding ENSO effects do not necessarily increase the risk of epidemic malaria in the highlands. This notwithstanding, the increasing trend in mean temperatures across the study sites in the lake basin region over the past three decades or so suggests that perhaps a critical threshold in this relationship could be breached in the near future if the warming trends continue, and that this could potentially lead to increased non-ENSO-related malaria epidemics in the highlands. From the flood frequency analysis, it can be deduced that the frequency of occurrence of conditions conducive to highland malaria epidemics could double in the future, assuming that the global warming trends do not significantly disrupt the current prevailing weather patterns in the region.
Climate-Related Trends in Malaria Outbreaks In East Africa, malaria causes 30–40 per cent of all hospital admissions. Available hospital-based morbidity records were reanalysed to indicate per-
3243 J&J Vulnerability
1/11/07
10:24 am
Page 383
Vulnerability to Climate-Induced Highland Malaria in East Africa 383
centage departure of mean monthly in-patient admissions from long-term means (six years) obtained from in-patient records from 1996 to 2001 (Figure 18.2). The data were also assessed for seasonal departures from the long-term mean and for long-term trends from 1996 to 2001. A monthly increase of 50 per cent in malaria admissions above the long-term mean from the respective study sites was taken as a threshold for malaria epidemic outbreaks. The first upsurge in malaria cases in Muleba was observed in May to July 1997, and that in Litein from June to July 1997. In Kabale, the number of cases during this period remained below normal (Figure 18.2).
Figure 18.2 Trends in malaria hospital admissions in Kenya, Tanzania and Uganda
Comparing trends in malaria cases for children under five and older individuals indicated that children under five were much more susceptible to malaria attacks (Figure 18.3). Children under 5 were 1.5 times more likely to be admitted than older individuals, which points to the fact that younger children have lower immunity. The most significant change in seasonal outbreaks was observed from January to March 1998 in Tanzania and Kenya, but the trends extended to May of the same year in Kabale, Uganda, which implies that the epidemic lasted for six months. In Tanzania, the epidemic caused a peak increase in cases of 146
3243 J&J Vulnerability
1/11/07
10:24 am
Page 384
384 Climate Change and Vulnerability
Figure 18.3 Trends in malaria in children <5 years in Muleba, Tanzania
per cent, while in Kenya and Uganda the increases were 630 per cent and 256 per cent respectively. The peak month for admissions in all countries was March. It should be noted that in Kenya, government hospital workers were on strike during this period, so most of the cases were treated in Mission hospitals, which probably accounts for the large increase, since the Kenyan hospital used in this study is a Mission hospital. It is more likely that the increases in malaria cases in Tanzania and Uganda give the true picture. Uganda had further malaria outbreaks in November–December 1999 (with epidemic increase of 63 per cent) and again in December 2000–February 2001, when the outbreak peaked at 312 per cent in January. A small outbreak was observed in Kenya (with an increase of 78 per cent) in February 2001. Data from Uganda indicate that outbreaks are more common after the short rainy season (September–November). The most significant anomalies in temperature and rainfall were observed during the El Niño period of 1997–1998, after which there were severe malaria outbreaks. In all cases, seasonal malaria outbreaks were associated with anomalies in temperature. This observation is consistent with the well-established biology of malaria transmission. For example, anomalies were observed at Kericho in the mean monthly maximum of 2.2–4.5°C in January–March 1997 and of 1.8–3.0°C in February–April 1998. The overall data did not show significant annual trends in malaria cases over the period 1996–2001. The data from Litein, Kenya, and Muleba, Tanzania, showed a declining trend in malaria cases, while data from Kabale
3243 J&J Vulnerability
1/11/07
10:24 am
Page 385
Vulnerability to Climate-Induced Highland Malaria in East Africa 385
showed a slightly increasing trend. Data from the three study sites were compared by regression analysis to determine the degree of association among sites. Data from Tanzania and Kenya had the best association of (R2 = 0.59), while R2 for Kenya and Uganda was 0.3 and R2 for Uganda and Tanzania was 0.29. These results indicate that there is likely a common effect modulating the outbreaks and this is most likely a climate phenomenon.
Non-Climate Factors that Influence Vulnerability to Malaria Outbreaks Self-medication Chloroquine was initially the drug of choice for the treatment of malaria in East Africa, but this was later replaced by sulfadoxine and pyremithamine (SP) combinations due to the development of drug resistance by about 70 per cent of the parasites by 1990. SP was in turn replaced by the more expensive artemisinin-based drugs, once again due to resistance. A fourth drug, quinine, which is still very effective, is only used in cases of hospitalization. Interviews with residents in Kericho, Kenya, revealed that many people (49 per cent) still relied on chloroquine for home treatment, despite 85 per cent of the malaria parasites being resistant to the drug. The next most popular drug, used by 39 per cent of the people, was Fansidar (an SP), to which about 50 per cent of the parasites are resistant. Other drugs used were quinine and antibiotics, which are prescription drugs. Thus people treating themselves in Kericho are at a high risk of developing severe and complicated malaria due to drug failure or under-dosage. This can contribute to high morbidity and mortality, particularly in populations with low immunity.
Knowledge of disease Knowledge of malaria among the communities and local health officials was found to be couched in myths. For example, the Public Health Act still requires clearing of bushes around houses to prevent yellow fever even though recent studies have demonstrated that such bush clearing instead creates a favourable microclimate for the mosquitoes that spread malaria (Walsh et al, 1993). A second misconception is regarding the role of climate in triggering the outbreak of malaria epidemics. For example, there was a malaria epidemic episode in 2002, which was a relatively hot year, followed by the absence of any outbreaks in 2003, when the temperature was 2ºC cooler. However, the Clinical Officer at Litein in Kericho attributed the low incidences of malaria in 2003 to the effectiveness of a government introduced campaign promoting the use of insecticide treated nets to protect against mosquito bites. In reality, this campaign has not been as successful as claimed due to a lack of supply of insecticide treated nets and also their unaffordability. Although the use of bed nets (treated or not) may have contributed to preventing malaria to some extent in 2003, the lack of association of the low disease incidence with a lower temper-
3243 J&J Vulnerability
1/11/07
10:24 am
Page 386
386 Climate Change and Vulnerability
ature points to vulnerability arising from inadequate knowledge about the disease. Generally, a significant proportion of the respondents (83.2, 94.5, and 52.7 per cent in Kabale, Kericho and Muleba respectively) could establish a link between the health of household members and weather conditions. The awareness of symptoms of malaria is also high. However, knowledge regarding the causes and prevention of malaria is once again largely based on myths. One such myth from Kenya associates consuming food cooked with an edible oil called Chipsy with immediate malaria outbreaks. This brand of edible oil was introduced in 1990, a year that coincided with the El Niño rains and malaria epidemics. In Muleba, Tanzania, people believe that eating maize meal instead of bananas causes malaria. But maize meal is generally consumed only during periods of food shortages that usually result from above and/or below average rains (for example, El Niño rains and/or La Niña droughts), periods when malaria is also more rampant. Similarly, in Uganda, malarial complications such as convulsions (neuropsychiatric events) are attributed to supernatural forces, and hence considered to be best treated with traditional medicine (Nuwaha, 2002). This often leads to delays in medical care, thereby increasing malaria morbidity, severity and mortality. Monthly household income, gender and levels of education also had a significant correlation with the level of awareness on prevention of malaria.
Socioeconomic characteristics The surveys do reveal that the interplay of poverty and other variables intensifies the vulnerability of a population to malaria. This is because of a lack of economic resources for investing in healthcare that can reduce susceptibility to the disease and help offset the costs of adaptation. Socioeconomic characteristics of the population suggest certain poverty indices that reflect the vulnerability of these communities to malaria epidemics. Most of the households in the survey area live below the poverty line (less than one dollar a day), relying predominantly on either farming or self-employment (Table 18.5). Only a privileged few are formally employed with a source of steady income, accounting for 19, 15 and 2 per cent of the populations of Kabale, Kericho and Muleba respectively. Indeed, when disaggregated by income group, formal employment is the most common source of income for the households found in the higher income brackets (US$91–100 and US$101+). In addition to poor incomes, these communities also experience household food security issues, with a significant proportion of the households in the study areas indicating days of household food shortages and the poorer households more likely to experience food shortages. For example, 50 per cent of households with a monthly income of less than US$30 experience days of food shortage (Table 18.5). Kabale and Muleba have a significantly higher proportion (54.5 and 46.7 per cent) of households experiencing food shortages than Kericho (20.5 per cent).
3243 J&J Vulnerability
1/11/07
10:24 am
Page 387
Vulnerability to Climate-Induced Highland Malaria in East Africa 387
Table 18.5 Selected indicators of vulnerability to malaria epidemics Monthly household income (US$)a
Proportion Predominant Average of source of household households income size
Days of food shortage
Households without bed-nets (%)
Household malaria mortality (1998–2002) (%)
Most common mode of transport (%)
(%) 8.0
50
76.2
30.0
7.4
47.2
69.0
16.7
Bicycle (46.7) Bicycle (32.4)
5.7
40
20.0
5.9
7.6
38.9
13.0
0
6.0
35
10.0
0
6.4
34.4
10.0
0
5.3
23.5
3.0
0
7.6
14.3
1.0
1
Bicycle (80)
7.0
0
1.0
1
Motor vehicle (56.3)
(%) HIGH ≤30
47.8
31–40
12.2
41–50
7.1
VULNERABILITY
51–60
6.4
61–70
7.1
71–80
2.4
81–90
3.2
91–100
2.0
LOW 101+b
Total
11.8
Farming (54.5%) Farming (73.5%) Selfemployment (50%) Farming (94%) Farming (60%) Farming (57.1%) Selfemployment (66.7%) Formal employment (60%) Formal employment (54.5%)
Bicycle (70.0) Bicycle (37.5) Bicycle (52.9) Bicycle (57.1) Bicycle (100)
100.00 (n = 450)
Notes: aThe average monthly income is US$50.2; the most common income (mode) is US$25.6. bThe highest monthly income in this class is US$580.3.
Lack of adequate healthcare systems coupled with persistent poverty greatly compromise the capacity of individuals and communities to cope with the consequences of malaria epidemics. Most households surveyed in Kericho indicated that they relied on local dispensaries rather than the provincial or district hospitals that are better equipped and have professional staff and inpatient facilities (Table 18.6). Only Kabale reported more use of the district hospital (59.2 per cent) and private clinics (28.7 per cent). Reliance on local dispensaries and private clinics for treatment often results in misdiagnosis due to lack of qualified staff or self-medication by the respondents. Quick treatment was one of the main reasons for the preference for private clinics over public health facilities, which are often overcrowded with long patient queues and have generally unfriendly staff. Accessibility to health facilities is also an issue since the predominant mode of transport is the bicycle for all income groups apart from the highest income group (Table 18.5), primarily due to the high cost of motorized transport. This is also reflected in the low frequency of
3243 J&J Vulnerability
1/11/07
10:24 am
Page 388
388 Climate Change and Vulnerability
visits to health facilities (Table 18.7). Consequently, the malaria mortality rate appears to be more prevalent among the low-income households: 30 per cent of households living below the poverty line had lost a household member in the past five years (Table 18.5). Poverty indices indicate that vulnerability increases as monthly household incomes decrease and days of food shortage and household mortality increase, while the proportion of bed-nets in use decreases (Table 18.5). Table 18.6 Type of health facility visited Health facility Provincial hospital District hospital Health centre Local dispensary Mobile dispensary Herbalist Private hospital Private clinic Total
Kericho
Kabale
Muleba
1.0% 5.5% 91.5%
0.6% 2.5% 59.2% 20.3%
1.0% 1.0% 100%
6.4% 11.0% 100%
0.7% 11.7% 15.3% 50.3% 5.3% 10.0% 6.7% 100%
Table 18.7 Visits to hospitals in the last three months by household members No of Visits 0 1 2 3 4 5 6 9 10+ Total
Kericho
Kabale
Muleba
44.4% 24.5% 15.9% 9.9% 2.0% 1.3% 1.3% 0.7%
31.4% 37.1% 21.4% 7.5% 1.3% 0.6%
28% 41.3% 16% 9.3% 2% 2% 0.7%
100%
100%
0.7% 100%
Existing coping and adaptation mechanisms The use of insecticide treated nets (ITNs) is one of the preventive measures advocated by the Malaria Global Control Strategy, as well as the national malaria control programmes in East Africa.4 However, the survey revealed that the use of ITNs is not very widespread, particularly among the households with monthly incomes of less than US$30 and US$31–40, of which 76.2 per cent and 69 per cent respectively do not have bed nets (Table 18.5). In Kabale, the proportion of household members sleeping under a bed net was observed
3243 J&J Vulnerability
1/11/07
10:24 am
Page 389
Vulnerability to Climate-Induced Highland Malaria in East Africa 389
to increase with increase in average income. The weak economic base of these highland communities, who rely largely on farming and self-employment for income generation, is a critical issue that leaves them vulnerable to external shocks and to seasonal and climatic variability and change.5 The World Health Organization’s (WHO, 2002) ‘Roll Back Malaria’ programme has been adopted by most countries in Africa. The Kenyan, Ugandan and Tanzanian governments actively promote this programme, whose objectives are to increase the use of ITNs, early diagnosis and treatment of malaria, and the use of effective antimalarial drugs. The programme has attracted several local and international civil society organizations, such as Population Services International (PSI), which is supported by the British and US governments and aims to increase the use, ownership, availability and accessibility of ITNs in the malaria-endemic areas of Kenya, Uganda and Tanzania. However, the cost of the subsidized ITNs (about US$1.50) is still beyond the reach of households living below the poverty line (Table 18.5). Even households that do own bed-nets are unable to provide them for all members. A household may have as many as 16 members with an average size of 3.7 persons. The number of bed-nets, on the other hand, may range from 1 to 6. Our analysis showed that for those who own bed-nets, only 37 per cent could afford to have more than three quarters of household members sleeping under them. Furthermore, a majority of those using bed-nets tend not to treat them with insecticides (about 75 percent of bed net users). Among those who do treat the nets, many do so only once or twice a year, which is considered insufficient for them to remain effective. These factors together tend to reduce the overall effectiveness of the ‘Roll Back Malaria’ programme in combating malaria. From the survey results we found that there are very few coping mechanisms available for the households. In the likely event of a malaria epidemic, the majority (75.5 per cent) sell their food crops to cover the cost of treatment. Other ways of coping include borrowing or relying on remittances from relatives. In Kabale, focus group discussions revealed that a number of people have resorted to selling land in order to cope with malaria. Out of the 30 participants in the group, 13 reported having sold land at some stage in the last 8 years in order to cope with malaria in the family. Such coping mechanisms deplete the resources of those affected and may lead to increased food shortage, debt and poverty. The common adaptation to highland malaria in Muleba, Kericho and Kabale is largely the use of traditional curative measures that rely on local herbs as insect repellents or antimalarial treatments, primarily because of their affordability to the large populations of poor people in these areas. Surveys carried out by the National Institute for Medical Research (NIMR) in Tanzania noted that traditional healers that give out these treatments do possess knowledge and skills useful for malaria disease management. Further, NIMR laboratory analyses of traditional herbs have also established their safety and efficacy, although this often varied depending on the herb (Mwisongo and Borg, 2002).
3243 J&J Vulnerability
1/11/07
10:24 am
Page 390
390 Climate Change and Vulnerability
Predicting Malaria Epidemics Using Climate Data The observed increase in maximum or both minimum and maximum temperatures and the relatively higher temperature variability in the highlands, as compared to the lowlands, have possibly increased mosquito productivity in vector habitats, thus leading to increased malaria transmission rates. This would partly explain why episodic malaria outbreaks have been increasing at higher altitudes in recent years. Githeko and Ndegwa (2001) have reported evidence to support this view, including improved larval survival at warmer temperatures and increased mosquito biting frequency in deforested areas of the western Kenyan highlands (in comparison to forested areas) due to a rise in temperature of 2°C (Githeko et al, 2006). The period March–May (MAM) receives more rainfall than the September–November (SON) or September–December (SOND) season throughout the three study sites. During MAM, when the highest rainfall occurs, rivers overflow their banks and flood their basins, while in much wider areas the soils get saturated, encouraging retention of standing water. There is a one-month lag between the peak rainfall and the peak river flow, as the rivers are largely recharged by land runoff and groundwater flow from their drainage basins. The malaria epidemics tend to occur from July to September, with a minimum lag of two-months after the peak rainfall in April; this is related to the one-month lag between peak rainfall and peak streamflow, and a further one-month lag necessary for the development of the malaria vector. During El Niño years, when the short rains (SOND) are anomalously heavy and temperature is high, a similar potential for malaria exists in January–February (JF) as the characteristic conditions of the MAM season are replicated. Figure 18.4 indicates that malaria outbreaks are sensitive to maximum temperature, with a lag of one to four months after the maximum peak to the onset of the malaria episode (Githeko and Ndegwa, 2001), which agrees with the hydrological data. Other factors, such as the shape of the valleys, which determine vector habitat availability and stability, can also account for large variations in transmission intensities and malaria prevalence at the same altitude (Githeko and Ndegwa, 2001). Thus, although the general principle of the model developed by Githeko and Ndegwa (2001) is valid, there is a need to fine-tune it to specific ecological zones. Furthermore, changes in malaria treatment policies can affect the outcome of the model, as the use of effective antimalarial drugs in primary healthcare can dramatically reduce the number of cases in hospitals. Analysis of trends in temperature data indicated that in Kabale, Uganda, there was an increase of 1.17°C in mean annual minimum temperature between 1960 and 2001. In Kericho, Kenya, the mean annual maximum temperature increased by 3.5°C. In Bukoba, Tanzania, the mean annual maximum and minimum temperatures were found to have increased by 0.21 and 0.49°C respectively between 1960 and 2001. While climate data are collected at regular intervals, the same is not true for malaria data, which is of relatively poor quality. The long-term climate data demonstrated a trend towards warming in the highlands, which suggests
3243 J&J Vulnerability
1/11/07
10:24 am
Page 391
Vulnerability to Climate-Induced Highland Malaria in East Africa 391
improved transmission conditions. Although the five-year malaria data available (1996–2001) cannot be used to detect long-term trends in transmission, it nevertheless demonstrated an association between malaria cases and climate variability. Additionally, every 1°C increase in temperature is equivalent to a reduction of 154 metres in altitude. Therefore transmission conditions at 1500 metres above sea level would be equivalent to conditions at approximately 1200 meters above sea level for a temperature increase of 2°C, thus making malaria at 1200 metres above sea level stable and hyperendemic due to the favourable temperature conditions. Such a situation could be reality for Bukoba located at 1100m above sea-level, even below the 1200m level. The risk of malaria epidemics is associated with positively anomalous temperatures in the months preceding and during the rainy season. Temperature controls the rate of larval and parasite development, with higher temperatures shortening the development time of the larvae as well as parasites in the mosquitoes. The logistic model for the effects of temperature and rainfall developed by Githeko and Ndegwa (2001) indicates that the rate of growth of a mosquito population is dependent on the initial population size before the rainy season. Rainfall increases the availability of mosquito breeding habitats, thus contributing to the size of the mosquito population, which directly influences the intensity of malaria transmission. Recent studies (Zhou et al, 2004) indicate that the availability and stability of mosquito breeding habitats and the initial vector population size before the rainy season are also a function of drainage efficiency and epidemic propagation and intensity. Githeko and Ndegwa (2001) also showed that malaria epidemics in Kakamega district in western Kenya could be predicted using simple temperature and rainfall data. The model was able to identify climatic conditions that enabled a rapid growth of mosquito populations leading to epidemics one or two months later. One of the problems with the model is its inability to account for the incidence of temperature and precipitation anomalies around January and February, which are associated with the Indian Ocean dipole reversal episodes that cause non-El Niño rains in East Africa (Nicholson, 1996; Conway, 2002). In the case of the 1997–1998 El Niño period, rainfall continued from November 1997 into January and February 1998 creating perfect breeding habitats for malaria vectors. We used the data from the three sites to determine whether the model of Githeko and Ndegwa was applicable to other sites in East Africa. Our preliminary results show that, with slight modification, the model was able to identify major epidemics in Kericho, Kabale and Muleba (Figure 18.4). In all cases the epidemics were associated with anomalies in the mean monthly maximum temperature one or two months before the epidemic. The other necessary condition was a significant increase in rainfall one month before the peak of the epidemic. The model is currently being further refined to take into consideration the drainage characteristics of individual sites as this affects the rainfall thresholds used in the model. The ability to forecast an epidemic about two months beforehand is critical for decision making and the logistics of putting preventative and curative measures in place in a timely fashion.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 392
392 Climate Change and Vulnerability
Figure 18.4 Modelled climate and malaria data for Litein, Kenya
Conclusion Changes in climate have been noted at the three study sites in Kenya, Tanzania and Uganda, and these changes are consistent with what has been observed and documented by previous research in other parts of the highlands of East Africa. The maximum and minimum temperatures have changed, with significant increases generally recorded at all sites. The temperature change has been more pronounced at the higher altitudes than in the lowlands. The observed temperature increase has enabled malaria vector mosquitoes to find new habitats in the highlands. Similarly, the rainfall pattern has changed. Generally, time series analyses for the 1961–2001 period show decreasing trends in rainfall for all of the stations except Kabale. Hydrological data show that for the Kericho site, the peak riverflow lags behind two of the three observed rainfall peaks (April and August) by one month, but is coincidental with the rainfall peak in November. Malaria epidemics often occur in the months of July to September; since peak rainfall occurs in April and there is a minimum two-month lag between the peak rainfall and the epidemics. If, for a given year, the maximum and minimum temperatures are consistently conducive for development and growth of the malaria vector, then the two-month lag between peak rainfall and the onset of the epidemics can largely be accounted for by the one-month lag in peak streamflow. During the 1997–1998 El Niño episode, malaria admission data from the
3243 J&J Vulnerability
1/11/07
10:24 am
Page 393
Vulnerability to Climate-Induced Highland Malaria in East Africa 393
study sites indicated that the epidemic months corresponded with the onset of abnormally high rainfall during the short rains season and abnormally high maximum temperatures during the months preceding the rainy season. This was confirmed with the observation of anomalies in the mean monthly maximum of 2.2–4.5°C between January and March 1997 and 1.8–3.0°C between February and April 1998. Other cases of malaria epidemics follow the trends described above, with the highest incidents in March–May and July– September associated with the long and short rainy seasons respectively. Poverty seems to play a very important role in determining the vulnerability of the communities to climate change and variations in the social system. Poor communities lack effective strategies for coping with climate-induced shocks such as disease and weather extremes. Household incomes in our study sites were generally very low and derived largely from insecure and uncertain sources, which exposes them to external shocks. The impacts of climate variability and change tend also to affect crop production and therefore the socioeconomic stability of the region. Shortages of food can lead to malnutrition, especially in poor households, and further increase vulnerability to diseases such as malaria. The inability to afford preventive and curative measures such as protective bed-nets and effective medical treatment can lead to high malaria mortality rates among this group. The absence of adequate early warning mechanisms for potential epidemic outbreaks and the lack of a good information system that can communicate predictable effects of climate change are some of the institutional shortcomings that further increase vulnerability to malaria outbreaks. The East African governments also have no comprehensive programmes or fiscal facilities to deal with climate variability and extremes. The few malaria programmes that exist are run by civil society or state governments with assistance from major external resources such as aid agencies, international institutions and donor country programmes. Therefore, the local capacity to develop adaptive strategies to cope with climate variations and extremes is still very poor at all levels and remains a major challenge. Future adaptation programmes should therefore take into account the diversity of factors that influence a society’s capacity to cope with these changes. Such programmes should have as major inputs demographic trends and socioeconomic factors, since these have an effect on land use, which may, in turn, accelerate or compound the effects of climate change. Positive trends in demographic, economic and social development would contribute towards better living conditions and therefore increase the ability to cope with the potential consequences of climate change. Diseases such as HIV/AIDS, malaria, diarrhoeal diseases and respiratory diseases are significant factors that affect not only people’s health but also their productivity and responsiveness to external threats. The trends in dealing with these diseases must therefore be factored into the analysis of climate change-related vulnerability. The implementation of policies and measures to enable the development of adequate early warning systems and sound communication systems with respect to climate variability and change would contribute towards the better management of disease epidemics. Institutional programmes that are geared towards antici-
3243 J&J Vulnerability
1/11/07
10:24 am
Page 394
394 Climate Change and Vulnerability
patory adaptation measures would be best able to address the climate-related vulnerabilities of the population.
Notes 1
2 3 4
Hay et al (2002) have disputed the association of malaria outbreaks with climate variability and change and reported finding no significant changes in temperature or vapour pressure at any of the highland sites that had reported high malaria incidences. However, these results have been challenged by Patz et al (2002) (and later others), who report a warming trend and claim that the use of a downscaled gridded climate data set by Hay et al (2002) ignores climate dependencies on local elevation, which compromises the accuracy of the results. The natural swamps in the valley bottoms contain papyrus, which, due to its cooling properties and its mosquito-inhibiting natural oil secretions, is believed to hamper mosquito development (Lindblade et al, 2000; Reiter, 2001). This sample represents 4, 5 and 7 per cent of the total number of households in the lowest level of administrative unit in Kabale, Kericho and Muleba. Malaria programmes and strategies in East Africa are guided by the overall health policy, whose goal is to provide universal primary healthcare. Such strategies seek to reduce malaria through the promotion of primary healthcare, increasing access to healthcare services and encouraging private sector participation in the delivery and financing of healthcare services. Coexisting with national health policies are international programmes like the Global Malaria Control Strategy, which advocates four technical measures: • • • •
5
sustainable preventive measures such as the use of ITNs; early diagnosis and treatment; early detection and prevention of epidemics; and strengthening local research capacities.
Subsistence farmers in Kabale, for instance, are worried that they can no longer accurately predict the onset of rains and that the rains have reduced in amount. This is affecting their agricultural productivity, income and nutritional status, and hence increasing their vulnerability to climate-related diseases.
References Afrane, Y. A., B. W. Lawson, A. K. Githeko, G. Yan (2005) ‘Effects of microclimatic changes caused by land use and land cover on duration of gonotrophic cycles of Anopheles gambiae (Diptera: Culicidae) in Western Kenya Highlands’, Journal of Medical Entomology, vol 42, pp974–980 Conway, D. (2002) ‘Extreme rainfall events and lake level changes in East Africa: Recent events and historical precedents’, in E. O. Odada and D. Olago (eds) The East African Great Lakes: Limnology, Palaeolimnology and Biodiverity, Kluwer Academic Publishers, Dordrecht, Germany, pp64–92 De Savigny, D., E. Mewageni, C. Mayombana, H. Masanja, A. Minhaji, D. Momburi, Y. Mkilindi, C. Mbuya, H. Kasale, H. Reid, and H. Mshinda (2004) ‘Care-seeking patterns in fatal malaria: Evidence from Tanzania’, Malaria Journal, vol 3, no 27, available at www.pubmedcentral.nih.gov/articlerender.fcgi?artid=514497 Fontaine, R. E., A. E. Najjar, J. S. Prince (1961) ‘The 1958 malaria epidemic in
3243 J&J Vulnerability
1/11/07
10:24 am
Page 395
Vulnerability to Climate-Induced Highland Malaria in East Africa 395 Ethiopia’, American Journal of Tropical Medicine and Hygiene, vol 10, pp795–803 Fowler, V. G. Jr., M. Lemnge, S. G. Irare, E. Malecela, J. Mhina, S. Mtui, M. Mashaka, and R. Mtoi (1993) ‘Efficacy of chloroquine on Plasmodium falciparum transmitted at Amani, eastern Usambara Mountains, Northeast Tanzania: An area where malaria has recently become endemic’, Journal of Tropical Medicine and Hygiene, vol 6, pp337–345 Garnham, P. C. C. (1945) ‘Malaria epidemics at exceptionally high altitudes in Kenya’, British Medical Journal, vol 11, pp45–47 Githeko, A. K. and S. Clive (2005) ‘The history of malaria control in Africa: Lessons learned and future perspectives’, in K. L. Ebi, J. Smith and I. Burton (eds) Integration of Public Health with Adaptation to Climate Change: Lessons Learned and New Directions, Francis and Taylor, London Githeko, A. K., J. M. Ayisi, P. K. Odada, F. K. Atieli, B.A. Ndenga, I. J. Githure and G. Yan (2006) ‘Topography and malaria transmission heterogeneity in the western Kenya highlands: Prospects for vector control’, American Journal of Tropical Medicine and Hygiene (in press) Githeko, A. K. and W. Ndegwa (2001) ‘Predicting malaria epidemics in the Kenyan Highlands using climate data: A tool for decision makers’, Global Change and Human Health, vol 2, pp54–63 Githeko, A. K., S. W. Lindsay, U. E. Confaloniero and J. A. Patz (2000) ‘Climate change and vector-borne disease: a regional analysis’, Bulletin of World Health Organization, vol 78, pp1136–1147 Greenwood, B. (2004) ‘Between hope and a hard place’, Nature, vol 430, pp926–927 Hay, S. I., M. Simba, M. Busolo, A. M. Noor, H. L. Guyatt, S. A. Ochola and R. W. Snow (2002) ‘Defining and detecting malaria epidemics in the highlands of western Kenya’, Emerging Infectious Diseases, vol 8, pp555–562 Khaemba, B. M., A. Mutani and M. K. Bett (1994) ‘Studies of anopheline mosquitoes transmitting malaria in a newly developed highland urban area: A case study of Moi University and its environs’, East African Medical Journal, vol 3, pp159–164 Kilian, A. H. D., P. Langi, A. Talisuna and G. Kabagambe (1999) ‘Rainfall pattern, El Niño and malaria in Uganda’, Transactions of the Royal Society of Tropical Medicine and Hygiene, vol 93, pp22–23 Lepers, J. P, P. Deloron, D. Fontenille and P. Coulanges (1988) ‘Reappearance of falciparum malaria in central highland plateaux of Madagascar’, Lancet, 12 March, p586 Lindblade, K. A., E. D. Walker, A. W. Onapa, J. Katunge and M. Wilson (1999) ‘Highland malaria in Uganda: Prospective analysis of an epidemic associated with El Niño’, Transactions of the Royal Society of Tropical Medicine and Hygiene, vol 93, pp480–487 Lindblade, K. A., E. D. Walker, A. W. Onapa, J. Katunge and M. L. Wilson (2000) ‘Land use change alters malaria transmission parameters by modifying temperatures in a highland area of Uganda’, Tropical Medicine International Health, vol 5, pp263–274 Lindsay, S. W. and W. J. M. Martens (1998) ‘Malaria in the African highlands: Past, present, and future’, Bulletin of World Health Organization, vol 76, pp33–45 Loevinsohn, M. E. (1994) ‘Climate warming and increased malaria in Rwanda’, Lancet, vol 343, pp714–748 Malakooti, M. A., K. Biomndo and G. D. Shanks (1998) ‘Reemergence of epidemic malaria in the highlands of western Kenya’, Emerging Infectious Diseases, vol 4, pp671–676 Matola, Y. G., G. B. White and S. A. Magayuka (1987) ‘The changed pattern of malaria endemicity and transmission at Amani in the eastern Usambara mountains, northeastern Tanzania’, Journal of Tropical Medicine and Hygiene, vol 3, pp127–134 McMichael, A. J., A. Hames, R. Scooff and S. Covats (eds) (1996) Climate Change and
3243 J&J Vulnerability
1/11/07
10:24 am
Page 396
396 Climate Change and Vulnerability Human Health: An Assessment Prepared by a Task Group on Behalf of the World Health Organization, World Meteorological Organization and the United Nations Environment Programme, Geneva, Switzerland Minakawa, N., S. Munga, F. Atieli, E. Mushinzimana, G. Zhou, A. K. Githeko and G. Yan (2005) ‘Spatial distribution of anopheline larval habitats in Western Kenyan highlands: Effects of land cover types and topography’, American Journal of Tropical Medicine and Hygiene, vol 73, pp157–165 Morse, S. S. (1995) ‘Factors in the emergence of infectious diseases’, Emerging Infectious Diseases, vol 1, pp7–15 Mouchet, J., S. Manuin, S. Sircoulon, S. Laventure, O. Faye, A. W. Onapa, P. Carnavale, J. Julvez and D. Fontenille (1998) ‘Evolution of malaria for the past 40 years: Impact of climate and human factors’, Journal of American Mosquito Control Association, vol 14, pp121–130 Munga, S., N. Minakawa, G. Zhou, E. Mushinzimana, O. O. Barrack, A. K. Githeko and G. Yan (2006) ‘Association between land cover and habitat productivity of malaria vectors in Western Kenyan highlands’, American Journal of Tropical Medicine and Hygiene, vol 74, pp69–75 Mwisongo, A. and J. Borg (eds) (2002) Proceedings of the Kagera Health Sector Reform Laboratory 2nd Annual Conference, Ministry of Health, United Republic of Tanzania Nicholson, S. E. (1996) ‘A review of climate dynamics and climate variability in Eastern Africa’, in T. C. Johnson and E. O. Odada (eds) The Limnology, Climatology and Palaeoclimatology of East African Lakes, Gordon and Breach, Australia, pp25–56 Nuwaha, F. (2002) ‘People’s perceptions of malaria in Mbarara, Uganda’, Tropical Medicine International Health, vol 7, pp462–470 Ogallo, L. J. (1989) ‘The spatial and temporal patterns of the East African rainfall derived from principal components analysis’, International Journal of Climatology, vol 9, pp145–167 Patz, J. A., K. Strzepek, S. Lele, M. Hedden, S. Greene, B. Noden, S. I. Hay, L. Kalkstein and J. C. Beier (1998) ‘Predicting key malaria transmission factors, biting, and entomological inoculation rates, using modeled soil moisture in Kenya’, Tropical Medicine International Health, vol 3, pp818–827 Patz, J. A, M. Hulme, C. Rosenzweig, T. D. Mitchell, R. A. Goldberg, A. K. Githeko, S. Lele, A. J. McMichael and D. Le Sueur (2002) ‘Regional warming and malaria resurgence’, Nature, vol 420, pp627–228 Reiter, P. (2001) ‘Climate change and mosquito-borne diseases’, Environmental Health Perspectives, vol 109, supplement 1, pp141–161 Roberts, J. M. D. (1964) ‘Control of epidemic malaria in the highlands of Western Kenya, Part I: Before the campaign’, Journal of Tropical Medicine and Hygiene, vol 61, pp161–168 Ropelewski, C. F. and M. S. Halpert (1987) ‘Global and regional-scale precipitation patterns associated with the El Niño/Southern Oscillation’, Monthly Weather Review, vol 115, pp1606–1626 Shanks, G. D., K. Biomondo, S. I. Hay and R. W. Snow (2000) ‘Changing patterns of clinical malaria since 1965 among a tea estate population located in the Kenyan highlands’, Transactions of the Royal Society of Tropical Medicine and Hygiene, vol 94, pp253–255 Some, E. S. (1994) ‘Effects and control of highland malaria epidemic in Uasin Gishu District, Kenya’, East African Medical Journal, vol 7, pp2–8 Walsh, J. F., D. H. Molyneux and M. H. Birley (1993) ‘Deforestation: Effects on vector-borne disease’, Parasitology, vol 106, pp55–75 WHO (2002) ‘Roll back malaria’, World Health Organization, Geneva, Switzerland, www.rbm.who.int WHO (1996) World Health Report: Fighting Diseases, Fostering Development, World
3243 J&J Vulnerability
1/11/07
10:24 am
Page 397
Vulnerability to Climate-Induced Highland Malaria in East Africa 397 Health Organization, Geneva, Switzerland Worrall, E., A. Rietveld, C. Delacollette (2004) ‘The burden of malaria epidemics and cost-effectiveness of interventions in epidemic situations in Africa’, American Journal of Tropical Medicine and Hygiene, vol 71, supplement 2, pp136–140 Zhou, G., N. Minakawa, A. K. Githeko and G. Yan (2004) ‘Association between climate variability and malaria epidemics in the East African highlands’, Proceedings of National Academy of Sciences, vol 101, pp2375–2380
3243 J&J Vulnerability
1/11/07
10:24 am
Page 398
19
Vulnerability to Dengue Fever in Jamaica Charmaine Heslop-Thomas, Wilma Bailey, Dharmaratne Amarakoon, Anthony Chen, Samuel Rawlins, David Chadee, Rainaldo Crosbourne, Albert Owino, Karen Polson, Cassandra Rhoden, Roxanne Stennett and Michael Taylor
Introduction Dengue fever is one of the most severe insect-borne viral infections; it is potentially fatal and is currently endemic in more than 100 countries in Africa, the Americas, the Eastern Mediterranean, Southeast Asia and the Western Pacific, with Southeast Asia and the Western Pacific being the most seriously affected (WHO, 1997). It is a flu-like illness but may develop into the more serious dengue haemorrhagic fever/dengue shock syndrome, which can result in death. In the Caribbean, virological evidence of dengue fever was first obtained in the 1950s, although the disease is believed to have existed there for the past 200 years (Ehrenkranz et al, 1971). The outbreak of dengue haemorrhagic fever in Cuba in 1981, which affected almost half the population, is considered to be one of the most important events in the history of dengue in the Americas (CAREC, 1997). Since this event there have been confirmed or suspected cases of dengue haemorrhagic fever almost every year in the American region. The last large epidemic in Jamaica occurred about ten years ago. Transmission of the disease from a sick to a healthy person occurs via the bite of the Aedes mosquito, of which the tropical and subtropical species, Aedes aegypti, is common in the Caribbean. This mosquito species has a large variety of breeding places and especially thrives in urban environments with poor sanitation. It can be found breeding in water storage containers and receptacles within the home; in water collected in blocked drains, improperly discarded tyres, bottles and coconut shells (World Resources Institute, 1998);
3243 J&J Vulnerability
1/11/07
10:24 am
Page 399
Vulnerability to Dengue Fever in Jamaica 399
and in rain water accumulated in tree holes and herbaceous plants. The eggs can resist desiccation for up to one year and hatch when sites are flooded with water. This results in the sudden emergence of mosquitoes at the end of long, dry spells. Vector abundance and the frequency of vector–human contact have been identified as the most significant factors that influence dengue outbreaks in the Caribbean region. The dengue virus itself exists as four distinct serotypes, which further adds to the risk of acquiring infection. Infection with one serotype leads to protection against homologous re-infection but provides only brief protection against heterologous infection (WHO, 1997). Currently all four serotypes of the dengue virus are in circulation in Jamaica. The risk of acquiring dengue haemorrhagic fever/dengue shock syndrome increases with sequential dengue virus infections (Valdes et al, 2000). The potential impact of climate change on the incidence of dengue in the Caribbean has emerged as an important cause for concern because the development, dynamics, abundance and geographical distribution of vectors as well as viruses tend to be affected by elements of climate (Martens et al, 1997). Temperature is an especially important factor, influencing mosquito development as well as viral multiplication and development within the mosquito (Wilson, 2001). It is expected that an increase in temperature will shorten the incubation period of the dengue virus within the mosquito and result in potentially higher transmission rates (Watts et al, 1987). It has already been demonstrated that increasing temperatures are altering the geographic range of A. aegypti, which is now appearing at higher elevations than before (Suarez and Nelson, 1981; Koopman et al, 1991). Previously limited to locations below 1000m above sea level, dengue appeared for the first time at 2000m sites in Colombia in 1980 (Suarez and Nelson, 1981) and at 1700m in Mexico in 1986 (Koopman et al, 1991). The situation is somewhat more complex for countries like Jamaica, where the disease is currently endemic and where conditions are already favourable for the vector and virus. The high level of poverty here, especially in rural areas and inner city regions, results in poor environmental sanitation and further encourages the existence and spread of the disease. Higher temperatures resulting from climate change are expected to cause Aedes aegypti to become infective faster, reproduce more rapidly and bite more frequently – factors that could increase transmission rates in dengue endemic areas (Focks et al, 1995) and also increase the risk of dengue haemorrhagic fever (Patz et al, 1998; WHO, 2002). Preliminary studies have also shown a link between El Niño Southern Oscillation phenomena and the incidence of dengue in some Pacific island nations (Hales et al, 1999). Given the important influence of climate on the status of dengue in the Pacific island nations, our objective here is to assess the existing vulnerability of the population of Jamaica to dengue outbreaks and to identify the specific factors that put communities at risk. This is examined at two levels: at the government/ministry level by means of stakeholder interviews with key individuals and at the community level by means of a local case study using survey questionnaires. The results of this assessment can then serve to inform further
3243 J&J Vulnerability
1/11/07
10:24 am
Page 400
400 Climate Change and Vulnerability
evaluation of the ability of the country to respond to a possible future increase in the transmission of dengue fever due to climate change as well as to stressful events in general. It will also help to lay the ground for the identification of potential adaptation options in successive studies.
Background A significant warming trend has been noted for the Caribbean region over the past two decades, and a similar, though less significant, increasing trend has been observed for rainfall over the past four decades (Peterson et al, 2002). Dengue data obtained from the Caribbean Epidemiological Centre (CAREC) in Trinidad and Tobago for the corresponding period also shows a higher incidence of the disease in the 1990s than in the 1980s.
Figure 19.1 Time series graph of reported cases of dengue with rainfall and temperature for Jamaica Source: Stennett (2004).
On a seasonal basis, dengue epidemics have been observed to peak in the latter part of the year, after a few warm months when maximum temperatures have been attained and usually when the rain is receding. There is thus a distinct lag between the attainment of maximums in temperature and rainfall and epidemic outbreaks (Figure 19.1), with a greater lag correlation observed for temperature than for rainfall. The stronger association of the disease with temperature can be explained by the fact that temperature not only results in shorter viral cycles within the mosquito (Koopman et al, 1991; Focks et al, 1995; and Hales et al, 1996) but also results in increased biting frequency of
3243 J&J Vulnerability
1/11/07
10:24 am
Page 401
Vulnerability to Dengue Fever in Jamaica 401
mosquitoes (McDonald, 1957), leading to greater vector–human contact. The association with rainfall, though weaker, is nonetheless important since either an increase in or lack of rainfall results in the increased availability of breeding grounds in the form of accumulated rainwater or stored water respectively, resulting in increased vector abundance. The majority of dengue outbreaks have also been found to occur in the El Niño and El Niño+1 years (Table 19.1 and Figure 19.2), which is possibly due to the existence of warmer temperatures and drier than normal conditions during the latter part of El Niño years and warmer and wetter than normal conditions during the first half of El Niño+1 years (Chen and Taylor, 2002; Taylor et al, 2002). These temperature and rainfall conditions are once again conducive to vector abundance and viral propagation.
Figure 19.2 Variation of annual reported cases and rate of change Note: Key: En – El Niño; W-En – Weak El Niño; En+1 – El Niño+1; Ln+1 – La Niña+1.
Table 19.1 Distribution of epidemic peaks among ENSO phases, 1980–2001 Region Caribbean Trinidad and Tobago Barbados Jamaica Source: CSGM (2004).
Total
El Niño and El Niño+1
La Niña
Neutral
8 8 6 5
7 6 5 4
– – – –
1 2 1 1
3243 J&J Vulnerability
1/11/07
10:24 am
Page 402
402 Climate Change and Vulnerability
Mean temperature projections for stations in Jamaica and Trinidad for 2020, 2050 and 2080, using statistical downscaling methods and based on the A2 and B2 scenarios from the Intergovernmental Panel on Climate Change’s Special Report on Emissions Scenarios (SRES), indicate a possibility of increased warming in the next century, with an increase of nearly 2ºC by 2080 (CSGM, 2004). The trend in rainfall is not as pronounced (CSGM, 2004). Some climate models also indicate an increase in the frequency of El Niño Southern Oscillation events due to the projected changes in climatic parameters (Timmermann et al, 1999). Given the association between dengue incidences and temperature, these future climate conditions may serve to enhance dengue transmission rates in the Caribbean and increase the possibility of epidemic outbreaks. The present overall public immunity to the disease in this region is also likely to be low since the last large outbreak occurred almost a decade ago, though there have been smaller outbreaks. There are also other individual and contextual circumstances that could modify vulnerability to dengue and make communities more or less susceptible. The various issues that affect vulnerability to dengue in the Caribbean region, especially in the context of climate change, are presented in detail in the sections that follow.
Institutional Capacity An important factor that determines vulnerability to a future increased incidence of dengue in Jamaica is the existing capacity of institutions in the country to deal with such health impacts. An effort was therefore made to assess institutional preparedness to handle climate change-related health challenges, the manner in which key individuals and decision makers interpreted their roles, and current efforts at sensitizing the public to climate change and its implications. For this purpose, at a more generic level, interviews were conducted with the heads of agencies responsible for preparing Jamaica’s First National Communication to the United Nations Framework Convention on Climate Change, namely the National Environmental Planning Agency, the National Meteorological Division, and the Office of Disaster Preparedness and Emergency Management. At a more specific level, six officials in key positions in the Ministry of Health were interviewed: the principal budget holder, two medical officers in the surveillance department, the environmental health officer, the chief public health officer and a health educator. These officers not only served as key informants but also as stakeholders who could become involved in the study right from the initial stages and could use the results for the benefit of their constituents. Additionally, scientists in the Climate Change Group of the University of the West Indies were also interviewed.
Generic institutional capacity The Office of Disaster Preparedness and Emergency Management has a mandate to manage all aspects of disaster management and risk reduction. It does so by an inclusive approach, working in partnership with other agencies, which
3243 J&J Vulnerability
1/11/07
10:24 am
Page 403
Vulnerability to Dengue Fever in Jamaica 403
allows it to pursue programmes that would otherwise be impossible to implement and also to influence national risk reduction. Sea-level rise and the inundation of coastal areas resulting in population displacement were stated as key areas of concern during interviews. However, climate change risk information was not included in the agency’s public education campaign, even though such communication was considered important. The topic was thought to be too complicated for the general population, which usually did not worry about issues that did not affect them immediately. The organization heads at this office also exhibited little awareness about the possible health effects of climate change, since health was not regarded as a part of their mandate. Resource constraints were also stated as a factor. The National Environment and Planning Agency’s mission is to promote sustainable development by ensuring protection of the environment (NEPA, undated). The issue of climate change was addressed indirectly in their public education programmes, though the specific term was not used since it was once again considered complicated. The attitude of this agency was found to mirror that of the Office of Disaster Preparedness and Emergency Management: instead of looking for ways to communicate climate change concerns to the public, they presumed a lack of intelligence and avoided the issue altogether. They, too, did not find the need for the inclusion of the health threats in their mandate. In contrast, the National Meteorological Service was the only agency that specifically included in its mission climate change research and communication of climate change-related risk information to stakeholders and the general public to enable adaptation by affected sectors. Unlike officials at the previous agencies, the expert interviewed here demonstrated a full appreciation of the health implications of climate change and considered public access to this kind of information to be important. There is some information on health impacts in the literature produced by the Meteorological Service, but implementation of public education programmes was only considered feasible if more resources were available, and even then addressing sea-level rise would always take precedence over other impacts. What was curious about the above agencies was the narrow interpretation of their mandates and the unwillingness to acknowledge health as a part of their purview. Since its establishment in 1980, the Office of Disaster Preparedness and Emergency Management has routinely been called on to deal with three types of hazards – hurricanes, landslides and floods caused by intense rainfall – and such extreme events have been noted to have increased in frequency and intensity in the recent past. Displacement of population resulting from hurricanes and floods has often been accompanied by outbreaks of communicable diseases in shelters managed by this office (Bailey, 1989), and this is one of the most critical areas in shelter management. Yet there was no acknowledgment of an interest in diseases resulting from hazards. The recent increase in the frequency and intensity of hurricanes and floods was also not viewed as a possible consequence of climate change and was not considered as a more immediate threat than the effects of sea-level rise. Concerning the National Environment and Planning Agency, it is interesting to note that its interpretation of its mission objective of sustainable
3243 J&J Vulnerability
1/11/07
10:24 am
Page 404
404 Climate Change and Vulnerability
development for environmental protection does not seem to include any aspect of population health, despite the direct linkage between environmental health and population health. The interviewee from the National Environment and Planning Agency completely denied any interest in health within the agency or any plans to incorporate health impacts into the education programme of the organization. The preoccupation with sea-level rise was observed to be at the forefront at all three organizations, and health issues were considered the responsibility of the Ministry of Health. Though the Meteorological Service was the only agency of the three that demonstrated a good awareness of climate-related health issues it was nonetheless reluctant to see itself in partnership with the Ministry of Health.
Specific institutional capacity Interviews with key officials at the Ministry of Health showed that there was overall a good awareness about the potential for increased temperature and precipitation in the Caribbean due to climate change. All but one official interviewed agreed that such climatic impacts could lead to an increase in dengue transmission. However, despite the relatively high level of awareness, there are no long-term strategies in place or under consideration to address the possible negative impacts of climate change. Dengue fever is classified as a Class 2 disease and given significantly less priority than Class 1 diseases,1 especially HIV/AIDS. The problem of inadequate financial resources, largely due to cuts in budgetary allocations, and the resulting need to establish priorities, was repeatedly cited as a major issue during interviews. There is a growing feeling, however, that the competition between HIV/AIDS and diseases such as dengue is not so much due to financial issues but rather stems from a lack of attention. For example, the Global Fund for HIV/AIDS brings large sums of money into the country and this has the unfortunate effect of diverting attention and manpower away from other health issues.2 The present approach to dengue prevention and control in the Caribbean therefore tends to be mostly reactive. There is also no vaccine against dengue or dengue haemorrhagic fever, and the only effective method of prevention is the elimination of vectors and their breeding places. The World Health Organization (WHO) has very clearly outlined priorities for the control of dengue, including epidemiological surveillance (WHO, 2002)3, but Jamaican health authorities often find these requirements difficult to meet, largely due to staff shortages of as much as 50 per cent at various levels in the ministry and to shortages of entomologists and public health inspectors. This situation is not expected to improve in the near future since many of these positions come with heavy workloads and unattractive salaries and are not considered respectable. There is also no routine vector surveillance and control programme in place, due to the expense involved and the lack of any budgetary allocation to such an exercise, according to one official from the Ministry of Health in the parish of St James. As a result, there are only knee-jerk responses to reports of heavy infestation in specific communities or special venues, such as important official meetings or events.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 405
Vulnerability to Dengue Fever in Jamaica 405
Identification of dengue cases was also cited as an issue, as there is only one under-equipped virology laboratory on the island and blood samples from sick persons must be sent to Trinidad and Tobago for disease identification. The position of the Ministry of Health is that environmental sanitation must be the responsibility of community members since their actions to a large extent determine the existence of mosquito breeding sites. It felt that government intervention could result in a shift of responsibility from the community to the state and generate a false sense of security among the public, who would then do nothing to control breeding sites, as has been the experience in Asia and the Americas (Gubler, 2002). Jamaican communities, on the other hand, considered dengue prevention and control to be entirely a government responsibility. On the positive side, the country does have a well-organized system of primary healthcare under a recently decentralized health services programme, which divides the island into four autonomous health regions (Figure 19.3) under four regional authorities. The system is based on a nested system of health centres, offering different levels of care. There is a health centre within five miles of every community on the island (Figure 19.4), which ensures greater sensitivity to local needs. This system could potentially be mobilized during emergency situations and could help facilitate greater responsiveness in the event of epidemic outbreaks, despite the financial constraints plaguing the Jamaican health system. There are only a few areas on the island that may experience difficulty in accessing health services due to geographic and socio-organizational factors.
Figure 19.3 Health regions in Jamaica Source: Ministry of Health, Jamaica (2006)
3243 J&J Vulnerability
1/11/07
10:24 am
Page 406
406 Climate Change and Vulnerability
Figure 19.4 Health centres in Jamaica Source: Ministry of Health, Jamaica (2006)
The Ministry of Health also has a long tradition of involvement in policy-oriented research and there have been occasions when research/policy collaborations have been initiated by the ministry to investigate the effects of its policies on vulnerable groups (Gordon-Strachan et al, 2005). This kind of interest and involvement on the part of the ministry could potentially be further developed to facilitate continued research on dengue in Jamaica and serve to inform policymakers to enable appropriate decision making.
Community Vulnerability To determine vulnerability to dengue fever at the community level and thus provide a more comprehensive assessment, a local case study was undertaken in a section of the parish of St James in the northwest of Jamaica. According to the dengue records for this parish, there had been a concentration of cases within the city of Montego Bay and sporadic cases within the vicinity of a permanent stream and its associated seasonal streams and gully banks (Figure 19.5). Three communities along this hydrological feature were selected: Granville/Pitfour, a suburb of the parish capital Montego Bay; Retirement, immediately beyond the boundaries of the urban area; and rural John’s Hall. A 10 per cent sample of heads of households in these communities (Table 19.2) was then surveyed using a questionnaire to determine socioeconomic conditions of householders, support systems available, their knowledge of the disease, and cultural practices that might have important implications for the spread of dengue fever.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 407
Vulnerability to Dengue Fever in Jamaica 407
Figure 19.5 Distribution of dengue cases in St James, Jamaica, 1998 Table 19.2 Sample size Community Granville/Pitfour Retirement John’s Hall Total
Number of households
Sample
1507 485 572 2564
151 49 57 257
Of the three communities surveyed, Granville/Pitfour is a low-income community of about 6300 people on the outskirts of the tourist centre of Montego Bay and comprises a mix of formal and informal structures. Most of the informal structures are one-roomed dwellings with a room density of 4.2 persons per room, among the highest of the three communities under study. Heads of households were either self-employed or worked in the service sector in Montego Bay. A few miles to the southeast is Retirement, a lower middleincome community of about 1780 people with few obvious informal dwellings. Heads of households here are employed in the service sector in Montego Bay and in public service. Further southeast is John’s Hall, a poor community of rural squatters. It consists of small, crudely built houses, most lacking basic amenities and scattered over rugged terrain. Roughly 60 per cent of the households here are headed by women, who typically have very little education or training and are employed in domestic service and petty trading activities with incomes at the minimum wage or below. Male heads share similar characteris-
3243 J&J Vulnerability
1/11/07
10:24 am
Page 408
408 Climate Change and Vulnerability
tics and are employed as gardeners and labourers on construction sites. Unemployment in this community stands at 33 per cent. Even for Jamaica as a whole, a little over 66 per cent of single parent households living in poverty in 2002 were in fact headed by females (PIOJ/STATIN, 2002). This is significant in light of the fact that gender is recognized globally as an extremely important factor in explaining poverty. In our survey, responses to questions such as the frequency with which households were forced to borrow money or take food on credit, had to rely on relatives or friends, and survived on restricted food intake were combined to form an index of coping, and the general picture obtained was one in which large numbers were ‘struggling to make ends meet’. Such conditions serve to greatly limit the coping capacity of the population to added health stresses. About 50 per cent of household heads in the survey also suffered from chronic illnesses, mainly hypertension and diabetes. It has been noted that households with disabled or ill members display greater vulnerability since this reduces the number of members available for productive labour and puts a strain on household resources (Kouri et al, 1989; Nyong et al, 2003). The strongest association was found with chronic illnesses, which are often incapacitating and require strict adherence to therapeutic and dietary regimes. Rapid access to health facilities can minimize vulnerability, and geographically Granville/Pitfour is the best positioned in terms of access to comprehensive public-sector services in Montego Bay, with John’s Hall the least. With respect to access to services, households in which there is no piped water are more at risk than those with a piped supply. In the three communities investigated, 23 per cent had no water piped into their homes or yards, and even when there was piped supply, it was often irregular, necessitating storage of some sort. This is especially true for the squatter communities in urban areas, which are, by law, not permitted piped water supply. The degree of risk varies with the mode of storage. Focks and Chadee (1997), working in Trinidad and Tobago, found that the outdoor drum was one of the most productive Aedes breeding containers and was most commonly found in homes with no access to piped supply. Four types of containers – outdoor drums, tubs, buckets and small containers – accounted for more than 90 per cent of all Aedes aegypti pupae discovered. Focks and Chadee concluded that the provision of an adequate water supply system and targeted source reduction had the potential to reduce pupal production in the country by more than 80 per cent. In our survey in St James, 54 per cent of the respondents stored water in drums, which were usually left uncovered to facilitate the entry of rain water and for easy access to the stored water. The overall awareness about dengue in the three communities surveyed was generally low. Many respondents were unfamiliar with the vector and had little knowledge about the habits of mosquitoes, which contributed to their inability to adequately protect themselves against the disease. The Aedes species typically bites in the early mornings and at dusk, so bed-nets that protect against mosquitoes at night are ineffective in controlling dengue fever. The best forms of protection have been found to be screens or mesh on windows and doors (Ko et al, 1992) to keep mosquitoes out of homes. Studies in Taiwan
3243 J&J Vulnerability
1/11/07
10:24 am
Page 409
Vulnerability to Dengue Fever in Jamaica 409
have shown that screens can eliminate as much as 63 per cent of dengue infection, and in Puerto Rico, the absence of screens showed a strong correlation with the occurrence of dengue fever (Dantes et al, 1988; Ko et al, 1992; Morens et al, 1978). Installation of screens or mesh is, however, expensive and lowincome households usually depend on repellents and mosquito destroyers, which often simply force mosquitoes into cupboards and other hiding places from which they later emerge. Screens were only used on some houses in the formal settlements in Granville/Pitfour and Retirement. Eight per cent of the sample used repellents, but the majority used no form of protection at all. Responsibility for vector control remained an issue, and it was observed that members of communities often do not appreciate the importance of their role in disease management. About 78 per cent of community members considered dengue control to be entirely a government responsibility. Community cooperation was also found to be low in insecticide spraying activities undertaken by the government in response to complaints of high mosquito infestation levels. This activity requires that all doors and windows be kept open during the spraying, but only about 44 per cent of individuals said they abided by this directive.
Ranking of communities Table 19.3 gives a ranking of the three communities using specific indicators of vulnerability. For each indicator, a score of 1 represents the lowest and a score of 3 the highest level of vulnerability. In tandem with poverty, vulnerability was found to increase outwards from the urban area. The score was the lowest for Granville/Pitfour and highest for John’s Hall, the rural squatter community. The difference between Granville/Pitfour and Retirement was small and this was attributed to the mix of formal and informal settlements in the Granville community, which resulted in depressed scores. In order to more accurately differentiate between the groups a vulnerability index was prepared based on survey responses and using indicators of vulnerability. Scores of 0 and 1 were assigned for the absence and presence of vulnerability respectively. A test of normality (mean ±3 standard deviations) was applied to the results and the data were observed to be normally distributed with a mean of 5.7 and standard deviation of 1.96. On this basis, five groups displaying varying degrees of vulnerability were created as shown in Table 19.4. The most vulnerable group (Group 5) comprised 24 respondents or 9 per cent of the sample, while the least vulnerable group (Group 1) accounted for 14 respondents or about 5 per cent of the sample. Most of those in the most vulnerable group, Group 5, lived in the community of John’s Hall, while 64 per cent of those classified as least vulnerable (Group 1) lived in Granville/Pitfour (Table 19.5). The overwhelming majority of the most vulnerable had no knowledge about the disease, its symptoms or mode of transmission. Together with the high level of poverty this means that they are not in a position to effectively protect themselves from the vectors of the disease. Most households headed by women were found to be at risk.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 410
410 Climate Change and Vulnerability
Table 19.3 Composite of ranking for communities in St James Vulnerability indicators
John’s Hall (%)
Score
Retirement (%)
Score
Granville/ Pitfour (%)
Score
No knowledge of dengue fever No knowledge of disease symptoms No protection Income minimum wage or less Inability to cope No pipe at home Water storage in drums Chronic illness Distance from health facility Female household head Believe dengue control is public health responsibility Total Score
52.6
2
53.1
3
42
1
72
3
69
2
59
1
95 68
3 3
92 33
2 1
89 61
1 2
63
3
51
2
50
1
46
3
12
2
11
1
65
3
53
2
44
1
53 70
2 3
37 49
1 2
54 13
3 1
60
3
47
1
55
2
56
3
47
1
51
2
31
19
16
Note: 1 = least vulnerable; 2 = vulnerable; 3 = highly vulnerable.
Finally, Table 19.6 shows the characteristics of Groups 4 and 2 (high and low vulnerability groups), both of which interestingly contain a majority of respondents from Granville/Pitfour. This exercise therefore brings out the effect of the mix of formal and informal settlements in Granville/Pitfour, which results in its dual nature in terms of vulnerability and highlights the situation of squatter settlements within this community and elsewhere on the island.
Discussion and Conclusion Vulnerability to dengue fever in Jamaica thus emerges as a multifaceted issue that stems from weaknesses at the social, economic and political levels. A potentially increased incidence of dengue in the future resulting from the projected continuation of the warming trend observed in Jamaica over the past
3243 J&J Vulnerability
1/11/07
10:24 am
Page 411
Vulnerability to Dengue Fever in Jamaica 411
Table 19.4 Identification of vulnerable groups Group
Vulnerability
Total
Measure
5
Most vulnerable
24
Mean +1.5SD >8.64 Mean + 0.5SD to Mean + 1.5SD 6.68–8.64 Mean ±0.5SD 5.72–6.68 Mean -0.5SD to Mean -1.5SD 2.76–5.72 < Mean -1.5SD <2.76 Normality Mean ±3SD Mean = 5.7 SD = 1.96 Mean ±3SD = 0.18–11.58
4
67
3
Average
2
51 101
1
Least Vulnerable
14
couple of decades raises serious concerns about the impacts of such outbreaks and the capacity of the island to cope. These findings are especially significant given that a substantial number of people in the Caribbean are poor and often live in situations conducive to the proliferation of the vector and the virus. Lack of decent living conditions and lack of access to infrastructure and services are at the forefront of issues that stem from poverty and greatly affect coping capacity in the face of climate-related stresses. In Jamaica, about 15 per cent of the population lived below the poverty line in 2002 (PIOJ/STATIN, 2002), earning the minimum wage or below. The majority of these people
Table 19.5 Characteristics of the most and least vulnerable groups Characteristics
Group 5 (%)
Group 1 (%)
1. No knowledge of vector 2. No knowledge of dengue symptoms 3. No Protection 4. Minimum wage or less 5. Inability to cope 6. Female head 7. Storage in drums 8. No piped water 9. Distance from health facility 10. Chronic disease 11. No personal acceptance of responsibility for dengue control Community with highest proportion
92 96 92 92 83 83 79 50 79 79 83
7 14 57 21 0 21 21 0 7 21 7
Johns Hall – 67%
Granville/Pitfour – 64%
3243 J&J Vulnerability
1/11/07
10:24 am
Page 412
412 Climate Change and Vulnerability
Table 19.6 Characteristics of Groups 4 and 2 Characteristics 1. No knowledge of vector 2. No knowledge of dengue symptoms 3. No protection 4. Minimum wage or less 5. Inability to cope 6. Female head 7. Storage in drums 8. No piped water 9. Distance from health facility 10. Chronic disease 11. No personal acceptance of responsibility for dengue control Community with highest proportion
Group 4 (%) 75 83 100 69 75 69 54 31 48 70 63 Granville/Pitfour – 48%
Group 2 (%) 21 48 89 38 37 44 44 10 16 36 39 Granville/Pitfour – 69%
resided in informal settlements Although the precise number of persons who live in such settlements in Jamaica is not known, it has been estimated that about 60 per cent of the population of the city of Montego Bay (about 55,000 to 60,000 people) live in informal settlements (Ministry of Environment and Housing, 1997). A significant number of these poor households were also headed by women, which serves to highlight gender-associated vulnerability to climate-related health stresses. In terms of infrastructure and services, access to piped water supply came across as a critical factor, with a lack of piped water increasing susceptibility to dengue fever by necessitating water storage, which, in turn, created additional breeding grounds for mosquitoes. Furthermore, there was a complete lack of awareness about the disease and its vector among the population surveyed. More than a half of those interviewed in the communities could not say what causes the disease, and the overwhelming majority had no knowledge of its symptoms (Table 19.3). The impact of the socioeconomic situation is compounded by the complete absence of any programme to address even the current threat of dengue fever at the policy and planning level. This is partly due to the classification of dengue as a Class 2 disease in comparison to the more serious Class 1 diseases like HIV/AIDS which warrant much global attention. Moreover, sea-level rise garners most of the attention in terms of climate change-related concerns, and many representatives from ministries responsible for articulating the country’s position on climate change are in fact unaware of the health implications thereof. As a result, the current capacity to deal with large outbreaks is quite low. Responses to outbreaks are typically reactive in nature and largely limited to spraying insecticide in mosquito-infested areas. There are also no public education programmes in place to increase awareness about the prevention and control of dengue. Distribution of responsibility for dengue control is itself a big question at the moment.
3243 J&J Vulnerability
1/11/07
10:24 am
Page 413
Vulnerability to Dengue Fever in Jamaica 413
This situation in Jamaica thus raises serious concerns about its capacity to cope with the future health impacts of climate change. While there is merit to the Ministry of Health’s position that communities must take some responsibility for vector control, at the same time communities cannot be expected to take action unless they are made aware about the disease. Public education that can adequately inform and empower communities therefore comes across as a strong necessity to address the knowledge gap revealed in the study and encourage collective action. Supporting initiatives to arrange provision of lowcost water storage options such as covered drums and for granting security of tenure to those who, because of their status, are denied access to running water, would also be important. Additionally, the narrow focus on sea-level rise must be revised to include a broader range of climate change impacts that can be equally devastating for affected communities. Health-related threats must be included among environmental hazards and addressed by institutions responsible for dealing with hazards, such as the Office of Disaster Prevention and Emergency Management. Similarly, the National Environment and Planning Agency and the National Meteorological Service must also specifically target climate-related health threats in their public education programmes. A concerted effort by the various ministries dealing with climate related issues would lend support to the Ministry of Health, which, by itself, is in no position to meet the challenge of increased disease transmission on the island. Strong government initiatives can help draw attention to the issue and enable the generation of additional resources for its management and control. Resources must also be targeted at the existing healthcare system in Jamaica, which is a definite strength and must be adequately equipped for mobilization in the case of epidemic outbreaks. Due recognition to the issue from national authorities would also be conducive to encouraging community participation in the implementation of measures to prevent and control dengue. A partnership approach that includes the government, communities and the private sector has the best potential to enable the development of appropriate policies and measures to reduce the vulnerability of Jamaica to the growing threat of increased dengue epidemics due to a changing climate in the Caribbean.
Notes 1 2 3
Class 1 diseases are those which must be reported immediately to the Ministry of Health and include HIV/AIDS, malaria and diseases preventable by immunization. Class 2 diseases are reportable weekly in line listing. This problem arising out of specific disease-focused funding programmes is currently being researched by the Alliance for Health Policy and Systems Research, an initiative of the Global Forum of Health Research and the WHO (2005). Priorities for the control of dengue outlined by the WHO include epidemiological surveillance, in other words both entomological surveillance and monitoring of human behaviours that contribute to larval habitats (WHO, 2002). Health authorities are also expected to improve emergency preparedness and response and strengthen national control programmes; promote behavioural change by developing guidelines for the prevention and control of vectors; and support dengue
3243 J&J Vulnerability
1/11/07
10:24 am
Page 414
414 Climate Change and Vulnerability research programmes that can also help build national and international partnerships (WHO, 2002).
References Alliance for Health Policy Systems Research/WHO (2005) Effects of Global Health Initiatives on Health Systems Development, Alliance for Health Policy Systems Research and World Health Organization, Geneva, Switzerland Bailey, W. (1989) ‘Disease outbreak in ODPEM shelters: Hurricane Gilbert 1988’, Disaster Report, no 5, Pan American Health Organization CAREC (1997) Epinote: An Update of Dengue Fever in the Caribbean, Caribbean Epidemiological Centre, Trinidad, West Indies Chen, A. A. and M. A. Taylor (2002) ‘Investigating the link between early season Caribbean rainfall and the El Niño+1 year’, International Journal of Climatology, vol 22, pp87–106 CGSM (Climate Studies Group Mona) (2004) ‘The threat of dengue fever in the Caribbean’, Project of the Climate Studies Group Mona, University of the West Indies, Jamaica Dantes, H., J. S. Koopman, C. Laddy, M. Zarate, M. Magin, I. Longini, E. Guttierez, V. Rodriquez, L. Garcia, and E. Mirelles (1988) ‘Dengue epidemics on the Pacific coast of Mexico’, International Journal of Epidemiology, vol 17, pp178–186 Ehrenkranz, N., A. Ventura and R. Cuadrado (1971) ‘Pandemic dengue in Caribbean countries and the Southern United States: Past, present and potential problems’, New England Journal of Medicine, vol 285, pp1460–1469 Focks, D., E. Daniels, D. Haile and J. Keesling (1995) ‘A simulation model of the epidemiology of urban dengue fever: Literature analysis, model development, preliminary validation, and samples of simulation results’, American Journal of Tropical Medicine and Hygiene, vol 53, pp489–506 Focks, D. and D. Chadee (1997) ‘Pupal survey: An epidemiologically significant surveillance method for Aedes Aegypti: An example using data from Trinidad’, American Journal of Tropical Medicine and Hygiene, vol 56, no 2, pp159–167 Gordon-Strachan, G., W. Bailey, S. Lalta, E. Ward, A. Henry-Lee and E. LeFranc (2005) ‘Linking researchers and policy makers: Some challenges and approaches’, Ministry of Health/ University of the West Indies, Mona, Jamaica Gubler, D. (2002) ‘Epidemic dengue/dengue hemorrhagic fever as a public health, social and economic problem in the 21st century’, Trends in Microbiology, vol 10, no 2, February Hales, S., P. Weinstein, Y. Souares and A. Woodward (1996) ‘Dengue fever epidemics in the South Pacific Region: Driven by El Niño Southern Oscillation’, Lancet, vol 348, pp1664–1665 Hales, S., P. Weinstein, Y. Souares and A. Woodward (1999) ‘El Niño and the dynamics of vector-borne disease transmission’, Environmental Health Perspectives, vol 107, pp99–110 Ko, Y., M. Chen and S. Yeh (1992) ‘The predisposing and protective factors against dengue virus transmission by mosquito vector’, American Journal of Epidemiology, vol 136, pp214–220 Koopman, J. S., D. Prevots, M. Mann and H. Dantes (1991) ‘Determinants and predictors of dengue infection in Mexico’, American Journal of Epidemiology, vol 133, pp1168–1178 Kouri, G. P., M. G. Bravo, M. Guzman and C. Triana (1989) ‘Dengue hemorrhagic fever/dengue shock syndrome: Lessons from the Cuban epidemic, 1981’, Bulletin of the World Health Organization, vol 67, pp375–380
3243 J&J Vulnerability
1/11/07
10:24 am
Page 415
Vulnerability to Dengue Fever in Jamaica 415 Martens, W., T. Jetten and D. Focks (1997) ‘Sensitivity of malaria, schistosomiasis and dengue to global warming’, Climate Change, vol 53, pp145–156 McDonald, G. (1957) The Epidemiology and Control of Malaria, Oxford University Press, London, UK Ministry of Environment and Housing (1997) ‘Low-income settlement policy design and development project’, draft final implementation plan, report IDB ATN/SF 5104-JA, Ministry of Environment and Housing, Kingston, Jamaica Ministry of Health, Jamaica (2006) Ministry of Health Annual Report: 2005, Ministry of Health, Kingston, Jamaica Morens, D., J. Rigan-Perez and R. Lopez-Correa (1978) ‘Dengue in Puerto Rico, 1977: Public health response to characterize and control an epidemic of multiple serotypes’, American Journal of Tropical Medicine and Hygiene, vol 35, pp197–211 NEPA (undated) ‘About NEPA’, National Environment and Planning Agency website, www.nepa.gov.jm/about/aboutnepa.htm#mission Nyong, A. (2003) ‘Vulnerability of rural households to drought in Northern Nigeria’, AIACC Notes, vol 2, no 2, p7 Patz, J. A., W. J. M. Martens, D. A. Focks and T. H. Jetten (1998) ‘Dengue fever epidemic potential as projected by general circulation models of global climate change’, Environmental Health Perspectives, vol 106, pp147–153 Peterson, T. C., M. A. Taylor, R. Demeritte, D. L. Duncombe, S. Burton, F. Thompson, A. Porter, M. Mercedes, E. Villegas, R. S. Fils, A. K. Tank, A. Martis, R. Warner, A. Joyette, W. Mills, L. Alexander and B. Gleason (2002) ‘Recent changes in climate extremes in the Caribbean region’, Journal of Geophysical Research, vol 107, p4601, doi: 1029/2002JD002251 PIOJ/STATIN (2002) Economic and Social Surveys of Jamaica, Planning Institute of Jamaica/Statistical Institute of Jamaica, Kingston, Jamaica Stennett, R. (2004) ‘Retrospective study: Climate and dengue fever in the Caribbean’, unpublished document, Climate Studies Group, Mona, Jamaica Suarez, M. F. and M. J. Nelson (1981) ‘Registro de altitude del Aedes Aegypti en Colombia’, Biomedica, vol 1, p225 Taylor, M. A., D. B. Enfield, and A. A. Chen (2002) ‘The influence of the tropical Atlantic vs. the tropical Pacific on Caribbean rainfall’, Journal of Geophysical Research, vol 107, p3127, doi: 10, 1029/2001JC001097 Timmerman, A., J. Oberhuber, A. Bacher, M. Esch, M. Latif, and E. Roeckner (1999) ‘Increased El Niño frequency in a climate model forced by future greenhouse warming’, Nature, vol 398, April 22 Valdes, K., M. Alvarez, M. Pupo, S. Vázquez, R. Rodríguez and M. G. Guzmán (2000) ‘Human Dengue Antibodies against Structural and Nonstructural Proteins’, Clinical and Diagnostic Laboratory Immunology, publication of the American Society of Microbiology, vol 7, no 5, September, pp856–857 Watts, D. M., D. S. Burke, B. A. Harrison, R. E. Whitmire and A. Nisalak (1987) ‘Effect of temperature on the vector efficiency of Aedes aegypti for Dengue-2 virus’, American Journal of Tropical Medicine and Hygiene, vol 36, pp143–152 WHO (1997) ‘World Health Report: Communicable disease surveillance and response’, in Dengue Haemorrhagic Fever: Diagnosis, Treatment, Prevention, and Control, 2nd edition, World Health Organization, Geneva, Switzerland WHO (2002) ‘Fifty-fifth World Health Assembly Provisional Agenda’, Item 13.14, March 4, Annecy, France Wilson, M. L. (2001) ‘Ecology and infectious disease in ecosystem change and public health’, in J. L. Aron and J. Patz (eds) Ecosystem Change and Public Health, Johns Hopkins University Press, Baltimore, MD WRI (1998) World Resources 1998–1999: A Guide to the Global Environment: Environmental Change and Human Health, World Resources Institute, Oxford University Press, New York
3243 J&J Vulnerability
1/11/07
10:24 am
Page 416
3243 J&J Vulnerability
1/11/07
10:24 am
Page 417
Index access food 175, 177, 178, 188–189 markets 232–233, 326, 327 resources 20–21, 267 services and infrastructure 408, 411, 412 adaptation strategies drought 254–255 food security 179, 193–194 forests 61–63 malaria 389–390, 393–394 tourism and small islands 164–168 watersheds 323, 326 adaptive capacity dengue fever 412–413 description of 89, 198, 240 droughts 241 estuarine fishing 144, 146 farmers’ social vulnerability 258, 260, 262–263, 266–267, 268–270, 271, 273, 274 floods 128–129, 131 national assessment 210–213, 216 rice production 342 river basins 98, 108, 109, 111 species level 44 tea production 368, 369, 370 see also coping capacity Africa biodiversity 33–48 droughts 239–256 food security 20, 179–180, 188, 192 malaria 25, 375–397 Nigeria 198–238 outcomes of concern 11 agriculture climatic threat spaces 279–306 droughts 239 expansion 244
farmers’ social vulnerability 257–275 food availability 176 mapping vulnerability 101–104 poverty 211 rice production 8, 21, 333–350 sensitivity 205–210, 214, 216 watersheds 312, 318–319 agro-pastoralism 228, 231, 242, 246, 249 AIACC see Assessments of Impacts and Adaptations to Climate Change alien invasive species 41–42 analytical hierarchy processes 271 Arba’at , Red Sea State, Sudan 246–247, 249, 250–251 Argentina 13, 22, 24, 117–133, 257–278, 279–306 artisanal fishing 17, 143–147 Assessments of Impacts and Adaptations to Climate Change (AIACC) 4, 7, 10, 12, 14–15, 18–19, 26 assets 228–229, 233, 238 autonomous response strategies 247–249 availability of food 175, 176, 193 awareness coffee production 283 disease 386, 403, 404, 408–409, 412 droughts 231, 249 floods 130 forest adaptation 63 Bara Province, Sudan 180–182 baseline scenarios 137 bed nets 386, 389, 408 Beja pastoralists 242, 246, 249
3243 J&J Vulnerability
1/11/07
10:24 am
Page 418
418 Climate Change and Vulnerability
Bhalme and Mooley Drought Intensity Index (BMDI) 220, 221 biocapacity 74 biodiversity 11–13, 33–48 biome impacts 13, 33–45 biophysical vulnerability 219, 321, 322 biotemperature 52, 53 BMDI see Bhalme and Mooley Drought Intensity Index bottom–up approaches 224, 226, 234, 245, 330 brushlands 50–51, 318, 321 Buenos Aires, Argentina 117–133 capacity 81, 167–168, 178, 402–406 see also adaptive capacity capital 182, 183, 243 carbon dioxide biomes 13, 39, 45 climate 218, 219 rice production 337, 338, 341, 347 tea production 362, 364, 367 yield sensitivity 207 carbon:nitrogen ratios 83, 84 carbon sequestration 181–182, 246, 251 Caribbean region 398–415 cash-crops 23–24, 213, 261, 262–263, 264, 265, 266, 267 cassava 209 causal chains 191, 193 cautious behaviour 145–146 CCAM see conformal cubic atmospheric model cereal production 333–350 CGCM model 356, 364, 365, 366, 367, 370 child dependency burden 212, 214 China 6–8, 88–114 Chingowa village, Magumeri Local Government State, Nigeria 183–185 climatic threat spaces 279–306 coastal areas 13–17, 115–169 coffee production 282–289, 296 collective approaches 79–80, 81, 129
commercial aspects 24, 259, 272, 274 communal approaches farmers’ social vulnerability 268–270, 271, 272, 273 food security 179, 180, 188 livestock herding 81 savanna biome 44 communication 130, 394 community level dengue fever 406–410, 413 droughts 233, 250, 250–251, 254 early strategic actions 249 food security 188 forests 62 organizations 233, 250, 267 rangeland rehabilitation 181–182, 246, 251 tourism 166 watersheds 22, 308–309, 319, 321–322 see also local level conflict agricultural expansion 244 droughts 184, 247, 252, 254 food security 22–23, 184 water use 97, 101, 104, 109 conformal cubic atmospheric model (CCAM) 337 consultations 95–96, 97–98 contingency plans 231 Cook Islands 16 coordination 130, 167–168 coping capacity disease 389–390, 412–413 rice production 342, 343, 344, 345, 348, 349 tea production 368 see also adaptive capacity coping deficit 244–245 coral reef systems 159, 160, 166, 167 Córdoba Province, Argentina 24, 260–267, 299–301 costs coral bleaching 160 extreme weather 16, 156 flood damage 125 tourism 164, 165, 168
3243 J&J Vulnerability
1/11/07
10:24 am
Page 419
Index 419
credit systems 210–211, 251 CSIRO-Mk2b 82 CSIRO model 356, 365, 366, 367, 370 cyclones 156–157, 159 dams 312 deforestation 54 degradation of land 8–9, 21–22, 23, 254 demographic level 211–212, 311–312, 327, 329 dengue fever 25, 398–415 dependence ratio indicator 107 dependency agricultural 207–209, 216, 342, 343, 345 child 212, 214 external assistance 329 household 233, 234 human and biodiversity 43–44, 45 markets 247, 248 relief 253 desertification 9–11, 181 determinants of vulnerability 189–190, 192, 193, 230, 237–238 developing countries 279 development 22, 88, 97, 165–166, 312–314, 330–331, 348 disaster preparedness 28, 402–403 disease dengue fever 25, 398–415 food security 20, 180, 185 HIV/AIDS 180, 213, 394, 404 malaria 25, 375–397 dispersers 35, 36, 38, 41, 44, 45 diversification 166–167, 168, 232, 258, 272, 345 domains of vulnerability 4–5 DPSIR index see driver-pressurestate-impact-response index driver-pressure-state-impactresponse (DPSIR) index 136, 141 drivers, climate coastal areas and small islands 14–15 ecosystems and biodiversity 12
human health 26 natural resources 7, 9, 10, 11–13 rural economies 18–19 droughts coffee production 284, 286 food security 20, 174, 175, 180–187, 188, 190–191, 194 health 27 household level 224–233, 237–238 impacts 6 indicators 100–101, 102–103 livelihoods 239–256 livestock herding 73–74, 75, 77–78 maize production 294 Mongolia 69–70 Nigeria 208, 218, 219, 220, 231, 233 resource systems 94, 107 sorghum 22 tea production 361–362 dry forests 13, 56, 57, 62, 63 early warning systems 91, 393, 394 East Africa 375–397 ECHAM4 82, 83, 147–148 economic aspects adaptive capacity 212 change 21, 24, 67–68, 85 coffee production 289 diversification 258 droughts 228, 231, 237–238, 254 estuarine fishing 145, 146 export-oriented 23–24, 351 farmers’ social vulnerability 258–259, 274 food security 17–24, 190 globalization 279 malaria 389 resource-based 94 rice production 342, 343, 349 storm surges 16 tea production 370–371 tourism 16, 155–169, 165 ecosystem pressures 11–13 ecotourism 43, 167 education 210, 212–213, 214, 233, 324–325, 403, 404, 413
3243 J&J Vulnerability
1/11/07
10:24 am
Page 420
420 Climate Change and Vulnerability
elevation malaria 377, 380, 390, 391 tea production 357, 358, 359, 360, 361, 362, 364, 365 El Niño Southern Oscillation (ENSO) climatic threat spaces 280–281 coffee production 288, 289 dengue fever 25, 399, 401, 402 estuaries 135, 136, 140, 150 forests 56 maize production 293 malaria 25, 376, 380, 381–382, 384–385, 390, 393 small island states 156 watersheds 320 emission scenarios see scenarios; Special Report on Emissions Scenarios employment 164, 165, 188, 191, 193, 209, 242, 248, 407–408 see also labour ENSO see El Niño Southern Oscillation environmental aspects agricultural development 274 estuaries 145, 148–150 food security 190 gated communities 126–127, 131 health 403–404, 405 maize production 186 pastoralism 72–78 risk 90 tourism 159 EPIC see Erosion Productivity Impact Calculator epidemics 25, 376, 379, 387, 390–392, 391, 392, 393, 398, 401 Erosion Productivity Impact Calculator (EPIC) 205 estuaries 13, 16–17, 117, 118–120, 134–154 eutrophication 16–17, 139, 141 evapotranspiration 104, 105 export-oriented economies 23–24, 351 exposure description of 89, 198, 240
droughts 228, 229, 234, 237, 241 farmers’ social vulnerability 258 floods 122, 123, 126, 127 malaria 382 resource systems 90, 95, 96–97, 98, 107 rice production 342, 344, 348 variability 213–214 external assistance 329 extreme weather climatic threat spaces 285, 287, 288, 293–294, 302 cyclones 156–157, 159 dengue fever 403 forests 63, 64 heat waves 69, 107 indices of vulnerability 316–317 rice production 341–342, 346–347 severe winters 21–22, 70–71, 74–78, 84–85 Southeast Asia 325 storm surges 16, 117–133 watersheds 319–321 see also droughts; El Niño Southern Oscillation; floods; La Niña famine 70, 239, 240, 241, 242, 253 Famine Early Warning System (FEWS) 91 farmers’ social vulnerability 257–278 farm size 323–324, 326, 327–328 FEWS see Famine Early Warning System Fiji 16 financial aspects adaptive capacity 210–211 dengue fever 404 droughts 228, 231, 237–238, 243 food security 180, 182, 183, 187 malaria 389 tea production 368, 371 tourism 168 see also costs fires 42, 318, 320 fisheries 17, 143–147, 312
3243 J&J Vulnerability
1/11/07
10:24 am
Page 421
Index 421
floods climatic threat spaces 290, 291 coastal areas and small islands 14–16 farmers’ social vulnerability 261, 263, 267 household vulnerability 20–21 maize production 294 rice production 341–342 rivers 8 urban 117–133 flow, river 138, 144, 148, 149, 382, 390 food droughts 240, 241–242 income 211 index of vulnerability 315–318 market dependence 248 resource systems 94 rice production 345 security 6, 17–24, 97, 171–372, 387, 393 watersheds 328 forests 13, 49–66, 319, 321, 329–330 fragmentation of land 40, 184 fynbos biome 11–13, 33, 34, 35, 38, 43, 44, 45 gated communities 126–127, 128, 131 GCMs see global circulation models gender 328, 408 see also women general circulation models 82, 281–282, 294 generic institutional capacity 402–404 Gireigikh rural council, Bara Province, Sudan 180–182 global circulation models (GCMs) 37–38, 147–148, 354–356, 370 globalization 274–275, 279 global warming 49, 174, 203, 325, 338–339 see also temperature González, Tamaulipas, Mexico 267–272, 272, 273, 274, 294–299 government level 67–68, 187, 193, 282, 283, 393, 402–406, 413
grant systems 180, 188 grasslands 50, 64, 72–73, 318, 321 groundwater depletion 222 growing seasons 100, 101, 102–103, 208 guinea corn 209 HAB see hypoxia and harmful algal blooms HadCM3 model 82, 83, 147–148, 354, 355, 356, 366, 370 health 24–27, 191, 193, 210, 213, 317, 373–415 heat waves 69, 107 Heihe river basin, China 6–8, 88–114 highland malaria 25, 375–397 HIV/AIDS 180, 213, 394, 404 Holdridge life zones 51–53, 54, 55, 56, 58–60 hospitals 379, 383–384, 388 hotels 165–166 household level access to resources 20–21 dengue fever 406–410 droughts 241–242 farmers’ social vulnerability 24, 257–278 food security 173–196 malaria 385–390 Nigeria 209–210, 212, 214, 215, 218–238 rice production 342, 343, 344, 346–347, 348 watersheds 309, 314–315, 322–323, 328 housing 126–127, 131 human assets 228–229, 233, 238 human capital 243 human health 24–27, 373–415 hydrology 382–383, 392–393 hypoxia and harmful algal blooms (HABs) 139, 141, 142–143 income adaptive capacity constraints 211, 212
3243 J&J Vulnerability
1/11/07
10:24 am
Page 422
422 Climate Change and Vulnerability
agriculture 97, 259, 270, 272 dengue fever 407 diversification 232, 272 droughts 232, 242 estuarine fishing 143, 144 food security 188, 193 malaria 386–387, 389, 393 rice production 21 Southeast Asia 335 watersheds 325, 326–327 indicators adaptive capacity 262, 269, 271 dengue fever 409–410 droughts 220, 221, 224–226, 234, 240–241 estuaries 136, 140–143 household 315–318, 326 livelihood assets 250 malaria 387 national 198, 213–215 resource vulnerability 91, 92, 96–111 rice production 342, 343–344 social vulnerability 122, 260, 262, 263–264, 269, 271 indirect sensitivity 209–210 information 129, 231, 403 infrastructure adaptive capacity 212 dengue fever 411, 412 droughts 232, 233 floods 122, 123, 124, 125, 131 food security 177, 188–189 health 27 population 220 small islands 155, 156 tourism 159–160, 165–166 insecticide treated nets (ITNs) 386, 389 institutions change 21, 67–68, 78–81 dengue fever 402–406 droughts 252–253, 254 estuarine fishing 145 floods 130, 131 watersheds 311, 329, 330 integrated approaches 168, 378–379
Intermediate Technology Development Group (ITDG) 252 interquartile range methods 280–281 Intertropical Convergence Zone (ITCZ) 156, 157, 247 irrigation 105, 106, 179, 180, 231, 313 ITDG see Intermediate Technology Development Group ITNs see insecticide treated nets Jamaica 398–415 karoo biome 13, 28, 33–34, 35–36, 38–39, 43, 44, 45 Kenya 375, 378, 385, 391, 392 Khor Arba’at , Sudan 246–247, 249, 250–251 knowledge 130, 385–386, 413 Kordofan State, Sudan 246, 247–248 Laboulaye, Cordoba Province, Argentina 260–267, 290–294, 299–301 labour 184–185, 188, 213, 232 see also employment Lake Victoria, East Africa 25, 377–394 land agricultural expansion 244 degradation 8–9, 21–22, 23, 254 food security 184, 188 forests 50–51, 54 holding size 369–370 ownership 21, 184, 249, 259, 328 pressures on 8–11 registration 251 transformation 38, 40, 376–377 La Niña 135, 156, 293, 380, 381–382 Lao People’s Democratic Republic 21, 333, 335, 337–338, 340, 341, 344, 345, 346, 347, 348–349 La Plata river basin 13–16, 16–17, 117, 118–120, 134–154 large farm holders 24, 262–267 liberalization of trade 185–186 life history characteristics 40–41
3243 J&J Vulnerability
1/11/07
10:24 am
Page 423
Index 423
life zone classification 51–53, 54, 55, 56, 57 literacy 213, 214 livelihoods agricultural dependency 207–209, 216 diversification 232 droughts 226, 228, 239–256, 242, 248 food security 173, 178–196 indices of vulnerability 317 pastoralists 72–78 rural 17–24 savanna biome 43–44 variability 3 watersheds 312 livestock droughts 183–184, 232, 241, 242, 248, 251–252 extreme weather 67–87 farmers’ social vulnerability 261, 263–264, 265, 266 food availability 176 karoo biome 43 land degradation 11, 21–22 ownership 24 local level 63, 166, 174–175, 254–255 see also community level logging 50, 61 long-term strategies 165–166, 193 Lotka–Volterra approach 37 Lower Mekong river basin 8, 21, 333–350 Magumeri Local Government Area, Nigeria 183–185 maize production 185–187, 205–207, 209, 289–294, 301 malaria 25, 375–397 Maldives 160 Mangondi village, South Africa 179–180 mapping 91, 98, 101–111, 318–319, 321–322 market economy transition 67–68, 80, 85
markets accessibility of 232–233, 326, 327 dependence 247, 248 food security 24, 179, 180 tourism 158 matrix management 42 medication, self 385 meteorological organizations 403 Mexico climatic threat spaces 279–306 food security 20, 185–187, 189, 192 land degradation 22 market integration 24 social vulnerability 257–278 migration droughts 23, 184, 229, 232, 248–249 estuarine fishing 144 food security 20, 185 health 27 livestock herding 77, 80–81 natural resources 9 watersheds 328 yield sensitivity 207–208 mixed farm holders 262–267 moist forests 13, 56, 57–61 Mongolia 9, 11, 21–22, 24, 67–87 morbidity 375, 379, 383 mortality 70, 72, 75, 77, 159, 375, 388 multidisciplinary approaches 120–123 multi-level indices 315–318 multiple destination, cross–section models 163 Namibia 13 national level 61–62, 63, 198–217, 244, 365–367, 369, 370–371 natural capital 243 natural resources 5–13, 23, 31–114, 228, 231, 237, 250 see also resources niche characteristics 40 Nigeria 6, 20, 23, 183–185, 188–189, 192, 198–217, 218–238
3243 J&J Vulnerability
1/11/07
10:24 am
Page 424
424 Climate Change and Vulnerability
nitrogen 83, 84, 139 non-climate factors climatic threat spaces 280, 281 coastal areas and small islands 14–15 ecosystems and biodiversity 12 human health 26 malaria 385–390 natural resources 6, 7, 9, 10 rice production 348 rural economies 18–19 North Darfur, Sudan 9, 23, 247, 252 ordinal logistic regression analysis 226–227, 230, 237–238 organizations, social 233, 250, 251, 267, 327, 328 outcomes of concern coastal areas and small islands 14–15 domains of vulnerability 5 ecosystems and biodiversity 11, 12 human health 26 natural resources 6, 7, 9, 10 rural economies 18–19 over-grazing 9, 11, 81 ownership land 21, 184, 249, 259, 328 livestock 24, 80 tea production 351, 358, 368, 369 see also tenure Pajas Blancas, Uruguay 143–147 Palmer droughts Severity Index (PDSI) 100–101, 102–103 Pantabangan–Carranglan watershed, The Philippines 8, 20–21, 22, 63, 307–332 Parana delta 16 participation dengue fever 413 droughts 226, 242–243 watersheds 313–314, 314–315, 318, 319, 321, 330 pastoralism agricultural expansion 244
agro- 228, 231, 242, 246, 249 corridors 221 droughts vulnerability 231–232 extreme weather 67–87 food security 184–185 pasture productivity 11, 21, 72–73, 82, 83, 85 PDSI see Palmer droughts Severity Index per capita water resources 106–107 perceptions 223–224, 226, 233–234, 263, 268–270, 270, 291, 341–342 PET see potential evapotranspiration ratio The Philippines 8, 11, 13, 20–21, 22, 49–66, 307–332 physical capital 243 planning 62, 165–166, 168, 244–245, 249–252, 255, 412 policy level agricultural 244, 259, 274 change 68, 78–81, 185–186 dengue fever 412 droughts 252–253, 254 food security 176, 185–186, 187, 193 forest adaptation 61–62 malaria 394 tourism 159 watersheds 329–330, 330–331 water use conflicts 101 political level 258–259 population adaptive capacity 210, 211–212 droughts 254 estuaries 134, 150 floods 124, 125 infrastructure 220 malaria 376–377 resource systems 92, 93, 106–107 potential evapotranspiration (PET) ratio 52–53 poverty adaptive capacity 210–211, 214 dengue fever 399, 408, 409, 411–412 food security 188, 189
3243 J&J Vulnerability
1/11/07
10:24 am
Page 425
Index 425
livestock herding 80 malaria 386, 387, 388, 393 watersheds 313 precipitation Africa 37–38, 203 Argentina 261, 290, 291, 299–301 China 94, 100–101, 104 climatic threat spaces 280, 282 coffee production 284, 285, 286, 287, 288 dengue fever 25, 400 forests 13, 49–66 Jamaica 402 Lao PDR 337–338 maize production 292, 293, 294 malaria 25, 376, 381–382, 390, 391, 392–393 Mexico 267, 294–296 Mongolia 68, 69, 70, 82, 85 Nigeria 200, 202, 203, 204, 208, 216, 218, 219, 220–221 Pantabangan–Carranglan watershed 8, 309–310 Río de la Plata 134–154 Seychelles 157, 159 Southeast Asia 325, 337, 339 Sri Lanka 353, 354, 356 Sudan 241, 246, 247 tea production 360–361, 367, 370 Thailand 338 tourism 160–161, 164 yields 205, 207, 341 private farmers 268–270, 271, 272, 273 probabilistic approaches 111 productivity agriculture 17, 244 droughts 182, 184 food availability 176 pasture 11, 21, 72–73, 82, 83, 85 tea 358–360 see also yield Proteaceae 35, 37, 38 public buildings 122, 124 public education 403, 404, 413 Puerto Rico 409 purchasing power 242, 259
rainfall see precipitation rangelands 72–78, 181–182, 183, 246, 251 rational runoff coefficients 104 real estate damage 123, 125, 131 Red Sea State, Sudan 246–247, 250–252, 253 reforestation 22, 310, 312–313, 330 refugia 40, 44 regression equations 298–299 relief 77, 253, 329 reservoirs 309, 312 residential development 126–127, 131 resilience 89, 144–146, 244 resources 20–21, 88–114, 174, 267 see also natural resources return periods 119, 122, 123, 124, 125 revenue 365–367, 370–371 rice production 8, 21, 333–350 rich farmers 324, 325, 326 Río de la Plata 13–16, 16–17, 117, 118–120, 134–154 risk 4–5, 8, 90, 126–129, 207–209, 223–224, 226, 333–350 rivers basins 6–8, 88–114, 134–154, 333–350 floods 117, 118–120 malaria 390, 392 rice production 21, 333–350 sea level rise 13–16 water quality 16–17, 134–154 River Uruguay 134, 138 Roll Back Malaria programme 389 Roque Sáenz Peña County, Argentina 289–294 runoff 104, 106–107, 247 rural economies 17–24, 171–372 Sahelian zone 22, 218–238 St James, Jamaica 406–410 sanitation 405 Santa Lucia river 139–140 savanna biome 11, 34–35, 36, 39, 43–44, 45
3243 J&J Vulnerability
1/11/07
10:24 am
Page 426
426 Climate Change and Vulnerability
Savannakhet Province, Lao PDR 21, 333, 335, 337–338, 340, 341, 344, 345, 346, 347, 348–349 scaled indicators 108, 240–241 scenarios baseline 137 climatic threat spaces 281–282, 294–302 dengue fever 401–402 estuarine fishing 146–150 farmers’ social vulnerability 274–275 floods 120–123 forests 53, 54–61 livestock herding 82–84 national 203 resource systems 95 rice production 337–339, 346 Sri Lanka 354–356 tea production 364–365 tourism 163–164 watersheds 315, 325–326 see also simulations screens 408–409 sea level rise 13–16, 117–133, 159, 412, 413 seasonal aspects biodiversity 43 coffee production 284 disease 383–385, 400 estuaries 149 food security 177–178 malaria 382, 383, 384 migration 24, 79, 81, 144, 221 rice production 21 seed availability 315 yields 208 see also variability seed availability 315–318, 323 self-assessed perceptions 223–224, 226, 234 self-medication 385 sensitivity description of 198, 240 droughts 241 estuarine fishing 144–146 farmers’ social vulnerability 258,
260, 262–263, 266–267, 268–270, 271, 273 malaria transmission 376 national assessment 205–210, 214, 216 resource systems 89, 90, 95–96, 98 rice production 342, 348 tea production 351–352 services 408, 411, 412 severe winters 21–22, 70–71, 74–78, 84–85 severity index 223–224 Seychelles 16, 155–169 simulations 51–53, 340–341, 362–365 see also scenarios small island states 13–17, 155–169 small-scale farms 8, 21, 24, 184, 262–267, 323–324, 326, 358, 368, 370 snowfall 67, 68, 70, 71, 76, 85 social level adaptive capacity constraints/Nigeria 212 change 67–68, 85 droughts 229, 233, 238, 243 estuarine fishing 145 farmer vulnerability 257–278 floods 126–129 food security 190 organizations 251 safety nets 24 tourism loss impacts 164, 165 social vulnerability 219 socioeconomic level 77, 311–312, 323–324, 326, 327, 328–329, 335, 386–388, 393 soil conditions 8, 39, 40, 100 sorghum 22, 209, 268, 269, 274 South Africa 11, 20, 33–48, 179–180, 188, 192 South America 8 Southeast Asia 325, 335, 338–339 southern Africa 33–48 soybean production 22, 259, 261, 272, 274
3243 J&J Vulnerability
1/11/07
10:24 am
Page 427
Index 427
spatial distributions household droughts vulnerability 227 mapping watersheds 318–319, 321–322 Nigeria 213–215 resource systems 91, 98, 104–105, 111 Spearman correlations 326–327 Special Report on Emissions Scenarios (SRES), IPCC 82, 95, 136, 147, 354, 402 species level 11–13, 35, 36, 37–45, 64, 142–143 specific institutional capacity 404–406 SRES see Special Report on Emissions Scenarios Sri Lanka 23, 351–372 stakeholder consultation 63, 95–96, 97–98 storm surges 16, 117–133 streamflow 138, 148, 149, 379, 390 subsistence vulnerability 250 succulent karoo see karoo biome Sudan 6, 9, 20, 23, 180–182, 189, 192, 239–256 sudestadas see storm surges sugar cane 289 sustainable development 97, 403–404 sustainable livelihoods approaches 245, 255 synoptic variability 240 Taiwan 408–409 Tamaulipas, Mexico 22, 24, 267–272, 273, 274, 294–299 Tanzania 375, 378, 384, 385, 391 tea production 23–24, 351–372 temperature Africa 203 Argentina 290, 291, 292, 299, 299–301 China 94, 95, 100–101 climatic threat spaces 280, 282 coffee production 284, 285, 286, 287, 288, 296, 297–299
dengue fever 25, 399, 400–402 forests 13, 51, 52, 53–54, 54–56, 57, 58–60 Lao PDR 338 maize production 291, 293, 294 malaria 25, 376, 379–380, 382, 385, 390, 391, 392, 393 Mexico 267, 294–296 Mongolia 68–69, 70–71, 81–82, 85 Nigeria 199–201, 203, 204, 208–209, 214, 215, 216 Pantabangan–Carranglan watershed 310 Río de la Plata 136, 147–148, 149 savanna biome 39 Southeast Asia 325, 337, 338–339 southern Africa 37 Sri Lanka 352, 353, 354–355 Sudan 247 tea production 358–360, 364–365, 367, 370 Thailand 336, 338 tourism 160, 161–162, 164 yields 207, 341 see also global warming tenure 9, 24, 259, 268–270, 270, 271, 272, 273 see also ownership Thailand 21, 333, 335–336, 338, 340, 341, 344, 345, 346, 347, 348–349 threat spaces 279–306 three step risk assessment protocol 4–5 time-slice modelling 37, 38 Tlaxcala, Mexico 22, 185–187, 189 tourism 16, 43, 44, 155–169 trade liberalization 24, 185–186, 258–259 traditional practices 78–79, 81, 181, 389–390 transmission of disease 375–376, 390, 391, 398, 400 transport 177, 188–189, 388 treatment of disease 385, 387–388, 389–390, 390, 405–406 trophic states 16–17, 139, 140–143
3243 J&J Vulnerability
1/11/07
10:24 am
Page 428
428 Climate Change and Vulnerability
two-dimensional hydrodynamic models 120–121 typhoons 319–320 Ubon Ratchathani Province, Thailand 335–336, 338, 340, 341, 344, 345, 346, 347, 348–349 Uganda 375, 378, 385, 391 urban areas 117–133 Uruguay 13, 17, 143–147 utilization of food 175–176, 177–178, 178 variability climate threat spaces 279–280, 290–291, 297–298 exposure 213–214 food security 173–174, 179 forests 63, 64 livelihoods 3 malaria 376 Nigeria 200–202 Sahel, Nigeria 219, 220–221 scales of 240–241 sensitivity 205–209 Sri Lanka 353 Sudan 239, 240–242, 253–254 water resources 8 watersheds 307–332 see also seasonal aspects variability of the mean 282 vegetative cover 53–61 Veracruz, Mexico 282–289, 296 water access to 408, 412
conflicts 97, 101, 104, 109 harvesting 247, 252 indicators 88–111, 316–317 maize production 292 pressures on 5–8 quality 134–154 storage 408 withdrawal ratios 99–100 watersheds 8, 20–21, 22, 63, 134, 307–332 weather 74–78, 107–108, 231 see also extreme weather; El Niño Southern Oscillation; La Niña West Africa 218–238 wet forests 13, 56 wind 71, 120, 137, 144, 146 winter severity 21–22, 70–71, 74–78, 84–85 women 181, 182, 184, 328, 408, 409 see also gender word pictures 242–243 world food demand 173 yield extreme weather 320–321 maize 292–294, 301 rice 340–341, 346 sensitivity 205–209, 214 tea 358–362, 363–365, 365–367, 368 see also productivity zud 21–22, 70–71, 74–78, 84–85
3232 J&J Climate Change
1/11/07
9:56 am
Page i
Climate Change and Adaptation
3232 J&J Climate Change
1/11/07
9:56 am
Page ii
3232 J&J Climate Change
1/11/07
9:56 am
Page iii
Climate Change and Adaptation Edited by Neil Leary, James Adejuwon, Vicente Barros, Ian Burton, Jyoti Kulkarni and Rodel Lasco
London • Sterling, VA
3232 J&J Climate Change
1/11/07
9:56 am
Page iv
First published by Earthscan in the UK and USA in 2008 Copyright © The International START Secretariat, 2008 All rights reserved Climate Change and Adaptation: ISBN 978-1-84407-470-9 Climate Change and Vulnerability: ISBN 978-1-84407-469-3 Two-volume set: ISBN 978-1-84407-480-8 Typeset by FiSH Books, Enfield Printed and bound in the UK by Antony Rowe, Chippenham Cover design by Susanne Harris For a full list of publications please contact: Earthscan 8–12 Camden High Street London, NW1 0JH, UK Tel: +44 (0)20 7387 8558 Fax: +44 (0)20 7387 8998 Email:
[email protected] Web: www.earthscan.co.uk 22883 Quicksilver Drive, Sterling, VA 20166-2012, USA Earthscan publishes in association with the International Institute for Environment and Development A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data Climate change and adaptation / edited by Neil Leary ... [et al.]. p. cm. Includes bibliographical references. ISBN-13: 978-1-84407-470-9 (hardback) ISBN-10: 1-84407-470-6 (hardback) 1. Climatic changes. 2. Climatic changes—Environmental aspects I. Leary, Neil. QC981.8.C5C5113456 2007 304.2’5—dc22 2007023424
The paper used for this book is FSC-certified and totally chlorine-free. FSC (the Forest Stewardship Council) is an international network to promote responsible management of the world’s forests.
3232 J&J Climate Change
1/11/07
9:56 am
Page v
Contents List of figures and tables
viii
Acknowledgements
xiii
Foreword by R. K. Pachauri
xv
1
A stitch in time: General lessons from specific cases Neil Leary, James Adejuwon, Vicente Barros, Punsalmaa Batima, Bonizella Biagini, Ian Burton, Suppakorn Chinvanno, Rex Cruz, Daniel Dabi, Alain de Comarmond, Bill Dougherty, Pauline Dube, Andrew Githeko, Ayman Abou Hadid, Molly Hellmuth, Richard Kangalawe, Jyoti Kulkarni, Mahendra Kumar, Rodel Lasco, Melchior Mataki, Mahmoud Medany, Mansour Mohsen, Gustavo Nagy, Momodou Njie, Jabavu Nkomo, Anthony Nyong, Balgis Osman-Elasha, El-Amin Sanjak, Roberto Seiler, Michael Taylor, Maria Travasso, Graham von Maltitz, Shem Wandiga and Mónica Wehbe
1
2
Adapting conservation strategies to climate change in Southern Africa Graham von Maltitz, Robert J. Scholes, Barend Erasmus and Anthony Letsoalo
28
3
Benefits and costs of adapting water planning and management to climate change and water demand growth in the Western Cape of South Africa John M. Callaway, Daniël B. Louw, Jabavu C. Nkomo, Molly E. Hellmuth and Debbie A. Sparks
53
4
Indigenous knowledge, institutions and practices for coping with variable climate in the Limpopo basin of Botswana Opha Pauline Dube and Mogodisheng B. M. Sekhwela
71
5
Community development and coping with drought in rural Sudan Balgis Osman-Elasha, Nagmeldin Goutbi, Erika Spanger-Siegfried, Bill Dougherty, Ahmed Hanafi, Sumaya Zakieldeen, El-Amin Sanjak, Hassan A. Atti and Hashim M. Elhassan
90
3232 J&J Climate Change
1/11/07
9:56 am
Page vi
vi Climate Change and Adaptation
6
Climate, malaria and cholera in the Lake Victoria region: Adapting to changing risks Pius Yanda, Shem Wandiga, Richard Kangalawe, Maggie Opondo, Dan Olago, Andrew Githeko, Tim Downs, Robert Kabumbuli, Alfred Opere, Faith Githui, James Kathuri, Lydia Olaka, Eugene Apindi, Michael Marshall, Laban Ogallo, Paul Mugambi, Edward Kirumira, Robinah Nanyunja, Timothy Baguma, Rehema Sigalla and Pius Achola
109
7
Making economic sense of adaptation in upland cereal production systems in The Gambia Momodou Njie, Bernard E. Gomez, Molly E. Hellmuth, John M. Callaway, Bubu P. Jallow and Peter Droogers
131
8
Past, present and future adaptation by rural households of northern Nigeria Daniel D. Dabi, Anthony O. Nyong, Adebowale A. Adepetu and Vincent I. Ihemegbulem
147
9
Using seasonal weather forecasts for adapting food production to climate variability and climate change in Nigeria James Oladipo Adejuwon, Theophilus Odeyemi Odekunle and Mary Omoluke Omotayo
163
10
Adapting dryland and irrigated cereal farming to climate change in Tunisia and Egypt Raoudha Mougou, Ayman Abou-Hadid, Ana Iglesias, Mahmoud Medany, Amel Nafti, Riadh Chetali, Mohsen Mansour and Helmy Eid
181
11
Adapting to drought, zud and climate change in Mongolia’s rangelands Punsalmaa Batima, Bat Bold, Tserendash Sainkhuu and Myagmarjav Bavuu
196
12
Evaluation of adaptation options for the Heihe river basin of China Yongyuan Yin, Zhongming Xu and Aihua Long
211
13
Strategies for managing climate risks in the lower Mekong river basin: A place-based approach Suppakorn Chinvanno, Soulideth Souvannalath, Boontium Lersupavithnapa, Vichien Kerdsuk and Nguyen Thuan
228
3232 J&J Climate Change
1/11/07
9:56 am
Page vii
Contents vii
14
Spillovers and trade-offs of adaptation in the Pantabangan–Carranglan watershed of the Philippines Rodel D. Lasco, Rex Victor O. Cruz, Juan M. Pulhin and Florencia B. Pulhin
247
15
Top–down, bottom–up: Mainstreaming adaptation in Pacific island townships Melchior Mataki, Kanayathu Koshy and Veena Nair
264
16
Adapting to dengue risk in the Caribbean Michael A. Taylor, Anthony Chen, Samuel Rawlins, Charmaine Heslop-Thomas, Dharmaratne Amarakoon, Wilma Bailey, Dave Chadee, Sherine Huntley, Cassandra Rhoden and Roxanne Stennett
279
17
Adaptation to climate trends: Lessons from the Argentine experience Vicente Barros
296
18
Local perspectives on adaptation to climate change: Lessons from Mexico and Argentina Mónica Wehbe, Hallie Eakin, Roberto Seiler, Marta Vinocur, Cristian Ávila, Cecilia Maurutto and Gerardo Sánchez Torres.
315
19
Maize and soybean cultivation in southeastern South America: Adapting to climate change Maria I. Travasso, Graciela O. Magrin, Walter E. Baethgen, José P. Castaño, Gabriel R. Rodriguez, João L. Pires, Agustin Gimenez, Gilberto Cunha and Mauricio Fernandes
332
20
Fishing strategies for managing climate variability and change in the estuarine front of the Río de la Plata Gustavo J. Nagy, Mario Bidegain, Rubén M. Caffera, Walter Norbis, Alvaro Ponce, Valentina Pshennikov and Dimitri N. Severov
353
Index
371
3232 J&J Climate Change
1/11/07
9:56 am
Page viii
List of Figures and Tables
Figures 2.1 2.2 3.1 4.1 4.2 5.1 6.1 6.2 6.3 6.4 6.5 7.1 8.1 8.2 8.3 8.4 8.5 8.6 8.7 10.1 10.2
The increase in conservation areas and the number of reserves 33 in seven southern African countries A decision tree for selecting adaptation strategies for different 35 surrogate species based on their response to climate change The Berg River Spatial Equilibrium Model (BRDSEM) 56 The Limpopo basin area of Botswana and case study sites: 73 Northeast District, Bobirwa Sub-District and Kgatleng District Mean annual rainfall over three stations in the Limpopo basin 74 area of Botswana: Francistown, Northeast District; Bobonong, Bobirwa Sub-District; and Mochudi, Kgatleng District Contour bunds for water harvesting and tree planting in Arbaat 97 Map showing the Lake Victoria highland malaria region and 112 the studied villages Map showing the Lake Victoria cholera region and the studied 113 villages Total death toll due to malaria for Ndolage Hospital, Tanzania, 115 in 2001 Total death toll due to malaria for Rubya Hospital, Tanzania, 116 in 2001 Sources of information on consequence of cholera 122 Analytical framework for assessing economic feasibility of 134 climate change adaptation Communities surveyed in northern Nigeria 150 Number of households classified as very vulnerable, vulnerable 153 and less vulnerable in the hamlets of Zangon Buhari, Dabai and Takwikwi Reasons for food storage 156 Reasons for planting early maturing crop varieties 156 Reasons for planting high yield crop varieties 156 Household vulnerability levels for the five livelihood capitals 157 Farmers’ willingness to change practices to reduce vulnerability 160 to drought Wheat yield and spring precipitation in Tunisia, 1961–2000 184 Simulated changes in irrigated wheat yields and 191 evapotranspiration under different climate conditions for (a) water conservation measures and (b) variations in sowing dates
3232 J&J Climate Change
1/11/07
9:56 am
Page ix
List of Figures and Tables ix
10.3 12.1 12.2 13.1 14.1 14.2 14.3 15.1 15.2 16.1 16.2 16.3 17.1 17.2 17.3 17.4 17.5 17.6 17.7 17.8
17.9 19.1 19.2 19.3 19.4
Simulated changes in rain-fed wheat yields for (a) different rainfall scenarios and (b) variations in sowing dates Framework for multi-criteria evaluation of climate change adaptation options Location of the Heihe river basin Study sites in Lao PDR, Thailand and Vietnam Location of the Pantabangan–Carranglan watershed Land use map of the Pantabangan–Carranglan watershed Summary of effects of adaptation strategies in one sector on other sectors Observed rainfall anomalies for Navua, 1960–2003 Maximum daily rainfall during March and April, 1960–2003 Annual variability of the reported cases of dengue and the rate of change (increase or decrease from previous year) for the Caribbean Monthly variability of the reported dengue cases, rainfall and temperature from 1996 to 2003 in Trinidad and Tobago Schematic of a possible early warning system Linear trends of annual precipitation (mm/year), 1959–2003 Isohyets in mm: 1950–1969 (solid line) and 1980–1999 (dashed line) Percentage change in the rate between standard deviation and mean value in the 1980–1999 period with respect to 1950–1969 Annual precipitation in Chile: La Serena (29.9ºS, 71.2ºW) (left); Puerto Montt (41.4ºS, 73.1ºW) (right) Mean annual streamflow (m3/s) of a representative river of the Cuyo region, Los Patos river, 1900–2000 Mean annual streamflow (m3/s) of rivers of the Comahue region, 1900–2000; note that the Negro river starts at the junction of the Limay and Neuquén rivers Number of events with precipitation greater than 100mm in no more than two days in periods of four years Annual frequency of cases with precipitation over 150mm in less than two days (left); for the same threshold (150mm), the ratio of the annual frequencies between the 1983–2002 and 1959–1978 periods (right) The Plata river estuary Study area and study sites Changes in monthly precipitation (%) projected by HadCM3 under SRES A2 and B2 for 2020, 2050 and 2080 Changes in irrigated maize and soybean yields (%) under different scenarios and CO2 concentrations Changes in rain-fed maize and soybean yields (%) under different scenarios and CO2 concentrations
193 215 218 229 249 249 262 268 269 280 282 292 298 299 300 305 306 307 308 309
310 334 336 338 339
3232 J&J Climate Change
1/11/07
9:56 am
Page x
x Climate Change and Adaptation
19.5
19.6
19.7
19.8 19.9 20.1 20.2 20.3 20.4
Changes in the duration of planting–flowering (P-F) and flowering–maturity (F-M) periods, expressed as mean values for the six sites, for maize and soybean crops under different SRES scenarios and time periods Maize: Yield changes (%) for different planting dates (ac = current, –20 and –40 days) in the six sites under different scenarios (A2 in grey, B2 in black for 2020, 2050 and 2080) and CO2 concentrations Soybean: Yield changes (%) for different planting dates (ac = current, ± 15, 30 days) in the six sites under different scenarios (A2 in grey, B2 in black for 2020, 2050 and 2080) and CO2 concentrations Adaptation measures for maize: Yield change (%) under optimal planting dates/nitrogen rates and supplementary irrigation for the six sites without considering CO2 effects Adaptation measures for soybean: Yield changes (%) under optimal planting dates and supplementary irrigation for the six sites without considering CO2 effects Vulnerability and adaptation framework River flow corridors and fronts of the Río de la Plata Long-term gross income of fishermen (local currency-1999) Unfavourable days for fishing activity on a monthly basis from October 2000 to March 2003, based on a lower threshold of 8m/s wind speed
340
342
343
344 344 355 357 358 362
Tables 2.1
2.2 2.3 2.4 3.1 3.2 3.3 3.4
The area as a percentage conserved in southern African countries in IUCN reserves (IUCN classes I–V), IUCN sustainable resource use areas (IUCN class VI), and other non-IUCN conservation areas The amount of conservation per ecoregion Extent of conservation versus ‘need’ for conservation Relative financial costs compared to the advantages and disadvantages of differing adaptation options Framework for estimating benefits and costs associated with climate change and climate change adaptation Net returns to water and optimal storage capacity of the Berg River Dam Adapting to development pressure and climate change under the existing water allocation system: net returns to water (present value, R billion) Adapting to development pressure and climate change by switching to water markets and adding storage capacity: net returns to water (present value, R billion)
31
32 32 36 57 61 64 66
3232 J&J Climate Change
1/11/07
9:56 am
Page xi
List of Figures and Tables xi
4.1 5.1 5.2 5.3 6.1 6.2 6.3 7.1 7.2 7.3 8.1 8.2 9.1 9.2 9.3 9.4 10.1 10.2 10.3 11.1 12.1 12.2 12.3 12.4 12.5 13.1 13.2 13.3
Examples of government policies relevant to vulnerability and 78 drought impact reduction Adaptation measures in Gireighikh Rural Council, Bara Province, 94 North Kordofan State Adaptation measures in Arbaat, Red Sea State 98 Adaptation measures in El Fashir Rural Council, 102 North Darfur State Percentage responses of how malaria is treated at the 118 household level by people with different levels of education Percentage of reasons/explanations for not treating/boiling 120 drinking water in Chato village Cholera control strategies suggested by stakeholders 123 Average millet yields (kg/ha) and variability (CV) for current 138 and future climates with business-as-usual and adaptive management strategies Climate change damages and costs and benefits of fertilization – 141 Average annual values in millions of US dollars Climate change damages and costs and benefits of irrigation – 142 Average annual values in millions of US dollars Indices and weights for vulnerability assessment in northern 152 Nigeria Respondents’ coping strategies 154 Skill assessment of quint forecast categories 171 Skill assessment of tercile forecast categories 171 Organizational skill performance assessment of the June, July, 172 August and September annual rainfall totals Regional disparities in forecasting skill 173 Average cereal yields during wet and dry years at four sites in 184 the Kairouan region of Tunisia Assumptions for simulations of irrigated wheat production in 189 the Nile Delta region Results of simulations for irrigated wheat in the Nile Delta 190 region (percentage changes are relative to current climate) Evaluation of adaptation options 203 Indicators used to evaluate adaptation options in the 217 Heihe river basin Water availability, water withdrawals and water withdrawal 219 ratio in the Heihe river region, 1991–2000 Water shortage/surplus in Heihe river basin under climate 220 change to 2040 Example of AHP comparison table 224 Overall rank and score of adaptation options for the 224 Heihe region Multiple orders of climate impacts on rain-fed farms in the 230 lower Mekong region Household-level on-farm measures for managing climate risks 232 Household-level off-farm measures for managing climate risks 235
3232 J&J Climate Change
1/11/07
9:56 am
Page xii
xii Climate Change and Adaptation
13.4 13.5 14.1 14.2 14.3 14.4 14.5 14.6 14.7 15.1 16.1 16.2 16.3 17.1 17.2 17.3 17.4 17.5 18.1 18.2 19.1 19.2
20.1 20.2
20.3
Community-level measures for managing climate risks National-level measures for managing climate risks Adaptation options for agriculture and forestry by land use category Options for adapting water resource supply and use in response to climate variations Adaptation strategies of different institutional organizations Cross-sectoral impacts of forest/agriculture sector adaptations on other sectors Cross-sectoral impacts of water sector adaptations on other sectors Cross-sectoral impacts of adaptations by institutions on other sectors Adaptation strategies common to multiple sectors Flood extent, duration and rainfall in five recalled flooding episodes in Navua Distribution of epidemic peaks among ENSO phases, 1980–2001 Socio-economic characteristics of three communities in Western Jamaica and survey sample size Adaptation strategies matrix Cultivated areas Density of rural roads in six provinces Major monthly streamflow anomalies (m3/s) at Corrientes Largest daily discharge anomalies (larger than 3 standard deviations) of the Uruguay river at the Salto gauging station, 1951–2000 Programmes funded by international banks to ameliorate and prevent damages from floods in Argentina Farmers’ socioeconomic characteristics Synthesis of adaptation options Projected changes in mean temperature (ºC) for the warm semester (October–March) according to HadCM3 under SRES A2 and B2 scenarios for 2020, 2050 and 2080 Length (days) of planting–flowering (P–F) and flowering– maturity (F–M) periods for maize at current planting date and 20 and 40 days earlier under SRES A2 scenario for 2020, 2050 and 2080 Freshwater inflow to the Río de la Plata from the River Uruguay and total Fishing activity, capture and income: Comparison between a good year (1988–89), long-term average and model results for a low-typical year (1), a bad year (2) and results with change in fishing behaviour (3) Type II adaptation measures by scale of implementation and objectives
237 239 253 254 255 256 258 259 260 267 283 288 288 299 301 303 303 304 318 325 335 341
357 365
366
3232 J&J Climate Change
1/11/07
9:56 am
Page xiii
Acknowledgements The two volumes Climate Change and Vulnerability and Climate Change and Adaptation are products of Assessments of Impacts and Adaptations to Climate Change (AIACC), a project that benefited from the support and participation of numerous persons and organizations. AIACC was funded by generous grants from the Global Environment Facility, the Canadian International Development Agency, the US Agency for International Development, the US Environmental Protection Agency and the Rockefeller Foundation. The initial concept for the project came from authors of the Third Assessment Report of the Intergovernmental Panel on Climate Change (IPCC) and was championed by Robert Watson, Osvaldo Canziani and James McCarthy, the IPCC chair and IPCC Working Group II co-chairs, respectively, during the Third Assessment Report. The productive relationship between AIACC and IPCC was continued and nurtured by Rajendra Pachauri, the current Chair of the IPCC, and Martin Parry, who joined Dr Canziani as cochair of IPCC Working Group II for the Fourth Assessment Report. The project could not have succeeded without the very capable and dedicated work of the more than 250 investigators who undertook the AIACC case studies, many of whom are authors of chapters of the two books. The project also benefited from the valuable and enthusiastic assistance of the many committee members, advisers, resource persons and reviewers. These include Neil Adger, Ko Barrett, Bonizella Biagini, Ian Burton, Max Campos, Paul Desanker, Alex De Sherbinin, Tom Downing, Kris Ebi, Roland Fuchs, Habiba Gitay, Hideo Harasawa, Mohamed Hassan, Bruce Hewitson, Mike Hulme, Saleemul Huq, Jill Jaeger, Roger Jones, Richard Klein, Mahendra Kumar, Murari Lal, Liza Leclerc, Bo Lim, Xianfu Lu, Jose Marengo, Linda Mearns, Monirul Mirza, Isabelle Niang-Diop, Carlos Nobre, Jean Palutikof, Annand Patwardhan, Martha Perdomo, Roger Pulwarty, Avis Robinson, Cynthia Rosenzweig, Robert Scholes, Ravi Sharma, Hassan Virji, Penny Whetton, Tom Wilbanks and Gary Yohe. Patricia Presiren of the Academy of Sciences of the Developing World (TWAS) and Sara Beresford, Laisha Said-Moshiro, Jyoti Kulkarni and Kathy Landauer of START gave excellent support for the administration and execution of the project. Finally, thanks are owed to Alison Kuznets and Hamish Ironside of Earthscan and to Leona Kanaskie for assistance with copy-editing and production of the books.
3232 J&J Climate Change
1/11/07
9:56 am
Page xiv
3232 J&J Climate Change
1/11/07
9:56 am
Page xv
Foreword Climate change is increasingly recognized as a critical challenge to ecological health, human well-being and future development, as underscored by the award of the Nobel Peace Prize for 2007 to the Intergovernmental Panel on Climate Change (IPCC). The award recognizes the substantial advances in our shared understanding of climate change, its causes, its consequences and its remedies, which have been achieved by more than 20 years of work by the thousands of contributors to the IPCC science assessments, and which draw from the research and analyses of an even larger number of scientists and experts. This work has culminated in the unprecedented impact of the Panel’s most recent report, the Fourth Assessment Report. The Fourth Assessment Report advances our understanding on various aspects of climate change based on new scientific evidence and research. A major contribution in this regard has come from the work promoted under the project Assessments of Impacts and Adaptation to Climate Change (AIACC). The AIACC project was sponsored by the IPCC to fill a major gap in the available knowledge about climate change risks and response options in developing countries that existed at the completion of the Third Assessment Report in 2001. Twenty-four national and regional assessments were executed under the AIACC project in Africa, Asia, Latin America and small island states in the Caribbean, Indian and Pacific Oceans. The two volumes Climate Change and Vulnerability and Climate Change and Adaptation present many of the findings from the AIACC assessments. The findings not only give us a fuller scientific understanding of the specific nature of impacts and viable adaptation strategies in different locations and countries, but have contributed to a much better appreciation of some of the equity dimensions of the problems as well. In simplified terms, the biggest challenge in confronting the negative impacts of climate change lies in the developing world, where people and systems are most vulnerable. Not only are these negative impacts likely to be most serious in the subtropics and tropics, where most developing societies reside, but the capacity to adapt to them is also limited in these regions. An important element in understanding vulnerabilities to climate change is in linking current and projected exposures to climate stresses with other existing stresses and conditions that are responsible for hardship and low levels of economic welfare. Climate change often adds to these existing stresses, increasing the vulnerability of such communities and ecosystems. Unfortunately, limited research is carried out in developing countries on likely impacts and appropriate responses related to climate change. This is where the knowledge provided by the assessments of the AIACC has been particularly valuable. There is considerable interest in interpretation of Article 2 of the United
3232 J&J Climate Change
1/11/07
9:56 am
Page xvi
xvi Climate Change and Adaptation
Nations Framework Convention on Climate Change (UNFCCC), the focus of which defines the ultimate objective of the Convention, namely that of preventing a dangerous level of anthropogenic interference with the climate system. Research on impacts of climate change focusing on specific parts of the world that are highly vulnerable enhances our understanding of what may constitute a dangerous level of anthropogenic interference with the world’s climate system. In the absence of such knowledge, any value judgement defining a dangerous level would apply to, and be determined by, knowledge only from particular regions of the world, primarily the developed nations. Understanding the critical nature of impacts in some of the most vulnerable parts of the world, which are largely in developing countries, will assist our determination of what might constitute a dangerous level of interference with the earth’s climate system. Such knowledge would help appropriately to include and consider those locations which are perhaps much closer to danger than was known earlier. The record and outputs of the AIACC are impressive. The project, funded by the Global Environment Facility and coordinated by the Global Change System for Analysis, Research and Training (START), the Academy of Sciences for the Developing World (TWAS) and the United Nations Environment Programme (UNEP), engaged investigators from more than 150 institutions and 60 countries to execute the assessments. The quality of the assessments is demonstrated by the more than 100 peer-reviewed publications produced, which benefited substantially the IPCC’s Fourth Assessment Report. In view of this success, it is imperative that we build on the experience and achievements of AIACC and develop the next phase of such work to help advance new knowledge for a possible Fifth Assessment Report of the IPCC. While the material contained in the two volumes from AIACC and the substantial amount of knowledge developed through the case studies presented in the following pages are valuable, the need for further work is enormous. There remain many countries in the developing world where very little is known about the nature and extent of the impacts of climate change, and these gaps would not permit the development of plans and programmes to address climate change risks or to put in place response measures that would help communities and ecosystems to adapt to the impacts of climate change. These clearly would get much more serious with time unless suitable mitigation measures are taken in hand with a sense of urgency. Yet, even with the most ambitious mitigation actions, the inertia of the system will ensure that the impacts of climate change will continue for centuries, if not beyond a millennium. Knowledge of impacts and the manner in which they would grow over time is therefore critical to the development of capacity and measures for adaptation to climate change. The work of the AIACC provides an extremely important platform to take such steps, but there is yet very far to go to meet the challenges ahead. It is hoped that the material contained in this volume is just the start of a process that must expand and continue in the future. R. K. Pachauri Director General, The Energy and Resources Institute (TERI) and Chairman, Intergovernmental Panel on Climate Change (IPCC)
3232 J&J Climate Change
1/11/07
9:56 am
Page 1
1
A Stitch in Time: General Lessons from Specific Cases Neil Leary, James Adejuwon, Vicente Barros, Punsalmaa Batima, Bonizella Biagini, Ian Burton, Suppakorn Chinvanno, Rex Cruz, Daniel Dabi, Alain de Comarmond, Bill Dougherty, Pauline Dube, Andrew Githeko, Ayman Abou Hadid, Molly Hellmuth, Richard Kangalawe, Jyoti Kulkarni, Mahendra Kumar, Rodel Lasco, Melchior Mataki, Mahmoud Medany, Mansour Mohsen, Gustavo Nagy, Momodou Njie, Jabavu Nkomo, Anthony Nyong, Balgis Osman-Elasha, El-Amin Sanjak, Roberto Seiler, Michael Taylor, Maria Travasso, Graham von Maltitz, Shem Wandiga and Mónica Wehbe
Introduction We can adapt to climate change and limit the harm, or we can fail to adapt and risk much more severe consequences. How we respond to this challenge will shape the future in important ways. The climate is already hazardous; indeed it always has been so. Variations and extremes of climate disrupt production of food and supplies of water, reduce incomes, damage homes and property, impact health and even take lives. Humans, in an unintended revenge, are getting back at the climate by adding to heat-trapping gases in the Earth’s atmosphere that are changing the climate. But the changes are amplifying the hazards. And we cannot in short order stop this. The physical and social processes of climate change have a momentum that will continue for decades and well beyond. This undeniable momentum does not imply, however, that efforts to mitigate climate change, meaning to reduce or capture the emissions of greenhouse gases that drive climate change, are wasted. Nor is a call for adaptation a fatalistic surrender to this truth. The magnitude and pace of climate change will determine the severity of the stresses to which the world will be exposed. Slowing the pace of human caused climate change, with the aim of ultimately
3232 J&J Climate Change
1/11/07
9:56 am
Page 2
2 Climate Change and Adaptation
stopping it, will enable current and future generations to better cope with and adapt to the resulting hazards, thereby reducing the damages and danger. Mitigating climate change is necessary. Adapting to climate change is necessary too. The challenges are substantial, particularly in the developing world. Developing countries have a high dependence on climate-sensitive natural resource sectors for livelihoods and incomes, and the changes in climate that are projected for the tropics and sub-tropics, where most developing countries are found, are generally adverse for agriculture (IPCC, 2001 and 2007a). Furthermore, the means and capacity in developing countries to adapt to changes in climate are scarce due to low levels of human and economic development and high rates of poverty. These conditions combine to create a state of high vulnerability to climate change in much of the developing world. To better understand who and what are vulnerable to climate change, and to examine adaptation strategies, a group of case studies was undertaken as part of an international project, Assessments of Impacts and Adaptations to Climate Change (AIACC). The studies span Africa, Asia, Central and South America, and islands of the Caribbean, Indian and Pacific Oceans. They include assessments of agriculture, rural livelihoods, food security, water resources, coastal zones, human health and biodiversity conservation. Results from the studies about the nature, causes and distribution of climate change vulnerability are presented in a companion to this volume (Leary et al, 2008). In this volume, we collect together papers from the AIACC studies that explore the challenge of adaptation. Comparison and synthesis of our individual contributions have yielded nine general lessons about adaptation, as well as many more lessons that are specific to particular places and contexts. The general lessons, formulated as recommendations, are as follows: (1) adapt now, (2) create conditions to enable adaptation, (3) integrate adaptation with development, (4) increase awareness and knowledge, (5) strengthen institutions, (6) protect natural resources, (7) provide financial assistance, (8) involve those at risk, and (9) use place-specific strategies. The lessons are briefly outlined below, followed by a more detailed examination of their nuances and supporting evidence from the case studies.
The Nine Adaptation Lessons Adapt now! The time-honoured proverb ‘a stitch in time saves nine’ means that immediate action to repair damage (to your clothing in the original context) can avoid the necessity to do much more later on. The expression captures one of the main findings of the AIACC’s programme of studies. It can simply be stated as the injunction to adapt now. Climatic variations and extremes cause substantial damage to households, communities, natural resources and economies. In many places the damage is increasing, giving evidence of an adaptation deficit, meaning that practices in
3232 J&J Climate Change
1/11/07
9:56 am
Page 3
A Stitch in Time: General Lessons from Specific Cases 3
use to manage climate hazards are falling short of what can be done (Burton, 2004). We find evidence in all our case study sites of an adaptation deficit that climate change threatens to widen. Acting now to narrow the deficit can yield immediate benefits. It will also serve as a useful, even essential, first step in a longer-term process of adapting to a changing climate. Failure to tackle adaptation vigorously now is likely to mean that many more than nine stitches will be required in the future.
Create conditions to enable adaptation In contrast to reducing emissions of the greenhouse gases that drive climate change, a policy that, in the parlance of economists, generates benefits that are substantially external, adaptation generates benefits that are largely internal. This means that the individuals, organizations, communities and countries that take action to adapt will capture for themselves most of the benefits of their actions, creating a strong incentive to adapt. This explains why we can see a wide range of practices being used to manage and reduce climate risks. But why then do we nonetheless observe adaptation deficits? Why doesn’t self interest motivate people to do more to protect themselves from climate hazards? Our case studies identify numerous obstacles that impede adaptation. Common obstacles include competing priorities that place demands on scarce resources, poverty that limits capacity to adapt, lack of knowledge, weak institutions, degraded natural resources, inadequate infrastructure, insufficient financial resources, distorted incentives and poor governance. Obstacles such as these severely constrain what people can do. Intervention by public sector entities, at levels from the local community to the provincial, national and international, can create conditions that better enable people to surmount the obstacles and take actions to help themselves. Enabling the process of adaptation is the most important adaptation that the public sector can make. Specific interventions to enable adaptation are addressed by some of the other lessons that follow.
Integrate adaptation with development The goals and methods of climate change adaptation and development are strongly complementary. The impacts of current climate hazards and projected climate change threaten to undermine development achievements and stall progress towards important goals. Adaptation can reduce these threats. In turn, development, if appropriately planned, can help to enable climate change adaptation. Integrating adaptation with development planning and actions can exploit the complementarities to advance both adaptation and development goals. To be effective, integration needs to engage ministries that are responsible for development, finance, economic sectors, land and water management, and the provision of public health and other services. It is in agencies such as these that key decisions are taken about the allocation of financial and other resources. And it is within these agencies and among their stakeholders that much of the sector-specific expertise that must be engaged resides.
3232 J&J Climate Change
1/11/07
9:56 am
Page 4
4 Climate Change and Adaptation
Increase awareness and knowledge Nearly all the case studies highlighted knowledge as a critical constraint on adaptation and rank efforts to increase and communicate knowledge as a high adaptation priority. Stakeholders in many of the study areas complained of inadequate or lack of access to information about climate history, projections of future climate change and potential impacts, estimates of climate risks, causes of vulnerability, technologies and measures for managing climate risks, and know-how for implementing new technologies. Uncertainty about the future and about the effectiveness and costs of adaptation options are common obstacles to action. Examination of these and other information problems in the case studies demonstrates the need for programmes to help advance, communicate, interpret and apply knowledge for managing climate risks.
Strengthen institutions Institutions are found to play important roles in enabling adaptation. Local institutions, including community organizations, farmer cooperatives, trade associations, local government agencies, informal associations, kinship networks and traditional institutions, serve functions in communities that help to limit, hedge and spread risks. They do this by sharing knowledge, human and animal labour, equipment and food reserves; mobilizing local resources for community projects and public works; regulating use of land and water; and providing education, marketing, credit, insurance and other services. Provincial, national and international institutions aid by providing extension services, training, improved technologies, public health services, infrastructure to store and distribute water, credit, insurance, financial assistance, disaster relief, scientific information, market forecasts, weather forecasts, and other goods and services. In many of our case study sites, key functions for managing risks are absent or are inadequate due to weak institutions that are poorly resourced, lacking in human capacity, overloaded with multiple responsibilities and overwhelmed by the demands of the communities that they serve. Strengthening institutions to fill strategic functions in support of adaptation is needed. In some instances, traditional institutions that have been diminished in role by socioeconomic changes and government policies provide a remnant framework that could be revitalized to facilitate adaptation and the management of climate risks.
Protect natural resources Developing countries typically are dependent on climate sensitive natural resources for a high proportion of their livelihoods, economic activities and national incomes. Too often these resources are in a degraded state from a combination of pressures caused by human use and climatic and environmental variation and change. Their degraded state makes these resources, and the people who are dependent on them, highly vulnerable to damages from climate change. Rehabilitating and protecting natural resources such as farm lands,
3232 J&J Climate Change
1/11/07
9:56 am
Page 5
A Stitch in Time: General Lessons from Specific Cases 5
grazing lands, forests, watersheds, wetlands, fisheries and biodiversity are a central focus of adaptation strategies in places as varied as the African Sahel, southern Africa, central Asia, southeast Asia, and south-eastern South America. Progress in many of these settings will require changes in incentives, reforms of tenure to land, water and natural products, education, training, and more vigorous enforcement of regulations. These, in turn, are dependent on strong institutions and access to financial resources.
Provide financial assistance Lack of financial resources is commonly cited as a major obstacle to adaptation. The constraint is particularly binding on the poor and the very poor, who typically are among the most vulnerable to climate change. Poor households and small-scale farmers and enterprise owners obtain finance through community and informal networks to recover from losses and make investments that reduce risks. But more adaptation could take place in impoverished localities and regions with greater financial assistance from provincial and national governments and international sources. Innovative ideas are needed to engage the private sector in financing adaptation. Internationally, some financial assistance is being provided and acts as a catalyst for raising awareness, building capacity and advancing understanding of risks and response options. But the magnitude of financial needs for adaptation is much greater than the current level of assistance. Increased financial assistance over and above normal development assistance is needed. Ultimately, however, financing will need to come from multiple sources, including those internal to developing countries.
Involve those at risk Involving persons at risk in the process of adaptation, the intended beneficiaries, can increase the effectiveness of adaptation to climate change. Many of our case studies involved at-risk groups in assessment activities. The experiences demonstrate the potential of participatory approaches to adaptation for focusing attention on risks that are priorities to the vulnerable, learning from risk management practices currently in use, identifying opportunities and obstacles, applying evaluation criteria that are relevant and credible to at-risk groups, and drawing on local knowledge and expertise for selecting and designing appropriate strategies, garnering support and mobilizing local resources to assist with implementation. A common result of involving those at risk is that it forces climate risks to be examined in context with other problems and gives emphasis to solutions that can be combined to attain multiple objectives.
Use place-specific strategies Adaptation is place-based and requires place-specific strategies. This fact has long been recognized in the climate impacts research literature. The general lessons outlined above conceal the much richer content of the case studies and risk presenting an oversimplified story. The ninth lesson is that there are many
3232 J&J Climate Change
1/11/07
9:56 am
Page 6
6 Climate Change and Adaptation
more lessons and that many are specific to particular contexts of particular places. For example, in the lower Mekong river basin, rice farmers face similar risks from floods but rely on different strategies for managing the risks that reflect differences in the level of economic development of their surrounding community, strength of community institutions, locally available natural resources and seasonal rain patterns (Chinvanno et al, Chapter 13). Pastoralists in Mongolia, Sudan and Botswana share some strategies for coping with drought that have general characteristics in common, but there are significant differences too that derive from different traditions, resources and climates (Batima et al, Chapter 11; Dube et al, Chapter 4; Osman-Elasha et al, Chapter 5). People living in the Caribbean and the highlands surrounding Lake Victoria both face health risks from mosquito-borne diseases that vary with the climate, but differences in public health infrastructure and access to health care contribute to differences in responses to the diseases (Taylor et al, Chapter 16; Yanda et al, Chapter 6). General lessons can be applied in these different settings to help guide adaptive strategies, but details of the local context will determine the specific approaches and measures that will be most effective in each place.
Adaptation Now and in the Future What is adaptation? The Intergovernmental Panel on Climate Change (IPCC) defines adaptation as adjustments in ecological, social or economic systems in response to actual or expected climatic stimuli and their effects (Smit et al, 2001). It includes adjustments to moderate harm from, or to benefit from, current climate variability as well as anticipated climate change. Adaptation can be a specific action, such as a farmer switching from one crop variety to another that is better suited to anticipated conditions. It can be a systemic change such as diversifying rural livelihoods as a hedge against risks from variability and extremes. It can be an institutional reform such as revising ownership and user rights for land and water to create incentives for better resource management. Adaptation is also a process. The process of adaptation includes learning about risks, evaluating response options, creating the conditions that enable adaptation, mobilizing resources, implementing adaptations and revising choices with new learning. We mean all these things by adaptation. But the conception of adaptation as a process is often the most important for formulating public interventions that will have lasting benefits.
Is adaptation new? Adaptation to climate is not new. People, property, economic activities and environmental resources have always been at risk from climate and people have continually sought ways of adapting, sometimes successfully and sometimes
3232 J&J Climate Change
1/11/07
9:56 am
Page 7
A Stitch in Time: General Lessons from Specific Cases 7
not. The long history of adapting to variations and extremes of climate includes construction of water reservoirs, irrigation, crop diversification, disaster management, insurance and even, on a limited basis, recent measures to adapt to climate change (Adger et al, 2007). The AIACC case studies document a variety of adaptive practices in use that have reduced vulnerability to climate hazards. In most cases these have been adopted in response to multiple sources of risk and only rarely to climate risk alone. One strategy commonly in use is to increase the capacity to bear losses by accumulating food surpluses, livestock, financial assets and other assets. Risks are hedged by diversifying crops, income sources, food sources and locations of production activities. Exposures to hazards have been reduced by relocating, either temporarily or permanently. Variability of production and incomes derived from natural resources have been reduced by restoring degraded lands, using drought-resistant seed varieties, harvesting rainfall, adopting irrigation and using seasonal forecasts to optimize farm management. Prevention of climate impacts through flood control, building standards and early warning systems is practised. Risk spreading is accomplished through kinship networks, pooled community funds, insurance and disaster relief. In many cases the capacity to adapt is increased through public sector assistance such as extension services, education, community development projects and access to subsidized credit.
Is adapting to climate change different? Is adapting to climate change different? Yes and no. People have always faced an uncertain future when coping with and adapting to climate. Human societies have long coped with floods, droughts and other climate hazards without knowing when the next event would occur, how big it would be or how long it would last. Past experience provided a basis, albeit an imperfect one, for approximating the frequencies of events of different magnitudes and the likely range of conditions that might be encountered in the coming season, year or decade. But climate change means that past performance of the climate is becoming a less reliable predictor of future performance. The frequency, variability, seasonal patterns and characteristics of climate events and phenomena will change. Phenomena once alien to a region could become regular features of its climate (for example, extra-tropical storm tracks are projected to move poleward, IPCC, 2007b). An important consequence of climate change for adaptation is that the future climate will be less familiar and in key respects more uncertain. However, some climate parameters will change with predictable trends as a result of human-driven climate change. Globally averaged surface temperatures are projected to rise by 1.1–6.4°C by the end of the 21st century relative to 1980–1999 temperatures (IPCC, 2007b). Annual and monthly average temperatures can be expected, with a very high degree of confidence, to increase virtually everywhere. Changes in average precipitation are also projected but vary from decreases to increases depending on location and season; while
3232 J&J Climate Change
1/11/07
9:56 am
Page 8
8 Climate Change and Adaptation
confidence in predictions of precipitation trends is less than for temperature trends, some broad patterns do seem to be robust across climate model projections. For example, precipitation is very likely to increase in highlatitudes while decreases are thought likely in most subtropical land areas. Likely trends for extreme weather include more frequent hot days, heat-waves and heavy precipitation events, more intense tropical cyclones with greater peak wind speeds and heavier precipitation, and increased summer drying and drought risk in continental interiors. The projected trends in temperature, precipitation and extremes will push future climate variations and extremes beyond the bounds of what people and places have been exposed to and had to cope with in the past. The implication is that current practices, processes, systems and infrastructure that are more or less adapted to the present climate will become increasingly inappropriate and maladapted as the climate changes. Fine tuning current strategies to reduce risks from historically observed climate hazards will not be sufficient in this dynamically changing environment. More fundamental adjustments will be needed. This will require recognizing what changes are happening, predicting the range of likely future changes, understanding the vulnerabilities and potential impacts, identifying appropriate adjustments, and mobilizing the resources and will to implement them. The experience of Argentina in the last decades of the 20th century is instructive of some of the challenges (Barros, Chapter 17). A number of climate trends are documented that began in the 1960s and 1970s. These include large increases in mean annual precipitation in southern South America east of the Andes Cordillera; increased flows and flood frequencies of the major rivers of the region, the Parana, Paraguay and Uruguay rivers; more frequent heavy rainfall events in central and eastern Argentina resulting in localized flooding; more frequent sudestadas, which bring winds from the southeast that cause high tides and flooding in Buenos Aires; and, in western Argentina, declining rainfall and stream flows. The speed and effectiveness of adaptive responses to these trends varied. In each case there was a lag between the onset of the climate trend and recognition by affected persons, government agencies and the public. The lag varied depending on the perception of impacts, their magnitude, natural variability of the climate phenomenon, adequacy of observational data, and the difficulty of detecting trends in low frequency events. The quickest response came in the case of increased rainfall east of the Andes but west of the traditional crop farming areas. Farmers recognized and acted on the new opportunity created by the greater rainfall, as well as by high soybean prices in international markets, to profitably cultivate lands that were previously too dry for crop farming. This resulted in significant westward expansion of crop farming, particularly of soybeans, but with roughly a ten year lag. Less quick to act was the government, which failed to provide road and other infrastructure to support the westward expansion of crop farming. Usually emphasis is placed on uncertainty of predicted climate change as a barrier to adaptation. Less appreciated is the barrier created by uncertainty in detecting changes that are already underway and likely to continue. The exam-
3232 J&J Climate Change
1/11/07
9:56 am
Page 9
A Stitch in Time: General Lessons from Specific Cases 9
ples from Argentina demonstrate how delays in recognition and limited awareness of climate trends by key stakeholders delayed adaptive responses. They also suggest that those who have a direct self-interest in adapting may be more astute and quicker to respond. Biodiversity conservation in southern Africa is an example where climate change will require a fundamental change in approach from current risk management (von Maltitz et al, Chapter 2). In 50 years’ time, up to half of South Africa will have a climate that is not currently found in that country. With the changes, many species will need to move across the landscape to track climates that are suitable to their requirements. It will no longer be adequate to protect species where they are currently found – conservationists will have to aim for a moving target. Some species will be able to tolerate the new climate in their current locations (persisters); some will thrive in new climate niches not currently available and expand their ranges (range expanders); some will no longer be viable in part or all of their current range and must disperse to new areas (partial and obligatory dispersers); and some will find no areas with suitable climate and will go extinct from the region (no-hopers). Modelling of climate change impacts on Proteaceae, a surrogate for the highly diverse fynbos vegetation of South Africa, yields estimates that in 50 years approximately 60 per cent of species would be persisters, 30 per cent partial or obligatory dispersers, and 10 per cent no-hopers. The no-hopers can be preserved only by ex situ conservation methods. Migration of the obligatory and partial dispersers over a mixed use, fragmented landscape to track a changing climate is not assured. And successful dispersal 50 years from now does not assure long-term survival, as the climate will continue to change. Multiple strategies will be needed to facilitate migration and minimize species loss. Adding to and reconfiguring land reserves is one element that will be needed, but it is a costly approach and the lands needing protection will change through time. New and more aggressive strategies will be needed to make the landscape more permeable and biodiversity friendly, including private and communal lands that are not in formal reserves. The terminology from the field of biodiversity conservation – obligatory dispersers and no-hopers, for example – is stark. But are there analogous cases in other contexts? Will climate change make inhabitants of some small islands, coastal areas and arid zones partial or obligatory dispersers? Is the hope for survival of some small island nation states and their cultures dependent on ex situ conservation? Do some livelihoods have no hope of persistence in a changing, more hazardous climate? The methods of adaptation to climate change will often be similar to, borrow heavily from and build on current adaptation practice. But as these questions suggest, the challenges and stakes are getting higher.
Is current adaptation enough? Adaptation to climate variation is a regular feature of our lives and, broadly speaking, we are adapted to cope with a wide range of climatic conditions.
3232 J&J Climate Change
1/11/07
9:56 am
Page 10
10 Climate Change and Adaptation
Indicators of successful adaptation include the increase in world food production in pace with population growth, increased life expectancy and decreased weather-related deaths in developed countries (Schneider et al, 2007; McMichael et al, 2001). But variations and extremes do regularly exceed coping ranges, too often with devastating effect. Weather-related hazards such as tropical cyclones, floods and droughts have caused more than one million deaths in the past 20 years, the overwhelming majority of which occurred in developing countries (Pelling et al, 2004). Individual events can cause billions of dollars in damages, and economic and insured losses from natural catastrophes increased more than 6-fold and 24-fold respectively since the 1960s (Munich Re, 2005). While climate impacts can never be reduced to zero, the heavy and rising toll of weather-related disasters and the burden of less severe variations indicate that we are not as well adapted as we might or should be. All of the AIACC case studies give evidence of an adaptation deficit and identify measures that could reduce current losses. For example, greater reforestation efforts and enforcement of forest protection laws would reduce soil erosion and flood risks in the Pantabangan–Carranglan watershed of the Philippines (Lasco et al, Chapter 14). In the Berg river basin of South Africa, allowing greater flexibility for water transfers or water marketing would enable water to be allocated more efficiently during periods of drought (Callaway et al, Chapter 3). A variety of underutilized options for reducing drought and flood risks are available to farmers in Argentina, Botswana, Cambodia, Egypt, Lao PDR, Mexico, Nigeria, Sudan, Thailand and Tunisia (Barros, Chapter 17; Dube et al, Chapter 4; Chinvanno et al, Chapter 13; Mougou et al, Chapter 10; Wehbe et al, Chapter 18; Dabi et al, Chapter 8; Osman-Elasha et al, Chapter 5). In Jamaica, management of dengue fever risks are largely reactive and could be improved by proactive steps for education, elimination of breeding sites and early warnings (Taylor et al, Chapter 16). Building sturdier houses raised above ground level, improved control of river siltation and more regular dredging of rivers would reduce flood losses in coastal towns of Fiji (Mataki et al, Chapter 15). The current deficit in adaptation makes it imperative to adapt now. Doing so would have immediate benefits in reduced weather-related impacts and increased human welfare. The need to adapt is made more urgent by climate change, which is now upon us and is widening the deficit. Adapting to current climate is an essential step towards adapting to future climates.
What are the obstacles to adaptation? People may not adapt, or adapt incompletely, for a variety of reasons. Climate may be perceived, rightly or wrongly depending on the context, to pose little risk relative to other hazards and therefore be given low priority. Knowledge of options to reduce climate risks or the means to implement them may be lacking. Or their expected costs may exceed the expected benefits. The means or capacity to adapt may be lacking. Uncertainty about the future may make it difficult to know what to do or when to do it. Irreversible consequences of some actions may delay choices until some of the uncertainty is resolved.
3232 J&J Climate Change
1/11/07
9:56 am
Page 11
A Stitch in Time: General Lessons from Specific Cases 11
Incentives may be distorted in ways that discourage choices that reduce risks, or even encourage risk-taking behaviour. Sometimes the action of others, or inaction of others, can be an obstacle. Some may believe that reducing their own risk is the responsibility of others. All these are found to impede adaptation in one or more of the case studies. The AIACC studies are all set in developing countries and most focus on places and households that are poor. Poverty, in human development as well as economic terms, is a major obstacle to adaptation in these study areas. Indicative of the constraint imposed by poverty is the high proportion of households in East Africa that do not use insecticide treated bed nets as a prevention against malaria, despite their effectiveness and seemingly low cost (Yanda et al, Chapter 6). The case studies of northern Nigeria (Dabi et al, Chapter 8) and the states of North Kordafan, North Darfur and Red Sea in Sudan (Osman-Elasha et al, Chapter 5) are illustrative of the constraints faced by poor rural households. Households in these study areas, located in the dry and drought prone SudanoSahel zone, typically have low capacity to adapt because of very limited financial, natural, physical, human and social capital. They have little or no cash income, financial savings or access to credit with which to purchase seed, fertilizer, equipment, livestock or food. The lands from which they derive their livelihoods have poor fertility, are highly erodable and are degraded from heavy use, clearing of vegetation, declines in average precipitation and increasing frequency of drought. Physical infrastructure for transportation, communication, water supply, sanitation, and other services are lacking. People have knowledge of many traditional practices for coping with drought and other stresses, but often have little knowledge of new or alternate methods due to poor access to education, training or extension services. Kinship networks provide a safety net for food and other necessities in times of crisis, but sometimes a crisis such as drought or violence will strike many members of a network simultaneously. Local institutions for providing community services are generally weak, governance at provincial and national levels is ineffective, and violence and conflict have heightened vulnerability – with devastating impact in Darfur. Lack of awareness, information and knowledge is a constraint on adaptation in all of the case studies. In Argentina, as noted previously, lags in recognition of climate trends that had begun in the 1960s and 1970s resulted in delayed and incomplete adaptive responses (Barros, Chapter 17). Tunisian farmers are reluctant to change from inherited traditional practices because they lack knowledge and education to evaluate and implement new methods (Mougou et al, Chapter 10). Similarly, in Tamaulipas, Mexico, ejidatarios and smallholder farmers lack know-how for adopting irrigation (Wehbe et al, Chapter 18). In Mongolia, herders voiced a strong need for education and training in methods for improving the condition and productivity of their rangelands and livestock (Batima et al, Chapter 11). Participants in the artisanal fishery of the La Plata estuary need better information about the effects of variations in climate on movements of fish stocks, forecasts of fishing conditions, and fishing methods and technologies for managing variability in the fishery (Nagy et al, Chapter 20).
3232 J&J Climate Change
1/11/07
9:56 am
Page 12
12 Climate Change and Adaptation
Seasonal weather forecasts and early warning systems are frequently suggested as useful for informing the management of climate risks. But, as shown by Adejuwon et al (Chapter 9), they require an effective knowledge network to deliver their promised benefits. Seasonal forecasts are made for West Africa and Nigeria, but few farmers use them. Their reliability is low, the variables forecast are not ones that are most relevant to farmers’ decisions, and the spatial resolution of the forecasts is coarse compared to what farmers need. The forecasts are poorly disseminated, are delivered only shortly in advance of the forecast period, do not regularly reach smallholder farmers and are in forms that are not readily understood by farmers. A number of recommendations are made by Adejuwon et al to improve this knowledge network and support an adaptation process that would provide farmers with more useful forecasts and the knowledge and skills to apply them. Success will be dependent on cooperation and coordination across the regional and national meteorological agencies, agricultural extension agency, local government units, and farmers’ associations, which may require changes in responsibilities, accountability and incentives. Scarce and degraded natural resources contribute to vulnerability and detract from the capacity to adapt in many of the case studies. Insufficient water supplies, and poor quality of existing supplies, prevent Tunisian farmers from expanding irrigation (Mougou et al, Chapter 10). In some instances, treatment of a resource as an open access commons has contributed to its degradation and created disincentives for adaptations to protect it. Following the transition to a market economy in Mongolia, livestock ownership was privatized while pastureland remained state owned and access largely unrestricted (Batima et al, Chapter 11). This has contributed to overstocking of animals, diminished seasonal migration of herds, and lack of investment in land improvements. This situation contrasts with earlier periods during which state collectives, and traditional family groups before that, controlled access to communal pastures. Social capital, an important resource for coping with risk, has been eroded in many places by social and economic changes and by government policies. In the Limpopo Basin of eastern Botswana, the Kgotla, or traditional institution for local decision making and administration of justice, played a central role in adapting the local community to climate variability by regulating resource use and maintaining and disseminating traditional knowledge for the use of veld products (Dube et al, Chapter 4). The mafisa system of lending cattle to poorer family members, the marriage institution and family-based user rights to land provided social security and income security that limited risks from climate extremes and other crises. These institutions were weakened during the 20th century, with the result that communities were alienated from decision making about local resource use, income poverty and capability poverty were deepened, and dependence on government interventions increased. The loss of social capital has reduced the capacity of communities to adapt and amplified their vulnerability to climate hazards. Governance can either constrain or enable adaptation. Financial constraints, already mentioned for households, is one factor that prevents
3232 J&J Climate Change
1/11/07
9:56 am
Page 13
A Stitch in Time: General Lessons from Specific Cases 13
governance from playing a more positive role. Government agencies are often poorly resourced relative to the demands placed on them. Other impediments to government support for adaptation include lack of awareness, knowledge and staff with relevant skills, ineffective administration, poor coordination across departments, inadequate accountability and corruption. Also important is the fact that persons who are most vulnerable to climate risks are often socially and politically marginalized and therefore unable to influence governments to act in their interest.
Climate and Development What are the impacts of climate on development? Billions of people in more than 100 countries are exposed to natural disaster risk, including weather-related disasters that take lives, damage infrastructure and natural resources, and disrupt economic activities (Pelling et al, 2004). Economic losses from natural catastrophes over the period 1996–2005 are estimated to be US$575 billion, with record losses of US$210 billion reported in 2005 (Munich Re, 2005). In the aftermath of disasters, human development in the impacted communities and wider region is set back and can take years to recover from the loss of housing, businesses, roads, water systems, schools, hospitals, farm fields and livestock. Events such as Hurricanes Mitchell, George and Katrina can cause economic losses that represent a significant percentage of national or regional income, and repairing the damage diverts scarce capital from new development projects. Recurrent climate anomalies that do not rise to the level of natural disasters also adversely affect supplies of food and water, incomes, livelihoods and health, reduce resilience to future shocks by depleting assets for coping, and place a drag on economic development. The projected changes in climate will have wide-ranging impacts on development. At risk are the productivity of agricultural lands, natural ecosystems and the livelihoods that are dependent on them. Also at risk are water supplies, human health and populations that inhabit low-lying coasts, floodplains, steep slopes and other exposed locations (McCarthy et al, 2001). The AIACC case studies illustrate these and other climate risks at national and local scales in a variety of developing country contexts. Not all impacts will be negative. For example, a number of studies find that climate change and higher concentrations of carbon dioxide in the atmosphere are likely to increase yields of important crops in parts of South America (Travasso et al, Chapter 19) and West Africa (Njie et al, Chapter 7; Adejuwon et al, Chapter 9). But most studies find that impacts will be predominantly negative in developing regions of the world (IPCC, 2001). Current climate hazards and the impacts of projected climate change threaten human development (African Development Bank et al, 2003). Climate is linked to all the Millennium Development Goals, but is most directly relevant to the goals to eradicate extreme poverty and hunger, reduce child mortality, combat disease and ensure environmental sustainability (MartinHurtado et al, 2002). Agriculture, which is highly sensitive to climate and
3232 J&J Climate Change
1/11/07
9:56 am
Page 14
14 Climate Change and Adaptation
which is projected to be negatively impacted by climate change in much of the tropics and sub-tropics, is the direct or indirect source of livelihood for about two-thirds of the population of developing countries and is a substantial contributor to their national incomes. About 70 per cent of the world’s poor live in rural areas. Progress on all the Millennium Development Goals will be dependent on progress in agricultural development and rural development. And management of climate hazards and climate change impacts in the agriculture sector and rural communities will be critical for success.
How does development affect vulnerability to climate? There is a clear link between development level and vulnerability to climate and other natural hazards. Disaster risk, measured in mortality from natural hazards, is significantly lower in high income countries than in medium and low income countries. Countries classified as having high human development represent 15 per cent of the population that was exposed to natural disasters in 1980–2000 but account for only 1.8 per cent of the deaths (Pelling et al, 2004). In comparison, countries with low human development represent 11 per cent of the exposed population but account for 53 per cent of the recorded deaths. The association of poverty and low levels of development with high levels of vulnerability are borne out in the AIACC studies. Failures of development to raise people out of poverty causes people to occupy highly marginal lands for farming and grazing, settle in areas susceptible to floods and mudslides, and live with precarious access to water, healthcare and other services. These conditions contribute to the high degree of vulnerability found among the rural poor of Botswana, Nigeria, Sudan, Thailand, Lao PDR, Vietnam, the Philippines, Argentina and Mexico. Squatter communities in Jamaica and the Philippines are more vulnerable than other communities because of lack of infrastructure, access to basic services and social institutions to support collective efforts for reducing risks (Taylor et al, Chapter 16; Lasco et al, Chapter 14). Although much of the world continues to live in poverty and at high risk from hunger and disease, human development has greatly reduced vulnerability to climate-driven risks by increasing agricultural productivity, food production and trade, water storage and distribution systems, housing quality, transportation and communication networks, healthcare, education and wealth. The Millennium Development Goals have set a challenge to expand the benefits of development to include those who continue to live in deep poverty. Moving forward, development that is focused on the poor can reduce vulnerability to climate and other stresses by improving the conditions and capacities of poor households, communities and countries so that they are more resilient to shocks and more capable of responding and adapting. If based on sound principles of resource management, development can improve resource-based rural livelihoods so that they are less sensitive to climate variations and more sustainable. Development can, however, exacerbate pressures that add to vulnerability. Past practice has given scant consideration to climate risks in planning development projects, resulting in greater vulnerability than otherwise could have
3232 J&J Climate Change
1/11/07
9:56 am
Page 15
A Stitch in Time: General Lessons from Specific Cases 15
been achieved, and even increasing vulnerability in some instances through maladaptive choices (Burton and van Aalst, 2004). The unevenly distributed benefits of development can also exacerbate vulnerability. Trade liberalization has brought general increases in economic activity, lower prices and greater overall wealth, but all do not share equally in the benefits and some have suffered harm. Smallholder farmers and livestock raisers in Argentina and Mexico have struggled to compete as output prices fell relative to the costs of inputs, making them more vulnerable to climate shocks (Wehbe et al, Chapter 18). Falling rice prices from greater productivity in Asia and liberalized trade caused rice farming to be abandoned in Navua, Fiji (Mataki et al, Chapter 15). The resulting loss of incomes and lack of maintenance of abandoned irrigation channels have raised vulnerability of inhabitants of the township to flood hazards. Development in the Heihe river basin of China has brought greater livelihood opportunities and incomes. But development has also increased water demand in this arid basin to the point where water withdrawals are 80 to 120 per cent of average annual flows and conflicts have arisen between competing water users (Yin et al, Chapter 12). Social and economic changes have driven rural-to-urban migrations, often concentrating poorer migrants in settlements that are prone to flooding, as is happening on the outskirts of metropolitan Buenos Aires (Barros, Chapter 17). Increasing market orientation, movements of population and government policies have weakened community institutions and diminished the use of collective strategies for managing climate risks in places such as Botswana (Dube et al, Chapter 4), countries of the lower Mekong (Chinvanno et al, Chapter 13), Mongolia (Batima et al, Chapter 11) and Sudan (Osman-Elasha et al, Chapter 5). Development projects intended to benefit one group can have spillover effects that harm others, as is the case with the Pantabangan dam in the Philippines (Lasco et al, Chapter 14) and the Khor Arbaat dam in Sudan (Osman-Elasha et al, Chapter 5).
Integrating adaptation with development Sometimes climate change adaptation is seen as competing with the human and economic development needs of the world’s poor. Development needs are immediate, the consequences of poverty in countries with low development are appalling, progress is less than desired and allocated resources too little. In comparison, climate change can be perceived as a problem distant in time, uncertain in its effects and less consequential than present day poverty. Adaptation may therefore seem less urgent and less compelling than increasing development efforts for the world’s poor. But, as argued above, climate hazards are immediate, they are growing, they threaten the quality of life and life itself, and they directly impact on the goals of development. In balancing needs for climate adaptation with those of development, it is critical to note that there is strong complementarity between their goals and methods. A society that is made more climate-resilient through proactive adaptation to climate variations, extremes and changes is one in which development achievements and prospects are less threatened by climate hazards and there-
3232 J&J Climate Change
1/11/07
9:56 am
Page 16
16 Climate Change and Adaptation
fore more sustainable. Development can repay the compliment by creating conditions that better enable adaptation. This complementarity implies that integration of adaptation efforts with development can yield synergistic efficiencies and benefits that advance the goals of both agendas. This is not to deny that tradeoffs and hard choices may be required. That is the reality of pursuing multiple goals with limited resources. But there are sufficient complementarities to make integration a workable and desirable strategy. Adaptation activities carried out in isolation from mainstream development, and external to the authorities normally responsible for managing economic sectors and natural resources, may be practical in some contexts. Adaptation carried out in this manner, while not ideal, can help raise awareness, allow experimentation with different methods and demonstrate effective strategies. But adaptation as a stand-alone function that is implemented without the collaboration of agencies responsible for economic and resource policy and management will fail to mobilize the resources and the full range of actors that are necessary for success. To create a climate-resilient society, adaptation as a process needs to be integrated into policy formulation, planning, programme management, project design and project implementation of the agencies that are responsible for human and economic development, finance, agriculture, forestry, land use, land conservation, biodiversity conservation, water, energy, public health, transportation, housing, disaster management, and other sectors and activities. At the most basic level, integration would avoid maladaptive actions by development and other agencies that fail to account for climate-related risks and thereby unintentionally increase risks or miss easy opportunities to reduce them. This could be achieved by subjecting policies, programmes and projects to initial scrutiny for exposure to climate risks and modifying them accordingly, similarly to assessments that are carried out for environmental impacts, gender equality and poverty reduction. A further step towards integration would be for public sector agencies to promote and support actions and behaviours by individuals, the private sector and civil society that would narrow the current adaptation deficit. Yet more ambitious, but ultimately essential, are development strategies that proactively create conditions to enable adaptation processes by enhancing the capacities of individuals, strengthening community institutions, advancing knowledge and creating knowledge networks, removing obstacles, and providing appropriate incentives. Many of the AIACC studies demonstrate the need for comprehensive approaches to adaptation that are integrated with broader development strategies and examine how this might be done. They highlight several characteristics of development that would be complementary to the goals of adaptation. These include development that targets highly vulnerable populations, diversifies economic activities, expands opportunities for livelihoods that are less climate sensitive, improves natural resource management, encourages the development and diffusion of technologies that are robust across a wide range of climate variations and extremes, directs development away from highly hazardous locations toward less hazardous ones, and invests in expanding knowledge that is relevant to reducing climate risks.
3232 J&J Climate Change
1/11/07
9:56 am
Page 17
A Stitch in Time: General Lessons from Specific Cases 17
An examination by Osman-Elasha et al (Chapter 5) of community development efforts in Sudanese villages of Bara Province in North Kordafan, El Fashir in North Darfur and Arbaat in the Red Sea State demonstrates that development and adaptation to climate risks can be strongly complementary. Community development projects implemented in the villages integrated multiple strategies to improve livelihoods, the quality of life and sustainability of resource use within a context of recurrent drought. Using measures of changes in household livelihood assets (human, physical, natural, social and financial capital), the holistic approach to development taken in the study areas is found to have succeeded in increasing the capacity of households to cope with the impacts of drought. Community participation in the projects and reliance on indigenous technologies for improving cultivation, rangeland rehabilitation and water management that are familiar to the communities are found to be important factors for success. The sustainable livelihood approach appears to be a viable model for integrating development and adaptation to climate hazards at the community scale. Rice farmers in Thailand, Vietnam and Lao PDR rely primarily on their own capacity to implement strategies for coping with floods and mid-season dry spells; this is strongly limited by the social and economic conditions and natural resources in the surrounding community (Chinvanno et al, Chapter 13). Collective strategies to pool resources within communities and provide buffers against food and income losses that were widely prevalent in the past are now much diminished, though still important in Lao PDR. National policies are in general not supportive of reducing the vulnerability of small rice farmers to climate hazards. A national strategy to integrate climate risk management with rural development, poverty reduction and farm policies is recommended for raising the capacity and resilience of farm households and rural communities. Opportunities for effective interventions by national governments include assisting farm households with financial resources, expanding off-farm income opportunities, marketing of farm products, improving access to water, protecting the natural resource base, developing and promoting new technologies to diversify farm incomes, improving seed varieties, and providing information about current and changing climate hazards. Revitalizing community institutions is seen as important for enabling communities to benefit from national interventions. An approach to integrating adaptation and development that is being embraced by Pacific Island Countries such as Fiji also combines top–down and bottom–up strategies (Mataki et al, Chapter 15). Top–down actions would be taken by the national government to create a climate-proof society by creating incentives, enforcing regulations, assisting with capital financing and implementing large projects that are beyond the means of local authorities. These actions would encourage and enable development and settlement away from hazardous locations, the building of flood-proof homes, purchase of insurance, better land-use practices, regular dredging of rivers, and maintenance of irrigation channels and floodgates. Bottom–up actions would draw on the communal traditions of Pacific Island societies to engage members of the community in pooling financial and human capital and other local resources,
3232 J&J Climate Change
1/11/07
9:56 am
Page 18
18 Climate Change and Adaptation
and channelling these in efforts to reduce climate related risks. The current political framework in Fiji does not provide an effective means for local communities to make their concerns felt at the national level and there is lack of communication and coordination across government departments. These obstacles will need to be overcome for the combined top–down and bottom–up integration to be effective.
Evaluating Adaptation Options What to do, how much and when Adaptation decisions are made in a context of uncertainty and change. While we can be confident that the climate will change in response to greenhouse gas forcing, there is uncertainty about how it will change and how fast, particularly at the spatial scales that are relevant for adaptation. The impacts are also uncertain, partly because the changes in climate are uncertain, partly because the sensitivities of systems to climate stresses are uncertain, and partly because there is uncertainty about future demographic, social, economic, technological and governance conditions that will shape future exposures, sensitivities, capacities and vulnerabilities. There is also uncertainty about the potential performance of different adaptation options, their costs and possible unintended consequences. Uncertainty makes it difficult to decide what to do, how much of it to do and when to do it. Many of the choices will have irreversible consequences, so choosing wrong can be costly, even deadly. This is just as true for deciding not to adapt, or to delay adapting, as it is for deciding to adapt now. Delaying adaptation will result in irreversible consequences that could be avoided by adapting now. But not all adaptations could or should be implemented now. Which are appropriate for immediate or near-term action and which should be delayed? A number of factors are relevant to the selection of options for immediate action. These include the timing of benefits, the dependence of benefits on specific climate conditions, irreversible consequences, option values and thresholds for adverse impacts (Leary, 1999). Characteristics of adaptation measures that warrant consideration for early action include expectation of significant near-term benefits (for example, in narrowing existing adaptation deficits), performance that would produce benefits under a wide range of possible future climates, low capital costs and minimal irreversible consequences. Also of interest for early implementation are actions that would preserve or expand options for future adaptation (for example, purchase of development easements, developing knowledge networks and capacity building), or counteract looming thresholds for adverse impacts (for example, facilitated migration of species that are obligatory dispersers). Characteristics that would suggest delay of some actions while uncertainties are resolved include little near-term benefit, future benefits that depend on a narrow range of climate conditions, high capital costs and large irreversible consequences.
3232 J&J Climate Change
1/11/07
9:56 am
Page 19
A Stitch in Time: General Lessons from Specific Cases 19
Evaluation of options by AIACC studies Decision-making criteria for evaluating and selecting adaptation options vary from context to context. Criteria can vary depending on who is making the decision, what stakeholders are affected by the decision, what role stakeholders have in the decision process, the objectives of decision makers and stakeholders, and characteristics of the decision such as the time horizon, uncertainty about outcomes, irreversibility of consequences and consequences of decision errors. Criteria applied in the AIACC studies include net economic benefit, timing of benefits, distribution of benefits, consistency with development objectives, consistency with other government policies, cost, environmental impacts, spill-over effects, capacity to implement, and social, economic and technological barriers. In some cases the criteria are chosen by the investigators, in other cases they are chosen by stakeholders or based on stakeholder input. Methods for their application include formal cost–benefit and multi-criteria analysis, expert judgement, and participatory exercises with selected stakeholders. Callaway et al (Chapter 3) apply formal cost–benefit analysis to decisions about building water storage and switching water allocation regimes for the Berg river basin in South Africa. The net benefits from choices of reservoir capacity are uncertain and vary depending on how the future unfolds with respect to climate, growth in water demand, and reliance on either the current regulatory regime or water markets for allocating water. The climate scenarios analysed include no change in surface water runoff and reductions of either 11 or 22 per cent. Under the current regulatory regime for water allocation and water demand growth of 3 per cent per year, climate change would cause estimated damages with a present discounted value of 13.4 billion to 27.6 billion rand, or roughly 15 to 30 per cent of the total net benefits of water use in the basin. Adapting by correctly anticipating and adjusting reservoir capacity to the optimal size corresponding to the change in climate would reduce the damages and yield net benefits, but the net benefits are modest and less than 2 per cent of the damages. In contrast, a switch from the current regulatory regime to allocation by water markets would yield net benefits of roughly 10 to 20 per cent by allowing efficient reallocation of scarce water. Njie et al (Chapter 7) also apply cost–benefit analysis to evaluate adaptations to climate change. They investigate increased use of fertilizers and adoption of irrigation for growing cereals in the uplands of The Gambia. Climate change would cause estimated annual damages to cereal production of roughly US$150 million in 2010–2039 and in excess of US$1 billion in 2070–2079. Increased use of fertilizers would yield net benefits that would reduce climate change damages by 10 per cent or more. Irrigation, however, is found to yield negative net benefits in the 2010–2039 time frame and mixed results in the more distant future. For cereal production, the high cost of pump irrigation relative to cereal prices make irrigation an inefficient adaptation for cultivation of cereals, at least in the near to medium term. Yin et al (Chapter 12) apply an analytic hierarchy process, a form of multicriteria analysis, to evaluate adaptation options for the water sector in the
3232 J&J Climate Change
1/11/07
9:56 am
Page 20
20 Climate Change and Adaptation
Heihe river basin of north-western China. Stakeholder meetings and surveys were used to elicit judgements about the effectiveness of different options with respect to four decision criteria and the relative importance of the criteria. The criteria include water-use efficiency, economic returns on water use, environmental effects and cost. The results rank intuitional options for managing water demand above engineering measures to increase water supply. Preferred options include economic reforms that would constrain sectors that are large water consumers, water user associations to share information and promote water conservation, and transferable water permits for allocating water use. Lasco et al (Chapter 14) perform a tradeoff analysis of effects of adaptations in one sector that spillover to and impact on other sectors in the Pantabangan–Carranglan watershed of the Philippines. Options are identified and examined for agro-forestry, water resources and local communities. They find that spillovers are common because the shared water resource creates a high degree of interdependence among people, livelihoods and biophysical resources located within the watershed. The spillovers include both positive and negative externalities. For example, many of the options identified for agro-forestry such as improving water use efficiency and controlling runoff and erosion have beneficial effects on the water sector and on local community institutions. But stricter enforcement of forest protection laws and reforestation to protect water resources can negatively affect incomes and livelihoods of some landowners and cause farmers in informal settlements with insecure land tenure to be forced from their farms. They find that these types of tradeoffs are seldom considered in planning new projects or revising policies, risking negative impacts on others, conflicts among stakeholders in the watershed and missed opportunities for mutually beneficial actions. In Mongolia, evaluation of adaptation options for the livestock sector applied a two-tiered screening process with participation from herders, scientific experts, and authorities from local, provincial and national offices (Batima et al, Chapter 11). In the first tier, options are screened for satisfying broad criteria for promoting both adaptation and development goals, consistency with government policies, and environmental impacts. Options that pass the first screening are then evaluated against a second tier of six additional criteria. These include capacity to implement, importance of climate as a source of risk, near-term benefits, long-term benefits, cost and barriers. Adaptation strategies that emerge as priorities from this process include measures that generate near-term benefits by improving capabilities for reducing the impacts of drought and harsh winters as well as measures that produce long-term benefits by improving and sustaining pasture yields. Some of the specific measures identified as warranting further consideration include improving pastures by reviving the traditional system of seasonal movement of herds; increasing animals’ capacity to survive winters by modifying grazing schedules and increasing use of supplemental feeds; enhancing rural livelihoods by strengthening community institutions to regulate use of pasture and provide local services such as education, training, access to credit and insurance; and research and monitoring to develop and improve forecasting and warning systems.
3232 J&J Climate Change
1/11/07
9:56 am
Page 21
A Stitch in Time: General Lessons from Specific Cases 21
In the study of dengue fever in the Caribbean, the investigators evaluate adaptation options for cost, effectiveness, social acceptability, environmental friendliness, promotion of local cooperation and technical/socioeconomic challenges (Taylor et al, Chapter 16). Three options of multiple measures are recommended based on these criteria. The first option would refocus current education, disease surveillance and vector control efforts to be more proactive and to address deficiencies in community involvement. Emphasis would be placed on education that stresses individual responsibility and community benefits of measures to reduce human–vector contact. The second option would combine the above measures with designing, producing and promoting the use of low-cost covered containers for storing rainwater. Discarded and uncovered oil drums are the most commonly used means of capturing and storing water and are ideal breeding sites for mosquitoes. The third option would include all the above plus development and implementation of an early warning system. Early warnings to give advance knowledge of the expected severity of possible disease outbreaks would enable responses to be calibrated to the anticipated threat level. Responses to an alert would include more frequent and extensive vector surveillance and control, stepped up education efforts tailored to the threat level, and more diligent efforts to eliminate breeding sites for mosquitoes.
Creating an Enabling Environment Many studies, including our own, identify numerous options for adapting to existing and changing climate hazards. Some are novel and untested, but many are based on current practices that have been amply demonstrated to reduce risks. As we noted earlier, individuals, communities and nations all have a strong self-interest in adapting. Yet many options go unused, or are used much less extensively or intensively than their benefits would seem to warrant. It is not for lack of options that adaptation lags. It is lack of determination, lack of cooperation and lack of means that impede it. Deliberate and sustained efforts are needed to create an enabling environment for overcoming these obstacles and facilitating the process of adaptation. The efforts need to engage the general public, as well as stakeholders and authorities, from the many different economic sectors and spheres of activity that are affected by climate and should link across local, provincial, national and international jurisdictions.
Creating the determination to adapt A primary obstacle is a lack of will, or determination, to adapt. This can happen at the individual level (for example people failing to take simple actions to limit their own exposure to malaria and dengue), the community level (local authorities allowing new development in hazardous locations), the national level (ministries failing to consider climate risks in new programmes and not being held accountable), and the international level (adaptation
3232 J&J Climate Change
1/11/07
9:56 am
Page 22
22 Climate Change and Adaptation
continuing to receive strong rhetorical support from international environmental and development communities but few resources). The reasons for lack of will are varied. One is a problem of awareness and understanding. People lack knowledge about, or are uncertain or sceptical about, current climate risks, climate change, options for adaptation and the effectiveness, and the feasibility and cost of adaptation. Another important reason is that people have other objectives that compete with adaptation for attention, priority and resources. In essence, determination to adapt will not gain acceptance unless people find the evidence that climate risks represent a substantial problem compelling, that addressing the risks warrants priority on a par with other objectives, that there are effective, feasible and affordable options, and that we know enough to make wise choices. Greater awareness and knowledge can help to create the determination to adapt. But it is not enough to simply create more knowledge. It needs to get into the hands, or the heads, of people facing decisions about how to allocate scarce resources to achieve their objectives, objectives that include, but are not limited to, reducing risks from climate and other sources. The knowledge needs to be relevant to the decisions being made and understandable to stakeholders and decision makers, who might be residents of hazardous places, resource users and owners, farmers, business operators, community leaders or government officials. The knowledge also has to be seen as credible and untainted by bias or intent to manipulate. The different types of knowledge, intended users and applications are too varied for the functions of knowledge creation, collection, communication, integration and interpretation to be done well by a single entity. Networks of knowledge institutions are needed that link the scientists, practitioners and public, the various economic sectors, and local, national and international actors. In each of the AIACC study areas, knowledge networks are very incomplete and not well coordinated, resulting in substantial gaps in the awareness and understanding of climate hazards, climate change and adaptation among many key stakeholders. This situation can be improved by strengthening knowledge networks. Investments are needed in scientific research, assessment and capacity in areas that are relevant to understanding climate risks and response options. Expanded efforts are needed to collect knowledge from the experiences and practices of at-risk groups, including traditional knowledge. Mechanisms are needed to integrate, interpret and communicate the created and collected knowledge and to assist stakeholders in applying the knowledge in decision making. Avenues are needed for stakeholders to give feedback about the information received and the information required, as well as to share their knowledge with other at-risk groups. Participatory processes that engage stakeholders and attempt to link the different functions and components of knowledge networks can be effective in generating and communicating knowledge that is relevant, understandable and credible. The AIACC project is one example of such a process and similar projects have been initiated and are underway. Ultimately, though, the generation and communication of knowledge for supporting adaptation needs to be
3232 J&J Climate Change
1/11/07
9:56 am
Page 23
A Stitch in Time: General Lessons from Specific Cases 23
connected with and embedded in ongoing processes of human development, economic planning, poverty reduction and resource management.
Creating cooperation to adapt What any one person or organization can do to adapt is very much constrained by what others do or do not do. Cooperation among members of a community can mobilize resources to reduce, hedge and spread risks beyond what individuals acting independently might achieve. Cooperation between local and national authorities can rationalize policies and plans so that they work toward common adaptation goals and not at cross purposes. Cooperation among stakeholders and authorities from different economic sectors can increase positive spillovers and avoid negative spillovers of their sector-based strategies. International cooperation can help to ensure that actions are based on the best available science, that information about best practices is shared, that financial resources can be pooled and directed toward common goals, and that efforts under different international agreements contribute to adaptation objectives where possible. Fostering cooperation on adaptation requires leadership within national governments. An environment or science ministry might play a useful role in raising awareness, sharing information about risks and adaptation options, supporting knowledge networks, assessing the implications of new legislation and policies for narrowing or widening the adaptation deficit, and monitoring overall progress on managing climate risks. But environment and science ministries typically lack the standing to marshal resources at the required scale or to compel other ministries to cooperate. The determination to adapt will need to permeate beyond environment and science ministries and be accepted by other ministries as important to their missions and objectives if there is going to be effective cooperation. The purpose of integrating or mainstreaming adaptation with development is to enlist the cooperation of these other ministries and associated stakeholders in making adaptation commonplace in economic and sector-, resource- and livelihood-based planning and programmes at national to local scales. Cooperation is not forthcoming when actors and stakeholders in these different spheres of activity view climate change as immaterial to their main objectives and adaptation as a potential new mandate that will divert resources from their priorities. The experience of the AIACC case studies is that stakeholders from varied perspectives often are aware of climate threats to their interests and that, when put in a broad context of managing current climate hazards and not limited to only climate change, are willing to engage with others to assess threat levels and possible responses. Through their participation in an assessment process, many accept, or at least are willing to consider seriously, the need to adapt to narrow the existing adaptation deficit, to limit vulnerability to climate change in the near- to medium-term future and to cooperate with others to move toward a climate-proof society.
3232 J&J Climate Change
1/11/07
9:56 am
Page 24
24 Climate Change and Adaptation
Creating the means to adapt Determination to adapt and cooperation are not sufficient by themselves, however. The means to adapt must also be available. Much of what needs to be done to adapt is at the level of the household and community. But for the most vulnerable households and communities, the means to adapt are in short supply. Often they do not have sufficient resources, knowledge and skills to implement measures that would reduce the risks that they face. Targeting development to highly vulnerable populations to provide expanded and diversified livelihood opportunities and access to services such as clean water, health care, education and credit can increase the assets of households and bolster their capacity to cope with and adapt to hazards of all types, including climatic hazards. Capacities that are specific to climate adaptation can be increased by providing information, training, technical advice and resources for adopting technologies and practices that can reduce climatedriven damages and variability of production and income. Strengthening and supporting community institutions can increase the capacity for collective action to reduce, hedge and spread risks.
Financing adaptation Financial resources are also an important part of the means to adapt. At the local level, many communities have been resourceful in operating village funds and other mechanisms to provide access to credit for small-scale farmers, enterprise owners and others that have proven useful for helping to finance risk-reducing investments or recovery from losses. Private sector finance markets play an important role in financing investments by larger enterprises, for example, for large-holder farmers to diversify farm operations, adopt new seed varieties, implement irrigation and provide insurance against losses. Insurance needs particular attention as it is far less prevalent in developing countries than in developed; premiums, already more than can be afforded by poor and vulnerable communities, are rising; and insurers are withdrawing from some markets where climate risks are high. Private sector innovations in micro-credit and micro-insurance can help to increase the access of the poor to financial resources. National governments also assist with direct financial payments and with subsidized credit and insurance, although in many places financial assistance from national governments to both rural and urban poor is diminishing. At the international level, financial assistance is being provided for adaptation through the Global Environment Facility under the United Nations Framework Convention on Climate Change (UNFCCC) as well as through development assistance from bilateral and multilateral aid agencies. This international funding is acting as a catalyst for raising awareness, building capacity, advancing understanding of risks and response options, and engaging developing country governments in prioritizing and assessing options. Recently, funding is also being made available for experimenting with and implementing selected measures for adapting to climate change.
3232 J&J Climate Change
1/11/07
9:56 am
Page 25
A Stitch in Time: General Lessons from Specific Cases 25
But the magnitude of the adaptation problem and the likely financial needs in developing countries are far greater than current funding can cover. Compelling arguments have been made that developed countries have a liability to help fund adaptation in developing countries that also exceed current contributions (see, for example, Baer, 2006). International financial assistance for adaptation does appear to be increasing. But it is not clear to what extent these are new resources or reallocations of limited development assistance funds, which is a source of tension for integrating adaptation and development. While the logic for integration is inescapable, there is legitimate concern that this will divert some funds away from critically important development objectives. Ultimately though, financing for adaptation will need to come from multiple sources, including developing country governments and their private sectors, as well as from foreign direct investment, international development assistance, and specialized funds under the UNFCCC and other multilateral sources.
A Final Word Climate hazards exact a heavy toll, impacting most strongly on the poor and acting as a drag on development. The toll is rising as climate change widens the gap between our exposures to risks and our efforts to mange them. National governments are increasingly aware of the growing risks and are cooperating in the UNFCCC and other processes to cautiously consider how to respond. But there is not yet widespread determination to adapt. The determination to adapt can be assisted by increasing recognition that closing the current adaptation deficit provides immediate benefits and is a first step toward adapting to climate change, that feasible, effective and affordable options are available, and that these options do not require certainty about how the climate will change to be effective. But beyond determination, the means to adapt need to be enhanced. Knowledge of climate risks and adaptation response strategies need to be increased. Capacities of atrisk households and community institutions need to be raised and access provided to improved technologies. Climate-sensitive natural resources need to be protected and rehabilitated. Financial resources are needed. Most of all, adaptation needs to be integrated with development so that it becomes commonplace in each sector of human activity. The time to act, to make a stitch in time, is now.
References Adger, W. N., S. Agrawala, M. Mirza, C. Conde, K. O’Brien, J. Pulhin, R. Pulwarty, B. Smit and K. Takahashi (2007) ‘Assessment of adaptation practices, options, constraints and capacity’, in M. Parry, O. Canziani, J. Palutikof and P. J. van der Linden (eds) Climate Change 2007: Impacts, Adaptation and Vulnerability, contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (forthcoming)
3232 J&J Climate Change
1/11/07
9:56 am
Page 26
26 Climate Change and Adaptation African Development Bank; Asian Development Bank; Department for International Development, UK; Directorate-General for Development, European Commission; Federal Ministry for Economic Cooperation and Development, Germany; Ministry of Foreign Affairs – Development Cooperation, The Netherlands; Organization for Economic Cooperation and Development; United Nations Development Programme; United Nations Environment Programme; and The World Bank (2003) Poverty and Climate Change, Reducing the Vulnerability of the Poor through Adaptation, The World Bank, Washington, D.C., US Baer, P. (2006) ‘Adaptation: Who pays whom?’, in W. N. Adger, J. Paavola, S. Huq and M. J. Mace (eds) Fairness in Adaptation to Climate Change, MIT Press, Cambridge, MA, US Burton, I. (2004) ‘Climate change and the adaptation deficit’, in Adam Fenech (ed) Climate Change: Building the Adaptive Capacity, papers from International Conference on Adaptation Science, Management, and Policy Options, Lijiang, Yunnan, China, 17–19 May 2004, Meteorological Service of Canada, Environment Canada, Toronto Burton, I., and M. van Aalst (2004) ‘Look before you leap: A risk management approach for incorporating climate change adaptation in World Bank operations’, Working Paper No 100, Environment Department, World Bank, Washington, D.C., US IPCC (2001) ‘Summary for policymakers’, in J. McCarthy, O. Canziani, N. Leary, D. Dokken and K. White (eds) Climate Change 2001: Impacts, Adaptation and Vulnerability, contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York IPCC (2007a) ‘Summary for policymakers’, in M. Parry, O. Canziani, J. Palutikof and P. van der Linden (eds) Climate Change 2007: Impacts, Adaptation and Vulnerability, contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York IPCC (2007b) ‘Summary for policymakers’, in S. Solomon, D. Qin, M. Manning, Z. Chen, M.C. Marquis, K. Averyt, M. Tignor and H. L. Miller (eds) Climate Change 2007: The Physical Science Basis, contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York Leary, N. (1999) ‘A framework for benefit–cost analysis of adaptation to climate change and climate variability’, Mitigation and Adaptation Strategies for Global Change, vol 4, pp307–318 Leary, N., C. Conde, J. Kulkarni, A. Nyong and J. Pulhin (eds) (2008) Climate Change and Vulnerability, Earthscan, London Martin-Hurtado, R., K. Bolt and K. Hamilton (2002) The Environment and the Millennium Development Goals, The World Bank, Washington, DC McCarthy, J., O. Canziani, N. Leary, D. Dokken and K. White (eds) (2001) Climate Change 2001: Impacts, Adaptation and Vulnerability, Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York McMichael, A., A. Githeko, R. Akhtar, R. Carcavallo, D. Gubler, A. Haines, R. S. Kovats, P. Martens and J. Patz (2001) ‘Human health’, in J. McCarthy, O. Canziani, N. Leary, D. Dokken and K. White (eds) Climate Change 2001: Impacts, Adaptation and Vulnerability, contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York Munich Re (2005) Topics GEO – Review on Natural Catastrophes 2005, Munich Re, Munich, Germany
3232 J&J Climate Change
1/11/07
9:56 am
Page 27
A Stitch in Time: General Lessons from Specific Cases 27 Pelling, M., A. Maskrey, P. Ruiz and L. Hall (2004) Reducing Disaster Risk: A Challenge for Development, United Nations Development Bank, Bureau for Crisis Prevention and Recovery, New York Schneider, S. H., S. Semenov, A. Patwardhan, I. Burton, C. Magadza, M. Oppenheimer, A. B. Pittock, A. Rahman, J. B. Smith, A. Suarez and F. Yamin (2007) ‘Assessing key vulnerabilities and the risk from climate change’, in M. Parry, O. Canziani, J. Palutikof and P. J. van der Linden (eds) Climate Change 2007: Impacts, Adaptation and Vulnerability, contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (forthcoming) Smit, B., O. Pilifosova, I. Burton, B. Challenger, S. Huq, R. Klein and G. Yohe (2001) ‘Adaptation to climate change in the context of sustainable development and equity’, in J. McCarthy, O. Canziani, N. Leary, D. Dokken and K. White (eds) Climate Change 2001: Impacts, Adaptation and Vulnerability, contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York
3232 J&J Climate Change
1/11/07
9:56 am
Page 28
2
Adapting Conservation Strategies to Climate Change in Southern Africa Graham von Maltitz, Robert J. Scholes, Barend Erasmus and Anthony Letsoalo
Introduction Global climate change is predicted to have substantial impacts on southern Africa’s biodiversity, including wide-scale extinctions over the next 50 years (Rutherford et al, 1999; Hannah et al, 2002a and b; Gitay et al, 2001 and 2002; Midgley et al, 2002a and b; MA, 2005). At a global scale, Thomas et al (2004) have predicted that 15–37 per cent of species in their sample (which covered 20 per cent of the Earth’s surface) may be at risk of premature extinction due to anthropogenically caused global change by 2050. The Millennium Ecosystem Assessment, using different models and assumptions based largely on habitat loss, reached similar conclusions (MA, 2005). Within South Africa, a reduction in size and an eastward shift for current biomes is predicted and up to half of the country will likely have a climatic regime that is not currently found in the country (Rutherford et al, 1999). The succulent karoo biome, (a succulent-dominated semi-desert located on the southwestern coast of southern Africa) is projected to be the most severely impacted, with the grassland and fynbos (a Mediterranean-climate sclerophyllous thicket that approximates to the Cape Floristic region) biomes also likely to suffer from high climate change impacts (Rutherford et al, 1999; Midgley et al, 2002a and b). Fynbos and succulent karoo are biodiversity hotspots of international importance (Myers et al, 2000), with the latter being one of only two globally important arid-climate biodiversity hotspots. Two climate parameters critical for animal and plant species distributions are temperature and water balance (a combination of precipitation and evaporation, which, in turn, is directly influenced by temperature) (Cubasch et al, 2001). The dynamics of plant and animal populations change at the edge of individual species’ distribution, as net mortality becomes larger than net fecundity, with a spatial gradient of declining population numbers as a result. In a scenario of climate change, the direct influence of temperature and water balance in combination with the indirect influence of interspecies competition,
3232 J&J Climate Change
1/11/07
9:56 am
Page 29
Adapting Conservation Strategies to Climate Change in Southern Africa 29
fire frequency, pollinator distribution, herbivory and predation, food availability, soil type, topography and so forth could lead to the progressive extinction of nonvagile species in their natural range, beginning with population dieback in the so-called ‘trailing edge’ of the historical distribution range (Davis and Shaw, 2001; Gaston, 2003).1 In southern Africa, global circulation models project the greatest increases in temperature (2–4°C this century) for the inland areas, while the coastal areas are predicted to experience somewhat lesser increases (1–3°C), due to the thermal buffering effect of the oceans (Cubasch et al, 2001; Scholes and Biggs, 2004). Changes to precipitation are more difficult to predict, and there is less agreement between models. For southern Africa the majority of models predict about a 10 per cent reduction in annual precipitation during the 21st century in the western two-thirds of the continent south of 15ºS , while the eastern onethird may see an increase of the same order (Scholes and Biggs, 2004; Hewitson and Crane, 2006). A combination of increased temperature (and thus increased evaporative demand) with decreased rainfall will increase the aridity of affected environments, notwithstanding the slight offsetting beneficial effect of elevated CO2 on plant water use efficiency (Scholes and Biggs, 2004). A combined increase in rainfall and temperature will increase primary plant production, but will still be detrimental to specific species (Gitay et al, 2001; Gelbard, 2003). Excluding evolutionary adaptations, species can be classified into four functional groups based on their response to climate change as follows: 1 2
3 4
Persisters: These species have tolerance for the new climate of their current location. Obligatory dispersers: These species will have to physically move with the changing climate to track areas with suitable climates (autonomous dispersers), or alternatively will have to be moved artificially to new areas with suitable climates if they are unable to move on their own (facilitated dispersers). Range expanders: These species may expand into new climatic envelopes that are not currently available, but to which the species are already well adapted. No-hopers: These species cannot do any of the above and will become prematurely extinct, although they may persist under unsuitable climates for some time.
Some species, referred to as partial dispersers, will experience range shifts causing them to persist in parts of their previous range while dispersing into new areas. The time span involved and the intensity of the climate change experienced (or modelled) will determine the extent to which species may persist or are obliged to disperse. Detailed modelling on the impacts of climatic change on individual species has been conducted in the fynbos and succulent karoo regions. The Proteaceae was studied as a surrogate for the fynbos vegetation to understand individual species response to changing climate over the next 50 years and thus to evaluate future conservation strategies. The model predicted that 57 per cent were
3232 J&J Climate Change
1/11/07
9:56 am
Page 30
30 Climate Change and Adaptation
persisters, 26 per cent partial dispersers, 6 per cent obligatory dispersers and 11 per cent no-hopers (Williams et al, 2005). In the karoo region, it was found that the riverine rabbit (Bunolagus monticularis) is likely to become extinct because of its specialized food and habitat requirements, while the tortoise (Homopus signatus), which is less selective, is likely to survive in the 50-year study timeframe (G. O. Hughes, personal communication, 2005). The combined impacts of climatic change and CO2 effects have been modelled for the lowveld savanna regions of South Africa (R. J. Scholes, personal communication, 2005).2 Preliminary results suggest that the decrease in soil moisture and the increase in temperature overwhelm the small elevated CO2 advantage that trees have, given that C3 and C4 plants respond differently to these factors.3 The model predicts that the structural and functional habitat suitability for browsers and grazers will likely remain relatively constant in the 50-year timeframe, provided that fire and elephant management are appropriate. Overall, the carrying capacity for large herbivores is projected to decrease by about 10 per cent. Although this study does not consider individual species, it suggests that the functional integrity of the savanna habitat can be maintained near to current conditions through appropriate management.
A Brief History of Conservation in Southern Africa Extensive tracts of land are managed as conservation areas in southern Africa (see Table 2.1). About half the countries in the region exceed the International Union for the Conservation of Nature (IUCN) guidelines of 10 per cent of land area under formal conservation. Over the entire region, approximately 10 per cent of land is conserved in IUCN categories I–V reserves (reserve categories set up strictly for conservation) and another 8 per cent conserved in areas managed for sustainable use, i.e. IUCN category VI areas. Some countries fall far short of the IUCN guidelines; for example, in Lesotho only 0.2 per cent of the surface area is conserved (Scholes and Biggs, 2004; WDPA, 2005). Even where countries have a relatively high level of land conserved, the fraction of biodiversity conserved may be substantially less (Rodrigues et al, 2004; Orme et al, 2005). This is because, historically, conservation has been based on the availability of land and in many instances the presence of big game species rather than strategic conservation objectives (Pressey et al, 1993; Heywood and Iriondo, 2003). As a result, of the 52 unique ecoregions identified in southern Africa (Olson et al, 2001), 23 per cent (15 per cent of land area) have less than 3 per cent conservation (see Table 22.). Forty per cent of ecoregions, representing 35 per cent of the land area, have less than 5 per cent formally conserved in IUCN reserves. Southern Africa has an exceptionally high biodiversity, including a number of centres of endemism and three biodiversity hotspots (Myers et al, 2003). The Madagascar hotspot has only 2.9 per cent of the area conserved in IUCN reserves with a further 1 per cent conserved outside of IUCN reserves. The succulent karoo hotspot has only 1 per cent conserved, although there are proposals to conserve an additional 19 per cent. The Cape floristic region is well conserved in the mountainous areas
3232 J&J Climate Change
1/11/07
9:56 am
Page 31
Adapting Conservation Strategies to Climate Change in Southern Africa 31
Table 2.1 The area as a percentage conserved in southern African countries in IUCN reserves (IUCN classes I–V), IUCN sustainable resource use areas (IUNC class VI), and other non-IUCN conservation areas Country
IUCN VI
IUCN I–V
Total IUCN
non-IUCN
Total
Angola
0.0
6.7
6.7
5.5
12.2
Botswana
0.0
18.0
18.0
12.7
30.7
Burundi
0.0
3.7
3.7
0.0
3.7
Congo
0.5
9.3
9.8
8.5
16.6
Congo (DRC)
3.6
4.7
8.2
3.1
10.6
Equatorial Guinea
0.0
17.2
17.2
0.0
17.2
Gabon
0.0
2.5
2.5
14.5
16.4
Kenya
1.6
5.6
7.1
2.5
9.6
Lesotho
0.0
0.3
0.3
20.8
21.0
Madagascar
0.6
2.4
2.9
1.0
4.0
Malawi
0.0
8.5
8.5
0.0
8.5
Mozambique
1.4
4.0
5.4
5.9
11.1
Namibia
0.7
13.2
13.8
3.6
16.7
Rwanda
0.0
11.1
11.1
0.0
11.1
Seychelles
0.0
59.2
59.2
0.0
59.2
South Africa
0.0
5.5
5.5
0.8
6.2
Swaziland
1.0
2.1
3.0
0.0
3.0
Tanzania
0.1
14.8
14.9
16.0
27.8
Uganda
12.6
7.4
20.0
6.1
23.8
Zambia
18.8
8.1
26.8
9.5
35.4
Zimbabwe
4.8
7.9
12.7
15.3
27.9
Total
2.6
7.6
10.2
6.0
15.6
Note: Non-IUCN conservation areas are mostly forest reserves. All data presented in this table are based on WDPA (2005).
but poorly conserved on the flats (see Table 2.3). By comparison, the mopane savanna regions (not a biodiversity hotspot) are well preserved, largely due to their low economic value for agriculture (see Table 2.3). Formal conservation began in the late 19th century. From about 1910 to 1970, there was a steady expansion of protected areas (see Figure 2.1), which fell into two categories: the forest reserves, managed for sustainable wood extraction and/or catchment protection, and the game and nature reserves for hunting, which originally tended to be centred in areas with high wildlife populations. Game and nature reserves are currently managed for biodiversity conservation and ecotourism (von Maltitz and Shackleton, 2004). During this period, reserves enjoyed strong state support and were well maintained. The postcolonial period saw a shift in government focus to social development
3232 J&J Climate Change
1/11/07
9:56 am
Page 32
32 Climate Change and Adaptation
Table 2.2 The amount of conservation per ecoregion Conservation in IUCN category I–VI reserves
Total conservation including IUCN and non-IUCN reserve areas (some of which are only in the planning stage)
Percentage Total Cumulative Cumulative conserved number percentage percentage per of ecoof ecoof total ecoregion regions regions land area
Total Cumulative Cumulative number of percentage percentage ecoof ecoof total regions regions land area
<3%
12
23.1
15.1
8
15.4
10.6
3–5%
9
40.4
35.1
5
25.6
19.5
5–10%
10
59.6
53.2
4
32.7
27.1
10–15%
10
78.8
83.6
12
55.8
60.4
15–20%
3
84.6
86.1
8
71.2
68.7
>20%
8
100.6
100.6
15
100.6
100.6
Note: All data presented are based on ecoregions studied in Olson et al (2001) and the WPDA (2005) database of protected areas. This is for the same set of southern and east African countries, including Madagascar, as listed in Table 2.1. Note that non-IUCN areas include some planned areas that have as yet not been proclaimed. Most of the non-IUCN areas are forest reserves.
issues and budgets for protected areas diminished. Due to population growth, there was increasing pressure on reserve borders and increasing conflict over resources. In some cases, local communities invaded the reserves and settled there (Fabricius et al, 2004; von Maltitz and Shackleton, 2004; Child, 2004). In response to the budget constraints and growing negative perception of conservation areas, government policies regarding resource conservation were altered. The trend from the 1980s was towards delegating ownership of wildlife and forestry resources from the state to those owning or resident on the land. This would make it possible for communities on communal land to enter into community-based natural resource management (CBNRM) programmes (Fabricius et al, 2004; Child 2004; Hutton et al, 2005), and for the establishment of private wildlife ranches on commercial land (ABSA, 2003). This approach
Table 2.3 Extent of conservation versus ‘need’ for conservation Vegetation type
Centre of endemism
Area in km2 thousands
Percentage transformed
Percentage conserved
Mopane Shrubveld
no
26
0
99.8
Mopane Bushveld
no
209
8
38
West Coast Renoster veld
yes
61
97
1.7
Mountain Fynbos
yes
247
11
26.2
Note: Two extremes shown are based on south African statistics; all data presented are based on Low and Rebelo (1996).
3232 J&J Climate Change
1/11/07
9:56 am
Page 33
Adapting Conservation Strategies to Climate Change in Southern Africa 33
Figure 2.1 The increase in conservation areas and the number of reserves in seven southern African countries Note: Only national parks and large reserves in South Africa have been included. Source: Based on Cumming (2004).
was expected to promote biodiversity conservation in the communal and private areas by creating an economic incentive for conservation (Fabricius, 2004; Child, 2004; ABSA, 2003). However, the programme has had mixed success, largely due to a lack of appropriate capacity, both in government departments and in communities (Hutton et al, 2005), resulting in increased criticism of the co-management and resource sharing strategy (Wilshusen et al, 2002; Hutton et al, 2005; Büscher, 2005). The most recent trend is towards international assistance for conservation in Africa, and millions of dollars have been contributed for this purpose by the Global Environmental Facility (GEF) and by first world countries. For the first time in decades, new areas are being proposed for conservation, and existing conservation is being strengthened. Strategic conservation planning tools such as Worldmap (www.nhm.ac.uk/science/projects/worldmap/index.html) and C-plan are making it possible to plan the location of reserves in a scientific and defensible manner to achieve agreed conservation targets (Pressey et al, 1993; Margules and Pressey, 2000; Pressey and Cowling, 2001), for example, in the fynbos, thicket and succulent karoo regions of South Africa (Cowling and Pressey, 2003). A number of transnational megaparks (sometimes referred to
3232 J&J Climate Change
1/11/07
9:56 am
Page 34
34 Climate Change and Adaptation
as ‘peace parks’) are also being developed such as the Limpopo, Kalagadi and Maluti-Drakensberg Transfrontier Parks (van der Linde et al, 2001). The possible consequences of climate change to biodiversity are also beginning to be considered (Hannah et al, 2002a and b; Midgely et al, 2003; Williams et al, 2005).
An Overview of Adaptation Options for Biodiversity Conservation in a Climatically Changing Environment Conservation becomes a moving target in a climatically changing environment, and although current reserve systems are a starting point, there is no clear end point. Biodiversity patterns in 50 years’ time represent only one period in an environment that is likely to see increasing temperature for at least 200 years because of the residual effect of CO2 increases (Cubasch et al, 2001). The following potential adaptation options were identified to prevent extinction of biodiversity given the predicted climate change: • • • • •
Do nothing (i.e. maintain the current conservation strategy). Reconfiguration of reserve system to strategically conserve areas that accommodate climate change. Matrix management, i.e. managing the biodiversity in areas outside of reserves. Translocation of species into new habitats. Ex-situ conservation, for example, gene banking, cryopreservation, zoos and botanical gardens.
Current understanding of ecosystem response to climate change, based both on historical data and modelled predictions, suggests that individual species will respond at different rates. As a consequence, entire ecosystems will not move in unison, but species will move independently, leading to altered community composition (Huntley, 1991; Graham, 1992; Gitay et al, 2001; Williams, 2005; Thuiller et al, 2006; Bush, 2002). It is therefore important that, in attempting to minimize losses, conservation strategies must also account for individual species in addition to the need to maintain entire habitats (ecosystems), which would be likely to have a different composition in the future, although in some instances the functional attributes may be similar (see description of lowveld savanna modelling study above). On the basis of individual species responses to climate change, a set of adaptation options are identified in Figure 2.2 and their relative constraints and benefits are compared in Table 2.4.
Conservation of species that persist or expand their range Where a species persists in large populations in an already-conserved area under future climates, there is no strong basis for concern. However, if the species becomes invasive and its range expands then it may become a threat to
3232 J&J Climate Change
1/11/07
9:56 am
Page 35
Adapting Conservation Strategies to Climate Change in Southern Africa 35
Figure 2.2 A decision tree for selecting adaptation strategies for different surrogate species based on their response to climate change
other species and may need control. If the species is already threatened under current conditions, even if it persists, it might warrant extra conservation attention, especially if it is not currently found in existing conservation areas.
Conservation of obligatory dispersers For the autonomous obligatory dispersers, a climatically and environmentally suitable migratory pathway must exist to allow the species to move through the landscape to track the changing climate. The extent of land transformation in dispersal corridors is a major concern (Hannah et al, 2002a). There are two
3232 J&J Climate Change
1/11/07
9:56 am
Page 36
36 Climate Change and Adaptation
Table 2.4 Relative financial costs compared to the advantages and disadvantages of differing adaptation options Relative financial cost
Advantages
Disadvantages
Do nothing, i.e., maintain the current conservation strategy
Zero additional cost but there is an existing high current cost of conservation management
The current reserve system is in place and funded. No new land needed. Easier to justify than new land acquisition. Will preserve a large percentage of current biodiversity. Maintains intact habitats and ecological interactions.
Not optimized for climate change. No provision is made for protection in a changing climate, so extinction of some species is inevitable. In most areas, the current reserves do not optimize biodiversity conservation, even for a static climate.
Reconfigure reserves
Very high additional cost if multiple small reserves are added, more cost-effective if existing reserves are expanded or realigned.
Ensures high conservation levels for a changing climate. Allows full state control and management of the land. If adequately funded reserves remain the most secure mechanism for ensuring biodiversity conservation. Maintains intact habitats, ecological process, and a large proportion of biodiversity. Most affordable when linked to existing reserves and for large areas. Best suited to land with high agricultural or development potential.
The high cost. The political aspects relating to acquiring land from private individuals or communities. Very difficult to acquire new land once the land is settled (as it is in many priority areas). Poor predictive capacity currently on how species will respond to climate change; therefore, it is difficult to know which land to include. Requires strategic planning to identify priority areas. Unlikely to ever conserve more than a small percentage of the total biodiversity.
Use contractual reserves
Less expensive per hectare than state-run reserves, especially if small areas involved.
No capital cost for land acquisition. A more cost-effective strategy to deal with small parcels of land than formal reserves. May be less detrimental to other land-based economic activities (e.g. it may be possible to mix agriculture with strategically configured migratory corridors. Similar benefits to reserve expansion, though slightly less secure. Does not require relocation of current land owners and therefore politically sounder option. Cheaper than reserve expansion, especially on agriculturally marginal land.
Less state control over the land. May require expensive administration and other infrastructure to administer. A recurring state budgetary item that may be cut in the future. May be difficult to secure longterm (indefinite) funding. This is still potentially an expensive option, particularly on land where high-value alternative land use options are available. Requires strategic planning to identify priority areas. Easier to implement on private land than on communal land. May be less effective at conserving some ecosystem processes than conventional reserves.
3232 J&J Climate Change
1/11/07
9:56 am
Page 37
Adapting Conservation Strategies to Climate Change in Southern Africa 37
Table 2.4 (continued) Relative financial cost
Advantages
Disadvantages
Matrix Management Conservation outside of reserves
Some options are very inexpensive. All options are less costly than formal reserves. Because of the land area involved (potentially 5 to 10 times greater than conservation areas), the overall cost may be high.
Ensures migratory pathways even if limited information is available on priority areas. Potentially conserves the greatest amount of biodiversity. May be relatively inexpensive.
State has limited control. Land conversion will continue to threaten some species. Some species cannot be accommodated in populated areas due to human–animal conflicts.
Translocation
Relatively cheap compared to the above options, but actual costs will depend on the number of samples translocated and the species involved.
The only option for facilitated dispersers, i.e., where habitat cannot be reached by natural distribution mechanisms. Far cheaper than ensuring migratory corridors. Will still require a conservation network into which the species can be reintroduced.
Only conserves a fraction of the genetic diversity within a species. Competitive interactions with other species will be an unknown element. Does not conserve ecosystem processes, but only species. Will need a sound understanding of individual species habitats. Will require extensive research and monitoring to know which species to move, where to move them to, and what species need to be moved jointly (e.g., pollinators or seed dispersers). Potential negative impacts of translocated species on the existing species in the new habitat.
Ex situ conservation
Relatively cheap once the infrastructure is in place, but varies between different types of species.
An ‘insurance policy’ when there is uncertainty as to how species will respond in the natural environment. The only option for ‘no-hoper’ species. The only option where there is total habitat loss. Relatively cheap (but the cost cannot be compared directly with in situ conservation as different objectives are achieved).
Conserves only a tiny fraction of genetic diversity. Conserves no ecosystem processes.
3232 J&J Climate Change
1/11/07
9:56 am
Page 38
38 Climate Change and Adaptation
options for protecting migratory pathways: expand the existing reserve network or ensure that the matrix (in other words those areas outside of formal reserves) is sufficiently protected by measures that do not require state ownership and exclusive use of the landscape for conservation objectives. The time-slice methodology of Williams et al (2005) provides a way of identifying key areas that need conservation to ensure the movement of these species and also for identifying those species that will require facilitated dispersal. For facilitated obligatory dispersers, the only option for maintaining wild populations is to physically move the species to the new suitable habitat (Hossell et al, 2003). While this has been undertaken to reintroduce large mammal and bird species to locations of their historic occurrence or to increase genetic exchange, the introduction of plant and invertebrate species to places where they probably did not exist within the recorded past is a new concept. Facilitated dispersal will have ethical and practical considerations such as follows: • •
• •
What is the number of individual organisms per species that need to be moved to establish a new viable population, and how should individuals for translocation be selected (Heywood and Iriondo, 2003)? Under what circumstances should a species be moved to an area where it did not historically exist, and what impact will this have on the species currently occurring in that area (or which will occur there naturally as a consequence of climate change) (Sakai et al, 2001; Hossell et al, 2003; Radosevich et al, 2003)? Which species need to be moved together, in order to preserve the community structure? How is the pattern of genetic variability within the population to be maintained?
Conservation of no-hopers For the no-hopers, the only nonfatalistic option is to maintain the biodiversity in artificial situations such as zoos, botanical gardens, seed banks and through cryopreservation, in the hope of perhaps introducing them to the wild at some distant future time. Such ex situ conservation practices are also a wise ‘insurance policy’ for species with some hope of surviving in the wild.
The threat of invasive species Some persisters, autonomous dispersers and facilitated dispersers are likely to become ‘weeds’, in other words overabundant in their new habitats, to the detriment of other species (McDonald, 1994). We will need to reconsider the concept of invader species given climatic change. The most likely candidates to invade are primary succession species that are well adapted to dispersal into new habitats. Weed outbreaks will be further encouraged by the disruption of communities in the receiving environment, directly or indirectly due to climate change, and by the possibility that the invasive species will travel faster than their natural competitors and controlling agents (Malcolm and Markham,
3232 J&J Climate Change
1/11/07
9:56 am
Page 39
Adapting Conservation Strategies to Climate Change in Southern Africa 39
2000). Range expansion is a potential threat to existing species in the new areas, and may indirectly prevent their survival in that habitat (even if they can persist from a climatic perspective). Additionally, climate change may well favour introduced exotic species, increasing their chance of becoming invasive. A more aggressive control of invasive species may therefore be needed.
Interventions to facilitate biotic adaptation From Figure 2.2, it is clear that no-hopers and facilitated obligatory dispersers require direct human intervention to prevent extinction. For the survival of the remaining species, conserving key areas of distribution both now and in the future and maintaining permeability of migratory pathways between protected areas would be necessary. For the autonomous obligatory dispersers the strategic combination of both ensuring conservation outside of protected areas (matrix management) and reconfiguring or expanding the conservation area would be the most effective.
Economic Considerations Relating to Adaptation Options The costs and benefits of the various adaptation options discussed above were investigated for the fynbos biome, and particularly for the conservation of members of the Proteaceae. A modelling process was used to identify areas critical for conserving migratory pathways, and to identify disjunct habitats and no-hoper species (Williams et al, 2005). Reserve expansion was found to be a very expensive option if it is used as the only mechanism of protection. Reserve costs comprise the costs of land acquisition and the annual cost of land management. Both operational costs and land management costs per unit area decrease substantially as reserve size increases.4 Therefore, from a cost-efficiency perspective, a few large reserves are better than many small reserves (Frazee et al, 2003; Balmford et al, 2003). In most cases contractual reserves (on private land) are more cost-effective than forming state reserves (Pence et al, 2003; A. Letsoalo, personal communication, 2005). The opportunity cost of managing the land for conservation is typically low where land is presently used for extensive rangelands or for dryland grain production but high where land is used for high value crops.5 In the latter case a formal reserve would be more cost effective (A. Letsoalo, personal communication, 2005). In many instances, rangeland management is already biodiversity-friendly to many species, and conservation may be achieved with little or no increased cost to the rancher. Where dryland cropping is involved, a spatially explicit strategic approach would be needed to ensure that viable biodiversity corridors are achieved. Where facilitated translocation becomes necessary, the cost is dependent on the number of organisms translocated and the establishment costs involved. Simultaneous translocation of communities of mutually interdependent organisms may have to be considered, including pollinators and seed dispersers in the case of plants.
3232 J&J Climate Change
1/11/07
9:56 am
Page 40
40 Climate Change and Adaptation
Gene-banking and other ex situ conservation will not achieve the same level of biodiversity conservation as in situ conservation, but will remain a fall-back position when other opportunities are not available, as well as an insurance measure when they are. Where in situ conservation typically targets the conservation of at least 10 per cent of the historical population, ex situ conservation only conserves a small number of organisms for each species. Gene fingerprinting to ensure that the collection represents the broader population is therefore a significant cost consideration. Table 2.4 compares the relative economic advantages and disadvantages of the different conservation strategies.
Planning and Design Considerations in Implementing Adaptation Options for Biodiversity Conservation Considerations for migratory corridors In general, the movement of species will be poleward or to higher altitudes in response to global warming, but it will also be affected at the local level by changes in precipitation and microclimatic influences (Gitay et al, 2001). Species are expected to respond individually, and gradually, per generation. Therefore, resources in any designated migratory corridor must be sufficient to sustain a lifecycle, not just an individual passing through (Simberloff et al, 1992). Both Halpin (1997) and Noss (2001) have emphasized the need for firm ecological evidence on which to base corridor and buffer zone design. Convincing ecological evidence can only be obtained by collating explicit studies on habitat use and habitat preference for a large number of species in any particular ecosystem. A key development in this field is the comprehension of the spatially explicit nature of habitat use. However, for effective corridor design, we need to understand fluxes of organisms and matter in the landscape in a spatially explicit manner. The intuitive ecological advantages of wildlife corridors suffer from a lack of empirical supporting evidence (Saunders et al, 1991; Simberloff et al, 1992).6 Connectivity and corridor design in a landscape with varying habitat suitability depends on the definition of habitat for a particular species. Any analysis must account for a large number of species, or groups of species, and the variables that influence habitat selection for each of them. An alternative approach is to use processes in landscapes as spatial planning units, and design reserves and corridors to maintain local and regional processes.7 The assumption is that if the processes thought to be responsible for the observed heterogeneity are preserved, then heterogeneity will be maintained in the face of climate change. However, apart from knowledge of previous disturbance events, the measure of the heterogeneity to be maintained is unknown. This level of heterogeneity has been termed functional heterogeneity in the context of savanna herbivore assemblages (Owen-Smith, 2004).
3232 J&J Climate Change
1/11/07
9:56 am
Page 41
Adapting Conservation Strategies to Climate Change in Southern Africa 41
Reconfiguring the reserve network The benefits of existing formal conservation areas can be enhanced by ensuring that they are well configured to best conserve biodiversity given the impacts of climate change, for example, by the conservation of potential refugia, environmental gradients and likely migratory corridors as adaptations to the current reserve network. Systematic conservation planning can now provide land-parsimonious algorithms to prioritize new areas quantitatively for addition to the existing reserve network (Pressey and Taffs, 2001; Pressey et al, 2000 and 2001; Reyers, 2004; Rodriguez et al, 2004). The inclusion of a climate change component is, however, still in its infancy (Cowling et al, 2003; Hannah et al, 2002a and b; Williams et al, 2005). In many situations, current reserve networks are poorly planned to conserve current biodiversity patterns, let alone the additional requirements stemming from climate change. In a first for southern Africa, Williams et al (2005) developed a method based on time-slice analysis of potential climate change-induced species migrations to understand how best to locate conservation areas in the fynbos biome. For the Proteacea species in this study area and a 50-year time frame, they recommend an approximate doubling of the current reserve network to achieve the required level of conservation in a changing climate, although some of this also reflects the inadequacy of existing reserve networks to conserve current biodiversity. Despite many limitations and assumptions, this study provides a powerful tool for objectively considering climate change impacts in reserve planning.
Managing areas outside of reserves (the matrix) Matrix management should be a complementary activity to formal conservation, and one way to do this is by the creation of contractual reserves outside of formal reserves (Pence et al, 2003). Changes to legislation in South Africa now make it possible for the state to enter into a contractual arrangement with landowners to ensure conservation (Pence et al, 2003). This is potentially cheaper than outright purchase of land and, for many landowners, non-agricultural activities such as ecotourism and wildlife ranching are economically attractive because they provide better returns, especially in drier areas. Areas outside formal reserves generally contain a significant portion of their biodiversity, often indeed more than in the reserves (Rodrigues et al, 1999). For instance, R. Biggs, B. Reyers, and R. J. Scholes (2006) estimated that 80 per cent of South Africa’s biodiversity is outside of formally protected areas, despite the high levels of degradation and land transformation. It has been shown that protecting 10 per cent of the land area in the savanna landscape, as per IUCN (1993) guidelines, may only represent 60 per cent of species in an area and exclude up to 65 per cent of rare and endangered species (Reyers et al, 2002). In fact, up to 50 per cent of the land area may be needed to preserve a representative portion of species (Soule and Sanjayan, 1998). Although South Africa has only 5.4 per cent of its land area under state conservation, it is estimated that an additional 13 per cent is currently managed
3232 J&J Climate Change
1/11/07
9:56 am
Page 42
42 Climate Change and Adaptation
as private wildlife ranches (Bond et al, 2004, updated from Cumming, 1999). Not all game-ranching practices automatically result in improved biodiversity conservation, but it is argued that, on balance, greater biodiversity benefits are achieved through this land use versus alternative agricultural practices (Taylor, 1974; Child, 1988; Bond et al, 2004). However, biodiversity is most threatened in the higher-rainfall areas where crop agriculture or forestry are more economically attractive options. In these areas, greater direct intervention may be needed to maintain biodiversity and migratory corridors. The landscape comprising the boundaries and edges between conserved areas and the matrix is considered dynamic and permeable to water, matter, species and energy fluxes (Saunders et al, 1991).8 These spatial linkages of energy, matter and species fluxes across edges provide additional support for biodiversity-friendly, matrix management as part of formal reserve management. When considering the likely impacts of climate change on biodiversity,9 matrix management practices need to anticipate an increased movement of species through the landscape, and therefore connectivity between suitable habitat patches is important. This connectivity may translate into buffer zones around existing suitable patches or linear corridor features that link suitable patches (in this chapter fragmented landscapes have been taken as a given and important component for consideration in conservation planning). For implementing matrix management for species movement an integrated procedure for determining land use, based on robust ecological evidence, is needed. Buyin from local stakeholders is also critical since the decision to use or not use any piece of land will affect individuals. To achieve an effective climate adaptation strategy for biodiversity using matrix management, both of the following options are considered important (adapted from Frazee et al, 2003): 1
2
Strategic conservation of critically important areas of the matrix: These are areas identified as critically important for conservation, but cannot be included into the formal conservation network for financial or other reasons. In these circumstances, the state can enter into a contractual agreement with the landowner that the land be managed for conservation purposes and the farmer could be compensated based on the opportunity cost of not undertaking the next best agricultural practice. General enhancements to biodiversity conservation on all non-reserve land: In this instance, less costly incentives could be used to promote more biodiversity-friendly farming practices. This could include incentives as discussed below for commercial land or the establishment of CBNRM in the communal areas.
Policy mechanisms for facilitating biodiversity conservation within the matrix Matrix management for biodiversity conservation while sustaining economic benefits may involve a variety of strategies such as setting aside riparian strips or woodland corridors, reducing the use of pesticides and fertilizers, reducing
3232 J&J Climate Change
1/11/07
9:56 am
Page 43
Adapting Conservation Strategies to Climate Change in Southern Africa 43
animal stocking rates, or reintroducing necessary disturbances such as fire. The wrong mix of land uses in the matrix can be inimical to conservation, for instance by increasing alien plant invasion, or by causing a retreating forest edge (Gascon et al, 1999). Due to poorly developed markets for ecological services, there is minimal incentive for landowners to promote biodiversity or maintain migratory corridors. Perverse policies for protecting threatened species that disadvantage landowners may even result in them deliberately reducing biodiversity on their land. Land tenure is important for developing matrix management interventions and a different set of incentives and approaches would be applicable for private as opposed to communal lands. Incentives for matrix management on privately held land Shogren et al (2003) and Doremus (2003) suggest the following policy and economic incentive systems for promoting biodiversity on private land: • • • •
• •
Education: Many landowners have a conservation ethic, and if educated about the pertinent issues, may change land management practices to meet biodiversity conservation needs, provided costs are low. Direct incentives: Positive incentives include cash payments, zero rating land tax on key conservation areas (Pence et al, 2003) or debt forgiveness. An example of negative incentives is taxes for poor land use. Approval and recognition: Competitive awards for conservation activities can be an incentive. An example, here is the landowner-targeted programme to promote raptor breeding in Kimberley, South Africa. Market creation or improvement: The state can create markets for environmental services, for example carbon credits, promotion of ecotourism, provision of information on markets, and the introduction of certification schemes (such as ‘badger friendly’ honey). Tradable development rights: Landholders are granted a fixed amount of tradable development rights. This creates a market value for resources. Regulatory control: The enactment and enforcement of laws, including the types of social prohibitions that served this function historically.
Inappropriate agricultural subsidies need to be removed. For example, large direct and indirect subsidies previously made cattle ranching economically viable (Child 1988; Bond et al, 2004), but now wildlife management is a better option for the arid and semi-arid areas in South Africa, Namibia and Zimbabwe. Matrix management on communal land The management of shared resources is referred to as ‘common property resource management’. A common property resource is defined as any resource that is subject to individual or group use but not to individual ownership and is used under some arrangement of community or group management (Mol and Wiersum, 1993). Despite concerns about overexploitation of such communal resources (Hardin, 1968), there is evidence that degradation is not an inevitable outcome of group management (Bromley and Cernea, 1989; Lawry, 1990; Ostrom 1992). A number of criteria have been identified under
3232 J&J Climate Change
1/11/07
9:56 am
Page 44
44 Climate Change and Adaptation
which group management is most likely to be successful (see, for example, Baland and Platteau, 1996; IFAD, 1995; Ostrom, 1992; Wade, 1987; Lawry, 1990; Cousins, 1996; Shackleton et al, 2002). Changes in human population density and resource use patterns have resulted in the evolution of the communal property resource management theory into a new paradigm of community-based natural resource management (CBNRM). Most southern African states now have some form of CBNRM programme (Murphy, 1997; Fabricius et al, 2004), partly due to support from official development aid agencies.10 Early CBNRM programmes emphasized the need to devolve ownership and management to the lowest possible level but it was later recognized that though important, this devolution of power on its own was not sufficient to initiate successful CBNRM. Even though the CBNRM programmes have not always been successful, Fabricius et al (2004) believe they can potentially achieve both community development and increased sustainability of natural resources. They identify seven principles paramount to a sustainable CBNRM: 1 2 3 4 5 6 7
a diverse and flexible range of livelihood options is maintained; the production potential of the resource base is maintained or improved; institutions for local governance and resource management are in place and effective; economic and other benefits to provide an incentive for wise use of resources exist; there are effective policies and laws, these are implemented, and the authority is handed down to the lowest level where there is the capability to apply it; there is sensible and responsible outside facilitation; and local-level power relations are favourable to CBNRM and are understood.
In southern Africa, the principles of CBNRM are presently being implemented in all of the transfrontier parks; the Wild Coast Initiative in South Africa; Administrative Management Design for Game Management Areas (Zambia); Communal Areas Management Program for Indigenous Resources (CAMPFIRE) (Zimbabwe); Community-Based Natural Resource Management Program in Conservancies (Namibia); and Community-Based Natural Resource Management Program in Controlled Hunting Areas (Botswana).
Conclusions The conservation network in southern Africa, although extensive, is poorly configured to adequately conserve biodiversity, even less so under a climatically changing environment. The largest proportion of biodiversity is still found outside of the reserve areas, despite the impacts of land transformation and degradation. With the anticipated impacts of climate change over the next 50 to 200 years, many species will have to move from their current locations to track areas with suitable climates. To facilitate this process and minimize species loss,
3232 J&J Climate Change
1/11/07
9:56 am
Page 45
Adapting Conservation Strategies to Climate Change in Southern Africa 45
a multitude of strategies will be necessary, such as realignment of reserves; ensuring that land use outside of reserves is biodiversity friendly; facilitated translocations, where species are unable to move on their own; and ex situ conservation as a precautionary measure and for species with no future habitats. In a climatically changing environment, strategic conservation becomes a shifting target, and it is therefore important to protect the migratory corridors and not simply a single end point. It is also important to realize that the entire habitat will not move, but rather individual species will move at different rates, creating new habitat structures. Cost considerations and difficulties in acquiring new areas for reserves may, however, pose barriers in reconfiguring the reserve network. The most cost-effective mechanism to both conserve biodiversity and allow species movement to new habitats is to ensure the permeability of areas between reserves to species migration. Economically, formal reserve expansion is viable where opportunity costs of alternative land use options are high, where large areas are involved, where there are clearly defined gradients needing protection, and where high levels of biodiversity loss can be prevented through reserve realignment. Contractual reserves should be considered for more marginal areas. Within southern Africa, the mechanisms to ensure biodiversity-friendly management of the matrix are likely to differ significantly between privately owned areas and communal lands. Direct incentives such as tax rebates, assistance with vegetation management and education may be sufficient to change behaviour on private land. Allowing private ownership of wildlife has greatly increased the extent of private game ranches. Contractual reserves, where the state compensates private land owners to manage portions of their farms as areas for biodiversity conservation, are also an option. On communal land, practices based on CBNRM principles are the likely solutions. Biodiversity-friendly practices may also be promoted by the removal of distorted market forces that encourage inappropriate agricultural practices. Use of management tools such as fire and grazing intensity (including grazing by mega-herbivores, such as elephants), can help maintain habitat functionality in a state similar to the present. It is the landscapes profoundly transformed for crop production that pose the greatest challenges for biodiversity conservation. Where the movement of species to new suitable habitats must be facilitated, of greater concern will be the movement of smaller animals, insects and plants. For species with no suitable future habitats, ex situ conservation is the only option to prevent extinction, although it can only conserve small populations of individual species and does not conserve ecological function. A radical change in current thinking about conservation planning will therefore be necessary since simply maintaining the status quo might result in species extinction under a climatically changed future. Ongoing monitoring, research and model improvement will be necessary to better understand the response of biodiversity to a changing climate and address the present uncertainties. Fortunately, there are many areas in which our current understanding is sufficient for us to begin planning for biodiversity conservation in a climatically changing environment.
3232 J&J Climate Change
1/11/07
9:56 am
Page 46
46 Climate Change and Adaptation
Notes 1
Species typically tend to occupy their ‘realized niche’, which is a subset of their ‘fundamental niche’ (the range determined by their physiological tolerance limits) resulting from the outcome of interactions with other species. The degree to which species distribution can be predicted based on their climatically defined habitat niche differs between species (see, for example, Thuiller et al, 2006). 2 Studies show that the fertilization effect of increasing atmospheric concentration of CO2 (primarily responsible for the current anthropogenic climate change) starts to saturate in natural ecosystems at around 500 ppm (Scholes et al, 1999). 3 The C3 plants produce a three-carbon compound in the photosynthetic process and include most trees and common crops like rice, wheat, barley, soybeans, potatoes and vegetables. The C4 category of plants produce a four carbon compound in the photosynthetic process and include grasses and crops like maize, sugar cane, sorghum and millet. Under increased atmospheric concentrations of CO2, C3 plants have been shown to be more responsive than C4 plants (see IPCC, 2001). 4 On the basis of South African National Park data, a 1km2 park has a US$104,793 annual operational cost, while a 100,000 km2 park only costs US$66/km2 (Martin, 2003). The land management cost per hectare decreases nonlinearly as the reserve size increases. 5 The cost of managing the land for conservation is the opportunity cost of lost income to the farmer for not using the land for the most profitable alternative land use activity. This cost will be very low for extensive rangelands, low for dryland grain production, but high for irrigated crops and speciality crops such as horticulture. 6 An often-stated example of the usefulness of corridors is riparian vegetation. However, Simberloff et al (1992) state that riparian vegetation does not constitute a typical corridor from a management point of view, as it is a unique habitat in itself that happens to be linear, and it does not connect discrete patches of like habitat. 7 An excellent example of using such processes in conservation planning is found in Rouget et al (2003a and b). 8 The process of forming such a landscape has been termed habitat variegation (McIntyre and Barrett, 1992) and Murphy and Lovett-Doust (2004) have expressed a similar viewpoint that a binary approach of suitable habitat versus the matrix is not a true reflection of landscape dynamics. 9 Biodiversity responses to climate change may take a variety of forms, and our current ability to predict this is limited due to uncertainties in both the climate scenarios and in how species will react to the change (reviewed by Walther et al, 2002; McCarty, 2001; Hughes, 2000; Parmesan and Yohe, 2003; Root et al, 2003). 10 The Communal Areas Management Programme for Indigenous Resources (CAMPFIRE) was initiated in Zimbabwe in the early 1980s as one of the first experiments in this regard.
References ABSA (2003) Game Ranch Profitability in Southern Africa, ABSA Group Economic Research, South Africa Financial Sector Forum, Rivonia, South Africa, available at www.finforum.co.za/absa/investment.html#game%20ranch%20profitability Baland, J. and J. Platteau (1996) Halting Degradation of Natural Resources. Is there a Role for Rural Communities?, Food and Agricultural Organization and Clarendon Press, Oxford, UK Balmford, A., L. Moore, T. Brooks, N. Burgess, L. A. Hansen, P. H. Williams and C.
3232 J&J Climate Change
1/11/07
9:56 am
Page 47
Adapting Conservation Strategies to Climate Change in Southern Africa 47 Rahbek (2001) ‘Conservation conflicts across Africa’, Science, vol 291, pp2616–2619 Balmford, A., K. J. Gaston, S. Blyth, A. James and V. Kapos (2003) ‘Global variation in terrestrial conservation costs, conservation benefits, and unmet conservation needs’, Proceedings of the National Academy of Sciences, vol 100, pp1046–1050 Biggs, R., B. Reyers and R. J. Schole (2006) ‘A biodiversity intactness score for South Africa’, South African Journal of Science, vol 102, pp277–283, July/August Bond, I., B. Child, D. de la Harpe, B. Jones, J. Barnes and H. Anderson (2004) ‘Private land contribution to conservation in South Africa’, in B. Child (ed) Parks in Transition: Biodiversity, Rural Development and the Bottom Line, IUCN, SASUSG and Earthscan, London and Sterling, VA Bromley, D. W. and M. M. Cernea (1989) ‘The management of common property natural resources: Some conceptual and operational fallacies’, World Bank discussion papers, no 57, World Bank, Washington, DC Büscher, B. (2005) ‘Conjunctions of governance: The state and the conservation-development nexus in southern Africa’, Journal of Transdisciplinary Environmental Studies, vol 4, pp1–15 Bush, M. B. (2002) ‘Distributional change and conservation on the Andean flank: A palaeoecological perspective’, Global Ecology and Biogeography, vol 11, pp463–473 Child, B. (1988) ‘The role of wildlife utilization in the sustainable development of semiarid rangelands in Zimbabwe’, D.Phil. thesis, Oxford University, Oxford, UK Child, B. (ed) (2004) Parks in Transition: Biodiversity, Rural Development and the Bottom Line, Earthscan, London Cousins, B. (1996) ‘Livestock production and common property struggles in South Africa’s agrarian reform’, The Journal of Peasant Studies, vol 23, pp166–208 Cowling, R. M., R. L. Presseyb, M. Rougetc and A. T. Lombarda (2003) ‘A conservation plan for a global biodiversity hotspot—The Cape Floristic Region, South Africa’, Biological Conservation, vol 112, pp191–216 Cowling, R. M. and R. L. Pressey (2003) ‘Introduction to systematic conservation planning in the Cape Floristic Region, Biological Conservation, vol 112, pp1–13 Cubasch, U., G. A. Meehl, G. J. Boer, R. J. Stouffer, M. Dix, A. Noda, C. A. Senior, S. Raper and K. S. Yap (2001) ‘Projections of future climate change’, in J. T. Houghton, Y. Ding, D. J. Griggs, M. Noguer, P. van der Linden, X. Dai, K. Maskell and C. I. Johnson (eds) Climate Change 2001: The Scientific Basis, Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, New York, pp525–582 Cumming, D. (1999) ‘Study on the development of transboundry natural resource management areas in Southern Africa, environmental context: Natural resources, land use, and conservation’, Biodiversity Support Program, Washington, D.C. Davis, M. B. and R. G. Shaw (2001) ‘Range shifts and adaptive responses to 36 Quaternary climate change’, Science, vol 292, pp673–679 Doremus, H. (2003) ‘A policy portfolio approach to biodiversity protection on private lands’, Environmental Science and Policy, vol 6, pp217–232 Dudley, N. and S. Stolton (2003) ‘Ecological and socioeconomic benefits of protected areas in dealing with climate change’, in L. J. Hansen, J. L. Biringer and J. R. Hoffmans (eds) Buying time: A User’s Manual for Building Resistance and Resilience to Climate Change in Natural Systems, World Wildlife Fund, Gland, Switzerland Fabricius, C., E. Koch, H. Maome and S. Turner (eds) (2004) Rights Resources and Rural Development: Community-based Natural Resource Management in Southern Africa, Earthscan, London and Sterling, VA Frazee, S., R. M. Cowling, R. L. Pressey, J. K. Turpie and N. Lindenberg (2003) ‘Estimating the costs of conserving a biodiversity hotspot: A case study of the Cape Floristic region, South Africa’, Biological Conservation, vol 112, pp275–290 Gascon, C., T. Lovejoy, R. O. Bierregaard, J. R. Malcolm, P. C. Stouffer, H. L.
3232 J&J Climate Change
1/11/07
9:56 am
Page 48
48 Climate Change and Adaptation Vasconcelos, W. F. Laurance, B. Zimmerman, M. Tocher and S. Borges (1999) ‘Matrix habitat and species richness in tropical forest remnants’, Biological Conservation, vol 91, pp223–229 Gaston, K. J. (2003) The Structure and Dynamics of Geographic Ranges, Oxford Series in Ecology and Evolution, Oxford University Press, New York Gelbard, J. L. (2003) ‘Grasslands at a crossroads: Protecting and enhancing resilience to climate change’, in L. J. Hansen, J. L. Biringer and J. R. Hoffmans (eds) Buying Time: A User’s Manual for Building Resistance and Resilience to Climate Change in Natural Systems, World Wildlife Fund, Gland, Switzerland Gitay, H., S. Brown, W. Easterling and B. Jallow (2001) ‘Ecosystems and their goods and services’, in J. J. McCarthy, O. F. Canzini, N.A. Leary, D.J. Dokken and K.S. White (eds) Climate Change: Impacts, Adaptations, and Vulnerability, Cambridge University Press, Cambridge, pp235–342 Gitay, H., A. Suarez, R. Watson and D. J. Dokken (eds) (2002) Climate Change and Biodiversity, IPCC Technical Paper V, Intergovernmental Panel on Climate Change, Geneva, Switzerland Graham, R. W. (1992) ‘Late Pleistocene faunal changes as a guide to understanding effects of greenhouse warming on mammalian fauna of north America’, in R. L. Peters and T. Lovejoy (eds) Global Warming and Biological Diversity, Yale University, New Haven, CT, pp76–87 Graham, R. W. and E. C. Grimm (1990) ‘Effects of global climate change on the patterns of terrestrial biological communities’, Trends in Ecology and Evolution, vol 5, pp289–292 Halpin, P. N. (1997) ‘Global climate change and natural-area protection: Management responses and research directions’, Ecological Applications, vol 7, pp828–843 Hannah, L., G. F. Midgley, T. Lovejoy, W. J. Bond, M. Bush, J. C. Lovett, D. Scott and F. I. Woodward (2002a) ‘Conservation of biodiversity in a changing climate’, Conservation Biology, vol 16, pp264–268 Hannah, L., G. F. Midgley and D. Millar (2002b) ‘Climate change-integrated conservation strategies’, Global Ecology and Biogeography, vol 11, pp485–495 Hardin, G. (1968) ‘The tragedy of the commons’, Science, vol 162, pp1243–1248 Hewitson, B. C. and R. G. Crane (2006) ‘Consensus between GCM climate change projections with empirical downscaling’, International Journal of Climatology, in press Heywood, V. H. and J. M. Iriondo (2003) ‘Plant conservation: Old problems, new perspectives’, Biological Conservation, vol 113, pp321–335 Hossell, J.E., B. Briggs and I. R. Hepburn (2003) Climate Change and UK Nature Conservation: A Review of the Impact of Climate Change on UK Species and Habitat Conservation Policy, Department of Transport, Environment and the Regions Wildlife and Countryside Directorate, Bristol, UK Hughes, L. (2000) ‘Biological consequences of global warming: Is the signal already apparent?’, Trends in Ecology and Evolution, vol 15, pp56–61 Huntley, B. (1991) ‘How plants respond to climate change: Migration rates, individualism and the consequences for plant communities’, Annals Botany., vol 67 (Supplement 1), pp15–22 Hutton, J., W. M. Adams and J. C. Murombedzi (2005) ‘Back to the barriers? Changing narratives in biodiversity conservation’, Forum for Development Studies, vol 32, pp341–370, Norsk Utenrikspolitisk Institutt, www.nupi.no IFAD (International Fund for Agricultural Development) (1995) Common Property Resources and the Rural Poor in Sub-Saharan Africa, Special Programme for SubSaharan African Countries Affected by Drought and Desertification, International Fund for Agricultural Development., IFAD, Rome IPCC (Intergovernmental Panel on Climate Change) (2001) ‘Glossary of terms’, in J. McCarthy, O. Canziani, N. Leary, D. Dokken and K. White (eds) Climate Change
3232 J&J Climate Change
1/11/07
9:56 am
Page 49
Adapting Conservation Strategies to Climate Change in Southern Africa 49 2001: Impacts, Adaptation, and Vulnerability, Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US IUCN (1993) Parks for Life, report of the Fourth World Congress on National Parks and Protected Areas, IUCN, Gland, Switzerland Lawry, S. W. (1990) ‘Tenure policy towards common property natural resources in subSaharan Africa’, Natural Resources Journal, vol 30, pp403–422 Lindenmayer, D. B. and H. Nix (1993) ‘Ecological principles for the design of wildlife corridors’, Conservation Biology, vol 7, pp627–630 Low, B. and A. G. Rebelo (eds) (1996) Vegetation of South Africa, Lesotho and Swaziland, Department of Environmental Affairs and Tourism, Pretoria, South Africa Malcolm, J. R. and A. Markham (2000) Global Warming and Terrestrial Biodiversity Decline, World Wildlife Fund, Washington, D.C. MA (2005) ‘Millennium ecosystem assessment’, www.millenniumassessment.org Margules, C. R. and R. L. Pressey (2000) ‘Systematic conservation planning’, Nature, vol 405, pp243–253 Martin, R. (2003) ‘Annual recurrent expenditure for conservation and management: A preliminary investigation of the use of a spreadsheet model to generate staff establishment and operating budget for individual parks’, South African National Parks, Pretoria, South Africa McCarty, J. P. (2001) ‘Ecological consequences of recent climate change’, Conservation Biology, vol 15, pp320–331 McDonald, I. A. W. (1994) ‘Global change and alien invasion, implications for biodiversity and protected area management’, in O. T. Solbrig, P. G. van Emden and W. J. van Oordt (eds) Biodiversity and Global Change, CAB International, Wallingford, Oxon, UK McIntyre, S. and G. W. Barrett (1992) ‘Habitat variegation: An alternative to fragmentation’, Conservation Biology, vol 6, pp146–147 Midgley, G. F., L. Hannah, D. Millar, W. Thuiller and A. Boot (2002a) ‘Developing regional species-level assessments of climate change impacts on biodiversity in the Cape Floristic region’, Biological Conservation, vol 112, pp87–97 Midgley, G. F., L. Hannah, D. Millar, M. C. Rutherford and L. W. Powrie (2002b) ‘Assessing the vulnerability of species richness to anthropogenic climate change in a biodiversity hotspot’, Global Ecology and Biogeography, vol 11, pp445–451 Midgley, G. F., L. Hannah, D. Millar, W. Thuiller and A. Boot (2003) ‘Developing regional and species-level assessments of climate change impacts on biodiversity: A preliminary study in the Cape Floristic region’, Biological Conservation, vol 112, pp87–97 Mol, P. W. and K. F. Wiersum (1993) ‘Common forest resource management: Asia’, in D. A. Messerschmidt (ed) Common Forest Resource Management: Annotated Bibliography of Asia, Africa and Latin America, Community Forestry Note No 11, Food and Agricultural Oganization of the United Nations, Rome Murphy, H. T. and J. Lovett-Doust (2004) ‘Context and connectivity in plant metapopulations and landscape mosaics: Does the matrix matter?’ OIKOS, vol 105, pp3–14 Murphy, M. W. (1997) ‘Congruent objectives, comparing interests and strategic compromise: Concepts and processes in the evolution of Zimbabwe’s CAMPFIRE programme’, Community Conservation in Africa Working Paper No 2, Institute for Development Policy and Management (IDPM),University of Manchester, Manchester, UK Myers, N., R. A. Mittermeier, C. G. Mittermeier, G. A. B. da Fonseca and J. Kent (2000) ‘Biodiversity hotspots for conservation priorities’, Nature, vol 403, pp853–858 Naidoo, R. and W. L. Adamowicz (2006) ‘Modeling opportunity costs of conservation in transitional landscapes’, Conservation Biology, vol 20, pp490–500 Noss, R. F. (2001) ‘Beyond Kyoto: Forest management in a time of rapid forest change’, Conservation Biology, vol 15, pp578–590
3232 J&J Climate Change
1/11/07
9:56 am
Page 50
50 Climate Change and Adaptation Olson, D. M., E. Dinerstein, E. D. Wikramanayake, N. D. Burgess, G. V. N. Powell, E. C. Underwood, J. A. D’amico, I. Itoua, H. E. Strand, J. C. Morrison, C. J. Loucks, T. F. Allnutt, T. H. Ricketts, Y. Kura, J. F. Lamoreux, W. W. Wettengel, P. Hedao and K. R. Kassem (2001) ‘Terrestrial ecoregions of the world: A new map of life on Earth’, BioScience, vol 51, pp933–938 Orme, C. D. L., R. G. Davies, M. Burgess, F. Eigenbrod, N. Pickup, V. A. Olson, A. J. Webster, T. S. Ding, P. C. Rasmussen, R. S. Ridgely, A. J. Stattersfield, P. M. Bennett, T. M. Blackburn, K. J. Gaston and I. P. F. Owens (2005) ‘Global hotspots of species richness are not congruent with endemism or threat’, Nature, vol 436, pp1016–1019 Ostrom, E. (1992) ‘The rudiments of a theory of the origins, survival, and performance of common property institutions’, in D. W. Bromley (ed) Making the Commons Work: Theory, Practice, Policy, Institute for Contemporary Studies, San Francisco, CA Owen-Smith, N. (2004) ‘Functional heterogeneity in resources within landscapes and herbivore population dynamics’, Landscape Ecology, vol 19, pp761–771 Parmesan, C. and G. Yohe (2003) ‘A globally coherent fingerprint of climate change impacts across natural systems’, Nature, vol 421, pp37–42 Pence, G. Q. K., M. A. Botha and J. K. Turpie (2003) ‘Evaluating combinations of onand off-reserve conservation strategies for the Agulhas Plain, South Africa: A financial perspective, Biological Conservation, vol 112, pp253–273 Pressey, R. L., C. J. Humphries, C. R. Margules, R. I. Vane-Wright and P. H. Williams (1993) ‘Beyond opportunism: Key principles for systematic reserve selection, Trends in Ecology and Evolution, vol 8, pp124–128 Pressey, R. L. and K. H. Taffs (2001) ‘Scheduling conservation action in production landscapes: priority areas in western New South Wales’, Biological Conservation, vol 100, pp355–376 Pressey, R. L., R. M. Cowling and M. Rouget (2001) ‘Formulating conservation targets for biodiversity pattern and process in the Cape Floristic region, South Africa’, Biological Conservation, vol 112, pp99–127 Pressey, R. L, T. C. Hager, K. M. Ryan, J. Scharz, S. Wall, S. Ferrier and P. M. Creaser (2000) ‘Using abiotic data for conservation assessments over extended regions: Quantitative methods applied across New South Wales, Australia’, Biological Conservation, vol 96, pp55–82 Pressey, R. L. and R. M. Cowling (2001) ‘Reserve selection algorithms and the real world’, Conservation Biology, vol 15, pp275–277 Radosevich, S. R., M. M. Stubbs and C. M. Ghersa (2003) ‘Plant invasions – Processes and patterns’ Weed Science, vol 51, pp254–259 Reyers, B. (2004) ‘Incorporating anthropogenic threats into evaluations of regional biodiversity and prioritisation of conservation areas in the Limpopo Province, South Africa’, Biological Conservation, vol 118, pp521–531 Rodrigues, A. S. L., R. Tratt, B. D. Wheeler and K. J. Gaston (1999) ‘The performance of existing networks of conservation areas in representing biodiversity’, Proceedings of the Royal Society, London, vol B 266, pp1453–1460. Rodrigues, A. S. L., S. J. Andelman, M. I. Bakarr, L. Boitani, T. M. Brooks, R. M. Cowling, L. D. C. Fishpool, G. A. B. Da Fonseca, K. J. Gaston, M. Hoffmann, J. S. Long, P. A. Marquet, J. D. Pilgrim, R. L. Pressey, J. Schipper, W. Sechrest, S. N. Stuart, L. G. Underhill, R. W. Waller, M. E. J. Watts and X. Yan (2004) ‘Effectiveness of the global protected area network in representing species diversity’, Nature, vol 428, pp640–643 Root, T. L., J. T. Price, K. R. Hall, S. H. Schneider, C. Rosenzweig and A. J. Pounds (2003) ‘Fingerprints of global warming on wild animals and plants’, Nature, vol 421, pp57–60 Rouget, M., D. M. Richardson, R. M. Cowling, J. W. Lloyd and A. T. Lombard (2003a) ‘Current patterns of habitat transformation and future threats to biodiversity in
3232 J&J Climate Change
1/11/07
9:56 am
Page 51
Adapting Conservation Strategies to Climate Change in Southern Africa 51 terrestrial ecosystems of the Cape Floristic region, South Africa’, Biological Conservation, vol 112, pp63–85 Rouget, M., D. M. Richardson and R. M. Cowling (2003b) ‘The current configuration of protected areas in the Cape Floristic region, South Africa – Reservation bias and representation of biodiversity patterns and processes’, Biological Conservation, vol 112, pp129–145 Rutherford. M. C., G. F. Midgeley, W. J. Bond, L. W. Powrie, R. Roberts and L. Allsopp (1999) Plant Biodiversity: Vulnerability and Adaptation Assessment. South African Country Study on Climate Change, National Botanical Institute, Cape Town, South Africa Reyers, B., D. H. K. Fairbanks, K. J. Wessels and A. S. Van Jaarsveld (2002) ‘A multicriteria approach to reserve selection: Addressing long-term biodiversity maintenance’, Biodiversity and Conservation, vol 11, pp769–793 Sakai, A. K., F. W. Allendorf, J. S. Holt, D. M. Lodge and K. A. J. Molofsky (2001) ‘The population biology of invasive species’, Annual Review of Ecology and Systematics, vol 32, pp305–332 Saunders, D. A., R. J. Hobbs and C. R. Margules (1991) ‘Biological consequences of ecosystem fragmentation: A review’, Conservation Biology, vol 5, pp18–32 Scholes R. J., E. D. Schulze, L. F. Pitelka and D. O. Hall (1999) ‘The biogeochemistry of terrestrial ecosystems’, in B. H. Walker, W. L. Steffen, J. Canadell and J. S. I Ingram (eds) The Terrestrial Biosphere and Global Change, Cambridge University Press, New York, pp88–105 Scholes, R. J. and R. Biggs (eds) (2004) The Regional Scale Component of the Southern African Millennium Ecosystem Assessment, CSIR, Pretoria, South Africa Shackleton, S., G. P. von Maltitz and J. P. Evans (2002) ‘Factors, conditions and criteria for the successful management of natural resources held under a common property regime: A South African perspective’, occasional report no 8, Plaas School of Government, University of the Western Cape, South Africa Shogren J. F., G. M. Parkhurst and C. Settle (2003) ‘Integrating economics and ecology to protect nature on private lands: Models, methods, and mindsets’, Environmental Science and Policy, vol 6, pp233–242 Simberloff, D., J. A. Farr, J. Cox and D. W. Mehlman (1992) ‘Movement corridors: Conservation bargains or poor investments?’ Conservation Biology, vol 6, pp6493–6505 Soule, M. E. and M. A. Sanjayan (1998) ‘Conservation targets: Do they help?’ Science, vol 279, pp2060–2061 Taylor, R. D. (1974) ‘A comparative study of land use on the cattle and game ranch in the Rhodesian lowveld’, MSC thesis, University of Rhodesia, Harare, Zimbabwe Thomas, C. D., A. Cameron, R. E. Green, M. Bakkenes, L. J. Beaumont, Y. C. Collingham, B. F. N. Erasmus, M. F. de Siqueira, A. Grainger, L. Hannah, L. Hughes, B. Huntley, A. S. van Jaarsveld, G. F. Midgley, L. Miles, M. A. OrtegaHuerta, A. Townsend Peterson, O. L. Phillips and S. E. Williams (2004) ‘Extinction risk from climate change’, Nature, vol 427, pp145–148 Thuiller, W., O. Broennimann, G. Hughes, J. R. W. Alkemade, G. F. Midgley and F. Corsi (2006) ‘Vulnerability of African mammals to anthropogenic climate change under conservative land transformation assumptions’, Global Change Biology, vol 12, pp12,424–12,444 van der Linde, H., J. Oglethorpe, T. Sandwith, D. Snelson and Y. Tessema (with contributions from A. Tiéga and T. Price) (2001) Beyond Boundaries: Transboundary Natural Resource Management in Sub-Saharan Africa, Biodiversity Support Program, Washington, DC von Maltitz, G. P. and S. E. Shackleton (2004) ‘Use and management of forests and woodlands in South Africa: Stakeholders, institutions and processes from past to present’, in M. J. Lawes, H. A. C. Ealey, C. M. Shackleton and B. G. S. Geach (eds)
3232 J&J Climate Change
1/11/07
9:56 am
Page 52
52 Climate Change and Adaptation Indigenous Forests and Woodlands in South Africa, Policy, People and Practice, University of KwaZulu-Natal Press, Scottsville, South Africa Wade, R. (1987) ‘The management of common property resources: Collective action as an alternative to privatisation or state regulation’, Cambridge Journal of Economics, vol 11, pp95–106 Walther, G. R., E. Post, P. Convey, A. Menzel, C. Parmesan, T. J. C. Beebee, J. M. Fromentin, O. Hoegh-Guldberg and F. Bairlein (2002) ‘Ecological responses to recent climate change’, Nature, vol 416, pp389–395 WDPA (2005) 2005 World Database on Protected Areas, Centre for Applied Biodiversity Science, Conservation International, Washington, DC Williams, P., L. Hannah, S. Andelman, G. Midgley, M. Araujo, G. Hughes, L. Manne, E. Marinez-Meyer and R. Pearson (2005) ‘Planning for climate change: Identifying minimum-dispersal corridors from the Cape Proteaceae’, Conservation Biology, vol 19, pp1063–1074 Wilshusen, P. R., S. R. Brechin, C. L. Fortwangler and P. C. West (2002) ‘Reinventing a square wheel: Critique of a resurgent “protection paradigm” in international biodiversity conservation’, Society and Natural Resources, vol 15, pp17–40
3232 J&J Climate Change
1/11/07
9:56 am
Page 53
3
Benefits and Costs of Adapting Water Planning and Management to Climate Change and Water Demand Growth in the Western Cape of South Africa John M. Callaway, Daniël B. Louw, Jabavu C. Nkomo, Molly E. Hellmuth and Debbie A. Sparks
Introduction The Berg river basin, located in the Western Cape Region of South Africa, provides the bulk of the water for household, commercial and industrial use in the Cape Town metropolitan region as well as irrigation water to the lower part of the basin to cultivate roughly 15,000 hectares of high-value crops. Since the early 1970s, water consumption in municipal Cape Town has grown at an average annual rate of about 3 per cent. As the population of the metropolitan Cape Town region grows, the competition for water in the basin has become intense, and farmers have responded by dramatically improving their irrigation efficiencies and shifting even more land into the production of high-value export crops. Meanwhile, over the past two decades, a number of national and regional commissions have been set up to investigate options for coping with the long-term water supply problems in the basin. One outcome of these efforts was the authorization of the Berg River Dam. In June 2004, after almost 20 years of debate about its economic feasibility and environmental impacts, final agreement was reached on construction of the dam. It will consist of a 130.1 million cubic metre storage reservoir and a pumping site to pump water from below the dam back to it. The dam is expected to be operational sometime during the period 2008–2010. The region has also recently experienced a number of unusually dry years, the most recent in the summer of 1994–1995, when peak storage in the upper basin was only about one-third of average. At the same time, concerns about the effects of global warming on basin runoff have been growing, along with suggestions that recent climatic anomalies may be associated with regional climate change. Not surprisingly, planning for the Berg River Dam and other
3232 J&J Climate Change
1/11/07
9:56 am
Page 54
54 Climate Change and Adaptation
water supply and demand options in the basin has, up until this point, failed to take into account the possibility that the build-up of greenhouse gases in the global atmosphere is already affecting and will continue to affect the regional climate by reducing runoff in the basin. The context for our analysis of water planning and management in the Berg river basin consists of three main elements. The first is the increasing competition for water between urban and agricultural water users because of growing urban water demands; second is the threat of unusual climate variability and/or climate change to exacerbate that competition; and third is the planning and policy responses to these issues. To address these three elements, we developed a policy-planning model that can be used to evaluate a wide range of structural, non-structural and technological measures for coping with basin water shortages due to demand growth, climate variability and climate change. The model we have developed to do this is called the Berg River Dynamic Spatial Equilibrium Model (BRDSEM) and is described in detail in Callaway et al (2006). In this chapter we describe very briefly the model; illustrate some of the ways the model can be used to assess the benefits, costs and risks of avoiding climate change damages by increasing the maximum storage capacity of the reservoir and/or implementing a system of efficient water markets; and present the results and major conclusions of our analysis for three deterministic climate scenarios. We also describe the limitations of the current version of the model and analysis methods and outline future plans for improving the model and analytical methods.
The Model BRDSEM is a dynamic, multiregional, nonlinear programming model patterned after the hydro-economic surface water allocation models developed by Vaux and Howitt for California (1984), Booker (1990) and Booker and Young (1991 and 1994) for the Colorado River Basin, and Hurd et al (1999 and 2004) for the Missouri, Delaware and Apalachicola-Flint-Chattahoochee river basins in the United States. This type of model is a more specific application of spatial and temporal price and allocation models that originated from the work of Samuelson (1951) and Takayama and Judge (1971), which have been widely applied in many natural resource sectors (McCarl and Spreen, 1980). Hydroeconomic models have been used by Hurd et al (1999 and 2004) to estimate the economic impacts of climate change in the four large US river basins. In these two studies, the authors provided estimates for the economic losses due to climate change that took into account only short-run adjustments to climate variability. They did not specifically assess the benefits of long-run measures for avoiding climate change damages, such as investments in storage capacity and changes in water allocation institutions and laws. Building on this past work, BRDSEM was designed specifically to estimate the economic value of the damages due to climate change with and without long-run adaptation measures and thereby to isolate the benefits and costs of avoiding climate change damages. This chapter represents the first attempt, to
3232 J&J Climate Change
1/11/07
9:56 am
Page 55
Benefits and Costs of Adapting Water Planning and Management 55
the best of our knowledge, to quantify the benefits of avoiding climate change damages in economic terms in a water basin context using such a model. BRDSEM is also an extension of a static spatial equilibrium model developed by Louw (2002) for the Berg river basin to examine the potential of water markets in the region (Louw and Van Schalkwyk, 2001). Much of the data from that model is used in BRDSEM, but is being updated on a continuous basis. Significant modifications were made to Louw’s model to add the spatial relationships between runoff, water storage, water conveyance, transfers, return flows, and water use in the natural and man-made hydrological systems and to account for the intertemporal aspects of reservoir operation in the upper and lower parts of the basin. One of the important features added to BRDSEM is that it can determine, endogenously, the optimal (in other words economically efficient) capacity of planned reservoirs and other structural works. The model does this by satisfying an economic efficiency (Kuhn-Tucker) condition, namely that the marginal capital cost associated with the optimal storage capacity is equal to the present value of the shadow price of storage in all periods where storage levels are at this maximum. The maximum capacity is determined in year one for all future periods (i.e. perfect foresight) and remains fixed thereafter. The structure of the BRDSEM model is shown in Figure 3.1. The core of the model consists of four main elements linked together in a nonlinear mathematical programming framework. These include a nonlinear (quadratic) objective function that characterizes the normative objectives of the agents in the model; an intertemporal, spatial equilibrium module that characterizes the spatially distributed flow of water and water storage in the basin; an urban water demand and supply module; and a regional farm irrigation demand module. External information inputs to BRDSEM include downscaled climate scenarios from a regional climate model; monthly runoff, surface water evaporation coefficients and crop water use adjustment factors from a regional hydrologic model (WATBAL, Yates, 1996); and information about policies, plans and technologies used to alter various parameters in the programming model to reflect alternative demand- and supply-side choices and constraints. Outputs of the model include measures of the economic value of water, optimal storage capacity of the reservoir, shadow prices for water transfers, monthly reservoir storage, releases and transfers, and monthly water diversions and consumptive use by urban and irrigation users by region.
Model Application: Methods and Scenarios Economic framework and methods The economic framework used in this chapter for evaluating the costs and benefits of measures to avoid climate change damages was first presented in Callaway et al (1998). It was extended to link adaptation to climate variability and climate change and to situations in which ‘regrets’ occur when the climate that is realized, ex post, is not the same as the climate planners and policy
3232 J&J Climate Change
1/11/07
9:56 am
Page 56
56 Climate Change and Adaptation
Figure 3.1 The Berg River Spatial Equilibrium Model (BRDSEM)
makers anticipated in formulating and implementing their plans and policies, ex ante (Callaway, 2003 and 2004a and b). Table 3.1 illustrates the basic framework for estimating various benefits and costs associated with climate change and adaptation for a simple case of two climate states, the existing climate (C0) and climate change (C1), and a single long-run adaptation option, investments in reservoir storage capacity, K(C), which is climate sensitive. The framework can be extended to multiple climate states and measures, including those that are not sensitive to climate in both deterministic and stochastic settings (Callaway, 2004b). The value of the net returns to water in each of the four cells, represented as W[C, K(C)], depends on the climate state and on the reservoir storage capacity, which is, in turn, determined, in part, by the climate state. From an ex ante planning perspective, the upper left cell characterizes welfare for the existing climate, with current water storage capacity adapted to the existing climate. The cell in the upper right corresponds to the case in which the
3232 J&J Climate Change
1/11/07
9:56 am
Page 57
Benefits and Costs of Adapting Water Planning and Management 57
climate changes but planners fail to plan for it. Economic agents make ex ante short-run partial adjustments to the climate change, but there is no new investment in reservoir storage (or institutional change). Note that water users and managers can adapt to many forms of extreme climate variability without knowing whether the climate is changing or not, by using short-run measures. For example, even if reservoir storage capacity is fixed at K0, operating policies can be changed, cities can institute water restrictions and farmers can change their management. But unless people can detect climate change or are confident that it will occur, or estimate that the costs of planning for climate change are small if it does not occur, generally they will not undertake long-run measures. The cell in the lower right depicts the ex ante long-run welfare consequences that take place when economic agents fully adjust using long-run measures to adapt to the expected climate. The cell in the lower left, depicts an ex ante situation where water storage capacity is increased in anticipation of climate change that does not occur. Table 3.1 Framework for estimating benefits and costs associated with climate change and climate change adaptation Ex Ante Climate Planning Options
Ex Post Climate States(C) Existing Climate (C0)
Climate Change (C1)
Plan for current climate, reservoir capacity = K(C0)
Existing climate with current reservoir capacity. Net returns to water = W[C0, K(C0)]
Do not plan for climate change that occurs. Net returns to water = W[C1 K(C0)]
Plan for climate change, reservoir capacity = K(C1)
Plan for climate change that does not occur. Net returns to water = W[C0,K(C1)]
Plan for climate change that occurs. Net returns to water = W[C1, K(C1)]
Table 3.1 can also be used to measure the economic costs, or regrets, of implementing planning decisions based on ex ante climate expectations that are wrong in terms of ex post climate outcomes. There are two kinds of regrets: those associated with caution, which involve not planning for climate change that is occurring, and those associated with precaution, which involve planning for climate change that is not occurring. In both cases, the net returns to water will be lower than if the ex post climate realization matched the ex ante climate expectation. In other words, regrets lead to costs. Estimates of the costs of regrets can be made without assigning any explicit probabilities to different climate states but can also be cast in a Bayesian framework when better information is available (Lindgren, 1968). From Table 3.1, we construct the following definitions of climate change damages, net benefits of adaptation, and ex post costs of errors of caution and precaution. Climate change damages (CCD) are the ex ante welfare losses caused by climate change when economic agents only make short-run adjustments to climate variability or climate change (in other words they do not adjust reser-
3232 J&J Climate Change
1/11/07
9:56 am
Page 58
58 Climate Change and Adaptation
voir capacity) to cope with climate change compared to the reference case and are calculated as: CCD = W[C1, K(C0)] – W[C0, K(C0)]. Net benefits of adaptation (NBA) are the ex ante welfare gains associated with reducing climate change damages by implementing long-run measures (in other words adjusting the reservoir capacity) compared to the partial adjustment case: NBA = W[C1, K(C1)] – W[C1, K(C0)]. Imposed climate change damages (ICCD) are the ex ante welfare losses relative to the reference case that cannot be avoided by implementing long-run measures. They correspond to the residual climate change damages that are not avoided by full adaptation and are calculated as: ICCD = W[C1, K(C1)] – W[C0, K(C0)]. Cost of precaution (CP) is the ex post welfare loss that would occur as a result of planning for climate change that does not occur: CP = W[C0, K(C1)] – W[C0, K(C0)]. Cost of caution (CC) is the ex post welfare loss that would occur as a result of not planning for climate change that does occur. It is equivalent to the net benefits of adaptation, with the sign reversed: CC = W[C1, K(C0)] – W[C1 K(C1)].
Policy, climate and water demand scenarios Two examples of adaptation policy scenarios or measures that are relevant and important to the basin are selected for analysis: building and optimally sizing the Berg River Dam and replacing the existing regulatory framework for allocating water in the basin with a system of efficient water markets. As already noted, the dam is under construction and expected to be in operation soon. Implementing water markets is a policy that is currently being investigated in the context of the new national water law. Existing water use regulation is represented in the BRDSEM model by placing upper bounds on summer and winter diversions by the seven regional farms as in Louw (2001 and 2002) and on water diversions from storage sites at Theewaterskloof and Wemmershoek, consistent with current water allocation policy. Efficient water markets are introduced to the model by removing both the agricultural and urban allocation constraints to simulate the economically efficient allocation of water to both urban and agricultural users.
3232 J&J Climate Change
1/11/07
9:56 am
Page 59
Benefits and Costs of Adapting Water Planning and Management 59
We simulated the welfare consequences of the different policy options under both full (optimal) and partial adjustment to three deterministic, transient climate scenarios. A detailed explanation of the climate scenarios, the downscaling from the global circulation models to the basin weather stations, and the generation of the monthly hydrologic data used by BRDSEM is discussed in detail in Hellmuth and Sparks (2005) and Nkomo et al (2006). The hydrology-climate scenarios are constructed with information provided by WATBAL and derived from downscaled outputs of the CSIRO general circulation model experiment for the SRES B2 scenario of greenhouse gas emissions. Three different time-slices of the CSIRO experiment are used to construct our scenarios: 1961–1990 for the reference case (REF), 2010–2039 for the near future case (NF) and 2070–2099 for the distant future case (DF). Relative to the reference case, average annual runoff in the basin is reduced by 10.7 and 22.0 per cent in the NF and DF scenarios respectively. While the climate scenarios are time dependent and correspond to specific years, we apply all of the scenarios to the same period, 2010–2039, to avoid having to develop even more uncertain future scenarios for the 2070–2099 period. Thus REF is a counterfactual reference case, assuming the same underlying runoff as in the period 2010–2039 as 1961–1990, while DF, instead of being a longer-term continuation of CSIRO B2, can be viewed as a more adverse climate scenario, producing lower runoff and higher evaporation, compared to NF for the same time period. CSIRO climate experiments for A2 and B2 emission scenarios project reductions in runoff for the region relative to the reference period for both the NF and DF time periods, with the B2 scenarios being slightly more severe. In comparison, projections from the Hadley Centre general circulation model are mixed: the A2 scenarios show virtually no changes in runoff for the region in the NF and DF time periods, whereas the B2 scenarios show reductions in runoff in the NF time period, but increases in the DF time period. Our initial economic analysis uses only the CSIRO projections for the SRES B2 emissions scenario. There are two reasons for this. First, these scenarios showed the most severe effects on basin runoff, yet appear to be more consistent with recent trends. Evaluating severe, yet plausible, scenarios is useful for benchmarking the upper range of potential impacts. Second, the long solution times of the model in conjunction with the large number of runs forced us to limit the analysis for the time being. Agricultural area in the basin has been relatively stable for the last halfdecade and is not expected to grow much more due to limited land availability (Louw 2001 and 2002). Most of the irrigated land in the basin is already under drip irrigation and the potential for efficiency gains is quite small. Dryland agriculture is basically limited to areas without access to irrigation water and should conditions become drier and hotter, this lack of access is likely to drive this land out of production. For these reasons, we did not make any exogenous changes to the irrigated land base in the model. The only changes in water demand for irrigation are those that are introduced by WATBAL for changes in crop water use factors in response to changes in potential evapotranspiration and changes in irrigation efficiencies in response to shadow prices that are determined endogenously by BRDSEM.
3232 J&J Climate Change
1/11/07
9:56 am
Page 60
60 Climate Change and Adaptation
However, urban water consumption in Cape Town has been growing rapidly at an average annual rate of about 3 per cent since the mid-1970s (Louw 2001 and 2002). In the real world and in our model, increases in water consumption over time will increase net returns to water as the sectoral water demand curves shift to the right. In other words, development increases welfare. But if we are to correctly estimate the costs and benefits associated with climate change and adaptation, we need to be able to separate not only the effects of climate from the effects of development, but also the adjustments to development from the adjustments to climate as a result of the implementation of various adaptation options. To illustrate how one can control for development and climate separately, we developed two urban water demand scenarios by adjusting the slopes of the monthly urban water demand functions using consumption data provided by the Cape Metropolitan Council (CMC). The first scenario is a no-demand growth case in which water demand curves are held fixed. The second scenario shifts water demand curves such that, at a constant water price, the quantity of water demanded by urban users grows at a rate of 3 per cent per year.
Results Table 3.2 shows the simulated present values for the total net returns to water for four policy options under three climate change and two urban water demand scenarios. The policy options are (A) water is allocated by existing regulations and no dam is constructed, (B) water is allocated by efficient water markets and no dam is constructed, (C) water is allocated by existing regulations and the dam is constructed and optimally sized for the ex post climate and water allocation regime, and (D) water is allocated by efficient water markets and the dam is constructed and optimally sized for the ex post climate and water allocation regime. The table also presents the optimal storage capacity of the Berg River Dam for options C and D. All of the net returns to water depicted in this table represent the optimum values that can be achieved by a policy option for each climate and urban demand scenario. Economic values in this chapter are calculated as constant South African rand, based on the year 2000 (Louw and Van Schalkwyk, 2001; Louw, 2002). This assumes that all input and output prices in the model are inflating at the same, constant rate. A constant, real discount rate of 6 per cent is used to convert future value flows into constant present values. In sensitivity trials, reducing (increasing) the discount rate had predictable effects on water use, increasing (reducing) future consumption and, thus, increasing (reducing) the endogenously determined maximum optimal storage capacity of the Berg River Dam. In all four policy cases, the net returns to water decrease as the annual average annual runoff in the climate scenarios decreases, holding urban water demand growth constant, and increase as urban water demand growth increases, holding climate constant. Moreover, the percentage reductions in the
3232 J&J Climate Change
1/11/07
9:56 am
Page 61
Benefits and Costs of Adapting Water Planning and Management 61
Table 3.2 Net returns to water and optimal storage capacity of the Berg River Dam Climate Scenarios No Urban Demand Growth
Ave. Annual Runoff (m3 – 106) (% Reduction from REF)
REF1
NF1
DF1
REF1
NF1
DF1
75.5
67.4 (–10.7)
58.9 (–22.0)
75.5
67.4 (–10.7)
58.9 (–22.0)
Present Value of Net Returns to Water2 (R 109)
Allocation/Reservoir Scenarios A. Existing water use regulation No Berg Dam (% Reduction from REF)
Climate Scenarios 3% Urban Demand Growth Per Year
58.7
55.8 (–5.1)
52.8 (–10.1)
B. Efficient water markets No Berg Dam 59.8 (% Reduction from REF)
56.3 (–5.9)
53.1 (–11.2)
C. Existing water use regulation Optimal storage for Berg Dam (% Reduction from REF)
55.8 (–5.0)
56.8 (–5.1)
58.7
D. Efficient water markets Optimal storage for Berg Dam 59.9 (% Reduction from REF)
58.0 (–22.6)
44.93 (–40.1)
92.0
83.9 (–8.7)
76.5 (–16.8)
53.6 (–8.9)
90.1
76.9 (–14.7)
62.7 (–30.5)
54.4 (–9.2)
96.3
89.7 (–6.9)
83.4 (–13.3)
74.9
Optimal Berg River Dam Capacity m3 106
Options C
0
15
69
151
272
240
D
6
84
109
138
128
178
REF: reference climate scenario, CSIRO model simulation for 1961–1990; NF: near future scenario, CSIRO model simulation for period 2010–2039 for SRES B2 emission case; DF: distant future scenario, CSIRO model simulation for period 2070–2099 for SRES B2 emission case. All climate scenarios are assumed to apply to 2010–2039 period. 2 All monetary estimates are expressed as present values in billions of constant rand (R) for the year 2000, discounting over 30 years at a real discount rate of 6 per cent. 3 This scenario was unfeasible because the urban regulation constraints were tight; consequently their right-hand side values were increased by 5 per cent in every year. 1
net returns to water become much sharper as urban demand growth increases. This directly raises the important question of how we can disentangle the effects of urban water demand growth from climate change in assessing these two options.
3232 J&J Climate Change
1/11/07
9:56 am
Page 62
62 Climate Change and Adaptation
The results in Table 3.2 also show a consistent relationship between the net returns to water and the type of allocation method. Allocation by water markets in cases B and D produces higher net returns than allocation by regulations in cases A and C. This should not be surprising, as BRDSEM is an optimization model and the only differences between these two pairs of scenarios is that A and C are more highly constrained than B and D. More importantly, as water demand increases, switching to efficient water markets reduces the economic impacts of climate change, compared to the existing regulated system. In the no urban demand growth cases, the reductions in the net returns to water from the base case are around 5 per cent for the DF climate and 10 per cent for the RF climate for both allocation systems. However, when urban demand growth increases substantially, allocation by markets serves to moderate the decreases in the net returns to water as climate change becomes more severe. Thus, simulating the substitution of efficient water markets in case B for the existing allocation system in case A reduces the percentage of welfare losses in basin-wide welfare from –22.6 per cent (NF) and –40.1 per cent (DF) in case A to –8.7 per cent (NF) and –16.8 per cent (DF) in case B. For cases C and D, the corresponding reductions in the percentage changes in basin welfare due to market substitution are –14.7 per cent (NF) and –30.5 per cent (DF) for case C (existing allocation system) and –6.9 per cent (NF) and –13.3 per cent (DF) for case D (water markets). The last conclusions demonstrate forcefully that simulated water markets have a substantial moderating effect on the impact of climate change on basinwide welfare as urban demand growth increases. But this is also true of dams, since the absolute values for the net returns to water are higher, and the percentage reductions due to climate change are smaller when storage capacity is added behind the Berg River Dam compared to the no dam cases, when urban demand growth is 3 per cent per year. The fact that both dams and water markets influence the effects of climate change on basin-wide welfare again raises the issue of how we separate out both the effects of climate change and economic development on basin-wide welfare and the effects of these measures on climate change and development. However, before we do this, we want to take one last look at the optimal reservoir capacity simulated for cases C and D due to climate change and development (see Table 3.3). One thing is clear: as urban development increases, the need for additional storage capacity not only increases, but is also economically feasible. Nevertheless, two partly counter-intuitive results need further explanation. First, under the no urban demand growth scenario, optimal reservoir capacity is higher for water markets (case D) than for the existing allocation system (case C). This is because, at low urban water demand levels in the upper basin, reducing consumption is expensive in terms of consumer welfare losses, and the implementation of markets increases the marginal present value of storage capacity (or the opportunity cost of the lack of storage) more than the capital cost in this capacity range, leading to higher capacity levels in the market situation. As urban demand growth increases, markets tend to become better substitutes for storage capacity, because the reductions in consumption (from very high prices) cost less, at the margin, than the
3232 J&J Climate Change
1/11/07
9:56 am
Page 63
Benefits and Costs of Adapting Water Planning and Management 63
additions to storage capacity at these very high storage capacity levels. Second, the pattern of storage capacity additions under high urban water demands is nonlinear with respect to climate change and the nature of this nonlinearity differs depending on the water allocation method. For option C, with higher water demand growth, the optimal storage capacity of the Berg River Dam peaks under the NF climate scenario and then falls from 272,000m3 to 240,000m3. This is because the marginal yield of the reservoir system due to increases in capacity falls sharply as runoff is reduced and development increases, leading to problems in filling the reservoir under the DF scenario and reducing the benefits of additional capacity. For option D at the same level of water demand, the optimal capacity decreases under the NF scenario and then rises. This is because lower basin runoff in the NF scenario varies in such a way relative to the upper basin that water can be moved around by transfers much more easily to satisfy demand than by adding storage capacity, but this is not the case for the existing water allocation system. Table 3.3 presents the welfare results for incrementally adapting storage capacity of the dam to development pressure and climate change under the existing regulatory system for allocating water. Adaptation is presented in two steps, first to development pressure and then to climate change by adding the economically efficient amount of water storage capacity in both steps. The upper part of Table 3.3 shows results for partially and fully adjusting reservoir capacity to development pressure if there is no change in climate. Development, represented by 3 per cent annual growth in urban water demand, with no addition to reservoir capacity would raise net returns to water from R58.7 billion to R74.9 billion. The net increase in welfare, R16.196 billion, represents the ex ante net benefits of adapting to development pressure, holding climate constant at REF, and only allowing short-run adjustments to take place in response to urban water demand growth. If basin planners correctly anticipated and fully adjusted to this development pressure by adding 151 million m3 of storage capacity, net returns to water would increase to R90.1 billion. This increase of R15.2 billion represents the ex ante net benefits of optimally (instead of partially) adjusting (or adapting) to development pressure. This same amount, but in minus also represents the ex ante, ex post cost of not planning for development that does occur, or the cost of caution associated with development. If, on the other hand, the Berg River Dam was built at this level of capacity, but no increase in water demand growth occurred, ex post basin-wide welfare would drop to R58.3 billion, which is R400 million less than the base case. This net welfare loss is the ex ante, ex post cost of planning for development pressure that does not occur, or the cost of precaution associated with development pressure. The second part of Table 3.3 shows results for partially and fully adapting to climate change for two climate scenarios, NF (by adding 272 million m3 of storage capacity) and DF (by adding 240 million m3 of storage capacity), under the existing water allocation system. The reference case used for measuring the effects of adapting to climate change is the post-development adjustment in reservoir capacity to 151 million m3 and the corresponding welfare level of
3232 J&J Climate Change
1/11/07
9:56 am
Page 64
64 Climate Change and Adaptation
Table 3.3 Adapting to development pressure and climate change under the existing water allocation system: net returns to water (present value, R billion) I. Adjustment to Development: Change Storage Capacity Ex Post Scenario: No Development and 3% Development Ex Ante Action: Change in Berg Reservoir Capacity Only
No Development, REF Climate 3%
Development, REF Climate
Partial adjustment to development Capacity = 0
58.7
74.9
Full adjustment to development Capacity = 151 106m3
58.3
90.1
II. Adjustment to Climate Change: Change Storage Capacity Ex Post Scenario: REF Climate and NF Climate Ex Ante Action: Change in Berg Reservoir Storage Capacity Only
3% Development, REF Climate
3% Development, NF Climate
Partial adjustment to NF climate change Capacity = 151106m3
90.1
76.7
Full adjustment to NF climate change Capacity = 272106m3
89.9
76.9
Ex Post Scenario: REF Climate and DF Climate Ex Ante Action: Change in Berg River Storage Capacity Only
3% Development, REF Climate
3% Development, DF Climate
Partial adjustment to DF climate change Capacity = 151 106m3
90.1
62.5
Full adjustment to DF climate change Capacity = 240 106m3
90.0
62.7
R90.1 billion. If the climate changes and the rate of urban demand growth is 3 per cent and reservoir capacity is not enlarged from 151 million m3, simulated net returns to water fall to R76.7 billion for the NF climate scenario (10.7 per cent reduction in average annual runoff) and R62.5 billion for the DF climate scenario (22.0 per cent reduction in average annual runoff). The resulting welfare losses of R13.4 billion (NF) and R27.6 billion (DF) represent the ex ante values of climate change damages. However, if basin storage capacity is increased from 151 million m3 to 272 million m3 (optimal for NF) or 240 million m3 (optimal for DF), ex ante net returns to water would increase to R76.9 billion and R62.7 billion respectively. The increases in net welfare of R0.209 billion (NF) or R0.198 billion (DF) represent the ex ante values for the
3232 J&J Climate Change
1/11/07
9:56 am
Page 65
Benefits and Costs of Adapting Water Planning and Management 65
net benefits of adaptation under the two different climate scenarios, with urban demand growth at 3 per cent. But these adaptation benefits are quite small relative to the value of climate change damages, so the simulated, ex ante imposed climate change damages are quite large, namely R13.2 billion (NF) and R27.4 billion (DF). Following on from this analysis, we can also estimate the ex ante, ex post cost measures associated with climate regrets. As already stated, the cost of not planning for climate change that does occur (caution) are the same as the net benefits of adaptation but in minus, –R0.209 billion (NF) or –R0.198 billion (DF). If the Berg River Dam is built but the climate does not change to NF or DF, then the ex post value of the net returns to water would be R89.9 billion (NF) or R90.0 billion (DF), which are R0.204 billion (NF) and R0.142 billion (DF) less than the reference case of R90.1 billion. These are the estimated costs of precaution associated with the two climate change scenarios. In Table 3.4 we show the results for adapting to development pressure and climate change by substituting efficient water markets for the existing allocation system and increasing the storage capacity. Adaptation is again represented in sequential steps, first for adjusting to development pressure alone and then for adjusting to the change in climate to either NF or DF scenarios. This time, the adjustment to development is decomposed into two separate steps: first by switching to efficient water markets and then by adjusting the storage capacity to be optimal for efficient water markets. Part I of Table 3.4 shows that, if urban water demand growth increases at 3 per cent per year, the ex ante net present value of welfare in the basin will increase from R58.7 billion to R74.9 billion, without any long-run adjustments by basin planners. Thus the ex ante net benefits of increased water demand growth are R16.2 billion. If basin planners and managers adjust to this development pressure by instituting a system of efficient water markets, but do not build a storage reservoir, ex ante basin welfare will further increase to R92.0 billion. If a storage reservoir is also built, and its capacity is optimal for the market allocation system and 3 per cent urban water demand growth, then ex ante basin welfare rises even further to R96.3 billion. The net benefits of adjusting to the development pressure by switching to a system of efficient water markets to cope with the development pressure are R17.1 billion in step A. Adjusting the reservoir capacity to 138 million m3 in step B so that it is optimal for the higher level of water demand growth and the new allocation system produces another R4.3 billion in ex ante basin welfare, or R21.4 billion compared to the reference case. Since the cost of caution is the cost of avoiding an error of caution (in other words not anticipating the higher level of development pressure), one just needs to reverse the signs on the estimates for the net benefits of development. If efficient water markets are implemented, ex ante, but the ex post level of demand growth is zero, net welfare in the basin would be R59.8 billion and, if 138 million m3 of storage capacity were developed, ex ante, but the ex post level of demand growth remains at zero, basin welfare would increase just slightly from that to R59.6 billion. These welfare levels are actually higher than the reference case. Thus, the resulting costs of precaution would be ex post
3232 J&J Climate Change
1/11/07
9:56 am
Page 66
66 Climate Change and Adaptation
Table 3.4 Adapting to development pressure and climate change by switching to water markets and adding storage capacity: net returns to water (present value, R billion) I. Adjustment to Development: Efficient Water Markets Ex Post Scenario: No Development and 3% Development Ex Ante Action: Change Allocation System Only Change Storage Capacity
No Development, REF Climate
3% Development, REF Climate
Partial adjustment to development: Existing Regulations Capacity = 0106m3
58.7
74.9
A. Full adjustment to development: Efficient Water Markets Capacity = 0106m3
59.8
92.0
B. Full adjustment to development: Efficient Water Markets Capacity = 138106m3
59.6
96.3
II. Adjustment to Climate Change: Efficient Water Market + Change in Storage Capacity Ex Post Scenario: REF Climate and NF Climate Ex Ante Action: Change in Berg Reservoir Storage Capacity with Efficient Water Markets in Place
3% Development, REF Climate
3% Development, NF Climate
Partial adjustment to NF climate change Capacity = 138106m3
96.3
89.7
Full adjustment to NF climate change Capacity = 128106m3
96.3
89.7
Ex Post Scenario: REF Climate and DF Climate Ex Ante Action: Change in Berg Reservoir Storage Capacity with Efficient Water Markets in Place
3% Development, REF Climate
3% Development, DF Climate
Partial adjustment to DF climate change Capacity = 138106m3
96.3
83.4
Full adjustment to DF climate change Capacity = 178106m3
96.3
83.5
benefits (not costs) of R1.1 billion and R0.9 billion, illustrating nicely the ‘no regrets’ character of these adjustments to development pressure. The second part of Table 3.4 presents results for partially and fully adapting to climate change for two climate scenarios, NF and DF, under the revised water allocation system via markets. The adjustments can be interpreted in the
3232 J&J Climate Change
1/11/07
9:56 am
Page 67
Benefits and Costs of Adapting Water Planning and Management 67
same way as part II of Table 3.3, and so we will highlight only the major differences in the two strategies as shown in the results from Tables 3.3 and 3.4. There are three main conclusions that can be drawn from these tables. First, the market-oriented strategy (Table 3.4) increases basin welfare by a larger amount over the whole range of development and climate impacts and adjustments than the strategy relying on the existing allocation system (Table 3.3). The welfare difference is roughly R12.8 billion (NF) to R20.8 billion (DF), in favour of the market-oriented strategy, depending on the climate scenario. The net benefits of adjusting to development pressure using the market-oriented strategy are R1.8 billion to R6.2 billion higher than those of the other strategy, depending on whether one just switches to efficient water markets or does this and also adjusts the reservoir capacity for the change in development pressure and water allocation system. Second, the development adjustment step in the market-oriented strategy also reduces the level of climate change damages by more than 50 per cent compared to the other strategy for both the RF and DF climates. After the adjustment to development pressure, but before the adjustment to climate change, the level of climate change damages is R6.8 billion (NF) to R14.8 billion (DF) lower than for the other strategy, depending on the climate change scenario. As a result, switching to water markets acts like a ‘no regrets’ insurance policy against having to adjust further to climate change. Third, once water markets are instituted, the net benefits of adapting to climate change (in the market-oriented strategy) are quite small both in absolute terms and compared to the other strategy. This is because of the large reduction in climate change damages that occurs as a result of adjusting to development pressure. For this reason, and because the costs of both caution and precaution associated with climate change adaptation for the marketadjustment strategy are very low in both absolute terms and compared to the other strategy, the least-risky and most economically efficient option would involve adjusting to development pressure and climate change by implementing a system of efficient water markets (step A) and increasing the reservoir capacity to 128 million m3 (step B), without developing any additional storage capacity until more information is available about climate change. Adding additional storage capacity above this level produces only small net benefits, but also has low costs of caution and precaution.
Limitations in the Analysis and Future Research BRDSEM, like many policy models, is a work in progress. Typically, models like this are never final and undergo numerous revisions as new data become available and new questions are asked. BRDSEM has reached the point in its development where it can be used to illustrate some of its policy uses, as in this chapter. Nevertheless, we would be remiss not to identify its current limitations and how we plan to remedy them. The current version of BRDSEM and its application in this chapter has the following limitations. First, the parameters of the urban water demand
3232 J&J Climate Change
1/11/07
9:56 am
Page 68
68 Climate Change and Adaptation
functions are not estimated from empirical time series and/or panel data, but are fit algebraically using assumed elasticities and estimates of base-level consumption. There is no upward sloping waterworks supply function to capture resource costs in purifying and distributing water to urban consumers, and we lack reliable data regarding the water supply elasticity to even ‘fit’ these functions algebraically. The model lacks a full set of demand- and supply-side options for coping with demand growth, climate variability and climate change. The transactions costs associated with markets are ignored, and the analysis is deterministic with regard to climate scenarios. With future support from all of the elements of the regional and South African water and agricultural resources community, we plan to address the first four of these issues as follows. Parameters of the urban demand and water works supply functions will be estimated with empirical observations, using suitable econometric estimators. We plan to extend the model to the larger Boland Region to include representations of all current and alternative water supply sources, reservoirs and conveyance structures, and users in the region, as well as a larger set of demand- and supply-side options. In the current analysis, we assumed that the transition from the current regulatory system to a system of efficient short- and long-run water markets would be frictionless and without cost. In the future, we plan to include a more realistic treatment of water transfers, including differences between long- and short-term water transfers and transaction costs. The final limitation, the deterministic nature of the illustrative analysis in this paper, requires further consideration and investigation. We used deterministic climate change scenarios because downscaled stochastic climate scenarios do not currently exist for the region. The use of deterministic climate scenarios is adequate for demonstrating the model and the economic framework for estimating the benefits and costs associated with adjusting to climate change. But the results are unreliable because they do not reflect the underlying means, variances and other moments of the partial and joint distributions of relevant meteorological variables at the appropriate spatial scales, nor does the runoff, also generated using the deterministic climate inputs. When such information becomes available, it will be possible to propagate the runoff, evaporation and crop water use distributions through BRDSEM by maximizing the expected value of net returns to water for a single or for mixed climate distributions using the methods illustrated in Callaway (2004b). This will also allow us to explore the economic and physical consequences of runoff sequences that depart from mean values, that are drier and wetter than average periods than reflected in mean runoff. Finally, it will allow us to explore more thoroughly the stochastic nature of ex ante, ex post regrets and the possibility of minimizing these regrets by policies and plans that are flexible over a wide range of mixed runoff distributions.
3232 J&J Climate Change
1/11/07
9:56 am
Page 69
Benefits and Costs of Adapting Water Planning and Management 69
References Booker, J. F. (1990) ‘Economic allocation of Colorado River water: Integrating quantity, quality, and instream use values’, PhD thesis, Department of Agricultural and Resource Economics, Colorado State University, Fort Collins, CO Booker, J. F. and R. A. Young (1994) ‘Modeling intrastate and interstate markets for Colorado River water resources’, Journal of Environmental Economics and Management, vol 26, pp66–87 Booker, J. F. and R. A. Young (1991) Economic Impacts of Alternative Water Allocations in the Colorado River Basin, Report 161, Colorado Water Resources Institute, Fort Collins, CO Callaway, J. M., L. Ringius and L. Ness (1998) ‘Adaptation costs: A framework and methods’, in J. Christensen and J. Sathaye (eds) Mitigation and Adaptation Cost Assessment Concepts, Methods and Appropriate Use, UNEP Collaborating Centre on Energy and Environment, Risø National Laboratory, Roskilde, Denmark Callaway, J. M. (2003) Adaptation Benefits and Costs – Measurements and Policy Issues, Report NV/EPOC/GSP (2003) 10/FINAL, Environment Directorate, Environment Policy Committee, OECD, Organisation for Economic Co-operation and Development, Paris Callaway, J. M. (2004a) ‘Adaptation benefits and costs: Are they important in the global policy picture and how can we estimate them’, Global Environmental Change: Human and Policy Dimensions, vol 14, pp273–282 Callaway, J. M. (2004b) ‘The benefits and costs of adapting to climate variability and change’, in J. C. Morlot and S. Agrawala (eds) The Benefits and Costs of Climate Change Policies: Analytical and Framework Issues, Organisation for Economic Cooperation and Development, Paris Callaway, J. M., D. B. Louw, J. C. Nkomo, M. E. Hellmuth and D. A. Sparks (2006) ‘The Berg River Dynamic Spatial Equilibrium Model: A new tool for assessing the benefits and costs of alternatives for coping with water demand growth, climate variability and climate change in the Western Cape’, AIACC Working Paper No 31, International START Secretariat, Washington, DC Hellmuth, M. E. and D. Sparks (2005) ‘Modeling the Berg river basin: An explorative study of impacts of climate change on runoff’, AIACC Project Completion Report, Project No 47, UNEP Collaborating Centre on Energy and Environment, Risø National Laboratory, Roskilde, Denmark Hurd, B. J., J. M. Callaway, P. Kirshin and J. Smith (1999) ‘Economic effects of climate change on US water resources’, in R. Mendelsohn and J. Neumann (eds) The Impacts of Climate Change on the US Economy, Cambridge University Press, London Hurd, B. J., J. M. Callaway, P. Kirshin and J. Smith (2004) ‘Climatic change and US water resources: From modeled watershed impacts to national estimates’, Journal of the American Water Resources Association, vol 22, pp130–148 Lindgren, B. W. (1968) Statistical Theory, Macmillan, New York Louw, D. B. (2001) ‘Modelling the potential impact of a water market in the Berg river basin’, PhD thesis, University of the Orange Free State, Bloemfontein, South Africa Louw, D. B. (2002) ‘The development of a methodology to determine the true value of water and the impact of a potential water market on the efficient utilisation of water in the Berg River Basin’, Water Research Commission Report (WRC) No 943/1/02, Pretoria, South Africa Louw, D. B. and H. D. Van Schalkwyk (2001) ‘Water markets: An alternative for central water allocation’, Agrekon, vol 39, pp484–494 McCarl, B. A. and T. H. Spreen (1980) ‘Price endogenous mathematical programming as a tool for sector analysis’, American Journal of Agricultural Economics, vol 62, pp88–102
3232 J&J Climate Change
1/11/07
9:56 am
Page 70
70 Climate Change and Adaptation Nkomo, J. C., J. M. Callaway, D. B. Louw, M. E. Hellmuth and D. A. Sparks (2006) Adaptation to Climate Change: The Berg River Basin Case Study, Energy Research Center, University of Cape Town, Cape Town, South Africa Samuelson, P. A. (1952) ‘Spatial price equilibrium and linear programming’, American Economic Review, vol 42, pp283–303 Smith, J. B. and S. S. Lenhart (1996) ‘Climate change adaptation policy options’, Climate Research, vol 6, pp193–201 Takayama, T. and G. G. Judge (1971) Spatial and Temporal Price and Allocation Models, North Holland Press, London Vaux, H. J. and R. E. Howitt (1984) ‘Managing water scarcity: An evaluation of interregional transfers’, Water Resources Research, vol 20, pp785–792 Yates, D. N. (1996) ‘WATBAL: An integrated water balance model for climate impact assessment of river basin runoff’, International Journal of Water Resources Development, vol 12, no 2, pp121–139
3232 J&J Climate Change
1/11/07
9:56 am
Page 71
4
Indigenous Knowledge, Institutions and Practices for Coping with Variable Climate in the Limpopo Basin of Botswana Opha Pauline Dube and Mogodisheng B. M. Sekhwela
Introduction Widespread poverty, high reliance on natural resources and low adaptive capacity contribute to conditions of high vulnerability to climate variability and climate change in developing countries such as Botswana (Kates, 2000; Desanker and Magadza, 2001; Mirza, 2003; Dube and Moswete, 2003). Lack of choice to meet basic needs such as nutrition, shelter and clothing at the household and individual levels in Botswana is attributed to income poverty and capability poverty (Jefferies, 1997). Income poverty is the inability to command the level of income or tangible resources needed to meet basic needs, while capability poverty involves the lack of human capabilities or intangible resources such as education and good health that enables one to escape poverty (Ministry of Finance and Development Planning, 1998). Both income poverty and capability poverty undermine the capacity to adapt to environmental and non-environmental stresses in Botswana. Despite the conditions of income and capability poverty that have been noted in recent years, rural communities of semiarid areas of Botswana evolved local institutions for managing natural resources and adaptable lifestyles marked by multiple livelihood strategies that enabled the communities to cope with variable climate and variable supplies of natural resources. This has been noted for similar natural resource-dependent communities (Burton, 2004; Hulme, 2004; Thomas and Twyman, 2005). The institutions and livelihood practices, supported by an indigenous knowledge system that evolved through accumulated experiences of changing environmental conditions, imparted a capacity or resilience to withstand drought and other climate variations over a relatively wide coping range. The indigenously developed institutions, livelihoods and knowledge base served well until recent years, when droughts exposed an increasing
3232 J&J Climate Change
1/11/07
9:56 am
Page 72
72 Climate Change and Adaptation
vulnerability among rural communities (Hulme, 2004; UNEP, 2001). The increasing vulnerability may indicate a limit to coping strategies for natural resource-dependent systems that are highly exposed to environmental stresses. However, other evidence suggests that the decline in resilience among natural resource-dependent communities signals major structural changes in their livelihood systems resulting from the way these systems interface with Westernoriented democratic systems. External interventions have been made in response to the increasing vulnerability, but their effectiveness has been minimal and they may even have inadvertently further weakened the adaptation capacity of communities in general (Sporton and Thomas, 2002; Warren, 2005). This chapter examines the effectiveness of past strategies used by rural communities in the Limpopo basin part of Botswana to reduce the negative impacts of climate variability. An assessment is made of the interplay between past strategies and the new socio-political and economic frameworks introduced since the beginning of the 20th century and the implications for evolving a capacity to adapt to climate change. In particular, attention is given to the role of the government’s rural development and disaster relief programmes in transforming coping capacity at the community level. The potential is explored for using remnant community coping strategies and institutions as bases for enhancing adaptation to climate change and stresses among rural communities. A synthesis of multiple factors that need to be considered when building adaptation capacity at the community level in Botswana, as well as recommendations for future actions, are presented in the closing section of the chapter.
The Study Area The study area, shown in Figure 4.1, covers the Botswana part of the Limpopo river basin. The basin as a whole extends over 3,720,000 square kilometers across semiarid lands in Botswana, Zimbabwe and South Africa, as well as humid and sub-humid areas in Mozambique (Sharma et al, 1996). In Botswana, the Limpopo river basin forms 20 per cent of the eastern section of the country known as the hardveld, an area of active erosion formed by igneous and metamorphic rocks overlaid by loamy soils. The hardveld has the highest density of surface drainage in the country, but tributaries have an average flow period of 10–70 days a year (B. P. Parida, 2005, personal communication). The Limpopo catchment yields, on average, about 10mm of surface water annually (Parida et al, 2005). The hardveld is subject to frequent climate extremes such as drought. Some of the worst droughts documented occurred in 1935, 1965, 1984 and 1991 and resulted in significant losses of livestock (Bhalotra, 1989). Average annual rainfall in the hardveld ranges from 400 to 450mm with 35 per cent variability (Figure 4.2). Temperatures can be as high as 38°C during the wet season, resulting in high potential evapotranspiration, with an estimated average of 1400mm per year, which reduces rainfall effectiveness (Parida et al, 2005).
3232 J&J Climate Change
1/11/07
9:56 am
Page 73
Indigenous Knowledge, Institutions and Practices for Coping 73
Figure 4.1 The Limpopo basin area of Botswana and case study sites: Northeast District, Bobirwa Sub-District and Kgatleng District
Most of the land in the hardveld is communally owned, but there are freehold farms along the main Limpopo river channel and in the Northeast District. The basin has a longer history of livestock rearing and dryland farming than the remaining 80 per cent of the western part of the country covered by the Kalahari aeolian sands. The Tswana pastoralists living in villages separated from cattle posts and fields originally occupied the drier southern part of the basin during the 17th century. In the wetter central to northern parts, the Ikalanga people lived in scattered settlements and practised mainly arable agriculture (Dube, 1984). The crops grown included millet, sorghum, beans, melons and maize. Mining of first gold, then coal and later copper and nickel have had a major impact on economic activities in the basin.
3232 J&J Climate Change
1/11/07
9:56 am
Page 74
74 Climate Change and Adaptation
Figure 4.2 Mean annual rainfall over three stations in the Limpopo basin area of Botswana: Francistown, Northeast District; Bobonong, Bobirwa Sub-District; and Mochudi, Kgatleng District Source: Data from Department of Meteorological Services, Botswana.
Projections of future climate change indicate a warmer and drier climate in the basin that may be accompanied by greater variability and more frequent extremes (Scholes and Biggs, 2004). Temperatures over southern Africa are expected to rise by 2 to 5°C by 2050, affecting most of the central land mass of the region occupied by Botswana (Desanker and Magadza, 2001). Precipitation patterns may shift and reduce growing season rainfall by up to 15 per cent (Hulme et al, 2001). The noted climate changes are likely to affect water availability with implications for plant, livestock and wildlife productivity and ultimately, human livelihoods (Desanker and Magadza, 2001). Not all of these changes will be new to the basin, but the magnitude may be different. Some of the relevant past experiences in coping with such variability will be crucial to efforts geared towards enhancing adaptation to these elevated impacts.
Traditional Institutions and Coping with Variability Past socioeconomic frameworks in the hardveld were in many ways born out of years of exposure to unpredictable semiarid climate and the resulting uncertainty in the supply of natural resources. Community coping measures were mostly integrated in the everyday social interactions and economic structures,
3232 J&J Climate Change
1/11/07
9:56 am
Page 75
Indigenous Knowledge, Institutions and Practices for Coping 75
allowing communities to be ready for difficulties rather than treating them with emergency measures introduced in times of difficulties and later forgotten until another disaster strikes. Consequently, institutions of local administration, marriage and extended family were adapted to be resilient to climate shocks. These practices have passed from one generation to another as part of a strategy to minimize the impacts of climate variability in the Limpopo river basin, as was the case for the rest of the country. However, different factors operating over the 20th century have seen most of these strategies fade away, and those that remain are distorted and becoming less effective. This section outlines examples of some of the past coping strategies, reasons for their success and factors that contributed to their weakening. Coping mechanisms were built into the traditional political and administrative framework of each settlement. Among the pastoral Tswana group, local decision making and the administration of justice were implemented through a centralized political institution formed by the chief (Kgosi), members of the royal family (Dikgosana) and other grades of council, such as adult men of the village (Schapera, 1951). Major decisions were made at the Kgotla, or assembly point, regarding user rights to communal lands, livestock and other resources, mobilization of labour, seasonal migration to cattle posts, villages and cultivation areas, settlement of disputes and security. This political and administrative system provided for local management of resources that helped to buffer drought impacts by securing access to land for women and poor households, producing and storing community food reserves and reducing grazing pressure. The institutions of the family and marriage shielded members from a variety of environmental stressors and also reduced the potential to degrade environmental resources (Khama, 1971). For example, the bride price entitled the bride to user rights to land and other household property such as cattle for subsistence purposes at her family of origin if her marriage failed to materialize. This reduced the possibility for the development of poor female-headed households, which is common nowadays as more and more women lose user rights for key resources such as land. The extended family structure still prevails, albeit subject to continuous weakening. At a broader community level, cattle were loaned by livestock owners to poorer households and relatives under a system called mafisa. The beneficiary family would take care of the cattle and in exchange would have use of the animals’ labor and milk, plus a payment at the end of the lease that would depend on the condition of the animals and whether the herd had increased in number. The mafisa system represented a win–win coping strategy that provided income and food security, as well as an opportunity to own cattle, to families that borrowed cattle, spatially diversified the risk to livestock owners from variable water supplies, pasture productivity and disease, and reduced grazing pressures (Tlou, 1990). The traditional institutions of the Kgotla and the extended family were weakened and the mafisa system was discontinued in the Limpopo basin as a result of political, social and economic changes in the 20th century. These changes have contributed to widespread poverty and increased vulnerability of communities to stresses.
3232 J&J Climate Change
1/11/07
9:56 am
Page 76
76 Climate Change and Adaptation
Following a policy common to southern Africa, the colonial administration in Botswana reduced the powers of traditional authorities such as the Kgosi and the Kgotla, a process that was intensified after independence (Sithole, 1993). The independence and community influence of these institutions, and their role in decision making about natural resource use and management, have been lost. As a result, communities have been alienated from decisions concerning their resources in their immediate environment and, consequently, unregulated use of their communal land, depletion of resources and loss of self-reliance. The role of the extended family and practices such as bride price and polygamous marriage have been lessened by growing influences of Christianity, the school system, Western social values, the market system, wage labour and urbanization. Results of these changes are increases in the number of female-headed households, loss of user rights and unrelenting poverty. Reasons for discontinuing the mafisa system include access to ground water as cattle spread to the western Kalahari sandveld, commercialization of cattle that diminished incentives to lend cattle to poorer farmers; expansion of wage labour that provided other means of acquiring cattle, adoption of the oxdrawn plough and subsequent use of tractors, which enabled the production of surplus grain in good years that could be exchanged for livestock, and veterinary services that reduced cattle mortality (Campbell, 1986; Dube, 1995; Dube and Kwerepe, 2000; Tlou, 1990). Termination of the mafisa system meant reduced milk and meat intake leading to protein-energy malnutrition in poorer families, particularly in drought periods.
Veld Products and Coping with Climate Stresses People of the hardveld harvested and used a variety of natural veld products that reduced their vulnerability to drought by diversifying their sources of food and income. Veld products such as meat from wild animals, edible insects, honey, roots, melons, seeds and wild fruits were an important part of the diet of all of the communities in the Limpopo river basin, as was the case for the rest of African societies (Campbell, 1986). Increased access to other food sources, as well as depletion of veld resources from increased population and use of new technologies, have diminished the contribution of these products to the local diet. But despite these changes, communities still resort to the harvesting of veld products for consumption and income generation during, for example, drought periods. Poorer households, the majority of which are female-headed households, currently dominate the harvesting of veld products. Fuel wood trade in the Limpopo river basin is common and is dominated by crop farmers who resort to this activity in response to failed crop yields and eventually adopt the trade as part of their strategy to diversify risks (Kgathi, 1984 and 1989a and b). Traders in the northern parts of the Limpopo river basin earn an average of US$29 and US$38 gross per week for rural and urban sellers respectively, with some wholesale distribution networks developing (White, 1999). However, the increase in fuel wood trade due to persistent drought and unemployment has raised environmental concerns. Normally,
3232 J&J Climate Change
1/11/07
9:56 am
Page 77
Indigenous Knowledge, Institutions and Practices for Coping 77
wood is sourced from old and dead trees. However, in areas of high demand, living trees are now being cut down for fuel wood, thus promoting deforestation. This has resulted in the government imposing bans on the sale of natural products without permit (Botswana Government, 2004). Phane caterpillar, the larvae of the Imbrasia belina moth, is a cherished food product of the veld that is harvested by local harvesters and traded commercially between Botswana and South Africa. On average, 150–250kg is produced by each harvester in one harvest period, 80 per cent of which typically is sold for cash to traders, at a price of roughly US$45 per 50kg bag, who resell the dried Phane for several times that price. Around 4,000 tons of airdried Phane caterpillar worth around US$9 million was exported to South Africa between 1991 and 1994 (Moruakgomo, 1996). Medicinal plants represent another set of natural resources closely linked to indigenous knowledge. A number of medicinal plants are traded locally by herbalists and traditional doctors in both rural and urban areas (Botswana Government, 2003). A few medicinal plants, such as the Grapple plant (Harpagophytum procumbens), also known as the Kalahari Devil’s Claw, have reached international markets. Extracts from the Grapple plant are used in the industrial production of drugs for the relief of arthritis-related pains. Further research is being conducted on the sustainability of the plant, as well as the domestication of the plant for agricultural production (Kgathi, 1989b).
Government Interventions Increasing vulnerability in rural areas in the past decades necessitated the intervention of the government to alleviate the impact of the stresses such as drought. One of the first well-documented droughts is that of 1964–1965, which resulted in large numbers of deaths of both humans and livestock and was associated with increased incidents of foot and mouth disease outbreaks (Campbell, 1978 and 1986; Dube, 1995). Numerous government interventions were made to alleviate the impact of the drought, including a borehole-drilling scheme to move more cattle to the Kalahari Desert, intensified veterinary services and establishment of a vaccine production center, and food relief through organizations from the UK and the US. Since then, the government has implemented a number of programmes in order to increase the resilience of rural areas to climate and other stresses by supporting and diversifying economic development, improving resource management and providing social security (see Table 4.1). The Drought Relief Programme (DRP) was implemented from 1979 to provide temporary supplements to rural incomes in times of drought. The programme accounted for 14 to 18 per cent of total government development expenditure between 1984 and 1987 (MFDP, 1998). In addition, vulnerable group feeding programmes targeting preschool and school-going children, as well as pregnant and lactating mothers, have existed since the 1960s (Ohiokpehai et al, 1998). A Destitute Policy was formulated in 1980 to reduce death due to hunger among those who had completely lost the means of sustenance. Despite all of these efforts,
3232 J&J Climate Change
1/11/07
9:56 am
Page 78
78 Climate Change and Adaptation
Table 4.1 Examples of government policies relevant to vulnerability and drought impact reduction Category
Policy
Year
Objectives
Weaknesses
Agriculture Production
Tribal Grazing Land Policy (TGLP)
1975
Increase livestock productivity; prevent overgrazing
Lack skill development in range management, had a weak link to climatic factors, and did not reduce overgrazing
Arable Lands Development programmes (ALDEP)
1977– 2000
Reduce food grain deficits and achieve food self-sufficiency in rural areas
Top–down approach, weak link to climatic factors, and less skills development in conservation tillage; no additional manpower to implement and monitor
Accelerated Rain-fed Agriculture Programme (ARAP)
1985– 1990
Reverse impacts of long period of drought on farmers
Destumping without condition for cultivating ultimately contributed to land degradation
National 2002 Master Plan for Agricultural Development (NAMPAAD)
Food security from a The manpower, expertise and diversified sustainable infrastructure required to production base implement is limited, and climatic issues are not adequately factored in
Game Ranching 2002 Policy
Economic use of wildlife through game ranching industry
Out of rural community reach due to high capitalization costs
1986
Wildlife conservation through sustainable utilization, including game ranching
Exclusion of communities in management of resources in their areas
CommunityBased Natural Resource Management (CBNRM)
2000 (draft)
Decentralized control of natural resources, greater local benefits from resources and improved management
Needs well-defined property rights for full re-empowerment and management skills
Tourism Policy
1990
Promote tourism
Currently dominated by foreign operated companies. Limited skill among locals.
Financial Assistance Policy (FAP)
1992– 2001
Citizen-owned businesses, employment creation and economic diversification
Lacked strong skills development in rural areas where business culture was nonexistent
Resource Wild Life Management Conservation Policy
Income and Employment Generation Support Policies
3232 J&J Climate Change
1/11/07
9:56 am
Page 79
Indigenous Knowledge, Institutions and Practices for Coping 79
Table 4.1 (continued) Category
Policy
Year
Objectives
Weaknesses
Small, Medium 2001 and Microenterprises (SMME)
Support creation of Strengths in target small-scale own incomeincome generation generating projects at micro level
Citizen 2002 Entrepreneurial Development Agency (CEDA)
Focused development Still new to evaluate of citizen-owned businesses
Rural National Policy 2002 Development for Rural Development
Enhance quality of life Mainly strengths of recognizing for all living in rural the diversity of potential areas economic pursuits in rural areas
Social Protection
Drought Relief Programme
1980
Temporary supplement Welfare relief rather than to rural incomes develop adaptation capacity during drought
Destitute Persons Policy
1980
Address plight of extreme poverty
Does not address capacity poverty
Financial assistance to elderly
A good start but more resources needed to improve the situation
Preparedness for timely and adequate response to emergencies
Lack skilled manpower, monitoring and ready resources to tackle quick-spreading disasters
Old-Age Pension Scheme Policy on Disaster Management
1996
Source: Botswana government (2002a–c), Ministry of Finance and Development Planning (1998 and 2002) and Gichangi and Toteng (2004).
however, Buchanan-Smith (1998) concluded that the Drought Relief Programme had not succeeded in preventing malnutrition in all parts of the country. Other programmes focused on agriculture. For example, the Tribal Grazing Land Policy (TGLP), introduced in 1975, allowed the demarcation of ranches for improved livestock production. However, the policy did not substantially improve the productivity of livestock or rangelands because, among other factors, it failed to account for climate variability (White, 1993). To compensate for this, ranch owners, who are mostly large cattle owners, graze their livestock in both communal areas and in their private ranches when convenient. It is ultimately the small farmers, confined only to communal areas, who are most disadvantaged (Dube and Pickup, 2001).
3232 J&J Climate Change
1/11/07
9:56 am
Page 80
80 Climate Change and Adaptation
The Arable Land Development Programme (ALDEP) was implemented to reduce food grain deficits and achieve food self-sufficiency in rural areas. Limitations of the programme include weak links to climatic factors, top–down approach, lack of skills development and competing agricultural programmes that resulted in manpower shortages for implementation and monitoring. The Accelerated Rain-fed Agriculture Programme was intended to reverse the impacts of drought but was implemented in a manner that aggravated land degradation. The Community Based Natural Resources Management (CBNRM) strategy seeks to decentralize the management of local resources to communities and thereby increase local control of resources, increase local benefits and improve their management (Phutego and Chanda, 2004). Implementation of CBNRM has been hampered by lack of well-defined property rights, full empowerment of local institutions to control resources and management skills. The Financial Assistance Policy (FAP), and the more recent Citizen Entrepreneurial Development Agency (CEDA), attempt to promote small business creation and employment to diversify the economy. In contrast to the agriculture sector programmes, FAP did not require financial contribution from beneficiaries. All these programmes, in addition to the recently introduced CEDA, aimed at diversifying the economy from the mining sector. However, in the majority of cases the initiatives survive only for as long as there is a government subsidy (MFDP, 1998). The contribution of all these programmes towards enhancing economic development and therefore adaptation to multiple stresses in rural areas, including climate change, remain unsatisfactory. For example, despite all of the support, agriculture’s contribution to GDP in the country has dropped from 40 per cent in 1966 to 2.6 per cent in 1999/2000 and the sector has grown to be less attractive to the young. The mushrooming of policies is evidence of a desperate and costly but futile effort on the part of the government to find a solution to address issues of low-income and capability poverty (CSO, 2003). The failure of government interventions to enhance the capacity of the community to cope with the stresses of climate variability is due to numerous complex and interlinked factors. Execution of government policies and programmes was based on a top–down approach that did not engage local institutions in decision making or mobilization of local resources. They depended on financial inputs, mainly from internal but also from external sources. Most are conceived as temporary relief measures despite the fact that climate variability and drought are inherent features of the region. The necessary executive capacity is lacking or inadequate, leading to failure to effectively implement the programmes. Moreover, there has been limited emphasis on conducting feasibility studies prior to programme implementation to assess supporting infrastructure and manpower resource needs. Consequently, there has been insufficient attention to develop local skills and capacities for the programmes to succeed. There has also been a tendency to implement policies from different sectors, some of them contradictory and or with no clear linkages between them. For the rural poor, capacity poverty continues to constrain access to the available assistance. The relief measures introduced during the 1980s drought were not with-
3232 J&J Climate Change
1/11/07
9:56 am
Page 81
Indigenous Knowledge, Institutions and Practices for Coping 81
drawn after the event, with the result that some of the programmes were incorporated into a burgeoning social welfare programme. It was obvious from this that there was an underlying problem of poverty in the country (Solway, 2002). Very poor female-headed households derived as much as 43 per cent of their total income from government programmes and just 18 per cent from their own initiatives, indicating a high level of dependence on social welfare assistance (BNA21, 2002). Drought relief measures and agricultural subsidies implemented in the country continue to be at the expense of self-initiated indigenous social welfare systems (Solway, 1994; Sporton and Thomas, 2002; Thomas and Twyman, 2005). This is also acknowledged in government reports (Ministry of Finance and Development Planning, 1998). Programmes such as the Drought Relief Programme address income poverty and exclude capacity poverty; hence they do not lift individuals from poverty’s vicious cycle. As a result of the foregoing, communities in rural areas are now more dependent on external intervention for both mild and severe stresses than was the case in the past. The increasing vulnerability to current climate variability is likely to make these communities more vulnerable to future climate change.
The Potential of Remnant Coping Strategies Despite the social transformation experienced over the 20th century and the subsequent government interventions, communities still display a rich knowledge of past systems and innovative strategies for coping that include traditional institutional frameworks and use of veld products. These remnants of traditional institutions and practices have potential for incorporation into strategies for sustainable community adaptation to climate change. Natural resources of the veld such as wood, medicinal plants and food products have a high potential to contribute to the livelihoods, incomes and food supplies of people in the Limpopo basin. By adding to and diversifying incomes and food sources, commercial development of veld products can play an important role in reducing the vulnerability of people who live in this semiarid region. There are good examples outside Botswana, such as the commercialization of Rooiboos and Honeybush tea plants (Gleason, 2004; Nofal, 2004) and medicinal plants that used to be gathered from the veld but are now cultivated and support modern industries in South Africa with large international markets (Cunningham and Milton, 1987; Theron, 2001; Trutter, 2001; Kupka, 2001; Gleason, 2004). The potential of veld products is indicated by the high economic value of wild plants in southern Africa, estimated to be roughly US$270 per household per year, with even higher direct use values (Shackleton and Shackleton, 2000; Twine et al, 2003). Examples of valuable natural products that are already commercialized in the Limpopo river basin include fuel wood and timber, food and medicinal plants, and insects (Phane caterpillar). These can be expanded to provide greater benefits and can also serve as models for commercializing other products of the veld.
3232 J&J Climate Change
1/11/07
9:56 am
Page 82
82 Climate Change and Adaptation
Fuel wood trade, already common in the basin, could benefit from increasing concentrations in the atmosphere of carbon dioxide, one of the principal gases driving global climate change. Experimental results in semiarid regions show a potential for woody plants to out-compete grass species at higher concentrations of carbon dioxide (Bassiri et al, 1998; Bond and Midgley, 2000). Other studies have linked the currently observed increase in woody plants in southern Africa to an increase in anthropogenic CO2 over the past 100 years (Bond et al, 2003; Joubert, 2004). On the basis of this hypothesis fuel wood trade might become an important economic activity in the future. Prospects for developing a coordinated fuel wood trade, not yet explored in Botswana (Sekhwela, 1997), warrant investigation. Development of valueadded products for purposes other than energy is also worth investigating. For example, fruits of the Morula tree (Sclerocarya birrea), found mainly in the Limpopo river basin, are already being used in the production of jam and oil by Kgetsi-ya-Tsie (KyT), a non-governmental organization with markets extending to the UK (Lapologa Weekend Gazette, 2003). With greater government support, such initiatives could be expanded to develop full fledged and diversified rural industries. As already noted, trade in dried Phane caterpillar earns a few million US dollars per year for Botswana. Much of the economic benefit is captured by traders and processors. The potential for more income realization by local harvesters lies in development of local value-added processing, market development and harvester mobilization under a common umbrella, such as that offered by KyT, and direct access to markets by producers. It is likely that the future availability of Phane would be affected by climate change. For example, late rainfall could result in caterpillars emerging when there are no Mophane leaves, resulting in a loss of the first (December) Phane emergence, while very wet conditions will result in high mortality of cocoons and young caterpillars. However, the development of the Phane industry now would enhance community rural industry skills that could be diversified to other products and broaden the income generation base, hence increasing resilience and adaptive capacity. The potential of medicinal plants of the Limpopo basin have been developed to only a very limited extent. The Devil’s Claw, one of the few examples of a commercially developed medicinal plant, thrives under desert to semiarid conditions and as a result has the potential to thrive under the future climate projected for Botswana. Many other plants could find commercial applications, and currently knowledge of the medicinal properties of plants is being plundered by scientists and pharmaceutical companies from developed countries (Moloi 2003a, b and c). This effectively denies communities their rightful potential income and benefits. For communities of the hardveld to benefit from commercial applications of medicinal compounds derived from local plants, changes are needed in intellectual property laws. Development of cultural heritage tourism in rural areas is another potential source for income and employment creation if practised under appropriate policy framework (Dube and Moswete, 2003). Proper measures need to be established, first to make communities aware of their culture and other
3232 J&J Climate Change
1/11/07
9:56 am
Page 83
Indigenous Knowledge, Institutions and Practices for Coping 83
resources as a potential source of income under tourism. Second, there is a need to provide appropriate infrastructure and build capacity to facilitate such activities. A culture of shared responsibility supported by available knowledge systems under defined traditional institutions such as the Kgotla and chieftainship is an intangible resource that could be developed as part of community adaptation strategies in the future. These traditional institutions for local administration and justice have persisted, albeit distorted, as the most viable basic community structure in rural settings, together with a wealth of human socio-cultural capital. Every Motswana is identified in legal police documents, applications for passport, government health cards and death notices in terms of their village, chieftainship and ward. The chieftainship structure is important for implementing government programmes and forms a communication link between rural communities and central government. The Kgotla still has roles to play in the administration of justice and serves as a platform for launching new programmes and the dissemination of information. The chieftainship institution is now part of the local government system, recognized among other things by the national flag flying at Kgotla. Government endorses the selection and election of leaders of these institutions, and they are paid a formal monthly salary. But instead of labour regiments, councils and headmen of wards, the chief has to work with a Village Development Committee (VDC), which is elected every two years. Chiefs are now ranked, and paramount chiefs sit in the House of Chiefs. Despite the institutional community structures described above, in reality, central government’s new framework has defined power relations that inadvertently sidelined the community institutions in decision making. For example, the Kgotla is not part of decision making on natural resource use and management, which was one of its traditional core functions. The independence and community influence of such institutions have been severely minimized. There is growing reluctance to attend Kgotla meetings, which shows their growing loss of value in terms of impacts on people’s livelihoods. However, as a system of organization, the Kgotla still offers some important lessons that could be updated and used in the development of community adaptation strategies. This is crucial, as in addition to natural products, community coherence based on its traditional governing structures offers wealth in terms of cultural heritage – another potential resource for sustainable adaptation that has remained untapped in the Limpopo river basin and in the country as a whole.
Synthesis of Community Coping Strategies The environmental and climatic conditions of the Limpopo river basin have shaped the adaptive lifestyles of rural communities. Drawing on accumulated knowledge and experiences with the local climate and resources, communities of the basin developed a strong institutional framework of collective decision making on resource use and management as well as innovative uses of natural
3232 J&J Climate Change
1/11/07
9:56 am
Page 84
84 Climate Change and Adaptation
products that provide a hedge against potential losses of cattle and crops. However, vulnerability to drought and other stresses increased as traditional institutions and practices were weakened by government policies both before and after independence, and by other social and economic changes. Attempts by the government to reduce rural poverty and vulnerability have largely been unsuccessful for reasons discussed earlier. One of the critical factors has been the failure to engage communities in decisions that affect their livelihoods. Through the CBNRM strategy, the government has sought to transfer back the management of local resources to communities. But thus far the strategy has struggled because of a lack of local institutions that are accountable to the communities and that have the capacity to effect equitable access and use of resources and to represent local interests in the implementation of the national policy (Thomas and Twyman, 2005). Remnants of the traditional institutions of the Kgotla and chieftainship could be supported and revitalized to facilitate decentralization of resource management, giving back some of the responsibilities these institutions held before but with a capacity to operate under a modern system. This would provide a framework at the community level that could also be effective in promoting adaptation to climate change. As a resource, indigenous knowledge has historically sustained precarious lifestyles in the shadow of a variable semiarid climate characterized by frequent droughts that have periodically reached catastrophic extremes, and this is evidence of its potential to further develop and enhance adaptation capacity to climate change. Some of the central issues to be addressed in providing a framework for successful utilization of indigenous resources and for implementing policies on decentralization of management of local resources include effective information dissemination, protection of intellectual property, development of modern rural industries where opportunities exist, cultivation of domestic and international markets, capacity building in entrepreneurship, and strategic provision of seed resources for community initiatives. These can be achieved with an enabling policy of technical and other resource support systems operating as much as possible within the existing institutional frameworks in rural areas. The prevention of loss of indigenous knowledge systems through biopiracy could be a major intervention that would empower the communities for a sustainable future, based on the commercialization of their knowledge systems (Hansen and Van Fleet, 2003; Phuthego and Chanda, 2004). Another important aspect of the development of commercialization of veld products is the domestication, cultivation and sustainable production of the resources. A series of stakeholder community workshops conducted in the Limpopo river basin showed that new programmes in rural areas need to take into account changes resulting from the interplay between traditional livelihood systems with Western institutions and values. Without the development of appropriate skills, income-generating community adaptation activities have a limited chance for success, as shown by the CBNRM initiatives (Phutego and Chanda, 2004). Many ongoing self-initiated women’s groups on income-generation projects, such as weaving, dance groups and other projects faced
3232 J&J Climate Change
1/11/07
9:56 am
Page 85
Indigenous Knowledge, Institutions and Practices for Coping 85
problems of organizational, management and financial accounting skills but had no access to assistance in their localities. There is a need to investigate the kind of management that would be suitable at these levels, where trust, confidence and transparency can easily become elusive, with the resulting failure to achieve the high potential of such noble initiatives. Also apparent from the stakeholder community workshops and the subsequent field surveys were the limited activities of non-governmental organizations (NGOs) in the Limpopo river basin rural areas, particularly in Bobirwa and the Northeast District. The categorization of Botswana as a middle-income country has resulted in less funding from donor agencies and a reduction in the activities of international NGOs in the country. NGOs play an important complementary role in the development of adaptation capacity in rural areas. In the past, the Kgotla used to provide a selling point for cattle in villages under the supervision of the chief. Combined efforts involving government, NGOs and local initiatives, such as community-based organizations, are required to develop the modern Kgotla under the village development committees into rural trade hubs where guidance could be provided – within a familiar setup and in a language that would be understood by communities – on issues of entrepreneurship or possibilities to partner with the private sector for the development of local industries. The prospects for this happening are increasing with the general rise in education levels in the country, which has also yielded increasing numbers of well-educated chiefs. It is worth noting that the economy of Botswana remains strongly driven by the government, because of its being almost wholly dependent on mineral exports. Government is the main source of revenues, and attempts to diversify have not been successful (BNA21, 2002). Industrial skills are limited in the country, and the role of the private sector is still emerging, partly explaining the failure of programmes such as the Financial Assistance Policy. Faced with issues of poverty and climate extremes, alleviation measures have had to be applied in the context of disaster management at the expense of productive initiatives. As a result, the spirit of self-reliance is destroyed and a society that expects nearly everything from the government is created. Under these circumstances, the main recommendation of this chapter, which is diversification of the economy through rural industries, remains a challenge within the framework of the current national economic structure. An economy that is heavily dependent on export earnings from a single non-renewable product (diamonds in this case), that is heavily dependent for a high proportion of its livelihoods on agriculture and pastoralism in a semiarid, drought-prone climate and that has high rates of income and capability poverty is highly vulnerable. This vulnerability is heightened by signals that the future climate may become more arid and unpredictable, with consequences for the natural resource base. A proactive approach is required to reduce vulnerability to variable climate, variable productivity of natural resources and climate change. Reducing vulnerability needs to be supported by an appropriate policy framework that engages the private sector, civil society and local institutions to facilitate economic development and diversification in rural areas and build
3232 J&J Climate Change
1/11/07
9:56 am
Page 86
86 Climate Change and Adaptation
local capacity in managerial, entrepreneurial and technical skills for the sustainable management and use of natural resources. Traditional knowledge, practices and institutions are intangible but critical resources that should be fully utilized to build sustainable adaptation capacity in rural communities.
References Bassiri, R., H. J. F. Reynolds, R. A. Virginia and M. H. Brunelle (1998) ‘Growth and root NO3 and PO43 uptake capacity of three desert species in response to atmospheric CO2 enrichment’, Australian Journal of Plant Physiology, vol 24, pp353–358 Bhalotra, Y. P. R. (1989) Nature of Rainy Season in the Chobe District, Department of Meteorological Services, Ministry of Works, Transport and Communication, Gaborone BNA21 (2002) Botswana National Report on the Implementation of Agenda 21 and Other Rio Earth Summit Decisions, Botswana National Agenda 21 Coordinating Committee, Gaborone, Botswana Bond, W. J. and G. F. Midgley (2000) ‘A proposed CO2-controlled mechanism of woody plants invasion in grasslands and savannas’, Global Change Biology, vol 6, pp865–869 Bond, W. J., G. F. Midgley and F. I. Woodward (2003) ‘The importance of low atmospheric CO2 and fire in promoting the spread of grasslands and savannas’, Global Change Biology, vol 9, pp973–982 Botswana Government (2002a) ‘National Master Plan for Arable Agriculture and Diary Development (NAMPAADD)’, Government White Paper No 1, Ministry of Agriculture, Gaborone Botswana Government (2002b) ‘Revised National Policy for Rural Development’, Government White Paper No 3, Ministry of Finance and Development Planning, Gaborone Botswana Government (2002c) ‘Game Ranching Policy for Botswana’, Government White Paper No 5, Ministry of Trade, Industry, Wildlife and Tourism, Gaborone Botswana Government (2004) ‘Agricultural Resources Conservation (Harvesting of Veld Products) Regulations’, Statutory Instrument No 28 of 2004, Agricultural Resources Board, Ministry of Environment, Wildlife and Tourism, Gaborone Buchanan-Smith, M. (1998) ‘Food and nutrition insecurity: The impact of drought’, in M. Mugabe, K. Gobotswang and G. Holmboe-Ottesen (eds) From Food Security to Nutrition Security in Botswana, National Institute of Research/Institute of General Practice and Community MedicineNIR/IASAM Collaborative Research Programme on Health, Population and Development, University of Botswana, Gaborone, Botswana and and University of Oslo, Oslo, Norway Burton, I. (2004) ‘Climate change and the adaptation deficit’, Occasional Paper 1, Adaptation and Impacts Research Group (AIRG), Meteorological Service of Canada, Environment Canada, Toronto, Ontario Campbell, A. C. (1978) ‘The 1960s drought in Botswana’, paper presented at the Drought Symposium, Botswana Society, Gaborone Campbell, A. C. (1986) ‘The use of wild food plants and drought in Botswana’, Journal of Arid Environments, vol 11, pp81–91 Cunningham, A. B. and S. J. Milton (1987) ‘Effects of basket-weaving industry on Mokola palm and dye plants in northwestern Botswana’, Economic Botany, vol 41, pp386–402 Dube, O. P. (1984) ‘Settlement pattern as a dynamic process: The Sebina Mathangwane area of North Central District, 1895–1959’, unpublished thesis, University of Botswana, Gaborone
3232 J&J Climate Change
1/11/07
9:56 am
Page 87
Indigenous Knowledge, Institutions and Practices for Coping 87 Dube, O. P. (1995) ‘Monitoring vegetation and soil degradation in Botswana’, Progress Report, Project No S010001, Australian Center for International Agricultural Research (ACIAR), Department of Geographical Sciences, University of Queensland, Brisbane, Australia Dube, O. P. and R. M. Kwerepe (2000) ‘Human induced change in the Kgalagadi sands: Beyond the year 2000’, in S. Ringrose and R. Chanda (eds) Towards Sustainable Management in the Kalahari Region—Some Essential Background and Critical Issues’, proceedings of the Botswana Global Change Committee – START Kalahari Transect Directorate of Research and Development, University of Botswana, Gaborone Dube, O. P. and G. Pickup (2001) ‘Effects of rainfall variability and communal and semi-commercial grazing on land cover in southern African rangelands’, Climate Research, vol 17, pp195–208 Dube, O. P. and N. Moswete (2003) ‘Tourism: Searching for adaptation options to climate change in southern Africa’, AIACC Notes 2, pp6–7 Gichangi, K. K. and E. N. Toteng (2004) ‘Policy and drought in the Kgatleng District of Botswana: Impact on communities’ vulnerability and adaptation’, Botswana First National Conference on Application of Science and Technology, 13–15 July, Botswana Environmental and Natural Resources Observation Network (BENRON), University of Botswana, Gaborone Gleason, G. (2004) ‘Cuttings are key to commercial cape honeybush production’, Farmer’s Weekly, 30 July, pp40–44 Hansen, S. A. and J. W. Van Fleet (2003) Traditional Knowledge and Intellectual Property: A Handbook on Issues and Options for Traditional Knowledge Holders in Protecting their Intellectual Property and Maintaining Biological Diversity, American Association for the Protection of Science, New York Hulme, M. (2004) ‘A change in the weather? Coming to terms with climate change’, in F. Harris (ed) Global Environmental Change Issues, John Wiley and Sons, Chichester Hulme, M., R. Dougherty, T. Ngara, M. New and D. Lister (2001) ‘African climate change: 1900–2100’, Climate Research, vol 17, pp195–208 Desanker, P. and C. Magadza (2001) ‘Africa’, in J. McCarthy, O. Canziani, N. Leary, D. Dokken and K. White (eds) Climate Change 2001: Impacts, Adaptation and Vulnerability, Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York Jefferies, K. (1997) ‘Poverty in Botswana’, in D. Nteta, J. Hermans and P. Jeskova (eds) Poverty and Plenty: The Botswana Experience, Botswana Society, Gaborone Joubert, L. (2004) ‘CO2-hungry super trees take over the savannas’, Farmer’s Weekly, 26 March, p34 Kates, R. W. (2000) ‘Cautionary tales: Adaptation and the global poor’, Climate. Change, vol 45, pp 5–17 Kgathi, D. L. (1984) ‘Aspects of firewood trade between rural Kweneng and urban Botswana: A socio-economic perspective’, NIR Working Paper No 46, National Institute of Development Research and Documentation, University of Botswana, Gaborone Kgathi, D. L. (1989a) Firewood Marketing in Urban Botswana, Urban Household Energy Strategy Project, The World Bank, Washington, DC Kgathi, D. L. (1989b) ‘The grapple trade in Botswana’, Botswana Notes and Records, vol 20, pp119–124 Khama, S. (1971) ‘Traditional attitudes to land and management of property with special reference to cattle’, Botswana Notes and Records, Special Edition No 1, Proceedings of the Conference on Sustained Production from Semi-arid Areas, October, Gaborone, pp57–61
3232 J&J Climate Change
1/11/07
9:56 am
Page 88
88 Climate Change and Adaptation Kupka, J. (2001) ‘Growing profits from organic medicine’, Farmer’s Weekly, 3 May, pp4–5 Lapologa Weekend Gazette (2003) ‘Kgetsi ya Tsie: Taking over the international cosmetics industry one cheru at a time’, Lapologa Weekend Gazette, 31 January–8 February, pp1–2 Ministry of Finance and Development Planning (1998) Study of Poverty and Poverty Alleviation in Botswana, stakeholder views on the conclusions and recommendations of the study reports by the National Institute of Development Research and Documentation, Phase 1, vol 3, University of Botswana, Gaborone Ministry of Finance and Development Planning (2002) ‘Revised national policy for rural development’, Government Paper No 3, Ministry of Finance and Development, Gaborone Mirza, M. M. Q. (2003) ‘Climate change and extreme weather events: Can developing countries adapt?’ Climate Policy, vol 3, pp233–248 Moloi, E. (2003a) ‘Traditional doctors enter the intellectual property fray’, The Botswana Guardian, Friday 26 September, p13 Moloi, E. (2003b) ‘USA linked to theft of Botswana herbs’, The Botswana Guardian, Friday 5 September, p3 Moloi, E. (2003c) ‘Traditional healers decry “half-hearted” government attitude’, The Botswana Guardian, Friday 21 November, p7 Moruakgomo, M. B. W. (1996) ‘Commercial utilization of Botswana’s veld products: The economics and dimensions of Phane trade’, in Proceedings of the First Multidisciplinary Symposium on Phane, Department of Biological Sciences, University of Botswana, Gaborone Nofal, J. (2004) ‘Honeybush hits the high road’, Farmer’s Weekly, 1 October, pp40–41 Ohiokpehai, O., J. Jagow, J. Jagwer and S. Maruapula (1998) ‘Tsabana: Towards locally produced weaning foods in Botswana’, in M. Mugabe, K. Gobotswang and G. Holmboe-Ottesen (eds) From Food Security to Nutrition Security in Botswana, National Institute of Development Research and Documentation, University of Botswana, and Department of General Practice and Community Medicine, University of Oslo, Oslo and Lentswe la Lesedi, Gaborone, pp179–195 Parida, B. P., D. B. Moalafhi and O. P. Dube (2005) ‘Estimation of likely impact of climate variability on runoff coefficients from Limpopo basin using artificial neural network (ANN)’, in Proceedings of International Conference on Monitoring, Prediction and Mitigation of Water-Related Disasters, Disaster Prevention Research Institute (DPRI), Kyoto University, Kyoto, Japan, pp443–449 Phuthego, T. C. and R. Chanda (2004) ‘Traditional ecological knowledge and community-based natural resource management: Lessons from a Botswana wildlife management area’, Applied Geography, vol 24, pp57–76 Schapera, I. (1951) The Ethnic Composition of Tswana Tribes, Monographs on Social Anthropology No 11, The London School of Economics and Political Science, London Scholes, R. J. and R. Biggs (2004) Ecosystem Services in Southern Africa: A Regional Assessment, Millennium Ecosystem Assessment, Southern Africa Millennium Ecosystem Assessment, Council for Scientific and Industrial Research, Pretoria, South Africa Sekhwela, M. B. M. (1997) ‘Coordinated monitoring and management of the use of natural woodlands in Botswana: A strategy for woody biomass harvesting’, Journal of the Forestry Association of Botswana, vol 1, pp32–44 Shackleton, C. M. and S. E. Shackleton (2000) ‘Direct use values of secondary resources harvested from communal savannas in the Bushbuckridge Lowvel, South Africa’, Journal of Tropical Forest Products, vol 6, pp28–47 Sharma, N., T. Damhang, E. Gilgan-Hunt, D. Grey, V. Okaru and D. Rothberg (1996) ‘African water resources. Challenges and opportunities for sustainable
3232 J&J Climate Change
1/11/07
9:56 am
Page 89
Indigenous Knowledge, Institutions and Practices for Coping 89 development’, Technical Paper No 33, African Technical Department Series, World Bank, Washington, DC Sithole, B. (1993) ‘Rethinking sustainable land management in southern Africa: The role of institutions’, in F. Ganry and B. Campbell (eds) Sustainable Land Management in African Semi-Arid and Subhumid Regions, Proceedings of the SCOPE workshop, 15–19 November, French Agricultural Research Centre for International Development (CIRAD), Dakar Solway, J. S. (1994) ‘Drought as a revelatory crisis: An exploration of shifting entitlements and hierarchies in the Kalahari, Botswana’, Development and Change, vol 25, pp471–495 Solway, J. S. (2002) ‘Navigating the “neutral” state: Minority rights in Botswana’, Journal of Southern African Studies, vol 28, pp711–729 Sporton, D. and D. S. G. Thomas (eds) (2002) Sustainable Livelihoods in the Kalahari Environments: A Contribution to Global Debates, Oxford University Press, Oxford, UK Theron, K. (2001) ‘Developing commercial products from wild plants’, Farmer’s Weekly, 7 December, pp30–32 Thomas, D. S. G. and C. Twyman (2005) ‘Equity and justice in climate change adaptation amongst natural-resource-dependent societies’, Global Environment Change, vol 15, pp115–124 Tlou, T. (1990) A History of Ngamiland – 1750 to 1906: The Formation of an African State, Macmillan Botswana Publishing Co. Pty, Gaborone Trutter, M. (2001) ‘Wild liquorice uplifts community’, Farmer’s Weekly, 1 June, pp52–53 Twine, W., D. Moshe, T. Netshiluvhi and V. Siphugu (2003) ‘Consumption and directuse values of savanna bio-resources used by rural households in Mametja, a semi-arid area of Limpopo Province, South Africa’, South African Journal of Science, vol 99, pp467–473 United Nations Environment Programme (UNEP) (2001) Vulnerability Indices: Climate Change Impacts and Adaptation, UNEP Policy Series, UNEP, Nairobi, Kenya Warren, A. (2005) ‘The policy implications of Sahelian change’, Journal of Arid Environments, vol 63, pp660–670 White, R. (1993) Livestock Development and Pastoral Production on Communal Rangelands in Botswana, Botswana Society, Gaborone White, R. (1999) ‘Fuelwood flow paths study in Francistown’, final report, Energy Affairs Division, Ministry of Mineral Resources, Energy and Water Affairs, Gaborone
3232 J&J Climate Change
1/11/07
9:56 am
Page 90
5
Community Development and Coping with Drought in Rural Sudan Balgis Osman-Elasha, Nagmeldin Goutbi, Erika SpangerSiegfried, Bill Dougherty, Ahmed Hanafi, Sumaya Zakieldeen, El-Amin Sanjak, Hassan A. Atti and Hashim M. Elhassan
Introduction Persistent and widespread drought is a recurrent feature of the climate of Sudan. Drought and highly variable rainfall severely impact rural populations in Sudan and contribute to poverty, hunger, water scarcity, dislocation and even famine. Sudan’s rural populations are also highly vulnerable to future climate change (Osman-Elasha and Sanjak, 2008). Still, there are examples of development strategies and resource management measures employed by rural populations of Sudan that have increased the resilience of communities to cope with drought and its effects. These examples provide models that can be applied to increase drought resilience more widely in Sudan. They also offer lessons that can guide the integration of climate change adaptation with community development strategies. Development and resource management activities are examined in three rural, drought-prone areas of Sudan where traditional farming practices are common: Gireighikh Rural Council in Bara Province of North Kordofan State, Arbaat in the Red Sea State, and El Fashir Rural Council in North Darfur State. People in these rural communities face multiple threats that include climate hazards as well as poverty, resource scarcity, food insecurity, disease, land degradation and violence. A variety of coping and adaptive strategies have been developed and employed in the communities to address these threats. Many came into use autonomously within the studied communities, meaning without external intervention or planning. Others were adopted, or expanded in use, as a result of community development projects that were planned and supported by external agencies. We examine the three cases to understand how development projects and resource management measures have affected community resilience to drought and other stresses, factors that enable or inhibit their effectiveness, and their potential as approaches to climate change adaptation. We briefly describe the
3232 J&J Climate Change
1/11/07
9:56 am
Page 91
Community Development and Coping with Drought in Rural Sudan 91
framework and methods of the case studies, present and compare experiences from the different cases, and close with general lessons for climate change adaptation.
Sustainable Livelihoods and Adaptation All of the case studies are sites of community development projects to enable and support sustainable livelihoods. Sustainable livelihoods is an approach to poverty reduction that focuses on the strategies and means by which people derive their livelihoods, the context of vulnerability to adverse outcomes in which they operate, and the assets needed to sustain and improve livelihoods, reduce vulnerability and move out of poverty (see, for example, DFID, 1999). Projects that apply a sustainable livelihoods strategy attack poverty and vulnerability holistically by increasing access to the assets needed to achieve positive livelihood outcomes. These livelihood assets include natural, physical, financial, human and social capital. In the rural settings of the three case studies, livelihoods are based predominantly on traditional farming and pastoral activities. Exposure to drought and highly variable rainfall, coupled with highly constrained and unequal access to assets, are defining features of the vulnerability of people pursuing farming and pastoral livelihoods in the region. The development projects in the study areas, which were implemented partly in response to the 1980–1984 droughts that severely impacted much of the region, consist of multiple measures implemented together to increase access to the five different classes of livelihood assets. Each of the projects are considered by leaders and members of the communities where they were implemented to have been successful in increasing household and community assets that have generated multiple benefits, including greater resilience to drought. The measures that compose the community development projects therefore can be considered to be adaptation measures for reducing climate risks. A variety of typologies have been applied to categorize adaptation measures (see Smit et al, 2001, for an overview of typologies). Adaptation measures have been categorized by their purposefulness (spontaneous or planned, anticipatory or reactive, and tactical or strategic; see Smit et al, 1996; Stakhiv, 1993), their function (reduce risks, diversify risks, spread risks and increase capacity to bear losses; see Burton et al, 1993), their form (technological, legislative, regulatory, financial, institutional and market-based instruments; see Bryant et al, 2000; Carter, 1996; Smithers and Smit, 1997) and implementing agent (individual, household, community, local government and national government; see Smit et al, 2000). While the above attributes of adaptation are relevant to adaptation in Sudan, we found that within the context of our case studies, it was useful to apply the concepts and framework of sustainable livelihoods to categorize and assess adaptation measures. Our methodological approach, which is adapted from Springate-Baginski and Soussan (no date), is briefly outlined below and is described in greater detail in Osman-Elasha (2006).
3232 J&J Climate Change
1/11/07
9:56 am
Page 92
92 Climate Change and Adaptation
Community members participated in identifying measures being used to increase drought resilience and in categorizing the measures according to the class of livelihood asset targeted by each measure. Structured interviews were used to gather information about climatic and other stresses and the adaptation measures and their effectiveness. Indicators of the effectiveness of adaptation measures were developed with community input and are based on the state of livelihood assets before and after implementation of the projects, as evaluated by community members. Changes in livelihood assets are evaluated in three dimensions: productivity, equity, and the sustainability of the measures and their benefits. The specific indicators differ across the case studies due to differences in the environments, livelihoods and priorities of community members. In Bara Province, indicators of resource productivity changes focus on the area of rangelands rehabilitated, carrying capacity of lands and forage production. In Arbaat and Darfur indicators of resource productivity changes included crop yields per unit land area, volume of crop production, diversification of crops and livestock breeds, numbers of animals, quantity and reliability of water supply, and water conservation. Assessment of productivity changes also takes into account financial indicators such as income changes, diversity of income sources and access to credit. The assessment of equity focused on the situation of minority groups and women. Indicators looked at changes in the access of women and minorities to land, water, social services and credit, their participation in training and production activities and their participation in decision making. Indicators of sustainability include improvement in environmental conditions, adoption of management practices that protect land and water, use of local knowledge, capabilities and technologies, formation and strengthening of community social institutions, raised awareness of climate hazards, and observed continuation of measures after discontinuation of externally funded projects.
Gireighikh Rural Council, Bara Province, North Kordofan State Bara Province lies in the North Kordofan State of Western Sudan. The climate of Bara Province is semiarid, average rainfall is quite low at roughly 250–300mm per year, and seasonal and interannual variability of rainfall is high. The lands are marginal, with sandy, low fertility soils that are becoming increasingly degraded under combined human and climatic pressures. Most of the province is covered by desert scrub vegetation on undulating sand dunes. Agropastoral and transhumant livestock grazing are the predominant livelihoods of the study area. The cumulative impact of recurring droughts, cultivation of marginal lands, fuel wood gathering and overstocking of livestock has drastically depleted the vegetation. As a result, soil erosion, desertification and dust storms have emerged as significant environmental challenges. The local resource base has been degraded, undermining livelihoods and leaving communities more
3232 J&J Climate Change
1/11/07
9:56 am
Page 93
Community Development and Coping with Drought in Rural Sudan 93
vulnerable to adverse effects of future drought. The province was severely impacted by the 1980–1984 droughts that hit the entire Sahel, affecting family and tribal structures and their autonomous traditional practices of resource management, and leading to thousands of people migrating from their villages to refugee camps around the towns and cities. In response to these devastating conditions, in 1992 the United Nations Development Program (UNDP) and the Global Environment Facility (GEF) initiated the project ‘Community-Based Rangeland Rehabilitation (CBRR) for Carbon Sequestration’. The project sought to promote both climate change mitigation and adaptation goals by implementing community-based natural resource management strategies in 17 villages of the Gireighikh Rural Council in central Bara Province. The objectives included prevention of overexploitation and degradation of marginal lands, rehabilitation of rangelands for the purpose of carbon sequestration, preservation of biodiversity, reduction of dust storms and reduction of risk of crop failure. The approach of the project was to increase livelihood opportunities and diversify local production systems; these are expected to have long-lasting benefits for improving socioeconomic conditions, decreasing out-migration and stabilizing the local population. The focus of the CBRR project was on measures to improve the productivity and sustainability of the natural resources that support agropastoral and transhumant livelihoods. But achievement of this goal was reinforced by an approach that also simultaneously targeted the four other livelihood assets: physical, financial, human and social. The project emphasized community participation in decision making, training of members of the community to better manage natural resources, and greater participation of women in economic and social activities. Community development committees were established in the villages through which community members participated in decision making, implementation of resource management measures and organization of social services. The measures implemented by the CBRR project, their benefits, actors who implemented them, the resources and capacities needed to implement them, and obstacles and risks impeding their implementation are listed in Table 5.1. Measures targeted at improving natural resource assets include land rehabilitation, controlling grazing pressures, introducing sheep as replacement for goats, planting trees and shrubs for shelterbelts, stabilizing sand dunes and developing women’s gardens. Measures targeted at physical assets include installing and maintaining wells and water pumps, building grain milling and storage facilities, and changing building practices to conserve wood. Financial assets were increased by improving access to local and national markets, diversifying production activities (for example raising, fattening and marketing sheep), and providing greater access to credit through revolving credit funds. Human capital was increased through training of farmers and workers, including women, and providing health, education and other services. Social capital was enhanced by the formation of community development committees, the participation of community members in decision making, greater participation of women in production activities, and education, information exchange and networking activities.
• Community Development Committees • Community members and elders
Improve land productivity Diversify production Improve environmental conditions Improve water supply and quality Household food security Food and financial security for women • Improve animal health and productivity • More reliable water supply • Increase local processing of farm products • Increase local food reserves • Increase vegetable and fruit production with supplemental irrigation • Improve nutrition • Reduce clearing of vegetation for fuel wood and building materials
• Install and maintain wells and water pumps • Build grain storage and milling facilities • Conserve energy with improved stoves, other methods • Mud-walled houses to replace wooden huts
• Expand and diversify on-farm income • Access to credit generating activities • Greater and less variable financial • Expand and diversify off-farm income employment opportunities • Greater financial resources for • Establish community revolving funds bearing losses, recovering from losses, maintaining food security and maintaining and improving other livelihood assets
Physical
Financial
Financial resources Spare parts Equipment for digging Clay soil for building mud houses Trained building workers Trained mechanics Informed pioneer women
Water source Rangeland Financial resources Farming knowledge and skills Training and extension
• Lack of financial resources • Out-migration of trained workers • Changing government policies
• Lack of financial resources • Changing state level policies
• Lack of financial resources • Out-migration of trained workers • Changing government policies • Competition among tribes • Conflict over resources
Obstacles and risks
9:56 am
• Financial resources • Trained workers • Training and extension facilities, services and materials • Rural development officers, extension officers • Trained community members
• • • • • • •
• • • • •
Needed resources and capacities
1/11/07
• Community development committees • Farmers, women and minority groups • Government development officers • Extension agents
Subsistence farmers Female farmers Project staff Government officials
• • • •
• • • • • •
Rehabilitate rangeland Shift from crop farming to livestock Diversify livestock breeds Water harvesting Stabilize sand dunes Construct and maintain windbreaks Women’s irrigated gardens Provision of veterinary services
• • • • • • • •
Natural
Actors
Benefits
Measures
Targeted Asset
Table 5.1 Adaptation measures in Gireighikh Rural Council, Bara Province, North Kordofan State
3232 J&J Climate Change Page 94
94 Climate Change and Adaptation
• Establish community development • Delivery of social services through committees in local villages a system in which the community • Empower community development has ownership committees to make decisions, allo- • Participation of community cate resources, implement measures members, including minorities and • Education, information exchange and women, in decision making networking • Early preparedness for climate hazards through dissemination of climate information and forecasts • Equitable participation in project activities and equitable distribution of benefits
Social
Actors
Community leaders Financial resources Communication network Updated climate information and forecasts • Extension materials • Teachers, community development officers
• • • •
• Schools • Training and extension facilities, services and materials • Clinics and health units • Veterinarian services • Teachers, trainers, extension workers and veterinarians
Needed resources and capacities
• Social conflicts • Changing government policies • Migration of trained workers • Diminished authority of traditional leaders and institutions
• Changing government policies • Migration of trained workers • Limited financial resources
Obstacles and risks
1/11/07
• Community Development Committees • Trained community members • School teachers, extension workers, veterinary officers • Women’s groups
• Community • Improve capabilities to manage Development natural resources Committees • Improve resource productivity • Better human health and nutrition • Trained community members and elders • Improve women’s livelihoods • School teachers, extension workers, veterinary officers
• Build on traditional knowledge • Training of farmers and workers • Provide health, education and other services • Training of women
Human
Benefits
Measures
Targeted Asset
Table 5.1 (continued)
3232 J&J Climate Change 9:56 am Page 95
Community Development and Coping with Drought in Rural Sudan 95
3232 J&J Climate Change
1/11/07
9:56 am
Page 96
96 Climate Change and Adaptation
Survey data indicate that community members observed substantial improvements in land rehabilitation, carrying capacity and forage production as a result of the various measures implemented by the project. The improvement of rangeland also led to a significant increase in animal numbers, particularly sheep. High levels of participation of women in irrigated gardens and of women and marginalized groups in training, extension and social services were attained. Indicators for the sustainability of livelihood assets and adaptation measures showed improvement as a result of the project. Key to the sustainability, according to survey results, are the efficiency of the community development committees, efforts of a Sudanese non-governmental organization to support and continue measures, dissemination of information on rainfall, new production inputs and technologies, and prices, and high loan repayment rates to the community revolving funds.
Arbaat, Red Sea State The Arbaat study area is the catchment of the Khor Arba’at, a small seasonal stream located in the Red Sea State of northeastern Sudan, about 50km north of Port Sudan, the state capital. The region is characterized by relative isolation, harsh terrain, highly variable rainfall, recurrent drought spells, small area of cultivable land and low population density. Surface runoff is the primary source of fresh water in the Red Sea area, where runoff rates are high due to the rocky and compact nature of soils, steep slope, high portion of rainfall from thunderstorms and the poor vegetation. Animals represent the main means of economic and social mobility, recognition and survival of the Beja pastoralists of the region. Aware of their environment’s vulnerability to drought and famine, they developed a primarily subsistence agropastoral system with a dispersed pattern of settlement and migration in pursuit of water, pasture and cultivable lands that helped maintain the carrying capacity of the land, reduce competition and conflict, and allow for recovery from shocks (Abdelatti, 2003). Adherence to Beja traditions of salif, a social code of conduct governing relations and resource use, helped to preserve land, animal and other resources. This included a strong social sanctioning system imposed by tribal leaders that helped to constrain overuse of land and wood, and rebuild of animal stocks after each drought cycle. Frequent occurrences of drought in the Red Sea hills have been the norm during the 20th century. But the previous pattern of short-term recovery was shattered after the long drought and famine of the mid-1980s and the traditional agropastoral system of the Beja failed to re-configure (Abdelatti, 2003). Since then the Red Sea State has been in an almost constant state of emergency and relief operations that only vary in scale, length and location from one year to the next. Although the term emergency implies a short or limited duration, whereby people are temporarily in need of relief (ODI, 1995), the reality in the area is that most emergencies last for longer than one year. According to Abdelatti et al (2003), a ‘relief culture’ has become established in which the population and local authorities are heavily dependent on the central govern-
3232 J&J Climate Change
1/11/07
9:56 am
Page 97
Community Development and Coping with Drought in Rural Sudan 97
ment and foreign aid organizations to address humanitarian crises. Under these conditions, long-term planning, including for building resilience to the impacts of drought, have become a low priority. The Khor Arba’at Rehabilitation Project (KARP) was initiated in 1994 by SOS Sahel in response to the syndrome created by the Sahelian drought in the 1980s. The main objectives of the project were to improve the livelihoods and food security of the local population. Measures implemented by the project are identified in Table 5.2. Measures to improve the management of natural resources include micro-catchment water harvesting using contour bunds for planting trees (see Figure 5.1) and terracing, establishing home gardens and establishment of a system for equitable distribution of water. Physical assets were increased by digging wells for irrigation, installing and maintaining equipment for applying fertilizers and pesticides, and introducing improved seed varieties and new crops, including date palm. Financial assets were increased by diversifying farm production, expanding off-farm income opportunities, increasing access to credit through revolving funds and providing access to markets. Agricultural extension services, training, adult literacy, education for women and education about the use of credit added to human capacity. Social capacity was enhanced by the formation of community organizations, involvement of community members, including women, in decision making, and information exchange activities.
Figure 5.1 Contour bunds for water harvesting and tree planting in Arbaat Benefits from the measures implemented by the KARP project include increased and less variable agricultural production, greater quantity, quality and reliability of water, enhanced livelihood opportunities, increased access to credit, reduced out-migration and stabilization of local population. Surveys also revealed that the project increased women’s participation in public life, production activities and community development activities, and provided them greater access to resources.
• Financial resources • Government policies • Trained workers • Limited financial • Training and extension facilities, resources services and materials • Rural development officers, extension officers • Trained community members
• Expand and diversify on-farm income • Access to credit generating activities • Greater and less variable • Expand and diversify off-farm financial income employment opportunities • Greater financial resources for • Establish revolving funds bearing losses, recovering from losses, maintaining food security and maintaining and improving other livelihood assets
Financial
9:56 am
• Community organizations • Women’s groups • Extension agents
• Lack of financial resources • Government policies, including construction of dam
1/11/07
Financial resources Equipment for Digging Spraying equipment Fertilizers and insecticides Spare parts Trained irrigation workers Trained mechanics Informed pioneer women
• Lack of financial resources • Government policies, including construction of dam
• • • • • • • •
• Water source (small stream or khor) • Agricultural land • Financial resources • Farming knowledge and skills • Technical capacity for irrigation • Training and Extension services
• Digging wells for irrigation • Improve water availability • Community organizations • Install and maintain equipment for • Increase quantity and reliability of • Farmers optimal application of fertilizers and farm output and income • Women pesticides • Improve nutrition • Introduce improved seed varieties, new crops and date palm
Subsistence farmers Women Project staff Government officials
Physical
• • • •
• More reliable and better quality water supply • Increase quantity and reliability of farm output and income • Diversify farm production • Stop out-migration • Improve household food security and nutrition
Obstacles and risks
• Earth bunds and terracing to harvest water • Increase recharge of ground water • Establish water distribution network for more equitable distribution • Establish home gardens
Needed resources and capacities
Natural
Actors
Benefits
Measures
Targeted Asset
Table 5.2 Adaptation measures in Arbaat, Red Sea State
3232 J&J Climate Change Page 98
98 Climate Change and Adaptation
• Social conflicts • Changing government policies • Diminished authority of traditional leaders and institutions
Community leaders Financial resources Communication network Updated climate information and forecasts • Extension materials • Teachers, extension workers, community development officers
• • • •
1/11/07
• Community Development Committees • Trained community members • School teachers and extension officers • Women’s groups
• Formation of community organiza- • Delivery of social services through tions a system in which the community • Promote women’s participation in has ownership public activities, decision making and • Participation of community production process members, including minorities and • Information exchange about climate women, in decision making hazards and forecasts • Ability of women to participate in public life • Early preparedness for climate hazards through dissemination of climate information and forecasts • Equitable participation in project activities and equitable distribution of benefits
Social
Obstacles and risks • Limited financial resources • Migration of trained workers
Needed resources and capacities
• Build on traditional knowledge • Improve capabilities to manage • Community organizations • Schools and literacy classes • Training of farmers and workers natural resources • Traditional leaders and • Training and extension facilities, • Training of community development • Increase productivity of farms and community elders services and materials committee members livestock • School teachers and • Teachers, trainers and extension • Education of children • Improve women’s livelihoods extension officers workers • Adult literacy classes, especially for women • Access to information and ability to predict environmental changes
Actors
Human
Benefits
Measures
Targeted Asset
Table 5.2 (continued)
3232 J&J Climate Change 9:56 am Page 99
Community Development and Coping with Drought in Rural Sudan 99
3232 J&J Climate Change
1/11/07
9:56 am
Page 100
100 Climate Change and Adaptation
Acquired skills and knowledge are considered to be the main sustainable benefits from the project. Elements of the project that have promoted sustainability of measures and benefits include government financial and technical support, establishment of community development organizations, continuous awareness campaigns, training programmes for different community groups, community participation in planning and contribution to the project costs and addressing community needs for health, education and other basic services.
El Fashir Rural Council, North Darfur State North Darfur is situated on the northern transitional margin of the Intertropical Convergence Zone. Consequently, the state is one of the most drought affected regions of Sudan. Most of the area is deficient in water even in the wettest months of July to September, which account for 80 per cent of annual rainfall. The drought years of 1983–85 greatly affected the demographic and socioeconomic conditions of the area. Many people lost over half of their cattle, as well as large numbers of sheep, goats and camels as a result of the prolonged drought. Large numbers left their homes due to famine and the environmental impacts of desertification and drought, resulting in the growth of shantytowns. Seeking water and forage for their animals, transhumant pastoralists encroached on farmers’ cultivated lands. Tribal conflicts and violence arose between pastoralists and farmers, and also between different groups of pastoralists. More recently, the region has been engulfed in civil war, the origins of which include conflicts over scarce resources in this harsh environment (Osman-Elasha and Sanjak, 2008). In the two previous case studies, adaptive measures were introduced by projects that were externally promoted in response to a specific climatic event. In contrast to those cases, adaptation measures were developed autonomously by the local communities of El Fashir as a means of coping with variable resource productivity, recurrent drought and other stresses. These locally developed measures were later supported and expanded by an externallyfunded food security project that was implemented in 1998 by the Intermediate Technology Development Group (ITDG). The ITDG food security project sought to build on indigenous knowledge of water-harvesting techniques with the involvement of local communities. Autonomous adaptation measures that were expanded in El Fashir by the ITDG project, as well as new measures introduced by the project, are identified in Table 5.3. The greatest share of family food production has typically come from cultivation of sandy soils to grow millet, sesame and groundnuts. But in recent years, rain-fed farming on sandy soils has become increasingly risky and unable to produce enough food for the family. In response to this risk, trus cultivation was developed to grow sorghum, vegetables and tobacco on clay soils. First implemented in 1964 when a large trus or embankment was constructed across the Wadi El Ku to harvest water runoff, this practice has been promoted by the ITDG project to provide an important supplement to household food production and incomes, and to relieve pressure on cultivated sandy soils.
3232 J&J Climate Change
1/11/07
9:56 am
Page 101
Community Development and Coping with Drought in Rural Sudan 101
Magun cultivation is another practice that is indigenous to the area that has been promoted by the ITDG project. Adopted as a response to sand encroachment onto fertile soil, magun cultivation calls for digging holes 10–30cm in diameter, 5–15cm deep and spaced 40–70cm apart, loosening the soil in the centre of the holes and planting tobacco and melon seedlings. Terracing is used to harvest water and grow vegetables such as okra, eggplant and tomatoes that can be harvested up to 5 months after the rainy season. Home gardens, or jubraka, operated mostly by women, have been promoted for the growing of fast maturing crops and vegetables including okra, pumpkins and cucumbers. Gum trees (Acacia senegal) were planted to restore lands, increase vegetation cover and supply fuel wood. Agricultural production was also improved by adding to physical assets with animal drawn ploughs, rental of tractors, expansion of traditional food storage facilities, construction of a central seed store, and use of improved seed varieties and new crop types. Financial assets were increased in El Fashir by diversifying and increasing agricultural production, improving access to credit and supporting marketing of farm products through unions. Training of farmers in techniques needed to diversify their production activities and improve resource management and the involvement of women in home gardening of vegetables has increased human capital. Social capital was increased through the formation of unions for traditional farmers, fruit growers and vegetable growers that have helped with organizing production, harvesting and marketing of products. The benefits to the villages around El Fashir from the implemented measures include substantial increases in farm productivity, diversity of income sources, income levels and stability of incomes. Survey results indicate some improvement in the status of women, but there remain sharp distinctions between the roles of women and men. Women bear a disproportionately large share of the workload for crop farming, harvesting and vegetable growing, yet men make decisions about land use and farm planning. Sustainability of the measures and their benefits are aided by emphasis on indigenous water harvesting and cultivation techniques, and the prominent role of the traditional system of administration and traditional leaders in the management of resources and project implementation.
Adaptation Opportunities, Obstacles and Risks Existing knowledge, skills, practices, resources and institutions that are indigenous to the case study sites provided opportunities to build on and increase resilience of the communities to drought and other hazards. But a variety of obstacles are identified by the case studies as impeding fuller adaptation. Furthermore, there are risks that threaten the sustainability of adaptation measures and benefits in each of the study areas. Effective adaptation requires strong community institutions, leaders and innovators. The sustainable livelihood projects that were implemented in the case study areas worked with traditional administrative systems and leaders
Benefits • Increase quantity and reliability of farm output and income • Diversify farm output and incomes • Stable water supply • Reduce pressure on sandy soils • Improve household nutrition and food security
• Reliable water supply • Increased soil fertility and agricultural production • Diversification of crops • Food reserves for added food security
Measures
• Expansion of water harvesting for trus cultivation on clay soils • Construct terraces to grow vegetables • Magun cultivation to grow tobacco and melons • Restocking of gum trees (Acacia Senegal) • Use of crop residues to increase soil nutrients
• Animal drawn ploughs for use in water harvesting • Rental of tractors • Traditional food storage technologies • Construction of central seed store • Use of improved seeds and new crop varieties
• Provide diversified income • Greater and less variable financial opportunities from selling fruits and income vegetable, gum garden, and • Greater financial resources for employment opportunities. bearing losses, recovering from • Provide access to credit losses, maintaining food security and maintaining and improving other livelihood assets
Targeted Asset
Natural
Physical
Financial
Subsistence farmers Women Small commercial farmers Cooperative societies Farmers and trade unions Extension workers
• • • • • •
Financial resources Collateral assets Farmers Union or Shail System Extension services and facilities Skilled farmers Trained community members
Domestic animals Local material for building stores Financial resources Plant and animal residues Farming knowledge and skills Training and extension services
• Inaccessible credit • Conflicts over resources • Civil war
• Conflicts over resources • Civil war • Changing Government policies • Financial resources
• Conflicts over resources • Civil war • Changing Government policies • Financial resources
Obstacles and risks
9:56 am
• • • • • •
• • • • • •
Agricultural land Tree seedlings Financial resources Farming knowledge and skills Indigenous knowledge of water harvesting • Technical capacity for irrigation • Training and extension services
• • • • •
Needed resources and capacities
1/11/07
• Subsistence farmers • Cooperative societies • Intermediate Technology Development Group (ITDG)
Subsistence farmers Women Small commercial farmers Merchants and urban farmers • Cooperative societies • Project staff • Government officials
• • • •
Actors
Table 5.3 Adaptation measures in El Fashir Rural Council, North Darfur State
3232 J&J Climate Change Page 102
102 Climate Change and Adaptation
Subsistence farmers Women Small commercial farmers Merchants and urban farmers • Community-based organizations • Cooperative societies • Government officials
• • • •
• Organize farmers unions • Organize women’s groups • Organize social networks and cooperatives
Social
• • • • •
Knowledgeable union members Communication networks Information sources Community leaders Development officers
• Training and extension facilities, services and materials • Community-based organizations • Farming knowledge and skills • Trainers, extension workers
Needed resources and capacities
• Conflicts over resources • Civil war • Diminished authority of traditional leaders and institutions
• Conflicts over resources • Civil war • Migration of skilled labourers
Obstacles and risks
1/11/07
• Mobilize resources for community projects • Participation of women in public life and production activities • Community participation in decision making
• Community-based organizations • Women • Small commercial farmers • Traditional leaders and community elders • Cooperative societies • ITDG
• Build on traditional knowledge • Improve capabilities to manage • Training of farmers to manage their natural resources resources and diversify production • Increase productivity of farms and • Involve women in production of livestock vegetables • Better health and nutrition
Actors
Human
Benefits
Measures
Targeted Asset
Table 5.3 (continued)
3232 J&J Climate Change 9:56 am Page 103
Community Development and Coping with Drought in Rural Sudan 103
3232 J&J Climate Change
1/11/07
9:56 am
Page 104
104 Climate Change and Adaptation
and local non-governmental organizations to gain community acceptance and ownership of the projects. The strengthening of institutions with training and resources, formation of new community institutions, empowering local institutions to plan and implement project activities, and promoting the participation of community members in the activities and decision-making processes of the institutions were central to the approach of the projects and are key factors contributing to their success. Local institutions have also been important for the continuation of development and adaptation activities after the termination of the externally funded projects. However, the authority and coherence of traditional institutions have been weakened in Sudan by changes in government policies that have shifted authority to new government structures, social and economic changes, and displacement and migration of people. Weakening of local institutions is degrading social capital that is important for adaptation. It is also eroding the traditional systems for managing and accommodating migrations of herders and their livestock and for resolving tribal disputes. A planned heightening of the Khor Arba’at Dam is another example of government policy that is jeopardizing adaptation gains. Heightening of the dam is expected to divert more water for urban use in the capital city of Port Sudan to alleviate severe water shortage during summer time. But increasing water storage in the dam and diverting more water to Port Sudan will reduce the volume of water spillover that supplies the Arbaat community. Potential adverse effects of reduced water supply to Arbaat include reduced cultivated area, displacement of families, spread and invasion of aggressive mesquite trees into fertile agricultural land, and reduced production of food for subsistence and marketing to urban dwellers in Port Sudan. Members of the Arbaat community are petitioning the government through traditional and religious leaders to either take necessary measures to mitigate the adverse impacts of the project or, preferably, to drop the idea. Lack of natural and physical resources of all kinds is an obstacle to adaptation in rural communities of this semiarid region. Lack of water and fertile lands are key constraints on livelihoods and most of the adaptation strategies are aimed at improving water supply and management, and improving the fertility and productivity of land. Shortage of financial resources is identified in all three case studies as threatening the sustainability of adaptation measures. Credit for the purchase of seed is particularly needed. Other resources repeatedly identified in the case studies as essential for adaptation and in short supply include facilities and material for education, health, marketing and agricultural extension services, food storage facilities, equipment and spare parts. Human capital is scarce and enhancement of human capital is an important component of the development projects. While most of the technologies promoted by the projects to improve the use and management of land and water, and to diversify production activities and incomes are indigenous to the areas, not all have the knowledge and skills needed to implement them effectively. Training and agricultural extension services are needed to disseminate the needed knowledge and skills. This, in turn, requires skilled rural development officers, extension workers and teachers. An important threat
3232 J&J Climate Change
1/11/07
9:56 am
Page 105
Community Development and Coping with Drought in Rural Sudan 105
highlighted in two of the three case studies is the out-migration of skilled workers and technicians. In Bara, in-migration to the region is also a threat, as people are being attracted by the improved condition of rangelands and availability of water and may come into conflict with residents of the region. Our case study of El Fashir Rural Council began prior to the current conflict and violence in Darfur, which pose substantial risks to the sustainability of adaptation efforts, as well as to the very survival of the people and communities of the region. Historically, conflicts in Darfur have arisen in response to scarcity of natural resources as farmers and pastoralists have clashed over water and land. The current crisis, which has reached a state of civil war, also has its origins in conflict over resources that have been exacerbated by the weakened state of local institutions. The context of vulnerability of the people of Darfur has deteriorated severely as a result. But despite the violence and displacement of large numbers of people in Darfur, within the case study villages of El Fashir Rural Council water harvesting and associated agricultural activities continue without external support, cultivated area has increased, crop productivity is up from 2005, and livestock production has increased such that the villages are the main meat suppliers to the town of El Fashir.1 Whether these activities and benefits can be sustained in El Fashir is doubtful as the crisis in Darfur continues. Extension of them to wider areas is impossible without cessation of hostilities and establishment of security in the region. The current conflict in Darfur deserves detailed assessment and in-depth analysis that go beyond the scope of this paper.
Lessons for Adaptation Processes In the three Sudanese case studies, strategic approaches were taken towards improving livelihoods, increasing food security and reducing vulnerability to multiple sources of risk, including but not limited to climatic risks. The community development projects each implemented multiple measures aimed at these goals. Some measures were motivated explicitly by the desire to reduce climate risks (for example, water harvesting), some to expand livelihood options (for example, introduction of new crops and types of livestock), some to improve and sustain the productivity of land (for example, rehabilitating rangelands, controlling access and use of rangelands) and some to increase individuals’ skills (training and extension services). Others were introduced to expand participation in decision making (for example, formation and engagement of village committees and women’s groups), empower marginalized persons (for example, literacy programmes and irrigated gardens for women) and improve financial conditions (for example, revolving credit fund and marketing assistance). Taken together, they had the effect of increasing resilience of the communities to recurrent drought. The strength of the development projects lay not in their individual measures for responding to drought, but in the holistic way in which the measures were planned and implemented to initiate and support adaptive processes and to build capacity to sustain the processes. They represent experiments in the
3232 J&J Climate Change
1/11/07
9:56 am
Page 106
106 Climate Change and Adaptation
integration of adaptation with community development. The result is an enhanced capacity for adaptation, or broader ability of the communities to manage climate and other risks to livelihoods and food security. The experiences from the studied communities suggest that, consistent with the findings of Smit et al (2001), future adaptation should focus on enhancing adaptive capacity, beginning with developing flexible management approaches and increasing current resilience to drought and other stresses. In the context of sustainable livelihoods, this takes the form of increasing access to natural, physical, financial, human and social assets that can be used to respond to climatic and other stresses. Community institutions played an important role in the development projects and implementation of adaptation measures. The sustainable livelihood projects that were implemented in the case study areas worked with traditional administrative systems and leaders and local non-governmental organizations to gain community acceptance and ownership of the projects. The strengthening of institutions with training and resources, formation of new community institutions, empowering local institutions to plan and implement project activities, and promoting the participation of community members in the activities and decision-making processes of the institutions were central to the approach of the projects and are key factors contributing to their success. Local institutions have also been important to the continuation of development and adaptation activities after the termination of the externally-funded projects. Each of the development projects relied on local institutions and local leaders to organize and implement their programmes. These institutions engaged persons at risk in decision processes, including farmers, herders, women and minorities. Using this bottom–up approach enables a community to have a better understanding of its own vulnerability, priorities and adaptation needs to shape the project’s objectives, design and implementation. It also facilitates cooperation within the community, mobilization of local resources and access to indigenous knowledge to achieve project objectives. Integration of indigenous knowledge and experience with climate variability and traditional practices for managing the effects of variability proved to be important in each of the case study sites. Human and social capacity building were critical to the effectiveness of the development projects in Bara, Arbaat and El Fashir, and appear to have produced long-lasting benefits. Each of the projects invested in creating and strengthening local institutions and providing training to community members on managing resources, diversifying livelihoods and responding to risks. The human and social capacities built by the projects are the mechanisms by which the adaptive processes in these communities can continue and become more effective. Looking to the future, climate change is likely to alter patterns of climate variability and extremes from those experienced in the past, possibly in ways that would exceed the existing tolerance ranges of Sudan’s rural communities. Traditional knowledge based on historical observations and experience will be less reliable as a guide for managing climate risks. While rural communities have adapted their livelihoods to variable climate conditions and developed
3232 J&J Climate Change
1/11/07
9:56 am
Page 107
Community Development and Coping with Drought in Rural Sudan 107
coping strategies to deal with local risks, events that are unusually severe, persistent or frequent can invalidate those strategies and increase risk (Pelling et al, 2004). Economic globalization and other economic and social changes add to the uncertainty about the viability of current livelihood strategies and adaptation measures. In this context of change and uncertainty, the advantages of building adaptive capacity, in contrast to focusing on implementation of specific adaptation measures, and integrating adaptation with holistic community development programmes are all the more compelling.
Note 1
Personal communication to N. Goutbi from Practical Action, 6 September 2006.
References Abdelatti, H. (2003) ‘Khor Arba’at livelihoods and climate variability’, AIACC-AF14 case study report, Higher Council for Environment and Natural Resources, Khartoum Bryant, C.R., B. Smit, M. Brklacich, T. Johnston, J. Smithers, Q. Chiotti and B. Singh (2000) ‘Adaptation in Canadian agriculture to climatic variability and change’, Climatic Change, vol 45, no 1, pp181–201 Burton, I., R. Kates and G. F. White (1993) The Environment as Hazard, Guilford Press, New York Carter, T. R. (1996) ‘Assessing climate change adaptations: The IPCC guidelines’, in J. Smith, N. Bhatti, G. Menzhulin, R. Benioff, M. I. Budyko, M. Campos, B. Jallow and F. Rijsberman (eds) Adapting to Climate Change: An International Perspective, Springer-Verlag, New York, pp27–43 DFID (1999) ‘Sustainable livelihoods guidance sheets’, Department for International Development, UK, available at www.livelihoods.org/info/info_guidancesheets.html Ministry of Environment and Physical Development (2003) ‘Sudan’s first national communication under the United Nations Framework Convention on Climate Change’, available at http://unfccc.int/resource/docs/natc Overseas Development Institute (ODI) (1995) ‘General food distribution in emergencies: From nutritional needs to political priorities’, Good Practice Review, vol 3 Osman-Elasha, B. (2006) ‘Human resilience to climate change: Lessons for eastern and northern Africa’, final report of AIACC Project No AF 14, International START Secretariat, Washington, DC, available at www.aiaccproject.org Osman-Elasha, B. and E-A. Sanjak (2008) ‘Livelihoods and drought in Sudan’, in N. Leary, C. Conde, J. Kulkarni, A. Nyong and J. Pulhin (eds) Climate Change and Vulnerability, Earthscan, London Pelling, M., A. Maskrey, P. Ruiz and L. Hall (2004) Reducing Disaster Risk: A Challenge for Development, United Nations Development Programme, Bureau for Crisis Prevention and Recovery, New York, www.undp.org/bcpr Smit, B., E. Harvey and C. Smithers (2000) ‘How is climate change relevant to farmers?’ in D. Scott, B. Jones, J. Andrey, L. Mortsch and K. Warriner (eds) Climate Change Communication, proceedings of international conference, 22–24 June, Kitchener-Waterloo, Environment Canada, Toronto Smit, B., D. McNabb and J. Smithers (1996) ‘Agricultural adaptation to climatic variation’, Climatic Change, vol 33, pp7–29
3232 J&J Climate Change
1/11/07
9:56 am
Page 108
108 Climate Change and Adaptation Smit, B., O. Pilifosova, I. Burton, B. Challenger, S. Huq, R. Klein and G. Yohe (2001) ‘Adaptation to climate change in the context of sustainable development and equity’, in J. J. McCarthy, O. F. Canziani, N. A. Leary, D. J. Dokken and K. S. White (eds) Climate Change 2001: Impacts, Adaptation and Vulnerability, contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York Smithers, J. and B. Smit (1997) ‘Human adaptation to climatic variability and change’, Global Environmental Change, vol 7, no 2, pp129–146 Springate-Baginski, O. and J. Soussan (no date) ‘A methodology for policy process analysis: Improving policy–livelihood relationships in South Asia’, Working Paper 9, , Leeds, UK, www.geog.leeds.ac.uk/projects/prp Stakhiv, E. (1993) Evaluation of IPCC Adaptation Strategies, Institute for Water Resources, US Army Corps of Engineers, Fort Belvoir, VA
3232 J&J Climate Change
1/11/07
9:56 am
Page 109
6
Climate, Malaria and Cholera in the Lake Victoria Region: Adapting to Changing Risks Pius Yanda, Shem Wandiga, Richard Kangalawe, Maggie Opondo, Dan Olago, Andrew Githeko, Tim Downs, Robert Kabumbuli, Alfred Opere, Faith Githui, James Kathuri, Lydia Olaka, Eugene Apindi, Michael Marshall, Laban Ogallo, Paul Mugambi, Edward Kirumira, Robinah Nanyunja, Timothy Baguma, Rehema Sigalla and Pius Achola
Introduction Malaria and cholera are potentially fatal diseases that affect millions of people every year and result in more than a million mortalities, particularly in the developing countries. Malaria itself causes about one million deaths per year globally, of which more than 90 per cent occur in Africa (WHO/UNICEF, 2003). Global pandemics of cholera have been recorded since the beginning of the 19th century and the 7th pandemic, beginning in 1961, affected Asia, Europe, Africa and Latin America, causing thousands of fatalities (WHO, no date). In the East African countries, malaria is ranked as the primary cause of morbidity and mortality in both children and adults. It causes about 40,000 infant deaths in Kenya each year; in Uganda annual cases of malaria range between 6 to 7 million, with 6500 to 8500 fatalities, and in Tanzania the annual death toll is between 70,000 and 125,000 and accounts for 19 per cent of health expenditure (De Savigny et al, 2004a and b). In the case of cholera, the first epidemic in Africa was reported as far back as 1836 (Rees, 2000). Major outbreaks were next reported in 1970 and affected West Africa (Guinea), the horn of Africa (Ethiopia, Somalia and Sudan) and Kenya (Waiyaki, 1996). The most severe cholera outbreak on the African continent was in 1998, accounting for more than 72 per cent of the global total number of cholera cases and acutely affecting the Democratic Republic of Congo, Kenya, Mozambique, Uganda and the United Republic of Tanzania. Cholera outbreaks in East Africa have been reported to the World Health Organization (WHO) since 1972. In
3232 J&J Climate Change
1/11/07
9:56 am
Page 110
110 Climate Change and Adaptation
the Lake Victoria region of East Africa both malaria and cholera are common, with malaria endemic in the lowlands and epidemic in the highland areas and cholera endemic in the basin since the early 1970s (Rees, 2000). Recently, there has been a marked increase in the severity and intensity of malaria outbreaks in the highlands and cholera outbreaks in the basin. Scientific studies have linked this resurgence in disease episodes to recent climatic anomalies such as above average temperatures and rainfall (Wandiga et al, 2006 and Olago et al, 2007), which in turn has raised serious concerns about the impact of climate change on human health. Further, research studies conducted in this area also suggest that incidences of highland malaria and cholera are likely to increase with future changes in climate (Githeko et al, 2000). In the light of the fact that the present capacity with respect to services, programmes and infrastructure is quite inadequate, the above observations and findings have only created an added uncertainty about the ability of this region to cope with increased disease occurrences in the future. This vulnerability of the Lake Victoria communities is further compounded by other pre-existing social and economic issues such as poverty, increasing population, lack of adequate infrastructure, human stresses on the local environment and the presence of other diseases (HIV/AIDS, diarrhoeal diseases and respiratory diseases). Non-health impacts of climate change on the environment, ecosystems, agriculture and livelihoods and infrastructure could also pose as additional stressors (see Confalonieri and Menne, 2007). Moreover, the absence of any policy interventions for implementing effective adaptive strategies to manage the increasing regional vulnerability to these diseases further reduces the ability of people to cope. This research study was therefore undertaken to attempt to address these climate change-related human health concerns in the Lake Victoria region. Our particular objective is to assess the existing vulnerability to highland malaria and cholera, and to evaluate the present adaptive capacity in terms of its strengths and weaknesses, using specific case studies from Tanzania, Kenya and Uganda.1 It is hoped that the results of this analysis will help to draw out important lessons that can inform the development and institution of appropriate policies and programmes to address the excess risks to communities in the Lake Victoria region of East Africa from future climate change.
Methodology for Assessment Various tools such as the review of published and unpublished material, participatory assessments through focus group discussions, household and key informant interviews, and field observations were employed for the collection and analyses of data for this study. Published and unpublished material was referenced to gather secondary data to facilitate the selection of study sites and to establish disease patterns. This included records from hospitals and health centres, government statistical abstracts, ministerial offices, local institutions and policy documents and the Weekly Epidemiological Review2 (WER-WHO).
3232 J&J Climate Change
1/11/07
9:56 am
Page 111
Climate, Malaria and Cholera in the Lake Victoria Region 111
Climate, hydrology and disease outbreak data were obtained from malaria and cholera study sites in Kenya, Uganda and Tanzania. The malaria study sites selected were Kericho (Kenya), Kabale (Uganda) and Muleba (Tanzania), all of which were highland sites (above 1100m above sea level) (see Figure 6.1). These sites have reported resurgence in malaria epidemics in the last two decades and have also experienced changes in climate since the turn of the 20th century. The cholera study sites in the Lake Victoria basin included Kisumu (Kenya), Kampala (Uganda) and Biharamulo (Tanzania) (see Figure 6.2). These sites also reported a history of cholera epidemics and recent changes in climate. It should be noted that it was important to include households from different elevations because previous studies (Githeko et al, 2006) have shown that prevalence of highland malaria is differentiated by elevation, with 70 per cent, 40 per cent and 30 per cent malaria prevalence among households in valley bottom, hillside and hilltop respectively. Other factors considered in the selection of sites were proximity to a hospital and a meteorological station with reliable data. Informal and formal interviews with individual households and key informants and focus group discussions served as primary data sources for this study. Both qualitative and quantitative methods were used and direct observations were made to supplement interview data (for detailed methodology see Chambers, 1992; Mettrick, 1993; Mikkelsen, 1995). Qualitative data were obtained from group discussions in selected villages, with participants drawn from all the sub-villages and representing various social groups. The aim was to capture the indigenous knowledge base on local vulnerability to climate impacts and adaptation mechanisms (curative and preventive) for malaria and cholera. For the quantitative data, a random sample of 900 households were interviewed, 150 from each of the 6 study sites. Information was collected on socioeconomic characteristics; infrastructure and services; climate-related human health problems and management strategies at the household level; exposure to diseases; and accessibility and availability of health services. Stakeholder workshops were conducted separately in Kenya, Tanzania and Uganda to present the study and research findings, and to obtain participant feedback.
Research Findings Malaria Malaria is caused by the Plasmodium parasite and is transmitted via the bite of the female Anopheles mosquito, which transfers parasite infected blood from a sick person to a healthy person (TDR, no date). Malaria causes more than a million deaths in Africa every year and in the highlands of the Lake Victoria region in East Africa it is known to be highly unstable and epidemic in nature (see Figure 6.1). Such zones of unstable malaria are shown to be more sensitive to climate variability and environmental changes than those where the disease is endemic (Mouchet et al, 1998).
3232 J&J Climate Change
1/11/07
9:56 am
Page 112
112 Climate Change and Adaptation
Figure 6.1 Map showing the Lake Victoria highland malaria region and the studied villages
A significant increase in the severity and frequency of highland malaria epidemics has been observed in recent decades in comparison to the early episodes of the 1920s and 1950s (Garnham, 1945; Roberts, 1964), with virtually no recorded epidemics from the 1960s to the early 1980s. Scientific studies
3232 J&J Climate Change
1/11/07
9:56 am
Page 113
Climate, Malaria and Cholera in the Lake Victoria Region 113
Figure 6.2 Map showing the Lake Victoria cholera region and the studied villages
have closely linked this resurgence of highland malaria epidemics with climate variability (Lepers et al, 1988; Khaemba et al, 1994; Lindsay and Martens, 1998; Malakooti et al, 1998; Mouchet et al, 1998; Some, 1994; Matola et al, 1987; Fowler et al, 1993) and El Niño events that lead to elevated temperatures
3232 J&J Climate Change
1/11/07
9:56 am
Page 114
114 Climate Change and Adaptation
and enhanced precipitation, which create optimal conditions for mosquito breeding and disease transmission (Kilian et al, 1999; Lindblade et al, 2000). It is also projected that climate change is likely to cause an increase in temperature and precipitation, above the minimum temperature and precipitation thresholds of malaria transmission, in various parts of the highland region (Githeko et al, 2000), thus encouraging greater incidences of disease outbreaks in the future.3 Our data for the study period 1996–2001 show the first upsurge in highland malaria cases in Tanzania during the period May to July 1997, and in Kenya from June to July 1997. In Uganda, the number of cases during this period remained below normal. The most significant surge in seasonal outbreaks was observed from January to March 1998 in Tanzania and Kenya, but the trends extended to May of the same year in Kabale, Uganda. In Tanzania the epidemic caused a peak increase in cases by 146 per cent, while in Kenya and Uganda the increases were 630 per cent and 256 per cent respectively. The peak month for hospital admissions in all countries was March. It is likely that the increases in malaria cases in Tanzania and Uganda reflect the true trends (Wandiga et al, 2006) since the Kenyan hospital used in this study is a Mission hospital and also includes cases that should have been admitted to the Kericho District Hospital. This period during 1997–98 when the malaria outbreaks occurred was also an El Niño period during which significantly higher than normal temperature and precipitation were recorded. The highland malaria outbreaks in the Lake Victoria region followed on the heels of this phase of anomalous climate. Data show that climate change is already affecting this area and that the highlands are warming at a greater rate than the lowlands, which has important implications for malaria transmission in the future. A warmer climate would not only reduce mosquito larvae and parasite development times but would also extend the duration of epidemics due to the longer duration of favourable climatic conditions. This could greatly increase mortality and morbidity due to the disease. Other factors, such as environmental and socioeconomic change, deterioration of health care and food production systems, and the modification of microbial/vector adaptation (McMichael et al, 1996; Morse, 1995; Epstein, 1992 and 1995), act as co-contributors in the emergence and spread of malaria. In the East African highlands, human exposure to malaria has also increased due to the increase in population density, which has significantly stressed limited productive land (Lindsay and Martens, 1998), forcing farmers to clear forests and reclaim swamps. Puddles and elevated temperatures result from lost tree and ground cover, providing ideal breeding sites for mosquitoes (Walsh et al, 1993). Even small increases in temperature can greatly hasten mosquito development, resulting in a greater density of mosquito population (Lindblade et al, 2000).4 The vulnerability of highland households is also significantly affected by the scarcity of economic resources, which prevents them from investing in health-coping mechanisms. Wealthy households can afford medical/health and other social services and are less susceptible to disease outbreaks in contrast to poorer households. For example, in Bugarama village in Tanzania, the richer
3232 J&J Climate Change
1/11/07
9:56 am
Page 115
Climate, Malaria and Cholera in the Lake Victoria Region 115
people, locally known as Washongole can better afford to meet health-related costs compared to the poor group, locally know as Abworu (Yanda et al, 2005). Children, pregnant women and the elderly are reported to be most likely to succumb to the disease (Greenwood and Mutabingwa, 2002; McMichael et al, 1996). Trends in malaria cases in our study sites show that children under five years of age display a higher vulnerability to malaria attacks compared to older individuals (see Figures 6.3 and 6.4). This is consistent with the fact that young children have lower immunity (Wandiga et al, 2006; Yanda et al, 2005). Household interviews at the Tanzanian study sites revealed that about 15 per cent of households had lost at least one member due to malaria, with 12 per cent of households reporting to have lost a child member and 3 per cent having lost an adult member (Yanda et al, 2005).
Figure 6.3 Total death toll due to malaria for Ndolage Hospital, Tanzania, in 2001 Source: Yanda et al (2005).
Other contributors that increase vulnerability to malaria in this region include drug resistance, home treatment of malaria and changes in vector biting behaviour (Mbooera and Kitua, 2001). The locals also harbour certain traditional beliefs about malaria, for example, the belief in Muleba, Tanzania, that eating maize meal instead of bananas causes the disease (Mbooera and Kitua, 2001).5 Existing capacity to cope Preparedness in terms of preventing and treating malaria was observed to be substantially better in the lowlands, where malaria is endemic, in comparison to our study sites in the highlands of the Lake Victoria region. In the lowlands
3232 J&J Climate Change
1/11/07
9:56 am
Page 116
116 Climate Change and Adaptation
Figure 6.4 Total death toll due to malaria for Rubya Hospital, Tanzania, in 2001 Source: Yanda et al (2005).
the number of health facilities is considerably higher; anti-malarial drugs and bed nets to protect against mosquito bites are readily available; and awareness about the disease and its prevention, diagnosis and treatment is generally high. In contrast, the capacity to deal with the disease is much lower in the highland areas surveyed, largely because the disease is not common to this region and there is a lack of local awareness about its prevention and treatment. The relative poverty in the highland region also further hampered the ability to deal with the disease. We identified some techniques commonly used by households in our study sites to overcome mosquitoes as an adaptation strategy for malaria. These include (number in brackets indicates percentage of respondents who used the particular technique): • • • • • • •
sleeping by a fire inside the house to avoid mosquitoes – burning of eucalyptus leaves and other herbs was reported to be very effective in chasing away mosquitoes (4 per cent); using mosquito coils (21 per cent); clearing of bushes around homesteads to destroy potential breeding grounds (18 per cent); house screening (16 per cent); draining stagnant water (15 per cent); treating bed nets with insecticides (mainly with Ngao), at most twice a year (15 per cent); and spraying insecticides inside the house (11 per cent).
3232 J&J Climate Change
1/11/07
9:56 am
Page 117
Climate, Malaria and Cholera in the Lake Victoria Region 117
In Tanzania, an increasing use of bed nets and insecticide-treated bed nets was observed, although most people could not afford to buy bed nets for the entire household. On average, each household of six or seven persons owned 1.5 bed nets, which were sufficient for only about 2.4 persons per household (Yanda et al, 2005). In Kenya, few people had access to bed nets, and the proportion of people using bed nets treated with insecticides was even lower. A household could have as many as 16 persons, the average size being 3.7 persons, and the number of bed nets per household ranged from 1 to 6. Only about 28 per cent of the households that use bed nets treat them with insecticides, that too at most twice a year, when the recommended frequency is four times per year (although some of the newer insecticides require only one annual treatment). The lack of use of bed nets, especially insecticide-treated bed nets tends to greatly reduce the efficacy of programmes designed to address malaria, for example, the World Health Organization’s ‘Roll Back Malaria’ campaign, which specifically promotes the use of insecticide treated bed nets for the prevention of malaria as a low-cost and effective measure.6 We also noted that many people in the study sites preferred traditional curative measures (local herbs) to treat malaria rather than visiting health facilities.7 Participants in the stakeholder workshop in Muleba town, Tanzania, estimated that about two-thirds of the malaria patients get cured with traditional medicines. A high reliance on herbal medicines was also reported at the other study sites in Tanzania (in Bukoba Rural District see Mwisongo and Borg (2002) and Bugarama village), with some people relying entirely on herbal remedies. According to Mwisongo and Borg (2002), more than 80 per cent of rural Tanzanians depend on herbal remedies for their primary healthcare. The preference for local herbs for the treatment of malaria over clinical medicines was found to be because (1) they were common, well known and familiar to most people; (2) they were easily available, less expensive, and effective as a first aid before taking the patient to a hospital or health centre; and (3) pregnant women using these herbs against malaria did not encounter problems during delivery, largely due to the fact that these herbs are also effective in reducing complications during pregnancy (Yanda et al, 2005). The herbs are usually prescribed by traditional healers who are reported to be familiar with the symptoms of malaria. Researchers from the National Institute for Medical Research (NIMR) have confirmed that traditional healers, like other cadres in medicine, do possess the knowledge and skills for malaria disease management (diagnosis, treatment and prevention). Laboratory analyses of the traditional herbs by the NIMR revealed that the majority were antimalarial in nature (with variable potency) and could treat other diseases as well (Mwisongo and Borg, 2002). Clinical treatment for malaria is available at the village health facility in every village. An annual contribution of 1500 shillings per household to the health facility ensures medical services (including medication). Treatment is available to non-contributors only under ‘emergency’ situations subject to the condition that they pay their dues on recovery. Only the extremely poor receive treatment free of charge. The receipt of contribution to the village health facility also serves as a guarantee for receiving treatment at government hospitals,
3232 J&J Climate Change
1/11/07
9:56 am
Page 118
118 Climate Change and Adaptation
such as Rubya, and health centres such as Kabare (in Biirabo village) and Chato (in Chato village). Data collected on the preferred means of treating malaria at the household level and analysed on the basis of the level of education of the surveyed people (Table 6.1) reveal that a majority of the people (78 per cent) do consider modern medicine to be the ultimate cure for malaria, despite the high reliance on traditional medicines. No variations in malaria treatment practices at household level could be discerned on the basis of education. A combination of modern and herbal/traditional medicine is locally perceived by some (about 16 per cent of respondent households) to be an important means of combating malaria. Only a small group of people with primary education (about 3 per cent of the total) considered herbal medicine alone to be sufficient for treating malarial disease. Given that the majority believe in the efficacy of modern medicine, the observed high reliance on indigenous medicines is likely to be due to financial barriers.
Table 6.1 Percentage responses of how malaria is treated at the household level by people with different levels of education Level of Education
How Malaria is Treated
Total
Modern Medicine
Herbal Medicine
Modern Medicine/ Tepid Sponging
Modern Medicine/ Herbal Medicine
Modern Medicine/ Prayer
None
12.0
0
0.3
3.3
0.7
16.3
Primary
56.3
3.0
0.7
11.7
1.4
73.0
Secondary
6.0
0
0
0.3
0
6.3
Tertiary
2.3
0.3
0
0
0
2.7
Others
1.4
0
0
0.3
0
1.7
Total
78.0
3.3
1.0
15.6
2.1
100
Local capacity to adapt is also constrained by the absence of adequate early warning mechanisms that could forewarn vulnerable communities about impending malaria epidemics. One such model linking temperature anomalies with malaria outbreaks in the highlands of East Africa has been developed by Githeko and Ndegwa (2001) and has been used with some success in Kenya. However, the current unavailability of such warning systems for the entire region means response to epidemics in most areas is typically reactive and often delayed.
3232 J&J Climate Change
1/11/07
9:56 am
Page 119
Climate, Malaria and Cholera in the Lake Victoria Region 119
Cholera Cholera is caused by the bacteria Vibrio cholerae (Waiyaki, 1996) and is transmitted through the faecal-oral route largely due to (1) the ingestion of contaminated water and food and (2) a lack of scrupulous attention to personal cleanliness (Snow, summarized in Waiyaki, 1996). Cholera causes thousands of fatalities in Africa every year and is endemic to the Lake Victoria basin. It is found to be more common among those living in villages bordering the lake compared to those who live in the hinterland (Shapiro et al, 1999). Specific risk factors associated with cholera outbreaks in the Lake Victoria basin include drinking water from Lake Victoria or from a stream, sharing food with a person with watery diarrhoea, and attending funeral feasts (Shapiro et al, 1999). Several environmental, social and demographic factors are believed to contribute to increasing the risk of cholera outbreaks in this region. Environmental factors include heavy rains and subsequent floods (Desanker and Magadza, 2001; Ulisses and Menne, 2007) and more recently, increased sea surface temperatures have also been implicated as a contributing cause (Desanker and Magadza, 2001; Rodo et al, 2002; Pascual et al, 2000; Colwell, 1996). According to Colwell (1996), increased precipitation increases nutrient laden discharge from rivers and streams into the sea. In combination with warmer sea-surface temperatures, this creates optimum conditions for phytoplankton growth, which subsequently leads to an increase in zooplanktons that feed on the phytoplanktons (Colwell, 1996; Desanker and Magadza, 2001; Cruz et al, 2007; Ulisses and Menne, 2007). Zooplanktons are the preferred host for Vibrio cholerae, and when zooplankton populations increase, so do Vibrio cholerae populations (Colwell, 1996; Patz, 2002; Desanker and Magadza, 2001; Ulisses and Menne, 2007). Contamination of coastal drinking water supplies by sea water containing zooplanktons (due to flood events) leads to the spread of the disease among humans (Colwell, 1996). Similar effects are also thought to be possible for lakes and other inland water bodies (Desanker and Magadza, 2001). Heavy rains and increased sea-surface temperatures are also associated with El Niño, and studies undertaken in Bangladesh have specifically correlated cholera epidemics with El Niño events (Rodo et al, 2002; Pascual et al, 2000).8 In the Lake Victoria basin cholera epidemics recorded during the El Niño years in 1982/83 and 1997/98 were found to coincide with high stream flows and well above normal temperatures. Above normal precipitation and flooding alone, without the above normal temperatures, did not appear to trigger cholera epidemics. Non-epidemic (hygienic) cholera outbreaks were found to be associated with the rainy season when there are above normal rainfall and temperatures, but not as intense as those experienced during El Niño. The morbidity due to such hygienic cholera outbreaks is several orders of magnitude lower than that due to epidemics. Typically cholera epidemics/outbreaks tend to occur anytime between April and December (Wandiga et al, 2006 and Olago et al, 2007), following periods of mainly sustained anomalous high temperatures in the months of January, February and March, and heavy rains. Of the socioeconomic and demographic factors that also play a critical role in encouraging cholera outbreaks in the Lake Victoria basin, poverty is the
3232 J&J Climate Change
1/11/07
9:56 am
Page 120
120 Climate Change and Adaptation
most significant. Most of the communities here rely predominantly on either farm earnings or self-employment, with very few people having a source of steady income. Poverty therefore proves to be a restrictive factor in the prevention of cholera outbreaks, for example, in the inability of local communities to construct durable sanitary facilities due to a lack of adequate financial resources. In the case of Chato village in Tanzania (see Figure 6.2), most people have access to toilet facilities, including pit latrines and a few flush toilets, although the extent of use of these facilities is hard to determine. Moreover, like many other rural areas in the country, there is no sewage removal system. This makes waste disposal difficult and is locally believed to contribute to cholera outbreaks (Yanda et al, 2005) due to the contamination of water supplies. We found that a good number of households in the study area used various treatments for drinking water to prevent cholera incidences: (1) boiling drinking water (56.7 per cent), (2) filtering drinking water (50 per cent), and (3) treating drinking water with chemicals (3.3 per cent). However, in certain villages, particularly those around the lake, the proportion of households drinking untreated water can be as high as 20 per cent. This is due to the perception that piped water or water from the pumped wells is safe for drinking (Yanda et al, 2005). The safety of this water depends, however, on the level of treatment at the source, which may not be adequate, and consuming untreated water greatly increases vulnerability to cholera. Moreover, field observations show that the tap water and pumped water supply is often unreliable; in such situations people are forced to collect water from the lake for drinking and other domestic purposes. This potentially exposes users to cholera pathogens in the lake water, as happened during the cholera outbreak of the 1980s. Table 6.2 provides a list of explanations offered by the surveyed households for not boiling or treating drinking water. Table 6.2 Percentage of reasons/explanations for not treating/boiling drinking water in Chato village Reason/Explanation
Percentage of Responses
Tap water considered safe
63.8
Not used to boiling drinking water
15.9
Lack of fuelwood for boiling water (boiling water is too costly)
10.1
No utensils for boiling water
5.8
Boiling water is tiresome
2.9
Fear of losing the taste of water Total
1.5 100
Finally, it must be noted that currently there are no traditional medicines/cures for cholera in the region. The disease is relatively new to the area and traditional herbalists have yet to invent any curative or preventive medication.
3232 J&J Climate Change
1/11/07
9:56 am
Page 121
Climate, Malaria and Cholera in the Lake Victoria Region 121
Existing capacity to cope Mortality due to cholera is generally observed to be lower compared to malaria in the Lake Victoria region. One possible reason is the relatively higher level of community awareness about the disease. This can be attributed to the awareness programmes instituted by the local administration largely to compensate for the insufficient capacity of public health facilities to handle epidemics. Local civil society organizations are also found to play an important role in helping construct sanitary facilities and water supply sources and in providing free medication during epidemics. According to our survey results, a significant proportion of the respondents (86.6 per cent, 67.2 per cent and 52.7 per cent in Kisumu, Kampala and Biharamulo respectively) exhibited awareness about the influence of weather conditions on the health of household members (Olago et al, 2007). They observed that cholera outbreaks occurred mostly during wet weather and during periods of low water supply associated with dry seasons. According to health workers from Biharamulo, this could be blamed on the sandy nature of the soil pit latrines that often collapsed during the rainy season and contaminated water supplies. During periods of low water supply there is increased dependency on water from the lake, which may be contaminated with cholera pathogens. Periods of low water supply may also lead to a reduced level of sanitary practices, such as hand washing. This knowledge about the association of specific weather patterns with the occurrence of the disease can, however, serve to warn people about possible outbreaks and enable them to be better prepared. Awareness about the need to treat or boil drinking water was also found to be high, despite the fact that some people drink untreated water. Ninety-four per cent of respondents reported that they were knowledgeable about cholera and the consequences of drinking untreated water. Ninety-one per cent knew that drinking untreated water could cause diarrhoeal diseases, such as cholera, and 49 per cent reported to have had at least one member of the household suffering from diarrhoeal diseases during the last 5 years (Yanda et al, 2005) as a result of having consuming untreated water from the lake. This awareness was attributed to various sources such as health service providers, formal sources such as schools, informal networks, media and community awareness campaigns (see Figure 6.5). Most households appeared to be sufficiently informed about medical treatment options for cholera such as the use of antibiotics and oral rehydration salts. However, costs of medical treatment are a barrier and the locals often have to rely on the free medications distributed during epidemics by civil society organizations. Then again, because cholera does not have any local/ indigenous treatment options, people do report to health facilities immediately when there are signs of the disease. This greatly contributes to reducing mortality in comparison to malaria, where many people attempt home treatment before reporting to health facilities. Viewpoints expressed by Chato village (Tanzania) residents on measures that are or should be taken to prevent or treat incidences of cholera include: 1
general cleanliness;
3232 J&J Climate Change
1/11/07
9:56 am
Page 122
122 Climate Change and Adaptation
Figure 6.5 Sources of information on consequence of cholera Source: Yanda et al (2005).
2 3 4
washing hands with soap; boiling drinking water; and using health facilities when sick.
Participants in the stakeholder workshops also identified and suggested strategies to reduce the vulnerability of the community to cholera epidemics, distinguishing between the allocation of responsibility at the village and district levels (Table 6.3). Some measures such as observing proper hygiene and protection of and proper management of water sources were identified for implementation on a routine basis, while other measures would have to be implemented specifically during cholera outbreaks, for example, promptness in reporting cholera outbreaks and sending sick people to health centres and hospitals for treatment.9 In the case of cholera, no early warning mechanisms about potential disease outbreaks have so far been instituted to allow residents to be better prepared to cope. Lipp et al (2002) have put forth the suggestion that the spatial and temporal correlation of Vibrio cholerae with El Niño, or its proxies and predictors, could help to create such a forewarning mechanism and provide an effective way to prevent exposure to cholera in vulnerable regions. However, any such system has yet to be developed and implemented on the ground.
Conclusions and Recommendations The findings of our assessment indicate that communities in the Lake Victoria region of Africa display a significantly high vulnerability to climate change/ climate variability-induced diseases such as malaria and cholera. The capacity of these communities for the prevention and treatment of such diseases is in contrast quite low, as a result of which mortality and morbidity in the event of disease outbreaks tends to be high.
3232 J&J Climate Change
1/11/07
9:56 am
Page 123
Climate, Malaria and Cholera in the Lake Victoria Region 123
Table 6.3 Cholera control strategies suggested by stakeholders Village Level
District Level
Construction and use of improved toilets
Awareness campaigns on how to prevent cholera outbreak
Use of clean and safe water (boiled)
Outbreak preparedness (districts need to have plans for controlling cholera in the event of outbreaks)
Use of clean and safe water (boiling cooking and drinking water)
Outbreak preparedness
Proper collection and disposal of wastes ● Collecting solid wastes in pits and burying the pits when they fill up ● Burning the wastes, whenever possible
● ●
Planning for cholera control strategies in cooperation with community leaders Providing equipment necessary to keep the environment clean and improve the hygienic conditions
Protection and proper management of water sources
Recruit more health staff
Cost-sharing in the management of water sources
Undertake environmental assessment to ascertain causes of problems and how to control the situation
Washing hands before taking any food Washing hands after every use of the toilet Cleanness of household utensils
Establish temporary camps for patients during cholera outbreaks
Community to report promptly when there is a cholera outbreak
Ensure prompt response to cholera outbreak
Sick people to report promptly at health centres and hospitals for treatment
Undertake laboratory analysis to confirm outbreak
Communities in this region are found to display a greater vulnerability to climate-induced highland malaria outbreaks in comparison to cholera outbreaks in the basin. This is largely due to a relative lack of awareness about the disease and a relative lack of preparedness in terms of availability of health care facilities and medication. The predominant reliance of communities on local herbs to treat malaria as a first and often only recourse (as opposed to clinical treatment) also contributes to increased mortality and morbidity, especially due to the herbs’ variable efficacy. In comparison, greater local knowledge about the prevention and treatment of cholera and the absence of any indigenous treatment options tends to ensure more prompt clinical attention. Vulnerability to malaria is also observed to be higher in the highland region, where the disease is epidemic, in comparison to the lowlands, where the disease is endemic. The endemicity of malaria in the lowlands has ensured
3232 J&J Climate Change
1/11/07
9:56 am
Page 124
124 Climate Change and Adaptation
greater awareness and better availability of health care services and medications, while in the highlands there is a lack of experience in dealing with the disease due to its recent resurgence, which translates into a lack of knowledge about the disease and a lack of accessibility to adequate health care and treatment. The poverty of the population in this region in comparison to the lowlands is also a critical factor that greatly impedes its capacity to invest in disease prevention (for example, use of insecticide-treated nets) or in disease treatment. The inability to pay also creates a high dependence on indigenous medicine despite the strong belief in the efficacy of modern medicine. In the case of cholera, although awareness about the disease and its treatment is relatively high and mortality is subsequently lower, many still tend to follow unsanitary practices and drink untreated water from the lake. Poverty is once again observed to be an important factor in the inability of the people to invest in better sanitation and appropriate medication to treat the disease. Many rural areas also do not have adequate sanitary services such as sewage removal services, which can cause contamination of drinking water supplies. Ignorance about the safety of water supplies among the rural population and the scarcity of well-equipped health centres tends to compromise the health of the people in the event of outbreaks and epidemics. Common to both highland malaria and cholera epidemics is the need for stronger and better public information campaigns to ensure that the rural population in this area is sufficiently informed about strategies for the prevention and treatment of these diseases. Second, better infrastructure in terms of sanitation and living conditions, access to basic services such as clean water supplies, and access to adequate health services and medication are critical for preventing and addressing disease epidemics in the region. Third, a better understanding of the correlation of climate phenomena with disease outbreaks would help to establish early warning mechanisms in order to allow for advance preparedness in the community in terms of strategies for prevention and treatment. Githeko and Ndegwa (2001) have initiated this process to some extent by applying such a model for malaria with some effectiveness in Kenya. The expansion of this model to other areas in the highland region could potentially help to greatly reduce the severity of malaria outbreaks in the future. No such system exists for cholera at the moment. Finally, strategies for poverty reduction, by developing and providing livelihood opportunities, could greatly help communities to invest in better health coping mechanisms, for example, the use of insecticide-treated nets in the case of malaria and effective sanitary facilities in the case of cholera. Better income security would also enable communities to seek prompt clinical treatment and help reduce mortality due to these diseases. Important lessons can be learnt from existing adaptation strategies that have been found to work. For example, the better adaptation to malaria in the lowlands of East Africa could instruct the development of better adaptation strategies in the highlands. In the case of cholera, community awareness campaigns can be further strengthened to ensure adequate sanitary precautions. Additionally, success stories from programmes dealing with other diseases in the region could help inform strategies for adaptation to climate-related
3232 J&J Climate Change
1/11/07
9:56 am
Page 125
Climate, Malaria and Cholera in the Lake Victoria Region 125
malaria and cholera outbreaks. In designing future adaptation programmes a range of socioeconomic factors and demographic trends would also have to be accounted for, because they influence social responses to climatic stresses and would be critical in the determination of strategies to cope with the impacts. National and local governments have a vital role to play in increasing the adaptive capacity of communities in the Lake Victoria region to climate related disease epidemics and outbreaks. In addition, various civil society organizations and relevant regional, national and international institutions and agencies could potentially make important contributions to this process. There are presently no comprehensive government programmes or fiscal facilities in place for climate-related disaster preparedness. Many of the current preventive and curative programmes for malaria and cholera that are run by governments or civil society organizations often predominantly rely on external sources of assistance, whose long-term sustainability is not always guaranteed. Therefore, finding ways to develop, institute and sustain programmes that would strengthen local capacity to adapt to the health impacts of climate change or climate variability is critical. In short, more effective adaptation to the health impacts of climate variability and future climate change in the Lake Victoria region would entail a combination of a multitude of strategies, including increased awareness, better health care, improved sanitation, adequate infrastructure, development of early warning mechanisms, creation of livelihood opportunities, and increased support from national and international governments and institutions.
Notes 1 2
3
4
Details of this study have been described in papers by Wandiga et al (2006), Olago et al (2007) and Yanda et al (2005). International health regulations require national health administrators to report the number of indigenous and imported cases of cholera and deaths to the World Health Organization (WHO) within 24 hours of receiving such information. This cholera data are then reported in the Weekly Epidemiological Review (WER) detailing the date and geographical location. Hay et al (2002) have disputed the association of malaria outbreaks with climate variability and change. They analysed climate data for the diurnal temperature range spanning the 1950–1959 period and found no significant changes in temperature or vapour pressure at any of the highland sites that had reported high malaria incidences. However, these findings have been challenged by Patz et al (2002), who claim that the use of a downscaled gridded climate data set by interpolating it to specific sites by Hay et al (2002) ignores climate dependencies on local elevation, which compromises the accuracy of the results. In opposition to Hay et al’s (2002) findings, Patz et al (2002) have reported a warming trend at highland sites in East Africa coinciding with an increasing trend in the incidence of malaria at those sites, using data obtained from the specific locations over the specific time period. Similar associations have also been reported by other scientists (Ulisses and Menne, 2007). In contrast to reclaimed swamps, the natural swamps in the valley bottoms contain Papyrus, which, due to its cooling properties and natural oil secretions, is believed to inhibit anopheles mosquito development (Lindblade et al, 2000; Reiter, 2001).
3232 J&J Climate Change
1/11/07
9:56 am
Page 126
126 Climate Change and Adaptation
5
6
7 8
9
Papyrus and other naturally occurring grasses tend to bring down temperatures in the swamp, which slows down mosquito development (Lindblade et al, 2000). Papyrus also produces natural oils, which are believed to inhibit anopheles mosquito development (see Reiter, 2001). In Muleba, Tanzania, maize meal is usually eaten only during periods of food shortages that occur due to above and/or below average rains (for example, the El Niño rains or La Niña droughts). Such climatic conditions are also conducive to the occurrence of malaria, which is therefore more rampant at such times. Additionally, poor nutrition due to the food shortage further makes people more susceptible to malaria, especially children who tend to become anaemic due to nutritional shortages (Mwisongo and Borg, 2002). As a result the traditional understanding that eating maize meal causes malaria has developed. In Kenya, the consumption of an edible oil called chipsy is associated with the occurrence of malaria simply because this oil was first introduced in Kenya in 1990, which was also the year of the El Niño rains and malaria episodes. Similarly, in Uganda, supernatural forces are considered to cause the convulsions that are associated with malarial complications and traditional medicines are considered to be the only solution (Nuwaha, 2002). This often leads to either no or delayed medical care and increases morbidity and mortality due to the disease. The World Health Organization’s (WHO) programme ‘Roll Back Malaria’ has been adopted by most countries in Africa. The three East African governments actively promote this programme, whose objectives towards malaria eradication are to increase the use of ITNs, early diagnosis and treatment of malaria, and the use of effective anti-malarial drugs. This programme has attracted several local and international civil societies. One such non-governmental organization active in East Africa is Population Services International (PSI), which receives financial support from both the British and American Governments. Its stated objective is to increase the use, ownership and availability of ITNs in Kenya, Uganda and Tanzania within 15 minutes’ walk of malaria endemic areas. Promotions of ITNs are prevalent in most market centres in East Africa. However, the cost of a subsidized ITN is still US$1.50, putting it beyond the reach of households living below the poverty line. The plants primarily used for this purpose include (using Haya names) Mbilizi, Kajule, Nkaka, Ikintuntumwa and Mwarobaini (Yanda et al, 2005), although the level of success may vary among the different varieties. A study in Chesapeake Bay though found the link between temperature and cholera in suboptimal environments (freshwater or high salinity) to be weak (Louis et al, 2003). Lake Victoria, which is a freshwater lake, has a salinity ranging between 3.9 and 7.0, much lower than the optimal ocean salinity. However, these research findings were based on a limited time record (2 years) and did not account for indicators associated with nutrient load and zooplanktons (for example, discharge and precipitation) (Louis et al, 2003). The Chato Health Center that caters for the entire Chato Division (and other neighbouring divisions) is conveniently located right within Chato village.
References Chambers, R. (1992) ‘Rural appraisal: Rapid, relaxed and participatory’, Discussion Paper No 311, Institute of Development Studies, Brighton, Sussex, UK Checkley, W., L. D. Epstein, R. H. Gilman, D. Figueroa, R. I. Cama, J. A. Patz and R. E. Black (2000) ‘Effects of El Niño and ambient temperature on hospital admissions for diarrhoeal diseases in Peruvian children’, The Lancet, vol 355, pp442–450
3232 J&J Climate Change
1/11/07
9:56 am
Page 127
Climate, Malaria and Cholera in the Lake Victoria Region 127 Colwell, R. (1996) ‘Global climate and infectious disease: The cholera paradigm’, Science, vol 274, pp2025–2032 Confalonieri, U. and B. Menne (2007) ‘Human health’, in S. Solomon and D. Qin (eds) Climate Change 2007: Climate Change Impacts, Adaptation and Vulnerability, Contribution of Working Group II to the Fourth Assessment Report, Intergovernmental Panel on Climate Change (IPCC), Cambridge University Press, Cambridge, UK Cruz, R. V., H. Harasawa, M. Lal and W. Shaohong (2007) ‘Asia’, in S. Solomon and D. Qin (eds) Climate Change 2007: Climate Change Impacts, Adaptation and Vulnerability, Contribution of Working Group II to the Fourth Assessment Report, Intergovernmental Panel on Climate Change (IPCC), Cambridge University Press, Cambridge, UK De Savigny, D., E. Mewageni, C. Mayombana, H. Masanja, A. Minhaji, D. Momburi, Y. Mkilindi, C. Mbuya, H. Kasale, H. Reid and H. Mshinda (2004a) Care Seeking Patterns in Fatal Malaria: Evidence from Tanzania, Tanzania Essential Health Interventions Project (TEHIP), Rufiji Demographic Surveillance System, Tanzania, Ifakara Health Research and Development Centre, Tanzania, Tanzania Ministry of Health and International Development Research Centre, Canada De Savigny, D., E. Mewageni, C. Mayombana, H. Masanja, A. Minhaji, D. Momburi, Y. Mkilindi, C. Mbuya, H. Kasale, H. Reid and H. Mshinda (2004b) ‘Highland malaria in Uganda: Prospective analysis of an epidemic associated with El Niño’, Transactions of the Royal Society of Tropical Medicine and Hygiene, vol 93, pp480–487 Desanker, P. and C. Magadza (2001) ‘Africa’, in J. J. McCarthy, O. F. Canziani, N. A. Leary, D. J. Dokken and K. S. White (eds) Climate Change 2001: Impacts, Adaptation and Vulnerability, Contribution of Working Group II to the Third Assessment Report, Intergovernmental Panel on Climate Change (IPCC), Cambridge University Press, Cambridge ,UK Epstein, P. R. (1992) ‘Cholera and environment’, Lancet, vol 339, pp1167–1168. Epstein, P. R. (1995) ‘Emerging diseases and ecosystem instability: New threats to public health’, American Journal of Public Health, vol 85, pp168–172, available at http://extdr/offrep/afr Fowler, V. G. Jr., M. Lemnge, S. G. Irare, E. Malecela, J. Mhina, S. Mtui, M. Mashaka and R. Mtoi (1993) ‘Efficacy of chloroquine on Plasmodium falciparum transmitted at Amani, eastern Usambara mountains, northeast Tanzania: An area where malaria has recently become endemic’, Journal of Tropical Medicine and Hygiene, vol 6, pp337–345 Garnham, P. C. C. (1945) ‘Malaria epidemics at exceptionally high altitudes in Kenya’, British Medical Journal, vol 11, pp45–47 Githeko, A. K., S. W. Lindsay, U. E. Confaloniero and J. A. Patz (2000) ‘Climate change and vector-borne disease: A regional analysis’, Bulletin of the World Health Organization, vol 78, pp1136–1147 Githeko, A. K., J. M. Ayisi, P. K. Odada, F. K. Atieli, B.A. Ndenga, I. J. Githure and G. Yan (2006) ‘Topography and malaria transmission heterogeneity in the western Kenya highlands: Prospects for focal vector control’, Malaria Journal, vol 5, p107 Greenwood, B. and T. Mutabingwa (2002) ‘Malaria in 2002’, Nature, vol 415, pp670–672 Hay, S. I., M. Simba, M. Busolo, A. M. Noor, H. L. Guyatt, S. A. Ochola and R. W Snow (2002) ‘Defining and detecting malaria epidemics in the highlands of western Kenya’, Emerging Infectious Diseases, vol 8, pp555–562 Hulme, M. (1996) ‘Recent climatic change in the world’s drylands’, Geographical Research Letters, vol 23, pp61–64 Intergovernmental Panel on Climate Change (IPCC) (2001). Climate Change 2001: Impacts, Adaptation and Vulnerability, J. McCarthy, O. F. Canziani, N. Leary, D.
3232 J&J Climate Change
1/11/07
9:56 am
Page 128
128 Climate Change and Adaptation Dokken and K. S. White (eds) contribution of Working Group II to the Third Assessment Report, Cambridge University Press, Cambridge, UK Khaemba, B. M., A. Mutani and M. K. Bett (1994) ‘Studies of anopheline mosquitoes transmitting malaria in a newly developed highland urban area: A case study of Moi University and its environs’, East African Medical Journal, vol 3, pp159–164 Kilian, A. H. D., P. Langi, A. Talisuna and G. Kabagambe (1999) ‘Rainfall pattern, El Niño and malaria in Uganda’, Transactions of the Royal Society of Tropical Medicine and Hygiene, vol 93, pp22–23 Lepers, J. P., P. Deloron, D. Fontenille and P. Coulanges (1988) ‘Reappearance of Falciparum malaria in central highland plateaux of Madagascar’, Lancet, vol 12, pp585–586 Lindblade, K. A., E. D. Walker, A. W.Onapa, J. Katunge and M. L. Wilson (2000) ‘Land use change alters malaria transmission parameters by modifying temperatures in a highland area of Uganda’, Tropical Medicine and International Health, vol 5, pp263–274 Lindsay, S. W. and W. J. M. Martens (1998) ‘Malaria in the African highlands: Past, present and future’, Bulletin of the World Health Organization, vol 76, pp33–45 Lipp, E. K., A. Huq and R. R. Colwell (2002) ‘Effects of global climate on infectious diseases: The cholera model’, Clinical Microbiology Review, vol 15, pp757–770 Louis, V. R., E. Russek-Cohen, N. Choopun, I. N. Rivera, B. Gangle, S. C. Jiang, A. Rubin, J. A. Patz, A. Huq and R. Colwell (2003) ‘Predictability of Vibrio cholerae in Chesapeake Bay’, Applied and Environmental Microbiology, vol 69, pp2773–2785 Malakooti, M. A., K. Biomndo and G. D. Shanks (1998) ‘Re-emergence of epidemic malaria in the highlands of western Kenya’, Emerging Infectious Diseases, vol 4, pp671–676 Matola, Y. G., G. B. White and S. A. Magayuka (1987) ‘The changed pattern of malaria endemicity and transmission at Amani in the eastern Usambara mountains, northeastern Tanzania’, Journal of Tropical Medicine and Hygiene, vol 3, pp127–134 Mbooera, L. E. G. and A.Y. Kitua (2001) ‘Malaria epidemics in Tanzania: An overview’, African Health Sciences, vol 8, pp17–23 McMichael, A. and A. Githeko (2001) ‘Human health’, in J. J. McCarthy, O. F. Canziani, N. A. Leary, D. J. Docken and K. S. White (eds) Climate Change 2001: Impacts, Adaptation and Vulnerability, contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change (IPCC), Cambridge University Press, Cambridge, UK McMichael, A. J., A. Hames, R. Scooffand and S. Covats (eds) (1996) ‘Climate change and human health: An assessment’, report prepared by task group on behalf of the World Health Organization, the World Meteorological Organization and the United Nations Environment Programme, Geneva, Switzerland Mettrick, H. (1993) Development Oriented Research in Agriculture: ICRA Textbook, The Centre for Development Oriented Research in Agriculture, Wageningen, The Netherlands Mikkelsen, B. (1995) Methods for Development Work and Research: A Guide for Practitioners, Sage Publications, New Delhi Mocumbi, P. (2004) ‘Plague of my people’, Nature, vol 430, p925 Morse, S. S. (1995) ‘Factors in the emergence of infectious diseases’, Emerging Infectious Diseases, vol 1, pp7–15 Mouchet, J., S. Manuin, S. Sircoulon, S. Laventure, O. Faye, A. W. Onapa, P. Carnavale, J. Julvez and D. Fontenille (1998) ‘Evolution of malaria for the past 40 years: Impact of climate and human factors’, Journal of the American Mosquito Control Association, vol 14, pp121–130
3232 J&J Climate Change
1/11/07
9:56 am
Page 129
Climate, Malaria and Cholera in the Lake Victoria Region 129 Mwisongo, A. and J. Borg (eds) (2002) Proceedings of the Kagera Health Sector Reform Laboratory 2nd Annual Conference, Ministry of Health, Dar es Salaam, United Republic of Tanzania Olago, D., M. Marshall, S. O. Wandiga, M. Opondo, P. Z. Yanda, R. Kangalawe, A. Githeko, T. Downs, A. Opere, R. Kabumbuli, E. Kirumira, L. Ogallo, P. Mugambi, E. Apindi, F. Githui, J. Kathuri, L. Olaka, R. Sigalla, R. Nanyunja, T. Baguma and P. Achola (2007) ‘Climatic, socio-economic and health factors affecting human vulnerability to cholera in the Lake Victoria basin, East Africa’, Ambio, vol 36, no 4, pp350–358 Pascual, M., R. Xavier, S. P. Ellner, R. Colwell and M. J. Bouma (2000) ‘Cholera dynamics and El Niño – Southern oscillation’, Science, vol 289, pp1766–1769 Patz, J. (2002) ‘A human disease indicator for the effects of recent global climate change’, Proceedings of the National Academy of Sciences, vol 99, pp12,506– 12,508 Patz, J. A., K. Strzepek, S. Lele, M. Hedden, S. Greene, B. Noden, S. I. Hay, L. Kalkstein and J. C. Beier (1998a) ‘Predicting key malaria transmission factors, biting and entomological inoculation rates, using modelled soil moisture in Kenya’, Tropical Medicine and International Health, vol 3, pp818–827 Patz, J. A., K. Strzepek, S. Lele, M. Hedden, S. Greene, B. Noden, S. I. Hay, L. Kalkstein and J. C. Beier (1998b) ‘Predicting key malaria transmission present and future’, Bulletin of the World Health Organization, vol 76, pp33–45 Patz, J. A., M. Hulme, C. Rosenzweig, T. D. Mitchell, R. A. Goldberg, A. K. Githeko, S. Lele, A. J. McMichael and D. Le Sueur (2002) ‘Regional warming and malaria resurgence. Brief communications’, Nature, vol 420, pp627–628 Rees, P. H. (2000) ‘Editorial – Cholera’, East African Medical Journal, vol 77, pp345–346 Reiter, P. (2001) ‘Climate change and mosquito-borne diseases’, Environmental Health Perspectives, vol 109, Supplement 1, pp141–161 Roberts, J. M. D. (1964) ‘Control of epidemic malaria in the highlands of western Kenya, Part I: Before the campaign’, Journal of Tropical Medicine and Hygiene, vol 61, pp161–168 Sachs, J. and P. Malaney (2002) ‘The economic and social burden of malaria’, Nature, vol 415, pp680–685 Shapiro, R. L., M. R. Otieno, P. M. Adcock, P. A. Phillips-Howard, W. A. Hawley, L. Kumar, P. Waiyaki, B. L. Nahlen and L. Slutsker (1999)‘Transmission of epidemic Vibrio cholerae in rural western Kenya associated with drinking water from Lake Victoria: An environmental reservoir for cholera?’ American Journal of Tropical Medicine and Hygiene, vol 60, pp271–276 Some, E. S. (1994) ‘Effects and control of highland malaria epidemics in Uasin Gishu District, Kenya’, East African Medical Journal, vol 71, pp2–8 TDR (Special Programme for Research and Training in Tropical Diseases) (no date) ‘Malaria: Disease information’, available at www.who.int/tdr/diseases/ malaria/diseaseinfo.htm, accessed 31 January 2007 Waiyaki, P. G. (1996) ‘Cholera: Its story in Africa with special reference to Kenya and other East African countries’, East African Medical Journal, vol 73, pp40–43 Walsh, J. F., D. H. Molyneux and M. H. Birley (1993) ‘Deforestation: Effects on vectorborne disease’, Parasitology, vol 106, ppS55–S75. Wandiga, S. O., M. Opondo, D. Olago, A. Githeko, F. Githui, A. Opere, P. Z. Yanda, R. Kangalawe, R. Kabumbuli, E. Kiramura, J. Kathuri, E. Apindi, L. Olaka, L. Ogallo, P. Mugambi, R. Sigalla, R. Nanyunja, T. Baguma, P. Achola, M. Marshall and T. Downs (2006) ‘Final report, Project AF91, Assessments of Impacts and Adaptations to Climate Change’, The International START Secretariat, Washington, DC WHO (World Health Organization) (no date) ‘Global epidemics and impact of
3232 J&J Climate Change
1/11/07
9:56 am
Page 130
130 Climate Change and Adaptation cholera’, available at www.who.int/topics/cholera/impact/en/index.html, accessed 3 April 2007 WHO/UNICEF (2003) ‘Africa malaria report’, WHO/UNICEF, Geneva, Switzerland, available at www.rbm.who.int/amd2003/amr2003/amr_toc.htm Yanda, P. Z., R. Y. M. Kangalawe and R. J. Sigalla (2005) ‘Climatic and socio-economic influences vulnerability on cholera in the Lake Victoria region’, AIACC Working Paper No 12, International START Secretariat, Washington, DC
3232 J&J Climate Change
1/11/07
9:56 am
Page 131
7
Making Economic Sense of Adaptation in Upland Cereal Production Systems in The Gambia Momodou Njie, Bernard E. Gomez, Molly E. Hellmuth, John M. Callaway, Bubu P. Jallow and Peter Droogers
Introduction The Gambia lies in the Sahel region, where rainfall is directly linked to the zonal position of the Inter-Tropical Discontinuity and is highly sensitive to perturbations of the global monsoon circulation. To cope with seasonal variability associated with such perturbations, Gambian farmers have traditionally used a number of strategies. But how successful these are is open to debate, considering rural–urban migration trends in the past three decades. In the face of imminent threats from climate change, adaptation strategies inspired and informed by past and current coping strategies are reported in The Gambia’s Initial National Communication (GOTG, 2003), but their performance remains to be evaluated in terms of economic viability or impact on national food security. In this chapter, the SWAP-WOFOST model is used in combination with CEREBAL to investigate the impact of climate change on The Gambia’s cereal balance under different management options. Most significantly, the analysis looks into the economic efficiency of specific management options, subject to social and political acceptability. Before and immediately after independence in 1965, agricultural policy in The Gambia was primarily driven by the need to generate foreign exchange to pay for goods and services required for economic development. Over the years, this paradigm was reinforced by buoyant world market prices for groundnuts and cheap food prices (Carney, 1986) and only began to lose ground after protracted drought, and economic hardship experienced by farmers in the last two decades. The gradual move away from cash to cereal crops is clearly shown in agricultural statistics (Department of Planning, 2001, 2003 and 2005). Cereal production is mainly for consumption, but surplus production by individual farming households, or dabada, is sold off in local grain markets. Cereals
3232 J&J Climate Change
1/11/07
9:56 am
Page 132
132 Climate Change and Adaptation
grown include rice (Oryza sativa), millet (Pennisetum typhoides), sorghum (Sorghum bicolor) and maize (Zea mays), with millet accounting for nearly 60 per cent of area planted and slightly more than 50 per cent of total cereal production (Department of Planning, 2005). The River Gambia divides the country into two strips of land no wider than 30km at any transect. Over 48 per cent of the total land area of The Gambia is below 20m above mean sea level, and nearly one-third of the country is at or below 10m above sea level. In general, elevation increases with axial distance from the river. Geomorphological units are described as lowland and uplands. Weathered tropical soils found in the uplands are not very fertile but are well drained. In contrast, the fine-textured soils of the lowlands are poorly drained. This juxtaposition of topography, pedology and hydrology leads to spatial differentiation of cereal cultivation areas. In general, the River Gambia valley and adjacent swamps are used to cultivate rice, whereas the plateau is the center for millet cultivation. The traditional agricultural system depends on extensive land use, using little agricultural input. To this effect, successful crop production is dependent on rainfall and favourable environmental conditions. Farmers’ vulnerability is systemic and inextricably linked to climate variability, natural soil fertility and economy-wide policy framework. A sharp and significant drop in average rainfall in The Gambia since the late 1960s has put tremendous pressure on crop production. Lowland rice production, in particular, is under heavy pressure from reduced flood duration and frequency, saline intrusion, soil acidification and deposition of sediments eroded from uplands. To fix some ideas, protracted drought and saltwater intrusion in cropland areas have resulted in a 50 per cent decline in the area under rice cultivation (Department of Planning, 2001). Declining soil fertility in uplands is forcing fundamental changes in production such as the use of marginal land, reduced fallows and deforestation to compensate for low productivity. Although tidal irrigation in lowlands and introduction of improved rice cultivars represent opportunities for increasing total cereal production, water still remains the limiting factor to expansion. Njie (2002a) and, more recently, Verkerk and van Rens (2005), demonstrate the environmental and economic risks of expanding irrigation schemes under natural flow conditions in the River Gambia.
Adaptation Strategies In response to climate hazards, farmers have experimented and adopted a raft of strategies to cope with erratic rainfall patterns. An insightful study by Jallow (1995) divides these into risk-aversion and risk-management strategies. The first category includes crop diversification, crop selection and plot dispersal. If these fail to provide adequate insurance, farmers, depending on their circumstances, sell off assets, use kinship networks, receive government assistance in the form of food aid and harvest natural forest food to get over the period of hardship.
3232 J&J Climate Change
1/11/07
9:56 am
Page 133
Making Economic Sense of Adaptation in Upland Cereal Production Systems 133
Cole et al (2005) report adaptation strategies at government level as including sustaining support for farmer risk-aversion strategies through appropriation and dissemination of high-yield cultivars, providing engineering and technical leadership in land rehabilitation/conservation and water control, providing scientific advice in the form of seasonal rainfall forecasting and, when all these fail, providing disaster relief with the assistance of relevant United Nations agencies, multilateral institutions and non-governmental organizations. Although these strategies worked sufficiently well in the past, a critical evaluation in the context of climate change and socioeconomic trends casts doubts on the prospects of some of them. The subsisting problem is how to maintain or increase production under adverse conditions. Limiting constraints include further decline in rainfall, per capita availability of land, land degradation, widespread poverty and social mutations. Our point of departure in this study is the understanding that crop production may be increased by increasing cultivated area and/or increasing crop yields. Through screening and integration of previously proposed adaptation options (Jallow, 1995; GOTG, 2003), we identify (1) crop breeding and selection, (2) crop fertilization and (3) irrigation as the most comprehensive, no-regrets, flexible strategies to improve crop yields. The main argument in favour of crop breeding and selection is that of probable decline in rainfall and increased variability. On the other hand, promotion of crop fertilization as an adaptation strategy is influenced by continuous decrease in available prime land and concurrent degradation of arable land. Moreover, land that requires some amendments represents the second largest class of agriculturally suitable land (National Environment Agency, 1997). Irrigation provides a sorely needed means of mitigating impacts of spatial and temporal variability of rainfall and offers the potential for extending the growing season and expanding total cultivated area. Except perhaps for irrigation of upland cereals, these strategies are not entirely new. What is novel about our restatement of these strategies is the systematic, rigorous and quantitative approach used in this study.
Analytical Framework The analytical framework used in this study (see Figure 7.1) is built around two key components: (1) crop modelling and (2) economic feasibility analyses using SRES A2 forcing and outputs from HadCM3 and ECHAM4 global circulation models (GCMs) adjusted to The Gambia’s climate. Every effort is made to ensure that socioeconomic scenarios prescribed are consistent with SRES A2.
Climate scenarios The climate change scenarios used in the analysis are derived from projections of the Max Planck Institute’s ECHAM4 general circulation model and the Hadley Centre’s HadCM3 model for the A2 emission scenario of the Intergovernmental Panel on Climate Change (IPCC). The A2 scenario of
3232 J&J Climate Change
1/11/07
9:56 am
Page 134
134 Climate Change and Adaptation
SRES GCMs Downscaling & Post-processing Climate scenarios SWAP WOFOST
Crop Modelling
Adaptation/ Management Strategies
Cereal Balance Computations CEREBAL
Economic Feasibility Analysis
Population Public Policy
Land and Water Resources
Technology
Socioeconomic scenarios
Figure 7.1 Analytical framework for assessing economic feasibility of climate change adaptation
greenhouse gas emissions is characterized by high population growth, slow and regionally oriented economic growth, and slow technological change (Nakicenovic and Swart, 2000). Details of the procedure for downscaling the climate projections of the global scale models to the study region can be found in Gomez et al (2005). Both the ECHAM4 and HadCM3 models project an average temperature rise of 3 to 4ºC by 2100 but differ significantly in their projections of precipitation changes. Whereas ECHAM4 shows no significant change in mean rainfall, and some increase in extreme values, HadCM3 shows a drastic drop of 400mm in annual rainfall in the distant future (2070–2099). This situation presents us with two scenarios: (1) global warming only and (2) global warming and increasing aridity. From recent changes in Sahel rainfall, we take a neutral position and assume both are plausible scenarios. To feed the climate model-derived information into the environmental and biophysical models used in this study, monthly data are transformed into daily values by interpolation and statistical modelling (Richardson, 1981; Racsko et al, 1991).
Socioeconomic scenarios The upper envelope of population projections from different growth models (Njie, 2002b) is used in this study. This corresponds to the cohort survival method that assumes unchanging fertility rates. Under this scenario, total
3232 J&J Climate Change
1/11/07
9:56 am
Page 135
Making Economic Sense of Adaptation in Upland Cereal Production Systems 135
population of The Gambia grows from under 2 million persons to over 11 million by the year 2100. Rural to urban population ratio, currently 50:50, is assumed to evolve linearly over time to 20:80 by the end of the century. In this scenario, however, absolute decline in rural population will occur late in this century. Bolstered by food security and poverty alleviation policies, agricultural production is, therefore, expected to be a dominant factor in the economy. Land availability, a crucial factor in cereal production, is assessed in light of other competing uses – economic, nature conservation and residential. Priorities of use, regeneration and degradation rates, together with suitability for agriculture, are also incorporated in the land availability calculus. In like manner, the feasibility of putting 20 per cent of millet production under irrigation is assessed by comparing projected water demand with renewable water resources. The reader may note that 20 per cent irrigation is a benchmark only so far achieved in developed countries. For the adaptation/management strategies analysed in this study, the issue of significant technological change and innovation is only considered for research and development outcomes of regional crop breeding programmes. No major revolution is expected in already mature irrigation technology, but there is room for improvement of water delivery efficiencies. Costs are also likely to change, but projections are not attempted because of large uncertainties and use of constant prices for other variables in the study. Little sophistication is required for fertilizer application. Public policy in food security and poverty alleviation seeks to ‘stay ahead of the curve’, so to speak. Policy variables used in this study include per capita cereal consumption of 250kg per year, based on the upper limit of local production and imports from 1995 to the present (Department of Planning, 2005). Strategic food reserves are defined as food reserves sufficient for one month up to a maximum of two months but not more. As already mentioned, we ensure at the problem specification stage that conflicts between land and water management and other policies are eliminated. At the analysis stage, changes in policy variables and food preferences are made to see what impact they might have on the economic performance of the adaptation strategies and food security.
Crop modelling Crop yields are simulated with the linked SWAP-WOFOST models (Feddes et al, 1978; Van Dam et al, 1997; Kroes et al, 1999). The SWAP (soil water atmosphere plant) model simulates one-dimensional water, solute and heat transport in saturated and unsaturated soils (Feddes et al, 1978; Droogers, 2000). WOFOST (the world food studies model) simulates the phenological development of a crop from emergence to maturity on the basis of the crop’s genetic attributes, and environmental conditions (Spitters et al, 1989; Supit et al, 1994). In the SWAP model, rainfall, irrigation water and solar radiation reaching the soil surface are related to the leaf area index (LAI) of the crop. Solute, heat and water transport within the soil, governed by laws of mass and energy conservation, is modulated by heat and moisture transmission and storage properties,
3232 J&J Climate Change
1/11/07
9:56 am
Page 136
136 Climate Change and Adaptation
concentration, temperature and pressure gradients, and fluxes at the boundaries of the study domain. While precipitation amounts are input from the downscaled climate scenarios, irrigation amount and scheduling is specified by the SWAP model user. Irrigation is triggered by soil in the root zone drying beyond a critically low value. The Penman-Monteith equation is used to compute evapotranspiration, or the sum of evaporation and transpiration. For crops with closed canopies, or densely planted crops, soil evaporation decreases but transpiration increases as crop development progresses. Density of foliage, characterized by the LAI, decreases the amount of rainfall and radiation that directly reaches the soil surface. Water not retained within the unsaturated zone, or taken up by the crop, flows to adjacent drains or groundwater or drains freely according to the boundary conditions specified by the model user. Runoff is generated when surface infiltration or storage capacity is exceeded. Crop water uptake is directly related to soil wetness, potential evapotranspiration and root length. Soil temperature exercises some influence on the bioavailability of nutrients, but less on water dynamics, especially when the crop in place has a well-developed root system. Essentially isothermal at depths below 100cm, the soil temperature regime depends on surface heating, soil thermal properties and wetness. Estimates of soil properties used in the study are obtained from the literature (Williams, 1979; Campbell, 1985; FAO, 2002). The SWAP model is linked to the WOFOST model through water and nutrient uptake by crop roots and LAI. The main processes simulated by WOFOST are the partitioning of assimilates from photosynthetic activity into root, stem, leaves and storage organs. In photosynthesis, CO2 from the air is transformed into glucose. The energy for this transformation originates from sunlight, or, more precisely, from the photosynthetically active radiation (PAR). Part of the glucose produced is used to provide energy for respiration and crop maintenance, depending on the amount of dry matter in the various living plant organs, the relative maintenance rate per organ and the temperature. The remaining assimilates are partitioned among roots, leaves, stems and storage organs in fractions depending on the phenological development stage of the crop. Time-dependent partitioning coefficients change with crop development stage (Van Diepen et al, 1989). For a grain crop such as millet, the dry weight of storage organs per hectare at the end of the crop cycle, equivalent to crop yield, is an important model output. The net increase in leaf structural dry matter and the specific leaf area determine leaf area development and, hence, the dynamics of light interception, except for the initial stage when the rate of leaf appearance and final leaf size are constrained by temperature, rather than by the supply of assimilates. Leaf senescence occurs because of water stress and shading, and also because of lifespan exceedance. The death rate of stems and roots is related to the development stage and crop genotype. Crop parameters in this study were taken from the literature (de Willingen and Noordwijk, 1987; Van Diepen et al, 1989). Some of these were changed in one of the adaptation strategies to see the impact on yields.
3232 J&J Climate Change
1/11/07
9:56 am
Page 137
Making Economic Sense of Adaptation in Upland Cereal Production Systems 137
Economic feasibility analysis CEREBAL, a simple spreadsheet model, is used to update and compute running totals of cereal stocks in The Gambia. Population size and per capita consumption constitute key variables on the demand side of the model. Cereal production in any year is obtained by summing up production from rain-fed/ irrigated rice with that from upland cereals, the latter derived from SWAPWOFOST crop yield and cultivated area. Variations in cultivated area are handled through a land-use submodel that incorporates competing land uses within the socioeconomic context of The Gambia. At the end of every year in the time window studied, CEREBAL compares demand for cereals with production and computes commercial grain imports and food aid requirement in line with national food security policy. The economic feasibility of different adaptation options is evaluated from net adaptation benefits and imposed climate change damages associated with a particular adaptation strategy. Climate change damages are measured as losses in the economic value of cereal production, as simulated with the SWAPWOFOST and CEREBAL models, for climate change scenarios of the near and distant future relative to a baseline reference case for current climate and current crop yields. The future climate change simulations include business-asusual management cases for low-intensity rain-fed cereal production and cases that incorporate changes in management to adapt to changes in climate. Net adaptation benefits are measured as the reduction in climate change damages, comparing the business-as-usual case with an adaptative management strategy, less the cost to implement the adaptation strategy. The residual damage after taking into account the net benefits of adaptation are referred to as the imposed climate change damages.
Results and Discussion Crop yields Statistics of simulated yields are presented in Table 7.1 for business-as-usual and adaptive management strategies combined with observed and future climates corresponding to the historical reference period 1961–1990 and projections of ECHAM4 and HadCM3 for the periods 2010–2039 and 2070–2099. The simulated yields reveal a systematic element of dependence on climate scenario. This is hardly surprising, especially in the case of distant future simulations with HadCM3, which we recall prescribes a 400mm decrease in rainfall relative to the period 1961–1990. The effects of climate change on yields under business-as-usual management vary depending on the time period and the climate model. For the near future period, ECHAM4 and HadCM3 alike project an increase in average yields of 13 per cent and 2 per cent respectively. Increasing yields could be explained by carbon dioxide fertilization and a shift in the climate toward optimum temperature for C4 crops (Wand et al, 1999). The advantages of the higher average yield estimated for the HadCM3 climate projection are coun-
3232 J&J Climate Change
1/11/07
9:56 am
Page 138
138 Climate Change and Adaptation
teracted by an increase in the variability of yields due to an 8 per cent decrease in rainfall, which, as we will see later, results in economic losses despite the marginally higher average yield. For the more distant future of 2070–2099, results for the ECHAM4 climate projection show a nearly 40 per cent increase in cereal yield for business-as-usual management. But for the severe rainfall decrease projected by the HadCM3 model, estimated yields collapse by almost 80 per cent. Table 7.1 Average millet yields (kg/ha) and variability (CV) for current and future climates with business-as-usual and adaptive management strategies Adaptation strategy
ECHAM4 Yield
HadCM3 CV
Yield
CV
1115
30
1141 1294 1517 1563
33 22 25 11
354 583 610 1811
167 135 125 13
Reference period: 1961–1990 Business-as-usual
923
23
Near future: 2010–2039 Business-as-usual High-yielding cultivar N-fertilizer (100kg/ha) Irrigation (150mm)
1046 1186 1450 1496
24 22 20 13
Distant future: 2070–2099 Business-as-usual High-yielding cultivar 1 N-fertilizer (200kg/ha) Irrigation (500mm)
1274 500 1733 1110
29 30 20 32
Table 7.1 also presents results for yields under three different adaptation strategies: substitution of an improved millet cultivar, fertilization with nitrogen and irrigation. As previously mentioned, there is nothing revolutionary about such an approach. The technology to implement some strategies already exists but has not been fully harnessed. In the future, one may also expect improved cultivars from crop breeding programmes. Mimicking the outcome of selective breeding and genetic engineering programmes, we make changes to the following attributes of P. typhoides: (1) increased drought tolerance, (2) increased yield and (3) shorter growing cycle. Nitrogen fertilization is introduced at rates of 100kg and 200kg per hectare per year and irrigation is introduced at rates of 150mm and 500mm per year for the time periods 2010–2039 and 2070–2099 respectively. For the period 2010–2039, average yields increase with adaptation relative to business-as-usual yields by 13 to 43 per cent and 13 to 37 per cent for the ECHAM4 and HadCM3 projected climates respectively. The substitution of a higher yielding cultivar produces the smallest gain in average yield of the three adaptation strategies, 13 per cent for both climate model simulations. In comparison, nitrogen fertilization and irrigation would increase yields by 33 to
3232 J&J Climate Change
1/11/07
9:56 am
Page 139
Making Economic Sense of Adaptation in Upland Cereal Production Systems 139
43 per cent. All the adaptation strategies would also reduce the variability of yields, ranging from 8 to 67 per cent. The greatest reduction in variability comes from irrigation. For smallholder farming households, stability in yields is one important aspect of poverty and survival. The consequences of a poor harvest year are not so devastating if the following year gives a normal harvest. Periods of two or more successive years of poor harvest, however, are rather difficult to overcome without external assistance. For the climate projection of the ECHAM4 model for the distant future (2070–2099), the adaptation strategies analysed result in average yield changes of between –12 and +36 per cent. Adoption of the improved cultivar and fertilization with nitrogen would amplify yield gains projected for the ECHAM4 climate with no adaptation. In these cases, adaptation and climate change would bring net benefits for cereal yields. In contrast, irrigation at the rate of 500mm would reduce yields by 12 per cent relative to business-as-usual management, although yields would still be higher with climate change than for the reference climate of 1961–1990. Waterlogged conditions caused by overirrigation explain the drop in yield relative to business-as-usual management under the ECHAM4 climate. Simulations based on the HadCM3 climate projection for the distant future show an increase of 64 to 411 per cent relative to the business-as-usual case, depending on the adaptation strategy simulated. Yet, except for the irrigation case, the gains are not enough to overcome the severely negative effects of climate change for the highly water stressed climate of the HadCM3 model and yields are less than for the reference case with no climate change. Implementation of irrigation at a rate of 500mm is projected to increase average yields by a factor of more than 5 and decrease variability by an order of magnitude. This more than compensates for the yield effects of the HadCM3 drop in mean annual rainfall relative to the reference period. But these benefits need to be weighed against the substantial costs of irrigation. In general, as crop yields increase, interannual variability decreases under all adaptation options. Using the criteria of reduced variability and increased average yield, irrigation, except when it is overdone, outranks other adaptation options. Crop fertilization ranks second best by these criteria. In the remainder of this paper, a pairwise comparison of the economic performance is made between irrigation and crop fertilization.
Economic performance of selected adaptation practices Economic analysis presented in this paper uses costs and benefits of adaptation strategies within a national cereal self-sufficiency/import substitution framework. The economic analysis focuses on the HadCM3 climate scenario as it is this scenario that poses the greatest threat to cereal production in The Gambia. However, it should be kept in mind that a wetter climate, as projected by the ECHAM4 model, would generate benefits which we have not estimated. Key variables in the analysis include the cost of inputs, market price of cereals, consumer preferences and food security policy. Constant cereal prices of US$150/ton are assumed. Other general assumptions include (1) cultivated
3232 J&J Climate Change
1/11/07
9:56 am
Page 140
140 Climate Change and Adaptation
area of millet increases proportionally with population growth, subject to a 0.15ha/capita limit in order to avoid conflict with other competing land uses; (2) food imports and aid are triggered by buffer stocks falling below a critical stock-to-utilization ratio (STU); and (3) food import and aid costs are expressed in constant dollar values. We further assume that economic agents with a central role in climate change adaptation, that is traders and dabadas, respond to market signals in a rational way that is influenced by government regulation, taxation, incentives and other economic instruments. Crop fertilization For analysis of crop fertilization, it is assumed that this strategy is applied to the entire area under millet cultivation. Two different assumptions are used for the stock utilization ratio that triggers food imports and aid: 10 and 20 per cent. Per capita cereal consumption is set at 250kg per person per year, closely matching current levels (Department of Planning, 2005). Results of the economic analysis of fertilization are shown in Table 7.2. Visual inspection of net adaptation benefits indicates the economic potential of fertilization as an adaptation option in the near future. Positive net benefits into the distant future also reinforce the evidence of economic viability under a changing climate. Note that while average crop yields are projected to increase in the near term for some of the scenarios (see Table 7.1), climate change damages are estimated to result due to increases in interannual variability that result in more frequent production shortfalls under a food security policy scenario. In essence, climate change damages are equivalent to commercial import of cereals and/or food aid required to maintain national food security. Average annual damages for the 2010–2039 period are roughly US$150 million for both stock utilization ratios and in excess of US$1 billion for the 2070–2099 period. Fertilization generates average annual benefits of US$29 million to 38 million in the near term period at an annual cost of US$6.3 million, for a net adaptation benefit of US$22 million to US$32 million. With adaptation, residual damages, hereafter referred to as imposed climate change damages, are the difference between climate change damages without adaptation and net adaptation benefits. Positive imposed climate change damages, such as the ones that appear in Table 7.2, indicate that fertilization alone is not sufficient to make up for cereal production shortfalls, under the combined effect of climate and demographic changes. The imposed climate change damages in the near term are US$123 million to US$130 million per year. For the more distant 2070–2099 period, results are more sensitive to assumptions about the stock utilization ratio. Estimated net adaptation benefits range from US$17 million to US$95 million and imposed climate change damages range from US$955 million to US$1.0 billion. Sensitivity of results in Table 7.2 to the percentage area treated with fertilizer is analysed by changing this fraction to several values between 5 and 100 per cent, without qualitative changes to the results.
3232 J&J Climate Change
1/11/07
9:56 am
Page 141
Making Economic Sense of Adaptation in Upland Cereal Production Systems 141
Table 7.2 Climate change damages and costs and benefits of fertilization – Average annual values in millions of US dollars Period
Damages, costs and benefits (million US$) Economic indicator
10% stock utilization ratio
20% stock utilization ratio
2010–2039
Climate change damages Adaptation benefits Adaptation costs Net adaptation benefits Imposed climate change damages
155.1 37.9 6.3 31.6 123.5
151.9 28.6 6.3 22.3 129.5
2070–2099
Climate change damages Adaptation benefits Adaptation costs Net adaptation benefits Imposed climate change damages
1049.8 28.0 10.7 17.2 1032.6
1049.8 105.4 10.7 94.6 955.2
The stock utilization ratio is an important food security variable, especially when natural hazards or disruption of supplies are anticipated. Results of increasing the stock utilization ratio shown in Table 7.3 are, however, ambivalent. An increase in imposed climate change damages in response to an increase in the ratio in the near future could be seen as a misallocation of resources. With higher interannual variability, however, an increased stock utilization ratio seems to have a positive payoff by stimulating an increase in net adaptation benefits from 17 million to nearly 95 million US dollars. In the long run, especially when individuals’ economic situations improve, it is reasonable to assume change in peoples’ food preferences. For the distant future, therefore, we posit and analyse the impact of a shift in food choices, marked by a reduction in cereal consumption from 250kg to 175kg per person per year. Observe that reducing consumption to 175 kg does not imply food rationing but simply reflects a change in dietary habits, born out of improved economic status. This results in a reduction in imposed climate change damages of from 65 to almost 80 per cent in the near term and 30 to 40 per cent in the more distant future, depending on the stock utilization ratio assumed to trigger food imports. Irrigation Assumptions for the analysis of irrigation are as follows: 1 2 3 4
irrigated area of coarse cereals increases linearly with time, from its current value of 2 per cent to 20 per cent by the end of the 21st century; rice irrigation from surface water is accelerated after commissioning of Sambangalou dam; rice yields increase to 4 metric tons/ha under controlled irrigation; and irrigated millet and rice are harvested twice a year.
3232 J&J Climate Change
1/11/07
9:56 am
Page 142
142 Climate Change and Adaptation
Table 7.3 shows results for economic performance indicators under irrigated conditions. A major point of observation in this table concerning the near future (2010–2039) is the negative net adaptation benefits, which range from 81 million to 88 million US dollars per year. While irrigation would increase yields rather substantially, the even more substantial costs of irrigation outweigh the value of the increased production. This clearly indicates that resources could be more efficiently allocated to procurement of food supplies on the world grain markets than to investment in irrigation of cereals, at least for the 2010–2039 time period. This situation may change in the more distant future when water becomes the major limiting factor for crop production, depending on the stock utilization ratio that triggers food imports. As shown in Table 7.3, the economic efficiency of irrigation is related to the policy stock utilization ratio variable, which governs cereal imports. Indeed, increasing the trigger level reverses the outcome of the net benefit calculus from a net benefit from irrigation of 52 million US dollars per year to an annual loss of 44 million. Sensitivity analysis shows that, for the latter case, net adaptation benefits from irrigation only become positive when water costs drop below US$0.09 per cubic metre. Considering, however, that this is 25 per cent less than the unit cost of water in the lowest tariff block, it is extremely unlikely that smallholder irrigation schemes can achieve economies of scale sufficient to bring down water costs to a profitable level. It suffices to point out that operation and maintenance account for 80–90 per cent of pump irrigation costs and how this fraction evolves would depend on future world energy markets, technological innovation and the state of The Gambian economy. An oblique approach to the problem of cost reduction is how to increase the market value of crops harvested. The conundrum essentially reduces the
Table 7.3 Climate change damages and costs and benefits of irrigation – Average annual values in millions of US dollars Period
Damages, costs and benefits (million US$) Economic indicator
10% stock utilization ratio
20% stock utilization ratio
2010–2039
Climate change damages Adaptation benefits Adaptation costs Net adaptation benefits Imposed climate change damages
155.1 43.3 124.5 –81.2 236.3
151.9 36.4 124.5 –88.0 239.9
2070–2099
Climate change damages Adaptation benefits Adaptation costs Net adaptation benefits Imposed climate change damages
1049.8 303.2 252.2 52.0 997.8
1049.8 207.5 251.2 –43.7 1064.8
3232 J&J Climate Change
1/11/07
9:56 am
Page 143
Making Economic Sense of Adaptation in Upland Cereal Production Systems 143
number of choices of crops. High value, non-food crops (for example, flowers) and non-staple crops (for example, vegetables) currently fetch higher prices than cereals on the market, but on the scale of production envisaged, this may no longer be the case. The problem, however, does not have to be articulated in such dichotomous terms. What one probably needs is to find an optimal mix of crops fetching the highest economic returns, subject to land, water, labour and other constraints. Analyses were also performed for a decrease in cereal intake from 250kg per person to 175kg. This reduction in per capita cereal consumption would reduce the imposed climate change damages by 85 per cent in the 2010–2039 time period and 30 to 35 per cent in the more distant time period. One way of interpreting such a reduction is foreign exchange savings if the bill for foreign goods and services relating to the production and/or importation of alternative foods is incorporated into the calculus.
Conclusions The estimated impacts of climate change on cereal crop yields range from increases to decreases, depending strongly on projected changes in rainfall. For a climate that is warmer but not drier than the present climate in The Gambia, simulated cereal yields would increase and generate economic benefits. But for warmer and drier scenarios, yields decrease and become more variable, resulting in economic losses, or climate change damages. For the severely dry climate derived from the HadCM3 model projection for the end of the century, cereal yields would collapse by an estimated 80 per cent. Crop selection and fertilization, used as insurance against climate variability by smallholder farming households, are shown to be effective in offsetting adverse climate change impacts in the near future, or amplifying beneficial impacts. The yet untested practice of irrigation gives the highest increase in average productivity and the greatest reduction in variability under the different climate projections. But the economic efficiency of adaptation options is strongly influenced by unit costs of implementation. High units costs of irrigation development result in negative net benefits from irrigation in the near term, making irrigation an unattractive option, at least in the near term. In contrast, adoption of fertilization practice would generate net benefits of roughly 20 million to 30 million US dollars per year. For the very dry climate projected by the HadCM3 for the end of the century, the economic performance of irrigation changes. Depending on government policies that determine food imports and aid, irrigation could generate substantial net benefits. Indeed, irrigation may become an imperative in the distant future if precipitation declines as sharply as indicated by the HadCM3 projections, or if world cereal markets become seriously affected by conditions in countries with historically surplus production. Whether expanded irrigation would be feasible in a much drier climate, however, has not been investigated. Essentially, there is no single best adaptation strategy, and instead of
3232 J&J Climate Change
1/11/07
9:56 am
Page 144
144 Climate Change and Adaptation
import substitution, one should be looking at complementing business-asusual (in other words food imports) with fertilization and irrigation of locally grown cereals. In the short run, expanding crop fertilization, in particular, has significant advantages. It requires no technological sophistication and promises high returns. These results call for an immediate response from government, the service sector and the farming community. Notwithstanding this, some challenges still remain. Food policy development and analysis require more sophisticated projections of commodity prices, costing of research and development in crop science, and technology trends. A commitment to food security, rooted in developing The Gambia’s agricultural potential, is indispensable to reducing the country’s sensitivity and vulnerability to climate risks. In this regard, it may be quite important to discuss the role of key stakeholders, as well as to examine the conditions under which adaptation options are most likely to be taken up by dabadas. Without any doubt, the government of The Gambia, responsible for social, economic and related policies, should take the first step to ensure that valid research findings get translated into tangible benefits. The government’s role is to pick up research results, demonstrate their validity and create incentives for integration of the options into current agricultural practice. Considering, however, the low level of returns on investments in cereal production compared to other crops, it is fair to say that farmers, when given the choice, will opt for irrigation of non-cereal crops in the dry season. There is already ample evidence of this in community and women’s gardens across the country. Even in some lowland environments, non-cereal, high-value horticultural crops, such as pepper (Capsicum annum), okra (Hibiscus esculentus) and tomatoes (Lycopersicon esculentum), are grown instead of rice under irrigation in the dry season. Two reasons may explain this practice. First, part of the incomes generated by households from horticultural production is used to purchase imported rice when their cereal stock gets depleted. Second, and perhaps overlooked, planting decisions also make sense in light of the agronomic practice of crop rotation. A future government outreach/extension programme therefore stands the best chance of success when dabadas pursue multiple objectives, including food security. In this scenario, discussed above as an optimization problem, a fraction of the area under full water control could be allocated to non-cereals, and part of the extra revenue generated used to expand cereal production in subsequent years. Dabadas have a good knowledge of these issues, experience of climate extremes and a strong stake in harnessing adaptation options to ensure their food security. From a broader perspective, The Gambia government’s national disaster reduction strategic framework, in the making, provides an attractive opportunity to build partnerships for food security. Community empowerment, efficient marketing structures, competitive prices and price stability are some of the factors that hold the key to the transformation of agricultural production in The Gambia.
3232 J&J Climate Change
1/11/07
9:56 am
Page 145
Making Economic Sense of Adaptation in Upland Cereal Production Systems 145
References Campbell, G. S. (1985) ‘A simple method for determining unsaturated conductivity from moisture retention data’, Soil Science, vol 117, pp311–314 Carney, E. (1986) ‘The social history of Gambian rice production: An analysis of food security strategies’, PhD thesis, Michigan State University, Ann Arbor, MI Cole, A., K. Sanyang, A. J. Marong and F. Jadama (2005) ‘Vulnerability and adaptation assessment of the agricultural sector of the Gambia to climate change’, consultancy report prepared for NAPA Project LDL 2328 2724 4699, Banjul de Willingen, P. and M. van Noordwijk (1987) ‘Roots, plant production, and nutrient efficiency’, PhD thesis, Wagenigen Agricultural University, Wagenigen, The Netherlands Department of Planning (2001) ‘Statistical yearbook on Gambian agriculture for the year 2000’, annual report, Department of State for Agriculture, Banjul Department of Planning (2003) ‘Statistical yearbook on Gambian agriculture for the year 2002’, annual report, Department of State for Agriculture, Banjul Department of Planning (2005) ‘Statistical yearbook on Gambian agriculture for the year 2004’, annual report, Department of State for Agriculture, Banjul Droogers, P. (2000) ‘Estimating actual evapotranspiration using a detailed agro-hydrological model’, Journal of Hydrology, vol 29, pp50–58 FAO (2002) ‘Digital soil map of the world and derived soil properties’, on CD-ROM, www.fao.org/ag/agl/agll/dsmw.htm Feddes, R. A., P. J. Kowalik and H. Zarandy (1978) Simulation of Field Water Use and Crop Yield, Simulation Monographs, Center for Agricultural Publishing and Documentation (PUDOC), Wageningen, The Netherlands Gomez, B. E., M. Njie, B. P. Jallow, M. E. Hellmuth, J. M. Callaway and P. Droogers (2005) ‘Adaptation to climate change for agriculture in The Gambia: An explorative study on adaptation strategies for millet’, AIACC Working Paper No 37, International START Secretariat, Washington, DC GOTG (Government of The Gambia) (2003) ‘First national communication of the Republic of The Gambia to the United Nations Framework Convention on Climate Change’, Government of The Gambia, Department of State for Fisheries, Natural Resources and the Environment, Banjul Jallow, S. S. (1995) ‘Identification of the response to drought by local communities in Fulladu West District of The Gambia’, Singapore Journal of Tropical Geography, vol 6, pp22–41 Kroes, J. G., J. C. van Dam, J. Huygen and R. W. Vervoort (1999) ‘User’s guide of SWAP version 2.0. Simulation of water flow, solute transport, and plant growth in the Soil-Water-Atmosphere-Plant environment’, Technical Document 48, Alterra Green World Research, Wageningen, Report 81, Department of Water Resources, Wageningen University and Research, Wageningen, The Netherlands Nakicenovic, N. and R. Swart (eds) (2001) Emissions Scenarios. A Special Report of Working Group III of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York National Environment Agency (1997) ‘State of the environment report – The Gambia’, National Environment Agency, Banjul Njie, M. (2002a) ‘National water security in the first half of the 21st century’, report prepared under GAM/93/003 as a contribution towards The Gambia’s National Water Resources Management Strategy, UNDP/DWR, Banjul Njie, M. (2002b) ‘Second assessment report of climate change induced vulnerability of Gambian water resources sector, and adaptation strategies’ report prepared for The Gambia’s National Climate Committee, Banjul Racsko, P., L. Szeidl and M. A. Semenov (1991) ‘A serial approach to local stochastic
3232 J&J Climate Change
1/11/07
9:56 am
Page 146
146 Climate Change and Adaptation weather models’, Ecological Modelling, vol 57, pp27–41 Richardson, C. W. (1981) ‘Stochastic simulation of daily precipitation, temperature and solar radiation.’ Water Resources Research, vol 17, pp182–190 Spitters, C. J. T., H. Van Keulen and D. W. G. van Kraalingen (1989) ‘A simple and universal crop growth simulator: SUCROS87’, in R. Rabbinge, S. A. Ward and H. H. van Laar (eds) Simulation and Systems Management in Crop Protection, Simulation Monographs, Center for Agricultural Publishing and Documentation (PUDOC), Wageningen, The Netherlands Supit, I., A. A. Hooijer and C. A. Van Diepen (eds) (1994) System Description of the WOFOST 6.0 Crop Growth Simulation Model Implemented in CGMS. Vol 1: Theory and Algorithms, European Commission, Luxembourg Van Dam, J. C., J. Huygen, J. G. Wesseling, R .A. Feddes, P. Kabat, P. E. V. Van Walsum, P. Groenendijk and C. A. van Diepen (1997) ‘Theory of SWAP version 2.0’, Technical Document 45, Wageningen Agricultural University and DLO Winand Staring Centre, Wageningen, The Netherlands Van Diepen, C. A., J. Wolf, H. Van Keulen and C. Rappoldt (1989) ‘WOFOST: A simulation model of crop production’, Soil Use and Management., vol 5, pp16–25 Verkerk, M. O. and C. P. M. van Rens (2005) ‘Saline intrusion in Gambia river after dam construction: Solutions to control saline intrusion while accounting for irrigation development and climate change’, research report, University of Twente, The Netherlands Wand, S. J. E., G. F. Midgley, M. H. Jones and P. S. Curtis (1999) ‘Responses of wild C4 and C3 grass (Poaceae) species to elevated atmospheric CO2 concentration: A meta-analytic test of current theories and perceptions’, Global Change Biology, vol 5, pp723-741 Williams, J. B. (1979) ‘Soil water investigation in The Gambia’ Technical Bulletin 3, Land Resources Development Centre, Ministry of Overseas Development, Surbiton, Surrey, England
3232 J&J Climate Change
1/11/07
9:56 am
Page 147
8
Past, Present and Future Adaptation by Rural Households of Northern Nigeria Daniel D. Dabi, Anthony O. Nyong, Adebowale A. Adepetu and Vincent I. Ihemegbulem
Introduction There is sufficient information now to indicate that the climate is changing with far reaching effects for the well-being and livelihoods of rural people in the northern, semi-arid areas of Nigeria. This part of the country has characteristics similar to the rest of the arid and semiarid regions of West Africa commonly known as the Sahel. The Sahel is characterized by scant rainfall, with an annual average of between 150 and 600mm. Annual rainfall levels have been decreasing in the region over the course of the last century, with an increase in inter-annual and spatial variability (Glantz, 1987; Tarhule and Woo, 1998; Ozer, 2003). This region has experienced fluctuations in rainfall on all time-scales – from decadal (ten-day cycle) to monthly, seasonal, annual and longer term. The longer-term fluctuations are caused by the oscillations of the climatic borders of the Sahara Desert. The arid and semiarid part of Nigeria experiences similar characteristics with droughts occurring due to rainfall variability (Tarhule and Woo, 1998). Drought is the main climatic hazard affecting the socioeconomic activities and livelihoods of rural households of our study region (Mortimore, 1989). For the purpose of this chapter, we have adopted the National Drought Mitigation Center’s conception of drought as a normal, recurrent feature of climate that originates from a deficiency of precipitation over an extended period of time relative to long-term average or normal conditions and results in a water shortage for some activity, group or environmental sector (National Drought Mitigation Center, 2005). Drought is also related to the timing of rainfall (principal season of occurrence, delays in the start of the rainy season, occurrence of rains in relation to principal crop growth stages) and the effectiveness of rain (rainfall intensity, number of rainfall events) and can be aggravated by other climatic factors such as high temperature, high wind and low relative humidity. In Nigeria, food shortages are attributable to poor socioeconomic conditions and government policies, increases in population and declining crop
3232 J&J Climate Change
1/11/07
9:56 am
Page 148
148 Climate Change and Adaptation
production following the oil boom of the 1970s, when the agricultural sector was abandoned for oil revenue. Problems of food shortages and insecurity are aggravated by drought, as occurred in 1973 to 1974. Since then, the agricultural sector has not lived up to expectations (Dabi, 2006). In its most recent assessment of regional climate change projections, the Intergovernmental Panel on Climate Change (IPCC) concludes that warming in Africa is very likely to be greater than the global average and that warming will be greater in the drier subtropical regions, which include northern Nigeria, than in the moister tropics (Christensen et al, 2007). Changes in rainfall are uncertain for the Sahel, with the 25–75 percentiles of model projections ranging from decreases to increases in the region. But despite the uncertainty in rainfall changes, governments and policy makers should plan for the possibility of more frequent and more severe dry spells and droughts and should develop strategies for coping and adapting accordingly. Preparing for a potentially more droughty future is prudent because drought will continue to be a major threat to food security, livelihoods and rural development in the Sahel, even if the climate becomes moister than at present, and because the current capacity to cope with drought is poorly developed. These efforts should emphasize enhancing the adaptive capacity of poor rural households, who are highly vulnerable to drought and other climate pressures (Nyong et al, 2003). Several definitions of climate change vulnerability, sensitivity and adaptation are found in the literature, for example, Smithers and Smit (1997), Pielke (1998), Smit et al (1999, 2000 and 2001) and Adger (2001), as summarized in Huq et al (2003). For the purposes of this chapter, we have adopted the definitions of Working Group II of IPCC (McCarthy et al, 2001): Vulnerability is ‘the degree to which a system is susceptible to, or unable to cope with, adverse effects of climate change, including climate variability and extremes. It is a function of the character, magnitude and rate of climate change and variation to which a system is exposed, its sensitivity, and its adaptive capacity’. Sensitivity is defined as ‘the degree to which a system is affected, either adversely or beneficially, by climate-related stimuli’, while the adaptive capacity is ‘the ability of a system to adjust to climate change, including climate variability and extremes, to moderate potential damages, to take advantage of opportunities, or to cope with the consequences’ (McCarthy et al, 2001). Increasing the adaptive capacity of a system represents a way of coping with changes and uncertainties in climate, reducing vulnerabilities and promoting sustainable development (Huq et al 2003).
Objectives and Methods Field observations have indicated that many poor rural households are already adapting to drought and water scarcity. Because there is general agreement that climate change adaptation can usefully begin with strategies to reduce present vulnerabilities (Downing et al, 1997; Adger, 2001), we seek to learn from the experiences of these households. The aims of this chapter are to examine how rural households of northern Nigeria have adapted to drought and water
3232 J&J Climate Change
1/11/07
9:56 am
Page 149
Past, Present and Future Adaptation by Rural Households 149
scarcity and to consider what lessons past and present strategies may hold for future adaptation. The specific questions we investigate are as follows: 1 2 3 4 5
Which members of the community are most vulnerable and what aspects of their livelihood are threatened? How have they coped with the problem of drought and water scarcity? How might they adapt in the future? What factors influence their capacity to adapt? What opportunities are available to them to enhance their adaptive capacity?
Our analysis is largely based on primary data collected directly in the field from households across the Sahelian belt of northern Nigeria. The field activities included a rapid rural appraisal, questionnaire surveys and focus group discussions. The rapid rural appraisal was carried out during the reconnaissance survey in October 2002. Observations were made, photographs taken, communities selected for data collection and poor rural households identified based on interactions with key informants. Twenty-seven communities from nine states were selected for administration of household surveys (see Figure 8.1). The selected communities include three settlement types: hamlets, villages and small towns, with populations of roughly 1000, 5000 and 10,000 respectively. The communities were selected in a ratio of 1:2:3 for small towns, villages and hamlets in that order to represent the first stratum in the sampling process. Thirty households were randomly selected from each of the communities for participation in the survey, giving a total of 810 households surveyed. A household is defined for our study as a group of people living in the same compound, consisting of an enclosed set of buildings or huts, eating from the same pot and recognizing one person as head, usually a husband and father or guardian. The household is chosen as our unit of analysis because livelihood decisions are usually made at the household level and vulnerability also resides at that level. The questionnaire is made up of a combination of open- and closed-ended questions related to aspects of drought, activities undertaken by households, problems they encounter, strategies they have adopted to cope with droughts and their opinions regarding future adaptation. Information generated by the questionnaires provided the basis for identifying vulnerable households, establishing how they have been coping (adaptation measures) and revealing factors that influence their capacity to adapt to droughts. The household survey was followed up with a series of focus group discussions. The focus group discussions used a bottom–up approach to adaptation strategizing and planning in which 10 key informants were invited from each of the 27 communities for in-depth discussions of present coping strategies identified by the survey, difficulties faced and possible strategies that might be useful for future adaptation. The informants include community and religious leaders, farmers’ representatives, and leaders and representatives of associations, women’s groups and local government. The bottom–up approach emphasizes initiatives developed locally at a micro or local community level that are based on indigenous knowledge and local perceptions, opinions and preferences. Aspects of indigenous knowledge
3232 J&J Climate Change
1/11/07
9:56 am
Page 150
150 Climate Change and Adaptation
1. 5. 9. 13. 17. 21. 25.
Kalalawa Shanga Auki Jinjimawa Kwarnar Gaki Zandam Damasak
2. 6. 10. 14. 18. 22. 26.
Marnona Dabai Dan Matamachi Zangon Buhari Sara Takwikwi Chingowa
3. 7. 11. 15. 19. 23. 27.
Kajiji Daki Takwas Tabanni Guruma Kubani Badrama Maimallamari
4. 8. 12. 16. 20. 24.
Andarai Maguru Kofin Soli Chanchanda Madara Buni Yadi
Figure 8.1 Communities surveyed in northern Nigeria
systems that are important for planning climate change adaptation include what the households did in the past to survive earlier droughts, what they are doing now to cope with current droughts, and the factors that influence their choices and activities. A bottom–up approach helps to ensure that adopted strategies are responsive to, and consistent with, local priorities, preferences
3232 J&J Climate Change
1/11/07
9:56 am
Page 151
Past, Present and Future Adaptation by Rural Households 151
and capacities and that the strategies are likely to benefit local people. This contrasts with a top–down approach in which measures are externally planned and imposed on a locality without the local people’s input or participation. Such projects run a high risk of failure. Data collected from the household survey and focus group discussions were analysed using qualitative methods and descriptive statistics in the form of frequencies and percentages presented in tables and graphs. We assessed the vulnerability and adaptive capacity of rural households using the livelihood systems approach (Carney et al, 1999; Davies, 1996). The livelihood systems approach examines five different categories of household resources or capital used to support a household’s livelihood strategy and provide it with resilience and capacity for coping with, and adapting to, shocks. These are natural, human, financial, physical and social capital. Davies (1996) distinguishes differential vulnerability and livelihood system vulnerability and argues that vulnerability research should focus more on livelihood system vulnerability because people in different livelihood systems are vulnerable in different situations and seasons. Different livelihood systems experience different trigger events that can cause food and livelihood stress. Following her approach, we identify three major livelihood strategies in the study region – crop farming, herding livestock and fishing – and distinguish the different ways in which these livelihood groups are vulnerable to drought.
Results Vulnerability of rural households Vulnerability is a relative term differentiating between socioeconomic groups or regions, rather than an absolute measurement of deprivation. The analyst or decision maker must assign the thresholds of vulnerability that warrant specific responses. For our study areas, we developed a methodology for classifying households based on their levels of current vulnerability using a vulnerability index constructed from 13 indicators that measure different aspects of vulnerability (Table 8.1). The indicators and their weights were determined based on factors identified by households and stakeholders as important determinants of vulnerability, as well as information from the published literature. Indicator values were normalized such that the highest value for an indicator was set at 100 per cent. Adding up the household scores on the 13 indices resulted in an overall vulnerability score for each household in the sample. Households were grouped into three levels of vulnerability: very vulnerable, vulnerable and less vulnerable. In the questionnaire, we had asked the respondents to place themselves in one of these classes of vulnerability. This allowed us to factor in their perceptions and self-reported assessments of vulnerability. We took all those who put themselves in each group, found the average scores for each group and used these as the midpoints of the various vulnerability classes and then built class intervals about them.
3232 J&J Climate Change
1/11/07
9:56 am
Page 152
152 Climate Change and Adaptation
Table 8.1 Indices and weights for vulnerability assessment in northern Nigeria Index
Index Measured/Calculated as Value
Range
Average
Acreage under cultivation
1
Hectares/consumer units
0.1–2.8
0.6
Dependency ratio
1
Labour units/consumer units (inverted)
0.3–0.8
0.5
Livestock ownership
1
Tropical livestock units/consumer units
0.0–8.2
3.7
Gender of household head
1
Value given to sex of household head
1.0–2.0
1.8
Livelihood diversification
1
Weighted number of non-agricultural income generating activities/consumer units
0.0–2.4
0.7
Annual cash income
1
In 1000 naira/consumer units
2.5–9.7
4.2
Drought preparedness
1
Value given to use of drought resistant crops and livestock, and receiving drought-related information and advice
0.0–2.0
1.1
Educational background of the household head
0.5
Value given to highest school level attained by the head of the household
0.0–4.0
1.8
Land tenure situation
0.5
Value given to land tenure situation
1.0–3.0
2.5
Type of house
0.5
Value given to type of house lived in
1.0–3.0
1.8
Self-sufficiency 0.5 in food production
Number of years surplus foodstuffs were sold minus number of years foodstuffs were bought in the past 10 years
0.0–20.0
11.2
Family and social networks
0.5
Value given to strength of family and social networks.
1.0–4.0
2.3
Quality of household
0.5
Number of able persons/number of disabled 1.5–12.0 and or sick persons in the household (inverted)
7.6
10
Sum of (index scores index value)
472.1
Overall vulnerability
235.1–833.9
All three livelihood groups of the Sahelian zone of northern Nigeria (crop farmers, herders and fishers) are vulnerable to drought and water scarcity. But within these groups, the poorer rural households are identified as the most vulnerable. The rural poor are highly vulnerable because their low asset base exposes them to high risk of impacts from climate stresses and limits their resilience and capacity to adapt. The landholdings and other natural resource assets of the poor are small in amount and degraded in quality, leaving a slim margin between meeting basic needs in good years and suffering severe
3232 J&J Climate Change
1/11/07
9:56 am
Page 153
Past, Present and Future Adaptation by Rural Households 153
depravation in poor years in this water scarce region. Their stock of human capital includes knowledge of traditional, indigenous practices as well as some new innovations that are used to cope with drought and scarcity. But educational attainment is low and constrains further innovation that requires new knowledge and skills. Food stores, financial savings, marketable assets and access to credit are too little to assure recovery from losses of crops and livestock that may be experienced due to drought or other shocks. Farm equipment and other physical assets of poor households are meagre and can act as a barrier to the adoption of some adaptive practices, while physical infrastructure in poor rural communities are also lacking and constrain options. Social networks exist and provide some security, but often these networks are weakest for the poorest households of a community Comparing across the three livelihood groups, we find that the proportion of households that is very vulnerable is greatest for crop farmers and lowest for fishing. We illustrate this for three hamlets, each dominated by a different livelihood strategy. Figure 8.2 shows the number of households classified as very vulnerable, vulnerable and less vulnerable for the hamlets of Zangon Buhari, Dabai and Takwikwi. The number of very vulnerable households is greatest in Zangon Buhari, a predominantly crop farming community located in Kano State in the north central region of Nigeria, followed by Dabai, a community of mostly livestock herders in Kebbi State in the northwest. The fewest number of very vulnerable households is found in Takwikwi, a fishing community located in Yobe State in the northeast. The lesser vulnerability of households in Takwiki may be attributed to the fact that the fishing community has access to the Nguru wetlands, a fairly reliable resource that supports fishing livelihoods. In Dabai, livestock farmers may have an advantage over crop farmers because of their mobility, moving cattle, sheep and goats to areas with greener pasture and greater availability of water, while the crop farmers depend on highly variable rainfall for their rainfed crops, which can easily be affected by droughts. The results for these hamlets are representative of the general pattern of vulnerability of the different livelihood groups.
Figure 8.2 Number of households classified as very vulnerable, vulnerable and less vulnerable in the hamlets of Zangon Buhari, Dabai and Takwikwi
3232 J&J Climate Change
1/11/07
9:56 am
Page 154
154 Climate Change and Adaptation
Strategies for coping with drought From the analysis of the survey data, a number of coping strategies have been identified, most of which are autonomous (implemented without external intervention) and responsive (implemented in reaction to climatic events and impacts) (Smit et al, 1999). We subdivide coping strategies into two categories, past and present. Past coping strategies are strategies adopted by the respondents during droughts that occurred in 1994 or earlier, while present coping strategies are more recent strategies adopted by the respondents after 1994. Table 8.2 identifies coping strategies reported by households and summarizes the number of respondents who adopted each strategy in the past and present. Past and present coping strategies are not mutually exclusive since a respondent might have adopted a strategy in the past and still use it in the present. Table 8.2 Respondents’ coping strategies Serial No
Strategy
Number of Respondents Past Present Total
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Drought resistant variety Crop diversification Livestock diversification Early maturing crop varieties High yield varieties Replanting Herd movement Herd supplementation Culling animals Labour migration Selling assets Herd sedentarization Farm relocation Herd/farm sizes Water exploitation methods Water use Water storage methods Food storage
206 201 164 183 128 198 89 54 57 105 142 70 230 132 120 204 161 361
168 235 150 325 312 128 104 127 56 95 150 75 96 82 93 115 166 203
374 436 314 508 440 326 193 181 113 200 292 145 326 214 213 319 327 564
Strategies that have been widely used in the past or present include food storage; cultivation of drought resistant, early maturing or high yield varieties of crops; diversification of crops or livestock; replanting when crops are lost; selling assets; relocating farms; and a variety of strategies for water use and storage. Food storage was the coping strategy adopted by the largest number of the respondents in the past. More recently, however, the number of people using food storage as a coping strategy has declined by 40 per cent. The most probable reason is that most households have had insufficient food even in good years to enable them to store food as a hedge against drought. Other strategies that have declined in use are cultivation of drought resistant varieties, replanting of
3232 J&J Climate Change
1/11/07
9:56 am
Page 155
Past, Present and Future Adaptation by Rural Households 155
crops, changes in farm locations, and changes in water exploitation methods and water use. The reasons for these decisions may be that benefits from the strategies are less than what was expected or that the costs of inputs needed to implement them, such as chemical fertilizers, water pumps and fuel to operate pumps, are not affordable to poor households. A number of coping strategies have increased in use, most notably the adoption of early maturing and high yielding crop varieties. Other strategies that have increased in use include diversification of crops and herd movement and supplementation. Respondents gave a series of reasons for the adoption of food storage and early maturing and high yield crop varieties. Figure 8.3 gives a summary of the reasons provided by the respondents for the adoption of food storage as a coping strategy. Food security was cited by the largest number of households as the reason for storing food (44 per cent), followed by lack of new information (20 per cent) and new method (19 per cent). The lack of new information may be attributed to the fact that most of the respondents had never been visited by government agricultural extension workers. The next most widely used past and present coping strategies are the cultivation of early maturing and high yield crop varieties. Figures 8.4 and 8.5 show the reasons for their adoption. The major reason for planting early maturing crops is to ensure early harvest (41 per cent), while the reason for planting high yield crops is, of course, to attain higher yields, as indicated by more than half (55 per cent) of the respondents who use this strategy. For both strategies, the second most important reason for the adoption of these strategies is lack of new information. The availability of early maturing and high yield crop varieties are recent developments, which explains why more respondents are adopting the strategies at present than in the past. This shift is also evidence that respondents are willing to change and adopt new coping strategies. Three groups of factors affect the adoption of the coping strategies discussed earlier: (1) the resource base for sustainable livelihoods, (2) government policy and (3) availability of information and warning signals. Households’ capabilities for adopting coping strategies are determined by their resource base, composed of their endowments of five types of livelihood capital: financial, human, natural, social and physical. We measure households’ endowments of these capitals using household income, level of education of household head and members, water availability for household use, community groups and social networks, and distance to road or market. Using the survey data and information from the focus group discussions, threshold values are established for these measures to classify the level of vulnerability of households along the dimensions of the five livelihood capitals. Figure 8.6 shows the number of households classified as very vulnerable, vulnerable and less vulnerable with respect to the five capital types. From Figure 8.6 it is evident that most households (80 per cent) are very vulnerable financially. Most households of the study sites have very low incomes, financial savings and saleable assets. They lack opportunities for offfarm labour income and the collateral necessary to access financial credit. This limits their ability to purchase farm inputs, equipment and other technology to improve agricultural production or adapt it to climate variability and extremes.
3232 J&J Climate Change
1/11/07
9:56 am
Page 156
156 Climate Change and Adaptation
Figure 8.3 Reasons for food storage
Figure 8.4 Reasons for planting early maturing crop varieties
Figure 8.5 Reasons for planting high yield crop varieties
Consequently, most farmers are restricted to their old traditions of low input farming practices, which bring about poor yields and high risk crop losses and failure. Unable to invest in improvements to their farms or earn off-farm labour income, poor rural households are trapped in poverty, placing them at risk of hunger and other adverse impacts of drought.
3232 J&J Climate Change
1/11/07
9:56 am
Page 157
Past, Present and Future Adaptation by Rural Households 157
Figure 8.6 Household vulnerability levels for the five livelihood capitals
A large proportion of households (68 per cent) are also very vulnerable with respect to physical capital. This is due to the absence of infrastructural facilities such as roads, which affects accessibility to sources of inputs and markets for products; water supply and sanitation, which affects human health and well-being; and energy in the form of electricity, which affects food storage and food processing. Many households (61 per cent) are highly vulnerable in terms of social capital, or family and community networks. Drought will often impact many or even all members of a social network simultaneously, weakening the ability of the network to assist and support its members during times of drought. This condition will affect self-help programmes such as food sharing, labour assistance and remittances, which serve as safety nets that can reduce vulnerability; religious beliefs and prayers, which give hope and relief to households during and after droughts but can also have negative effects (for example, the purdah system which restricts women may affect the search for work and food for the household); and the possibilities of conflict (for example, conflicts between arable farmers and herdsmen over grazing land and watering points, especially during and immediately after droughts). Many households (55 per cent) are also very vulnerable with respect to their endowment of human capital. This is attributed to the fact that most households have had no formal education, which may constrain access and use of information about climate risks, risk management strategies and new technologies, which affects in turn the adoption of coping strategies. The majority of households (55 per cent) are less vulnerable with respect to natural capital, for example, water availability. The most probable reason for this perception is because no major drought events, such as the 1973/74 and 1982/83 drought periods, have been recorded recently. There have been many policies on agriculture and rural development in Nigeria since the pre-colonial through the independence and post independ-
3232 J&J Climate Change
1/11/07
9:56 am
Page 158
158 Climate Change and Adaptation
ence periods up to the end of the last century. Unfortunately, these policies have not been favourable to the rural poor (Dabi, 2006). More recently, however, with the advent of the current democratic dispensation, government policies have been targeted at rural areas and for poverty alleviation, although most of the rural poor are yet to see the dividends of such policies. For example, agricultural inputs are not readily available to farmers. Whenever fertilizers are provided or subsidized to farmers, farmers either do not have access to the commodity or the quantity available does not reach them at the appropriate time. About 90 per cent of households did not receive fertilizers during the last planting season. Extension workers are based at local government authorities to assist farming households in rural areas, and more than 90 per cent of surveyed households indicate awareness of extension services. But many extension workers are not adequately trained to assist farmers to improve their farm practices. In most cases, there is little or no aid coming from government during and after droughts. The extension service and workers represent a potentially valuable resource that could help to create the right environment for appropriate adaptation to climate change by raising farmers’ knowledge and skills for improved management and use of new methods and technologies. Climatic information for rural areas of Nigeria is greatly limited by the sparse distribution of weather stations. The number of stations is greater now than 20 years ago, but the number is still insufficient to meet the needs for information about recent climate trends or to prepare forecasts at scales useful for adaptation. Furthermore, many of the stations are not adequately maintained due to lack of commitment from government. Farmers generally have poor access to what little climatic information is collected. Occasionally, some information may be disseminated via the electronic media, but few if any rural farmers possess television sets. The radio is the only dependable source of information for most of the rural poor, who own small transistor radios that use dry cell batteries. However, the language of transmission (English) is not usually understood by the majority of the rural poor. Issues translated into local languages are normally far from climatic; rather they are political. From the standpoint of the poor rural households, weather and climatic issues may not necessarily interest them due to ignorance and lack of awareness regarding the use of this kind of information.
Enhancing Future Opportunities Although the coping capacity of poor rural households is limited by a variety of factors as discussed in the preceding section, opportunities for future adaptation can be enhanced. Three strategies for enhancing opportunities are explored below: (1) development of past and present coping strategies; (2) introduction of alternative strategies; and (3) improvements in government policies and assistance. A wide range of strategies are in use for coping with drought and water scarcity that have helped to limit vulnerability, as demonstrated by our field-
3232 J&J Climate Change
1/11/07
9:56 am
Page 159
Past, Present and Future Adaptation by Rural Households 159
work. Still, these practices are not as widely or effectively used as they might be. Consequently, there exists a significant adaptation deficit (Burton, 2004) among rural households of the dry northern areas of Nigeria. Support and encouragement of selected current strategies would help to close this deficit by expanding their use to more households and improving the effectiveness with which they are applied. Efforts should focus in particular on four coping strategies that have been used by more than half of the surveyed households, an indication of their wide acceptance and effectiveness. These include crop diversification, cultivation of early maturing and high yield crop varieties and food storage. Other strategies that have been less widely used may nonetheless be effective and should also be encouraged as part of a diversified portfolio of coping mechanisms (see Table 8.2). Extension services could support coping strategies by increasing awareness of them and their utility for reducing climate related risks and helping farm households to develop the knowledge and skills needed to adopt and apply them effectively. Better distribution of early maturing and high yielding seed varieties and other farm inputs is needed, as is access to credit to enable farmers to purchase them. Assistance is also needed for marketing to support diversification of farm products. In addition to supporting the use and expansion of existing strategies, the introduction of alternative or new strategies should also be promoted. There exist alternate strategies that are not used presently by farmers in the study area but that have been used elsewhere (Dabi and Anderson, 1999) and which can be implemented within the household livelihood structures of crop farming, livestock herding and fishing. Coping strategies for crop farming can be expanded through introduction of new water sourcing and water use practices. Examples include sinking of boreholes and tube wells, cultivating Fadama or floodplain areas where water can be accessed via shallow wells or groundwater, and water management techniques such as conjunctive use. Farming practices to conserve water resources and reduce environmental degradation can also be introduced. These changes include reducing irrigation practice in water scarce areas, changing the timing of farm operations such as planting, irrigation and harvesting, and adopting new crop varieties and types. Livestock farming can also be enhanced with changes in grazing methods or reduction of herd sizes and animal diversification, herd sedentarization and supplementation. These changes will help to reduce the risk of animal malnutrition or even death. In our earlier analysis of vulnerability, fishing households were found to be the least vulnerable. Improvements in fishing techniques and the introduction of aquaculture can help them to cope even better and to serve as an adaptation strategy in the future. Success in introducing new strategies will depend on the willingness of farmers to change their practices, which has been explored in our survey. Farmers were asked their views on the adoption of selected options. As shown in Figure 8.7, most households are willing to adopt the selected options. Households expressed the greatest willingness to change the type of crop they cultivate and the timing of cultivation practices. They also indicated a willing-
3232 J&J Climate Change
1/11/07
9:56 am
Page 160
160 Climate Change and Adaptation
ness to change grazing practices and to relocate their farms. Farmers are less willing to reduce the size of their herd, as indicated by a high number of respondents who ‘disagree’ and ‘strongly disagree’ with this option. Livestock often are a family’s major asset and serve as security during droughts and famines. Consequently, many are reluctant to reduce their herds, even if the animals are vulnerable to lack of food and water during droughts.
Figure 8.7 Farmers’ willingness to change practices to reduce vulnerability to drought
Besides the use of existing and new coping strategies, supporting government policies are also critical for creating an enabling environment for adaptation. Important policy areas include education, research and development, institutional change and political will, religious and traditional institutions, and provision of infrastructural facilities. In the area of education, government should provide formal and informal educational institutions in order to improve literacy levels. Education will enable households to be informed and benefit from any strategy that may be introduced. Education will also enable households to access and utilize weather information. However, government must improve on information dissemination, provision of extension services and capacity building. Research and development are needed to better understand the risks to rural livelihoods from climate variability and change, as well as to develop and test risk management strategies and technologies. Important areas for research and development include climate change, water resources management, agricultural practice and development, land use and land use changes, farm inputs, and harvest and post harvest activities. These efforts will facilitate the monitoring of weather elements and the provision of warning signs as well as the development of good practice. Institutional change and political will are needed to ensure that policy and decision makers develop measures that will assist households in their adaptation process. Examples of such measures include credit facilities for
3232 J&J Climate Change
1/11/07
9:56 am
Page 161
Past, Present and Future Adaptation by Rural Households 161
households, favourable pricing and market policies, storage facilities, and the distribution of food commodities. Religious and traditional institutions can help to ensure community cooperation, assistance and participation through self-help. Finally, there must be an improvement in the provision of infrastructure facilities in rural areas for roads, electricity and communication networks, and development of industries such as food processing and marketing.
References Adger, N. W. (2001) ‘Scales of governance and environmental justice for adaptation and mitigation of climate change’, Journal of International Development, vol 13, pp921–931 Burton, I. (2004) ‘Climate change and the adaptation deficit’, Occasional Paper 1, Adaptation and Impacts Research Group (AIRG), Meteorological Services of Canada, Environment Canada, Toronto, Ontario, Canada Carney, D., M. Drinkwater, T. Rusinow, K. Neefjes, S. Wanmali and N. Singh (1999) ‘Livelihoods approaches compared: A brief comparison of the livelihoods approaches of the UK Department for International Development (DFID), CARE, Oxfam and the United Nations Development Programme (UNDP)’, November, Department for International Development (DFID), London, UK Christensen, J. H., B. Hewitson, A. Busuioc, A. Chen, X. Gao, I. Held, R. Jones, R. Koli, W. Kwon, R. Laprise, V. Rueda, L. Mearns, C. Menendez, J. Raisanen, A. Rinke, A. Sarr and P. Whetton (2007) ‘Regional climate projections’, in S. Solomon, D. Qin, M. Manning, Z. Chen, M.C. Marquis, K. Averyt, M. Tignor and H. L. Miller (eds) Climate Change 2007: The Physical Science Basis, contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York Dabi, D. D. (2006) ‘Agricultural and rural development policies in the third world: Lessons for Nigeria in the new millennium’, Abuja Journal of Geography and Development Sciences, vol 1, no 3 Dabi, D. D. and W. P. Anderson (1999) ‘Water use for commodity production in Katarko village, northern Nigeria’, Applied Geography, vol 19, no 1, pp105–122 Davies, S. (1996) Adaptable Livelihoods: Coping with Food Insecurity in the Malian Sahel, Macmillan Press, UK Downing, T. E., L. Ringius, M. Hulme and D. Waughray (1997) ‘Adapting to climate change in Africa’, Mitigation and Adaptation Strategies for Global Change, vol 2, pp19–44 Huq, S., A. Rahman, M. Konate, Y. Sokona and H. Reid (2003) Mainstreaming Adaptation to Climate Change in Least Developed Countries(LDCs), The International Institute for Environment and Development, Russell Press, Nottingham, UK Mortimore, M. (1989) Adapting to Drought, Cambridge University Press, Cambridge, UK McCarthy, J., O. Canziani, N. Leary, D. Dokken and K. White (eds) (2001) Climate Change 2001: Impacts, Adaptation and Vulnerability, contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York National Drought Mitigation Center (2005) ‘What is Drought? Understanding and defining drought’, Drought Impact Reporter, University of Nebraska–Lincoln, Lincoln, NE, US, www.drought.unl.edu/whatis/concept.htm
3232 J&J Climate Change
1/11/07
9:56 am
Page 162
162 Climate Change and Adaptation Nyong, A., A. Adepetu, V. Ihemegbulem and D. Dabi (2003) ‘Vulnerability of rural households to drought in northern Nigeria’, AIACC Notes, November 2003, vol 2, issue 2, pp6–7 Pielke, R. A. (1998) ‘Rethinking the role of adaptation in climate policy’, Global Environmental Change, vol 8, pp159–170 Smit, B., I. Burton, R. J. T. Klein and R. Street (1999) ‘The science of adaptation: A framework for assessment’, Mitigation and Adaptation Strategies, vol 4, pp199–213 Smit, B., I. Burton, R. J. T. Klein and J. Wandel (2000) ‘An anatomy of adaptation to climate change and viability’, Climatic Change, vol 45, pp223–251 Smit, B., O. Pilifosova, I. Burton, B. Challenger, S. Huq, R. Klein and G. Yohe (2001) ‘Adaptation to climate change in the context of sustainable development and equity’, in J. McCarthy, O. Canziani, N. Leary, D. Dokken and K. White (eds) Climate Change 2001: Impacts, Adaptation and Vulnerability, contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York Smithers, J. and B. Smit (1997) ‘Human adaptation to climatic variability and change’, Global Environmental Change, vol 7, no 2, pp129–146 Ozer, P., P. Erpicum, G. Demaree and M. Vandiepenbeeck (2003) ‘The Sahelian drought may have ended during the 1990s’, Hydrological Sciences, vol 48, no 3, pp489–496 Tarhule, A. and M. Woo (1998) ‘Changes in rainfall characteristics in northern Nigeria’, International Journal of Climatology, vol 18, no 11, pp1261–1271
3232 J&J Climate Change
1/11/07
9:56 am
Page 163
9
Using Seasonal Weather Forecasts for Adapting Food Production to Climate Variability and Climate Change in Nigeria James Oladipo Adejuwon, Theophilus Odeyemi Odekunle and Mary Omoluke Omotayo
Introduction An overarching anxiety among peasant farmers concerns the unpredictable onset and cessation of the rainy season. A foreknowledge of the weather for an upcoming growing season can enable farmers to plan with greater confidence to forestall negative consequences of poor or late rains and exploit beneficial opportunities when more favourable weather is in the offing. Extended-range or seasonal weather forecasts are made for West Africa and are a potential tool that could be used to great advantage for adapting farm decisions to climate variability. Successful application of seasonal forecasts in farming would also increase resilience to climate change. But seasonal forecasts are little used by Nigerian farmers. The reasons for this include inadequacies of the forecasts themselves and also failures in the communication of forecasts to farmers. In this chapter we describe the effects of climate variability on peasant farmers, strategies that are used by farmers to cope with the variability that do not depend on seasonal weather forecasts, and decisions that could benefit from the use of forecast information. Current efforts and capacity for extended-range forecasting in West Africa and Nigeria are examined and the skill and utility of the forecasts are evaluated. Based on our analysis, recommendations are made for improving forecasts and the communication of forecasts so that the potential benefits offered by this tool might be realized. The situation in Nigeria is relevant more broadly as a case study for subSaharan West Africa. The country encompasses climatic zones of the subcontinent from very wet to semiarid and all the indicator vegetation types of these zones are present, including evergreen rainforests, Southern Guinea Savannah, Northern Guinea Savannah, Sudan Savannah and Sahel Savannah.
3232 J&J Climate Change
1/11/07
9:56 am
Page 164
164 Climate Change and Adaptation
Of all the countries in West Africa, Nigeria alone is able to produce the complete range of foodstuffs characteristic of the subcontinent. Apart from the presence of the major tropical climate and vegetation types, the range of latitude, the varied relief and soils, together with differing peoples with their contrasted methods and crops, make this possible (Harrison-Church, 1956). This contribution is a synthesis of several investigations of climate variability, climate change and food security in sub-Saharan West Africa, the details of which have been provided in the technical report to the Assessment of Impacts and Adaptation to Climate Change (AIACC) Project (Adejuwon, 2006). Some of the work, including Adejuwon (2002, 2005 and 2006), Adejuwon and Odekunle (2004 and 2006), Odekunle (2003 and 2004) and Odekunle et al (2005), has been published in academic journals, in which details of the methods used can be found. The review of the farm operations and possible choices for responding to forecasts is based mainly on data collected during field surveys. The field surveys were conducted in five major ecological zones in Nigeria, including: rainforest (Atakumosa Local Government Area of Osun State), Southern Guinea savannah (Irepodun Local Government Area of Oyo State), Northern Guinea savannah (Oorelope Local Government Area of Oyo State), Sudan savannah (Askira Local Government Area of Bornu State) and Sahel savannah (Konduga Local Government Area of Bornu State). The assessment of the capacity represented by the major forecasting organizations (see Existing Capacity and Practice for Extended-range Weather Forecasting, on page 168) is based on a variety of sources, including Folland et al (1986 and 1991), Ward et al (1990), Philippon and Fontaine (2000), Colman and Richardson (1996), Colman et al (1997, 2000), websites of relevant organizations and climate records, and other information collected from the offices of the Nigerian Meteorological Agency at Oshodi, Lagos State. The methods used for the assessment of forecasting skills are described in detail in Adejuwon and Odekunle (2004).
Farm Operations and Variable Weather Farmers, whether in the humid or semiarid zones, realize the need to time farm operations to correspond with specific weather patterns. The dry season is used to prepare the land for cultivation. If the rains come too early, preparation operations could be adversely affected. The main anxiety about the onset of the rainy season concerns the planting date. Farmers like to sow their crops as early as possible. The earlier the crops are sown, the earlier the food products will be made available to end the annually occurring period of food deficiency. Moreover, crop yields are typically higher when planted early in the rainy season (Adejuwon, 2002; Fakorede, 1985) and farmers who get their crops early to the market are likely to enjoy better prices. The higher yields are explained by the early season nitrogen flush from farm residues of the previous season and higher incident solar radiation levels in the period before the heavy rains come. All of these factors explain why farmers are anxious about when the rains will come.
3232 J&J Climate Change
1/11/07
9:56 am
Page 165
Using Seasonal Weather Forecasts to Adapt Food Production 165
Sometimes there are false starts of the rainy season and farmers rush to plant their crops. This can result in disaster, as the seedlings may be completely lost. Replanting could be expensive if the seeds have to be purchased. If replanting seeds are sourced from the farmers’ stores, they could deplete food needed by the household during the period of low food supply. The amount of resources wasted in this way could be considerable when the crop concerned is yam. The yam seed is cut from the same tuber used as food, and could be up to 25 per cent of production. The farmers would, therefore, benefit considerably from a foreknowledge of the onset of the rainy season. Crop yield response to rainfall variability was investigated in the semiarid zone of Nigeria (Adejuwon, 2005). June rainfall turned out to be a more powerful predictor of crop yield than any of the other monthly rainfall variables. This is explained by the fact that June is the month of the onset of the rainy season. Low or insufficient June rainfall implies a delayed onset and a rainy season not long enough for the needs of most crops. September rainfall also served as a powerful predictor of yield. September is the month of cessation of the rainy season. Low or inadequate rain in September results in inadequate moisture for crops during the critical phases of grain filling, truncating the growing season prematurely. Crops such as cowpeas or late maize are planted so that they could be ready for harvesting after the rains have ceased. A late cessation of the rains means that the crops would be harvested under wet conditions, and much of the crop could be lost to mould and other pests and diseases. On the other hand, if the rains cease too early, the entire crop could fail due to inadequate moisture. There is a system of yam production practised at the drier margins of the rainforest zone which requires the seeds to be planted at the end of the rainy season, just before the rains cease. The seed remains dormant for the whole period of the dry season. As soon as the rains come during the following year, yam vines shoot up and harvestable tubers are produced two or three months ahead of the normal yam harvest season. New yam produced in this way commands very high prices, as it is preferred to the old yam, which would by then be losing its taste. The critical weather requirement of this system is that at least one heavy downpour must fall on the planted seeds before the dormancy period. In the absence of this, as much as 50 per cent of the crop could be lost. Thus the farmers can also benefit considerably from a foreknowledge of when the rains would cease. The length of the rainy season and the amount of rain that falls during the peak rainfall period are also watched with anxiety by farmers. Palm fruits are harvested by climbing the tree to cut the fruit. This is a very hazardous operation, especially during the rainy season when the trunks are slippery. Fewer climbers are available during the rainy season, and for this reason, palm products such as palm oil are in short supply and expensive. The fruits are left to waste and the longer the rainy season the greater the loss. Furthermore, during the rainy season rural roads become impassable and crops like cassava are unable to reach the market. Heavy tropical rainstorms can make all of the difference between a good harvest and crop failure. Farmers complain that such storms could cause a
3232 J&J Climate Change
1/11/07
9:56 am
Page 166
166 Climate Change and Adaptation
heavy loss to flowers before they become fruits. Cowpeas are an example of crops that could be damaged in this way. Heavy rainfall also reduces the number of pods per stand of cocoa, while it increases the degree of infestation by black pod disease (Thoroid, 1952; Adejuwon, 1962). Years with heavy rainfall, therefore, usually correspond to years with low yields of the crop. Moreover, cocoa harvested at the height of the rainy season has low grades and may not be able to make the export market. This is due to the prevailing heavy clouds and the little sunlight available for drying the produce. A prolonged little dry spell midway into the rainy season is a blessing to cocoa farmers. A minimum amount of rainfall during the dry season is needed for the establishment phase of tree crops. Cocoa is usually planted as seeds or seedlings during the rainy season. The new crop plants will die during the first dry season if no rain falls or if there is a spell of desiccating harmattan winds from the Sahara. Thus in the forest zone farmers meet with varying degrees of success in developing a new plot for the crop, depending on how much rain falls during the first dry season after planting (Adejuwon, 1962). Farmers’ needs from the weather forecaster thus include statements on whether or not the dry season will be completely dry.
Coping Without the Benefits of Weather Forecasts It needs to be noted that the peasant farmers are not altogether helpless in the absence of weather forecast information. Traditional agricultural practices include a number of options designed to mitigate the negative consequences of unfavourable weather which are applicable each year, whether or not the weather turns out to be unfavourable. One good example of such practices is the storage of water and the use of shallow wells to extend the period of adequate water supply at the end of the rainy season. Such practices will help to cushion the impacts of abrupt termination of the rainy season. Mulching is another practice which helps to protect seedlings against dry spells during the earlier parts of the growing season. The use of wetlands, whether extensive floodplains or local valley bottoms, is also part of traditional agricultural practice. Wherever such lands are available, they help to reduce the impacts of milder droughts on peasant communities. As of now, farmers use a number of hedging strategies while expecting the worst of weather and hoping for the best. Such strategies include multiple cropping, relay cropping and intercropping. These are recognizable features of traditional farming practices evolved over time all over the country and are designed to make one crop serve as an insurance against the failure of another. In the case of multiple cropping, crops that occupy different ecological niches are planted together on the same plot. For example, maize may be planted with melon. While maize is a standing crop, melon is a creeper. In the case of relay cropping, several crops are planted in succession on the same plot to make use of different parts of the growing season. For example, in the Guinea Savannah Zone of south-western Nigeria, maize is planted in April, harvested in July; sorghum is planted in June and harvested in November. In the case of inter-
3232 J&J Climate Change
1/11/07
9:56 am
Page 167
Using Seasonal Weather Forecasts to Adapt Food Production 167
cropping, some crops are planted at low density among the major crop, which is planted on every heap. For example, maize could be planted at low density on a farm plot primarily meant for yam production. In the forest zone, tree food crops are maintained as insurance against the failure of field crops. In normal years, the fruits of the tree crops are not harvested. However, during years of inadequate rainfall, when the yields of the field crops are not able to sustain the peasants’ livelihood, they fall back on the tree crops whose food products are considered to be of lower quality. Another example: cocoyam, planted under cocoa, is treated as a weed during normal years. However, whenever the major food crops fail, it becomes a dependable source of food. There are also measures for damage control as soon as it is realized that an abnormally dry season is in the offing. One such measure is replanting with crops with shorter growing seasons or that are more tolerant of arid conditions. Water yam with a maturity period of 3 months could be planted to replace white yam, whose maturity period lasts seven months; millet, with a maturity period of two months, could be planted to replace sorghum, which takes six months to mature.
Decisions that Might Benefit from Forecast Information In a typical tropical region like West Africa, rainfall is the principal controlling element of crop productivity (Nieuwolt, 1982; Stern and Coe, 1982). The crop plants are sensitive to the moisture situation both during their growth and development, and especially as they reach maturity. This is reflected in a definite soil and atmospheric moisture range in which field preparations are expected to commence and also in which such farm operations as sowing, thinning; transplanting, weeding, irrigation, insecticide and fertilizer application, as well as harvesting are scheduled to take place. Thus, as seasonal weather changes from one year to another, the most suitable crops, cropping systems and operations schedules also change. The need thus arises to choose from a range of options. As an adaptation strategy, extended-range weather forecasting represents an early warning system that could be used at the farm level for decision making. A foreknowledge of seasonal weather affords the farmer the opportunity to make decisions that could enhance the productivity of his farm and maximize returns on his inputs of land, labour and capital. Ex ante decisions that might benefit from a foreknowledge of seasonal weather include the timing of farm operations such as land preparation, tillage, planting, transplanting, thinning, weeding, irrigation and harvesting. Whether, when and how much insecticide, herbicide, fungicide and fertilizer to apply can also be optimized with forecast information. Choices of crops, crop varieties, tillage method and depth, and density of planting can be adapted to the anticipated weather. Whether or not to adopt water conserving practices, which type to adopt, and whether to store water for irrigation and the mode of irrigation to use are other decisions that can benefit from forecasts. If drought conditions
3232 J&J Climate Change
1/11/07
9:56 am
Page 168
168 Climate Change and Adaptation
are expected, poorly drained fadama soils can be productively cultivated, or deep, loamy soils can be cultivated if abundant rainfall is anticipated. The best choice of farming system, for example, single, multi-cropping or intercropping, depends on the weather. How much credit to secure and how much of the harvest to store or sell are also decisions that can profit from forecast information.
Existing Capacity and Practice for Extended-range Weather Forecasting West Africa depends to a large extent on organizations based in Europe and North America for its operational weather forecasting. The UK Meteorological Office (UKMO or Met Office), the French Centre de Recherche de Climatologie of the Centre National de la Recherche Scientifique (CNRS) and the National Oceanic and Atmospheric Administration (NOAA) in the US, routinely make forecasts directed at the West African subcontinent. Within Africa, the Africa Centre for Meteorological Applications for Development (ACMAD) is responsible for gathering, collating and disseminating forecast information, while the national meteorological services are responsible for issuing weather forecasts for their own countries and sub-regions. The statistical forecasting methods of the Met Office, CNRS and NOAA are described in Adejuwon (2007). The Met Office has been making experimental forecasts of seasonal rainfall in the Sahel since 1986. Forecasts are made for four regions for the months of June, July, August and September using ocean and atmospheric information that is available in early May (Colman et al, 1996 and 1997; Graham and Clark, 2000).The predictions use seasonally averaged sea surface temperature anomalies (SSTA) with a spatial resolution of 10° 10°. The CNRS forecasts cumulative rainfall for June to September in West Africa, also using sea surface temperature as well as other indices, but aggregated to a spatial resolution of 5° 5° (Philippon and Fontaine, 2000). In order to make the forecasts available before the beginning of the growing season, only information available by the end of April is employed. NOAA produces experimental forecasts for rainfall anomalies for July, August and September in the Sahel using predictors with a spatial resolution of 2° 2° (Thiaw and Barnston, 1996, 1997 and 1999). So far, the experiments confirm that the global SSTA field is the best predictor, a view shared by the European and African forecasting teams. Additional fields such as upper air geo-potential heat, tropical low level wind and outgoing long-wave radiation could enhance forecast skill. However, the data do not extend far enough into the past (minimum of 25 years for an adequate control period). ACMAD (www.acmad.ne), based in Niamey, Niger Republic, was created in 1987 by the Conference of Ministers of the United Nations Economic Commission for Africa and the World Meteorological Organisation to meet the challenge of weather prediction for the continent. The strategy is for the centre to lead both in training personnel for capacity building within the continent and implementing operational activities, such as the issuance of weather infor-
3232 J&J Climate Change
1/11/07
9:56 am
Page 169
Using Seasonal Weather Forecasts to Adapt Food Production 169
mation products. In the execution of its programmes, ACMAD operates within a network with international partners, regional partner institutions and national focal points. The international partners include the CNRS, the Met Office and NOAA, among others. These organizations provide the primary weather information to be collated and transmitted to the end users within the various African countries. The regional partner institutions have specific sector and regional responsibilities that require weather and climate information. Among such institutions are ICRISAT and AGRHYMET. The focus of ICRISAT is on agricultural development in semiarid tropics while AGRHYMET deals with agriculture and hydrology in the Sahel region of West Africa. These organizations are expected to help in broadcasting information to sector end users. To reciprocate, they also use the website of ACMAD to advertise their products for the benefit of their stakeholders. The primary focal points are the national meteorological services of 53 African countries. Focal points for ACMAD have also been established within the operational structure of sub-regional economic groupings, such as the Economic Community of West African States and the South Africa Development Community. The focal points are the primary recipients of the products emanating from ACMAD, meant for end users in agriculture, energy, water resources and other sectors within the various countries. The operational products of the Numerical Weather Prediction Unit within ACMAD include the daily Meteorological Bulletin for national meteorological services and the daily continent-wide 24-hour public significant weather forecasts. These products give relatively accurate forecasts but do not go far enough into the future for our purposes. The numerical weather production unit of ACMAD also issues 5-day guidance bulletins on specific days of the week (Mondays, Wednesdays and Fridays). On other days (Tuesdays and Thursdays) the unit issues 3-day guidance bulletins to monitor and follow up the previous day’s 5-day guidance bulletin. The terms ‘guidance’ and ‘forecast’ are understood in the context that ACMAD issues weather guidance to African National Meteorological Services, which have the responsibility to issue the weather forecasts for the specific countries or sub-regions. The conventional network of surface and upper air observations within ACMAD’s area of responsibility is generally too poor to make longer-term relevant numerical forecasting feasible. However, ACMAD, through its African Climate Watch page (www.acmad.ne), advertises two relevant products for short-term, mediumterm and seasonal weather prediction: ‘El Niño/La Niña Update on Impact over Africa’ and ‘Rainfall Onset over West Africa’. The former gives earlywarning type of forecasts extending over three to nine months. However, this is a very new product that has yet to be tested. The other product predicts the onset of the rainy season within three to six weeks (Omotosho, 1990). However, the predictions are confirmed only within three weeks. The Cotonou workshop on ‘Climate variability prediction, water resource and agricultural productivity: Food security in tropical sub-Saharan Africa’ led to ACMAD being charged with organizing the Seasonal Climate Prediction Forum for West Africa, PRESAO, with assistance from START to arrange
3232 J&J Climate Change
1/11/07
9:56 am
Page 170
170 Climate Change and Adaptation
funding (Fleming et al, 1997). PRESAO pools expertise in the subcontinent for the purpose of weather prediction and assembles producers and users of predictions for two-way interactions to improve the forecasts and the ability of the agricultural and water resource sectors to use forecast information. The PRESAO forecast map appears on the webpage of ACMAD and is meant to be accessed and downscaled by the national meteorological services. The Nigerian Central Climate Forecasting Office (CFO) of the Nigerian Meteorological Agency (NIMET) is responsible for short-term and extendedrange weather forecasts in Nigeria. Ahead of each cropping season, the CFO issues a bulletin on the weather outlook for the season. The CFO makes forecasts based on analogue, statistical and dynamic methods as well as forecasts downscaled from European and American forecasting organizations. While making its local forecasts, the CFO makes use of SSTA data of 2° 2° resolution (Toure, 2000), compared with SSTA data on a resolution of 5° 5° employed by the CNRS (Philippon and Fontaine, 2000) and the 10° 10° grid SSTA data used by the Met Office (Colman et al, 2000). For its own local forecasts, the CFO makes use of the current trend of the weather, the pressure systems, the position of the Intertropical Convergence Zone (ITCZ) and global SSTA. The CFO also makes forecasts of the dates of the first rains (which is not the same thing as the date of onset of the rainy season as defined by Ilesanmi (1972a). Very often, the first rains represent only a false start.
Assessment of Seasonal Forecast Skill The forecasts assessed have been prepared by various meteorological organizations, including: the Met Office, the CNRS, NOAA and the CFO. The forecasts of the international organizations were obtained from articles published in Experimental Long-Lead Forecasts Bulletin. The forecasts of the Nigerian CFO were obtained directly from their offices in Lagos. The stations selected to test the forecasting skills of the various tools include Benin City, Lagos, Ibadan, Ilorin, Enugu, Minna, Jos, Kaduna, Lokoja, Maiduguri and Kano, all located within Nigeria. The stations have been selected to represent the various climatic and ecological zones between the Gulf of Guinea in the south and Sahara Desert in the north. Their selection is also based on the availability of rainfall data from 1961 to 2000. The forecasts assessed were for the five years from 1996 to 2000. The scheme used for the assessment of forecast skill is described in detail in Adejuwon and Odekunle (2004). Skilful forecasts are those that are subsequently confirmed by observations. High skills are demonstrated when forecasts are very close to observations, whereas low skills are recorded when the two are substantially different. One practical problem in assessing the skills is the fact that observations and forecasts are not presented in the same units of measurement. Observations are usually presented on an interval scale with the amounts of rainfall given in millimetres. On the other hand, forecasts are stated using ordinal categories. The most common are quint categories varying from very wet to wet, average, dry and very dry. Determination of what is very
3232 J&J Climate Change
1/11/07
9:56 am
Page 171
Using Seasonal Weather Forecasts to Adapt Food Production 171
wet, wet and so on in this exercise was based on the records from 1961 to 1995. Rainfall values for each year, whether annual, seasonal or monthly, were arranged in descending order of magnitude and divided into five groups. The resulting highest quint consists of the values for the seven wettest years and the lowest those of the seven driest years. The years with rainfall values falling within the range in the highest quint are classified as very wet, while years with values falling within the range of rainfall in the lowest quint are classified as very dry. Other years are similarly classified as wet, average or dry. Sometimes tercile categories are used by simply forecasting near normal, above normal or below normal. Determination of what is above normal, near normal or below normal was also based on the records from 1961 to 1995. Rainfall values for each year, whether annual, seasonal or monthly, were arranged in descending order of magnitude and divided into three groups. The resulting highest tercile (above normal) consists of the 12 wettest years and the lowest (below normal) those of the 12 driest years. The 11 middle years define the near-normal range. The quint and the tercile limits provide the framework for converting both observations and forecasts to the same units of measurement (Adejuwon and Odekunle, 2004). In assessing the skills of forecasts, the same criteria were used to classify observations as were used for the forecast categories. Where observations and forecasts fell within the same category, skill was assessed as high. Where there was a one-category difference – for example, forecast was average but observation was very wet – skill was assessed as moderate. In situations of more than one category disparity between observation and forecasts, the skill was assessed as low. Tables 9.1 and 9.2 provide the framework for the assessment of the skills of the forecasts. Table 9.1 Skill assessment of quint forecast categories Forecast
Very Wet
Wet
Average
Dry
Very Dry
High Moderate Low Low Low
Moderate High Moderate Low Low
Low Moderate High Moderate Low
Low Low Moderate High Moderate
Low Low Low Moderate High
Observations Very Wet Wet Average Dry Very Dry
Table 9.2 Skill assessment of tercile forecast categories Forecast
Above Normal
Near Normal
Below Normal
High Moderate Low
Moderate High Moderate
Low Moderate High
Observations Above Normal Normal Below Normal
3232 J&J Climate Change
1/11/07
9:56 am
Page 172
172 Climate Change and Adaptation
The skill of the forecasts for June, July, August and September annual rainfall totals, as determined in the analysis, is 26 per cent rated as high, 45 per cent moderate and 29 per cent low. There is thus considerable room for improvement. The results ranked NOAA and the CFO ahead of the CNRS and the Met Office in weather forecasting skill over West Africa (Table 9.3). A careful look at the background of the methods of data collection and analysis appears to explain the relative successes of the forecasting organizations. Virtually all of the forecasting organizations made use of the SSTA as a predictor among others. However, a noticeable difference in the nature of the sea surface temperature data used by various organizations is with respect to the spatial resolution. Although the CFO and NOAA made use of SSTA data of 2° 2° resolution (Tourre, 2000), the CNRS used seasonally averaged 5° 5° grid SSTA data (Philippon and Fontaine, 2000) and the Met Office used seasonally averaged 10° 10° square SSTA data (Colman et al, 2000). It thus appears that the finer the SSTA resolution, the better the forecasting skill. Table 9.3 Organizational skill performance assessment of the June, July, August and September annual rainfall totals Forecasting Organization UKMO (all) NOAA CNRS CFO
Percentage Contribution of Each Skill Category High
Moderate
Low
21% 56% 18% 20%
44% 22% 55% 70%
35% 22% 27% 10%
The number of predictor variables used in the forecast models also seems to have played a role in determining the level of skill. Although the Met Office, whose tools seem to be less skilful than the others, made use of SSTA data alone in the construction of their prediction models, other forecasting centres made use of additional rainfall formation-related factors. For instance, the CFO, which came first on the basis of having the smallest low-skill score, made additional use of synoptic data, including current weather, pressure systems, equivalent potential temperature and the position of the ITCZ. NOAA, which ranked first on the basis of the highest high-skill score made additional use of upper air geo-potential heat, tropical low level wind and outgoing long-wave radiation (Thiaw and Barnston, 1999). The CNRS, which was third, used additional factors, such as geo-potential indexes, describing near surface humidity and moist static energy values (Philippon and Fontaine, 2000). Note that the ITCZ is one major factor not used by the CNRS in the construction of their prediction model. The comparison indicates that the inclusion of these additional predictor variables in the construction of a forecast model improves the skill of the forecasts. The study clearly demonstrates regional disparities in the skills of the forecasting tools (Table 9.4). The prediction skill is generally higher for southern
3232 J&J Climate Change
1/11/07
9:56 am
Page 173
Using Seasonal Weather Forecasts to Adapt Food Production 173
coastal locations than for the northern continental locations. It is well known that the Atlantic Ocean is the major, if not the only, source of moisture into the West Africa subcontinent. The moisture is brought to the land areas by the southwesterly winds moving in after the northward migrating ITCZ. The characteristics of the southwesterly winds, which bring the moisture to the continent, are, in turn, determined by the nature of the sea surface temperature of the Gulf of Guinea (Adedokun, 1978). It is thus logical that the conditions of the southwesterly winds, as determined by the SSTA and its associated ITCZ, would be least modified near the coast. As the ITCZ advances and the southerly winds progress further inland, their thermodynamic transformations become more pronounced. The changes in the nature of southwesterly winds and other rainfall-associated factors are thus a function of space and time. The space connection between rainfall over the land and the activities over the Atlantic Ocean weakens with distance between a location over the land and the sea. It is, therefore, not surprising that a prediction model based on SSTA is more skilful in the south, near the ocean, than in the interior of the continent. Table 9.4 Regional disparities in forecasting skill Stations
Maiduguri Kano Kaduna Minna Jos Ilorin Lokoja Ibadan Enugu Ikeja Benin
Percentage Contribution of Each Skill Category
Sahel Sudan Northern Guinea Northern Guinea Plateau Southern Guinea Southern Guinea Dry Forest Dry Forest Rainforest Rainforest
High
Moderate
Low
17% 17% 20% 25% 17% 17% 50% 29% 33% 29% 33%
33% 17% 60% 25% 50% 83% 25% 71% 33% 57% 67%
50% 66% 20% 50% 33% 0% 25% 0% 34% 14% 0%
Source: Adejuwon and Odekunle (2004)
Inadequacies in the Existing Forecast Capacity A number of inadequacies are associated with the forecasts made between 1996 and 2000. First, the skill was not sufficiently high, and there was no evidence to the effect that it was improving (Adejuwon and Odekunle, 2004). Second, the forecasts were directed at the total rainfall amount, whereas the more relevant determinants of crop yield were neglected. As noted by Odekunle and Gbuyiro (2003), all of the recent studies on rainfall predictability in West Africa, including those from the Department of Meteorology, University of Oklahoma (Berte and Ward, 1998), Nigerian Central Forecasting Office, (Omotosho et al 2000), Centre de Recherche de Climatologie
3232 J&J Climate Change
1/11/07
9:56 am
Page 174
174 Climate Change and Adaptation
(Philippon and Fontaine, 2000) and UK Meteorological Office (Colman et al, 2000), Gbuyiro et al (2002) and Wilson (2002), were directed at rainfall amounts only. The number of rain days, which is a measure of rainfall effectiveness, was hardly ever included. The same total amount of rainfall is expected to benefit crops more coming as drizzle over several days than when it comes in one heavy downpour (Odekunle and Gbuyiuro, 2003). Other important parameters of rainfall that were not considered as predictands are the rainfalls of onset and retreat periods. These are of paramount importance in the subregion because they affect regional economies (Walter, 1967; Olaniran, 1983; Adejuwon et al, 1990). A failure in early establishment of rainfall onset, for example, usually indicates a drought in the early part of the rainy season, as noted earlier. A third inadequacy, as noted by Odekunle et al (2005), is that the forecasting tools currently deployed are regional in approach and general in perspective. Models developed at such coarse spatial scales often fail at individual farm-level sites. Within the same zone, in respect of which the forecasts are usually made, a wet season at one station may be a dry season at another (Adejuwon et al, 1990). The disparities in forecasting skills between stations lying in the same ecological zones, as demonstrated in Table 9.4, are probably a result of such intrazonal variability. The fourth and most serious inadequacy of the existing capacity is the absence of an effective means of communicating the forecasts with any end user in the agricultural sector. During stakeholder field surveys conducted in all ecological zones, from the rainforest belt in the south to the Sahel savannah zone in the north, in 50 village communities not a single respondent affirmed receiving either directly or indirectly weather-related information from the meteorological services (Adejuwon, 2005). Extension personnel routinely advise the farmers in their charge on the timing of farming activities. It turned out that such advice is based not on weather information received from the meteorological services, but on the extension agents’ knowledge of the climate of the area.
Measures for Upgrading Forecasting Effectiveness New attempts are being made to generate extended-range weather forecasting models in Nigeria that address the inadequacies noted above (Odekunle and Gbuyiro, 2003; Odekunle, 2004, 2006; Odekunle et al 2005; Adejuwon and Odekunle, 2006). The models, which are yet to be incorporated into the seasonal weather forecasting programmes of the country, are designed to improve the skills, credibility and applicability of the existing capacity at the local level. For example, Odekunle et al (2005) examined the rainfall onset and retreat dates between 1962 and 1996 in Nigeria, and generated models for their prediction. The study adopted a composite of rainfall-promoting factors: the sea surface temperature of the tropical Atlantic Ocean, land-sea thermal contrast between selected locations in Nigeria and the tropical Atlantic Ocean, surface location of the Intertropical Convergence Zone, and the land surface
3232 J&J Climate Change
1/11/07
9:56 am
Page 175
Using Seasonal Weather Forecasts to Adapt Food Production 175
temperature in selected locations. Rainfall and temperature data were collected from Ikeja, Benin, Ibadan, Ilorin, Kaduna and Kano, in Nigeria. Cumulative percentage mean rainfall was employed to generate the rainfall onset and retreat dates series, while the method of stepwise multiple regression analysis was used to construct the required prediction models. The results obtained showed that the hypothesized rainfall-promoting factors are efficient in predicting rainfall onset and retreat dates. Both the statistical goodness-of-fit (using both the R2 and a-values) and the actual goodness-of-fit (comparing the observed rainfall onset and retreat dates with the predicted values using 1962–1969 and 1995–1996 data) of the models obtained in this study support the reliability of the models for predicting rainfall onset and retreat dates in Nigeria.
Measures for Communicating Forecasts to End Users More than forecasting skill is needed if forecasts are to be useful to, and used by, farmers. Much depends on the communication of forecasts, how they are perceived and how they are understood. First, the forecasts must be credible (Patt and Gwato, 2002). Where previous forecasts were perceived as incorrect, the tendency would be for subsequent ones to be ignored. Credibility is determined primarily by the level of skill, but it is also a function of the difference between what was promised by the forecasts and what was realized. At their best, forecasts are probabilistic. By couching forecasts in deterministic language, skill will invariably be perceived as low whenever there is an incorrect forecast. On the other hand, a probabilistic forecast is accorded the benefit of the doubt if there are infrequent, incorrect forecasts. Considerable efforts, therefore, need to be put into how the forecasts are interpreted before they are transmitted. It is not always easy to find a simple expression to convey the concept of probability, especially in a non-mathematical context such as that in which peasant farmers operate. Another requirement for a credible forecast is that it must be presented at a spatial resolution or scale with which individuals can identify. As indicated above, forecasts are made for extensive zones, within which there could be considerable differences in the actual weather observed. The CFO divides Nigeria into five zones and makes its forecasts for each zone. In order to achieve a high skill level for each zone, the skill levels for the localities within the zone, with which the farmers are familiar, have been compromised. Second, the forecasts must be presented in simple, understandable language. Forecasts need to be presented in the local languages. Where forecasts are rendered in strange and confusing language, users will either not incorporate them or they will do so in a way that is counterproductive (Patt and Gwato, 2002). However, forecasters cannot be expected to be experts in all the local languages, and leaving the translation entirely in the hands of local agricultural extension personnel may introduce considerable distortions. Sometimes, to increase farmers’ acceptance of the forecasts, doses of exaggeration are introduced. Apart from this, technical terminologies in English may
3232 J&J Climate Change
1/11/07
9:56 am
Page 176
176 Climate Change and Adaptation
not have local language equivalents. Hence, there is the need for the forecasters to work repeatedly with users or the intermediary conveyers of forecasts to be able to arrive at an appropriate forecasting language for each locality. Third, the meteorological department should also bear the responsibility of issuing forecasts early enough to be useful in planning the following season’s operations. Planting normally starts as soon as the rainy season is established. The decision as to what farm operations to adopt for the season, therefore, must have been made before then. While the rainy season may come as late as June in the semiarid zone, it could come as early as February in the humid coastal zone. Thus, weather forecasts made in April or May are practically useless to half of the farming community in Nigeria – the expected end users will simply not be in a position to make use of them. Fourth, the forecasts should include information about the consequences of forecasted weather, for example, effects on yields of different crops, that can enable farmers to understand potential outcomes of decisions based on tradition or observations and compare them to results that might be attained if the forecasts are acted on. If forecasts do not contain enough new information to alter specific decisions, the intended users will not organize their activities in response to them (Patt and Gwato, 2002). As indicated previously, the peasant farmers are not altogether helpless in the absence of weather forecast information. Traditional agricultural practices include a number of no-regret options designed to mitigate the negative consequences of unfavourable weather, applicable each year, whether or not the weather turns out to be unfavourable. These cover a considerable proportion of the risks posed by interannual climate variability. One way of providing the type of information that could offer credible choices is to indicate the changes in crop yield that could be expected given the forecast in question. The Agricultural Meteorology Department of NIMET must be equipped in terms of personnel to issue such forecasts. Even though there is an Agricultural Meteorology Department within NIMET, it appears that the organization is not explicitly charged with activities that could expedite the transmission of the forecasts. Its extended-range weather forecasts are consigned to the archives as soon as they are made. Either by amendment to the Act Establishing the Agency or amendment to the relevant statute or regulations, the responsibility of NIMET in this regard should be clearly stated. The Agricultural Meteorology Department should be strengthened and charged with the responsibility of translating the forecasts into forms that could be understood and are relevant to the needs of farmers. The government at the federal level should also accommodate budgetary provisions for communicating the forecasts to the agricultural extension services. A related problem concerns the channel of communicating the forecasts with the farmers. Our suggestion is that the existing administrative structure for local government be used. Just as ACMAD recognizes the national meteorological services as the focal points within each member country for the dissemination of its products, NIMET should identify focal points close to the farmers for purposes of passing on weather forecast information to them. Because NIMET is a federal agency, the logical location of such focal points is within the administrative structure of the states. However, this would tend to
3232 J&J Climate Change
1/11/07
9:56 am
Page 177
Using Seasonal Weather Forecasts to Adapt Food Production 177
prolong the line of communication and reduce its efficiency. Therefore, we suggest the Local Government Administration (LGA) as the location of NIMET focal points. The precedent for this is the direct transfer of the LGA component of federally generated revenue to the local governments without passing through the states. There are more than 700 LGAs in the country. Each LGA is structured to have a department of agriculture, whose political head is the Supervisory Councillor for Agriculture. There are also technical staff within such departments with qualifications comparable to those of holders of the Higher National Diploma, obtained after four years’ post-secondary education and one year of industrial training. Focal points established within such departments will be close enough to the farmers for effective and efficient transmission of forecasts. The LGA represents government at the grass roots and their technical staff in the agricultural departments are expected to operate among the rural populace, especially among farmers. They speak the same language as the farmers and are well placed to ensure that vital weather information gets to them within a few days. As part of the bottom–up approach, we suggest that the forecasts be based on the existing 28 synoptic weather stations in the country. On average, each station will be expected to represent the weather of an area of some 30,000 km2. This, it is hoped, will enhance the skill of the forecasts, as they could be accessed from each farm site. This means that 28 weather forecast pamphlets will be produced to cover the country and that it is the forecasts for the weather station closest to each focal point that will be sent to the farmers who need them. However, communication should not just be one-way. There should be opportunities for feedback from end users about the skill of the forecasts. This could be in the form of specific requests for forecasts or in the form of complaints about the level of skills demonstrated.
Conclusions Extended-range weather forecasting is a basic step for creating an early warning system for farmers in West Africa. Skilful, timely and effectively communicated forecasts of climate variables that matter to farmers can be applied in comprehensive adaptation strategies to forestall the negative consequences of climate variability and climate change. However, the existing forecast system in Nigeria, and West Africa generally, has a number of inadequacies in both the prediction skill and the usefulness of the forecasts to the end users. A high percentage of the forecasts are found to have moderate or low skill, while a low percentage of forecasts exhibit high skill. In addition to problems with forecast skill, the usefulness of forecasts to the end users is limited by lack of forecasts of rainfall characteristics that are important to farmers such as onset, cessation, length of the growing season and number of rain days; lack of forecasts for specific localities instead of extensive zones; and lack of forecasts for the coastal and middle belts of West Africa by some of the
3232 J&J Climate Change
1/11/07
9:56 am
Page 178
178 Climate Change and Adaptation
forecasting organizations. Seasonal forecasts of rainfall characteristics of West Africa can be made more skilful and useful. As described in this chapter, we have made skilful predictions of the onset and cessation dates of the rainy season using SST and land/sea thermal contrast alone. In other published works, we have also generated prediction models for the length of the rainy season, the length of the ‘safe’ period and the number of rain days (Odekunle and Gbuyiro, 2003; Odekunle, 2004). The models generated represent an improvement on the existing weather forecasting tools, as they could be used to make forecasts for specific locations (in other words the areas around the weather stations) and for all the zones from the coast to the Sahel. As well as improving the skills of the forecasts, constraints that prevent the use of forecasts by peasant farmers need to be addressed. These include inadequate or lack of credibility, late publication of forecasts, cognitive deficiencies, and absence of clear choices associated with specific forecasts. The existing agricultural extension system has an extensive local presence throughout the farming communities of Nigeria that could be used as channels for communicating forecasts. To be effective, close cooperation between forecasters and extension personnel is needed to ensure that forecasts are timely, are rendered in forms that are relevant to farmers’ decisions, are presented in ways that make their probabilistic nature understood, are in languages used by the farmers and have credibility.
References Adedokun, J. A. (1978) ‘West African precipitation and dominant atmospheric mechanisms’, Archives for Meteorology Geophysics and Bioclimatology, Series A, vol 27, pp289–310 Adejuwon, J. O. (1962) ‘Crop-climate relationship: The example of cocoa in Western Nigeria’, Nigerian Geographical Journal, vol 5, pp21–31 Adejuwon, J. O. (2002) ‘Extended weather forecasts as management tool for the enhancement of crop productivity in sub-Saharan West Africa’, Nigerian Meteorological Society Journal, vol 3, pp25–38 Adejuwon, J. O. (2005) ‘Food crop production in Nigeria: I. Present effects of climate variability’, Climate Research, vol 30, pp53–60 Adejuwon, J. O. (2006) ‘Climate variability, climate change and food security in subSaharan West Africa’, technical report of AIACC Project No AF23, International START Secretariat, Washington, DC (www.aiaccproject.org/Final%20Reports/ final_reports.html) Adejuwon, J. O., T. O. Odekunle and M. O. Omotayo (2007) ‘Extended-range weather forecasting in Sub-Saharan West Africa: Assessing a potential tool for adapting food production to climate variability and climate change’, AIACC Working Paper No 46, International START Secretariat, Washington, DC (www.aiaccproject.org) Adejuwon, J. O., E. E. Balogun and S. A. Adejuwon (1990) ‘On the annual and seasonal patterns of rainfall fluctuations in sub-Saharan West Africa’, International Journal of Climatology, vol 10, pp839–848 Adejuwon, J. O., and T. O. Odekunle (2004) ‘Skill assessment of the existing capacity for extended-range weather forecasting in Nigeria’, International Journal of
3232 J&J Climate Change
1/11/07
9:56 am
Page 179
Using Seasonal Weather Forecasts to Adapt Food Production 179 Climatology, vol 24, pp1249–1265 Adejuwon, J. O. and T. O. Odekunle (2006) ‘Variability and severity of the Little Dry Season in south-western Nigeria’, Journal of Climate, vol 19, pp1–8 Berte, Y. and M. N. Ward (1998) ‘Experimental forecast of seasonal rainfall and crop index for July–September 1998 in Cote D’Ivoire’, SODEXAM, Meteorologie Nationale de Cote D’ivoire, Abidjan, Department of Meteorology, University of Oklahoma, Norman, OK Colman, A.W. and D. Richardson (1996) Multiple Regression and Discriminant Analysis Predictions of July–Aug–Sept 1996: Rainfall in the Sahel and other Tropical North Africa Regions, Ocean Application Branch NWP (National Weather Prediction) Division, UK Met Office, Bracknell, UK Colman, A. W., D. M. Harrison, T. Evans and R. Evans (1997) Multiple Regression and Discriminant Analysis Predictions of July–Aug–Sept 1999 Rainfall in the Sahel and other Tropical North Africa Regions, Ocean Application Branch NWP Division, UK Met Office, Bracknell, UK Colman, A. W., M. K. Davey, R. Graham and R. Clark (2000) Experimental Forecast of 2000 Seasonal Rainfall in the Sahel and other Regions of Tropical North Africa, Using Dynamical and Statistical Methods, Met Office, Bracknell, UK Fakorede, M. A. B. (1985) ‘Response of maize to planting dates in a tropical rainforest location’, Experimental Agriculture, vol 21, pp19–30 Fleming, C., M. Hoepftner, A. Freise and M. Sobtj (1997) Climate Variability, Climate Prediction, Water Resource and Agricultural Productivity: Food Security in Tropical Sub-Saharan Africa, proceedings of the Cotonou-Benin workshop, July, Cotonou, Benin Folland, C. K., T. N. Palmer and D. E. Parker (1986) ‘Sahel rainfall and worldwide sea temperature’, International Journal of Climatology, vol 13, pp271–285 Folland, C. J., J. Owen, M. N. Ward and A. Colman (1991) ‘Prediction of seasonal rainfall in the Sahel region using empirical and dynamical methods’, Journal of Forecasting, vol 10, pp21–56 Gbuyiro, S. O., M. T. Lamin and O. Ojo (2002) ‘Observed characteristics of rainfall over Nigeria during ENSO years’, Nigerian Meteorological Society Journal, vol 3, pp1–17 Graham, R. and R. Clark (2000) Experimental Forecast of 2000 Seasonal Rainfall in the Sahel and Other Regions of Tropical North Africa Using Dynamical and Statistical Methods, Meteorological Office, Bracknell, UK Harrison-Church, R. J. (1956) West Africa, Longman Green and Co, London Nieuwolt, S. (1982) Tropical Climatology: An Introduction to the Climates of the Low Latitudes, John Wiley and Sons, New York Odekunle, T. O. (2004) ‘Rainfall and the length of the growing season in Nigeria’, International Journal of Climatology, vol 24, pp467–477 Odekunle, T. O. (2006) ‘Determining rainy season onset and retreat over Nigeria from precipitation amount and number of rainy days’, Theoretical and Applied Climatology, vol 83, pp193–201 Odekunle, T. O. and S. O. Gbuyiro (2003) ‘Rain days predictability in south-western Nigeria’, Nigerian Meteorological Society Journal, vol 4, pp1–17 Odekunle, T. O., E. E. Balogun and O. O. Ogunkoya (2005) ‘On the prediction of rainfall onset and retreat dates in Nigeria’, Theoretical and Applied Climatology, vol 81, pp101–112 Olaniran, O. J. (1983) ‘The onset of the rains and the start of the growing season in Nigeria’, Nigerian Geographical Journal, vol 26, pp81–88 Omotosho, J. B. (1990) ‘Theory of rainfall onset prediction’, The African Climate Watch, ACMAD (www.acmad.ne/uk) Omotosho, J. B., A. A. Balogun and K. Ogunjobi (2000) ‘Predicting monthly and seasonal rainfall, onset and cessation of the rainy season, in West Africa, using only
3232 J&J Climate Change
1/11/07
9:56 am
Page 180
180 Climate Change and Adaptation surface data’, International Journal of Climatology, vol 20, pp865–88 Patt, A. and C. Gwato (2000) ‘Effective seasonal climate applications: Examining constraints for subsistence farmers in Zimbabwe’, Global Environmental Change, vol 12, pp185–195 Philippon, N. and B. Fontaine (2000) West African June to September Rainfall: Experimental Statistical Forecasts Based on April Values of Regional Predictors, Centre de Recherche de Climatologie, 21000 Dijon, France Stern, R. D. and R. Coe (1982) ‘The use of rainfall models in agricultural planning’ Agricultural Meteorology, vol 26, pp35–50 Thiaw, W. and A. Barnston (1996) CCA Forecast for Sahel Rainfall in Jul–Aug–Sept 1996, Climate Prediction Centre, NOAA, Camp Springs, MD Thiaw, W. and A. Barnston (1997) CCA Forecast for Sahel Rainfall in Jul–Aug–Sept 1997, Climate Prediction Centre, NOAA, Camp Springs, MD Thiaw, W. and A. Barnston (1999) CCA Forecast for Sahel Rainfall in Jul–Aug–Sept 1999, Climate Prediction Centre, NOAA, Camp Springs, MD Thoroid, C. A. (1952) ‘The epiphytes of Theobroma cacao in Nigeria in relation to the incidence of the black pod disease’, Journal of Ecology, vol 40, pp126–142 Toure, Y. M. (2000) CLIMAB: 2000. Version 1.0., International Research Institute for Climate Prediction, Columbia University, New York Walter, M. W. (1967) ‘The length of the rainy season in Nigeria’, Nigerian Geographical Journal, vol 10, pp123–128 Ward, M. N., J. A. Owen, C. K. Folland and G. Farmer (1990) ‘The relationship between sea surface temperature anomalies and summer rainfall in Africa 4–20°N’, Long-Range Forecasting and Climate Memorandum, no 32, available from the National Meteorological Library, Met Office, Bracknell, UK Wilson, S. M. (2002) ‘Empirical prediction of seasonal rainfall in Nigeria’, Nigerian Meteorological Society Journal, vol 3, pp25–29
3232 J&J Climate Change
1/11/07
9:56 am
Page 181
10
Adapting Dryland and Irrigated Cereal Farming to Climate Change in Tunisia and Egypt Raoudha Mougou, Ayman Abou-Hadid, Ana Iglesias, Mahmoud Medany, Amel Nafti, Riadh Chetali, Mohsen Mansour and Helmy Eid
Introduction Agriculture in North Africa is both the main use of the land in terms of area and the principal water-consuming sector, accounting for over 70 per cent of total water consumption (Iglesias, 2003; Iglesias et al, 2003). Rainfall is low and characterized by large year-to-year variation. Water is scarce and the countries of North Africa are considered to be water stressed, withdrawing large proportions of available surface waters for agricultural, domestic and other uses. Climate change is projected to bring warmer temperatures and changes in rainfall that, together, have the potential to reduce water availability. The impacts that climate change will have on agriculture and water supplies, and the potential for conflicts over water, are critical concerns in the region. Under current conditions, North African countries import significant amounts of grain and it seems likely that climate change will lead to an expansion of this import requirement. The effects of sea level rise in North Africa, especially on the coast of the Delta Region of Egypt, would reduce the area under cultivation and probably reduce agricultural production (Desanker and Magadza, 2001; El-Raey, 1999; Abd-El Wahab, 2005). In addition to the effects of climate change, agricultural production in the region is expected to change rapidly due to technological advancements, economic changes, such as potential trade agreements with the European Union, and projections of high population increase (Iglesias et al, 2004). Over 80 per cent of the cropland in the region is used for rain-fed production, which produces very low and variable yields (FAOSTAT, 2005). In Tunisia, an even larger percentage of cropland, 93 per cent, is used for rain-fed agriculture, yet irrigation is the largest consumptive use of water (Mougou and Salem, 2003). Egypt is an exception in the region, with almost 90 per cent of
3232 J&J Climate Change
1/11/07
9:56 am
Page 182
182 Climate Change and Adaptation
cropland being irrigated (FAOSTAT, 2005). Water demand for irrigation is expected to increase in all countries of North Africa and it is important to define adaptation strategies that take into account the possible deficit of water for irrigation in the future (Strzepek et al, 1995; Iglesias et al, 2003). Northern Africa’s adaptation capacity to climate change is challenged as it is compounded by high development pressure, increasing population, water management that is already using most of the available water resources, and agricultural systems that are often not well adapted to current conditions (Abdel et al, 2003; Eid, 1994; Iglesias and Moneo, 2005). Evidence of limits to adaptation of socioeconomic and agricultural systems in the North African region has been documented in recent history. For example, water reserves were not able to cope with sustained droughts in the late 1990s in Morocco and Tunisia, causing many irrigation dependent agricultural systems to cease production (Iglesias and Moneo, 2005). Adaptation to climate change in North Africa is a major issue from the perspectives of food production, rural population stabilization and distribution of water resources. Previous studies have addressed adaptation in a top–down approach, evaluating theoretical options with little relation to current agricultural management. But there is also a need to incorporate management knowledge for formulating adaptation measures in a bottom–up approach (Abdel et al, 2003; Eid, 1994; Eid et al, 1993, 1995 and 1996). The aim of our study is to evaluate adaptation measures for rain-fed and irrigated agriculture in Tunisia and Egypt. Using surveys of farmers and interviews with technical resource managers, information is collected about the sensitivity of farm operations to climate and available adaptation measures. Information was also provided to farmers and other stakeholders to increase their understanding of the interactions between climate and agriculture. Selected measures are evaluated for Tunisia using the expert judgements of agricultural managers and farmers, while in Egypt, adaptation measures are evaluated quantitatively using agricultural simulation models.
Case Study of Rain-fed Cereal Farming in Kairouan, Tunisia Tunisia belongs to the group of water scarce countries. Rainfall is characterized by its scarcity and spatial and temporal variability (Mougou et al, 2002). The average of total annual rainfall varies from 1500mm in the north to 100mm in the south. High temperatures and solar radiation result in high rates of evapotranspiration and an environment of water stress. Variability and scarcity of water resources and high temperatures negatively affect agricultural production, most particularly rain-fed agriculture. The Tunisian case study focuses on cereal farming Kairouan in central Tunisia. The central region is a climatic transition zone between the Mediterranean zone and the Sahara region. Severely dry years with a deficit of evapotranspiration of over 50 per cent are frequent in this region and represent a recurrent risk for rain-fed agriculture, which is practiced on nearly 90 per cent of farmlands in Kairouan. High temperatures during the grain filling
3232 J&J Climate Change
1/11/07
9:56 am
Page 183
Adapting Dryland and Irrigated Cereal Farming to Climate Change 183
period can negatively affect crop growth and development. Another major risk to crops is the very dry wind blowing from the Sahara known as the Sirocco, which occurs between May and September. If the Sirocco occurs during the grain filling stage, an irreversible loss of water from the plant can take place (Mougou and Henia, 1998). Tunisian farmers apply a variety of strategies to lessen the negative effects of high temperatures, high rates of evaporation and periods of below normal rainfall. Some of these practices may be useful as strategies for adapting to future climate change. To identify and evaluate adaptation measures in use, a survey was conducted of farmers in the Kairouan region. One hundred farmers of rainfed cereals in the region were selected at random and surveyed (Mougou et al, 2003). The sample represents 5 per cent of the 2000 cereal farmers in the study area. The survey sought to identify farmers’ behaviours for coping with climate variability to characterize the capability of the farmers to adapt to climate variability, define the ‘know-how’ of the farmers in relation to predicted climate change, list the adaptation methods already in use, and identify factors that prevent farmers from adapting to current and future climate variability.
Survey results Cereal yields fluctuate in the surveyed farms, especially for rain-fed cereals. There is also variation in yields across farms due to differences in soils, farm inputs and management practices. For example, yields of durum wheat and barley range between 24 and 47 quintals (100kg) per hectare and 38 and 54 quintals per hectare for irrigated durum wheat and barley respectively, whereas for rain-fed wheat and barley the respective ranges are 0–33 and 0–40 quintals. However, scarcity of water and financial considerations prevent irrigation on 88 per cent of farms in the study area. On rain-fed farms, variability in cereal production is explained by the variability of rainfall, with values of 78 per cent for the north, 50 per cent for the centre and 40 per cent for the south regions (see Figure 10.1 and Mougou et al, 2003). For the study region of Kairouan, variability of rainfall explains 56 per cent of the cereal yield variability. Differences in average cereal yields for wet and dry years over the period 1995–2003 are shown in Table 10.1 for sites at Kairouan, Sbikha, Haffouz and Ouslatia. The average yield in dry years at the four sites is 6.9 quintals per hectare, while in wet years the average is 15.6 quintals, or 132 per cent more. The survey highlighted the fact that agriculture in the region is mainly practised by old farmers with an average age of 58 years and with low levels of schooling. Illiterate farmers represent 55 per cent of the sample; 36 per cent have primary education, 5 per cent have secondary education and 3 per cent have had schooling at higher levels. Average farm size is 29.8ha, which is large by Tunisian standards. Approximately one third of farmers have farms smaller than 10ha, while 54 per cent have farms between 10 ha and 50ha. Only 12 per cent of farmers have farms larger than 50ha. The participation of family members in agricultural activities is high; we found that in 62 per cent of the farms family members participate in farming.
3232 J&J Climate Change
1/11/07
9:56 am
Page 184
184 Climate Change and Adaptation
Figure 10.1 Wheat yield and spring precipitation in Tunisia, 1961–2000 Source: Mougou et al (2003).
Table 10.1 Average cereal yields during wet and dry years at four sites in the Kairouan region of Tunisia Rainy years Dry years (1995/1996 and 2002/2003) (1996/1997 to 2001/2002)
Yield increase during rainy years compared Mean rainfall Mean yields Mean rainfall Mean yields to dry years (%) (mm) (qt) (mm) (qt)
Kairouan Sbikha Haffouz Ouslatia Average
231.55 275.2 269.15 317.7 273.4
15.5 17.5 14 15.5 15.6
114.2 112.6 99. 130.9 114.2
6.5 6.3 16 8.6 6.9
138 178 133 80 132.3
Surface ploughing is used by almost all farmers since it is relatively cheap. Only 29 per cent of the farmers practise crop rotation, fertilizers are used by only 17 per cent, and this is done only during years with favourable weather. While a majority of farmers use commercial seeds, nearly 40 per cent use seeds that they produce themselves or buy from other farmers. Supplemental irrigation, defined as the application of a limited amount of water to the crop when rainfall fails to provide sufficient water for plant growth to increase and stabilize yields (Oweis et al, 1999), is used by roughly one quarter of the farmers. Supplemental irrigation is applied mainly to fodder crops for livestock. The irrigated area represents less than 4 per cent of the total cultivated area because of the small amount of water available and the financial constraints on purchasing irrigation equipment and materials. All farmers who use supplemental irrigation are conscious of the advantages of fertilization and its management. It is important to note that rain-fed cereal production is the principal activity even for the farmers who have access to water.
3232 J&J Climate Change
1/11/07
9:56 am
Page 185
Adapting Dryland and Irrigated Cereal Farming to Climate Change 185
Nearly three quarters of surveyed farmers supplement their cereal production by raising sheep. This serves to diversify farm households’ income and food sources and acts as a hedge against crop losses. In poor years, livestock are sold to provide income, while in good years the number of livestock is increased. Nearly 30 per cent of farmers reported liquidating their herds because of the drought experienced in 1997–2002. All farmers expressed their suffering from difficult climatic conditions. Indeed they suffered from long years of drought. Even during the rainy years rainfall distribution can be inadequate for the crops. Dry conditions in March decrease cereal production and result in a loss of income for the farmer. According to farmers’ responses, over 95 per cent of agricultural output is determined by climate conditions and by water scarcity. This perception results in a feeling of frustration, and it is obvious that this is the result of the drought during recent years (1997 to 2002). Concerning the farmers’ understanding of climate change, only 52 per cent know about climate change phenomena and only 12 per cent are aware of the necessity to adapt to climate change in a different way than to climate variability. Although only a small number of farmers understand the concept of adapting to climate change, over 90 per cent employ measures to adapt to current climate variability. Practices currently used include digging a well (48 per cent), shifting sowing date if the autumn is drier than normal (48 per cent), storing fodder to ensure supply of food for livestock (57 per cent), changing cultivation techniques to decrease management costs (75 per cent) and applying supplemental irrigation when rainfall is not enough to ensure a minimum water requirement (44 per cent). Other practices that are used yearly regardless of climate are breeding sheep, which are resilient to adverse climate conditions and consume a range of fodder resources, and growing cactus and other crops for use as fodder. Despite the diversity of adaptation methods used by Kairouan farmers, yields are low and variable and adaptation is incomplete and inefficient. Factors that limit adaptation include the low level of education of farmers, difficulties for the extension services to change the farmers’ behaviours and difficulties for farmers to adopt new techniques even if they agree with them. Survey results indicate that only 20 per cent of the sampled farmers adopt the advice of extension services. Others use some adaptation methods but do them in their own way and do not follow exact prescriptions for implementing recommended technologies.
Implications for adapting to climate change Farmers in the survey area appear to overestimate the effect of climate variability on crop yields and so underestimate the potential effectiveness of adaptive strategies for improving yields and reducing the variability of yields. While nearly all farmers use adaptation measures to reduce risks from climate variability, many confirm that they do not use them consistently with the recommendations of extension service agents. This may reduce their effectiveness. In addition, some investments by farmers may not be implemented
3232 J&J Climate Change
1/11/07
9:56 am
Page 186
186 Climate Change and Adaptation
because low prices for agricultural products result in low economic returns on the investments. Our results show that only 20 per cent of farmers adopt the advice of extension services. Indeed, it arises also from the survey data that the extension services are effective only in medium and large farms. Only one third of farmers that follow extension service advice are smallholder farmers with farms less than 10ha in size. This is confirmed by Chennoufi and Nefzaoui (1996), who find that ‘technologies generated appear to be more readily adopted by largescale farmers, agricultural development agencies, rural development societies and cooperative farms, rather than by the majority of medium-and small-scale farmers’. This behaviour goes contrary to the strategies of rural development in Tunisia. In fact the strategies of rural development are directed towards a participative approach and providing technical aid mainly for small farmers, who represent about 80 per cent of farmers in Tunisia. In the Tunisian private sector the primary agricultural areas, 24 per cent of farm land, are cultivated by 1 per cent of farmers with farms of at least 100ha (Souki 1994). Their large incomes allow large-scale farmers to use new techniques and to be advised by qualified people. New agricultural techniques need to be developed and introduced to farmers, with a focus on smallholder farmers, to build their capacity for coping with and adapting to climate change. The involvement of the rural population is essential to ensure that appropriate technologies are being developed and that farmers are provided with the knowledge and skills necessary for effective adoption of the technologies. Research and development on improvement of cereal varieties started in Tunisia in the 1970s. Packages of agronomic varieties and techniques have been developed mainly for wet regions. But attempts to diffuse new technologies to farmers have not been entirely successful. New approaches that engage farmers are being used to improve dissemination of new technologies and to transfer to farmers knowledge and skills that are necessary for their effective adoption. Cooperation with development agencies has been improving through short-term training, field days, joint research programmes, research contacts, utilization of libraries and databases, use of laboratories and research stations, and joint publications and reports. The most innovative move has been the establishment of seven ‘regional development poles’ for research, one per large agro-ecological zone, which offer a good framework for bringing together all partners in research at the regional level, including development agencies and farmers’ representatives (Lasram and Mekni, 1999). During the last ten years, the agronomic research programmes in Tunisia have mainly been focused on the assessment of drought impacts on agricultural production. Because of the potential for the climate to become drier and more drought-prone, results of this research are expected to provide more options for adapting to climate change. Research on water resources management, the improvement and the genetic selection of cultivars, the improvement of techniques already used, and the use of new techniques such as fertilization and irrigation methods are particularly promising.
3232 J&J Climate Change
1/11/07
9:56 am
Page 187
Adapting Dryland and Irrigated Cereal Farming to Climate Change 187
Case Study of Irrigated and Rain-fed Wheat Farming in Egypt Precipitation in Egypt is only significant in the northern Mediterranean coast, where average annual rainfall is roughly 180mm, and is extremely low in the rest of the country’s desert territory. Consequently, agriculture is restricted to the fertile lands of the narrow Nile Valley from Aswan to Cairo and the flat Nile Delta north of Cairo. Together this comprises only 3 per cent of the country’s land area. The study in Egypt focuses on the Nile Delta, represented by Sakha in the Kafr El-Shikh Governorate and Mersa Matrouh on the northwest coast of Egypt. The objective of the Egyptian case study is to detect important adaptation measures of the wheat crop production system under irrigated and rain-fed agriculture systems. The study was conducted in three steps: stakeholders engagement, evaluation of adaptation measures and simulation of effectiveness of selected adaptation measures using crop models. The Nile Delta region is characterized by high-production irrigated smallholder agriculture, high urban water demands and rapidly growing population. Water for irrigation, which comes entirely from the River Nile, varies due to changes in freshwater availability and to competition among water users. The total area cultivated for wheat in Egypt reached nearly one million hectares in 2004, just over half of which are located in the Delta region. The region also accounts for just over half of the wheat production of Egypt, 7.2 million tons, and has an average yield of 6.6 tons of wheat per hectare (MALR, 2004). The Mersa Matrouh region is one of the few regions of rain-fed agriculture in Egypt. The region is characterized by low population and low agricultural productivity. The total cropped area in the governorate is about 95,000ha, of which approximately 20 per cent is cultivated for wheat in the winter season (MALR, 2004). Wheat yields average only two tons per ha in Mersa Matrouh, or about 30 per cent of the national average. The low yields are due to scarce and highly variable water availability for rain-fed agriculture in this arid region and to low levels of farm inputs and management. The stakeholders engaged in the study represent the smallholder farmers, commercial farmers and strategic resource managers. Training programmes, field demonstration and workshops were organized by the study to orient stakeholders with the issues of climate change and adaptation, and to build an information exchange between stakeholders and the study team. The impacts of climate change on agriculture and possible adaptation measures were discussed with stakeholders. Adaptation measures proposed by stakeholders include changing cropping patterns, adopting new drought-tolerant cultivars, reducing or constraining the cultivation of high water consuming crops, changing sowing dates, changing fertilization practices, and improving the efficiency of water application and water use.
3232 J&J Climate Change
1/11/07
9:56 am
Page 188
188 Climate Change and Adaptation
Simulation of Adaptation Measures Simulation models were used to quantify some of the adaptation measures proposed by the stakeholders. The modelling study considered on-farm adaptation techniques such as the use of alternative varieties of wheat, optimizing the time of planting and optimizing the application of irrigation water. Simulations were performed for irrigated wheat production in the Sakha district of the Kafer El-Sheikh governorate in the Delta region and for rain-fed wheat production at the Sidy Baraney district in the Marsa Matrouh governorate on the northern coast. The CERES-Wheat model developed by the International Benchmark Sites Network for Agrotechnology Transfer and included in the DSSAT 3.5 package was used for the simulations (Tsuji et al, 1995). The simulations were performed for current climate conditions and for two scenarios of future climate change that were constructed using the MAGICC/SCENGEN software of the University of East Anglia (UK) and input data from climate projections of the HadCM3 general circulation model for the A1 and B2 SRES emissions scenarios (Nakicenovic and Swart, 2001; Eid et al, 2001). The regional temperature increases corresponding to the A1 and B2 scenarios are 3.6°C and 1.5°C respectively. Precipitation changes in most of the territory are about ±10 to 20 per cent, depending on the season. Data for daily maximum and minimum temperatures, precipitation, solar radiation, soil properties and water quality were collected for the period 1975–1999 for Sakha and 1993–1995 for Sidy Baraney.
Irrigated wheat in the Nile Delta region of Egypt For irrigated wheat, the simulation experiments are designed to investigate variations in crop yield and crop water demand in response to higher temperatures for different choices of wheat cultivars (Sim 1), water conservation measures (Sim 2), and sowing dates (Sim 3). The irrigation requirements of the crop, applied by flood irrigation, and application of fertilizer are set according to recommendations of the agricultural extension services. Other assumptions incorporated into each of the simulation cases are summarized in Table 10.2. Two outputs of the DSSAT model are used as indicators of the effects of climate and adaptation measures: the percentage change in wheat yield and the percentage change in evapotranspiration (Etcrop). For Sim 1, three cultivars common to Egypt and the Delta region are selected. The selected cultivars are Giza-168, which is planted on 24 and 22 per cent of land cultivated for wheat in Egypt and the Delta respectively, Sakha-8, which is planted on 2.0 and 1.6 per cent of land cultivated for wheat in Egypt and the Delta respectively, and Sakha-69, which is planted on 10 and 13 per cent of land cultivated for wheat in Egypt and the Delta respectively. Simulations were performed for each cultivar for the current climate, a climate that is 1.5°C warmer on average and a climate that is 3.6°C warmer on average. In each of these cases irrigation water is applied at a level of 600mm and the crop is planted on 20 November, which reflect the current recommendations for these management variables.
3232 J&J Climate Change
1/11/07
9:56 am
Page 189
Adapting Dryland and Irrigated Cereal Farming to Climate Change 189
Table 10.2 Assumptions for simulations of irrigated wheat production in the Nile Delta region Sim 1
Sim 2
Sim 3
Change cultivars
Change water amount
Change sowing dates
Climate scenario
Current +1.5°C +3.6°C
Current +1.5°C +3.6°C
Current +1.5°C +3.6°C
Wheat cultivar
Giza-168 Sakha-8 Sakha-69
Sakha-8
Sakha-8
Sowing date
20 November (recommended)
20 November (recommended)
20 November (0 shifting days) 1 December (11 shifting days) 10 December (20 shifting days) 20 December (30 shifting days)
Irrigation amount
600mm/season (recommended)
300mm/season (50% water saving) 400mm/season (33% water saving) 500mm/season (17% water saving) 600 mm/season (0% water saving)
600mm/season (recommended)
Adaptation measure
Sim 2 varies the amount of irrigation water over a range of 300 to 600mm for the Sakha-8 cultivar. For Sim 3, the effects of delaying planting by 11, 20 and 30 days are investigated for the Sakha-8 cultivar and an irrigation application of 600mm. Results for the three irrigated wheat simulation cases are presented in Table 10.3. In Sim 1, the productivity of all three wheat cultivars is reduced by higher temperatures relative to current yields, while Etcrop is increased. Under current climate conditions, Giza-168 provides the highest yields of the three and is less water demanding than Sakha-8. However, the effects of higher temperatures are more pronounced on Giza-168 than on the two other cultivars, with the exception that Sakha-69 is not viable for the 3.6°C warming case. Yields of Giza-168 are reduced by 30 and 37 per cent for warming of 1.5 and 3.6°C respectively. In comparison, yields of Sakha-8 are reduced by 4 and 26 per cent for 1.5 and 3.6°C warming. These results suggest that the Sakha-8 cultivar is more stable under climate change conditions than Giza-168 and Sakha-69. For this reason, Sakha-8 was selected for use in the two other simulation cases. Water conservation is an important objective for irrigated agriculture in the Nile Delta because of high national agricultural water demands. The joint effects of climate change and water conservation measures are simulated in the Sim 2 cases, the results of which are illustrated in Figure 10.2(a). Under the
3232 J&J Climate Change
1/11/07
9:56 am
Page 190
190 Climate Change and Adaptation
Table 10.3 Results of simulations for irrigated wheat in the Nile Delta region (percentage changes are relative to current climate) Indicator Climate Sim 1 (cultivars) Sim 2 (water amounts) scenario Giza- Sakha-Saka- 300 400 500 600 168 8 69 mm mm mm mm
Sim 3 (sowing dates) 20 01 10 20 Nov Dec Dec Dec
Yield (t/ha)
Current +1.5°C +3.6°C
6.8 4.7 4.2
4.7 4.5 3.5
6.1 4.7 na
3.6 3.0 2.3
4.5 4.2 3
4.7 4.5 3.6
4.7 4.5 3.9
4.8 4.5 3.1
4.7 4.5 3.1
4.7 4.5 3.4
Yield change (%)
+1.5°C +3.6°C
–30 –37
–4 –28 –26 na
–16 –35
–6 –32
–4 –25
–5 –18
–6 –35
–5 –34
–4 –5 –28 –27
Etcrop (mm)
Current 354 +1.5°C 357 +3.6°C 503
464 352 489 358 498 na
313 316 310
403 409 405
464 489 498
476 513 579
469 498 507
462 490 501
467 490 498
476 491 494
Etcrop change (%)
+1.5°C +3.6°C
1 –1
1 0
5 7
8 22
6 8
6 8
5 7
3 4
1 42
5 7
2 na
4.7 4.5 3.4
current climate, reducing flood irrigation from the recommended rate of 600mm to 500mm or even 400mm have small effects on yields while generating significant water conservation benefits. At 300mm, a 50 per cent reduction in water use, yields would be reduced 25 per cent relative to what can be attained with 600mm in the current climate. But this is considered an acceptable trade-off in this water scarce region. In comparison, under climates that are 1.5° and 3.6°C warmer, a 50 per cent reduction in water use would result in unacceptably high losses of wheat yield of 33 per cent and more. In the warmer climate scenarios, water conservation measures would probably be limited to less aggressive targets due to economic considerations. With 3.6°C warming, a reduction to even 500mm of water, which has negligible effects on yields in the current climate or in a climate that is 1.5°C warmer, would cut yields by nearly 10 per cent. Another adaptation option proposed by farmers is to shift sowing dates. This strategy is represented by the Sim 3 cases, for which the date of sowing is delayed by 10, 20 and 30 days compared to the current norm. The results are illustrated in Figure 10.2(b). Delaying planting dates results in negligibly changed yields in the current climate and under a climate 1.5°C warmer. But for 3.6°C warming, delaying sowing by 20 to 30 days would increase yields by 10 to 11 per cent. Changes in sowing dates had only small effects on evapotranspiration. The simulation results indicate that changes in wheat cultivars can be an effective strategy for adapting to a warmer climate. Of the cultivars tested for the selected climate change scenarios, the Sakha-8 cultivar appears to have advantages over the Giza-168 cultivar, which is more widely used at present. But the performance of different cultivars is controlled by local environmental conditions such as the local climate, soil condition and water availability, and
3232 J&J Climate Change
1/11/07
9:56 am
Page 191
Adapting Dryland and Irrigated Cereal Farming to Climate Change 191
Figure 10.2 Simulated changes in irrigated wheat yields and evapotranspiration under different climate conditions for (a) water conservation measures and (b) variations in sowing dates
the best choice of cultivar will vary by location and will differ according to the change in climate. The results also show that climate change may constrain water conservation efforts, as yields are more sensitive to reduced flood irrigation at higher temperatures. Changes in planting dates are found to have little benefit except for the warmest climate scenario studied.
3232 J&J Climate Change
1/11/07
9:56 am
Page 192
192 Climate Change and Adaptation
Rain-fed wheat in Mersa Matrouh on the northwest coast The impacts of climate change on yields of rain-fed wheat grown in Mersa Matrouh on the northwest coast are simulated for warming of 1.5°C combined with assumed changes in rainfall that include no change from the current average of 240mm per year, ±10 per cent and ±20 per cent. Simulations are also made for changes in sowing dates as an adaptation to the changes in climate. Yields of rain-fed wheat in this dry climate are highly sensitive to variations in rainfall. At current temperatures, a 10 per cent reduction in rainfall would reduce yields by 26 per cent, while a 20 per cent reduction in rainfall would reduce yields by nearly 50 per cent (Figure 10.3a). Yields respond positively but less strongly to rainfall increases. Warming of the climate would amplify the yield losses for rainfall reductions and would erode the yield gains from increases in rainfall. Yields can be improved 10 to 35 per cent by delaying planting dates, both in the current and warmer climates, with the best results being achieved with a delay of 40 days (Figure 10.3b). These results indicate that a change in sowing dates could be utilized to reduce the negative effects of water shortage from reduced rainfall.
Conclusions Climate is perceived by farmers in the study areas of Tunisia and Egypt as a major risk to agricultural production. A variety of practices are in use to help manage climate-related risks, but their adoption is incomplete and their use is often not consistent with the advice of agricultural extension services. The survey of Tunisian farmers found that they tend to over-estimate the impact of climate variability and consequently fail to recognize the potential of farm management techniques to offset the effects of adverse climate conditions. The extension services are found to be effectively used primarily by large-scale farms and some medium-scale farms. Therefore, new agricultural techniques and extension programmes need to be developed and introduced to small- and medium-scale farmers to build up their capacity to manage climate risks. Changes in wheat cultivars, irrigation practices and sowing dates were suggested by farmers and other stakeholders in Egypt as options for adapting to climate risks and were evaluated using crop simulation models. The results demonstrate that it is possible to adapt and partly offset the effects of climate change on wheat yields using these strategies. The simulations also pointed to the potential of climate change to constrain water conservation efforts as wheat yields are more sensitive to reductions in irrigation water at higher temperatures. Further work is needed to explore synergistic effects of different combinations of farm management strategies to identify promising options, both through model simulations and in field tests with farmers. But a key challenge will be to improve the communication and diffusion of agricultural methods for managing climate risks.
3232 J&J Climate Change
1/11/07
9:56 am
Page 193
Adapting Dryland and Irrigated Cereal Farming to Climate Change 193
Figure 10.3 Simulated changes in rain-fed wheat yields for (a) different rainfall scenarios and (b) variations in sowing daes
3232 J&J Climate Change
1/11/07
9:56 am
Page 194
194 Climate Change and Adaptation
Dedication The authors of this work dedicate this chapter to the memory of Professor Helmy Eid, who passed away in 2005. Professor Eid, a valued colleague and friend, helped to advance the understanding of climate change issues in Egypt and contributed to the research reported in this publication.
References Abdel, H., N. G. Ainer and H. M. Eid (2003) ‘Climate change impacts on Delta crop productivity, water and agricultural land’, Journal of Agricultural Science, special issue: ‘Scientific symposium on problems of soils and water in Dakahlia and Damietta governorates’, 18 March, pp15–26 Abd-El Wahab, H. M. (2005) ‘The impact of geographical information system on environmental development’, MSc thesis, Faculty of Agriculture, Al-Azhar University, Cairo Chennoufi, A. and A. Nefzaoui (1996) The Role of Universities in the National Agriculture Research Systems of Egypt, Jordan, Morocco, Sudan and Tunisia. Case Study: Tunisia, Food and Agriculture Organization, Rome Desanker, P. and C. Magadza (2001) ‘Africa’, in J. McCarthy, O. Canziani, N. Leary, D. Dokken and K. White (eds) Climate Change 2001: Impacts, Adaptation and Vulnerability, contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York Eid, H. M. (1994) ‘Impact of climate change on simulated wheat and maize yields in Egypt’, in C. Rosenzweig and A. Iglesias (eds) Implications of Climate Change for International Agriculture: Crop Modeling Study, US Environmental Protection Agency, Washington, DC Eid, H. M., A .A. Rayan, K. A. Mohamed, M. M. A. El-Refaie, M. M. Attia, H. A. Awad, K. M. R. Yousef and M. M. A. El-Koliey (1995) ‘Impact of climate change on yield and water requirements of some major crops’, in 2nd Conference of OnFarm Irrigation and Agroclimatology, 2–4 January, Agricultural Research Center, Egypt, pp492–507 Eid, H. M., N. M. Bashir, N. G. Ainer and M. A. Rady (1993) ‘Climate change crop modeling study on sorghum’, Annals Agricultural Science, Special Issue 1, pp219–234 Eid, H. M., S. M. El-Marsafawy, N. G. Ainer, M. A. Ali, M. M. Shahin and N. M. ElRaey (1999) ‘Impact of climate change on Egypt’, special report, EEMA, Cairo, Egypt Eid, H. M., S. M. El- Marsafawy and N. G. Ainer (2001) ‘Using MAGICC and SCENGEN climate scenarios generator models in vulnerability and adaptation assessments’, Conference on Meteorology and Environmental Issues, March, Cairo, Egypt El-Mowelhi, N. M. and O. El-Kholi (1996) Vulnerability and Adaptation to Climate Change in Egyptian Agriculture, Country Study Report, Country Studies Program, Washington, DC Iglesias, A. (2003) ‘Climate, drought and prediction in the Mediterranean: Opportunities for agricultural adaptation’, Revista de Ingenieria Civil, vol 131, pp25–31 Iglesias, A. and M. Moneo (2005) Drought Preparedness and Mitigation in the Mediterranean: Analysis of the Organizations and Institutions, Options Mediterranean, Series B, No 51,Universidad Politécnica de Madrid, Madrid, Spain
3232 J&J Climate Change
1/11/07
9:56 am
Page 195
Adapting Dryland and Irrigated Cereal Farming to Climate Change 195 Iglesias, A., M. N. Ward, M. Menendez and C. Rosenzweig (2003) ‘Water availability for agriculture under climate change: Understanding adaptation strategies in the Mediterranean’, in C. Giupponi and M. Shechter (eds) Climate Change and the Mediterranean: Socioeconomic Perspectives of Impacts, Vulnerability and Adaptation, Edward Elgar Publishers, Northampton, MA, US Iglesias, A., N. X. Tsiourtis, D. A. Wilhite, A. Garrido, L. Garrote, M. Moneo, A. Gomez-Ramos, M. J. Hayes and C. Knutson (2004) ‘Terms of reference for drought risk management: Drought identification studies, drought risk analysis and best practices’, MEDROPLAN working paper, Mediterranean Agronomic Institute of Zaragoza, Zaragoza, Spain Lasram, M. and M. Mekni (1999) The National Agricultural Research System of Tunisia, WANA NARS study, International Center for Agricultural Research in the Dry Areas (ICARDA), Aleppo, Syria MALR (Ministry of Agriculture and Land Reclamation) (2004) Agricultural Statistics, vol 1, Economic Affairs Sector, Ministry of Agriculture and Land Reclamation, Cairo, Egypt Mougou, R. and M. Ben Salem (2003) ‘Meteorological conditions in arid regions and effects of climate change in dryland crops’, Proceedings of the Training on Agricultural Techniques for Rain-fed Agriculture and Communication to Farmers, Arab Center for Studies in Dry land Agriculture (ACSAD), Tunis, Tunisia Mougou, R. and L. Henia (1998) ‘Contribution à l’étude des phénomènes à risques en Tunisie cas du sirocco’, Les Publications de l’Association Internationale de Climatologie, vol 9 Mougou. R., S. Rejeb and F. Lebdi (2002) ‘The role of Tunisian gender issues in water resources management and irrigated agriculture in Tunisia’, in ‘The First Regional Conference on Perspectives of Arab Water Cooperation: Challenges, Constraints and Opportunities’, workshop on gender and water management in the Mediterranean countries, Cairo Nakicenovic, N. and R. Swart (eds) (2001) Emissions Scenarios, special report of Working Group III of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York Oweis, T., A. Hachum and J. Kijne (1999) ‘Water harvesting and supplementary irrigation for improved water use efficiency in dry areas’, SWIM Paper 7, International Water Management Institute, Colombo Souki, K. (1994) ‘Analysis of the effects of water and nitrogen supply on the yield and growth of durum wheat under semiarid conditions in Tunisia’, doctoral thesis, University of Reading, Reading, UK Strzepek, K., D. Yates, G. Yohe, R. Tol and N. Mander (2001) ‘Constructing “not implausible” climate and economic scenarios for Egypt’, Integrated Assessment, vol 2, pp139–157 Tsuji, G. Y., J. W. Jones, G. Uhera and S. Balas (1995) Decision Support System for Agrotechnology Transfer, V 3.0 (three volumes), International Benchmark Sites Network for Agrotechnology Transfer (IBSNAT), Department of Agronomy and Soil Science, University of Hawaii, Honolulu, HI
3232 J&J Climate Change
1/11/07
9:56 am
Page 196
11
Adapting to Drought, Zud and Climate Change in Mongolia’s Rangelands Punsalmaa Batima, Bat Bold, Tserendash Sainkhuu and Myagmarjav Bavuu
Introduction When climatic events adversely impact Mongolia’s grasslands and livestock herds, the effects reverberate throughout the country. The people and economy are highly dependent on pastoral livestock herding, a livelihood that is extremely sensitive to climate variations and extremes. Roughly 80 per cent of Mongolia’s 127 million hectares is used as open pasture for year-round grazing of sheep, goats, cattle (including yaks), horses and camels. Herding these animals and processing livestock products engages nearly half of the Mongolian population, provides food and fiber to the other half, and generates about one-third of the country’s foreign exchange earnings (Mongolian Statistical Yearbook, 2004). The arid to semiarid climate of Mongolia supports extensive grasslands that, while fragile, have sustained pastoral herding for centuries. In recent decades the climate has become warmer and drier. Partly due to this trend, the productivity of Mongolia’s pastures has declined by 20 to 30 per cent. Another observed trend is an increase in the frequency and intensity of climatic extremes such as drought and severe winters, or zud. Drought and zud events have caused livestock deaths, hardship for herders and, in some instances, large rural to urban migrations, unemployment, deep poverty and economywide losses of income. The pastoral system of Mongolia has also been strongly impacted by the change from a socialist system to a market economy. The change is not complete in the livestock sector as pastureland remains state owned and, with both traditional and socialist systems for livestock management disintegrated in the wake of the transition, is treated as an open access commons. The changes have increased the vulnerability of the pastoral system to climate and other stresses and led to practices that have placed heavy grazing pressures on many pasture areas. These pressures, combined with the trend toward a drier climate, have contributed to the degradation of some of Mongolia’s pasturelands.
3232 J&J Climate Change
1/11/07
9:56 am
Page 197
Adapting to Drought, Zud and Climate Change in Mongolia’s Rangelands 197
Projections of future climate change indicate that Mongolia will become warmer still, potentially drier in summer and wetter in winter. There is also a threat of even more frequent and intense droughts and zud in the future. Our study of projected climate change finds that the rangelands, livestock herds and pastoral livelihoods of Mongolia would be strongly impacted (Batima, 2006; Batima et al, 2008). Finding strategies to adapt to the changes will be critical if pastoral livelihoods are to be sustained and adverse impacts on Mongolia’s economy and future development avoided. We briefly review below the effects of social and economic changes on the livestock sector and the vulnerability of the sector to climate extremes and change. We then describe the stakeholder process used to identify and evaluate options for adapting to current and future climatic risks, present some of the options that emerged as high priorities, and explore barriers, opportunities and responsibilities for adaptation.
Social and Economic Changes Up until the 1960s, Mongolia’s pastures and livestock herds were managed according to traditional pastoral practices that included four seasonal migrations each year, maintenance of emergency pasture reserves, and grazing schedules that consider vegetation growth phases and the recovery periods needed by previously grazed pasture. This system provided for efficient and sustained use of pasture in the arid and semiarid climates of Mongolia’s grasslands, limitation of grazing pressures, and management of risks from drought and zud. In the 1960s, the livestock industry was collectivized. The land, animals and tools of production became the property of collectives called negdels that managed livestock production. Inputs and services such as fodder, transport, veterinary services, marketing, maintenance of wells, improved breeds and fencing were subsidized and provided by the central government through the negdels. With these subsidies, herders became highly dependent on the state. Seasonal migration of herds was still practised but it become more restricted with centralized management and provision of services. In 1990–1992, with the transformation of the country from a socialist system to a market economy, the collectives were disbanded and ownership of livestock was privatized. Land, however, was not privatized and remains the property of the state. State-subsidized inputs and services to the livestock sector were cut sharply and infrastructure such as water supply, electricity, schools and hospitals collapsed in the rural areas. The state-owned land effectively became an open access commons in which herders lacked incentives to conserve pastures or make investments in the land such as maintenance of water supplies. Herders responded by increasing the numbers of animals stocked on pastures. Many reduced their seasonal migrations to twice a year, migrated shorter distances because of lack of infrastructure and services near distant pastures, moved their herds closer to urban centres with services that were increasingly unavailable in rural areas, or even stayed in one place to graze
3232 J&J Climate Change
1/11/07
9:56 am
Page 198
198 Climate Change and Adaptation
the same pasture year-round (Chuluun and Enkh-Amgalan, 2003; Batima et al, 2005). The higher animal stocking rates and more limited movement of herds raised grazing pressures on many pastures and disrupted ecological balances. This has contributed to land degradation and reduced pasture productivity. Some areas have been abandoned because of lack of water supply or overuse (Tserendash et al, 2005). In the current situation, traditional and socialist systems for managing livestock activities are no longer operative, while incentives in the market system are distorted and imperfect with privatized livestock but open access land. The resulting practices and behaviours have heightened vulnerability in the livestock sector to climate and other stresses.
Vulnerability to Climate Extremes and Change Changes in the climate of Mongolia observed in recent decades have impacted the ecosystems and livestock sector of the country (Batima, 2006; Batima et al, 2005; Tserendash et al, 2005). From 1940 to 2003, annual mean temperature rose by 1.8°C, precipitation increased slightly in autumn and winter and decreased slightly in spring and summer, potential evapotranspiration rose by 7 to 12 per cent, the duration of ice cover on lakes and rivers decreased by 10 to 30 days, maximum ice thickness decreased by 40 to 100cm, and the date of snow cover has shifted earlier by 3 to 10 days (this last since 1981). The warming has been most pronounced in winter, during which average temperatures have risen by 3.6°C. Peak above ground biomass on Mongolia’s pastures, which is usually attained in August, has declined by 20 to 30 per cent over the past 40 years. In the forest-steppe and steppe zones, the emergence of pasture plants comes earlier but above ground biomass has decreased by 10 to 20 per cent in April and by 30 per cent in May. The warmer and drier spring and early summer have contributed to the decline in biomass. Also observed is a shift in the species composition of pastures towards low nutrient plants such as Carex duriuscula Artemisia sp, which has become dominant in pasture communities. These changes have reduced forage production for livestock animals. Livestock grazing is also directly impacted by rising summer temperatures. Mongolia’s livestock animals are accustomed to relatively low temperatures. When mid-day summer temperatures exceed thresholds, animals cease grazing. One impact of the warming of past decades is a reduction in the time spent grazing by livestock. Over the past 20 years, the time spent grazing during the months of June and July has reduced by an average of 0.8 hours per day and the number of days with more than 3 hours less grazing time increased by 7 days. The reduced pasture productivity and grazing time show their effects in reduced animal weights observed since 1980 (Bayarbaatar et al, 2005). Changes in extremes have also been observed. Drought has increased significantly in the last 60 years. The worst period was 1999–2002, during which droughts affected 50 to 70 per cent of the land area for four consecutive summers. In normal years, animals start to gain weight in early summer, attain-
3232 J&J Climate Change
1/11/07
9:56 am
Page 199
Adapting to Drought, Zud and Climate Change in Mongolia’s Rangelands 199
ing their maximum weight by the end of autumn and having built up the capacity to withstand winter and spring weather. But in years of summer drought, reduced forage limits animal weight gain and leaves them more vulnerable to the stresses of winter and spring. Mongolian winters are harsh and livestock losses occur every winter. But unusually severe winters, called zud, can cause conditions of livestock famine that result in abnormally high rates of animal deaths. There are several forms of zud. Tsagaan, or white zud, the most common and disastrous form if it affects a large area, results from high snowfall that covers the grass and prevents animals from grazing. Har, or black zud, occurs when a lack of snow causes water supply shortages. An impenetrable ice cover that forms when snowmelt refreezes and prevents access to grass is called Tomor, or iron zud. Extreme cold temperatures and strong winds that prevent animals from grazing and cause them to expend most of their energy to maintain body heat is called Huiten, or cold zud. Havarsanzud results when two or more of the above phenomena occur together. Very severe zud in the winters of 1944–45, 1967–68, 1978–79 and 1999–2002 each resulted in millions of animal deaths. The harsh winters during the 1999–2002 period, which also coincided with severe and extensive summer droughts, were particularly devastating. An estimated 12 million animals died, representing almost one-third of Mongolia’s livestock. More than 12,000 families lost all their animals, while thousands more lost substantial portions of their herds and were forced into poverty. Many migrated from rural areas to urban centres, raising the urban population in Mongolia from just under 50 to over 57 per cent and increasing unemployment in the urban areas. Agricultural output declined by 40 per cent in 2003 relative to 1999 and the contribution of agriculture to national gross domestic product decreased from 38 per cent to 20 per cent (Mongolian Statistical Yearbook, 2004). Projections of future climate from five different general circulation models, interpolated to 2.52.5 degree grid cells in Mongolia, indicate that the climate will become warmer, in both summer and winter, and that precipitation is likely to increase in winter but change only slightly in summer (Natsagdorj et al, 2005; Batima, 2006). Winter temperatures are projected to warm by 2.8 to 8.7°C by the 2080s relative to 1961–1990, while summer temperatures are projected to warm by 3.2 to 6.9°C. Projected changes in precipitation range from +6.9 to +67.0 per cent in winter and –2.3 to +10.9 per cent in summer by 2080. The relatively small precipitation changes projected for summer imply that, with higher temperatures, evapotranspiration will increase and soils will become drier in summer. It is also estimated that drought incidence will increase (Natsagdorj and Sanjid, 2005). Estimates of net primary productivity for scenarios of future climate change project decreases in pasture biomass in the forest-steppe and steppe zones of up to 37 and 20 per cent respectively by 2080 and increases in the high mountains and desert steppe of up to 28 and 52 per cent respectively (Tserendash et al, 2005; Batima, 2006). Carbon:nitrogen ratios are projected to decrease for all zones, indicating a decrease in nutritional quality of forage. Weight gain of livestock would be negatively affected in all but one
3232 J&J Climate Change
1/11/07
9:56 am
Page 200
200 Climate Change and Adaptation
zone. For example, weight gain by ewes would decrease by up to 58 per cent in the forest-steppe, 39 per cent in the steppe and 9 per cent in the high mountains, while the positive change in the dessert steppe would be less than 1 per cent. With warming, winter temperatures will be milder in the future, which may bring some benefits to herders. The effect of climate change on the occurrence of zud, however, is not clear. With greater winter precipitation, more snow will fall, which can cause white zud events. Short periods of high temperatures followed by a return to sub-freezing temperatures can cause iron zud. Wind storms in the coldest months of December and January, which were rare in the past but have become more frequent in recent years, can bring cold zud. Despite the warming observed in past decades, the frequency of zud that covered more than 25 per cent of the land rose between the 1960s and 1990–2000 (Batima, 2006). Thus it cannot be assumed that zud will become less common with climate change, and increases in zud frequency are possible. As demonstrated by past events, the livestock sector and Mongolian society as a whole are highly vulnerable to zud. Continued increase in their frequency, particularly if in combination with summer droughts, would have detrimental impacts on the sector and on development in Mongolia. The impacts of climate variability and climate change on the pastoral livestock system of Mongolia are thus both near term and long term. In the near term, extreme events such as droughts and zud hold greater significance. Over the longer term, climate change will bring changes in temperatures, precipitation, snowfall and the duration of snow cover, as well as changes in the frequency of droughts and zud. These long-term trends in combination with socioeconomic changes in the country could potentially degrade pasturelands and significantly impact on human livelihoods.
Identifying and Screening Adaptation Options Recognizing the near- and long-term risks due to climate variability and change, adaptation planning should take into account both the need to increase the current ability of pastoral communities to lessen and cope with the impacts of extremes as well as the need to conserve and improve the resilience of pastureland. The near-term productivity and longer-term resilience of the pastoral system of Mongolia depends on three primary components of the system: the stock and condition of natural resources, primarily pasture, which are strongly affected by climatic conditions; animals’ biocapacity to cope with environmental stresses; and the human element that manages and depends on livestock and pasture lands. Our investigation of adaptation strategies, performed with the participation of stakeholders in Mongolia’s pastoral system, focuses on measures to bolster these three components. The identification and evaluation of options for adapting the Mongolian pastoral livestock system to climate change followed a two-step process. In the first step, a team of technical experts identified a large number of adaptation options, screened the options against a small number of broad criteria, and
3232 J&J Climate Change
1/11/07
9:56 am
Page 201
Adapting to Drought, Zud and Climate Change in Mongolia’s Rangelands 201
selected a subset for further evaluation with stakeholders. In the second step, workshops and consultations were held with three different groups of stakeholders. Technical experts from our case study team prepared a preliminary list of 89 adaptation options. The options were drawn first from responses to a household survey that was carried out during 2002–2004. The survey involved more than 700 herders’ households in 16 of Mongolia’s 21 provinces, called aimags, eliciting information about their perceptions of major risks to their livelihoods and strategies used to cope with problems caused by climatic phenomena. Additional options were identified from the findings and recommendations of previous studies of climate change in Mongolia (see, for example, Batima et al, 2000) and expert judgements of the team. The preliminary list of options was then screened to select options that warrant further consideration. A number of factors were judged to be important in the context of Mongolia’s livestock sector for screening options. As in many developing countries, the people and government of Mongolia give priority to immediate and pressing domestic problems such as economic development, poverty, public health, education and environmental degradation. Consequently, emphasis is given to adaptation measures that are consistent with and therefore might be more easily integrated into existing policies, plans and programmes in these areas. More specifically, the preliminary list of options was screened to identify those that satisfy three criteria: (1) Would the option advance climate change adaptation as well as existing objectives for development, poverty reduction, public health and education? (2) Is the option consistent with government policy, plans and programmes for agriculture? and (3) Would the option cause any adverse impacts on the environment? Options that were judged to satisfy at least two of these criteria were passed for further evaluation. This shortened the list of adaptations to be considered from 89 to 56. The shortened list of adaptation options that passed the initial screening was evaluated by three different levels or groups of stakeholders in a series of workshops and consultations. These included workshops with local community stakeholders, with scientists, and with policy and decision makers from national ministries. Participants in the workshops applied six additional criteria to evaluate the potential of each option: 1 2 3 4
Current adaptive capacity: What is the existing capacity to implement the option successfully? Importance of climate as driver of outcomes: How important is climate relative to other exogenous factors as a driver of the risk that is targeted by the adaptation option? Near-term effectiveness: How effective is the option expected to be for reducing negative near-term impacts of drought, zud and other extremes that are important sources of current climatic risks? Long-term benefits: Will the option produce long-term benefits for reducing vulnerability to climate change by, for example, improving the condition and sustainable use of pastureland?
3232 J&J Climate Change
1/11/07
9:56 am
Page 202
202 Climate Change and Adaptation
5 6
Cost: What are the expected investment, operation and maintenance costs of the option? Barriers: Are there significant technical, social, financial or institutional obstacles that could impede the implementation or performance of the option?
The workshop participants qualitatively rated the adaptation options for each of the six criteria, giving rankings of high (H), medium (M) and low (L). The aggregated results are shown in Table 11.1 for those adaptation measures that emerged as high priorities. The participation of local stakeholders, scientists, and policy and decision makers in the evaluation of adaptation measures is described below.
Workshops with local stakeholders Much of the actual implementation of adaptation measures will be carried out at the level of the household or community and key factors for consideration include (1) local community and herders as the primary beneficiaries of successful adaptation, (2) the substantial indigenous knowledge and experience regarding pasture and animal management accumulated within local communities, and (3) the devolution of responsibility for the management of the livestock sector to local authorities, including the governors’ offices of the aimags, or provinces, and sums, or subregions of the aimags. The incorporation of the perspectives and priorities of local actors in the implementation of climate change adaptation is considered critical. Four local community workshops were held, one each in the eco-regions of the Gobi-steppe, steppe, high-mountain and forest-steppe. More than 200 participants attended the workshops, including local governors, herders, environmentalists, climatologists and animal experts such as veterinarians. Almost all of the adaptation options prepared by the expert team were accepted as feasible by the local stakeholders, including the herders. An important exception are measures that would promote private ownership of pasture and water supply, which were rejected by more than 98 per cent of participants. Seasonal migration to access pasture, water and shelter are considered essential for optimal and sustainable management of Mongolia’s pastoral system. Private land ownership, in the opinion of herders and other community stakeholders, is not conducive to maintaining the traditional or even current pastoral livestock system of Mongolia. Almost all the participants expressed the view that individualized, private ownership of pastureland, under Mongolian conditions, is likely to increase conflict and jeopardize environmental stability. But despite the opposition to private ownership of land generally, herders expressed openness to the possibility of private ownership limited to small pastures for emergency use near winter and spring camps. Some of the major constraints to the implementation of adaptation measures identified in the local workshops include financial and material shortage, inadequate information, and remoteness from markets. A shortage of educated people in the rural areas to take care of rural education, grazing management,
3232 J&J Climate Change
1/11/07
9:56 am
Page 203
Adapting to Drought, Zud and Climate Change in Mongolia’s Rangelands 203
Table 11.1 Evaluation of adaptation options Evaluation criteria Importance of climate drivers
Near-term effectiveness
Long-term benefits
Cost
Barriers
Conserve natural resources
M L L L M L M
H M M H M M M
H H M H M M M
H H H H M M M
L H H H L L H
S& I F T&F T&F S& I S S
Strengthen animal biocapacity
Improve shelter for animals Increased supplementary feed Improve per animal productivity Introduce genetic engineering Improve veterinary services Introduce high productive cross breeds
M M L L M L
H H H H H H
H H H H H H
M H H H H H
M M H M H H
F R&F T&F T&F T&F T&F
Promote collective communities Develop/transfer new technologies Expand access to credit and generate alternative income Expand the supply of renewable energy applications to herders Promote and support the establishment of different kinds of enterprises Establish insurance system for animals Establish risk fund Prepare educated herders Training of young herders
M M L
H H H
M H M
H H H
M H H
S&I T&F I&F
M
H
H
H
M
T&F
L
M
M
H
H
T&F
L
M
M
M
H
T&F
L L H
H M H
M M M
H H H
H M M
F&I F F
Expand dairy and meat farms close to big cities to meet the demand of milk and other dairy products Promote and expand other food supply farms (e.g. eggs and vegetables)
M
H
H
H
H
T&F
M
H
H
H
H
T&F
Establish climate change monitoring stations Improve forecasting system of extreme events
H
H
H
H
H
T&F
H
H
H
H
H
T&F
Improve Increase food security understanding and supply
Improve grazing management Introduce cultivated pasture Improve pasture yield Improve pasture water supply Legislate possession of pasture Introduce taxation of pasture Livestock population control according to the pasture capacity
Enhance rural livelihoods
Current adaptive capacity
Objective Adaptation measures
Note: H-high, M-medium, L-low, F-financial; T-technical, S-social, I-institutional, R-resource
3232 J&J Climate Change
1/11/07
9:56 am
Page 204
204 Climate Change and Adaptation
and social and economic issues was also raised as a serious concern. Participants of the local workshops emphasized the importance of education and training, the return of students after graduation to their home provinces, and financial and management support in the implementation of adaptation measures. Most important, the participants stressed the need for the integration of scientific and local indigenous knowledge to cope with climate variability and to move from study, assessment and discussion to the actual implementation of adaptation measures.
Consultations with scientists Successful implementation of many adaptation options requires the support of and interaction with the scientific community to advance knowledge and understanding, develop new technologies and communicate information. Results from the local workshops were presented and discussed with leading scientists in the fields of animal husbandry and pasture management. The discussions focused on the following issues: • What is the role of scientists in the implementation of adaptation measures? • What should be done to introduce new varieties of drought-resistant pasture plants and to develop cultivated or irrigated pasture? • What should be done to improve the productivity of animals? • How should know-how on mechanical and automatic equipment or appliances to facilitate the manual labour of herders be transferred? • What is the existing scientific, technical and financial capacity of the institutions to support implementation of adaptation measures? • What should be done to improve the capacity of institutions and how should the barriers, if any, be overcome? The discussions revealed serious concern among scientists about the problems of the livestock sector, awareness of the vulnerability of the sector to climatic stresses, and a willingness to cooperate in the development and implementation of adaptation measures. The scientists emphasized technologies for increasing livestock and pasture productivity, for example, with the development of new animal breeds and grass species through genetic engineering, and the use of experiments to test grassland adaptations for a warmer climate. Financial constraints and lack of collaboration among different institutions were identified as major obstacles for such work. The collaboration between local people and scientists established by our study team was acknowledged as a useful model and the creation of an institutional mechanism to continue this collaboration was suggested.
Consultations with policy and decision makers The shortened list of adaptation options that passed the initial screening were evaluated by three different groups of stakeholders in a series of workshops and consultations. These included workshops with local community stake-
3232 J&J Climate Change
1/11/07
9:56 am
Page 205
Adapting to Drought, Zud and Climate Change in Mongolia’s Rangelands 205
holders and consultations with scientists, and with policy and decision makers from national ministries. The focus of the meetings with policy makers was to draw their attention to the impacts of climate change and urge their action in the implementation of adaptation measures and their integration in national planning. Participants from the different ministries demonstrated good understanding of the current vulnerability of the livestock sector to climate extremes such as severe drought and zud, and also of the low technological advancement of the sector. Less well understood by the ministries were human caused climate change and the potential effects of climate change on the livestock sector. Knowledge of the design and implementation of adaptations to climate variability and climate change differs markedly across the ministries. Common to most is the perspective that adaptation measures can be synergistic with development policies in the livestock sector and that they should be implemented within that context. But the policy makers made no clear distinction between adaptation measures implemented at a project level and broader policy and institutional measures that could be enabling mechanisms for implementing adaptation across different sectors and different policy and decision-making contexts.
Priority Adaptation Measures The adaptation measures that were discussed and evaluated in the stakeholder workshops can be classified into five main groups: 1 2 3 4 5
conserve natural resources; strengthen animal biocapacity; enhance capacities and livelihood opportunities of rural communities; increase food security; and improve understanding and forecasting of climate variations and extremes.
Mongolia’s pastures are the key natural resource input to livestock production. Climate change threatens to reduce the production of forage grasses by this resource and may, in combination with heavy grazing pressures, degrade the land itself. Sustaining and improving Mongolia’s pastures will be critical for future livestock production. A variety of options to conserve and improve pastures and pasture yields were identified. Better management of grazing by limiting livestock numbers and returning to the traditional rotation of herds to different pasture each season would control grazing pressures, prevent degradation of pastures and help degraded pastures to recover. Degraded pastures can also be restored by reforestation and increasing vegetation cover. Drought-tolerant perennial species can be used to increase vegetation cover and increase pasture yields. Establishment of cultivated pasture would reduce dependency on nature and climate. A successful implementation of this measure would greatly reduce not only the expected impact of climate change but also the vulnerability to drought and harsh
3232 J&J Climate Change
1/11/07
9:56 am
Page 206
206 Climate Change and Adaptation
winters (zud). Expansion and rehabilitation of pasture water supply is also viewed as a promising measure for improved pasture utilization and stock survival, as well as ecosystem conservation and rural development. Biocapacity is important for reducing the vulnerability of animals to drought and winter zud. Options identified for strengthening animal biocapacity include modifying the daily grazing schedule, increasing use of supplemental feeds, and increasing feed and pasture reserves. When temperatures get very high in the middle of the day in summer, livestock cease to graze and this can have a large impact on weight gain. Modification of the grazing schedule to put animals to pasture in early morning or late evening can partly compensate for the reduced grazing in the middle of the day due to high summer temperatures. But to fully compensate for the simulated effects of warming would require extending grazing at times other than the middle of the day by an impractical 6 hours per day by 2020 and even longer by 2050 and 2080. Thus modification of the grazing schedule needs to be considered in combination with other measures. Supplemental feeds to increase daily feed intake could be given to livestock not only in winter but also in summer to fatten animals and increase their capacity to survive the winter. Simulation results show that 1.9–3.3kg/day of supplemental feed would be required per sheep in the summer of 2020 in order to compensate for the projected weight decrease. Feed reserves for emergency supply in case of drought or zud can be improved by increasing haymaking, sowing fodder crops, and increasing feed preparation and manufacture. Pasture can also be allocated as reserve not to be used except in the event of a harsh winter. An important option for enhancing rural livelihoods is to promote traditional pastoral systems such as khot ail and neg golynhon that would revive traditional pasture management. The traditional way of life of Mongolian herders is in itself an effective indigenous measure to cope with climate extremes. Herders typically lived in groups of two or more households called khot ail, which served a variety of functions, including informal regulation of pasture use by member households and seasonal movement of herds. The neg golynhon, which means ‘people from one river area,’ is also a similar informal social network. Such community supervision was effective in protecting pasturelands and mitigating the impacts of droughts and zud. Demonstrations of community-based adaptation, in which the community decides on how best to share the limited common resource based on its indigenous knowledge about the livestock and the environment, hold promise as a potential adaptation measure for climate change. A number of other options can complement collective organization of the pastoral system to enhance rural livelihoods. These include development and transfer of new technologies, education and training of herders, establishment of rural enterprises, access to credit for financing investments in land improvements, new technologies and enterprises, and establishment of insurance mechanisms for spreading risks. Extension services, education, training and other mechanisms are needed to enable herders to engage with experts and groups that can help find solutions and means to conserve natural pasture,
3232 J&J Climate Change
1/11/07
9:56 am
Page 207
Adapting to Drought, Zud and Climate Change in Mongolia’s Rangelands 207
improve livestock management, increase and improve the use of feed supplements, and increase the know-how of herders for managing and adapting to changing climate risks. Monitoring and research are needed to improve understanding of climate extremes and change and their consequences, and to develop forecasting and warning systems. Forecasts of drought and zud could be used to issue warnings and prompt actions by herders, local authorities and national ministries to prepare for and mitigate the impacts.
Barriers, Opportunities and Responsibilities The implementation of adaptation measures to reduce climate change vulnerability in Mongolia faces many barriers. The legal framework for access to land and water treats these resources as common property and does not provide an effective means for regulating and managing their use. Lacking secure tenure to land and water, herders do not have appropriate incentives for conserving, improving and investing in pasturelands and water systems. Many of the identified adaptation options are therefore unlikely to be widely used by herders under the current framework. Privatizing ownership of pastureland and water would correct this failure, but would give rise to other problems and is widely and strongly opposed by herders, with the possible exception of small pastures for emergency use. Revival of a traditional collective system for regulation of access to pasture and water is attractive to many stakeholders but will require changes in legal and institutional frameworks to be successful. The current institutional framework also poses barriers because responsibility for different sectors, resources, functions and programmes that are relevant to the livestock sector and the management of climate risks are divided among many different ministries and institutions. Successful implementation of a comprehensive adaptation strategy will require coordination and cooperation across these institutions. Such coordination does not occur at present. The participation of different ministries and institutions, as well as local stakeholders, in programmes such as AIACC and The Netherlands Climate Change Studies Assistance Program is helping to bring together the different stakeholders that are needed for successful adaptation. But more remains to be done. Lack of modern technologies for animal husbandry, pasture improvement, cultivation, irrigation, processing, storing, packaging and transporting animal products results in low productivity of the sector, low incomes and inability to compete in the global market. The limitations include lack of infrastructure, machinery, improved seeds and animal breeds, and know-how. Rectifying these limitations is a priority for development of the livestock sector and would also enhance the capacity of the sector to adapt to climate change. Advances are being made but they are slow in coming and are sometimes reversed by events such as the droughts and zud of 1999–2002. Many of the options have financial costs that are high relative to the financial means of most individual herders. Implementation at the level of the community, sum, aimag or national government allows for pooling of resources
3232 J&J Climate Change
1/11/07
9:56 am
Page 208
208 Climate Change and Adaptation
and may yield scale economies that reduce costs. But due to economic difficulties in Mongolia, financial resources are very limited at these levels, which would limit the implementation of adaptation measures. A range of possible actors is likely to be involved in the implementation of the identified measures. Long-term concerns with respect to sustainable use of pasture resources are generally the responsibility of the national government, as the pasture is state owned. Hence the implementation of adaptation measures should begin at the level of national planning organizations. Planning for adaptation at the national or regional level is also justified when impacts and interventions cross local boundaries and/or economic and financial implications are beyond the capacity of local communities. But most measures are to be implemented at the local level and by herders themselves. Thus success requires coordination between central and local levels of management to implement a combination of measures, allocate appropriate resources, provide support services, and create an institutional and legal framework that enables adaptation. Participation of national, provincial and local governments, scientists and herders is equally important in the implementation of any of the adaptation measures. Behavioural modifications with respect to pasture use and management, and livestock management among the herder communities could be a key factor in safeguarding the natural resource base. Implementation of most of the identified adaptation measures requires substantial investment. At the moment Mongolia faces many other socioeconomic problems and financial constraints. Therefore, it is important that, at the national planning level, the available funds are more clearly prioritized and allocated based on the objectives of the selected adaptation strategies. There are also many regional and global financing programmes related to the implementation of the United Nations Framework Convention on Climate Change (UNFCCC) goals. The Global Environment Facility (GEF) provides financial support to cover the incremental cost in developing countries and in countries with economies in transition of protecting and managing the global environment, including dealing with climate change impacts. Therefore, adaptation projects could be developed with GEF funding through its implementing agencies like the United Nations Development Programme, United Nations Environment Programme and the World Bank. Mongolia could also participate in regional, sub-regional and bilateral cooperation activities, and initiatives on climate change-related issues, so that it can gather more experience and knowledge on adaptation to its impacts.
Conclusions The Mongolian livestock sector engages almost half the population, is a major contributor to national income and export earnings, and is highly vulnerable to climate change impacts. Observed increases in the incidence and impacts of severe weather events like drought and zud, and the potential for further increases as a result of climate change, are a particular concern for Mongolia.
3232 J&J Climate Change
1/11/07
9:56 am
Page 209
Adapting to Drought, Zud and Climate Change in Mongolia’s Rangelands 209
Adaptation to reduce the impacts is vital for ensuring livelihood security and for promoting social and economic development. Herders, technical and scientific experts, and representatives from local, provincial and national authorities give priority to adaptation strategies that would generate near-term benefits by improving capabilities for managing the extremes of drought and zud, and long-term benefits by improving and sustaining pasture yields. Specific measures identified as advancing these broad goals and warranting further consideration include: 1 2 3
4 5
improving pastures by reviving the traditional system of seasonal movement of herds, restoring degraded pasture, expanding and rehabilitating water supply, and developing cultivated pasture; strengthening animal biocapacity by modifying grazing schedules, increasing use of supplemental feeds, and increasing feed and pasture reserves; enhancing rural livelihoods by promoting traditional pastoral networks and herders’ communities to regulate access and use of pasture and water, developing and transferring new technologies, educating and training herders, establishing rural enterprises, and providing access to credit and insurance; improving food security by improving and diversifying food production and distribution systems; and research and monitoring to develop and improve forecasting and warning systems.
Administrative decisions and actions are necessary to remove or ease barriers to the implementation of many of these adaptation measures. This includes revising the legal and institutional frameworks to provide incentives for conserving, improving and investing in pastureland and water systems, and promoting the organization of the herding population into local associations for managing resources and reviving practices of the traditional pastoral system. A sensible first step to initiate the process of adapting to longer-term climate change would be to facilitate existing adaptation strategies used by the herders to deal with climate variability and extreme events.
References Batima, P. (2006) ‘Climate change vulnerability and adaptation in the livestock sector of Mongolia’, final report, Project AS06, Assessments of Impacts and Adaptations to Climate Change. International START Secretariat, Washington, DC, www.aiaccproject.org Batima, P., B. Bolortsetseg, R. Mijiddorj, D. Tumerbaatar and V. Ulziisaihan (2000) ‘Impact on natural resources base’, in Climate Change and Its Impacts in Mongolia, NAMHEM and JEMR publishing, Ulaanbaatar Batima, P., L. Natsagdorj, P. Gombluudev and B. Erdenetsetseg (2005) ‘Observed climate change in Mongolia’, AIACC Working Paper No 12, International START Secretariat, Washington, DC, www.aiaccproject.org Batima, P., B. Bat, S. Tserendash, D. Tserendorj, S. Shiirev-Adya, N. Togtokh, L.
3232 J&J Climate Change
1/11/07
9:56 am
Page 210
210 Climate Change and Adaptation Natsagdorj and T. Chuluun (2005) ‘Adaptation to climate change’, in P. Batima and D. Tserendorj (eds) Climate Change Impacts on Extremes – Vulnerability and Adaptation Assessment for Grassland Ecosystem and Livestock Sector in Mongolia, ADMON, Ulaanbaatar (in Mongolian) Batima, P., L. Natsagdorj, N. Batnasan and M. Erdenetuya (2008) ‘Vulnerability of Mongolia’s pastoralists to climate extremes and changes’, in N. Leary, C. Conde, J. Kulkarni, A. Nyong and J. Pulhin (eds) Climate Change and Vulnerability, Earthscan, London Bayarbaatar, L., G. Tvuaansuren and D. Tserendorj (2005) ‘Climate change and livestock’, in P. Batima and B. Bayasgalan (eds) Climate Change Impacts on Extremes – Vulnerability and Adaptation Assessment for Grassland Ecosystem and Livestock Sector in Mongolia: Impacts of Climate Change, ADMON, Ulaanbaatar (in Mongolian) Chuluun, T. and A. Enkh-Amgalan (2003) ‘Tragedy of commons during transition to market economy and alternative future for the Mongolian rangelands’, African Journal of Range and Forage Science, vol 20, no 2, p115 Mongolian Statistical Yearbook (2004) Mongolian Statistical Yearbook 2003, National Statistical Office, Ulaanbaatar, Mongolia Natsagdorj, L. and G. Sanjid (2005) ‘Climate change and drought’, in P. Batima (ed) Climate Change Impacts on Extremes – Vulnerability and Adaptation Assessment for Grassland Ecosystem and Livestock Sector in Mongolia: Livestock Sector Vulnerability to Climate Change, ADMON, Ulaanbaatar (in Mongolian) Natsagdorj, L., P. Gomboluudev and P. Batima (2005) ‘Future climate change’, in P. Batima and B. Myagmarjay (eds) Climate Change Impacts on Extremes – Vulnerability and Adaptation Assessment for Grassland Ecosystem and Livestock Sector in Mongolia: Current and Future Climate Change, ADMON, Ulaanbaatar (in Mongolian) Tserendash, S., B. Bolortsetseg, P. Batima, G. Sanjid, N. Erdenetuya, T. Ganbaatar and N. Manibasar (2005) ‘Climate change and pasture’, in P. Batima and B. Bayasgalan (eds) Climate Change Impacts on Extremes – Vulnerability and Adaptation Assessment for Grassland Ecosystem and Livestock Sector in Mongolia: Impacts of Climate Change, ADMON, Ulaanbaatar (in Mongolian)
3232 J&J Climate Change
1/11/07
9:56 am
Page 211
12
Evaluation of Adaptation Options for the Heihe River Basin of China Yongyuan Yin, Zhongming Xu and Aihua Long
Introduction The Heihe basin, located in northwestern China, is a region of predominantly arid and semiarid climate with extremely fragile ecological systems, few financial resources, poor infrastructure, low levels of education, and restricted access to technology and markets. The region suffers from climate variations and may experience more severe impacts of climate change on water resources, food production and ecosystem health in the future. Moreover, the region’s adaptive capacity is much less than that in the coastal region of China. People in the Heihe region are facing substantial and multiple stresses, including rapidly growing demands for food and water, poverty, land degradation and other issues that may be amplified by climate change. In the Heihe region, various water use policies and measures have been implemented or designed to limit or prohibit the utilization of water by different sectors or regions. Controversies have occurred, of course, as these policies are redistributive in nature, making some sectors or regions worse off and others better off. It is this redistributive nature of policies that often aggravates water use conflicts. Measures for improving the adaptation of the water system to climate variability were introduced recently to the region, including water supply control, water permits, water right certificates, farmer water use associations, water pricing adjustment, and a better water allocation policy. What seems to be missing, however, is an overarching strategy that brings the climate change concern into water use decision-making process. For the most part, the impact of climate change on the water system has received scant attention from government agencies and others responsible for water resource management and planning. A partial explanation for the limited response to take consideration of climate change in water use management might be the lack of knowledge or awareness of the issue by policy makers and the general public. Because of the limited knowledge and awareness, we undertook a study of the Heihe basin to improve understanding of the potential effects of climate
3232 J&J Climate Change
1/11/07
9:56 am
Page 212
212 Climate Change and Adaptation
change on water resources, the vulnerability of water users and options for adapting to the changes. Findings from the study about vulnerability to climate change are described in Yin et al (2008). Here we focus on our examination of adaptation, which applies multi-criteria evaluation methods. Our purpose is to provide decision makers with the information needed to improve the capacity of the water resource system for coping with and adapting to climate change in the Heihe region.
Current Status of Adaptation Science and Evaluation Tools Research on developing well-designed adaptation strategies and options can provide the information and understanding necessary for establishing efficient adaptation options or policies to deal with climate vulnerability. In general, there are two broad approaches used for adaptation assessment. The first, developed by the IPCC (Carther et al, 1994) and applied within the context of climate change impact assessment, evaluates the effects of adaptation to lessen the impacts of climate change. The second is a policy analysis approach that seeks to understand the processes by which adaptation occurs, the multiple objectives of adaptation, the factors and conditions that enable or impede adaptation, the resources needed, and the consequences. Policy analysis of climate change adaptation aims to support decisions to improve the adaptability, resilience and sustainability of various systems with respect to climate change. A wide range of tools has been used in the assessment of adaptation across and within different natural resource and socioeconomic sectors. The United Nations Framework Convention on Climate Change (UNFCCC) has collected and compiled information about methods and tools for evaluating climate change adaptation from member governments and other organizations (see Stratus Consulting, 1999). Their compendium provides basic information about adaptation evaluation tools, as well as related and complementary tools for constructing climate change scenarios and assessing climate change impacts and vulnerability. While the compendium is useful as a reference document, it is not a manual that describes how to implement each tool. It is rather a survey of possible tools that can be applied to a broad spectrum of situations. We review below the main groups of tools included in the UNFCCC compendium that are relevant for policy analysis of adaptation. The compendium includes relatively few examples of such tools, indicating that there is a need for new research approaches and tools that can be used specifically in adaptation option evaluation, selection and decision making. Adaptation tools need to be able to evaluate alternative options or policies, a capacity, which many impact assessment methods lack. Appropriate decision tools exist, having been developed originally for policy evaluation in contexts other than climate change decision making. There are numerous applications of these tools in disciplines such as decision theory, management science, resource management and systems engineering. Introducing these tools into climate adaptation study can aid in developing more effective climate change adaptation strategies or policies.
3232 J&J Climate Change
1/11/07
9:56 am
Page 213
Evaluation of Adaptation Options for the Heihe River Basin of China 213
Adaptation option evaluation tools in relation to impact assessment As noted previously, the science of adaptation usually applies two approaches in evaluation of adaptations: impact assessment and policy analysis. Carter et al (1994) provided guidelines for climate change impact assessment in which impacts are estimated for cases in which agents are assumed to either adapt or not adapt. A comparison of cases with and without adaptation gives a measure of the potential effectiveness of the adaptation options considered, usually short-term or autonomous measures. A variety of assessment methods are presented by Carter et al (1994), but they are rather general and the guidelines do not offer specific recommendations or prescriptions. Two concurrent research programmes, the US Country Studies Program and a parallel programme of the United Nations Environment Programme (UNEP), were developed to fill the gap. The former programme assisted more than 50 developing and transition economy countries to develop capacity for assessing vulnerability to climate change and evaluating adaptation options; it produced a guidebook of specific methods and approaches for impact assessment and adaptation evaluation (Benioff et al, 1996). The UNEP programme also sponsored the writing of a handbook on assessing climate change impacts and adaptation, to serve as a supporting resource for its own country studies programme. The handbook presents an overview of different methodologies and covers several sectors (Feenstra et al, 1998). While these guidebooks and the approaches that they recommend have provided a useful framework for climate change research and assessment, they focus mainly on the impacts of climate change and give relatively little attention to the evaluation of adaptation options. It has been a common experience in applying these research approaches in climate change impact studies at country, regional and sectoral levels that the overwhelming part of the time, effort and resources were devoted to the selection and application of climate scenarios and impact assessments. It has been invariably noted that insufficient time and effort were left to develop the adaptation component. There are some shortcomings associated with these scenario-driven approaches from the point of view of the need to improve our understanding and evaluation of adaptation in a policy analysis context. The problem is more fundamental than simply a matter of available project time and financial resources. There are many important reasons why a range of applications of the scenario-based approaches have not yielded useful results for the purposes of adaptation option evaluation and policy analysis (Lim et al, 2005).
The adaptation evaluation tools for policy analysis The UNFCCC compendium (Stratus Consulting Inc, 1999) introduced a range of general decision tools that are applicable in evaluating adaptation policies in multiple sectors. The compendium groups these tools into three broad categories: initial survey, economic analysis and general modelling. The initial survey tools include expert judgement, screening of adaptation options and the
3232 J&J Climate Change
1/11/07
9:56 am
Page 214
214 Climate Change and Adaptation
adaptation decision matrix, which are useful for identifying potential adaptation strategies or narrowing down the list of appropriate options. They are relatively informal, inexpensive, and utilize qualitative judgement rather than quantitative data. General tools of economic analysis that can be applied to policy evaluation of climate change adaptation include financial analysis, cost–benefit analysis, cost-effectiveness analysis and risk–benefit/uncertainty analysis. These are typically used to determine which options are most economically efficient and to assist the user in deciding which adaptation option is the most appropriate once a final list of adaptation options has been compiled. General modelling tools include TEAM and CC:TRAIN. These address different adaptation strategies across a number of sectors and are used to evaluate several sectors of concern in a particular region. More detailed information on these methods is available from Stratus Consulting Inc (1999). To deal with the multi-criteria and multi-stakeholder nature of the adaptation evaluation process, multi-criteria evaluation technologies or tools can be adopted as effective evaluation instruments by which alternative adaptation policies can be compared and evaluated in an orderly and systematic manner. Given a set of adaptation policies available to deal with climate vulnerabilities or impacts on biophysical and socioeconomic aspects of our society, the evaluation tools can be used to help identify the policies that best satisfy selected decision criteria. A range of methods/tools developed in decision theory, multicriteria evaluation and systems analysis can be adopted for adaptation option evaluation (Zadeh, 1965; Holling, 1978; Yin and Xu, 1991; Yin et al, 2000; Yin, 2001a).
An Integrated Assessment of Adaptation Evaluation in the Heihe Basin The IPCC (2001) suggested a list of high priorities for narrowing gaps in vulnerability and adaptation research. Among these is integrating scientific information on impacts, vulnerability and adaptation in decision-making processes, risk management and sustainable development initiatives. In this section, an integrated approach is introduced as an example for discussion. The approach which was applied in the Heihe river basin case study, bringing together partners from the private sector, the public sector policy community and the academic research community, demonstrated how to meet the challenge of linking climate change adaptation and sustainable development. The analytic hierarchical process (AHP), a method of multi-criteria analysis, was applied in the case study to identify desirable adaptation options in dealing with climate change vulnerabilities. Figure 12.1 illustrates the integrated assessment framework for adaptation option evaluation, linking impact assessment with local sustainability indicators, and using tools such as multi-criteria policy analysis and multi-stakeholder consultation in the Heihe region. In the following discussion, not all the components shown in Figure 12.1 are covered in the same
3232 J&J Climate Change
1/11/07
9:56 am
Page 215
Evaluation of Adaptation Options for the Heihe River Basin of China 215
detail. Rather, focus is on the two main concerns of this chapter: sustainability indicators/goals identification and adaptation policy evaluation. The case study presents the importance of indicator/goal setting in regional sustainability research and the approach to identify indicators/goals. The adaptation option evaluation system shown in Figure 12.1 represents a participatory approach to integrated assessment. Working in partnership with multiple stakeholders, alternative adaptation measures and water sustainability indicators are selected. Based on water vulnerability information provided by researchers of the project (Yin et al, 2008), the evaluation system identifies practical adaptation options to deal effectively with water vulnerabilities likely to become more severe in the Heihe river basin due to the impacts of climate change.
Figure 12.1 Framework for multi-criteria evaluation of climate change adaptation options
The first step in implementing the framework is to identify existing and possible adaptation options to deal with vulnerabilities of climate variation and change. This was done through multi-stakeholder consultations and using multi-criteria evaluation techniques. Numerous potential adaptation options are available for dealing with vulnerabilities to climate change. An initial screening process was conducted to reduce the number of options for further detailed evaluation. The multi-stakeholder consultation yielded a collective recommendation of adaptation options for further multi-criteria evaluation. To link climate change impact analysis, adaptation option evaluation and sustainability evaluation, water system sustainability indicators must be set and
3232 J&J Climate Change
1/11/07
9:56 am
Page 216
216 Climate Change and Adaptation
the performance of adaptation options must be measured in a manner that integrates social, environmental and economic parameters that may be influenced by climate. In this study, indicators are evaluation criteria or standards by which the efficiency of alternative adaptation options can be measured. There are several general frameworks that can be adopted for developing sustainability indicators. The first is a domain-based framework that groups indicators into three main dimensions of sustainability – economic, environmental and social. The three-dimensional nature of sustainability and the need to make trade-offs (for example, between economic growth and environmental quality) require maintaining these three components in a dynamic balance. Sustainability indicators thus should include economic, social and environmental information in an integrated manner. Another important framework is a goal-based indicator system. Goals usually reflect the major development concerns of a nation or a region. For example, some concerns represent national or regional objectives of economic viability, maintenance of the resource base, and minimizing the impacts of climate change on natural ecosystems. Each goal is composed of a number of attributes or indicators which are measurable in most cases using existing sources of information. Other types of indicator frameworks include sectorbased, issue-based, cause–effect and combination frameworks. No single indicator would be sufficient to determine sustainability or non-sustainability of a region or a system: a set of goals and/or indicators is required in sustainability evaluation. Indicators are not a new concept and have been used to measure the performance of regional development policies or plans, to identify growth trends, to monitor the social and economic conditions of regions or nations, to inform the general public, to define planning goals or objectives, to guide strategic development options, and to compare different regions. For example, gross domestic product (GDP), housing price indices, unemployment rate and stock indices are commonly used to measure the social or economic performance of a society or economy. While these indicators strongly influence decision making by governments, other policy makers and the general public, they have shortcomings when used for measuring sustainable development. Recently, research has been initiated into developing indicators of societal sustainability. Indicators may be conflicting in that the achievement of one target precludes the achievement of another. Possible trade-offs between indicators therefore need to be identified. Very often the trade-off relations are nonlinear, creating situations of dramatic changes in the attainment of certain indicator levels once a threshold has been surpassed. Other indicators, however, are complementary – in other words by increasing the attainment of one indictor target it is possible to increase the attainment of other indicators. It has been suggested, for example, that development and environment are complementary up to some level of resource use. Indicators are also considered compatible when the attainment of one does not compromise the attainment of others. Based on information gathered from stakeholders through householder surveys and consultation meetings, four indicators for evaluating adaptation
3232 J&J Climate Change
1/11/07
9:56 am
Page 217
Evaluation of Adaptation Options for the Heihe River Basin of China 217
options to reduce water system vulnerability were selected for our study of the Heihe basin (Table 12.1). The selected indicators corresponded to four broad water use sustainability goals. A multi-criteria options evaluation approach (MCOE) using these four indicators was implemented through stakeholder meetings and surveys to identify desirable adaptation measures by which decision makers can alleviate vulnerabilities and take advantage of positive impacts associated with climate change. The survey was mainly carried out in one-onone interviews and in small group/workshop settings with a wide range of experts and stakeholders. Table 12.1 Indicators used to evaluate adaptation options in the Heihe river basin Sustainability goal
Indicator
Maximize water use efficiency Maximize economic return to society Minimize harm to natural environment Minimize economic costs to society
Reduce per unit production water use Increase economic return per unit water Ecological and environmental benefit Costs of adaptation options
Workshops were held to present climate impacts and vulnerability information to policy makers and stakeholder representatives for their review and comments. In the evaluation process, alternative options are evaluated by relating their various impacts to a number of relevant indicators. The results of impacts generated in various impact assessments are used as references for ranking the performance of each adaptation option against each sustainability indicator. Climate impact assessments of this project are discussed in detail in the project final report (Yin, 2006). When multiple criteria are relevant to the evaluation of options, indicators associated with the different criteria must be combined if options are to be ranked. We use the analytical hierarchy process (AHP) method to combine our four indicators and rank adaptation options. The AHP method has been employed to evaluate alternative policies, allocate resources, conduct sensitivity analysis for resource-use planning, and select desirable project locations for both developed and developing countries (Saaty, 1980 and 1982). When applied to adaptation option evaluation, the AHP method requires decision makers to provide judgements on the relative importance of each option in relation to each criterion. In this exercise, a decision maker compares options two at a time (pairwise comparison). Then decision makers specify their judgements about the relative importance of each option in terms of its contribution to the achievement of the overall goal, in our case to alleviate the adverse consequence of climate change (Yin and Cohen, 1994; Yin, 2001b). The result of the AHP is a prioritized ranking indicating the overall preference for each of the adaptation options.
3232 J&J Climate Change
1/11/07
9:56 am
Page 218
218 Climate Change and Adaptation
Applications and Results The study area The Heihe river basin covers an area of approximately 128,000 square kilometres and is located in a region with latitude of 35.4–43.5°N and longitude of 96.45–102.8°E. Figure 12.2 shows the location of the study region. The study area is the second largest inland river basin in the arid region of north-western China. The basin includes parts of two provinces (Qinghai and Gansu) and the Inner Mongolia Autonomous Region. The region is composed of diverse ecosystems including mountain, oasis, forest, grassland and desert. The River Heihe flows from a headwater in the Qilian Mountain area to an alluvial plain with oasis agriculture, and then enters deserts in Inner Mongolia, representing the upper, middle and lower reaches of the basin.
Figure 12.2 Location of the Heihe river basin
There is an increasing concern about water use conflicts in the Heihe region. The limited water supplies have to provide a number of economic sectors and communities in different jurisdictions with a range of different and often conflicting functions to meet the various demands. While the demands for resources increase as populations and economies grow, the availability and the inherent functions of water resources are being reduced by water pollution, environmental degradation and climate change. Competition over access to water resources in the Heihe region has led to disputes, confrontation and in many cases violent clashes. The growing water use conflicts have posed a big challenge for government agencies to implement effective water allocation policies.
3232 J&J Climate Change
1/11/07
9:56 am
Page 219
Evaluation of Adaptation Options for the Heihe River Basin of China 219
Climate change trends in northwest China in the past 50 years have been investigated by analysing temperature and rainfall from 1951 to 2004 (Wang, 2005). The results of the analysis show a significant rise in daily mean temperature with linear trends in most areas ranging from +0.2ºC to +0.4ºC per decade. The Qilian Mountain glaciers are already undergoing rapid retreat, at a rate of about one metre per year. The region depends on the glaciers as important natural reservoirs for water supply. The water supply mainly comes from the spring melting of glaciers (Yin, 2006 and Yin et al, 2008). One important water vulnerability indicator is the water withdrawal ratio, defined as the ratio of average annual water withdrawal to water availability. The WMO (1997) suggests that water withdrawal ratios that exceed 20 and 40 per cent be considered as indicators of medium and high water stress respectively. In northern China, however, where water withdrawal ratios typically exceed these thresholds, the government suggests that a 60 per cent threshold is a more practical indicator of high water stress given the severe water shortage situation in the area. Table 12.2 lists water withdrawal ratios in the Heihe region under current climate conditions for the period 1990–2000. The current water withdrawal ratios in the region are extremely high (80–120 per cent), far exceeding the critical threshold levels set by both the WMO and the Chinese government. Conflicts over water use, including violent fighting for water, have been increasing in the basin over the past decade. The trend of this social indicator suggests that water shortage in the growing season is becoming more and more serious because of decreased water supply and increasing population and per capita water use (Yin, 2006). Table 12.2 Water availability, water withdrawals and water withdrawal ratio in the Heihe river region, 1991–2000 Year
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
Water availability (108m3)
35.9
34.3
35.5
356
34.0
34.6
34.6
34.3
34.7
34.8
Total water withdrawal (108m3)
29.0
27.4
35.4
28.8
29.6
35.8
28.0
41.4
35.5
32.3
80% 100%
81%
87% 103%
81% 120% 102%
93%
Water withdrawal 81% ratio
Water stress in the region might intensify in the future because of growing water withdrawals related to population and economic growth, and decreasing water availability related to climate change. The National Climate Center of the Institute of Atmospheric Physics has developed climate change projections for western China for the 21st century using outputs from eight coupled global atmospheric and oceanic circulation models. These projections were calculated using the NCC/IAP T63 (National Climate Center/Institute of Atmospheric
3232 J&J Climate Change
1/11/07
9:56 am
Page 220
220 Climate Change and Adaptation
Physics) model (Xu et al, 2003). Ding et al (2005) applied a regional climate model (Ncc/RegCM2) nested with a coupled GCM (NCCT63L16/T63L30) and Hadley Centre model (HadCM2) for climate change studies. Outputs of the Chinese regional-scale climate model were used to design scenarios of climate change for our study (Li and Ding, 2004). The climate change scenarios up to 2040 were combined with scenarios of socioeconomic change to calculate future water supply and demand for the study region. Results of projected water shortage under climate change and socioeconomic change show that water shortage in almost all the municipalities of the Heihe river basin will be worse than at present (Table 12.3).
Table 12.3 Water shortage/surplus in Heihe river basin under climate change to 2040 (108m3) Municipality Suzhou Jinta Jiayuguan Shandan Minle Sunan Ganzhou Linze Gaotai or county Year 2000
–0.20 –0.07
–0.36
–0.12 –0.08
0.00
–0.80 –0.62 –0.77
Year 2010
–0.67 –0.41
–0.23
–0.27 –0.61
–0.01
–2.39 –1.99 –1.10
Year 2020
–1.26 –0.76
–0.18
–0.09 –0.24
0.04
–1.32 –1.33 –0.61
Year 2030
–1.01 –0.62
–0.05
0.09
0.15
0.09
–0.01 –0.52 –0.01
Year 2040
–0.19 –0.12
0.16
0.19
0.39
0.12
Average
–0.67 –0.39
–0.13
–0.04 –0.08
0.05
0.80 –0.26
0.29
–0.74 –0.94 –0.44
Note: Negative numbers indicate water demand exceeds water supply; positive numbers indicate water supply exceeds demand.
Most policy makers and communities across the Heihe river basin have very limited knowledge about the current adverse effects and impacts associated with climate change in water resource management and planning. It is also uncertain whether the region’s water infrastructure and measures have the capacity to respond quickly and effectively to future climate change. Effects of climate change on water shortage may be so significant that a comprehensive adaptive action or strategy is required, involving the participation and coordination of national, provincial and local authorities, and other stakeholders engaged in water resource planning and management.
Water system adaptation options for evaluation Numerous potential measures or options are available to alleviate negative consequences of extreme climate events or climate change for the water system. In general, the different options can be grouped into two categories: engineering and non-engineering measures. The former involve construction works that attempt to supply more water resources to various users. These structural measures include reservoirs, irrigation systems and wells. Options in the latter category do not involve construction and include demand manage-
3232 J&J Climate Change
1/11/07
9:56 am
Page 221
Evaluation of Adaptation Options for the Heihe River Basin of China 221
ment, water use policy, pricing, water trade and permits, and other institutional and governance measures. In the Heihe region, both engineering and non-engineering measures have been implemented or designed to address the regions problems with water shortage and water stress. Controversies have occurred, of course, as a result of such policies. In 2001 the State Council of China enacted a compulsory water division regulation to limit water withdrawals by Zhangye City on the middle reach of the River Heihe, requiring the city to provide 0.95 billion m3 of water over a period of three years for the lower reach region in the Inner Mongolia Autonomous Region. The purpose of this government policy was to protect the extremely fragile ecological conditions in the lower reach, which had become a major source of sandstorm hazards affecting the Beijing region of China. While the policy helped improve the ecosystem condition in the lower reach, 40,000ha of cropland annually in Zhangye City could not obtain water for irrigation. As a result, farmers suffered considerable crop losses. Production from about 160,000ha of cropland was reduced due to water shortage, and some of the area produced no harvest at all. Farmers in the upper and middle reaches of the river argue that the policy’s reduction of water for irrigation has led to detrimental consequences in the agricultural sector, while others have indicated that the new policy has enabled the revival of dried lakes located in the downstream region. Obviously, water policies or regulations may make some sectors or regions worse off and others better off because of their re-distributive nature. It is this re-distributive nature of policies that often aggravates water use conflicts. There have also been a growing number of local and provincial initiatives and programmes to address various aspects of the water shortage problem. For example, farm water user associations were established to engage farmers in water use decision-making processes. Each association elects a chairperson based on a ‘one family-one vote’ basis. The association appoints an irrigation inspector to measure the volume of water inflow to the association, together with water managers from the local water resource bureau, and then to allocate water to each individual farm within the association. The municipal water resource bureau allocates water quotas to the farm water user associations. The chairperson organizes the ‘collective activities’ of constructing and maintaining on-farm irrigation systems and water tanks, protecting irrigation facilities, purchasing irrigation water and storing extra deliveries, catching return flow, and harvesting rainwater (Li, 2006). Meanwhile, the government has established policies for water recycling, pollution control and water-efficient technology to improve water use efficiencies in industries and farming. For example, the Ministry of Water Resources introduced the first pilot project to incorporate practices of water conservation in the Heihe River region communities in 2001. Some adaptation measures were also introduced to the region, including overall water supply control, water permits, water right certificates, water pricing adjustment and better water allocation policy. Effective water pricing mechanisms have been implemented to encourage water saving. Higher water prices reduce water demand and increase water
3232 J&J Climate Change
1/11/07
9:56 am
Page 222
222 Climate Change and Adaptation
supply. The Provincial Finance and Pricing Control Bureau determines the price per unit of water for each alternative water use, including irrigation, industry, domestic and hydropower. In Zhangye City, the water tariff per unit of water increases step-wise from one block to the next, depending on the amount of water used relative to a water quota. The municipal Water Resource Bureau sets a water quota to supply water users with their basic water needs at a price that is affordable. When water use exceeds the specified water use quota, water prices then shift to higher levels. The water price increases 50 per cent when water use exceeds 1–25 per cent of the quota and increases 100 per cent when water use exceeds 26–50 per cent of the quota. In addition, the water price is not constant in different seasons. The price increases during the dry season and decreases during peak rainy season. Preliminary effects of the pilot projects have shown some positive effects in dealing with the water shortage problem. The water price reform pilot projects in the region have provided water consumers with an incentive to reduce their water consumption (Liu et al, 2006). Numerous potential measures or options are available to alleviate negative consequences associated with climate change. Based on government documents and existing literature on water resource management, the project researchers prepared a list of existing and potential options (Zhangye Government, 1998; Jia et al, 2004; Liu et al, 2006; Ministry of Water Resources of China, 2006; China Water Saving Irrigation Net, 2007). A primary screening process was conducted by the research team to select a limited number of adaptation options for further evaluation using the multi-criteria evaluation process. Eight options were selected for evaluation: 1 2 3 4 5 6 7 8
reform the economic structure to promote activities that are less water consuming (for example, promote drought-tolerant crops in place of high water consuming crops); establish a water use permit and trading system; construct water works (for example, irrigation systems); establish water users’ associations to engage farmers in water use decision making; adopt advanced water use technologies (for example, water saving irrigation methods); government pricing of water to control water demand; involve multiple stakeholders in improving water allocation policies; and increase awareness and education about water conservation.
The above list of adaptation options was evaluated using the AHP method to produce a ranking of options that reflects stakeholders’ expectations of their effectiveness for reducing vulnerability as measured by four criteria: economic efficiency, environmental quality, equality and feasibility. The AHP application was facilitated by a series of workshops with participation of a broad range of stakeholders and policy makers from water sectors to identify sustainability indicator priorities, as well as a series of desirable adaptation policies. These policy workshops, which were carried out in Zhangye City, provided forums
3232 J&J Climate Change
1/11/07
9:56 am
Page 223
Evaluation of Adaptation Options for the Heihe River Basin of China 223
for the research team to present research findings to local officials, experts and policy makers, and for stakeholders to review the study results and provide comments on the findings and their relevance to government decision making in water resource management. During the policy workshops, participants completed evaluation forms for the set of eight adaptation options. Survey questions on the evaluation forms were designed according to the principles of AHP so that the responses could be input into a software program, Expert Choice 2000, for compilation and analysis. The software computes an overall score for each alternative option by distributing the importance of the indicators among the adaptation options, thereby dividing each indicator’s priority into proportions relative to the percentage rankings of alternative options. Using the four indicators listed in Table 12.1 and the above set of adaptation options, a decision hierarchy model was created. This decision hierarchy is quite simple because it includes a single overall goal, to reduce vulnerability, with two levels below it in the hierarchy: a set of criteria or indicators and a list of alternative adaptation options. Once the relative importance of individual criteria is determined, decision makers need only think about their preference for each alternative adaptation option in terms of achieving a single criterion. The survey was designed as a series of tables. Respondents were given a pair of indicators or a pair of options, and asked to compare them using a numerical sliding scale. The comparison scale ranged from 1 to 5, with 1 representing options that are equally effective (or indicators that are equally important) and 5 representing options where one is very strongly more effective than another (see Table 12.4). The purpose of providing these comparison tables is to facilitate the AHP pairwise comparison process. Respondents select the relative effectiveness numeral based on their preferences and having been given certain impact information about the adaptation options. In this exercise, a stakeholder compares two options, given in the far left and far right columns of the table, against each criterion at a time. In the example given in Table 12.4, the respondent considers the option ‘reform economic structure’ to be equally effective as ‘adopt advanced water use technologies’ (row 5: numeral 1). Comparing ‘reform economic structure’ with ‘water use permit and water trade system’ in row 3, the respondent’s cross indicates a view that the second option is strongly more effective.
Results of the AHP analysis Reform of the economic structure was ranked as the most desirable adaptation option for the Heihe region, with establishment of farm water users’ associations ranked fairly high as well (see Table 12.5). The moderate performance levels for the options ‘improve water allocation policies’, ‘establish water permits and trading’ and ‘increase awareness and education’ are probably due to the fact that these are relatively new measures in water resource management in the study region. The scores for the options ‘adopt advanced water use technologies’ and ‘implement water price system to control demand’ are near the bottom of the list for most participants, especially from an economic
3232 J&J Climate Change
1/11/07
9:56 am
Page 224
224 Climate Change and Adaptation
Table 12.4 Example of AHP comparison table Adaptation option
Relative effectiveness scale 5
4
3
2
1
2
Reform economic structure Reform economic structure
X
5 Water use permit and water trade system Construct water work
X
Reform economic structure
X
Government sets water price X
X
Establish farm water users’ society or committee Adopt advanced water use technologies
X
Reform economic structure Reform economic structure
4 X
Reform economic structure
Reform economic structure
3
Adaptation option
Improve water allocation policies Increase water saving awareness and education
Note: Relative effectiveness scale: 1 – equally effective; 2 – marginally more effective; 3 – moderately more effective; 4 – strongly more effective; 5 – very strongly more effective.
perspective, and are not considered to be desirable adaptation options. It appears that regional stakeholders consider the two options to be expensive alternatives for dealing with watershed management and farmers do not want to pay higher water prices. Constructing water works is judged to be the most inefficient option from an economic perspective and it is ranked at the bottom overall by regional respondents. The above information is useful for government decision making in selecting efficient options, especially when considering the goal of Heihe river basin sustainability. The results suggest that institutional options (reform economic
Table 12.5 Overall rank and score of adaptation options for the Heihe region Water resource adaptation option Reform economic structure Form farm water user society Improve water allocation policies Establish water permits and trade Increase awareness and education Apply water saving equipment and technology Implement water price system Construct water works
AHP result
Rank order
0.26 0.18 0.14 0.13 0.12 0.08 0.05 0.04
1 2 3 4 5 6 7 8
3232 J&J Climate Change
1/11/07
9:56 am
Page 225
Evaluation of Adaptation Options for the Heihe River Basin of China 225
structure, water user associations and water permits) are considered by a wide range of stakeholders to be effective and desirable with respect to the four evaluation indicators. Implementation of these water-use adaptation options in the Heihe river basin might be able to reduce water vulnerabilities associated with climate change in the context of regional sustainability. These adaptation options can be incorporated into a comprehensive sustainable development plan for the basin.
Conclusions Working in partnership with local, provincial and national governments, and other key stakeholders such as water use professionals, farmers and other organizations, the study has demonstrated the use of a multi-criteria evaluation approach to analyse adaptation options for the Heihe basin. The application has identified adaptation measures considered by stakeholders to be potentially effective for reducing vulnerability of water users by increasing economic efficiency of water use, improving environmental quality, promoting equality and reducing water costs. The highest ranked options, which emphasized institutional approaches, could become practical options to deal with water vulnerabilities likely to become more severe in the Heihe region due to the impacts of climate change as well as pressures from population and economic growth. Through public consultation activities, stakeholders’ understanding of adaptation options and their possible effects was greatly enhanced. The increased awareness among local officials will certainly increase the effectiveness of implementing alternative adaptation policies. A properly developed and implemented adaptation action plan consisting of various effective measures could have positive benefits to the well-being and productivity of all people living in the Heihe region. These effective adaptation measures could help reduce water system vulnerability and water use conflicts. Indirectly, a reduction in water system vulnerability will mitigate the impacts of climate change on agricultural systems and protect the livelihoods of farmers. Water system sustainability can also improve ecosystem health and reduce sandstorms, which have a global environmental impact. In addition, a successful adaptation action plan could become a useful model for communities across China and around the world.
References Benioff, R., S. Guill and J. Lee (eds) (1996) Vulnerability and Adaptation Assessments: An International Guidebook, Kluwer Academic Publishers, Dordrecht, The Netherlands Carter, T. R., M. L. Parry, H. Harasawa and S. Nishioka (eds) (1994) IPCC Technical Guidelines for Assessing Climate Change Impacts and Adaptations, Department of Geography, University College, London China Water Saving Irrigation Net (2007) China Irrigation and Drainage Development Center, www.jsgg.com.cn/Index/Index.asp?ClientScreen=1024
3232 J&J Climate Change
1/11/07
9:56 am
Page 226
226 Climate Change and Adaptation Ding,Y. H., Q. P. Li and W. J. Dong (2005) ‘A numerical simulation study of the impacts of vegetation changes on regional climate in China’, Acta Meteorological Sinica, vol 63, no 5, pp604–621 (in Chinese) Feenstra, J., I. Burton, J. Smith and R. Tol (eds) (1998) Handbook on Methods for Climate Change Impact Assessment and Adaptation Strategies, Version 2.0, United Nations Environment Programme, Nairobi, and Institute for Environmental Studies, Vrije Universiteit, Amsterdam Holling, C. S. (ed) (1978) Adaptive Environmental Assessment and Management, John Wiley, Chichester, UK IPCC (2001) ‘Summary for policymakers’, in J. McCarthy, O. Canziani, N. Leary, D. Dokken and K. White (eds) Climate Change 2001: Impacts, Adaptation, and Vulnerability, contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York Jia, S. F., S. F. Zhang, J. Xia and H. Yang (2004) ‘Effect of economic structure adjustment on water saving’, Journal of Hydraulic Engineering, vol 3, pp111–118 (in Chinese) Li, Q. P. and Y. H. Ding (2004) ‘Multi-year simulation of the East Asian monsoon and precipitation in China using a regional climate model and evaluation’, Acta Meteorologica Sinica, vol 62, no 2, pp140–153 (in Chinese) Li, Y. H. (2006) ‘Water saving irrigation in China’, Irrigation and Drainage, vol 55, pp327–336 Lim, B., E. Spanger-Siegfried, I. Burton, E. Malone and S. Huq (eds) (2005) Adaptation Policy Frameworks for Climate Change: Developing Strategies, Policies and Measures, Cambridge University Press, Cambridge, UK Liu, W., H. Huang, W. Zhang and M. Zhang (2006) Evaluation Report of Water Saving Society Pilot Projects, Research Center for Hydraulics Development, Nanjing, China (in Chinese) Ministry of Water Resources of China (2006) ‘Hydraulic works’, available from water information website, www.cws.net.cn/ Saaty, T. L. (1980) The Analytic Hierarchy Process, McGraw-Hall, New York Saaty, T. L. (1982) Decision Making for Leaders: The Analytical Hierarchy Process for Decisions in A Complex World, McGraw-Hall, New York Stratus Consulting Inc (1999) Compendium of Decision Tools to Evaluate Strategies for Adaptation to Climate Change Final Report, FCCC/SBSTA/2000/MISC.5, UNFCCC Secretariat, Bonn, Germany Wang, Z. Y. (2005) ‘Climate change analysis for western China: 1951–2004’, PhD thesis, China Meteorological Administration, Beijing WMO (1997) Comprehensive Assessment of the Freshwater Resources of the World, World Meteorological Organization, Geneva, Switzerland Xu, Y., Y. H. Ding, Z. C. Zhao and J. Zhang (2003) ‘A scenario of seasonal climate change of the 21st century in northwest China’, Climatic and Environmental Research, vol 8, no 1, pp19–25 (in Chinese) Yin, Y. (2001a) ‘Flood management and water resource sustainable development: The case of the Great Lakes Basin’, Water International, vol 26, no 2, pp197–205 Yin, Y. (2001b) ‘Designing an integrated approach for evaluating adaptation options to reduce climate change vulnerability in the Georgia basin’, final report submitted to Adaptation Liaison Office, Climate Change Action Fund, Ottawa, Canada Yin, Y. Y. (2006) ‘Integrated assessments of vulnerabilities and adaptation to climate variability and change in the western region of China’, final technical report of the AS25 Project, International START Secretariat, Washington, DC Yin, Y. and S. Cohen (1994) ‘Identifying regional policy concerns associated with global climate change’, Global Environmental Change, vol 4, no 3, pp245-260 Yin, Y. and X. Xu (1991) ‘Applying neural net technology for multi-objective land use
3232 J&J Climate Change
1/11/07
9:56 am
Page 227
Evaluation of Adaptation Options for the Heihe River Basin of China 227 planning’, Journal of Environmental Management, vol 32, pp349-356 Yin, Y., S. Cohen and G. Huang (2000) ‘Global climate change and regional sustainable development: The case of Mackenzie basin in Canada’, Integrated Assessment, vol 1, pp21–36 Yin, Y. Y., N. Clinton, B. Luo and L. C. Song (2008) ‘Resource system vulnerability to climate stresses in the Heihe river basin of western China’, in N. Leary, C. Conde, J. Kulkarni, A. Nyong and J. Pulhin (eds) Climate Change and Vulnerability, Earthscan, London Zadeh, L. A. (1965) ‘Fuzzy sets’, Information and Control, vol 8, pp338-353 Zhangye Government (1998) Heihe Basin Ecological and Environmental Protection and Construction Plan: 1998–2010, Zhangye, China
3232 J&J Climate Change
1/11/07
9:56 am
Page 228
13
Strategies for Managing Climate Risks in the Lower Mekong River Basin: A Place-based Approach Suppakorn Chinvanno, Soulideth Souvannalath, Boontium Lersupavithnapa, Vichien Kerdsuk and Nguyen Thuan
Introduction Climate risks are not new to farmers of the lower basin of the Mekong river. For smallholder farmers of rain-fed rice, a dominant economic activity of the region, flood, drought and other climate hazards pose substantial threats to their livelihoods (Chinvanno et al, 2008). A variety of strategies and practices are employed to cope with and manage climate risks, which we document through field studies of farming villages in Lao PDR, Thailand and Vietnam. The strategies and specific measures for managing climate risks are broadly similar across the villages, but there are also important differences, despite the similar hazards being faced and the livelihood patterns held in common. In this chapter we examine these similarities and differences and their implications for promoting effective strategies for adapting to climate change.
Farmers’ Concerns about Climate Our study was conducted through household interviews and focus group meetings in farm communities of the Vientiane Plain and Savannakhet Province in Lao PDR, Kula Field and Ubonratchathani Province in Thailand, and the Mekong river delta of Vietnam. More than 1600 households plus local officials participated in the interviews and meetings, which were conducted in 2004 and 2005 and are detailed in Kerdsuk and Sukchan (2005) and Boulidam (2005). The locations of the study sites are shown in Figure 13.1. The interviews and focus group discussions explored farmers’ perceptions of climate hazards, the risks to their farming activities, observed changes in climate and the impacts, strategies and measures used to cope with climate risks, and options for improving the management of climate risks. The climate
3232 J&J Climate Change
1/11/07
9:56 am
Page 229
Strategies for Managing Climate Risks in the Lower Mekong River Basin 229
Figure 13.1 Study sites in Lao PDR, Thailand and Vietnam
risks found to be major concerns for farmers of the lower Mekong basin vary from location to location, depending on the geographical characteristics of the farmland, farming practices of the community and local features of the climate. However, two climate phenomena are identified by farmers at most of our study sites as significant threats to their livelihoods. These are prolonged midseason dry spells coming after sowing rice seeds or transplanting seedlings and flooding near the end of the crop cycle before harvest time. With the limited extent of irrigated area in the region, most farmers rely mainly on natural rainfall for growing crops (Barker and Molle, 2004). In most parts of Thailand and Lao PDR, farmers of rain-fed rice practise single wet-season cropping, which normally starts in May and ends in October to November. These farmers start sowing rice at the beginning of the rainy season. Farmers who use a transplanting technique begin the process in mid-June to mid-July and harvest in October to November (Boulidam, 2005). The farmers of the Mekong river delta in Vietnam, where the rainy season is longer due to the influence of two monsoon systems, the southwest monsoon and northeast monsoon, are able to grow two rice crops per year (N. T. H. Thuan, personal communication, 2004). A midseason dry spell typically occurs after seeding and/or transplanting. If prolonged, the midseason dry spell can seriously damage young rice plants. Such events increase the cost of production, as farmers may have to replant their rice. However, in some cases of delayed or prolonged dry spell, replanting may not be feasible because the rainy season would end before the replanted rice could reach maturity. Floods that occur late in the rainy season, in October or November, pose serious risks for rice cultivation and farmers’ livelihoods. The lower Mekong river basin experiences floods from the major tributary of the river, most
3232 J&J Climate Change
1/11/07
9:56 am
Page 230
230 Climate Change and Adaptation
commonly towards the end of the rainy season, when water flow is high and water from tributaries cannot flow into the main stem of the river. Sometimes the situation is made worse when water from the Mekong river is backed up into the tributaries (Mekong River Commission, 2005). This period of high flood frequency is close to harvesting time and for most farmers there would be no time to replant rice for that year if the crop were destroyed or damaged by a late-season flood. Only farmers cultivating areas close by the river or major tributaries and using short-cycle rice varieties have the possibility to replant after a late-season flood. In the discussions with farmer communities in Lao PDR and Thailand, the possibility of increasing flood risk in the future due to climate change raised high concerns among the farmers. Direct and indirect impacts of floods and midseason dry spells reported to be major concerns by rice farmers in the lower Mekong are presented in Table 13.1. These have been categorized as first-order impacts (biophysical consequences of meteorological events), second-order impacts (crop production consequences of the biophysical impacts) and higher-order impacts that affect human well-being. Table 13.1 Multiple orders of climate impacts on rain-fed farms in the lower Mekong region Order of impact
Description
Impacts
First-order impacts
Biophysical consequences Drying of soil due to midseason dry spell, of meteorological events particularly after seeding or transplanting Flooding due to heavy rain, particularly toward the end of the rainy season
Second-order impacts
Crop production consequences of the biophysical impacts
Damage to immature plants Reduced harvest Loss of harvest
Third-order impacts
Consequences of the second-order impacts
Increase in cost of production Food scarcity Decline in household income
Fourth-order impacts
Consequences of the third-order impacts
Degradation in household livelihood and socioeconomic condition (e.g. reduced financial and other wealth, reduced food reserves, malnutrition, increased debt) Migration of member(s) of the household (temporary or permanent) Migration of entire household and exit from farming Change in social status (e.g. change from independent farmer to contracted farmer or hired labour) Conflict among villages
Fifth order impacts
Consequences of the fourth-order impacts
Reduced labour force in farming communities Greater costs for hired labour, machinery to replace labour
3232 J&J Climate Change
1/11/07
9:56 am
Page 231
Strategies for Managing Climate Risks in the Lower Mekong River Basin 231
Managing Climate Risks: Current Practice and Potential Adaptation Farmers surveyed in Lao PDR, Thailand and Vietnam identified numerous practices currently in use in their communities which they believe lessened their vulnerability to present-day climate variability and hazards. Some of the measures are motivated by climate risks. Others are primarily motivated by different concerns, yet nonetheless reduce climate risks by increasing the resilience of farmers’ livelihoods to multiple sources of stress. They include onfarm and off-farm measures that are implemented at the household level (Tables 13.2 and 13.3), the community level (Table 13.4), and the national level (Table 13.5). Although none of the measures are motivated by perceived needs to adapt to human-induced climate change, many measures that are focused on near-term climate risks could be developed further for longer-term climate change adaptation (Kates, 2001). Implementation and the effectiveness of the measures in the different countries, some of the enabling and limiting factors that give rise to differences across the countries, and their potential as adaptations to climate change are examined below.
Vientiane Plain and Savannakhet, Lao PDR Most farmers in Vientiane Plain and Savannakhet Province are subsistence farmers, producing rice mainly for their own consumption. They have farms of moderate but sufficient size for producing rice to support the annual consumption of the farm household. They produce a single rice crop each year, and their use of mechanized and advanced farm technology and formal institutional organizations (for example, cooperatives) is limited. The communities are still surrounded by intact natural ecosystems from which natural products can be harvested. This strengthens livelihoods by supplementing and diversifying the farm household’s food and income sources (Boulidam, 2005). Farmers of the Lao PDR study sites tend to rely mostly on farm-level measures for adapting to climate hazards and, to a lesser degree, on collective actions at the community level. Measures at the national level are very limited. Consequently, the capacity of the individual farm household to adapt is a key limiting factor at present for managing climate risks. The responses to climate hazards aim mainly at basic household needs, primarily food security of the household. Common measures implemented by rice farmers include seasonal changes in seed variety, cultivation methods, and timing of farm management tasks based on seasonal climate forecasts made with indigenous knowledge. Also common are raising livestock and harvesting natural products for additional food and income, which are considered major and primary adaptation measures in Lao PDR. The use of indigenous knowledge to make seasonal climate predictions is still popular. Indigenous knowledge based on observations and interpretations of natural phenomena, for example, the height of ant nests in trees, the colour of frog’s legs, the colour of lizard’s tails and various indicators of the dry season weather pattern, is used to make forecasts of the
Objective
Avoid productivity loss from adverse climate conditions, improve food security
Reduce variability of crop yield and income
Diversify exposures to climate hazards
Avoid productivity loss from adverse climate conditions, improve food security
Water source during dry spells; control flooding
Measure
Change rice variety – seasonal
Change rice variety – permanent
Multiple, spatially separated farm plots
Match method and timing of cultivation practices to seasonal climate
Manage water with small-scale irrigation, embankments
Lao PDR: Limited use Thailand: Moderate use Vietnam: Moderate use
Lao PDR: Common practice. Use traditional knowledge, not constrained by market considerations. Thailand: Moderate use; change seedling technique. Crop calendar constrained by the market. Vietnam: Moderate use. Long rainy season allows more flexibility in crop calendar.
Moderate to high if sufficient resources
Low
High
Geographical features; financial resources for investment and operating costs
Forecast accuracy; length of rainy season; market constraints on crop calendar
Land availability and characteristics; population growth
Development of new seed varieties; market acceptance; consumption preference
Forecast accuracy; market acceptance of seed varieties; consumption preference
Enabling and Limiting Factors
9:56 am
Lao PDR: Limited use Thailand: Limited use Vietnam: Limited use
Moderate
Moderate
Effectiveness
1/11/07
Lao PDR: Limited use Thailand: Common practice Vietnam: Common practice – commercial farming
Lao PDR: Common practice. Rice grown for own and local consumption, market acceptance not a factor Traditional knowledge used for seasonal forecasting Thailand: Limited use. Local seed varieties not accepted by market Vietnam: Moderate use. Short-cycle seed variety accepted by the market, but at a lower price
Current Implementation
Table 13.2 Household-level on-farm measures for managing climate risks
3232 J&J Climate Change Page 232
232 Climate Change and Adaptation
Objective
Increase and diversify food supply and income
Reduce variability of food supply and income
Reduce variability of income, food security
Measure
Grow alternate crops between rice seasons
Grow crops resilient to wider range of climate conditions than rice
Livestock
High
High where feasible
Moderate
Effectiveness
Financial reserves; farm size and condition
Know-how; markets for other crops; financial reserves; farm size and soil condition; local culture
Water availability in dry season; market for alternate crops; size and condition of farm land
Enabling and Limiting Factors
1/11/07
Lao PDR: Common practice at a small scale Thailand: Common practice at a small scale Vietnam: Not available
Lao PDR: Limited use Thailand: Limited to moderate use Vietnam: Limited use
Lao PDR: Limited to moderate use Thailand: Limited to moderate use Vietnam: Limited to moderate use; two crop seasons for rice is the normal practice
Current Implementation
Table 13.2 (continued)
3232 J&J Climate Change 9:56 am Page 233
Strategies for Managing Climate Risks in the Lower Mekong River Basin 233
3232 J&J Climate Change
1/11/07
9:56 am
Page 234
234 Climate Change and Adaptation
onset and cessation of the rainy season, quantity of rain and other climate parameters (Boulidam, 2005). The forecasts are used for seasonally adjusting choices of seed varieties and time and methods for soil preparation, seeding, planting, fertilizing, weeding, harvesting and other tasks (Grenier, 1998). Because farmers in Vientiane Plain and Savannakhet Province grow rice mainly for their own consumption (and selling excess production to the local market for local consumption), they have flexibility to select the seed variety to match local climate conditions without regard for the requirements of the commercial markets of other regions. Changing seed varieties in accordance with indigenous seasonal climate predictions is considered to be moderately effective by the surveyed farmers; adjusting the methods and timing of farming practices can be effective up to a point, but implementation has been patchy. Performance of these measures for adapting to climate change could potentially be enhanced by implementation of an early warning system based on modern inter-annual and seasonal climate forecasting, coupled with risk communication techniques to reach the populations at risk. Constraints on this measure include the precision of seasonal climate forecasts, ability and institutional network to communicate the forecasts in ways that are useful to farmers, acceptance of the forecasts by farmers, availability of suitable seed varieties, and flexibility for changing the crop calendar for their cultivation. There is less flexibility for farmers in the Lao PDR sites to change the rice variety on a semi-permanent basis to one that is more climate-resilient or switching to an alternative crop. Constraints on these measures include lack of appropriate seed types, consumption preferences, national dependence on rice for food security, market conditions, lack of know-how and lack of required financial reserves. Consequently, these measures have limited current use. Where they have been used, these measures are considered by farmers to have moderate to high effectiveness for reducing vulnerability to climate and so are potential options for adapting to climate change. But the factors that constrain current use would need to be overcome. Growing a crop other than rice during the dry season is another moderately effective measure that is practised to a limited or moderate degree and can be an effective adaptation to climate change. But its use is restricted to areas where there is access to water and suitable markets. The community still has an important role in the management of climate risks in the study areas of Lao PDR. For example, in the case of severe loss of rice production, the village leader would establish a cooperative network with other villages located near a river or stream or with irrigation systems, where supply of water is available for dry season crop. Shared farmland would be used for the cultivation of short-cycle rice varieties during the dry season to supplement the community’s food supply. In addition, shared resources, such as a community rice reserve contributed to by households in the village or a community fish pond, also act as buffers to climate hazards that sustain the livelihoods and food security of the community. However, some of these collective actions are becoming obsolete, or will be in the near future, because of changes in socioeconomic conditions. Forces that have reduced the role of community-level
Objective
Increase and diversify food supply and income
Increase and diversify income
Increase and diversify income
Increase and diversify income
Measure
Harvest natural products
Produce and market non-farm products
Seasonal migration for off-farm labour
Permanent migration by family member
Low in Lao PDR and Vietnam; high in Thailand
Labour demand in urban areas; access to labour market; reduced farm labour for family
Labour demand in urban areas; access to labour market; networks for job search
Know-how; access to market; market conditions
Productivity, diversity and condition of natural ecosystems near villages
Enabling and Limiting Factors
9:56 am
Lao PDR: Limited use Thailand: Common practice Vietnam: Limited use
Low in Lao PDR and Vietnam; high in Thailand
Low to moderate in Lao PDR and Vietnam; moderate in Thailand
High in Lao PDR; low to moderate in Thailand and Vietnam
Effectiveness
1/11/07
Lao PDR: Limited use Thailand: Common practice Vietnam: Not available
Lao PDR: Limited use Thailand: Moderate use Vietnam: Not available
Lao PDR: Common practice Thailand: Limited use Vietnam: Not available
Current Implementation
Table 13.3 Household-level off-farm measures for managing climate risks
3232 J&J Climate Change Page 235
Strategies for Managing Climate Risks in the Lower Mekong River Basin 235
3232 J&J Climate Change
1/11/07
9:56 am
Page 236
236 Climate Change and Adaptation
actions include population growth and expansion in the use of credit as an alternative to village rice reserves for coping with crop losses. To date, national-level measures to manage climate risks are reported by surveyed farmers to be limited in scope and scale in Lao PDR. National action on climate risks has been constrained by local culture, lack of institutional arrangements to address climate risks, and limited know-how, resources and investment. Looking to the future, climate change is magnifying climate risks and increasing the amount of resources, technology and know-how that will be needed to manage the risks. Farmers have very limited capacity to adapt to the changes, and the diminishing role of communities is widening the gap between needs and capacities for managing risks. Consideration should be given to measures at the national level that would enhance capacity and enable actions for managing and adapting to climate risks at the farm level and at the community level.
Kula Field and Ubonratchathani Province, Thailand Rice farmers in Thailand, particularly in the study areas in the northeast, are mostly commercial farmers who live in a monetary-oriented society and grow rice primarily for national and international markets. They have farms of moderate size on which they produce a single rice crop each year using mechanized and modern technologies, and formal organizations to support farm operations. The sale of rice is their main source of income, which is used primarily to purchase household basic needs, including rice for consumption, which could be cheaper in price and of different quality and texture than the rice the farm household grows. Only a small portion of farmers with larger farms are able to divide their farmland to grow both commercial rice variety for sale and a local rice variety for their own consumption or sale in the local market. The farming communities are closely linked to urban society. The surrounding land area is populated and used for settlements or is deteriorated natural forest that can provide only limited natural products as a supplement or alternative source of food and income (Kerdsuk and Sukchan, 2005). According to the field assessment, farmers at the study sites in Thailand tend to rely on household and national-level measures for reducing climate risks, whereas the role of community-level measures has declined or been neglected. The household-level measures focus on income diversification, primarily from off-farm sources, which are not as sensitive to climate variations as income from rice (Kerdsuk and Sukchan, 2005). The main practice is seasonal migration to work in the cities, which can lead to the permanent migration of some members of the family in order to secure fixed income for the household. Wage income from city employment is less sensitive to climate and helps to insulate the farm household from climate-driven variations in farm income. Seasonal and permanent migration to diversify and supplement household incomes are more common in the Thai study sites than in Lao PDR and Vietnam and are made possible by close links between the rural villages and urban areas where there is demand for labour.
Finance investments to improve farms, livelihoods
Spread risks by sharing rice Lao PDR: Moderate use production, food supplies Thailand: Limited use and labour with other villages Vietnam: Not available
Increase and diversify income
Village fund
Cooperative network among villages
Cooperative processing and marketing of farm and natural products
Know-how; financial reserves; market access; market conditions
Relationship between village leaders; cultural practices
Guaranteed repayment by borrower
Cultural practices; strength of community institutions; guaranteed replenishment of rice reserve
Enabling and Limiting Factors
9:56 am
Moderate
Low to moderate in Lao PDR; low in Thailand and Vietnam
Moderate in Lao PDR and Thailand
High in Lao PDR; low in Thailand and Vietnam
Effectiveness
1/11/07
Lao PDR: Limited use Thailand: Limited use Vietnam: Not available
Lao PDR: Limited use; use expanding under community management Thailand: Common practice; managed by government Vietnam: Not available
Lao PDR: Common practice Thailand: Limited use Vietnam: Not available
Spread risks by creating food reserve; increase income for community
Shared resources – rice reserve/fish pond
Current Implementation
Objective
Measure
Table 13.4 Community-level measures for managing climate risks
3232 J&J Climate Change Page 237
Strategies for Managing Climate Risks in the Lower Mekong River Basin 237
3232 J&J Climate Change
1/11/07
9:56 am
Page 238
238 Climate Change and Adaptation
Unlike the studied communities in Lao PDR, where seasonal changes in rice variety and the crop calendar made in response to seasonal climate forecasts is common practice, these measures are little used by rice farmers in Kula Field and Ubonratchathani Province. Because they grow rice for national and international markets, they are limited in their ability to use local seed varieties, which fetch lower prices than commercial rice varieties, or to alter their crop calendar. In contrast, semi-permanent changes in seed variety to commercial varieties that are more resilient to climate stresses is common practice for farmers at the Thai study sites. This is made possible by the greater financial resources of commercial farming and by research and development programmes that provide new rice varieties that are both accepted in the market and more resistant to stress. This option could be moderately effective for adapting to climate change. Limitations on wider use are financial, technological and environmental. Other on-farm measures for reducing climate risk practised by rice farmers in Thailand include changing seedling technique, using hired machinery, growing alternative crops between rice seasons and raising livestock. Some farmers make investments to increase and sustain the productivity of their farms in ways that make them more resilient with respect to climate variations and changes. For example, they construct small-scale irrigation systems to provide an alternative source of water for midseason dry spells or for growing a crop during the dry season. They may also build embankments to protect their fields from flood damage. Such measures are more common than in Lao PDR. But greater use is limited by financial requirements for investment and maintenance. A small number of farmers with large landholdings implement mixed-farming practices or switch part of their farmland from rice to a crop that is more resistant to climate stresses. Harvesting of natural products from forests, a common practice in Lao PDR, is limited at the study sites in Thailand because of high population densities and the degraded nature of forests that are adjacent to farm lands. National-level policies and measures that serve to reduce vulnerability to climate hazards are more prevalent in Thailand than in Lao PDR and Vietnam. These policies and measures are not motivated by concerns about climate stress, especially climate change, but mainly by poverty reduction goals. Yet national measures in Thailand have supported financial needs, infrastructure development, transitions to more diversified farming systems, marketing of local farm products and farm planning, which have helped to improve the livelihoods of farmers and increase their resilience to climatic stresses. For example, an initiative of the Ministry of Agriculture and Cooperatives in 2004 (Department of Livestock Development, 2004) diversifies farming activity by promoting and providing support to farmers to raise livestock. Another initiative promotes transition from rice cultivation to other plantation crops that are more resistant to climate stresses, such as rubber trees. Research and development by government research facilities have provided new varieties of rice that are more resilient to climate variations, while maintaining the quality that is required by the market.
Assist investments to improve farms, livelihoods
Diversify and improve farm livelihoods; increase resilience and sustainability of rural economy
Increase and diversify incomes
Increase farm productivity and incomes; decrease variability of farm productivity and incomes; increase sustainability of farming Reliable water supply for irrigation; flood control
Enable improved farm management
Financial support to farmers
Support transition to other crops and more diversified farming systems
Support marketing of village products
Research and development of new seed varieties
Provide information for farm management, including seasonal climate forecasts
Lao PDR: Limited use Thailand: Moderate use Vietnam: Not available Lao PDR: Nonexistent Thailand: Limited use Vietnam: Limited use
Moderate
Moderate
Low in Lao PDR; moderate in Thailand and Vietnam
Low in Lao PDR and Vietnam; moderate in Thailand
Low in Lao PDR and Vietnam; moderate in Thailand
Low in Lao PDR; moderate in Vietnam; high in Thailand
Effectiveness
Markets; national financial conditions; know-how; mechanism to develop sustained market National financial condition; time lag between research and availability of new seeds to farmers; technology; transfer knowledge to farmers National financial condition; geographical conditions; technical feasibility Communication channel; know-how to apply information; technology; forecast accuracy
National financial condition; mechanism to allocate funds to the farmers; terms and conditions of loan National financial condition; know-how; mechanism to transfer know-how to farmers; markets; food security; soil properties
Enabling and Limiting Factors
9:56 am
Lao PDR: Moderate use Thailand: Common practice Vietnam: Common practice
Lao PDR: Limited use; rice farming central to food security Thailand: Limited to moderate use; market driven trend toward mono-cropping Vietnam: Limited use; rice farming central to food security Lao PDR: Limited use Thailand: Moderate use Vietnam: Not available
Lao PDR: Limited use Thailand: Common practice Vietnam: Moderate use
Current Implementation
1/11/07
Develop rural infrastructure
Objective
Measure
Table 13.5 National-level measures for managing climate risks
3232 J&J Climate Change Page 239
Strategies for Managing Climate Risks in the Lower Mekong River Basin 239
3232 J&J Climate Change
1/11/07
9:56 am
Page 240
240 Climate Change and Adaptation
Community-level measures are diminishing in Kula Field and Ubonratchathani Province, with the exception of village funds for local investments to support farm livelihoods, which are managed by the government. But community or local administration units could play an expanded role to assist in planning as well as implementing future adaptation to climate change. The advantages of involving local institutions are that they are more aware of local risks, priorities and resources than national authorities and can be more flexible and timely in implementation.
The Mekong river delta, Vietnam Rice farmers of the Mekong river delta in Vietnam are mainly commercial farmers. Unlike farmers at the study sites in Lao PDR and Thailand, they are able to grow two rice crops each year because of a longer rainy season. Farmers of the delta can grow sufficient rice to both supply the annual consumption of the household and sell rice to the market. They make moderate use of modern farm technology and formal institutional organizations in farming practice. The household relies heavily on income from rice production. The farm communities are surrounded by populated areas and are not tightly tied to the urban economic system (field interviews in Long An, Can Tho, Dong Thap and An Giang Provinces, Vietnam, 2004). The farmer of rain-fed rice in Vietnam tends to rely on measures implemented at the household level and aimed mainly towards on-farm actions to protect against climate hazards. Community- and national-level measures play a very limited role in reducing their climate risks. The farm-level solutions include efforts and investments to increase and sustain the productivity of their farms, such as construction and maintenance of small-scale irrigation systems or embankments to protect their farmland from flooding. But investment costs and the limited financial capacity of farmers limit wider use of these measures. Using an alternative strategy, some farmers in the study sites have adapted to floods by accepting them as part of the ecosystem of their farmland, adjusting their crop calendar accordingly and allowing their lands to be flooded, thereby gaining advantages from nutrients being deposited that enhance soil fertility and pollutants being washed from their farmland. The use of alternative crops and seed varieties are also common adaptation measures used by farmers in the Mekong river delta. Changing the variety of rice grown, both seasonally in response to climate forecasts and semi-permanently in response to markets and technological changes, is practised by Vietnamese farmers, even though they are commercial farmers and grow rice to match market demand. Because the rainy season in the delta region is usually seven months long, two crop cycles of rain-fed rice can be grown in one year. A two-crop cycle is also facilitated by the availability of short-cycle rice varieties that are suitable for growing in Vietnam and that are accepted by the market. This gives additional flexibility to farmers in Vietnam to select varieties of rice so as to balance the risk of losses from climate events against expected market returns according to their preferences regarding risk. Consequently, seasonal changes of rice variety are more commonly observed among rice farmers in Vietnam than in Thailand.
3232 J&J Climate Change
1/11/07
9:56 am
Page 241
Strategies for Managing Climate Risks in the Lower Mekong River Basin 241
Community-level measures at the study sites in Vietnam are limited and have low effectiveness. Some measures that are implemented at a national level in Vietnam are considered by farmers to be moderately effective. National research and development programmes have facilitated changes in rice varieties by farmers that lessen vulnerability to climate extremes. Also being implemented, but on a limited scale, are national support for transition to alternative crops and provision of climate forecast information to farmers to assist with farm planning efforts.
Commonalities and Differences: A Matter of Context Rice farmers are shown by the surveys to be experienced at managing climate risks, employing a variety of highly place- and time-specific measures to reduce their vulnerability. Many measures for managing climate risks are common to all of the study sites, at least in general characteristics. But there are also significant differences in the specific measures chosen and in the degree to which farmers rely on farm-level, community-level and national-level actions. These differences are apparent despite our focus on farmers who all make their livelihood primarily from growing rain-fed rice in a common river basin of Southeast Asia and who are exposed to similar climate hazards. The differences demonstrate the strong influence exerted by the local context on climate risk management. They arise from local differences in the specific climate hazards faced, physical and environmental constraints, available technologies, social and economic condition of the farm household and community, vitality of community institutions, degree of engagement in the market economy, market conditions, and the priorities and objectives of the farm households. Even so, some commonalities do emerge from the experiences of farmers across the study sites. Some of the commonalities and differences are summarized in the following sections. In interpreting the findings, it should be borne in mind that the exploratory assessment surveyed farmers at only two sites each in Lao PDR and Thailand and only one site in Vietnam. While for convenience of exposition, we write about farmers in Lao PDR, Thailand or Vietnam, it would be misleading to extrapolate from farmers at the selected sites to characterize the condition and practices of farmers nationwide in any of the three countries. Differences in local context within a country can yield different risk management approaches and performance between communities of the country, just as they do in our comparisons of study sites from different countries. At all of the study sites farmers rely primarily on their own capacity for implementing farm-level measures. But the context for farm-level action is shaped by what is done at community and national levels. Community-level measures are most prevalent in the farm communities of Vientiane Plain and Savannakhet Province in Lao PDR, where they play an important role in providing food security buffers and strengthening livelihoods. Farmers from the study sites in Thailand and Vietnam report that community-level measures are used only to a limited degree and are much diminished relative to the past. This too is the trend at the Lao PDR sites. The diminishing role of collective
3232 J&J Climate Change
1/11/07
9:56 am
Page 242
242 Climate Change and Adaptation
action at the community level may be an important deficit in the capacity of these communities to adapt to future climate change. Our evaluation of national-level measures is based on the perspectives reported by farmers and community leaders at the study sites and does not reflect a comprehensive evaluation of national policies and programmes that are related to climate risks. But this is an important perspective, as it gives a sense of what is happening on the ground, at least in the communities surveyed. In none of the three countries can the national-level measures of which farmers are aware be described as constituting a national strategy for managing climate risks. The actions are not coordinated and typically are not designed specifically to combat climate risks. National-level measures in Thailand, as perceived and reported by farmers in the Thai communities of Kula Field and Ubonratchathani Province, are greater than those reported by farmers surveyed in the other two countries and are an important complement to farm-level measures there. National-level actions in Thailand provide financial and other support for investments in farming infrastructure and expansion of farming technologies, including climate-resilient varieties of rice and other crops, sustainable farming practices, and diversified farm incomes. These efforts help to strengthen farm livelihoods and make them more resilient to climate and other shocks. In Vietnam, the national government supports research and development of seed varieties and provides financial support for investment in farm sector infrastructure, but other measures by the national government are reported by farmers to be limited. National-level measures are the least prevalent in Lao PDR and do not presently play a strong role in making farm households in the study areas climate resilient. Farmers’ objectives, priorities and capacities for using farm-level risk management measures vary between the study sites, and this influences their choice of measures. At the Lao PDR sites, most farmers practise subsistence agriculture and depend primarily on their own rice production for their food supply. Their choice of which rice variety to cultivate only needs to satisfy their own preferences and is not constrained by market requirements. They have access to healthy forests, from which they can harvest products to supplement their food supply. There are opportunities to earn monetary income, but these are little used. Consequently, their choices emphasize providing and protecting basic household needs, most particularly household food security, and employ strategies that have little financial cost and rely on household labour, indigenous knowledge and use of natural products. Rice farmers in Kula Field and Ubonratchathani Province in Thailand are very much oriented to the market economy. They grow rice for cash income and have opportunities to participate in nearby urban labour markets. Their participation in commercial activities provides them with important financial resources and capacity, but their income can be volatile due to climate and market events, and market requirements for commercial rice can limit options for changes in rice cultivation. Consequently, their choices emphasize diversifying household income, particularly from off-farm labour, adoption of rice varieties that are more climate resilient and thus less variable in the income
3232 J&J Climate Change
1/11/07
9:56 am
Page 243
Strategies for Managing Climate Risks in the Lower Mekong River Basin 243
they provide, and investments such as small-scale irrigation and flood control that improve the productivity and resilience of their farmland. In the Mekong River delta of Vietnam, farmers grow rice commercially but have little opportunity to participate in urban labour markets and so are highly dependent on the cash income from the sale of their rice. They have some financial resources and benefit from a longer rainy season than that at the Thai and Lao PDR sites, which allows them to grow two rice crops each year. The availability of short-cycle rice varieties that are suitable for growing on their farms and are accepted by the market also gives them greater flexibility to vary their rice cultivar and crop calendar if the season is expected to be unusually short or dry. Choices of the surveyed Vietnamese farmers emphasize varying cultivation practices to reduce the risk of damage or loss to the rice crop, and investments to improve the productivity and resilience of their farms. Taken as a whole, the survey results suggest a pattern of climate risk management choices by farm households that is shaped by the socioeconomic condition of their surrounding communities. Farmers in communities with less developed socioeconomic conditions tend to pursue simple strategies targeted at increasing coping capacity and sustaining basic needs that can be implemented at the household or community level with limited financial and other resources. Farmers in communities with more developed socioeconomic conditions tend to pursue strategies targeted at reducing the variability of income and at improving the productivity and resilience of their farms. The measures that they adopt tend to depend more on market and other institutions, improved technologies and financial resources than is the case for farmers in less developed communities.
Climate Change in the Lower Mekong Rain-fed agriculture is the dominant economic activity of countries in the lower Mekong basin and engages a high proportion of the population (Schiller et al, 2001; UN-ESCAP, 2006). Despite the efforts made to manage climate risks, farmers of rain-fed crops remain highly vulnerable to climate variations and extremes. Today, human-caused climate change threatens to magnify climate risks in the region and expose farmers to conditions that are outside of the range of current experience. Many farmers over the age of 40 surveyed in Thailand and Lao PDR report noticeable changes in climate patterns over the past 25 to 30 years. These include increasing variability in the dates of onset and end of the rainy season, changes in wind direction, changes in the rainfall distribution pattern throughout the season, and an increase in thunderstorm activity. Thunderstorms, according to farmers at many of the study sites, have increased in frequency, and their occurrence has extended throughout the rainy season. In the past, they only occurred during the beginning and towards the end of the rainy season. This observed phenomenon may be an indicator of changes in the regional high–low pressure front during the rainy season, which no longer moves to a higher latitude after the beginning of the rainy season and south-
3232 J&J Climate Change
1/11/07
9:56 am
Page 244
244 Climate Change and Adaptation
ward again at the end of the rainy season. The front now seems to stay within the region throughout the rainy season. Some farmers also noticed a change in the wind direction pattern, which now varies throughout the season, unlike in the old days, when farmers observed that clouds and rain always came from a certain direction, which was thus more predictable. The future climate of the lower Mekong, like much of the world, will be warmer. It is also likely to be wetter. Climate model simulations project trends toward greater precipitation and higher intensity precipitation in the region during the rainy season, which would increase the magnitude and possibly the frequency of floods in the Mekong basin. The greater precipitation during the rainy season suggests the potential for reduced frequency of prolonged dry spells in the middle of the growing season, but this will depend on how daily variability in rainfall changes. Mathematical modelling simulations performed with the high resolution Conformal Cubic Atmospheric Model (CCAM) from McGregor and Dix (2001) are used to construct scenarios of future climate change for our assessment of impacts and vulnerability (Chinvanno et al, 2008). The scenarios correspond to increases in the atmospheric concentration of carbon dioxide from 360ppm to 540 and 720ppm, which are projected to be reached roughly by the 2040s and 2070s respectively in the IPCC’s A1FI scenario (Nakicenovic and Swart, 2000). Projected changes in annual precipitation in subcatchments of the region range from no change to increases of more than 500mm per year (an increase of up to approximately 25 per cent), with the greatest increases projected for Lao PDR. Higher precipitation within a rainy season of approximately the same length as for the baseline scenario is projected by CCAM, implying potentially greater intensity of rainfall in the rainy season. Because of the potential for increased flood risk, as well as other changes in climate that would impact agriculture, there is a need to evaluate current practice for managing climate risks to the farm sector and strategies for adapting to future climate change.
Enabling Adaptation to Climate Change The measures that are in use in the surveyed communities of Lao PDR, Thailand and Vietnam address current climate risks. They are not deliberate attempts to adapt to climate change. But they provide a basis of experience, knowledge and skills on which to build a climate change adaptation strategy. They also demonstrate a history of farmers in the region acting effectively, within their constraints, in their self-interest to reduce their vulnerability to climate hazards. Despite these efforts, however, farmers in the study area, particularly those who rely on rain-fed crops, are still strongly impacted by prolonged dry spells, floods and other climate events. They are highly vulnerable to climate hazards now and so can be expected to be highly vulnerable to climate change in the future. Their vulnerability is partly due to lack of capacity of farm households, lack of capacity of rural communities, and lack of coordinated national strategies to support farmers and their communities in managing climate risks. An
3232 J&J Climate Change
1/11/07
9:56 am
Page 245
Strategies for Managing Climate Risks in the Lower Mekong River Basin 245
effective starting point for a national strategy of climate change adaptation would be to integrate policies to raise the capacities of farm households and rural communities for managing present climate risks into farm policy, rural development and poverty reduction efforts. Some national policies in the region already do this to a limited extent, though not explicitly. Farm households need help with financial resources, opportunities for offfarm income, marketing of farm products, access to water and healthy ecosystems, information about current and changing climate hazards, knowhow to diversify their farming practices and apply new farming methods and technologies, and access to improved varieties of rice and other crops. They also need buffers to protect their food security, health and livelihoods when they suffer severe crop or financial loss. Delivering this assistance to bolster the capacity of farm households requires community-level institutions with vitality and high capacity. Community institutions can also play a role in coordinating collective actions that require pooled resources to implement. Sadly, community-level institutions in the surveyed communities are in decline, and some community-level measures are becoming obsolete. A reversal of this trend will be important for maintaining existing capacity and raising capacity to the levels that will be needed to address the challenges of climate change. An important concern for adaptation measures in the basin is that measures taken in one locality may have significant ‘spillover’ effects on neighbouring or downstream communities. A holistic approach to national policy and strategic planning for managing climate risks is needed in order to address concerns about potential spillovers. In addition, coordinated regional action by the countries of the lower Mekong river basin should also be considered as the countries share a common resource, the Mekong river, and some adaptation measures may only be feasible with regional collaboration. Climate change will alter water availability, water quality, flood risks, and the performance and sustainability of river-dependent livelihood systems throughout the basin. The actions taken within any of the countries to adapt to these changes are also likely to have spillover effects that cross national borders. In this context, the countries of the lower Mekong river region should explore the potential for trans-boundary effects of their actions, options for reducing negative trans-boundary effects, and options for collective actions that may yield higher effectiveness of the adaptation measures and positive trans-boundary effects.
References Barker, R. and F. Molle (2004) ‘Evolution of irrigation in South and Southeast Asia’, Comprehensive Assessment Research Report 5, Comprehensive Assessment Secretariat, Colombo Boulidam, S. (2005) ‘Vulnerability and adaptation of rain-fed rice farmers to impact of climate variability in Lahakhok, Sebangnuane Tai, Dong Khamphou and Khudhi villages of Songkhone District, Savannakhet, Lao PDR’, Mahidol University, Salaya, Nakhon Pathom, Thailand Chinvanno, S., S. Boulidam, T. Inthavong, S. Souvannalath, B. Lersupavithnapa, V.
3232 J&J Climate Change
1/11/07
9:56 am
Page 246
246 Climate Change and Adaptation Kerdsuk and N. Thuan (2008) ‘Climate change risk and rice farming in the lower Mekong river basin’, in N. Leary, C. Conde, J. Kulkarni, A. Nyong and J. Pulhin (eds) Climate Change and Vulnerability, Earthscan, London Department of Livestock Development (2004) Website of Department of Livestock Development, Ministry of Agriculture and Cooperatives, Thailand, available at www.dld.go.th/transfer/Mcattle/index.php?option=com_content&task=view&id =15&Itemid=1, accessed 9 October 2006 Grenier, L. (1998) Working with Indigenous Knowledge: A Guide for Researchers, International Development Research Centre, Ottawa, Canada Kates, R. W. (2001) ‘Cautionary tales: Adaptation and the global poor’, in S. Kane and G. Yohe (eds) Societal Adaptation to Climate Variability and Change, Kluwer Academic Publishers, Dordrecht, The Netherlands Kerdsuk, V. and S. Sukchan (2005) ‘Impact assessment and adaptation to climate change: The study of vulnerability and adaptation options of rain-fed farms in Tung Kula Ronghai’, Khon Kaen University, Khon Kaen, Thailand McGregor, J. L. and M. R. Dix (2001) ‘The CSIRO conformal-cubic atmospheric GCM’, in P. F. Hodnett (ed) IUTAM Symposium on Advances in Mathematical Modelling of Atmosphere and Ocean Dynamics, Kluwer Academic Publishers, Dordrecht, The Netherlands, pp197–202 Mekong River Commission (2005) Overview of the Hydrology of the Mekong Basin, Mekong River Commission, Vientiane, Lao PDR Nakicenovic, N. and R. Swart (eds) (2000) Emissions Scenarios, Intergovernmental Panel on Climate Change special report on emissions scenarios, Cambridge University Press, Cambridge, UK Rothman, D. S., D. Demeritt, Q. Chiotti and I. Burton (1998) ‘Costing climate change: The economics of adaptations and residual impacts for Canada’, in N. Mayer (ed) The Canada Country Study: Climate Impacts and Adaptation, vol VIII, Environment Canada, Montreal, Canada. Schiller, J. M., S. A. Rao, Hatsadong and P. Inthapanya (2001) Glutinous Rice Varieties of Laos: Their Improvement, Cultivation, Processing and Consumption, Food and Agriculture Organization, Rome United Nations Economic and Social Commission for Asia and the Pacific (UNESCAP) (2006) ‘Annual indicators for millennium development goals’, http:// unescap.org/stat/data/goalIndicatorArea.aspx, accessed 9 October 2006
3232 J&J Climate Change
1/11/07
9:56 am
Page 247
14
Spillovers and Trade-offs of Adaptation in the Pantabangan–Carranglan Watershed of the Philippines Rodel D. Lasco, Rex Victor O. Cruz, Juan M. Pulhin and Florencia B. Pulhin
Introduction Watersheds are a critical aspect of the economy and the environment in the Philippines. Approximately 18 to 20 million people inhabit the uplands of many watersheds and depend on their resources for survival. It is estimated that no less than 1.5 million hectares of agricultural lands presently derive irrigation water from these watersheds. However, despite their tremendous value, it has been observed that many watersheds are now in varying stages of deterioration (Cruz et al, 2000), largely due to population stresses. This could potentially serve to increase the vulnerability of their inhabitants to the impacts of future climate change and could have important implications for the viability of the economy and the environment that depend on them. One of the most important watersheds in the Philippines is the Pantabangan–Carranglan watershed, which houses the Pantabangan Dam and supplies water for irrigation and power generation. It is also home to thousands of people largely dependent on agriculture as a source of livelihood. It has been projected that this watershed is likely to see important increases in temperature and precipitation by 2080 due to climate change, which could cause an increase in the frequency of extreme events such as floods. This could, in turn, significantly impact agricultural systems and endanger the survival of the local population dependent on farming. Additionally, such changes can also endanger the viability of the forest, grassland and brushland ecosystems and threaten the stability of this critical watershed. With this in mind, in this chapter we evaluate potential adaptation options to address the vulnerability of natural and social systems in the Pantabangan–Carranglan watershed, namely forest and upland agriculture, water resources, and local institutions and communities. The shared water resource creates a high degree of interdependence among people, livelihoods
3232 J&J Climate Change
1/11/07
9:56 am
Page 248
248 Climate Change and Adaptation
and biophysical systems located within the watershed. Because of this interdependence, projects or policies implemented for the benefit of one sector often create spillover effects, both positive and negative, for other sectors. Yet crosssector spillover effects or trade-offs are seldom considered. In our analysis of climate change adaptation strategies, we emphasize cross-sectoral effects and trade-offs to better understand how strategies implemented in one sector may conflict with or reinforce strategies of other sectors, and to identify potential win–win options. We anticipate that these findings will serve to better inform policymakers at the local and national levels in the Philippines and enable more effective decision making with respect to addressing the impacts of climate change on the Pantabangan–Carranglan watershed.
Characterization of the Study Site and Existing Vulnerability The following description of the study site was partly based on the Watershed Atlas of Philippine Watersheds (Bantayan et al, 2000), as well as our primary data collection activities. The Pantabangan–Carranglan watershed is located in central Luzon between the 15º44’ and 16º88’ north latitudes and the 120º36’ and 122º00’ east longitudes (Figure 14.1). It is bounded by the Caraballo Mountains on the north, northwest and northeast and by the Sierra Madre ranges on the south, southeast and southwest. The region has a complex topography varying from nearly level, undulating and sloping land to steep hilly and rugged mountain landscapes. It is largely characterized by the Philippine Climatic Type I, with two pronounced seasons – dry from December to April and wet the rest of the year (Bantayan et al, 2000). A small portion of the watershed falls under Climatic Type II, which has no dry season and very pronounced maximum rainfall from November to January. Average annual rainfall at stations in the watershed ranges from roughly 1780mm to 2270mm and the mean monthly temperature ranges from 25.7ºC to 29.5ºC. The watershed lies within the typhoon belt, with most typhoons occurring between September and October (Bantayan et al, 2000). The major land cover types, shown in Figure 14.2, are natural forests (predominantly secondary forests), grasslands, reforestation areas, and alienable and disposable (A and D) lands, the latter including residential and barangay or community sites and cultivated areas. Grasslands occupy the largest portion, followed by secondary forests. Primary forests have largely disappeared since the logging boom of the 1960s (Saplaco et al, 2001). There has been a subsequent significant increase in reforested area, although it too is presently under increasing population pressure. The Philippine government has classified the Pantabangan–Carranglan as a critical watershed since it houses the multipurpose Pantabangan Dam, built in 1974, which supports irrigation and hydroelectric generation and supplies domestic and industrial water needs. Construction of the dam and reservoir resulted in the submergence of the old Pantabangan town and seven adjacent
3232 J&J Climate Change
1/11/07
9:56 am
Page 249
Spillovers and Trade-offs of Adaptation 249
Figure 14.1 Location of the Pantabangan–Carranglan watershed
Figure 14.2 Land use map of the Pantabangan–Carranglan watershed
3232 J&J Climate Change
1/11/07
9:56 am
Page 250
250 Climate Change and Adaptation
villages (Saplaco et al, 2001). All the displaced residents were resettled in the upper portion of Pantabangan and land grants were provided in place of submerged properties. The watershed is presently home to more than 60,000 people comprising of about 12,500 households (National Statistics Office, 2000a, b). Irrigation water from the reservoir is distributed to farmers outside the watershed in adjacent areas of central Luzon. Within the watershed, farmlands are largely unirrigated and dependent on rainfall. Rain-fed agriculture is the main occupation, with rice, vegetables, corn, cassava and onion as primary crops. Fishing is the second most important source of livelihood because of the reservoir and other occupations include grazing and reforestation activities, cottage industries and small businesses. Many residents are also employed in the labour force. Unemployment, however, remains a significant problem as a result of the limited opportunities, and many people resort to slash-and-burn farming and charcoal making at the expense of the local environment (Municipality of Pantabangan, undated; Municipality of Carranglan, undated). The government has instituted several livelihood projects with support from various agencies and institutions to compensate for the residents’ losses and to increase local productivity (Municipality of Pantabangan, undated; Toquero, 2003). These projects also include several reforestation programmes, integrated social forestry programmes and soil erosion control projects. One example is the RP-Japan reforestation project, with assistance from the Japanese government, which has reforested denuded portions of the watershed and created several local jobs. Another example is the World Bank-funded Watershed Management and Erosion Control project implemented by the National Irrigation Authority (NIA). However, despite these efforts, poverty remains a major issue and highlights the government’s failures in providing viable resettlement alternatives. The government programmes have also had the negative effect of creating dependency such that once the programmes conclude, those relying on them for livelihoods resort to charcoal making for survival, thus destroying the very areas they reforested. It is estimated that more than 50 per cent of the residents in the watershed now follow such practices (F. D. Toquero, personal communication). Added to these existing stresses, it is projected that climate change could cause important changes in temperature (about a 5 per cent increase as compared to 1960–1990) and precipitation (about a 13 per cent increase as compared to 1960–1990) in this area by 2080, which could increase the likelihood of floods in the wet season and the possibility of droughts in the dry season (Lasco and Boer, 2006). More than 25 per cent of the watershed is estimated to be highly vulnerable to the impacts of climate change, this including forests, grasslands and brushlands located on steep slopes, at high elevations and by roadsides. The moderately vulnerable areas constitute more than 65 per cent of the watershed and include grasslands, brushlands and forests in other locations. Of these ecosystem types, the dry forests are the most affected in simulation studies and face the possibility of elimination with a 50 per cent increase in rainfall. On the other hand, the wet forest and rainforest zones are expected to increase (Lasco and Boer, 2006).
3232 J&J Climate Change
1/11/07
9:56 am
Page 251
Spillovers and Trade-offs of Adaptation 251
From a socioeconomic perspective, the farming community is likely to be the most impacted by the changes in climate. Focus group discussions with vulnerable communities have led to the identification of areas most susceptible to the impacts of climate variability and change from the perspective of the local inhabitants. These include low-lying flood-prone and fire-prone settlements, agricultural areas prone to floods and droughts, dying streams and rivers, farmlands at the end of the irrigation canals, highly erosive areas along river banks, unstable areas with steep slopes that support infrastructure and grasslands, and forested areas and plantations near roadways (Lasco and Boer, 2006). Significant losses of life, property, infrastructure and livelihoods have been associated with past climate events in this area, with the more vulnerable population unable to ever fully recover from these setbacks. Incidences of illnesses such as diarrhoea, dysentery, dehydration, dengue, malaria and typhoid have also been blamed on climate variability and change. Given this existing vulnerability of the population and ecosystems in the Pantabangan–Carranglan watershed, it therefore becomes extremely important to identify appropriate adaptive measures that can address the negative impacts of future climate change and also help sustain the local environment and economy. It is also critical that the selected sectoral adaptation measures are complementary and do not interact in a manner that reduces their effectiveness or causes undesirable outcomes in order to ensure their successful implementation.
Identification of Sectoral Adaptation Options In attempting to determine and evaluate adaptation strategies for the Pantabangan–Carranglan watershed we focused on three sectors: the agro-forestry, water and institutional organizations of the region. Several different methods were used to identify climate change impacts, vulnerability and adaptation options, including GIS analysis, computer modelling, household surveys, focus group discussions, multi-stakeholder workshops and key informant interviews (for details see Pulhin et al, 2008 and Cruz et al, 2005). Key informant interviews were conducted with government officials from different barangays or communities within the watershed and representatives from different local institutions, while focus group discussions targeted representatives from various institutions concerned with the Pantabangan– Carranglan watershed. A survey was conducted of 375 households from 25 barangays of the municipalities of Pantabangan and Carranglan in Nueva Ecija Province, Alfonso-Castañeda in Nueva Vizcaya, and Maria Aurora in Aurora. The overall objective of these participatory processes was to determine the services provided by the watershed to its different stakeholders; to obtain the perspective of participants on climate variability and extremes experienced in the area, their impacts and the coping strategies adopted; to assess the assistance provided by local institutions; and to solicit recommendations on improving responses to climate impacts.
3232 J&J Climate Change
1/11/07
9:56 am
Page 252
252 Climate Change and Adaptation
Lastly, a major stakeholders’ workshop was held in March 2004 to validate initial results of the study, conduct consultations on climate change impacts on forest, water and land use, and solicit feedback from various stakeholders. Thirty participants from different organizations within the Pantabangan– Carranglan watershed, particularly the National Power Corporation, the National Irrigation Authority, local government units, non-governmental organizations and people’s organizations, were invited to attend this workshop. As anticipated, most of the adaptation options identified via the consultative processes described above were in response to the climatic variability observed in the study site. A brief description of these responses is provided below. Tables 14.1, 14.2 and 14.3 show options for adapting to the impacts of climate variability and extremes that were identified during the multi-stakeholder workshop. The range of options identified by participants suggests that there is a high degree of awareness on adjusting to climate variability and extremes, which could serve to provide solid building blocks for adaptation to the impacts of future climate change. In general, the adaptation options identified are consistent with those recommended in the Philippines Initial National Communication (Government of Philippines, 1999). Agro-forestry options for different land-use categories (Table 14.1) focus on the use of appropriate species and crops, scheduling of activities, technical innovations (for example, in water conservation), capacity building and law enforcement. Options for adapting water supply and use in response to different climate variations (Table 14.2) highlight water conservation practices, choice of crops/species, technical innovations and livelihood options. Options for institutional organizations (Table 14.3) include varied strategies used by the National Irrigation Administration, the Department of Environment and Natural Resources, the National Power Corporation and local government units.
Trade-off Analysis of Adaptation Options The concept of trade-offs arises from the idea that resources are scarce. As a general principle, trade-off analysis shows that, for a given set of resources and technology, to obtain more of a desirable outcome for any given system, less of another desirable outcome is obtained (Stoorvogel et al, 2004a). Although there can be win–win outcomes in two dimensions, even such a win–win outcome must come at the expense of some other desired attribute. Trade-off analysis has been used in exploring the effects of changes in land use, policies and scenarios in agricultural production systems in Ecuador and Peru (Antle et al 2003; Stoorvogel et al, 2004b). In these studies, the researchers developed a simulation model called TOA (trade-off analysis) for conducting an integrated analysis of trade-offs between economic and environmental indicators using biophysical as well as econometric process simulation models. Another form of trade-off analysis has been used in marine protected area management in the West Indies (Brown et al, 2000), which allowed decision makers to consider trade-offs between different criteria to evaluate alternative management options.
3232 J&J Climate Change
1/11/07
9:56 am
Page 253
Spillovers and Trade-offs of Adaptation 253
Table 14.1 Adaptation options for agriculture and forestry by land use category Land Use
Adaptation Options
Lowland farms
Late rains: • Use of early maturing crop varieties • Shift to drought-resistant crops • Use of adaptable species • Supplemental watering Early rains: • Installation of small water impounding facilities
Upland farms
• • • • • • •
Tree plantation
• • • • • •
Adjust silvicultural treatment schedules and practices Plant species that can adjust to variable climate situations Proper timing of tree-planting projects or activities Construction of fire lines Control burning Supplemental watering
Grasslands
• • • • • •
Supplemental feeding Reforestation Adaptation of SALT method of farming in combination with organic farming Promote integrated social forestry and community-based forest management Increase funds for forest protection and regeneration from national government Increase linkages among local government, national government and nongovernmental organizations Introduction of drainage measures Controlled burning Introduction of drought-resistant species Intensive information dissemination campaign among stakeholders
Use of appropriate planting materials Shift to more tolerant crops Use of drought-resistant crops Use of prescribed fungicides/pesticides Installation of fire lines Strict implementation of forest laws Adoption of modern method of farming suited for uplands (e.g. sloping agricultural land technology (SALT)) • Greater visibility of enforcement agencies • Delay of planting
• • • • Natural forest
• Improve safety net measures for farmers • Coordination between local government units • Cancellation of the total ban on logging
In this chapter, we apply a less technically demanding approach for analysing the trade-offs between adaptation options for various sectors, which policymakers and stakeholders can use for first-order estimates. Specific adaptation options were initially analysed individually by sector (forest/agriculture, water resources and institutions and local communities), the details of which are reported in Cruz et al (2005) and Pulhin et al (2008). The trade-offs were
3232 J&J Climate Change
1/11/07
9:56 am
Page 254
254 Climate Change and Adaptation
Table 14.2 Options for adapting water resource supply and use in response to climate variations Climate Variation Adaptation Options General
• • • • • • • • •
Adaptation of SALT method of upland farming Implementation/intensification of reforestation programme Strict implementation of forest laws Programmes and research to increase groundwater utilization by households More funds from national and local government Construction of small water impounding facilities Cloud seeding Introduction of water conservation measures Stabilization of watershed
Water shortage
• Use of shallow tube wells • Planting of new varieties of rice (e.g. Gloria rice) and other crops with less water requirements • Rotation method for scheduling irrigation • Planting early maturing varieties of crops and vegetables • Use the direct seeding method, which requires less water • Development and use of other water sources (e.g. the Atate and Penaranda rivers)
Floods
• • • • • •
None (wait for the next cropping season to respond) Repair the damages Close the main canal Switch to other crops that can survive floods and heavy rainfall Switch to early maturing varieties of crops (e.g. from palay to corn) Explore other livelihoods through the Farmers’ Business Resource Cooperative (e.g. pig rearing, squash and saluyot farming, canton (noodle) making and fruit juice making) • Construct fish ponds • Obtain training in farm management
then determined by means of matrix analysis to bring out the positive and negative interactions between sectoral adaptation options. The positive and negative ratings attributed were based on the expert judgement of the researchers. Next, mitigation measures were identified to minimize or eliminate adverse spillovers. Adaptation strategies common to all sectors were also identified.
Trade-off effects of adaptation strategies for forests and agriculture Table 14.4 presents the analytical matrix describing the positive and negative impacts of the forest and agriculture sector on the other sectors. The effects of adaptation strategies for the forest and agriculture sector on water resources were generally found to be positive. This suggests a synergistic relationship
3232 J&J Climate Change
1/11/07
9:56 am
Page 255
Spillovers and Trade-offs of Adaptation 255
Table 14.3 Adaptation strategies of different institutional organizations Institutions
Adaptation Strategies
National Irrigation Administration
• Reforestation and forest protection • Physical rehabilitation • Manage water releases from the reservoir to reduce flooding
Department of Environment and Natural Resources
• • • • • • • • •
Reforestation and forest protection Adjust programme priorities and schedules Monitoring Shading of seedlings in reforestation sites Deploy forest guards to patrol the forest Plant fire breaks Information, education and communication Hire additional manpower, especially casual labourers Promote Integrated Social Forestry
National Power Corporation
• • • •
Reforestation Information, education and communication Proper choice of species Adjust schedules and implementation
Local Government Unit
• Tree planting and reforestation • Information, education and communication, especially during typhoon season • Creation of El Niño/La Niña, task force formation of disaster brigade • Provision of relief goods • Provision of health services and medicines through extension workers • Bridge construction • Construction, maintenance and repair of roads • Construction, operation, maintenance and repair of water infrastructure (small impounds, wells, diversion canals, water tanks, pumps, etc) for irrigation and other uses • Provision of solar dryer facilities for drying palay • Obtain and distribute free seedlings from the Department of Agriculture • Visits by Barangay Tanod to the community to help with their problems • Small organizations that buy palay (rice) at higher prices and sell rice at lower prices • Training of People’s Organizations
between adaptation for forest/agriculture and water. In contrast, adaptation strategies for forests/agriculture have a mixed effect on the various institutions in the Pantabangan–Carranglan watershed. Most of the recommended adaptation strategies require additional investments, which could pose a significant hurdle in their implementation given the tight budget constraints of many Philippine agencies. Likewise, the effects on local communities are mixed – in some cases there are positive effects but in others quite the opposite. For example, it is possible for farmers to obtain higher yields and incomes as a result of adaptation options such as the use of appropriate crop varieties; however, some
3232 J&J Climate Change
1/11/07
9:56 am
Page 256
256 Climate Change and Adaptation
Table 14.4 Cross-sectoral impacts of forest/agriculture sector adaptations on other sectors Adaptation Strategy Effect on Water for Forests and Resources Agriculture
Effect on Institutions
Effect on Local Communities
Use of early maturing crops Use of droughtresistant crops Supplemental watering
0
+ Higher income
0
+ Higher income
Proper scheduling of planting
Soil and water conservation measures Establishment of fire lines Construction of drainage structures Controlled burning Tree planting fuelwood Enhance communitybased organizations Total logging ban
Use of appropriate silvicultural practices
Better coordination between local government units Information campaign Better implementation of forest laws
+ Lower water demand + Lower water demand – Higher demand for water
– Increase cost of developing alternative sources of water 0 – Increase cost for training, technical assistance, research and development + Conservation of – Increase cost for water training, technical assistance, research and development + More vegetative + Less expense for cover promotes good fire fighting hydrology + Better water quality – Increase cost of (less sediment load) implementation + Less damage to 0 watershed cover + Better hydrology – Increase cost of implementation + Better conservation + Better participation of water in the political process + More forest cover – Increase cost of enforcement and protection +/– Could promote – Increase cost of or impair hydrology implementation depending on the practice + Promotes better + Greater collaboration watershed among local management government units + Better conservation + Increase awareness of water and competence + Promotes better – Increase cost of watershed implementation management
– Greater labour demand + Higher income 0
– Cash expenses
– More labour demand + Less damage to crops from fire; more income + Less soil erosion in the farm; greater yield 0 + Steady supply of – Less area for farming + Better participation – Less income from timber – Fewer sources of income – Increase cost of implementation
+ Better delivery of services to farmers + Increase awareness and competence +/– Could adversely affect current livelihood of farmers that are deemed ‘illegal’
3232 J&J Climate Change
1/11/07
9:56 am
Page 257
Spillovers and Trade-offs of Adaptation 257
adaptation activities such as supplemental watering could require more labour. Among other technical adaptation strategies, the use of early maturing crops and drought-resistant crops has the most positive effects on other sectors. The establishment of fire lines also has a generally positive effect but requires more labour time to establish. Similarly, social adaptation strategies (for example, community organizing) have positive effects on the other sectors. The negative effect in many cases is the additional cost required to implement adaptation strategies. This hurdle may prove daunting, considering the lack of resources of many Philippine government and non-governmental agencies. In order to minimize the negative impacts, adaptation strategies for the forest/agriculture sector could be prioritized on the basis of their effects on other sectors (in addition to their effectiveness for forestry/agriculture). In general, those that have positive effects on other sectors should receive higher priority and attempts could be made to possibly alleviate the negative effects of the remaining options.
Trade-off effects of adaptation strategies for water resources Adaptation strategies for water resources have overwhelmingly positive effects on forest and agricultural crop production as can be seen in Table 14.5. This is understandable considering that proper water management is essential to crop growth and development. Consequently, farmers can also earn a greater income if appropriate adaptation strategies for water are implemented. On the other hand, the effect on institutions is mainly negative as a result of the additional expenditures associated with the implementation of measures such as re-engineering, retooling, training programmes, technical assistance programmes and other related services. This implies that in the face of limited financial resources, adaptation strategies for water may not be fully implemented. One possible approach is to determine which of the recommended adaptation options are economically affordable given the current budget constraints of the implementing agencies.
Trade-off effects of adaptation strategies for institutions Adaptation strategies identified by various institutions have generally positive effects on the other sectors (Table 14.6). This shows that the identified strategies are holistic in nature. (Of course, the underlying assumption is that financial resources are available to implement these strategies, which is the major constraint identified above.) There are a couple of exceptions to this. First, the strategy of stricter enforcement of forest protection rules could adversely affect farmers with no clear land tenure instruments. Many farmers in the watershed are informal settlers, and forest protection officers could compel them to leave their farms. Second, the strategy of releasing water from the Pantabangan Dam to prevent overflow could lead to flooding in downstream communities. Overall, most of the sectoral adaptation strategies identified are found to have mixed interactions (both positive and negative), which suggests that generally, adaptation strategies are not neutral. Of the various adaptation
3232 J&J Climate Change
1/11/07
9:56 am
Page 258
258 Climate Change and Adaptation
Table 14.5 Cross-sectoral impacts of water sector adaptations on other sectors Adaptation Strategy for Water Resources
Effect on Forest Resources/ Agriculture
Effect on Institutions
Effect on Local Communities
Reforestation + Greater tree cover Soil and water + Increased yield conservation measures
– Higher investment cost + More income – Higher investment cost + More income
Water impoundment
+ Increased yield
– Greater expenses
+ More income – Greater expenses
Well construction
+ Increased yield
– Greater expenses
+ More income – Greater expenses
Cloud seeding
+ Increased yield
– Greater expenses
+ More income
Use of appropriate crops/varieties
+ Increased yield
– Greater expenses for research and development, technical assistance and information, education and communication
+ More income
Irrigation management
+ Increased yield
– Greater expenses for implementation
+ Increased income
Tap other water sources (e.g. rivers)
+ Increased yield
– Greater expenses
+ Increased income
Fishponds in flooded areas
+ Decreased pressure – Additional expenses on forests and for technical agricultural resources assistance
+ Increased income – Greater expenses
Repair of damaged infrastructure
0
– Greater expenses
0
Shift in livelihood
+ Less use of land
– Additional expenses for technical assistance, training
+ Increased income
Strict implementation of forest laws
– Could affect crop production in areas deemed for forest
+ Strengthen role of regulatory agencies
+/– Promote peace but possibly lower income
Research on groundwater
0
– Greater expenses for research and development, technical assistance and information, education and communication
0
Capacity building activities
+ Build up of mass of – Greater expenses for competent players research and development, technical assistance and information, education and communication
Note: Key: + positive impact; – negative impact; 0 no effect.
+ Build up of mass of competent players
3232 J&J Climate Change
1/11/07
9:56 am
Page 259
Spillovers and Trade-offs of Adaptation 259
Table 14.6 Cross-sectoral impacts of adaptations by institutions on other sectors Adaptation Strategy for Water Resources
Effect on Forest Resources/ Agriculture
Effect on Institutions
Effect on Local Communities
Reforestation
+ Increased tree cover + Better watershed cover
+ Source of fuelwood/tree products
Forest protection
+ Reduce forest destruction
+ Better watershed cover
+/– Could affect source of forest products
Physical rehabilitation
0
0
+ Better facilities
Release of water from the dam
0
0
– Flooding in low-lying areas
Adjustment of schedule
0
0
0
Fire break establishment
+ Reduced fire loss
+ Better watershed cover
+ Reduced fire loss
Community-based management
+ Better forest land management
+ Better watershed cover
+ Empowerment of local people
Development of water sources
+ Increased crop yield + Stable water supply
+ Increased crop yield
Hiring additional personnel
+ Better forest protection
+ Improved water quality and regimen
+ Additional sources of income
Proper choice of species
+ Increased yield
0
+ Increased income
Provision of relief goods
+ Reduction of + Reduction of pressure on forest and pressure on water agricultural resources
+ Relief goods supplied
Creation of task forces
+ Better coordination + Better coordination
+ Better coordination
Infrastructure repair and construction
+ Increased farm yield + Stable and more efficient water supply
+ Increased income
Information, education and communication
+
+
+
Training of People’s Organization
+ Better forest/farm management
+ Better water management
+ Skills developed
Note: Key: + positive impact; – negative impact; 0 no effect.
options recommended, four were common to all the sectors: tree planting/ reforestation, selection of appropriate species/crops and better implementation of laws and information campaigns (Table 14.7). This reflects the high degree of consciousness among stakeholders of the importance of forests in the watershed. Also, the use of soil and water conservation strategies was explicitly identified by the forest and water sectors and was also implied by the institu-
3232 J&J Climate Change
1/11/07
9:56 am
Page 260
260 Climate Change and Adaptation
tional sector. These results suggest that individual adaptation strategies could address more than one sector, allowing for greater synergy and cost-efficiency. Table 14.7 Adaptation strategies common to multiple sectors Adaptation Strategy
Forest/Agriculture
Water
Institutions
Tree planting/reforestation
X
X
X
Selection of appropriate crops/varieties
X
X
X X
Better implementation of forest laws
X
X
Soil and water conservation measures
X
X
Establishment of fire lines
X
Construction of drainage
X
X
Controlled burning
X
Enhance community-based organizations
X
Total logging ban
X
Appropriate silvicultural practices
X
X
Better coordination between local government units
X
X
Information campaign
X
X
X
Water impoundment
X
Well construction
X
Irrigation management
X
Cloud seeding
X
Develop other water sources
X
Research
X
Capacity building
X
X
X X
While we have restricted ourselves to a qualitative approach to examine the trade-offs between various sectoral adaptation strategies, trade-offs can also be quantified. For example, the expense associated with reforestation, which has been identified as a desirable adaptation strategy by all sectors, is estimated by the Department of Environment and Natural Resources (DENR) to cost about US$900 per ha for three years (DENR, 1999). This level of investment may prove limiting for the organizations bearing the cost of reforestation such as the National Irrigation Administration, the National Power Corporation and the DENR. One possible way to reduce expenses is by encouraging more community participation to lower the labour cost, which constitutes the biggest fraction (about 70 per cent) of the total cost of reforestation. A somewhat less expensive adaptation strategy identified by stakeholders is fire line construction for forest/grassland fire prevention in the watershed, which costs about US$20/ha. Although this is typically a part of reforestation work, it can
3232 J&J Climate Change
1/11/07
9:56 am
Page 261
Spillovers and Trade-offs of Adaptation 261
be undertaken as a separate project. Some of the other adaptation options are even less costly, for example, proper scheduling of planting to coincide with the late or early onset of the rainy season involves practically no financial cost to farmers and institutions and yet the positive effect on farm yields and income could be very high. While a detailed quantification of trade-offs in sectoral adaptation strategies is beyond the scope of this chapter, this is a potential area for further exploration in future studies.
Policy Implications The findings discussed above have important implications for policy planning and management with respect to addressing the impacts of climate change in watersheds in the Philippines. By identifying the trade-offs between adaptation strategies for different sectors in the Pantabangan–Carranglan watershed, we have highlighted the significance of a cross sectoral analysis at the watershed scale, which can inform the design of effective policy and management responses to climate change. A cross-sectoral analysis of adaptation strategies will enable decision makers and managers to anticipate potential conflicts between sectors early on, thus providing greater opportunities for finding solutions. Cost stands out as the most significant limiting factor in the implementation of adaptation options. The most common trade-off identified for all sectors is the additional cost of implementing adaptation strategies such as the construction of a water-impounding structure or tree planting. In developing countries like the Philippines, where priority for climate change adaptation is low, strategies that meet other, sometimes more important, goals may therefore stand better chances of implementation. For example, reforestation and tree planting programmes are already ongoing in the watershed, irrespective of climate change considerations, due to the other benefits they offer. Besides highlighting the negative interactions, this type of trade-off analysis also helps to pinpoint those strategies that have synergistic effects across sectors and can be prioritized for implementation since they provide greater chances of stakeholder acceptance. A good example is tree planting/reforestation, which was identified as an adaptation strategy by all the sectors investigated. It is also important that the positive and negative interactions between different adaptation strategies and their implications outlined above must be presented to policymakers in a manner that is easily comprehensible in order to enable effective decision making. One way of visually presenting the results of the study in a nutshell to policymakers is by means of the chart in Figure 14.3, which summarizes the impacts of adaptation strategies in one sector on other sectors and thus captures the potential synergies and conflicts in a simplified manner. For example, adaptation measures in water resources have mostly positive effects on forests and agriculture. These include tree planting and provision of irrigation water, which directly benefits forestry and agriculture. However, adaptation in water resources will have a largely negative effect on institutions. This can be attributed to the high cost of many of the adaptation measures identified, such as construction of shallow tube wells and impound-
3232 J&J Climate Change
1/11/07
9:56 am
Page 262
262 Climate Change and Adaptation
ing structures (Table 14.2). On the other hand, the effects of adaptation measures in forest resources and agriculture on local communities are mixed, in other words there are both positive and negative interactions. For example, use of more resistant varieties could lead to higher incomes but establishment of fires lines means higher labour costs.
Figure 14.3 Summary of effects of adaptation strategies in one sector on other sectors
This kind of visual presentation would thus enable policymakers and stakeholders to flag those adaptation measures that deserve attention and further study. Those with negative interactions could also be prioritized for quantitative analysis of trade-offs in order to determine the degree of impact and evaluate possibilities to minimize the impact.
References Antle, J., J. Stoorvogel, W. Bowen, C. Crissman and D. Yanggene (2003) ‘The trade-off analysis approach: Lessons from Ecuador and Peru’, Quarterly Journal of International Agriculture, vol 42, pp189–206 Bantayan, N. et al (2000) Philippine Watershed Atlas, College of Forestry and Natural Resources, University of the Philippines, Laguna, Philippines Brown, K., W. N. Adger, E. Tompkins, P. Bacon, D. Shim and K. Young (2000) ‘Tradeoff analysis for marine protected area management’, CSERGE Working Paper GEC 2000-02, Centre for Social and Economic Research on the Global Environment, University of East Anglia, Norwich, UK Cruz, R. V. O., R. D. Lasco, J. M. Pulhin, F. B. Pulhin and K. B. Garcia (2005) ‘Assessment of climate change impacts, vulnerability and adaptation: Water
3232 J&J Climate Change
1/11/07
9:56 am
Page 263
Spillovers and Trade-offs of Adaptation 263 resources of the Pantabangan–Carranglan watershed’, Environmental Forestry Programme, College of Forestry and Natural Resources, University of the Philippines Los Baños College, Laguna, Philippines Cruz, R. V. O., S. R. Saplaco, R. D. Lasco, M. M. B. Avanzado and F. B. Pulhin (2000) ‘Development of water budget models for selected watersheds in the Philippines: Assessment of the impacts of ENSO on the water budget of selected watersheds’, unpublished research report, Philippine Council for Agriculture, Forestry and Natural Resources Research and Development, Department of Science and Technology, Manila, Philippines Government of Philippines (1999) Philippines Initial National Communication to the UN Framework Convention on Climate Change, Government of Philippines, Manila Lasco, R. D. and R. Boer (2006) ‘An integrated assessment of climate change impacts, adaptations and vulnerabilities in watershed areas and communities in Southeast Asia’, final report, Assessment of Impacts and Adaptation to Climate Change Project No AS21, International START Secretariat, Washington, DC Municipality of Carranglan (undated) Development Master Plan of the Municipality of Carranglan, Nueva Ecija, 2003–2007, Carranglan, Nueva Ecija, Philippines Municipality of Pantabangan (undated) Master Plan of the Municipality of Pantabangan, Nueva Ecija, 1998–2000, Pantabangan, Nueva Ecija, Philippines National Statistics Office (2000a) ‘Census 2000: Philippines population by Barangay’, CD-ROM, National Statistics Office, Makati, Philippines National Statistics Office (2000b) Philippine Statistical Yearbook, National Statistics Office, Makati, Philippines Pulhin, J. M., R. J. Peras, R. V. O. Cruz, R. D. Lasco, F. B. Pulhin and M. A. Tapia (2008) ‘Climate variability and extremes in the Pantabangan–Carranglan watershed of the Philippines: An assessment of vulnerability’, in N. Leary, C. Conde, J. Kulkarni, A. Nyong and J. Pulhin (eds) Climate Change and Vulnerability, Earthscan, London Saplaco, S. R., N. C. Bantayan and R. V. O. Cruz (2001) GIS-based Atlas of Selected Watersheds in the Philippines, Department of Science and Technology, Philippine Council for Agriculture, Forestry and Natural Resources Research Development, and Environmental Remote Sensing and GIS Laboratory, University of the Philippines Los Baños College of Forestry and Natural Resources, Laguna Stoorvogel, J. J., J. M. Antle, C. C. Crissman and W. Bowen (2004a). ‘The trade-off analysis model: Integrated biophysical and economic modeling of agricultural production systems’, Agricultural Systems, vol 80, pp43–66 Stoorvogel, J. J., J. M. Antleb and C. C. Crissman (2004b) ‘Trade-off analysis in the Northern Andes to study the dynamics in agricultural land use’, Journal of Environmental Management, vol 72, pp23–33 Toquero, F. D. (2003) ‘Impact of involuntary resettlement: The case of Pantabangan resettlement in the province of Nueva Ecija’, PhD thesis, Central Luzon State University, Science City of Muñoz, Nueva Ecija, Philippines
3232 J&J Climate Change
1/11/07
9:56 am
Page 264
15
Top–Down, Bottom–Up: Mainstreaming Adaptation in Pacific Island Townships Melchior Mataki, Kanayathu Koshy and Veena Nair
Introduction Climate change, whether due to natural variability or human activity, is one of the most pressing issues for the Pacific island countries. The impacts of climate variability and extreme events such as cyclones, floods, droughts and sea level rise are rapidly pushing people beyond their coping range. The already strained economies are being drained trying to keep up with the impacts of these stresses on livelihoods. In the 1990s alone, the Pacific island region bore up to US$1 billion costs related to climate extremes (Campbell, 1999; Feresi et al, 2000), and the costs are expected to rise even further with a rise in the frequency and intensity of extreme events. Climate projections for the South Pacific indicate warming of 0.8 to 1.8°C and precipitation changes that range from -8 to +7 per cent by mid-century (Ruosteenoja et al, 2003). By the end of the century, projected warming is 1.0 to 3.1°C and precipitation changes range from -14 to +14 per cent. Projections of globally averaged sea level rise range from 0.18m to 0.58m in 2090–2099 relative to 1980–1999, while tropical cyclones are likely to become more intense, have higher peak wind speeds and bring heavier rainfall (IPCC, 2007). Small islands share a number of characteristics that increase their vulnerability to climate variability and change, including small land area, proneness to natural disasters and climate extremes, limited water supplies, high concentrations of population and infrastructure close to coasts, open economies, low adaptive capacity, and adaptation costs that are high relative to national incomes (Mimura et al, 2007). Analyses by Feresi et al (2000) predict up to 14 per cent loss of coastal lands in Fiji due to sea level rise and flooding by 2050, lands that are prime areas for economic activities and human settlements. In some areas, the demand for water resources is expected to outstrip supply by 5–8 per cent by 2050 (Feresi et al, 2000). Agriculture, human health and fisheries are also expected to be impacted negatively because of climate change, which, in turn, will have a negative impact on the economies of Pacific island countries. Following a ‘do-nothing’ option, a small island such as Viti Levu (Fiji) could
3232 J&J Climate Change
1/11/07
9:56 am
Page 265
Top–down, Bottom–up 265
incur a cost equivalent to 2–4 per cent of Fiji’s gross domestic product, or US$23–52 million, by 2050 in damages associated with climate-related disasters (World Bank, 2000). The capacity to mitigate the impacts of climate change and extreme events is beyond the island countries of the Pacific. Even if the developed nations reach the target of reducing emissions as proposed by the Kyoto Protocol, climate will continue to change, and the Pacific islands, among other poor countries of the world, will have to bear the consequences. The only logical option for Pacific island countries is to proactively learn to adapt to climate variability and extreme events. Several adaptation options have been implemented in the Pacific islands through the actions of individuals, national governments and externally-funded climate change adaptation projects. The most common step being taken is the construction of seawalls to protect settlements against coastal erosion and storm surges. However, some options have proven to be unsuccessful in solving the underlying problems. For example, in Qoma, Fiji, the community reported experiencing frequent inundation further downstream after the construction of a sea wall upstream (World Bank, 2000). Given the uncertainties regarding impacts and adaptation strategies, varying approaches have been experimented with in the islands. For example, the Secretariat of Pacific Regional Environment Program (SPREP) carried out the project Capacity Building for the Development of Adaptation Measures in Pacific Island Countries (CBDAMPIC), which focused on community and national capacity building, and identification and implementation of adaptation measures through community participation (Nakalevu et al, 2005). The Asian Development Bank- and Canadian Cooperation Fund for Climate Change-funded project Climate Change Adaptation in the Pacific (CLIMAP) has focused on developing case studies that demonstrate climate change adaptation through risk reduction. The case studies cover the spectrum from immediate project-level actions to longer-term national-level development planning. This chapter presents lessons learned from a case study of vulnerability to river flooding and adaptation in Navua township of Viti Levu in Fiji, a typical community of the Pacific islands. The Navua case study is one component of a larger study implemented as part of the Assessments of Impacts and Adaptation to Climate Change (AIACC) programme. The project expanded an integrated framework for assessing climate change vulnerability and adaptation to incorporate both natural and human systems, and applied the framework to Viti Levu in Fiji and Aitutaki in the Cook Islands (Koshy, 2007).
Navua Township Navua, characterized by rapid urbanization, meagre economic activities and low to middle incomes, is prone to recurrent flooding. The 1996 census recorded the residential population as 4220, 52 per cent higher than the population in 1986 (Sinclair Knight Merz, 2000) and the current population is estimated to be near
3232 J&J Climate Change
1/11/07
9:56 am
Page 266
266 Climate Change and Adaptation
7000. Urbanization and resettlement of displaced sugar cane farmers following the expiry of their land leases has contributed to this rapid growth. Our surveys found that, on average, a Navua resident earns $US35–46 per week, which is comparable to the average weekly earnings recorded by a consulting firm in 2000 (Sinclair Knight Merz, 2000). This indicated that the socioeconomic status of average Navua residents has not improved over the past five years. Consequently, residents also rely on subsistence farming and fishing for sustenance and to supplement their incomes. The land area of Navua is 16.7km2, with a maximum elevation of 31.4m and a minimum elevation of less than 0.6m above sea level. The Navua floodplain is characteristically low lying, increasing the potential for flooding during intense and/or prolonged rainfall episodes. A section of the Navua river measuring about 163m wide and 5.8km in length runs along the town, with the central business district and some homes only a few metres from the river banks. The greater Navua area is crisscrossed by a network of irrigation channels and floodgates at the coast, previously used to distribute and control water needed for commercial rice farming. Before 1990, commercial rice farming was an important economic activity in the area, but it was abandoned because of competition from cheaper rice imports from Asia, floods and pest infestation (Sinclair Knight Merz, 2000). Today, the main agricultural activities are small-scale commercial and subsistence farming of root crops such as cassava, dalo or taro, and vegetables, and raising livestock such as cattle and goats. Logging in the upper catchment of the Navua river and aggregate mining in the river are also significant activities. In 2003, 113 properties comprising 45 business properties (private and government) and 68 residential properties in the project site were surveyed. All interviewees were adults present at the properties during the survey. Information and data were gathered concerning socioeconomic variables (population, economic activities and income level); building types, recollections of past floods, views about factors contributing to flooding, adaptive measures taken by residents to cope with flooding, the barriers to implementing adaptive measures, and perceptions of climate change in general. A second survey was carried out in 2004 following a flash flood that affected the study area in April of that year. Sixty-five per cent of properties initially surveyed in 2003 were surveyed again to record data relating to the recent flood. Interviews were also held with officials from the Navua Rural Local Authority and persons familiar with flooding in Navua. Results of these surveys are discussed below.
Flood Risks Navua has experienced frequent flooding in recent decades (Fiji Meteorological Service, 2004), and flooding is the major threat to livelihoods of Navua residents. In the recent flooding episode of April 2004, the national government incurred costs of approximately US$65,000 for emergency food rations for a 30-day period for the greater Navua area. Damage to homes was estimated at over US$100,000 (SOPAC, 2003). More than 2700 people, repre-
3232 J&J Climate Change
1/11/07
9:56 am
Page 267
Top–down, Bottom–up 267
senting about 40 per cent of Navua’s population, were displaced from their homes because of the flooding and temporarily relocated to evacuation centres (Central Division Disaster Management Council Operation Centre, 2004). Apart from businesses and the houses of a few middle-class residents, most of the properties in Navua are not insured because of financial constraints and inability to meet basic insurance requirements. And the full social and economic impacts of the most recent and previous floods are not known. However, the impacts are deemed substantial, taking into consideration destruction of crops, loss of income and properties, diseases and, in some cases, deaths. Many may not be able to improve their standard of living if they are to sustain significant damages to life and property on an annual basis. In the surveys of residents, five significant local flood events were recalled since 1972, three of which flooded more than 80 per cent of the land area (Table 15.1). All five of the flood events came during the wet season of November to April, also the season of tropical cyclones, and four of the events were initiated by intense and prolonged rainfall associated with tropical cyclones. However, the most recent event, which caused the most extensive flooding of the five, was due to intensive rainfall associated with two consecutive tropical depressions. Because of strong variability in daily rainfall, there is also potential for flash floods in the dry season (May–October). Table 15.1 Flood extent, duration and rainfall in five recalled flooding episodes in Navua Climate Event
Dates and Duration of Event
Total Rainfall Average Daily (mm) Rainfall (mm/day)
Area Flooded (% of study site)
Bebe
19 Oct–6 Nov 1972 19 days
652
34
86
Wally
1–6 Apr 1980 6 days
682
113
22
Oscar
28 Feb–2 Mar 1983 3 days
412
19
21
Kina
26 Dec 1992–5 Jan 1993 11 days
537
6
89
Two tropical depressions
6–15 Apr 2004 10 days
592
59
90
Flood frequency has been observed to have increased in the past decade compared to earlier periods. Survey participants were asked their opinion on the factors that contribute to increased potential for flooding in Navua. Contributing factors identified by the respondents include increased sediment input to the river, which raises the river bed; build up of sediments at the river mouth, impeding movement of water out of the basin; the presence of aban-
3232 J&J Climate Change
1/11/07
9:56 am
Page 268
268 Climate Change and Adaptation
doned irrigation channels previously used in rice farming; non-functioning floodgates, especially those at the coast; lack of regular dredging of the Navua river; and changes in rainfall patterns. While survey respondents report their perception that rainfall patterns have changed, this is not corroborated by our analysis of weather station data for Navua obtained from the Fiji Meteorological Service. Normal rainfall for Navua, based on the average for the 1961–1990 period, is 3500mm per year, but with large interannual variations of as much as +/-40 per cent (see Figure 15.1). Analysis of observed rainfall for the 43 year period 1960–2003 does not show any discernable increasing or decreasing trend. This result is consistent with similar analyses of rainfall patterns for Suva and Nadi carried out by Mataki et al (2006). The large interannual variations are driven mainly by movements of the South Pacific Convergence Zone, the main rain-producing system of the region, and the presence and absence of the El Niño Southern Oscillation and La Niña (Mataki et al, 2006). El Niño conditions are associated with droughts while La Niña episodes are associated with enhanced rainfall across the Western Equatorial Pacific, including Fiji. We also analysed the average return periods for extreme rainfall events in the
Figure 15.1 Observed rainfall anomalies for Navua, 1960–2003 (percentage variation from normal rainfall); the trendline is not statistically significant
two wettest months of the year, March and April, based on changes in the time period between peak rainfall events. The average return period of intense rainfall decreased from approximately 3 years in the period up to 1994 to 2 years since that time (see Figure 15.2). However, the reduction was found to be
3232 J&J Climate Change
1/11/07
9:56 am
Page 269
Top–down, Bottom–up 269
statistically insignificant. This further reinforces the conclusion that rainfall in the most recent decade has not diverged from the established norm of 1961–1990. Nevertheless, the study site had been flooded in recent decades more
Figure 15.2 Maximum daily rainfall during March and April, 1960–2003
frequently than in the past. The explanation lies in the complex interplay between climatic factors and human activities in the Navua watershed. The Navua river is silted more intensely than before because of intensified human activities such as logging, aggregate mining and agricultural practices in the upper Navua river catchment (Sinclair Knight Merz, 2000; Ba, 1993; SOPAC, 2003). Siltation raises the riverbed and increases the river’s potential to burst its banks during prolonged and/or intense rainfall episodes (Central Division Disaster Management Council Operation Centre, 2004; National Institute of Water and Atmospheric Research Ltd, 2004). The homeowners interviewed also recognized this as a major contributor to increasing flooding potential. Moreover, they stated the view that accumulated silt at the river mouth acts as a barrier to the free flow of water during floods. Logging in the Navua catchment, mainly of mahogany, takes place in both natural and planted forests. Logging practices are regulated by the Fiji Forestry Decree of 1992 (Government of the Republic of Fiji, 1992) and the National Code of Logging Practice (NCLP), as well as by other environmental legisla-
3232 J&J Climate Change
1/11/07
9:56 am
Page 270
270 Climate Change and Adaptation
tion and forestry regulations. These instruments have provisions to reduce the environmental impacts of logging, including soil erosion, blockage of waterways and sediment input to streams and rivers (Strehlke, 1996). Despite the regulations, however, logging operators often knowingly ignore requirements, engaging in excessive bulldozing and logging within waterway buffer zones and on steep slopes (Chand and Prasad, 2005). Such practices have been cited as contributors to recent floods in the Northern Division of Fiji that caused damages estimated at 10 million Fijian dollars (Fiji Times, 2007). Some of Fiji’s largest mahogany forests are found in the upper catchment of the Navua river. Logging of these forests is contributing to problems of sedimentation of the river and is expected to increase in future years. Aggravating the problem of sedimentation of the river is the failure of the national government to implement a programme of regular dredging of waterways. Dredging can facilitate the flow of rain-waters out to sea and dampen the severity and extent of flood events. However, the Navau river was last dredged in 1994 and funds that were earmarked for dredging the river in 2000 were diverted to other purposes (Auditor General of the Republic of Fiji, 2001). Navua residents also attributed the extensive nature of flooding in April 2004 to dysfunctional irrigation channels and floodgates. Examination of the data in Table 15.1 indicates that the two most extensive floods occurred in 1993 and 2004, after commercial rice farming was abandoned. After the abandonment of rice farming, the irrigation channels and floodgates were poorly maintained, giving rise to blockages and uncontrolled movement of floodwater. Elevation above the ground, location relative to the river and abandoned irrigation channels, other topographic features and structural strength all influence the vulnerability of buildings to flooding. Seventy-five per cent of surveyed homes were raised above ground level, while the remaining 25 per cent were built on ground level. Homes that were raised on pillars or had concrete porches were observed to be less affected by flooding, excluding factors such as the location of the home, the intensity of flooding and the strength of the building. The depth of water in homes that were flooded varied considerably. In most cases, both raised and unraised homes were flooded, but as expected, unraised homes experienced higher levels of water within them during floods, which also suggests that these homeowners generally sustain greater property losses. Nevertheless, even some homes raised nearly 2m above the ground level were also flooded during the five flooding episodes. These observations suggest that houses in Navua, especially those within a few metres of the river or confluence points of irrigation channels and those near previous flood water routes, may need to be raised by more than 2m to reduce the potential of being flooded in the future.
Adapting to Climate Change Nearly 95 per cent of the interviewed residents and local authorities of Navua have heard about climate change, mainly through the media, though most are unaware of the human and natural factors responsible. They also associate the
3232 J&J Climate Change
1/11/07
9:56 am
Page 271
Top–down, Bottom–up 271
recent floods with climate change, which is an indication of their limited understanding of climate change science. Most are aware of measures that can be taken to reduce vulnerability to flooding and accept that as individuals, they have a role to play in reducing their vulnerability to flooding in collaboration with the government and its agencies. Nearly all survey respondents recommended dredging of the river as an important measure to reduce flood risks, in addition to building raised and sturdy homes. Other adaptive measures identified by respondents include taking insurance coverage, relocation, and maintaining irrigation channels and floodgates. Adaptive capacity is defined by the IPCC as the ability of a system to adjust to climate change, including climate variability and extremes, to moderate potential damages, to take advantage of opportunities and to cope with the consequences (McCarthy et al, 2001). In the context of Pacific island countries, adaptive capacity is dependent on the net resources (financial, human and technological) available to national governments, communities and individuals to implement adaptation measures. Resource allocation and use are strongly influenced by factors such as governance, fiscal policies, tradition and culture, poverty, hardship, and prevailing socioeconomic and environmental conditions. On the basis of our studies, it appears that a majority of the residents in Navua lack sufficient net resources and the capacity to enable them to autonomously adapt to climate stresses and shocks to any significant extent without the government’s intervention. Adapting to present climate variability and extreme weather events is an early opportunity to enhance the resilience and the adaptive capacity of Navua residents to future climate change. Moreover, adaptation to climate change in a socioeconomically disadvantaged community such as the one in Navua is better approached from a broader development framework. Within such an approach, the government would oversee implementation of adaptation measures and incorporate adaptation measures in the development plans for the region. In addition, autonomous adaptation by individuals and communities should be encouraged with appropriate incentives and clear demarcation of responsibilities of government and communities in planning and implementing adaptation measures. Furthermore, government can also engage development partners (including funding agencies) through bilateral and multilateral arrangements to provide technical and financial support for the implementation of adaptation options. This is quite crucial for Pacific islands given the diminishing overseas development aid and the tight donor situation despite the continued requests of Pacific Islanders for donor contributions to satisfy various development objectives. A lesson from our study is that a system embracing both top–down and bottom–up approaches to the adaptation process has the best chance of improving the adaptive capacity of towns in the Pacific with geographic features and socioeconomic backgrounds similar to those of Navua. This dual approach is also aligned with current regional efforts under the CBDAMPIC project to mainstream climate change adaptation into national development planning and concurrently engage and empower local communities and nonstate actors to develop and implement effective and appropriate adaptation
3232 J&J Climate Change
1/11/07
9:56 am
Page 272
272 Climate Change and Adaptation
options. The two approaches are discussed separately in this chapter for clarity purposes only.
The top–down approach The top–down approach recognizes the weak adaptive capacity of the Navua population and similar island communities, and places the onus on the national government and its agencies to mainstream adaptation by developing a framework for national adaptation policy within which to implement, promote and support adaptations that help to make society more climate-proof. National level actions should include developing and improving regulation of climatesensitive sectors and geographic areas, investing in measures to reduce risks, and creating incentives for adaptation by local authorities, private entities and individuals. For example, incentives could be developed in consultation with local residents to enable them to afford flood-proof homes, take out insurance policies and relocate to sites that are less prone to flood risks. Climate proofing of building codes, tourism and land-use plans can yield immediate positive results. As previously noted, Fiji has in place a variety of forestry regulations, but they are not rigorously enforced and compliance is imperfect. The result is that forestry activities have negative environmental impacts that could be avoided, including siltation of rivers that aggravate flooding. Improved enforcement of the regulations by the national government would help to climate-proof downstream communities. Dredging of the river and proper maintenance of irrigation channels and floodgates would dampen the severity and extent of flooding in Navua. But these measures are beyond the financial capacity of the residents and require investment and action at the national level to bring the needed resources to bear. There is a national programme for dredging of rivers; however, dredging projects are implemented in an ad hoc manner and allocated funds can be diverted without proper consultation. The result is that some rivers are not dredged even several years after scheduled dates. The lack of adequate dredging equipment and faulty equipment are also partly responsible for delays, a situation that is prevalent in Pacific island countries for projects requiring specialized equipment. The government should revitalize the national dredging programme and prioritize and schedule dredging projects based on a comprehensive assessment of the vulnerability of communities and localities to flooding.
The bottom–up approach A bottom–up approach to adaptation is a community-based approach that engages local stakeholders to identify and prioritize risks, select appropriate responses, and implement selected measures using local institutions, resources and knowledge. To be effective, individuals and local institutions need to be encouraged with proper support and incentives by the government. The approach should be underpinned by recognition and use of the positive aspects of cultural and communal-based traditions of Pacific island societies. The approach also recognizes the need to engage all stakeholders, including
3232 J&J Climate Change
1/11/07
9:56 am
Page 273
Top–down, Bottom–up 273
non-government organizations (NGOs) and intergovernmental regional organizations such as SPREP and the University of the South Pacific. An important aspect of Pacific island communities, including those in Fiji, is the strong communal nature of living and working together, although at times this is also viewed by some as an obstacle to development (Duncun and Toatu, 2004). The communal nature of living in the Pacific, especially in the rural and urban hinterlands, brings forth an opportunity to pool and mobilize resources (for example, finance and local expertise) that are required for adaptation. Therefore, the adaptation process should be taken within the context of the community as a whole, although consideration should also be given to the needs of individuals. NGOs and regional organizations with mandates in line with reducing climate-related risks have played significant roles in community-based development and advocacy for climate change in the Pacific. Their experience with community-level development, provision of technical advice and carrying out research on climate change issues will complement efforts by the national government and individuals to promote climate change adaptation. The engagement of regional organizations has been proven to be effective in the sharing of adaptation lessons learned elsewhere in the Pacific and to engender a consolidated stand on climate-related issues in international forums. Autonomous adaptation has been observed within Navua town, especially in the construction of homes. Discussions held with officials of the Navua Rural Local Authority (RLA) indicated that the number of newly-approved buildings raised above the ground has been on the rise, especially since the recent floods of 2004. Residents were encouraged by Navua officials to build higher than the previous flood level. Such autonomous adaptation needs to be properly encouraged, as it will contribute to reducing the vulnerability of the Navua residents to flooding and reduce financial obligations of the national government during flooding disasters. Our assessment of the situation in Navua, and the results from the CBDAMPIC project, indicate that local communities should be actively engaged in the full adaptation process, from planning to implementing and monitoring adaptation measures. Their involvement in this process is important, whether the technical advice on adaptation to climate change originates from local or international experts (Nakalevu, 2005). This approach will also contribute to heightening the community’s responsibility to sustain adaptation to change and to proactively internalize the adaptation process. It is anticipated that by internalizing and sustaining the adaptation process, the communities’ dependence on external assistance to implement adaptation options will progressively reduce over time.
Challenges to Implementing Adaptation in Pacific Island Countries Four challenges to implementing adaptation to climate change in the Pacific are identified: (1) perceptions and competing government and individual priorities, (2) weak governance and institutional framework, (3) weak socioe-
3232 J&J Climate Change
1/11/07
9:56 am
Page 274
274 Climate Change and Adaptation
conomic conditions, and (4) lack of technical capacity. We elaborate on each of these below.
Perceptions and competing government and individual priorities Perceptions of the public and decision makers in the Pacific about climate change will influence the actions they take to deal with climate change risks. On the basis of our surveys, most Navau residents, local officials and national government officials have only a low level of awareness of climate change, in most cases influenced by media reports, which are seldom accurate. Only a few of them acknowledge the influence of human activities on the climate. This implies that many people are unable to perceive concrete links between climate change and the contribution humans make to aggravating climate change and variability. Consequently, when the implementation of climate change adaptation is advocated, it is often perceived as an attempt to prepare for a future ‘unlikely adversity’, which is not as pressing as the need to meet basic daily needs such as food and shelter. Climate change is often viewed as a futuristic phenomenon and does not align well with the decision timeframes of individuals and governments, which are invariably short, at 1–5 years for governments, depending on the duration of the national parliament. Consequently, the notion of adapting to climate change is seldom regarded as a high priority and thus loses out in terms of funding and institutional support. In some cases, such perceptions are reinforced by the limited climate change awareness. A study in the Cayman Islands of the Caribbean also showed that policymakers seldom regard climate change as a priority environmental concern and therefore see little need to make policy responses to cater for it (Tomkins and Hurlston, 2003). Perceptions that climate change is a distant and low-priority concern necessitate discussion of climate change adaptation in the context of climate variability and extreme weather events. People are better able to visualize the link between extreme weather events and climate variability and their livelihoods, and thus strategies for managing risks in this context. However, this approach to climate change discussion must be taken with care. Extreme weather events are frequent in Pacific islands and communities may perceive them as normal events. For example, Fiji is affected by an average of two tropical cyclones per year and numerous tropical depressions. Consequently, if human-caused climate change is associated with climate extremes, climate change may come to be regarded as remaining within the norms with which islanders presently have to cope. This can lead people to downgrade the importance of adaptation to climate change. The need for caution is also pertinent because national governments usually provide relief assistance during and after tropical cyclones and severe tropical depressions. Association of climate change with extremes could create expectations that individuals and communities should rely on government relief to cope with climate change. This could accentuate the local community’s dependence on the national governments while also dissuading national
3232 J&J Climate Change
1/11/07
9:56 am
Page 275
Top–down, Bottom–up 275
governments from actively participating in the adaptation process. Proper public awareness about climate change adaptation should aim to unravel the above misconceptions.
Institutional framework and governance Governments in most Pacific island countries are often challenged internally and externally to demonstrate good governance by establishing appropriate institutions with proper checks and balances to optimize the delivery of goods and services to the country as a whole. National governments can no longer afford to maintain rigid decision-making structures if they are to be effective and efficient in working towards the goal of enhancing the adaptive capacity of the population to climate change. The need to promote participatory approaches to planning and decision making in the context of climate change adaptation is pertinent to ensure the internalization and sustainability of the adaptation process. However, such changes by national governments towards participatory and decentralized decision making should be judiciously implemented with national interests at their core to avoid unnecessary delays and the continued dominance of decision making by a few stakeholders. Good governance is needed to enable climate change concerns to permeate all levels and sectors of the society, including the local communities. The case of Navua demonstrates some of the problems with institutional frameworks in Fiji that hinder complementary decision making at local and national levels. The government activities within Fiji are undertaken through four distinct systems: the National Government Administration, the Fijian Administration (which exclusively looks after indigenous Fijian affairs), the Municipal Administrations (incorporated towns and cities) and Rural Local Authorities (RLA). Navua has not been incorporated as a town under the Local Government Act and is therefore governed as a rural local authority. The RLAs are essentially public health authorities responsible for public health, building construction and other matters coming under the Public Health Act. However, most functions and services are consolidated on a national basis for efficiency and economy of scale, and as a result RLAs have relatively limited powers. Within this framework, as an RLA the residents of Navua do not have local-level political representatives, as is the case for incorporated town and city councils, although the Navua RLA officials work tirelessly to provide services and represent Navua residents with minimal financial and human resources. This ultimately means that local-level concerns about river flooding are often inadequately dealt with at the political and administrative levels. For example, the absence of a stronger political framework in Navua (Duncan and Toatu, 2004) made possible the easy diversion of funds earmarked for dredging in 2000 (Auditor General of the Republic of Fiji, 2001). Certain officials with the Navua RLA interviewed expressed their intention to legally incorporate Navua as a town as a means to have local-level political representation and improve the services and economic activities in Navua. As mentioned earlier, the national government needs to establish a
3232 J&J Climate Change
1/11/07
9:56 am
Page 276
276 Climate Change and Adaptation
top–down or national framework of climate change adaptation policy and planning within which bottom–up strategies can be implemented in localities such as Navua. To drive this process, the institutional and governance structures need to be reinvigorated and strengthened. Awareness needs to be raised within these structures of the significance of adapting to climate variability and extreme weather events as preparation for changing climate patterns. There is also a lack of communication and coordination between relevant government departments (Duncan and Toatu, 2004; Raj, 2004), an institutional setback apparent in many government departments in Pacific island countries. The fragmented jurisdictions over related areas reinforce the lack of communication in some circumstances. For example, in Fiji, the Land and Water Resources Management of the Ministry of Agriculture is responsible for river engineering, drainage and irrigation, while the Public Works Department is responsible for flood control, watershed management and flood forecasting. Although there is an amicable working relationship between the two government departments (Raj, 2004), regular communication on matters of mutual interest cannot be guaranteed as they go about their day-to-day operations. Furthermore, there is no central authority for flood management and the National Disaster Management Office only plays a coordinating role during disasters.
Weak socioeconomic conditions and lack of capacity Large adaptation projects are often costly, especially for socioeconomically disadvantaged communities in the Pacific, such as that in Navua. An average income of $US35–46 per week barely meets basic needs let alone affords floodproof homes, relocation or insurance. On the other hand, there are also certain individuals who are already implementing autonomous adaptation measures to the risks posed by flooding in Navua and coastal erosion and storm surges in Samoa (Nakalevu, 2005). The lack of capacity for climate science to evaluate changing climate trends and risks and to predict future changes is pervasive throughout Pacific island countries. The lack of capacity is also evident at the systemic and institutional levels and therefore affects the ability to properly plan and implement climate change adaptation within a development framework. Also lacking is the technical capacity to formally evaluate the potential performance, costs and impacts of adaptation measures. For example, under the CBDAMPIC project, the cost–benefit analysis of adaptation options identified in the project sites had to be contracted out to a consultant because of the lack of expertise at the national level to carry out such analysis.
Conclusions Climate-related disasters put a lot of strain on the sustainable livelihoods of communities in the Pacific islands. It is anticipated that with climate change, ongoing climate variability coupled with extreme weather events will increas-
3232 J&J Climate Change
1/11/07
9:56 am
Page 277
Top–down, Bottom–up 277
ingly threaten people’s livelihoods. A significant and growing risk to livelihoods in the township of Navua is river flooding. The climate driver of river flooding, rainfall, has not shown any significant increase in its pattern or intensity to suggest that it is the dominant driver of vulnerability in the study region. Instead, non-climatic drivers such as increased sedimentation in the river from logging, mining and other activities, degraded infrastructure such as irrigation channels and flood gates, failure to regularly dredge the river, and locating housing and other structures in flood-prone locations have been responsible for increasing the vulnerability of Navua residents to flooding. In the future, however, climatic factors could amplify the risks as sea level rises and, possibly, rainfall from tropical cyclones intensifies. Navua typifies local communities in the Pacific islands, which are locked in a vulnerable situation because of their poor socioeconomic conditions coupled with limited input to government decision-making processes and access to financial resources, expertise and technical knowledge. A way forward is to implement climate change adaptation embracing a connective top–down and bottom–up approach that provides a national framework for climate-proof development, and resources and incentives to enable local level action. Stakeholders from national to local levels, underpinned by lessons learned through experiences with climate variability and extreme weather events, should be involved in planning, implementing and monitoring adaptation measures.
References Auditor General of the Republic of Fiji (2001) ‘Report of the Auditor General of the Republic of Fiji Islands’, Suva, Fiji Ba, T. (1993) ‘Wai-Magiti Na Kena Iyau: Rivers and river basins in Fijian development’, Geography, University of the South Pacific, Suva, Fiji Campbell, J. (1999) ‘Vulnerability and social impacts of extreme events’, in PACCLIM Workshop: Modeling Climate Change and Sea Level Change Effects in Pacific Island Countries, International Global Change Institute, Waikato University, Hamilton, New Zealand, pp1–6 Central Division Disaster Management Council Operation Centre (2004) ‘Report of the floods encountered in the Central Division on the 8th and 15th April 2004’, Suva, Fiji Duncun, R. and T. Toatu (2004) Measuring Improvements in Governance in the Pacific Island Countries, Pacific Institute of Advanced Studies in Development and Governance, University of the South Pacific, Suva, Fiji Feresi, J., G. Kenny, N. Dewet, L. Limalevu, J. Bhusan and L. Ratukalou (eds) (2000) Climate Change and Vulnerability and Adaptation Assessment for Fiji, International Global Change Institute, Waikato University, Hamilton, New Zealand and Pacific Island Climate Change Assistance Programme, Government of Fiji, Suva, Fiji, p135 Fiji Meteorological Service (2004) The Climate of Fiji, Fiji Meteorological Service, Climate Division, Nadi, Fiji Fiji Times (2007) ‘Rehabilitation won’t cost $10m’, Fiji Times Online, www.fijitimes.com/story.aspx?id=57082, accessed 15 February 2007 IPCC (2007) ‘Summary for policymakers’, in S. Solomon, D. Qin, M. Manning, Z.
3232 J&J Climate Change
1/11/07
9:56 am
Page 278
278 Climate Change and Adaptation Chen, M. C. Marquis, K. Averyt, M. Tignor and H. L. Miller (eds) Climate Change 2007: The Physical Science Basis, contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York Koshy, K. (2007) ‘Modeling climate change impacts on Viti Levu (Fiji) and Aitutaki (Cook Islands)’, final report of AIACC Project No SIS09, International START Secretariat, Washington, DC, www.aiaccproject.ort/FinalReports/final_reports.html Mataki, M., M. Lal and K. Koshy (2006) ‘Baseline climatology of Viti Levu (Fiji) and current climatic trends’, Pacific Science, vol 60, pp49–68 McCarthy, J., O. Canziani, N. Leary, D. Dokken and K. White (2007) Climate Change 2001: Impacts, Adaptation and Vulnerability, contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York Mimura, N., L. Nurse, R. McLean, J. Agard, L. Briguglio, P. Lefale, R. Payet and G. Sem (2007) ‘Small Islands’, in M. Parry, O. Canziani, J. Palutikof and P. J. van der Linden (eds) Climate Change 2007: Impacts, Adaptation and Vulnerability, contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (forthcoming) Nakalevu, T. (2005) ‘CIDA/SPREP CBDAMPIC project in the Pacific’, paper presented at the Samoa Training Institute on Climate and Extreme Events, Apia, Samoa, Secretariat of the Pacific Regional Environment Programme, Apia, Samoa National Institute of Water and Atmospheric Research Ltd (2004) ‘The island climate update no 44’, www.niwa.co.nz/ncc/icu/2004-05/icu-2004-05.pdf/view_pdf, accessed 20 May 2005 Raj, R. (2004) ‘Integrated flood management (Case study: Fiji islands flood management – Rewa river basin)’, www.apfm.info/pdf/case_studies/cs_fiji.pdf, accessed 28 June 2006 Ruosteenoja, K., T. R. Carter, K. Jylha and H. Tuomenvirta (2003) Future Climate in World Regions: An Intercomparison of Model-based Projections for the new IPCC Emissions Scenarios, The Finnish Environment 644, Finnish Environment Institute, Helsinki, Finland Sinclair Knight Merz Pty Ltd (2000) Environmental Impact Assessment for the Navua River Mouth Dredging Project, Ministry of Agriculture, Fisheries and Forests, Suva, Fiji SOPAC (South Pacific Applied Geoscience Commission) (2003) Proceedings of the 4th National Fiji Consultations: Reducing Vulnerability of Pacific ACP States, South Pacific Applied Geoscience Commission, Suva, Fiji Tompkins, E. L. and L. A. Hurlston (2003) ‘Report to the Cayman Islands’ Government. Adaptation lessons learned from responding to tropical cyclones by the Cayman Islands’ Government, 1988–2002’, Tyndall Centre Working Paper No 35, Tyndall Centre for Climate Change Research, www.tyndall.ac.uk/publications/ working_papers/wp35.pdf, accessed 24 April 2005 World Bank (2000) Cities, Seas and Storms: Managing Pacific Island Economies: Vol IV, Adapting to Climate Change, East Asia and Pacific Region, Papua New Guinea, and Pacific Island Countries Management Unit, World Bank, Washington, DC, p48
3232 J&J Climate Change
1/11/07
9:56 am
Page 279
16
Adapting to Dengue Risk in the Caribbean Michael A. Taylor, Anthony Chen, Samuel Rawlins, Charmaine Heslop-Thomas, Dharmaratne Amarakoon, Wilma Bailey, Dave D. Chadee, Sherine Huntley, Cassandra Rhoden and Roxanne Stennett
Dengue – A Caribbean Health Problem Dengue fever is a potentially serious vector borne viral disease transmitted by the Aedes aegypti mosquito.1 It is generally associated with common viral symptoms but may progress to the more severe dengue haemorrhagic fever, which can prove fatal, particularly for children, young adults and the elderly. Dengue is endemic in several tropical and subtropical countries, particularly affecting Southeast Asia and the Western Pacific (WHO, 2002). In the Caribbean the existence of dengue has been reported for well over 200 years (Ehrekranz et al, 1971) with sharply decreasing intervals between subsequent epidemics since 1970. Four closely related viral strains (dengue serotypes 1–4) are responsible for the dengue epidemics in the Caribbean region either due to their introduction or reintroduction. For example, a devastating pandemic reported in 1977 and lasting until 1980 (PAHO, 1997) was caused by the reintroduction of the dengue-1 serotype. Similarly the dengue-4 strain emerged in this region in 1981, and currently all 4 serotypes coexist here (Rawlins, 1999)2, with the number of recorded cases continuously increasing since 1981 (Figure 16.1), drastically so since 1991. Between 1981 and 1996, 42,246 cases of dengue hemorrhagic fever and 582 deaths were reported by 25 countries in the Caribbean and wider Americas (CAREC, 1997). The various factors that influence dengue outbreaks in the region include the presence of the disease in a territory, the immunity of the population, and socioeconomic conditions that impact (among other things) on vulnerability and the ability of the disease to spread and do so quickly. The free movement of people between and beyond Caribbean borders also aids the spread of the disease. For example, the reintroduction of dengue-1 in the region in 1977 was
3232 J&J Climate Change
1/11/07
9:56 am
Page 280
280 Climate Change and Adaptation
Figure 16.1 Annual variability of the reported cases of dengue and the rate of change (increase or decrease from previous year) for the Caribbean
initially detected in Jamaica, from where it spread to virtually all other islands of the Caribbean. Presently, however, the factors that have emerged as the most significant are vector abundance and the frequency of vector–human contact. There is reason to believe that climate change is creating favourable conditions that are contributing to the increasing frequency of dengue epidemics in the region. Warmer temperatures result in shorter viral cycles within the vector, thus causing faster transmission to humans. This is aided by the increased vector abundance brought on either by excess rainfall (breeding sites resulting from pools of stagnant water) or by the lack of it (breeding sites resulting from stored water). Variability in these two climate stimuli, temperature and rainfall, probably accounts for some of the outbreak patterns seen since 1981 (Figure 16.1), and further changes in both variables, for example, a warmer wetter scenario, would similarly influence future outbreaks. Added to these risk factors is the fact that there is currently a deficit in the capacity of regional public health programmes to deal with the existing level of threat. A future increase in the frequency of dengue outbreaks due to climate change would only further overwhelm these already inadequate and resource-constrained programmes. Keeping these issues in mind, in this chapter we attempt to examine the potential for adaptation to the increased future risk of dengue fever due to climate change in the Caribbean. The dengue–climate link is first explored to determine whether experience from past outbreaks can establish the increased future threat of dengue epidemics. Thereafter, the effectiveness of current adaptive strategies to contain dengue at present risk levels and their ability to cope with the possible increased future risk is also evaluated. Finally, options for expanding, strengthening or combining existing programmes or imple-
3232 J&J Climate Change
1/11/07
9:56 am
Page 281
Adapting to Dengue Risk in the Caribbean 281
menting new ones to better respond to the challenge of increased risk, given the constraints of the Caribbean reality, are suggested.
Data and Methodology Dengue records were obtained from the Caribbean Epidemiological Center (CAREC). Climate data were obtained from the Workshop Database of the Climate Studies Group, Mona (CSGM), and consisted of daily and monthly station data (maximum and minimum temperatures and rainfall) for 27 Caribbean territories. Based on the available data the study period was selected as 1980–2001; climate and dengue incidence data from this period were then statistically analysed to indicate associations between the variables and to identify target countries for detailed analysis. Geographic and socioeconomic data and survey instruments were used to identify vulnerable subgroups within target countries and to isolate common characteristics among the subgroups. The identification and classification of potential adaptive strategies was done in collaboration with CAREC and the health ministries of the target countries; these were then assessed using matrix analysis. Next, a limited set of best practices was isolated with the help of expert judgement in order to recommend a course of action that can inform appropriate policy formulation to address the increased risk of dengue fever within the region.
Dengue and Climate – A Cause for Concern Our analysis indicates that there is a clear seasonality to dengue outbreaks within the Caribbean, with a tendency to peak towards the end of the year (see Figure 16.2; the pattern shown here, for Trinidad and Tobago, can be generalized for the entire region3). On average, outbreaks were observed to lag the second and principal maximum in rainfall by 3–4 weeks and the maximum in temperatures by 6–7 weeks. Similar observations have been made for Trinidad and Tobago (Campione-Piccardo et al, 2003; Wegbreit, 1997) and for Barbados (Depradine and Lovell, 2004). The climatological sequence is therefore as follows: warm temperatures (with the climatological maximum being attained) ➞ abundant rainfall (in most cases the late rainfall season has begun to end) ➞ dengue outbreaks. The association between dengue outbreaks and climate was confirmed by statistically significant lag correlations between weekly dengue cases and temperature and precipitation in given years, with the relationship with temperature being stronger than that with precipitation. Strong associations with temperature are not unexpected, as studies by Koopman et al (1991), Focks et al (1995) and Hales et al (1996) have all shown that there is a shortening of the extrinsic incubation period of the dengue virus within the mosquito with warmer temperatures (in other words a shortening of the interval between the acquisition of the virus by the mosquito and its transmission
3232 J&J Climate Change
1/11/07
9:56 am
Page 282
282 Climate Change and Adaptation
Figure 16.2 Monthly variability of the reported dengue cases, rainfall and temperature from 1996 to 2003 in Trinidad and Tobago Source: Chen et al (2007).
to other susceptible vertebrate hosts). An increase in the intake of blood meals by the vector has also been associated with warmer temperatures (McDonald, 1957), which means an increased vector–human contact. The linear associations with rainfall (though weaker) are also not unexpected since both excess or lack of rainfall can contribute to increased vector abundance due to either rainwater accumulation or water storage. Note, however, that despite being weaker, the link between rainfall availability and vector abundance should not be dismissed and rainfall variability must be monitored and/or accounted for in adaptation strategies. Besides seasonality, there are also changes on longer timescales, as well as an interannual variability in the number of dengue cases reported in the Caribbean. Reported cases increased dramatically in the 1990s, and there is marked variability from year to year superimposed on the increase. Stratifying dengue epidemics according to the El Niño Southern Oscillation (ENSO) phases reveals a bias towards El Niño or El Niño+1 years (Amarakoon et al, 2003); this is shown in Table 16.1. Similar links between upsurges of dengue fever and ENSO events have been noted for the South Pacific (Hales et al, 1996) and Colombia (including its Caribbean coast) (Poveda et al, 2000). The correlation between reported four-weekly cases and monthly temperatures was also found to be significant over the period of analysis, with a stronger correlation for El Niño and El Niño+1 years (correlations are found to be weaker and insignificant with rainfall).
3232 J&J Climate Change
1/11/07
9:56 am
Page 283
Adapting to Dengue Risk in the Caribbean 283
Table 16.1 Distribution of epidemic peaks among ENSO phases, 1980–2001 Region
Total
El Niño and El Niño+1
La Niña
Neutra
Caribbean Trinidad and Tobago Barbados Jamaica
8 8 6 5
7 6 5 4
– – – –
1 2 1 1
Source: Chen et al (2007).
We suggest that the link with ENSO events results from their impact on both the temperature and rainfall regimes of the Caribbean. The rainy season is drier and hotter during El Niño occurrences, with warmer temperatures continuing into the El Niño+1 year (Chen and Taylor, 2002; Taylor et al, 2002). The early part of the rainy season also tends to have wetter than normal conditions during El Niño+1 years (Taylor et al, 2002). These climatic factors create conditions conducive to the propagation of the dengue virus. Past weather records for the region show that average temperatures (maximum and minimum) have increased over the past 40 years (Peterson et al, 2002; Taylor et al, 2002), and this trend is projected to continue in the near future. Projections for 2080 suggest a Caribbean that is warmer by 2 or 3°C (Santer, 2001), with statistical downscaling confirming increases of this magnitude for Trinidad and Tobago, Jamaica and Barbados (Rhoden et al, 2005). Additionally, the El Niño phenomenon is also expected to respond to climate change (the exact manner of the response is still unclear), and some models indicate an increase in the frequency of ENSO events in a climate forced by future greenhouse warming (Timmermann et al, 1999). Both the projected temperature increases and the possible increase in El Niño frequency would impose additional stresses on an already fragile and vulnerable health system by favouring increased vector abundance and vector–human contact. This poses a serious cause for concern in the region and underscores the urgent need to develop appropriate policies and programmes to address this significant health threat.
Retirement, Pitfour and John’s Hall – Profiles of Vulnerable Communities In order to determine appropriate adaptation strategies to the increased threat of dengue fever, it is first important to identify the communities that are the most vulnerable. A local level study was conducted for a section of the parish of St James in western Jamaica, based on the pattern of dengue occurrences in 1998, to isolate the common characteristics of those most vulnerable to dengue fever (Heslop-Thomas et al, 2006). In that year, there was a concentration of cases within the parish capital city of Montego Bay and sporadic cases along a permanent stream with associated seasonal streams and gully banks. Three
3232 J&J Climate Change
1/11/07
9:56 am
Page 284
284 Climate Change and Adaptation
communities, Retirement, Pitfour and John’s Hall (Table 16.2) were selected along this hydrological feature, and a questionnaire was administered to a 10 per cent sample comprising heads of households. The questionnaire solicited information on socioeconomic conditions, support systems, knowledge of the disease and cultural practices that might have implications for the spread of dengue fever. Table 16.2 Socioeconomic characteristics of three communities in Western Jamaica and survey sample size Granville/Pitfour
Retirement
John’s Hall
No of households
1507
485
572
Infrastructure
Mixture of formal and informal structures (ratio 50:50)
Few informal dwellings
Rural squatters
Economic bracket
Low income Heads of families self-employed or in the service sector in Montego Bay
Lower middle income Heads of families in the service or public sector in Montego Bay
Poor Rural squatters; primarily female heads of household engaged in domestic service or petty trading
Sample/households
151
49
57
Survey responses were then used to construct a vulnerability index based on a number of indicators identified in the literature, which include group immunity; knowledge of symptoms and vectors of the disease; the use of protective measures; source of water and water storage devices; distance from the nearest health facility; chronic illnesses; and measures of resilience and stress – education, employment, income, female household headship, room densities, coping strategies and integration into the community. The index was determined for the community in general and for the surveyed households and subgroups consisting of the most and least vulnerable within each community. Some of the important findings from the assessment of index values and survey responses are as follows: • Seventy-eight per cent of the respondents felt that dengue control should be the responsibility of the government, in sharp contrast to the Jamaican Ministry of Health’s view that dengue control be a community responsibility. More than half the people interviewed in the communities were unaware of the causes of dengue fever, and the overwhelming majority had no knowledge of its symptoms. • The most vulnerable subgroups, comprising 14 per cent of the entire sample, had the largest representation in John’s Hall and Granville/Pitfour – the two communities with squatter settlements. Prominent characteristics of the most vulnerable group were that 95 per cent of the household heads
3232 J&J Climate Change
1/11/07
9:56 am
Page 285
Adapting to Dengue Risk in the Caribbean 285
had no skills, 87 per cent earned the minimum wage or less and 84 per cent had female heads. The least vulnerable were the employed urban dwellers (97 per cent), those who possessed marketable skills (78 per cent) and those who earned wages that were above the national minimum (81 per cent). • In the surveyed communities, 23 per cent had no access to piped water and water supply was irregular for all respondents. Fifty-four per cent of the respondents stored water in drums, which in most cases were uncovered to facilitate rainwater collection and/or easy access to the stored water. Such outdoor drums have been found to be the most productive Aedes mosquito breeding containers (Focks and Chadee, 1997). Significantly, 97 per cent of the least vulnerable subgroup had water piped into their dwellings. In summary, the study suggested that the most vulnerable to dengue fever were the poor, who lived in informal or squatter settlements and who, as a result, lacked basic community infrastructure, including access to piped water and adequate garbage disposal. The common practice of storing water in open drums creates breeding grounds for the mosquito vector. These settlements were also found to lack any community structure to facilitate collective action on any issue, including the threat of a dengue fever outbreak. The most vulnerable were also, to a large extent, unaware of the cause of dengue fever and how their actions might contribute to the disease and felt that it was the responsibility of the government to contain the disease or any threat of its outbreak.
Current Adaptation Strategies Assessed A marked similarity can be observed in the strategies adopted for handling dengue fever in the islands of the English-speaking Caribbean. This is probably due in part to a common heritage that has led to common social structures, the active involvement of the PAHO/WHO in supporting anti-dengue programmes in member countries and the work of CAREC in defining regional strategies. Because of the similarities, it is possible to generalize about the strategies used, highlighting, wherever necessary, the exceptions. In this study we have assessed the measures of the health ministries of Trinidad and Tobago and Jamaica since both countries had a high percentage of dengue cases in our data set and both have expended resources in the recent past on the management of the disease. The strategies generally fall into one of three categories: health promotion and education, surveillance, and vector adult control. Health promotion and education: This strategy generally targets the entire population but only after the risk is detected. Media resources, posters, printed flyers, booklets and, in limited instances, health education teams in outbreak communities are used for this purpose. Various messages regarding the identification and elimination of breeding sites, the description of symptoms, and disease treatment options are conveyed. However, despite the existence of such education and promotion units within both the Trinidad and Tobago and Jamaican health ministries, neither has staff specifically dedicated to dengue fever and nor are media resources employed before the occurrence of an
3232 J&J Climate Change
1/11/07
9:56 am
Page 286
286 Climate Change and Adaptation
outbreak. Effectiveness of promotion strategies is measured by the reduction/increase in reported cases rather than an assessment of knowledge gained or changed behavioural practices. The assumption is that the former implies the latter. Surveillance: As with education and promotion, the health ministries of both Jamaica and Trinidad and Tobago possess surveillance units staffed by government-employed public health inspectors. In Jamaica, staffing is deemed inadequate and there is no routine surveillance of communities and dwellings except in response to outbreaks or to directives from senior health officials. Trinidad and Tobago, on the other hand, employs in excess of 600 inspectors who conduct surveillance on a quarterly cycle, but these cycles are not necessarily synchronized with dengue outbreaks, nor do they help anticipate or reduce them. Surveillance programmes in both countries emphasize the identification and treatment of breeding sites, as well as recording epidemiologically important indicators such as vector abundance – adult, larval or pupal – particularly in regions with recorded or suspected cases. Control of adult mosquitoes: This is carried out on request or directive and involves the use of ultra low-volume (ULV) or thermal fog sprays of an appropriate insecticide. Although the consensus of the health ministries is that its effectiveness is short term (a few days), it is an often demanded solution to intolerable adult mosquito levels. Fogging is an expensive exercise and is limited to areas where dengue outbreaks have occurred or where risk is high because of the abundance of adult mosquitoes. Other vector adult control practices are usually initiated by individuals, largely to alleviate the nuisance of mosquito bites. These include the use of repellents to reduce vector–human contact or the use of physical methods such as the installation of mesh screens on buildings. In Jamaica, there is no sustained programme aimed at the reduction of larval or pupal abundance, although on request there is limited chemical control with Abate granules. In Trinidad and Tobago, inspectors from the surveillance unit carry a supply of temephos (Abate) for the active treatment of breeding places on discovery. As is obvious, a primary weakness of current adaptation strategies is that they are reactive rather than anticipatory. In general, reactive strategies fail to engender long-term behavioural change and instead help to institutionalize the idea of the government being solely responsible as opposed to fostering community or personal responsibility. Even the best education programmes therefore fail to translate into sustained dengue preventive actions at the community level since their suggested strategies for vector control are merely viewed as necessary only when mosquito levels become intolerable. The general lack of anticipatory action suggests that the Caribbean population has developed a high tolerance for mosquitoes and mosquito bites, and any existing riskreducing activities therefore lose some of their inherent effectiveness because they are often deployed late in the vector and/or viral development cycle. Cuts in budgetary allocations to the health ministries and an overall resource problem are the primary reasons for the reactive approach to dengue control and prevention. In Jamaica, dengue fever is classified as a Class 2 disease and is given significantly less priority than Class 1 diseases, especially
3232 J&J Climate Change
1/11/07
9:56 am
Page 287
Adapting to Dengue Risk in the Caribbean 287
HIV/AIDS (Class 1 diseases are those which must be reported immediately to the Ministry of Health and include HIV/AIDS, malaria and diseases preventable by immunization; Class 2 diseases are reportable weekly in line listing). There is correspondingly a lack of adequate manpower and facilities to tackle the problem on an ongoing basis; for example, there is only one understaffed virology lab in Jamaica and excess samples must therefore be sent to Trinidad and Tobago, resulting in delays in outbreak identification. Despite the financial constraints, the Jamaican health system does possess a strong network, which can be mobilized in emergency situations. Primary health care is well organized and based on a nested system of health centres offering different levels of care. There is a health centre within five miles of every community, and the decentralization of health services has resulted in the division of the island into four health regions. This enhances the delivery of primary care by allowing autonomy in meeting the identified health needs of a region. There is, therefore, greater sensitivity to local needs and the potential for greater responsiveness in the event of epidemic outbreaks. In summary, the current adaptive strategies are clearly ineffective in addressing the needs of the most vulnerable. Limited financial resources constrain the ability of these strategies to address the prevention of dengue as opposed to providing a reactive response after an outbreak has occurred. They are also unable to rally or facilitate collective action, nor can they assist in transferring responsibility from governments to communities. The relatively well-organized health system is the only plus, and though it does not serve as an adaptive strategy, it can allow for targeted resource allocation to the affected community’s health centre in the event of an epidemic and ensure some level of primary care to all affected.
Assessment of Potential Future Adaptation Options Table 16.3 presents a matrix of possible adaptation options available for coping with an increased threat of dengue fever. The methods listed include those currently employed in the Caribbean region (as discussed in the previous section), other options practised elsewhere in the world or on a very limited scale within the region, and options that present themselves as future (though not too distant) possibilities, as a result of ongoing research in the region. These options are assessed by six characteristics rated high, medium and low. For example, cost is a serious adaptive constraint and so each proposed adaptation option is rated on the likely cost of implementation within the context of the Caribbean region. The assessments are a best guess (expert opinion) and are guided by the views and knowledge of the region’s environmental health officers. The six assessment characteristics are: 1 cost of implementation; 2 effectiveness (as measured by its long-term ability to reduce risk or address vulnerability); 3 social acceptability;
3232 J&J Climate Change
1/11/07
9:56 am
Page 288
288 Climate Change and Adaptation
Table 16.3 Adaptation strategies matrix
Cost
Effectiveness
Social Acceptability
Friendly for Environment
Neighbour Effects
Technical Challenges and Socioeconomic Change
Score
Measures
H
L
L
L
L
H
6
2. Education (disease symptoms, sanitizing the environment)
M
M
H
H
H
M
24
3. Surveillance for vector or larval/ pupal control
H
M
M
M
M
L
18
H
M
M
M
L
L
16
2. Community education and involvement
M
H
H
H
H
M
26
3. Chemical control
H
M
M
L
M
L
16
4. Biological control
H
H
M
H
M
M
20
Short term 1. Adulticide (ULV or thermal fog sprays) in truck or air
Long term 1. Surveillance for vector or larval/ pupal control and environmental sanitation
5. Adult control – Mesh windows
M
H
H
H
H
H
24
– Personal protection
M
M
M
M
M
H
16
6. Use of physical control; low-cost secure drums
H
H
M
H
H
H
20
7. Granting security of tenure to squatters
H
H
H
M
H
H
20
8. Early warning system
M
H
H
H
H
H
24
Note: Columns 2 through 7 indicate assessment criteria. Column 8 gives a composite score based on the ranking in columns 2–7.
3232 J&J Climate Change
1/11/07
9:56 am
Page 289
Adapting to Dengue Risk in the Caribbean 289
4 environmental friendliness; 5 promotion of neighbourliness; and 6 technical and/or socioeconomic challenges to implementation. A simple composite score is offered in the final column for comparison purposes. In compiling the score, high is given a score of 5, medium a score of 3 and low a score of 1, except for categories 1 and 6, in which the scoring allocation is reversed. The maximum possible score is 30. The adaptation strategies are again categorized under the headings of health education and promotion, surveillance and adult vector control. They are also divided into short-term and long-term practices, in other words, by whether their intent is to immediately alleviate the threat associated with dengue fever or to do so gradually. The short-term strategies are those currently adopted in the region (see previous section), namely public education aimed at encouraging individuals to identify and eliminate current breeding sites and to identify dengue symptoms; surveillance in outbreak communities for the purpose of environmental sanitization; and adult mosquito control with an appropriate insecticide (fogging). Of the three, public education received the highest composite score, whereas adulticidal fogging received the lowest score. Education benefits from the fact that in the present framework it is generally medium to high ranked in each category. Its effectiveness is medium ranked because of the seasonal nature of the campaign, while the presence of established units to handle education accounts for the medium (as opposed to high) ranking with respect to cost and technical challenges. Insecticidal fogging, though often demanded and practised, received a low score because of limited long-term effectiveness, inability to promote neighbourliness (people shut their windows during fogging) and limited social acceptability (the commonly used insecticide, malathion, has an unpleasant odour and needs specialized equipment for distribution). Of the long-term strategies assessed, the education strategies again achieved the highest composite ranking (though only marginally), with the focus on sustained campaigns aimed at community education (as opposed to targeting individual behavioural practices) and community involvement. Chemical control, surveillance practices and strategies relying on individuals to personally protect themselves received the lowest scores. Surveillance as a long-term approach does not engender neighbourliness (general suspicion) while the best personal protective measures come at a cost, thereby limiting their possible use by the poorest, who are the most vulnerable. Generally, however, most strategies fall in the mid range of scores (16–24), suggesting that relative advantages in one area are offset by disadvantages in another. Physical control of mosquitoes via the use of low-cost covered drums would address vulnerability issues associated with water storage, but such drums or drum covers are yet to be designed and would have to be subsidized or made available free to the most vulnerable. Even then, much would depend on householders being vigilant in covering containers. Granting security of tenure to squatters would promote community structure and increase the
3232 J&J Climate Change
1/11/07
9:56 am
Page 290
290 Climate Change and Adaptation
possibility of the eventual implementation of appropriate infrastructure for regular water supply. Such a move, however, is costly and fraught with social tensions. Biological control, for example, using fish to control mosquito larvae, is an environmentally friendly option, but is not suited for community practice unless the community could be persuaded of the benefits of proper implementation. Finally, an early warning system for action would require the coordination of a number of agencies (for example, climate research and monitoring agencies and health ministries) and necessitate the development of appropriate thresholds for action and coordinated action plans.
Best Practice Recommendations At present there is no single ‘best’ adaptation measure to counteract the threat of increasing dengue fever within the Caribbean. As suggested by Table 16.3, the various strategies have their relative merits and demerits. In light of this, we offer three potential options for tackling the dengue adaptation problem. Each option represents a combination of selected strategies outlined in Table 16.3, with due consideration given to their relative strengths and weaknesses. The options also give primacy to the need to address vulnerability issues, namely the lack of knowledge about dengue fever, the lack of community structure to facilitate collective action and the issues of water storage. The options are in increasing order of human and economic investment required and all assume that the currently practised strategies (outlined above) are at least maintained.
Option 1 – Refocusing current strategies Option 1 advocates that currently used strategies at least be maintained at their present level of activity and funding, but that approaches to them be refocused and relatively minor modifications be made. The key emphasis would be on education with a leaning towards personal and community education derived from the environmental sanitation and vector control strategies proposed in the campaign. This is as opposed to merely providing information about the disease and the steps to be taken to reduce mosquito abundance. A proposed modification would also be to engage communities before the rainy season through organized activities in local churches, schools, and youth and service clubs, and using competitions to test knowledge and community cleanliness. Involvement before dengue onset would promote long-term behavioural change (not restricted to the dengue season) and community responsibility. Behavioural change strategies have been advocated by the WHO as an important mechanism for prevention and control of dengue fever (WHO, 2000) and community mobilization has shown recent success in reducing cases of dengue fever during outbreaks (WHO, 2005). Vector surveillance in its current form would provide added support for the educational activities, particularly approaching the dengue season. Option 1 calls for the least additional investment, though a capacity
3232 J&J Climate Change
1/11/07
9:56 am
Page 291
Adapting to Dengue Risk in the Caribbean 291
upgrade of the education and promotion units of the health ministries would be required to initiate and sustain activities outside the dengue season. The possibility of cost sharing with the engaged community groups should also be explored.
Option 2 – Plus proper water storage Option 1 does not address the vulnerability issues surrounding proper water storage. The proposed adaptation strategies for this purpose given in Table 16.3 (design of drums and covers and security of tenure) are, however, costly; consequently, option 2 would require an even greater investment by the ministries of health. In option 2, the refocusing actions of option 1 are still undertaken, as they address education deficiencies and community involvement and responsibility, which are the two identified characteristics of the most vulnerable. In addition, however, the design of a suitable low-cost water storage drum or drum cover would be actively pursued. Currently, water is stored in discarded oil drums that are left open to catch rainwater runoff from roofs, which encourages mosquito abundance by creating breeding places. A covered low-cost unit which allows water in and whose cover is easily removable but secure, or from which water can easily be otherwise removed, would be ideal. There is also the option to design only a drum cover that meets the above requirements since the commonly used storage drums are fairly standard in size. Such storage units/ covers do not currently exist and might be costly to design and manufacture with little guarantee of their eventual use by the community. To ensure the latter, incentives would have to be offered (for example, in the form of subsidies) and an intensive public education campaign emphasizing the value of covered drums/drum covers would have to be undertaken. Incentives may also have to be offered to ensure the appropriate use of drum covers and efforts would have to be made to ensure that other habitats are made vector-free.
Option 3 – Plus an early warning system Like option 1, an early warning system has the advantage of anticipatory action. However, whereas option 1 promotes education simply based on the knowledge that there is a dengue season, an early warning system attempts to gauge the severity of any possible outbreak. Consequently, responses can be appropriately tailored based on the anticipated level of threat. Option 3 therefore proposes the actions of option 1 but coupled with an early warning system. An example of the structure of a simple early warning system is given in Figure 16.3 and would involve multi-sectoral cooperation. Monitoring of climatic indices would be undertaken by the meteorological services, the regional universities and/or the regional climate research institutes. Monitoring would involve the continuous close tracking of dengue-related temperature indices (for example the MAT index discussed in Chapter 2 of Chen et al, 2007), especially immediately following the onset of El Niño, when the risk of a dengue epidemic is greatest. If climate (tempera-
3232 J&J Climate Change
1/11/07
9:56 am
Page 292
292 Climate Change and Adaptation
Figure 16.3 Schematic of a possible early warning system
ture) thresholds are being quickly approached, then an alert or watch would be issued. On this basis, epidemiological surveillance by the ministries of health would be initiated or its frequency altered, and the education campaign tailored to meet the perceived level of threat. If the surveillance data confirm the presence of the pathogen or an increase in its abundance and if climate thresholds are exceeded then subsequent warnings would be issued, as needed. One benefit of this multi-staged early warning approach is its cost effectiveness since continuous monitoring of climate threshold is less costly than continuous epidemiology surveillance. With this approach response plans can be gradually ramped up (for example, by the inclusion of other strategies such as chemical or biological control) as forecast certainty increases, allowing public health officials several opportunities to weigh the costs of response actions against the risk posed to the public. The implementation of option 3, however, requires a memorandum of understanding between the cooperating institutions, a definition of roles, a focal point, some investment in research and the possibility of staging a pilot project. A similar framework for a health early warning system has been provided in US National Research Council (2001).
Conclusions
3232 J&J Climate Change
1/11/07
9:56 am
Page 293
Adapting to Dengue Risk in the Caribbean 293
There has been a significant increase in dengue cases in the Caribbean since 1991, and presently there is concurrent circulation of all four dengue serotypes in the region. The increase can be linked to climate since both the abundance of vectors and the transmission rate are modulated by temperature and rainfall. A marked seasonality in dengue outbreaks has been observed and extreme changes in climate stimuli (for example, as occur during an El Niño or El Niño+1 year) also appear to increase the risk of severe outbreaks. Current adaptation strategies within the region are limited as they are reactive rather than anticipatory, primarily due to cost constraints, and give priority to reducing vector abundance over reducing vector–human contact. Though both strategy foci are important, the current strategies also fail to address the needs of the poor, who are the most vulnerable to the disease. Water storage is a critical issue for this vulnerable population, which lacks both an adequate knowledge of the disease and a sense of neighbourliness that can promote community action. Adaptive strategies that target these issues, particularly appropriate water storage that discourages vector breeding, would be ideal, though not necessarily easy or practical to implement within the Caribbean context. Admittedly, however, even if proper water storage facilities were provided, there would still remain the uphill task of convincing people to remove the additional breeding sites in the vicinity of residences, including pools of water collected in old tyres, in plant pot bases and in other domestic receptacles lying around. Consequently, of the three adaptation options offered in the previous section, the implementation of an early warning system (option 3) might hold significant potential. As its basis, option 3 suggests that it might be easier to achieve behavioural change during an emergency than its induction as an everyday practice. Because it is possible to predict the likelihood of outbreaks on the basis of climatic changes, for example, El Niño events, public health authorities could try issuing emergency warnings and urging prompt action during such critical periods that are favourable to vectors. If this temporary behaviour change is successful, it may eventually become widespread and permanent. Over time, option 3 could then facilitate the transfer of information, the transformation of behavioural practices, (hopefully) the engendering of community spirit and action, and the gradual shift of responsibility for alleviating the dengue threat to government–community partnerships. Of course, the meteorological services and public health authorities must first be persuaded to cooperate in operationalizing this option. We finally note again that even at current levels of threat, there is an inability to cope, and reported cases are continuing to increase. Inaction is therefore not an option for the Caribbean, especially in the light of increased future threat due to anticipated climate change.
3232 J&J Climate Change
1/11/07
9:56 am
Page 294
294 Climate Change and Adaptation
Notes 1
2 3
The Aedes egypti mosquito acquires the dengue virus when it bites a sick person. The virus then undergoes a 8–10 day period of incubation within the mosquito after which it is can be transmitted to any other human the mosquito bites for the purpose of obtaining a blood meal. A mosquito carrying the dengue virus is capable of transmitting it for life (WHO, 2002). Dengue 2 and dengue 3 were previously introduced during the early 1950s (Downs, 1964) and 1960s (Ehrekranz et al, 1971) respectively. The Trinidad data also show a bimodality to the dengue case pattern, which, however, is not consistent in other territories.
References Amarakoon, D., A. A. Chen and M. A. Taylor (2003) ‘Climate variability and patterns of dengue in the Caribbean’, AIACC Notes, vol 2, no 2, p8, Assessment of Impacts and Adaptation to Climate Change, International START Secretariat, Washington DC Campione-Piccardo, J., M. Ruben, H. Vaughan and V. Morris-Glasgow (2003) ‘Dengue viruses in the Caribbean. Twenty years of dengue virus isolates from the Caribbean Epidemiology Centre’, West Indian Medical Journal, vol 52, pp191–198 CAREC (1997) ‘Epinote: An update of dengue fever in the Caribbean’, Caribbean Epidemiological Centre, Trinidad, West Indies Chen, A. A. and M. A. Taylor (2002) ‘Investigating the link between early season Caribbean rainfall and the El Niño +1 year’, International Journal of Climatology, vol 22, pp87–106 Chen A. A., D. D. Chadee and S. C. Rawlins (eds) (2007) ‘Climate change impact on dengue: The Caribbean experience’, Climate Studies Group, Mona, Kingston, Jamaica Depradine, C. A. and E. H. Lovell (2004) ‘Climatological variables and the incidence of dengue disease in Barbados’, International Journal of Environmental Health Research, vol 14, pp429–441 Downs, W. (1964) ‘Immunity patterns produced by arthropod borne viruses in the Caribbean area’, Annals of the Institute of Tropical Medicine, (Lisbon), vol 16, no 9, pp88–100 Ehrenkranz, N. J. et al (1971) ‘Pandemic dengue in Caribbean countries and the southern United States – Past, present, and potential problems’, New England Journal of Medicine, vol 285, pp1460–1469 Focks, D. A., E. Daniels, D. G. Haile and L. E. Deesling (1995) ‘A simulation model of the epidemiology of urban dengue fever: Literature analysis, model development, preliminary validation, and samples of simulation results’, American Journal of Tropical Medicine and Hygiene, vol 53, pp489–506 Focks, D. A. and D. D. Chadee (1997) ‘Pupal survey: An epidemiological significant surveillance method of Aedes aegypti: An example using data from Trinidad’, American Journal of Tropical Medicine and Hygiene, vol 56, pp159–167 Hales, S., P. Weinstein and A. Woodward (1996) ‘Dengue fever in the South Pacific: Driven by El Niño Southern Oscillation?’ Lancet, vol 348, pp1664–1665 Heslop-Thomas, C., W. Bailey, D. Amarakoon. A. Chen, S. Rawlins, D. Chadee, R. Crosbourne, A. Owino, K. Polson, C. Rhoden, R. Stennett and M. A. Taylor (2006) ‘Vulnerability to dengue fever in Jamaica’, AIACC Working Paper No 27, International START Secretariat, Washington, DC, www.aiaccproject.org Koopman, J. S., D. R. Prevots, M. A. V. Marin, H. G. Dantes, M. L. Z. Aquino, I. M.
3232 J&J Climate Change
1/11/07
9:56 am
Page 295
Adapting to Dengue Risk in the Caribbean 295 Longini Jr and J. S. Amor (1991) ‘Determinants and predictors of dengue infection in Mexico’, American Journal of Epidemiology, vol 133, pp1168–1178 McDonald, G. (1957) The Epidemiology and Control of Malaria, Oxford University Press, London National Research Council (2001) Under the Weather: Climate, Ecosystems and Infectious Disease, Commission on Geosciences, Environment and Resources, Board on Atmospheric Sciences and Climate, National Research Council, National Academy Press, Washington, DC, US PAHO (Pan American Health Organization) (1997) ‘Re-emergence of dengue and dengue hemorrhagic fever in the Americas’, Dengue Bulletin, vol 21, pp4–10. Peterson, T. C., M. A. Taylor, R. Demeritte, D. Duncombe, S. Burton, F. Thompson, A. Porter, M. Mercedes, E. Villegas, R. Semexant Fils, A. Klein Tank, R. Warner, A. Joyette, W. Mills, L. Alexander, B. Gleason (2002) ‘Recent changes in climate extremes in the Caribbean region’, Journal of Geophysical Research, vol 107(D21):4601, doi:10.1029/2002JD002251 Poveda, G., N. E. Graham, P. R. Epstein, W. Rojas, M. L.Quiñones, I. D. Valez and W. J. M. Martens (2000) ‘Climate and ENSO variability associated with vector-borne diseases in Colombia’, in H. F. Diaz and V. Markgrtaf (eds) El Niño and the Southern Oscillation, Cambridge University Press, New York Rawlins, S. (1999) ‘Emerging and re-emerging vector-borne diseases in the Caribbean region’, West Indian Medical Journal, vol 48, pp252–253 Rhoden, C. L., A. A. Chen and M. A. Taylor (2005) ‘Scenario generation of precipitation and temperature for the Caribbean’, in Proceedings of the 7th Conference of the Faculty of Pure and Applied Sciences, Kingston, Jamaica Santer, B. D. (2001) ‘Projections of climate change in the Caribbean basin from global circulation models’, in Proceedings of the Workshop on Enhancing Caribbean Climate Data Collection and Processing Capability, Kingston, Jamaica Taylor, M. A., D. B. Enfield and A. A. Chen (2002) ‘The influence of the tropical Atlantic vs the tropical Pacific on Caribbean rainfall’, Journal of Geophysical Research, 107(C9):3127, doi:10.1029/2001JC001097 Timmermann, A., J. Oberhuber, A. Bacher, M. Esch, M. Latif and E. Roeckner (1999) ‘Increased El Niño frequency in a climate model forced by future greenhouse warming, Nature, vol 398, pp694–697 Wegbreit, J. (1997) ‘The possible effects of temperature and precipitation on dengue morbidity in Trinidad and Tobago: A retrospective longitudinal study’, PhD thesis, University of Michigan, Ann Arbor, MI WHO (2000) ‘Strengthening implementation of the global strategy for dengue fever/dengue haemorrhagic fever prevention and control’, report of informal consultation, 18–20 October 1999, World Health Organization, Geneva, Switzerland WHO (2002) ‘Dengue and dengue haemorrhagic fever’, World Health Organization Fact Sheet No 117, revised April 2002, available at www.who.int/mediacentre/ factsheets/fs117/en/ WHO (2005) ‘Dengue outbreak in Timor-Leste’, Communicable Disease Newsletter, vol 2, no 1, p12
3232 J&J Climate Change
1/11/07
9:56 am
Page 296
17
Adaptation to Climate Trends: Lessons from the Argentine Experience Vicente Barros
Introduction Adaptation to climate change is not usually a priority issue in most countries; this has largely been attributed to a lack of adequate and unambiguous information on the specific future impacts of climate change, especially at regional and local scales. In developing countries, adaptation is further hindered by the relative absence of social and institutional practices of long-range planning. The only exceptions are cases where there is some certainty about the direction of future trends; for example, sea level rise. In contrast, the issue of climate variability is better understood and so far much scientific research has been directed towards understanding adaptation to climate variability in the hopes of drawing important lessons that can inform adaptation to climate change in the future. Another aspect of the global climate system that is even less well understood is that of long-term climate trends (of 20 to 50 years) that have in the past substantially altered the climate in some regions of the world. These trends could be the result of either global warming or interdecadal natural variability or both. Attribution of these trends to a specific cause is a complex issue because they may also be influenced by many natural or local human-driven processes. However, irrespective of their relation to global climate change, these trends do produce important social and economic impacts, adaptation strategies for which could also provide important insights into adapting to future climate change. Since such trends result in entirely new climatic conditions, adapting to them is much more complex than adapting to interannual climate variability. Sometimes it may be decades before there is public or even technical awareness of such trends and their associated costs or opportunities, while in other cases responses may be relatively fast. Over most of southern South America, there were important long-term trends in precipitation during the second half of the last century (Barros et al, 2000), which were also reflected in the trends in mean river streamflows (Camilloni and Barros, 2003). These trends were positive on the eastern side of the continent and negative on the Andes Cordillera and west of it (Giorgi, 2003). As a result, there are a number of areas and socioeconomic systems in
3232 J&J Climate Change
1/11/07
9:56 am
Page 297
Adaptation to Climate Trends: Lessons from the Argentine Experience 297
southern South America that are now under entirely new climatic conditions. This region therefore offers very recent experiences in adaptation to such longterm climate trends and could potentially provide valuable lessons for adapting to the impacts of climate change in the future. In this context, this chapter analyses five such experiences: the rapid adaptation of the agriculture system; the increasing frequency of floods along the flood valleys of the great rivers of eastern Argentina (Camilloni. and Barros, 2003; Barros et al, 2004); the increasing frequency of extreme precipitation events in the central and eastern part of that country (Re et al, 2006); the decreasing trend in the discharges of rivers fed by the melting of snow and ice from the Andes and rainfall in central Chile; and the recurrent floods on the coastal zones of the Río de la Plata caused by wind storms (Barros et al, 2003). These five case studies vary in terms of climate aspects, socioeconomic systems, and responses to the new climate conditions or climate threats. However, all of them indicate that awareness about the climatic changes plays a central role in the adaptation process. They also highlight the role of scientific knowledge in enabling timely and effective adaptation by building awareness and guiding appropriate responses. An in-depth discussion on each of these case studies is offered in the following sections, categorized on the basis of the nature of responses to climate trends; the nature of the trends themselves; and changing attitudes in adaptive responses.
Autonomous and Planned Adaptation Responses to Current Long-Term Trends Autonomous adaptation is usually triggered by climate-driven changes in natural or human systems (McCarthy et al, 2001). It is typically the result of reactive responses to current climate impacts (rather than preventive measures) by individuals or groups acting independently that could nonetheless result in large cumulative effects. On the other hand, planned or anticipatory adaptation takes place before impacts of climate change are observed and is also referred to as proactive adaptation (McCarthy et al, 2001). This adaptation is necessarily based on scientific knowledge that anticipates how changes in climate will evolve and is usually undertaken under collective and organized social policies or measures, more likely directed by a government agency or by a large public or private organization. An example of each of these adaptation types is offered in the two case studies that follow.
A case of autonomous adaptation: The expansion of the agriculture frontier Southern South America, east of the Andes Cordillera, is one of the regions of the world that has seen the largest positive trend in mean annual precipitation during the 20th century (Giorgi, 2003), especially after 1960 (Figure 17.1).
3232 J&J Climate Change
1/11/07
9:56 am
Page 298
298 Climate Change and Adaptation
Figure 17.1 Linear trends of annual precipitation (mm/year), 1959–2003
Until recently, technical knowledge about such long-term climate trends was quite modest (Castañeda and Barrros, 1994; Barros et al, 2000), with almost no public awareness. As a result, adaptation responses to the changing precipitation trend in southern South America have so far primarily been reactive in nature and limited to the agricultural sector, where the effects of these changes could be more easily perceived and managed. The increased precipitation therefore prompted a rapid expansion of agriculture into the former semiarid areas of this region, towards the west of what was known as the humid plains, aided by new technologies in crop production. Figure 17.2 shows this increased precipitation zone as a 100-km westward shift of the isohyets that are considered the boundary of extensive agriculture, namely those of 600mm annual rainfall in the south and 800mm in the north. This positive trend in precipitation was first observed in the provinces of Buenos Aires, La Pampa and Córdoba, south of subtropical Argentina, in the 1960s (Barros et al, 2000). In response, there was significant agricultural extension in the provinces of La Pampa and Córdoba during the 1970s and in Buenos Aires province there was an increase in cultivated area from 8.3 to 9.6 million hectares between 1971/1972 and 1982/1983. Between 1992 and 2004, the total cultivated area in Argentina expanded from about 20 million hectares to about 29 million hectares, 4.6 million hectares of the newly added 8.7 million hectares being located in the provinces of La Pampa, Córdoba, Santiago del Estero and Chaco, in other words in the western and northwestern border of the traditional humid plains. Much of the change in the La Pampa region happened prior to 1982/1983, but in the Córdoba, Santiago del Estero and Chaco provinces a steady increase in cultivated area continued,
3232 J&J Climate Change
1/11/07
9:56 am
Page 299
Adaptation to Climate Trends: Lessons from the Argentine Experience 299
Figure 17.2 Isohyets in mm: 1950–1969 (solid line) and 1980–1999 (dashed line)
with soybean being the primary crop of choice. These newly extended agricultural areas also saw an increase in crop yields, which were partly due to the positive precipitation trends and partly due to the geographic extension of the growing area (Magrin et al, 2005). Table 17.1 shows the cultivated areas for these four provinces. Table 17.1 Cultivated areas 1971/1972 La Pampa Córdoba S. del Estero Chaco Total four provinces Country Share of the four provinces in %
1.6 3.5 0.3 0.7 6.6 20.1 32.8
1982/1983 2.2 4.6 0.3 0.7 7.8 23.9 32.6
1992/1993 2.1 3.9 0.3 0.7 7.0 20.2 34.7
2003/2004 1.9 6.6 1.2 1.5 11.2 28.8 38.9
Note: Values are in millions of hectares.
Initially, crops other than soybean were cultivated, but from the 1970s onwards much of these new agricultural areas were taken up by soybean farming. Besides the positive trend in precipitation between the second half of the 1970s and the 1990s that made soybean cultivation possible in the former semiarid regions, other factors such as favourable international prices and new techno-
3232 J&J Climate Change
1/11/07
9:56 am
Page 300
300 Climate Change and Adaptation
logical packages that included minimum or no tilling practices contributed to the selection of soybean as the crop of choice. This adaptation of agriculture to changing climatic trends was entirely autonomous and not planned by the government or any organization. It resulted from a large number of individual decisions that were taken even before technical specialists became aware that new climate conditions allowed successful crop production in lands that previously did not support farming. The lag time prior to this adaptation or adjustment was about one decade, this being the period before the farmers realized that the new climate conditions were persistent. The economic returns from the agricultural expansion were, however, not always positive. In northern Argentina, some farmers suffered severe losses after switching from livestock to crop farming because there has not only been an increase in the mean annual precipitation in this region over the past 30–40 years, but also an increase in its interannual variability. Figure 17.3 shows this increase in interannual variability as the intensification of the rate between the interannual standard deviation and the mean of annual precipitation. This climatic trend has thus simultaneously increased the vulnerability of farming, especially in the case of modern agriculture, where the costs of inputs are considerable. While the rise in the mean annual precipitation was rapidly noticed by farmers, the increasing interannual variability was not always perceived, leading to huge losses in some cases.
Figure 17.3 Percentage change in the rate between standard deviation and mean value in the 1980–1999 period with respect to 1950–1969
3232 J&J Climate Change
1/11/07
9:56 am
Page 301
Adaptation to Climate Trends: Lessons from the Argentine Experience 301
The financial benefits to the farmers were additionally constrained by the lack of an adequate network of rural roads in the region, which prevented them from fully exploiting the newly available cultivable area under the new climatic trends. Table 17.2 shows the low density of rural roads in the meridional axis of the expansion of the agriculture frontier that runs from La Pampa in the south, through Córdoba, to Santiago del Estero and Chaco in the north. The lack of roads has deterred the further expansion of agriculture in these provinces despite favourable climatic conditions. Table 17.2 Density of rural roads in six provinces Province
Density of Rural Roads (km/km2)
Santa Fe Buenos Aires Córdoba Chaco La Pampa Santiago del Estero
0.90 0.48 0.34 0.22 0.16 0.11
Source: Adapted from Escofet and Menendez (2006).
Besides the economic issues, the rapid expansion of agriculture has also negatively impacted the natural environment in certain regions. Deforestation to convert land to crop farming use is causing ecosystem losses and affecting biodiversity in the northern region of the country. Moreover, if this long-term precipitation trend is associated with global warming impacts, it is possible that it could reverse in the future, since its relation with temperature may not necessarily be linear. In such a situation, the loss of the natural vegetation cover due to deforestation would favour a desertification process and further add to the economic losses of farmers. These considerations have now led to a government-issued moratorium forbidding further deforestation in the province of Santiago del Estero. Yet another drawback of this agricultural expansion is associated with the very nature of the Argentine Pampas, which are characterized by large plains with very small slopes, hindering water runoff. As a result, the positive trend in precipitation here has been accompanied by a greater frequency and extension of floods. However, public initiatives to address this flooding were absent for decades and only recently have some projects been undertaken to facilitate and manage the runoff. In the meantime, a chaotic network of unauthorized private channels for water drainage has developed, leading to numerous conflicts as water drained away from one field often floods the neighbouring one. In addition, existing roads also lack adequate drainage provisions and further add to the problem of waterlogging. These undesirable consequences of the rapid adaptation in the Argentine agricultural sector indicate that autonomous adaptation to climate change is a process that requires more attention and research. The lesson learnt here is
3232 J&J Climate Change
1/11/07
9:56 am
Page 302
302 Climate Change and Adaptation
that adaptation in the form of anticipatory and inexpensive regulation, and its effective enforcement is not only convenient but highly recommended as it is difficult to impose restrictions after the impacts of climate change have occurred. This must be supported by technical advice that can help moderate the negative impacts and guide the adoption of better choices, both from the environmental and the economic point of view.
A case of planned adaptation: The great river floods Consistent with the precipitation trends described above, streamflows and flood frequency of the great rivers of the Plata basin, namely the Paraná, Paraguay and Uruguay rivers, have also considerably increased since the mid 1970s (García and Vargas, 1998; Genta et al, 1998; Barros et al, 2004). The percentage amplification of their streamflows was two to three times greater than that of precipitation over their respective basins. Berbery and Barros (2002) have shown that this amplification of streamflows was primarily due to the features of the basin and that streamflow trends were largely a result of the precipitation trend. This has also been shown, with the help of various examples of sub-basins, by Tucci (2003), who estimated that at least two-thirds of the increasing trends in the Plata basin streamflows were caused by the precipitation trends. This large amplification of the streamflow response to changes in precipitation implies that activities based on or affected by these streamflows would likely be highly vulnerable to climate variability and to climate change. While the positive consequence of this increase in streamflows has been the increase in hydroelectric power generation, which was greatly favoured by this regional climate trend, the negative consequence was the increased frequency of major floods, which caused huge social and economic damages. Though the large streamflows originate in the Brazilian and Paraguayan territories, the greatest floods tend to occur in Argentina along the banks of the Paraná. In 1998, the flooded area reached 45,000km2, and similar areas were affected by floods of comparable magnitude in 1992 and 1983.1 There are also clear indications of a recent change in the frequency of occurrence of these floods: four out of the five greatest discharges of the 20th century of the major river in the Plata basin, the Paraná, occurred in the past 20 years (Table 17.3). A similar situation occurred with the Uruguay river (Table 17.4), although the floods from this river extended over less territory and had smaller socioeconomic impacts than in the case of the Paraná. It can be seen that during the second half of the last century, there was a consistent increase in the frequency of large discharges: one in the 1950s, two in the 1960s and 1970s, five in the 1980s, and six in the 1990s. The magnitude of the area flooded by the great rivers and the large number of people affected resulted in the generation of rapid awareness about the change and the need to cope, which in turn created the conditions for government action. Moreover, the worst flood events were found to be related to the El Niño phenomena (Table 17.3), which helped to identify a cause and facilitate the availability of international credit (Table 17.5). As a result, a public
3232 J&J Climate Change
1/11/07
9:56 am
Page 303
Adaptation to Climate Trends: Lessons from the Argentine Experience 303
Table 17.3 Major monthly streamflow anomalies (m3/s) at Corrientes Date and ENSO Phase June 1983 El Niño June 1992 El Niño June 1995 El Niño May 1998 El Niño September 1989 Neutral
Streamflow Anomaly (m3/s) 38,335 26,787 24,153 22,999 16,698
Note: Mean flow is 18,000m3/s.
adaptation policy was implemented after the great flood of the great tributaries of the Plata river in 1983 and its recurrence in 1992. After the 1983 flood, a hydrologic alert system was also implemented with a focus on the floods of the great rivers of the Plata basin. This system was improved after the 1992 flood, and several programmes related to reconstruction and the building of structural measures (defences) were executed with credit from international banks (Table 17.5).
Table 17.4 Largest daily discharge anomalies (larger than 3 standard deviations) of the Uruguay river at the Salto gauging station, 1951–2000 Date 9 June 1992 17 April 1959 21 July 1983 7 January 1998 16 April 1986 5 May 1983 8 March 1998 15 June 1990 24 October 1997 20 June 1972 24 April 198 7 9 September 1972 1 May 1998 16 November 1982 19 November 1963 20 September 1965 Note: Mean flow is 4500m3/s.
Discharge Anomaly (m3/s) 31,784 30,575 27,831 27,677 26,779 25,678 25,302 24,355 23,967 20,660 20,187 18,664 18,089 17,317 16,867 15,913
3232 J&J Climate Change
1/11/07
9:56 am
Page 304
304 Climate Change and Adaptation
Table 17.5 Programmes funded by international banks to ameliorate and prevent damages from floods in Argentina Programme
Funds in millions Purpose of US$
Programme of reconstruction for the emergency caused by floods (PREI)
293.4
434 works of infrastructure and housing
Programme of defence of floods (PPI)
420
155 works of infrastructure and improvement of the hydrologic alert system
Programme El Niño Argentina (defences)
60 (25 for the great rivers area)
Works of defence
PREI (second phase)
17.3
Works of defence, housing and infrastructure studies
The benefits of this adaptation process are obvious from a comparison of the flood events of 1983 and 1998. Although streamflow and flood duration were different in the two cases, the areas affected were about the same. However, because of the defences built, the number of people evacuated was considerably lower in 1998, about 100,000 versus 234,000 in 1983. Unfortunately, this is the only overall objective measure of the successful, although yet incomplete, adaptation process, besides the qualitative assessment made after the 1998 flood. A national assessment of economic losses was conducted for the 1983 flood but a similar assessment for the 1998 flood is still lacking. What this case shares in common with the preceding example was awareness. This was favoured by the fact that in both cases the changes were perceived by the key sectors. In the first example, the farmers initially became aware of the changes and their reactions came before the government’s response, which is still lagging behind in many cases. On the other hand, in the example of the great river floods, the change was easily perceived by the entire society due to the nature of its impacts, and this led to prompt public action.
Masked Climate Trends It is evident that the first requirement for adaptation to climate change or a changing climatic trend is awareness and perception of the associated advantages or threats. However, certain climate trends may be masked due to various natural and socioeconomic factors and their impacts may not be easily perceptible to the public. In this section two such examples of masked climate trends are explored: slow but very long trends and changes in extreme but infrequent events.
Slow trends Climate trends can be slow enough to pass unperceived or get masked by interannual or interdecadal climate variability. An example is the situation with
3232 J&J Climate Change
1/11/07
9:56 am
Page 305
Adaptation to Climate Trends: Lessons from the Argentine Experience 305
water resources in the regions of Cuyo and Comahue in western Argentina and in central Chile. Western Argentina is mostly arid and the region of Cuyo (between 27° and 38°S), near the Andes Mountains, receives an annual precipitation of about 100mm, with population and economic life largely concentrated along the river oasis. The rivers are primarily fed by snow melting in the mountains and therefore have a pronounced annual cycle with a maximum in summer. In the Comahue region (between 36° and 42°S), annual precipitation is also quite small, less than 200mm in the plains, although considerably greater over the Andes. Population and economic activity in the plains, as in Cuyo, are dependent on the rivers, which in this case are fed by both snow melt and rainfall in the Andes. A long-term decreasing trend in river flows exists in these two regions, but for various reasons the potential dangers of this trend have so far not been noticed by the public or by the key sectors involved in water administration. One problem is the lack of adequate monitoring of precipitation over the Andes and the absence of long series of snowfall data. Therefore precipitation trends over the Andes can only be indirectly assessed using data from central Chile (the same synoptic systems are responsible for producing precipitation in Chile and snowfall over the Andes). Figure 17.4 shows annual precipitation for two stations that are at the extremes of the latitudes between which Cuyo and Comahue are located. Both stations show definite downward trends that are also present in all the other stations (not shown) at intermediate latitudes. In addition, most climate models project a continuation of this downward trend during the present century, with some variation in intensity. This agreement in the direction of climate trends between different socioeconomic scenarios and models suggests that this signal is robust and very likely to occur.
Figure 17.4 Annual precipitation in Chile: La Serena (29.9ºS, 71.2ºW) (left); Puerto Montt (41.4ºS, 73.1ºW) (right) Consistent with the precipitation trends in Chile, river flows in Cuyo and Comahue have a general downward trend that is illustrated by two examples in Figures 17.5 and 17.6. In the case of Cuyo, there were two downward trends
3232 J&J Climate Change
1/11/07
9:56 am
Page 306
306 Climate Change and Adaptation
during 1920–1938 and 1945–1970, followed by periods of recovery (Figure 17.5). During the second negative trend, there was widespread public concern about the future viability of the oasis economy that led to a better administration of water. However, given the recovery after each negative trend, particularly after 1970, the key sectors have become sceptical about the dangerous implications of this long-term trend. This is significant in view of the fact that a new downward trend began in the mid-1980s (Figure 17.5). The interannual and interdecadal variability in this case tends to mask the significance of the long-term climate trend in public memory and contributes to the lack of risk perception.
Figure 17.5 Mean annual streamflow (m3/s) of a representative river of the Cuyo region, Los Patos river, 1900–2000 This low level of concern is the cause (and also partially the consequence) of the lack of systematic monitoring of snow in the mountains and of the evolution of the glaciers. Indeed, the little documentation available indicates a general receding of glaciers, which, if correct, would indicate a decline in the water stock in the mountains. This, together with the downward trend in precipitation and in the river flow, would indicate a matter of great concern, especially given that future trends are projected to continue in the same direction. In the case of Comahue, the downward trend in river flows was ignored until recently, when technical specialists began to analyse climate change data. The reason behind this attention on Comahue is that it possesses 25 per cent of the installed hydroelectric power projects in the country, providing 15 per cent of the country’s generated power. A technical analysis of the generation capacity found that if the existing dams were to be operated with 1940 river flows, they would produce about 30 per cent more energy than they do now,
3232 J&J Climate Change
1/11/07
9:56 am
Page 307
Adaptation to Climate Trends: Lessons from the Argentine Experience 307
Figure 17.6 Mean annual streamflow (m3/s) of rivers of the Comahue region, 1900–2000; note that the Negro river starts at the junction of the Limay and Neuquén rivers
given the same level of investment. This clearly pointed to a marked decrease in river flows since 1940 (Devoto, 2005). The other major user of water in this region is the farming sector, but since it uses only a small fraction of the available water resource, no shortage has yet been experienced. What sets Cuyo and Comahue apart is the fact that Cuyo has a long-established population, whereas Comahue has been occupied relatively recently and most of the inhabitants are immigrants from other provinces or countries. There is thus a weak collective memory about natural conditions and hazards in this area, which could account for the lack of awareness among the general public about long-term trends in the river flows. In comparison to Cuyo and Comahue, central Chile is more developed and populated. The lack of rainfall during summer makes irrigation necessary for most crops. In addition, power generation is also largely based on hydropower. However, despite Chile’s heavy reliance on its water resources, little public and scientific attention has been accorded to the slow but persistent downward trend in precipitation during the past century (Figure 17.4) and its projected continuation over the present century, as predicted by almost all climate scenarios. This is possibly because, once again, this long-term declining trend in precipitation has been masked by interannual and interdecadal variability. It can be concluded that slow but steady trends, which could be related to climate change and are likely to continue in the future based on the projections of climate change scenarios, can be masked by the interannual and interdecadal variability that mislead the population and the key sectors, thus preventing awareness about the impacts of such trends.
3232 J&J Climate Change
1/11/07
9:56 am
Page 308
308 Climate Change and Adaptation
Changes in the frequency of extreme events Changes in extreme but infrequent events can take time before they are noticed by the public. Although extreme events can cause severe damage and loss, including loss of lives, the more extreme ones are generally infrequent, occurring possibly only once in many years. Thus even a considerable increase in the frequency of such events may not be perceived until a catastrophe captures the public attention, as happened in Argentina. Trends towards a greater frequency of extreme precipitation have been observed in the central and eastern part of Argentina over the last few decades, as shown in Figure 17.7. It can be seen that the number of events of precipitation exceeding 100mm in no more than two days began to increase around 1980, and by the end of the century such extreme precipitation events were three times more frequent than observed in the 1960s and 1970s. Such trends can be expected due to the impacts of increased atmospheric concentration of greenhouse gases (Watson et al, 2003) and have been observed in many other regions of the world.
Figure 17.7 Number of events with precipitation greater than 100mm in no more than two days in periods of four years (16 stations in central and eastern Argentina)
In Argentina, these trends emit a robust signal, and one that does not depend on the threshold. For instance, for a 150mm threshold, Figure 17.8 shows the annual frequency of events in the 1983–2002 period and the ratio of the annual frequencies between the 1983–2002 and the 1959–1978 periods. In the eastern part of the country, this ratio is greater than 1:1 almost everywhere, and in some areas in the northeast the ratio is as high 4:1 or even 7:1. A ratio of 7:1 means that where extreme precipitation events could be expected once every 7 years in the past, they now occur every year.
3232 J&J Climate Change
1/11/07
9:56 am
Page 309
Adaptation to Climate Trends: Lessons from the Argentine Experience 309
Figure 17.8 Annual frequency of cases with precipitation over 150mm in less than two days (left); for the same threshold (150mm), the ratio of the annual frequencies between the 1983–2002 and 1959–1978 periods (right)
Such extreme precipitation events tend to produce devastating local floods in Argentina that affect both rural and urban environments. Cultural habits and infrastructure developed during a period with a different climate are now proving to be a burden under these new climate conditions. The inadequate infrastructure (drainage, bridges, roads and so on), which was not designed to withstand such circumstances, tends to further enhance the damages caused by these flood events. In spite of this, new infrastructure continues to be developed without accounting for the changed situation. The population in the area where the 150mm precipitation events increased twofold is 2.5 million, while in the area where these events are now four times more frequent it is 1 million. This population has only a diffuse and unclear knowledge of this change. The poor population, which bears much of the negative impacts, was probably aware of the changing precipitation trends but assumed the burden fatalistically without demanding adaptive measures. This attitude has, however, begun to rapidly change of late, especially after the event of April 2003 that flooded half the city of Santa Fe and took many lives. The people have now begun to press for solutions and no longer accept the standard excuses from public officials about the event being extraordinary or unexpected. Thus national and provincial governments have, in response, begun to institute some adaptive measures, including implementing new earlywarning systems at the provincial level (in Santa Fe) or enforcing land zoning (in Chaco). In addition, the Institute for Water Research has also started a programme to develop new standards for the design of water management. The delay in the initiation of adaptive activity in this case has been attributed to several factors, including the lack of technical knowledge, the lack of an appropriately dense network of pluviometers (to measure precipitation),
3232 J&J Climate Change
1/11/07
9:56 am
Page 310
310 Climate Change and Adaptation
and the difficulties encountered by some officials at the National Weather Service in disseminating meteorological data. This highlights the fact that the weak scientific and technical infrastructure of developing countries often tends to act as a barrier to the acquisition of information and to guiding the implementation of adaptation measures. However, it is likely that even if sufficient technical information was available, adequate and timely preventive measures may not have been implemented since, as is often typical of such events, it is only the big disasters that draw attention to their causes.
Forgetting Adaptation Attitudes The banks of the Plata river in Argentina have nearly 14 million inhabitants, mostly living in the metropolitan region that includes Buenos Aires. In this area, high tides associated with the inward drag of strong winds from the southeast are common, especially if they occur simultaneously with high astronomic tides. These events are locally known as sudestadas and cause floods along the low coasts of the Argentine margin. Because of the shape and other characteristic features of the Plata (Figure 17.9), the sea level rise resulting from climate change is also expected to propagate inwards into the river. The results of modelling studies2 indicate that higher sea levels are likely to cause an increase in the frequency of storm surge floods along the coast of metropolitan Buenos Aires.
Figure 17.9 The Plata river estuary
3232 J&J Climate Change
1/11/07
9:56 am
Page 311
Adaptation to Climate Trends: Lessons from the Argentine Experience 311
In coastal neighbourhoods such as La Boca and Avellaneda, there exists a long tradition of coexistence with floods. Here, informal networks among neighbours support local practices and strategies that aid in anticipating the arrival of floods (including an early warning system and self-help and evacuation strategies), and tend to diminish local vulnerability. However, it has lately been observed that, in both these areas, the increasing influx of newcomers is gradually eroding this collective cultural adaptation to floods. Another factor that tends to compromise the shared knowledge of cultural adaptation to floods, despite its importance, is the construction of structural defences against flooding. The coastal defence structure in the city of Buenos Aires, built in 1998, has successfully mitigated recent floods and cultural adaptation practices no longer assume as much importance. This defence was, however, designed without taking into consideration future river level rise, which may reduce its effectiveness in the coming years. By this time the knowledge of practical flood prevention and coping strategies will probably have been diminished or forgotten, and even institutional mechanisms of response to floods dismissed. One traditional adaptation strategy employed until recently was the avoidance of living in areas exposed to frequent floods. These areas are typically located away from the city and such an adaptation strategy would also have ensured minimal social impacts from the greater frequency of future floods in this region. However, since the 1980s people have increasingly begun to favour living in private gated communities over living in the metropolitan area. The rising demand for private gated towns is making the natural areas near the water attractive for the upper middle class (Ríos, 2002). It has thus become very common for gated communities to be sited on lands that are frequently flooded, though often at an artificially elevated level of 4.4m above sea level, which is the assumed secure height against flood risks. Besides the fact that this massive modification of the environment affects ecosystems and creates drainage problems, this height of 4.4m above sea level may not be adequate protection against future more intense flood threats (Barros et al, 2005). This trend favouring an increased number of gated communities near the coast, in the margins of the Paraná or even in the front of the Paraná delta, is expected to increase in the coming years (Ríos, 2002). At the same time, a growing occupation of the low lands in the valleys of some tributaries of the Plata river by low income people is also observed. This current tendency of occupying lands at risk of flood, by both very poor settlements and upper middle class gated communities, thus negates customary strategies for adaptation to present and future scenarios of recurrent floods. The changing relationship of society with the Plata river and its hazards in this case study indicates that new habits, greater resources or a large percentage of newcomers may lead a social group to forget its collective adaptive attitudes towards climate hazards. In this context, the dissemination of the findings of the AIACC project (Barros et al, 2005) in collaboration with a local non-governmental organization, Fundacion Ciudad, proved to be opportune, as these results can serve to guide governments in more effective decision making regarding potential adaptation strategies for climate change impacts in
3232 J&J Climate Change
1/11/07
9:56 am
Page 312
312 Climate Change and Adaptation
this region. They can also assist individuals and private developers in making appropriate decisions regarding future property development. This case study thus also helps to highlight the importance of scientific research in helping develop and maintain the collective awareness of both present and future climate hazards.
Discussion and Conclusions The case studies examined in this chapter show that adaptive strategies to current regional climatic trends can offer important lessons for designing socioeconomic responses to future climate change. The important trends in climate that were observed over the past 40 years in Argentina provide a wealth of valuable experiences that could help direct future adaptation planning. It is noted that in some sectors like agriculture, autonomous adaptation can be relatively fast, as has happened in Argentina over the past few decades. This adaptation was facilitated by the relatively short production cycle in agriculture and the independent process of decision making of farmers, which produced quick, results-oriented experiences and choices. However, autonomous adaptation may also have undesirable implications for the environment, other sectors, society as a whole and sometimes even the very people who take the adaptive actions. Therefore, autonomous adaptation needs to be guided by the public and research sectors to improve its benefits and reduce its negative impacts. When decisions cannot be individually taken, but depend on large entities like governments or big organizations, public awareness of the changing climatic conditions becomes a key issue in initiating the process of adaptation. This can be observed in the case of adaptation to the increasing flood frequency of the great rivers of the Plata basin, where public awareness of the impacts led to the implementation of a public adaptation policy by the government. Often this kind of awareness initially begins in the technical spheres, but its dissemination among the wider public is important to ensure the feasibility of political and economic decisions, even where the required funds come from international agencies, as is often the case in developing countries. Sometimes, climate trends may have features that make them difficult to be noticed by society. This lack of awareness is more likely to occur when the local climate observation system lacks the capacity to provide the necessary information. Two cases in this chapter show that slow but steady trends, which could be related to climate change, can often be masked by interannual and especially by interdecadal variability. This variability can confuse the population and key sectors, thus preventing their perception of the risks associated with the long-term trend. The first case showed that a downward trend in precipitation and river streamflows has been taking place for about a century in two regions of Argentina and in central Chile without much public awareness, largely because it is masked by the effects of interannual and interdecadal variability. In the second case, there was a trend towards an increasing frequency of extreme, but occasional, precipitation events in eastern Argentina
3232 J&J Climate Change
1/11/07
9:56 am
Page 313
Adaptation to Climate Trends: Lessons from the Argentine Experience 313
about which the local population was largely unaware. It was only after a catastrophic flood event in the city of Santa Fe in April 2003 that the attention of the public, authorities and specialists was drawn to the issue. Various factors, such as changes in social attitudes and habits, new resources like structural defences or highways, a large percentage of immigrants, and even new technologies, may also reduce local adaptive capacity by diminishing communal memory of collective adaptation attitudes to climate hazards as can be observed in the final example of the coastal area of the Plata river. This process could have significant social and economic repercussions, especially if these hazards are further enhanced by climate change. All the above experiences on adaptive responses to long-term climate trends in Argentina offer important lessons that can serve to guide the process of adaptation to climate change, especially in developing countries. What is common to these cases is that they all point to the significance of awareness about the changing climatic trends in order for adaptive strategies to be effectively implemented. However, various factors, such as the lack of technical knowledge, lack of an appropriate monitoring system, and difficulties in the dissemination of data and information, may act as barriers in the development of public awareness and delay adaptation in developing countries. This could have important implications in terms of future ability to adapt to the impacts of climate change. Scientific and technical organizations therefore have a key role to play in providing the necessary information for designing and implementing effective adaptive responses when necessary. Nonetheless, the weak scientific and technical capacity of developing countries could prove to be a big drawback for such countries and increase their vulnerability to the impacts of climate change.
Notes 1 2
A study made for the World Bank indicates that Argentina ranks 14th among the countries affected by floods, with economic losses that sometimes reach more than 1 per cent of annual GNP (World Bank, 2001). These were developed as part of the Assessments of Impacts and Adaptations to Climate Change (AIACC) project ‘Global Climate Change and the Coastal Areas of the Río de La Plata’.
References Barros, V. (2005) ‘Global Climate Change and the Coastal Areas of the Río de la Plata, final report’, Assessment of Impacts and Adaptation to Climate Change Project No LA 26, International START Secretariat, Washington, DC Barros, V., E. Castañeda and M. Doyle (2000) ‘Recent precipitation trends in South America East of the Andes: An indication of climatic variability’, in P. Smolka and W. Volheimer (eds) Southern Hemisphere Paleo and Neoclimates: Key Sites, Methods, Data and Models, Springer-Verlag, Berlin and Heidelberg, Germany, pp187–206 Barros, V., L. Chamorro, G. Coronel and J. Baez (2004) ‘The major discharge events in
3232 J&J Climate Change
1/11/07
9:56 am
Page 314
314 Climate Change and Adaptation the Paraguay river: Magnitudes, source regions and climate forcings’, Journal of Hydrometeorology, vol 5, pp1061–1070 Barros, V., I. Camilloni and A. Menéndez (2003) ‘Impact of global change on the coastal areas of the Río de la Plata’, AIACC Notes, vol 2, pp9–12 Barros, V., A. Menéndez and G. Nagy (eds) (2005) El Cambio Climático en el Río de la Plata [Climate Change in the Plata River], CIMA, Buenos Aires Berbery, E. and V. Barros (2002) ‘The hydrologic cycle of the La Plata basin in South America’, Journal of Hydrometeorology, vol 3, pp630–645 Camilloni, I. and V. Barros (2003) ‘Extreme discharge events in the Paraná river and their climate forcing’, Journal of Hydrology, vol 278, pp94–106 Castañeda, E. and V. Barros (1994) ‘Las tendencias de la precipitación en el Cono Sur de América al este de los Andes’, Meteorológica, vol 19, pp23–32 Devoto, J. (2005) personal communication Escofet; H. and A. Menéndez (2006) ‘Vulnerabilidad de campos productivos a mayor intensidad y frecuencia de grandes precipitaciones’ [‘Vulnerability of farming to the greater intensity and frequency of extreme precipitations’], in Vulnerability to Climate Change in Argentina, Di Tella Foundation, Buenos Aires, pp411–465 García, N. and W. Vargas (1998) ‘The temporal climatic variability in the Río de la Plata basin displayed by the river discharges’, Climate Change, vol 38, pp359–379 Genta, J., G. Perez Iribarne and C. Mechoso (1998) ‘A recent increasing trend in the streamflow of rivers in southeastern South America’, Journal of Climate, vol 11, pp2858–2862 Giorgi, F. (2003) ‘Variability and trends of subcontinental scale surface climate in the twentieth century. Part I: Observations’, Climate Dynamics, vol 18, pp675–691 Magrin, G. O., M. I. Travasso and G. R. Rodríguez (2005) ‘Changes in climate and crop production during the 20th century in Argentina’, Climatic Change, vol 72, pp229–249 McCarthy, J., O. Canziani, N. Leary, D. Dokken and K. White (eds) (2001) Climate Change 2001: Impacts, Adaptation and Vulnerability, Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK, and New York Re, M., R. Saurral and V. Barros (2006) ‘Extreme precipitations in Argentina’, in Proceedings of the 8th International Conference on Southern Hemisphere Meteorology and Oceanography, Foz de Iguazú, Brazil, American Meteorological Society, Boston, MA, pp1575–1583 Ríos, D. (2002) ‘Vulnerabilidad, urbanizaciones cerradas e inundaciones en el partido de Tigre durante el período 1990–2001’ [‘Vulnerability, gated communities and floods in the Tigre District during the period 1990–2001’], Tesis de Licenciatura en Geografía [Thesis of Degree in Geography], Facultad de Filosofía y Letras, University of Buenos Aires Tucci, C. E. (2003) ‘Variabilidade climática e o uso do solo na bacia brasileira do Prata’ [‘Climate variability and land use in the Brazilian Plata basin’], in C. Tucci and B. Braga (eds) Clima and Recursos Hidricos no Brazil [Climate and Water Resources in Brazil], ABRHA, Porto Alegre, Brazil Watson, R. T. and the Core Writing Team (eds) (2001) Climate Change 2001: Synthesis Report, Contribution of Working Groups I, II and III to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK World Bank (2001) ‘Gestión de los recursos hídricos: Elementos de política para su desarrollo sustentable en el siglo XXI’ [‘Water resource management: Elements for a sustainable development policy in the 21st century’], Report no 20729-AR, World Bank, Washington, DC, US
3232 J&J Climate Change
1/11/07
9:56 am
Page 315
18
Local Perspectives on Adaptation to Climate Change: Lessons from Mexico and Argentina Mónica Wehbe, Hallie Eakin, Roberto Seiler, Marta Vinocur, Cristian Ávila, Cecilia Maurutto and Gerardo Sánchez Torres
Introduction The municipio of González, Tamaulipas, in northern Mexico and the South of Córdoba Province in the Argentinean Pampas are regions strongly dependent on agriculture and therefore highly vulnerable to climate variability and change. Adverse climatic events such as floods, droughts and frosts can negatively impact the economy of these regions and also affect social composition and stability. In the absence of conscious efforts to adapt, potential increases in the frequency or magnitude of adverse climate events or changes in climate averages (Houghton et al, 2001) may make it more difficult for some producers to participate in the agricultural market. This may be particularly true for small-scale commercial farmers with limited capital, who are not always able to recover from recurrent crop failures. The centralized nature of sector policy in Mexico and Argentina also implies that adaptation strategies may be assessed and planned without taking into account specific local needs. On the other hand, the availability of technology, information and other resources at the local level are what determines the socioeconomic characteristics of production and the performance of farmers and communities, and their capacity to cope with adverse climate impacts. Understanding the existing coping strategies of farmers in specific geographic contexts is thus the first step towards the identification and prioritization of appropriate options to increase the adaptive capacity of particular farmer groups to future climate change and facilitate the creation of a more sustainable and equitable production environment (IUCN/IISD, 2004; Wehbe et al, 2005a and b). With this objective of focusing on the local context, we therefore present two case studies of grain and cattle producers in the above mentioned regions of Mexico and Argentina. The cases are distinguished by important differences
3232 J&J Climate Change
1/11/07
9:56 am
Page 316
316 Climate Change and Adaptation
in their respective socio-productive structures, which point to the significance of the geographic context and socioeconomic circumstance in understanding the challenge of adaptation to climate change. Yet, in presenting the cases together, we illustrate that farmers in both locations have experienced similar processes of institutional and policy reforms that have had important implications for adaptive capacity. We therefore suggest that any interventions intended to enhance adaptation to climate risk need to be considered in the context of the opportunities and constraints posed by the broader institutional environment and, conversely, that there is also a need to examine how institutions and policy explain differential adaptive capacities at the farm level.
Geographic Background González, Tamaulipas (Mexico) The municipio of González lies in the watershed of the River Panuco (or Guyalejo) in southern Tamaulipas and covers an area of 3491km2. It receives an annual rainfall of about 850mm, largely concentrated in the months of June to September, with a midsummer period of diminished rainfall (la canícula) in July and August. Drought is the most common natural hazard, although there has been occasional flooding due to cyclones and even hurricanes. Crops and pasture together cover 50 per cent of the municipio’s land, and nearly 60 per cent of the economically active population is involved in these primary sector activities (INEGI, 2000). In the southwest of the municipio, irrigated production is supported by the Las Animas Dam, which allows for the cultivation of vegetables in addition to grain crops. In the rain-fed area, sorghum is the principal crop, followed by safflower, maize and soybeans (according to the Secretary for Agriculture, Livestock, Rural Development, Fisheries and Food, SAGARPA, these crops represented 47 per cent, 17 per cent, 13 per cent and 12 per cent respectively of the planted area in 2002). Many farmers manage two annual harvests: sorghum during the summer rainy season and safflower or additional sorghum in winter with the residual soil moisture. The majority of the farmers in the municipio are ejidatarios, or farmers who received land as part of the land distribution programme after the 1910 Agrarian Revolution. Approximately 20–30 per cent of the farmers have a form of private tenure that generally permits larger landholdings (pequeños propietarios). Although sorghum is one of Tamaulipas’s most important commercial crops, the local economy is not particularly prosperous. Approximately 71 per cent of the economically active population reported receiving less than two minimum salaries in 2000, 10 per cent less than in 1990 (one minimum salary in 2000 in González was approximately US$100/month) (INEGI, 2000). Over half of the adult population in the latest population census reported having had no or incomplete primary school education, and 13 per cent were illiterate (INEGI, 2000).
3232 J&J Climate Change
1/11/07
9:56 am
Page 317
Local Perspectives on Adaptation to Climate Change 317
South of Córdoba Province (Argentina) The South of Córdoba Province in central Argentina occupies approximately 9 million hectares or (90,000km2), with 75 per cent dedicated to agricultural activities. It lies in the western portion of the Argentinean Pampas and is a transitional area between the humid and arid regions. The average annual precipitation ranges from 700 to 900mm (SMN, 1992) and is mostly concentrated in the months of September to March (Ravelo and Seiler, 1979). This region is home to more than 800,000 inhabitants but there exists an established trend of rural–urban migration. In the period between the last two National Agricultural Censuses (1988 and 2002), the number of farm units here declined from 21,645 to 14,299. An accelerated process of land concentration, particularly during the 1990s, has left 50 per cent of the agricultural land in the province in the hands of 9 per cent of landowners (each with more than 1100 hectares). The other 50 per cent of the land is distributed among the remaining 90 per cent of farmers; this is a highly heterogeneous group with a wide range of landholding sizes (INDEC, 2002) and varying levels of prosperity. The majority are family farmers that primarily depend on agricultural activities for their livelihoods. Much of the commercial grain and cattle production in the area is rain-fed, although a few farmers have incorporated groundwater irrigation systems. Because the soils rarely freeze, most farmers manage two annual harvests: wheat and other fodder crops in winter, and soybean, maize, peanuts, sorghum and, to a lesser extent, sunflower (among other less important cash crops) in summer. The area is historically characterized by mixed cash crop and livestock production; however, declining relative prices of beef have resulted in a reduction in herd size over recent decades. Similar declines have been noted in the pork, lamb and poultry industry, which prior to the MERCOSUR (the Southern Common Market) trade agreement were complementary activities within the farm. Crop production has also benefited in recent years from an increase in the exchange rate and the high international prices of soybeans and maize.
Methods Various methods, such as surveys, interviews, and workshops with farmers and other agricultural actors (public officials, leaders of farmer unions, rural infrastructure specialists and academics) were used to obtain information in both regions. The farm level survey was designed to collect data on farm characteristics (for example, type of production system, landholding size, agricultural practices and income sources), farm-level resources hypothesized to be associated with adaptive capacity (for example, education, age, technology use, climate information use, risk perception and finances) and indicators of the farm households’ sensitivity to climate impacts (for example, frequency and extent of crop losses) (Table 18.1; see also Gay, 2006). Farmers’ interviews and stakeholder workshops (for farmers and public officials) in both regions were used to determine their climate risk perceptions, the adaptive strategies employed and the principal constraints faced in their implementation.
3232 J&J Climate Change
1/11/07
9:56 am
Page 318
318 Climate Change and Adaptation
Table 18.1 Farmers’ socioeconomic characteristics Capacity Attribute
Variable
Total South of Córdoba
Total González
Number of cases
227
234
36.5 (SD 14.1)
n/a
10.1
3.4 (SD 1.77)
Age (years)
52.6 (SD 12.0)
51.6 (SD 14.3)
Landholding size (ha)
649.5 (SD 716,6)
69.9 (SD 285.2)
Machinery index
1.91 (SD 1.03)
1.62* (SD 1.86)
213,075 (SD 329,509)
n/a
38.2 (SD 34.5
n/a
n/a
13.9 (SD 54.9)
18.43
n/a
Other source of income (non-farm income as % of total income)
n/a
45.5 (SD 28.3)
Information
Official technical assistance (% beneficiaries)
30.9
26.9
Diversity
Number of crops
2.4 (SD 0.79)
1.7 (SD 1.0)
% of land dedicated to cash crops
71.5 (SD 42.3)
n/a
% livestock income
12.8 (SD 21.8)
12.7 (SD 21.6)
% of land dedicated to soybeans relative to cash crop area
64.8 (SD 25.1)
n/a
Social/human resources Potential experience (years) Education (years)
Material resources
Income (Arg$) Management capacity
Rented land (as % of worked area ) Rented land (ha/household)
Financial resources
Other sources of income (% of cases)
Note: *In Mexico, the machinery index is the sum of six binary variables, representing the ownership of six different farm machines. N/a refers to the fact that the particular variable in question was not measured in the case study. SD is standard deviation. Source: Survey data.
Together these data were integrated in a vulnerability assessment framework (Eakin et al, 2006) to identify the primary resources and characteristics of farms in each region that were considered necessary for adaptation (adaptive capacities) and the degree to which those characteristics were either present or absent in the population. Resources associated with adaptive capacity were
3232 J&J Climate Change
1/11/07
9:56 am
Page 319
Local Perspectives on Adaptation to Climate Change 319
assigned weights in consultation with farmers, based on their importance for facilitating adaptation (Eakin et al, 2006). This allowed adaptive capacity to be evaluated as a product of resources and attributes, in which no one resource or attribute is a substitute for another, but rather different combinations of resources can provide households with similar degrees of flexibility in the face of climate risk. The evaluation thus provided the basis for examining potential obstacles to adaptation in each case study.
Vulnerability to Climate Threats Farmers in both regions face frequent conditions of climate variability and extremes, which continue to exert a toll on production, indicating a lack of success in adapting to climate risk. In González, rainfall is highly variable and climate extremes tend to follow a pattern of decadal oscillation (Conde, 2005). There is also a possibility that the winter rainfall here is associated with the Pacific North American Oscillation and the El Niño-Southern Oscillation (ENSO) (Cavazos, 1999) phenomena. According to experts, excessive rainfall, floods and cyclonic activity were common in the 1970s, while in the 1980s and 1990s drought and high temperatures were more common (Gay, 2006), with both resulting in crop losses. In general, a greater overall climate variability and a decrease in precipitation have been observed since the 1970s, (Sánchez Torres and Vargas Castilleja, 2005). The year 2000 was reported to be the worst year by a majority (64 per cent) of the farmers surveyed, particularly for sorghum and safflower production. Pest outbreaks are also particularly problematic and believed to be associated with the magnitude of the midsummer drought (July–August). In interviews, farmers reported experiencing recent changes in climate in terms of increasing temperatures and some associated an increase in precipitation in September with greater moisture available for winter planting. Although climate models for the region are inconclusive in terms of future rainfall changes, they do consistently indicate that the area will likely experience higher temperatures and, possibly, a consequent decrease in soil moisture availability (Sánchez Torres and Vargas Castilleja, 2005; Gay, 2006). The city of González already experiences deficits in water availability, which is expected to increase even without climate change according to recent modelling studies (Sánchez Torres and Vargas Castilleja, 2005). In the South of Córdoba, thermal and water conditions are important variables affecting crop yields, with soil moisture being the primary limiting factor. Winters are mild and short, characterized by frost events coupled with soil moisture deficits. Although there is a water surplus in the average balance of the region, the inter-annual variability of precipitation generates occasional droughts of varying frequency and severity (Rivarola et al, 2004a; Vinocur et al, 2004). According to farmers, such droughts, followed by hail storms, have the biggest impact on production. The impact of climate events in the region is further complicated by the soil properties and topography in certain areas (depressed areas and flood-
3232 J&J Climate Change
1/11/07
9:56 am
Page 320
320 Climate Change and Adaptation
prone basins, salty soils, drainage difficulties, soil water capacity, and so on), which cause varying levels of risk of drought or flood and generate varying environmental responses. Drought risk increases from the east of the region to the west and south (Rivarola et al, 2004b), whereas floods are more common in the south. In that area three major flood episodes have occurred during the past 25 years, affecting agricultural production and the economy for several years after each episode (Seiler et al, 2002). Linkages of the present climate variability with the El Niño Southern Oscillation are a good possibility and strong evidence exists of a La Niña signal associated with significant decreases in rainfall in the west and east of the region over the rainfall periods analysed (Seiler and Vinocur, 2004). On the other hand, there was insufficient evidence linking increases in rainfall during El Niño years with a clear El Niño signal, as compared to neutral years. Climate change scenarios for the region project an overall increase in mean temperature for all seasons and an increase in precipitation during summer, spring and autumn. The greater rainfall projected during the summer and autumn may contribute to an increased flood risk in the flood-prone basin of the south of the region (Gay, 2006).
Other factors that contribute to climate risks It is important to recognize that climate is only one of several factors to which farmers are making intra-seasonal, inter-annual and longer-term adjustments in their production strategies (Risbey et al, 1999; Smit et al, 1996). In Mexico, the price of grains – principally maize, but also wheat and sorghum – declined during the 1990s and further declines are projected in the future (Claridades Agropecuarias, 2004). The liberalization of Mexico’s grain import markets during this period resulted in increased competition for González farmers and necessitated financial support from the government to help farmers commercialize their sorghum harvests. A restructuring of public agricultural institutions has paralleled market liberalization, reducing the availability of publicly subsidized credit, insurance and technical assistance for smallholders (Appendini, 2001; de Janvry et al, 1995). Nevertheless, current agricultural plans for the region focus on facilitating farmers’ risk management through the promotion of contract farming (to provide greater price stability) and private insurance schemes (to address climatic risk). In Argentina, trade liberalization and a retrenchment of state roles in the agriculture sector were instituted to revive national economic growth. As a result, farmers’ resources and their production decisions now have greater weight in determining their economic stability. In the 1990s, a fixed exchange rate translated into declining relative prices of traded to non-traded goods and high real interest rates, producing a 60 per cent decline in farmers’ purchasing power. Despite devaluation of the Argentinean peso in 2001 and the consequent economic recovery of farmers, rising costs of production and finance and newly incorporated export taxes have prevented smaller agricultural enterprises from maintaining agricultural equipment and acquiring sufficient capital to finance their production. This situation, in a sector characterized by greater
3232 J&J Climate Change
1/11/07
9:56 am
Page 321
Local Perspectives on Adaptation to Climate Change 321
competition and economic consolidation, has increased the economic vulnerability of lower-scale producers (Lattuada, 2000) and reduced the demand for locally sourced agricultural inputs and services, thus producing a drop in local economic activity. These socioeconomic and institutional changes in Mexico and Argentina have overall greatly impacted the sustainability of farmers’ livelihoods, especially affecting the small producers, and further compounded their vulnerability to the impacts of climate variability and change.
Current Farm-level Adaptations and Public Sector Support In both regions, farmers currently practise a variety of production strategies that represent their different capacities to manage risk and to take advantage of new opportunities in their respective agricultural sectors. From the farmers’ perspective, production, income and investment decisions are rarely made in response to a single stressor such as drought risk, but are rather the outcome of simultaneously considering a wide variety of stressors, including, but not limited to, climatic factors. The degree to which households are able to and do respond to a specific climatic threat is, in part, determined by their perception of the threat, as well as the relative importance they place on climatic risk compared to other sources of stress and the range of choice and opportunity available from their particular socioeconomic conditions. To assess farmers’ climate adaptation options, we evaluated the factors considered in their production decisions and the specific role of climate and climate information in those decisions. We also recorded their current climate risk management strategies, such as crop diversification and seasonal crop switching, the potential of cattle-raising and pasture as a more sustainable alternative under drier conditions, the use of irrigation, and financial mechanisms such as insurance.
González, Mexico In González, farmers reported a range of adjustments to the drought conditions of the 1990s, including changing crop-planting dates, switching crops, changing crop varieties or livestock breeds, modifying infrastructure or inputs, or a combination of these strategies. On an inter-annual basis, over one third of the surveyed farmers reported adjusting their crop choice, according to their observations of the timing and quantity of the initial rains of the season. The total range of crops planted in González is, however, relatively small (averaging between 1 and 2 crops per household per year in the survey), and because of this homogeneity of production, local markets are often saturated with the same crop variety, resulting in problems with their commercialization. The government is therefore now promoting diversification into non-traditional crops and livestock, to address both environmental challenges to production (for example, soil degradation due to sorghum mono-cropping) and the lack of commercial opportunities in grain farming (Secretaría de Desarrollo
3232 J&J Climate Change
1/11/07
9:56 am
Page 322
322 Climate Change and Adaptation
Económico y del Empleo, 2001). Some of the alternatives include crops suited to drier and warmer climates such as tequila ágave (Agave tequiliana Weber azul) and aloe (Aloe barbadensis Miller). The planting of buffle grass is also being encouraged through a national programme of crop conversion (PIASRE, Programa Integral de Agricultura Sostenible y Reconversión Productiva en Zonas de Sinestralidad Recurrente) (Yarrington Ruvalcaba, 2004). As a result the area under planted pasture increased by 63 per cent between 1999 and 2002 in the Rural Development District of González (which includes González’s neighbouring municipios of Altamirano and Mante), although only a handful of farmers reported receiving support through this programme. The smaller-scale farmers were particularly dissatisfied and perceived the government’s support for these alternatives to be inadequate and found the investment necessary to be prohibitive. Even much of the land planted with ágave and aloe was rented out by ejidatarios to outside investors because the ejidatarios often found it difficult to obtain credit and commercialize their harvests due to the small scale of their production and the variable quality of their products. Nevertheless, an increasing number of farmers, mostly smallholder ejidatarios, are now investing in livestock in response to repeated crop losses and problems in commercializing their harvests. The government’s Program of Incentives for Livestock Productivity (PROGAN) partly supports such livestock rearing activity. However, not all experts interviewed favoured the livestockpasture strategy as an appropriate response to drought given that livestock farmers have been the most affected by drought in recent years. This is because, in the event of drought, livestock farmers have to resort to either culling their animals or selling them at very low prices due to a scarcity of feed or buying sorghum as feed from their neighbours, which is again expensive. Some local farmers interviewed also agreed with this assessment of the liability of owning cattle in the event of drought. According to our survey farmers who had planted pasture reported some of the highest losses due to drought in 2002, and many sold cattle as a result. Additionally the influx of live cattle from the US due to the liberalization of the cattle market has also further driven down local cattle prices. In response to drought, some wealthy farmers, usually with larger landholdings and private tenure, have constructed small earthen dams to capture rainwater for auxiliary irrigation during dry spells. Interviews with some ejidatarios who had constructed such dams revealed much scepticism about their effectiveness because in the event of insufficient rain little water is captured, meaning that the investment has been in vain. In an effort to help farmers address climatic contingencies and price volatility, the state and federal governments are actively promoting crop insurance and contract farming to reduce the financial burden of crop loss compensation programmes (Yarrington Ruvalcaba, 2004). Very few (9 per cent) of the surveyed farmers, however, had crop insurance and the majority of farmers with insurance were pequeños propietarios (large landholders), although a handful of ejidatarios in the irrigation district also had insurance. Lack of affordability, lack of information and general distrust were commonly cited reasons for not having contracted insurance. Some ejidatarios who had purchased insurance in the 1980s under a government scheme reported repeated difficulties in receiving
3232 J&J Climate Change
1/11/07
9:56 am
Page 323
Local Perspectives on Adaptation to Climate Change 323
insurance payments. The recent declining value of smallholder harvests also leaves little incentive for purchasing insurance. Other factors that inhibit farmers from experimenting with new tools such as crop insurance or new commercial crops include the relatively low education levels coupled with the absence of extension services (either private or public). Less than one quarter of farmers reported being members of agricultural organizations through which they could conceivably acquire information on public and private agricultural services and opportunities, as well as lobby for programme changes to meet their common goals.
South of Córdoba, Argentina The farmers in South of Córdoba differed from those in González not only in their exposure to specific climate stressors but also in terms of production activities, soil conditions and use, material assets, perception of risk, and landholding size (and therefore income). As a result, their specific responses to climate impacts also differ, although direct relationships are difficult to quantify (Eakin et al, 2006). From the survey data, the most common agronomic adaptations of farmers were adjusting planting dates (36 per cent of surveyed farmers); spatially distributing risk through geographically separated plots (52 per cent); changing crops (12 per cent); accumulating commodities as an economic reserve (85 per cent); and maintaining livestock (70 per cent). Many of these strategies were not always specifically in response to climate conditions but rather as economic responses to general changes in the agricultural sector. Drought is perceived not only as an event in and of itself but also the result of a combination of climate events, namely increasing temperature, decreasing precipitation and wind, with livelihood impacts that potentially last over a year. Although irrigation is an obvious technical solution for drought risk mitigation, its exorbitant cost (especially for smaller farmers) and the quality of available groundwater are important barriers. Only 1 per cent of farmers in the region, mostly large landholders, have installed irrigation systems. Cattle-rearing activities, on the other hand, are generally perceived to offer greater security than cash cropping because cattle rearing is considered to be less sensitive to climatic anomalies (for example, hail storms) and cattle also serve as an economic reserve. The provincial government has also introduced programmes, including credit support, to promote livestock rearing. Commercial hail insurance is typically used to address the impact of hail storms, although only 65 per cent of farmers surveyed reported having contracted hail insurance and, of these, only 53 per cent contracted insurance annually. Another type of insurance, ‘climate risk insurance’, is even less used, largely because it is costly and also poorly implemented. Public officials interviewed reported that the implementation of subsidized climate risk insurance was often problematic due to oligopolistic practices by insurance companies, which have created a general sense of distrust among farmers. In response, a group of farmers have recently established a new cooperative programme called Seguro Solidario, wherein participating farmers contribute a fixed sum
3232 J&J Climate Change
1/11/07
9:56 am
Page 324
324 Climate Change and Adaptation
to a collective fund for coping with climatic events. This local insurance mechanism is, however, not widespread in the regions studied and its effectiveness depends largely on the severity of climate impacts. The programme is now being promoted at the provincial level as a pilot programme. The primary source of government support for dealing with climate impacts is from a highly controversial mechanism under the Agricultural Emergency Law (AEL) of 1983 under which farmers may publicly declare their losses. With the objective of diminishing impacts from climatic, telluric, biological, physical and unforeseeable or inevitable events, the AEL allows farmers access to benefits such as delaying fiscal obligations, acquiring tax extensions or exemptions and accessing credit, among others. However, opinion about this mechanism among farmers is once again generally low, largely because it is a tedious process and in most cases only serves to delay payments. In contrast, participating in farmers’ organizations or associations is generally viewed as highly positive, necessary, useful and powerful because of its ability to promote common interests. However, farmer interviews revealed that the advantages of organization depend largely on personal experience and member attitudes. Moreover, with the improved economic situation due to the devaluation of the peso, organizations are now considered necessary only as a temporary response in periods of difficulty. In our analysis we found that only 50 per cent of farmers participated in such organizations, while the rest considered them either not useful (13 per cent), associated with bad experiences (12 per cent), of little interest (27 per cent) or lacking capacity (39 per cent). Besides formal mechanisms, adaptation to climate risks in the South of Córdoba is also facilitated by the use of climate information from various sources. However, climate information, especially seasonal forecasts, is usually only accessed through private seed or chemical providers, the internet and special workshops/seminars organized by farmers’ organizations. It is largely used to inform short-term decisions such as determining planting and harvesting dates, while the major production decisions are based primarily on market signals, soil conditions and the availability of working capital. Overall, the common opinion among farmers is that any action necessary to resolve local problems such as repeated negative climate impacts would benefit more from local action rather than interventions from the national government. Recent changes in macroeconomic and sector policy have made farmers distrustful of any support or protection from the national government, and the added burden of export taxes as a result of these market changes represents a fundamental concern. Thus, both factors, state interventions and climate, are commonly perceived as very unpredictable by the majority of the farmers.
Opportunities for Intervention The determination of the resources and attributes of adaptive capacity specific to each region discussed above enabled the identification of possible priorities for public sector interventions that could enhance farm-level capacities (a systematization of these priorities is presented in Table 18.2).
Low to medium*
Cost of equipment, cost of maintenance, economies of scale, scepticism (Mex)
Improved yields, reduced drought impacts, additional subsistence benefits (aquaculture, Mex), reduced risk in new crop investment (Mex)
Capacity to implement
Potential obstacles
Benefits
Medium (Arg)
Public funds
Enables cost recovery after loss (Mex and Arg), facilitates agricultural diversification (Mex)
Reduced uncertainty over production in flood-prone areas
Political will (Arg), scepticism, Competition for public distrust, low value crops funds, regional (Mex) priorities
Guarantees of contracts, market transparency, information, high value crops Medium (Arg) Low (Mex)
Before – general
Information
Climate trends, variability, forecasts; markets; prices; new technologies All levels of government, farmer associations, extension services Time for technology Information networks and development, institutional intermediaries, extension coordination services, human resources High (Arg) Medium** Medium-low (for those technologies that require public investment) Cost, decline in public Lack of organizational investment in research, capacity, lack of funding, lack lack of explicit demand of interest (Mex), lack of from social sector ‘culture of information’ (Mex) Reduces productivity gap Better risk management and between farmer groups, improved decision making, increases economic improved dissemination of margins technology, greater access to public support programmes
Inputs (seeds, fertilizers, etc) and management (conservation tillage etc) Local government, public research institutions
Before – general
Technology
9:56 am
Notes: *With financial and technical support; **Information is available but the network for distribution is not established.
Hydrological studies, credit
Government (national or state)
Drainage containment infrastructure, roads
Before – flood
Infrastructure
1/11/07
Conditions
Farmer
Commercial, publicly subsidized or cooperative
Individual or system development; groundwater or surface water Farmer
Who would implement?
After – hail, drought, floods
Before – drought
Timing of measure (before/after and for what hazard) Type of measure
Insurance
Irrigation
Measures
Table 18.2 Synthesis of adaptation options
3232 J&J Climate Change Page 325
Local Perspectives on Adaptation to Climate Change 325
3232 J&J Climate Change
1/11/07
9:56 am
Page 326
326 Climate Change and Adaptation
Critical components of adaptive capacity in González include financial resources, such as credit and insurance; material resources, such as land, irrigation and equipment; the degree of economic and agricultural diversification of the farm; and access to resources, such as by means of technical assistance (Eakin et al, 2006). In South of Córdoba, adaptive capacity was observed to be more a function of material and financial resources (such as credit, soil quality, landholding size and type of activity) and to a lesser extent of human/social resources (for example, personal experience, availability of technical assistance and participation in organizations). Other coping strategies such as maintaining crop diversity and alternative non-farm income sources were also important, together with specific climate adaptations, such as the use of insurance and climate information. The prioritization of potential interventions by the public sector to boost this existing farm-level adaptive capacity and address the various issues that contribute to the increasing vulnerability to future climate impacts are discussed below.
Mexico The majority of the adaptation options described above require finance and, although there are limited credit windows for smallholders, the support is generally not extensive. Most households therefore increasingly depend on alternative income sources to finance their agricultural activities. Public intervention to improve and expand access to credit, especially for the small farmers, could help ease their financial burden and enable more effective adaptive responses to climate risks. For instance, according to some of the larger-scale producers in the region, diversification into alternative commercial crops is only possible with appropriate tools and capital, for example, the construction of greenhouses, mayas de sombra (artificial shade cover) and private auxiliary irrigation networks for supporting new crops under warmer and drier conditions. Obtaining public support for such projects is tedious and farmers still need to invest substantially in terms of financial and labour contributions. As a result, few small-scale farmers risk investment in infrastructure projects given the uncertainty of their harvests. Adaptation in this case can thus be facilitated with targeted public support for specific infrastructure projects at the farm level, combined with a guaranteed living wage for farmers and greater security in marketing their harvests. Another element critical for adaptation in agriculture is irrigation, but the future availability of irrigation water is uncertain given the increasing urban and industrial demand (Sánchez Torres and Vargas Castilleja, 2005). Currently, ejidatarios with irrigation generally belong to the irrigation districts of the Rio Guayalejo, depending on infrastructure (often unlined canals) from the 1970s. Investment in auxiliary irrigation or improved irrigation efficiency is only viable if the crop is sufficiently high value (for example, ágave, onions and other vegetables, and fruit trees), yet such high-value crops are again associated with new and often high economic risks (Eakin, 2003). Many farmers are thus increasingly renting their land to outside investors and have little personal
3232 J&J Climate Change
1/11/07
9:56 am
Page 327
Local Perspectives on Adaptation to Climate Change 327
interest in investing in their water works. Improved management of current irrigation networks and greater efficiency of new infrastructure are thus likely to be critical adaptation interventions. Furthermore, many farmers also do not possess any insurance against climate risks, once again due to the high costs as well as the overall lack of confidence in insurance mechanisms. However, investing in new crops such as aloe without insurance also represents a high financial risk, despite the fact that the crop may be better adapted to warmer or drier conditions. Public initiatives to make insurance services more affordable and reliable would therefore enable more farmers to be protected against investment risks in the implementation of adaptive strategies to cope with climate impacts. Finally, farm organizations and producer associations could potentially play a key role in climate adaptation by facilitating access to markets and providing services and information. Public support for the development of their administrative and technical capacity would, however, be important. The trend of using such associations for meeting political ends would also need to be discouraged since it prevents the achievement of any technical and productive goals by participating farmers (Eakin, 2004).
Argentina In the case of Argentina, under the current policy environment, access to credit is likely to continue to be restricted to the private banking system and input suppliers, despite the fact that most farmers believe that increasing credit availability and diminishing export taxes could resolve their problems. Important investments in fixed capital would possibly be required in the near future for the installation of supplementary irrigation systems to address the impacts of drought caused by climate change (Wehbe et al, 2005; Eakin et al, 2006; Gay, 2006). Public interventions (tax incentives or interest rate subsidies) to overcome the high cost of private banking credit as well as educating farmers to use supplementary irrigation technologies would therefore assume importance. There is also need for an accurate analysis of the capacity of regional surface and groundwater as potential sources of irrigation. Smallholder farms in Argentina also face significant barriers in obtaining insurance to cover climate-related agricultural losses due to the lack of guarantees by insurance companies and the exorbitant premiums. Government interventions to address these issues and facilitate insurance use could include supervising the completion of contract obligations, providing information and subsidizing insurance for lower-scale farmers. Unfortunately, a lack of political will and the absence of the necessary political infrastructure have so far prevented any control over the industry. To date, the primary government interventions have been restricted to limited subsidies of premiums and the declaration of an agriculture emergency only when an event affects an important geographic area. Some farmers in the region, mainly in the south of the region, are highly vulnerable to an increased risk of floods. Floods are also a principal source of conflict among neighbouring farmers and between rural and urban areas.
3232 J&J Climate Change
1/11/07
9:56 am
Page 328
328 Climate Change and Adaptation
Interventions to support flood risk management such as infrastructure works (additional drainage or containment structures, the diversion of excess water, and road construction) as well as improved rezoning of crops and improvements in land use practices would entail substantial investments that would require support from the national or provincial government. In addition, support from local governments for smaller flood management initiatives and maintenance work would also be essential. Last but not least, the use of advanced agricultural technology could additionally play a very important role in adapting to climate risks in this region. Unfortunately, most agricultural technology is presently only commercially available and is therefore unaffordable for many small farmers. Public sector support for technology, research and development could help increase the accessibility of technology for such farmers and increase the likelihood of adaptation. However, this sort of public intervention currently faces significant barriers which would first need to be addressed, for example, the high investment costs of technology, the lack of institutional coordination, and a lack of participation by farmers in producer associations to help articulate their technology needs.
Comparing the two regions Despite significant differences in the scale of production and agricultural histories between the municipio of González in Mexico and the South of Córdoba Province in Argentina, there are also important similarities in the opportunities and constraints for adaptation. In both cases, one could argue for improved access to climate, market and technological information as an important means for enabling farmers to respond rapidly to economic and environmental change. Enhancing the accessibility of information entails supporting farmers’ social and professional networks, as well as investment from public sector institutions in the synthesis and systematization of available information. It can also be concluded that targeted interventions from the public sector and from farm associations are necessary to ensure capacity building for adaptation. For smaller-scale farmers to be able to sustain their agricultural livelihoods under a potentially more variable future climate, specific technical support would be necessary to facilitate their access to appropriate technological packages, markets for alternative cash crops and formal insurance mechanisms, and to support improvements in irrigation, drainage and other productive infrastructure. In the absence of such support, it appears probable that the ability to adapt to future climatic and economic changes will be restricted to larger-scale farmers or external agribusinesses with the capital to acquire credit, technology and insurance. Many of these producers will likely be outside investors renting or purchasing land. Small-scale farmers struggling with crop losses and commercialization issues today may choose adaptive options entirely outside the agricultural sector, a move which would stimulate not only a social transformation of the sector but might also entail important landscape and ecological changes in both regions.
3232 J&J Climate Change
1/11/07
9:56 am
Page 329
Local Perspectives on Adaptation to Climate Change 329
Conclusions Our investigation of the agricultural sector in González, Mexico and the South of Córdoba Province, Argentina, has primarily highlighted the importance of the local context, existing practices and farmers’ perceptions in the determination of potential adaptation options to the impacts of climate change in these regions. We have shown that although some sections of the population are currently engaging in a variety of agricultural practices that could be helpful in mitigating climate risk, such as crop and economic diversification, insurance coverage and irrigation development, widespread adoption of these practices and technologies are limited by a lack of access to finance, poor information networks and market failures. Of particular concern is the differential access to specific coping strategies between large- and small-scale farmers and, in the case of Argentina, between smaller family-run farms and large agribusinesses. Current policy trends in both countries indicate little government support for resolving specific agricultural sector problems. Instead, the focus of public policy is on the development of an enabling environment for private investment and economic growth, with less attention to the distributive implications of such policies. Although improved economic conditions will undoubtedly increase the flexibility of some farmers in responding to environmental change, problems with resource access and technology adoption in vulnerable subsectors demand more specific local action. Ideally such local action would be the result of collaboration between farmers, producer associations, the private sector and local government. Given the significance of economic and political obstacles to the implementation of various adaptation options, the possible interventions identified above require rigorous evaluation within a participatory and collaborative local context. This would ensure the selection of adaptive strategies that have the greatest potential to foster the sustainability of the farm sector and thus positively impact the economic, social and environmental conditions of communities.
References Appendini, K. (2001) De la Milpa a los Tortibonos: La Restructuracion de la Politica Alimentaria en Mexico [From Milpa to Tortibonos: Food Policy Restructuring in Mexico], El Colegio de Mexico, Mexico City, Mexico and United Nations Research Institute for Social Development (UNRISD), Geneva, Switzerland Claridades Agropecuarias (2004) ‘Perspectivas agrícolas de la OCDE 2003–2008’ [‘Agriculture Perspectives from the OCDE 2003–2008’], Claridades Agropecuarias, vol 133 (September), pp16–30 Cavazos, T. (1999) ‘Large-scale circulation anomalies conducive to extreme precipitation events and derivation of daily rainfall in northeastern Mexico and southeastern Texas’, Journal of Climate, vol 12, pp1506–1522 Conde, C. (2005) ‘Climate change and variability in southern Tamaulipas’, paper presented at Stakeholder Workshop, Project Closure, AIACC Project LA29, González, Tamaulipas, 6 May de Janvry, A., M. Chiriboga, H. Colmenares, A. Hintermeister, G. Howe, R. Irigoyen, A. Monares, F. Rello, E. Sadoulet, J. Secco, T. van der Pluijm and S. Varese (1995)
3232 J&J Climate Change
1/11/07
9:56 am
Page 330
330 Climate Change and Adaptation Reformas del Sector Agricola y el Campesinado en Mexico [Agriculture Sector Reforms and Peasantry in Mexico], Fondo Internacional de Desarollo Agrícola y Instituto Interamericano de Cooperacíon para la Agricultura, San José, Costa Rica Eakin, H. (2003) ‘The social vulnerability of irrigated vegetable farming households in Central Puebla’, Journal of Environment and Development, vol 12, pp414–429 Eakin, H. (2004) ‘Waiting to recover: Adaptation to the coffee crisis in two Mexican coffee communities’, paper presented at the Annual Meetings of the Association of American Geographers, Philadelphia, PA, 14–19 March Eakin, H., M. Wehbe, C. Avila, G. Sánchez Torres and L. Bojórquez (2006) ‘A comparison of the social vulnerability of grain farmers in Mexico and Argentina’, AIACC Working Paper No 29, International START Secretariat, Washington, DC, www.aiaccproject.org Gay, C. (2006) ‘Vulnerability and adaptation to climate change: The case of farmers in Mexico and Argentina’, final report, Project LA29, Assessments of Impacts and Adaptations to Climate Change, International START Secretariat, Washington, DC, www.aiaccproject.org Houghton, J. T., Y. Ding, D. J. Griggs, M. Noguer, P. J. van der Linden and D. Xiaosu (eds) (2001) ‘Technical summary’, in Climate Change 2001: The Scientific Basis, contribution of Working Group I to the Third Assessment Report, Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK INDEC (Instituto Nacional de Estadística y Censos) (2002) ‘Censo nacional agropecuario 2002’ [‘National Agriculture Census 2002’], preliminary data, available at www.indc.mecon.ar INEGI (Instituto Nacional de Estadística Geografía e Informática) (2000) ‘Censos económicos 1999’ [‘Economic Census 1999’], Government of México, available at www.inegi.gob.mx/ IUCN (World Conservation Union)/IISD (International Institute of Sustainable Development) (2004) ‘Sustainable livelihoods and climate change adaptation. A review of phase one activities for the project “Climate Change, Vulnerable Communities and Adaptation”’, available at www.iisd.org/pdf/2004/envsec_ sustainable_livelihoods.pdf Lattuada, M. (2000) ‘El crecimiento económico y el desarrollo sustentable en los pequeños productores agropecuarios argentinos de fines del siglo XX’ [‘Economic growth and sustainable development of small Argentinean agriculture producers at the end of the 20th Century’], Conferencia electrónica sobre Políticas Públicas, Institucionalidad y Desarrollo Rural en América Latina, available at www.rlc.fao.org/foro/institucionalidad Ravelo, A. C. and R. A. Seiler (1979) ‘Agroclima de la provincia de Córdoba: Expectativa de precipitación en el curso del año’ [‘Agroclimate of the Province of Córdoba: expected amount of annual rainfall’], Revista de Investigaciones Agropecuarias, INTA, vol 14, pp15–36 Risbey, J., M. Kandlikar and H. Dowlatabadi (1999) ‘Scale, context and decision making in agricultural adaptation to climate variability and change’, Mitigation and Adaptation Strategies for Global Change, vol 4, pp137–165 Rivarola, A., R. Seiler and M. Vinocur (2004a) ‘Vulnerabilidad y adaptación de los productores agropecuarios al cambio y a la variabilidad climática: El uso de la información agrometeorológica’ [‘Vulnerability and adaptation of farmers to climate variability and change: The use of agroclimatic information’], Revista Reflexiones Geograficas, vol 11, pp109–120 Rivarola, A., R. Seiler and M. Vinocur (2004b) ‘Vulnerabilidad agroclimática a las sequías en el sur de la provincia de Córdoba’ [‘Agroclimatic vulnerability to droughts in the south of Córdoba, Argentina’], paper presented at Actas X Reunión Argentina y IV Reunión Latinoamericana de Agrometeorología, Mar del Plata, Argentina, 13–15 October
3232 J&J Climate Change
1/11/07
9:56 am
Page 331
Local Perspectives on Adaptation to Climate Change 331 Sánchez Torres, G. and R. Vargas Castilleja (2005) ‘Integrated assessment of social vulnerability and adaptation to climate variability and climate change among farmers in Mexico and Argentina. Case Study: Municipality of González, Tamaulipas, Mexico’, final report, Graduate Division, School of Engineering, Universidad Autónoma de Tamaulipas, Tampico, Tamaulipas, Mexico Secretaria de Desarrollo Económico y del Empleo (2001) Programa Regional de Conversión de Cultivos [Regional Programme for Crop Conversion], Ciudad Victoria: SAGARPA y Gobierno de Tamaulipas, Tamaulipas, Mexico Seiler, R., M. Hayes and L. Bressan (2002) ‘Using the standardized precipitation index for flood risk monitoring’, International Journal of Climatology, vol 22, pp1365–1376 Seiler, R. and M. Vinocur (2004) ‘ENSO events, rainfall variability and the potential of SOI for the seasonal precipitation predictions in the south of Cordoba, Argentina’, 14th Conference on Applied Climatology, Seattle, WA, 11–15 January, available at http://ams.confex.com/ams/84Annual/techprogram /paper_71002.htm Smit, B., D. McNabb and J. Smithers (1996) ‘Agricultural adaptation to climatic variation’, Climatic Change, vol 33, pp7–29 SMN (Servicio Meteorológico Nacional) (1992) ‘Estadísticas climatológicas’ [‘Climatic statistics’], Servicio Meteorológico Nacional, Fuerza Aérea Argentina, Buenos Aires Vinocur, M., A. Rivarola and R. Seiler (2004) ‘Use of climate information in agriculture decision making: Experience from farmers in central Argentina’, Second International Conference on Climate Impacts Assessment, SICCIA, Grainau, Germany, 28 June–2 July, available at www.cses.wash ington.edu/cig/ outreach/ workshopfiles/ SICCIA /program .shtml Wehbe M. B, R. A. Seiler, M. R. Vinocur, H. Eakin, C. Santos and H. M. Civitaresi (2005a) ‘Social methods for assessing agricultural producers’ vulnerability to climate variability and change based on the notion of sustainability’, AIACC Working Paper No 19, International START Secretariat, Washington, DC, www.aiaccproject.org Wehbe, M. B., H. Eakin and A. Geymonat (2005b) ‘Macroeconomic reforms and agricultural policies in developing countries: Impacts on the social vulnerability of family farmers in Argentina and Mexico’, paper presented at the IHDP Workshop on Human Security and Climate Change, Oslo, 21–23 June Yarrington Ruvalcaba, T. (2004) V Informe de Gobierno 1999–2004 Tamaulipas [Fifth Report of the Government of Tamaulipas 1999–2004], Government of the State of Tamaulipas, Ciudad Victoria, Mexico
3232 J&J Climate Change
1/11/07
9:56 am
Page 332
19
Maize and Soybean Cultivation in Southeastern South America: Adapting to Climate Change Maria I. Travasso, Graciela O. Magrin, Walter E. Baethgen, José P. Castaño, Gabriel R. Rodriguez, João L. Pires, Agustin Gimenez, Gilberto Cunha and Mauricio Fernandes
Introduction Several important changes in climate and crop production trends have been detected in Southeastern South America since the late 20th century. The climatic changes are characterized by increases in precipitation (up to 50 per cent in some areas); decreases in maximum temperature, especially during spring and summer; and increases in minimum temperature during most of the year (Castañeda and Barros, 1994; Barros et al, 2000; Pinto et al, 2002; Bidegain et al, 2005; Magrin et al, 2005). In response to the favourable climate trends, a subsequent significant increase in crop production, especially in the yield of rain-fed crops, has been noted. Magrin et al (2005), in comparing the period from 1950 to 1970 with that from 1971 to 1999, calculated a 38 per cent increase in soybean yields and an 18 per cent increase in maize yields attributable to the changes in climate (isolated by using crop simulation models with the same production technology). Added to the influence of changing temperature and precipitation, changes in land use over the past few years have further contributed towards the increase in crop yields. Encouraged by favourable climatic and economic conditions, farmers in the region have begun to devote more and more land to agriculture, especially soybean farming. The recent expansion of soybean cultivation is particularly remarkable, with an increase of 133 per cent (from 6 to 14 million hectares, Mha) in the area devoted to this crop between 1995 and 2003 in Argentina alone (SAGPyA, 2005). A similar trend has also been observed in Brazil, where, in the traditional soybean growing areas such as Rio Grande do Sul, lands devoted to soybean have increased by 38 per cent over the past five years (IBGE-LSPA, 2005). More recently, a huge expansion of this crop has also been observed in Uruguay,
3232 J&J Climate Change
1/11/07
9:56 am
Page 333
Maize and Soybean Cultivation in Southeastern South America 333
with an increase in cultivated area from 12,000ha in 2000 to 278,000ha in 2004 (MGAP-DIEA, 2005). According to future projections, the total area under soybean cultivation in South America is expected to grow from 38Mha in 2003/2004 to 59Mha in 2019/2020. Thus an increase of 85 per cent (172 million tons) in the total production of Argentina, Brazil, Bolivia and Paraguay is expected (Maarten Dros, 2004). Some previous studies have used crop production models1 based on climate scenarios from general circulation models (GCMs) to assess future changes in crop yields specifically for the southeastern South America region. According to model projections, when CO2 enrichment effects on crop growth stimulation are accounted for, soybean production in the Pampas region of Argentina during the 21st century could range between a reduction of 22 per cent and increases of 3 per cent, 18 per cent and 21 per cent under the UKMO (Wilson and Mitchell, 1987), GFDL (Manabe and Wetherald, 1987), GISS (Hansen et al, 1989) and MPI-DS (Magrin et al, 1998) scenarios respectively.2 Maize production, on the other hand, is expected to decline under most GCM scenarios; results for the overall Pampas region were -8 per cent, -16 per cent, -8 per cent and +2 per cent under the UKMO, GFDL, GISS and MPI-DS respectively (Magrin and Travasso, 2002), although an earlier study for the main maize production zone estimated reductions of between 20 per cent and 25 per cent based on UKMO, GFDL and GISS projections (Paruelo and Sala, 1993). For Brazil (de Siqueira et al, 2000), a 16 per cent reduction in maize yields and a 27 per cent increase in soybean yields were reported under the GISS scenario when CO2 effects are taken into consideration. Finally, for Uruguay, reductions of 14 per cent and 25 per cent in maize yields have been reported for increases of 2ºC and 4ºC in mean temperature respectively (Sawchik, 2001), although in this case CO2 effects were not considered. Some degree of uncertainty is invariably associated with such crop model projections, largely because the interactions between various climatic and nonclimatic elements and their impacts on crop production are not fully understood. However, regardless of the level of uncertainty, it appears that future climate conditions will probably be more favourable for soybeans than for maize in southeastern South America. Therefore, based on the current trend, a further expansion of soybean growing areas in this region can be anticipated. Such large-scale expansion of soybean farming, with a subsequent decline in maize production, could have significant negative consequences in the medium and long term. Potential ecological implications of this trend in soybean monoculture could include reductions in soil organic matter (García, 2003), soil compaction in superficial layers (Diaz-Zorita et al, 2002) and a noticeable increase in the use of herbicides like glyphosate (for example, in Argentina, use was 28Ml in 1997 and 100Ml in 2002) (Joensen and Ho, 2004). Such impacts could seriously threaten the system’s sustainability if current practices are not reconsidered. Effective adaptation strategies to sustain the viability of agricultural production systems in this region have therefore also become a critical necessity for the immediate future.
3232 J&J Climate Change
1/11/07
9:56 am
Page 334
334 Climate Change and Adaptation
Keeping these findings in mind, our specific objective in this study was to assess the impacts of future climate scenarios (based on the latest GCM projections – HadCM3 – under different SRES scenarios and time periods) on crop production systems in southeastern South America, specifically the Pampas region in Argentina, Uruguay and the states of Rio Grande do Sul in Brazil (Figure 19.1).3 Six sites in this region were selected representing areas with contrasting environmental conditions (from the humid subtropics in Brazil to the humid and semiarid Pampas), namely Azul, Pergamino, Santa Rosa and Tres Arroyos in Argentina; La Estanzuela in Uruguay; and Passo Fundo in
Figure 19.1 Study area and study sites Note: AZ = Azul; PE = Pergamino; SR = Santa Rosa; TA = Tres Arroyos; LE = La Estanzuela; PF = Passo Fundo.
Brazil (see Figure 19.1). Crop simulation models were used to examine the influence of increasing temperature and precipitation on future crop yields. The effect of CO2 enrichment was also assessed since some studies have shown that increased CO2 concentration in the atmosphere promotes plant growth by boosting photosynthetic activity and increasing water use efficiency (see Kimball et al, 2003). The crop yield results generated were next used in the same crop models to evaluate potential adaptation measures that could help reduce the negative impacts of climate change on maize and soybean production in this region. This study thus played an important role not only in determining the vulnerability of agriculture in this region, but also in identifying strategies that can potentially address this vulnerability.
3232 J&J Climate Change
1/11/07
9:56 am
Page 335
Maize and Soybean Cultivation in Southeastern South America 335
Future Climate Scenarios in Southeastern South America Future climatic scenarios for each of the six study sites were determined with the help of the HadCM3 (Johns et al, 2003) GCM model, which was found to reproduce local conditions in this region with greater confidence than other GCMs (Camilloni and Bidegain, 2005). Model runs were conducted for the two socioeconomic scenarios, A2 and B2, from the IPCC Special Report on Emissions Scenarios (SRES). Monthly values for the variables maximum temperature, minimum temperature and precipitation were obtained from the model results for three 30-year time periods (centred on 2020, 2050 and 2080), and the monthly rate of change of each variable was obtained by comparison with the baseline period 1960–1990. These coefficients of change were next applied to the observed data (1971–2000) to obtain the future climatic scenario on a daily basis. It was observed that larger increases in temperature and precipitation, particularly for 2050 and 2080, could be expected for the SRES A2 scenario (which assumes a higher CO2 concentration) than for the SRES B2 scenario (Table 19.1 and Figure 19.2). Increases in mean temperature for the warm semester (October–March) ranged from 0.8ºC to 4.1ºC under SRES A2 and from 0.7ºC to 2.9ºC under SRES B2, depending on site and time period (Table 19.1).
Table 19.1 Projected changes in mean temperature (ºC) for the warm semester (October–March) according to HadCM3 under SRES A2 and B2 scenarios for 2020, 2050, and 2080 Mean temperature changes (October–March) HadCM3 A2 HadCM3 B2 SR TR AZ PE LE PF Mean
2020 0.9 0.8 0.8 0.9 0.8 0.9 0.9
2050 2.1 1.9 1.9 2.1 2.0 2.4 2.1
2080 3.4 3.1 3.1 3.4 3.2 4.1 3.4
2020 0.7 0.7 0.7 0.8 0.8 0.9 0.8
2050 1.7 1.6 1.6 1.7 1.5 1.8 1.7
2080 2.5 2.4 2.4 2.7 2.6 2.9 2.6
Note: AZ = Azul; PE = Pergamino; SR = Santa Rosa; TA = Tres Arroyos; LE = La Estanzuela; PF = Passo Fundo.
With respect to precipitation (Figure 19.2), the general pattern showed an increasing trend during the warm semester (October–March), with up to 253mm and 172mm rainfall increases for SRES A2 and B2 respectively. A decreasing trend was observed for the coldest months (May–August), with up to 46mm and 34mm reductions in rainfall for SRES A2 and B2 respectively.
3232 J&J Climate Change
1/11/07
9:57 am
Page 336
336 Climate Change and Adaptation
Figure 19.2 Changes in monthly precipitation (%) projected by HadCM3 under SRES A2 and B2 for 2020, 2050 and 2080
Impact on Crop Production The expected impacts of these climate scenarios on crop yields in each of the six study sites were next assessed using crop simulation models included in the Decision Support System for Agrotechnology Transfer (DSSAT) computer program4 (Tsuji et al, 1994). The crop models in DSSAT (including CERES for maize and CROPGRO for soybean) are detailed biological simulation models of crop growth and development that operate on a daily time step. They simulate dry matter production as a function of climate conditions, soil properties
3232 J&J Climate Change
1/11/07
9:57 am
Page 337
Maize and Soybean Cultivation in Southeastern South America 337
and management practices. These models are also able to simulate crop growth for variable atmospheric CO2 concentrations. The inputs required to run the models are daily weather variables, crop and soil management information (planting date, fertilizer use, irrigation and so on), cultivar characteristics, and soil profile data. Output from the models includes final grain yield, total biomass and biomass partitioning between the different plant components at harvest, among others. These models have previously been exhaustively tested in the study region at the plot and field levels with relatively low estimation errors (Guevara and Meira, 1995; Meira and Guevara, 1995; Travasso and Magrin, 2001; Sawchik, 2001; de Siqueira et al, 2000). Subsequently, they have also been used to assess the impacts of interannual climate variability and climate change in the agricultural sector of this region (Magrin et al, 1997 and 1998; Travasso et al, 1999; Magrin and Travasso, 2002; de Siqueira et al, 2000; Sawchik, 2001). Agronomic model inputs used in our study included initial water and nitrogen content in the soil profile, date of planting, plant density, sowing depth, date and rate of fertilizer application, and cultivars. Climatic inputs for the crop simulation models included observed daily maximum and minimum temperatures, precipitation, and solar radiation corresponding to the period 1971–2000 and the climate change scenarios obtained from the HadCM3 runs described above. Crop models were run under rain-fed and irrigated (water and nutrients non-limiting) conditions for different atmospheric CO2 concentrations: 330ppm (current) and those corresponding to each SRES scenario (417, 532 and 698ppm for A2, and 408, 478 and 559ppm for B2 in 2020, 2050 and 2080 respectively).
Changes in crop yields The results show a decrease in irrigated maize yields at almost all sites and under all scenarios in comparison to the baseline (1971–2000) when the direct effects of CO2 on crop growth were not considered (Figure 19.3). Yield reductions were larger for the later time periods, and were stronger under SRES A2 (up to –23 per cent) than under B2. A significant correlation was observed between changes in maize yield and temperature increases during the crop growing season (R2 = 0.74), resulting in a reduction of 5 per cent in yields per ºC temperature increase. Under the same conditions, irrigated soybean yields were, however, less affected, with yield changes varying between –8 per cent and +5 per cent (Figure 19.3). The correlation between yield changes and temperature increases was weaker (R2 = 0.4) than in the case of maize and therefore the yield reduction was also smaller (a decrease of 1.8 per cent per ºC temperature increase). When the direct CO2 effects on crop growth were considered, maize yields under irrigated conditions were somewhat higher than those obtained in the absence of CO2 effects, but the increase was insufficient to offset the negative temperature effects (Figure 19.3). In contrast, huge increases in irrigated soybean yields were detected under both SRES scenarios (of up to 43 per cent and 38 per cent for A2 and B2 respectively).
3232 J&J Climate Change
1/11/07
9:57 am
Page 338
338 Climate Change and Adaptation
Figure 19.3 Changes in irrigated maize and soybean yields (%) under different scenarios and CO2 concentrations
Under rain-fed conditions and without considering the direct CO2 effects, maize yield changes ranged between –9 per cent and +9 per cent for SRES A2, and –12 per cent and 3 per cent for SRES B2. Rain-fed soybean yield changes varied between –22 per cent and 10.5 per cent for SRES A2, and between –18 per cent and 0.5 per cent for SRES B2 (Figure 19.4). When the direct effect of CO2 on crop growth was taken into account, grain yields increased for both crops but once again a greater impact was observed on soybean yields (up to 62.5 per cent increase). Thus, under the A2 scenario, when the direct effects of CO2 were considered, irrigated and rain-fed soybean yields and rain-fed maize yields were higher than current climate yields: the direct effects of high CO2 concentration and the higher spring and summer precipitation more than compensated for the negative effect of increased temperature. As expected, the changes in crop yield under the B2 scenario were in the same direction as those under the A2 scenario but smaller in magnitude.
3232 J&J Climate Change
1/11/07
9:57 am
Page 339
Maize and Soybean Cultivation in Southeastern South America 339
Figure 19.4 Changes in rain-fed maize and soybean yields (%) under different scenarios and CO2 concentrations
The differences in the response between soybean and maize can be attributed to the differences in the influence of temperature and CO2 concentration on crop growth. In soybean (a C3 plant) CO2 effects on photosynthesis are greater than in the case of maize (a C4 plant) (Derner et al, 2003).5 The soybean simulation model used in our research assumes about a 40 per cent increase in photosynthesis efficiency at a CO2 concentration of 660ppm, while the corresponding value for maize is only about 10 per cent. Consequently, the effect of temperature is more dominant in irrigated maize crops than the effect of CO2 concentration, which explains the obtained yields. On the other hand, the effect of CO2 on stomatal resistance is known to be higher in C4 than in C3 plants (Kimball et al, 2003), which contributes to a higher water use efficiency in rain-fed maize and explains the increase in yield, especially when the direct effects of CO2 are not considered.
3232 J&J Climate Change
1/11/07
9:57 am
Page 340
340 Climate Change and Adaptation
Changes in crop phenology The projected increases in temperature were also observed to lead to a shortening of the crop growing seasons (Figure 19.5). For both soybean and maize, the worst impacts were observed with the highest temperature increases (A2, 2080). Impacts were much more severe in the case of maize, since the most affected phases were planting–flowering and flowering–maturity. Under the A2 scenario, in 2080 the maize crop growing season was reduced on an average by 27 days. In the case of soybean the worst case scenario resulted in growing seasons that were only 2–7 days shorter, mostly due to reductions in the planting–flowering period. The bigger shortening of the crop growing season observed in maize is coincident with the greater reductions in grain yields.
Figure 19.5 Changes in the duration of planting–flowering (P–F) and flowering–maturity (F–M) periods, expressed as mean values for the six sites, for maize and soybean crops under different SRES scenarios and time periods
Assessment of Potential Adaptive Measures The DSSAT crop models used to assess crop yield vulnerability to changes in temperature and precipitation were also capable of evaluating climate adaptation measures. We therefore used the same crop models to examine the impact of several adaptive management practices on crop yields, which include changing planting dates, supplementary irrigation and increasing nitrogen application rates. At each site, alternate planting dates were tested for maize and soybean by advancing/delaying planting from the actual dates. Supplementary irrigation was added to both crops during the reproductive period, beginning 20 days before flowering at a rate of 20mm every 20 days. Incremental nitrogen application rates were tested in all sites for maize only
3232 J&J Climate Change
1/11/07
9:57 am
Page 341
Maize and Soybean Cultivation in Southeastern South America 341
since nitrogen application is not a current practice for soybean in this region. These measures were tested both in the presence and absence of CO2 enrichment effects.
Considering CO2 effects Changing planting dates for maize On average, earlier planting dates were observed to lead to increased maize yields under both SRES scenarios, especially for 2050 and 2080, although there were differences between individual sites (Figure 19.6). Earlier planting dates contributed to longer planting–flowering periods (Table 19.2) and earlier maturity dates. This measure thus allows maize crops to develop under more favourable thermal conditions, increasing the duration of the vegetative phase, which in turn increases grain yield. An additional possible advantage of earlier planting dates relates to the corresponding earlier crop maturity dates and therefore the earlier harvest period. Under current planting dates, maize crops are usually harvested during March–April or later, depending on the region. Future climate scenarios project important increases in rainfall for these months (see Figure 19.2), which could lead to excess water episodes that could, in turn, affect harvest and cause yield losses. The CERES model is unable to account for this effect and therefore the impact of earlier sowing dates could possibly be even higher under the climate conditions predicted by the HadCM3 GCM. Table 19.2 Length (days) of planting–flowering (P–F) and flowering–maturity (F–M) periods for maize at current planting date and 20 and 40 days earlier under SRES A2 scenario for 2020, 2050 and 2080 Current
P–F F–M
20 days before
40 days before
1971–2000
2020
2050
2080
2020
2050
2080
2020
2050
2080
86 59
82 54
77 50
73 46
91 53
85 49
81 46
101 53
94 50
88 47
Changing planting dates in soybean Even though soybean yields were less affected by temperature increases than maize, changing planting dates did result in higher yields. In three of the six sites evaluated (Azul, Santa Rosa and Passo Fundo), earlier planting dates were found to be beneficial, while in the others, delayed planting dates were found to be the best option under future conditions (Figure 19.7). Effects of nitrogen fertilization (maize only) Changes in nitrogen application rates along with optimal planting dates resulted in increased maize yields in only two of the six study sites (Passo Fundo and Santa Rosa). At these sites increases in nitrogen application rates of
3232 J&J Climate Change
1/11/07
9:57 am
Page 342
342 Climate Change and Adaptation
Figure 19.6 Maize: yield changes (%) for different planting dates (ac = current, -20 and -40 days) in the six sites under different scenarios (A2 in grey, B2 in black for 2020, 2050 and 2080) and CO2 concentrations
20 and 45kg N/ha respectively were found to be optimum for encouraging increased yields, possibly due to the more favourable future environmental conditions allowing a positive crop response to increases in nitrogen. Therefore, in the case of maize, given optimal planting dates and nitrogen application rates, mean yield increases of 14 per cent, 23 per cent and 31 per cent for 2020, 2050 and 2080 respectively are possible under the SRES A2 scenario and mean yield increases of 11 per cent, 15 per cent and 21 per cent are possible under the SRES B2 scenario. For soybean, when optimal planting dates are considered, mean yield increases of 35 per cent, 52 per cent and 63 per cent for 2020, 2050 and 2080 respectively are possible under the SRES A2 scenario, while under the SRES B2 scenario mean yield increases of 24 per
3232 J&J Climate Change
1/11/07
9:57 am
Page 343
Maize and Soybean Cultivation in Southeastern South America 343
Figure 19.7 Soybean: yield changes (%) for different planting dates (ac = current, ± 15, 30 days) in the six sites under different scenarios (A2 in grey, B2 in black for 2020, 2050 and 2080) and CO2 concentrations
cent, 38 per cent and 47 per cent are possible for the same time periods.
Without considering CO2 effects Maize When CO2 effects were not taken into consideration, yields decreased between 1 per cent and 5 per cent under future scenarios in the absence of adaptation measures. Under optimal planting dates and nitrogen application rates, the overall yield response was higher but differed from the case where CO2 effects were considered, even though the adaptation measures were exactly the same. In this case, changes in maize yields were positive under all scenarios and time periods only in Passo Fundo, Santa Rosa and Azul. Conversely, in Tres Arroyos, a generalized yield decrease was obtained, while in Pergamino and La Estanzuela both positive and negative changes were found depending on the scenario (Figure 19.8). Mean changes for the six sites ranged between 4 per cent (B2 2020) and 12 per cent (A2 2050 and 2080). These results suggest that without CO2 fertilization simple measures such
3232 J&J Climate Change
1/11/07
9:57 am
Page 344
344 Climate Change and Adaptation
Figure 19.8 Adaptation measures for maize: Yield change (%) under optimal planting dates/nitrogen rates and supplementary irrigation for the six sites without considering CO2 effects
as changes in planting dates or nitrogen rates will not be sufficient in some places. When supplementary irrigation was applied, an overall yield increase was observed with changes close to 20 per cent under all scenarios (Figure 19.8). Soybean In the absence of adaptation measures, soybean yields decreased under all scenarios (1–12 per cent). Changing planting dates led to a modest increase in yields (2–9 per cent) only for 2020 and 2050 (Figure 19.9), but the addition of supplementary irrigation strongly reverted this situation, increasing yields
Figure 19.9 Adaptation measures for soybean: Yield changes (%) under optimal planting dates and supplementary irrigation for the six sites without considering CO2 effects
3232 J&J Climate Change
1/11/07
9:57 am
Page 345
Maize and Soybean Cultivation in Southeastern South America 345
between 30 per cent (A2 2080) and 43 per cent (A2 2020) (Figure 19.9). From these results it is obvious that soybean crops would benefit much more in comparison to maize from the application of simple adaptation measures under the projected future climate conditions. This is true even when CO2 effects are not taken into consideration. However, the positive crop responses to CO2 enrichment must be treated with caution since such responses have yet to be fully understood under field conditions. Though the positive effect of increasing atmospheric CO2 concentration on photosynthesis and water use efficiency has been demonstrated for several crops (Kimball et al, 2003), such simulation studies have also been criticized by some scientists (Long et al, 2005; Morgan et al, 2005) on the grounds that they are carried out under controlled or semi-controlled conditions, which could lead to an overestimation of the effects, in particular, for soybeans, where photosynthetic efficiency was assumed to be higher than 30 per cent in model simulations. Similarly, Leakey et al (2006) have reported that under field conditions, maize crops growing under ample water and nutrient conditions showed a lack of response to increased CO2. In addition, there is also uncertainty about the response of crops to environments slowly enriched with CO2 (as would happen in the case of climate change) because of the likely acclimation (Ainsworth and Long, 2005). Past research has suggested that the initial stimulation of photosynthesis observed when plants grow at elevated CO2 concentrations may be counterbalanced by a long-term decline in the level and activity of photosynthetic enzymes as the plants acclimate to their environment, a phenomenon referred to as ‘down-regulation’ (Ainsworth and Long, 2005). Acclimation to CO2 enrichment is not included in the crop models used in our study. Irrespective of the uncertainties, the climatic changes so far have proven to be favourable to the expansion of soybean cultivation in the study area, as is obvious from current trends. However, the present high growth rate of soybean cropping and its possible continuation in the future, provided continued favourable climatic conditions, is raising important concerns with respect to its impacts on the long-term sustainability of agriculture in the region. The ecological impacts that could potentially affect the growth trends in soybean farming are discussed in greater detail in the following section.
Ecological Impacts of Soybean Cultivation The significant growth of agricultural production in southeastern South America, especially of soybean, is unfortunately occurring at the expense of the local environment in the region. It has been observed that the large scale expansion of agriculture in Argentina has created negative soil nutrient balances (nutrient removal exceeding nutrient application) and although crop fertilization is a common practice, only 37 per cent of nutrients extracted are actually replenished (García et al, 2005). Soybean is a highly nutrient extractive crop with a low level of crop residues, and the current trend of extensive soybean monoculture not only
3232 J&J Climate Change
1/11/07
9:57 am
Page 346
346 Climate Change and Adaptation
creates a deficit in soil nitrogen content (nitrogen fertilization is not a current practice for this crop) but also leads to negative soil carbon balances because of low carbon (C) inputs to the soil. Experiences in Argentina have shown that for a crop yielding 4000kg/ha, some 120kg N/ha/year and 950kg C/ha/year are lost from the system (García, 2003). Modern agricultural technologies such as direct seeding and herbicide resistant soybeans (Pengue, 2001) have also contributed to the ecological degradation in the region by further encouraging intensive soybean monoculture. The direct seeding technology was initially introduced to address issues of serious soil erosion and the subsequent loss of soil fertility since it requires no tillage and allows residues from the previous crop to remain on the ground. Although this technique has successfully reduced the rate of erosion, other problems have cropped up from the further intensification of agriculture it encourages, namely the emergence of new diseases and pests, a marked reduction in the levels of nitrogen and phosphates in the soil, and, most recently, the emergence of herbicide-resistant weeds.6 In the Pampas, there are already several types of weeds that are suspected of being tolerant to the recommended doses of glyphosate herbicide. Some of these now require a doubling of the application, leading to increased herbicide use. This further endangers the environment since herbicide run-off from the farms is already affecting adjacent ecosystems, for example, aquatic ecosystems. In order to address these issues several alternatives have been suggested, one option being the use of fertilizers to restore soil carbon and nitrogen balances. Other potential measures to improve soil nutrient levels include the use of grasses as cover crops and crop rotation using a higher proportion of corn and wheat in the rotation. The traditional method of crop–pasture rotations is also recommended for correcting soil organic matter balances and thus restoring soil carbon and nitrogen levels (García, 2004). Such crop–pasture rotations were previously the primary crop rotation method in the Pampas region of Argentina (García, 2004). Similarly in Uruguay, traditional rotations included 3–4 years of crops alternating with 3–4 years of pasture. The recent expansion of soybean cultivation has unfortunately led to a decrease in the pasture component of this system. For southern Brazil, Costamilan and Bertagnolli (2004) recommend a three-year crop rotation, including the sequence oats/ soybean, wheat/soybean and spring vetch/maize. Besides restoring soil nutrient levels, rotating crop varieties also helps to reduce the incidence of diseases, pests and weeds by breaking their cycle. A somewhat innovative option for maintaining the viability of agriculture in the region is the ‘transformation in origin’ approach, promoted by Oliverio and Lopez (2005). The transformation in origin approach suggests the cultivation of a mix of oilseeds (for example, soybean) and cereals (for example, maize) with a part of the production (for example, of maize) remaining at the place where it was cultivated. This is either used as animal feed or in local industry. This practice adds value to the primary product, as opposed to the traditional approach in which it is sold as a commodity, which, in some cases, implies important costs of transportation to ports and fiscal retentions, among other things. Additionally, the cultivation of a mix of oilseeds and cereals under
3232 J&J Climate Change
1/11/07
9:57 am
Page 347
Maize and Soybean Cultivation in Southeastern South America 347
this method also ensures that the nutrient content of the soil is better maintained. In order to test this approach, Oliverio and Lopez (2005) analysed two possible scenarios to estimate Argentina’s crop production in 2015, assuming that the trend of increasing agricultural production will continue, regardless of climate change. In the first case, they extrapolated the actual trend in planted areas (with increasing importance of oilseeds, especially soybean), and in the second, they proposed a maximum ratio of 2.5:1 between oilseeds and cereals to test the transformation in origin method. They found that even if half the cereal portion of the crop yield (for example, maize) were to be transformed in origin, economic benefits could be more than doubled. The future outlook for agricultural expansion in southeastern South America will therefore depend not only on future climatic conditions in the region but also on the viability of crop production systems, especially the maintenance of soil nutrient content. Although important remedial strategies have been suggested in studies, further research in this area could help to identify options that effectively address the combined implications of crop responses to future climatic conditions and the ecological impacts of intensive agriculture, specifically soybean cultivation, for the long-term sustainability of farming in the region.
Conclusion Future climate scenarios based on the runs of HadCM3 suggest that mean temperatures for the entire study region would increase by an average of 0.9, 2.1 and 3.4°C by 2020, 2050 and 2080 respectively, for the SRES A2 scenario. Corresponding figures for SRES B2 are 0.8, 1.7 and 2.6°C. Precipitation projections show an increasing trend during the warm semester (October– March), which encompasses the growing seasons of both maize and soybeans, and a decreasing trend during the coldest months (May–August). Changes in precipitation were stronger for the 2050 and 2080 time periods. These climatic changes are likely to have important implications for agriculture in the region according to crop model results. In the case of maize, the increased temperatures would result in shorter growing seasons and consequently in lower grain yields. However, if the effect of CO2 enrichment is accounted for, this negative impact could be greatly diminished by adjusting the crop sowing time to earlier dates. When the effect of CO2 enrichment was not considered, changes in planting dates or nitrogen application rates were not enough to improve yield, but moderate supplementary irrigation during the reproductive growth phase did lead to significant yield increases (of up to 20 per cent). The crop that would benefit the most under future climate conditions was found to be soybean. Assuming the effect of CO2 enrichment, yield increases of greater than 60 per cent could be obtained simply by modifying planting dates. In the absence of CO2 effects, supplementary irrigation was found to be necessary to ensure yield increases and increases of about 30–40 per cent were still possible.
3232 J&J Climate Change
1/11/07
9:57 am
Page 348
348 Climate Change and Adaptation
The positive impact of the future climate on soybean production, nonetheless, has important implications for the local environments. A rapid expansion of soybean farming is already underway in this region, a trend that is only expected to strengthen in the future. However, intensive soybean cultivation tends to drain the soil of critical nutrients such as nitrogen and carbon. Additionally, the use of modern technological tools such as direct seeding and herbicide resistant seeds have generated further problems of new diseases and pests, a reduction in soil nutrient levels and the emergence of herbicide resistant weeds. Contamination of aquatic ecosystems with herbicide run-off from farms is also an issue. Given the inevitability of continued large-scale expansion in soybean cultivation, the timely implementation of appropriate adaptive measures to tackle its harmful impacts becomes critical. Adaptation measures should thus promote adequate nutrient supply, crop and soil management practices, as well as weed, pest and disease control to sustain the environment and the welfare of farmers in the region. Suggested strategies include the traditional methods of crop rotation, crop–pasture rotation, mixed cropping, the use of grasses as cover crops and transformation in origin, in addition to the more common method of fertilizer application. A failure to initiate timely adaptive responses in the commercial farming sector in southeastern South America could translate into significant economic and ecological losses that could negate the benefits arising out of the new climatic conditions. Finally, the further improvement of crop models is also important in order to better understand the impact of CO2 enrichment on crop production, the interaction with weeds, pests and diseases, and the impact of excess water/flooding. This would ensure better accuracy in determining yield estimations under future conditions and help to identify more effective adaptation strategies.
Notes 1
2
3
Such models estimate agricultural production as a function of weather and soil conditions, as well as crop management by integrating current knowledge from various disciplines to predict growth, development and yield (Hoogenboom, 2000). United Kingdom Meteorological Office (UKMO); Goddard Institute of Space Studies (GISS); Geophysical Fluid Dynamics Laboratory (GFDL); and Max Plank Institute – Downscaling model (MPI-DS). It is important to note that reductions in soybean production are only expected under UKMO, which projects for the main production area a huge increase in mean temperature (7ºC) without changes in precipitation during the most sensitive period for the crop (December–February). In the Pampas region of Argentina, arable lands cover approximately 34Mha and mean annual precipitation varies from 600mm in the southwest to more than 1200mm in the northeast. Mean annual temperature displays a north–south gradient from 13.5 to 18.5°C. In Uruguay, the total area under agriculture is approximately 1.5Mha, precipitation is more or less evenly distributed, and the mean annual rainfall ranges from 1050mm in the southwest and west to 1450mm
3232 J&J Climate Change
1/11/07
9:57 am
Page 349
Maize and Soybean Cultivation in Southeastern South America 349
4
5
6
in the north. Mean annual temperature varies from 18.2ºC in the north to 16.8°C in the south and southeast. In Rio Grande do Sul, cultivated area covers about 7Mha. This region is characterized by subtropical humid conditions with rainfall evenly distributed throughout the year. Mean annual rainfall in Passo Fundo is 1788mm and mean annual temperature is 17.5ºC. DSSAT is a microcomputer software program combining crop soil and weather data bases and programs with crop models and application programs to simulate multi-year outcomes of crop management strategies. It allows users to ask ‘what if’ questions and is capable of simulating results in a mater of minutes (www.icasa.net/dssat/). C3 plants produce a three-carbon compound in the photosynthetic process and include most trees and common crops like rice, wheat, barley, soybeans, potatoes and vegetables. C4 plants produce a four-carbon compound in the photosynthetic process and include grasses and crops like maize, sugar cane, sorghum and millet. Under increased atmospheric concentrations of CO2, C3 plants have been shown to be more responsive than C4 plants (see IPCC, 2001) Intensive monoculture is associated with several known issues: due to the absence of rotation with other crop varieties or pasture, the cycle of diseases, pests and weeds cannot be broken, which then necessitates the intensive application of chemical pesticides, herbicides and so forth, which, in turn, can eventually result in the development of resistance to these chemicals over successive crop cycles. Monocropping also depletes soil nutrient levels that otherwise would have been restored under the traditional methods of crop rotation or mixed cropping.
References Ainsworth, E. A. and S.P. Long (2005) ‘What have we learned from 15 years of free-air CO2 enrichment (FACE)? A meta-analytic review of the responses of photosynthesis, canopy properties and plant production to rising CO2’, New Phytologist, vol 165, pp351–371 Barros, V., M. E. Castañeda and M. Doyle (2000) ‘Recent precipitation trends in southern South America to the east of the Andes: An indication of a mode of climatic variability’, in Southern Hemisphere Paleo- and Neo-climates. Concepts, Methods, Problems, Springer Verlag, New York Bidegain, M., R. M. Caffera, F. Blixen, V. Pshennikov, J. J. Lagomarsino, E. A. Forbes and G. J. Nagy (2005) ‘Tendencias climáticas, hidrológicas y oceanográficas en el Río de la Plata y costa Uruguaya’ [‘Climatic, hydrological and oceanographical trends in the la Plata River and Uruguayan coast’], Chapter 14 in V. Barros, A. Menéndez and G. Nagy (eds) The Climate Change in the Plata River, Centro de Investigaciones del Mar y la Atmósfera (CIMA), Buenos Aires, pp137–143 (in Spanish) Castañeda, M. E. and V. Barros (1994) ‘Las tendencias de la precipitación en el Cono sur de America al este de los Andes’ [‘Precipitation trends in the Southern cone of America East of the Andes’], Meteorológica, vol 19, no 1–2, pp23–32 Camilloni, I. and M. Bidegain (2005) ‘Climate scenarios for the 21st century’, Chapter 4 in V. Barros, A. Menéndez and G. Nagy (eds) The Climate Change in the Plata River, Centro de Investigaciones del Mar y la Atmósfera (CIMA), Buenos Aires, pp33–39 (in Spanish) Costamilan, L. M. and P. F. Bertagnolli (2004) Indicações Técnicas para a Cultura de Soja no Rio Grande do Sul e em Santa Catarina – 2004/2005 [Technical Recommendations for Soybean Crops in Rio Grande do Sul and Santa Catarina – 2004/2005], Embrapa Trigo, Passo Fundo, Rio Grande do Sul, Brazil
3232 J&J Climate Change
1/11/07
9:57 am
Page 350
350 Climate Change and Adaptation Derner, J. D., H. B. Johnson, B. A. Kimball, P. J. Pinter Jr, H. W. Polley, C. R. Tischler, T. W. Boutton, R. L. Lamorte, G. W. Wall, N. R. Adam, S. W. Leavitt, M. J. Ottman, A. D. Matthias and T. J. Brooks (2003) ‘Above- and below-ground responses of C3 C4 species mixtures to elevated CO2 and soil water availability’, Global Change Biology, vol 9, no 3, pp452–460 de Siquiera, O., S. Steinmetz, M. Ferreira, A. Castro Costa and M. A. Wozniak (2000) ‘Climate changes projected by GISS models and implications on Brazilian agriculture’, Revista Brasileira de Agrometeorologia, vol 8, pp311–320 Díaz-Zorita, M., G. A. Duarte and J. H. Grove (2002) ‘A review of no-till systems and soil management for sustainable crop production in the subhumid and semiarid Pampas of Argentina’, Soil Tillage Research, vol 65, pp1–18 García, F. (2003) ‘Agricultura sustentable y materia orgánica del suelo: Siembra directa, rotaciones y fertilidad’ [‘Sustainable agricultura and soil organic matter. Direct seeding, rotations and ferility’], www.INPOFOS.org García, F. (2004) ‘Fertilizers to sustain the production of 100 million tonnes of grain in Argentina’, www.INPOFOS.org García, F. O., G. Oliverio, F. Segovia and G. López (2005) ‘Fertilizers to sustain production of 100 million metric tons of grain’, Better Crops, vol 89, no 1 Guevara, E. and S. Meira (1995) ‘Application of CERES-Maize model in Argentina’, Second International Symposium on Systems Approaches for Agricultural Development (SAAD2), International Rice Research Institute, Los Baños, Philippines Hansen, J. I., I. Fung, A. Lascis, D. Rind, S. Lebedeff, R. Ruedy and G. Russell (1989) ‘Global climate changes as forecasted by Goddard Institute for Space Studies three-dimensional model’, Journal of Geophysical Research, vol 93, pp9341–9364 Hoogenboom, G. (2000) ‘Contribution of agrometeorology to the simulation of crop production and its applications’, Agricultural and Forest Meteorology, vol 103, pp137–157 IBGE-LSPA (2005) Levantamento Sistemático da Produção Agrícola [Systematic Survey of Agricultural Production], Instituto Brasileiro de Geografia e Estadistica, www.ibge.gov.br IPCC (Intergovernmental Panel on Climate Change) (2001) ‘Glossary of terms’, in J. McCarthy, O. Canziani, N. Leary, D. Dokken and K. White (eds) Climate Change 2001: Impacts, Adaptation, and Vulnerability, Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, UK and New York, US Joensen, L. and M. W. Ho (2004) ‘La paradoja de los transgénicos en Argentina’ [‘The paradox of transgenics in Argentina’], Revista del Sur, no 147–148, available: at www.redtercermundo.org.uy Johns, T. C., J. M. Gregory, W. J. Ingram, C. E. Johnson, A. Jones, J. A. Lowe, J. F. B. Mitchell, D. L. Roberts, D. M. H. Sexton, D. S. Stevenson, S. F. B. Tett and M. J. Woodage (2003) ‘Anthropogenic climate change for 1860–2100 simulated with the HadCM3 model under updated emissions scenarios’, Climate Dynamics, vol 20, pp583–612 Kimball, B. A., K. Kobayashi and M. Bindi (2003) ‘Responses of agricultural crops to free-air CO2 enrichment’, Advances in Agronomy, vol 77, pp293–368 Leakey, A. D. B., M. Uribelarrea, E. A. Ainsworth, S. L. Naidu, A. Rogers, D. R. Ort and S. P. Long (2006) ‘Photosynthesis, productivity and yield of maize are not affected by open-air elevation of CO2 concentration in the absence of drought’, Plant Physiology, vol 140, pp779–790 Long, S. P., E. A. Ainsworth, A. D. B. Leakey and P. B. Morgan (2005) ‘Global food insecurity: Treatment of major food crops with elevated carbon dioxide or ozone under large-scale fully open-air conditions suggests recent models may have overestimated future yields’, Philosophical Transactions of the Royal Society, vol 360, pp2011–2020
3232 J&J Climate Change
1/11/07
9:57 am
Page 351
Maize and Soybean Cultivation in Southeastern South America 351 Maarten Dros, J. (2004) ‘Managing the soy boom: Two scenarios of soy production expansion in South America’, commissioned by WWF Forest Conversion Initiative, AID Environment, Amsterdam, The Netherlands Magrin, G., M. I. Travasso, R. Díaz and R. Rodríguez (1997) ‘Vulnerability of the agricultural systems of Argentina to climate change’, Climate Research., vol 9, pp31–36 Magrin, G. O., R. A. Díaz, M. I. Travasso, G. Rodriguez, D. Boullón, M. Nuñez and S. Solman (1998) ‘Vulnerabilidad y mitigación relacionada con el impacto del cambio global sobre la producción agrícola’ [‘Vulnerability and mitigation related to global change impacts on agricultural production’], in V. Barros, J. A. Hoffmann and W. M. Vargas (eds) Proyecto de Estudio Sobre el Cambio Climático en Argentina [Country Study Project on Climate Change in Argentina], Project ARG/95/G/31PNUD-SECYT, Secretaría de Ciencia y Tecnología (SECYT), Buenos Aires, Argentina Magrin, G. O. and M. I. Travasso (2002) ‘An integrated climate change assessment from Argentina’, Chapter 10 in Otto Doering III, J. C. Randolph, J. Southworth and R. A. Pfeifer (eds) Effects of Climate Change and Variability on Agricultural Production Systems, Kluwer Academic Publishers, Boston, MA Magrin, G. O., M. I. Travasso and G. R. Rodríguez (2005) ‘Changes in climate and crop production during the 20th century in Argentina’, Climatic Change, vol 72, pp229–249 Manabe, S. and R. T. Wetherald (1987) ‘Large-scale changes in soil wetness induced by an increase in carbon dioxide’, Journal of the Atmospheric Sciences, vol 44, no 8, pp1211–1235 Meira, S. and E. Guevara (1995) ‘Application of SOYGRO model in Argentina’, Second International Symposium on Systems Approaches for Agricultural Development (SAAD2), International Rice Research Institute, Los Baños, Philippines MGAP-DIEA (2005) Ministerio de Ganadería, Agricultura y Pesca, Dirección de Estadísticas Agropecuarias, Montevideo, Uruguay, www.mgap.gub.uy Morgan, P. B., G. A. Bollero, R. L. Nelson, F. G. Dohleman and S.P. Long (2005) ‘Smaller than predicted increase in above-ground net primary production and yield of field-grown soybean was found when CO2 is elevated in fully open air’, Global Change Biology, vol 11, pp1856–1865 Oliverio, G. and G. Lopez (2005) ‘El desafío productivo del complejo granario argentino en la próxima década: Potencial y limitantes’ [‘The productive challenge of the Argentine’s grain complex for the next decade: Potential and limitations’], Fundación Producir Conservando, available at www.producirconservando.org.ar Paruelo, J. M. and O. E. Sala (1993) ‘Effect of global change on maize production in the Argentinean Pampas’, Climate Research, vol 3, pp161-167 Pengue, W. (2001) ‘The impact of soya expansion in Argentina’, Seedling, vol 18, no 3, June, GRAIN Publications Pinto Hilton, S., E. D. Assad, J. Zullo Jr and O. Brunini (2002) ‘Mudanzas climáticas: O aquecimento global e a agricultura’ [‘Climate change: Global warming and agriculture’], www.comciencia.br/ SAGPyA (2005) website of Secretaría de Agricultura, Ganadería, Pesca y Alimentos [Secretariat of Agriculture, Cattle, Fisheries and Food], Buenos Aires, Argentina, www.sagpya.mecon.gov.ar Sawchik, J. (2001) ‘Vulnerabilidad y adaptación del maíz al cambio climático en el Uruguay’ [‘Vulnerability and adaptation of maize to climate change in Uruguay’], available at www.inia.org.uy/disciplinas/agroclima/publicaciones/ambiente/ Travasso, M. I., G. O. Magrin and M. O. Grondona (1999) ‘Relations between climatic variability related to ENSO and maize production in Argentina’, Proceedings of 10th Symposium on Global Change Studies, American Meteorological Society, Boston, MA, pp.67–68
3232 J&J Climate Change
1/11/07
9:57 am
Page 352
352 Climate Change and Adaptation Travasso, M. I. and G. O. Magrin (2001) ‘Testing crop models at the field level in Argentina’, in Proceedings of the 2nd International Symposium Modelling Cropping Systems, 16–18 July, Florence, Italy, pp89–90 Tsuji, G. Y., G. Uehara and S. Balas (1994) Decision Support System for Agrotechnology Transfer (DSSAT v3.0), International Benchmark Sites Network for Agrotechnology Transfer (IBSNAT), Department of Agronomy and Soils, University of Hawaii, Honolulu, HI Wilson, C. A. and J. F. B. Mitchell (1987) ‘A doubled CO2 climate sensitivity experiment with a global climate model including a simple ocean’, Journal of Geophysical Research, vol 92, pp13,315–13,343
3232 J&J Climate Change
1/11/07
9:57 am
Page 353
20
Fishing Strategies for Managing Climate Variability and Change in the Estuarine Front of the Río de la Plata Gustavo J. Nagy, Mario Bidegain, Rubén M. Caffera, Walter Norbis, Alvaro Ponce, Valentina Pshennikov and Dimitri N. Severov
Introduction The Río de la Plata estuary supports important artisanal and coastal fisheries in Uruguay and Argentina. The estuary and fisheries have been substantially influenced by human activities in recent decades and are vulnerable to climate extremes and changing precipitation patterns caused by climate change and variability (Camilloni and Barros, 2000; Nagy et al, 2002a; Nagy et al, 2006a). A self-sufficient artisanal fleet based on the Uruguayan shore at Pajas Blancas exploits fisheries within the estuarine frontal system. The fishermen of this artisanal fleet are impacted by changes in the fisheries that are driven by shifts in the location of the front, which are related to climate variations, and by symptoms of eutrophication, which are associated with human activities in the watershed but can be triggered by climatic events. Though subject to highly variable fish catch and incomes, until recently, the fishermen showed resilience over a wide range of conditions. In more recent years, however, they have been less resilient to the stresses to which they have been exposed. We examine in this chapter strategies used by the artisanal fishing fleet to cope with past and current variability, called Type I adaptation. In a previous paper (Nagy et al, 2006a) we analysed the vulnerability of the fishery and the fishing fleet to the effects of the El Niño Southern Oscillation (ENSO). Here, we focus on strategies for coping with and adapting to the effects of ENSO. Our motivation for examining Type I adaptation is a belief that observed responses to past socioeconomic and environmental changes can serve as analogues for social adaptation to future climate change (Stockholm Environment Institute, 2001; Easterling et al, 2004). Based on our assessment of experiences with Type I adaptation, we suggest strategies for adapting to future climate change, or Type II adaptation.
3232 J&J Climate Change
1/11/07
9:57 am
Page 354
354 Climate Change and Adaptation
The main questions addressed in the chapter are as follows: • • •
Is Type I adaptation adequate to cope with current climate variability in the coastal fishery? Do temperature increases of 1–2°C and precipitation and river flow changes of ~5–20 per cent, which are plausible scenarios by 2050, represent threats to the sustainable livelihoods of coastal fishermen? Are information and knowledge generation, access and communication adequate to enable Types I and II adaptation?
Adaptation, Sustainable Livelihood and Local Knowledge Adaptation is the process by which stakeholders reduce the adverse effects of climate on their livelihoods. This process involves passive, reactive and anticipatory adjustments of behaviour and economic structure in order to increase sustainability and reduce vulnerability to climate change, variability and weather/climate extremes (modified from Burton, 1996 and 1997; Smit, 1993; Smit et al, 2000). An action is effective when it avoids a potential impact (Ionescu et al, 2005). In the context of our study of an artisanal fishery of the Río de la Plata estuary, adaptation occurs in response to two physical processes: changes in river flow and winds, and associated displacement of the frontal zone. We examine two categories of adaptation called Type I and Type II by Burton (2004). The former refers to strategies for managing past and current climate-related stresses without considering future climate change. Most of the adaptation that is presently done is Type I. Type II adaptation refers to strategies that explicitly take into account potential future changes in climate. Because climate change risks, future scenarios and uncertainty have still not been factored into many development decisions, not much Type II adaptation has taken place. A sustainable livelihood is a dynamic set of capabilities, assets and activities required for a means of living (DFID, 2000). The assets are human, physical, social, financial and natural capitals. Livelihoods are considered to be sustainable when they are resilient in the face of external shocks and stresses, are not dependent on external supports, and maintain the long-term productivity of natural resources. The sustainable livelihood of small-scale and subsistence fisheries is strongly associated with local and traditional knowledge, the way they organize themselves to manage natural resources, and the improvement of participatory processes and governance (FAO, 2000). Knowledge is at the heart of economic growth and sustainable development. Understanding how people and societies acquire and use it – and why they sometimes fail to do so – is essential to improving people’s lives, especially the lives of the poor. The fisheries sector is particularly rich in custom, tradition and local knowledge, reflecting these in its communities and their established beliefs and practices. The location and seasonality of fishing grounds and fishing are all facets of this knowledge fund. The proximity to the natural resource base has
3232 J&J Climate Change
1/11/07
9:57 am
Page 355
Fishing Strategies for Managing Climate Variability and Change 355
a dominating influence on the culture and thinking of the fishing community (FAO, 2000). Conservation and management decisions for fisheries should be based on the best scientific evidence available, also taking into account local knowledge of the resources and their habitat (FAO, 2000). Some of the characteristics of local knowledge (Studley, 1998) which are attributes of the Pajas Blancas community of fishermen are 1) it is linked to a specific place, 2) it is dynamic in nature, and 3) it belongs to a group of people who live in close contact with natural systems.
Framework and Methods Our approach is based on a coupled human–environment system framework developed at Clark University and the Stockholm Environment Institute for assessment of vulnerability and adaptation (Kasperson and Kasperson, 2001; Kasperson et al, 2002). The approach, shown schematically in Figure 20.1, has been called a second-generation method for climate change assessment (Leary and Beresford, 2002). It is based on six concepts: 1) to determine what responses reduce risks, 2) to investigate causes of vulnerability, 3) to take into account social causes of vulnerability, 4) to consider multiple stresses, 5) to use recent experiences as analogues and 6) to treat adaptation measures centrally.
Figure 20.1 Vulnerability and adaptation framework Source: Leary and Beresford (2002), adapted from Kasperson et al (2002).
3232 J&J Climate Change
1/11/07
9:57 am
Page 356
356 Climate Change and Adaptation
We also took into account the FAO Report No 639 on fisheries (FAO, 2000), Adaptation Policy Frameworks for Climate Change (Lim et al, 2005), our previous assessment of vulnerabilities (Nagy et al, 2006a) within the context of the multilevel indicator of vulnerability to climate variability and change developed by Moss (1999), and a classical driver-Pressure-State-Response (d-PSIR) framework (Nagy et al, 2006a and b). Likewise, we assume that reducing vulnerability requires enhancement of adaptive capacity. We used a combination of primary and secondary information described in Nagy et al (2006a), including: 1 in situ and remote observations to follow the location of the estuarine fronts; 2 climate scenarios from global circulation models; 3 expert judgement of the authors based on primary and secondary information, indicators such as ENSO SST 3,4, salinity, fish yields and wind speed; and 4 cost–benefit valuation of direct use of artisanal fisheries/fishing activity. Social vulnerability was assessed from secondary information (Hernández and Rossi, 2003; Spinetti et al, 2003; Norbis et al, 2005; Nagy et al, 2006a).
Variability of the Estuarine Front and Fishery Resources The estuary of the Río de la Plata is characterized by a circulation and stratification pattern formed by the penetration of saltier and denser marine waters riverward along the bottom while freshwater inflow moves seaward along the surface. The salt intrusion limit is located within two front lines, shown in Figure 20.2: the main turbidity front (MTF) and the secondary main front (SMF) (Lappo et al, 2005; Severov et al, 2003). This feature sustains ecological processes (for example, nutrient assimilation and denitrification), services (for example, carbon dioxide fixation and fish reproduction) and fisheries. The estuarine frontal system displaces anywhere between 10 and 200 kilometres either towards the river or towards the sea, as a function of variations in river inflow at seasonal to interannual timescales and variations in onshore (S–SE) and offshore (W–NW) winds at synoptic (1–10 days) and seasonal timescales (Framiñán and Brown, 1996; Severov et al, 2003, 2004; Nagy et al, 2006a). Frequency patterns of these winds have changed over the last few decades, with an increase in onshore (E–SE) winds (Pshennikov et al, 2003; Escobar et al, 2004; Bischoff, 2005). Variations in river inflow are associated in part with ENSO variability. Strong El Niño (in other words >2.0 Sea Surface Temperature-SST 3.4) sustained for three or more months induces high rainfall and increased river flow. This increase in freshwater displaces the frontal system seaward away from the fishermen’s land site. During strong La Niña (<2.0 SST 3.4), rainfall and river flows are below average and the front displaces riverward. Freshwater inflow to the estuary, from the River Uruguay and total are shown
3232 J&J Climate Change
1/11/07
9:57 am
Page 357
Fishing Strategies for Managing Climate Variability and Change 357
Figure 20.2 River flow corridors and fronts of the Río de la Plata Note: The estuarine frontal system is demarked by MTF and SMF, the shown locations of which are based on a composite image from Sea-WIFS satellite observations and in situ measurements for 17 November 2003. The Pajas Blancas and San Luis fishing sites are located 140 and 240km downriver from Colonia respectively.
for selected years in Table 20.1. Usually, the percentage of inflow from the River Uruguay increases with total freshwater inflow. Flooding of the River Uruguay occurs mainly in the northern corridor in October–November, followed by low water from December to February. Flood events were triggered by El Niño in 1997 and 2002, and low flow events by La Niña in 1988 and 1999, which respectively pushed the frontal zone far downriver and upriver from the land base. Table 20.1 Freshwater inflow to the Río de la Plata from the River Uruguay and total Period 1987–1988 1988–1989 1990–1999* 2000 2002
ENSO state
River Uruguay flow (103m3/s)
Total Inflow (103m3/s)
Percentage from River Uruguay
2.9 1.5 5.6 4.7 9.2
19.2 14.0 26.0 20.1 26.8
15 11 21 23 35
La Niña La Niña El Niño
Note: *Average annual flows for 1990–1999.
Fish populations track movements of the front. For example, whitemouth croakers (Micropogonias furnieri), the main fish resource, migrate to the
3232 J&J Climate Change
1/11/07
9:57 am
Page 358
358 Climate Change and Adaptation
estuarine front to spawn in the bottom waters of the front from October to January. Movements of the front affect the spatial distribution of fish and recruitment of juveniles, which, in turn, impact the artisanal fleet through changes in navigation distances to fishing grounds, fishing cost and fish catch. The estuarine waters are also subject to environmental changes such as oxygen deficit (hypoxia) and harmful algal blooms from eutrophication. These changes are associated with human activities within the watershed and can be triggered by climatic stimuli, such as floods and droughts, which are partly associated with ENSO variability on an interannual timescale (Nagy et al 2002b and 2006a). The symptoms of eutrophication, as well as changes in wind climate, water temperature and the vertical structure of the frontal system, impact water quality, the availability and accessibility of fish resources, and the benefits and costs of fishing (Nagy et al, 2006a and b).
Impacts of Variability on Fishing Activity and Incomes The artisanal fleet exploits the croaker and other fisheries within 3–4km of the Uruguayan coast at the Northern corridor close to the MTF. The main fishing community is based at Pajas Blancas, about 140km downriver from Colonia. Fishing activity is carried out year-round, the peak being associated with the croakers’ migration to the estuarine front to spawn during spring and early summer. This period accounts for more than 80 per cent of annual catch and incomes (Figure 20.3), after which, from January to September, many fishermen migrate to seaward fishing sites at San Luis, 240km downriver from Colonia, or look for alternative income sources.
Figure 20.3 Long-term gross income of fishermen (local currency, 1999) Note: Average: black; strong ENSO years: light grey; maximum: dark grey. 1 = October, 12 = September (from Nagy et al, 2003). The primary peak, October through December, is associated with croakers; the secondary peak in May is associated with other species.
3232 J&J Climate Change
1/11/07
9:57 am
Page 359
Fishing Strategies for Managing Climate Variability and Change 359
The first studies of the Pajas Blancas fishery were carried out from 1987 to 1989. During the 1988–1989 fishing period, both the length of the reproductive period and the number of captured fishes were less than in 1987–1988 and total catch was 60 per cent down (Acuña et al, 1992). Norbis (1995) analysed captures from the 1987–1988 fishing period and concluded that the catch was greater after southeast winds (postfrontal period) and when north/northeast winds blow, but diminished when south/southeast winds greater than 8m/s or west/southwest winds, which oppose river discharge, prevailed. These studied fishing periods were characterized by moderate (1987–1988) and very low river discharge (1988–1989), especially of River Uruguay flow, the latter associated with the strong La Niña in 1988 (see Table 20.1). As described previously, movements of the front impact the fisheries and the economic returns on fishing. This is most evident and dramatic in strong ENSO years. During these years, net incomes of the fishermen are reduced to about 40–70 per cent, or 60 per cent on average, relative to the long-term average for the 1988–2001 period (Norbis et al, 2005; Nagy et al, 2006a and b). Decreases in net income are mainly the result of a shortened peak fishing period, due both to the inaccessibility and small size of fish. Strong La Niña events such as 1988–1989 and 1999–2000 are severe shocks that pose a big threat to fishermen’s adaptive capacity and sustainable livelihood (Norbis et al, 2005). Until 2001–2002, fishermen had shown that they were resilient to stresses within a wide coping range and maintained the long-term productivity of resources. However, they were partly dependent on external support (they received some money in advance for fuel in exchange for future catch). According to the criteria of Pittaluga et al (2005) for poverty and vulnerability of livelihood systems the Pajas Blancas community of fishermen can be placed between moderately poor and self-sufficient groups.
Coping with Current Variability: Type I Adaptation Fishermen have acquired local knowledge of the interactions between the environment and resources, as well as adjusted their own mode of behaviour in response to such environmental and resource interactions. This accumulated human capital is the basis of their adaptive capacity. Fishermen have developed adaptation strategies to cope with the increase in ENSO-related river flow variability and related locations of front lines, evolving from autonomous and reactive actions to planned private ones, without any participation of public managers or local authorities. These strategies include migration to follow movements of fish stocks and cautious fishing behaviour to avoid fishing effort on days when conditions are unfavourable. But, in spite of their human capital and good availability of resources, the fishermen lack sufficient social and financial capital to cope with climate extremes and remain vulnerable to changes in the spatial distribution and quantities of fish stocks (Nagy et al, 2006a). Typically, fishermen do not notice fluctuations of resources until they perceive changes of availability as a consequence of river flow and/or winds.
3232 J&J Climate Change
1/11/07
9:57 am
Page 360
360 Climate Change and Adaptation
Thus, extreme events drive adaptive prevention of loss (Norbis, 2003). It is probable that the very low catch attributable to La Niña in 1988 induced autonomous behavioural changes. Migration can be divided into seasonal (temporary, once a year), relocation of fishing site (longer-term) and spontaneous. The first two are planned actions and the third one is a reactive response that is forced once each 3–5 years by the lack of resources usually associated with ENSO-related river flow variability. The main causes of migration are changes in the availability of resources and migration decisions are influenced by the structure of the fishing unit. Because of the spatial and temporal changes of the estuarine front and resource availability, as well as the increasing trend of river flows, many fishermen have migrated seasonally or permanently along the coast following resources in order to reduce their long-term vulnerability to hydroclimatic fluctuations and avoid bad catch years (Hernández and Rossi, 2003; Norbis et al, 2005; Nagy et al, 2006a). A typical fishing unit is a family business composed of 4 or 5 persons with the head of family acting as the crew skipper and sometimes the owner. The boat, radio, engine and fishing gear belong to the owner, who receives 40 per cent of gross income, while the skipper and sailors receive 60 and 40 per cent of net income respectively (Spinetti et al, 2003). Fuel is supplied in advance by middlemen and the owner is in charge of maintenance costs, about 2.3 per cent of mean gross value. Cash flow is managed domestically, which leads in certain times to the fishermen not having enough working capital to perform fishing activity and needing funds in advance from the middleman. During bad years, the fishermen exchange labour and raw materials without having any contact with money, a condition called pre-economy by Spinetti et al (2003). About 50 per cent of fishermen migrate by February or March of each year as an anticipatory response to the expected seasonal decrease in croakers (Spinetti et al, 2003). Most fishermen migrate 60–100km seaward of the front following resources along the coast, going mainly to San Luis, 80km to the east of Pajas Blancas, to start the sea trout season (Hernández and Rossi, 2003; Spinetti et al, 2003). This behaviour of seasonal migration began, we believe, from fishermen looking for new resources following the very bad season of 1988/89, a strong La Niña year with extremely low river flow. During bad fishing years more than 50 per cent of fishermen migrate spontaneously riverward or seaward of the front, sometimes as early as December, as a reaction to an unexpected decrease in fish capture often associated with ENSO extremes or at other times due to very low river flows that are not associated with ENSO. Some fishermen that used to migrate seasonally to San Luis relocated permanently there during the 1980s and 1990s, as well as more recently. The causes of this planned, long-term relocation were the search for resources the whole year, lower dependence on hydro-climate variability, and increased experience and skills among sailors and skippers. The cost of this relocation was not quantified but seems to be small as there are no big obstacles to migration and many migrants have relatives and friends who host them. They do not need port facilities, only a small bay with a large beach to place their fishing boats and gear. Spinetti et al (2003) report that net monthly income at Pajas
3232 J&J Climate Change
1/11/07
9:57 am
Page 361
Fishing Strategies for Managing Climate Variability and Change 361
Blancas is 66 per cent higher than at San Luis during the peak of the fishing season (three months) but several times lower the rest of the year. Only 9 per cent of San Luis fishermen have alternative jobs, whereas this figure reaches 45 per cent at Pajas Blancas (Spinetti et al, 2003). Generally, those that migrate are young, unmarried and have less economic assets. Many of them are fishing boat sailors who receive only 20 per cent of net boat income and some of them are skippers; they are seldom owners. Their opportunity is to have resources available the whole year and a lower exposure to the effects of hydro-climatic variability. Aside from these seasonal and climate-induced migration strategies, fishermen have learned that cautious behaviour to avoid weather-related risks also reduces their vulnerability (Nagy et al, 2006a). Most skippers began their fishing careers very young; they have an average experience of 21 years and all of them took courses on navigation and safety (Spinetti et al, 2003). They have acquired the knowledge that the availability of croakers decreases when there are fresh southern to eastern winds, even before navigation is made difficult or impeded by these winds. Skippers developed and widely adopted practices of not fishing the day after fresh breezes come from the south or southeast. This behaviour, which sacrifices income but also reduces fishing costs and risks, is interpreted by Norbis (1995) and Norbis et al (2003 and 2005) as a ‘cautious behaviour’ regarding weather. The frequency of unfavourable winds has increased over the last few decades. For instance, strong E/SE winds at Buenos Aires increased in frequency by 80 per cent from 1961–70 to 1991–2000 (Bischoff, 2005). But the impact of this increase was not clearly perceived by fishermen since they regard changes in fish catch as a consequence of factors other than weather (Wells and Daborn, 1997; Spinetti et al, 2003; Nagy et al, 2006a). Because of cautious fishing behaviour, the average time spent fishing during the fishing season is 15 days per month (Spinetti et al, 2003). However, analyses of wind conditions and fish availability indicate that fishermen could increase their net incomes by increasing the number of fishing days. Norbis (1995) found that 8m/s is the lower threshold at which fish availability is decreased by southern to eastern winds. Applying this threshold to the period 1977–1986, the frequency of days with unfavourable winds during the fishing season is 24 per cent, with a range of 6–9 days and 4–6 events per month (see Figure 20.4). Based on this, we believe that the cautious practice of the fishermen could be modified by increasing the average fishing days from the current 15 (Spinetti et al, 2003) to about 18 (Norbis; 1995; Norbis et al, 2003). An early scouting of sea conditions and fish availability based on real time fishingoriented forecasting and flow of information should allow keeping safe navigation and increasing net income by at least 15 per cent. Adaptation is a risk-management strategy that is not free of cost (Easterling et al, 2004). However, there is evidence that, up to the year 2002, fishermen could afford the residual loss that occurred even after the use of reactive and anticipatory adaptation measures (Norbis et al, 2005; Nagy et al, 2006a). The latter suggested that the balance between climate and socioeconomic drivers of impacts and autonomous adaptation lay between the coping ranges. However,
3232 J&J Climate Change
1/11/07
9:57 am
Page 362
362 Climate Change and Adaptation
Figure 20.4 Unfavourable days for fishing activity on a monthly basis from October 2000 to March 2003, based on a lower threshold of 8m/s wind speed
adaptation failed and cautious behaviour disappeared during the bad fishing season of 2002/03 due to three different pressures. First, increased river flow due to a moderate to strong El Niño events displaced the estuarine front seaward by up to 150–200km from Colonia. Second, days with unfavourable winds increased by 57 per cent from 11 days per month in 2001/02 to 17.6 days per month in 2002/03 (Norbis et al, 2005; Nagy et al, 2006a). Compared to the 1970s, which averaged 7.2 unfavourable days per month, this increase is 150 per cent. The third factor was a deterioration of socioeconomic conditions due to a regional economic crisis during 2001–2003 that reduced economic welfare, increased fuel price and limited alternative sources of income (Nagy et al, 2003). Thus, many fishermen took the risk of fishing under unfavourable conditions (southeast and eastern winds), which ultimately resulted in maladaptive practice (Norbis et al, 2005; Nagy et al, 2006a). Fish yields were very bad because of several factors – displacement of the front; increased cost of navigation (due to both fuel price and distance), shortened fishing period, lower number of good fishing days and reduced availability of resources – all of which led to a high number of ‘non-fishing trips’ at the beginning of the season and forced many fishermen to migrate to eastern sites such as San Luis before January (non-fishing trips are defined as those where the cost of navigation equals or exceed the benefits of catch). Was this failure of human capital an anticipation of future difficulties to adapt (Type II adaptation) to increasing changes? It seems that fishermen will need to adopt further adaptation strategies in the event of climate variability and change, as well as increases in non-climate stresses. A key question for future research is thus whether adaptation benefits from avoiding damage
3232 J&J Climate Change
1/11/07
9:57 am
Page 363
Fishing Strategies for Managing Climate Variability and Change 363
losses are equal or greater than adaptation costs and loss of benefits. The potential adaptation options available to fishermen are discussed in the forthcoming sections using the reference period 2002–2003.
Future Climate and Environmental Scenarios To examine whether the increases in temperature, precipitation and river flow pose a threat to the sustainable livelihood of coastal fishermen, we examined both recent experiences with unfavourable climatic conditions and projections of the future from the literature. In order to construct scenarios of future climate for the southeastern South America region, we use outputs from climate experiments of general circulation models that are available from the Intergovernmental Panel on Climate Change (IPCC) Data Distribution Center. The selected models are HadCM3 (UK) and ECHAM4/OPYC3 (Germany), which have acceptable agreement with the observed sea level pressure field and are able to represent the position and intensity of the pressure systems of the region, both on monthly and annual bases (Bidegain and Camilloni, 2004; Nagy et al, 2006a). Although observed climate fields indicate that both models underestimate precipitation within the Río de la Plata basin, monthly and annual temperature fields show that, in general, both models have acceptable agreement with the observed fields. The selected climate experiments of the HadCM3 and ECHAM4/OPYC3 models are forced with the A2 and B2 scenarios of greenhouse gas emissions, which are characterized by a globalized world with high emissions and a regionalized world with low emissions respectively. They correspond to middle–high and middle–low views of future emissions respectively. Future regional climate scenarios for precipitation and temperature were constructed for 2050 and 2080 based on the range of values generated by the aforementioned models and socioeconomic scenarios. Current climate and future scenarios for the Río de la Plata basin and estuary for 2050 suggest a change in precipitation within the range +5 to +20 per cent and in temperature from +1 to +2°C. In comparison, during the last few decades precipitation increased by 20 to 25 per cent, river flows increased by 25 to 40 per cent and average temperatures rose by 0.5 to 0.8°C. Trends for future river flows are very difficult to estimate because of both the uncertainty of regional human drivers and because of the varied regional scenarios from different GCMs. Under a future scenario in which stream flow remains similar to or slightly lower than at present (no change to -10 per cent), we do not expect a significant increase in current environmental stresses on the estuarine system from the already moderately high level, with the exception of nutrient inputs (a pressure indicator for eutrophication). Of greater concern is a future scenario where river flow increases by 10 to 25 per cent, together with projected temperature increases and economic and population growth, for which significant impacts are expected in the estuarine system. Considering the fact that seasonal temperature, precipitation and stream flow cycles are not superposed, any changes should modify seasonal
3232 J&J Climate Change
1/11/07
9:57 am
Page 364
364 Climate Change and Adaptation
circulation, stratification and estuarine front location, inducing further environmental shifts with a probable increase in the degree and occurrence of symptoms of eutrophication such as hypoxia (oxygen deficit) in bottom waters and harmful algal blooms (Nagy et al, 2002a and b and 2006a).
Scenarios of Sustainability and Future Adaptation A model of fishing activity was developed by Nagy et al (2006b) based on direct use costs estimated from secondary information for 1998–1999, a good and long fishing season. Usually, the fishing period lasts 3–5 months, average favourable fishing days are 15–16/month and total number of boats is about 31 (Norbis et al, 2005; Nagy et al, 2006a). We assume, on the basis of climatic conditions and seasonal yield compared with long-term yields and income figures that the studied case – the 1998/99 fishing season – allowed the fishermen’s livelihood to be sustainable. These figures are 923 fishing navigations in 64 days with an average catch of 22 boxes, that is to say about 20 per cent greater than the long-term average capture. Modelled scenarios presented in Table 20.2 are based on observed seasons and show the long-term catch, observations for the 1998/99 fishing season, and model simulations for the minimum activity of a typical season and a bad season (as would be typical in a strong ENSO year) and for improved performance with changes in fishing behaviour. This latter case corresponds to a typical season with normal flow and wind conditions but with an adjustment in fishing behaviour to increase the number of fishing days to 18 per month over a 4 to 5 month fishing season. The model is based on the hypothesis that the actual number of fishing days is about 80 per cent of the maximum on a monthly basis within a 5 months season and estimates the number of boxes that fishermen would have under increasing number of fishing days on both monthly and seasonal bases. Both the availability of resources and current demand for fresh fish by local and regional markets allow increasing fish catch during typical years. The increase in economic return from changes in fishing behaviour is an easy low-cost adaptation. The increase in capture estimated as the maximum possible with low-cost measures should allow fishermen to recover both investment capital and losses during bad years. An information system to support such changes in behaviour is described in the next section. It must be noted that average boxes and navigation days are not independent because usually a few leading fishermen decide to navigate and are followed by others, and this number should increase if adaptation measures were taken. The model, which takes into account conservative numbers of catch per day (20 boxes), suggests that an increase in the number of fishing navigations is a key factor (provided the number of non-fishing trips remains stable). Thus, because neither weather nor climate conditions can be managed, both real-time forecasting and communication with fishermen, which should allow an increase in the number of navigations under favourable fishing conditions, seem to be wise and low-cost adaptation practices. For this, an increase
3232 J&J Climate Change
1/11/07
9:57 am
Page 365
Fishing Strategies for Managing Climate Variability and Change 365
Table 20.2 Fishing activity, capture and income: Comparison between a good year (1988–89), long-term average and model results for a low-typical year (1), a bad year (2) and results with change in fishing behaviour (3) Observed (1998/99)
Long-term average
Model 1
Model 2
Model 3
Number of boxes per boat per day
22
21
20
16
20
Fishing days
64
57
50
40
70–75
Fishing navigations
963
850
640
500–600
1150–1300
Boats per day
15
15
10
10
17–18
Income (% relative to 1999)
100
85
70
45
115–125
Source: Nagy et al (2006b).
in security of navigation and good relations with the Coast Guard and the Directorate of Aquatic Resources are needed. Regarding the question of whether the coastal fisheries’ Type I adaptation is adequate in the face of climate variability, we can say that it is adequate for coping with current variability, but it is not adequate for coping with anticipated changes in climate, climate variability and other factors that affect the fishery and livelihoods of fishermen. For this, Type II adaptations are needed.
Adapting to Future Climate Change: Type II Adaptation Past experience shows that under severe climatic pressures, fishermen are strongly impacted and reactive measures have poor results independent of socioeconomic and institutional responses. Under projected climatic, environmental and economic changes, the artisanal fishery is likely unsustainable without policy and behavioural changes to adapt to climate change. Yet nothing has been done to mobilize or enable Type II adaptations. The lack of a social network is identified as a main structural vulnerability and an obstacle to climate change adaptation (Nagy et al, 2006a). Lack of information, knowledge, capacity and financial resources are also obstacles to adaptation. Measures to enable Type II adaptation are needed at national, regional and local levels to increase resilience to climatic and environmental threats, decrease economic vulnerability, increase adaptive capacity, and decrease physical vulnerability and exposure. Potential measures, identified from a field survey by Spinetti et al (2003) and the expert judgment of the authors, are presented in Table 20.3.
3232 J&J Climate Change
1/11/07
9:57 am
Page 366
366 Climate Change and Adaptation
Table 20.3 Type II adaptation measures by scale of implementation and objectives Measures and Policies to Level
Increase resilience to natural threats
National
Decrease economic vulnerability
Increase adaptive capacity
Decrease physical vulnerability and exposure
Weather and Increase economic climate forecasting growth and employment
Integrate adaptation into planning; coastal zone management
Integrate adaptation into planning; coastal zone management
Regional
Fishing-oriented monitoring and early warning
Subsidize fuel price; increase access to fresh fish markets; refrigeration; insurance
Promote awareness, Facilitate migration association, education and flow of information
Local
Scouting
Increase access to credit and fishing activity
Stakeholder participation
Cautious fishing behaviour; improve safety, communications, boats and engines
Source: Based on priorities of fishermen (Spinetti et al, 2003), authors’ expert knowledge, vulnerability analysis (Nagy et al, 2006a and b) and UNDP-GEF criteria (Ebi et al, 2005).
Regarding the question of whether access to information and knowledge (generation, demand, information and outreach) is adequate to enable Types I and II adaptation, fishermen using wise fishing practices have had an adaptive potential that has proven to be sufficient to cope with past climate and nonclimate scenarios, and they have shown that they neither depend on the flow of information from managers and scientists nor demand it. However, empirical evidence suggests that if business-as-usual management scenarios continue, it is likely that current adaptation will not be sufficient under increased climate pressures. Management strategies will need to be periodically revised and adapted to the dynamic conditions of the climate, fish stocks, the environment and resource users, as well as to changes in the intertemporal preferences of the fishing sector. In this dynamic and uncertain environment, knowledge and information will become increasingly important. Fishermen and policymakers will need to adopt an adaptive management strategy that is supported by a system to integrate, communicate and apply multiple sources of information. Such a system, which we call an adaptation control information system (ACIS), should prioritize generation and access to knowledge and information on weather and climate forecasts, fishery resources, frontal dynamics (satellite data-based), and water quality; education, learning processes and participatory processes; early warning systems; and realtime flow of information to fishermen in an appropriated language. Local and scientific knowledge are shared between managers, institutions in charge of observations, scientists and fishermen, but there is a lack of inter-
3232 J&J Climate Change
1/11/07
9:57 am
Page 367
Fishing Strategies for Managing Climate Variability and Change 367
action and trust between them. Furthermore, some coastal observations have been discontinued in Uruguay (daily salinity in several sites). Knowledge is accumulated through observation, monitoring and analysis, which degrades if the data collection system collapses, if literacy and education levels diminish, or if basic societal infrastructure diminishes (Easterling et al, 2004). States should assign priority to undertake research and data collection in order to improve scientific knowledge of fisheries (FAO, 2000). Adaptation measures will only be effective if education, the generation of and access to information, and communications among stakeholders (fishermen, managers, local authorities and scientists) are improved. Neither the acquired scientific and local knowledge nor the improvement of early warning systems will be enough until fishermen are able to make effective use of them. An important constraint is the failure of fit between time and space scales between institutions responsible for management and actors (Norbis, 2003). Translation of climate scenarios and forecasting to advise appropriate action are not simple matters. Most of the success will depend on the adaptive potential of stakeholders, that is to say the ability to innovate and create new strategies and actions outside the actor’s customary network (Downing et al, 2004). It is imperative that fishermen participate in this process from the very beginning. These measures should be taken with the agreement of all stakeholders, including fishermen, the coast guard, the directorate of aquatic resources, the directorate of meteorology and the EcoPlata Program of Coastal Management.
Conclusions The last two decades have been characterized by increasing trends in means and variability of river flow and temperature, changes in front location and salinity, and resource availability. These environmental factors have led to an increase in interannual fluctuations of fish yields and fishermen’s income. All of these facts have imposed new challenges and threats to subsistence fishermen. Past experiences indicate that subsistence fishermen have successfully adopted autonomous adaptation options summarized as follows: 1 seasonal, definitive and reactive migrations along the coast following the resources associated with frontal displacement; and 2 cautious fishing behaviour under non-favourable wind conditions and acceptance of income loss. However, in spite of these wise Type I adaptation practices, fishermen remain highly vulnerable to severe weather and climatic conditions and their livelihood is likely to be unsustainable under increasing climate variability, environmental changes and economic pressures in the absence of proactive adaptations. Projected scenarios for 2050 will increase vulnerabilities of the artisanal fisheries, which would be heavily impacted.
3232 J&J Climate Change
1/11/07
9:57 am
Page 368
368 Climate Change and Adaptation
The current level of uncertainty about near-future climatic change and socioeconomic trends does not seem to be the main constraint to adaptation, rather it is the lack of both access to information and knowledge and public awareness about the impacts of current climate variability and extremes. This statement could be extended for the coastal zone as a whole. Education and training, participatory processes, dialogue and communication between stakeholders are needed to implement effective measures to take advantage of the generation of knowledge and information, forecasting and early warning systems. ENSO events are recurrent, and once the first indicators are known (SST 3.4), anticipatory adaptation measures should start (in other words realtime communication and early warning). Adaptive management should emphasize the integration of local and scientific knowledge, training, enhancement of data collection systems, weather and climate forecasting, and real-time communication to users (fishermen and the coast guard). The implementation of the suggested easy adaptation measures will improve livelihood quality and should augment the financial and social capitals fishermen need to access credit and markets at the national level. As a consequence of research on recent severe ENSO events and the National Communications to the UNFCCC, public awareness has been increased and new regulations on practices and plans have been planned in several sectors in order to adapt to the new climate variability conditions.
References Acuña, A., J. Verocai and S. Márquez (1992) ‘Aspectos biológicos de Micropogonias furnieri (Desmarest 1823) durante dos zafras en una pesquería artesanal al Oeste de Montevideo’ [‘Biological aspects of Micropogonias furnieri (Desmarest 1823) during two fishing years of an artisanal fisheries at Western Montevideo’], Revista de Biología Marina, University of Valparaíso, vol 27, pp113–132 Bidegain, M. and I. Camillloni (2004) ‘Performance of GCMs and climate baselines scenarios for southeastern South America’, paper presented at the Latin America and Caribbean AIACC Workshop, Buenos Aires Bischoff, S. (2005) ‘Sudestadas’, in V. Barros, A. Menéndez and G. J. Nagy (eds) El Cambio Climático en el Río de la Plata [Climate Change in the Río de la Plata], Project Assessments of Impacts and Adaptation to Climate Change, CIMACONICET-UBA, Buenos Aires, pp53–67 Burton, I. (2004) ‘Climate change and the adaptation deficit’, Occasional Paper 1, Adaptation and Impacts Research Group, Environment Canada, Ottawa, Canada Burton, I. (1996) ‘The growth of adaptation capacity: Practice and policy’, in J. Smith, N. Bhatti, G. Menzhulin, R. Benioff, M. Campos, B. Jallow, F. Rijsberman, M. I. Budyko and R. Dixon (eds) Adapting to Climate Change: An International Perspective, Springer, New York, pp55–67 Burton, I. (1997) ‘Vulnerability and adaptive response in the context of climate change’, Climate Change, vol 36, pp185–196 Camilloni, I. and V. Barros (2000) ‘The Paraná river response to El Niño 1982–1883 and 1997–1998 events’, Journal of Hydrometeorology, vol 1, pp412–430 DFID (Department for International Development) (2000) ‘The sustainable livelihoods approach (SLA) and the code of conduct for responsible fisheries (CCRF)’, Sustainable Fisheries Livelihoods Programme, www.sflp.org/eng/007/pub1/ bul1a_art1.htm
3232 J&J Climate Change
1/11/07
9:57 am
Page 369
Fishing Strategies for Managing Climate Variability and Change 369 Downing, T., S. Bharwani, S. Franklin, C. Warwick and G. Ziervogel (2004) ‘Climate adaptation: Actions, strategies and capacity from an actor-oriented perspective’, unpublished manuscript, Stockholm Environment Institute, Oxford, UK Easterling, W. A., B. H. Hurd and J. B. Smith (2004) Coping with Global Climate Change. The Role of Adaptation in the United States, Pew Center on Global Climate Change, Arlington, VA Ebi, K., B. Lim and Y. Aguilar (2005) ‘Scoping and designing an adaptation project’, in B. Lim, E. Spanger-Sigfried, I. Burton, E. Malone and S. Huq (eds) Adaptation Policy Frameworks for Climate Change, Developing Strategies, Policies and Measures, Cambridge University Press, Cambridge, UK, and New York Escobar, G., W. Vargas and S. Bischoff (2004) ‘Wind tides in the Río de la Plata estuary: Meteorological conditions’, International Journal of Climatology, vol 24, pp1159–1169 FAO (2000) ‘Fisheries Report No 639 FIPL/R639’, report of the Third Commission of the Advisory Committee on Fisheries Research, 5–8 December, Food and Agricultural Organisation, Rome Lappo, S. S., E. Morozov, D. N. Severov, A. V. Sokov, A. A. Kluivitkin and G. Nagy (2005) ‘Frontal mixing of river and sea waters in Río de la Plata’, Transactions of the Russian Academy of Sciences (Earth Science Section), vol 401, pp267–269 Leary, N. and S. Beresford (2002) Vulnerability of People, Places and Systems to Environmental Change, International START Secretariat, Washington, DC Kasperson, J. X. and R. E. Kasperson (2001) ‘International Workshop on Vulnerability and Global Environmental Change: A workshop summary’, Stockholm Environment Institute, Stockholm, Sweden Kasperson, J. X., R. E. Kasperson, B. L. Turner II, W. Hsieh and A. Schiller (2002) ‘Vulnerability to global environmental change’, in A. Diekmann, T. Dietz, C. Jaeger and E. Rosa (eds) The Human Dimensions of Global Environmental Change, MIT Press, Cambridge, MA Lim, B., E. Spanger-Siegfried, I. Burton, E. Malone and S. Huq (eds) (2005) Adaptation Policy Frameworks for Climate Change: Development Strategies, Policies and Measures, Cambridge University Press, Cambridge, UK, and New York Moss, R. (1999) ‘Vulnerability to climate variability and change: Framework for synthesis and modeling: Project description’, Battelle Pacific Northwest National Laboratory, Richland, WA Nagy, G. J., M. Gómez-Erache and A. C. Perdomo (2002a) ‘Río de la Plata’, in T. Munn (ed) The Encyclopedia of Global Environmental Change. Volume 3: Water Resources, John Wiley & Sons, New York Nagy, G. J., M. Gómez-Erache, C. H. López and A. C. Perdomo (2002b) ‘Distribution patterns of nutrients and symptoms of eutrophication in the Río de la Plata estuarine system’, Hydrobiology, vol 475–476, pp125–139 Nagy, G. J., G. Sención, W. Norbis, A. Ponce, G. Saona, R. Silva, M. Bidegain and V. Pshennikov (2003) ‘Overall vulnerability of the Uruguayan coastal fishery system to global change in the estuarine front of the Río de la Plata’, The Open Meeting Human Dimensions of Global Environmental Change, Montreal, Canada Nagy, G. J., M. Bidegain, R. M. Caffera, F. Blixen, G. Ferrari, J. J. Lagomarsino, C. H. López, W. Norbis, A. Ponce, M. C. Presentado, V. Pshennikov, K. Sans and G. Sención (2006a) ‘Assessing climate variability and change vulnerability for estuarine waters and coastal fisheries of the Río de la Plata’, AIACC Working Series Paper No 22, International START Secretariat, Washington, DC Nagy, G. M. M., Bidegain, R. M. Caffera, J. J. Lagomarsino, W. Norbis, A. Ponce and G. Sención (2006b) ‘Adaptive capacity for responding to climate variability and change in estuarine fisheries of the Río de la Plata’, AIACC Working Paper No 36, International START Secretariat, Washington, DC Norbis, W., J. Verocai and V. Pshennikov (2003) ‘Activity of the artisanal fishing fleet
3232 J&J Climate Change
1/11/07
9:57 am
Page 370
370 Climate Change and Adaptation in relation to meteorological conditions during the October 1998 to March 1999 fishing season’, in Vizziano et al (eds) Research for Environmental Management, Fisheries Resources and Fisheries in the Saline Front, vol 14, EcoPlata, Montevideo, Uruguay, pp191–197 Norbis, W., A. Ponce, D. N. Severov, G. Saona, J. Verocai, V. Pshennikov, R. Silva, G.Sención and G. J. Nagy (2005) ‘Vulnerabilidad y capacidad de adaptación de la pesca artesanal del Río de la Plata a la variabilidad climática’ [‘Vulnerability and adaptive capacity of the artisanal fisheries of the Río de la Plata to climate variability’], Chapter 18 in V. Barros, A. Menéndez and G. J. Nagy (eds) El Cambio Climático en el Río de la Plata [Climate Change in the Río de la Plata], Project Assessments of Impacts and Adaptation to Climate Change, CIMA-CONICETUBA, Buenos Aires, Argentina, pp173–187 Norbis, W. (1995) ‘Influence of wind, behaviour and characteristics of the coraker (Micropogonias fournieri) artisanal fishery in the Río de la Plata’, Fisheries Research, vol 22, pp43–58 Norbis, W. (2003) ‘Adaptive management of fisheries in the Río de la Plata’, paper presented at the AIACC Latin America and Caribbean Workshop, San José de Costa Rica, May, available at www.aiaccproject.org/meetings/San Jose_03 Pittaluga, F., N. Salvati and C Seghieri (2005) ‘Livelihood systems profiling: Mixed methods for the analysis of poverty and vulnerability’, LSP working series paper, FAO. Rome Pshennikov, V., M. Bidegain, F. Blixen, E. A. Forbes, J. J. Lagomarsino and G. J. Nagy (2003) ‘Climate extremes and changes in precipitation and wind-patterns in the vicinities of Montevideo, Uruguay’, paper presented at the AIACC Latin America and Caribbean Workshop, San José de Costa Rica, May, available at www.aiaccproject.org/meetings/San Jose_03/Session 6/ Severov, D. N., G. J. Nagy and V. Pshennikov (2003) ‘SeaWiFS fronts of the Río de la Plata estuarine system’, Geophysical Research Letters, vol 5, p1914 Severov, D. N., G. J. Nagy, V. Pshennikov and E. Morozov (2004) ‘Río de la Plata estuarine system: Relationship between river flow and frontal variability’, paper presented at 35th COSPAR Scientific Assembly, Paris, France, July Smit, B. (1993) Adaptation to Climatic Variability and Change, Environment Canada, Guelph, Ontario, Canada Smit, B., I. Burton, R. J. T. Klein and J Wandel (2000) ‘An anatomy of adaptation to climate change and variability’, Climatic Change, vol 45, no 1, pp233–251 Spinetti, M., G. Riestra, R.Foti and A. Fernández (2003) ‘Activity of the artisanal fishery in the Río de la Plata: Socio-economic structure and situation’, in Vizziano et al (eds) Research for Environmental Management, Fisheries Resources and Fisheries in the Saline Front, vol 17, EcoPlata, Montevideo, Uruguay, pp231–259 Studley, J. (1998) ‘Dominant knowledge systems and local knowledge’, Mountain Forum Online Library document, www.mtnforum.org/resources/library/ stud98a2.htm Wells, P. G. and G. Daborn (1997) ‘The Río de la Plata: An environmental overview’, EcoPlata Project background report, Dalhosie University, Halifax, Nova Scotia, Canada
3232 J&J Climate Change
1/11/07
9:57 am
Page 371
Index acceptability, social 288, 289 adapt now 2–3 administrative level 75, 83, 101–104, 105, 176–177 Africa cereal production 19, 131–146, 181–195 conservation strategies 28–52 disease 109–130 drought 9, 90–108, 147–162 obstacles to adaptation 11, 12 water management 53–70 weather forecasting 163–180 agriculture cereal production 19, 131–146, 181–195 drought 78, 79–80, 101 frontier expansion 297–302 rice production 228–246 variable weather 164–166 vulnerability 153 water resources 59, 181, 221, 253, 254–257, 262 agropastoralism 96 AHP see analytic hierarchical process AIACC see Assessments of Impacts and Adaptations to Climate Change alternative strategies 159, 234, 238, 240 analytic hierarchical process (AHP) 19–20, 214, 217, 222, 223–225 anticipatory strategies 286, 291, 293, 360, 368 see also planned adaptation Arba’at, Red Sea State, Sudan 96–100 Argentina 8, 11, 15, 296–314, 315–331, 332, 333, 345
artisanal fishing 11, 353–370 Assessments of Impacts and Adaptations to Climate Change (AIACC) 2, 7, 10, 13, 16, 19–21 assets, livelihood 92, 93, 94–95, 97 autonomous adaptation 100–101, 154, 271, 273, 297–302, 312, 359, 361–362, 367–368 autonomous dispersers 29, 35, 39 awareness cereal farmers 185 disease 121–122, 123, 124 estuarine fishing 368 increasing 4 lack of 11, 22 long-term trends 302, 304, 307, 312–313 Pacific Islands 270–271, 274, 276 risk 25 see also knowledge banks 304 Bara Province, Sudan 92–96 bed nets 11, 117 behavioural change strategies 290, 293 Beja pastoralists 96–100 benefits and costs see cost–benefit analysis Berg river basin, South Africa 10, 19, 53–70 biocapacity 203, 206, 209 biodiversity conservation 9, 28–52 biotic adaptation 39 Bolivia 333 Botswana 12, 71–89, 72–86 bottom–up strategies 17–18, 106, 149–151, 177, 264–278 Brazil 332, 333 breeding and selection 133, 138, 143
3232 J&J Climate Change
1/11/07
9:57 am
Page 372
372 Climate Change and Adaptation
Buenos Aires, Argentina 15, 298, 310, 311 business-as-usual strategies 136, 137, 138, 143 capability poverty 71 capacity biocapacity 203, 206, 209 building 84–85, 106, 328 disease 115–118, 121–122 drought 148 estuarine fishing 359 floods 271 increasing 7, 17, 24 lack of 80, 244–245, 276 Pacific Islands 265 poverty 81 reservoir 19, 55, 56, 57, 60, 61, 62–63, 64, 65, 66 water resources 12, 211 weather forecasting 168–170, 173–174 Cape Town, South Africa 53 capital 11, 12, 93, 95, 97, 99, 103, 155–157 see also human capital; social capital carbon dioxide fuel wood trade 82 predictions 30 yields 333, 334, 335, 337, 338, 339, 341–342, 343–344, 345, 347 The Caribbean 21, 279–295 cautious behaviour 361, 367 CBNRM see Community Based Natural Resources Management CCD see climate change damages cereal production 19, 131–146, 181–195 CEREBAL 131, 136 Chile 307 China 15, 19–20, 211–227 cholera 109–110, 111, 113, 119–122, 122, 123 civil society organizations 125 climate change damages (CCD) 19,
54–58, 64–65, 67, 136–137, 139–142, 140, 143 cocoa 166 collective strategies 15, 17, 234, 237, 241, 245, 311, 313 commercial development 81–82, 236, 238, 240, 242–243, 321, 322 common property resources 12, 44–45, 45–46, 197–198 communication 169, 174, 175–177, 276 Community Based Natural Resources Management (CBNRM) 44–45, 80, 84 community level development 17, 90–108 indigenous knowledge 71–86 Pacific Islands 273 rice production 234, 237, 238–240, 240, 241, 245 watersheds 256, 258, 259, 262 see also local level complementary indicators 216 conflict 100, 105, 218, 219 conservation 4–5, 32–33, 45, 189–190, 191, 203, 205–206 see also biodiversity conservation consultations 201–202, 204–205, 215, 225, 251–252 consumption 62, 63, 67–68, 141, 142 contract farming 320, 322–323 contractual reserves 37, 40, 42, 45 Córdoba Province, Argentina 315, 317, 319, 323–324, 326, 327–328, 329 costs caution and precaution 58, 65, 67 conservation strategies 36–38 dengue fever 287, 288, 291 extreme weather 264, 266–267 irrigation 141–142 reforestation 260–261 cost–benefit analysis 19, 53–70, 139–142, 143 credit systems 11, 24, 326, 327 crop–pasture rotations 346, 348 cross-sectoral effects 247–263
3232 J&J Climate Change
1/11/07
9:57 am
Page 373
Index 373
dams 53, 58, 60, 62–63, 104, 248–250, 322 decision making AHP method 217 community engagement 84 consultations 204–205 evaluation options 19 participation 275 tools 212 traditional 12, 75, 76, 83 water resources 221, 223, 224–225 weather forecasting 167–168 decision trees 35 deficit, adaptation 2–3, 10, 159 deforestation 301 delayed actions 18 dengue fever 10, 21, 279–295 dependency, export 85–86 determination to adapt 21–23, 25 deterministic scenarios 68 developing countries 2, 4–5 development capacity 84–85 community 17, 90–108 economic 77–81 human 11, 13–14 integrating adaptation 3, 15–18, 23, 25, 105–106 research 134, 160, 186, 209 urban water demand 60, 63, 64, 65, 66, 67 direct seeding 346, 348 disaster risk 14 disease dengue fever 21, 279–295 malaria and cholera 11, 109–130 dispersers 9, 29, 35–38, 39 diversification 159, 185, 236, 242, 321–322, 326 domain-based frameworks 216 do nothing approaches 34, 36 dredging 270, 272, 277 drinking water 120, 121, 124 droughts Africa 147, 148 coping strategies 90–108, 154–158 crop production 132
El Niño 268 government intervention 77–81 local actions 322 pastoralists 6, 196–210 vulnerability 152–153 dryland cereal farming 181–195 dynamic spatial equilibrium models 54–68 early maturing crop varieties 155, 156, 159 early warning systems disease 118, 122, 124, 291–292, 293 weather forecasting 12, 167–168, 169, 177, 234 East Africa 11, 109–130 ECHAM4 model 133, 134, 137, 138, 139, 363 economic aspects agricultural expansion 300–301 analysis tools 214 cereal production 131–146 conservation 39–40, 43–44 development 77–81 disease 114–115 estuarine fishing 364 export dependency 85–86 natural disasters 13 policy-planning models 54–68 reform 223–224 transition 197–198 education biodiversity conservation 43 cereal production 183, 185 disease 118, 285–286, 288, 290 estuarine fishing 367, 368 rangelands 11, 202–204 rural households 160 vulnerability 153 Egypt 181–195, 187–192 El Fashir Rural Council, North Darfur, Sudan 100–101, 102–103, 105 El Niño Southern Oscillation (ENSO) disease 113–114, 119, 282–283
3232 J&J Climate Change
1/11/07
9:57 am
Page 374
374 Climate Change and Adaptation
drought 268 estuarine fishing 353, 356, 357, 359, 360, 362, 368 floods 302, 303 rainfall 320 emissions scenarios see scenarios; Special Report on Emissions Scenarios, IPCC employment 78–79 enabling conditions 3, 4, 16, 21–25, 84 engineering options 220, 221 ENSO see El Niño Southern Oscillation environmental aspects autonomous adaptation 301, 312 disease 114, 119, 288, 289 estuarine fishing 358, 363–364 soybean cultivation 333, 345–347, 347–348 epidemics 112–113, 114, 119, 279, 283 equity assessment 92 estuarine fishing 353–370 Expert Choice 2000 223 export dependency 85–86 ex situ conservation 9, 34, 38, 39, 40, 46 extended-range weather forecasting 168–170, 177 extension services 158, 159, 178, 185, 186, 192 extreme weather 8, 10, 264, 274, 308–310 see also droughts; El Niño Southern Oscillation; floods; La Niña facilitated dispersers 18, 29, 36–38, 39, 40, 45, 46 family institutions 75, 76 farming see agriculture feasibility analysis 136 fertilization 19, 133, 138, 139–141, 143, 158, 341–342, 343, 344, 346 Fiji 15, 17–18, 265–266 financial level
assets 93, 94, 97, 98, 101, 102, 104, 155–156 conservation strategies 36–38 constraints 12–13, 207–208 credit access 326, 327 dengue fever 286–287 international 5, 24–25, 33, 208, 304 rice production 238, 239, 240, 242, 245 fishing 11, 153, 159, 353–370 floods Argentina 8, 301, 309 Lower Mekong river basin 6, 229–230 Pacific Islands 266–270, 271, 276, 277 planned adaptation 302–304 vulnerability 327–328 water resources 254 food security 41, 100–101, 134–135, 140, 143, 144, 155, 203, 209, 242 shortages 147–148 storage 154, 155, 156, 159 veld products 81, 82 weather forecasting 163–180 forage production 198, 199–200 forecasts see weather forecasting forests 20, 253, 254–257, 258, 259, 260–261, 262, 269–270, 272 fuel wood trade 76–77, 82 fynbos biome 29, 41 The Gambia 19, 131–146 game-ranches 31, 42, 45, 78 gated communities 311 GEF see Global Environment Facility general circulation models 59, 133, 188, 199, 333, 363 Gireighikh Rural Council, Bara Province, Sudan 92–96 global circulation models 29, 59, 133, 134, 356 Global Environment Facility (GEF) 208 goal-based indicators 216
3232 J&J Climate Change
1/11/07
9:57 am
Page 375
Index 375
González, Tamaulipas, Mexico 315, 316, 319, 321–323, 326–327, 329 governance 12–13, 75, 76, 83, 84, 85, 96, 275–276 government level cereal production 133, 143–144 cooperation leadership 23 disease 125 floods 271 intervention 77–81 obstacles to adaptation 12–13 planning 208 priorities 274–275 support 160, 242, 322, 324, 327, 328, 329 top–down approaches 272 see also national level; policy level grasslands 196 growing seasons 340 HadCM3 models 133–134, 137, 138, 139, 143, 334, 335, 347, 363 hail storms 323 hardveld, Botswana 72–86 hazards, adapting to 1–2 health care 6, 114–115, 117–118, 285–287 Heihe river basin, China 15, 19–20, 211–227 herbicide resistance 346, 348 heterogeneity 41 horticultural production 144 household level Nigeria 11, 147–162 rice production 231, 232–233, 235, 236, 240, 241 housing 10, 270, 311 human capital drought 93, 95, 97, 99, 101, 103, 104, 106 estuarine fishing 359, 362–363 vulnerability 153, 157 human development 11, 13–14 human–environment system frameworks 355 incentives 43–44, 272, 291, 322
income 71, 78–79, 84–85, 358–359, 360, 365 indicators adaptation options 216–217, 219, 223 drought resilience 92 economic 140, 141, 142 vulnerability 151, 152, 284 indigenous knowledge 71–89, 149–150, 231–234 information accessibility of 328 cereal production 186 disease 122, 124 estuarine fishing 366, 368 interventions 325 lack of 11 long-term trends 313 weather forecasting 158, 169, 174, 175–177, 324 infrastructure 11, 124, 161, 309, 310, 325 initial survey tools 213–214 insecticides 286, 288 insecticide treated nets 11, 117 institutions change 160, 161, 207 governance 275–276 strengthening 4, 104, 105 traditional 12, 15, 71–89 watersheds 252, 255, 256, 257–261, 262 insurance 24, 267, 322–323, 323–324, 325, 327 integrated approaches 3, 15–18, 23, 25, 105–106, 214–225 international level finance 24–25, 33, 208, 304 weather forecasting 168, 169 International Union for the Conservation of Nature (IUCN) 30, 31, 32 Intertropical Convergence Zone (ITCZ) 170, 172, 173 intervention, government 77–81 irrigation cereal production 19, 133,
3232 J&J Climate Change
1/11/07
9:57 am
Page 376
376 Climate Change and Adaptation
138–139, 141–142, 143, 144, 181–195 flooding 270 maize and soybean yields 337, 338, 340, 344 water management 12, 59–60 ITCZ see Intertropical Convergence Zone IUCN see International Union for the Conservation of Nature Jamaica 10, 283–284, 284, 285–287, 288 Kairouan, Tunisia 182–186 KARP see Khor Arba’at Rehabilitation Project Kenya 114, 117 Kgotla system 12, 75, 76, 83, 84, 85 Khor Arba’at Rehabilitation Project (KARP) 97–99 kinship networks 11 knowledge increasing 4, 25 indigenous 71–89, 149–150, 231–234 lack of 11–12, 22 local 354–355, 366, 367, 368 scientific 297, 367, 368 traditional 12, 354–355 water resources 220 see also awareness Kula Field, Thailand 236–240 Lake Victoria, East Africa 109–130 land availability 134 ownership 197, 202, 207, 316, 317 use 253, 297–302, 332 language 158, 175–176 La Niña 255, 268, 283, 320, 356, 357, 359, 360 Lao People’s Democratic Republic 17, 231–236, 232, 233, 241–245 La Plata river basin 11, 297, 302–304, 353–370 leadership 23, 101–104, 105
leaf simulations 135, 136 Limpopo river basin, Botswana 12, 71–89 livelihood assets 92, 93, 94–95, 97 livelihood systems approaches 71–89, 151–161 livestock changing practices 159, 160 diversification 185, 322 herders vulnerability 153 options evaluation 20, 196–210 ownership 12, 75, 76, 79, 179 see also pastoralism local level 177, 221, 241, 275, 315–331, 354–355, 366, 367, 368 see also community level logging 269–270 long-term climate trends 296–307, 312–313 Lower Mekong river basin 6, 228–246 mafisa system 12, 75, 76 maize cultivation 332–352 malaria 11, 109, 110, 111–118, 122, 123–124 markets biodiversity conservation 43, 46 crop value 142 efficient water 58, 60, 62, 65, 66, 67 liberalization 320 marriage 75, 76 masked climate trends 304–310 matrix management 34, 37, 41–45 means to adapt 24, 25 medical treatment 117–118, 121 medicinal plants 77, 81, 82 Mersa Matrouh, Egypt 192 meteorological organizations 168–170, 176 Mexico 11, 315, 316, 319, 320 midseason dry spells 229 migration corridors 40–41, 45 estuarine fishing 359, 360–361, 367
3232 J&J Climate Change
1/11/07
9:57 am
Page 377
Index 377
obligatory dispersers 35–36 obstacles to adaptation 105 seasonal 202, 236, 360, 367 Millennium Development Goals 13–14 millet production 137, 139–141 modelling, crop 135–136 Mongolia 11, 12, 20, 196–210 monitoring 291–292 mortality 109, 115, 116, 121, 199 mosquitos 116–117, 286 multi-criteria evaluation approaches 214, 215, 217, 225 multiple cropping 166–167 multiple measures approaches 21, 105 multi-stakeholder approaches 251–252 national level 236, 238, 239, 240–241, 241–242, 366 see also government level; policy level natural resources community-based management 44–45, 80, 84, 93, 94, 97, 98, 102 conservation 4–5, 203, 205–206 degradation 12, 92–93 dependent communities 71–72, 85–86 lack of 104 Navua township, Fiji 15, 265–266 NBA see net benefits of adaptation neighbour effects 288, 289 net benefits of adaptation (NBA) 58, 136–137 networks 11, 22, 206, 311, 365 NGOs see non-governmental organizations Nigeria 11, 12, 147–162, 163–180 Nile Delta region 187–192 nitrogen fertilization 138, 341–342, 343, 344 no-hopers 9, 29, 39 non-cereal production 144 non-engineering options 220–221
non-governmental organizations (NGOs) 85, 273 non-reserve land 42, 43 North Africa 181–195 North Darfur, Sudan 11, 100–101, 102–103, 105 Northern Nigeria 147–162 obligatory dispersers 9, 29, 35–38 obstacles to adaptation 10–13, 21, 101–105 off-farm measures 235 on-farm measures 232–233, 236–238, 240 opportunities for adaptation 101–105, 158–161 ownership conservation areas 32–33, 45 land 197, 202, 207, 316, 317 livestock 12, 75, 76, 79, 197 Pacific Island 17–18, 264–278 palm fruits 165 Pantabangan–Carranglan watershed, The Philippines 10, 20, 247–263 partial dispersers 9, 29 participation bottom–up approaches 273 decision-making 275 determination to adapt 22–23 estuarine fishing 367 farmers organizations 324 integrated assessment 215 natural resource management 93 people at risk 5 watersheds 251 pastoralism 6, 96–100, 196–210 see also livestock peaceparks 33–34 perceptions 118, 120, 121, 228–229, 274–275, 306, 323 persisters 9, 29, 34–35, 39 Phane caterpillars 77, 82 phenology of crops 340 The Philippines 10, 20, 247–263 photosynthetic activity 135–136, 345
3232 J&J Climate Change
1/11/07
9:57 am
Page 378
378 Climate Change and Adaptation
physical assets 93, 94, 97, 98, 101, 102, 104, 153, 157 place-based approaches 5–6, 228–246 planned adaptation 297, 302–304 see also anticipatory strategies planning 40–45, 46, 176, 208 planting dates 340, 341, 342, 343, 344 policy level analysis 212, 213–225 consultations 204–205 drought 77–79 estuarine fishing 366–367 flooding 302–303 lack of support 329 local needs 315 matrix management 43–44 obstacles 104 planning models 54–68 rice production 238 rural poverty 157–158 scenarios 134–135 support 160 watersheds 261–262 see also government level; national level political level 75, 160, 196, 197–198 population 114, 134 poverty development 14, 15 disease 114–115, 119–120, 124, 285 obstacles to adaptation 11 relief measures 81 rural policies 157–158 sustainable livelihoods 91–92 vulnerability 5, 14, 71, 152–153 precipitation Argentina 8, 305, 308, 319 China 219 crop production 184, 188, 192, 193, 338 disease 119, 280, 281, 282, 283 global changes 7–8 Lower Mekong 244 Mexico 319, 320 Mongolia 198, 199, 200
Nigeria 148 Pacific Islands 264, 267, 268–269 Pantabangan–Carranglan watershed 250 Río de la Plata 363, 364 South America 297–298, 332, 335, 336, 347 southern Africa 29, 74 weather forecasts 173–174, 174–175 yield 165, 166 pre-economy 360 private land 43–44, 45 productivity 92, 101, 189, 199–200 Proteaceae 9, 29–30, 41 protected areas 30–32, 33, 257 see also reserves public information campaigns 124 public sector support 321–324 quint forecast categories 170–171 rainfall see precipitation range expanders 9, 29 rangelands 11, 12, 20, 196–210 rapid rural appraisal 149 reactive strategies 286, 293, 360 reconfiguration of reserves 34, 36, 39–40, 41 Red Sea State, Sudan 11, 96–100 reforestation 250, 260–261 regional level 173, 174, 366 regrets 55–56, 57, 65, 68 regulations 44, 60, 221, 269–270, 272 relief 77–79, 80–81, 96–97, 274 relocation 360 research and development 134, 160, 186, 209 reserves 34–40, 41, 45 see also protected areas reservoir storage capacity 19, 55, 56, 57, 60, 61, 62–63, 64, 65, 66 resettlement 250 residential development 10, 270, 311 rice production 6, 17, 229–230, 231–245, 266, 270 Río de la Plata 11, 297, 302–304, 353–370
3232 J&J Climate Change
1/11/07
9:57 am
Page 379
Index 379
risk adaptation and development 16 awareness of 25 capacity to cope 7 development 13 disaster 14 drought 101–105 estuarine fishing 361, 362 floods 6 involving those at 5 malaria and cholera 109–130 management 12 place-based approaches 228–246 rivers basins 6, 10, 11, 12, 19, 53–70, 71–89, 228–246, 297, 302–304, 353–370 dredging 270, 272 flow 305–306, 356–357, 362, 363–364, 367 River Uruguay 8, 302, 303, 356–357, 359 road density 301 runoff 59, 60–61, 62, 63, 68 Sahel 147 St James, Western Jamaica 283–284 sanitary systems 120, 121, 124 Savannakhet, PDR Lao 231–236 scenarios cereal production 133–135 driven approaches 213 estuarine fishing 363–365 floods 244 maize and soybean cultivation 333–334, 335–336 water management 19, 58–67 scientific knowledge 204, 297, 367, 368 screening options 201 sea level rise 264, 310 seasonal aspects disease 281–282, 293 migration 202, 236, 360, 367 weather forecasts 12, 163–180, 234, 324 sea surface temperature anomalies
(SSTA) 168, 172 sedimentation 270, 277 selective breeding 133, 138, 143 sensitivity 148 siltation 269 simulations 54–68, 135–139, 188–192, 252–254, 334 skill assessments 170–173 slow trends 304–307, 312 social level acceptability 288, 289 capital 12, 93, 95, 97, 99, 101, 103, 104, 106, 157 change 197–198 networks 206, 365 protection 79 vulnerability 153 socioeconomic level cereal production 134–135 change 320–321 dengue fever 289 estuarine fishing 362 farmer characteristics 318 flooding 276 rice production 243 watersheds 250–251 Western Jamaica 284, 288 soil quality 345–346, 347, 348 soil water atmosphere plant model (SWAP) 131, 135, 136 South Africa 10, 19, 30, 53–70 South America 296–297, 297–298, 332–352 southern Africa 9, 28–52, 74 southwesterly winds 173 sowing dates 164, 165, 190, 191, 193 soybean production 8, 299–300, 332–352 Special Report on Emissions Scenarios (SRES), IPCC 59, 133, 188, 334, 335–336, 337–338, 340, 341, 342, 347 species level 9, 28–29, 29–30, 34–39 spillover effects 20, 245, 247–263 SRES see Special Report on Emissions Scenarios, IPCC
3232 J&J Climate Change
1/11/07
9:57 am
Page 380
380 Climate Change and Adaptation
SSTA see sea surface temperature anomalies stock utilization ratios 139, 140–141, 141 streamflow 302, 303, 305–307, 363 subsidies 44 substitution of cultivars 138, 143 Sudan 11, 17, 90–108 supplementary irrigation 184, 340, 344 surveillance 286, 288, 292 sustainability 92, 216, 217, 364–365 sustainable livelihood approaches 17, 91–92, 101–102, 155, 354–355 SWAP see soil water atmosphere plant model Tamaulipas, Mexico 11, 315, 316, 319, 321–323, 326–327, 329 Tanzania 114, 115, 116, 117, 120 temperature cereal production 134, 188, 189, 190 China 219 disease 113–114, 119, 280, 281–282, 283 global change 7 maize and soybean production 337, 339, 340 Mexico 319 Mongolia 198, 199, 200 Nigeria 148 Pacific Islands 264 Pantabangan–Carranglan watershed 250 Río de la Plata 363, 364 South America 332, 335, 347 southern Africa 29 tercile forecast categories 171 Thailand 17, 232, 233, 236–240, 241–245 top–down approaches 17, 151, 272, 275–276, 277 tourism 82–83 tradable development rights 44 trade 15, 76–77, 320 trade-offs 20, 216, 247–263, 252–261
traditional level disease treatment 117, 118, 120 farming practices 166–167 knowledge 12, 71–89, 354–355, 359 pastoral systems 197, 205, 206, 209 trans-boundary effects 245 transformation in origin approaches 346–347, 348 translocation 34, 38, 40 transnational megaparks 33–34 treatment of disease 117–118, 121 tree food crops 167 Trinidad and Tobago 282, 285–287 tropical rainstorms 165–166 Tunisia 11, 12, 181–195 two-tiered screening processes 20 Ubonratchathani Province, Thailand 236–240 Uganda 114 uncertainty 8–9, 18, 107 UNEP see United Nations Environment Programme UNFCCC see United Nations Framework Convention on Climate Change United Nations Environment Programme (UNEP) 213 United Nations Framework Convention on Climate Change (UNFCCC) 212, 213–214 upland cereal production 131–146 urban water demand 53, 60–63, 64, 65, 66, 67 Uruguay 332, 333 variability Argentina 319, 320 cereal production 137, 138, 139, 183 disease 113–114, 280, 282 estuarine fisheries 356–363, 365 food production 163–180 long-term trends 305–307, 312–313
3232 J&J Climate Change
1/11/07
9:57 am
Page 381
Index 381
Lower Mekong 243–244 Mexico 319, 320 Mongolia 200 North Africa 182 Pacific Islands 274 traditional strategies 74–76 vulnerability 300 variety changes 232, 234, 240 veld products 12, 76–77, 81 Vientiane Plain, Lao PDR 231–236 Vietnam 17, 232, 233, 240–241, 241–245 visual presentations 262 vulnerability agricultural expansion 300 cereal production 132 changing practices 159–160 definitions 148 development 14–15 disease 110, 114–115, 122, 123–124, 277, 281, 284–285 drought 90, 91 estuarine fishing 355–356, 359, 368 farm level 318–321 floods 270, 271, 327–328 livelihood systems approaches 151 livestock herding 198–200, 207 Lower Mekong 244–245 natural resources 12, 71–72, 85–86 poverty 5, 84 rural households 151–153, 244–245 small islands 264, 265 water resources 219, 225, 248–251 water conservation 189–190, 191 cross-sectoral impacts 258, 259, 260, 262 demand 15 management 19–20, 53–70, 97 scarcity 152–153
spillovers and trade-offs 254, 257 storage 166, 285, 291, 293 transport 135 use practices 159, 211, 218, 219–225 see also irrigation watersheds 10, 20, 247–263, 269, 316 weather cautious behaviour 361, 367 extreme 8, 10, 264, 274, 308–310 forecasting 12, 158, 163–180, 234 predictions 7–8 stations 158 see also droughts; El Niño Southern Oscillation; floods; La Niña welfare 63, 64, 65, 67 West Africa 12, 168–170 wheat 184, 187–192, 193 will to adopt 21–23, 159–160 winds 173, 361, 362, 367 winter severity 199, 200 withdrawal ratios 219 WOFOST see world food studies model world food studies model (WOFOST) 131, 135, 136 yam production 165 yield cereal production 137–139, 183, 184 high yield varieties 155, 156, 159 fertilization 139–140 irrigation 141 maize and soybean 336–345, 347 weather forecasts 165, 166, 176 wheat cultivation 190, 192, 193 zud 199, 200
