Developments in Soil Science 20
SOILS ON A WARMER EARTH
Further Titles in this Series 1. I. VALETON BAUXITES 2. IAHR...
142 downloads
1290 Views
14MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Developments in Soil Science 20
SOILS ON A WARMER EARTH
Further Titles in this Series 1. I. VALETON BAUXITES 2. IAHR FUNDAMENTALS OF TRANSPORT PHENOMENA IN POROUS MEDIA 3. F.E. ALLISON SOIL ORGANIC MATTER AND ITS ROLE IN CROP PRODUCTION 4. R. W. SIMONSON (Editor) NON-AGRICULTURAL APPLICATIONS OF SOIL SURVEYS
5A. G.H. BOLT and M.G.M. BRUGGENWERT (Editors) SOIL CHEMISTRY. A. BASIC ELEMENTS 5B. G.H. BOLT (Editor) SOIL CHEMISTRY. B. PHYSICO-CHEMICAL MODELS 6. H.E. DREGNE SOILS OF ARID REGIONS
7. H. AUBERT and M. PINTA TRACE ELEMENTS IN SOILS
8. M. SCHNITZER and S. U. KHAN (Editors) SOIL ORGANIC MATTER
9. B.K.G. THENG FORMATION AND PROPERTIES OF CLAY-POLYMER COMPLEXES 10. D. ZACHAR SOIL EROSION 11A. L.P. WILDING, N.E. SMECK and G.F. HALL (Editors) PEDOGENESIS AND SOIL TAXONOMY. I. CONCEPTS AND INTERACTIONS 11B. L.P. WILDING, N.E. SMECKand G.F. HALL (Editors) PEDOGENESIS AND SOIL TAXONOMY. 11. THE SOIL ORDERS 12. E.B.A. BISDOM and J. DUCLOUX (Editors) SUBMICROSCOPIC STUDIES OF SOILS 13. P. KOOREVAAR, G. MENELIK and C. DIRKSEN ELEMENTS OF SOIL PHYSICS
14. G.S. CAMPBELL SOIL PHYSICS WITH BASIC-TRANSPORT MODELS FOR SOIL-PLANT SYSTEMS 15. M A . MULDERS REMOTE SENSING IN SOIL SCIENCE 16. I.B. CAMPBELL and G.G.C. CLARIDGE ANTARCTICA: SOILS, WEATHERING PROCESSES AND ENVIRONMENT 17. K. KUMADA CHEMISTRY OF SOIL ORGANIC MATTER 18. V. VANCURA and F. KUNC (Editors) INTERRELATIONSHIPS BETWEEN MICROORGANISMS AND PLANTS IN SOIL 19. L.A. DOUGLAS (Editor) SOIL MICROMORPHOLOGY: A BASIC AND APPLIED SCIENCE
Developments in Soil Science 20
SOILS ON A WARMER EARTH effects of expected climate change on soil processes, with emphasis on the tropics and sub-tropics Edited by
H.W. SCHARPENSEEL Institute of Soil Science, University of Hamburg, Allende-Platz, 0-2000 Hamburg 13, F.R. Germany and
M. SCHOMAKER and A. AYOUB UNEP, P.O. Box 30552, Nairobi, Kenya Proceedings of a n International Workshop on Effects of Expected Climate Change on Soil Processes in the Tropics and Sub-tropics, 12-14 February 1990, Nairobi Organized by United Nations Environment Programme (UNEP) International Society of Soil Science (ISSS) Sponsored by United Nations Environment Programme (UNEP ) UNEP International Society of Soil Science (ISSS) EC Technical Centre for Agriculture and Rural Development (CTA)
(ISSS)
ELSEVIER Amsterdam - Oxford -New York - Tokyo
1990
ELSEVIER SCIENCE PUBLISHERS B.V. Sara Burgerhartstraat 25 P.O. Box 211,1000 AE Amsterdam, The Netherlands Distributors for the United States and Canada: ELSEVIER SCIENCE PUBLISHING COMPANY INC. 655, Avenue of the Americas New York, NY 10010, U S A .
ISBN 0-444-88838-1
0Elsevier Science Publishers B.V.. 1990 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the publisher, Elsevier Science Publishers B.V./ Physical Sciences & Engineering Division, P.O. Box 330,1000 AH Amsterdam, The Netherlands. Special regulations for readers in the U S A . - This publication has been registered with the Copyright Clearance Center Inc. (CCC ), Salem, Massachusetts. Information can be obtained from the CCC about conditions under which photocopies of parts of this publication may be made in the USA. All other copyright questions, including photocopying outside of the USA, should be referred to the copyright owner, Elsevier Science Publishers B.V., unless otherwise specified. No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. This issue is printed on acid-free paper. Printed in The Netherlands
V
CONTENTS Forewords Pn=face Conclusions and recommendations Chapter 1
Chapter 2
Chapter 3 Chapter 4
Chapter 5
Chapter 6
Chapter 7
Overview of the greenhouse effect. Global change syndrome, general outlook H .W . Scharpenseel and P . Becker-Heineman
Chapter 9
1
Inputs to climatic change by soil and agriculture related activities. Present status and possible future trends A.F. Bouwman and W.G. Sombroek
15
Processes that affect soil morphology R .W . Arnold
31
Influence of climatic change on soil moisture regime, texture, structure and erosion G.Y. Varallyay
39
Resilience against climate change? Soil minerals, transformations and surface properties, Eh, Ph R. Brinkman
51
Impact of climatic change on soil attributes. Influence on salinization and alkalinization I . Szabolcs
61
Soil organic matter and biology in relation to climate change P.B. Tinker and P . Ineson
Chapter 8
ix xiii xv
71
Influence of climatic change on development of problem soils, especially in the alluvial domains W.R. Fisher Addendum: Methane formation in waterlogged paddy soils and its controlling factors Y.Takai and E . Wada
101
Potential influence of climate change on soil organic matter and tropical agroforestry E.H. Franz
109
89
vi
Contents
Chapter 10 The use of models of soil pedogenic processes in understanding changing land use and climatic conditions 121 J.W.B. Stewart, D.W. Anderson, E.T. Elliott and C.V. Cole Addendum: Modeling nitrous oxide production by denitrification 133 J.R.M. Arah and K.A. Smith Chapter 11 Impacts of anthropogenic atmospheric pollution on soils, with special relevance to tropical and subtropical soils, and possible consequences of the greenhouse effect N . Van Breemen
137
Chapter 12 Changes in soil resources in response to a gradually rising sea-level H . Brammer and R . Brinkman
145
Chapter 13 Soils on a warmer earth: tropical and subtropical regions W.G. Sombroek
157
Chapter 14 Soils on a warmer earth: subtropical and Mediterranean regions 175 D. Yaalon Chapter 15 Impact of climate warming on arid region soils D.H. Dregne
177
Chapter 16 Soils of the subboreal region on a warmer earth B.G. Rozanov and E.M. Samoilova
185
Chapter 17 Climate induced changes of the boreal and subpolar soils S.V. Goryachkin and V.O. Targulian
191
Chapter 18 Approaches to mitigate tropical deforestation by sustainable soil management practices P.A. Sanchez, C.A. Palm and T.J. Smyth
21 1
Chapter 19 Managing global change by curtailing emission sources and creating new sinks R. Grantham
22 1
Chapter 20 Implications of the greenhouse effect for African agriculture R.S. Odingo
23 1
Contents
vii
Chapter 21 The agricultural environment of Latin America and the Caribbean and the greenhouse effect N . Ahmad
249
Chapter 22 The Asian agricultural environment and the greenhouse effect L. Venkutaratnam
267
This Page Intentionally Left Blank
ix
FOREWORDS The motivation for organizing the International Workshop ‘Effects of Expected Climate Change on Soil Processes in the Tropics and Subtropics” was threefold. Soil Science is heavily involved in the problems of trapped radiation in the atmosphere by greenhouse effect (GHE) promoting gases. The e€fect of soils on the GHE syndrome was analyzed in a preceding conference: the International Conference on Soils and the Greenhouse Effect, Wageningen, the Netherlands, August 1989, organized by the International Soil Reference and Information Centre (ERIC). An answer was needed regarding the effect of climate change on sustainable productivity of soils, with emphasis on tropical and subtropical regions. The confusing dispute in scientific and popular science journals regarding the predictions of effects of temperature and eustatic sea-level rise, as well as a wide array of possible advantages and disadvantages due to especially C02-rise, had to be thoroughly evaluated against the background of impacts on soil changes. A state of the art of predictive approaches, recognized by the majority of scientists, had to be elaborated. In this volume some 24 scientists contributed to such an evaluation and inventory, each in their own specific field. Though differences in opinion on assumptions, approaches and predictions still exist, these proceedings clearly bring the discussion a great step forward, and it is hoped that they will be of assistance to the soil science community in deciding on concepts for soil related core programmes of the forthcoming International Geosphere Biosphere Programme. The editors
x
Forewords
What once were local incidents of environmental damage, confined to one area and region now involve the whole world. Concern on the global environment has indeed become more and more pressing. All people concentrate their attention to this matter, regardless of their wealth. The United Nations Congress, the Alsh Summit and the Tokyo Conference on Global Environment and Human Response Towards Sustainable Development in 1989, issued the statements that the human race may be making our planet uninhabitable. They are concerned about the rapid decrease in land which should be set aside for forests and other vegetation, essential for maintaining the balance of atmospheric gases, and about the increasing deterioration of soil, water and air, that support the continued existence of all life on the earth. These threatening developments call for an increased emphasis on studies specifically aimed at problems of global change. In a broader perspective the need to expand the role of soil scientist's approaches to global environmental and resources utilization is evident. May these proceedings become an important milestone, which will amongst others contribute to a fruitful discussion during the 14th International Society of Soil Science Congress and its symposium "Global Soil Changes Under Influence of a Changing Environment" (Kyoto, Japan, August 1990). Professor Y. Takai Vice-president, ISSS
Fore words
xi
The theme of this book is a far reaching one and should attract great interest, both in industrialized as well as in less developed countries. Although the question of climate change and its effects on soil productivity is overshadowed by so many current events, it is almost certain that this subject will remain on the international agenda for the years to come; not only as science and research are concerned but also in relation to development policies, agricultural policies and others, especially since poor, rich, developed and developing countries are equally threatened by these problems. The Technical Centre for Agricultural and Rural Co-operation (CTA), provides ACP states with better access to information, research and innovations in the spheres of agricultural and rural development. CTA gives great attention to increasing agricultural production on a sustainable basis and CTA is prepared to strengthen and expand these activities in the years to come. Within this context the recommendations in these proceedings and the subsequent follow-up will be of great interest to us and we are certainly prepared to continue our collaboration in this respect with the United Nations Environment Programme (UNEP) and the International Society of Soil Science (ISS S ) . Dr. W. Treitz Deputy Director, CTA/ACP-EEC Lorn6 Convention
xii
Forewordr
Each day, we are getting a clearer understanding of the growing pressures that threaten our planetary biosphere. Each day, pressures are mounting, due to demographic momentum and rising industrialization. Though uncertainties remain about the magnitude of climate change, scientific evidence confirms human activity has undoubtedly altered, and continues to alter, the atmosphere. While the principle greenhouse gas is carbon dioxide from fossil fuels, agricultural practices are having important impacts on possible changes in global climatic regimes. Recent studies suggest burning of savannahs for agriculture contributes three times more carbon dioxide per hectare to the atmosphere as does burning of tropical forests. Savannah clearing and increase in cattle and rice paddies are just three greenhouse gas sources from agriculture. Studies suggest soil processes involving inputs and outputs of water, gases, soluble salt and organic matter are likely to be influenced by global mean temperature increases. Such changes could in turn reduce land productivity, further threaten biodiversity, exacerbate soil loss and disrupt sediment budgets. Productive cropping areas could turn into deserts, while coastal storms could inundate or degrade fertile lands. Restoring the balance between the sources and the sinks of greenhouse gases will likely not be achieved solely through carbon emission reductions. Conserving existing carbon sinks (oceans, rainforests, grasslands, mangroves, coral reefs, etc.) and significantly increasing the productivity on degraded soils would be cost effective measures to achieve both environmental and economical sustainability. Massive reforestation, regrassing and amelioration of degraded lands are needed. A growing challenge to soil experts and agriculturalists is to produce more food for our growing world-wide population (currently already at 5.2 billion) from finite and shrinking productive lands. Of the total world-wide area of potential arable land, nearly half (about 1.5 billion ha) is already cultivated. There are very few virgin lands easy to exploit for additional food production. With these proceedings, a serious effort has been made to: 1) discuss latest knowledge on the expected impact of climate change on soil processes, with emphasis on the tropical and sub-tropical regions; 2) agree on recommendations for meaningful and useful future research and monitoring programmes; and 3) formulate policy responses to the expected impact in the form of preventive and adaptive measures. I hope that with these proceedings, we have moved forward to a better understanding of soil processes and their relationship to climate change. Mostafa K. Tolba Executive Director, UNEP
...
Xlll
PREFACE The Plan of Action for the implementation of the World Soils Policy was endorsed by UNEP’s Governing Council in 1984. The document identified the international action required to promote the sound use of land and soil resources. It also highlighted the need for internationally acceptable methods for assessing and monitoring the existing status and risk of soil degradation. The importance of ensuring that practices to prevent soil degradation, improve land productivity and reclaim degraded areas are appropriate to Iocal physical, economic, social and cultural conditions had been recognized in the document. At the same time, it is also recognized that serious gaps exist in knowledge of the extent, mechanisms and economic consequences of soil degradation as well as in knowledge of costeffective means for controlling erosion and other forms of land degradation. Consequently, there is an equally urgent need for a long term commitment by international organizations and Governments to support research on mechanisms and effects of soil degradation on the development for combatting soil erosion, as well as for the identification of critical areas and for a standardized assessment of soil resources and of the seriousness of soil degradation, through surveys and inventories. Methods are required which can reliably detect significant changes in those soil and land characteristics which directly or indirectly affect the quality and quantity of the land and its liability to produce food, fibre and timber. The manipulation of the pedosphcre by man is substantial and is likely to increase strongly until at least the year 2050, whcn the world population is expected to peak. This manipulation certainly has an impact on global climate and should be taken into account in connection with the increasing greenhouse effect on climate. The present status and future trends concerning the effects of soils and land cover on the fluxes of greenhouse gases, the surface energy balance and the water balance are discusscd in Bouwman (1990). The current publication explores and assesses the concept, the trends and impacts of the possible global warming on soil processes in the tropics and sub-tropics. Hierarchical research requirements and response stratcgies nccessary to meet the ncgative impacts of climate change on soils are also outlined. UNEP sees that the great task for the world community is therefore three fold: 1) to reduce greenhouse gascs; 2) to prepare for the impacts of global warming; and 3) to provide technical and financial assistance to developing countries so that their strive towards legitimate economic expansion is not hampered by measures addressing global warming. UNEP welcomes thc concern expressed on global warming at the highest levels. Thcse and other priorities, such as protection of land resources by combatting dcsertification and
x iv
Preface
deforestation, have bcen constant pre-occupations of UNEP for well over a decade. The preparatory work for global conventions on climate change and biological diversity are but a few of the many actions of UNEP to safeguard our planet. A.T. Ayoub, Senior Programme Officer, UNEP
xv
CONCLUSIONS AND RECOMMENDATIONS OF THE INTERNATIONAL WORKSHOP ON THE EFFECTS OF EXPECTED CLIMATE CHANGE ON SOIL PROCESSES The workshop discussed how soils in specific ecosystems respond to changes of climate and the needfor soil scientists to work with other disciplines.It also listed research priorities and discussed how the use and management of soil resources can mitigate climate change and at the same time adapt to it.
ECOSYSTEM DYNAMICS, INTERACTIONS WITH THE ATMOSPHERE AND THEIR EFFECT ON SOIL AND LAND USE A scenario and some predictions
We assumed: over a time span of about 50 years a gradual warming of the atmosphere of about 3°C with 2°C in tropics and 5°C in sub-polar areas; 10% increase of precipitation, a sea level rise of 50 cm, and an equilibrium world population of 10 billion; by about the year 2100 a doubling of the atmospheric C 0 2 content. The results of geophysical, geochemical, and biogeochemical processes vary widely in current soil ecosystems. Within a decade changes of temperature and precipitation will influence soil temperature and moisture regimes, pH, base saturation, fertility status, surface litter and biological activity, and the presence of salic and fluvic soil properties where appropriate. Over a period of several decades changes of climate are manifested in soils through the depth and kind of humus in the topsoil, relative fertility, erosion, and in peatiness, swell-shrink features, degree of bleaching and calcareoumess. We believe that soil ecosystem responses to changes in land cover and climate differ mainly due to different degrees in sensitivity of soils to disturbance and modification. It is recognized that there are likely to be very large regional and local variations in the manifestation of warmer earth conditions. Nevertheless the following general statements can be made for some major ecosystems. In a Tropical Rain Forest ecosystem (hot, humid) one might expect increases of forest growth, nutrient cycling, and mineral weathering. Ferralsols are common in such biomes and with more percolation through the soil we would expect a decline of the fertility status, slight decreases of pH and base saturation, more litter production but also more biological activity with little change in humus content in the topsoil. Additional Weathering of minerals would occur, however, due to the great thickness of many saprolites there would be little change in relative fertility over time. Although soil processes would respond to the changes of climate and vegetation the overall impact would be minimal compared to the present ecosystcm. Thcse kinds of soils have low sensitivity to the assumed
xvi
Conclusions and recommendarions
climatic changes. In a Steppe or Grassland ecosystem (warm, semiarid) the increase of precipitation would increase biomass production and movement of water through and on the soil. In such ecosystems, Chernozemic soils are common; they are moisture sensitive and the humus content in the topsoil would increase, carbonates and other soluble salts would leach deeper and salinization might occur in adjacent lower areas. Soil pH and base saturation would decrease slightly but the nutrient supply would remain high. Increased erosion would occur on slopes where vegetation would become stressed. In a subpolar Boreal Forest ecosystem (cold, subhumid) we would expect an increase of biopmductivity and biogeochemical cycling. For the dominant soils in this ecosystem, Podzols, the humus content, base saturation, relative fertility status and pH would become higher. The litter and liistic horizon, if any, would become thinner, leaching would tend to intensify, and the spodic horizon would likely become thicker. The vegetative cover would minimize erosion of the soil. The soil processes in this ecosystem are fairly sensitive to the assumed climate changes, especially temperature. We conclude that the changes in soil processes will be substantially more pronounced in cold and temperate regions than in the tropics because the soils outside the tropics are more temperature sensitive and the temperature changes are expected to be greater there. As one approaches the transitional zones between ecosystems ("ecotones") it is more difficult to make general predictions about soil changes because of the strong interactions of climate and biota that often occur in such zones. In coastal swampy areas and tropical mountains a global warming can have dramatic effects on the ecosystems. Rising sea levels would generate a host of events such as flooding, salinization, and new erosion and sedimentation patterns that would affect many soil ecosystems. In mountainous areas the ecological zones respond to elevation, thus a warming would push the climatic belts up the mountains, creating new conditions. The essential contribution of soil science to climate change studies The impact of climate change will be extremely diverse and difficult to predict. The consequence for soils and their associated vegetation will be even more complex, and will have the most wide ranging and diverse consequences for agriculture, forestry, ecology, conservation, and all forms of land use. Progress in understanding and pedicting these effects can only be made by multidisciplinary studies. Soil science forms an essential component of such work. Soil is the basis for plant growth, and the full effects of climate change will be mediated by complex interaction between the two. There will also be direct effects of increase in atmospheric C02.At the highest level, full scale ecosystem studies are essential to understand changes resulting from broad shifts of vegetation bands and biomes. Major modeling programmes will be needed at
Conclusions and recommendarions
xvii
several different scales of size and complexity. Apart from these vital studies on impact, it must not be forgolten that soils have major sources and sinks for important greenhouse gasses: carbon dioxide, water vapor, methane, nitrous oxide. These fluxes require much more accurate measurement, and the understanding of all underlying processes. We also need to know how the expectcd climate changes will interact with these greenhouse gas processes in positive or negative feedbacks. Major research programmes are bcing dcvcloped on these themes in a number of countries, many with a strong international component. It is critically important that the central position of soil science in such studies is recognized at an early stage, so that these programmes are fully balanced and complete. This also demands from the soil science community that it is prepared to collabotate freely and widely with other essential disciplines having a bearing in this issue. These studies require new techniques, which are only just being introduced into soil science, such as molecular biology, natural abundance stable isotope studies, and the newest methods of remote sensing. A crucial group of soil processes arc tcrmed "anthropogenic", referring to those brought about by the activitics of man. Thcsc processes likely override all others. The implications of man's direct influence on soil processes, on ecosystems, and on the earth systcm as a whole are such that they tend to overshadow our attempts to deal with natural ecosystems and climate-induced changes. Such anthropogenic processes have the potential to change the course of civilization especially if they are ignored.
RESEARCH PRIORITIES ON EFFECTS OF EXPECTED CLIMATE CHANGE ON SOIL PROCESSES, WITH EMPHASIS ON THE TROPICS AND SUBTROPICS The soil research priorities identified during the workshop can be categorized in 6 main groups: Collection of baseline data; Study of soil processes; Study of land use dynamics; Modelling of soil processes; Long-term monitoring of global soil change; Management and mitigation measures of global soil change. Furthermore, to carry on many of thc soil studies, data have to be collected and processed in a whole hcarted collaborative effort with other disciplines, such as those dealing with climatology, ecology, and crop production. No priorities were identified for research on emission of greenhouse gases, because these were considered during the recent conference on "Soils and the
xviii
Conclusions and recommendafions
Greenhouse Effect" in August 1989 in Wageningen (see Bouwman 1990). The following specific subjects in each of the six categories were identified as priority research subjects. Base line data A large number of data bases pertaining to the nature and state of the worlds soil cover exist. To prcdict effects of climatic change on soil properties and on quantitative and qualitative aspects of ground water recharge and on surface water, it is necessary to identify gaps in these data bases. A first brief survey indicated the following gaps (some of which may overlap with each other): Soil information on a scale of 1:1 million, or even smaller (such as the FAOAJnesco Soil Map of the World at a 1 : 5 million schale), of large parts of Africa, South America, and Asia is not accurate enough for predictive and interpretative purposes; larger scales (up to 1:250,000) are needed. The state of soil/land degradation is generally poorly understood and quantified, although an overview will be available soon through the UNEPDSRIC Global Assessment of Soil Degradation (GLASOD) Project. Special attention should bc paid to mapping of (potentially) saline, alkaline and acid sulfate soils. Global data bases on actual land use are either out-of-date or too general. These baseline data should be organizcd and made available through the development of Land Information Systems (LIS) such as the ISSSDSRIC Soil and Terrain Digital Database (SOTER). Paleopedological maps of the world (1:2.5 million) of 2 or 3 climatic periods in the Pleistocene and Holocene will be very hclpful in analogue studies to forecast the effect of climatic change on present-day soils and for estimating the carbon cycle changes. As a first priority it would be sufficient to prepare such maps for a numbcr of selected, promising areas, whcre more background data are available. The Unesco-initiated project of mapping continental soil salinization and potential soil salinization should be cornpletcd. Soil processes and their dynamics In studying soil processes, priorities should be given to processes that operate on a time scale of 1 to 100 years, including those that may change suddenly or catastrophically. We bclieve that these are (in order of decreasing
Conclusions and recommendations
xix
potential rate of change): Changes in soil salinity and alkalinity (one month to 10 years); Changes in soil meso- and macro fauna important for bioturbation and homogenization (1 to 10 years); Changes in structural stability and moisture characteristics (1-10 yrs); Changes in amount and quality of organic C and N levels (1-10 yrs); Changes in nutrient status, acidity, redox regime (10 to 100 years); Changes in susceptibility to erosion (10 to 100 years); Changes in iron and amorphous minerals (10 to 100 years). A number of methods are available for this purpose: Analogue studies, using soil chronosequences or using situations where the soil climate has actually changed substantially (due to deforestation, artificial drainage, ponding, etc.) Manipulation studies in the field (small chamber, small watershed), greenhouse or laboratory. Field studies on soil processes should preferably be camed out in areas involved in a network for long-term monitoring (see below). An important problem to be addressed is the methodology of scaling down experimental (as well as modeling) results to values that are relevant at less detailed scales.
Changes in land use For a number of other disciplines (agronomy, animal husbandry, economics, general politics, medicine) it will be important to predict the changes in land use resulting from climatic change and related changes in soil properties. Soil scientists will be asked to provide data needed to help forecast optimal land use on the short (0-10 y) medium (10-30 y), and long (40-60 y) run. GIs- or LIS-based expert systems should be very helpful in this regard.
Mo dei i n g Process oriented simulation models describing changes in soil properties, as a function of temperature and precipitation (time scale of days, weeks, months, or years) are urgently needed. Research on soil processes and collection of longterm monitoring that should provide parameters and opportunities for calibration and validation of these models.
Long term monitoring To detect ecologically relevant changes related to climate change, and to provide opportunities for research, a network of stations for long-term
xx
Conclusions and recommendations
monitoring of vegetation-soil properties - topography-hydrology-hydrochemistryclimate of a number of globally representative,natural, rangeland and agricultural ecosystems should be set up. Priority should be given to: Transitional regions where rapid climate change is expected; Coastal areas where sea level changes are expected; Irrigated areas in semiarid regions; Deforested areas; Natural wetlands. Besides detailed monitoring at field/small catchment scale on the ground, detailed remote sensing monitoring of the general region, representative for the ecosystem, should be carried out. Remote sensing should be directed to estimating vegetation (cover), net primary production, evapotranspiration, surface temperature, IR radiation and, where applicable, surface soil properties. The monitoring research sites should be run by local organisations and coordinated internationally to promote the use of common methodologies and exchange of data. Data from other disciplines Climatologists
In addition to forecasts about mean annual temperature and precipitation, it will be very important to obtain estimates of the ranges of year to year and month to month variability, especially in rainfall, and the expected frequency of high intensity, highly erosive rainfall. Ecologists and crop production physiologists Effects of increased CO;!on organic matter return to the soil, estimated on the basis of crop production models. In addition to quantity, qualitative aspects of the organic matter (possible changes in biodegradability) should be considered.
ADAPTATION TO CLIMATE CHANGE: SOIL RESOURCES USE AND MANAGEMENT The workshop participants discussed strategies to counteract increases in greenhouse gases and defend present soils and their uses against adverse effects of climate change. The resulting recommendations are directed at policy makers and at the users of the soils. Management strategies were identified that are desirable in their own right as well as in the context of the greenhouse effect. Four sets of concerns with accompanying recommended management strategies are reported:
Conclwionr and recommenda~iom
xxi
Increasing C 0 2 sinks The following measures are advocated, recognizing that soil- and land management-related methods of C02 sequestering on-their-own would only remove a fraction of the increased atmospheric carbon. Besides removing CO2, each of these measures would provide direct production benefits to the land user. Conditions that favour C retention in soil and biomass include wetness, low temperatures, increased mineral nutrition (soil fertility), deep root development as well as biological activity. Management strategies should aim to preserve and, where possible, increase standing biomass and residues in soils. On many soils this will involve appropriate use of mineral fertilizers. On acid soils, selection of aluminium-tolerant crops or cultivars and liming of the acid subsoil to encourage deep root development are recommended. Irrigation in arid areas, either to produce grassland or annual field crops or to establish perennial vegetation, will increase standing biomass and soil organic matter. In the case of hardpans or other root barriers, subsoiling would enable deeper root development.
Reducing N2O and CH4 emissions Conditions favouring N2O emission from soils include high concentrations of mineral N species and alternation of reduction and oxidation. Conditions favouring methane emission include deep reduction in the absence of sufficient active iron oxides as well as of sulfate. Management measures to avoid such conditions include the following: Proper dosage. timing and placement of nitrogen fertilizers in accordance with crop growth stages, to minimize periodic excess N. Drainage or cultivation or other measures to avoid topsoil compaction and periodic reduction in cropland or pastures. In wetland soils, efficient water management so as to avoid periodic oxidation. Under rainfed conditions, this may involve land shaping, bunding, puddling, supplementary irrigation where feasible. Methane emission from wetland soils can be rninimizcd by avoiding incorporation of readily decomposable organic matter, and by addition of material rich in ferric iron, where practical, on soils low in active iron oxides. Also the use of fertilizers with the sulfate anion suppresses methanogenesis.
xxii
Conclusions and recommendations
Adapting to sea-level rise Sea-level rise causes a destabilized coastline, inland extension of areas affected by salinity and increased flooding depth and wetland conditions in inland parts of coastal plains. These problems can be combatted by one or more of the following measures: Sea defense including structural works such as embankments, as well as planting or preservation of protective mangrove forest belts. Embankment and pump drainage, or tidal drainage where feasible, of the impoldered land. Changing land use to fish farming where protection of the land against fresh-water or saline flooding is not practical. Changing to more salt-tolerant crops or cultivars in salt-affected areas. Selection of cultivars for higher yield under deep flooding or brackish water conditions.
Adapting to climatic variability The main problems to be considered are high-rainfall events and drought. Soil management measures to mitigate their effects in dryland soils should aim to maintain a complete soil cover, minimizing rainfall impact and non-beneficial evaporation; assure good macroporosity, infiltration and aeration; promote deep root penetration. Such measures include mulching, use of cover crops, relay cropping to improve cover and bioporosity; and subsoil liming where needed and appropriate deep ferlilization, especially with P, to increase rooting depth. Additionally, water harvesting methods and terracing and other recognized water conservation methods would increase amounts of water available to crops. Response farming techniques including, for example, late topdressing in amounts will capture the benefits of years with good rainfall or minimize costs in poor rainfall years. Finally, it should be mentioned that a rise in atmospheric C02 can have a positive effect on plant growth through increased photosynthesis ("CO2 fertilization"), and on the water use efficiency of plants due to reduced stomata1 openings. Global temperature rise will moreover result in higher precipitation due to greatly increased evaporation over the oceans - which in places will significantly enlarge the amount of fresh water that can be used for irrigation purposes.
1
Chapter 1
OVERVIEW OF THE GREENHOUSE EFFECT Global change syndrome; general outlook H.W. Schapenseel and P . Becker-Heidmann Institute of Soil Science, University of Hamburg Allende-Platzz, D-2000 Hamburg 13. Federal Republic of Germany
ABSTRACT Accumulation of cosmic dust and planetesimals was most likely the mechanism that created our planet. Due to dominance of hydrogen, the extruded gases produced a primordial reducing atmosphere. enriched with methane and ammonia. Then, after a slow start, continued oxidation with oxygen, released from photolysis of water, and the later development of life from photosynthesis caused the atmosphere to become dominated by COz, water vapor and N. The two former components were able to trap IR radiation and to produce a warming greenhouse effect of 33"C, shifting the surface temperature to +15"C. Oxygen from photosynthesis (at present yearly c a 330 bil t from terrestrial photosynthesis) was used over at least 2 billion years, for sustaining respiration of the various facets of life and for iron oxidation in marine and terrestrial sediments. During the last billion years oxygen began to enrich in the atmosphere, parallel to reducing CO2 concentration, due to its consumption by photosynthesis, chemical weathering and the carbonate precipitating pumping effect of the oceans. C02 replenishment occurs via volcanism and release from subduction zones. The faster biochemical cycle of smaller pool size (organic matter production, respiration, humification, kerogene formation, and biotic-abiotic-photochemical organic matter turnover) and over longer geological periods especially the slow but very large geochemical cycle (exchange of carbon between atmosphere, ocean, biosphere, and sediments), are decisive for CO2 concentration and its contribution to temperature. Some features of the biochemical cycle against the background of climate changes, including those due to Pangaea/Gondwana shifting, are discussed. Life is on a carbon trip. Wasteful consumption of fossil C based fuel, due to rising living standard and population explosion in conjunction with increasing release of greenhouse active (radiatively active) gases - which are fingerprinted - threatens to exert climate changes detrimental to our life conditions and civilization. Arguments to characterize the situation are assessed, also those expressing potential advantages of increasing C02 concentration for crop yields and expansion of the farmland area, doomed to shrinking at the present level of population explosion. The need for a change from the carbon trip to a mixed carbon - hydrogen trip is evident.
INTRODUCTION Environmental consciousness, especially watchfulness with focus on all anthropogenic activities causing pollution, tend to deviate our attention from tf-e dominant natural processes, underlying the whole web of contributing factors,
2
H.W. Schorpenseeiand P.Beckr-Heidmann
actions and feed-back systems in our unique earthly environment. A predictive analysis of the possible effects of a global climate change on soil processes and land degradation should be preceded by a short review of the scene as it existed, before a steadily growing human population created the syndrome of changing climate, basic to our worries. Its background is population explosion in conjunction with carbon-based energy sources and technologies, accompanied by steadily increasing release of nitrogen oxides (N20,NO), that absorb IRradiation or consume ozone in the stratosphere and produce ozone in the troposphere, as well as by increased infrared radiation trapping and stratospheric ozone destroying CFCs (chloro-flouro carbon compounds). These compounds, C02, CH4, N 2 0 , 0 3 , CFCs are expected by the majority of atmospheric chemists to lead to a further indirect temperature increase at the earth surface and decrease in the stratosphere. The estimated temperature rise of 3 to 5°C in the next 50 to 100 years may cause an eustatic sea level rise of 0.7 to 3 m due to water expansion and melting of polar ice masses. How did it all dcvelop and finally become a problem ?
OVERVIEW Basic facts, relating to this question in a nutshell expose the following tableau: Within the sun 700 mil t of hydrogen are fused per second into helium, i.e., ca 4.3 mil t of solar mass are converted into radiation energy, equal to 1.2 x 1015t per year (Wunderlich 1968). From the total solar mass of 2.2 x t , ca 1/40,000 has so far been consumed. The share of the solar radiation hitting the earth, the solar constant, amounts to 2.0 k 0.04 cm-2min-1.This solar radiation with wave lengths of less than 3,000 is absorbed in the ozone layer. Besides, light waves of 3,000 - 20,000A, till near-IR and radiowaves of 1 - lo3 cm wave length enter the atmosphere. The energy invested in the sun radiation is the origin and source of all important features of climate and environment, such as temperature, wind, clouds, precipitation and autotrophic organic matter production. The fact that the earth possesses an atmosphere, is taken as indication, that the origin of the earth is unlikely the result of a cosmic catastrophe, e.g. a collision of the sun with another cosmic body. I may have been formed however, by contracting dust and planetesimals with gaseous inclusions, giving rise after its compaction to extrusion of gases. Provided the gravitational forces are strong enough, those gases will be retained by the planet to form its atmosphere. The very light elements, such as H, He and Ne dissipated into space. This is revealed by comparison of the remaining atmospheric concentration with the share of these elements of the matter in the universe. Estimates are, that about 1 of 50 bil original Ne atoms in the primordial gas cloud is still left; He of the atmosphere is held 10 be almosr entirely radiogenic. The very wasteful atmospheric H, the major
Ilistorical overview of the greenhouse effect
3
cosmic element, may be representing 1 out of 5 mil H atoms in the original dust cloud; the even more reactive 0 about 1 of former 6 atoms; the less reactivc N about 1 of 800,000 N atoms (Asimov 1981). The high cosmic excess of the element H, also early earth, led to an initially reducing atmosphere of chiefly methane (CH4, carbon plus hydrogen) and Ammonia (NH3, nitrogen plus hydrogen). Depending on the amount of oxygen available, water (H20, oxygen plus hydrogen) was formed, which howcvcr was progressively precipitated and collected in depressions and marine basins together with the water vapor emitted by volcanic exhalations, thus leavingCH4, N H 3 , and water vapor as dominating atmospheric gases (Urey's work). Photodissociation of water (H2O + h.v = H+ + OH- ) led to slow oxidation of methane and ammonia into C02 and N2, producing an N2 and C02 atmosphere. With progressing integration of N2 into nitrates, C02 gained dominance till its rising conccntration increasingly blocked the photodissociation of water. Furthcrmorc, ozone formation from free 0 2 in the higher atmosphere absorbcd the UV-radiation and prevcntcd its penetration into the lower atmosphere and action of photolysis. As a rcsult, a stable C02 dominated atmosphere came into existcncc (scc also Habcr 1965). The high C02 concentration could have strongly promoted the greenhouse effect. Due to a rising temperature, water evaporation would have bcen further enhanced, with its additional promotion of the GHE and atmosphcric tcmpcrature rise until a hot earth would have emerged, envclopcd by a water vapor cloud and C02 dominated atmosphere. (For comparison, planct Venus built up a hot and stable COz atmosphere of ca 450" C). But planct earth took a completely different turn in the development of life, probably already slowly bcginning under the reduced CH4/NH3 atmosphere, whcre NH3 was decomposed, releasing N2 into thc aunosphcre, whilc excessive C 0 2 precipitated with Ca, Mg or Fc, which were dissolved by weathering procedures (without oxygen participation) in thc marine basins (not the least enclosed in phytoplancton). Thus, only a moderate GHE occurred due to water vapor and C02 built up, increasing carth's mean temperature by 33" C from - 1 8" C to +15" C (Arrhcnius 1896). This is a temperature level suitable for the liquid state of water and thc colloidal state within living organisms. Other conditions supported the sustainability of life as well, such as: 1 ) the Van Allen belt (the magnetosphere); 2) the shield against cosmic radiation; 3) similarly, the ozone shield for absorption of UV light; 4) the earth magnctic field, although changing its polarity rcpcatcdly in the course of earth history, giving furlher radiation protcction and orientation; 5 ) the high altitude cirrus clouds, heating the atmosphere; 6) the lower altitudc vapor saturated clouds exerting a cooling effect; 7) the inclined earth axis (23.5"), producing annual seasons of climate; 8) the earth rotation, causative for day and night change for regeneration of the metabolisms; and 9) the atmospheric currents for transportation of moisture, heat and dissipation of products of pollution.
4
H.W. Scharpenseel and P . Becker-Heidmann
Soil and humus formation, CO, and 0, trends
&M 0 Quaternary Ter t iar y
.
70
Cretaceous plus humus of Angiosperms
. 140
Jurassic
.
180
-
225
Triassic Permian . 275
first complete soil cover Carbonian plus humus of Gymnosperms
345
'
Devonian . 400
.
Silurian 440 Ordovician
. 490
Cambrian '
580
C02 trend decreasing 0 trend increasing
Fig. 1.1
Soil and humusformation in earth is history
Life on earth was slowly turning the N/C02 atmosphere into a N/O atmosphere. The 0 2 concentration increased by almost one order of magnitude (Fig. 1.1) during the last ca 600 million years, that is since the beginning of terrestrial plant growth (the Phanerozoic; flowering plants, the angiosperms since just ca 150 million years) . This is mainly the effect of oxygen release during the photosynthesis process of organic matter production from CO2, and 0 and H from H20, previously dissociated by sunlight energy (light reaction of photosynthesis). The slow development of our oxygen rich atmosphere after exhaustion of the enormous demand for marine and terrestrial Fe-oxidation is shown in Fig. 1.2. Meanwhile the total free oxygen pool in the atmosphere and dissolved in the oceans is estimated to amount to 1.3 x 1021 g. Most of the biologically produced oxygen, at present ca 3.3 x 1017g of 0 2 per year (corresponding with ca 1.2 x lOI7 g of C per year by terrestrial photosynthesis) plus ca 1.3 x 1017 g of 0 2 per year (corresponding with ca 5 x 10l6 g of C consumed by marine photosynthesis), is bound in the earth crust as metal oxide, sulfate, silicate, and carbonate and represents about 6 x g of oxygen (Chem. Ind. 1987). Considering the high reactivity of oxygen, its existence as free 02-gas in the atmosphere is possibly only the result of constant new 0 2 production and
5
Historical overview of the greenhouse effect
addition. Without replenishment by photosynthesis, our atmospheric oxygen may
be consumed in about 3000 years due to oxidation processes in the earth crust (Haber 1965, 1971). But also atmospheric C02 needs replacement outside the biochemical cycle of photosynthesis and respiration due to consumption of C02 by silicate weathering, where from 2 molecules of C02 involved in the bicarbonate reaction always only one is returned to the atmosphere, whereas the second one is precipitated as carbonate, which would use up the present atmospheric CO2-pool in about 10.000 years (2 COz + H 2 0 + CaSi03 t Ca2' + 2 HCO3- + Si02) (Berner and Lasaga 1990). Similarly the gas exchange pump of the oceans induce C02 intake to replace C02 of precipitated carbonate. After longtime involvement of the geochemical cycle, these carbonates may under high pressure and temperature be subjected to metamorphic processes and eventually release the C02 through volcanism or expulsion by subduction zones.
4 B I L L I O N Y E A R S AGO
3 TO
Primordial m l m o ~ p h e r e
Formallon 01
co2
S o l u l l o n O d 0, I " w a t e r
S O l U l l O " I " "11*,
4
B I L L I O N Y E A R S AGO
0, I n
water
CA
2 B I L L I O N Y E A R S AGO
Formillon 01
0, I n w a t e r
T B ~ l d S l I 1 ~l 1l x a l l o n 01
0,
......~
CA
0 5 B I L L I O N Y E A R S AGO
T e r r e $ t r i a l o r m ~ t i a n0 1
0, R I S E
0,
I N ATMOSPHERE
IIIIII 0 2
0 A C T E A IA
W E AT H E R I N G
CaCO, A
I
Fig. I .2
Fe*O 3
-
MUD
,
History of oxygen formation and dynamics
This gas exchange process would exhaust atmospheric C02 in about 300,000 years (Bemer and Lasaga 1990). Planet earth, its atmosphere and biosphere become vitally predetermined by the consequences of the vast but slow geochemical and faster biochemical cycle, which however represents a much smaller carbon compartment.
6
H.W. Scharpenseel and P . Becker-Heidmann
FEATURES OF THE BIOCHEMICAL CYCLE Only carbonaceous materials, produced by abiotic processes till ca 3 billion years ago are exclusively geochemical (Rankama 1948). All others also have a biochemical component. The organic matter residues of living organisms are preserved almost exclusively in aquatic sediments as carbonates or in contact with shales and clay minerals. The latter as clay domains provide also the matrix for organo-mineralcomplexation of younger or even today's terrestrial organic matter (Aylmore and Quirk 1960; Theng and Scharpenseel 1975; Theng 1979) (Fig. 1.3). 70-
60-
50I
P
E" 40-
1
H m i c acid concentration (mg/rnl)
Fig. 1.3
Isotherms at 20°C f o r the adsorption of 14 C-humid acid by montmorillonite saturated with diflerent cations (Theng and Scharpenseel 1975)
Most of recent as well as ancient sediment's organic matter stcms from phytoplancton and bacteria (Bordowsky 1965; Murphy et al. 1966); this forms the major sink of organic C and of CO2. Sediments, produced by precipitation, such as evaporites and carbonates, rarely contain large amounts of allochthonous organic matter. Detrital rocks like sandstone or shales engulf, usually diagenetically formed, relatively stable, secondary polymerized compounds, such as humic acids as oxidative or kerogene as reductive products (Tappan and Loeblich, ref. Welte 1963). Both together represent the major organic carbon sink
Historical overview of the greenhouse effect
7
at a level of ca 3.6 x loi5t, compared with petrol or coal with stocks of the order of 10l2- l O I 3 t only (Degens 1967). Finally, climatic and tectonic events have a great influence on the organic compound production and preservation, e.g. bituminous sequences often seem to be related to orogenic phases or epimgenic oscillations with corresponding eustatic lifts, trans- and regressions (Bitterli 1963). The C02 and 0 2 balance in ocean water and in the atmosphere changed with the organic matter production in the course of earth history, with its carbonate precipitation as well as the emergence of higher plants and animals (Tappan and Loeblich, ref. Welte 1969). During sediment diagenesis organic matter supports the microbial metabolism and it exerts influence on chemical reactions through pH and Eh changes, especially those involving C02 - SO2 CH4. After microbial activity terminates, chemical interactions with the inorganic matrix occur, leading to complexation and chelation, and reactive chemical groups like carboxyl, hydroxyl, and amino groups are released. The origin of life is believed to have occurred in an aquatic milieu, which provides more continuity due to less zoographic isolation than terrestrial life which developed later. According to Schilder (1956), 63 among 68 animal classes live in a marine environment, but, due to geographic/ecologic isolation in terrestrial environments, the differentiation in species is more pronounced, comprising ca 83 % of all known animal species. The transition from marine to land based life, that contributed most to an oxygenation of the atmosphere and that became a major sink for C02 excess in the biochemical cycle, must have begun preferably in marginal shelf fringes of the epipelagial, the euphotic zone, mostly under tidal influence, such as in marshy or mangrove environments and in shallow littorals. Since organisms sustaining the biochemical cycle by photosynthesis (C02 consumption) and respiration (C02 release) survive in evolutionary processes, due to their capability to adopt flexibly to environmental changes and to find ecological niches or refuges, the eco/geographical boundaries governing the distribution of species are constantly shifting due to climate changes as well as to tectonic effects. This applies particularly to stenothermic animal species requiring a narrow temperature regime. In biomes, animal life usually is more flexible and stretches further into critical environments than growth of vegetation (Wurmbach 1971). The spread of terrestrial life has also been largely influenced by plate tectonics in conjunction with the dissolution of Pangaea, the Gondwana and Europe-Angara drifting with the corresponding climate changes. In this context, the most striking process of soil degradation, a land fossilization and lateritic cuirass formation, occurred due to changed erosion and drainage patterns. These cuirasscs represent an extreme form of humid tropical weathering under the changing climate of the floating Gondwana subcontinents (mainly Jurassic to Oligocene) (Valeton 1984). A sporadic soil blanket (pers. comm. Dudal 1990)
8
H . W . Scharpenseel and P . Becker-Heidmann
existed probably since the end of thc Silurian, a soil continuum came into being some time after the cold spell during the Permian.
IMPACTS OF GEOCHEMICAL AND BIOCHEMICAL CARBON CYCLE A comparison of the magnitudes of the compartments of carbon in the (bio)geochcmical (exchange of C between sediments, atmosphere, biosphere and ocean) and biochemical (organic matter production, respiration, turnover) cycle shows the dominance in pool size of the former, which over geological time pcriods is all decisive with its enormous buffer capacity (Table 1.1). Shortcuts in the biochemical cycle and short-term excessive inputs may produce a flicker, strong enough though for consequences on the GHE, the earth temperature, the prccipitation and the circulation, which may damage or even exterminate species causativc of the disturbance. Thus, in our short historical span we can not rely on the buffcr capacity of the large but slow gcochcmical cycle to neutralize the consequcnces of our mistakes. Table 1.1
Comparison of carbon pool sizes in biochemical and geochemical cycle*
Componcnts CaC03 in sediments CaMg (CO& in sediments Organic sediments (kerogcne) HCOf and C032- dissolved in sea Fossil fuel (coal, gas, oil) Dead soil biomass (humus) C02 in atmosphere Living biomass (plant, animals) * Data slightly modified from Berner and
C-amount in 10'2 t of c 35,000 25,000 15,000
42 5
2 0.72 0.56 Lasaga 1990.
LIFE ON EARTH, A CARBON TRIP Lifc on earth is on a carbon trip. We must get aware of the need for changing course in time to avoid an erratic trip back into chaos, from which evolution made us ascend. Looking at the biochemical carbon cycle (Fig.1.4). about 115 to 120 x l O I 5 g of C are turned over annually in the terrestrial ecosystems by photosynthesis and inversely by respiration through the bio- and pedosphere. Worrysome is the surplus due to respiration of ca 1.5 x lOI5 g of C from annual land clearing (slash and bum) as well as the 5.5 x l O I 5 g of C from
9
Historical overview of the greenhouse effecr
combustion of fossil fuel, which are adding up at a ca 50% rate to the 720 x 10'5 g of C in the atmospheric carbon pool. However, recent results of Esser (1990) suggest that the C-sink due to C02 fertilization is already overcompenzating the C02 source from forest clearing (ca. 1.5 Gt C y - l from 10 to 15 mil. ha of clearing a year). Principally, carbon oxidation products, C02 and CH4, although being greenhouse active trace gases, are minor in importance compared with water vapor (1:5) in generating the temperature rising greenhouse effect of 33"C, lifting the surface temperature from -18°C to +15"C. According to Ramanathan (1989) as well as Raval and Ramanathan (1989), the total natural 33°C greenhouse effect generated by water vapor plus C02 equals 155 Wm-2 (ca 145 W at clear sky, 180 W at cloudy sky). Doubling of the C 0 2 concentration would add ca 4 Wm-2 only; human activities so far have enhanced the GHE of the atmosphere by ca 1.5 % only.
CARBON CYCLE
l
120
I
n
A
2'
B'OSPHERE + PEDOSPHERE
i...ca.?3P..!!.!v!na.! .......
$.
C6 2000 (dead)
ca 100
ca 100
ca 60
01
5"
,
DEPOSITS 5000 i ......COAL, . ...........OIL, .........GAS ............................................ SEDIMENTS/LITHOSPHERE 66 000 000 o,2 from l o r e s t a n d sol1 d e g r a d a t i o n
OCEAN SURFACE WATER c a 700 ................................................................... DEEPER OCEAN WATER c a 37000
l r o m c o a l , 011.
gas combusllon
a m o u n t s ~n G I ( I O " ~ ) , f l u x e s i n G I I ~
Fig. 1.4
Carbon compartments in biochemical and geochemical cycle
But the increments of C02, CH4, as well as the other radiatively active trace gases, such as the CFCs, N20 and tropospheric ozone, are wholly or to greatest extent the product of anthropogenic activities. Human civilization was in its earlier phases sustained by the direct products of photosynthetic C reduction and
10
H.W. Schorpenseel and P.Becker-Heidmann
photolysis of water (Calvin 1962). Since beginning craftsmanship and industrialization it progressively built up a dependency on minable and consumable forms of organic matter, produced mainly throughout the last 350 million years, such as coal, kerogene/petrol and methane. While the first scientists, expressing the vision of an expanding greenhouse effect with temperature and in consequence an eustatic sea level rise (Arrhenius 1896; Callendar 1938; Flohn 1941, cited in Lausch 1989) found little attention or enthusiastic response, at present an almost overheated scientific climate and preferably apocalyptic predictions thtcaten to misinterpret the real facts, needs and situative, curative potentials. We also should not ignore the main trend of decreasing C02 concentration over most of the geochemical and biochemical evolutionary phases of the past, with at least five times the present level still during Tertiary times and about 200 ppm after fluctuations during glacial and interglacial periods at the end of last Wisconsin-, Wurmian-, Weichselian glacial. Since even the present 350 ppm are at the lower end of the plant physiologically suitable C02 range, could there have been an acute C02 deficiency developing, and the C02 rise since begin of industrialization in reality being beneficial in our endeavour to increase food production for the growing world population ? The temperature rise of 0.7"C since 1860 and a sea level rise of about 17 cm are acknowledged by most climate research units, but a highly rated group of scientists from the G.C. Marshall Institute in Washington, D.C. draws attention to a better correlation between the observed trend in temperature and the solar activity/sun spot activity than the more commonly considered increasing C02 concentration (ref. Economist Vol. 313, No. 7633, 1989). R.S. Lindzen from M.I.T. Boston a.0. reported to be expecting nominal temperature increases only. Other critical observations focus in many different ways on the overestimation of the C02 concentration change and neglect of the importance of changes in atmospheric moisture level, which, as mentioned before, is undisputed the major promoter of the basic greenhouse effect. While the determination to reduce the release of the other greenhouse active trace gases, such as CH4, N 2 0 , CFCs, tropospheric ozone (see overview in Table 1.2) is apparently worldwide accepted, a production oriented pedologist/plant nutritionist can argue, whether the rise of C02, which is expected to enhance especially C3 photosynthetic efficiency and to improve the water economy of plants due to closing of stornatal aperture with increasing C02 concentration, which even under rising temperature and water evaporation may also conserve precious water (Schleser and Kirstein 1990), would be really all that bad. Fig. 1.5 shows the wide range of C02 concentration versus light intensity capable of increasing photosynthetic efficiency. The relative effect of increasing C02 on the major cultivated plants as well as the most obnoxious weeds requires thorough attention.
Historical overview of the greenhouse effect
Table 1.2
Greenhouse active and other (polluting)trace gases, basic pool sizes
Trace gases
% share of antrogenic
GHE co2 CH4 CFCs 0 3 (tropospheric)
co2
50 19 17 8 6 Residence time Concentration 100 y 350 ppm
co
1 - 6 months
CH4 CFCs (ClOx -radicals)
10 Y 50 - 150 y
100-150ppb N 40-80 ppb S (N, S. hemisphere) 1.7 ppm 0.2 - 0.3 ppb
N20
170 y few days only
0.31 ppm 0 - 100 ppb
N20
Components
NOx (NO, NO21 0 3 (stratospheric) 0 3 (tropospheric)
OH (atmospheric)
so2 Atmospheric C-pool Photosynthesis (terrestr.) Soil organic matter pool Ocean C-cycle flux
Fossil fuel C
11
few seconds only
10 ppm (35 km) 0.02 ppm (0.1 ppm max.) < O.oooO1 ppb
GHE-rising potential of trace gases rel. to c02 1 32 14 - 17,000 2000 150 Increase per year (1800->280, 1950>310); ca. 0.5% variable
1.1 % (18 ppb) 5% 4% 0.3 % (ca 1 ppb) 0.2 - 0.3 % (stratospheric)
0.5 %
50 ppb (max.) 740 bil t 115 bil t (ca 1/2 respired, ca. 1/2 Wmpo@ 1.8 x 10l2 t
38.5 x 10l2 t (3x 109tpery precipitated as carbonate) 5 - 10 x 1012 t
The vegetation belt, which may be particularly affected by increasing heat and drought, the inigation-dependent and responsive sub-tropics, may hopehlly
12
H.W. Scharpemeel and P . Becker-Heidmann
replenish water deficits by stepped up sea water distillation, which may become technically feasible in nearer future. Besides, the leading Russian climatologist from Leningrad, M.I. Budyko (ref. Spiegel 1990) expects even for the (semi)arid tropics more rainfall and a vegetation cover similar to that of the Pliocene. Increase of temperature and precipitation in the boreal belt could bear grave consequences due to enhancement of organic matter decomposition and consequent COs release, but may also hold unforeseeable opportunities to increase agricultural production in vast areas of so far low productive boreal lands, which may even become important carbon sinks (see for more details Chapters 16 and 17). The catalogue of pros and contras of effects of climate change is certain to expand. I
Net photosynthetic rate for different light intensities and C 0 2 concentrations Rp (mg C O P / q m
I Fig. 1.5
s)
I light Intens.. Rp net photosynth. r a t e
Influence of light intensity and CO;, on photosynthesis. (modified from Mengel and Kirkby 1979,acc. to Warren-Wilson 1969)
We should be critically alert towards the climate change syndrome, but as critically open to the more constructive and may be even optimistic arguments before destroying productive structures without having replacements in panic, through poorly conceived legal actions as well as preventive or curative measures, . The most important reaction, though, should be to prepare t k follow-up phase of a mixed hydrogen plus carbon trip. In geological times the geochemical cycle is absolutely dominating and decisive for CO2 concentration and its influence on temperature. However, the anthropogenic shortcut of the
H k l o r i c a l overview of the greenhouse effect
13
biochemical cycle that we practice with wasteful consumption of fossil fuel reserves, with rising population and living standards, may suffer under t k unsentimental regime of natural processes, such as temperature and sea-level rise, to destroy our species and civilization (only a fast forgotten flicker in earth history), We must therefore concentrate all our efforts on the development of the photovoltaic hydrogen technology for replacement of C-based fossil fuels, as well as on the methodology for use of sun energy in desalination of sea and brackish ground water to stabilize life and productivity in the (semi)arid lands.
REFERENCES Asimov, 1. (1984). Asimov’s new guide to science. Basic Books Publ., New York. Aylmore, L.A.G. and J.P. Quirk (1960). Domain or turbostratic structure of clays. Nature 187. 1046- 1048. Berner, R.A. and A.J. Lasaga (1990). Simulation des geochemischen Kreislaufs. Spektrum der Wissenschaft 5, 56. Bitterli, P. (1963). Aspects of the genesis of bituminous rock sequences. Geol. Mijnbouw. 42, 183-201. Bordovskij, O.K. (1965). Accumulation and transformation of organic substances in marine sediments. Marine Geology 3, 3-1 14. Calvin, M. and J.A. Bassham (1962). The photosynthesis of carbon compounds, W.A. Benjamin Inc., New York. Degens, E.T. (1967). Diagenesis of organic matter. In: Diagenesis in sediments, Larsen, G. and G.V. Chilingar, Eds., Elsevier, Amsterdam, chap. 7. Dudal, R. (1990). Global Soil Change, report of an IIASA-ISSS-UNEP Task Force Meet. on the Role of Soil Global Change, Chap. 3 (in print). Esser, G. (1990). Modeling global terrestrial sources and sinks of COz with special reference to soil organic matter. In: Soils and the greenhouse effect, A.F. Bouwman (Ed.) (1990). John Wiley and Sons, Chichester. Fond der Chemischen Industrie, Umweltbereich Luft (1987). Vol 22, p. 15. Haber. H. (1965. 1971). Die Entwicklungsgeschichte der Erde, Deutsche Verlags Anstalt, Stuttgart, p. 79, p. 208. Lausch, E. (1989). Treibhaus Erde. CEO (Gruner u. Jahr, Hamburg) 37,46-49. Mengel, K. and E.A. Kirkby (1979). Principles of Plant Nutrition. Int. Potash Institute, Bern, Switzerland, 233. Murphy, M, B. Nagy, G. Rouser, and G. Kritchevsky (1965). Analysis of sulphur compounds in lipid extracts from the Orguiel meteorite. J. A. Oil Chem. Soc. 43, 189.196. Ramanathan, V. (1989). Spurengase, Treibhauseffekt und weltweite Erwrmung, In: Das Ende des blauen Planeten ? Crutzen and Muller (Eds) Beck, Federal Republic of Germany, 6576. Rankama, K. (1948). New evidence of the origin of pre Cambrian carbon. Bull. Geol. Soc. Am. 59, 389-416. Raval. A. and V. Ramanathan (1989). Observational determination of the greenhouse effect. Nature 342, 758-761. Schilder, F.A. (1956). Lehrbuch der Allgemeinen Zoogeographie, Jena, German Democratic Republic. Schleser, G. and W. Kirstein (1989). Der Treibhauseffekt. Ursachen und Konsequenzen fur
14
H.W. Scharpenseel and P . Becker-Heidmonn
Klima und Biosphre. Seminar Technik und Gesellschaft, KFA Jlich, Federal Republic of Germany (preprint). Spiegel (1990). M.I. Budyko - Interview. Der Spiegel, Hamburg, 1, 143. Theng, B.K.G. (1979). Formation and Properties of Clay Polymer Complexes, part 3, Chap. 12, 283-314. Theng. B.K.G. and H.W. Scharpenseel (1975). The adsorption of 14-C labelled humic acid by montmorillonite. Proc. Internatl. Clay Conf., Mexico City, 643-653. Valeton, I. (1983). Klimaperioden, lateritische Verwitterung und ihr Abbild in den synchronen Sedimentationsrumen. Z. Dtsche Geol. Ges. 134, 2. Warren-Wilson, I. (1969) Maximum yield potential. In: Transition from extensive to intensive agriculture with fertilizers, Proc. 7th Coll. Intern. Potash Institute, Bern, Switzerland, 34-56. Welte, D.H. (1969) Organic matter in sediments. In: Organic Chemistry, Springer, Berlin, 262-264. Wunderlich, H.G. (1968). Einfhrung in dic Geologie, Vol. 1, Exogene Dynamik. Bibliographisches Institut, Mannheim. Federal Republic of Germany. Wurmbach, U. (1971). Zoologie. Vol. 2, G. Fischer Verlag, Stuttgart.
15
Chapter 2
INPUTS TO CLIMATIC CHANGE BY SOIL AND AGRICULTURE RELATED ACTIVITIES Present status and possible future trends A.F. Bouwman and W.G. Sombroek International Soil Reference and Information Centre P.O. Box 353, 6700 AJ Wageningen, the Netherlands
ABSTRACT The most important soil borne and land use related greenhouse gases are carbon dioxide
(COz), methane (CHd and nitrous oxide (N20). The present annual increase of atmospheric CO2 is 0.5%. The total emission of C 0 2 is 6.5 to 7.5 Gt C y-'. Fossil fuel combustion contributes 5.7 Gt C y-*. The present global rate of deforestation of 10 to 20 million ha y-l causes an emission of 1 to 2 Gt C y-' including the release of C 0 2 from soil organic matter oxidation. There is uncertainty about the sinks of C02. The oceanic uptake is less than 1 Gt C y-1, while the atmospheric accumulation accounts for approximately 3.5 Gt C. Increasing net primary production and other terrestrial sinks balance the budget. The emissions from fossil fuel use will probably increase, and projections of efficiency and magnitude of future energy use are rather uncertain. Deforestation will continue in the coming decades to satisfy the growing demand for agricultural land. The atmospheric concentration of CH4 is rising at a rate of 1% y-'. The major part of this fast increment is caused by increasing emissions, while a minor part can be attributed to decreasing destruction in the atmosphere. The total annual emission from all sources is 400 to 640 Tg CH4. Biotic sources make up about 80% of the total annual emission, the rest is from fossil sources. The biotic sources of CH4 are: wetland rice cultivation (20%of the total source), natural wetlands (20%). ruminating animals (15%). landfills (10%). oceans and lakes (5%) and biomass burning (15%). The contribution by termites is very uncertain. Most sources are increasing at present. Nitrous oxide is increasing at a rate of 0.2 to 0.3% and its sources are almost exclusively biogenic. Nitrous oxide is inert in the troposphere, but it destroys stratospheric ozone ( 0 3 ) . The causes of the increase in atmospheric N 2 0 are not well known. Fossil fuel combustion is a minor source of N20. Increasing use of N-fertilizers in agriculture is a growing source of N2O. The emissions from natural ecosystems are not well known at present .
INTRODUCTION Atmospheric gaseous constituents which are able to absorb thermal radiation and thus contributc to Lhe trapping of heat in the atmosphere are water vapor (H,O), carbon dioxidc (CO,), methane (CH,), nitrous oxide (N,O), ammonia (NH,), halocarbons (the most important are the chlorofluorocarbons, CFCs), and ozone ( 0 3 ) . Of Lhcse, CO;?, CH4, N 2 0 and NH3 are produced in natural and
16
A T . Bouwman and W . G . Sombroek
agroecosystems. To assemble a complete picture of the role of soils and land use in the greenhouse effect, it is necessary to attend to indirect effects of other gases as well. Indirect effects of a source change (i.e. soil change or land use change) are induced changes in a chemical species which affect a given valued atmospheric component through an intermediate influence on another chemical species. Atmospheric constituents with biotic sources having indirect effects on concentrations of the greenhouse gases listed above are carbon monoxide (CO), non-methane hydrocarbons (NMHC, isoprene, C5H8, and terpenes), nitric oxide (NO) and nitrogen dioxide (NO,; NO and NO, as a group are denoted by NO,), ammonia (NH3), oxides of sulphur (SO,), and organic S compounds. Ammonia has indirect effects as well: about 10% of atmospheric ammonia is oxidized to NO and NOz, which play a catalytic role in various photochemical reactions in which 03,CH,, CO and OH are involved. Carbon monoxide is oxidized to CO, thereby affecting many other atmospheric constituents, such as ozone (03). hydroxyl radicals (OH) and CH4. Non-methane hydrocarbons are oxidized by O3 to CO in the atmosphere.
Table 2.1 Residence time of the major greenhouse gases, their rise, atmospheric concentration, radiative adsorption potential and contribution to global warming Type
co2 co CH4 N20 03 CFCs'O
Residence time (y)
Annual
1985
rise (%)
concentration
0S3 0.6-1.07 15 0.256 2.02 3 .02
100' 0.24 8-123 100-2006 0.1-0.32 65-1102
Radiative Contribution absorption to global potential 9 warming 9 345 ppm3 1 50 90 ppbz n.a. n.a. 32 19 1.65 ppm7 300 ppb6 150 4 n.a.8 2,000 8 0.18-0.28 ppb > 10,000 15
-',
Total sources Q is 5-7 Gt C y and atmospheric burden B is 720 Gt. The following relation exists: B = Q x T (see Cicerone and Oremland 1988). The lifelime of C 0 2 must thus be f 100 y (strictly speaking the residence time of C 0 2 is shorter when exchanges between atmosphere and biosphere on the one hand and atmosphere and oceans are allowed for. In that case however, the true importance of C 0 2 is underestimated; Ramanathan et al. (1985) (data for 1980); Bolin (1986); The residence time of CO is not a single constant number, because spatial and temporal variations of sources and sinks are not identical. Total sources and average atmospheric concentration are not known accurately, bu1 accounting for the uncertainties Cicerone (1988) estimates the residence time of CO at 2-2.5 months; Cicerone and Oremland (1988); Crutzen and Graedel (1986); Bolle et al. (1986); 0, vanes from 25 ppb at surface to 7 0 ppb at 9 km (Ramanathan et al. 1985); Enquete Kommission des 11. Deutschen Bundestages "Vorsorge zum Schutz der Erdatmosphare" (1988); "Data presented are for the two major CFCs. Reprinted, with adaptations from Bouwman (1990). by permission from Wiley & Sons, Chichester.
D~
~, ....
-O.
. I . . I -
......
C"
I,
The role of water vapor will not be discussed here. By far the most important other atmospheric greenhouse gases from soil and agriculture related sources are C02, CH4 and N20. These will be discussed in more detail in this paper. We will discuss some chemical aspects of CO in the context of atmospheric chemical reactions of methane. For an overview of the sources and sinks of CO, the reader is referred to Cicerone (1988). A review of the other trace gases not discussed here is presented in Mooney et al. (1987). Due to their relatively long atmospheric residence times C02, CH4 and N20 have important effects on the radiative balance. Some of their characteristics are compared with a number of non-biogenic gases in Table 2.1. The information on the sources and sinks of C02, CH4 and N20 and the possible causes of their increase will be summarized below. In Table 2.3 the global sources and sinks of the three major gases discussed in this paper are summarized.
CARBON DIOXIDE Sources of CO,
The major sources of CO;! are fossil fuel combustion, gas flaring and cement manufacturing, which produced 5.3 Gt CO2-C y-1 in 1984 (Rotty 1987) (Gt = gigaton; 1 Gt = 1015 g), while the 1987 emission was 5.7 Gt C y - 1 (CDIAC 1989). The C02 emission from these sources will amount 2 to 20 Gt C y-l (Keepin et al. 1986), depending on hture efficiency of energy use and shifts to non-carbonaceous sources of energy. Past release of C02 from land use modifications has contributed significantly to the present atmospheric CO;? concentration. Today deforestation and increasing shifting cultivation are responsible for only 1-2 Gt C y-' (Table 2.2). The disagreement between the different suweys of deforestation shown in Table 2.2 is in the fate of fallow lands in shifting cultivation, which at present is being replaced by non-sustainable systems called "sedentary shifting cultivation" or "forest farming" (Houghton et al. 1985), whereby the forest is often not able to re-establish on fallow lands due to soil nutrient depletion. Data presented by INPE (1989) suggest that the deforestation in the Brazilian Amazon region is much slower than the rates presented by the authors quoted in Table 2.2. Part of the CO;?sources shown in Table 2.2 stems from soil organic matter oxidation. Detwiler and Hall (1988) estimated that this soil C 0 2 emission is 0.1 10.25 Gt C y-I for the tropics. Armentano and Menges (1986) estimated that the net release from wetlands as a consequence of drainage, calculated as the sum of loss of sink strength and gain of source strength as a result of wetland drainage, is 0.15 to 0.1 8 Gt C y-l. The role of weathering on fluxes of C02, which often is totally neglected, is discussed in Van Breemen and Feijtel (1990).
18
A P . Bouwman and W.G.Sombroek
Sinks of C O ,
The release of C02 from fossil sources and from land use modifications is being balanced by atmospheric accumulation uptake of C02 in the world's oceans and by the terrestrial biosphere. The increase of 0.5 % y - l in the C 0 2 concentration in the atmosphere accounts for about 3.5 Gt C y-l. The uptake of C02 by the oceans may be less than 1 Gt y - l (Tans et al. 1990). Increased atmospheric C02 concentrations stimulate net primary productivity considerably, and moreover plants will utilize water more efficiently (see Sombroek, Chapter 13 of this volume, for some details). Modelling results (Esser 1987, 1990) suggest that the C02 fertilizing effect is responsible for a growth in net primary productivity which since around 1970 would overcompensate the losses from clearings. The terrestrial biosphere was a small sink of CO;! of about 0.3 (Goudriaan and Ketner 1984) to 0.5 Gt C y-l (Esser 1990) in the early 1980s. The world's wetlands may sequester carbon, while formation of caliche (calcium carbonates) in desert areas is proposed to be a small but important sink. An additional sink of carbon may be the formation of charcoal or inert C , one of the products of biomass burning, which may even amount up to 0.5 - 1.7 Gt C (Seiler and Crutzen 1980). Table 2.2 Annual release of COZ for the early 1980sfrom terrestrial biota including soils caused by deforestation and increasing areas of shifting cultivation, demonstrating uncertainties in estimates. Release rates in Gt C y-I; areas cleared in 106 hay"
2 References
World
Tropics
Houghton et at. (1987)
1.0 - 2.6
0.9 -2.5
Detwiler and Hall (1988)
0.3 - 1.7
0.4
- 1.6
Area of permanent clearing
in the tropics low: 6.0 a high:15.2 low: 3.0 medium: 10.6 high: 15.2
F A 0 (1983); Myers in Houghton et al. (1985); Seiler and Crutzen (1980); Lanly (1982); ' Detwiler and Hall (1988) concluded from results in other studies that temperate forests contributed -0.1 to +0.1 Gt C y - ' in 1980; their own analysis considers only tropical deforestation. B o ~ hlow and high figures of forest biomass were used in the presented estimates. The low estimate for the tropics is based on forest volume figures: 90 t C h a - ' for closed and open forests, respectively (Brown and Lug0 (1984). The other estimate which is based on destructive sampling yields higher values of between 164 and 40 I ha-' for different types of tropical and subtropical forests (Brown and Lugo, 1982). Reprinted with adaptaoons form Bouwman (1990b). by permission of Butterwonh Scientific Ltd, UK.
Inputs to climatic chonge by soil and agriculture related activities
Table2.3
19
Global emissions and sinks of major greenhouse gases. G t = gigaton, 1 Gt = g, Tg = teragram, I Tg = 10l2 .
co2 Global annual 6.5- 7.5 Gt C emission* % Biotic 20-30
CH4
N20
400-640 Tg CHq
11-17 Tg N
70-90
90-100
Sources*
fossil fuel use (5.7 Gt) deforestation and shifting cultivation (1-2 Gt)
rice paddies (60-140 Tg) cultivated soils (S Tg?) wetlands (40-160 Tg) natural soils (?) ruminants (65-100 Tg) fossil fuel (?) termites (10-100 Tg) landfill sites (30-70Tg) oceansflakes(15-35) biomass burning (50100 Tg) fossil (50-95 Tg), coal mining/gas exploitation
Sinks*
atmospheric accumulation (3.5Gt) Oceans (<1 Gt) biosphere (? Gt)
atmospheric accumu- atmospheric accumulation (50Tg) lation (2.8) soil oxidation (32 Tg) atmospheric chemistry atmospheric chemistry (10.5) (300-650 Tg) soils (?)
charcoal formation (?)
* The total budget or total sources is not necessarily the s u m of the indicated individual sources. The total of sources can be derived independently from the atmospheric burden, residence time and measured increase (see also Table 2.1) or from models of atmospheric chemical processes. Reprinted with adaptations from Bouwman(l990b), by permission from Butterworth Scientific Ltd, UK. Future developments Possible future C 0 2 concentrations were analyzed using the Osnabriick Biosphere Model which includes C 0 2 fertilization (Esser 1990). The results suggest that depending on different energy scenarios (0.1% and 0.5% increase annually) and progressive clearing of 50% to 100% of the area of natural vegetation in the year 2400, the 2 x pre-industrial C 0 2 level (2 x k 285 ppmv = k 570 ppmv) will be reached in the year 209 1 (worst case) or 2 181 (best case). A similar analysis was camed out by Swart and Rotmans (1990), who used the dynamic biosphere model developed by Goudriaan and Ketner (1984). Based on a slow increase of 0.7% y-l of the energy use from 5.4 in 1985 to 12.5 Gt C y-l in the year 2100, the difference in 2100 between total deforestation and gradually
20
A.F. Bouwman and W . G . Sombroek
decreasing or even reversed deforestation was 60 ppm, being only 10% of the projected 2x pre-industrial CO, levels. This illustrates the modest role of the terrestrial biosphere. Actively growing trees accumulate carbon. Hence, forest plantations can be implemented to absorb part of the annual atmospheric C 0 2 injection. Approximately 465 x lo6 ha (which is equal to about the total land area of Europe) of forest plantations would sequester the present 3Gt C0,-C surplus of sources over sinks, based on very high accumulation rates of 6240 kg C ha-ly-l in new forests (Sedjo 1989). The future demand for different kinds of wood and the feasibility of afforestation and the actual productivity of planted forests was considered by Wiersum and Ketner (1989). On the basis of literature data these authors estimated that in the year 2000 there will be a worldwide need for 55 lOOx lo6 ha of afforestation to cover fuelwood deficits, 20-40x106 ha for industrial wood needs and 127-190x106 ha for environmental protection. The latter plantations would occur mainly in waste lands and regions in process of desertification. Part of the environmental plantations would also serve for fuelwood or industrial wood production, so that the authors arrive at their final estimate of 1 6 0 ~ 1 ha 0 ~for long rotation forest plantations with fast growing species and 4Ox1O6ha of plantations for fuelwood (mainly subsistence). It was assumed that the long rotation plantations (environmental + industrial wood) have an effective period of 30 years during which carbon accumulation is 3000 kg y - l (note that this accumulation is much lower than the rate reported by Sedjo 1989). Fuel wood forests sequester 2000 kg C ha-ly-' during an effective period of 5-10 y. The 1 6 0 ~ 1 ha 0 ~of long rotation forest and 4Ox1O6 ha of new forest for fuelwood would sequester 0.5 and 0.1 Gt C y-l, respectively. In addition an assessment was made of the feasibility of high-productive wood energy plantations, with rotation periods of 8 years and carbon fixation of 4500 kg C y-'. These plantations would replace areas of cropland, primarily in the USA and Europe, that may be taken out of production in future. The projected 30x lo6 ha of highly productive energy wood forest plantations wouId substitute 0.12 Gt C y-l of fossil fuel consumption. Wiersum and Ketner (1989) did not consider effects of such massive plantations on the partitioning of C 0 2 between atmosphere, oceans and biosphere, and this requires further analysis with dynamic models. Planting of trees in agroforestry systems, which are an alternative to shifting cultivation, can also contribute in this respect. A secondary effect of reduction of slash and bum practices is reduction of emissions of other gases and aerosols (see discussion of CH4 from biomass burning below). Currently the worldwide reforestation rate (on previously forested lands) is 1 4 . 5 ~ 1ha 0 ~(World Resources Institute 1988). Increasing tropical afforestation to achieve extents in the order of 2 0 0 ~ 1 0ha~ (half of the total land area of Europe) will meet many problems (Wiersum and Ketner 1989). This area of forest plantations would have to be increased by a factor 4 to 5 to sequester the
Inputs to climatic change by soil and agriculture relaled activities
21
total annual C 0 2 increase and this is only effective during the forest's active growth period. Large scale intensification of agriculture is therefore required to release land for afforestation and at the same time satisfy the growing demand for food. Populations in rural areas in the developing countries will not easily cooperate with long term afforestation programmes if their immediate needs are not fulfilled. Moreover, the major producers of C02 are the industrialized nations.
METHANE General The methane concentration in the atmosphere has showed an accelerated increase during the past 300 years (Khalil and Rasmussen 1985; Steele el al. 1987), and its current increase is 1% y - l or 40-90 Tg CH4 y-I (Tg = teragram; 1Tg = 1012g). Methane's atmospheric residence time is relatively short (Table 2.1). Its concentration in the atmosphere and its increase depend primarily on short term fluxes and removal processes. The major portion of the observed rise over the past 300 years is caused by increasing sources (Khalil and Rasmussen 1985). Hydroxyl radicals (OH)' are responsible for the removal of CH4 and CO. Parallel to increasing CH4 there is a strong increase of CO in the atmosphere, caused primarily by anthropogenic sources2 . This increase in part also depresses oxidation of CH4 by depleting OH radicals. Since CH4 is a source of CO itself, there is an instability in the CHJCOIOH system: increases of CH4 and CO lead to depletion of OH, and this leads to further perturbations of CH4 and CO. Khalil and Rasmussen (1985) estimated that 30% of the increase of CH4 is attributable to OH depletion, whereas Cicerone and Oremland (1988) suggested that the OH depletion induces an increase of atmospheric CH4 of only 0.5 to 3 Tg y-l (which
'
OH radicals are the cleaning agents of the atmosphere. They oxidize CH, and in that process CO is formed. What occurs during further oxidation of C O to CO, depends on the concentration of nitrogen monoxide (NO). With sufficient NO (generally in industrialized regions and tropical areas where biomass burning takes place), OH radicals and O3 F are formed. O3 is increasing at Northern mid latitudes and in the upper troposphere. In the absence of NO (e.g. in remote marine areas) O3 is consumed in the oxidation process of CO (Crutzen and Graedel 1986). CO has an atmospheric residence time of only 2-3 months (Cicerone 1988). Therefore determination of a secular trend is very difficult. CO has probably increased in the northern hemispheric troposhere since 1950. Its increase is not uniform or steady in space. No positive trend is observed in the Southern hemisphere.
22
A P . Bouwman and W.G. Sombroek
is much less than the observed total increase of 40-90 Tg y-I), illustrating the uncertainty of these atmospheric processes. Methane is produced during microbial decomposition of organic material under anaerobic conditions. Natural wetlands, wet rice cultivation and landfill sites for solid waste dumping are places where anaerobic conditions prevail. Methane is also formed in the digestive tract of ruminating animals and in the guts of various insects, the major species being termites. The a-biogenic process of biomass burning is a further source of CH4. Recent investigations showed that 20 to 30% of total sources is fossil or 'dead' CH4. This fraction includes approximately 30 Tg of dead CH4 from wetlands and old peat layers (Cicerone and Oremland 1988); emissions from fossil sources therefore amount to 50-95 Tg y-I (Table 2.3). A further source of methane is release of radioactive CH4 by nuclear power plants (see Grantham, Chapter 19 in this volume) which is a negligible source in amount, but with important effects on the isotopic composition of atmospheric CH4. Sinks for CH4 are accumulation in the atmosphere, oxidation by OH, and oxidation in soils. Their relative sizes are highly uncertain (Table 2.3). Land use related sources which will become increasingly important in the future are: wetland rice cultivation, natural wetlands, landfills, ruminants and biomass bu mi ng . Wet rice cultivation Paddy rice fields currently constitute approximately 20% of the total CH4 budget. The harvested area of paddy rice, about 10% of the world's total cultivated area, has shown a steady increase of more than 1% y-I during the last decades (Bouwman 1990a). By the year 2020 the global rice production may have to be increased by a factor 1.6 to meet the growing demand. Extrapolation of available flux measurement data to the global harvested area of paddy rice of each type (dryland, rainfed and irrigated) yields a global emission of 60-140 Tg CH4 y-* (ASehaM and Cmtzen 1989). The uncertainty in this estimate is caused by lack of flux measurements, gaps in the knowledge of the process of methanogenesis and lacunae in the geographic database of the harvested area of the various rice ecologies, soil types used, and soil, water and crop management practices. The use of straw and compost stimulate soil reduction and methanogenesis, while the application of sulphate containing fertilizers (such as ammoniumsulphate) may reduce methanogenesis. Specific agronomic practices such as adapting levels, application mode and timing of fertilization (including chemical as well as organic fertilizers), water management and new rice varieties may help reducing the methane emissions from flooded rice fields. Flooded rice soils with dry fallow periods between rice growing seasons, and fields that are likely to dry up during the cultivation period, are not prone to high CH4 production (Neue
lnpufs to climatic change by soil and agriculture related activifies
23
1990). Specific combinations of soil characteristics and agronomic practices are not likely to show high methane productions. Heavy textured soils with high content of montmorillonitic clays and high content of organic carbon, will probably show high rates of CH4 formation. In leached tropical soils and other soils with fenitic, gibbsitic, fermginous or oxidic mineralogy, the methane production will probably not be high if dry fallow periods are included in the crop rotation. It appears feasible to reduce CH4 emission rates at field level by 10 to 30% (Neue, pers. comm.). However, the global emissions may still increase due to intensification of cultivation and expansion of the cultivated area.
Natural wetlands The global wetland area is 530x106 ha (Matthews and Fung 1987) to 5 7 0 ~ 1 0ha~ (Aselmann and Crutzen 1989), but the various estimates show regional discrepancies and definitional differences. Most of these occur between 50 ON and 70 ON and between 10 ON and 10 "S. The global emission from wetlands ranges between 40 and 160 Tg y-l (Matthews and Fung 1987; Aselmann and Crutzen 1989). Better geographic information coupled with more flux measurements are therefore required to narrow the range of estimates. Methanogenesis is a temperature dependent process. It may even show a 5 fold increase with every 10 "C temperature increase (Burke et al. 1990). It is suggested that - based on past temperature anomalies and estimates of future warming - by the year 2080 CH4 emissions from wetlands will amount to 3 times the 1880 values and more than 2 times the present emissions (Burke et al. 1990). This estimate may be too high, because CH4 oxidation will also be stimulated, possibly even more than CH4 formation. Predictions of direction and magnitude of changes in the area of wetlands due to climatic change are hampered by the uncertainty about future climatic conditions at high latitudes and of temperature enhanced evaporation. In addition to the stimulation of methanogenesis, the possibility of destabilization of methane in hydrates by warming has been raised by various reports (see review by Cicerone and Oremland 1988). These hydrates are solid structures, composed of strong cages of water molecules surrounding CH4 molecules. Methane hydrates are most prevalent at depth in permafrost and in sea sediments. A tentative estimate of the CH4 release from hydrates is 5 Tg CH4 y-l (Cicerone and Oremland 1988). The total area of artificially drained wetlands in temperate regions in the period 1795-1980is 23x106 ha (being 6% of the temperate wetlands), of which 60% in Finland-USSR, 25% in Western and Central Europe and 7% in North America. Drainage of tropical wetlands during that period was only 4%. In the future large areas of natural wetlands will be affected by land reclamation. Large parts of Amazonian wetlands may vanish, while in other places new wetlands
24
A.F. B o u w m n and W . G . Sombroek
will be created in the realization of hydro-electric power schemes. Extension of cattle farming in seasonally flooded or submerged savannas and reclamation of wetlands is proceeding in many parts of South America, while in other tropical regions wet rice cultivation often occurs in reclaimed wetlands. Landfills Anaerobic decay of collected municipal and industrial organic matter that is dumped in landfills add 30-70 Tg CH4 y-l to the atmosphere globally (Bingemer and Crutzen 1987). By far the largest emission currently comes from landfills in the industrialized world. This source has been increasing in recent decades but at present its rise is stagnating. A growing urban population in the developing countries will produce more solid wastes. Projections for the year 2000 suggest a doubling of waste generation in developing countries (Bingemer and Crutzen 1987). Reduction of emissions from landfills may be achieved by gas harvesting or by recycling of solid wastes. Ruminants and termites The global CH4 emission by ruminant animals is 65-100 Tg y-1. Approximately 80% of this amount is produced by cattle and buffaloes, the rest is from sheep (7 Tg y-I), wild ruminants (2-6 Tg y-1) and others (Crutzen et al. 1986). About 40% of the CH4 produced by domestic animals is emitted in the developing countries, mainly in Asia, followed by South America and Africa. The major source regions for CH4 from cattle are North America (1 1 Tg y-'), Europe (8 Tg y - l ) and the USSR (7 Tg y - l ) (Crutzen et al. 1986). This source deserves priority for further research. Changing from animal protein to more plant proteins in human nutrition is a possible way of reducing these CH4 emissions. Modifying feeding strategies and rumen processes and development of CH4 inhibitors also merits attention (EPA 1989). The methane production by termites ranges from 10 to 100 Tg y-1 (Cicerone and Oremland 1988). Most measurements were carried out in vitro. Field measurements are required here, since living conditions and type of nutrition of the numerous families of termites are completely different. It is suggested that the total population of termites shows an increasing trend, caused by extensions of the global cultivated area, burning and forest clearing activities and possibly by increasing biomass production through CO, fertilization. Biomass burning A source of CH4 which is closely related with land use is the burning of biomass, such as burning of agricultural wastes, savannah fires, slash and bum agriculture (shifting cultivation) and buming of fuelwood. On the basis of the
lnpuls lo climatic change by soil and agriculture related oclivilies
25
increasing population of shifting cultivators, increasing production of industrialand fuelwood and of food production, Seiler (1984) estimated that the CH4 production from biomass burning increased from 41-74 Tg y-' in 1950, to 47-84 Tg y-l in 1960 and 53-97 Tg y-l in 1975. These estimates have recognized the large uncertainties and sources of variability, such as amounts of biomass burnt, types and conditions of burning. Growing populations will increase the demand for agricultural land and fuelwood and therefore this source will probably continue to increase in the future. Possible ways to reduce this source are replacement of slash and bum agriculture by agroforestry and alternative noncarbonaceous sources of energy (such as solar energy) to substitute fuelwood.
NITROUS OXIDE Nitrous oxide destroys ozone in the stratosphere (above 10-20 km height) but it is inert in the troposphere. This atmospheric photolysis is a sink of 10.5 Tg N20-N y-1, while atmospheric accumulation of 0.2-0.3% is equal to 2.8 Tg y-l (Table 2.3). Analysis of antarctic ice core samples suggests that the atmospheric N 2 0 concentration has risen from about 270 ppb 400 years ago to about 293 ppb in the beginning of the 20th century (Zardini et al. 1988). Records from longer periods indicate a slight increase in N 2 0 when climate was warming. Nitrous oxide is formed during nitrification and denitrification processes in soils. Nitrification is microbial transformation of ammonia (NH4+) to nitrate (NO3-) while denitrification is transformation of NO3- to molecular nitrogen (N2). Nitrification is dominant in aerobic soils. Denitrification occurs in anaerobic environments and in anaerobic microsites in well aerated soils. Nitrous oxide is an intermediate product of these microbial transformations and, being a gas, it may escape from the soil system to the atmosphere. The turnover rates of N in ecosystems are important for N20 losses. Factors influencing the speed of microbiological Nturnover are soil temperature, soil moisture, soil fertility, availability of organic substrate and soil drainage. N20 fluxes show an enormous spatial and temporal variability. The extrapolation of flux measurements to smaller scales is therefore fraught with potential errors. There are a number of possible explanations for the current 0.2-0.3% increase of the atmospheric N20 concentration. The expanding area of agricultural land and the increasing use of nitrogen fertilizers are identified as major causes. Part of the losses from N fertilizer in the form of N2O occurs directly from soils. Another pathway is through losses from N containing water leached from cultivated fields. The world's cultivated lands, about 1500 x 106 ha, may emit 2.4 to 3.6 Tg of N2O-N annually (Bouwman 1990a). The N20 emission can also be estimated as the fertilizer induced loss, which is the fraction of applied N-fertilizer lost as N20 exclusive of 'natural' emissions. This loss is in
26
A P . Bouwman and W.G. Sombroek
the range of 0.5 to 2.0% of applied N fertilizer (Bolle et al. 1986). In recent years the consumption of fertilizer increased from 67 Tg in 1983/84 to 72 Tg in 1986/87 (FA0 1988). Fertilizer induced N 2 0 emission was thus 0.4 to 1.4 Tg N20-N y-l in 1987. The fertilizer consumption figures suggest that this source is increasing. Literature data show that under certain conditions the timing of Nfertilizer application and choice of fertilizer type can strongly influence N20 emissions (Bouwman 1990a). Legumes may play an important part in the production of N2O. The symbiosis between leguminous host plants and Rhizobial bacteria can add nitrogen to the soil-vegetation system through fixation of atmospheric N2. Freeliving Rhizobia may be important denitrifiers. Free living Rhizobia can persist in infertile environments, thanks to a combination of properties such as ability to assume dormancy, slow growth, ability to denitrify and slow metabolism. It is also possible that symbiotically living Rhizobia bactemids in root nodules are able to denitrify. If symbiotic denitrification is important, the cultivation of leguminous crops is a potential increasing source of atmospheric N20. More on this subject can be found in O'Hara and Daniel (1985). Further potential causes of increasing N20 emissions are increasing amounts of N reaching the soil through wet and dry NO, deposition and NH4+in rainwater, which add to soil nitrogen and may in part be lost as N20, and N20 produced in surface waters and oceans. Tropical rainforests and tropical savannas emit high amounts of N20 per unit area. High rates of N20 emission in tropical savanna areas were measured by Hao et al. (1988). In the case of savannas the occurrence of alternating wet and dry conditions may enhance N20 fluxes. Leguminous plants occumng in these ecosystems contribute to soil-N through N-fixation. At the end of the dry season in savanna regions the availability of organic substrate in the soil is high. The first rains initiate a very fast mineralization of this organic matter and nitrification of N compounds. Since high temperatures favour fast turnover rates, under these conditions high N2O losses may be expected (see e.g. Keller et al. 1988). The emissions of N 2 0 may increase after replacing the original tropical forest vegetation by grassland (Matson, pen. comm.), possibly caused by an increased frequency of rewetting soils after periods of drought. Another source is N2O from fossil fuel combustion and from biomass burning, but their annual emissions are highly uncertain. Recently Muzio and Kramlich (1988) reported that estimates published so far of N20 fluxes by fossil fuel combustion are much lower than the 1 to 2 Tg y-l (Seiler and Conrad 1987) postulated in previous global budgets. On the basis of new data, Crutzen (pers. comm.) estimated that N 2 0 emission by biomass burning is less than half the 1.5 Tg y-l assumed by Seiler and Conrad (1987). Clearly, the global N20 budget needs further investigations.
Inputs
10 climaric
change by soil and agriculture related activities
27
CONCLUDING REMARKS With regard to COz the rate of deforestation still forms an uncertain factor. Definitional controversy on land use and vegetation types causes great disagreement between estimates of C02 emissions. A further cause of unreliability in the estimates is the biomass in tropical forests, but its effect on the net flux of C02 is small compared to the effect of uncertainties in the rate of deforestation. The accumulation in the atmosphere is well known, while the uptake in the Oceans is probably lower than assumed until recently. The increase of net primary production caused by increasing atmospheric C 0 2 concentrations needs further research to solve the uncertainties in the C02 budget. The production and oxidation of methane in soils and wetlands is extremely dependent on environmental conditions. As for the nitrous oxide, the variability of fluxes both in time and space of fluxes is extremely high. A number of specific fields require research attention: 1). the geographic distribution of rice ecologies, soil, crop and water management, and soil types used for wetland rice cultivation; 2) the soil processes responsible for the production and emission of methane; 3) the geographic distribution of the different types of salt and fresh water wetlands and the relation between the type of wetland and methane fluxes; and 4) fluxes of CH4 from landfill sites, especially quantities of organic waste which is decomposed anaerobically in landfill sites are virtually unknown. Methods to reduce the CH4 release from landfills merits research attention. The above investigations will also improve the capability to assess consequences of climate change for methane fluxes, especially those from natural wetlands. The allocation of N20 emissions to different sources is still very uncertain. As shown above, there is no good explanation for the atmospheric increase, which may in the future even be amplified by changing concentrations of other atmospheric constituents. Much more process oriented research parallel to flux measurements is needed, especially in heavily fertilized areas and in ecosystems where major changes occur. As for land use, an option which needs consideration is stimulation of agronomic practices to depress N 2 0 losses from agricultural lands and surface waters, stimulation of efficient N-fertilizer use and prevention of over-fertilization,and use of nitrification inhibitors. The currently used methods of measuring trace gas fluxes are point measurements. Extrapolation of such measurements to smaller scales is fraught with potential emrs. The development of methods to estimate trace gas fluxes for larger areas is needed to improve the quantification of the global and regional sources. A geographic database of present-day land cover and land use types, the latter including management practices, is urgently needed.
28
A . F . Bouwman and W.G. Sombroek
REFERENCES Armentano, T.V. and E.S. Menges (1986). Patterns of change in the carbon balance of organic soil wetlands of the temperate zone, Journal of Ecology 75:755-774. Aselmann, I. and P.J. Crutzen (1989). Freshwater wetlands: Global distribution of natural freshwater wetlands and rice paddies: their net primary productivity, seasonality and possible methane emissions, Journal of Atmospheric Chemistry 8:307-358. Bingemer, H.G. and P.J. Crutzen (1987). The production of methane from solid wastes, Journal of Geophysical Research 92:2181-2187. Bolin, B. (1986). How much C 0 2 will remain in the atmosphere? The carbon cycle and projections for the future. In: B. Bolin, B.R. Doos, J. Jager, and R.A. Warrick (Eds.) The greenhouse effect, climatic change and ecosystems, pp 93-156, SCOPE Vol. 29, Wiley and Sons, New York. Bolle, H.J., W. Seiler and B. Bolin (1986). Other greenhouse gases and aerosols. Assessing their role in atmospheric radiative transfer. In: B. Bolin, B.R. Doos. I. Jager and R.A. Warrick (Eds.), The greenhouse effect, climatic change and ecosystems, pp 157-203, SCOPE Vo1.29, Wiley and Sons, New York. Bouwman, A.F. (1990a). Exchange of Greenhouse Gases between terrestrial ecosystems and the atmosphere. In: A.F. Bouwman (Ed.), Soils and the Greenhouse effect, pp 61-128, Wiley and Sons, Chichester (1990). (Adapted version of ISRIC Working Paper 89/4, International Soil Reference and Information Center, Wageningen, the Netherlands). Bouwman, A.F. (1990b). Land use related sources and sinks of greenhouse gases. Present emissions and possible future trends, accepted in Land Use Policy. Brown, S . and A.E. Lug0 (1982). The storage and production of organic matter in tropical forests and their role in the global carbon cycle, Biotropica 14:161-187. Brown, S. and A.E. Lug0 (1984). Biomass of tropical forests: a new estimate based on forest volumes, Science 223:1290-1293. Burke, M.K., R.A. Houghton and G.M. Woodwell (1990). Progress toward predicting the potential for increased emissions of CH4 from wetlands as a consequence of global warming. In: A.F. Bouwman (Ed.), Soils and the greenhouse effect, pp 451-456, Wiley and Sons, Chichester (1990). CDIAC (1989). CDIAC Communications, CDIAC's corner, Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, Summer 1989. Cicerone, R.J. and R.S. Oremland (1988). Biogeochemical aspects of atmospheric methane, Global Biogeochemical Cycles 2299.327. Cicerone, K.J. (1988). How has the atmospheric concentration of CO changed ? In: F.S. Rowland and I.S.A. Isaksen, The changing atmosphere, Dahlem Workshop Reports, pp 49-61, Wiley and Sons, Chichester. Crutzen, P.J. and T.E. Graedel (1986). The role of atmospheric chemistry in environmentdevelopment interactions. In: W.C. Clark and R.E. Munn (Eds.), Sustainable development of the biosphere, pp 213-250, IIASA, Laxenburg, Austria, Cambridge University Press. Crutzen, P.J., I. Aselmann and W. Seiler (1986). Methane production by domestic animals, wild ruminants, other herbivorous fauna and humans, Tellus 38B:271-284. Detwiler, R.P. and C.A.S. Hall (1988). Tropical Forests and the Global Carbon Cycle, Science 239142-47. Enquete-Kommission Vorsorge zum Schutz der Erdatmosphare ( I 989). Schutz der Erdatmosphiire: eine internationale Herausforderung; Zwischenbericht der EnqueteKomm. des 11. Deutschen Bundestages Vorsorge zum Schutz der Erdatmosphae, Dt. Bundestag, Referat Offentlichkeitsarbeit, Bonn.
lnpuls to climatic change by soil and agriculture relared activities
29
EPA (1989). Reducing methane emissions from livestock: opportunities and issues (by M.J. Gibbs, L. Lewis and J.S. Hoffman), US Environmental Protection Agency, Washington, D.C., USA. Esser, G . (1987). Sensitivity of global carbon pools and fluxes to human and potential climatic impacts, Tellus 39B:245-260. Esser, G. (1990). Modelling global terrestrial sources and sinks of C 0 2 with special reference to soil organic matter. In: A.F. Bouwman (Ed.), Soils and the greenhouse effect, pp 247-262, Wiley and Sons, Chichester (1990). F A 0 (1988). Fertilizer Yearbook. F A 0 Statistics series No. 83. 123 p., FAO, Rome. F A 0 (1983). Production yearbook Vol. 37. F A 0 Statistics series No. 55, F A 0 Rome. Goudriaan, J. and P. Ketner (1984). A simulation study for the global carbon cycle, including man's impact on the biosphere, Climatic Change 6:167-192. Hao, W.M., D. Scharffe and P.J. Crutzen (1988). Production of N20, CH4 and C 0 2 from soils in the tropical savanna during the dry season, Journal of Atmospheric Chemistry 7:93105. Houghton, R.A., R.D. Boone, J.M. Melillo, C.A. Palm, G.M. Woodwell, N. Myers, B. Moore I11 and D.L. Skole (1985). Net flux of carbon dioxide from tropical forests in 1980, Nature 3 16:6 17-620. INPE (1989). Avaliacao de alteraqao da cobertura florestal na Amaz6nia legal utilizando sensoriamento remoto orbital, Instituto de Pesquisas Espaciais, Roberto Pereira da Cunha (coordinator), LNPE, Sao Jost dos Campos, May 1989. Keepin, W., I. Mintzer and L. Kristoferson (1986). Emission of C 0 2 into the atmosphere. The rate of release of C 0 2 as a function of future energy developments. In: B. Bolin, B.R. DOOs, J . Jager and R.A. Warrick (Eds.), The greenhouse effect, climatic change and ecosystems, pp 35-91, SCOPE Vo1.29, Wiley and Sons, New York. Keller, M., S.C. Wofsy and J.M. Da Costa (1988) Emission of N 2 0 from tropical soils: response to fertilization with NH4+, NO3- and P043-, Journal of Geophysical Research 93: 1600-1604. Khalil, M.A.K. and R.A. Rasmussen (1985). Causes of increasing atmospheric methane: depletion of hydroxyl radicals and the rise of emissions, Atmospheric Environment 191397-407. Lanly, J.P. (1982). Tropical forest resources. F A 0 Forestry Paper No. 30. 106 pp. FAO, Rome. Matthews, E. and I . Fung (1987). Methane emission from natural wetlands: Global distribution, area, and environmental characteristics of sources, Global Biogeochemical Cycles 1:61-86. Mooney, H.A., P.M. Vitousek and P.A. Matson (1987). Exchange of materials between terrestrial ecosystems and the atmosphere. Science 238:926-932. Muzio and Kramlich (1988). Artifact in the measurement of N 20 of combustion sources, Geophysical Research Letters 15:1369-1372. Neue, H.U., P. Becker-Heidemann and H.W. Scharpenseel (1990). Organic matter dynamics, soil properties and cultural practices in rice lands and their relationship to methane production. In: A.F. Bouwman (Ed.), Soils and the greenhouse effect, pp 457-466, Wiley and Sons, Chichester (1990). OHara, G.W. and R.M. Daniel (1985). Rhizobial denitrification: a review, Soil Biology and Biochemistry 17, 1985, pp 1-9. Ramanathan, V., R.J. Cicerone, H.B. Singh & J.T. Kiehl (1985). Trace gas trends and their potential role in climate change, Journal of Geophysical Research 90, pp 5547-5566. Rotty, R.M. (1987). A look at 1983 C 0 2 emissions from fossil fuels (with preliminary data for 1984). Tellus 39B:203-208.
30
A.F. Bouwman and W.G. Sornbroek
Sedjo. R.A. (1989). Forests to offset the greenhouse effect, Journal of Forestry, pp 12-15, July 1989. Seiler, W. (1984). Contribution of biological processes to the global budget of CH4 in the atmosphere In: M.J. Klug and C.A. Reddy (Eds.), pp 468-477, American Society for Microbiology, Washington, D.C.. Seiler, W. and R. Conrad (1987). Contribution of tropical ecosystems to the global budgets of trace gases, especially CH4, H2, CO and N 2 0 . In: R.E. Dickinson (Ed.), The geophysiology of Ammonia. Vegetation and climate interactions, pp 133-160, Wiley and Sons, New York. Seiler, W. and P.J. Crutzen (1980). Estimates of gross and net fluxes of carbon between the biosphere and the atmosphere from biomass burning, Climatic Change 2:207-247. Steele, L.P.. P.J. Fraser, R.A. Rasmussen, M.A.K. Khalil, T.J. Conway, A.J. Crawford, R.H. Gammon, K.A. Masarie and K.W. Thoning (1987). The global distribution of methane in the troposphere, Journal of Atmospheric Chemistry 5:127-171. Swart, R.J. and J. Rotmans (1990). Food or forest? Can the tropical forests survive along with continuing growth of population and economy? In: A.F. Bouwman (Ed.), Soils and the Greenhouse Effect, pp 431-440, Wiley and Sons, Chichester (1990). Tans, P.P., I. Y. Fung and T. Takahashi (1990). Observational constraints on the global atmospheric carbon dioxide budget, accepted in Science. Van Breemen, N, and T.C.J. Feijtel (1990). Soil processes and properties involved in the production of greenhouse gases, with special relevance to soil taxonomic systems. In: A.F. Bouwman (Ed.), Soils and the greenhouse effect, pp 195-224, Wiley and Sons, Chichester (1990). Wiersum, K.F. and P. Ketner (1989). Reforestation, a feasible contribution to reducing the atmospheric carbon dioxide content. In: P.A. Okken, R.J. Swart and S. Zwerver (Eds.), Climate and Energy, the feasibility of controlling C 0 2 emissions, pp 107-124, Kluwer Publishing Company, Dordrecht. World Resources Institute (1988). World Resources 1988-89. An assessment of the resource base that supports the global economy, with data tables for 146 countries, World resources Institute, International Institute for Environment and Development. Zardini, D., D. Raynaud, D. Scharffe and W. Seiler (1988). N 2 0 measurements of air extracted from Antarctic ice c b r e s : implication on atmospheric N 2 0 back to the last glacialinterglacial transition, Journal of Atmospheric Chemistry 8:189-201.
31
Chapter 3
PROCESSES THAT AFFECT SOIL MORPHOLOGY Richard W. Arnold USDA-Soil Conservation Service P. 0. Box 2890, Washington, D.C. 20013, USA
INTRODUCI'ION Soil morphology consists of the features and properties of soils that ~IC commonly visible when looking at a profile or handling samples of soil. The more usual features are associated with colors and their variation in kind, amount, location and distribution. The presence of organic matter, soluble salts, carbonates, iron compounds, and colors inherent in some parent materials are all examples of variations in color that we recognize as differences of morphology. In addition to color patterns there are aggregates of soil, peds, that give rise to patterns of organization and distribution throughout a profile. When a sample of soil is examined we can feel the differences in resistance to crushing which we call consistence and also feel the texture which is the combination of sand, silt and clay particles. The processes that give rise to variations in soil morphology can be grouped (Fridland 1976) into: Those that translocate the particles without significant alteration chemically; Those that translocate solutes, cause chemical alteration of compounds through reduction and oxidation, effect degrees of hydration, and so forth; Those that result in mixing of soil materials by animals or alter the soil through biotic influences; Those that are anthropically induced and may be planned such as imgation or terraces, or may be unplanned such as enhanced erosion due to grazing, cultivation, or other disturbances. It is also common and very likely that combinations of these processes are the major influences giving rise to variations of soil morphology.
GEOPHYSICAL PROCESSES Where the physical particles and the soil aggregates are moved without much alteration, the processes may be considered to be geophysical. Water
32
R . W . Arnold
erosion is an important geophysical process that gives rise to variations in soil morphology. Locally some soil materials are removed, transported across a slope and deposited elsewhere. The removal and additions of these particles and their subsequent arrangements are recognized as changes in the color and suucture patterns of a soil's morphology. Where the processes have been responsible for major accumulations of parent materials e.g. volcanic ash, the initial state of the profile is influenced, in contrast to the micro variations that occur in the surface layers of soils. Most landscapes in the world have been influenced markedly by the action of running water; thus landscape evolution and soil morphology tend to go hand in hand. Small amounts of sheet and rill erosion are usually obliterated by tillage operations and thus the impact on soil morphology tends to be transient in the minds of the observers. Wind deflation is a companion process that removes soil particles from one place which are transported and deposited in adjacent areas. Similar microstratification is observed in alluvial environments along streams and on terraces where sediments are deposited from the overflow waters. In places erosion also takes place. The alternating deposition and removal of sediments on alluvial fans gives rise locally to recognizable patterns of soil mophology. Within soil profiles the process of translocation of physical particles, often clay and/or organic matter, result in easily recognizable differences of morphology. In fact such features are usually identified as specific kinds of soil horizons, such as argillic for the clay enriched ones, and spodic (podzolic) for the humus enriched ones. Other processes that move soil particles include landslides and other forms of mass movement resulting in mixing and disturbance of the soil materials. The melting of ice and freezing of water in profiles and the variability of permafrost also moves soil materials, often giving rise to unique patterns of soil morphology. The churning of clayey soils due to wetting and drying also produces characteristic patterns of soil morphology. Fig. 3.1 illustrates an extreme case of vertic properties giving rise to short range variability in a Vertisol from Texas, U.S.A.
GEOCHEMICAL PROCESSES Where the major differences in morphology are related to the chemistry of compounds, whether translocated solutes or chemically altered materials, the processes can be thought of as being geochemical. In humid areas where relief gives rise to moisture differences across the landscape (catena) the patterns of soil morphology shift in definite patterns from drying on the lower slopes resulting in oxidation-reduction phenomena and the more freely draining uplands and upper slopes to the fluctuating wetting and
Processes
that
affecis soil morphology
33
influence on color patterns in soils. In drier areas the presence and absence of ground water at relatively shallow depths modifies the soil morphology as soluble salts are translocated upwards, downwards and laterally. Within the capillary fringe oxidation and reduction often occurs. Karst like suffusion alters minerals and removes some of the compounds, thus a sequence of changes of soil morphology may be observed in such environments. Variations in rocks that make up the parent rock of soils are often different in chemical and physical properties that influence how and where features will be observed in soils. In some environments there are paleo soils and parent materials that become exposed and blend into the current stages of soil evolution, consequently part of the observable soil morphology is relict. Where mineral transformations occur and the components are translocated, and then reorganized or synthesized into new compounds, the resulting accumulations are recognized as differences of morphology. Some argillic horizons and spodic horizons form this way as do some of the pans. Cementing materials of pans, like the petrocalcic, petrogypsic, placic and duripan are examples.
Fig. 3.1
"Bow structure" (Dudal and Eswaran 1989) in vertisols illustrating morphological microvariability
34
R.W. Arnold
BIOGEOCHEMICALPROCESSES Plants and animals interact with soils in some situations giving rise to mixing and chemical alteration of the materials. These processes can be thought of as biogeochemical. In and and semiarid areas the soils under certain shrubs or trees have characteristics that contrast with adjacent areas without the vegetation. It may show as more darkened aggregates or structures that differ due to increased moisture supply. In more humid areas there may occur distinct eluvial zones due to the chemical compounds released from trees or other types of vegetation. The mixing of soil is common where winds and storms uproot trees and micropits and mounds are produced with strikingly different morphologies. In cold, and many wet, environments there is an accumulation of biologic materials such as moss mounds or blanket bogs which modify or develop morphologies of their own. Zoogenic processes that mix and alter the soil materials are common in many parts of the world. In the tropics termites build mounds from the soil particles moved from the subsoil and substratum. Prairie dogs, marmots, moles and other animals build nests and burrows in soils altering the observable morphology. Crayfish build castle like tubes in some wet soils. Earthworms, beetles, cicada and many other types of fauna exert an influence on soils. Their excrement and gastric compounds significantly alter the chemistry on a microscale.
ANTHROPIC PROCESSES -AGRICULTURE Man seldom uses soils the way they occur naturally. Rather they become a medium for plant growth or an earthy construction material. Because of the almost limitless possibilities, there are many kinds of changes and alterations of soils that show up in morphology. Direct effects Tillage of land after the existing vegetation has been removed, mixes the surface materials together and tends to homogenize the morphology. This is common where the soils are disked or plowed, although changes of morphology will be noted even where the disturbance of the soil is less systematic. Due to mechanical alteration of the landscape such as hillside terraces and many rice paddies, there is a change in the moisture regime, which induces chemical and biological changes of the system, and in the resulting morphology. Fields that have been levelled and imgated have similar kinds of alterations. Land disturbance due to logging or other types of intermittent harvesting
Processes that affects soil morphology
35
also alter the morphology of soils. Gouging, slipping, and compacting of the surface may be noted within the soil.
Indirect effects As a consequence of overgrazing rangelands and of mismanaging cultivated lands, water and wind erosion are common. The processes are often accelerated relative to the same kinds of processes under natural conditions. The effects on morphology are about the same.
ANTHROPIC PROCESSES - INDUSTRY Some of the more striking disturbances of the soil landscape are those associated with industrial development and use of land. Cities cover the surface and bury pipes and other features under the surface. In the search for economic minerals the mining industries disturb soils to varying degrees depending on the procedures used to obtain the resources.
Direct effects The changes of soil morphology are associated with the kind and degree of disturbance of the natural soil. Soils under roads, streets, parking lots and other surface structures show disturbance and often have an increase of moisture and a decrease of biological activity. Park lands associated with cities often have been modified drastically to produce an externally pleasing view. Gardens and green spaces have anthropic soils of varying degrees of expression. Strip mining removes the soil and when the area is reclaimed the replacement usually is a mixture of new parent materials. The morphology often includes a random assortment of prior features. Dredging from stream channels that are deposited on land gives rise to new areas of artificial parent materials from which new soils can form. Industries that process agricultural products often dispose of their wastes on land. The wastes may vary widely in pH, have lots of organic debris, or contain caustic compounds which alter the chemistry and biology of the soils so treated. These additions are noted in changes of soil morphology. Archeology contains evidence of many kinds of uses and abuses of soils throughout man's history. Removal, burial, mixing or simply covering are all recorded in the morphology of the effected soils.
Indirect effects The burning of high sulfur content coal has contributed to an acidification of the precipitation falling in the down wind areas from major zones of such
36
R.W. Arnold
industries (see also Van Breemen, Chapter 11 of this volume). Other kinds of pollutants have been deposited on lands near smelting and refining industries. Some compounds deposited on soils, usually as aerosols, do not readily affect the existing morphology. For example, radiation nuclides are not easily detected in the gross features of morphology. They may eventually affect other processes that will result in visible changes.
TIME SCALE OF SOIL FEATURES A list of the estimated amount of time needed for processes to make a significant imprint on soil morphology has been proposed by Targulian (1989). The time scale is by powers of 10 and so those with characteristic times of 1-10 years and 10-100 years are of more immediate interest.
Characteristic time: 1
.
-
10 years
Root distribution of annual plants; Surface litter, Fluvic properties; Gleyic and stagnic properties; Slickensides; Permafrost presence; Inundic (temporary flooding); Salic features; Yermic features of fine earth (desert crust).
Characteristic time: 10
-
100 years
Tree root distribution; Humus content and thickness in topsoil; Salic properties; Calcareous properties; Sodic properties; Vertic properties; Histic epipedon, immature; Ochric epipedon; Gypsic horizon; Albic horizon in Spodosols; Natric horizon - immature; Spodic horizon - immature; Gilgai; Placic horizon - weak. It is interesting that many of the soil features that we associate with soil
Processes
affects soil morphology
that
37
morphology are thought to take more than 100 years and many others more than a 1000 years to develop their expression (Birkeland 1974). The concept of Birkeland (1974) is illustrated in Fig. 3.2. Profile DeveloPment . -
/I
A
/
/
10
6
0 0
1
3
2
4
5
Time (years as exp. of 10)
=
Oxisols
Fig. 3.2
Ultisols
0Mollisols
6
10) Spodosols
Schematic diagram to show time required by different kinds of soils to attain steady state (adaptedfrom Birkeland 1974)
Examples of the more than 100 years are high chroma colors, iron concretions, histic, mollic and umbric epipedons, calcic, natric, cambic, and spodic horizons, and desert pavement. In the latter group are argillic and oxic horizons, petrocalcic and petrogypsic horizons, duripans and fragipans. Carbonate accumulations in soils have been subject to dating and appreciating the errors associated with the dating procedure. Carbonate accumulation in soils is a good example of soil morphologic changes associated with time. Fig. 3.3 gives an example from New Mexico, USA, with data from Gile et al(1971). Even in a simple system such as the carbonate system, there are many processes (dissolutions, translocations, precipitations) acting with various intensities and in different parts of the profile, and these cumulatively contribute to the final profile morphology at any given time.
CONCLUDING REMARKS Pedology is based on a morphogenetic understanding of the pedons and profiles that exist in a landscape. It is not surprising therefore that most of the processes that are thought to affect the development of soils are also thought to be reflected in the observable features known as soil morphology. The grouping of processes into physical, chemical, biological, and anthropic is artificial in that most actions taking place in soils are complex combinations of processes.
38
R. W.Arnold
There has not been sufficient study of the time characteristics of processes that modify soil features and properties. If we wish to contribute to unravelling the earth's surface history, pedologists have much to offer and a long way to go.
SOIL CALCIUM CARBON
Om
0
10
Im
2m
> 1,000 < 2,500
> 10,000 < 15,000
>25,000
<50,000
APPROXIMATE SOIL AGE
Fig. 3.3
v
Evolution of depth function of carbonate profiles in an aridic environment (after Gile et al. 1971)
REFERENCES Birkeland, P.W. (1974). Pedology, weathering, and geomorphological research. Oxford University Press . New York. 285p. Dudal. R. and H. Eswaran (1989). Distribution. properties and classification of Vertisols. In Eds. L. P. Wilding, and R. Puentes. Vertisols: Their distribution, properties, classification, and management. Tech. Mono. 18.. Soil Management Support Services, SCS-USDA, Washington D.C., USA., I - 22. Fridland, V.M. (1972). Pattern of the soil cover. Translated from Russian, 1976. Keter Publ. House, Jerusalem. Gile, L.H., J.W. Hawley, and R.B. Grossman (1971). The identification, occurrence and genesis of soils in an arid region of Southern New Mexico. Training Bull. Soil Conservation Service, USDA. 177 pp. Targulian, V.O. (1989). Personal communication. IIASA, Laxenburg, Austria. influence on color patterns in soils.
39
Chapter 4
INFLUENCE OF CLIMATIC CHANGE ON SOIL MOISTURE REGIME, TEXTURE, STRUCTURE AND EROSION G.Y. Varallyay Research Institute for Soil Science and Agricultural Chemistry of the Hungarian Academy of Sciences Herman Orio UI 15, Budapest, Hungary
Human activities are currently leading to changes in the global environment challenging us to describe and understand the interactive physical, chemical and biological processes that regulate the entire earth system, the unique environment for life (IGBP 1988; Toward ...1988; Earth System Science 1988). Climate is a primary factor influencing these processes regarding direction, rate, spatial distribution, time variability and further consequences. In the last years several international scientific programmes were initiated for the registration, modelling and forecast of global climatic changes (Bolin et al. 1986; WCAP-4 1987; IGBP 1988; IGBP 1989). Due to natural processes and human activities such as increasing energy consumption, industrialization, agriculture, urban and rural development, considerable changes took place in the gas composition of the atmosphere. The C02 concentration, which was about 180200 ppm after the last glaciation and 270 ppm at early industrial times, increased to the present 350 ppm, a 50 % increase during the last 100-120 years. A doubling of the concentration of radiative gases in the atmosphere is expected to lead to a rise of global temperature of 3"C+/-1,5"C, which means a rapid, 0.103°C average temperature increase per decade. The spatial and temporal patterns of temperature increases will be heterogeneous and are expected to be largest in the northern mid-continent region of North America and Eurasia. The various Global Circulation Models (GCMs) predict a greater temperature increase in the polar regions than in the equatorial regions; and relatively higher temperature increase during summer, and lower during winter periods in both hemispheres. A temperature rise of 5 ° C for example, would imply northward displacements of isotherms in North America by 500 km, and such a change may occur within 100 years (Toward ... 1988). GCMs forecasting the temperature regime (WCAP-4 1987; WCRP-22 1988) are still rather uncertain, because in addition to solar radiation, the influences of circulation, changes in vapor content, cloudiness, albedo and surface roughness have to be evaluated more quantitatively. Warming up of mountain glaciers (resulting in changes in the waterflow dynamics, such as additional floodwaves), melting in the permafrost soil zone and the polar ice caps and increasing volume of the warmer oceans, will result in a rise of the eustatic sea-level. Its magnitude, assuming a 2-5°C temperature increase in the next 50-
40
C.Y. Varallyay
100 years, may vary between 0.20-1.40 m. This sea level rise is threatening low lying, man protected lands, including settlements, agricultural areas and extended sea shores with low slope angles. Another consequence will be the further extension of salt-affected areas under the direct temporal seawater inundation or indirect rise of the sea-level connected water table of saline or brackish groundwaters, as influences of the higher sea level (Szabolcs and Redly 1989). The change of sea-level will also raise the erosion basis in the affected catchment area which may result a non significant reduction of the water erosion potential. The forecast of precipitation characteristics (quantity of rain and snow; rainfall distribution pattern; intensity) is even more uncertain. Within the World Climate Programme, the World Meteorological Organization (WMO) takes serious efforts to improve the complex GCMs and quantify at least some basic plausible, hypothetical climate-change scenarios (Washington and Meehl 1983; Gleick 1986 and 1987; WCAP-4 1987; WCRP-5 1988). From the state-of-the-art of these GCMs it can be concluded that in the first period of the forecasted warming-up the average global precipitation will decrease, with high spatial variability and considerable regional redistribution. But later these tendencies will be counterbalanced by the increasing evaporation from water surfaces (first of all from the oceans), which leads to higher air humidity and probably more precipitation, again in a rather uneven spatial and time distribution, due to the changing circulation phenomena and vegetation pattern. The changing climate will result in considerable changes in the natural vegetation and in the land use practices (IGBP 1989; Bolin et al. 1986; Parry et al. 1990). The major vegetation belts will move northwards in the Northern and southwards in the Southern Hemisphere. Manabe and Wetherald (1 986) forecasted a rate of 25-200 km/IOO year northward "movement" of the coniferous forest zone on the North American continent. Vegetation in many cases cannot tolerate this "velocity", it will not be able to change its geographical distribution as fast as the changes in suitable habitat, there would be decades in length in the adjustment of ecological systems to rapidly changing climatic conditions. This leads to considerable changes in the species distribution, dynamics, diversity and production capacity of various ecosystems, consequently in their ecosystem functions. Similar changes are forecasted in agricultural (Blasing and Solomon 1984; Parry et al. 1990) and managed forest ecosystems, in spite of human activities, as land use, agrotechnics, amelioration, etc. The changes in the vegetation pattern result in feedback effects on climate, modifying albedo, surface roughness, micro-circulation processes, the heat and energy balance of the nearsurface atmosphere, the characteristics of both temperature and precipitation. Vegetation changes will considerably influence the field water cycle and soil formation proccsscs. The changes in tempcrature, precipitation and vegetation patterns (as soil forming Factors) will result in significant changes in soil conditions. These
41
Soil m i n u r e . [exlure, SIruclure and erosion
impacts and their relationships are summarized in Fig. 4.1.
J
I
--
S
E
T
wapo- Ironroiion Slaage ration spira -
Wilt-
*t
Fig. 4.1
tion
.
'7'
-
-
water
air
heat
r.
r.
r.
subsiancer.
- tl
The influence of climate change on soil moisture regime and soil erosion
It expresses in a selfexplanatory way why the quantitative evaluation of any climatic change on soil conditions is so difficult, and far from being satisfactory. The uncertainty of long-term global temperature and precipitation forecasts are combined here with the complex, integrated influences of changing vegetation and land use pattern and the changing hydrological cycle. Consequently, the global soil change prognosis can only be a rough, qualitative estimation and only
42
G.Y. Varallyay
allows to draw some rather general conclusions. Most of the forecasted soil changes are related to the field water balance (Vituki 1989; WCAP-4 1987). Its components are shown in Fig. 4.2. Because of the uncertainty of temperature and precipitation prognoses four plausible, potential climate change scenarios were selected to present their possible impacts on the elements of field water cycle and soil moisture regime. The moisture regime consequences of these four scenarios are also summarized in Fig. 4.2.
COLD
COLD DRY
Too cold for clay illuvis 1'ion
I
WET
-
frost heaving destroys any
texture differentiation Becoming too cold for clay illuviation
-
Maxirrwm clay illuviation
Minimum clay illuviation profiles high in sesauioxide c l a p
HOT DRY
'
semiarid
HOT WET
sub-
humid
humid
r -
~
~~~~
The effect of climate on texture differentiation in soils
little vegetation growth
I
Humification is slow so organic matter accumulates throughout profile, impeding podzolisation
I
Some humus
in soils as cooler conditions slow the destruction of humus
Humification fairly intense so lulvic acids orrn resultiiig in illuviation of iron and humus
in their profiles tion
profiles
HOT
WET
The role of organic matter in soil formation in response to climate
Fig. 4.2
The effect of four potential climate scenarios on texture differentiation and organic matter (from IGBP Report No. 5,1989)
Soil moisiwe, iexiwe, s i r u i w e and erosion
43
A general conclusion can be drawn, that an increase in the average annual precipitation will be followed by an increase in:
Surface runoff (R) in hilly lands with undulating surfaces and without permanent and dense vegetation, if the infiltration rate, permeability and water storage capacity of soil is limited; Infiltration (I) and water storage (S) within the soil, if they are not limited, such as in flat lands; Groundwater recharge (G) if the soil profile has good vertical drainage, permeability is not limited, especially in low lying areas; Evaporation (E) if infiltration is limited; Transpiration (T) in the case of well developed plant canopies. The decrease in precipitation results in adverse changes. Increasing temperature will result in: Increases in the potential E and T, if the plant canopy is not suffering from limited water supply due to climatic or soil induced drought, e.g. low precipitation or limited water storage capacity; Decreases in R, I, S and G, especially if it is accompanied by low precipitation; Decreases in the intensity (depth) of permafrost, it will modify the geographical boundaries of permafrost, opening possibilities for increasing water storage and water movement, biological activity and soil formation processes within the unfrozen part of the soil. Decreasing temperature will result adverse changes. These general influences are modified with the impact of vegetation characteristics (type, density, dynamics, species composition, biomass production, litter and root characteristics), and depend greatly on the type, intensity, spatial and temporal distribution of atmospheric precipitation. Man's influence is still more complex. Land use, cropping pattern, agrotechnics, amelioration (including water and wind erosion control, chemical reclamation, imgation and drainage) and other activities sometimes radically modify the field water balance and its components. In regimes under agricultural utilization the influence of global climatic change on the soil moisture regime is partly affected through these human actions (Parry et al. 1990). Within the World Climate Programme of the WMO a series of international projects are dealing with the assessment, modelling and forecast of the impact of climate variations and climatic changes on the field water cycle. In these studies proper spatial and time scaling has special significance,determining the minimum and optimum data sets required, the applicable models and approaches as well as the interpretability of outputs for global and regional, long-term, yearly, seasonal
44
G.Y. Varallyay
or month to months applications (Solomon et al. 1987; WCAP-4 1987; WCRP-5 1987- 1988 ; WCW-22 1988). Based on general circulation models (GCMs) and hypothetical scenarios several prognoses were elaborated for the assessment of the influence of forecasted climatic changes on soil moisture regime. Washington and Meehl (1983) prepared a map on the geographical distribution of soil moisture differences for the periods December to February and June to August (see Fig. 13.1). These differences were calculated by subtracting the soil moisture content at the present atmospheric C02 concentration from the moisture content at an assumed double atmosphere CO;?concentration (Washington and Meehl 1983; Manabe and Holloway 1975). The map indicates a substantial drying in spring and summer in the middle to upper mid latitudes and significant high latitude increase in precipitation, and hence runoff. In Fig. 4.3 the monthly distribution of these soil moisture differences are indicated, with longitudinally averaged values. The plot shows increases in continental soil moisture between 30 ON and 60 ON with greatest increases in the winter and spring seasons. Decreased soil moisture is found for most seasons in tropical latitudes and the Southern Hemisphere. Manabe and Holloway prepared a map in 1975 indicating the estimated global distribution of the annual mean rate of runoff. SOIL MOISTURE DIFFERENCES
60N
30N
EQ
305
60s
J
F
M
A
M
J
J
A
S
O
N
0
J
Fig. 4.3 Monthly distribution of soil moisture differences in em (longitudinal averages)for 2 x C02 - 1 x C02 (afer Washington and Meehll984)
In addition to the global scale approaches, numerous authors analyzed the regional hydrological impacts of global climatic change. Gleick (1986) draws the conclusion that in North California in the case of the forecasted climatic changes
Soil moirture, texture, struclure and erosion
45
the soil moisture reserve will decrease with 8-44%, resulting in serious drought induced yield reduction. There will be a certain decrease in summer runoff but a dramatic increase of winter runoff, leading to serious environmental and socioeconomic consequences. The majority of the investigations has pointed out that in most cases 10% decrease of rainfall results in more than 10% reduction in surface runoff. Since the water storage capacity of the soil is relatively constant, most part of the reduced precipitation can be stored in the soil. However, a 10% increase of precipitation may result in more than 10% increase in surface runoff, because of the relatively constant, and limited water storage capacity of the soil. Spatial and temporal variation of surface runoff can be as high as (or even higher) the variation in precipitation, because of the modifying effects of soil cover and canopy structure. Climate, vegetation and water regime influence soil properties. It depends on the character and intensity of climatic, vegetation and moisture regime changes and the changeability of soil characteristics, which are summarized in Table 4.1. As can be seen from this table, soil texture (particle size distribution) is a rather constant soil parameter: characteristic response time is lo3 years. Physical and chemical weathering are slow processes. Climatic changes result in only slight and slow process changes, and only moderately influence the "biological weathering". The influence is more expressed in the texture differentiation within a soil profile and in the organic matter turnover. The effects of the four basic potential climate scenarios on these processes are illustrated in Fig. 4.4. The influence of climatic change on soil structure (type, spatial arrangement, stability) is a more complex process with numerous direct and indirect impacts. The most important direct impact is the aggregate destructing role of raindrops, surface runoff and filtrating water. The rate of structure damage depends on the intensity of the destroying factor and the stability of soil aggregates against these actions. The indirect influences act through the vegetation pattern and land use practices. The consequence of vegetation changes on soil structure can be both favorable (tundra-forest; forest-grassland) and unfavorable (desertification, waterlogging, salinity-alkalinity). The same is true for land use. The impact of overgrazing, irrational land use, misguided agricultural utilization (cropping pattern, crop rotation) and improper agrotechnics (heavy machinery, over tillage, over imgation) is unfavorable, practically non reversible and hardly correctable. However, rational land use, proper agrotechnics and amelioration practices may help the maintenance or restoration of good soil structure. Pedogenic inertia will cause different time-lags and response rates for different soil types. Soil changes will be more rapid and profound in the younger or less weathered sediments of the glaciated or desert fringe region of the Northern Hemisphere and slower or less profound on the stable, continental shields of the equatorial region.
Table 4.1
P
Time changebility of various soil characteristics
Time changeability categories and their characteristic time 1 < 10-l yr
Soil parameter
bulk density; total porosity; compaction moisture content; infiltration rate; permeability; composition of soil air; nitrate content
aeration; heat regime
2
10-l-lOo y r
microbiota
microbial activity; human controlled plant nutrient regime
3
10°-lO' yr
total water cap. field capacity; hydraulic conductivity; pH; nutrient status; composition of soil solution wilting percentage; soil acidity; cation exchange capacity; exchangeable cations; ion composition of extracts
type of soil stmcture; annual roots biota; meso-fauna; litter, fluvic, gleyic. stagnic properties; slickensides
sulphuric h o r i m ; gelundic, inundic, salic, yennic phases (fine earth properties only)
4
101-102 yr
specific surface clay mineral association; organic matter content
soil biota tree roots; salic, calcareous, sodic, vertic properties
5
102-103 y r
primary mineral composition; chemical composition of mineral
tree roots; colour (yellowish/ reddish); iron concretions; soil depth; cracking; soft powdered lime; indurated sub-soil parent material; depth; abrupt textural change
histic (<20cm), ochric, gypsic. albic. and immature natric and spodic horizons (Podsols); gilgai, placic, sodic. takyric phase histic, mollic, umbric, calcic, albic, natric, cambic, spodic, and nitic horizons; plinthite, placic, yermic phases (stone surfaces) argic, oxic. petrocalcic, petrogypsic horizons; duripan, fragipan, skeletic, petroferric. lithic,
Part
6 > 103yr
texture; particle-size distribution; SP. hy; particle density
Properties and characteristics
o\
Horizons and phases
Regimes
moisture; natural fertility; salinityalkalinity; permafrost
P
?
Soil moisture, texture. structure and erosion
47
= precipitation = irrigation water = surface runoff (in and out) = filtration in the unsaturated
zone (in and out) = groundwater flow (in and out)
= rise of water table = lowering of the water table = infiltration into the soils = infiltration into the groundwater = storage within the soil (recharge) = filtration to the plant roots,
uptake by plants
T E
= transpiration
C
=
= evaporation
capillary transport from groundwater
Climate Factors
Cold, wet
Cold, wet
Hot. wet
Hot, wet
P
I I
D dD
I
D
R
1
D
G
i
1
I
I
d d
1
i i
D
(i) (i)
D D
E E
D D D D I I
S
E T F Gr GS
d
1
E i (i)
1 1
I
i and I = slight and great increase; d and D = slight and strong decrease; E = no change (equilibrium).
Fig.4.4 The components of the field water balance and soil moisture regime and the influence of four potential climatic scenarios on thesefactors
48
G . Y . Varallyay
If temperature increases, a warming up of the Northern Eurasian and North American permafrost plains with their loamy to silty sediments, all the shallow, imperfectly to poorly drained soils of the tundra and the Northern boreal forest biomass, will be radically changed by the melting of huge amounts of ice. The peat soil of the polar and boreal zones will shrink and slowly disappear due to the increased rates of decomposition of the organic matter. The podzolized soils of the tundra and boreal forests, which are derived from (peri)glacial sands and coarse crystalline rocks will turn into more acid and more leached variants. If precipitation increases, the heavy textured soils of present day tundra, boreal/ and humid temperate regions (some Luvisols, Podzoluvisols) will develop gleyic features in their topsoil, turning them into pseudogleyic/stagnic variants. As was described before, there are no linear relationships between mean annual precipitation, surface runoff and the rate of denudation/erosion. The rate, type and extension of soil erosion depends on climate (primarily the quantity and intensity of rainfall), relief, vegetation (type, continuity, density), and soil erodibility characteristics (La1 1988). The starting point of a comprehensive erosion-risk assessment can be the rate of surface runoff (Washington and Meehl 1983) or the sediment losses in the various major river basins (Lal et al. 1989). Some general conclusions on the possible influences of climatic changes on soil erosion can be summarized as follows:
Higher precipitation may cause an increasing rate of erosion; higher runoff, if it is not balanced by improved soil conservation impacts of stronger vegetation due to better water supply; Lower precipitation generally reduces the rate of erosion, but it can be counterbalanced by the less intensive soil conservation influence of poor vegetation due to the non adequate water supply to plants; this can be the consequence of increasing temperature as well; Lower precipitation (or higher temperature) may intensify wind erosion; Increasing temperature may reduce the erosion hazard moderating the permafrost influence (limited infiltration rate and water storage capacity of the soil), but may considerably increase the erosion-risk reducing the snowhain ratio in the cold regions and in high mountains. More detailed studies are required for the quantitative characterization of these relationships and for the improvement of prediction models expressing the influence of climatic changes on soil erosion processes.
REFERENCES Blasting, T.J. and A.M. Solomon (1984). Response of the North American corn belt to climatic warming. Progress in Biorneteorology, 3:311-321.
Soil moirture. texture, structure and erosion
49
Bolin. B, B.R. Doos, J. Jager and R.A. Warrick (Eds.) (1986). The greenhouse effect, climatic change, and ecosystems. SCOPE 29.. Wiley, New York. Earth System Science. A closer view. NASA, Washington, 1988. Gleick, P.H.( 1986). Methods for evaluating the regional hydrologic impacts of global climatic change. Journal of Hydrology 88:97-116. Gleick, P.H.( 1987). Regional hydrologic consequences of increases in atmospheric C 0 2 and other trace gases. Climate Change 10: 137-161. IGBP (1988). A study of global change. A plan for action. Global Change IGBP Report No.4. IGBP (1989). Effects of atmospheric and climate change on terrestrial ecosystems. Global Change ICBP Report No.5. Lal, R. (Ed.) (1988). Soil erosion research methods. Soil and Water Conservation Society. Ankeny, Iowa. Manabe, S a n d J.L. Holloway Jr.(1975). The seasonal variation of the hydrologic cycle as simulated by a global model of the atmosphere. J. Geophys. Res. 80. 1617-1649. Manabe, S.and R.T. Wetherald (1986). Reduction in summer soil wetness induced by an increase in atmospheric carbon dioxide. Science 232626-628. Parry, M.1.. T.R. Carter and N.T. Konijn (1990). The impact of climatic variations on agriculture. Vol. 1.: Assessment in cool temperate and cool regions. V01.2. Assessment in semi-arid regions. Reidel, Dordrecht. Solomon, S.I., M. Beran and W. Hogg (Eds.) (1987). The influence of climate change and climatic variability on the hydrologic regime and water resources. IASH Publication, No. 168. Szabolcs, 1 and M. Redly (1989). State and prospects of soil salinity in Europe. Agroktmia 6s Talajtan, 38: 537-558. Toward an understanding of global change (1988). National Academy Press, Washington, D.C.. Vituki (1989). The influence of climate variability and potential climatic change on the hydrologic cycle and water regime. A Hungarian Case Study. Budapest. Washington, W.M. and G.A. Meehl (1983). General circulation model experiments on the climatic effects due to a doubling and quadrupling of carbon dioxide concentration. J. Geophys. Res. 88:6600-6610. Washington, W.M. and G.A. Meehl (1984). Seasonal cycle experiments on climate sensitivity due to a doubling of C 0 2 with an atmospheric general circulation model coupled to a simple mixed-layer ocean model. J. Geophys. Res. 89:9475-9505. WCAP-4 (1987). Water resources and climatic change: sensitivity of water-resource systems to climate change and variability, World Climate Programme Applications, WMO, Norwich, U.K. WCRP-5 (1988). Concept of the global energy and water cycle experiment. World Climate Programme Research. WMO, Montreal 1987 and Pasadena 1988. WCRP-22 (1988). The global water runoff data project. Workshop on the global runoff data set and grid estimation. World Climate Programme Research. WCRP-22. WMO. Koblenz.
This Page Intentionally Left Blank
51
Chapter 5 RESILIENCE AGAINST CLIMATE CHANGE? Soil minerals, transformations and surface properties, Eh, pH Robert Brinkman AGLS, FAO, Via delle Terme di Caracalla. Roma 00100, I d y
ABSTRACT The main soil changes directly resulting from climate change would be in soil temperature regimes and soil hydrology. The minor soil temperature increases in the tropics to moderate increases in temperate and cold climates would modify organic matter dynamics, and may increase soil reduction in high latitudes where part of the permafrost would disappear. Increased variability and greater incidence of high-intensity rainfall events would entail higher leaching rates in most soils, with more runoff on sloping soils (and increased erosion and sedimentation). There would also be greater extents of periodic soil reduction in humid climates and salinization in semiarid or arid climates. The more rapid leaching, mainly in the tropics and in high latitudes, could accelerate the processes of hydrolysis and cheluviation; larger soil areas becoming periodically reduced could also become subject to ferrolysis. The influence of climate change on soil reaction would be generally minor. Soil changes would be least in a temperate climate. Soils with maximum resilience against degradation by these processes have a high infiltration rate and high structural stability, at least moderate cation exchange capacity and anion sorption, and moderate or good external drainage. Most of these properties are found in certain Nitisols as found near Nairobi, which have oxidic surfaces of the clay fraction. Soils least resilient against such changes have amorphous siliceous material coating the surfaces of the clay fraction, as in certain thixotropic Andosols or in quick clays in Canada or Scandinavia. The majority of the world's soils has intermediate properties. Management measures to increase soil resilience against the effects of climate change should increase vegetative cover and soil faunal activity, the same measures as against several other perturbations and stresses caused by human exploits. These human impacts are potentially more damaging than the direct effects of climate change on soils.
INTRODUCTION The main potential changes in soil forming factors (forcing variables) directly resulting from climate change would be in soil temperature regimes and soil hydrology, the latter because of shifts in rainfall zones as well as changes in potential evapotranspiration. Soil changes because of a potential rise in sea-level resulting from a net reduction in Antarctic ice cap volume and ocean warming are discussed in Brammer and Brinkman (1990). The biggest single qualitative change in soils expected as a result of these postulated forcing changes would be the poleward shift of the permafrost boundary, discussed in Chapter 17 by Goryachkin and Targulian (1990). Other
52
R . Brinkman
changes would be in degree rather than in kind. The changes in temperature but particularly in rainfall to be expected as a result of global warming are subject to major uncertainties, because different global circulation models do not lead to mutually consistent results (an example for Europe in Santer 1985), they are not yet adequately verified, and the interaction with changes in location and intensity of major ocean currents and resultant possible modifications in sea surface temperatures still is most uncertain, as well as the interaction with possible major changes in cloudiness and land cover and the resulting changes in albedo and actual evapotranspiration. Indirect effects of climate change on soils through sea-level rise, decrease or increase in vegetative cover, or a change in human influence on soils because of the changes in options for the farmer, for example, may each well be greater than direct effects of higher temperatures or greater rainfall variability and larger or smaller rainfall totals.
POSSIBLE CHANGES IN FORCING VARIABLES With these caveats, one could stipulate the following changes in forcing variables as likely to materialize sometime during the next century. Minor increases in soil temperatures in tropics and subtropics; moderate increases and extended periods in which soils are warm enough for microbial activity (warmer than about 5°C) in temperate and cold climates, parallel to the changes in air temperatures and vegetation zones summarized by Emanuel et al. (1985); Minor increases in potential evapotranspiration in the tropics to major increases in high latitudes caused both by temperature increase and by extension of the growing period; Increases in amount and in variability of rainfall in the tropics; possible decrease in rainfall in a band in the subtropics poleward of the present deserts; minor increases in amount and variability in temperate and cold regions; Increased frequency and severity of cyclonic storms in the present cyclone belts and some poleward widening of these belts because of increased sea-surface temperatures, giving rise to greater frequencies of high-intensity rainfall events.
EFFECTS ON SOILS Tropical areas In the tropics, increased intensities of rainfall events and increased rainfall
Soil resilience against climate change?
53
totals would increase leaching rates in well drained soils with adequate infiltration rates, and would cause temporary flooding or water saturation in many soils in level or depressional sites. They would also give rise to greater amounts and frequency of runoff on soils in sloping terrain, with sedimentation down slope and, worse, downstream. There would be increased chances of mass movement in the form of landslides or mudflows in certain soft sedimentary materials, discussed below. Soils most resilient against such changes would have adequate cation exchange capacity and anion sorption to minimize nutrient loss during leaching flows, and have a high structural stability and a strongly heterogeneous system of continuous macropores to maximize infiltration and rapid bypass flow through the soil during high intensity rainfall.
Subtropical areas In subtropical areas, decreasing rainfall totals with increasing intra and interannual variability could lead to less dry matter production and hence in due course, lower soil organic matter contents. Periodic leaching during high intensity rainfall with less standing vegetation could desalinize some soils in well drained sites, cause increased runoff in others, and lead to soil salinization in depressional sites or where the groundwater table is high. Soils most resilient against the effects of such increasing aridity and rainfall variability would have a high structural stability and a strongly heterogeneous system of continuous macropores, hence a rapid infiltration rate, as well as a large available water capacity -but less than the total for a high rainfall event expected with a frequency of about once per year-; and a deep groundwater table.
Temperate climates In temperate climates, minor increases in rainfall totals would be expected to be largely taken up by increased evapotranspiration of vegetations or crops at the expected higher temperatures, so that net hydrologic or chemical effects on the soils might be small. The negative effect on soil organic matter contents of a temperature rise might be almost compensated by the greater organic matter supply from vegetations or crops growing more vigorously because of the greater potential evapotranspiration and the higher transpiration efficiency in a high CO2 atmosphere. The temperate zone would thus be likely to have the smallest changes in soils, even in poorly buffered ones, directly caused by the effects of global warming. A minor and probably slow, but very visible, change could be a reddening of presently brown soils where increased periods with high summer temperatures would coincide with dry conditions, so that haematite would be stable over the presently dominant goethite. This mineralogical change might decrease the intensity and amount of phosphate fixation. An overview of such changes, with emphasis on temperate climate zones, is given by Buol et al.
54
R. Brinkman
(1990).
Boreal climates In boreal climates, the gradual disappearance of large extents of permafrost and the reduction of frost periods in extensive belts adjoining former permafrost are expected to improve the internal drainage of soils in vast areas, with probable increases in leaching rates. The appreciable increase in period when the soil temperature is high enough for microbial activity would lead to lower organic matter contents, probably not fully compensated by increased primary production through somewhat higher net photosynthesis and a longer growing period. Paradoxically, the extent of soils subject to periodic reduction could well increase in level areas, in spite of the greater leaching, because of increased periods when the soils are water saturated but also sufficiently warm for microbial activity. Soils most resilient against such effects, including the leaching of nutrients and periodic soil reduction, would have similar characteristics as the most resilient ones in the tropics: adequate cation exchange capacity and anion sorption to minimize nutrient loss during leaching flows, a high structural stability and a strongly heterogeneous system of continuous macropores to maximize rapid bypass flow during periods with excess meltwater.
PROCESSES IN SOILS Loss of salts and nutrient cations is the most rapid process of the various chemical and mineralogical changes that can take place with increased leaching. Normally, the clay mineral composition of a soil and, particularly, the mineralogy of the coarser fractions, would change little even over periods measured in centuries. Only in rare cases would the bulk of the clay fraction change appreciably over periods of a few years, as for example, when the largely amorphous clay fraction in a perennially moist, young volcanic soil under rain forest is suddenly subjected to periodic drying after deforestation. Then, halloysite (a member of the kaolinite group) may be formed very rapidly, with a consequent decrease in available water holding capacity and cation exchange capacity. Another relatively rapid change would be the dehydration of goethite to haematite mentioned above. However, changes in the surface properties of the clay fraction of soils may be much faster than changes in its bulk composition or crystal structure. Such changes in surface properties leave the bulk composition substantially unchanged but have a dominant influence on the physical and chemical soil characteristics. These changes, as well as the much slower changes in the bulk clay mineralogy, are brought about by a small number of transformation processes (Brinkman 1982, 1985), summarized below in order of decreasing areal importance:
Soil resilience against climate change?
55
Hydrolysis by water containing carbon dioxide, which removes silica and basic cations; Cheluviation, which dissolves and removes especially aluminium and iron by chelating organic acids; Ferrolysis, a cyclic process of clay transformation and dissolution influenced by alternating Fe reduction and oxidation, which decreases the cation exchange capacity by A1 interlayering in swelling clay minerals; Dissolution by strong mineral acids, which attack all clay minerals, producing acid A1 salts and amorphous silica; Clay transformation under neutral to strongly alkaline conditions (reverse weathering), which creates minerals such as montmorillonite, or in extreme cases palygorskite or analcime. Each of these processes changes the buffer system of the soils. Except for the last process, they decrease the cation exchange capacity of the clay fraction; hydrolysis and acid dissolution tend to increase phosphate sorption capacity and intensity (fixation). Hydrolysis and cheluviation, in soils that were already subject to these processes, may be accelerated by increased leaching rates. Ferrolysis may occur wherever soils are seasonally reduced and oxidized while there is leaching during the reduced phase. This combination of circumstances may, in a warmer world, occupy larger areas than at present, especially in high latitudes and in humid monsoon climates. Each of these three processes may thus impoverish or degrade the soils in which it occurs - however, even at somewhat faster rates than at present, the effects would be appreciable only after centuries, or millennia for hydrolysis. Dissolution by strong acids would be expected if and as low lying coastal areas with pyrite rich clays would dry out more deeply during the dry season than at present - but this is not likely in view of the expected gradual sea-level rise concurrent with, or closely following, global warming. Reverse weathering may take place in areas drying out during global warming, and would continue in soils in most of the presently arid areas. This process is of limited practical importance, however, since the overriding limitation on use of the soil resources in these areas is lack of moisture rather than any chemical or mineralogical problem. Most of these processes occur both in aerated, oxidized and in reduced conditions. Ferrolysis is limited to periodically reduced soils. Cheluviation occurs in somewhat different forms: with accumulation of iron(II1) compounds in the illuvial horizon in well drained soils, but with virtually complete elimination of iron in poorly drained conditions. Similarly, hydrolysis eventually leads to soils dominated by iron(II1) and some aluminium oxides in well drained conditions, but to soils dominated by A1 oxides (bauxite) in conditions of poor drainage. The five processes mentioned above normally influence the clay mineral
56
R. B r i n h n
composition over long periods only, and modify surfaces of the clay fraction over intermediate periods of time (generally, centuries). Direct human action, rather than climate change, may vastly accelerate some of these processes, however. In parts of Europe, for example, acid rain has extracted the basic cations from large areas of sandy soils and, through acid dissolution, replaced these by amounts of exchangeable and soluble aluminium ions that are toxic to trees and undergrowth of the natural or planted forests, as discussed by Van Breemen (1990). A less extensive example is the extremely rapid ferrolysis in soils seasonally inundated by water level fluctuations in a large reservoir, described by Amatekpor (1989).
SOME PROPERTIES OF CLAY SURFACES Soils with a naturally high structural stability are relatively rare, even though they happen to be prevalent near the workshop site, Nairobi. The clay fraction in such soils generally has oxidic surfaces: mainly iron(II1) and A1 oxides or hydroxides, while the bulk of the clay fraction may have different compositions. The oxidic surfaces could form from parent materials with moderate or high iron contents under long-continued hydrolysis by water (containing carbon dioxide). At the other extreme are soils with a very low structural stability, or with a severe hazard of failure under load or shock (socalled quick clays). The surfaces of the clay minerals in these soils are generally covered by amorphous, gel-like material with a high silica content (McKyes et al. 1974). Such material may have originated in earlier periods when the soils were strongly saline and, presumably, subject to processes of reverse weathering. Examples are the quick clays of the Champlain Sea sediments in Ontario and Quebec and of parts of Scandinavia. Such soils are most likely to generate high proportions of runoff and suspended sediment, but also most liable to mudflows once sloping sites are water saturated to appreciable depth. Some Andosols are thixotropic and have similarly low stability because of their similar composition, derived from volcanic materials (tuff). Most soils fall somewhere between these extremes. Vertisols, for example, have moderate or low structural stability, and clay surfaces that are mainly silica, but with generally small amounts of amorphous coating. In Planosols, if formed by ferrolysis, the clay fraction in the upper, eluvial horizons has been partly decomposed with a residue of amorphous silica, but the remaining smectite or illite has been interlayered with aluminium hydroxide polymers, which has decreased the swell-shrink potential and the cation exchange capacity of the clay fraction. Concurrently,part of the free iron oxides have been reduced and leached out. The net effect of these changes generally is a decrease in structural stability.
Soil resilience againsf climafe change?
57
RESLIENCE AGAINST PHYSICAL SOIL DEGRADATION As discussed, most soils do not have a high intrinsic resilience against physical soil degradation by, e.g., high intensity rainfall. Under natural conditions in humid climates, it is the complete soil cover near ground level combined with the perforating activity of the soil fauna that makes the soilvegetation system resilient against physical degradation. In the Rhine river plain in the Netherlands, for example, most of the originally calcareous alluvial soils have been decalcified within a millennium or so. Only in small areas on the highest levees of that age, continually under forest, soils have remained calcareous, even with lime pseudomycelia (filaments) indicative of less humid soil conditions, and full of vertical macropores produced by earthworms. In these soils, faunal activity is high because of the adequate litter supply; the resulting macropores remain open, protected against rain impact by litter and undergrowth; and heavy rain passes to the substratum through the macropores without leaching lime from most of the soil mass. Soil and crop -including pasture- management that maintains soil cover and organic matter supply to soil biota, while minimizing mechanical disturbance by heavy traffic, cultivation or excessive grazing intensity, makes soils most resilient against the effects of climate change and other extreme events. Such physical management may also help conserve plant nutrients through decreased leaching, as illustrated in the natural example above. This does not imply that a single management recipe would be applicable in different conditions. Certain termite species harming crop performance may necessitate a period without residues on the soil; or crop residues may be needed for feed or fuel. Such factors, and others, are taken into account in designing an optimum management strategy for any specific natural and cultural environment.
RESILIENCE AGAINST SOIL REDUCTION Soil reduction, which would limit land suitability for dryland crops, or strong reduction which would be liable to produce toxins even for wetland crops, may take place once the soil is water-saturated long enough for microbial action to exhaust the oxygen remaining in the soil when water-saturation started. Another necessary condition is the presence of sufficient readily decomposable organic matter as an energy source for the microbial activity. In most soils, during reduction the redox status is stabilized at an Eh about 100-200 mV near neutrality by the Fez+ - Fe(OH)3 equilibrium, except where the content of readily decomposable organic matter is very high or the content of free iron(II1) oxides very low. In such cases, negative Eh values may occur, and toxic hydrogen sulfide or low molecular organic compounds -including methane- may be formed. Resilience against soil reduction in practice depends on the drainage
58
R. B r i n h n
conditions, since most soils have sufficient organic matter for reduction to start within about a week after water saturation. Soils most resilient against reduction in conditions of increased rainfall variability and incidence of high intensity rainfall have similar properties as those resilient against other effects of climate change or other perturbations: high infiltration rate, high structural stability and a permanent heterogeneous system of tubular macropores, good external drainage.
SOIL REACTION Most soils would not be subject to rapid pH changes resulting from climate change; exceptions might be found in potential acid sulfate soils subject to increasingly long dry seasons. Even though most of such soils are clays with moderate or high cation exchange capacity, the amounts of acid liberated in such soils upon oxidation generally exceed this rapid buffering capacity. Therefore, pH values may temporarily reach 2.5 to 3.5 and a small part of the clay fraction may be decomposed as indicated under 'processes', above. This then buffers the pH generally between 3.5 and 4 in the long run.Depending on the efficiency with which the excess acid formed can be leached out, the period of extreme acidity and aluminium toxicity may last between less than a year and several decades. In calcareous soils, soil reaction may range between about 8.5 and 7 depending on the C02 partial pressure in the soil; this range is maintained against leaching of basic cations by the different soil processes as long as a few percent of finely distributed lime remain. Buffering in noncalcareous soils is less strong, but depends on the cation exchange capacity at soil pH. In soils with variable charge surfaces of the clay fraction, this decreases with acidification. It should be noted that the simple modelling of accelerated CaC03 leaching under a doubled atmospheric C02 concentration generally does not hold true. In most soils, the ongoing decomposition of organic matter maintains C02 concentrations in the soil air far above atmospheric concentration even now, and C02 in soil air determines CaC03 solubility. Leaching of lime is thus positively related to rate of organic matter decomposition, negatively to gas diffusion rate, and positively to amount of water percolating through the soil. In conditions where leaching is accelerated by climate change, it would be possible to find relatively rapid soil acidification after a long period with little apparent change, as has been the case -but more rapidly- in some soils in Europe that have been subject to acid rain for several decades. The soil might in fact be steadily depleted of basic cations, but a pH change may start, or may become more rapid, once certain buffering pools are nearly exhausted. Such non-linear and time-delayed effects have been discussed in the context of soil and water pollution by Stigliani (1988); they are also expected to occur in various ways at different times after increased temperatures and changed rainfall patterns will have been operative.
Soil resilience against
climate change?
59
CONCLUSION The soil changes expected as a direct result of climate change are generally minor or relatively well buffered by the mineral composition, the organic matter content or the structural stability of most soils. Major soil changes would be expected where permafrost would disappear. Climate induced changes such as major decreases in cover by vegetation or annual or perennial crops, could degrade structure and decrease porosity in most soils, as well as increase runoff and erosion on sloping sites and aggravate the concomitant sedimentation. Changes in options available to land users because of climate change may have similar major effects. In most cases, changes in soils by direct human action, whether intentional, on site or off site (and mostly unintended), are far greater than the direct climateinduced effects. Soil management measures designed to optimize the soil's sustained productive capacity would therefore be generally adequate to counteract degradation of agricultural land by climate change. Soils of nature areas, or other land with a low intensity of management such as seminatural forests used for extraction of wood and other products, are less readily protected against the effects of climate change but such soils, too, are threatened less by climate change than by human actions - off site, such as pollution by acid deposition, or on site, such as excessive nutrient extraction under very low-input agriculture. To armour the world's soils against the effects of climate change, or other extremes in external circumstances such as nutrient depletion or excess (pollution), or drought or high intensity rains, the best that land users could do, would be: Managing their soils to give them maximum physical resilience through a stable, heterogeneous pore system by maintaining a closed ground cover as much as possible; Using an integrated plant nutrient management system to balance the input and offtake of nutrients over a cropping cycle or over the years, while maintaining soil nutrient levels low enough to minimize losses and high enough to buffer occasional high demands. An analogous philosophy, at lower levels of external inputs, could be formulated for extensive grazing land and production forest, whether planted or managed natural forest. Human action and management has been emphasized in these conclusions because most of the world's land is used and, to different degrees, managed rather than under natural conditions.
60
R. B r i n h n
ACKNOWLEDGEMENTS Critical and constructive comments on a drafi of this paper by J.R. Benites, G.M. Higgins, W. Klohn, F. Nachtergaele and D. Norse are gratefully acknowledged.
REFERENCES Amatekpor, J.K. (1989). The effect of seasonal flooding on the clay mineralogy of a soil series in the Volta lake drawdown area, Ghana. Land degradation and rehabilitation 1: 89- 100. Brammer, H. and R. Brinkman (1990). Changes in soil resources in response to a gradually rising sea-level. (Chapter 12, this volume). Brinkman, R. (1982). Clay transformations: aspects of equilibrium and kinetics. Ch. 12, p 433-458 In: Bolt, G.H. (ed.) 1982. Soil Chemistry, B. Physicochemical models. Developments in Soil Science 5B, 2d ed. Elsevier, Amsterdam. Brinkman, R. (1985). Mineralogy and surface properties of the clay fraction affecting soil behavior and management. p 161-182. In: Woodhead, T. (ed.) 1985. Soil physics and rice. International Rice Research Institute, Los Banos, Philippines. vi + 430 p. ISBN 97 1- 104-146-4. Buol, S.W., P.A. Sanchez, J.M. Kimble and S.B. Weed (1990). Predicted impact of climatic warming on soil properties and use. In: American SOC.Agron. Special Publ. (in press). Emanuel, W.R., H.H. Shugart and M.P. Stevenson (1985). Climatic change and the broadscale distribution of terrestrial ecosystem complexes. Climatic change 7: 29-43. Goryachkin, S.V. and V.O. Targulian (1990). Climate induced changes of the boreal and subpolar soils (Chapter 17, this volume). McKyes, E., A. Sethi and R.N. Yong (1974). Amorphous coatings on particles of sensitive clay soils. Clays and clay minerals 22: 427-433. Santer, B. (1985). The use of general circulation models in climate impact analysis - a preliminary study of the impacts of a COz-induced climatic change on West European agriculture. Climatic change 7: 71-93. Stigliani, W.M. (1988). Changes in valued "capacities" of soils and sediments as indicators of nonlinear and time-delayed environmental effects. Environmental monitoring and assessment 10: 245-307. Van Breemen, N. (1990). Impact of anthropogenic atmospheric pollution on soils (Chapter 11, this volume).
61
Chapter 6
IMPACT OF CLIMATIC CHANGE ON SOIL ATTRIBUTES Influence on salinization and alkalinization I. Szabolcs Research Institute for Soil Science and Agricultural Chemistry of the Hungarian Academy of Sciences Herman Otto ut 15. Budapest, Hungary.
INTRODUClTON If the predicted climatic changes due to an increase of atmospheric carbon dioxide and other "greenhouse gases" develop, they will definitely alter the entire environment, including the soils of our globe. The expected changes will surpass human impact on the soil in the past and will affect many soil attributes, both directly and indirectly. Part of the changes will manifests itself in soil morphology, in physical, chemical, physico-chemical and biological soil properties and eventually in the fertility and productivity of soils. Several changes are unpredictable, while others are more or less precisely predictable. Soil salinization and alkalinization, influenced to a high degree by the greenhouse effect and subsequent climate changes, belong to such predictable processes (Szabolcs 1989).
PRESENT GLOBAL STATUS OF SOIL SALINITY AND ALKALINITY About 10% of the total surface of the continents is covered by different kinds of salt affected soils, mostly by saline and alkali types. No continent is free from such soils; even in the Antarctic vast areas of saline soils have been discovered (Szabolcs 1979). In recent years salt affected soils have been described in many countries where they were formerly unknown. The global area of these soils is increasing partly due to the extension of irrigation, overgrazing, etc. Salt affected soils occur in different climatic conditions though their appearance is most frequent in arid and semiarid areas, and in bottomlands and river valleys (Polynov 1956; Szabolcs 1987) (see Table 6.1).
POSSIBLE INFLUENCE OF PREDICTED CLIMATIC CHANGES The predicted climatic changes will increase the areas with saline and alkali soils mainly for two reasons:
62
I . Szobolcs
The direct effect of climate on salt movement and the salt balance of soils by increasing the temperature and aridity, on the water and salt balance of the watershed and on the extent of coastal salinization due to the rise of the sea level. The indirect effect on salt dynamics by changing the land use pattern e.g. by extending the area of imgation as a consequence of the climate changes. Table 6.1 Salt affected soils on the continents and subcontinents (without the Antarctic) (Sub) Continent
North America Mexico and Central America South America Afnca South Asia North and Central Asia South East Asia Australia Europe Total
Thousand ha
15 755 1 965 129 163 80 538 87 608 211 686 19 983 357 330 50 804 954 832
It is well known that the cause of salinization and alkalinization is the accumulation of salts in soil horizons and, as its consequence, the development of soil properties unfavorable for production. Salt movement and the salt balance of a given territory is closely related to the geochemistry, hydrology and climate of the given area. As a rule, soil salinity is common in dry areas and seldom occurs in well-drained humid places (Fersman 1934; Kovda 1947). Where the climate gets hotter and/or drier, accumulation of salts can be expected and their leaching from the soils will decrease. The consequence of this phenomenon will be soil salinization and/or alkalinization in many places where they were unknown before (Dregne 1976). If the predicted climate changes indeed occur, substantial alterations will take place in the practice of land use, agro-, sylvi- and horticulture, etc. One of the main consequences of climatic changes will be the necessity to increase irrigated agriculture because the higher temperatures and subsequent increased aridity will necessitate the use of additional water for agricultural production (Alekseevsky 1971).
Influence on salinization ond alkalinization
63
It is also well known that one of the most dangerous side effects of imgation is secondary salinization and alkalinization of soils, due to the alteration of the natural salt balance of the environment. UN data show that more than half of the existing irrigation systems of the world are affected by this adverse process and every year several million hectares of irrigated land are abandoned from agricultural production because of salinization and alkalinization. It is evident that in case of the extension of irrigation the hazard of soil salinization and alkalinization will also increase (Kovda 1980). While it is difficult to predict quantitatively many of the possible adverse soil processes which can develop under the influence of climatic changes, we do have the knowledge to elaborate more or less reliable predictions on the possible occurrence and extension of soil salinization and alkalinization in most of the areas that may be affected. It is a pity that studies for such predictions started quite reluctantly and between very narrow limits. However, it is necessary to characterize as soon as possible the anticipated processes, together with their geographical distribution, extension and necessary measures. In the past few years we developed some methods for the prediction of the extension of soil salinization and alkalinization as a consequence of predicted climatic changes for a large part of Europe, though this continent is comparatively less affected by salt than any of the others (Szabolcs 1974).
SCENARIOS FOR THE CHARACTERIZATION OF POTENTIAL SOIL SALINIZATION IN EUROPE For the studies of potential soil salinity in Europe we assumed that, due to different factors both natural and man-made, several major processes can result in salinization in different parts of Europe in the next half century. We selected the following three scenarios which represent the most important processes of potential salinization in different parts of our continent: Potential soil salinity caused by climatic changes (Scenario 1); Potential soil salinity caused by sea-level rise (Scenario 2); Potential soil salinity caused by extension of imgation (Scenario 3). Different regions of Europe were selected to test the above scenarios. For Scenario 1 the Mediterranean Region of Europe was selected where possible changes in climate, such as increasing temperature and decreasing precipitation may result in secondary salinization. For Scenario 2 certain coastal areas of Northwestern Europe were chosen where possible sea-level rise may cause remarkable soil salinity. For Scenario 3 regions with continental and semiarid climates were chosen where the major hazards of secondary salinization are present and imgation is extending.
64
I . Szabolcs
The location of the areas of the three Scenarios is shown in Fig. 6.1. .It can be seen that the three Scenarios cover a large part of Europe, with potential salinity caused by different direct or indirect man-made salt accumulation processes.
Fig. 6.1
Location of Scenarios I 2 and 3 in Europe. ~
SCENARIO 1 Potential soil salinity caused by climatic changes For Scenario 1 we selected the major Mediterranean areas of Europe (see Fig. 6.1), where at present salt affected soils occur; mainly in the Iberian Peninsula and only to a smaller extent in Southern France, Italy, Sicily, Sardinia and Corsica, as well as in the Dalmatian coast of the Balkan Peninsula. The total area of this Scenario is 1,979,959 km2, with 885,826 km2 land surface. In comparison with the salinity conditions of Scenario 3, the extension of salt affected soils in Scenario 1 is much lower, it is nearly half of the percentage. We assumed that the possible climatic changes, due to C 0 2 accumulation and other causes, will increase the average annual temperature of the area by about 1°C in the next 50-70 years. As a consequence, the aridity index will also increase, which creates progressive salinity in those marginal areas where at
65
Influence on salinizaiion and alkulinizarion
present salinity does not exist or can be found only in a latent form in soils or in waters. The following assumptions were taken into consideration: The currently irrigated area will not change substantially; No natural disasters or tectonical changes will occur, The present riverbeds and major hydrological conditions will remain unchanged. Fig. 6.2 shows the area of existing salinity, and potential salinity as predicted in Scenario 1.
Mediterranean Sea
Iexisting
salinity potential salinity
Fig 6.2
0
200
400
kin
Area of salt affected soils and potential salinity as a consequence of climatic changes (Scenario I )
The area of potential salinity is estimated at about 122,000 km2 (13.8% of the total land surface), which is twice as much as the area of existing salinity. The dry areas of the Iberian Peninsula (like Castilia, the Ebro Valley), southwestern France and also several areas in the Italian and Balkan Peninsulas are particularly exposed to potential salinity due to the increasing aridity. It must be noted that if the increase in the average annual temperature will surpass IoC, the increase in salinity will not be linear, but rather exponential compared to the information in Fig. 6.2. The salinity in those areas threatens very fertile agricultural lands and must be predicted in order to instigate necessary preventive measures in good time, before the climatic changes occur. It should be noted that the potential salinity in Scenario 1 equally endangers river valleys and estuaries (Ebro, Neretva, Rhone, Guadalquiver, Tajo, etc.) as plains and plateaus (Castilia, Aragonia, Umbria).
66
I . Szabolcs
SCENARIO 2 Potential soil salinity caused by sea-level elevation For Scenario 2 we selected a part of North-Westem Europe (Fig. 6.1) where the following two conditions are favorable for the study: Measurable extension of present salt affected soils; Good probability of sea-level rise due to global climatic changes in the next 50 years. The total area of Scenario 2 is 344,942 km2, with 226,393 km2 land surface, including the southeastern part of England, the Western part of the Netherlands, Belgium and the Northeastern part of France bordering, the North Sea, the Atlantic Ocean and the English Channel. In this Scenario mainly coastal saline soils with high sodium chloride content occur in several seashores and adjacent areas. The area of existing and potential salinity in Scenario 2 is illustrated in Fig. 6.3. It was assumed that - due to global climatic changes - the expected sea-level rise will be 1 cm per year on average, due to which, the effect of seawater on land will provoke further salinization. Changes in aridity and humidity were not considered. However, the following points were taken into consideration: Irrigation will not be extended; No natural disasters or tectonical changes will occur, Present riverbeds and major hydrological conditions remain the same.
Mexisting salinity a potential salinity
Fig. 6.3
0
.
200
400
km
Area of salt affected soils and potential salinity as a consequence of sea-level rise (Scenario 2 )
As shown in Fig. 6.3, the area of potential salinity caused by sea-level rise again substantially surpasses that of existing salinity. This is also valid for the two other scenarios, proving that all over Europe potential salinity constitutes a much greater hazard than existing salinity.
Influence on salinizarion and alkaliniealion
67
For Scenario 2 - as it was mentioned above - those areas were selected where salinity caused by seawater is essential (10,520 km2) and constitutes 4.65% of the total land surface. The area of potential salinity is 24,977 km2, which is 11.03%of the total land surface. As observed in Scenario 1, potential salinity in Scenario 2 is predicted in areas surrounding the existing salt affected soils. It should also be noted that if sea-level rise surpasses the values taken for granted in this study, the extension of potential salinity will again not be linear, but will increase in an even higher degree.
SCENARIO 3 Existing and potential soil salinity caused by irrigation For Scenario 3 we selected a part of Europe (Fig. 6.1) where, due to both climatic and economic conditions, irrigation has been practised for a long time. More than half of all imgated areas in Europe are situated in the area of Scenario 3. As can be seen from Fig. 6.4, salt affected soils are rather extensive. A large part of existing salt affected soils is situated in the vicinity of or even within irrigation systems . In nearly all countries included in Scenario 3 further extension of imgation has been envisaged. In countries with less precipitation, larger increases in irrigated areas have been assumed (e.g. the USSR, Bulgaria, Rumania) than in countries which do not suffer from aridity (Austria, Poland). At present, irrigated lands cover less than 20%, in most cases less than lo%, of the total agricultural land. The envisaged increase in irrigated areas is different in the various countries, but in no case will it be more than 100% for the next fifty-sixty years. In most countries concerned one of the main limiting factors of the extension of irrigation in the future is the shortage of good quality irrigation water. In Scenario 3, taking the above described circumstances into consideration, we assumed that up to the middle of the 21st century the irrigated area will at the most double and calculations were made accordingly. There are two main processes of salinization caused by irrigation: Salt accumulation in the soil originating from the salt in irrigation water, Salt accumulation in the soil originating from the salt in rising salty groundwater. As far as soil salinization from irrigation water is concerned there m acceptable regulations for quality control of imgation water. Subsequently, only a small part of present secondary salinization occurs as a result of the use of poor quality irrigation water. In the case of soil salinization from groundwater, the
68
1. Szobolcs
table of the salty groundwater (which sharply rises in most irrigated areas), unfortunately results in a large part of the current soil salinity and this process can be predicted as the most hazardous one also for the future. Considering climatic, physico-geographic, hydrological, agricultural and pedological aspects, we have compiled a map of Scenario 3, using the following assumptions: Until 2050 the present irrigated areas will increase, but not more than double; There will be no substantial changes in climate and no elementary disasters will occur; The present quality requirements for applied irrigation water will remain the same. Fig. 6.4 illustrates the extension of existing and potential soil salinity, following the above described conditions and assumptions. The area indicated on Fig. 6.4 is about 20% of the total land area of Europe, and represents areas intensively affected by both present and potential salinization, including large areas of the European part of the USSR, Rumania, Hungary, Austria, Yugoslavia, Czechoslovakia an Poland.
Iexisting
salinity
0
200
400
km
apotential salinity Fig. 6.4
Areas of salt affected soils and potential soil salinity as a consequence of extended irrigation (Scenario3)
Although Scenario 3 covers only about 20% of the land area of Europe, about 50% of all salt affected soils of the continent are found here. As shown in
Influence on solinizorion and olkolinizotion
69
Fig. 6.4, the area of potential soil salinity exceeds that of existing salinization by almost two times. The three scenarios, which have been discussed and demonstrated above, represent the main processes leading to the hazard of secondary salinization in Europe. However, they do not exhaust all the possibilities of this phenomenon, which may develop due to other factors like: changes in cropping pattern, intensive use of chemicals, changing farm management, etc. In order to predict the adverse salinization processes in more detail and more accurately, more appropriate methods for special surveys need to be developed. However, we already are in the possession of a number of such methods, as well as of methods for the prevention of salinization mainly in imgated agriculture.
REFERENCES Alekseevsky, E.E. (1971). Irrigation and drainage of the world. Izd. Kolos, Moscow (In Russian). Dregne, H.E. (1976). Soils of arid regions. Elsevier, Amsterdam. Fersman, A. E. (1934). Geochemistry. Leningrad (In Russian). Kovda, V.A. (1947). Origin and regime of salt affected soils. Vols.1. and 11. Izd Akad. Nauk SSSR. Moscow (In Russian). Kovda, V.A. (1980). Problems of combating salinization of irrigated soils. UNEP. Nairobi. Polynov, B.B. (1956). Selected papers. Izd. Akad. Nauk, SSSR. Moscow. (In Russian). Szabolcs, I. (1974). Salt affected soils in Europe. Research Institute for Soil Science and Agricultural Chemistry of the Hungarian Academy of Sciences. Martinus Nijhoff, the Hague Szabolcs, I. (1979). Review on research of salt affected soils. UNESCO, Paris. Szabolcs, I. (1987). The global problems of salt affected soils. Acta Agron. Budapest, Hung. 36. (1-2). 159-172. Szabolcs, I. (1989). Salt affected soils. CRC. Boca Raton, Florida, U.S.A.
This Page Intentionally Left Blank
71
CHAPTER 7
SOIL ORGANIC MATTER AND BIOLOGY IN RELATION TO CLIMATE CHANCE P.B. Tinker and P . Ineson Natural Environment Research Council Polaris House, North Star Avenue, Swindon SN2 IEU,United Kingdom
ABSTRACT Climate change, a.0. due to high atmospheric C 0 2 concentrations, is expected to produce higher temperatures, and largely unpredictable changes in rainfall. The temperature changes are likely to be least in the tropical zones and largest at high latitudes. Soil carbon changes are implicated both as part of the global carbon cycle, and as a main area of impact of the climatic changes. Increased C 0 2 may lead to more rapid plant growth when this is the currently limiting factor, but it seems unlikely to be so to an important extent except under intensive agriculture. Primary productivity and hence the input of plant residues may then become larger. In addition, higher concentrations of CO2 may change the quality of residues, and the water use efficiency of vegetation. The C 0 2 increase is unlikely to have any significant direct effect within soils, where the COz level is usually enhanced already. Primary productivity will respond to temperature change only where the latter is limiting, which is least likely in the tropics, but will have a major effect at high latitudes. Rainfall is much more likely to be important in the tropics. The most pervasive effects on soils will be where the whole character of the vegetation changes. Increased temperature will usually lead to more rapid breakdown of soil organic matter, and these processes have been modeled with promising results. The relatively small changes expected in the tropics appear unlikely to induce key changes, but the effects in high latitudes are expected to be much larger. Soils with high organic matter content due to low temperature and waterlogging could be strongly affected due to both temperature and hydrology changes. The combined effect of changes in organic matter input and in rates of organic material breakdown will set the final new equilibrium level of soil organic matter. The effects that are most difficult to predict are those resulting from the effects of rainfall changes, and the behavior of soils which are not in equilibrium with the climate. Changes in ecosystems and migration of vegetation zones are likely in some areas, and the speed with which this happens will be of critical importance. Soil flora and fauna, including symbiotic organisms, may be seriously affected by any such changes, as their migration rates are likely to be small.
INTRODUCTION The general outline of processes leading to climate change is now well known. However the climate predictions from the Global Circulation Models (GCMs) are still uncertain; in particular there are serious doubts about the effects
12
P . B . Tinker and P . lneson
of cloudiness, and the way in which land-atmosphere interactions should be included in the modeling. There is a reasonable agreement about global temperature changes, but the impacts of climate change will be felt more locally, and the current predictions of regional or national climate change are extremely uncertain. We are, therefore, still very much in the situation of working out the consequences of scenarios that may or may not happen. Here we start from the most recent predictions of the GCMs (Fig. 7.1) (Dickinson 1986; Trabalka 1985), which predicts temperature increases averaging 1.5-4.5OC. Such changes are likely to be accompanied by major alterations in winds, storms and rainfall. Despite the fact that these proceedings focus most on tropical areas, we cannot ignore the fact that all predictions agree that the temperature effects will be largest at the poles and least at the equator. The higher latitudes must therefore also be considered. We take as a reasonable range for a possible scenario that rainfall may change by about 20% at any point, though some changes may well be larger than this (Department of the Environment 1988). A point of general agreement is that the increase in temperature will be most marked in winter, so that cold winters, sea, ice, and snow cover are all likely to diminish.
SOIL AS PART OF THE CARBON CYCLE Soil organic matter occupies a particularly critical position in this topic. The carbon cycle is affected by climate change, and it has an important impact itself upon the changes via the COz in the atmosphere. The carbon in soil organic matter is a significant component of the earth's carbon reservoirs, with around 1 . 5 ~ 1 g0C~ (Solomon ~ et al. 1985) (Table 7.1). This is small compared to the carbon stored in the deep ocean (ca. 38x10 I8 g C and in the deep earth, but most of this is only in extremely slow equilibrium with the atmosphere, and the amount in the more rapidly equilibrating surface ocean layer is only some 0 . 6 ~ 1 0g' ~C. The turnover time of carbon in the soilvegetation system is variable, ranging from less than a year for annual vegetation, to thousands of years for organic matter in the deeper parts of soil profiles. The loss of carbon from both terrestrial biomass and soil organic matter has been considerable over the last 200 years. The calculation of the net release, over this whole period or currently per year, is very difficult due to the heterogeneity of the land surface, the variation in treatment or cover, and the uncertainty of the areas of particular ecosystems.
The current net release of C from soils and plants is estimated to be around l O I 5 g C yr-l, but the uncertainties in the data and the variable assumptions and
definitions used by different authors provide a very large margin of error.
73
Impact on soil organic matter and biology
(0
1
(b)
SURFACE AIR TEMPERATURE DIFFERENCES. DJF
SURFACE AIR TEMPERATURE DIFFERENCES, JJA 180
9
Fig. 7.1
mw
9oE
60 9oN
Predicted air temperature changes ("c)for a steady-state doubling of C02, showing winter (a) and summer (b) surface air temperature @-omBolin et al. 1986)
Estimates of the net annual flux between the atmosphere and terrestrial ecosystems (Houghton et al. 1985) range from -2.0 to 20.0 x 1015g C yr - l , a range that clearly indicates the gross uncertainties that persist in this area. The net release is about 20% of the annual flux of C02 from fossil fuel burning, of ca 5 x 1015g C y r l . At present it is estimated that some 40% of the total increase in
74
PB . Tinker and P. lnesan
atmospheric C02 to date has arisen from changes in the soil-vegetation system (Trabalka 1985). The total release since 1880 is estimated at 120 x 1015g C (Solomon et al. 1985). Table 7.1
Estimated values and uncertainties of key parameters in the global carbon cycle (from Solomon et al. 1985)
parameter
Net carbon flux from land biosphere since 1800 Gross annual terrestrial plant Co;!uptake Net primary production Annual tropical forest area conversion (1970-1980) Annual net carbon flux from land conversion (1970-1980) Contemporary soil carbon Contemporary biomass
Value 120 1015 c
Uncertainty range 90-180 x 1015 g c
120 1015 g c
90-120 1015 g c
0.004
45-62 1015 g c 0.003-0.006
1.3 1015 g c
0.0-2.6 1015 g c
1.5 x 10l8 g C 560 1015 g c
1.2-1.8x 10l8 g C 420-660 1015 g c
60
1015 g c
During the last century the major component of this was probably the loss of soil organic carbon from the cultivation of virgin grasslands, but now it is from the felling of forests, mainly in the tropics. A major complication occurs in peatlands, where the storage of C can be regarded as part of a soil reservoir, or as a separate compartment (subfossil C). Indeed, the whole question of carbon changes in wetland ecosystems is extremely interesting (Olson et al. 1985), because of the potential changes in coastal wetlands due to sea-level rise, as well as to temperature change and the production of CH4, another greenhouse gas, as well as C02 from such systems. With an annual net flux from soil and plants of some 1 x loi5g C and a total soil plus plant biomass carbon pool of around 2 x 10l8g C, there is plenty of scope for continuation or even acceleration of the present release of C02 from the terrestrial biosphere. The prediction of such changes is a matter of urgency.
THE ANALOGY APPROACH There appear to be two ways of approaching the problem of predicting the impact of climate change on soils, which we define as the 'analogy' and the 'process' methods. The analogy approach depends upon identifying a soil elsewhere which has developed under a climate that is very close to that expected at the site considered. There are several assumptions involved; the first is that climate is the dominant
75
Impact on soil organic matter and biology
factor in determining both vegetation and soil type. This may not be true if there are marked differences in parent material. Secondly, it assumes a fairly rapid reestablishment of equilibrium between soil and climate after a perturbation of l i ~ latter. In fact, different soil characteristics will alter at different rates (Fig 7.2), and the full development of a zonal soil in equilibrium with a climate takes thousands of years. Thirdly, this approach tends to assume that there is a relatively sharp change from one climate to another. In fact it is likely that this will be progressive, so that climate changes gradually, with soil and vegetation changes lagging behind to varying degrees. Nevertheless, with all these weaknesses, it is clearly a step forward to identify a soil that is the 'equilibrium' soil for a given climate, and then estimate the rate of progression towards this. NEAR EQUILIBRIUM
1 steady slate
-
Slowly adjusting features SOIL-P
Initial stale
Id 5
Fig 7.2
Id 6
Generalized diagram of the time taken to achieve steady state in soil properties after the initiation of soil development (from Walker and Graetz 1989)
The vegetation classification of Holdridge (1947 1964) has been used to predict the alteration in vegetation zones that would follow climate change (Bolin et al. 1979). This analysis is based solely upon a projected temperature change scenario, and any additional changes, such as in rainfall, would certainly cause larger effects. However, given these vegetation changes, their impact upon soils can be estimated at least qualitatively. One of the most interesting extensions of this system is the one of Zinke ct al. (1984), who plotted the carbon content in 3400 soil profiles on the Holdridge triangular diagram (Fig. 7.3), to produce a set of contours. In principle, at least, the equilibrium amounts of organic matter in a soil with a defined climate can be determined, though it is hard to believe that other factors, such as the texture of the soil, are not involved. The predicted temperature changes will produce shifts on this diagram that imply a smaller total pool of soil carbon; it would be possible
76
P.B. Tinker and P . Ineson
to sum this change over the terrestrial surface. However, the structure of the diagram shows clearly that the soil carbon store is much more sensitive to changes in rainfall than to changes in temperature.
2
Fig. 7.3
L 6 8
10
1 L 18
22
Soil carbon content (kg rn-2) plotted as contours on a Holdridge (1967)life-zone chart (source Zinke et al. 1984)
It should be remembered that very large areas contain soils that cannot be appropriately delimited simply by the biome in which they are contained. In particular, cultivation, management and cropping change both the profile and the carbon content in all farmed soils.
THE SOIL ORGANIC MATERIAL BALANCE - THE PROCESS APPROACH The components of the carbon cycle that determine soil organic matter levels (including subfossil remains such as peat) are: Net primary productivity of organic matter (NF'P); Carbon partition above and below ground; Breakdown rate of organic debris;
Impact on soil organic matter and biology
77
Formation of humus; Breakdown rate of humus. The soil organic matter level in any soil will settle at an equilibrium level characteristic of its permanent characteristics, the climate and the land use or vegetation cover. This concept is most clearly shown in the studies of the soil organic matter levels in the long-term experiments at Rothamsted (Johnston 1987), in which any change in treatment caused a corresponding change in the progression to a new equilibrium level. We draw a clear distinction between the decomposition of litter or debris, in which plant or animal structures can still be recognized, and soil organic matter, or humus. The latter is ultimately bound up with the mineral soil, has no recognizable structure, and has a fairly constant C/N ratio.
NET PRIMARY PRODUCTIVITY - INPUTS In general, increased atmospheric C 0 2 is expected to increase net productivity by more rapid photosynthesis. The present C02 concentration in the atmosphere, of ca 350 ppm, is suboptimal for plant growth in certain circumstances (Goudriaan and Ajtay 1979). The benefit to plants with the C4 photosynthetic system is very small or zero, because of their lack of photorespiration. In C3-plants there is undoubtedly the potential for increased growth, of perhaps 30%, if the C02concentration rises to 600 ppm. For a canopy or stand smaller values may be found than for single plants, because selfshading is less in the latter. In practice, growth of most crops or vegetation stands are limited by factors other than C02, e.g. drought, nutrient deficiency or disease. Even when such constraints are not overtly present, only a small fraction of even high yielding crops appear to reach the optimum yield set by radiation income (Tinker 1984). Any increase in crop yields is therefore uncertain, and major increases in net primary productivity in natural vegetation seem unlikely. However, if the growth constraint is due to water shortage, there may be a benefit from the more efficient use of water in plants whose stomatal behavior is controlled by CO2 concentration. The increased C02 concentration in the atmosphere should then increase the ratio of the flux of C02 into the leaf to the flux of water out of the leaf, unless the relatively lower use of water raises the leaf temperature sufficiently to counteract most of the potential effect. For a doubling of the C02 concentration, the water use per unit of photosynthesis could possible be halved (Squire and Unsworth 1988). Where water is limiting, this interaction of CO2 level and water supply is likely to be the most important effect in practice. There are other effects of increased atmospheric C 0 2 which may be important in some circumstances. The morphology of plants may be changed, in
78
P B . Tinker and P. Ineson
particular the root-shoot ratio. There is also evidence that the quality and composition of vegetation, such as the C/N ratio (Van Cleve et al. 1983), and also mot exudation and mycorrhiza formation (Norby et al. 1987) is altered. The problem in making any forecast concerning the climatic variables is the unpredictability of the latter. In general, temperatures will increase, and this will normally give a higher productivity, but the reverse is true if rainfall is not adequate. It is therefore impossible to predict net primary productivity at a designated spot in response to climate change, except to conclude that on average dry matter formation is likely to increase. However, two general situations can be distinguished. Firstly, in most areas the general vegetation type will remain constant. There is a strong relationship between IWP and net annual evapotranspiration; the latter measures temperature, but with an inbuilt correction for any deficiency of water (Lieth 1975; Whittaker 1975) (see Fig 7.4). This can be used as a guide to the behavior of NPP in areas where the general vegetation type remains constant. Moderate changes in NPP in such conditions are unlikely to alter soil organic matter greatly.
Annual actual evaporation (rnrn)
Fig. 7.4
Relationship between actual evapotranspiration and productivity (-), decomposition of deciduous (--- j and coniferous litter (.-.----j
Secondly, the most striking effects on soils and vegetation will certainly occur near the boundaries of the present vegetation zones, where the whole character of the vegetation, and the type and amount of litter, may change (Table 7.2). Thus, Emanuel et al. (1985) calculated that the temperature rise from a doubling of CO2 could result in a decrease in the global area of boreal forest and tundra of 37% and 32% respectively, this area changing into grassland or cultivated land. Soils in such conditions will then have quite new and different equilibrium states, with very different organic matter levels. The organic cycle
79
lmpocl on soil organic mailer and biology
differs significantly in grassland and forest soils, the organic matter in the former being both greater in quantity relative to the plant weight, and more deeply distributed than in forest soils. The quality of the litter, and the type of soil fauna, also differ considerably (Anderson, in press). The 'analogy' approach must be used in predicting these. Table 7.2
Plant biomass, litter input, soil organic matter and microbial biomass of major global vegetation types.
Subject
Tropical forest
Tem-
Boreal Savannah Tem-
perate
forest
perate grass
forest Area x 1012 m2 Plant biomass g C m-2 Litter input g C m-2 Soil c g m-2 Soil N g m-2 Microbial biomass C
12.0 12.5 14,000 9,000 368 250
Tundra
15
land 9
8 250 1,440 667 75 23,000 22,000 2,100 1,125 215 20
24.5 18,000 710 13,000 816 50
640 110
1,800 360 15,000 5,400 1,100 333 60 35
2
14
2.5
8.7
51
1
0.07
0.30
0.14
0.17
0.32
0.27
9,000
g m-2
Microbial biomass N g m-2
Microbial turnover yrs
Plant biomass and net primary productivity from Whittaker and Likens (1973). Litter from Ajtay et al. (1979). Soil C and N from Zinke et al. (1984). Microbal biomass from E.A. Paul ( p e n comm.)
THE BREAKDOWN OF SOIL ORGANIC MATTER The Qlo coefficient for the response of microbes to temperature change is normally around 2, but may vary from 1.6 to 3.2 (Schlesinger 1977; Singh and Gupta 1979). However, such figures are not necessarily a good indication of what happens when soil temperature is permanently increased in the field. Microbial species, moisture content and resource quality will change, and all affect microbial reproduction and respiration rates. A laboratory measurement of the effects of a simple short-term temperature shift in a soil is therefore of rather little use, and only gives a very general indication of direction and magnitude of change. The rate of decomposition of organic materials is sensitive to both temperature and moisture, and can therefore be related to the actual evapotranspiration rate in that environment. Hence, both this and the effect on NPP can be compared as in Fig. 7.4. Generally the effect of evapotranspiration on decomposition is greater than on NPP, hence as evapotranspiration increases,
80
P B . Tinker and P. Ineson
there is a tendency for the ratio of vegetation biomass to litter to increase. This is very clear in a study of biomass and soil carbon in various biomes (Fig. 7.5).
t
4
1
I
I
I
300
200
100
0
1
400
I
I
I
Tropical forest
I
1
I
100
Temperate I forest
I
200
I
I
300
400
I 500
Carbon pool (t ha-’)
Fig. 7.5
Comparison of carbon pools in plant biomass and soil organic matter (including litter) across different ecosystems Cfrom Anderson in press)
The breakdown of organic material and humus has been successfully modeled. The clearest situation is in agricultural soils, in which the same crop is grown regularly year after year, so that input, in quality and quantity, remains constant. Jenkinson et al. (1987) have modelled this situation for the Rothamsted long-term experiments. The structure of the model is illustrated in Fig. 7.6.
Fig. 7.6
Diagram showing the flow of carbon through the soil model of Jenkinson et al. (1987). DPM = decomposable plant material; RPM = resistant plant material; BlOZ and BlOA = zymogenous and autochthonous soil microbial biomass, respectively; HUM = hwnijied organic matter
lmpaci on soil organic mafler and biology
81
The ratio of biomass to humus produced is always the same in all soils, but differs between A and B pathways. The decay rate is (l-e-k), where k is the rate constant, that varies with temperature and moisture (Van Veen and Paul 1981). The effect of soil texture is included by amendments of the ratio of CO2 to biomass plus humus produced at each decomposition step. It was also found necessary to postulate a separate class of 'biologically inert' organic matter. This model has been applied to Nigeria Alfisols and Ultisols (Jenkinson and Ayanaba 1977), where it has successfully modelled the breakdown of 14-C labelled maize and rye grass residues, using the original parameters derived in Rothamsted. A set of runs of this model are illustrated in Fig. 7.7 (Jenkinson, personal communication), with a direct comparison of the effect of two different temperatures, of 9.22"C (normal mean at Rothamsted) and 19.22"C. The higher temperature leads to more rapid decomposition of residues, as expected. Broadly, the amount remaining at any time at 19.2"C is roughly equivalent to what is left at the lower temperature, but after about 3 times the time. Using the effects of change of biome, of changes in C02 and of actual evapotranspiration on NPP, and the existing models of organic decomposition, it should be possible to make prediction on the soil carbon levels following climate change. The uncertainties and errors will however be very large.
SOIL ORGANISMS The general soil heterotrophic population will behave in a way largely characterized by the speed of decomposition of soil organic matter. Temperature changes and precipitation changes will have effects as discussed above. Species of micro-organisms will change, but the likely speed of change, and $e acceptable climatic limits for any one group pf organisms, are such that no major unexpected effects are likely. The temperature optima for soil organisms are usually quite broad, and a shift of one or two degrees will normally make little difference. The most interesting problem in this topic will be in the wetlands of the high latitudes of the northern hemisphere. Temperature changes will be largest here, perhaps 6-10°C, with sharp changes in water relations. The question of whether the microflora will change in line with the conditions, or will lag well behind these, does not appear to have been addressed yet. The long-distance transport of bacteria on migrating birds, animals or man will normally be fairly rapid; this is less certain with the larger soil organisms. Previous comparisons of microbial communities across different biomes have been largely mycological, and show that species composition is clearly related to biome type, even for the same biome in different continents (Kjoller and Struwe 1982). Simple biomass comparisons by these authors suggest that upper levels for fungal biomass are an order of magnitude higher in temperate grasslands than in tundra, and that turnover rates are similarly higher in the more
P . B . Tinker and P . lneson
82
productive systems. Bacterial populations have been less well studied in this respect, yet Sundman (1970) has demonstrated that the functional attributes of the soil bacterial population are influenced by plant cover type.
0
12
24
36
48
60
I 72
I 84
I
96
I 108
I 120
Time in Months.
Fig 7.7
Runs of the Jenkinson et al. (1987) model, showing a direct comparison of the predicted rates of decomposition at 9.22 "C(-) and 19.22 "C(-----)
There is no direct evidence relating faunal biomass or productivity to organic matter quality; these seem to reflect responses to the environment directly. Thus, for example the distribution of termites appears to be limited by low minimum temperatures (Hams 1971), and the increases in temperature predicted as a consequence of climatic change can be expected to extend their range. Soil organisms with more specific characteristics may well be slower in adjusting to the new conditions. These are: Plant pathogens; Symbiotic organisms; Soil fauna. For plant pathogens and symbiotic organisms there is the additional complication that they are, to varying degrees, dependent upon specific types of vegetation. A vegetation zone shift caused by temperature/precipitation changes will only have reached full equilibrium when both the vegetation and the appropriate lower organisms have established themselves together.
Impact
on
soil organic mauer and biology
83
In terms of temperature only, the North-South displacement equivalent to the predicted climate change is up to a few hundred kms, depending upon latitude. With a C02 doubling time of 40-50 years, the rate of movement of organisms may need to be up to 10 km yr -1 to keep up with this shift. There is some information on the rates of spread of soil borne diseases and the recent rapid movement of the barley yellow mosaic virus across England provides an example of the time scales involved (Hill and Walpole 1989). In the seven years from the initial reports of the occurrence of this soil borne virus in England, the organism is now established over much of the country, and has spread at a rate which clearly exceeds those anticipated for climatic change. However, there are no precise analogies with the proposed scenario, in that all studies of organism dispersion have been within a broadly static environment, following an introduction of the organism. There has been no real analogy since the end of the last Ice Age, when the rate of spread and change was probably (though not certainly) slower. It will be essential for the appropriate symbionts to be available to the invading vegetation. Amongst the myconhizas, there is likely to be a decrease in the amounts of ectomycorrhizal fungi, as the large boreal forest belts move northwards towards the poles. Tropical and subtropical trees are almost wholly symbiotic with vesicular-arbuscular mycorrhizal fungi (Harley and Smith 1983). The rate of movement of mycorrhizal infection from root to root in grassland is slow, at ca 50 cm/yr (Sparling 1976). Fortunately these have low specificity, and some common species are found very widely distributed. Except in very high latitudes or in previously bare soils acceptable mycorrhizal symbionts are likely to be present, though the best equilibrium population of fungi may take a considerable period to establish. The N-fixing Rhizobia, actinomycetes or freeliving bacteria are more specific in their hosts, in particular the first two. There will certainly be situations in which there is a delay in introduction of these symbionts, and the spread of some leguminous plants may be delayed because of that. Heal and Ineson (1984) provide evidence to support the argument that the pattern of primary productivity in different biomes is reflected in the microflora and fauna, mainly through variation in resource quality, but also particularly for the fauna, through external climatic factors for the fauna. Again, it is when climatic change results in vegetation changes that associated shifts in below ground populations will become most pronounced. The time taken for the soil fauna to migrate into new areas is expected to be slower than for soil microorganism, and introductions of, for example, earthworms have shown rates of migration restricted to a few meters p e r year, in the absence of appropriate vectors (Van Rhee 1969). The net result is that the rate of change in climate may exceed the rate of colonization by the soil organisms usually associated with the new climate, and this lag will be greater for the larger components of the soil biota. The longer term picture is shown in Fig.7.2,
84
P.1)
T'inker and
P lneson
suggesting that equilibrium conditions will ultimately be reached; il is the predicted speed of change of climate which will result in imbalance.
LAND CAPABILITY EFFECTS Over most of the tropical and subtropical parts of the world, the restraint on plant production is water or fertility rather than temperature or radiation. As temperature will change relatively little with climatic change in these latitudes, the overwhelmingly important effect is likely to be in rainfall changes, which are the most difficult to predict. Any rise in temperature would be expected to produce more intense rain, even if the mean precipitation were unchanged or were lowered, with consequential effects on infiltration rates and erosion. In the higher latitudes in particular, there may be major changes in land use, with the movement of crop belts to the North (Warnck et al. 1986). If this involves the opening up of previously uncropped grassland or forest, the organic carbon contents of the soils will decrease, which has already contributed to global CO2 increase (Trabalka 1985). The greatest potential for catastrophe must be in impacts on the dryland areas of the world. These are soils which are already infertile, often with serious additional constraints of poor drainage, wind or water erosion, stoniness or shallowness and salinity (Dregne 1982). Their agriculture is already often barely viable, and dryland areas contain some of the most disastrous agroecological problems in the world (Steiner et al. 1988). Any decrease in mean precipitation, or any increase in rainfall irregularity, could be very damaging in many areas. Even now, the rainfall distribution is normally skewed, with many more years below the mean than above. The risks associated with this type of farming can easily become so large that the whole system loses its viability. The degradation that already occurs in many of these soils, will be accelerated by and associated with any decrease in organic matter (Lal 1987). The lowering of dry matter yields by lower rainfall will cause lower residue returns to the soil. A temperature increase will lead to some acceleration in organic matter decomposition, and therefore soil fertility (most probably nitrogen supply), water retention, infiltration and erosion resistance will gradually diminish. Changes of this type may be catastrophic in such areas, and the effects on soil organic matter will be strongly implicated.
CONCLUSIONS 1)
Climatic changes may have massive consequences for the total vegetationlsoil system , and it is quite unprofitable to consider soils alone in this context. The impacts will vary widely with the actual climatic change
lmpacl on soil organic matter and biology
85
experienced, and also with the sensitivity of the vegetation-soil system at any point. The changes are likely to result in a significant net release of carbon dioxide to the atmosphere, and it is important to be able to predict this as soon as possible. Hydrological and vegetation changes may be more important than simple temperature changes in this regard. Scientists dealing with the terrestrial part of the global system need to develop a more integrated approach, such as is particularly evident in the atmospheric sciences. Single site studies may be useful in defining principles, but are of little value on their own in developing a global picture. Soil scientists in particular need to emphasize the global nature of soil science, not so much in taxonomic pedology, but in process studies. There is a fundamental need to improve our methods of extending data from point measurements to global fluxes and processes. We need to look very carefully at our use of current classification systems, that are mostly based on semipermanent, slowly changing characteristics. We also need to improve our use of remote sensing technology. Where new technology or instrumentation an: relevant, we should have no hesitation in identifying this as an urgent need. This short review has shown quite clearly the weaknesses in soil ecology, and that additional effort and resources are needed if better predictions are to be made. The comment applies particularly to microbial ecology, in which the vast majority of soil microbes cannot even be identified, much less studied. It also applies to the ecology of soil fauna. More work should be done on soils that are at the boundaries of vegetation zones, and are thus sensitive to change; in general, soils not in equilibrium with their environment warrant more interest. Better methods of studying organic decomposition are needed including standard materials for testing rates (Anderson, in press). We must recognize the weaknesses of GCM predictions, but be ready to seize upon every advance as it appears.
ACKNOWLEDGEMENTS We thank Dr. I2 Jenkinson for the model runs in Fig. 7.7, and Dr. J. Anderson for allowing us to see a preprint of his paper in Functional Ecology.
P.B. Tinker and P. lneson
REFERENCES Ajtay, G.L, P. Ketner and P. Duvigneaud (1979). Terrestrial primary production and phytomass. In: The Global Carbon Cycle. Scope 13, Bolin B., E.T. Degens, S . Kempe and P. Ketner (Eds). The Global Carbon. Scope 13, 129-182. Wiley: New York. Anderson, J.M. (in press). The effects of climate change on decomposition processes in grassland and coniferous forest. Functional Ecology Bolin, B., E.T. Degens. S. Kempe and P. Ketner (Eds.) (1979). The global carbon cycle. SCOPE 13. Wiley: New York. Bolin, B., B.R. Doos, J. JSiger and R.A. Warrick (Eds.) (1986). The greenhouse effect, climate change and ecosystems. SCOPE 29. Wiley and Sons, Chichester, U.K.. Department of the Environment. (1988). Possible impacts of climate change on the natural environment in the United Kingdom. DOE: London. Dickinson. R.E. (1986). Evapotranspiration in global climate models. Paper presented atthe 26th COSPAR Meeting, July 2-10 1986, Toulouse, France. Dregne, H.E. (1982). Dryland soil resources. Science & Technology Agriculture Report. Washington: US Agency for International Development. Emanuel, W.R., Shugart, H.A. & Stevenson, M.P. (1985). Climatic change and the broadscale distribution in terrestrial ecosystem complexes. Climatic Change, 7, 29-43. Goudriaan, J. and G.L. Ajtay (1979). The possible effects of increased COz on photosynthesis. In: Bolin, B., E.T. Degens, S. Kempe and P. Ketner (Eds). The Global Cycle. Scope 13, 237-249. Wiley: New York. Harley, J.L. and S.E. Smith (1983). Mycorrhizal symbioses. Academic Press: London. Harris, W.V. (1971). Termites, their recognition and control. Tropical Agriculture Series, Longman: London. Heal, 0.W. and P. Ineson (1984). Carbon and energy flow in terrestrial ecosystems: relevance to microflora. In: M.J. Klug and C.A. Reddy (Eds). Current Perspectives in Microbial Ecology. 394-404. American Society for Microbiology: Washington. Hill. S.A. and B.J. Walpole (1989). National and local spread of barley yellow mosaic virus in the United Kingdom. Bulletin OEPPEPPO Bulletin, 19, 555-562. Holdridge, L.R.( 1947). Determination of world plant formations from simple climatic data. Science, 105, 367-368. Holdridge, L.R. (1964). Life Zone Ecology. Tropical Science Center: Costa Rica. Houghton, R.A., W.H. Schlesinger, S. Brown and J.F. Richards (1985). Carbon dioxide exchange between the atmosphere and terrestrial ecosystems. In: Trabalka, J.R. (Ed). Atmospheric carbondioxide and the global carbon cycle. (DOEER-0239). , 113-140. U S . Department of Energy: Washington. Jenkinson, D.S. and A. Ayanaba (1977). Decomposition of carbon-14 labelled plant material under tropical conditions. Soil Science Society of America Journal, 41, 9 12-915. Jenkinson , D.S., P.B.S. Hart, J.H. Rayner and L.C. Parry (1987). Modelling and turnover of organic matter in long-term experiments at Rothamsted. INTECOL Bulletin, 15, 1-8. Johnston, A.E.(1987). Effects of soil organic matter on yields of crops in long-term experiments at Rothamsted and Woburn INTECOL Bulletin, 15, 9-16. Kjoller, A. and S. Struwe (1982). Microfungi in ecosystems: fungal occurrence and activity in litter and soil. Oikos, 39, 389-422. Lal, R. (1987). Managing the soils of the sub-Saharan Africa. Science, 236, 1069-1076. Lieth, H. (1975). Modeling the primary productivity of the world. In: Lieth. H. and R.H.
Impaci on soil organic matier and biology
87
Whittaker (Eds). Primary Productivity of the Biosphere, 237-263. New York: Springer Verlag. Norby, R.J., E.G. O'Neill, W.G. Hood and R.J. Luxmore (1987). Carbon allocation, root exudation and mycorrhizal colonization of Pinus echinata seedlings under carbon dioxide enrichment. Tree Physiology 3, 203-210. Olson, J.S., R.M. Carrels, R.A. Berner, T.V. Armentano, M.I. Dyer and D.M. Yaalon (1985). The natural carbon cycle. In: Trabalka, J.R. (Ed). Atmospheric carbondioxide and the global carbon cycle. (DOE-0239). 175-213. U.S. Department of Energy: Washington. Schlesinger, W.H. (1977). Carbon balance in terrestrial detritus. Annual review of Ecological Systematics, 8. 51-81. Singh, J.S. and S.R. Gupta (1979). Plant decomposition and soil respiration in terrestrial ecosystems. Botanical review, 43,449-528. Solomon, A.M., J.R. Trabalka, D.E. Reichle and L.D. Voorhees (1985). The global cycle of carbon. In: Trabalba, J.R. (Ed). Atmospheric carbondioxide and the global carbon cycle.(DOE/ER-0239), 1-13. U.S. Department of energy: Washington. Sparling, G.P. (1976). Effects of vesicular-arbuscular mycorrhizas on pennine grassland vegetation. PhD Thesis, University of Leeds. Squire, G.R. M.H. Unsworth (1988). Effect of C 0 2 and climate change on agriculture. Contract report to U.K. Department of the Environment. Steiner, J.L., J.S. Day R.T. Papendick, R.E. Mayer and A.R. Bertrand (1988). Improving and sustaining productivity in dryland regions of developed countries. Advances in Soil Science, 8, 79-150. Sundman, V. (1970). Four bacterial soil populations characterized and compared by a factor analytical method. Canadian Journal of Microbiology, 16. 455-464. Tinker, P.B. (1984). Site specific yield potentials in relation to fertilizer use. 18th Coll. Int. Potash Institute, Bern 193-208. Trabalka, J.R. (Ed) (1985). Atmospheric carbondioxide and the global carbon cycle. (DOEER-239). US. Department of Energy: Washington. Van Veen, J.A. and E.A. Paul (1981). Organic carbon dynamics in grassland soils. 1. Background information and computer simulation. Canadian Journal of Soil Science, 61. 185-201. Van Cleve, K.. L. Oliver, R. Schlentner, L.A. Viereck and C.T. Dyrness (1983). Productivity and nutrient cycling in taiga forest systems. Canadian Journal of Forest Research, 13, 747-766. Walker, B.H. and R.D. Graetz (1989). Effects of atmospheric and climatic change on terrestrial ecosystems. IGBP Report No 5. Warrick, R.A.. R.M. Gifford and M.I. Parry (1986). Assessing the response of food crops to the direct effects of increased C 0 2 and climatic change. In: B. Bolin, B.R. Doos, I. JLger and R.A. Warwick. The greenhouse effect, climatic change, and ecosystems, SCOPE 29, 393-473. Wiley: New York. Whittaker, R.H. (1975) Communities and ecosystems. MacMillan: New York. Whittaker R.H. and G.E. Likens (1973). The primary production of the biosphere. Human Ecology, 1, 299-369. Zinke, P.J.. Strangenberger, A.G., Post, W.M., Emanuel. W.R. & Olson J.S. (1984). Worldwide organic soil carbon and nitrogen data. (ORNLnM-8857). Tennessee: Oak Ridge National Laboratory.
This Page Intentionally Left Blank
89
Chapter 8
INFLUENCE OF CLIMATIC CHANGE ON DEVELOPMENT OF PROBLEM SOILS, ESPECIALLY IN THE ALLUVIAL DOMAINS Walter R . Fisher Universitat Hohenheim Institut fur Bodenkunde und
Standortslehre
D-7000 Stuttgart 70, Fed. Rep. Germany
ABSTRACT Among the factors which change during a global climatic change, mean annual precipitation and increased temperature are supposed to have the strongest influence on the ecological properties of alluvial soils. Whereas increased precipitation would enforce leaching, or moderate salt problems, increased biological activity due to higher lernperature would accelerate microbial turnover in the soil and enlarge Lones of anaerobiosis.
THE DEFINITION OF "PROBLEM SOILS" - GENERAL CONSIDERATIONS The response of soil ecology on any changes in environmental conditions depends on the buffering ability of a given soil ecosystem. This is shown in Fig. 8.1, where a non linear relationship between the strength of the influencing factor and the respective soil properties is supposed. In this diagram, the buffering activity of the soil is documented by the inverse slope of the response curve. If one external factor is less strong compared to the buffering capacity of the soil system, it would be buffered, and the corresponding changes of the soil properties would be very low. At higher levels of external loads, a threshold may be exceeded, and irreversible changes of the ecological properties may occur. Depending on the actual land use of a given agricultural soil, problems can occur if such soil is in a condition where it is sensitive to external influences, and can be changed irreversibly. With respect to alluvial soils, such dominant factors may be acidification by rain or by soil organisms, the height of the permanent water table or the reduced horizon. In general, the water regime has a strong influence on the ecological soil properties, because nearly all nutrients or toxicants are transported by the soil water. Three environmental factors which are supposed to change during the global climatic change may influence the ecological properties of a given soil ecosystem: Composition of the atmosphere;
90
W.R. Fisher
Mean annual precipitation, and position of the phreatic groundwater table; Temperature of the atmosphere just above the soil surface. Influence of external factors on soil ecology
intensity of external factors (e.9. acid rain, temperature)
F i g . 8.1
Schemuiic drujl of the sensitivity of a soil 10 external factors
COMPOSITION OF THE ATMOSPHERE The CO, content of the atmosphere is about 0.03% by volume, whereas in biologically active topsoils 1 - 10% of CO, are typical concentrations. So the difference between the CO, concentration of the soil air and atmospheric air is relatively high, even if the atmospheric CO, concentration is increased by a factor of two. By consequence, the CO, concentration gradient is less influenced by the global C02 increase, and a biologically active soil is acting as a C 0 2 source. For the genera1 relations between CO, concentrations, light intensity, and photosynthetic efficiency, see Chapter 1 , Fig. 1.5 (Scharpenseel). This is not the case for soils which are very low in biological activity, or for rock surfaces not covered by soils. In these cases, especially for carbonaceous rocks, the weathering rate is increased due to the lower pH of the rain water. Among the consequences of this change is the accelerated release of nutrients from primary minerals as well as the intensified leaching by percolating water. With respect to acid soils, a marked impact of higher CO;? content on soil forming processes is not assumed. This is because the dissociation constant of the first step of H2C03:(into H+ + HC03‘) is very low (pk = 6.35) and should therefore not influence soils with pH values below 4. The two other contaminants which play a role in the global “greenhouse effect”,
Influence of climalic change on soils in the alluvial domains
91
methane (CHJ and nitrous oxide (N,O) are not assumed to have an influence on the more important soil processes (for NH,, see Addendum to this Chapter, by Takai).
MEAN ANNUAL PRECIPITATION After the most reliable models, the mean annual precipitation will increase globally. For estimating the appropriate influences upon a given soil, we have to distinguish between two hydrological situations (Fig. 8.2):
I PRECIPITATION
increasing
decreasing
A. groundwater level high:
-
zone of accumulation
6.groundwater level low: infiltration rate _ _
I1 - 1
Zoneofleaching (nutrients)
Fig. 8.2
a)
n
Soil ecological consequences due to groundwaterjluctuatiom
High groundwater level; with increasing precipitation, the level of the permanent groundwater table should increase slowly. As a consequence, the position of the capillary rise should also move upward in the soil. Problems
92
W.R. Fisher
could arise, if at high actual evapotranspiration,zones of accumulation (salt, iron oxides) would form within the rhizosphere of agricultural plants. This is shown in Table 8.1, where the Fe(II)/Fe(III) redox system is used as an example. Due to the higher mobility of the reduced phase, Fe can move along the redox gradient, and Fe oxide is accumulated in the contact zone. In contrast, no acidification occurs if the Fe(I1) comes from hydrolyzed species such as Fe(OH)2,because protons are released only if the degree of hydrolysis for Fe is changed during oxidation. As a result of the oxidation of Fe sulfides, acids (sulfuric acid) are formed. Prolon release during Fe and S oxidation
Table 8.I Reduced species Fe ion
Oxidized species Fe(ll1) oxide
Fe( 11) compound Fe( 11) sulfide
Reaction
Effect
4 Fe2++ 6H2O + 0 2 = 4 FeOOH + 8 H+
acidification due to Fe hydrolysis
Fe(l11) oxide
4 Fe (OH):! + 0 2 = 4 FeOOH + 2 H20
none
Fe(ll1) oxide sulfuric acid
4 FeS + 9 0 2 + 6 H 2 0 = 4 FeOOH + 4 H2SO4
acidification due to sulfur oxidation
Low groundwater level; in this case, increasing precipitation will enhance the rate of percolation and can therefore step up nutrient leaching or moderate salt problems, depending on whether it is a highly permeable or salt affected soil, respectively.
b)
Table 8.2
influence of changing soil humidity on pedological problems Humidity
Problem ~~
Increasing
Decreasing
moderated if water balance goes to positive increased in permeable soils increased in permeable soils (low buffering capacity) increased
increased
~
high salt concentration nutrient deficiency acidification reducing zones in topsoil
not influenced or increased due to low availability not affected moderated
93
Influence of climalic change on soils in [he alluvial domains
Table 8.2 summarizes briefly the expected changes in soil ecology due to changing annual precipitation. Not included in these considerations are soils which are going to be influenced directly by the (increasing) sea level, though this might be of great importance for large coastal areas.
TEMPERATURE OF THE ATMOSPHERE Possibly the most important impact on soil ecology (from a global point of view) would come from the increasing temperature of the air just above the soil surface. This may have three reasons: a)
Direct influence on chemical equilibria. Table 8.3 shows the temperature function of the equilibrium constant of a chemical reaction following the Van 't Hoff equation. As is demonstrated by the given example the variation of dk/dT with changing temperature is more or less negligible within the regarded temperature range.
Table 8.3
Theoretical influence of temperature on chemical equilibrium constants
Influence of temperature on chemical equilibria AGO = -RT In k [AGO = GIBBS energy of reaction, k = equilibrium constant, AH = free enthalpy of reaction] d(AG/T3 - -&
Crr T2 d(In k) - AHo R T ~
(Van't Hoff)
Example:
Assumption:
Increasing temperature favors endothermic reaction (for AHo/R = 10K) T[KI WdT 293 1.000116 298 1.000113
real effect negligible, because most reactions are not in a sensitive range Possible exceptions: formation of secondary minerals (oxides, clay minerals)
W . R . Fisher
The only exception from this model would occur if an equilibrium is extremely sensitive to temperature variations, e.g. the goethitehematite system in model experiments. However, under pedogenetic conditions, this equilibrium is superimposed by the strong influence of soil organic matter, and the direct influence of temperature variations is supposed to be very low. Increase of the reaction rate of most physical and chemical processes. Due to the relatively slowly moving water in the pores of a soil, this is supposingly of less importance (except some events of very high precipitation) because the chemical equilibrium between the solid surfaces and the soil solution is obtained in at least several hours. Table 8.4 indicates three different cases of initially varying soil humidity. Table 8.4
Ecological consequences of the acceleration of physical processes
Increased temperature makes most physical processes faster: evaporation - transpiration water transport in soil surface transport of dissolved salts to surface "dry case" "moist case" "wet case" increasing salinity increasingacidification increasing bioactivity decreasing biological first phase: increasing importance of activity increasing biomass reduced (anoxic) zones production (increasing osmotic potential) c)
second phase: toxic effects (of AI)
mot zone influenced
by anoxia
Increase of' biological activity. This is expected to be most sensitive to temperature variations; it possibly has the strongest influence on soil ecology. In the following paragraphs some examples are given on what the expected influence of increased biological activity might be.
The total biological activity, as documented by CO, production, depends strongly on temperature. As shown in Fig. 8.3, the slope of the cuwe increases with increasing temperature, so the expected influence of elevated temperature would be higher in the tropics than in moderate climates. This is also reflected by the electrolyte concentration of the pore water. As illustrated in Fig. 8.4, the electrical conductivity of the pore solution was increased dramatically during incubation at high temperature, whereas at 5°C it did not. By consequence, at elevated temperature, all chemical exchange
Influence of climairc change on soils in the alluvhl domains
95
processes are accelerated, and are continued to a higher extent. Biological
activity
(soil respiration)
5 t
0 0
10
20
30
-_
40
temperature PC
Fig. 8.3
Soil respiration as a function of temperature (after Anderson and Domsch 1986)
Electrical conductivity of soil solutions
0
I
I
I
2
4
6
8
period submerged 1 days
Fig. 8.4
Electrical conductivity of the soil solution at various temperatures (after Yu 198.5)
Under waterlogged conditions, the actual redox potential depends on two counteracting processes.
96
W . R . Fisher
Sulfide formation
Fig. 8.5
Sufide formation during incubation at various temperatures for two soils (after Uuo and Mikkelsen 1981)
The oxygen consumption by organisms, and the oxygen supply from the atmosphere, or from plant roots with aerenchyms. Whereas the latter process is less influenced by temperature variations, the increased biological activity t higher temperatures leads to a faster decrease of Eh, even at slightly permeable soils. This is shown in Fig. 8.5: in that case the lower diffusion rate of oxygen in the clay soil supports a rapid sulfate reduction at higher biological activity. Therefore the border line of the reduced horizon of a waterlogged or subaquatic soil is elevated at higher temperature (Fig. 8.6). low temDerature
n
l
I = Q)
8-
oxidized horizon
reduced horizon
Fig. 8.6
Schematic drafl of the height of the reduced horizon in a waterlogged soil due to varying temperature
InJluence
of climatic change on soils in the alluvial domains
97
Even if the permanent groundwater level would not be affected, a rise of the reduced layer could occur, and therefore, the zone of strong acidification could expand to the rhizosphere. This is shown in Fig. 8.7, which demonstrates that, at higher temperature, the zone of major Eh changes is both narrower and higher in the soil profile. So, the expected acidification concentrates on a narrower soil compartment and could therefore influence the soil ecology stronger.
-
with increasing temperature increase:
--
rate of biomass mineralization rate of oxygen consumption elevated zone of reduction
elevated zone of sutfide oxidation
redox
tential
predicted
5 a
a, -0
capillary
N
s (D
0, P, 9.
a =f? 0
%. 0 3
I Fig. 8.7
groundwater level
Schematic draft of the redox profile combined with soil acidification at increasing temperature (high groundwater level)
At permanently low redox potentials, the microbial metabolism is quite different compared to soils with air filled pores, and different compounds are formed. One of these compounds is acetaldehyde, the rate of formation depending strongly on temperature (Fig. 8.8). This, besides a direct impact on bacterial growth and selection, can influence the composition of the soil microflora community, due to toxic or competitive effects.
98
W.R. Fisher
I
Formation of acetaldehyde 100
,Q.
80-
-0' -
60-
63' :
: 20
.-
1
.*a. .a. .
'-....-.P I
.*w
'0
L2O"cJ
,
0.
Fig. 8.8
Formation of acetaldehyde in a soil during incubation at different temperatures (after Tsutsuki and Ponnamperuma 1987)
CONCLUSIONS The direct influence of a global climatic change upon soils in alluvial areas with a relatively high permanent groundwater level are presumably not very drastic. The increasing temperature as well as the increasing annual precipitation will lead to an increase in the biological activity in the soils. This, in general, is favorable but, at high water level or for submerged soils, zones of reduction might become more important and could influence the rhizosphere. For relatively dry soils, the increasing precipitation might moderate salt problems, but might also enhance nutrient leaching and soil acidification. So the ecological significance of the expected consequences of global climatic changes depends mainly on the actual ecology of the interesting soil system. Two other items which are not taken into consideration are the large area sedimentation of materials eroded upstream, and the salinization by seawater of flat coastal areas. These factors might be of greater importance than the direct influences mentioned above, especially in highly productive areas of the tropics.
REFERENCES Anderson, T. - H and Domsch, K.H. (1986). Carbon assimilation and microbial activity in the soil. Z. Pflanzenern&r. Bodenkd. 149: 457-468.
Influence of climatic change on soils in the alluvial domains
99
Kuo, S . and Mikkelsen, D.S.(1981). The effects of straw and sulfate amendments and temperature on sulfide production in two flooded soils. Soil Sci. 132: 353-357. Tsutsuki, K . and F.N. Ponnamperuma. (1987). Behavior of anaerobic decomposition products in submerged soils. Soil Sci. Plant Nutr. 33: 13-33. Yu, Tian-ren (1985). Physical chemistry of paddy soils. Science Press, Beijing.
This Page Intentionally Left Blank
101
Addendum to Chapter 8 METHANE FORMATION IN WATERLOGGED PADDY SOILS AND ITS CONTROLLING FACTORS Yasuo Takai* and Eitaro Wada**
* Tokyo University of Agriculture, Sakuragaoka. SeLagaya, Tokyo 156. Japan ** Mitsubishi Kasei Institute of Life Sciences, Minamiooya, Machida, Tokyo 194, Japan A distinguishing feature of waterlogged paddy soil lies in the creation of a reduced state in the soil due to the limited oxygen supply by the flood water. First, the authors show the successive reduction process in waterlogged paddy soils (Takai, Koyama and Kamura 1956a and 1956b; Takai and Kamura 1966; Takai 1984). As shown in Table 8.5, it could be concluded that each type of microbial metabolism proceeds successively according to the oxidationreduction condition from aerobic respiration using molecular oxygen, which is the most efficient energy- yielding reaction, to methane fermentation, which is a comparatively inefficient energy-yielding one. Table 8.6 is prepared on the basis of the idea of oxidative and reductive capacities in the soil using the results of incubation experiments under a closed condition. In the left column, eight paddy soils are listed in decreasing order of rice productivity. Total oxidative capacity is represented as 0 2 ml 100 g-’ of dry soil, after the conversion of N03-, bio-active Mn(1V) and Fe (111) contents to equivalents of molecular oxygen. The mineralized nitrogen content is assumed to be a reductive capacity. Both the relative values are normalized to the greatest yielding paddy soil Mimaki as 100%. The order of the ratio of oxidative to reductive capacities corresponds well to that of the ratio of COdCH4 formed during the incubation under waterlogging. This reveals that the content of bioactive ferric iron may be the most important controlling factor for methane fermentation. The mechanism of methane fermentation was investigated using a 14-C tracer technique in the 1960s (Takai 1970). Two metabolic pathways were found, firstly the transfer of the methyl radical in acetic acid and secondly the carbon dioxide reduction pathway. 14CH3COOH- l4Co4+Cq 14C02+4H2A- I4CH4+2H2+4A The first way was more predominant than the second way in waterlogged paddy soils. In the temperate paddy field and the rice plot experiment, mid-summer drainage increased soil Eh and suppressed methane formation (Takai and Koyama 1956).
Table 8.5
Succession of microbial metabolism in waterlogged paddy soils
Period of Stage of incubation reduction
Earlier
Later
1st stage
2nd stage
Chemical transformation
Initial Eh in soil V
Expected pattern of energy metabolism
Disappearance of molecular oxygen
+ 0.6
+ 0.5
Oxygen respiration
Disappearance of nitrate
+ 0.6 - + 0.5
Nitrate reduction
Formation of Mn2+
+ 0.6
Formation of Fe2+
+ 0.6 - + 0.3 (Ferric ion
-
-
+ 0.4
Y
i3 Formation of N Q - N
Formation of Co;!
Rapid progress
Rapid progress
(Manganate reduction)
Formation of organic acid
Hypothetical pattern of O.M. decomposition
Usually not Aerobic and accumulated semi-anaerobic without fresh decomposition organic process matter (0.M) Y
reduct.)
- -0.19
Sulphate reduction
-
Fermentation
Formation of sulphide
0
Formation of hydrogen
-0.15
Formation of methane
-0.15 - -0.19 Methane fermentation
-0.22
2 Early stage: rapid accumulation Slow progress
Slow progress
Advanced stage: rapid decrease
s D
K
Anaerobic decomposi tion process
tp;
5
Table 8.6
The oxidative and reductive capacities, and the ratio of CO2lCH4 formed during the incubation for 42 days under waterlogged conditions in various soils with different rice productivity
i
; I
Soils Nagano Mimaki Nagano Noushi Hitokuwata Aichi Noushi Aichi Arumi Nigata Yahiko Aichi Sanage Aichi Teruo
Rice Prductivity High
T
1 Low
Microbial Metabolism I stage Aerobic
T
I' stage Anaerobic
Oxidative capacity 0 2 m1/100 g Redu- Redu02 N Q cible cible Total Mn Fe ml % 2.4 0.6 4.3 56 63.3 100 2.4 1.0 0.7 63 67.1 106 3.8 1.2 0.3 63 68.5 108 1.1 1.2 0.2 20 22.5 36 3.8 1.4 2.1 35 42.3 67 3.4 1.9 0.9 70 76.2 120 1 4 0.8 0.2 4.2 6.6 10 217 1.0 0.4 7.0 11.1 18
Reductive capacity = Mineral N mg/lOOg 8.8 17.4 16.7 13.0 25.8 39.4 7.8 16.8
% 100 198 190 148 293 448 89 191
Ox.cap/ Red.cap
Cod
XlOO 100 54 57 24 23
% 100
27 11 9
k
CHq
30 37 19 17 14 11 7
$' 5' f
; 2 00
b h
0. i;-
104
Y.Takai and E. Wa&
In the tropical paddy field, the authors revealed that the content of methane gas in the rice plot was significantly lower than in the unplanted plot (Yoshida, Takai and Rosario 1975). It has been elucidated that rice planting suppressed methane formation due to the aeration by the diffusion of oxygen from the rice rhizosphere which was transported from the soil into the roots. At that time, a hydrophyte transport release of methane gas was unknown by the authors. Last decade, carbon isotope ratios of methane, C 0 2 and related organics in long-term fertilizer trials at Konosu, Japan were studied with emphasis on methane formation and emission (Nakamura, Takai and Wada 1990). Fig 8.9 summarizes carbon dynamics in a plot, supplied with organic manure (OM plot) at the above-mentioned field. The annual input of organic carbon accounts for 308 kg C/lOa/year with an average 613C value of -26./8%0, whereas the accumulation of organic matter in the plot is 40 Kg C/lOa/year. Consequently, the bulk 613Cvalue of decomposed organic matter (308-40= 268 K gC/lOa/year) is calculated through an isotope mass balance calculation to be 26.4%0.
Fig. 8.9
Summary of carbon cycling in the OM plot at Konosu paddy field
Assuming C02 (-18.5%0) and CH4 (-52.9%0)as the final products, the production rate of COdCH4 = f/( 1-0is calculated by the following mass balance: -26.4 = -18.5ft + (-52.9) (1-f)
f = 0.77
The ratio of COdCH4 in paddy soils under anaerobic condition is around 3.3 with an average 6I3C (CH4)value of -53 %o. The production rate of CH4 (268 x 0.23 Kg C/lOa year = 62 Kg C/lOa/year) in the OM plot is similar to that
Methane formalion in
water logged paddy soils
105
(60 Kg C/lOa/year) estimated from the incubation experiments of other paddy soils in Japan (Koyama 1964). The present soil methane 613C (-53 Ym) is significantly higher than those (from -58 to -67960) reported in the literature (Quay et al. 1988). However, an average 613C value for bubbles collected from Japanese paddy fields (-54O/w, unpubl.) is similar to the present data. The discrepancy might be explained as follows: recently the importance of methane transport through rice plants has been emphasized as opposed to difhsion and bubble transport across the water-air interface (Cicerone and Shetter 1981; Minami and Yagi 1988). Although contributions of both pathways to methane release from paddies have not yet been quantified, the relatively high occurrence of the C o r n 2 process near roots of rice plants might lower the values of methane transported via rice plants. Summing up current knowledge on the methane production in paddies a possible schema is presented in Fig. 8.10 with emphasis on the 613C variation of methane (Wada and Nakamura 1989). In Japan we are at present considering that bubble methane is important during the early stage (June - July) of flooding, while plant transporting methane becomes high from August to October.
A : 6°C ( ~ 0C ~ H ~- )
Fig. 8.10
Schematic illustration of methane production and its release. Possible
Sl3C values of methane at direrent sites are also included.
106
Y . Takai and E . Wada
Difference in 613Cvalues of substrata, reaction mechanism and its kinetics, and oxidation of methane in oxidized layers, are the three major factors that govern the variation of 6I3C values of methane in paddy fields. A difference in reaction mechanism may be the most probable factor for variations in methane 613C in bubbles evolved in paddies, since source organic matter in rice fields is mostly rice plants of which 613Cchanges within a very narrow ranges only. Stable isotope approaches will be promising not only for the study of methane emission but also nitrous oxide formation. Taking the above-mentioned data into consideration, factors for controlling methane formation in waterlogged paddy soils are: Contents of reducible iron and manganese compounds in soils; Water management such as mid-drainage, intermittent irrigation and water percolation; Enhancement of oxidative decomposition of organic debris such as rice straw, stubbles and weeds before waterlogging.
REFERENCES Cicerone, R.J., J.D. Shetter (1981). Sources of atmospheric methane: Measurements in rice paddies and a discussion. J. Geophys. Res., Vol. 86, 7203-7209. Koyama, E. (1964). Biogeochemical studies on lake sediments and paddy soils and the production of atmospheric methane and hydrogen. In: Recent researches in the fields of hydrosphere, Atmosphere and Nuclear Geochemistry, Maruzen Co. Ltd., Tokyo, pp. 143- 177. Minami, K., K. Yagi (1988). Method for measuring methane flux from rice paddies. Jpn. J. Soil Sci. Plant Nutr., vo1.59, 458-463. Nakamura, K., Y. Takai, E. Wada (1990). Carbon isotopes of soil gases and related organic matter in an agroecosystem with special reference to paddy field. In 'Geochemistry of gaseous elements and compounds', Theophrastus Publications, s.a., Athens, Greece, in press. Quay, P.D.. S.L. King, J.M. Lansdown and D.O. Wilbur (1988). Isotopic composition of methane released from Wetlands: Implications for the increase in atmospheric methane. global biogeochemical cycles, V01.2. 385-397. Takai. Y . (1970). The mechanism of methane fermentation in flooded paddy soil. Sci. Plant Nutr. m. Vol. 16, 238-244. Takai, Y. (1984). Microbial dynamics of paddy soil under waterlogged condition. Jour. Korean SOC.Soil Sci. & Ferti., Vol. 17, 187-199. Takai, Y., and T. Kamura (1966). The mechanism of reduction in waterlogged paddy soil. Folia Microbiologica, Vol. 11, 304-313. Takai, Y., T. Koyama (1956). Composition of gases and organic acids contained in soil of paddy field. Jour. Japanese Soil & Ferti., Vol. 26, 509-512. Takai. Y., T. Koyama and T. Kamura (1956a). Microbiological studies on the reduction of paddy soils. Trans. 6th Int. Congr. Soil Sci., Paris, 527-531.
Methane formation in water logged paddy soils
107
Takai, Y., T. Koyama and T. Kamura (1956b). Microbial metabolism in reduction of process of paddy soils, Part 1. Soil & Plant Food, V01.2. 63-66. Wada, E., K . Nakamura (1989) Carbon isotopic studies on global methane production with emphasis on paddy fields In: Proceedings of the fifth working meeting on isotope in nature, Leipzig. in press. Yoshida, T., Y. Takai, D.C.D. Rosario (1975). Molecular nitrogen content in a submerged rice field. Plant & Soil, Vol. 42, 653-660.
This Page Intentionally Left Blank
109
Chapter 9 POTENTIAL INFLUENCE OF CLIMATIC CHANGE ON SOIL ORGANIC MATTER AND TROPICAL AGROFORESTRY Eldon H . Franz Environmental Research Center and Program in Environmental Science and Regional Planning Washington State University, Pullman, Washington 99164-4430, USA
INTRODUCTION Soil organic matter (SOM) has a general range of "response times" on the order of 10' to lo2 years and is a sensitive indicator of climatic patterns (see Chapter 17 by Goryachkin and Targulian for general discussions of the "response time" concept). Thus, SOM can be expected to manifest measurable responses to global warming over the next 20 to 80 years, assuming that the global mean annual temperature will increase by 2 to 4°C at 0.5 to 1 "C per decade. This chapter introduces a simple model for estimating the temperature sensitivity of SOM and calculating the change in global SOM distribution that would be likely to accompany global warming. The implications of the predicted changes are outlined in terms of the global carbon cycle and the management of SOM in tropical agroforestry systems.
THE DYNAMICS AND GLOBAL PATTERN OF SOM The total amount of organic matter in the soil depends on the balance of biomass production and decomposition, and on the soils capacity for organic matter storage. It is, therefore, sensitive to temperature and precipitation and is quite variable at local scales. At continental scales, the highest levels of SOM occur in climates that are not at the extremes of either temperature or precipitation and decline toward hot/dry and cold/wet climates (Kononova 1975). At global scales, however, a more general pattern is evident: SOM increases along gradients of increasing precipitation and decreasing temperature (Post et al. 1982, 1985; Zinke et al. 1984). The role of temperature in determining the global pattern of SOM follows directly from its differential effects on production (additions to SOM) and decomposition (losses of SOM). Fig. 9.1 shows relative values for production and decomposition as a function of mean annual temperature from -15 to 30°C.
110
Fig. 9.1
EH.Franz
Relative rates of production and decomposition as a function of mean annual temperamre
The decomposition function was derived from data for cellulose humification (Sorenson 198 1) as described by Parton et al. (1987). It represents the temperature influence on decomposition in a complex simulation model (see Chapter 10 by Stewart et al.) that gives good predictions of steady state SOM on a broad regional basis (Parton et al., 1987). It agrees closely with the field observations that decomposition coefficients for IITA (Nigeria, mean annual temperature = 261 " C ) and Rothamsted (England, mean annual temperature = 8.9"C) differ by a factor of 4 (Jenkinson and Ayanaba 1977). The production function was calculated from Lieth's (1973) net primary productivity/mean annual temperature equation by scaling relative to the maximum value of 3,000 g m-2 y-l. Using it to represent relative rates of additions to SOM requires the assumption that the influence of temperature on SOM production parallels its influence on net primary productivity. For a soil at steady state with SOM assumed to comprise a single pool, dividing the annual additions to the SOM pool by the decomposition coefficient gives an estimate of the steady state SOM level. The basic first order model of this process using soil nitrogen as the measure of SOM (following Stevenson 1986) is:
Climatic change, soil organic matter and tropical agroforestry
111
N = A/r - (A/r - No)e-fi
where
N A r t
No A/r
= % of N in
the soil = annual rate of addition of N = decomposition coefficient = time = the initial level of N = the steady state level of N
The significant mathematical properties of function (1) relate to the dynamics of SOM change when SOM is not at steady state. Under such conditions, the gains or losses of SOM are determined by the difference between steady state and initial levels and constitute a negative exponential series for constant intervals of time. It is, therefore, possible to characterize the response of SOM to any perturbation involving a change in A and/or r by calculating the time required for a half-change in the SOM increment, the relaxation time or t 1,2. The relaxation time for SOM is an explicit measure of the "response time" of the system as SOM levels approach the steady state following a perturbation. It is an inverse power function of r and takes values between 10' to lo2 years for r ranging from 0.069 and 0.0069. This range of r values encompasses the majority of decomposition coefficients determined from field studies (Young, 1989) thus verifying that the general "response time" for SOM is on the order of 10' to lo2 years. Fig. 9.2 shows the relative steady state values of SOM with respect to temperature as estimated by taking the quotient of the two functions in Fig. 9.1. Since the decomposition function is 0 at OOC, the base line case calculation is from 1"C to 30°C. An increase in the mean annual temperature at any base case location has the effect of shifting the SOM curve 2 or 4 "C to the left as shown. As the model shows, on a warmer Earth, SOM levels decline. The curve suggests that SOM increases accelerate as temperature decreases, a finding that is consistent with the pattern revealed by aggregated data arrayed to show variations with both temperature and precipitation if precipitation is held constant (Post et al. 1982, 1985; Zinke et al. 1984). The similarity is particularly apparent for the precipitation interval of 650mm to 2000mm, approximately the range of values in the latitudinally averaged profile of the Earth's precipitation (see Fig. 7.3 for a visual comparison). Fig. 9.3 shows the percentage changes in SOM calculated from the shifts in Fig. 9.2. The largest changes are projected for mean annual temperatures below 10 "C at which point the changes increase exponentially with decreasing temperatures. Using the latitudinally averaged profile of the Earth's mean annual temperature (Fig. 9.4), the SOM/temperature curve is transformed to a SOMAatitude curve (Fig. 9.5). Although the temperature profile is approximately symmetrical around 5 ON, the "thermal equator", rather than 0 ", the differences
112
E H . Frau
between the profiles for the Northern and Southern hemispheres are not significant at the scale of the figure and only the Northern hemisphere curve is shown. The curve terminates just below 60"N where the temperaturecurve crosses 1 O C .
;
100
I
I
?
I 3
,
:.
I
..
.....
.
I
:
0
075
,
i
I
/I
, t
, I
'
,
'
Warmer
.
I1
.
.
'.
\
\
'.. \
0.25
7 D e g r e e s Wormer
. ~4 . Degrees . ._
,
.
\
0.50
I
- Bose Line
\
.... .. .-. .. . '....
-
-. _. ._ --:I-.. _,
- - -,:_.. . ..1:. .. . - - _ _---.-.
0 00
Figure 9.2
,
1
-__ ,
Relative SOM at steady state as a function of mean annual temperature for the base line case. For two degrees warmer and four degrees warmer at any base line temperature, steady state SOM decreases
The transformation from mean annual temperature to latitude (Fig. 9.4) is approximately linear at temperatures less than 20°C and latitudes greater than 30 degrees (0.75"C per degree latitude). Assuming that the shift in the SOM profile would be proportional to the temperature gradient, a 2OC warming would shift the SOM profile poleward by 2.67 degrees latitude (approximately 300 km) and a 4OC warming by 5.33 degrees (approximately 600 km). The predicted shift in steady state SOM in the Northern hemisphere is therefore on the order of 150 km per "C. The poleward migration of boreal forest during interglacial periods has been estimated to be 100 km per "C (Davis, 1989). In addition, the Holocene rates of expansion of hemlock, a particularly well documented species, appeared to have averaged 20 km per century (Davis 1989). Such rates are generally considered to be consistent with steady state tracking of glacial cycles by vegetation (Webb 1986; Prentice 1986). Since rates of interglacial tree expansion were probably limited by rates of soil development (Pennington 1986), estimates of the geographical extent of change for vegetation and soils should agree whenever climatic change is slow enough to allow for a steady state adjustment. Potential
CIimlic change, soil organic
matter
and tropical agroforesrry
113
rates of warming (between 0.5 and 1.0 degree per decade) over the next few decades suggest that trees at the higher latitudes may be displaced at 25 to 50 times the average rate for the Holocene.
100% 90% 80%
70% 60%
50% 40%
30% 20% 10% 0%
10
I
Base Line Mean Temperature (degrees C)
Percentage decrease in steady state SOM for 2 "c and 4 "C warmer as a function of base line mean annual temperature
Fig. 9.3 90N 80N 70N 60N 50N 40N
30N 20N 1 ON
0 10s
20s 305 40s
50s 60s
70s
aos 90s 4 0
-30
-20
-10
0
10
20
30
Mean Annual Temperature (degrees C)
Fig. 9.4
Latitudinal profile of mean annual temperature
114
E H.From
00%
I -
I
4 Degrees Warmer
._.
2 Degrees Warmer I
60%
40%
I
/
20%
, , ,
- -
_._ __ __ .__
0%
,
_ -. ___ - ---- - .-
___ I
I
The change in mean annual temperature per degree latitude is much less in the tropics (Fig. 9.4). This implies that the SOM profile would shift over a wider range but with much smaller absolute and percentage changes than at higher latitudes. Although the resultant distribution of changes in steady state SOM is generally the same as that predicted for the distribution of the temperature change by GCMs smaller increments in the tropics than at high latitudes, changes in the distribution of precipitation could produce shifts in vegetation zones that are at least a great as those in the north. For example, the steep gradient of summer precipitation recently located between 12 and 17'N was displaced 450 to 550 km to the North during the Holocene pluvial 5000 to 10,000 years BP (Ritchie and Haynes 1987). Vast improvements in GCMs will be required to adequately predict such patterns.
IMPLICATIONS FOR THE GLOBAL CARBON CYCLE The differential temperature sensitivities of production and decomposition processes, reflected in the SOM model, will apparently actuate decreases in the carbon density of the world's soils as global temperature increases. This implies a shift in global carbon inventories from SOM to atmospheric carbon dioxide and a potential positive feedback to greenhouse warming (Woodwell 1989; Houghton and Woodwell 1989).The loss of SOM has been occumng due to cultivation for at least the last 2000 years. Buringh (1984) estimated that the cumulative losses over that period now total 27% of the total SOM originally stored in soils. Due to
Climatic change. soil organic m i t e r and tropical agroforesiry
115
global warming, future losses will accrue from uncultivated soils as well. There is abundant evidence that the rate of decomposition changes more rapidly as a function of temperature than does productivity. However, it has been argued that the stimulation of photosynthesis, the "fertilization effect" of increasing carbon dioxide, could more than compensate for these temperature effects per se. This is proving to be a more complicated problem than it might appear at first glance. How the temperature response of net primary productivity might be modified by increased CO, concentrations is still difficult to predict. It is now clear, however, that the effects of the expected simultaneous increases in CO, concentrations and temperature will have interactive effects on plant photosynthesis and respiration, perhaps combining to lower net photosynthesis at temperatures below 18°C and increasing them above (Allen et al. 1990). If this proves to be a general pattern it may mean that tropical latitudes will benefit more from the fertilization effect than temperate and boreal zones. Experiments have been planned for the IGBP that would provide some definitive answers. In terms of the global atmospheric CO, pool however, it is the balance of effects on decomposition and production that are of concern. If the stimulation of photosynthesis by CO, is the predominant effect, then CO, concentrations should decrease as temperature increases and thus generate a negative feedback to CO,. If the stimulation of decomposition and loss of SOM is the predominant effect, then CO, concentrations should increase with temperature and thus generate a positive feedback to CO,. Several lines of evidence indicate that decomposition with positive feedback is the more likely of the two, and may have already been detected. First, temperature and CO, increase and decrease in tandem throughout the 160,000 year record provided by the Vostok core (Houghton and Woodwell 1989). While cause and effect cannot be established with certainty by such a correlation, the fact that CO, tracks temperature increases closely during deglaciation, but lags when the temperature trend reverses, strongly disconfirms a negative feedback effect. The pattern is consistent, however, with positive feedback increasing the rate of release of carbon dioxide from SOM during warming and decreasing it upon cooling. Second, the rate of accumulation of CO, in the atmosphere has recently increased. According to Houghton and Woodwell (1989) the rate has risen from about 1.5 ppm per year to 2.4 ppm year since the middle of 1987. They suggest that this reflects positive feedback from the loss of SOM in Northern hemisphere soils warmed during the decade of the 80s. Third, the latitudinal dependence of the seasonal amplitude of carbon dioxide peaks in the northern hemisphere. It rises sharply from the equator to a maximum between 50 and 60"N, suggesting that the ecosystems there are the source of carbon dioxide driving the seasonal amplitude increase (Gammon et al.
116
E H . From
1985). Taken together with the latitudinal dependence of change in SOM, which also shows a peak above 50"N (Fig. 9 . 3 , this observation gives additional weight to Houghton and Woodwell's attribution. It is also likely that much of the potential for loss of SOM has already been exhausted by cultivation of soils at latitudes below 50"N. Such losses are estimated to average approximately 30% over the first 20- 50 years of cultivation (Schlesinger 1986), an amount greater than the predicted loss due to a warming of 4°C at those latitudes.
IMPLICATIONS FOR TROPICAL AGROFORESTRY Management of SOM for productivity maintenance or enhancement is a unifying feature of tropical agricultural and agroforestry practices (Young 1976, 1989). Such practices rest on the knowledge that the availability of plant nutrients in tropical soils is largely a function of SOM (e.g. Sobulo and Jaiyeola 1977; Nair 1984). Climatically induced changes in SOM, therefore, have particular implications for the management of tropical soils. Although the percentage changes in SOM that are predicted for the tropics are small relative to those at higher latitudes, decreases in SOM by any amount will only increase the intrinsic challenges of agriculture in the tropics. Although temperate agriculture has performed satisfactorily in many locations for some time by returning only the annual crop residues to the soil, systems that also involve perennial woody plants are more generally practiced in the tropics. In fact, shifting cultivation, an indigenous form of agroforestry based on alternating periods of cultivation and fallow, is currently the largest category of land use in the tropics and subtropics (Detwiler and Hall 1988). In 1980 shifting cultivation was practiced by an estimated 500 million people on 21% of the total area of tropical forests (Lanly 1985). Each year 3.4 million hectares of closed forests and 1.7 million hectares of open forests are converted to shifting cultivation (Lanly 1982). Shifting cultivation is also the only form of land use that has been explicitly recognized as a source term in global cycles (Ha0 et al. 1989). Recent calculations based on the area estimates of Lady (1985) suggest that the annual conversion of primary closed forests to shifting cultivation releases 252541 x 10l2g C ha-* and conversion of open forests 38-60 x 10l2g C ha-l (Ha0 et al. 1989). Agroforestry systems are thus central to the assessment of effects on climate as well as of climate's effects. Shifting cultivation produces a mosaic of patches in the forest or savannah matrix: cultivated patches that have been recently cleared and burned, and fallow patches that are in different stages of succession. When practiced by populations at low densities, in its traditional and culturally integrated forms, the soil fertility that is lost during cultivation is regained during the fallow interval (Nye and Greenland 1960; Aweto 1981a, b; Young 1989).
117
Climatic change, soil organic matter and lropical agroforeslry
Nye and Greenland (1960) estimated that the decomposition coefficient under forest fallow in West Africa was generally on the order of 0.03 and under cultivation 0.04. The calculated relaxation time for a decomposition coefficient of 0.03 is 23 years, for 0.04, 17 years. It is therefore logical that the period of fallow must be longer than the period of cultivation for the SOM balance to be maintained. The effect of global warming on the decomposition coefficient will be to reduce the cultivation period even more and increase the time required for fallow, thus reducing the intensity of land use that can be sustained. This will increase the urgency of research to develop alternatives that can sustain more intensive use. A recent series of papers by La1 (1989a-e) provides a comprehensive evaluation of the changes in stocks of soil carbon and nitrogen that occur with different land use systems with the potential to replace shifting cultivation by providing a greater intensity of land use. The experiment included four land use systems which were studied over a period of four years: plow-till, no-till, and two agroforestry systems, alley cropping with Gliricidia at 4 m spacings, and alley cropping with Leuceana at 4 m spacings. The annual crops in each system were maize and beans. C and Nlevels declined in all cropping systems but increased, by comparison, in nine years of fallow (Table 9.1). Table 9.1
Differences in carbon and nitrogen densities between 1982 and 1986for the land use systems studied by Lal. Data comparing year 1 with year 10 from the study of Aweto (1981a. b ) are given for comparison. Both sets of data are for the surface 10 em of soil
Land use
Change in C (kg rn-2)
Change in N (kg m-2)
Plow-till
- 1.594
- 0.131
No-till
- 0.607
- 0.085
Leucaena 4m
- 0.653
- 0.167
Gliricidia 4m
- 1.208
- 0.145
9 Years of fallow
+- 0.657
+ 0.056
The data clearly indicate that these systems were not sustainable alternatives to the fallow systems they might replace. For the maintenance of SOM without fallow in the tropics, more research will be needed (see Chapter 18 by Sanchez el al.). The studies of Aweto and Lal exemplify research designed to assess soil changes. For detection of climatic effects, however, controls will be needed to differentiate effects of changes in land use from confounding changes due to climatic change. In addition, the intrinsic variability of soils means that sampling
118
EN. Frau
designs must be adequate to detect expected changes at predetermined levels of significance. For the purpose of monitoring climatic change, knowledge of soil changes during fallow could provide a more or less consistent basis for evaluating soil changes for other land uses. Much of the research in agroforestry in fact, is an explicit search for systems that improve on rotational tree fallows as a basis for soil conservation (Young 1989). Improvements may also be achieved through species selection. For example, the ratio of plant carbon bound in long lived tissue per year to plant carbon bound in short lived tissue per year, the wood to litter production ratio, is variable among species. To compensate for the effect of climatic change on the decomposition coefficient, tree species that allocate greater amounts to litter product and have lower wood to litter production ratios could be selected for planting in an "improved fallow" or on cropland. Selecting species for litter quality and timing of litter additions could be effective in synchronizing mineralization rates with the nutrient requirements of the growing crop in spatial agroforestry systems. The macroclimatic scenario is based on a deviation from the base line mean. Specific microclimates deviate substantially from that mean. Isolated mature trees, for example, have a pronounced effect on the microenvironment beneath the canopy. Compared to open areas, the canopies of Acacia rurtilis and Adansunia digirara in Tsavo National Park, Kenya, reduce incident solar radiation at the soil surface by 45-65%and soil temperatures by 5 1 1 ° C (Belsky et al. 1989). The soils under these trees are also higher in SOM than the surrounding grassland soils by a factor of 1.8 (3.6%compared to 2.0%), because additions tend to be greater and decomposition losses less than in the open (Cunningham 1963). One of the best known agroforestry systems in West Africa is based on the cultivation of crops under the canopy of Faidherbia (formerly Acacia) albida (Felker 1978). In addition to microclimatic and chemical effects associated with the subcanopy microsite, Faidherbia soils also have higher water holding capacities. With the potential €or microclimatic effects along with the other benefits of trees in cropland, planting trees could help to ameliorate the local effects of global warming and provide a sink for global atmospheric carbon dioxide as well (Franz 1989).
REFERENCES Allen, S.G., S.B. Idso, and B.A. Kimball (1990). Interactive effects of COz and environment on net photosynthesis of water lily. Agriculture, Ecosystems and Environment 30:8188. Aweto, A.O. (1 98 la). Secondary succession and soil fertility restoration in Southwestern Nigeria: I. Succession. Journal of Ecology 69:601-607.
C ’ l i m [ i c change, soil organic m u e r and fropical agroforeslry
119
Aweto, A.O. (1981b). Secondary succession and soil fertility restoration in Southwestern Nigeria: 11. Soil fertility restoration. Journal of Ecology 69:609-614. Belsky, A.J., R.G. Amundson, J.M. Duxbury, S.J. Riha, A.R. Ali, and S.M. Mwonga (1989). The effects of eees on their physical, chemical, and biological environments in a semiarid savanna in Kenya. Journal of Applied Ecology 26:1005-1024. Ruringh, P. (1984). Organic carbon in soils of the world. pp 91-109 IN: G. M. Woodwell (Ed). The Role of Terrestrial Vegetation in the Global Carbon Cycle: Measurement by Remote Sensing. John Wiley & Sons, Ltd., Chichester. Cunningham, R.K. (1963). The effect of clearing a tropical forest soil. Journal of Soil Science 14:334-345. Davis, M.B. (1989). Lags in vegetation response to greenhouse warming. Climatic Change 15:75-82. Detwiler, R.P. and C. A.S. Hall. (1988). Tropical forests and the global carbon cycle. Science 239:42-47. Felker. P. (1978). State of the art: Acacia albida as a complementary permanent intercrop with annual crops. Report to USAlD. University of California, Riverside. Franz, E.H. (1989). Tropical agroforestry. Background paper for the international workshop on the contribution of greenhouse gas emissions from agricultural systems to climate change, Intergovernmental panel on climate change. Gammon, R.H., E.T. Sundquist, and P.J. Fraser (1985). History of carbon dioxide in the atmosphere. pp. 25-62. IN: J. R. Trabalka (ed). Atmospheric carbon dioxide and the global carbon cycle. United States Department of Energy, DOE/ER-0. National Technical Information Service, Springfield, Virginia. Hao, W.M., M.H. Liu and P.J. Cruuen (1989). Estimates of annual and regional releases of C 0 2 and other trace gases to the atmosphere from fires in the tropics, based on the F A 0 statistics for the period 1975-1980. Paper presented at the Third International Symposium on Fire Ecology, Freiburg University. Federal Republic of Germany, 16-20 May 1989. Houghton, R.A. and G.M. Woodwell. (1989). Global climatic change. Scientific American 260(4):36-44. Jenkinson, D.S. and A. Ayanaba. (1977). Decomposition of C-14 labeled plant material under tropical conditions. Soil Science Society of America Journal 41:912-9 15. Kononova, M.M. (1975). Humus of virgin and cultivated soils. pp. 475-526. In: J.E. Gieseking (Ed). Soil Components Volume I: Organic Components. Springer-Verlag. Berlin. Lal, R . (1989a). Agroforestry systems and soil surface management of a tropical Alfisol: I. Soil moisture and crop yields. Agroforestry Systems 8:7-29. Lal, R. (1989b). Agroforesrry systems and soil surface management of a trop. Alfisol: 11. Lal, R . (198%). Agroforestry systems and soil surface management of a tropical Alfisol: 111 Changes in soil chemical properties. Agroforestry Systems 8:113-132. Lal, R. (1989d). Agroforestry systems and soil surface management of a tropical. Alfisol: IV. Effects on soil physical and mechanical properties. Agroforestry Systems 8:197-215. Lal, R. (1989e). Potential of agroforestry as a sustainable alternative to shifting cultivation: Concluding remarks. Agroforestry Systems 8:239-242. Lanly, J.-P. (1982). Tropical Forest Resources. F A 0 Forestry Paper 30. Food and Agriculture Organiz.ation of the United Nations. Rome. xii+lO6pp. Lanly. J.P. (1985). Defining and measuring shifting cultivation. Unasilva 37(147):17-21. Lieth, H. ( 1 973). Primary production: terrestrial ecosystems. Human Ecology 1:303-332. Nair, P.K.K. (1984). Soil productivity aspects of agroforestry. International Council for Research in Agroforestry, Nairobi.
120
E H . Franz
Nye, P.H. and D.J. Greenland (1960). The soil under shifting cultivation. Commonwealth Agricultural Bureaux. Farnham Royal, Bucks, England. viii+l56pp. Parton, W.J., D.S. Schimel, C. V. Cole and D. S. Ojima. (1987). Analysis of factors controlling soil organic matter levels in great plains grasslands. Soil Science Society of America Journal 51:1173-1179. Pennington, W. (1986). Lags in adjustment of vegetation to climate caused by the pace of soil development: Evidence from Britain. Vegetation 17: 105-118. Post, W.M., W.R. Emanuel, P.J. Zinke, and A.G. Stangenberger (1982). Soil carbon pools and world life zones. Nature 298:156-159. Post, W.M., J.Pastor, P.J. Zinke, and A.G. Stangenberger. (1985). Global patterns of soil nitrogen storage. Nature 317:613-616. Prentice, I.C. (1 986). Vegetation responses to past climatic variation. Vegetation 67: 131141. Ritchie, J.C. and C.V. Haynes. (1987). Holocene vegetation zonation in the Eastern Sahara. Nature 330:645-647. Schlesinger, W.H. (1986). Changes in soil carbon storage and associated properties with disturbance and recovery. pp 194-220. IN: J. R. Trabalka and D. E. Reichle (eds). The Changing Carbon Cycle: A Global Analysis. Springer Verlag, New York. Sobulo, R.A. and K.E. Jaiyeola (1977). Influence of soil organic matter on plant nutrition in Western Nigeria. pp. 105-1 15. In: Soil Organic Matter Studies, Volume I. International Atomic Energy Agency, Vienna. Sorenson. L.H. (1981). Carbon-nitrogen relationships during the humification of cellulose in soils containing different amounts of clay. Soil Biology and Biochemistry 13:313321. Stevenson, F.J. (1986). Cycles of Soil Carbon, Nitrogen, Phosphorus, Sulfur, Micronutrients. John Wiley & Sons, New York. Webb, T. 111. (1986). Is vegetation in equilibrium with climate? How to interpret latequaternary pollen data. Vegetation 67:75-9 1. Woodwell, G.M. (1989). The warming of the industrialized middle latitudes in 1985-2050: Causes and Consequences. Climatic Change 15:31-50. Young, A. (1976). Tropical Soils and Soil Survey. Cambridge University Press. Cambridge, UK. Young, A. (1989). Agroforestry for Soil Conservation. CAB International, Wallinford. UK. Zinke, P.J., A.G. Stangenberger, W.M. Post, W.R. Emmanuel, and J.S. Olson (1984). Worldwide Organic Soil Carbon and Nitrogen Data. ORNLflM-8857. Oak Ridge National Laboratory, Oak Ridge, Tennessee.
121
Chapter 10
THE USE OF MODELS OF SOIL PEDOGENIC PROCESSES IN UNDERSTANDING CHANGING LAND USE AND CLIMATIC CHANGE J.W.B. Stewart*, D.W. Anderson*, E.T. Elliott**, and C.V. Cole**
*
**
College of Agriculture, University of Saskatchewan, Saskatoon, Saskatchewan S7h' OWO, Canada NREL, Colorado State Univ., Fort Collins, CO. U.S.A.
INTRODUCTION This paper addresses the functioning of agroecosystems as one aspect of changing land use in varying atmospheric conditions. Agroecosystems are the most intensively managed ecosystems, located on the most productive land with the result that farmers have custody of more environment than any other group (Paarlberg 1980, Elliott and Cole 1989). Agroecosystems provide us low food and fiber costs but also have contributed to the degradation of the environment. Problems such as soil erosion, soil depletion, concurrent increased fertilizer use, and environmental pollution are actively being considered in areas of North America, Asia and Europe with quite different cultivation histories. The critical issue is: "Do we understand the functioning of the agroecosystems well enough to manage them for productive and sustainable agriculture given the cyclical nature of the weather patterns and the intensive use of the land?" An agroecosystem has been defined as an interactive group of biotic and abiotic components, some of which are under human control, that form a unified whole (ecosystem) for the purpose of producing food and fiber (Elliott and Cole 1989). The major driving variables are tillage, crop management and changing climate patterns. Our objective was to determine: 1) the current status (i.e.soi1 quality etc.) of the agroecosystems; 2) how this current state was reached; and 3) to predict long-term effects of existing and new management practices, changing weather patterns and impending climate change on soil quality and agricultural productivity. We illustrate this approach in the North American Great Plains, a fragile agroecosystem, which has already been subjected to considerable degradation and change as the result of a combination of management decisions and climatic change.
BACKGROUND Using concepts and procedures of quantitative pedology on a small number of soils selected along climo-, topo- and cultivation chronosequences we were able to expand on the use of driving variables or soil forming factors (cf. Jenny
122
J.W.B. Stewart
el a1
1980) and to integrate them (Fig. 10.1) with hierarchical perspectives used to describe ecosystems (Reiners 1986, Elliott and Cole 1989). We studied the processes and properties that characterize the highly dynamic nature of former grassland and bordering forest soils of the Great Plains of North America, especially those involved in organic matter stability, in nutrient cycling, and in the solution and movement of soluble components (Anderson 1987, Roberts et al. 19891. REGIONAL INFORMATION
-
climate change
- new management practices Impact of genetically altered organisms
GEOGRAPHIC
--
Experiment\ Field management treatments - Field manipulations - Greenhouse experiments Laboratory microcosms
I
Validations
I
Olhers
-
Process .Models - Photosynrhesis, plant growth - .Microbial decomposition - <;as fluxes - Water and solute flow - Orhers
Fig. 10.1
REGIONAL SITE NETWORKS
..U S D A - ~ S -
--
sites ,+r,cu~tural Canada sites State Experiment Station sites ( U S . ) Provincial research sites (Canadian) Private institute sites Olhers, worldwide
SCS-National Soil Survey Area Database * USDA-National Resource inventory * Climate databases * SCS-pedon databases Biological surveys * Other databases
Process level information is integrated and used to develop the agroecosystem model or modifr existing models. Sets of driving variables generated from geographic information and organized with CIS are used us model input. Model output is assigned to corresponding subregions using GlS to obtain regional prediction Cfrom Elliott and Cole 1989)
These studies emphasize that soils must not be viewed as relatively static bodies that are developed gradually over very long periods of time. True, they do reflect the properties of the original parent material, past climate and vegetation, topography and time, but more accurately within the time frame of global change concerns, soil development is presented (Fig. 10.2) as the result of basic processes of additions, removals, transformations and translocations (Anderson 1989). All of these processes can occur simultaneously, therefore the present
Use of soil models 10 undersfand land w e - and climale changes
123
characteristics of a soil represents an integration over the recent past of a large number of temporarily variable processes (e.g. Solonetz, Solodized-Solonetz and Solod soils occurring in landscapes of similar age usually in close juxtaposition) .
Fig. 10.2
Soil forming factors based on the concepts of Simonson (1959) as described by Anderson (1988)
Soils, in common with many ecosystems, have variables that operate at slow, intermediate and fast rates and it is important to recognize the nature of the variables studied. These may not always move soil propenies or development in the same direction. Thus, many soil properties are highly dynamic whereas others are static (see Table 10.1). Variables such as soluble salts are highly dynamic, varying over season and reaching tentative equilibrium in a few years, whereas organic matter levels have a time dimension of decades to centuries with carbonate and particularly clay weathering having a scale of millennia in semiarid climates (Anderson 1977). Interestingly, despite a medium time dimension for organic mattcr buildup in soils, at another level of detail one can differentiate fast (mainly microbial processes), intermediate components where turnover is dampened by physical sorption to clay and slow or chemically stabilized humus components in soils (Anderson 1979; Jenkinson and Rayner 1977).
124
J.W.B. Sfewarf el al.
Table 10.1
Grouping of soil related processes and components based on time
Highly Dynamic Soluble nutrients
Dynamic Adsorbed nutrients
More static, slow Nutrient reserves in minerals
Active or soluble organic matter
Labile organic matter adsorbed to clay
Chemically stabilized organic matter
Solution and movement of soluble components
Weathering of carbonate minerals
Weathering of silicates and clay minerals
Microbial growth
Micro-fauna and meso-fauna Plant growth
Vegetation, i.e. forest
Related to time is the concept of threshold controlled responses or mechanisms. Threshold controlled responses are considered general in the understanding of the development of above ground ecosystems and geomorphic surfaces, but have only recently been discussed in soil science (Stoner and Ugolini 1988). Episodic events that initiate a threshold response result in considerable change in properties over short time spans. In this aspect soils must be considered as being highly dynamic.
DATA BASES - SOURCES AND INTEGRATION A number of ongoing, long-term studies of cropping management systems studies at Agriculture Canada and the United States Department of Agriculture Agricultural Research Service research stations (Campbell et al. 1989, Haas et al. 1957) provide a valuable data base. They made it feasible to integrate and synthesize considerable information on long-term dynamics of agroecosystems without having to wait many years for research sites to mature. Soil inventories at 1:100,000 scale or better are available for the North Great Plains area within Canadian or U.S. soil survey data bases. For instance, the National Soil Survey Area Database (NSSAD) currently contains soils information for approximately 2500 of 3000 soil survey areas in the U.S. in a single database. NSSAD has two components of significance, the first is the map unit component which consists of site specific descriptions of individual soil types (series). These descriptions include information pertaining to the soil physical setting, horizonation, areal extent, predominant land use, crop types and crop yields. The second component is the soils interpretation record, which contains interpretive information for each individual soil type. A Soil Conservation Service database, which contains chemical and morphological attributes (organic C, N, sand/silt/clay percentage and bulk density are included, by horizon, for each soil) for approximately 300 cultivated and 500 uncultivated
Use of soil models
10
understand land me -and climate changes
125
soils of the Great Plains, was used in validation of models (Cole et al. 1989). Extensive climate data are available which includes mean annual and growing season (April through September) precipitation, temperature and potential evapotranspiration for approximately 560 weather stations in the U.S. Great Plains. A measure of site productivity (Sala et al. 1988) and decomposition (Parton et al. 1987) are included for each station. Similar data exist in Canada. The individual "map layers" described above have been used to form a composite GIS that integrates soils, climate, and management data to a single scale of resolution. Overlay procedures in ARC/INFO ((Environmental Systems Research Institute, Redlands, CA) are being used to combine the component map layers, producing a dense polygon map. The resulting composite polygon attributes can then be extracted and subjected to an external classification procedure, producing 100-500 classes of polygons, each with corresponding climate, management, and soils attributes. The ARC/INFO geographic information system seems well suited for the purpose of integrating these multiple data sets. More importantly, it should be possible to utilize a variety of existing data bases using this system.
PROCESSES, INTERACTIONS AND PREDICTIVE SIMULATION MODELS Progress in understanding the processes which are important in elemental interactions and which are important in soil quality considerations have recently been reviewed (Anderson 1988; Stewart and Cole 1989). Generalized agroecosystem models are used to integrate information on driving variables, processes and properties (e.g. Parton et al. 1989a). Processes already studied and integrated into models include above and below ground primary production, decomposition and nutrient cycling. Major driving variables include climate (temperature and water), parent material (soil texture), base status, total S and P, topography and management. The effects of these controlling factors are expressed over a wide range of resolutions, from the global down to regional landscape and field plot levels. We have been using the Century model (Parton et al. 1983), which was developed to simulate soil organic matter dynamics and plant production in grazed grasslands and agroecosystems. The model simulates the dynamics of C, N, and P in the soil-plant system using monthly time steps (Fig. 10.3a, 10.3b, 10.3~). The input data required for the model include soil texture, monthly precipitation, maximum and minimum air temperature, and plant lignin content. The Century model has been used to simulate regional patterns of soil C, N, and P and plant production for the U.S. central grasslands region (Parton et al. 1988, 1989a) and the impact of management practices on agroecosystems (Cole et al. 1989).
J . W . B . Slewart
126
el
al.
PLANT RCSIDUE
L O W
STRUCTURAL N (GIN = 150)
-
AtmOSpherE Fixatiun Deposition
METABOLIC N (GIN = 10-25)
WINERAL N (NO3
*
NH4)
Leaching Gaseous Erosion
PASS1 VE (VN
Fig. 10.3
=
9)
continued on next page
I = Immobilization = nineralizatlon
n
Use of soil models
10
understand land use
STRUCTURAL D (CIP 500)
- and
climale changes
127
OUllCl Tert~ I 1rer Depos I t i o n PLANT RESIDUC
flFTABOLlC P (C/P = 80 150
ACTlVt SOIL P (C/P = 30 8 0 )
LABILE P
tn)
Leaching Erosion Crop
OCCLUDtD P PASSIVE SOIL P ( C I P = 20-200)
Fig. 10.3
I = lmmoblii~ation M = Mineralization ontrols D = Temp 8. l(20 T = Texture h p~
Flow diagram for the a ) carbon submodel, b) nitrogen submodel and c) phosphorm submodel (adaptedfrom Parton et al. 1987, 19896)
This model has successfully simulated the impact of cultivation on agroecosystems in the United States (Cole et al. 1987), Canada (Parton et al. 1988, 1989a), and Sweden (Parton et al. 1983) over time periods ranging from 10 to 300 years. The model includes the direct effect of cultivation events on nutrient cycling in soil organic matter dynamics (Elliott 1986). The direct impacts include incorporation of standing and surface residues into soils of different texture, modification of soil temperature and soil water patterns, and increasing the turnover rate of soil organic matter. Comparison of simulated effects with detailed soil organic matter, nutrient cycling and plant production data from regional rcsearch sites validated the model. Larger areas are currently being investigated using this regional modeling approach as this procedure had been applied to Northeastern Colorado to produce maps of net primary production, net N mineralization and nitrous oxide production and to provide summed estimates of these output variables for the entire region (Burke et al. in press). Other more mechanistic models such as the Grassland Ecosystem Model (GEM) have also been used to predict the effects of climatic change on grasslands
128
J.W.B. Sfewort el 01
(Hunt et al. in press), and to represent important feedbacks among primary production, photosynthetic pathways, water use, decomposition, and nutrient cycling processes. For example, it includes mycorrhizae and fauna, and uses more mechanistic translocation, nutrient uptake, shoot and root death and decomposition processes. This model has been applied to analyzing the difference in primary production and nutrient cycling patterns between native shortgrass prairie and introduced crested wheatgrass in Wyoming. The GEM model is able to serve as an aid for interpreting differences in production and nutrient cycling between different crops and among sites. Another important cropping systems model is the nitrogen, tillage and residue management simulation model (NTRM) (Shaffer and Larson 1987) which can be used to quickly determine environmental impacts and provide direct management assistance to farmers in an interactive version, COFARM (Shaffer et al. 1984). I t has been used to simulate many processes such as the decay of crop residues and cycling of C and N. The crop growth submodel consists of a series of subroutines capable of simulating the growth and development of field maize, sweet corn, sorghum, soybean, spring and winter wheat, oat, barley, rye, sunflower, alfalfa, pasture grass, sugarbeet, cotton, peanut, tomato, field pea, sugarcane, sweet potato, and carrot. NTRM is capable of simulating any combination of these crops in rotational sequence, and several years of simulation can be run to estimate the impacts of crop rotation on various parts of the system. It can also be used in studies involving NO,-N loading of groundwater. NTRM model inputs include climate data such as daily maximum and minimum air temperatures, pan evaporation, precipitation, wind run, and solar radiation (optional). Values must be provided for soil physical, chemical, and biological properties of user specified soil horizons. The NTRM model contains crop submodels for growth of tops and roots as a function of climate and soil variables. Processes simulated include photosynthesis, respiration, leaf area growth, grain filling, transpiration, and N uptake. The root growth submodel (Shaffer and Larson 1987) includes root extension, branching, and death. The impacts of soil tillage on the physical, biological, and chemical properties of the soil are simulated using both tillage submodels and specific relationships within other subroutines (Shaffer 1985). Changes in certain soil "macro" properties such as bulk density, percent organic matter, and texture are translated into changes in other properties such as the soil water characteristic curves, C and N transformalion rates, soil water content, soil strength, soil aeration, and nutrient and salt concentrations.
THRESHOLD CONTROLLED RESPONSE A threshold controlled response occurs when a low frequency, high intensity impact or combinations of impacts impinge on a system resulting in a
Use of soil models lo understand land use -and c l i m l e changes
129
new course of development for a previously stable system (Anderson 1989). Intense wind or precipitation events are examples. We can use our knowledge of wind and water erosion phenomena (e.g. Universal Soil Loss Equation) in conjunction with organic matter management to predict the change in soil properties such as erodibility, and to calculate the potential for erosion in a variety of climates and topography. Future long term studies of soil ecosystems must recognize the possibility of rare or intense threshold controlled responses, in that they may be critical to understanding the system.
USE OF THIS APPROACH IN MORE WEATHERED SYSTEMS We have tried, using this approach, to model organic matter dynamics in tropical soils (Parton et a1.1989b) without great success. In part, this is because several of the important processes and controls are not well known. Fewer data are available for tropical soils and soil organic matter dynamics are less well understood. The results from attempts to simulate soil organic matter dynamics in a tropical forest suggest some modifications are needed. Specific research is needed mainly including N and P compounds, to clarify: 1) the effect of soil texture, mineralogy and Fe and A1 chemistry on soil organic matter formation and decomposition; 2) the proportion of soil organic matter in active, slow and passive pools; 3) the dynamics of phosphorus forms in tropical soils, particularly organic P forms; and 4) the extent of the leaching of organics, including N and P compounds in tropical soils of varying mineralogy and texture. These processes are mainly the controlling factors of dynamic soil properties in weathered soils.
CONCLUSION This paper discusses a few of the models that have been developed to understand the functioning of agroecosystems in relatively young unweathered soils. These models or combinations of these models can be used to simulate the effects of a range of climatic conditions on crop productivity and soil quality. We plan to continue this approach concentrating on the technology transfer mechanisms that will unite theoretical approaches with practical action. Extrapolation of this approach to more weathered soils will depend on a better understanding of several major processes. These include the effect of soil inorganic constituents and selected minerals on abiotic and biotic formation and stabilization of organic matter, especially in weathered soils.
REFERENCES Anderson, D.W. (1977). Early stages of soil formation on glacial till mine spoils in a
130
J.W.B. Slewarl
el a1
semiarid climate. Geoderma 19:ll-19. Anderson, D.W. (1979). Processes of humus formation and transformations in soils of the Canadian Great Plains. J.Soil Sci. 30:77-84. Anderson, D.W. (1987). Pedogenesis in the grassland and adjacent forests of the Great Plains. Adv. Soil Sci.7:53-93 Anderson, D.W. (1988). The effect of parent material and soil development on nutrient cycling in temperate ecosystems. Biogeochem. 5:71-97. Anderson, D.W. (1989). Long term ecological research - a pedological perspective p 394-417 in Final Rep. of Int Workshop, Berchtesgaden on "Long Term Ecological Research - a Global Perspective" Publ. by the German Nat. Comm. for UNESCO/MAB Bonn, FRG. Burke, I.C., D.S. Schimel, C.M. Yonker. W.F. Parton and L.A. Joyce (1990). Regional modelling of grassland biogeochemistry using CIS. Landscape Ecology (in press) Campbell, C.A., et al. (1989). Crop rotations on the Canadian prairies - an Agriculture Canada perspective. Agric. Can. Bull.. Ottawa. Cole, C.V., J.W.B. Stewart, H.W. Hunt, and W.J. Parton (1987). Cycling of carbon, nitrogen, sulfur and phosphorus : Controls and interactions. Trans.XI1Ith Int. Society of Soil Sci., Hamburg Symposia and Joint Symposia, VI ; 636-643,. Cole, C.V., J.W.B. Stewart, W.J. Parton, and D. Ojima (1989). Modelling land use effects of organic matter dynamics in the Great Plains of N. America.p89-98 in M. Clarholm and L. Bergstrom (eds.) Ecology of Arable Land - Perspectives and Challenges. Martinus Nijhoff Publishers Dordrecht. Elliott, E.T. (1986). Aggregate structure and carbon, nitrogen and phosphorus in native and cultivated soils. Soil Sci. SOC..h e r . J. 50:627-633. Elliott, E.T., and C.V. Cole (1989). A perspective on agroecosystem science. Ecology 70:1597-1602. Haas, H.J., C.E. Evans, and E.F. Miles (1957). Nitrogen and carbon changes in Great Plains soils as influenced by cropping and soil treatments. USDA Tech. Bull. 1164. Hunt, H.W., M.J. Trlica, J.K. Detling, E.F. Redente, C.V. Cole. J.C. Moore, D.E. Walter, M.C. Fowler, D.A. Klein, and E.T. Elliott (in press). Simulation model for the ecosystem level effects of climate change in temperate grasslands. Ecol. Model. Jenkinson, D.S. and J.H.Rayner (1977). The turnover of soil organic matter in some Rothamsted classical experiments. Soil Sci. 123:298-305. Jenny, H. (1980). The soil resource. Ecol. Studies Vol. 37. Springer Verlag, New York. Paarlberg, D. (1980). Farm and Food Policy. Univ. of Nebraska Press, Lincoln, NE, USA. Parton, W.J., D.W. Anderson, C.V. Cole, and J.W.B. Stewart (1983). Simulation of soil organic matter formation and mineralization of semi-arid agroecosystems. In: R K. Lowrance,ed.,Nutrient cycling in agricultural ecosystems, pp. 533-550, University of Georgia, Col.of Agr., Spec. Pub. No. 23, Athens, GA. Parton, W.J., D.S. Schimel, C.V. Cole, and D. Ojima (1987). Analysis of factors controlling organic matter levels in grassland soils Soil Sci. SOC.Amer. J. 51:1173-1179. Parton, W.J., J.W.B. Stewart, and C.V. Cole (1988). Dynamics of carbon, nitrogen, phosphorus and sulfur in cultivated soils: a model. Biogeochem. 5:109-131. Parton, W.J.. C.V. Cole, J.W.B. Stewart, D.S. Schimel, and D. Ojima (1989a). Simulating the long-term dynamics of C, N and P in soils. pp. 99-108. In: M. Clarholm and L. Bergstrom (eds.) Ecology of Arable Land - Perspectives and Challenges. Martinus Nijhoff Publishers Dordrecht. Parton, W.J., R.L. Sanford, P.A. Sanchez and J.W.B. Stewart (1989b). Modelling soil organic matter dynamics in tropical soils. pp. 153.171. In D.C. Coleman, J.M. Oades and G. Uehara (eds.) Dynamics of Soil Organic Matter in the Tropics. Univ. Of Hawaii Press. Manoa.
Use of soil models 10 undersland land use - and climale changes
131
Reiners, W. A. (1986). Complementary Models for Ecosystems. Amer.Nat. 127: 59-73. Sala, O.E., W.J. Parton, L.A. Joyce, and W.K. Lauenroth (1988). Primary production of the central grassland region of the United States. Ecology 69:40-45. Shaffer, M.J. (1985). Simulation model for soil erosion-productivity relationships. J. Envir. Qual.. 14:144-150. Shaffer, M.J. and W.E. Larson ed. (1987). NTRM, a soil-crop simulation model for nitrogen tillage. and crop-residue management. U.S. Dept, Cons. Res. Rept. No. 34-1. Shaffer, M.J., J.B. Swan, and M.R. Johnson (1984). Coordinated farm and research management (COFARM) data system for soils and crops. 1. Soil and Water Cons. 39:320-324. Simonson, R.W. (1959). Outline of a generalized theory of soil genesis. Soil Sci. SOC.Amer. Proc. 23: 152-156. Stewart, J.W.B., and C.V. Cole (1989). Influence of elemental interactions and pedogenic processes on soil organic matter dynamics. Plant and Soi1,115:199-209. Stoner, M.G. and F.C. Ugolini (1 988). Arctic pedogenesis: Threshold controlled subsurface lcaching eposidcs. Soil Sci. 145:46-51. Roberts, T.L., J.K. Bettany and J.W.B. Stewart (1989). A hierarchical approach to the study of organic C, N, P and S in Western Canadian soils. Can. J. Soil Sci. 69:739-750.
This Page Intentionally Left Blank
133
Addendum to Chapter I0
MODELLING NITROUS OXIDE PRODUCTION BY DENITRIFICATION J.R.M. Arah and K.A. Smith The Edinburgh School of Agriculture West Mains Road, Edinburgh EH9 3JG. UK.
INTRODUCTION Developing rational responses to anticipated climate change depends on understanding the processes by which this change is brought about. Increasing concentrations of nitrous oxide (N20) in the atmosphere contribute both to the degradation of the stratospheric ozone layer and to the socalled "greenhouse effect" (Bouwman 1990). Much of this N 2 0 derives from soils and land use, a significant fraction of it from denitrification, which is the dominant source of N20 emission from poorly aerated, nitrate rich soils. Denitrification produces both N2O and molecular nitrogen (N2), in highly variable proportions; here the term '"20 fraction" is used to denote that fraction of gaseous products of denitrification which reaches the atmosphere as N20. Denitrification fluxes and N 2 0 fractions measured in the field using the acetylene inhibition technique (Ryden et al. 1979) show extreme spatial and temporal variability. High fluxes are usually associated with low N 2 0 fractions; beyond that it is difficult to see any pattern (Arah and Smith 1990). This indicates that the relationship depends on a number of parameters, the interaction of which must be modelled in order to predict current rates and N 2 0 fractions, together with any alterations which might occur in response to anticipated climatic changes.
THE MODEL A deterministic model, based on knowledge of the actual processes
occurring in the soil rather than on simple curve fitting, should apply to all soils under all climatic regimes. Arah and Smith (1989) and Arah (1990) describe such a model, developed for aggregated temperate soils but applicable in principle to all. Denitrification occurs when nitrate is present in anaerobic microsites, which develop wherever the microbial demand for oxygen ( 0 2 ) exceeds the diffusion controlled supply. This may occur where 0 2 diffusion is impeded by water, either at the centers of soil aggregates (Smith 1980), or in small saturated regions
134
JR.M. Arah and K A . Smith
within a structureless soil (Arah 1988), or wherever the local 0 2 demand (reduction potential) is exceptionally high (Parkin 1987). Whatever the precise mechanism causing local anaerobism is, denitrification in such a microsite may be modelled as a process of simultaneous diffusion and reaction of 0 2 , nitrate and N20. Enzyme mediated microbial reactions are assumed to follow dual substrata Michaelis Menten kinetics, nitrate reduction is inhibited by 0 2 , nitrate and N20 is inhibited both by O2 and by nitrate (McConnaughey and Bouldin 1985). Intramicrosite concentration profiles of 0 2 , nitrate and N 2 0 are calculated by an iterative method implemented on a microcomputer. The only requirement for convenient mathematical simulation is some degree of symmetry; linear, cylindrical or spherical. Linear symmetry is appropriate to homogeneous, structureless soils and sediments; the rhizosphere best approximates to a cylinder, and both organic "hot spots" and physical aggregates are approximately spherical. Denitrification rates and N20 fractions of larger volumes of soil (containing many microsites differing in size, moisture content and reduction potential) may be obtained by integrating microsite rates over appropriate probability distributions. The model is thus deterministic at a microsite scale, and stochastic in its macroscopic predictions.
MODEL RESULTS Analysis of the response times of microsite denitrification rates and N20 fractions to a major perturbation, such as instantaneous saturation suggests that spherical soil microsites under about 1 cm in diameter may usually be treated as if they were in a state of pseudo equilibrium (Arah 1990). Linear and cylindrical model microsites of similar dimensions respond equally quickly. This considerably simplifies the application of the model. A preliminary prediction is that, in soils where nitrate supply is not limiting, the denitrification rate of macroscopic soil samples should be proportional to the anaerobic fraction calculated according to the method of Smith (1980); in experiments conducted by Parkin and Tiedje (1984) this was found to be the case. Similar experiments performed on different soils in Scotland (unpublished data) support this finding. The calculated N 2 0 fraction of spherical model microsites is extremely sensitive to microsite radius and reaction potential, and considerably less so to the intra microsite gaseous diffusion constant and the external 0 2 concentration (Arah and Smith 1990).
CONCLUSION The model provides a means for rationalizing the otherwise extremely confusing interactive effects of the parameters known to influence N20 emission
DenilriJcation: model nilrow oxide production
135
by denitrification. The problem of relating microscopic parameters to macroscopic properties can be approached form either end: process modelers must attempt to "coarsen" their model parameters to fit in with the information available in soil mapping and GIS databases; compiles of the latter must be prepared to take account of internal inhomogeneity within their mapping units (Burrough 1989), especially where the phenomenon of interest is as variable as N20 emission. It is to be hoped that the relatively precise parameters of this model (microsite diffusion constants, size distributions and reduction potentials) can be related to more readily measured macroscopic properties, some relatively stable (such as soil texturc), and others likely to alter at various rates under the impact of climatic change (including soil organic matter content, pH, rainfall and temperature). If this can be achieved, thcn the model should permit large area predictions of N 2 0 emissions in a changing environment.
REFERENCES Arah J.K.M. (1988). In: Nitrogen efficiency in agricultural soils (D.S Jenkinson and K.A. Smilh, Eds), Elsevier. London pp. 433-444. Arah J.R.M. (1990). Biology and fertility of soils 9:71-77. Arah J.K.M. and Smith K.A. (1989). Steady state denitrification in aggregated soils: a mathematical model. Journal of Soil Science 40: 139-149. Arah J.R.M. and Smith K.A. (1990). Factors influencing the fraction of the gaseous products of soil denitrification evolved to the atmosphere as nitrous oxide. In: A.F. Bouwman (Ed) Soils and the greenhouse effect. Wiley. Chichesler, pp 475-480. Bouwman A.F. (Ed) (1990). Soils and the greenhouse effect, Wiley Chichester. Burrough P.A. (1989). Journal of Soil Science 40: 477-492. McConnaughey P.K. and Bouldin D.R. (1985). Transient microsite models of denitrification 1: Model development. Soil Science Society of America Journal 49:886-891. Parkin T.R. (1987). Soil microsites as a source of denitrification variability. Soil Science Society of America Journal 51:1194-1199. Parkin T.R. and Tiedje J.M. (1984). Soil biology and biochemistry 16:331-334. Ryden J.C., L.J. Lund, J. Letey and D.D. Focht (1979). Direct measurement of denitrification loss from soils 11: Development and application of field methods. Soil Science Society of America Journal 43: 110-118. Smith K.A. (1980). A model of the extent of anaerobic zones in aggregated soils, and its potential application to estimates of denitrification. Journal of Soil Science 31:263277.
This Page Intentionally Left Blank
137
Chapter I 1
IMPACT OF ANTHROPOGENIC ATMOSPHERIC POLLUTION ON SOILS, WITH SPECIAL RELEVANCE TO TROPICAL AND SUBTROPICAL SOILS, AND POSSIBLE CONSEQUENCES OF THE GREENHOUSE EFFECT N . van Breernen Department of Soil Science and Geology Agricultural University, P. 0. Box 37, 6700 A A Wageningen. the Netherlands
INTRODUCTION Widespread recognition of an appreciable effect of atmospheric pollutants on soils has come only recently, with most emphasis on the socalled acid rain problem. Hitherto, acid atmospheric deposition is a problem of the industrialized Northern countries, and there is little evidence of a serious occurrence in tropical and subtropical countries. However, with the expected increase in the use of fossil fuel for industrial purposes in a number of tropical countries, the problem could become greater in the future (Fig. 11.1). In addition to acid precipitation, I will discuss the related problem of nitrogen deposition, and will discuss possible consequences of the greenhouse effect for soil acidification caused by atmospheric deposition.
?/A Sensitive
SOIIS
Present emissions
Fig. I 1 .I
0Present problem areas
.-.
I.-;
Potential problem areas
Schematic map showing regions that currently have acidification problems, and regions where acidification might become severe in the 1988, by permission of SCOPE Publications) future (Rohde et d.,
138
N. van Breemen
RECOGNITION OF THE PROBLEM Acid atmospheric deposition was widely recognized as a problem in relation to acidification of lakes and streams in Scandinavia and Northeastern America in the early seventies (Likens and Bormann 1974). Decline in numbers of salmonid fish were soon found to be associated with low pH and high levels of dissolved aluminium (Schofield and Trofhar 1980). Although some argued that acidification in relation to changes in land use were more important (Rosenqvist 1983), it is now generally recognized that high acidity in precipitation caused by increased industrial emissions of SO2 was the main cause of water acidification and fish decline. Dramatic improvements of surface water quality by decreasing ambient atmospheric acid loads have been demonstrated in small watersheds in Southern Norway where precipitation was intercepted by transparent roofs, and replaced by preindustrial precipitation (Wright et al. 1988). Ulrich et a]. (1979) were among the first to call attention to acidification of forest soils from acid atrnosphcric deposition and its potentially harmful effects on forest ecosystems. The first comprehensive effort to make an inventory of acid rain problems in tropical countries was provided by Rohde and Herrera (1988).
NATURE OF ACID ATMOSPHERIC DEPOSITION AND EFFECTS O N SOILS IN NORTHERN TEMPERATE COUNTRIES Acid atmospheric deposition may consist of dissolved, gaseous and particulate matter that is cither acid or potentially acid. The source of the (potentially) acid substances include SO2 (from burning fossil fuel), NO, (from combustion processes), and NH3 (mainly from animal manure). The potential acidity of NH3 stems from its conversion to H N 0 3 - in soils by nitrifying microorganisms. Potentially acid components are also formed naturally: SO2 from volcanic aclivity, dimethylsulfide formed in marine environments, NO, in atmospheric processes involving other N gases as precursors, and NH3 from organisms and organic rcmains. However, the concentrations of most acidic precursors have increascd dramatically in many areas due to man’s activities. The prccursor gascs arc rather reactive, and have atmospheric residence times in the ordcr of hours to a weck or so. As a result, most of the S and N emitted is dcpositcd within several hundrcd to a few thousand kilometer of their source area. So, contrary to the greenhouse problem which is caused by more inert gascs, acid dcposition is a regional, not a global phenomenon. H2S04 or HN03in acrosol form or present in rainwater or snow is the main component of acid deposition in areas that arc relatively remote from the source areas. Closer to source arcas, dry deposition of the precursor gases is often more imponant. Due to higher rates of wet plus dry deposition of potentially acidic substanccs, prcsent day ratcs of soil acidification in most Western and Central
Impact of anrhropogenic atmospheric pollution on soils
139
European forests are much higher (2-6.5 kmol ha-1 yr-l) than the rates of natural soil acidification (generally 0.1 to 1 kmol ha-' y r l ) and in parts of the USA and Canada affected by acid deposition (1-1.5 kmol ha-' yr-') (van Breemen et al. 1984). pH Values in rain water do not differ much between the two continents: the main reason for the difference is higher dry deposition of sulfur and nitrogen in Europe (due to higher concentrations of gaseous N and S). Ammonia is an important potentially acid substance in the atmosphere in many European areas. Nitrification of deposited NH, may be intense even in acidic (pH 3.5-4.5) soils (De Boer 1989; Stams et al. submitted) and contributes to soil acidification. The high levels of atmospheric ammonium result from N H 3 volatilized from animal manure, produced in large quantities by widespread intensive stockbreeding activities, a pollution source that is of particular concern in the Netherlands and surrounding areas (Asman and Drukker 1987). The high acid loads in many European forest soils have resulted in ecologically relevant chemical changes in the soil such as decreased pH, decreased levels of exchangeable bases, and increased levels of dissolved Al, as shown by comparison of soil samples taken decades ago and taken in the 1970s and 1980s from the same sites in the FRG (Ulrich et al. 1980; Butzke 1988), Austria (Glatzel and Kazda 1985) and in South Sweden (Hallbacken and Tamm 1986; Fakengren-Grerup et al. 1987). These chemical changes are in accordance with those expected from various simulation models that take into account acid atmospheric deposition, soil hydrology, nutrient cycling, weathering, soil microbial processes etc. (Van Grinsven 1988). Sulfate sorption capacities of most European forest soils have been saturated, and mineral soil horizons now not only frequently have pH values (pH in aqueous extracts or in the soil solution) in the order of 3.3 to 4,and base saturations lower than 5 percent, but also soil solutions are often dominated by inorganic A13+ (which is potentially phytotoxic), with sulfate and nitrate as the accompanying anions (Mulder et al. 1987; Mulder 1988; Matzner and Ulrich 1987; Tyler et al. 1987). Concentrations of A13+ are usually in the order of 1-10 mg l-l, and seasonal peak concentrations of up to 50 mg 1-1 and 500 mg 1-1 have been reported in the most acidified soils in the FXG (Raben 1988) and in the Netherlands (Kleijn et al. 1988) respectively. In fact, present-day pH buffering of acid atmospheric deposition in many soils is mainly by dissolution of amorphous A1 compounds that have been accumulated in the soil by centuries to millennia of soil formation. The pools of these compounds in the surface soil are now being depleted rapidly (Mulder et al. 1989). As a consequence, continued acid atmospheric deposition at ambient levels, may cause soil pH values to drop to values closer to those of unbuffered dilute sulfuric-nitric acid mixtures at concentrations determined by the local acid atmospheric deposition and the degree of evaporative concentration of incoming meteoric water. At deposition rates as given above, and percolation rates between 200 and 500 mm per year, theoretical minimum pH values of the percolating water are
140
N.van Breemen
between 2.5 and 3.4. Atmospheric deposition of nitrogen in Western and Central European forest areas is generally between 10 to 70 kg ha-l yr-' (Moseholm et al. 1988). The net annual N uptake in growing timber usually varies from 6 kg ha-' yr-'(in slowly growing coniferous forests) to 15 kg ha-' y r l (in deciduous forests on productive sites). As a result soils tend to become enriched in N, as evidenced by elevated levels of dissolved nitrate in soil solutions, by elevated levels of exchangeable ammonium in surface soils, and by an increase in the proportion of nitrophilous ground vegetation in many forest areas (Van Brcemen and Van Dijk 1988). Where the capacity of the vegetation and soil microorganisms to tie N in organic form has decreased as a result of forest decline, mineralization and nitrification of soil organic N can further contribute to soil acidification and nitrate pollution of ground water. This so-called N saturation problem is of great concern at present in Western Europe. It is somewhat paradoxal that the effects of acid atmospheric deposition were discovered in the downstream parts of the hydrological cycle, i.e. in lakes and streams, before they were noticed higher up in the cycle, i.e. in soils. The main reasons are probably: 1) that certain highly visible and highly prized aquatic organisms (salmon, trout) are very sensitive to water acidification; and 2) that stream water acidification is particularly important when residence time of water in soils, and thus the possibility for buffering of acidity in soils, is limited. For instance, a relatively large fraction of atmospheric acidity is transfercd unbuffered to surface water during snow melt events, in areas with shallow, acid soils, and in higher, steeper reaches of watersheds. The findings discussed above refer mainly to to forested and heathland areas with naturally acidic soils: in agricultural soils and calcareous forest soils, soil acidification (or alkalinization) from other causes (COz, crop removal, fertilization, liming) generally overwhelm the effect of acid atmospheric deposition.
RISK OF SOIL AND WATER ACIDIFICATION FROM ACID DEPOSITION IN THE TROPICS AND SUBTROPICS Mean areal and per capita emissions and depositions of acidic and potentially acid substances are much lower in most countries outside Europe and North America, but exact data are relatively scarce. Although the composition of air and rainwater in the tropics is generally still close to natural background values, local higher emissions may be important in certain areas. At higher and lower latitudes, acidification can be a serious problem, as e.g. in southwestern China and in southeastern Brazil (Moreira-Nordemann et al. 1988; Zhao and Xiong 1988). A future projection of anthropogenic sulfur emissions by Graedel and
lmpacf of anfhrupugenic afmuspheric pollution un soils
141
Cmtzen (1989) suggests that in areas such as the Gangetic plain in India, with rapid ongoing industrialization and abundant supplies of (high sulfur) coal, acid atmospheric deposition could become an increasingly important problem in the future. In view of the fairly low inputs of N in agriculture, ammonium deposition is not likely to become a problem in most tropical countries. Because many soils in the wet tropics have a relatively low base status, increasing acid deposition could quickly result in ecologically harmful changes in the soils. However, because many tropical soils have relatively high sulfate adsorption capacities, the acidification front caused by sulfuric acid will move relatively slow, so that effects of acid deposition on aquatic systems will be delayed considerably.
POSSIBLE EFFECTS OF CLIMATE CHANGE ON SOIL ACIDITY FROM ATMOSPHERIC DEPOSITION A warming of the climate in areas where soils have been exposed to acid atmospheric deposition for a considerable time could aggravate the soil acidity status. First, N saturation and associated soil acidification and ground- and stream water eutrofication will be stimulated if decomposition of soil organic matter is accelerated by higher temperatures. The same may be true for mineralization of organic S. However, at higher temperatures, the sulfate sorption capacity of podzol Bs horizons may increase, while the solubility of (potentially toxic) aluminum may decrease (Slomp 1989), which may partly offset the aggravating effects noted above.
ESTIMATING CRITICAL LOADS OF ACID ATMOSPHERIC DEPOSITION Recently the concept of critical loads has been developed (Nilsson 1986; Nilsson and Grennfelt 1988). A critical load has been defined as "a quantitative estimate of an exposure to one or more pollutants below which significant harmful effects on specified sensitive elements of the environment do not occur according to present knowledge". The elements considered may be soils or soils plus vegetation, groundwater, or surface water. Criteria can be based on sensitivity of biota (vegetation, fish, man) to certain chemical aspects of their environment. Examples are pH, base saturation, Al/Ca ratio in soil solution, A1 concentration in surface water, NH4K ratio at the exchange complex, nitrate concentration in ground water used as drinking water, or combinations thereof. Critical loads have been set for sulfur and nitrogen deposition for specific ecosystems. While critical loads can be determined experimentally (Boxman et al. 1988), process oriented simulation models are being used increasingly for that purpose. This approach is very promising for pollution control strategies by
142
N . van Breemen
policy makers, if the models are designed for regional application (De Vnes 1987) and are combined with emission-atmospherictransport-depositionmodels for large areas (Kauppi et al. 1985; Alcamo et al. 1987).
REFERENCES Alcamo, 1.. M. Amman, J.P. Hetteling, M. Holmberg, L. Hordijk, L. K m a r i , L. Kauppi, P. Kauppi, G. Kosrnai and A. Miikela (1987). Acidification in Europe: A simulation model for evaluating control strategies. Ambio 16: 234-245. Asman, W.A.H., and B. Drukker (1987). Estimated historical concentrations and depositions of ammonia and ammonium in Europe and their origin (1870-1980). IMOU-Report R87-2. Boxrnan, D., H. Van Dijk and J. Roelofs (1988). Critical loads for nitrogen, with special emphasis on ammonium. In: Critical loads for sulphur and nitrogen. Nordic Council of Ministers, Milj~rapport1988: 15, 417 p. (Nilsson and Grennfeld eds.) pp. 295-322. Butzke, H. (1988). Zur zeitlichen und kleinraumigen Variabi1it;it des pH-Werten in Waldboden Nordrhein-Westfalens. Forst und Holz, 43(4): 81 -85. De Boer. W. (1989). Nitrification in Dutch heathland soils, PhD Thesis, Agr. Univ. Wageningen, 96 p. De Vries. W. (1987). The role of soil data in assessing the large-scale impact of atmospheric pollutants on the quality of soil water. Proc. Int. Conf. on the vulnerability of soil and groundwater to pollutants, Noordwijk aan Zee, the Netherlands. Falkengren-Grerup. U., N. Linnermark, G. Tyler (1987). Changes in acidity and cation pools of South Swedish soils between 1949 and 1985. Chemosphere. Vol. 16, Nos. 10-12, p. 2239-2248. Glatzel. G., and M. Kazda (1985). Wachstum und Mineralstoffernhung von Buche (Fagus s y l vatica) und Spitzahorn (Acer platanoides) auf versauertem und schwermetallbelastetem Bodenmaterial aus dern Einsickerungsbereich. Z. Pflanzenernahr. Bodenk. 148: 429-438. Graedel, T.E. and P.J. Crutzen (1989). The changing atmosphere. Scientific American, 261: 28-36. Hallbacken, L., C.O. Tamm (1986). Changes in soil acidity from 1927 to 1982-1984 in a forest area of Soulh-West Sweden. Scand. J. For. Res. 1 (2). 219-232. Kauppi, P., J. Kamari, M. Posch, L. Kauppi and E. Matzner (1985). Acidification of forest soils: a model for analyzing impacts of acidic deposition in Europe, version 11. Collaborative Paper 27 of the International Institute for Applied Systems Analysis, Laxenburg, Austria: 28p. Kleijn, C.E., G. Zuiderna and W. De Vries (1988). De indirecte effecten van atmosferische depositie op de vitaliteit van Nederlandse bossen. 2. De bodemvochtsamenstelling van 8 Douglas opsranden. Soil Survey Inst. Wageningen, the Netherlands, Rapt. nr. 2050, 67 p. Likens. G.E., and F.H. Bormann (1974). Acid rain: a serious regional environmental problem. Science 184: 1176-1179. Matzner, E. and B. Ulrich (1987). Results of studies on forest decline in Northwest Germany. In: Effects of atmospheric pollutants on forests, wetlands and agricultural ecosystems (T.C. Hutchinson, and K.M. Meema, Eds.), Springer Verlag, Berlin. Moreira-Nordemann, L.M., M.C. Forii, V.L. Di Lascio, C.M. do Espirito Santo and O.M. Danelon (1988). Acidification in Southeastern Brazil. p 197- 255 In: Rohde, H. and R.
lmpacl of anthropogenic almospheric p o i l u i o n on soils
143
Herrera (Eds.). Acidification in tropical countries. SCOPE 36, John Wiley & Sons, 405 P. Moseholm, L., B. Andersen and I. Johnson (1988). Acid deposition and novel forest decline in Central and Northern Europe. Nordic Council of Ministers, Milj~rapport,no. 9. Mulder, J., J.J.M. van Grinsven and N. van Breemen (1987). Impacts of acid atmospheric deposition on woodland soils in The Netherlands: 111. Aluminum Chemistry. Soil Science of Am. J., Vol. 51, no. 6 Nov.-Dec. Mulder, J. (1988). Impact of acid atmospheric deposition on soils: Field monitoring and aluminum chemistry. PhD thesis, Agric. Univ. Wageningen, the Netherlands, Department Soil Science and Geology,. Mulder, J., N. van Breemen and H.C. Eijck (1989). Depletion of soil aluminium by acid deposition and implications for acid neutralization. Nature 337: 247-249. Nilsson, J. (1986). Critical loads for nitrogen and sulphur. Nordisk ministerrld MiljOrapport 1986: 11. Nilsson, J., and P. Grennfelt (Eds.) (1988). Critical loads for sulphur and nitrogen. Nordic Council of Ministers, Milj~rapport1988: 15, 417 p. Raben, G.H. (1988). Untersuchungen zur raumzeitlichen Entwicklung boden- und wurzelchemischer Stressparameter und deren EinfluR auf die Feinwurzelentwicklung in bodensauren Waldgesellschaften des Hils. Berichte des Forschungszentrums Waldokosysteme/Waldsterben, Reihe A, Bd. 38. Rohde, H., and R. Herrera (Eds.) (1988). Acidification in tropical countries. SCOPE 36, John Wiley & Sons, 405 p. Rohde, H., E. Cowling, I. Galbally, J. Galloway and R. Herrera (1988). Acidification and regional air pollution in the tropics, p. 3-39. In: Rohde, H. and R. Herrera (Eds.). Acidification in tropical countries. SCOPE 36, Wiley & Sons, 405 p. Rosenqvist, I.T. (1983). Acidification of freshwaters in Europe. Water Quality Bulletin 8: 137-142. Schofield, C.L., and J.R. Trofnar (1980). Aluminum toxicity to Brook trout (Salvelinus fontinalis) in acidified waters. In: Polluted rain. (T.Y. Toribara, M.W. Miller and P.E. Morrow, Eds.). Environmental Science Research: vol. 17, 341-366. Slomp, C. ( I 989). The influence of temperature on aluminum precipitation and dissolution in Spodosol Bs horizons. Unpubl. report, Department of Civil and Environmental Engineering, Syracuse University, Syracuse NY, USA, 40 pp. Stams, A.J.M., H.W.G. Booltink, I.J. Lutke-Schipholt, B. Beemsterboer, J.R.W. Wottiez and N. van Breemen (in press). 15N field study on the fate of atmospheric ammonium in an acid forest soil. Submitted for publ. Biogeochemistry. Tyler, G., D. Berggren, B. Bergkvist. U. Falkengren-Grerup, L. Folkeson and A. Ruhling (1987). Soil acidification and metal solubility in forest of Southern Sweden. In: Effects of atmospheric pollutants on forests, wetlands and agricultural ecosystems (T.C. Hutchinson and K.M. Meema, Eds.). Springer Verlag, Berlin. Ulrich, B., R. Mayer and P.K. Khanna (1979). Deposition von Luftverunreinigungen und ihre Auswirkungen in Waldokosystemen im Solling. Schriften aus der Forstlichen Fakultat der Universitat Gottingen und der Niedersachsischen Forstlichen Versuchsanstalt. Band 58, 291 p. Ulrich. B.. R. Mayer and P.K. Khanna (1980). Chemical changes due to acid precipitation in a loess-derived soil in Central Europe. Soil Sci. 130: 193.199. Van Breemen, N., C.T. Driscoll and J. Mulder (1984). Acid deposition and internal proton sourccs in acidification of soils and waters. Nature 307: 599-604. Van Breemen, N., and H.F.G. van Dijk (1988). Ecosystem effects of atmospheric deposition of nitrogen in the Netherlands. Environmental pollution, Vol. 54: 249-274.
144
N.
van Breemen
Van Grinsven, J.J.M. (1988). Impact of acid atmospheric deposition on soils: Quantification of chemical and hydrologic processes. PhD thesis, Department of Soil Science and Geology, Agric. Univ. Wageningen, the Netherlands. Wright, R.F., E. Lotse and A. Semb (1988). Reversibility of acidification shown by wholecatchment experiments. Nature 25 August, vol 334: 670-675. Zhao, D. and J. Xiong (1988). Acidification in Southwestern China. p.317-345. In: Rohde, H. and R. Herrera (Eds.) Acidification in tropical countries. SCOPE 36, Wiley & Sons, 405 p.
145
Chapter 12 CHANGES IN SOIL RESOURCES IN RESPONSE TO A GRADUALLY RISING SEA-LEVEL H . Brammer* and R . Brinkman**
*
**
Formerly F A 0 Agricultural Development Adviser in Bangladesh. Present address: 37 Kingsway Court, Hove, East Sussex BN3 2LP, United Kingdom. Land and Water Development Division, FAO. Via delle Terme di Caracalla, Roma 00100, Italy.
ABSTRACT A gradual eustatic rise in sea-level of 0.5 to 1 m over the next century could erode some coastal land, impede soil drainage on coastal plains and salinize soils further inland than at present. However, the effects will vary considerably from place to place depending on such local or regional factors as climate, sediment supply to estuaries and coasts, concurrent land subsidence or elevation resulting from natural or human induced causes, and human interventions to protect land from the effects of a rising sea-level. Location specific monitoring and research are needed to understand environmental change processes, identify appropriate responses and estimate their technical, ecological and socioeconomic feasibility. They should include studies of recent and subrecent palaeo-environments which developed during past periods when the relative sea-level was rising gradually, world wide or locally.
INTRODUCTION This paper examines the probable effects of an eustatic rise in sea-level on coastal soils. It assumes, for the sake of analysis, that such a rise in sea-level will take place, will occur gradually and might amount to approximately 0.5 to 1 m over the next century. The paper ignores possible direct effects of climatic change on soils and land use that might be taking place concurrently with a rising sealevel. It also ignores the possible effects of a sudden, catastrophic rise in sea-level of up to 6 meters which could occur if the West Antarctica ice sheet were to separate and slide off into the sea (Hekstra 1989). A sea-level rise of 1 meter, which could affect some land even up to the present 5 m contour by storm surge flooding and salt water intrusion up river estuaries, could potentially affect a gross area of about 5 million km2, equivalent to about 3 percent of the world's total land area (Hekstra 1989). However, the coastal lowlands potentially at risk include about one-third of the world's total cropland and have a present population of about 1 billion - a figure which might be doubled within the next one hundred years. Thus, the potential effects of a 1 m rise in sea-level could be very much greater than would be suggested by the small proportion of the world's land and soils that might be affected.
146
H . Brammer and R . Brinkman
It is important to recognize at the outset that these potential figures grossly exaggerate the soil areas and population that might actually be at risk. 'The populist scenario that has been portrayed of sea-level rising contour by contour to overwhelm 20% of Bangladesh and the Nile delta (e.g. Broadus et al. 1986) is far too simplistic. Such a scenario ignores dynamic changes that will be taking place in coastal lowlands concurrently with or in response to a gradual rise in sealevel. Some of these changes are natural, such as continuing sediment deposition in some deltas and estuary areas which would reduce or totally offset the effects of a rise in sea-level. Others are the result of human action, such as the building of embankments to protect land and people from the effects of a rising sea-level. Little attention appears to have been given to date to soil changes that might take place with a climatically induced rise in sea-level. However, prediction of the likely consequences on soil properties and productivity need not be entirely speculative. There is abundant geological, historical and contemporary evidence of the effects of a rise in sea-level on adjoining land areas. Examples include: The effects of the 100 m eustatic rise in sea-level following the last glacial period on the coastal areas bordering the North Sea as described by Jelgersma (1988); Land subsidence in relation to sea-level of about 1 m during the last 2000 years in the Netherlands, attributed to tectonic subsidence of the North Sea basin and eustatic sea-level rise by geoidal changes following the last glacial period (Jelgersma 1988); Geological and archaeological evidence of land subsidence in the Mississippi, Nile, Rhine and other deltas resulting from tectonic warping or compaction of alluvial sediments (Morgan 1970; Stanley 1988; Jelgersma 1988); Recent evidence of local land subsidence occumng as a consequence of water, natural gas or oil extraction in the Northern Netherlands, Venice, Bangkok, Osaka and elsewhere (Jelgersma 1989; UNEP 1988; Volker 1989); and world-wide evidence of a 10 to 15 cm eustatic rise in sea-level during the lasl 100 years (Hekstra 1989). There is also evidence from a number of places that human interventions could aggravate the effects of a rising sea-level: e.g. by interrupting sediment supply to deltas by dams across rivers upstream, as has happened in the Nile delta following the construction of the Aswan high dam in 1964 (Coutellier and Stanley 1987), and by embanking rivers to channel flood flows directly to the
Effec! of euslalic sea-level rise on soil resources
147
sea, as in the Mississippi delta, thus depriving basin areas of sediment supplies which formerly offset land subsidence due to compaction of the deltaic sediments (Day and Templet 1989). In several localities, rates of land subsidence by natural causes or directly attributable to human activities are much greater than the rates of sea-level rise estimated to occur in response to global warming during the next century.
PROCESSES OF CHANGE IN A COASTAL AREA RESULTING FROM GRADUAL SEA-LEVEL RISE A gradual sea-level rise may cause a complex of changes in different parts of an estuary or coastal plain, depending on a range of external factors and on the nature of the area itself. These factors include the length of the dry season, the tidal range, the incidence of cyclonic storms and storm surges, the exposure of the coastline and the nature of the shelf offshore, and the supply of sandy or fine textured sediments to the area by longshore currents or from rivers. The nature and extent of the changes also depend on the nature of the sediments in the coastal area (clayey, sandy, peat), on the closed or open nature of the coast and the presence of coastal ridges or dunes, on the presence and density of any network of tidal tributaries, and on the presence of any coastal defences such as sea-walls, or other modifications to the natural hydrology by human engineering, such as polder embankments. Increased frequency of storm surges combined with rising mean sea-level may cause lateral erosion of exposed coasts, with landward movement of any sand, shell or clayey beach ridges. Changes inland in saline, brackish water and fresh water tidal zones and the lower parts of river plains will vary with the climatic conditions and are described separately below.
Changes in seasonally wet-dry climates In the tidal or estuarine floodplains, storm surges may bring intermittent salt water flooding, inducing soil salinization. Spring tide flooding during a rising sea-level may also cause seasonal salinization, particularly in the dry season or near equinoxes. In a broad band along saline or brackish tidewater, the natural rates of vertical accretion will be increased by a slow rise in sea-level, the rates also depending on the sediment supply, storm surge incidence and extent of spring tide flooding. In these areas, as well as in the clay plains further away from tidewater, the increased incidence and duration of brackish water flooding will also extend and intensify reducing conditions in soils. This increasing incidence of salinity would reduce the available length of growing period in the remainder, which would tend to decrease potential yields.
148
H . Brammer and R . Brinkman
In the widening belt influenced by brackish water flooding, fresh sediment and increased reduction, mangrove may begin to dominate again (generally Rhizophora species in the tropics; or reeds such as Phragmites in temperate climates). These vegetation types cause increased accumulation rates of soil organic matter and, in combination with sulphate from the brackish water, accumulation of pyrite and organic S compounds. Potential acid sulphate conditions thus may become more widespread. A rising sea-level will cause river beds to aggrade and levees to increase in height along lower river courses, thus impeding drainage from interior basins with consequent raising of water tables and increasing depth and duration of seasonal flooding. Soils subject to fresh water flooding or seasonal saturation will be reduced for longer periods and their organic matter contents will tend to increase. Land that becomes perennially wet will be liable to accumulation of a peat cover, or existing peat areas will increase in thickness and extent. Such changes in soil conditions, unless counteracted by protective embankments or artificial drainage, would generally reduce soil productivity or potential. Specifically: Land suitability of levee soils for perennial dryland crops would decrease; Areas suitable for annual dryland crops would decrease and this land might become better suited for wetland crops such as rice or jute; Areas presently shallowly flooded and well suited to high yielding wetland rice cultivars might only be suited for traditional, tall varieties; and Land now deeply flooded but dry for part of the dry season and used for floating rice and perhaps a short season oilseed crop, may become permanently flooded and unsuitable for cultivation. The aggradation of river courses and raising of levee levels along lower river courses associated with a rising sea-level will also tend to increase channel instability. This could reduce the intervals between potentially catastrophic shifts in river courses in major deltas such as that of the Ganges and Brahmaputra rivers in Bangladesh (Brammer 1990) or those of the Hoang Ho and Yangtze rivers in China. Changes in arid and semiarid climates Low lying coastal areas in tropical and or semiarid climates generally have narrow bands of mangrove vegetation along tidewater; in very arid environments, there may be as little as a single line of low trees. On the landward side, this is bordered by dry, hypersaline, barren or virtually barren land which is shallowly flooded by seawater during occasional spring tides or rare storm surges. Much of
Effecf of euslalic sea-level rise on soil resources
149
this water subsequently evaporates, adding further amounts of salts to the surface and the upper soil layers. A gradual rise in sea-level would appreciably increase the flooding frequency in such areas and somewhat increase the flooding depth and the proportion of the saline floodwater flowing back into tidewater. This would entail a partial desalinization of the outer parts of the hypersaline land and a broadening of the band suitable for the growth of mangroves (mostly Avicenniu in this saline to hypersaline environment). In this climate zone, therefore, mangroves would extend further inland, colonizing former barren hypersaline land rather than freshwater swamp or agricultural land. There would be some increase in organic matter content and some pyrite accumulation, but both to much lower levels than in more humid climates. Seawater intruding further inland than at present and reaching present rainfed wetland rice areas during the short wet season, would severely affect their suitability for this land use. Changes in perhumid climates In perhumid, mostly equatorial, climates without a significant dry season, large proportions of coastal and estuarine plains tend to be covered with a layer of eustatic peat (with its surface at about high tide level), except for relatively narrow, low levees along tidal channels and coastal ridges along exposed coasts. Natural vegetation grades from mangrove forest near saline or brackish tidewater to freshwater swamp forcst on the peat. Wetland rice, or oil palm or rubber on drained land (from which any shallow peat layer would have gradually disappeared) may extend to near tidewater, with a narrow protective belt of mangrove remaining. A gradual sea-level rise would tend to widen the strip influenced by brackish water flooding and vertical accretion of fine textured sediment. The latter would be colonized by mainly Rhizophora mangrove and subject to accumulation of pyrite. Much of the ncwly deposited material would thus become potential acid sulphatc soil. This would be at the expense of land presently suitable, or actually used, for wetland rice, coconuts, oil palm, sugarcane, pineapple or rubber. Further from tidewater, peat growth would tend to accelerate under natural conditions and the peat surface would thus keep pace with the rising groundwater level, linked to the rising mean sea-level. On cultivated peat soils, the maintenance of satisfactory drainage conditions would become increasingly difficult and costly with a rising sea-level, which might lead eventually to abandonment, especially of land under tree crops. The land would then revert to swamp vegetation under which peat growth would resume. Where the protcctive strip of mangrove forest is removed or is inadequate in width, higher storm surges associated with a higher sea-level could inject enormous volumes of seawater onto the peat land. This would kill the freshwater
150
H . Brammer and R . Brinkman
vegetation in the area affected. The peat would then become highly vulnerable to wave erosion and open water might form. This would generally be a shallow freshwater lake, subsequently invaded by a floating herbaceous vegetation and, in the long run, recolonized by freshwater swamp forest. A breach in the strip of land separating the lake from the coast or from a wide tidal channel would, however, convert it into a saline or brackish lagoon and thus change the coastline. Changes on coral islands Coral islands provide a special case. It seems to be popularly expected that a 1 meter rise in sea-level would submerge or destroy the homelands of several Pacific island nations (UNEP 1988). Such islands are formed of coral rubble thrown up by typhoon storm waves to a height of a few meters above high-tide level. Apart from any artificial defences that might be built, the extent to which islands would be submerged or destroyed by a rise in sea-level will depend on whether or not the growth of the living coral at individual locations can keep pace with a rising sea-level and continue to supply sufficient coral debris to maintain storm ridges at the same relative height above sea-level as at present. Living corals can grow vertically at rates up to ca. 1 cm annually, except where retarded by muddy or polluted water.
HUMAN ACTIVITIES AGGRAVATING OR COUNTERACTING EFFECTS OF SEA-LEVEL RISE Human activities may influence the response of low lying coastal areas to sea-level rise in different ways. The interventions may be remote, as in the case of dam construction, or local, as in the case of subsidence because of drainage.
Effects of upstream interventions The construction of dams in upstream areas, for hydroelectric power, irrigation supplies or protection of land downstream from river floods, blocks sediment supply downstream and to coastal areas. This prevents or slows down vertical accretion, thus further aggravating salt water intrusion and impairing drainage conditions in riverine, deltaic or estuarine areas. It also diminishes or blocks sediment supply to the coast itself, which may give rise to retreat of the coastline by wave erosion and longshore transport of the eroded material, as has occurred in the Nile delta following the completion (1964) of the Aswan high dam (Coutellier and Stanley 1987). In the Northeastern Nile delta, the problem of sea-level rise would be compounded by the ongoing, probably neotectonic, subsidence of about 0.5 m per century estimated by Stanley (1988); until 1964, this subsidence was compensated by the deposition of sediments from the annual
Effect of eustatic sea-level rise on soil resources
151
Nile floods. Ongoing natural, geological erosion, or human induced erosion caused by large scale deforestation, constitute major sources of sediment for low lying coastal areas. In the Netherlands, the upper 4 m of mainly clayey sediment in coastal areas, deposited over the last 4000 years, largely originated from central European sources. This sediment supply prevented major changes in the coastline during the slow sea-level rise of about 4 meters over the same period. Major soil conservation activities over large upland areas could therefore have effects on coastal lowlands similar to those of large dams. River training works, such as are under consideration for the Brahmaputra river in Bangladesh, can increase the flow of water and sediment through the main channel at high river stages. This tends to increase the sediment supply and deposition downstream, which could offset the effects of a rising sea-level, partly or wholly. Increased water abstraction for irrigation development in upstream areas, particularly from rivers during the dry season, tends to increase the distance of sea-water penetration in water-courses in coastal areas. This would aggravate the effects of sea-level rise in cutting off existing irrigation inlets from fresh water supplies and by extending the area of coastal saline soils. Effects of local interventions Pumping of groundwater or natural gas or oil may accelerate land subsidence to rates up to several times those envisaged for the sea-level rise over the next century. Natural gas extraction, for example, has lowered part of the Northern provinces in the Netherlands by about 0.6 m over about 30 years, and parts of the Mississippi delta are subsiding more rapidly than in the 19th century, probably in part because of oil and gas extraction. Parts of Bangkok city have subsided by more than 1 m, and are currently subsiding by about 0.12 m per year, by large scale pumping of water for urban and industrial supply. Parts of Taipei have subsided more than 2 m during the last 20 years for the same reasons (Poland 1984). This subsidence is creating serious land drainage and saline intrusion problems in affected areas. Groundwater extraction from coastal aquifers may also induce sea water intrusion into these aquifers at rates far exceeding those envisaged as a result of sea-level rise alone, particularly in arid and semiarid climates. Sea-level rise would therefore compound an already growing problem. The fresh groundwater in coastal dunes and coral islands often occurs in the form of a lens, in dynamic equilibrium with approximate vertical isostasy at the boundary between fresh and saline water. This requires that the groundwater level be above mean sea-level by an amount of about one-fortieth of the total thickness of the fresh water lens. If the sea-level rises, the volume of the fresh water lens would remain dynamically stable only if its surface would also rise by
152
H . Brammer and R . Brinkman
a similar amount. If the groundwater surface were to remain constant, the depth of fresh water storage would be reduced in due course by up to 40 times the sealevel rise, which could drastically reduce or eliminate local water supplies in coral islands and coastal dune areas. A narrowing of dunes or coral islands by coastal erosion would entail similarly serious consequences for the capacity of a fresh water lens. Sediments in low lying coastal areas normally subside by slow compaction, which can be compensated by vertical accretion of fresh sediment. The Mississippi delta, for example, was subject to compaction and sedimentation at a rate of about 1 m per century before human influence became dominant (Titus 1987). Embankment against sea water intrusion or fresh water flooding stops vertical accretion in the embanked area, but subsidence by compaction or by oxidation of peat continues, so that drainage conditions tend to deteriorate. Eventually, this leads to loss of land to shallow fresh water lakes in basin areas. Artificial drainage of coastal soils by ditches or subsurface drains, with evacuation of the drainage water through sluice gates at low tide or by low lift pumping, will accelerate land subsidence. This might be very rapid in peat soils, which are compacted as well as oxidized at high rates after drainage and air entry, but it can occur also in clayey soils. This entails a need for progressively higher embankments and greater pumping heads of the drainage water, with consequently increasing capital and running costs, aggravating the effects of a sea-level rise. On and around coral islands, as in other cases, human activities tend to compound the dangers from sea-level rise: e.g. by effluents, intensive fishing and tourism damaging or slowing the growth of the coral, and thus decreasing the source of material essential for the continued existence of coral islands in a dynamic equilibrium with the sea.
LESSONS FROM THE PAST About two thousand years ago, extensive areas in the southwestern and Central parts of the Netherlands were peat swamps, protected from the sea by levees, coastal ridges and, locally, dunes, as well as by their own elevation: the peat had grown up somewhat above high tide level (Zagwijn 1986). During the post Roman transgression, the sea breached the coast in several places, and the peat was naturally drained and subsided to near sea-level along the tidal inlets. Since about 1000 A.D., farmers colonized the peat areas by cutting shallow ditches for gravity drainage and by repeated burning of the soil surface to accumulate enough nutrients for a crop. By the time the general land level had gradually subsided through these causes to about mean sea-level, and locally lower, a series of major storm surges broke through the protective bands of mineral soils, generally at the sites of estuaries or inlets. The sea-water destroyed
Effecr of eurlalic sea-level rise on soil resources
153
crops and natural vegetation, leaving the peat soils an easy prey to wave action. Large areas in the Northcentral and Southwestern Netherlands thus became part of the North Sea. It took nature and the people centuries to repair the damage. The present IJsselmeer (the former Zuiderzee) is a gash that remained - albeit presently as a fresh water lake embanked against the sea and with polders reclaimed from part of the lake bottom. Coastal peat areas in the tropics that are used for agriculture need protection against such catastrophes in case of a rise in sea-level. Since the end of the last glacial, about 12 to 10 thousand years ago depending on the criteria adopted, there has been a eustatic rise of sea-level of about 100 m , very rapid initially (almost all of the rise took place before 6000 years B.P.), diminishing to rates of the order of 1 m per century around that time and essentially zero since then. Sedimentary landscapes from the time of the gradual sea-level rise, about 6000 to 5000 years B.P. (mid Holocene) have been recognized in different parts of the world. In the coastal plain of the Guyanas, for example, as well as in the Mekong delta (Brinkman and Pons 1968; Brinkman et al. 1990), there are extensive areas of soft clay soils which contain much finely distributed mangrove root material and high proportions of pyrite. These soils occur inland from younger clay plains where soils have less organic matter and much lower pyrite contents. The two landscapes are separated from each other by a beach ridge or major levee landform; in other places, they grade into each other. The soil material in the older, inland part of the coastal plain extends down for several meters, or down to still older, generally very firm,Pleistocene sediments. The most inland parts of this landscape, furthest away from the rivers, are covered by deep peat which grades laterally into the soft clay. The soft, pyrite rich, peaty clays or clays with large amounts of fine root material have been deposited from brackish water in a very broad band of mainly Rhizophora mangrove forest while the sea-level rose at rates similar to those speculatively estimated for the next century. Where 14-C dates are known, they indicate ages between about 6000 and 5000 years B.P., as for a shell ridge separating the older landscape from the younger one, and for the upper soil layers of the older landscape. The younger landscape has a sediment sequence clearly indicating its genesis under conditions of essentially constant sea-level. The upper clay layers are underlain by silty clay material with very thin silt layers at depths ranging between 1 and 3 m; and broad or narrower clay plains are interrupted by sand (or locally shell) ridges running roughly parallel to the present coast or to present estuary levees. Unless there is human intervention, for example in the form of embankment, it would be expected that this young Holocene landscape would again be buried by a very broad band of brackish water swamp forest, slowly
154
H . Brammer and R. Brinkman
rising by vertical accretion with a stipulated slowly rising sea-level and accumulating soft, pyrite-rich clay with high organic matter contents. Landward from this broad landscape, renewed peat accumulation would be expected to occur. The primary productivity of a swamp forest would be ample for raw peat accumulation to keep pace with a sea-level rising at a rate of 1 m per century. Because dead plant material is decomposed much more rapidly above groundwater or floodwater level than when submerged, present day peat accumulation rates are generally smaller.
MONITORING AND RESEARCH REQUIRED Monitoring in carefully selected type locations with very good baseline information will be needed to quantify changes in soils resulting from sea-level rise, including: Disappearance (by lateral erosion); Rejuvenation by reworking (lateral transport); Rejuvenation by vertical accretion; Salinization; Desalinization (from hypersaline conditions); Reduction (increasing period and intensity); Increasing organic matter content; Peat formation; Accumulation of pyrite and organic S compounds. Monitoring will be needed in natural environments as well as where t k present day nature and rates of these changes are influenced by human actions: for example, by embankment or extraction of deep groundwater, natural gas or oil. Such monitoring will not only increase our understanding of the processes themselves, but also aid in understanding and modelling changes by a stipulated sea-level rise. Any corrective action that may need to be taken can then be identified and its technical, ecological and economic feasibility assessed. Research does not start at zero as far as the response of low lying coastal areas to a sea-level rise is concerned. There have been many instances of gradual sea-levcl rise in the recent and more distant past as well as, locally, ongoing at present. Some of these cases have been studied and described or monitored in some detail, as indicated above. Only in a few cases, however, have soils been taken explicitly into account. Further studies are needed on subrecent and recent palaeo-environments where the sea-level was rising, eustatically or for tectonic or other reasons, at rates of about 0.2 to 1 meter per century. Cases could include: the North Sea coast between Denmark and Belgium about 4000 years B.P.; the shallowly
EfJeci of euslaiic sea-level rise on soil resources
155
buried mid Holocene landscape South of Dhaka, Bangladesh, recognized by Brammer and Brinkman in 1969; the Nile delta (Stanley 1988); and the Musi delta area North of Palembang, South Sumatra, Indonesia (Brinkman 1987).
ACKNOWLEDGEMENTS We thank S. Jelgersma, D. Norse and G.M. Higgins for critical and helpful comments and suggestions on a draft of this paper.
REFERENCES Brammer, H . (1990). Geographical complexities of detailed impact assessment for the Ganges-Brahmaputra-Meghna delta of Bangladesh. In: Wanick, R.A. and T.M. Wigley (Eds). Climate and sea-level change: observations, projections and implications. Cambridge University Press (in press). Brinkman, R. (1987). Sediments and soils in the Karang Agung area. Ch. 4, p 12-22. In: Best, R., R. Brinkman and J.J. van Roon. Some aspects of tidal swamp development with special reference to the Karang Agung area, South Sumatra Province. Mimeo, World Bank, Jakarta, Indonesia. Brinkman, R., and L.J. Pons (1968). A pedo-geomorphological classification and map of the Holocene sediments in the coastal plain of the three Guyanas. Soil Survey Paper No. 4. 40 p. separate map. Soil Survey Institute (Staring Center), Wageningen. Brinkman, R., Nguyen Bao Ve, Tran Kim Tinh, Do Phuoc Hau and M.E.F. Van Mensvoort (1990). Sulfidic materials in the Western Mekong delta, Viet Nam. In press. Paper presented at the Dakar Symposium on Acid Sulphate Soils, January 1986. Broadus, J . , J. Milliman, S . Edwards, D. Aubrey and F. Gable (1986). Rising sea-level and damming of rivers: possible effects in Egypt and Bangladesh. p 165-189. In: Titus, J.G. (Ed.). Effects of changes in stratospheric ozone and global climate. Vol. 4: Sea level rise. UNEP/U.S.EPA. Coutellier, V., and D.J. Stanley (1987). Late Quaternary stratigraphy and paleogeography of the Eastern Nile delta, Egypt. Marine Geology 77: 257-275. Day, J.W. and P.H. Templet (1989). Consequences of sea level rise: implications from the Mississippi delta. Coastal Management 17: 241-257. Hekstra, G.P. (1989). Global warming and rising sea levels: the policy implications. Ecologist 19: 4-15. Jelgersma, S . (1988). A future sea-level rise: its impacts on coastal lowlands. p 61-81 in: Geology and urban development. Atlas of urban geology, Vol. 1 . UN-ESCAP. Morgan, J.P. (1970). Depositional processes and products in the deltaic environment. p. 3147. In: Morgan, J.P. and R.H. Shaver (Eds). Deltaic sedimentation, modem and ancient. Special Publ. No. 15, SOC.of economic paleontologists and mineralogists, Tulsa, Oklahoma. Poland, J.F. (Ed.) (1984). Guidebook to studies of land subsidence due to groundwater withdrawal. Studies and Reports in Hydrology No. 40. Unesco, Paris. Stanley, D.J. (1988). Subsidence in the Northeastern Nile delta: rapid rates, possible causes, and consequences. Science 240: 497-500. Titus, J.G. (1987). Causes and effects of sea level rise. p 125-139. In: Preparing for climate change. Proc. of the first North American conference on preparing for climate change:
156
H . Brammer and R . Brinkman
a cooperative approach. October 27-29 1987, Climate Institute, Washington, D.C. Government Institutes, Inc., Rockville, MD. UNEP (1988). Report of the joint meeting of the task team on implications of climatic changes in the Mediterranean and the co-ordinators of task teams for the Caribbean, Southeast Pacific. South Pacific, East Asian seas and South Asian seas regions, Split, 3-8 October 1988. UNEP(OCA)/WG.2/25. Volker, A. (1989). Impacts of climatic changes on hydrology and water resources of coastal zones. Vol. 1. p 114-127. In: Conference on climate and water, Helsinki 11-15 Sept. 1989. Publ. Acad. Finland, Govt. Printing Centre, P.O.Box 516, SF 00101 Helsinki. Zagwijn, W.H. (1986). The Netherlands in the Holocene. (Dutch; Nederland in het Holoceen. Geologie van Nederland, Dee1 1). Geological Service of the Netherlands, Staatsuitgeverij, 's Gravenhage. 46 p, separate map sheet.
157
Chapter 13
SOILS ON A WARMER EARTH: THE TROPICAL REGIONS W.G. Sombroek International Soil Reference and Information Center, P. 0. Box 353. 6700 AJ Wageningen. the Netherlands.
ABSTRACT After discussing the vagary of the trend in global warming during the past hundred years, an overview is given of the several predictive models of climatic change in the forthcoming century at a doubling of the atmospheric C 0 2 content, with special reference to the tropics and subtropics. The positive effects of rising atmosphere CO, and near surface temperatures on plant growth are outlined; more freshwater for irrigation due to higher evaporation over the oceans; more biomass production, especially of C3-plants; more efficient water consumptive use of plants because of lower stomata1 conductance. Then a discussion follows on the effects of climatic change on soil development and conditions in various upland regions of the tropics and subtropics, taking the outcome of the NCAR modelling of 1984 as starting point (2°C temperature increase; varying seasonal increase or decrease of soil moisture availability), superimposing that on present day climatic conditions and soil patterns. The coastal lowlands and the mountainous cases, where impact of climatic change may be largest, are treated separately.
INTRODUCTION The greenhouse function of the earth's atmosphere can be described as the trapping of part of the thermal emission from the earth's surface, because of the presence of a number of gases. The most important greenhouse gases are water vapor (H20), carbon dioxide (COz), methane (CH4), nitrous oxide (NzO), ozone ( 0 3 ) and chlorofluorocarbons (CFC's). The greenhouse function is an essential part of the life support system of planet Earth. In its absence there would be no life as we know it, because mean global annual temperatures would be minus 18°C. An enhanced greenhouse function of the atmosphere, caused by anthropogenic increase of the above greenhouse gases, is commonly known as the GreenHouse Effect (GHE), or the Arrhenius Effect after its first postulator in 1896. A definite increase since 1850 has been measured in the atmospheric concentration of COz (from 290 to 345 ppm) and CH4 (from 0.85 to 1.7 ppm). This has coincided with a trend towards global warming of about 0.5"C (global mean annual near-surface temperature) in the period 1880-1990 (Hansen et al.
158
W . C . Sombroek
1981 and 1987). Part of this gradual increase, viz. 0.2"C, has recently been ascribed to "urban warming" at some key observation points (Karl et al. 1988). The series of temperature measurements may also be misleading due to changes in the methods of ocean surface temperature measurements in the past hundred years (for instance a change-over from "bucket" to "engine" measurement of ocean surface temperature). Long-term measurements in parts of Europe (Scandinavia, Holland) show no increase at all or even a slight decrease in local temperatures over the period considered. At global level there is a peculiar interruption in the period 1940-1975 when a slight downward trend occurred. This may well be caused by changed solar activity (sun spots) as already mentioned by Scharpenseel in Chapter 1; during the period concerned many researchers speculated on the imminence of a new ice age. Notwithstanding the observed increase in atmospheric C02 and CH4, the rise in global temperatures over the past 100 years is slow and irregular and even non significant statistically; it would be "noise" rather than a "signal". Hansen et al. (1988) estimated that only by the mid nineties a definite signal on temperature rise from the combined effect of all anthropogenic greenhouse gases can be expected.
CLIMATIC CHANGE MODELS There is little doubt that the concentration of anthropogenic greenhouse gases will increase substantially in the forthcoming 50-100 years. A doubling of the atmospheric preindustrial C02 content to about 580 ppm by the year 2050 is not unlikely. Therefore, atmospheric scientists and climate modelers have combined the 0.5"C overall temperature rise in the past 100 years, uncertain as it may be, with the measured atmospheric C02 rise of 55 ppm in the same period, using algorithms of atmospheric processes, to make projections of future climates. A doubling of the atmospheric C02 has been taken as scenario for the development of a number of General climatic Circulation Models (GCMs, see Table 13.1 for the results of five of them). The temperature rises predicted by these models vary between 2.8 and 5.2"C worldwide. However, there is much discrepancy between the various modelling results. Even the present day climate is not simulated satisfactorily (Mitchell 1989). The scientists concerned use varying estimates of the influence of feedback mechanisms by oceans (surface phenomena and deep water upwelling), by cloudiness, by ice coverage, by land cover and land use, by soils, etc. Terms used in this respect are: "negative feedback" which is mitigation of the GHE, and "positive feedback" which is strengthening of the GHE. Most models are of the "steady state" type, which simulate a new equilibrium situation after a sudden increase of atmospheric C02 from 300 ppm to 600 ppm. Models of the "transient response" type which include a year-by-year cumulative time path, with due account of temporal lag
Soils on a warmer earth: tropical regions
159
effect (such as delayed ocean warming) as well as short term effects (such as increasing plant assimilation), are more complicated than steady state models. Comprehensive reviews of the GCM results and their drawbacks have been given by Dickinson (1986) and Mitchell (1989). Table 13.1 Global m a n changes infive recent C02-doubling studies
Study GISS NCAR GFDL
UKMO
osu
Source Hansen et al. (1984) Washington and Meehl(l984) Wetherald and Manabe (1986) Wilson and Mitchell (1987a) Schlesinger and Zhao (1987)
Surface Temperature Change, K 4.2
Precipitation Change, % 11.0 7.1 8.7
4.0 4.0 5.2
15.0
2.8
7.8
Goddard Institute for Space Studies (Hansen et al. 1984). Seasonal data on a 7.83" latitude x 10" longitude grid: mean surface air temperature and precipitation anomalies for (2 x C 0 2 - 1 x C02). GFDL Geophysical Fluid Dynamics Laboratory (Wetherald and Manabe 1986). A 4.5" x 7.5" grid: mean surface air temperature and precipitation anomalies for (2 x C 0 2 - 1 x C02). NCAR: National Center for Atmospheric Research (Washington and Meehl 1984). Seasonal data on a 4.5" x 7.5" grid: mean surface air temperature and precipitation anomalies for (2 x C 0 2 - 1 x C02). OSU: Oregon State University (Schlesinger et al. 1985). Seasonal data on a 4" x 5" grid: mean surface air temperature and precipitation anomalies for (2 x C 0 2 - 1 x C02). UKMO: UK Meteorological Office (Wilson and Mitchell 1987a). Monthly data on a 5" x 7.5" grid: mean surface air temperature and precipitation for control run (1 x C@) and perturbed run (2 x C02). from: Hulme and Mitchell (1989). GISS:
All GCMs predict a much stronger temperature rise at higher latitudes in the respective summers than near the equator. The maximum steady state responses at C02 doubling are upward from 5°C annual mean surface air warming for the high latitudes, but only 2°C for the equatorial regions. The latest models, presumably including not only C02 doubling but also the longer and stronger action of increases of trace gases such as methane and nitrous oxide, are said to predict a global rise in temperature of 5°C. with increases up to 12°C in polar regions (Press 1989). The magnitude of such a change would be comparable to the change from a glacial to an interglacial period in the Pleistocene. On the other hand, within the Intergovernmental Panel on Climate Change (IPCC) a consensus emerged at a recent meeting that mean global temperatures will have gone up by only 2°C by the year 2050 (Anderson 1990).
160
W.G. Sombroek
Associated with such a global warming would be a sea level rise of 0.5-1 m, caused by temperature induced expansion of surface water volume and strongly increased melting of ice in polar regions. These predictions are also fraught with uncertainties, as discussed by Pirazzoli (1989). A rise of only about 0.3 m would be the best estimate according to a recent review in Nature (Meier 1990). The global precipitation would increase by 7-15% (Mitchell 1989), though the effect on freshwater availability would be largely offset by increased evapo(transpi)ration. The precipitation pattern would also change: a smng increase is predicted for the Northern high latitudes; non change or decrease in the mid latitudes and minor increases in the tropics. Estimates on the geography of annual and seasonal precipitation vary widely between the GCMs. Often the predicted changes would not be more than "noise" (Schlesinger and Mitchell 1985), because of problems in modelling of the seemingly chaotic regime of the weather, particularly in the Northern higher latitudes (North America, Western Europe). For the tropics the scarcity of reliable long term weather records is a serious drawback. Some models give the net effect per region of changing seasonal rainfall patterns and increased evapo(trans)piration as expressed in soil moisture availability, though not taking into account changed runoff conditions. An example is the NCAR model (see Fig. 13.1) described in Washington and MeeN (1984). It gives a degree of winter and strong summer drying in the Northern mid
latitudes (Southern USA, Indochina) but more soil moisture in the Northern high latitudes, especially in the winter. In the tropics, the model outputs on seasonal soil moisture changes are complicated: Little or no change in the year-round soil moisture conditions in the monsoon climate of Southeast Asia; Up to 30 mm increase in soil moisture in December-January-February (DJF) for Eastern Amazonia but up to 30 mm decrease for Western Amazonia and Central Brazil; a 10-20 mm increase in eastern Amazonia and central Brazil in June-July-August (JJA) which is at present a dry season in that region, but no change in Westem Amazonia (no dry season at present); Little change in DJF for Westcentral Africa but a decrease of up to 40 mm in JJA, nowadays the middle of the wet season; and Up to 40 mm decrease in available soil moisture in Eastern Africa for DJF (at present already the "long" dry season), and 20-30 mm increase in JJA (at present the "short" dry season). The implication of this is the attenuation of the bimodal pattern of rainfall effectiveness in this region ("long rains" moving more smoothly into the "short rains") but a more pronounced (long) dry season - not considering effects of
Soils on a warmer earth: tropical regions
161
local orography; in Southeastern Africa however, with a single rainy season in DJF, this rainy season would be less pronounced.
SOIL MOISTURE DIFFERENCES, JJA
_,._ -,\.,-.-..---’
________*
Fig. 13.1
Geographical distribution of soil moisture differences in cm for DecemberFebruary and June-August obtained by Washington and Meehl(l984)for 2 x CO2 - I x C02 (NCAR model)
Whatever the merit of the NCAR model, on its own and in comparison with
162
W . G . Sombroek
the other models, - for instance the one by Manabe and Wetherald (1986), as reproduced in Chapter 4,- the above predicted regional changes in soil moisture will be taken as starting point for the discussion of the effects on tropical soil formation and conditions in this paper. First, however, the effect of changing atmospheric C02 and temperature on tropical plant growth will be discussed.
EFFECTS OF RISING TEMPERATURE, ATMOSPHERIC CO2 AND SEA-LEVELON PLANT GROWTH IN THE TROPICS AND SUBTROPICS An increase of global mean annual near surface temperatures will stimulate plant growth in the world-at-large, other conditions remaining the same. A 1°C increase would result in 5 1 0 % extra biomass production worldwide (Idso et al. 1987). The model-surmised higher temperatures will be reflected in the Holdridge Classification of biotemperatures, resulting in changed patterns and acreages of its "life zones" or schematized biomes. The modeling results of Emanuel et al. (1985) show only minor shifts in the main tropical life zone belts ("tropical wet forests", "tropical dry forests", etc.) but it should be noted that any changing rainfall patterns have not been taken into account in that study. In the tropics and subtropics, where temperatures are normally not a limiting factor, the increase in annual biomass production will be modest and grain yields of some major agricultural crops may decline if temperatures happen to surpass 24"-27°C (see Chapter 22). Only in mountain areas such as the Andean Cordillera and the highlands of Eastern Africa the change will be spectacular: the successive natural vegetation or crop growing belts will shift upwards by 1000 feet at an increase of 2°C. In these areas more land would be vegetation covered, though the net acreage of individual altitudinal crop-belts, such as coffee and tea, will become smaller (Ackland 1971 for East Africa, and Braun's report on the Agroclimatic Zone map of Kenya, in Sombroek et al. 1982). A preliminary case study by UNEP-GEMS-GRID for Uganda has shown that suitable conditions for Robusta coffee in Uganda would decrease in acreage to near zero (GRID 1987) assuming no gradual adaptation by introduction of new crop cultivars. The changing acreages are exemplified in Fig. 13.2. Increase in near surface temperatures over oceans will increase evaporation substantially. The present-day annual evaporation of all oceans is 4.4 x 1020 cm3y-', at a global mean annual surface temperature of I1.5"C (Holland 1978). An increase in temperature of only 1°C worldwide would result in about 20% extra evaporation (Handbook of Chemistry and Physics, 1987 edition) or 1.0 x 1020cm3y-', (10,OOO km3 water; compare with 85 km3 annual discharge of the Nile river). The resulting extra water vapor will induce more cloud formation and ultimately more rain- and snowfall. Part of the increased rain- or snowfall will be on the land surface, enlarging the present body of fresh water in rivers and lakes
Soils on a warmer earth: tropical regions
163
and the availability of water for plant growth. Models of the world hydrologic cycle are, however, not yet able to predict how much of this increased fresh water body will accumulate in subhumid and semiarid tropical land areas where it could contribute to increased crop production through irrigation. The effect of a rising sea level will be dramatic, especially in the tropics. Many parts of tropical coastal plains and river deltas such as those of the Mekong, the Sunderbans, the Congo, the Nile, the Amazon and the Orinoco will become flooded to such a depth that natural vegetation and crop growth will be impossible - unless huge empoldering works are executed (see Brammer and Brinkman, Chapter 12, for details). This loss of landbased plant growth will only partially be offset by plankton and algae growth in coastal waters. Interior parts of deltas will have poorer drainage conditions.
Fig. 13.2
Upward changes in altitudinal cropping c.q. vegetation belts in the tropics at a 2 "C increase of annual temperature (a = mountains of Eastern Africa: relatively small decrease in acreage; b = high lands of Uganda: near disappearance of the belt concerned)
In contrast, the positive effect of rising atmospheric C02 levels on landbased biomass production ('TO2fertilization") will be significant. This effect has not yet been duly incorporated in the GCMs. As already mentioned by Scharpenseel (Chapter l), laboratory experiments on a range of crops at doubled atmospheric C 0 2 level have shown an average increase of about 30% in vegetative productivity (Kimball 1983; Goudriaan and Unsworth 1988; Schleser and Kirstein 1989). In natural environments, light and C02 are normally suboptimally present. Hence, photosynthesis of green plants can be stimulated by an increase of ambient C02, by suppression of photorespiration particularly in C3-plants. These plants include nearly all leguminosae and woody plants, and many crops such as cotton, rice, wheat, soya beans, barley, sunflower, and tuber crops (cassava, potatoes). The reaction to increased CO2 is less strong, about lo%,in C4-plants, such as many tall tropical grasses, halophytes and the crops maize, sugar cane, sorghum and millets. CAM-plants such as pineapple, Opuntia and agaves show an intermediate reaction. Another effect of higher atmospheric C02 concentrations is a reduction of
164
W . G . Sombroek
plant stomatal openings, hence smaller water vapor diffusion (higher stomatal resistance). This will increase the plant's water use efficiency, and the process would be particularly strong in some C4-plants such as maize (37% smaller stomatal conductance, leading to a 26% lower transpiration -Cure 1988). The combined result of modest extra assimilation and much smaller transpiration of this crop would also be in the order of 30%, comparable to the pure assimilation effect on C3-plants (see also Schleser and Kirstein 1989). The likelihood of increased dark respiration (dissimilation) because of higher night temperatures may, however, have a mitigating effect. In this context, it is likely that the measured gradual increase in atmospheric C 0 2 over the past 100 years of about 20% is responsible for part of the approximately doubling of the agricultural crop productivity worldwide over that period (Allen et al. 1985, estimated the C02-induced increase to be 15% for soya). Such a CO2-induced increase of terrestrial net primary production would also account for the "missing" sink in the global carbon cycle, resulting from the smaller reported CO;! uptake in ocean surface waters of the Northern hemisphere than assumed previously (Tans et al. 1990). Doubling of C02 in combination with a 3°C mean global air temperahm rise would result in another 25% extra biomass production (Idso et al. 1987). For the tropics the combined effect would be smaller because of:
Less than average predicted temperature rise; The chance that plants become overheated because of too limited evapotranspiration - especially in dry season regions; The likelihood that daylight will become a limiting factor - especially in the below canopy growth environment of standing vegetation and crops; and The likelihood that soil nutrients become limiting - especially in the humid tropics where strongly weathered soils are widespread. Higher C02 concentrations would, however, also stimulate root growth which, combined with more organic matter production and its even stronger decomposition, would intensify rock weathering (see below) - hence deeper soils with extra availability of some nutrients such as potassium from the freshly weathered substratum. There is yet another factor: higher COz concentrations would induce shorter growing periods to maturity for many annual plants (Oecher and Strain 1985), which implies increased possibilities for plant growth in (sub)tropical regions with a short rainy season (plus less sensitivity to soil salinity), and more crops per year in humid areas. Finally, several indirect effects should be mentioned: 9
Increased biomass production leads to better soil cover, hence less erosion (unless the seasonal or annual ratio between rainfall and
Soils on a warmer earlh: tropical regions
165
evaporation changes for the worse); The competitive force of perennial grassy weeds - predominantly C4plants - relative to many crops would become smaller at higher C 0 2 levels, hence weed control would be less costly (the situation becomes more complex in case the crops concerned are C4-plants too, such as sorghum and millets in the Sudano-Sahelian zone of West Africa or semiarid India). In summary, atmospheric C02 doubling and related temperature rise on their own are likely to have a definite positive effect on plant biomass production in tropical forests (C3-plants), less so in tropical savannas (C4-plants), and a significant increase of production potential of most agricultural crops. The demand for fertilizers, especially the nitrogen and phosphorus based ones will, however, strongly increase. It also remains to be established whether other biological influences such as the activity of pathogens, pests, decomposers and symbionts will change at the same or higher rate as biomass productive capacity per se. In contrast with C02, the rising concentrations of atmosphere trace gases CH4 and N20 have no effect on plant growth. In view of their stronger and longer term forcing of the surmised global temperature rise, and their chemical reactivity, they may even turn out to be the most worrisome eventually. One can however imagine systems of "harvesting" CH4 gas in paddy fields and garbage piles for local fuel supply.
EFFECTS OF AN ENHANCED GREENHOUSE EFFECT ON SOIL CONDITIONS AND SOIL DEVELOPMENT IN THE TROPICAL AND SUB-TROPICAL UPLANDS A distinction should be made between soil features that have a short "response time" (RS) and those that change only very gradually (rapidly adjusting soil features vs. slowly adjusting ones). Walker and Graetz (1989) give three RS classes: short ( 4 0 years), long (50-1000 years) and very long (>lo00 years). A more detailed scheme is given in Chapter 4 by Varallyay. It is obvious that in soils the moisture, temperature and the organic matter and nutrient/salts dynamics are easily changed. Within the time frame of Global Change, say 50-100 years, this would not normally affect pedological soil classification, since the current systems are based on more stable/static soil characteristics (a partial exception is the US Soil Taxonomy system, where soil moisture and partly also soil temperature regimes figure at a high categoric level). The modest surface air temperature rise surmised for the tropics (maximum of 2°C) is likely to be even smaller in the soil, because of the denser vegetative covers (higher biomass production because of higher atmospheric CO;?, see
166
W.G. Sornbroek
above). With regard to soil moisture and salinity, much depends on the regional changes in the rainfall and evaporation patterns, but here the models are vague and contradictory.
Changes in dominant soil forming processes Soil forming factors acting slowly, dominant soil forming processes within the Global Change time frame of 50-100 years would not change significantly. There are, however, some fragile ("ecotonal" or "threshold") soil situations, where a slight change in one of the factors would induce a major change in the dominant process ("ferralitization" vs. "podzolization"; "illuviation" vs. "homogenization"; "salinization" vs. "leaching"), without even considering landscape processes such as wind and water erosion. Examples of ecotonal soil situations where a transition to a different genetic soil group may occur within a time span of tens of years, upon only slight changes of climatic conditions or somewhat higher frequency of extreme events are the following: a)
The lighter textured yellowish ferralitic soils of eastern Amazonia, the Congo basin and Kalimantan (xanthic Ferralsols and ferralic Arenosols in the FA0 terminology) may become rapidly podzolized upon small changes in precipitation (total annual amount or more seasonal concentration) and/or increasing acidic organic matter production, resulting in Podzols or albic Arenosols ("Giant Podzols") of which gleyic variants are already present in large areas in the Rio Negro-Rio Branco area of Northwestern Amazonia. The fragility of the present ferralitization process in the Amazon region has been amply demonstrated by detailed studies on short distance sequences of ferralitic soils-Podzols by French pedologists (for instance Turenne (1975) for French Guyana; Dubroeucq and Volkoff (1988) for the Brazilian Rio Negro area; Lucas et al. (1987) for the Amazon in general). The predicted increase of 30 mm effective soil moisture in the NCAR model during the dry season of Eastern and Southern Amazonia may be just enough to induce podzolization over large areas of sandy Pleistocene terraces - where up till now Podzols are of patchy occurrence only (Sombroek 1966; Camargo 1981).
b)
The imperfectly drained loamy soils on the flat water divide areas of the Western Amazon with their plinthitic subsoil (the Madeira-Purus area Southwest of Manaus) may dry out, resulting in an irreversible hardening of this subsoil into "laterite" (change from Plinthosol or plinthic Acrisol into Acrisol, petroplinthic phase - see Camargo 1981; Sombroek 1984) (Fig. 13.3). The decrease in soil moisture availability in DJF predicted by NCAR may induce such drying out, especially if the overall river hydrology would become more erratic.
167
Soils on a warmer earth: tropical regionr
Norlhweslern region (Rlo Negro) Southwesternregion (Rlo Purus)
Eastern region (Lower Amarmas)
(A)
t + t
......, (i'..
----
--a-
_-........, ---
grsnlle oulcrop
LP
sandy sediments
FRX
xanfhlc Ferralsols a 0
loamy to clayey sedimenls
ARa
albic Arenosols
pllnlhlle
PZh
deep humic Podzols
---
pelropllnlhlle
Leplosols
I
'Giant' Podzols
or15IeIn
PZg
gleylc Podzols
hlgh (ground) water level
PTd
dystrk Pllnlhosols
GL
Gleysols undiHerenlialed
seasonally low (ground) water level
--
expanslon of area at cllmallc change
(c)
Fig. 13.3
area wlth palchy occwenca 01 'Gianr Pcdrols
Schematic presentation of some Amazonian ecotonal soil situations
c)
The deeply weathered reddish, loamy to clayey and porous soils of the forest-savannah transition zones of Eastern Africa, which are stable under their present day natural vegetation, may be leached to an extent that a relatively dense clay illuviated horizon develops below an unstable topsoil low in organic matter (changing the soil from rhodic or orthic Ferralsol into an orthic Acrisol or Lixisol). This in fact is happening already at present day land clearing in parts of the region, because of the sudden diminution of the homogenizing action of soil biologic life. The NCAR predicted changes in soil moisture conditions for this area would therefore stimulate the replacement of ferralitization by illuviation as a dominant soil forming process, especially in the present bimodal rainfall areas (Sombroek et al. 1982).
(d)
Certain silty sedimentary deposits in the wide riverine valleys of the Sudano-Sahelian zone of West Africa ("fadama's"; the interior delta of t k Niger river) ma) develop from Fluvisols into saline and/or sodified soils at an even minimal change in precipitation and flooding regimes - as exemplified by current human actions with the same soil-hydrological implications (Sombroek and Zonneveld 1971).
168
W . G . Sombroek
The NCAR model for this region predicts a decrease in precipitation during the rainy season, hence less suppression of salinity.
Changes in soil weathering rates In non ecotonal tropical upland soil situations, the pathway of soil forming processes would not change, but the weathering potential and rate may increase somewhat. The weathering potential at a doubling of the atmospheric C02 per se, as measured by the increase of the solubility of calcite will be discussed first. The present C02 concentration of the atmosphere is taken as 300 ppm or 0.03%. The C02 concentration in the soil gaseous phase is 10 to 100 times higher, because of root respiration. The low values would occur under dry and cool climatic conditions, the high values in continuously warm and moist soils such as in tropical rain forests where root activity is very high (when the soil condition is anaerobic, such as in paddy fields the value of soil C02 may be up to 300 times the air C02; Novozamski and Beek 1976). At a 10 times higher CO2 concentration in the topsoil ( ~ 0 . 3 % )the equilibrium pH of water would be 5.22 (in the absence of any mineral component). With a doubling of the C02 content such as in the GCM scenarios, the C02 concentration in the topsoil would become: 0.30 (original) + 0.10 (33% increase of biomass production) + 0.03 (increase of atmospheric C02) = 0.43%. The corresponding equilibrium pH would be 5.14 l , meaning 20% more H-ions available for protolytic weathering (Fig. 13.4). This can be illustrated by the increased solubility of calcite. In the present situation 113 mg CaC03 1-l can be dissolved by soil water whereas in the future this would be 128 mg 1-l (Fig. 13.5) which means an increase of about 15%.In practice, however, the increase may be smaller than lo%, because the 33% increase in biomass production may not be reached, due to likely limitations of nutrients and light - as discussed before. When the C02 content of the topsoil is taken to be 100 times the atmospheric content (e.g. tropical humid forest) then the solubility of CaC03 increases from 251 mg 1.' to 277 mg 1-I, which is an increase of ca. 10%. Thus, taking into account a slight increase of the effect of rate limiting factors at a higher rate of reaction (e.g. diffusion of components), an average increase of 10% in the rate of weathering due to a doubling of atmospheric CO;!can be expected. The increase in temperature will also lead to a higher rate of weathering.
'
The lowering of pH to 5.14 includes the slight effect of a 2°C temperature increase. If buffering
minerals are present (which is usually the case in soils) then the equilibrium pH will of course be higher.
On h e other hand, this will
be
counteracted by an increase of h e content of organic acids.
Soils
on
a warmer earih: lropical regions
169
%
co,
Fig. 13.4 p l l of water as a function of 3' 6 C02 in air
% co,
Fig. 13.5 Solubility of CaC03 in water of 25 "c as a function of % C02 in air
There are indications that a temperature rise of 10°C results in an increase of velocity of mineral weathering reactions by a factor of approx. 1.5 (Qlo or "Van 't Hoffs factor"; Loughnan (1969), p.69; Zivkovic et al. 1983). A 2°C temperature increase would thus result in a 10% higher rate of weathering.
170
W . G . Sombroek
It follows that the combined effect of the increase of both the C02 content and the temperature on the rate and therefore the depth of Weathering can be estimated at a minimum increase of 20%. Of paramount importance, however, is the influence of the change in climate on the effective rainfall i.e. the water available for the reactions and transport of components. The estimated minimum rate increase of 20% will only occur if the effective rainfall does not decrease. The higher biomass production inevitably leads to a higher amount of organic decomposition products. In view of the somewhat higher soil temperatures, this decomposition of fresh organic matter is likely to be more complete, hence there may in fact be a smaller rather than a higher production of stable humus (see also Tinker in Chapter 7). At the same time the deeper soil profiles offer extra space in the subsoil for fixation of the humus on the newly formed silicate clay minerals. The net result may be a slow change in sesquioxide rich clayey tropical soils such as Nitisols and rhodic Ferralsols with humic topsoils (rnollic or umbric A horizons) into units with the stable humus distributed more evenly in the whole soil profile. A more prominent presence of organic acids by decomposition may favour organo-metallic complexing. This will particularly be the case where nutrient supply cannot keep pace with the increase in biomass production and a poorer type of organic matter (with a higher C/N ratio) is produced. The result may be a degree of podzolization where the soil is sandy and of relatively low sesquioxide content (sandy xanthic Ferralsols; see above). All above considerations refer to the (sub)humid tropics. In the semi-arid tropical and sub-tropical uplands the weathering potential, its rate and direction, depend very much on the seasonal and annual changes in soil moisture availability - for which the GCMs, as stipulated above, give unreliable predictions. Higher surface temperatures in open savannah like conditions (predominantly Cq grasses hence less increase in biomass production) would lead to faster combustion of soil humus already present, while soil fauna such as termites would carry more litter to their cool underground nests (which may lead to more methane emissions as a positive feedback to the Arrhenius effect). Whcre free salts are available at some depth in the subsoil these may be transported upwards due to stronger evaporation and evapotranspiration, leading to a degree of soil salinization. The likely combined result of a greenhouse warming on the semiarid tropics is a strengthened hazard of desertification. The statements given apply to whole continents or regions. The local pattern of soils in the tropics, often a reflection of former climatic conditions different from the present ones, coupled with the local implications of any anthropogenic climatic change in the near future, offer a large and intricate scala of possibilities of future soil development. This cannot be discussed in the framework of this general review on tropical soils on a warmer earth. The modeling of soil genetic
Soils on a warmer earlh: lropical regions
171
processes as proposed by Stewart in Chapter 10 may, however, be put to good use for individual soil profiles and their present day or surmised future environmental conditions.
EFFECTS ON SOIL CONDITIONS OF TROPICAL COASTAL LOWLANDS AND OF TROPICAL MOUNTAIN AREAS The tropical coastal lowlands and wetlands, as well as the tropical mountain areas (roughly all areas above 2000 m altitude), merit special attention. Changes are likely to be substantially more dramatic than in tropical uplands. The coastal wetlands will be affected principally by the expected rise of sea level of 0.5 to 1 m. Brackish water will penetrate in many coastal marsh lands and swamps. The brackish water, in combination with the high organic matter accumulation of these lands, will result in soil forming processes favouring the formation of potential Acid Sulphate Soils (thionic Fluvisols, - Gleysols, and Histosols). These are real problem soils, because they turn very compact and extremely acid upon artificial drainage, becoming unsuitable for most plant growth. Coastal lands at present only a few meters above flooding would have shallower levels of ground water, often saline, resulting in salinization and/or sodification of soil profiles especially where the climate is subhumid or semiarid. For details, also as regards low level islands with coral reefs such as prevalent in the Pacific, see Brammer and Brinkman in Chapter 12. Tropical mountain areas, as discussed before, will see a substantial shift of natural or agroecological zones in an upward direction, because of the predicted 2°C temperature increase. The intensity of rock weathering will increase and soils will become deeper. This implies more agricultural potential where the adjoining uplands remain humid, such as is the case on the Eastern slopes of the Andean Cordillera and the Northeastern slopes of the Sumatran mountain ridge. If however, the associated uplands become dryer, as may be the case in Eastern Africa, then the net agricultural area in the region may diminish. Whatever be the case, the surface and subsurface hydrology on the mountain slopes will change substantially, with a potentially strong effect on the drainage conditions of the lands and soils of the lower parts of the catchment areas concerned. The details of soil changes in mountain areas depend not only on differences in parent materials, but also on past shifts, upwards and downwards, of the belts of climate conditions and associated plant communities. Systematic and detailed paleoecological studies, such as those already carried out on several altitudinal transects in the Colombian Andes (Van der Hammen et al. 1983, 1984 and 1989) will provide important clues as to the future soil development in tropical mountain regions under anthropogenic climatic change.
172
W . G . Sombroek
ACKNOWLEDGEMENTS The author is indebted to Dr. Pieter van Reeuwijk, soil chemist of ISRIC, Wageningen, for his contribution and constructive comments.
REFERENCES Ackland, J.D. (1971). East African Crops. F A 0 - Longman. Allen, L.H., K.J. Boote, P.H. Jones et al. (1985). Response of vegetation to rising C02: Soybean Photosynthesis and Yield. In: Direct effects of increasing carbon dioxide on vegetation. DOE/ER-0238 Report p.175. US Dept. of Energy. Anderson, G.C. (1990). Climate Change; towards a 2°C consensus. Nature 343:401. Arnold. R.A., I. Szabolcs and V. Targulian (1990). Concept of Global Soil Science. Report of 1989 rIASA/ISSS/UNEP task force meetings. IIASA, Vienna (in press). Bolin, B., B.R. DOOs. J. Jager and R.A. Warrick (eds)(1986). The greenhouse effect, Climatic Change and Ecosystems. SCOPE 29. John Wiley & Sons, Chichester, 541p. Camargo, M.N. et al. (1981). Mapa de solos do Brasil, escala 1:5.000.000. EMBRAPASNLCS, Rio de Janeiro. Cure, J.D. (1988). Carbon dioxide doubling responses. In: The use of statistical climate-crop models for simulating yield to project the impacts of C 0 2 induced climate change. DOE/ER/60444-H1, July 1988, p.31. Dickinson, R.E. (1986). How will climate change? The climate system and modelling of future climate. In: Bolin et al. op.cit., 207-270. Dubroeucq, D. and B. Volkoff (1988). Evolution des couvertures pddologiques sableuses i podzols giants d'Amazonie (Bassin du Haut Rio Negro). Cah. ORSTOM, Strie P6dologie 24(3):191-214. Emanuel, W.R., H.H. Shugart and M.P. Stevenson (1985). Climatic change and the broad scale distribution of terrestrial ecosystem complexes. Climatic Change 7:29-43. Esser, G. (1990). Modeling global terrestrial sources and sinks of CO, with special reference to soil organic matter. In: A.F. Bouwman (Ed), Soils and the greenhouse effect, pp. 247-262, Wiley and Sons, Chichester (1990). F A 0 (1988). FAO/Unesco soil map of the world, Revised legend. World soil resources report 60, FAO, Rome. Goudriaan, J., and M.H. Unsworth (1988). Implications of increasing carbon dioxide and climatic change for agricultural production and water resources. Proceedings annual ASAE meeting, Los Angeles, USA (in press). GRID (1987). Uganda Case Study: A sampler atlas of environmental resource data sets within GRID. UNEP-GEMS, Nairobi. Handbook Chemistry and Physics (1987). 68th ed., CRC Press. Hansen, J., D. Johnson, A. Lacis, S. Lebedeff, P. Lee, D. Rind and G. Russell (1981). Climatic impact of increasing atmospheric carbon dioxide. Science 213:957-966. Hansen. J . and S. Lededeff (1987). Global trends of measured surface in temperature. J. Geophys.Res. 92:13345-13372, 1987. Hansen, J., I. Fung. A. Lacis, D. Rind, S. Lededeff, R. Ruedy and G. Russell (1988). Global climate as forecast by Goddard Institute for Space Studies; three dimensional model. J. Geophys. Res. 93:9341-9364. Holland, H.D. (1978). The chemistry of the atmosphere and oceans. Wiley & Sons, New York. p. 56-63.
Soils on a warmer earth: tropical regions
173
Hulme, M., and P.D. Jones (1988). Climatic change scenarios for the UK. (in press). Idso, J.B., B.A. Kimball, M.G. Anderson and J.R. Mauny (1987). Effects of atmospheric CO2 enrichment on plant growth: the interactive role of air temperature. Agric. Ecosys. and Environ. 20:1-10. Karl, T.R., H.F. Diaz and G.Kukla (1988). Urbanization: its detection and effect in the United States. Climate Record. J. Climate 1:1099-1123. Kimball, B.A. (1983). Carbon dioxide and agricultural yield: An assemblage and analysis of 430 prior observations. Agronomy Journal 75:79-88. Loughnan, F.C. (1969). Chemical weathering of the silicate minerals. American Elsevier Publ.Co. New York. Lucas Y., R. Boulaine, A. Chauvel and L. Veillon (1987). Systbmes sols ferralitiques-podzols en rtgion amazonienne. In: Podzols et podzolisation. D. Righi and A. Chauvel (Eds). AFES-INRA, Paris. 53-65. Manabe, S. and R.T. Wetherald (1986). Reduction in summer soil wetness induced by an increase in atmospheric carbondioxide. Science 232:626-628. Meier, M.F. (1990). Reduced rise in sea-level. Nature 343:115-116. Mitchell, J.F.B. (1989). The "Greenhouse" effect and climate change. Review of Geophysics 27(1): 115-139. Mooney, H.A., P.M. Vitousek and P.A. Matson (1987). Exchange of materials between terrestrial ecosystems and the atmosphere. Science 238:926-932. Novozamski I. and J. Beek (1976). Common solubility equilibria in soils. In: Bolt, G.H. and M.G.M. Bruggenwerth (Eds). Soil chemistry. a basic element. Developments in soil science 5A. Elsevier Amsterdam. p.69-125. Oechel, W.C. and B.R. Strain (1985). Native species responses to increased atmospheric carbon dioxide concentration. In: Direct effects of increasing carbon dioxide on vegetation. DOE/ER/0238 Report, p. 117.154. US Dept. of Energy Pirazzoli, P.A. (1989). Present and near future global sea-level changes. Global and planetary change 1(4), Elsevier Science Publishers, Amsterdam. Press, E. (1989). What I would advise a head of state about global change. Earthquest 3(2):13. UCAR, Bouldcr, U.S.A. Schleser, G and W. Kirstein (1989). Der Treibhaus Effect - Ursachen und Consequenzen fur Klima und Biosph5re. Seminar Technik und Gesellschaft fiir Journalisten, K.F.A Jiilich, 26-4- 1989. Schlesinger M.E., and J.F.B. Mitchell (1985). Model projections of the equilibrium climatic response to increased carbon dioxide. In: The potential climatic effort of increasing carbon dioxide, J.F.B. Mitchell, U.S. Dept. of Energy. Sombroek, W.G. (1966). Amazon Soils. A reconnaissance of the soils of the Brazilian Amazon, 292p. Pudoc, Wageningen. Sombroek, W.G. (1984). Soils of the Amazon Region. In: Sioli H., The Amazon, limnology and landscape ecology of a mighty tropical river and its basin. Dr.W. Junk Publishers. Dordrecht/Boston. p. 521 -536. Sombroek, W.G., H.M.H. Braun and B.J.H. van der Pouw (1982). Exploratory soil map and agroclimatic zone map of Kenya 1980. Exploratory soil survey report E l . Kenya Soil Survey, Nairobi. Sombroek, W.G., and I S . Zonneveld (1971). Ancient dune fields and fluviatile deposits in the Rima-Sokoto river basin (N.W. Nigeria). Stiboka, Soil Survey Papers 5. Wageningen. p.109. Tans, P.P., I.Y. Fung and T. Takahashi (1990). Observationed constraints on the global carbon dioxide budget. Accepted by Science. Turenne, P. (1975). Mode dhumification et de diffkrentiation podzolique dam deux stquences
174
W.G. Sombroek
guyanaises. Mem. ORSTOM 84. Paris. pp.173. Van der Hammen. Th., A.P. Preciado and P. Pento (1983). La Cordillera Central Colombiana, transecto parque Los Nevados (Introduccion y datos iniciales). Studies on tropical Andean ecosystems Vol.1. A.R. Gantner Verlag K.G., Berlin-Stuttgart. Van der Hammen, Th., and P.M. Ruiz (1984). La Cordillera Central Colombiana, transecto parque Los Nevados (Introduccion y datos iniciales). Studies on tropical Andean ecosystems Vol.11. A.R. Gantner Verlag K.G.. Berlin-Stuttgart. Van der Hammen, Th., S. Diza-Piedrahita and V.J. Alvarez (1989). La Cordillera Central Colombiana, transecto parque Los Nevados (Introduccion y datos iniciales). Studies on tropical Andean ecosystems Vol.III. A.R. Gantner Verlag K.G., Berlin-Stuttgart. Walker, R.H., and R.D. Graetz (1989). Effects of atmosphere and climate change on terrestrial ecosystems. Global Change Report no 5. Stockholm. Washington, W.M., and G.A. Meehl (1984). Seasonal cycle experiment on the climate sensitivity due to a doubling of C 0 2 with an atmospheric general circulation model coupled to a simple mixed-layer ocean model. J. Geophys.Res. 89:9475-9505. Zivkovic, D.Z..J.J. Crnko and G. Sretenovic (1983). Kinetics and mechanism of the acid kaolin leaching process. In: Augusthitis, S.S. (ed.), Leaching and diffusion in rocks and their weathering products. Theophrastus Publications. Athens, Greece. p.63-77.
175
Chapter 14
SOILS ON A WARMER EARTH Projecting the effect of increased CO, and gaseous emissions on soils in Mediterranean and subtropical regions
Dan H. Yaalon Institute of Earth Sciences, Hebrew University. Givat Ram Campus Jerusalem 91904, Israel
ABSTRACT The subtropical region is defined here as roughly the zone of 10' to 35" N and S of the equator, with a pronounced seasonal rainfall, with about 2 to 4.5 months in which precipitation exceeds evapotranspiration by 75%. Rain falls mainly in summer in contrast to the Mediterranean Region with a winter rainfall season. The area covered by the subtropics is around 18 million km2, mainly in Africa South of Sahara and smaller regions in Asia (India), Australia and South and North America. The area of the Mediterranean region cover close to 9 million km2 mostly in Southern Europe, North Africa and the Near East. General circulation models for the condition of a double C02 concentration are not consistent in predicting the moisture conditions. Some models predict a decrease in precipitation and soil moisture, especially in the summer months, which would result in a considerably change in the moisture regime both on cotenary slopes and in individual pedons. In general these are very fragile environments, sensitive to climatic change, partly because of the high variability and seasonality of the climatic parameters. The major soil groups in the two regions are Luvisols, Acrisols, Vertisols, Cambisols, Arenosols and Fluvisols. Six major soil responses can become evident within a short time of less than 50 years: 1) organic matter reduction; 2) carbonate and salt regime changes; 3) erosion; 4) crusting; 5 ) waterlogging; and 6) fire frequency. Among these the reduction of organic matter due to increased temperature should be most easily modelled and is of special importance because of a positive feed back effect on additional temperature increase. The soil carbonate regime and salt balance may be affected by both decreasing and increasing precipitation. Salinization on footslopes may be caused by increased rainfall upslope. Changes in intensity and seasonality of rainfall will result in increased runoff and erosion, in other cases surface crusting. Reduced infiltration in clayey soils may cause waterlogging downslope. Prolonged drought periods may cause an increased fire
176
D.H . Yaalon
frequency and result in subsequent severe erosion. The response action will depend not only on more precise prediction of the climatic parameters but also on the nature of the soils, their texture, structure, depth and spatial variability. Much detailed research on soil distribution and sensitivity to changes is needed. The basic soil processes are well known and understood, but their actual rates and possible feedbacks are known to a much lesser extent.
Note from the editors: Unfortunately Dr. Yaalon could not deliver the final paper in time due to unexpected hospitalization. The reader may wish to contact the author directly for more detailed information.
177
Chapter 15 IMPACT OF CLIMATE WARMING ON ARID REGION SOILS H.E. Dregne Texas Tech University, Lubbock Texas 79409. USA
INTRODUCTION Global atmospheric warming is generally expected to occur as a consequence of the increases in carbon dioxide and other greenhouse gases such as nitrous oxide, methane, chlorofluorocarbons, and tropospheric ozone. However, the imprecision of general circulation models, the uncertainty of the effect warming will have on ocean evaporation and global precipitation, and doubts about the validity of numerous assumptions that must be made, have raised questions about what is going to happen and where it will happen. Nevertheless, it is prudent to consider the impact of climate warming on natural resources, of which soil is an essential component. For the purpose of this discussion, the maximum global temperature increase is set at 3°C by 2050 and the mean annual precipitation change is set at plus or minus 10% of the current mean. The climate change is expected to be least in equatorial zones and greatest at high latitudes. Arid regions are commonly divided into three climatic zones: hyperarid, arid, and semiarid (Fig 15.1). Together, the three zones cover about one-third of the global land surface. Not shown on the map are the dry regions of the Arctic and Antarctic. In the latter regions, surface soils generally have properties similar to those of surface soils in the hot arid regions, but subsoils differ markedly. Changes in properties of soils in the hyperarid regions, due to climate warming, will not be discussed since any changes are expected to be insignificant due to the very low rainfall and near absence of natural vegetation. Just as profile development increases, when time is held constant, as one goes from the dry end of the arid climatic zone to the end of the semiarid climatic zone, so would the impact of climate change be expected to vary with mean annual rainfall as well as with temperature.
SOILS And region soils belong to six of the 11 orders of the U.S. Soil Taxonomy
178
2
I i 3 'tI PJ? . A
H . E . Dregne
2 0
3
2 ,
e
K
Fig. 15.1
Distribution of arid region climates
Impact of climale warming on arid region soils
179
(Aridisols, Alfisols, Entisols, Inceptisols, Mollisols, and Vertisols) and 14 of the 26 orders of the F A 0 classification system (Arenosols, Cambisols, Chemozems, Fluvisols, Kastanozems, Lithosols, Luvisols, Regosols, Solonchaks, Solonetz, Vertisols, Xerosols, and Yermosols). Most arid region soils are calcareous somewhere within the top 2 meter of soil; alkaline, neutral or slightly acid in the surface horizon; coarse textured sandy or gravelly on upland sites, sometimes covered with a desert pavement of gravels and stones (gobi, reg); fine texture in closed basins that frequently are saline as in the chotts and sebkhas of North Africa and the Middle East; variably textured in stream flood plains; and usually are only weakly developed morphologically. Within the dry regions large areas occur of gypsiferous soils, such as in Iraq, the Soviet Union, Syria and the United States. The most notable exception to the generality about arid region soils having calcareous horizons and an alkaline or slightly acid pH occurs in the semiarid African Sahel. There, where upland soils are permeable and rain is concentrated in a short period during the hot season, soils tend to be strongly acid, particularly in the subsoil. Similar conditions are said to exist in the Eastern part of Rajasthan State in India. One characteristic arid region soils have in common is a low level of organic matter. Organic matter content varies with rainfall and vegetative cover as well as with texture. Clay soils tend to be considerably higher in organic matter (humus) than sandy soils in the same climatic zone and organic matter usually is higher when the vegetative cover is relatively heavy. This means that upland sandy soils in the hot arid climatic zones are very low in organic carbon, typically containing less than 0.4%, whereas upland clay soils in the semiarid zones commonly have more than 1 % organic carbon and may contain 2% or more.
MEASURING SOIL CHANGES The principal problems with demonstrating that atmospheric temperature increases alter soil properties are sampling errors and difficulties in measuring soil erosion. Sampling errors in rangeland soils poses a special problem due to differences in density of vegetative cover during the year and the kind of vegetation present. Sampling a uniform cover of annual or perennial grasses during the growing season is quite different from sampling hummocky land or soils under shrubs and trees. Analytical methods are at least two or three orders of magnitude more accurate than sample collection methods. Independent sampling of one hectare for soil organic matter, for example, would not be expected to give results that are reproducible to within 20%. Salinity measurements are even harder to reproduce. In the case of water erosion, fairly elaborate and costly experimental plots must be constructed to provide reasonable accurate measures of soil loss from a
180
H.E. Dregne
small area. Determining water erosion on large fields with a good degree of accuracy is nearly impossible. Gully erosion commonly is estimated by measuring gully length as shown on aerial photographs. No one has yet devised a satisfactory technique for measuring removal and deposition of wind eroded soil on a field scale.
DATA COLLECI'ION METHODOLOGY A logical way to estimate the effect a 3°C rise in air temperature would have on soils would be to compare two areas where the physiography, parent materials, and rainfall were similar but there was a 3°C difference in the atmospheric temperature. Many such areas can be found in upland plains. One of them is the Texas High Plains, where the difference in mean annual temperature between Big Spring and Amarillo is 33°C. The two cities are approximately 400 km apart. Mean annual precipitation is 450 mm in Big Spring and 485 mm in Amarillo, the end of the semiarid climatic zone. If a 3°C increase in air temperature has readily detectable effects on soils, they should be observable here. A thick, fairly uniform eolian mantle covers most of the Texas High Plains. Immediately below the mantle is the Ogallala Formation, of Pliocene age. The Ogallala consists of calcareous fluviatile material composed of sand and gravel. The High Plains between Big Spring and Amarillo give the appearance of being absolutely flat. They are broken only by numerous closed basins known as playas. Soils are so similar that several belong to the same series and textural class. An extensive group of soils consists of Paleustalfs (Luvisols) of the Amarillo soil series. A comparison of those soils would seem to be ideal for determining the effect a 3°C temperature change would have on soil properties. Unfortunately, virtually all of the land in this part of Texas is cultivated. Half is rainfed and half is irrigated. Although the principal crops are only two, cotton and grain sorghum, many different management systems are practiced. Some farmers rotate their crops, some fertilize, some employ minimum tillage systems, some use terraces, some do not plant on the contour, some apply cotton gin trash (organic waste) to their fields, a few apply feedlot manure, etc. With all that management variability, it is almost impossible to find fields where everything is nearly the same except temperature. A critical observer could find several plausible reasons why soil organic matter differences, for example, were due to something other than higher or lower temperatures. Unger and Pringle ( 1 98 1 ) made intensive comparisons of the variability in one soil, Pullman clay loam, a Paleustalf, in the vicinity of Amarillo, Texas. Among the five samples from adjoining countries, organic matter content in the cultivated surface horizon ranged from 1.4% to 2.2%. Comparable differences were measured in the subsoils.
Impacl of climate warming
OR
arid region soils
181
Similar problems in lack of uniformity in land management can be expected in most semiarid regions, whether they are cropped or grazed. If topography, exposure, erosion, and other factors also vary, field comparisons may be of little value in arriving at definitive conclusions about climate change impacts until after many decades have passed. Small experimental plots will have to be monitored.
SOIL CHANGES AND LAND USE Approximately 90%of the arid and semiarid climatic zones of the world are uscd for grazing. Only about 6% arc utilized for growing rainfed crops. The remainder is used for irrigated agriculture, urban and rural settlements, and a number of other activities (Dregne 1983). Climate changes should have little effect on imgated soils. A slight effect on hastening organic matter decomposition may occur. It should not significantly alter soil properties, so those soils will not be discussed. According to Emanuel et al. (1985), a doubling of atmospheric carbon dioxide leads to increasing the area of global deserts from 20.6% to 23.8% of the land surface and a large increase in grasslands from 17.7% of the land to 28.9%. The estimates of changed area are based upon temperature increases that one of the general circulation models predicts from the doubling of carbon dioxide. Temperature increases predicted from the model are about 1.25"C at the equator and 3.75"C or more near the poles. In the Holdridge Life-Zone maps employed in estimating the desert and grassland area changes, a desert corresponds roughly to the arid climatic zone and the grasslands are similar to semiarid lands. In the Holdridge classification, desert precipitation is less than 250 mm and grassland precipitation is between 250 and 500 mm. The increase in grasslands would be at the expense of forcsts. The accuracy of that prediction is highly questionable.
Grazing lands Pastoralism is by far the dominant land use in the dry regions. It is practically the sole land use in the arid climatic zone and is dominant in the dry part of the semiarid zone, which it shares with rainfed cropping (dryland farming). Plant cover consists of a mixture of annual grasses, perennial grasses, forbs, shrubs, and trees. An estimated 80% of the world rangelands are at least moderately overgrazed (Dregne 1983). Of that 80%, threefifths is moderately degraded and two-fifths is severely degraded. Along with a reduction in perennial grasses by overgrazing may go a slight reduction in annual grasses, a large reduction in tree cover, an increase in unpalatable shrubs, and an increase in bare land. While total green biomass production may not decrease, the organic matter content of the soil very likely would decrease due to the lower production of perennial grasses and the increased bare space.
182
H.E. Dregne
An atmospheric temperature change of 3°C without a change in rainfall probably would have effects similar to overgrazing. Biomass production would be likely to decline significantly over the area as a whole, as the combined result of overgrazing and increased temperatures. Soil organic matter levels would also drop considerably on a percentage basis. The organic matter decline should occur nearly everywhere grazing is practiced. The decline would be difficult to detect in a few dccades because of the wide natural variability in daily and seasonal temperatures in the dry regions. A simultaneous 10%increase in mean annual precipitation would change things very little because the higher temperature would increase organic matter decomposition rates. A 10% decline would aggravate the organic matter loss. Runoff would increase, bringing more water erosion, probably in the form of gullies, on most soils as the vegetative cover at the soil surface became less. Wind erosion also would increase on upland sandy soils. Salinity should increase, as- well, although the effect on the salt content of chotts, sebkhas, salinas, and salares would be too small to detect against the large amount already present. Surface soil crusting should increase as the result of lower soil organic matter and more rapid surface soil drying. All of these effects would be minimal in the drier part of the arid climatic zone and greatest in the wetter grazing lands. This probably means that any soil changes would be undetectable in the drier arid regions as air temperature gradually became 3°C higher by the year 2050. In the semiarid rangelands, detectable declines in soil organic matter may occur. Increase in gully formation and wind erosion in drought years are likely. However, validating the increases would be difficult, given the normal high temperatures and dry conditions during droughts. Salinity changes and increases in surface crusting would be undetectablc. Rainfed cropland
Global warming, in the absence of any change in mean annual precipitation, would be expected to increase evaporation and reduce soil moisture (Manabe and Wetherald 1986), according to one scenario. That scenario also predicts enhancement of winter soil moisture. Changes would be greatest in certain regions in the middle and high latitudes. These regions include the Great Plains of Canada and the United States, Western Europe, Central Asia, and North Africa. Rainfed croplands commonly have lower organic matter contents than their uncultivated counterparts. The reason is less surface insulation by plant cover, summer fallow, harvesting of crops, little or no addition of fertilizers or manures, and the absence of a soil building crop in a rotation. A temperature rise in the cool or cold climatic zones in areas such as the Northern Great Plains, however, should increase production of cultivated crops and of native plants in rangelands
Impact of climate warming on arid region soils
183
(Williams et al. 1987). Increased plant production on rangelands should increase soil organic matter in these semiarid climatic zones. There probably would be a less pronounced increase, if any, in the rainfed croplands because organic matter decomposition rates would also increase. Of the other soil properties, dryland salinity may decrease slightly, the pH would change very little, and there probably would be no measurable change in base saturation and in calcium carbonate content or the depth at which the calcic horizon appears. There should be a slight decrease in surface soil crusting on both croplands and rangelands. Wind erosion would be substantially worse. Water erosion may increase slightly. In the hot semiarid regions, temperature rises will reduce soil organic matter levels, whether or not the rainfall changes upward or downward by 10%. There would be more surface crusting but no significant change in other characteristics. Polar arid regions Climatic conditions on the Antarctic continent probably are the most severe of any arid land area in the world. They are more severe than conditions in the Arctic, where most of the area is ocean. A major difference between Antarctic and Arctic soils is the virtual absence of organic matter in Antarctic soils. In fact, Tedrow and Ugolini (1966) note that, by common soil definitions which emphasize the biotic element, there are no soils in Antarctica. Other than the absence of organic matter, the soils of the dry (ice-free) valleys resemble temperate and region soils in the upper part of the profile. They are alkaline, are well supplied with basic elements, have very low amounts of nitrogen and phosphorus, generally are saline, and contain free calcium carbonate. A global 3°C rise in air temperature could increase the organic matter level by some small amount, in contrast to the decrease expected in temperate zone soils. In the warmer Arctic, soils commonly are alkaline, have a high base saturation, contain free calcium carbonate, are sometimes saline, and are very low in organic matter content (Tedrow 1974). They are underlain by permafrost at about 1 m. Global warming, which is supposed to be greater at the high latitudes than in the tropics, should increase plant growth and soil organic matter levels by a significant amount. Soils probably would have a somewhat lower salinity, base saturation, and calcium carbonate content. The soils would be thawed to a greater depth during the summer. Williams et al. (1987) estimated from one of the general circulation models (GISS) that a doubling of carbon dioxide would raise mean annual air temperature in Southern Saskatchewan, Canada (50"N latitude) by 4.7"C, a large amount. Presumably, the rise would be even greater in the Arctic.
184
H.E. Dregne
CONCLUSIONS In the tropical and temperate and climatic zone, where mean annual precipitation is less than about 250 mm and grazing is the dominant land use, a global warming of 3°C by the year 2050 should cause a modest decline in soil organic matter, whether precipitation increases or decreases by 10%. There should be a slight to moderate increase in surface soil crusting and, perhaps, a slight increase in salinity. Water erosion may increase somewhat but wind erosion should increase significantly. Conditions in the polar and regions will be different in the Arctic than in the Antarctic. In the latter region, warming may bring a slight increase in soil organic matter. By contrast, there should be a sharp increase in organic matter in Arctic soils, as well as a lowering of the permafrost layer. Semiarid grazing lands should experience a decrease in organic matter. Hot rainfed croplands will show less soil organic matter and no change in salinity or other properties. Cold rainfed croplands will have a modest increase in organic matter but no change in other properties. Wind erosion will increase more in the cold regions than in the hot regions. Water erosion may not change.
REFERENCES Dregne, H.E. (1983). Desertification of arid lands. Harwood Academic Publishers. New York. 242 p. Emanuel, R., H.H. Shugart, and M.P. Stevenson (1985). Climatic change and the broad-scale distribution of terrestrial ecosystem complexes. In Martin L. Parry (Ed.), The sensitivity of natural ecosystems and agriculture to