PERSPECTIVES FOR AGRONOMY Adopting Ecological Principles and Managing Resource Use
Developments in Crop Science Volume 1 Oil Palm Research, edited by R.H.V. Corley, J.J. Hardon and B.J. Wood Volume 2 Application of Mutation Breeding Methods in the Improvement of Vegetatively Propagated Crops, by C. Broertjes and A.M. van Harten Volume 3 Wheat Studies, by H. Kihara Volume 4 The Biology and Control of Weeds in Sugarcane, by S.Y. Peng Volume 5 Plant Tissue Culture" Theory and Practice, by S.S. Bhojwani and M.K. Razdan Volume 6 Trace Elements in Plants, by M.Ya. Shkolnik Volume 7 Biology of Rice, edited by S. Tsunoda and N. Takahashi Volume 8 Processes and Control of Plant Senescence, Y.Y. Leshem, A.H. Halevy and C. Frenkel Volume 9 Taigu Genic Male-Sterile Wheat, edited by Deng Jingyang Volume 10 Cultivating Edible Fungi, edited by P.J. Wuest, D.J. Royse and R.B. Beelman Volume 11 Sugar Improvement through Breeding, edited by D.J. Heinz Volume 12 Applied Mutation Breeding for Vegetatively Propagated Crops, by C. Broertjes and A.M. van Harten Volume 13 Yield Formation in the Main Field Crops, by J. Petr, V. Cern~, and L. Hrugka Volume 14 Origin of Cultivated Rice, by H. Oka Volume 15 Nutritional Disorders of Cultivated Plants, edited by W. Bergmann Volume 16 Hop Production, edited by V. Ryb~t~ek Volume 17 Principles and Methods of Plant Breeding, by S. Borojevi~ Volume 18 Experimental Morphogenesis and Integration of Plants, by J. Seb/mek, Z. Sladl~ and S. Proch~tzka Volume 19 Plant Tissue Culture: Applications and Limitations, by S.S. Bhojwani Volume 20 Weather and Yield, edited by J. Petr Volume 21 Plant Physiology, edited by J. Sebfinek Volume 22 Reproductive Adaption of Rice to Environmental Stress, by Y. Takeoka, A.A. Mamum, T. Wada and P.B. Kaufman Volume 23 Natural Rubber: Biology, Cultivation and Technology, edited by M.R. Sethuraj and N.M. Mathew Volume 24 Irrigated Forage Production, by A. Dovrat Volume 25 Perspectives for Agronomy, edited by M.K. van Ittersum and S.C. van de Geijn v
Developments in Crop Science 25
PERSPECTIVES FOR AGRONOMY Adopting Ecological Principles and Managing Resource Use Proceedings of the 4th Congress of the European Society for Agronomy, Veldhoven and Wageningen, The Netherlands, 7-1 1 July 1996 Edited by
M.K. VAN ITTERSUM Wageningen Agricultural University, Department of Theoretical Production Ecology, P.O.Box 430, 6700 AK Wageningen, The Netherlands
S.C. VAN DE GEIJN Research Institute for Agrobiology and Soil Fertility (AB-DLO), P.O. Box 14, 6700 AA Wageningen, The Netherlands
ELSEVIER
Amsterdam - Lausanne - New York - Oxford - Shannon - Singapore - Tokyo 1997
ELSEVIER SCIENCE B.V. Sara Burgerhartstraat 25 P.O. Box 21 l, 1000 AE Amsterdam, The Netherlands
Partly reprintedfrom the European Journal of Agronomy, Vol. 7/1-3
ISBN 0 444 82852 4 © 1997, ELSEVIER SCIENCE B.V. 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 B.V., Copyright & Permissions Department, P.O. Box 521, 1000 AM Amsterdam, The Netherlands. Special regulations for readers in the U.S.A.-This publication has been registered with the Copyright Clearance Center Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923. Information can be obtained from the CCC about conditions under which photocopies of parts of this publication may be made in the U.S.A. All other copyright questions, including photocopying outside of the U.S.A., should be referred to the copyright owner, Elsevier Science 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 book is printed on acid-free paper Transferred to Digital Printing 2006
Preface Since the Second World War, agricultural development has been characterised by a strong increase in land and labour productivity in large parts of Europe (Porceddu and Rabbinge, 1997), strongly stimulated by policy (e.g., Common Agricultural Policy of the EU). The level of self-sufficiency has been surpassed for nearly all agricultural products. The strong increase in productivity leading to the present situation characterised by over-production, has been attended by changes in farm structure, a decrease in agricultural employment, social, budgetary and economic problems at the regional and European Union level and by environmental deterioration. This situation contrasts with a situation of relatively low productivity, socio-economic and environmental problems in mainly Southern and Eastern parts of Europe, and with a major problem and at the same time a challenge for the next century at a global level: food production and security for a doubled population with a more affluent diet. Agricultural development has evolved from an activity with mainly one-dimensional, productivity aims, into a multi-dimensional issue with environmental, agricultural, economic and social objectives. Agriculture should adopt ecological principles and it should optimize the use of resources, i.e., agriculture should 'ecologize', and at the same time it should meet social and economic objectives. Agricultural development is an important issue within the framework of what is called sustainable development. The notion of sustainable development calls for explicit consideration of each of the mentioned objectives and for consideration of the problem at several aggregation levels. The situation in (a part of) Europe cannot be isolated from the situation at country or global scale; problems and solutions at field level should also consider, for instance, plant and crop rotation level.
Agriculture with such broadened objectives requires new systems at a range of aggregation levels. It requires different analyses, synthesis of knowledge and a different type of intervention at the policy level. Agriculture with broadened objectives requires a different agronomy. It calls for detailed knowledge concerning the functioning and production of agricultural plants and crops and their ecological relationships. In addition, it calls for synthesis and design of new ideotypes and genotypes, new production technologies, cropping systems, farming systems and agroecological land use systems. Basic knowledge at field, plant and lower levels of integration should be used and synthesized in the design of new systems at higher levels that meet a set of explicit objectives. These new systems should then be evaluated for their effectiveness at the various levels. This type of agronomic research will often be of an interdisciplinary nature with agronomists working together with breeders, physiologists, soil scientists, economists and sociologists. To fulfil this new role, agronomy has a range of sophisticated tools at its disposal. To fully exploit the potential of these tools, they should not be used separately, but in combination. A new agronomy should tailor the tools to the type of questions and benefit from the synergism of: empirical and experimental research, be it in laboratories, climate chambers, greenhouses or in the field to diagnose and analyse problems and to test new designs; mathematical modelling techniques to summarize knowledge, to test hypotheses and to identify knowledge gaps; - prototyping on experimental and commercial farms to design and implement new crop rotations and farming systems; and - model-based explorations to improve systems -
-
vi
understanding and identify a wide range of options. During the Fourth Congress of the European Society for Agronomy, held in Veldhoven-Wageningen, The Netherlands, 7-11 July 1996, the new perspective for agronomy emerged. Various keynote addresses, session themes, and oral and poster contributions demonstrated the need for a new role of agronomy and its tools (Van Ittersum et al., 1996). The special issue of the European Journal of Agronomy and the Proceedings Book of the Fourth ESACongress (Van Ittersum and van de Geijn, 1997) present a set of case studies illustrating the various agronomic tools that can be used for specific questions. The case studies are grouped in sections illustrating relevant subquestions in developing an agriculture with broadened objectives. The papers were selected such that the various subquestions were represented in the Proceedings. This implies a non-random sample from the contributions during the Congress, since the number of contributions addressing the level of cropping system, farm and agricultural land use was limited. Nevertheless, we think that agronomy should consider these levels of scale in its analysis and design because questions of stakeholders often concern these levels. After an introductory paper on the role of agronomy in research and education in Europe, the second section presents case studies addressing issues concerning agricultural land use, food security and environment. The next set of papers addresses crop physiological aspects in relation to growth factors such as radiation, CO 2, temperature and water. Experimental research and simulation modelling are used in mutual interaction. One important outlet for the generated and integrated knowledge is the ideotyping of crops. Improving resource-use efficiency in agriculture positively affects economic, environmental and agricultural objectives. Many papers presented during the Congress have directly or indirectly addressed this issue. A set of papers particularly focusing on nutrients and organic matter is presented in this volume. Again, a combination of experimental and modelling research is used to enhance understanding of the system and identify options for improvement. The final section addresses the design of integrated and ecological arable farming systems. Prototyping is
put forward as a promising tool to design and implement new farming systems. Indicators are presented that support evaluation of farming systems. Finally the contribution of model-based explorations in developing new farming systems is considered. A discussion follows on the notion that development of sustainable farming systems is not a matter of one-way research delivery, but rather a process in which researchers and target groups should cooperate, learn and develop in true interaction. This applies especially as the operationalisation of sustainability requires value-driven choices calling for a continuous interaction between society, its organisations and farmers on the one hand, and the scientists and designers on the other. We hope that the activities of the European Society for Agronomy and the Proceedings of its Fourth Congress will stimulate to serve the new perspectives of agronomy, i.e., to adopt ecological principles, to optimally manage the use of resources and to meet social and economic objectives.
Martin K. van Ittersum Wageningen Agricultural University Department of Theoretical Production Ecology P.O. Box 430 6700 AK Wageningen The Netherlands
Siebe C. van de Geijn Research Institute for Agrobiology and Soil Fertility (AB-DLO) P.O. Box 14 6700 AA Wageningen The Netherlands
References Porceddu, E. and Rabbinge, R., 1997. Role of research and education in the developmentof agriculture in Europe. Eur. J. Agron., 7: 1-13. Van Ittersum, M.K., Venner, G.E.G.T., van de Geijn, S.C. and Jetten, T.H. (Editors), 1996. Book of Abstracts, Volume I and II. Fourth Congress of European Society of Agronomy, 7-11 July, 1996, Veldhoven, The Netherlands, 736 pp. Van Ittersum, M.K. and van de Geijn, S.C. (Editors), 1997. Perspectives for Agronomy - Adopting Ecological Principles and Managing Resource Use. Elsevier Science, Amsterdam, in press.
vii
Acknowledgements The editors gratefully acknowledge logistic and technical assistance of Loes Helbers, Irene Gosselink and Guido Venner in processing the flow of manuscripts. We also sincerely thank the reviewers of the manuscripts for contributing to the maintenance of the scientific standard of the papers: D. Auclair L. Bastiaans F. Bonciarelli C.J.H. Booij M.G.R. Cannell J.G. Conijn J.B. Dent P. Dijkstra M. Donatelli P.A.I. Ehlert B. Gerowitt J. Goudriaan D.J. Greenwood P. Gregory J.J.R. Groot J. Hassink A.J. Haverkort G. Hoogenboom C. Jambert B.H. Janssen S.C. Jarvis J. Kubat P.J. Kuikman E.A. Lantinga D.W. Lawlor J.F. Ledent D.K.L. MacKerron L. 't Mannetje H. Meinke K. Mengel J.M. Meynard S. Mikkelsen M.I. Minguez
G.M.J. Mohren J.D. Mumford H. Naber J.J. Neeteson B. Nicolardot S.E. Ogilvy J.E. Olesen F.W.T. Penning de Vries J.R. Porter G. Russell M. Schenk H. Schnyder J.J. Schr6der C.A. Shand A.L. Smit J.-F. Soussana J.H.J. Spiertz E.A. Stockdale P.C. Struik H.F.M. Ten Berge N. Van Breemen P. Van Halteren H. Van Keulen E.N. Van Loo M.A. Van Oijen W.H. Van Riemsdijk J.A. Van Veen J. Vos D.C. Whitehead A.P. Whitmore F.G. Wijnands D. Younie J.C. Zadoks
This Page Intentionally Left Blank
ix
Table of contents Preface ............................................................................................................................................................. v Acknowledgements ................................................................................................... ................................ vii Section I INTRODUCTION
Role of research and education in the development of agriculture in Europe E. Porceddu and R. Rabbinge .......................................................................................................................
3
Section 2 AGRICULTURAL LAND USE, FOOD SECURITY AND ENVIRONMENT
Land use transformation in Africa: three determinants for balancing food security with natural resource utilization P.A. Sanchez and R.R.B. Leakey .................................................................................................................... 19
Agro-ecological characterisation, food production and security P. Bullock. ..................................................................................................................................................... 29
The potential benefits of agroforestry in the Sahel and other semi-arid regions H. Breman and J.J. Kessler. .......................................................................................................................... 39
Chemical crop protection research and development in Europe R. Neumann .................................................................................................................................................... 49
Emissions of CO2, CH4 and N20 from pasture on drained peat soils in the Netherlands C.A. Langeveld, R. Segers, B.O.M. Dirks, A. van den Pol-van Dasselaar, G.L. Velthof and A. Hensen .......... 57
Section 3 CROP PHYSIOLOGY AND IDEOTYPING
Effects of CO2 and temperature on growth and yield of crops of winter wheat over four seasons G.R. Batts, J.I.L. Morison, R.H. Ellis, P. Hadley and TR. Wheeler. ........................................................... 67
Use of in-field measurements of green leaf area and incident radiation to estimate the effects of yellow rust epidemics on the yield of winter wheat R.J. Bryson, N.D. Paveley, W.S. Clark, R. Sylvester-Bradley and R.K. Scott. ................................................... 77
Simulating light regime and intercrop yields in coconut based farming systems 3'. Dauzat and M.N. Eroy ................................................................................................................................. 87
Improving wheat simulation capabilities in Australia from a cropping systems perspective: water and nitrogen effects on spring wheat in a semi-arid environment H. Meinke, G.L. Hammer, H. van Keulen, R. Rabbinge and B.A. Keating. .................................................... 99
Comparison of CropSyst performance for water management in southwestern France using submodels of different levels of complexity C. O. Stockle, M. Cabelguenne and P. Debaecke ......................................................................................... 113
Root growth of three onion cultivars A.D. Bosch Serra, M. Bonet Torrens, F. Domingo Oliv~ and M.A. Melines Pagbs ...................................... 123
Interspecific variability of plant water status and leaf morphogenesis in temperate forage grasses under summer water deficit J.-L. Durand, F. Gastal, S. Etchebest, A.-C. Bonnet and M. Ghesqui~re .................................................... 135 Evaluation of sunflower (Helianthus annuus, L.) genotypes differing in early vigour using a simulation
model F. Agiiera, F.J. Villalobos and F. Orgaz ...................................................................................................... 145
Options of breeding for greater maize yields in the tropics A. Elings, J.W. White and G.O. Edmeades .................................................................................................. 155
Section 4 MANAGING RESOURCE USE
Nitrogen budgets of three experimental and two commercial dairy farms in the Netherlands J.J. Neeteson and J. Hassink ........................................................................................................................ 171
Resource use at the cropping system level P. C Struik and F. Bonciarelli ...................................................................................................................... 179
The efficient use of solar radiation, water and nitrogen in arable farming: matching supply and demand of genotypes A.J. Haverkort, H. van Keulen and M.I. Minguez .......................................................................................... 191
Soil-plant nitrogen dynamics: what concepts are required? E.A. Stockdale, J.L. Gaunt and J. Vos ......................................................................................................... 201
Modeling crop nitrogen requirements" a critical analysis C 0. Stockle and P. Debaeke ....................................................................................................................... 217
Maize production in a grass mulch system - seasonal patterns of indicators of the nitrogen status of maize B. Fell, S.V. Garibay, H.U. Ammon and P. Stamp ........................................................................................ 227 Nitrogen transformations after the spreading of pig slurry on bare soil and ryegrass ~5N-labelled ammonium T. Morvan, Ph. Leterme, G.G. Arsene and B. Mary ...................................................................................... 237 Size and density fractionation of soil organic matter and the physical capacity of soils to protect organic matter J. Hassink, A.P. Whitmore and J. Kub6t ....................................................................................................... 245 Characterization of dissolved organic carbon in cleared forest soils converted to maize cultivation L. Delprat, P. Chassin, M. Lin~res and C. Jambert. ...................................................................................... 257 Analysis of impact of farming practices on dynamics of soil organic matter in northern China H.S. Yang and B.H. Janssen .......................................................................................................................... 267 Agronomic measures for better utilization of soil and fertilizer phosphates K. Mengel. ................................................................................................................................................... 277 Section 5 DESIGNING FARMING SYSTEMS
A methodical way of prototying integrated and ecological arable farming systems (I/EAFS) in interaction with pilot farms P. Vereijken ................................................................................................................................................. 293
The Logarden project: development of an ecological and an integrated arable farming system CA. Helander. ............................................................................................................................................. 309
xi
Integrated crop protection and environment exposure to pesticides: methodes to reduce use and impact of pesticides in arable farming F. G. Wijnands ..............................................................................................................................................
319
Use of agro-ecological indicators for the evaluation of farming systems C. Bockstaller, P. Girardin and H.M.G. van der Werf ................................................................................ 329
Model-based explorations to support development of substainable farming systems: case studies from France and the Netherlands W.A.H. Rossing, J.M. Meynard and M.K. van lttersum ............................................................................... 339
Learning for substainable agriculture B.M. Somers .................................................................................................................................................
353
Author Index ............................................................................................................................................... 361 Subject index ............................................................................................................................................... 363
This Page Intentionally Left Blank
Section 1 INTRODUCTION Role of research and education in the development of agriculture in Europe E. Porceddu and R. Rabbinge ........................................................................................................................ 3 Reprinted from the European Journal of Agronomy 7 (1997) 1-13
This Page Intentionally Left Blank
© 1997 ElsevierScience B.V. All rights reserved Perspectives for Agronomy - Adopting Ecological Principles and Managing Resource Use M.K. van lttersum and S.C. van de Geijn (Editors)
Role of research and education in the development of agriculture in Europe E. Porceddu a'*, R. Rabbinge b aDepartment of Agrobiology and Agrochemistry, University of Tuscia, Via S. Camillo De Lellis, O1100 Viterbo, Italy bDepartment of Theoretical Production Ecology, WageningenAgricultural University, P.O. Box 430, 6700 AK Wageningen, The Netherlands
Accepted 16 June 1997
Abstract Agricultural research and education in Europe has played a major role in the advancement of agriculture and land use during the last century. The scientific basis of agriculture has been strengthened and the use of insight, knowledge and expertise in farmers' fields is widely adopted. As a result of this development, productivity per hectare, efficiency and efficacy of use of external inputs has increased considerably. During the last decades, objectives of agriculture and land use have broadened and this illustrates the need for further ecotechnological knowledge and insight to reach, in a balanced way, multiple goals of agriculture, productivity, protection of the environment, nature conservation and development. Research and education have to be developed in that direction to make agricultural science and technology more responsive to changing societal demandsl The disciplinary scientific quality and depth should develop in tandem with integrating problem-oriented multidisciplinary research activities. Systems approaches may serve as an instrument to that goal. © 1997 Elsevier Science B.V. Keywords: Systems approaches; Reorientation agricultural research; Broadened objectives
I. Introduction Since mankind started to exploit plants and other organisms to fulfill the changing demands for food and other needs, agriculture has been an important activity of mankind. The continuous evolution and manipulation of organisms and their function has been based on empiricism, knowledge, insight and expertise. The development has accelerated during the last century, when the scientific basis for agriculture strengthened and its simultaneous implementation became possible. The last decades have seen a broadening of aims and objectives of agriculture and
an increased importance of various environmental and socio-economic constraints. The ways that development took place, the change from old concepts to new perspectives and the challenges of agricultural research in general and more specifically of agronomy, are described in this introductory article to the Proceedings of the 4th ESA conference. The possibilities to reach those aims, and the new institutions and concepts which recently were introduced and adopted, are described.
2. Historical setting
* Corresponding author. Reprinted from the European Journal of Agronomy 7 (1997) 1-13
Scientific developments in agriculture during the
19th century were dominated by two scientists: Justus von Liebig and Gregorius Mendel. Their discoveries in the field of plant nutrition and the laws of heredity opened entirely new roads. Even today they still inspire scientific progress in fields very distant from the ones in which they originally made their contributions. A careful examination of their scientific experiences leads us to understand that, contrary to what is often reported, they worked in a lively and stimulating environment. It laid the basis for agricultural sciences and the institutionalization of research in agriculture; an event that Whitehead (1925) considered as the greatest innovation of the 19th century. The experiments carded out by private farmers, including the introduction of new crops, the establishment of chairs of agriculture at several university level teaching institutions, and the initiation in 1786 of the first public experimental farm near Braunschweig and a farm school in Hamburg, gave birth to agricultural research and experimental institutes in most European countries. The forestry school and the pomology institute of Weihenstephan were set up in 1803, followed by the research and teaching institutes at Hohenheim in 1818 and at Moglin in 1819. By the end of the century, Germany could boast as many as 87 research institutes, many of them in Prussia (Nomisma, 1996). The movement soon spread to France, where, at the initiative of private societies, farm schools were set up in Roville, near Nancy in 1822, Le Saussaie in 1828, and Bechelbon in Alsace in 1834. The first public institution became eventually operational in Versailles in 1848. In total there were 61 experimental stations in France at the end of the century (Nomisma, 1996). In the United Kingdom, the first experimental institution was set up in Rothamsted in 1844 as a private foundation, a form it was to retain until the end of the century; it was later joined by similar organizations in Woburn and Pumpherstone. By the end of the century agricultural colleges had been established at Bangor in 1889, Leeds in 1890, Aberystwyth in 1891, Nottingham in 1893, Reading in 1893, Cambridge in 1894 and Wye in 1894 (Speedy, 1994). In Italy, a private school opened in Tuscany in 1834 with a program of theoretical and practical lectures. Six years later it was transformed into a university level institution attached to the University of Pisa
(Coppini and Volpi, 1991). Within a few years of the unification of the country, in 1866, the University of Naples started planning a Faculty of Agriculture. It eventually became operational in 1872, though by that time it had been preceded by a Higher School of Agriculture, set up in 1870 by the University of Milan. Three agricultural experimental stations were set up in the same year. The last institution set up during last century in Italy was the Higher School of Agriculture at the University of Perugia in 1896. In Spain teaching started in 1855 at the Agricultural Central School in Aranjuez, close to Madrid, which had been promoted by the Ministry of Agriculture (J.A. Cubero, in litteris). In many countries of Eastern Europe similar developments took place, although in most cases much later. During 18501900 many National Agricultural Research Systems (NARS), including academic education and extension, were developed. However, during the last decade of the 19th century, many of these systems suffered due to considerable contraction of the agricultural sector. The story of the agricultural research institutes was very different in Denmark and the Netherlands, where the stimulus for extensive teaching and experimentation activities was provided by the great agricultural crisis of the last quarter of the 19th century. While other countries adopted protectionistic measures, these two countries reacted by rising the level of cultivation and agronomic techniques of their farmers. As a result, their competitive capacity increased. Agricultural research in Denmark was entrusted to four experimental stations, while teaching was concentrated at the Higher School of Agriculture in Copenhagen. In the Netherlands, the Faculty of Agriculture, then still a Higher School, was founded at Wageningen in 1876. Very quickly it gave rise to four distinct bodies: the Higher School of Agriculture, the School of Horticulture, the Agricultural Secondary School, and the Agricultural Colonial School. Agricultural extension was started and an intensive system of experimental stations and farmer field experiments initiated (Eveleens and Rabbinge, 1994). This is not the place for attempting a full-scale historical review of Europe's agricultural research, experimentation, and teaching institutions, but the few words devoted to their origin demonstrate not only the fervor that existed in Europe for agricultural
research and teaching, but also that this fervor made it possible to develop and exploit certain processes. In no more than a century these lead to greater transformations in agriculture than the ones that had occurred in the two preceding millennia. Both in arable farming and dairy farming, increases in land and labor productivity in the last 100 years were dramatic, compared with all ages before. In Italy, wheat production, which at the beginning of the century was of the order of 1 t/ ha, is today four times as great; the cultivation of 1 ha of wheat today absorbs 4 man days or 30 h of labor, while as many as 70 man days were needed only 70 years ago. Similar developments occurred in other industrialized countries. Average yield level or wheat in the Netherlands increased from 1.5 t/ha to 8.5 t/ha and labor requirement decreased from 300 to 15 man hours/ha (de Wit et al., 1987). At present yield levels of 10 t/ha are not exceptional. Similar developments can be observed in other crops. The increased productivity coupled with an expansion of the agricultural land allowed most of the European countries to produce enough food commodities and meet the increasing demand through population growth and diet change. Severe famine crisis could thus be prevented. Also, emigration, which drained
more than 40 million Europeans during the 19th century, could be prevented in this century.
3. Recent developments in agricultural productivity and employment Events during the last quarter of the present century, a period even the youngest scientists have experienced at first hand, are well documented. At the end of the 1960s the number of people served by a single European Union farmer was only half the present number (Table 1), and the amount of produce provided by such a farmer was half of what it is today (Table 2). This has left an ever larger number of people free to be engaged in other activities, to leave the rural areas and settle in towns (Table 1). In the same 25 years, agricultural land has decreased and within this acreage the arable crop land has increased, the yield per unit of land has increased by approximately 50%, and yield per unit of labor has tripled (Table 2). These facts made it possible for agricultural commodity prices to lag behind the increase in the general cost of living, enabling an ever larger number of people to enjoy
Table 1 Total and active population in EU countries. Situation in 1973 and trends 1970-1993. (Source: Eurostat, 1994; FAO, 1995) Country
Denmark Finland Sweden Austria Belgium France Germany Ireland Netherlands United Kingdom Greece Italy Portugal Spain
1993
1993 vs. 1970
Total popul, (1000)
Rural Rural/ T o t a l Agric. Agric./ popul, total labor labor total (1000) popul. (1000) (1000) popul. labor labor (%) (%)
Inhab/ arg. labor (#)
Total popul, (%)
Rural Total popul, labor (%) (%)
Agri./ tot. % labor (%)
Inhab./ arg. labor (%)
5155 5017 8633 7846 10251 56848 63694 3555 15 158 57908
644 2011 1356 3199 243 14400 7704 1512 1716 6341
12 40 16 41 2 25 12 43 II II
2567 2030 3912 3570 3744 21908 36 Ill 1149 6640 25 348
131 174 139 245 99 1195 1272 157 256 522
5l 8.6 3.6 6.9 2.6 5.5 3.5 13.7 3.9 2.1
39 29 62 32 104 48 50 23 59 Ill
105 109 107 105 103 100 105 121 ll6 104
64 88 89 89 37 98 68 106 95 97
111 93 100 Ill 98 105 134 103 145 100
48 34 41 41 55 41 62 52 78 73
202 320 260 270 190 270 170 230 150 140
10055 57 812 10300 39390
3671 17604 6732 8134
37 30 65 21
3715 20267 4464 11 868
791 1488 516 1212
21.3 7.3 11.6 10.2
13 39 20 33
ll4 107 ll4 117
88 92 100 71
98 105 139 96
45 40 53 32
260 260 220 360
Table 2 Total land and utilized agricultural area in EU countries. Situation in 1973 and trends 1970-1993. (Source: Eurostat, 1994; FAO, 1995) Country
1993 Total area (1000 ha)
Denmark Finland Sweden Austria Belgium France Germany Ireland Netherlands United Kingdom Greece Italy Portugal Spain
1993 vs. 1970 (2)/(1) (%)
Crop area/ (2) (%)
1
Utilized agric, area (1000 ha) 2
(2) (%)
(4) (%)
(3) 1970 (%)
(5) (%)
Output/ area (%)
Output/ labor (%)
4
(2)/ labor (ha) 5
3
4306 33 699 44 759 8385 3050 54 883 24871 7031 3694 24419
2751 2610 3359 3482 1412 30 217 17 162 4450 1997 17 178
64 8 8 42 46 55 69 63 54 70
92 96 83 41 58 59 68 17 46 35
19,2 11,6 19,2 15,6 17,8 21,3 10,1 29,1 8,3 30,3
92 87 97 89 88 91 126 92 91 89
94 94 91 85 96 94 145 65 105 84
90 91 89 43 53 58 60 24 40 37
183 190 173 208 207 207 177 162 128 121
153 137 135 144 154 136 143 158 187 152
282 263 232 297 320 280 255 255 239 165
13 196 30 098 9202 50 508
5785 16 800 3829 26 398
43 56 42 52
51 54 58 58
9,3 9,3 5,6 18,2
65 87 78 73
81 64 51 75
41 73 89 56
155 202 140 194
151 145 102 180
233 292 143 348
higher food standards and a better lifestyle. Today there are problems of overproduction. The increasing yield per unit of land, as result of research innovations and use of external inputs, which have been progressively introduced and adopted by farmers, is continuing. Greater income has also led to a greater demand for goods and services of high elasticity with respect to income, among them a cleaner environment, stricter requirements for food quality and the way food is produced. Environmental quality is nowadays considered as a 'superior good'. Its demand rises in proportion to income growth, while food production is an 'inferior good', the demand for which falls in proportion to increases in income (Engel' s Law). Demand on food quality increasingly dominates food quantity while concern for environmental health and safety grows. Natural resources are no longer perceived by policy communities as merely the medium to produce more and cheaper food but rather in terms of local and global ecosystem functioning. Ecosystem maintenance, bio-diversity, recharge of ground water, clean air and bequeath value are important topics, as agronomists well know. Agriculture is also a service to society, for example, as a means to preserve the landscape. However, there
is no doubt that such a service to society requires greater attention in planning and financing structures, and can only play an additive role for the farmer community. It cannot replace farming as an economic activity. Interventions using farmers income compensation, as it was recently experimented in the Netherlands, where a number of farmers in certain areas are subsidized to maintain landraces and primitive varieties, utilizing traditional technologies, may be seen as examples. National attitudes are nowadays dominated by the effects of market globalisation, which modifies trade flows, reallocates production and consumption at the world level, thus influencing agricultural activities and natural environment and, quite generally, setting the scene for any development in rural areas.
4. The influence of agricultural policy and trade agreements At present, the situation in EU agriculture can be depicted as follows: agricultural employment decreased by 2% per year during the last 25 years, but in a number
of member states agricultural employment still exceeds 5% (Table 1); land and labor productivity are still limited in some member countries (Fig. 1), in spite of progress during the last 25 years (Table 2); land per unit of labor is still limited in a number of countries (Table 1), and a high percentage of agricultural laborers remains (Table
•
•
25
• Gr
20 ,P • IRL
-110
,E
,I **
2); •
pro capite GPD (Gross Domestic Production) in member states is inversely proportional to the percentage of labor in agriculture (Fig. 2).
Taking into account these facts, the restructured EU-CAP (Common Agricultural Policy) envisages an agriculture with: •
more technology, but also with greater diversity; a smaller cultivated acreage, but also with more land dedicated to special uses; a more stable production, but also with a greater variety of produce; a smaller number of farmers, who, however, perform several and more diversified activities.
• • •
*Dk
*UK,B
The reform of the EU-CAP was followed by a further extension of trade agreements in the form of General Agreement on Tariffs and Trade (GATT). The most recent economic analyses (Anania et al., 1996) indicate that these agreements would not distort the EU-CAP, although the impact may not be negligible. Analyses also underscore the need that the EU
--
0
I ....
5000
I
I
10000
15000
20000
25000
30000
pro cap/h) GDP
Fig. 2. Relationbetween pro capite GrossDomesticProductionand labor force in agriculture in EU countries during 1993. agricultural system maintains its high competitiveness in terms of low cost in combination with high product quality, this in the domestic market as well as in the international one. The need represents the nodal point of the European agriculture in the near future: lower costs, including environmental costs, better quality and toxicologically safe products. These involve the efficiency of production processes and farm technologies. That requires a policy presently adopted by the European Union. Yet, it is not sufficient. The relative weight of various objectives has to be reconsidered. The choice of such a weight factor may result in considerable differences in land use in the future, as shown in an explorative study on reformulation of the Common Agricultural Policy of the European Union (Rabbinge and van Latesteijn, 1992; WRR, 1992). Using modem simulation and systems approaches, and incorporating appropriate technical information, a series of land use scenarios were generated, based on different priorities for objectives that dominate agricultural development: o
jA
B
k,
~m
•
O I&l i
]mooo
F
•
i
o 1000
2000
~
•
4000
Output per hectare (ECU's per ha)
Fig. 1. Trends in agricultural productivityin EU countries during the period 1970-1993. Lower left points 1966-1970, upper fight points 1986-1990. (Source: Eurostat, 1994; FAO, 1995).
•
free market and free trade, based on economic efficiency, i.e. maximum yield at minimum cost; regional development, with the aim of maintaining employment in agriculture at the highest possible level but within the constraints of a productive agriculture; nature and landscape conservation, by the creation of natural reserve areas separated from the agricultural ones and thus minimiz-
ing land use for agriculture and maximizing land productivity; environmental protection, minimizing the negative effects of agriculture; in other words, minimum use of pesticides and emission of nitrate, and other environmental side effects of agriculture, per unit of product or per unit of land. The results provided a framework for assessing the strategic options on which policy makers may base their decision. Indeed, the exploration of options showed very marked differences: •
•
•
as far as land use is concerned, the maximum acreage used by agriculture in the case of regional development is three times as great as the minimum associated with the free market hypothesis, but 40 out of the present 127 million hectares will nevertheless have to be taken out of use in all scenarios (Fig. 3a); as far as employment is concerned, all scenarios give rise to a further downturn of agricultural employment, but the size of the exodus may vary between 3 and 4.5 million man power units depending on the chosen scenario (Fig. 3b); quite independent of the scenario actually adopted, policy measures can successfully promote more environmentally friendly production methods by limiting the use of fertilizer and reducing the large scale use of crop protection agents, avoiding adverse effects on the environment (Fig. 3c).
The outcomes of the scenarios demonstrated the enormous challenge and chances for reorientation of the CAP, and for agricultural research. It becomes even more important to do such explorative studies when the European Union will be further extended in the near future. The considerable increase of land area, the tremendous possibilities for agriculture and the needed readjustment of the common agricultural policy in those types of situations require extensive explorations. Therefore, explorative studies on land use and agrotechnologies should become major issues on the future research agenda.
5. Agronomy and multiple goals in agriculture and land use The changes of agriculture from a purely production-oriented activity into a science based production sector, trying to meet productivity, efficiency and efficacy aims, has been of considerable importance during the last decades. Agriculture has broadened and diversified its objectives, rendering explicit and important a number of economic, environmental, social and nature-protection roles that in the past were simply passed over in silence. The concept of a good agricultural practice - where farmers produce in an economic and socio-political environment conducive to long-term conservation and use of natural resources, and do not want their source of survival and generation of income to disappear - has always been the foundation of agronomic teaching. Indeed, the same Justus von Liebig, often blamed as the father of chemistry in agriculture, way back in 1855 wrote as follows: 'The task of the farmer is not to achieve high crop yield to the detriment of the field, which only causes it to impoverish earlier. Rather, it should be in his own interest, as well as society's, to achieve high yields that are constantly increasing on a permanent basis.' (von Liebig, 1855). The concept is echoed in the 'sustainability' idea, entered into a common use thanks to Our Common Future, the report of the World Commission on the Environment and Development (1987), and defined by the Consultative Group for International Agricultural Research (TAC, 1989) as the successful management of resources for agriculture to satisfy changing human needs while maintaining or enhancing the quality of the environment and conserving natural resources. The question of sustainability, consequently, arises when the resources used for production are placed under stress, as is now widely happening in industrialized countries due to excessive irrigation, fertilizer and pesticide application, and in developing countries where increasing population pressure continues to strain product on resources. An agricultural production system is not sustainable if it leads to declining productivity, degrades the resource base, or is not economically viable. The given description of unsustainability does not necessarily lead to a precise definition of sustainabilty. This is not easy as
.AND USE 150-
"~" 100 r,-
o..,
E t~
2 <:
50
0
A
B
C
D TODAY
SCENARIOS BdR.OYIdB~ ,,,
A
_,: E
.~
f
,.,
6-
W C
|
4"
g C
J
2-
, _ t~2~.~'.-:
I i F~//~ A
B
,~,~~-',',;.~,,:~|~!:..:~::: C
D TODAY
SCENARIOS CROP PROTECTION
E
v
300-
O C ¢g
]~ 200oo .~ 1000A
B
C
D TODAY
SCENARIOS
Fig. 3. Land use, employment, and crop protection in different scenarios (WRR, 1992): (A) Free market and free trade. (B) Regional development. (C) Nature and landscape. (D) Environmental protection.
sustainability comprises both normative and scientific notions that may be weighted differently, but in all cases the 'unsustainability spiral' (Rabbinge, 1997) should be broken. Agriculture, indeed, makes use of nature's productive capacity, tapping outputs from the system in the form of harvest, and the very continuity of this process presupposes that certain inputs must be added to the system to compensate the tapped outputs. It is in any case well known that inputs, no matter what they might be, cannot be 100% converted into outputs or, put differently, that the efficiency of the system cannot be 100%. Part of the inputs is always lost into the environment as leakage. These losses can be reduced but not entirely avoided, irrespective to the farmer aiming at maximizing the output or seeking input efficiency. In this context now, the role of the agronomist (literally: he who governs the fields) is to optimize the system in such a way as to obtain a maximum efficiency. The modern agronomist is no longer a pure empiricist. More and more science based interventions and measures are developed and implemented. White peg agronomy is replaced by production ecology (Rabbinge, 1997). Poor agricultural efficiency, and the environmental pollution that may derive from it, is often associated with the law of diminishing returns. It states that the relationship between the amount of a production factor and the yield level is not linear, but levels off and more and more external inputs are needed in order to push up the yields to their potential level. High input doses are blamed to be responsible for inefficiency and pollution. While the law is certainly valid, it should be noticed that it has been formulated for situations where all the other production factors remain constant. This is obviously not the case when changes over time occur. Using data of maize yield and nitrogen fertilizer use in the United States for the period 1945-1982 (Fig. 4), de Wit (1992) showed that at the high end of the yield range the fertilizer efficiency is at least equal to, if not greater than the efficiency at the low end. The relationship flattens for some years in the 1960s, as indicated by the law of diminishing returns, but when agricultural practices improved again in the 1970s - possibly under influence of the oil shock, but there may also have been an impact of the change of the varieties' cytoplasm, as a consequence of susceptibility of the T cytoplasm to Hel-
10
Grain yield (kg ha -1) 8000 -
1982 OO
•
6000 •• •
4000
2000
•
1970
O o 00 ° -O
1945 I
0
40
,,
I
I
I
80
120
160
Nitrogen (kg ha -1) Fig. 4. Plot of annual average maize yields versus nitrogen fertilizer rate in USA during the period 1945-1982. (From: Sinclair, 1990).
the fertilizer response returned to its previous level. These relationships become even more evident when expressed in energy terms (de Wit, 1979). Elaborating on the law of optimum (Liebscher, 1895) - a production factor, which is in minimum supply, contributes more to production, when the other production factors are closer to their optimum - de Wit (1992) contended that most production resources are used more efficiently with increasing yield level due to optimization of growing conditions. According to this vision, strategic agronomic research, that is to serve both agriculture and its environment, should be directed towards the search for the minimum of each production resource that is needed to allow maximum utilization of all other production resources in the farming system considered. The problems - inefficiency and pollution - therefore derive from the fact that the processes governing agricultural production are not yet properly understood. This may lead to the adoption of measures and practices that eventually do not contribute towards efficient use of resources and control of environmental effects. The broadened objectives of agri-
minthosporium
-
culture have also broadened the mission of agronomy as a scientific discipline. Agronomy is the science that, taking advantage of ecological principles, devices and tests new approaches, rules and means to govern the relationships between the different production factors in order to obtain an appropriate harvest. The study of single production factors is entrusted to such propaedeutic disciplines as agrometeorology, soil physics, crop physiology, genetics, pathology, entomology, and land use, while the results are offered to other disciplines, such as economy, sociology, production ecology, and human nutrition and health, that provide the strategic options of the agricultural production. Agronomy, therefore, plays an altogether central role in agricultural research. It has a revitalized and crucial role after some decades of obsolescence and crisis due to the difficulties it faced in devising innovative solutions and/or sustainable options.
6. Approaches in agronomic research
The study of agronomic problems has been traditionally faced by means of two different types of research that could be called general and specific or specialist approaches (L. Cavazza, in litteris). The general approach consists essentially in verifying the reaction of the entire agro-ecological system i.e. soil, plant, atmosphere, other organisms - to well defined experimental conditions containing one or a combination of few changed technological factors or doses of factors, where a factor may be an input (ploughing, crop rotation, fertilization, irrigation, weed, pest, disease control etc.), its dose, and/or its application time. The qualifying components of the general approach rest on the identification of the factors capable of modifying the system and the analysis of the responses to the doses of the factor(s) under test. The results are strongly in line with the experimental conditions, rich in unsuspected information about the system behavior in response to the specific experimental practices and of considerable technical value. Nevertheless, the system is considered as a single entity, or as a set of a limited number of single entities, and the approach does not provide any direct, simple information on what occurs inside the system or the
11
system components. The theoretical knowledge of single components does not enter into the logic of the experiment; rather it is utilized to explain and justify the results. Findings are, therefore, essentially descriptive, site specific, closely tied to experimental conditions, without wide opportunities for generalization. Real innovations are rare, although they are promptly entered into agricultural practice. It should be pointed out, however, that most of the progress in crop and animal production has been attained in this way and no new technique should be entered into actual agricultural practice without having been tested in this way. The specific approach has its starting point in the analysis of the system components, such as root system, canopy structure and hormonal system. One or few of those will then become the subject of research to gain as much insight as the state of scientific knowledge will enable. The smaller and more clearly defined the component, the better it can be analyzed, the more readily principles and rules can be formulated, even in deterministic terms. The results are not bound to the experimental conditions. The difficulty in this type of approach derives from having to assess the behavior of the system from some detail. The risk of a loose re-assembling of various components, that have nevertheless been very precisely analyzed, is always there. Another risk is that only part of the system is analyzed, considering the other parts as fixed. One relies on either a single or just a few technological factors to push up the entire system. Failure in optimizing the system, promotion of diminishing returns and pollution of the environment may become real consequences. Both the general and specific approaches were very productive in scientific terms. The first by the implementation and derivation of practical rules, the second by gaining understanding in some of the basic processes involved. The distinction between the two approaches has its counterpart in education and training. A scholar who has shared his energies over extensive scientific bases, though without getting insight in any one of them, will be inclined to the general approach, but will not get the results typical of the other approach, unless he enrolls on specific postgraduate preparation, like a Ph.D. or similar project. On the other hand, a scholar who, with identical commitment, concentrates his education on some specific
aspect, will prefer the specific approach, even though he will find it difficult to insert his findings into the system and to translate them directly into advice and recommendations at the system level. The characterization of the two approaches does not seek to place them in opposition to one another, or to underscore their own relative shortcomings. It rather aims at highlighting the opportunities for integrating them, on the ground that the thorough scientific insight and the creativity of the specific approach, joined with the description and understanding of the system behavior and its problems typical of the general approach, show to be very productive and will ultimately lead to real progress in agronomic science. Research and education approaches are really modem and productive when they combine and utilize in a proper balance the most positive components of the general and specific approach. Actually, in the last decades we have witnessed the integration of process specific knowledge into very precise, widely accepted relationships between processes in the system and driving factors from outside the system. Systems approaches and simulation were very functional in that integration. Simulation models, capable of synthesizing different pieces of knowledge, were developed. These models are simplifications of the real system functioning. They function in the soil, plant, atmosphere, or other organisms systems, but only represent the major elements of the system. The fact that models require well defined hypotheses on systems behavior and an adequate interpretation of the results needs to be stressed. Once formulated, the model helps in a heuristic way to gain insight that may be used to derive general insights and rules. The model, when confronted with experimental evidence, is, therefore, the true reviewer of our knowledge. When a model produces unacceptable results, it means that input data were not reliable or, even worse, hypotheses and assumptions underlying the model were not correct, that knowledge is insufficient, and that we must consequently review the simplifications made about the system to better identify its character and define the relations existing among the different components. In this way, systems analysis and simulation constitute an excellent opportunity in agricultural research, in the advancement of knowledge and the optimization of relations among factors. Optimi-
12
zation itself requires creativity and innovation, in addition to a well structured education, where deepening in basic science is embedded in the broadened perspective of agriculture and agricultural problems. Modem agricultural research means finding solutions to whole-system deficiencies as well as to component problems. It is futile to attack specific technical problems without addressing the overall pattern. Modem agronomic research must aim at all levels, including farmers' fields, cropping systems, farming systems, and regional and supra-regional levels. Systems analysis enables the agronomist to go beyond the farm bounds. It enlarges his field of activity to policy, social and economic grounds, that may sometimes heavily affect his research activity. Crop and/or farm systems are, in fact, embedded in the economic-social-political system, and this requires a widening of those considerations, levels, constraints, and results that guide farmers in implementing agronomic rules and technologies. Two new dimensions, among those in front of us, call for immediate and adequate attention: (i) the shifting of interests from farm to the regional dimension, and (ii) the widening of effects and aims of farm activity to include, in addition to production, also those of labor conditions, commodity quality, agricultural impact at farm and district levels, and more generally all those interactions between farm activities and the surrounding environment. The new dimensions require the commitment of scientists from different fields, well beyond those traditional in agricultural research. It represents an additional real challenge for agronomic research.
7. Challenges for agronomic research
Agriculture has to meet at global level a rising demand for marketable outputs, while satisfying ever tighter constraints with respect to safety of products and impact of production techniques on man, nature, environment and landscape. The attainment of these goals requires a comprehensive and integrated research approach. It is the synthesis of knowledge from various disciplines into a coherent framework, subsequently used to develop, implement and evaluate location specific farming and management options.
Crop harvest is the result of interactions among components such as growth-determining factors, growth-limiting factors, and growth reducing factors (Fig. 5), where: •
•
•
growth-determining factors are those which determine the growth potential of a crop under proper supply of water and nutrients, such as radiation and crop characteristics; growth-limiting factors comprise abiotic resources, such as water and nutrients, the suboptimal supply of which limits crop growth and yield; growth reducing factors are those, like pests, diseases, weeds, that reduce attainable growth to actual growth (van Ittersum and Rabbinge, 1997).
Three research themes at the interaction between growth determining, growth limiting, and growth reducing factors deserve urgent attention and may represent useful ground for international co-operation among European Countries. The first research theme concerns precision farming. Smart farming technologies aiming at both productivity and efficiency gains are vital under either well endowed or marginal conditions. There is a considerable need to manipulate the defining, limiting and reducing factors, in such a way that all individual factors are used with nearly maximal efficiency. Actually, the role of growth limiting factors and agronomic practices - such as fertilizers, irrigation and crop density - has been extensively investigated in many different parts of the world to determine actual yield. Yet, much research is still of a descriptive nature at systems level and does not consider the causal factors behind it. A clear cut distinction between cause and effect is absent while processes with long time coefficients are not used to understand systems behavior. The analysis of systems will result in a better understanding of the role of specific components and their interactions, and will allow to more appropriate farm technologies to be devised. In particular, improvements in efficiency under suboptimal growth conditions would produce a tremendous impact on agriculture in Southern Europe. The optimization of external inputs and the way they affect the growth and production of agro-ecosystems under suboptimal conditions is a major task for production ecologists.
13
I.
AttainableOrowll~//~, i
(
n~n~gen phosi~xxus "
•d i ~ s ~
:~o , •
g:
,
~ 1
!
i
Production level
Fig. 5. Production levels and corresponding growth factors. (From: Rabbinge, 1993; van Ittersum and Rabbinge, 1997).
The second theme is closely linked to these aspects and concerns the water resource use efficiency. In a situation dominated by an ever increasing competition between agriculture and urban areas for the use of water, agricultural research has the duty of optimizing the utilization efficiency of water. Optimization of irrigation aimed at maximizing water use efficiency, performing irrigation in crop critical development and production stages would greatly help in economizing this important resource (Bonciarelli, 1996). Studies at field level and the underlying processes of water flow, water uptake and use may lead to better use of this important external input. However, studies at the watershed level are also needed. Suboptimal use at the field or farm level does not necessarily lead to low efficiency at the watershed level. The contrary may be true in some situations (Seckler, 1996). In combination, field and watershed studies may help to design optimal irrigation systems, techniques and policies at watershed, district, farm and field levels. The third theme concerns the soil. In Mediterranean as well as in Central European areas the study of interactions among soil components and system management is an urgent issue, an absolute imperative. Processes such as microbial activity, mineralization and immobilization events may be studied within a well defined framework of cropping systems research. Old findings on no tillage agriculture may also find their place within this framework. Examples of studies at various integration levels are scarce. The majority of research is at field level or at lower integration levels, limiting the knowledge basis for changes in the agro-ecosystem management. Through wider rotations, growth and yield reduction,
due to soil born diseases, pests and weeds, are limited and better use of mineralization processes and a reduction in nitrate or other plant nutrient losses are possible. Cropping systems research and farming systems research are still in their infancy. Results from descriptive research are available, but well defined explanatory research in connection with agronomic practices in general is lacking. A renaissance of mixed farming systems may even be considered. In some places in Europe, mixed farming systems never disappeared, but they are in general low in efficiency and much less economically feasible than specialized farms. However, new techniques, increased insight and better agronomic methods may help to redesign such farming systems that may lead to a technology jump in efficiency. Some preliminary results of such systems are very encouraging (Lantinga and Rabbinge, 1997). 8. C o n c l u s i o n s a n d p e r s p e c t i v e s
The chances and challenges for agronomy as a discipline are increasing. New perspectives, and new problems are identified. The solution of the problems mentioned is directly connected to the capability of generating scientific knowledge and technological innovations. The systems approach, identified as suitable to provide valuable solutions to the existing problems, requires that innovations proceed in a very organized and structured way. The success of the agricultural knowledge innovation systems depends, among others, on: •
•
•
national cohesion and institutional infrastructure; homogeneity in objectives among parties (farmers' organizations, research institutions, agricultural and environmental authorities, consumers etc.); incorporation of system-oriented, instead of single production-oriented, objectives in research policies; careful monitoring of signals from the outside world; more output-financing research; interest in basic science; co-operation among institutions in different European countries, extended to Eastern and Southern (Mediterranean) countries.
14
The list highlights very clearly that the present role of agricultural research is different from its classical role. The empirical approach, that has been so successful in the past decades, has to be complemented with a systems approach, which bridges the gap between detailed process-oriented research explaining the functioning of crops on basis of the physical, chemical and physiological processes, and the application of that insight and knowledge at systems level. The new approach requires that research teams have a multi-disciplinary nature and they need to work together in a problem-oriented approach; traditional boundaries between disciplines are counterproductive. Disciplinary contributions are fully used only when they are of high quality and tailored to the specific questions, dictated by the problem formulated at system level. An example of such a research organization is the C.T. de Wit Graduate School of Production Ecology of the Wageningen Agricultural University. It has made the systems approach and its development an important part of its mission. A research approach using old concepts and not using the new possibilities and opportunities is inadequate in answering problems raised in a modem society. The described systems approach may help to fulfill the new mission of agronomy as an interdisciplinary field. Productivity rise, efficiency increase, efficacy of input use, maximization of biological control mechanisms, and protection of land and natural resources require a reorientation of agronomic research. The techniques and possibilities are there, why not use them?
Acknowledgements We would like to thank Luigi Cavazza for the useful discussion on the subject, Martin van Ittersum and Siebe van de Geijn, the editors, and Alexandra Nagel for the valuable editorial comments on the manuscript.
References Anania, G., De Filippis, F. and Scoppola, M., 1996. Implicazioni delraccordo GATT per l'agricoltura e le politiche agrarie deU'Unione Europea. In: G. Anania and F. De Filippis (Editors),
L'accordo GATT in Agricoltura e l'Unione Europea. Collana RAISA, Franco Angeli Editore, Milano, pp. 218-339. Bonciarelli, F., 1996. L'evoluzione degli obiettivi ed il ruolo della ricerca agronomica. La Rivista di Agronomia (in press). Coppini, R.P. and Volpi, A., 1991. La Fecolt/i di Agronima dell Universit~t di Pisa. Istruzione agraria e transformazione economicail ruolo delle scuole di agricoltura nella Toscana della prima parte dell'Ottocento. In: A. Benvenuti, R.P. Coppini, R. Favilli and A. Volpi (Editors), La Facolt/l di Agraria delrUniversit~t di Pisa, Pisa, pp. 43-87. de Wit, C.T., 1979. The efficient use of labour, land and energy in agriculture. Agric. Systems 12: 279-287. de Wit, C.T., 1992. Resource use efficiency in agriculture. Agric. Systems, 40: 125-151. de Wit, C.T., Huisman, H. and Rabbinge, R., 1987. Agriculture and its Environment: Are there other ways? Agric. Systems, 23: 211-236. Eurostat, 1994. Serie storiche 1960-93. Eurostat. Eveleens, K.G. and Rabbinge, R., 1994. Current trends in post graduate education at Wageningen Agricultural University: innovation, organisation and internationalisation. Eur. J. Agric. Ed. Extens., 1: 9 l - 102. FAO (Food and Agricultural Organisation of the United Nations), 1995. Country Tables. FAO, Rome. Lantinga, E. and Rabbinge, R., 1997. The renaissance of mixed farming systems: a way towards sustainable agriculture. In: Proceedings of Conference on Nitrogen Emission from Grasslands', 20-22 May, 1996. Institute of Grassland and Environmental Research, North Wyke, United Kingdom, in press. Liebscher, G., 1895. Untersuchungen uber die Bestimmung des Dungerbeduffnisses der Ackerboden und Kulturpflanzen. J. Landwirtsch., 43: 49. Nomisma, 1996. Rapporto 1995 sull'Agricoltura Italiana. Agra Editrice. Rabbinge, R., 1993. The ecological background of food production. In: D.J. Chadwick and J. Marsh (Editors), Crop Protection and Sustainable Agriculture. Ciba Foundation Symposium 177, Wiley, Chichester, UK, pp. 2-29. Rabbinge, R., 1997. Integrating policy and technical issues for international research on agriculture and the environment using systems approaches. In: Teng, P.S., Kropff, M.J., ten Berge, H.F.M., Dent, J.B., Lansigan, F.P. and van Laar, H.H. (Editors): Applications of Systems Approaches at the Farm and Regional Levels. Kluwer, Dordrecht, pp. 249262. Rabbinge, R. and van Latesteijn, H.C., 1992. Long term options for land use in the European Community. Agric. Systems, 40: 195210. Seckler, D., 1996. The new era of water resources: from 'dry' to 'wet' water savings. IIMI, Colombo, 17 pp. Sinclair, T.R., 1990. Nitrogen influence on the physiology of crop yield. In: Rabbinge, R., Goudriaan, J., van Keulen, H., Penning de Vries, F.W.T. and van Laar, H.H. (Editors), Theoretical Production Ecology: Reflections and Prospects. Simulation Monographs 34. Pudoc, Wageningen, pp. 41-55. Speedy, A.W., 1994. Development and future prospects for agricultural Education in Great Britain. In: International Conference
15 on Integration of Agricultural Science Education. EC Countries. Bologna, Italy, October 27-29, pp. 73-86. TAC (Technical Advisory Committee)CGIAR, 1989. Sustainable Agriculture production. Implications for International Agricultural Research. FAO Research and Technology Paper 4, Rome. van Ittersum, M.K. and Rabbinge, R., 1997. Concepts in production ecology for analysis and quantification of agricultural input-output combinations. Field Crops Res., 52: 197-208. yon Liebig, J., 1855. Die Grundsaetze der Agricultur-Chemie: mit Ruecksicht auf die in England angestellten Untersuchungen - 2.
Durch einem Nachtrag verm. Auflage. Vieweg, Braunschweig. Whitehead, A.N., 1925. Science and the modern world: Lowell Lectures, 1925. Glasgow, 252 pp. World Commission on the Environment and Development, 1987. Our common Future. Report by the Bruntland Commission. Oxford University Press, Oxford. WRR (Netherlands Scientific Council for Government Policy), 1992. Ground for choices: Four perspectives for the rural areas in the European Community. The Hague, The Netherlands, p. 144.
This Page Intentionally Left Blank
Section 2 AGRICULTURE LAND USE, FOOD SECURITY AND ENVIRONMENT Land use transformation in Africa: three determinants for balancing food security with natural resource utilization P.A. Sanchez and R.R.B. Leakey .................................................................................................................... 19 Reprinted from the European Journal of Agronomy 7 (1997) 15-23
Agro-ecological characterisation, food production and security P. Bullock ......................................................................................................................................................
29
The potential benefits of agroforestry in the Sahel and other semi-arid regions H. Breman and J.J. Kessler .......................................................................................................................... 39 Reprinted from the European Journal of Agronomy 7 (1997) 25-33
Chemical crop protection research and development in Europe R. Neumann .................................................................................................................................................... 49
Emissions of CO2, CH4 and N20 from pasture on drained peat soils in the Netherlands C.A. Langeveld, R. Segers, B.O.M. Dirks, ,4. van den Pol-van Dasselaar, G.L. Velthof and A. Hensen .......... 57
Reprinted from the European Journal of Agronomy 7 (1997) 35-42
This Page Intentionally Left Blank
t~ 1997 ElsevierScience B. V. All rights reserved Perspectives for Agronomy - Adopting Ecological Principles and Managing Resource Use M.K. van lttersum and S.C. van de Geijn (Editors)
19
Land use transformation in Africa" three determinants for balancing food security with natural resource utilization Pedro A. Sanchez*, Roger R.B. Leakey International Centerfor Research in Agroforestry, P.O. Box 30677, Nairobi, Kenya Accepted 8 April 1997
Abstract
The continued threat to the world's land resources is exacerbated by the protracted food crisis in sub-Saharan Africa. Per capita food production continues to decrease even though this region compares favorably with other tropical regions in terms of climate and soil resources. The main determinant of this situation is the widely recognized need for an enabling policy environment that favors smallholder rural development. However, there are two other key determinants to food security and environmental sustainability in Africa that have not received sufficient attention in the past and are the focus of this contribution: (1) the need to tackle soil fertility depletion as the fundamental biophysical constraint to food security and (2) the need for more intensive and diverse land use, based on the domestication of indigenous trees to produce high value products while increasing agroecosystem resilience. Approaches that include these three issues will transform smallholder farming in Africa into productive and sustainable enterprises and will contribute greatly to food security and environmental conservation, in a win-win situation. © 1997 Elsevier Science B.V. Keywords: Food security; Environmental sustainability; Soil fertility depletion; Domestication of indigenous trees; Agroindustry policy; Profitable tree crops; Soil fertility replacement; Nutrient recapitalization; Irvingia gabonensis; Prumus africana
1. Introduction
The continued threat to the world's land resources is now exacerbated by looming food shortages in the Third World, increased food prices globally and wider climatic variations resulting from global warming. During the last decade food security was not a global priority, but recent studies such as 2020 Vision (IFPRI, 1996) show that food security should be one of the main global concerns of our time. Food inse* Corresponding author. E-mail:
[email protected]
curity encompasses food scarcity as well the inability to purchase food, a poverty-related issue. Although food insecurity occurs throughout the developing world, it is most acute in Sub-Saharan Africa (hereinafter referred to as Africa), where the attainment of food security is intrinsically linked with reversing agricultural stagnation, safeguarding the natural resource base and reducing population growth rates (Cleaver and Schreiber, 1994). The purpose of this contribution is to raise three determinants of food security and environmental sustainability in Africa for the next 25 years.
Reprintedfrom the European Journal of Agronomy 7 (1997) 15-23
20
2. The problem Per-capita food production continues to decrease in Africa, in contrast to increases in other parts of the developing world (Fig. 1). Analysis of trends indicate that the green revolution and enabling policies drastically increased per-capita food production in Asia and Latin America, where the race between food production and population growth is steadily being won. In contrast, the race is being lost in Africa, where percapita food production has been declining at about 2% per year since 1960 (World Bank, 1996a). In addition, Africa has about half of its population classified as absolute poor (those subsisting on per-capita incomes of less than one US dollar per day), as well as the highest proportion of undernourished children and the highest rates of population growth of any region in the world (Badiane and Delgado, 1995; World Bank, 1995a). Most of the poor in Africa live in rural areas. A 4% annual sustained growth rate of agricultural production in Africa is an absolute requirement for reversing this situation by the year 2020 (Badiane and Delgado, 1995). Improved governance, macroeconomic stabilization, and more support to the agricultural sector are widely considered the necessary conditions to reversing the trend of per-capita agricultural production (Cleaver and Schreiber, 1994). Attempts to bypass the development trajectory of Europe, North America and Asia which started by developing a strong agricultural base have simply not succeeded in Africa or elsewhere. Agriculture (in its broadest sense, which includes food crops, tree crops, livestock, forestry and fisheries) is the engine of economic growth (Brown and Haddad, 1994). Agriculture meets the country's basic needs for human 130 125 120 115 110
S. America
105 100
Africa
95 90 1980
1985
1990
1995
Fig. 1. Per capita food production indices (1979-81 = 100).
survival and provides rural employment which leads to the development of an agroindustrial sector that processes food, timber, paper etc. This in turn leads to marketing enterprises, and often to an export sector. Industrialization and service sectors arise out of this foundation. Mining of natural resources, tourism and other sectors contribute significantly to growth but are not the engine that agriculture is because these sectors do not address the basic needs of the majority of the population, nor do they provide sufficient employment opportunities for the expanding rural labor force. African countries that are relatively better off like South Africa and Zimbabwe have made agricultural development a high priority, in spite of their abundant mineral resources and a strong tourism sector. In addition to the general need for an enabling policy environment, other factors affecting African agriculture's productivity over the last 20 years are low crop productivity, low growth rate of agricultural production (2% per year), the highest rate of population growth of any region in the world (2.9% per year), and the highest rates of land degradation (30%) of usable land (Cleaver and Schreiber, 1994). Indeed, the population to land resource ratio in Africa has been highlighted as a major factor contributing to Africa's low per-capita food production, but is widely debated. While at the continental scale, Africa is not densely populated, large areas of Africa have population densities approaching that of Asia, but with a major difference in that Africa does not have the irrigation infrastructure Asia has (John Lynam, personal communication). In addition, Africa is certainly faced with the large problem of increasing population that Asia no longer has, which exacerbates the decline in per-capita agricultural production. However, Africa compares favorably with other tropical regions in terms of climate and soil resources (Sanchez and Logan, 1992; World Resources Institute, 1992). The biophysical endowment of tropical regions of Africa, Latin America and Asia is summarized in Table 1. In terms of rainfall, tropical Africa has a similar proportion of areas with drought stress as tropical Asia, while tropical America is better endowed. Tropical Africa compares favorably with the other two regions in terms of the proportion of its soils with fertility problems, such as high soil acidity resulting in
21
Table I Areal extent of soil-related constraints in the tropics of Africa, Asia and Latin America (calculated from a data base by W. Couto, North Carolina State University, based on Sanchez et al., 1982) Soil-related constraints
Tropical Africa
Tropical Asia
Tropical America
Total Tropics
Total area (million ha) % long drought stress % low in weatherable minerals % aluminium toxicity % poor drainage % steep (> 30% slope) % low cation exchange capacity % high phosphorus fixation (by Fe) % sandy % vertic (black cotton soils) % gravel layer % calcareous (Fe and Zn deficiencies) % saline % alkaline % shallow to rock (<50 cm)
1555 67 31 26 22 22 II 11 7 3 3 2 I I 1
1205 72 27 24 16 51 i 20 2 4 7 8 2 1 2
1879 45 46 43 20 32 4 32 4 I 1 I 1 I 3
4638 60 36 32 19 34 5 22 4 3 3 3 1 1 2
aluminum toxicity, high phosphorus fixation, micronutrient deficiencies in calcareous soils, salinity and alkalinity. It compares less favorably in terms of its higher proportion of soils low in weatherable minerals and low cation exchange capacity, which are largely located in the desert margins of Africa. In terms of physical soil constraints such as poor drainage, shallow soil depth, steep slopes and gravelly soils, tropical Africa ranks well. At the continental scale therefore, the climate and soils of tropical Africa compare well with other tropical regions where development has advanced further. The main determinant of Africa's position at the bottom of the development scale, therefore is the need for an enabling policy environment, rather than an inferior biophysical endowment. Two other determinants are not so widely recognized in Africa, but the authors believe to be necessary conditions for sustainable development. They are: (1) the need to tackle soil fertility depletion as the fundamental biophysical constraint and (2) the need for more intensive and diverse land use based on profitable tree crops.
declining per-capita food production of the continent (Sanchez et al., 1995; World Bank, 1995b, 1996b; IFPRI, 1996). The magnitude of nutrient mining is huge, as evidenced by nutrient balance studies. Table 2 shows large per-ha losses of nitrogen, phosphorus and potassium during the last 30 years in about 100 million ha of cultivated land in Africa. In contrast, commercial farms in North America and Europe have averaged even larger but positive nutrient balances during the last 30 years is over four times the cultivated land of Africa. This has often resulted in groundwater and stream pollution in the developed world. Nutrient mining in Africa, therefore, contrasts sharply with nutrient buildups in temperate regions. How did this situation come about? Everywhere in the world, people have settled first in high potential areas with fertile soils, adequate rainfall and moderate Table 2 Net nutrient balances in cultivated land of Africa and Europe + North America (calculated by Sanchez et al., 1995 based on studies by Smaling, 1993 and Frissei, 1978) Region
3. Soil fertility r e p l e n i s h m e n t
Soil fertility depletion in smallholder farms of Africa is beginning to be recognized as the fundamental biophysical limiting factor responsible for the
Nitrogen (kg N/ha per 30 years)
Africa -700 Europe + North +2000 America
Phosphorus Potassium (kg P/ha (kg K/ha per 30 years) per 30 years) -100 +700
--450 +1000
22
temperatures, such as in parts of the highlands of Eastern and Central Africa and the plateau of Southern Africa with fertile soils derived from basic rocks. These areas, such as the Lake Victoria basin, now have some of the highest population densities in the world. Originally such populations were supported by the high soil nutrient capital of the predominant high fertility soils (classified as Nitisols in the FAO legend and as Alfisols and eutric Oxisols in soil taxonomy). This nutrient capital has gradually been depleted, primarily through successive crop harvest removals. Leaching and soil erosion have also contributed to these losses which have exceeded nutrient inputs such as biological nitrogen fixation, manures and inorganic fertilizers (Smaling, 1993). A similar situation is now occurring in the inherently less fertile, sandy soils of the West and Southern African subhumid savannas and the Sahel, because of population increases in these more marginal areas. Static or decreasing crop yields in Africa are largely due to nutrient depletion (Borlaug and Dowswell, 1994). In addition to marked reductions in crop productivity, nutrient depletion triggers several negative side effects on farm such as less fodder for cattle, less fuelwood for cooking, smaller amounts of crop residues, and less manure from the cattle. These in turn further increase runoff and erosion losses because there is less plant cover to protect the soils from wind and water erosion. There are also major environmental externalities. Deforestation in search of the few remaining pockets of fertile land in densely populated areas often results in an almost total removal of trees from the landscape, as seen in parts of Ethiopia. Lack of tree protection at the upper parts of watersheds severely affect their functioning. Increasing soil erosion from unproductive cropland, communal grazing lands and denuded watersheds leads to silting of reservoirs, lakes and coastal areas and can lead to the eutrophication of fresh waters. Food shortages and famines become more acute during drought years. Lack of opportunities for cash income pushes people off the land and into urban areas where many cannot find productive jobs are, further taxing the limited urban infrastructure. Urban poverty, crime, and political unrest commonly follow (Homer-Dixon et al., 1993).
Fertilizers have been the traditional tool to overcome soil fertility depletion, and indeed fertilizer use is responsible for a large part of the food production increases worldwide, including in the commercial farm sector in Africa (Mokwunye and Vlek, 1986; Borlaug and Dowswell, 1994). Fertilizer use is viewed as a recurring cost of production, that must be paid for by the increased crop yields farmers obtain. Attempts to introduce this approach to smallholder farming in Africa have met with limited success, even with input subsidies and credit schemes. Current thinking on natural resource management, however, leads us to propose an alternative approach for situations where the traditional one has not worked. Development activities that supply water for agriculture, such as reservoirs and irrigation systems have long been considered as capital investments, paid for by governments and development banks. Users pay the recurrent costs such as maintenance of canals and drainage ditches on the farm. Replenishing plant nutrients cansh also be viewed as a capital investment (Sanchez et al., 1995). Restoring nitrogen and phosphorus, the two most limiting nutrients, to their original levels in the soil, in a way that maintains them and allows them to be used for many years, is a capital investment. Nutrient capital refers to the soil reserves of nutrients that will be released gradually over a time scale of years or decades (Sanchez and Palm, 1996). Nitrogen capital consists of the active and slow pools of soil organic nitrogen. Phosphorus capital consists of the same active and slow pools of soil organic phosphorus plus inorganic phosphorus fixed at the surface of iron and aluminum oxides on clay particles. Replenishment is not feasible with nutrients that are not held by soil organic matter or clay particles, such as potassium in most soils, but mechanisms exist to build up the nitrogen and phosphorus capital in soils where these elements have been depleted. The 'interest' from such capital is used for crop production for years, and with good management, the 'principal' can remain at a high level. Nitrogen needs of most cereal crops at a grain yield level of 4 t/ha can be met by the appropriate use of organic inputs in Africa (Palm, 1995). Nitrogen capital can be built-up gradually through additions of organic inputs, because much of the nitrogen not taken up by the crop plus the carbon present in the
23
organic inputs can be converted into soil organic matter. For higher yields, inorganic fertilizer nitrogen must be added (Sanchez, 1995). Phosphorus capital, however, can only be built through the additions of inorganic phosphorus fertilizers, including rock phosphates and more soluble forms, because organic sources are unable to meet more than half the crop needs (Palm, 1995). The World Bank is beginning to incorporate soil fertility replenishment as a capital investment into their lending policies in Africa (World Bank, 1996b). There is, however, a need for a substantial research agenda that addresses new issues such as the interactions between organic and inorganic sources of nutrients, and the role of trees as recyclers of the 'interest' not utilized by crops, to further increase the returns from the investment. The type of policies required to enable nutrient replenishment need also to be determined. They include conventional returns to investment studies as well as how to determine the social and environmental benefits at the national and global scales, and therefore who should pay for such capital investments. The various benefits of recapitalization resulting from the restoration of all the service flows of nutrient capital are in essence multilayered. One layer consists of the on-farm benefits, another of the national benefits and a third of the global benefits. The issue of who should pay for this recapitalization thus becomes relatively straightforward, in principle at least. Using the fundamental principle that those who benefit from a course of action should incur the costs of its implementation, three layers of costs can be distinguished, corresponding to the three layers of benefits (Sanchez et al., 1995). On-farm, maintenance costs should be borne by farmers, whereas the national and global societies should share the more substantial costs of actual phosphorus applications. This sharing should reflect the ratio of national to global benefits. Research is needed for evaluating these various layers of costs and benefits on the basis of actual field measurements.
4. Intensifying and diversifying land use Soil fertility replenishment can go a long way in boosting agricultural production and restoring food
security in Africa, but it is a necessary but not sufficient condition for sustained development. Numerous other factors have to come together as well. For example, the reduction of post-harvest losses, pests and disease attacks, soil losses by water and wind erosion, as well as the declining size of land holdings and declining human health. The last two impact on the availability of field labor which is also a consequence of family members moving to the town to make offfarm income. Clearly what is needed is a paradigm shift from policies directed only at increasing yields of the few staple food crops. This Green Revolution approach has played, and will continue to play, an important part in meeting the needs of the rural poor, but additional steps also have to be taken. Dewees and Scherr (1996) have indicated that policy scientists need 'to stretch their conceptual framework.., to consider more carefully the links between markets, the environment, household production and household welfare'. It is in this vein that the authors suggest that a further transformation is needed in the long run; intensifying and diversifying land of smallholder farms in Africa as the next determinant. President Yoweri Museveni of Uganda, in his opening address to the SPAAR meeting in Kampala, February 6th, 1996 articulated this idea very clearly. He stated 'it does not make sense to grow low-value products (maize and beans) at a small-scale; instead highvalue products should be grown at a small-scale, while low value products should be grown on a large-scale'. The obvious implication is that small-scale farming in Africa must diversify by producing a combination of high-value, profitable crops along with the basic food crops. Examples of this strategy occur in Western Kenya, where small patches, in the order of 100 m 2, of french beans are grown by smallholders contracted by an exporting company for fresh consumption in Europe. The market is assured and farmers intensively water, fertilize and weed these islands of wealth among their lower value crops. But the largest opportunities for diversification come from tree products. Traditionally people throughout the tropics have depended on their indigenous plants for fruits and everyday products of the household from medicines to fibres. The products have also provided the essential vitamins and minerals for family health and
24
through local and regional trading have generated the cash to meet the needs of the household for purchased products. Maybe it is here, in peoples' own backyard, that the solution lies. But sadly, through deforestation, the forest or woodland that used to be in the backyard is now all but disappeared for the vast majority of people in Africa. In 1992, a conference pulled together the growing amount of biophysical information on the techniques available to domesticate the wide range of the wild and overlooked species, many of which are trees (Leakey and Newton, 1994a,b; Newton et al., 1994). These 'Cinderella' species (so called because their value has been largely overlooked by science although appreciated by local people) include indigenous fruit trees and other plants that provide medicinal products, ornamentals or high-grade timber. Examples are the bush mango (Irvingia gabonensis) a nutritious fruit from the humid tropics of West Africa (Ladipo et al., 1996), Uapaca kirkiana and Sclerocarya birrea, from the Miombo woodlands of Southern Africa. Techniques are being developed to convert some of these wild species into domesticated crops in agroforestry systems including vegetative propagation and clonal selection that capture genetic diversity (Leakey and Jaenicke, 1995). Domestication involves the formulation of a genetic improvement strategy for agroforestry trees and a strategy on the use of vegetative propagation to capture the additive and non-additive variation of individual trees in tree populations (Simons, 1996). Furthermore, guidelines have been developed for determining the species priorities of farmers (Franzel et al., 1996; Jaenicke et al., 1996). The domestication strategy for these indigenous fruit tree species, as well as for Prunus africana and Pausinystaliajohimbe, two priority trees for medicinal products, is to conserve their genetic resource in living-germplasm banks and subsequently to develop cultivars for incorporation into multistrata agroforests. High-value trees can fit in specific niches on farms while leaving more open land to staple food crops, or other profitable crops such as vegetables. Timber trees can also be grown on farm boundaries with leguminous fodder trees under them. Similarly fuelwood trees can be grown on field boundaries or as contour hedges on sloping lands. In such a scheme, improved fallows become a crucial part of the crop rotation
scheme. In such farms, income is increased and diversified, providing resiliency against weather or price disruptions. Soil erosion is minimized, nutrient cycling maximized and above- and below-ground biodiversity enhanced. The farm truly approximates a functioning ecosystem. The latest definition of agroforestry summarizes this approach: a dynamic, ecologically-based, natural resource management system that, through the integration of trees in farm and rangeland, diversifies and sustain smallholder production for increased social, economic and environmental benefits (Leakey, 1996). Through domestication these tree crops could become higher yielding, have higher quality products, be more attractive commercially and diversify diets. Such progress could improve household welfare by providing traditional food and health products, boosting trade, generating income and diversifying farming systems, both biologically and economically, beyond the production of basic food crops. Generally tree crops have lower labor requirements than basic food crops and could thus allow farmers to also seek offfarm income. A new paradigm for smallholder farming in Africa emerges: one which instead of being based on a limited number of highly domesticated crops, often grown in monoculture, is based on a much greater diversity of plants that together produce food and high-value products (Leakey and Izac, 1996). What evidence is there that it will work? The African experience with homegardens provides some examples of traditional intensification. In Nigeria, for example, they are important areas with high population densities (1000 people/km2). In such areas up to 29% of the cultivated area can be in compound gardens and these produce 59% of the crop output (Okafor and Fernandes, 1987). In monetary terms the output of these gardens is 5-10 times greater than crop fields, with returns to labor also 4-8 times greater (Watson, 1990). But the concept we are advocating is a total farm approach, not just a comer of it as in the case of homegardens. A better model is found in Southeast Asia, where the productive diversity is extended from homegardens to complex agroforests. In south-west Sumatra, near Krui, farmers have for more than a century been planting damar trees (Shoreajavanica) for resin production, in mixture with a range of crops and indigen-
25
ous fruit tree species. Originally, farmers cleared the forest by slash and burn methods and planted upland rice, along with coffee, fruit and damar trees. Lowland rice patches are also part of the farming system. The area in complex damar-based agroforests is contiguous and now exceeds 10000 ha and 65-80% of the households are involved in damar production each averaging 1-5 has of these agroforests. The system insures that the farm is productive at all stages of its growth with food crops in years 1-3, coffee and bananas in years 3-8, fruits and fuelwood in years 2-20, resins, fruits and timber from year 20 onwards (De Foresta and Michon, 1994), and paddy rice all the time. In the case of damar, the resins are utilized by industries in Indonesia or exported worldwide. In 1984, the export market represented one third of the harvested volume, a trade rising from 250-400 t/year between 1972 and 1983 (Michon and de Foresta, 1994). In 1994, the damar production was expected to reach 10000 t (Dupain, 1994), at a value of US $300-400/t. Eighty percent of this trade is met by the damar agroforests. The economic value of the damar trade and its associated activities is of major significance to the villages around Krui. In 1993, the profits from damar production were US $7.2 million from sales, US $2.6 million from added value and US $1.4 million from wages. To this is added US $ 0.3 million in profits made by Krui traders (De Foresta and Michon, 1994). This analysis excludes the locally consumed products from these agroforests, e.g. fruits, vegetables, spices, fuelwood, timber, palm thatching, rattan, bamboo, fibres, as well as paddy rice. Systems similar to the damar agroforests are also practiced in Sumatra with rubber and cinnamon. The 'jungle rubber' agroforests covered more than 2.5 million ha, while cinnamon agroforests covered 42 600 ha in 1989 (Aumeeruddy, 1994). These damar, cinnamon and jungle rubber complex agroforests are probably the ultimate example of an alternative to slash and burn agriculture, being highly productive, biodiverse, and very similar in structure to natural tropical high forest. In short, they are the nearest thing to a domesticated ecosystem (Michon and de Foresta, 1996). While land use intensification caused by demographic pressure is generally associated with envir-
onmental degradation, the long-term relationship between land resource degradation and demographic pressure is not necessarily negative and linear. Harwood (1994) has described a U-shaped curve, with an initial fast rate of land degradation with subsistence agriculture, increasing with increasing population pressure. With further increases in population pressure, however, a point is reached where degradation is reversed with further land intensification and incorporation of trees within the farm. This has happened in the semiarid Machakos District of Kenya, where despite increasing population pressure since the 1930s farmers were able to reverse land degradation through an indigenous soil conservation technology that improved both crop and livestock productivity (Pagiola, 1994; Tiffen et al., 1994). Furthermore, in the more heavily populated areas of the Central Province of Kenya, where farm size is extremely small, the number of trees on farms is also expanding as farmers increasingly recognize their value (Cleaver and Schreiber, 1994).
5. Enabling policies Current policy recommendations place high priority on the revitalization of the agricultural sector in Africa (IFPRI, 1996), and some success stores are beginning to emerge (Cleaver and Schreiber, 1994). The fact that most food in Africa is produced by smallholders, often female farmers is frequently considered a major constraint to agricultural development. In contrast, these authors believe that smallscale farms can be an asset rather than a liability when supported by appropriate policies, the agricultural production boom in Asia is a product of smallholder farms and not of a shift from small to largescale farming. Such policies include improvements in land tenure, infrastructure, marketing information, credit, research, extension and access to inputs and markets at reasonable prices. Public investments to expand access to primary education to girls and improve public health services in rural areas also play an important role in this transformation process. Policy reform to seize opportunities for smallholder development and to eliminate policies that discriminate against the smallholder agricultural sector therefore remains top priority. Policy reform, therefore, is a
26 necessary but not a sufficient condition for food security and environmental conservation. In order for enabling policies to work in most of Africa the twin issues of soil fertility depletion and land-use intensification and diversification have to be tackled. Thus the vision now is of agroforestry as an integrated land use policy that combines increases in productivity and income generation, with environmental rehabilitation and the diversification of agro-ecosystems. Such a vision can be fitted to the range of situations found in the major ecoregions of Africa. According to Cooper et al. (1996), the realization of this vision, however, is going to be dependent on: •
•
•
the appreciation by the international community and the donors, of the importance of high value indigenous species in the lives and welfare of local people, as well as incentives (or the removal of disincentives) for local people to plant trees on their farms, the domestication of commercially-important indigenous tree species producing high-value products, and the development of processing and marketing infrastructure.
For this latter step, it will be important for scientists involved in domestication to work closely with the food and pharmaceutical industries since the agroforester needs to know that there will be a market, while the industry which develops the market wants to know that there is a regular supply of a uniform and high quality product, before committing capital to developing that market (Leakey and Izac, 1996). Commercialization is both necessary and potentially harmful. It is necessary because without it the market for products is small, and the opportunity for rural people to make the money would not exist. A degree of product domestication is therefore essential. On the other hand commercialization is potentially harmful to rural people, if it expands to the point that outsiders with capital to invest, come in and develop large-scale moncultural plantations. However, from the experience of the Southeast Asian complex agroforests, smallholder units producing non-timber forest products, which are also biologically-diverse, do seem to be economically viable indicating that the intensification and diversification of land uses is not a pipe-dream.
6. Conclusion
The dynamic tension between increasing population, declining farm-size, declining farm labor supply, diversification of income sources and the management of natural resources varies with regions. For SubSaharan Africa, where most of the food is produced by smaUholders this tension should encourage farmers and policy makers to replenish the fertility of exhausted soils and plant high-value trees as a diversified but intensive agroecosystem, based on profitable tree crops, as the way forward socially, economically and environmentally.
References Aumeeruddy, Y., 1994. Local representations and management of agroforests on the periphery of Kerinci Seblat National Park, Sumatra, Indonesia. People and Plants Working Paper 3, UNESCO, Paris, 46 pp. Badiane, O. and Delgado, C.L., 1995. A 2020 vision for food, agriculture and the environment in sub-Saharan Africa. Food, Agriculture and the Environment Discussion Paper 4, IFPRI, Washington, 56 pp. Borlaug, N. and Dowswell, C.R., 1994. Feeding a human population that increasingly crowds a fragile planet. Supplement to Transactions 15th World Congress of Soil Science, Acapulco, Mexico. International Soil Science Society, Chapingo, Mexico, 15 pp.. Brown, L. and Haddad, L., 1994. Agricultural growth as the key to poverty alleviation. 2020 Brief No. 7, IFPRI, Washington, 2 pp. Cleaver, K.M. and Schreiber, G.A., 1994. Reversing the spiral; the population, agriculture and environment nexus in sub-Saharan Africa. World Bank, Washington, 293 pp. Cooper, P.J.M., Leakey, R.R.B., Rao, M.R. and Reynolds, L., 1996. Agroforestry and the mitigation of land degradation in the humid and sub-humid tropics of Africa, Exp. Agric., 32: 000-000. Cunningham, A.B. and Mbenkum, F.T., 1993. Sustainability of harvesting Prunus africana bark in Cameroon. A medicinal plant in the international trade. People and Plants Working Paper 2, Paris, UNESCO, 28 pp. De Foresta, H. and Michon, G., 1994. Agroforests in Sumatra: where ecology meets economy. Agrofor. Today, 6: 12-13. Dewees, P.A. and Scherr, S., 1996. Policies and markets for nontimber tree products. EPTD Discussion paper, IFPRI, Washington, 76 pp. Dupain, D., 1994. Une regione traditionaliment agroforestiere en mutation: Le Pesisir. CNGARC, Montpellier, France. Franzel, S., Jaenicke, H. and Janssen, W., 1996. Choosing the right trees: setting priorities for multipurpose tree improvement. ISNAR Research Report No. 8., ISNAR, The Hague, The Netherlands, 87 pp.
27
Frissel, M.J., 1978. Cycling of Mineral Nutrients in Agricultural Ecosystems. Elsevier, Amsterdam, 356 pp. Harwood, R.R., 1994. Agronomic alternatives to slash-and-burn in the humid tropics. In: P.A. Sanchez and H. van Houten (Editors), Alternatives to Slash and Burn Agriculture. Symposium ID-6, 15th World Congress of Soil Science Acapulco, Mexico. International Society of Soil Science, Chapingo, Mexico, pp. 93106. Homer-Dixon, T.F., Boutwell, J.H. and Rathfens, G.W., 1993. Environmental change and violent conflict. Sci. Am., 268 (2): 16-23. IFPRI, 1996. Feeding the word, preventing poverty and protecting the Earth: a 2020 vision. International Food Policy Research Institute, Washington, WA, 28 pp. Jaenicke, H., Franzel, S. and Boland, D.J., 1996. Towards a method to set priorities among species for tree improvement research: a case study from West Africa. J. Trop. For. Sci., 7: 490-506. Ladipo, D.O, Fondoun, J.M. and Nganga, N., 1996. Domestication of bush mango (lrvingia spp. ): some exploitable intraspecific variations in West Africa. In: R.R.B Leakey, A.B Temu and M. Meinyk (Editors), Domestication and Commercialization of Non-timber Forest Products in Agroforestry Systems. NonWood Forest Products 9, FAO, Rome, (in press). Leakey, R.R.B. and Newton, A.C., 1994a. Domestication of 'Cinderella' species as a start of a woody plant revolution. In: R.R.B. Leakey and A.C. Newton (Editors), Tropical Trees: Potential for Domestication and the Rebuilding of Forest Resources, HMSO, London, pp. 3-6. Leakey, R.R.B. and Newton, A.C., 1994b. Domestication of Timber and Non-timber Forest Products, MAB Digest 17, UNESCO, Paris, France, 94 pp. Leakey, R.R.B. and Jaenicke, H., 1995. The domestication of indigenous fruit trees: opportunities and challenges for agroforestry. In: K. Suzuki, S. Sakurai, K. Ishii and M. Norisada (Editors), Proc. 4th International BIO-REFOR Workshop, BIO-REFOR, Tokyo, Japan, pp. 15-26. Leakey, R.R.B., 1996. Definition of agroforestry revisited. Agrofor. Today, 8 (l): 5-7. Leakey, R.R.B. and Izac, A.-M.N., 1996. Linkages between domestication and commercialization of non-timber forest products: implications for agroforestry In: R.R.B. Leakey, A.B. Temu and M. Melnyk (Editors), Domestication and Commercialization of Non-timber Forest Products for Agroforestry. NonWood Forest Products, 9, FAO, Rome, Italy, (in press). Michon, G. and de Foresta, H., 1994. Damar agroforests in the Passisir, Sumatra. Rainforest Alliance, May 1994, New York, 45 pp. Michon, G. and de Foresta, H., 1996. Agroforests as an alternative to pure plantations for the domestication and commercialization of NTFP's. In: R.R.B. Leakey, A.B. Temu and M. Melnyk (Editors), Domestication and Commercialization of Non-timber Forest Products in Agrofor. Syst. Non-Wood Forest Products 9, FAO, Rome, (in press). Mokwunye, A.U. and Vlek, P.L.G., 1986. Management of Nitrogen and Phosphorus Fertilizers in Sub-Saharan Africa. Martinius Nijhoff, Dodrecht, 362 pp. Newton, A.C., Moss, R. and Leakey, R.R.B., 1994. The hidden
harvest of tropical forests: Domestication of non-timber products, Ecodecision, 13 (July): 48-52. Okafor, J.C. and Fernandes, E.C.M., 1987. Compound farms of southeastern Nigeria: a predominant agroforestry homegarden system with crops and small livestock. Agrofor. Syst., 5: 153168. Pagiola, S., 1994. Soil conservation in a semi-arid region of Kenya: rates of return and adoption by farmers. In: T.L. Napier, S.M. Camboni and S.A. EI-Swaify (Editors), Adopting conservation on the farm. Soil and Water Conservation Society, AIkeny, IA, pp. 171-187. Palm, C.A., 1995. Contribution of agroforestry trees to nutrient requirements of intercropped plants. Agrofor. Syst., 30: !05124. Sanchez, P.A., 1995. Science in agroforestry. Agrofor. Syst., 30: 5-55. Sanchez, P.A. and Logan, T.J., 1992. Myths and science about the chemistry and fertility of soils in the tropics. SSSA Special Publication, Soil Science Society of America, Madison, Wl, 29: 35-46. Sanchez P.A. and Palm, C.A., 1996. Nutrient cycling and agroforestry in Africa. Unasyiva, 185 (47): 24-28. Sanchez, P.A., Couto, W. and Buol, S.W., 1982. The fertility capability soil classification system: interpretation, applicability and modification. Geoderma, 27: 283-309. Sanchez, P.A., lzac, A.-M.N., Valencia, I. and Pieri, C., 1995. Soil fertility replenishment in Africa: a concept note. ICRAF, Nairobi, 5 pp. Simons, A.J., 1996. ICRAF's strategy for domestication of nonwood tree products. In: R.R.B. Leakey, A.B. Temu and M. Melnyk (Editors), Domestication and Commercialization of Nontimber Forest Products in Agroforestry Systems. Non-Wood Forest Products 9, FAO, Rome, (in press). Smaling, E., 1993. An agroecological framework for integrated nutrient management with special reference to Kenya. Doctoral thesis, Agricultural University, Wageningen, The Netherlands, 250 pp. Tiffen, M., Mortimer, M. and Gichuki, F., 1994. More People, Less Erosion. Environmental recovery in Kenya. Wiley, Chichester, UK, 311 pp. Watson, G.A., 1990. Tree crop and farming systems development in the humid tropics. Exp. Agric., 26: 143-159. World Bank, 1995a. Towards Environmentally Sustainable Development in Sub-Saharan Africa: a World Bank Agenda. Report No. 15111-AFR, World Bank, Washington, WA, 114 pp. World Bank, 1995b. Feasibility of Phosphate Rock Use as a Capital Investment in Sub-Saharan Africa: Issues and Opportunities. Africa Technical Division, World Bank, Washington, WA, 72 PP. World Bank, 1996a. African Development Indicators 1996. World Bank, Washington, WA, 431 pp. World Bank, 1996b. Natural Resource Degradation In Sub-Saharan Africa: Restoration of Soil Fertility. Africa Region. World Bank, Washington, WA, 12 pp. World Resources Institute, 1992. World resources 1992-1993, A Guide to the Global Environment. Oxford University Press, New York, pp. 28 ! - 284.
This Page Intentionally Left Blank
© 1997 ElsevierScience B.V. All rights reserved Perspectives for Agronomy - Adopting Ecological Principles and Managing Resource Use M.K. van Ittersum and S.C. van de Geijn (Editors)
29
Agro-ecological characterisation, food production and security P. Bullock * Soil Survey and Land Research Centre, Cranfield University, Silsoe, Bedford MK45 4DT, United Kingdom Abstract
There has been a large increase in food production in the last 30 years which has kept pace with the growing population. There is, however, concern that increased production has been at the expense of increasing degradation of the natural resources and hence is not sustainable. There is growing belief that agricultural development has paid insufficient attention to basic ecological principles and that agriculture needs to be subject to a paradigm change which incorporates these principles into current and future developments while at the same time being cognisant of the continued need for external inputs to achieve adequate production. The FAO Agro-Ecological Zones Project is a step towards agro-ecological harmony. It provides a basis for informing planners about land potential, the suitability and yield potential for some 11 major crops important in the developing world and the potential sustainability of the various cropping systems. It was one of the first schemes also to quantify the amount of potentially cultivable land in developing countries. New developments, which now allow the agroecological zones approach to be strengthened and made more available, include the availability of well organised national and international land-related databases, the development of crop suitability and production models of varying sophistication, models for environmental risk assessment including risks from the adoption of particular agricultural practices, and geographical information systems that now allow information to be put to policy makers and others in a simple and useable format. To feed the predicted future world population is a challenge on an unprecedented scale. It needs to be met with a combination of a sound agro-ecological approach to food production, biotechnological development, carefully controlled and understood inputs, improved technology, more investment in deprived areas and well thought out strategies for food production, access and trade.
Keywords: Agro-ecological zones; Crop growth models; Environmental risk assessment; Sustainable land use and management; Food production; Food security
1. Introduction
The last few decades have seen a number of major changes in agriculture and food production. Prominent among these has been the Green Revolution which has had a major impact on agriculture worldwide. Using improved plant and animal breeding, large fertiliser inputs, improved technology and other external inputs, productivity has increased greatly in many parts of the world enabling the population to
* Fax: +44 1525 863253. Tel: +44 1525 863251. E-mail:
[email protected]
be better fed than at any time in history. Agricultural production grew faster than population with per caput production about 18 per cent above that of 30 years ago (Alexandratos, 1995). However, this success is moderated by two facts: (i) approaching one billion people suffer from hunger, largely because of problems of getting surplus food to where it is needed, drought and warfare and (ii) the increased food production in some areas has been at the expense of increasing degradation of our natural resources. Thus, it is estimated that some 6-7 million hectares of land are lost each year to soil erosion, about one quarter of the world's land is suffering from desertification and some 20 million hectares
30
Table 1 Sustainable land use and management principles for future food production Economic viability Strong productivity Working with indigenous land users In harmony with the soil and prevailing climate According to ecological principles Environmentally benign Preserving and preferably enhancing soil quality Careful water use and management have been largely lost through salinisation (World Resources Institute, 1992). Not all agriculture has been marked by increased productivity. Several countries and whole regions, mainly in Africa, failed to make progress. Many subsistence farmers of the developing world are experiencing decreasing yields, caught in a poverty trap which precludes them from using fertiliser to at least keep pace with the loss of nutrients from the land (Stoorvogel and Smaling, 1990; van der Pol, 1992; FAO, 1996). Most such farmers are farming in extremely fragile circumstances with, in addition to poverty, an unreliable climate and soils of depleting fertility. In Africa, food production per person has fallen by 20% since the 1960's (Pretty, 1996). Thus with both intensive and subsistence agriculture there is concern over the sustainability of agricultural production. The historical association of expansion and intensification of agriculture with pressures leading to resource degradation and adverse impacts on the wider environment is now well documented (FAO, 1996). Such pressures are likely to increase in the future. How to minimise the negative effects on the land resources will be a major challenge for future agricultural production. It is time to review, learn from history and develop a strategy for a more soundly based agriculture. That
there is a degree of will to do this is suggested by the Rio Summit and Agenda 21. Sustainable development issues are fundamental to a strategy for sound adequate food production and food security, guided by the principles outlined in Table 1. Meeting food and fibre demands of a future population will require that land currently in production produces more, that land currently not in production is brought into production and, where reasonable that degraded land is rehabilitated to a reasonable productive capacity (Pierce and Lal, 1991). There is a growing belief that agricultural development needs to be more mindful of ecological principles and rather than pursuing external inputs alone as the panacea for increased production, a better balance involving external inputs, natural processes and resource-conserving technologies is required. The concept of agro-ecological zones, if followed, would lead to a more balanced, more harmonious approach to agricultural production, necessary for future sustainability.
2. Agro-ecological principles Pierce and Lal (1991) and others have highlighted the importance of applying ecological principles to agricultural systems as part of the sustainable development strategy. Cox and Atkins (1979) and Pierce and Lal (1991) point out that agricultural systems, though managed systems, function with the same group of physical and biological laws as other ecosystems but differ from the latter in both structure and function in a number of ways (Table 2). A better understanding of the importance and implications of ecological processes in the sustainable agricultural paradigm needs to be developed. The concept of the agro-ecological approach recognises that crop
Table 2 Comparisons of Agricultural and Natural Ecosystems(Pierce and Lal, 1991) Agrosystems
Natural Ecosystems
Highly productive with huge energy and material subsidies Lower productivity: few or no subsidies Simple ecosystems, resembling early stages of succession; hence unstable More complex, higher diversity and higher stability Artifically regulated; little continuity Self-regulating and self perpetuating; considerable continuity Not adaptive Adaptive Not very diverse Diverse Artificial management Nature's management
31
production worldwide is influenced by a complex interaction of several factors, particularly the species and cultivar of the crop being grown, the nature of the climate, the soil type in which the crop is grown and the use of inputs, such as fertilisers and pesticides (FAO, 1991). It does not preclude the use of external inputs to maintain production levels but attempts to follow ecologically sound concepts and cultural practices in doing so. In the view of Sehgal et al. (1990), the need for increased food production to support the growing population 'demands an appraisal of our soil and climatic resources to recast an effective and alternative land use plan. Since soils and climatic conditions of a region largely determine the suitability of different crops and their yield potential, efforts in mapping agro-ecological regions may go a long way in identifying optimum cropping patterns for increasing agricultural production on a sustainable basis.' The food security of future populations will be dependent on developing sustainable agriculture. A combination of an agro-ecological approach and a better understanding of the significance of ecological processes and their application in crop production systems are essential if this is to be achieved.
3. The Agro-Ecological Zone (AEZ) concept
0FAO 197S, ~991) The Agro-Ecological Zone concept is an important one in bringing world agriculture back towards ecologically based, more sustainable principles. FAO has been one of the chief developers of the concept and some 15 years ago introduced the AEZ system in support of developing countries. The types of question that the establishment of the FAO agro-ecological zones aimed to address were (FAO, 1991): how much potentially arable land is available and its location what crops can be grown, and what are their potential yield what level of inputs are required to produce the necessary yields what is the risk of land degradation and how can it be minimised
which crops in which areas would produce the greatest return given more resources what would be the cost of realising these returns what are the maximum returns available where should future research be concentrated
3.1. The parameters for delimiting agro-ecological zones
A number of sets of information were assembled for the FAO agro-ecological zone delimitation: Crop Requirements: To define, with respect to climate and soil, the needs of the following crops: wheat, paddy rice, maize, pearl millet, sorghum, soybean, cotton, phaseolus bean, white potato, sweet potato and cassava (Fig. 1). Climate Information" Emphasis was given to temperature and water availability, the combination of which is expressed in the growing period. The length of the growing period is assumed to be the continuous period from the time when rainfall is greater than half the potential evapotranspiration until the time when the rainfall is less than the full potential evapotranspiration, plus a number of days required to evaporate an assumed 100 mm of soil moisture reserve when available (FAO, 1978, Sehgal et al., 1990). Soil Information: The AEZ project used the FAO 1 "5 million soil map as its information base. Four basic variables are included on the soil map. The soils themselves are divided into 26 major groups and further subdivided to give a total of 106 individual soil units. The map also specifies three categories of texture and of slope. There is also additional information in the form of different soil phases, such as stoniness and salinity. The Land Inventory: In the FAO project, the climatic inventory was combined with the soil map to produce a land inventory map which contains the accumulated data on climate, thermal regime, length of growing period zones, soil associations, slope, phase and texture. The map represents the information base of the study. Calculations were then made of the maximum potential yield of each crop in each zone, taking into consideration how ideal the soil conditions were, problems of soil management where
32
Choice of crops and identification of their climatic and soil requirements Assembly of agroclimatic data Assembly of soil data Assessing agroclimatic suitability Assessing land potential for particular crops
Potential population supporting capacity
Fig. 1. Agro-ecological zones production stages.
climatic conditions were less than ideal, to produce a land suitability assessment for each particular crop. The AEZ approach can provide answers to questions that are an important aid to planning and development for food security. These include (FAO, 1991):which crops are best suited to the soil and growing conditions in any country? what agricultural policies would make best use of land resources of a particular country? where are the potential food-deficit areas that cannot attain self-sufficiency? where are the best prospects for examining production including cash crops? which areas, with high production potentials, could provide food for, or support people moving from food-deficit areas? what levels of farming inputs and soil conservation need to be developed to meet specific targets for sustainable self-sufficiency and crop exports? what are the future requirements for seed, fertilizer, pesticide and agricultural power to meet future population demands? The AEZ project was one of the first to quantify systematically the extent of potentially cultivable land in developing countries, to provide estimates of crop yields under varying levels of inputs and helped to provide a basis for identifying the popula-
tion-supporting capacity of developing lands (Higgins and Kassam, 1981 ; FAO, 1991. It has been used to inform planners about land potential, which crops are most suitable for a given area and what levels of sustainable production can be expected under different levels of input. It also provided a balance sheet to test whether agricultural strategies were compatible with land resource potentials. The conclusions reached by the agro-ecological zone project were that in 1975 some 55 out of the 117 countries studied could not support their populations from their own land resources with low-input production systems. The area of these countries was 32.5% of the entire developing world. However, the situation changed dramatically if the level of inputs was improved. By the year 2000 the number of countries unable to support their own populations was calculated to be 65 if low inputs were applied, 36 at intermediate inputs, 19 even if high inputs were applied. The project was important also in estimating that there was potential for about 300 per cent increase in the existing arable area in the developing world, but more importantly, it enabled the potential land available to be determined on a country or regional basis. In particular, it showed that while some countries have large amounts of land currently unused for agriculture but suitable for cultivation, e.g. many African and south America countries, in other parts of the world such as southern Asia and the Near East/North Africa, the amount of currently unculti-
33
Table 3 Land With Crop Production Potential in Developing Countries - Land in Use and Land Balance (million ha) (Alexandratos, 1995)
Sub-Saharan Africa Near East/North Africa East Asia (exc. China) South Asia Latin America and Caribbean
Land in Use 1988-90
Balance
Land in Use 2010
Balance
213 77 88 191 190
797 16 97 38 869
255 81 103 195 217
754 13 82 34 842
vated land available to feed increasing populations is relatively small. Predictions of the land use in 19881990 and in 2010 and the balance of land available have been given by Alexandratos (1995), (Table 3).
sustainable agricultural development and this involves adopting the principles of the agro-ecological zones.
3.2. Adoption of the AEZ approach
4. Developing on the AEZ Foundation
The AEZ approach has much to offer in terms of assessing the capability of a country to feed its current and future population. It has been promoted by FAO who have applied the system in, for example, Mozambique (FAO, 1982), Bangladesh (FAO, 1988) and Kenya (FAO 1991). The Technical Advisory Committee of the Consultative Group on International Agricultural Research (CGIAR) has adopted the FAO AEZ approach to subdivide some of the main climatic zones into rainfed moisture zones and thermal zones (TAC/CGIAR, 1992). Their approach was weakened by the exclusion of soils and landform data on the grounds of keeping the number of subdivisions to manageable proportions but provides valuable information on land use patterns and major food crops in the different regions of the developing world. Recently, Sivakumar and Valentin (1997) have examined the crop production potential of different developing region AEZs including soil and landform constraints. However, despite the information it is able to provide and the message it carries, few countries in the developing world have integrated the agro-ecological zone approach into their strategic agricultural development. In view of the fact that the AEZ approach is in harmony with strategies for sustainable development, surely it is only a matter of time before its approach becomes commonplace. To date, rather little attention has been given to using the AEZ approach in developed countries, the implication being that much more is known about the carrying capacity and potential of land there. But developed countries also need to pay attention to
In the last decade there have been at least four major developments which permit significant advances beyond the AEZ project.
4.1. National and international databases The first of these advances relates to improvements in national and international databases. Generally, climatic databases are of long standing and although the detail and types of measurement varies from country to country, the databases are well established. Developing countries fall short of adequate climatic information and this and the vagaries of the climate make crop suitability modelling and yield predictions difficult. Concern in the last decade about climate change and the need to improve General Circulation Models (GCMs) is impacting positively on the collection and management of climatic data worldwide (Houghton et al., 1996) and better datasets are likely to become available. Soils databases have a much shorter history than those relating to climate but there have been major developments in these in recent years. At an international level there are now several databases" the FAO soil database, the one used in the AEZ project; GRID, a UNESCO-organised database including both soils and climatic information; and the International Geosphere-Biosphere Programme is also promoting an international soil database involving inputs of data from FAO, the US Soil Conservation Service, ISRIC and SSLRC on behalf of the European Soil Database. Within Europe there is spatial information about
34 soils available at a scale of I'IM, including a digital version of the map supported by an analytical database (Breuning-Madsen and Jones, 1994), which together constitute the European Soil Database. The amount and scale of information about national soils is variable. That for many countries of Europe is among the most comprehensive and best organised in the world. Le Bas and Jamagne (1996) provide an update of the mapping and supporting databases available for each of the countries of the European Commission.
4.2. Crop growth and crop suitability models The second significant improvement of recent years has been the development of models of crop growth and crop suitability. Models are now available for most of the major crops and these continue to be refined. Several such models predict the growth and production of annual crops according to crop species, soil type, moisture conditions and weather during the growing season. Some are concerned with a single crop, others are broadly based across a number of crops. Some of the principal models include:CRIES"
Comprehensive R_.esourceIndustry and E_valuation System. Includes a yield model which allows prediction for some 30 different crops (Schultink, 1987). DSSAT (D...ecisionSupport S_ystem for Agrotechnology Transfer). Contains a number of crop models (Singh et al., 1990). WOFOST (W...~orldF..~oodStudies). Models the growth and production of annual crops (van Diepen et al., 1989). Now in Version 6.0. ALMANAC (Agricultural Land Management _Alternatives with Numerical Assessment Criteria). A predictive and process orientated model simulating hydrology and erosion; able to simulate competition between 2-10 species (Jones and O'Toole, 1987). EPIC (Erosion Productivity Impact Calculator). Predicts the long term effects of various components of soil erosion on crop production (Williams et al., 1990). ACCESS (Agroclimatic Climate Change and European Soil Suitability). Models the impact of climate change on the suitability of land for strategic European crops (Loveland et al., 1994).
A recently produced Register of Agro-Ecosystems Models (Plentinger and Penning de Vries, 1996) lists
some 200 different models currently available, the number giving a clear indication of the efforts being put into the development of simulation and other models. The potential and attainable levels of food production in the different agroclimatic regions using a modelling approach has been discussed by Penning de Vries (1997).
4.3. Environmental risk assessment Although problems of land degradation associated with some agricultural systems have been long acknowledged, a much wider group of environmental problems are now recognised e.g. water quality, air pollution, soil contamination. There are now a large number of environmental risk assessment models available with which a particular agricultural land use and management can be assessed from the point of view of likely damage to the environment (Plentinger and Penning de Vries, 1996). Examples include:LEACH M
EPIC EUROSEM SUNDIAL
(Leaching Estimation and Chemistry Model). The model includes sub models for water flow, pesticide movement, nitrogen dynamics and salinity (Wagenet and Hutson, 1989). (see above) for soil erosion. (European Soil Erosion Model) (Morgan et al., 1994). (Simulation of Nitrogen D_.ynamicsin Arable Land) (Bradbury et al., 1993).
4. 4. Use of Geographical Information Systems (GIS) The fourth major advance is that of the development and widespread use of GIS in predictive modelling for both crop suitability and environmental risk. The beneficial effects of being able to integrate many different types and sources of information and provide output according to selected models is huge. One of the reasons that uptake of the AEZ approach by national planners has been disappointing is the lack of availability in the past of suitable inexpensive tools to handle multiple databases. Perhaps more than any other single tool in recent years, GIS has revolutionised the opportunities for expert systems which can be integrated into decision support systems through which agroecological zones can be better defined, crops and their management systems se-
35
recognition that food production is not only about economics. While economics are very important, agricultural systems will need to be managed in the future in a different way to what they have been in the past, with more consideration of the land resource itself and the wider environment (Fig. 2). It needs to be recognised that much land in the future that is marginal to agriculture may need to be used for food production. Such lands are very vulnerable to damage and will need to be managed with care to achieve both adequate productivity and protection of the land resource. a better understanding of the needs and inputs of agricultural systems within the environment in which they are being applied. Thus for a given crop, multiple crops or rotational systems, it is reasonable to need to know their impact on the land resource after 5 years, a decade and so on, i.e. what is the anticipated lifespan of the system in that particular agro-ecological zone and what changes can be made to prolong that life and still
lected and by which a sustained land use strategy can be developed. Not only has GIS improved the technological handling and management of data but has provided a means of information transfer to the policy maker on the one hand and the land user on the other in a simple, understandable form.
5. An agro-ecological approach to future food production The Agro-Ecological Zones (AEZ) project of FAO provides an important contribution to the attempt to build sustainable agricultural production for the future. The following are now needed:delimitation of agro-ecological zones world wide using the latest databases, models and information about the agricultural systems likely to be involved in future production. Policy makers in every country need to be made aware of the importance and urgency of developing food production strategies based on agro-ecological principles.
MANAGING THE LAND RESOURCE
MAXIMISIN PROFIT AND PROFITABILITY
PREVENTING SOIL DEGRADATION
'REDUCING ENVIRONMENT POLLUTION
REDt rCING EMISSIONS FROM SOIL TO ATMOSPHERE
•
Economic viability
•
Appropriate land use
•
Soil erosion control
•
Monitoring land use change
•
Technology development
•
Conservation technology
•
Crop nutrient management
•
Carbon sequestration
•
Growing the soil to grow the crop
•
Soil enhancing cropping systems
•
Reducing industrial emissions
•
Managing CH4 source and sink
•
Farming by soil, landscape and climate
•
Managing the noncrop period
•
Catchment management
•
Catchment management
•
•
Fanning by soil type, landscape and climate
Fig. 2. Managing the land resource (after Pierce and Lal, 1991).
Managing fertiliser USe
36
maintain a suitably productive system. a research and development programme that is strongly focused on improving the quality and productivity of agricultural systems. There will be great dependence on biotechnology to play a major part in this, perhaps in the course of time to develop crops specifically for given agro-ecological zones. Advances in biotechnology will need to be supported by research into managing water movement and water storage in soils, manipulation of nutrient dynamics, harnessing micro-organisms, seeking more efficient fertiliser use and improving organic matter levels. greater awareness on behalf of the policy makers, development of national strategies for food production and security and linking of these world wide to manage food production and security. At a global strategic level, special attention and recognition will need to be given to sub-Saharan countries, and to others caught in the trap of subsistence farming. Strategies will need to be developed for increasing their productivity in a sustainable manner. Much of this will depend initially on large scale investment and on adequate mechanisms for technology transfer.
6. Overview of food production and food security
There have been enormous increases in food production in the past 50 years in the developed countries and this has more than kept pace with the population increase in those countries. In some situations there has been overproduction and storage problems of unused food have tended to make the headlines. The problems of transferring the excess food from one area to another with need for food have not been adequately solved. Dominated by subsistence farming, productivity appears to be declining in sub-Saharan countries rather than increasing (Harrison, 1987; FAO, 1993) although pockets of improvement are the basis for restrained optimism (Pretty, 1996). The rehabilitation and development of suboSaharan agriculture is a critical regional problem with international dimen-
sions. The problem of improving food consumption levels in sub-Saharan countries needs to be given centre stage in any debate on the future of the world food economy (Alexandratos, 1995). A number of studies suggest that the lands of the world could provide increased productivity and food production provided current trends in food production supported by improved technology, plant breeding and new agricultural systems can be sustained. There is much concern, however, about the sustainability of current agriculture (e.g. Kendall and Pimentel, 1994), because of the problems of increasing land degradation, dependence on fertiliser and limited water supplies which may ultimately prevent agricultural production keeping pace with the needs of a hugely increasing world population. This challenge to increase food production must be met with a combination of biotechnological development, improved technology, more investment in currently deprived areas and well thought out strategies for food access, trade and production. To feed the predicted future world population is a challenge on an unprecedented scale. Some comfort can be taken from the progress that has been made in the last 30 years but the next 30 years will require an even greater advance. Agricultural production needs to move hand in hand with sustainable development. The principles of farming in harmony with soil and climate rather than against them are fundamental to a future sustainable agricultural production and to food security.
References Alexandratos, N. (Editor), 1995. World Agriculture Towards 2010 - An FAO Study. FAO and John Wiley and Sons, Chichester, 488 pp. Bradbury, N., Whitmore, A.T., Hart, P.B.S. and Jenkinson, D.S., 1993. Modelling the fate of N in crop and soil in the years following application of 15N labelled fertiliser to winter wheat. J. Agric. Sci., 121: 363-379. Breuning-Madsen, H. and Jones, R.J.A., 1994. A Soil Profile Analytical Database for the European Union. Report for the Joint Research Centre, ISPRA, Italy. Cox, G.W. and Atkins, M.D., 1979. Agricultural Ecology: An Analysis of World Food Production Systems. W H Freeman and Co., San Franciso. 721 pp. FAO, 1978. Report of the Agro-Ecological Zones project. Volume 1. FAO, Rome.
37
FAO, 1982. Assessment of land resources for rainfed crop production in Mozambique. FAO, Rome. FAO, 1988. Land resources appraisal of Bangladesh for agricultural development. FAO, Rome. FAO, 1991. How Good the Earth. FAO, Rome. 33 pp. FAO, 1993. Agriculture: Towards 2010. Conference C93/94. FAO, Rome. FAO, 1996. World Food Summit. Volume 1. Technical background documents 1-5. FAO, Rome. Harrison, P., 1987. The Greening of Africa. Paladin Grafton, London. 380 pp. Higgins, G.M. and Kassam, A.H., 1981. The FAO agro-ecological zone approach to determination of land potential. Pedologie, XI: 147-168. Houghton, J.J., Meiro Filho, L.G., CaUander, B.A., Harris, N., Kattenberg, A. and Maskell, K. (Editors), 1996. Climate Change 1995-The Science of Climate Change. Contribution of Working Group I to the Second Assessment Report of the Intergovernmental Panel On Climate Change. Cambridge University Press. Jones, C.A. and O'Toole, J.C., 1987. Application of crop production models in agro-ecological characterization: simulation models for specific crops. In: A.H. Bunting, (Editor), Agricultural Environments, CAB International, pp 199-210. Kendall, H.W. and Pimentel, D., 1994. Constraints on the expansion of global food supply. Ambio, 23: 198-205. Le Bas, C. and Jamagne, M., 1996. Soil Databases to Support Sustainable Development. Joint Research Centre, European Commission, EUR 16371 EN. 149 pp. Loveland, P.J., Legros, J.P., Rounsevell, M.D.A., de la Rosa, D. and Armstrong, A.C., 1994. A spatially distributed soil, agroclimatic and soil hydrological model to predict the effects of climate change on land use within the European Community. Transactions 15th World Congress of Soil Science. International Society of Soil Science, Commission V, Symposia Volume 6a, 83-100. Morgan, R.P.C., Quinton, J.N. and Rickson, R.J., 1994. Modelling methodology for soil erosion assessment and soil conservation design: the EUROSEM approach. Outlook Agric., 23: 5-9. Penning de Vries, F.W.T., 1997. (in press). Potential and attainable food production levels in different regions. In: D.J. Greenland, P.J. Gregory and P.H. Nye (Editors), Land Resources: On the Edge of the Malthusian Precipice? Philosophical Transactions of the Royal Society, Series A, 355. Pierce, F.J. and Lal, R., 1991. Soil management in the 21st century. In R. Lal and F.J. Pierce (Editors), Soil Management for Sustainability. Soil and Water Conservation Society, USA. pp. 175-179 Plentinger, M.C. and Penning de Vries, F.W.T., (Editors), 1996.
CAMASE - Register of Agro-ecosystems Models. Version II. DLO Research Institute of Agrobiology and Soil Fertility, Wageningen. 411 pp. Pretty, J.N., 1996. Could sustainable agriculture feed the world? Biologist, 43: (3). 130-133. Schultink, G., 1987. The CRIES resource information system: computer aided spatial analyses of resource development potential and development policy alternatives. In: Beek, K., Burrough, P.A., and McCormack, D.E. (Editors), Quantified Land Evaluation Procedures. ITC Publication No. 6. Enschede, The Netherlands. pp 95-99. Sehgal, J., Mandal, D.K, Mandal, C. and Vadivelu, S., 1990. Agro Ecological Regions of India. National Bureau of Soil Survey and Land Use Planning, Nagpur, India. Publication no. 24, 73 Pp. Singh, U., Tsuji, G.Y. and Godwin, P.C., 1990. Planting new ideas in DSSAT: the CERES-Rice model. Agrotechnoi. Transf., 10: 1-6. Sivakumar, M.V.K. and Valentin, C., 1997. (in press). Agro-ecological zones and the assessment of crop production potential. In: D.J. Greenland, P.J. Gregory and P.H. Nye (Editors), Land Resources: On the Edge of the Malthusian Precipice? Philosophical Transactions of the Royal Society, Series A, 355. Stool'vogel, J.J. and Smaling, E.M.A., 1990. Assessment of soil nutrient depletion in Sub-Saharan Africa, 1983-2000. Report 28, The Winand Stating Centre for Integrated Land, Soil and Water Research, Wageningen. TAC/CGIAR (Technical Advisory Committee of the Consultative Group on International Agricultural Research), 1992. Review of CGIAR Priorities and Strategies. Part I. CGIAR, Washington DC. van der Pol, F., 1992. Soil mining, an unseen contributor to farm income in southern Mali. Bulletin 325: Royal Tropical Institute, Amsterdam. van Diepen, C.A., Wolf, J., van Keulen, H. and Rappoldt, C., 1989. WOFOST: a simulation model of crop production. Soil Use Management, 5: (1) 16-24. Wagenet, R.J., and Hutson, J.L., 1989. LEACH-M - A process based model of water and solute movement, transformations, plant uptake and chemical reactions in the unsaturated zone. Continuum - Water Resources Institute, Vol. 2. Version 2. Cornell University, USA. 148 pp. Williams, J.R., Dyke, P.T., Fuchs, W.W., Benson, V.W., Rice, O.W. and Taylor, E.D., 1990. EPIC-Erosion/Productivity Impact Calculator. 2. User Manual. US Department of Agriculture, Technical Bulletin No. 1768. 127 pp. World Resources Institute 1992. World Resources 1992-93. Towards Sustainable Development. Oxford University Press, Oxford. 385 pp.
This Page Intentionally Left Blank
© 1997 ElsevierScience B.V. All rights reserved Perspectives for Agronomy - Adopting Ecological Principles and Managing Resource Use M.K. van lttersum and S.C. van de Geijn (Editors)
39
The potential benefits of agroforestry in the Sahel and other semi-arid regions
H. Breman a,l'*, J.J. Kessler b aAB-DLO, Research Institutefor Agrobiology and Soil Fertility, P.O. Box 14, 6700AA Wageningen, The Netherlands bAlD Environment, Donker Curtiusstraat 7-523, 1051 JL Amsterdam, The Netherlands
Accepted 1 July 1997
Abstract This article summarises results of three related studies. The first study is a synthesis and analysis of available knowledge on woody plants in semi-arid regions with an emphasis on the Sahel region, investigating the potential 'added value' of woody plants to improve nutrient and water availability for agricultural crop and pasture production goals. The other two are field studies aimed to validate key issues identified by the synthesis study, through measurements on the influences of woody plants on water and nutrient availability, and investigations of the relationship between soil organic matter quantity and quality and the recovery rate of nitrogen and phosphorus from fertilisers. The field studies support the conclusion that processes leading to an added value of woody plants in agroforestry systems are mainly related to reduced losses of water and nutrients. Such added values are therefore lowest where they are most needed, in resource-poor environments. Specific farmers goals, agroecological and socio-economic conditions have to be taken into account to make optimal use of the potentials of agroforestry. The potential benefits of agroforestry systems are mainly in terms of the improved efficiency of nutrient inputs than as an alternative for fertilisers. This has major implications for design and management of agroforestry systems. © 1997 Elsevier Science B.V. Keywords: Agroforestry; Soil fertility; Water availability; Sahel; Systems analysis
1. Introduction Although drought years brought the Sahel countries into publicity in the early seventies, water scarcity appears not to be the main bottleneck for rural devel-
* Corresponding author. ~Present address: IFDC-Africa, P.O. Box 4483, Lom6, Togo, Africa.
opment in this region. Extremely poor soil fertility, extreme aridity of a long dry season causing high fragility of perennial plant communities, rapid turnover of organic matter, and unfavourable socio-economic conditions are more important constraints (e.g. Penning de Vries and Djit~ye, 1982; Van Keulen and Breman, 1990). A Malian-Dutch research project called 'Production Soudano-Sah61ienne (PSS)' aimed to alleviate these constraints (Production Soudano-Sah61ienne, 1991), recognising:
Reprinted from the European Journal of Agronomy 7 (1997) 25-33
40
• • •
the urgent need to improve soil fertility (Penning de Vries and Djit~ye, 1982); the relative land scarcity and shortage of organic fertilisers (Breman et al., 1990); the unfavourable price relations of chemical fertilisers and agricultural products, given the low average recovery rate of nutrient inputs (Wooning, 1992).
Agroforestry is defined as a land-use system in which woody plants are grown in association with agricultural crops, pastures or livestock. Apart from diversification of the system with tree crops, agricultural crop and/or livestock production are expected to be improved by the interaction with woody plants, mainly through their capacities of deep rooting, nitrogen fixation and soil conservation, all leading to improved soil fertility at the benefit of agricultural crops and pastures (e.g. Young, 1989). Agroforestry has been widely promoted as a more sustainable agricultural production system, and would be particularly attractive for developing countries where the use of external inputs is not feasible (e.g. Winterbottom and Hazelwood, 1987). More recently, evidence has been generated to support the hypothesis that 'ecological agriculture technologies' such as agroforestry are important mainly in view of their potential to improve soil organic matter status (improved quantity and quality), which will improve the economic feasibility of using fertilisers through a higher recovery rate. Improved soil management is crucial for sustainable intensification of agriculture in the Sahel region (Kon6 and Groot, 1996; T6m6 et al., 1996; Breman and Sissoko, 1997). The PSS project also analysed the potential 'added value' of woody plants in terms of improved water and nutrient availability for plant production, in other words their increased accessibility within the rooting zone. Potential benefits of agroforestry in the Sahel region were identified, based on farmers' goals and Sahelian site conditions. Two aspects were characteristic for the studies executed: •
•
the role of woody plants as an option to increase the recovery rate of fertilisers, not as an alternative for fertilisers; the use of a systems approach to quantify the potentials of woody plants to improve water
and nutrient availability for primary production in relation to variable site conditions. Following presentation of the methodology of the synthesis study and field validations, this article presents the results, the consequences of the findings for potentials of agroforestry, and conclusions in terms of constraints and opportunities. As a result of the systems approach, using the absolute and relative availability of nutrients and water as main variables to determine potentials for specific regions, the conclusions are not only applicable to the Sahel region, but to semi-arid regions in general.
2. Methodology This article summarises results of three related studies. The first concerns a synthesis and analysis of available knowledge on woody plants in semi-arid regions (Breman and Kessler, 1995). More than 500 publications on woody plants in semi-arid agro-ecosystems were analysed, with an emphasis on the Sahel region, using insights on eco-physiological processes of primary production as a guiding framework (Penning de Vries and Djit~ye, 1982). This study investigated the potential 'added value' of woody plants to improve nutrient and water availability in rangelands and agricultural lands. Estimates were made of the maximum possible benefits from woody plants, by quantifying all relevant eco-physiological processes influenced by woody plants (Tables 1 and 2). Maximum use was made of data from the Sahel region, and other (semi-)arid regions (Australia, India and Northern America mainly) to fill in gaps of knowledge, while taking into account the differences in ecological conditions, e.g. through modelling and simulation. Amongst others, simulation models were used to assess the influence of woody plants on the microclimate, to quantify light interception by woody plants, and to quantify organic matter dynamics. Nitrogen (N) availability for woody plants and herbaceous plants in a mixed vegetation, was estimated in relation to variable woody plant canopy cover and the number of leaf layers of their crowns, using a formula which had been developed for a vegetation with annual and perennial plant species (Breman and De Ridder, 1991).
41
Table 1 The maximum effects of woody plants on water availability in a mixed vegetation in the Sahel zone (150--600 mm) and the Sudan zone (600-1200 mm), in comparison to a vegetation without woody plants Process
Sahel zone
Sudan zone
Stem flow: deep infiltration Improved soil structure: less run-off
+ +
+4++÷
Improved soil storage capacity Micro-climatic changes Hydraulic lift uptake by deep roots: less percolation losses
0 + 0 +
+ + + ++
0, negligible effects; +, ++, ++, water availability increases 10-50, 50-100 or > 100 mm/year.
The synthesis study was necessary to focus the other two field studies as part of the PSS project. One of the research clusters studied the possibility to increase the recovery of fertiliser nutrients in agriculture through the improved organic matter status of the soil, amongst others through the integration with woody plants. Since the duration of the project was too short to develop comparable situations with and without well developed woody plants, two alternative study approaches were applied. Study A aimed to quantify the processes by which woody plants can influence water and/or nutrient availability for primary production and soil fertility changes. Study B aimed to quantify the recovery of N and phosphorus (P) for soils with variable organic matter status. Study A was executed on the Niono ranch, 15 km East of Niono in central Mali (Southern Sahel climate zone, 550 mm annual rainfall, altitude 300 m, deep sandyloam soils); study B was executed on the same ranch and more to the South on the Cinzana and the N'Tarla research stations (Northern and Southern Sudan climate zones, 800 and 1000 mm annual rainfall, respectively, similar altitude and soils). Study A included measurements of key tree parameters of Acacia seyal and Sclerocarya birrea (Soumar6, 1996; Breman and Sissoko, 1997). Studies on the area where water and nutrient availability were influenced by woody plants, were executed for Acacia
senegal, Acacia seyal, Balanites aegyptiaca, Combreturn ghasalense, Commiphora africana and Sclerocarya birrea. Within a circle with a radius of 2 m
around the stem the whole root system was excavated. In addition, some lateral roots were entirely excavated; the total length of all lateral roots was determined using the ratio between base diameter and length of excavated roots. The water balance was monitored during 1 year, using eight rain gauges: four under the crown (in each principal direction) to measure through fall, and four at a distance of twice the crown diameter to measure real rainfall. Stem flow was measured by a stem spanned collector. Water interception by the crown was calculated as rainfall minus through fall and stem flow. Evaporation under the crown and beyond the crown area was measured by a microlysimeter, and the calculated water balance was verified by monitoring soil humidity (weighing soil samples before and after drying). The influence of woody plants on soil fertility was determined by measuring foliage and litter production (Soumar6, 1996) and by taking soil samples to determine the enrichment factor for carbon (C), N and P (Radersma, 1996; Soumar6, 1996; Breman and Sissoko, 1997). For the six woody species mentioned, soil samples were taken (at 0 - 2 0 , 2 0 - 5 0 and more than 50 cm soil depth) under the tree crowns (three in each principal direction) and at a distance of two and
Table 2 The maximum effects of woody plants on nutrient availability in a mixed vegetation in the Sahel zone and the Sudan zone, in comparison to a vegetation without woody plants Process Spatial concentration Uptake by lateral roots Capture of wind blown material Deposition by animals Reduction of losses Decreased wind erosion Decreased water erosion Decreased leaching Decreased fire (volatilisation) Internal recycling External recycling Enrichment of the system Uptake by deep tap roots Nitrogen fixation P-uptake through mycorrhiza
Sahel zone
Sudan zone
+ + +
++ + +
+ + + + + +
+ ++ ~,~, + ++ ++
0 0 0
+ + +
0, negligible effects; +, ++, +++, N increase I-5, 5-10 or > 10 kg/ ha per year; P increase 10% of N increase.
42
three times the crown radius (one in each principal direction). The enrichment factor is one way to express the influence of woody plants on soil fertility, and is defined as the ratio of nutrient concentration under a woody canopy and in the open field. Per tree species, three trees were studied, but eight trees for Acacia seyal and Sclerocarya birrea. Light interception was measured with a photometer (trade-mark DECAGON), starting in April, when trees are without leaves, until maximum tree foliage biomass in August. Foliage biomass was estimated using the allometric relations established per species by Hiernaux et al. (1992). Measurements were done for four trees of both Acacia seyal and Sclerocarya birrea, twice a month, at 8, 10, 12, 14 and 16 h, with 44 observations per tree (11 in each principal direction at equal distances from each other: five under the crown, five beyond the crown and one in the open field). Study B investigated the relationship between soil organic matter quantity and quality and the recovery rate of N and P from fertilisers. In each of the climate zones mentioned above, three soil types were studied: sand, loam and clay (Kon6 and Groot, 1996; Breman and Sissoko, 1997). The 'three quadrant methodology' (Van Keulen and Wolf, 1986) was used to measure the recovery of N and P. Comparisons between sites with a low and a relatively high soil organic matter status were made in three ways: (1) using land that had just been abandoned by farmers after years of production, and a more than 10 year old fallow; (2) using a degraded crop land, and a productive land densely covered for more than 40 years with the perennial grass Andropogon gayanus; and (3) following annual fertiliser application, monitoring changes in N and P by recovery by grasses (the annual Pennisetum pedicellatum and the perennial Andropogon gayanus) and leguminous herbs (Stylosanthes hamata and Vigna unquilata).
increase nutrient and water availability, as compared to a herbaceous vegetation only. Tables 1 and Table 2 present results for the Sahel zone (150-600 mm/year of rainfall) and for the subhumid Sudan ('savannah') zone (600-1200 mm). The effects by each indicated process refer to the maximum effects for a mixed vegetation with woody plants as a whole, and not to the area covered by woody plants (fraction of the soil surface occupied by the projection of their crowns). These effects are expected for a cover of woody plants as found in the 'natural vegetation' of good condition before the drought of the seventies, generally considered as the maximum cover possible: 2-20% in the Sahel zone and 15-35% in the Sudan zone (variation according to soil type). On the basis of the results of the analyses carried out as part of the synthesis study, it was concluded that the added value of woody plants largely depends on the processes that reduce losses of nutrients and water, mainly through soil improvement. The added value increases with increasing rainfall, from the Sahel to the Sudan zone. The following woody plant species characteristics were identified as having maximum potential to increase water and nutrient availability for primary production by the processes indicated:
3. Results
As regards site conditions, maximum benefits in terms of improved nutrient availability can be expected on well-drained, deep and fertile soils. Maximum benefits on water availability can be expected on fine-textured soils with high moisture contents and where potential evapotranspiration is high (e.g. in fine-textured valleys in arid zones).
3.1. Synthesis study To analyse the potential 'added value' of woody plants, Breman and Kessler (1995) defined and quantified the processes through which woody plants can
• • • • • • • •
the capacity to generate stem flow; deep rooting, preferably reaching the ground water table; a tall, closed canopy and an evergreen phenology; unpalatibility of foliage and fruits, for maximum contribution to soil improvement; a slow growth rate and long life span; high efficiency of internal nutrient cycling; the capacity for nitrogen fixation and presence of mycorrhiza; foliage properties allowing maximum contribution to soil organic matter status (quantity and quality).
43
The effects of the separate processes cannot be simply added to obtain an overall effect of the mixed vegetation, because several processes are more or less incompatible in view of the required properties of the woody species. Examples of incompatible processes are 'increased moisture retention of the soil through improved soil structure' and 'uptake by deep roots of percolated water' (Table 1), or 'decreased leaching of nutrients' and 'nutrient uptake by deep roots' (Table 2). Based on the relative availability of water and nutrients, water is the factor limiting plant growth in the Northern Sahel zone, in areas with run-off in the Southern Sahel zone, and elsewhere in areas with strong run-off. In other words, in the Sahel zone the limited improvement of water availability, in absolute terms, due to the influences of woody plants, may be relatively important. On the other hand, in the Sudan zone the impacts of any improved water availability will be limited as long as nutrients are the main factor limiting primary production. Here, the most important added value of woody plants is the influence on nutrient availability, to be benefited from in agroforestry systems. The situation becomes more complicated where irrigation or fertilisers are being applied. Where water availability is adequate following irrigation, several processes improving nutrient availability (Table 2) could become important, but, unfortunately, most woody species do not survive and produce well when inundated. In addition, their shading effects and the supply of nesting sites for graniverous birds are generally considered unacceptable. On the other hand, where water is the main limiting factor the processes that improve water availability (Table 1) become useful, and synergistic effects can be expected when applying fertilisers. The nutrient improving processes (Table 2) will only remain effective as long as they do not show a negative feedback with increased soil fertility (e.g. N-fixation and internal recycling). Some of these processes, particularly those decreasing nutrient losses become more important under conditions of improved fertilisation. From the synthesis study, it can be concluded that agroforestry has the highest potential (i.e. to improve the potential to use available nutrients and water for primary production purposes in a sustainable way) where production conditions are already most favourable!
Table 3 Comparison of agroforestryparameters from the synthesis study of available knowledge and form field studies in the Southern Sahel zone
Synthesis study Fieldstudies Water balance Stem flow (%) Growing season Nutrient balance Foliage production (g/m2) Enrichment factor Ca Enrichment factor Na Enrichment factor P~ Light interception Isolated tree (%) Cereal yield increase Faidherbia x milletc (kg/ha)
0--20 Extension rare
2-5 No extension
average230 1-3 1-3 l-3x
310, 155b Average 1.3 0.9-1.3 0.8-3.1
Average 47
50, 65b
Maximum 200--400
Maximum300
aln 20--30 cm topsoil under woody plant canopy in comparison to open field; bAcaciaseyal and Sclerocaryabirrea,respectively;CFor Sudan zone.
3.2. Validation by field work The two field studies, aimed to test some of the above conclusions on the added value of woody plants, showed the positive influence of soil organic matter status on recovery rate of fertilisers (Kon6 and Groot, 1996). Soil organic matter quality was improved by annual phosphate inputs on a cowpea crop on different textured soils during successive years, in the 800 mm rainfall zone. The P recovery rate increased from 15% in the first year to 50% in the fourth year. The highest effects were obtained on a sandy soil, particularly in relatively humid years. The perennial grass Andropogon gayanus showed a much higher N recovery rate (70%) than the annual Pennisetum pedicellatum (40%). The N recovery rate of Pennisetum pedicellatum was higher when cultivated on a plot where Andropogon gayanus had previously been grown (45%), as compared to a plot with previous cultivation of annual crops (35%). The root systems of the studied woody species vary strongly, determining their capacity to utilise nutrients and water from extended areas and deep soil layers (i.e. redistribution of nutrients and water towards the topsoil and the woody plant). This influences competition for water and nutrients with herbaceous plants. Combretum ghasalense had the most extensive root
44
system (in width and in depth), Sclerocarya birrea had the longest lateral roots (up to 40 m from the trunk), Acacia seyal had the longest tap root (6 m), while Commiphora africana had a very shallow root system. The study confirmed that roots of woody plants are concentrated in the topsoil, where strong overlap occurs with those of herbaceous plants, and where most water and nutrients are found (Soumar6, 1996). The field results on water and nutrient availability and light interception (Table 3) support the conclusions from the synthesis study (Breman and Kessler, 1995). Based upon the enrichment factor, which expresses the importance of soil improvement by woody plants, soil improvement by woody plants on the Niono ranch appears to be very limited, if not negligible, for C, N and P. The only exceptions were the high P enrichment factors (P-Bray, i.e. available P), up to 3.1, for Acacia senegal and Balanites aegyptiaca, but the average enrichment factor for total P was only 1.1. The combination of drought and heavy exploitation rates (cropping and grazing mainly) on the ranch must be the explanation for the overall limited enrichment effects of woody plants. As a result, perennial grasses have disappeared, there has been massive mortality of woody plants, and a decline of C content in the topsoil from 2.5 to 4.0 g/kg in the seventies (Penning de Vries and Djit6ye, 1982) to 2.0-2.5 at present (Radersma, 1996; Breman and Sissoko, 1997). Annually, as much as 90% of the annual production of the herbaceous layer and of litter from woody plants disappears by grazing and fire, while maximum return to the soil of organic material produced is a condition to make optimal use of the added value of woody plants in agroforestry systems (Breman and Kessler, 1995). De Ridder and Van Keulen (1990) predicted an increase of the cation exchange capacity (CEC) of 4.3 mmol/kg per 1 g/kg increase of soil organic matter. The difference in N recovery between Andropogon gayanus and Pennisetum pedicellatum, as mentioned above, was associated with a difference in topsoil C of 3.9 and 2.4 g/kg, respectively, and a CEC of 1.9 and 1.6 mmol/kg, respectively, which is as predicted (Breman and Sissoko, 1997). For woody plants, the average enrichment factor of 1.3 for C in the topsoil (Table 3) would thus imply a CEC increase of only 1.3 mmol/kg. However, the field dynamics of
organic matter appear to be more complicated. Radersma (1996) observed only a slight increase of the CEC in relation to Sclerocarya birrea and Combretum ghasalense, while one Acacia seyal tree showed an important increase and another a slight decrease, without correlation with the C enrichment factor. Breman and Kessler (1995) predicted the maximum increase of cereal production through increased availability of nutrients by woody plants in the Sudan zone to be 200-400 kg/ha (depending upon nutrient use efficiency), equivalent to about 5 kg N/ha. This maximum can be reached with a canopy cover of about 20%, limited shade effects (using homogeneously distributed trees with a high ratio of trunk height to crown diameter, or no foliage during the growing season), and a return to the soil of all organic material produced by the agroforestry system, except for cereal grain yield. If one or more of these conditions is not met, the increase of cereal production will be less. This probably explains why the maximum cereal yield increase of 300 kg/ha, observed for well managed Faidherbia albida park lands in the Sahel region, is nowadays seldomly reached.
4. Consequences The results of the field validations on the role of woody plants in plant production systems by the two field studies show that, in spite of several processes through which woody plants can improve the availability of nutrients and water, the added values of woody plants are generally even less than expected on the basis of the synthesis study, and thus do not easily represent benefits for farmers, because: the added values are lowest where they are most needed, in resource-poor environments; the competition between woody plants and crops or pastures is strong; conditions are generally not suitable for maintenance of an optimal canopy cover and tree canopy structure, and for maximum return to the soil of organic material produced, mainly because of the high exploitation levels and drought events. The field evidence strengthens the arguments of
45
Breman and Kessler (1995) to limit agroforestry (research) to on-farm testing of options that have potentials to meet specific farmers goals, for specific soil-climate and socio-economic conditions. The following represent the best potentials for effective use of agroforestry. 4.1. In sylvopastoral systems
In case of water limited production, woody cover is good for security aims (to stabilise fodder availability during drought) and to reduce resource losses by erosion. In case of nutrient limited production, a canopy cover of 15-20%, of homogeneously distributed trees, with unpalatable foliage, and a high ratio of trunk height to crown diameter, supports animal production goals. To maximise animal production goals, fodder banks of highly palatable woody species are required, at locations representing optimum growth conditions and high niche differentiation for woody plants (e.g. valleys and depressions with deep and fertile soils). 4.2. In cropping systems
Windbreaks are useful to improve crop establishment on sandy soils in dry areas characterised by nutrient limited primary production (Southern Sahel zone), and where shallow ground-water is available. In the Sudan zone, maximum benefits for crop production are obtained where canopy cover is 15-20%, trees are homogeneously distributed, with a high ratio of trunk height to crown diameter (possibly achieved by pruning) and restricted exploitation. Here, the potential benefits are more important for farmers using fertiliser inputs than those without? 4.3. In relation to economic conditions
There is a positive correlation between the potentials for agroforestry and wood and fruit prices, and a negative one with labour wages. Stimulating agroforestry through specific subsidies is justified on slopes and in the upper course of river basins in view of the crucial role in soil stabilisation by woody plants, and in desert margins for protection
of more productive land against desertification. Such subsidies will be more effective and sustainable than to tackle the symptoms of desertification and of fiver basin degradation, provided that the woody plants are not exploited, but used for soil protection. In view of the improved efficiency and recovery rate of the use of water and nutrients (naturally available and external inputs), there will be a less need for subsidies in the agricultural sector in well designed and managed agroforestry systems.
5. Constraints and opportunities Even a focus on the potential options outlined above will not easily generate participation and enthusiasm by farmers, for two main reasons. (1) Farmers should learn how to manage agroforestry systems in such a way that maximum benefits are obtained in terms of increased water and/or nutrient availability for crops or pastures, instead of maximum use of tree crops only. This implies finding the optimum balance between (i) benefits of the added value by woody plants in terms of water and nutrient availability for crops and pastures, and (ii) increased competition between woody plants and crops or pastures as woody cover increases. Based upon farmers goals of productivity (of crops, pastures or woody plants), security and/or sustainability, careful decisions on the design and management of agroforestry systems should involve four interrelated issues: • • • •
selection of tree species with desirable properties (see list of properties in Section 2); woody plant configurations (densities and orientation); tree management issues (pruning, lopping, controlled harvesting, etc.); selection of appropriate site conditions for specific goals.
It will be difficult to find the fight 'package' as regards these four issues, but the amount of 'trial and error' can be substantially reduced by taking the situations with highest potentials for effective use of agroforestry (as indicated above) as starting points for on-farm trials. Benefits can be further improved through judicious use of fertilisers in well designed
46 agroforestry systems, thus making use of the potentials for improved recovery of nutrients. (2) The organic matter status of soils in most semiarid regions is poor, whereas possibilities to improve the soil organic matter status are limited. Such improvements will require a long 'transition period' during which investments (in terms of labour and inputs mainly) are not matched by increased outputs, due to the slow establishment of a woody plant community, and the rapid turn-over of organic matter under tropical conditions, particularly in the Sahel region (De Ridder and Van Keulen, 1990). Land tenure security is one pre-requisite, and subsidies during this transition period may be another. Using simulation models, Groot et al. (1997) estimated that the process of soil improvement through an increased annual supply of organic matter takes 5-15 years. Woody cover required to maintain C concentrations on cropland at a level of 6 g/kg (considered as a minimum for agricultural use of a soil with 20% clay) was estimated at an average of 34% for the Southern Sahel, provided return to the soil of all organic material produced. This unrealistic high level underlines the limited potentials for agroforestry in this zone, and the need for careful management decisions as indicated above. The required 17% woody cover in the Sudan zone is more realistic to achieve. Two specific opportunities for agroforestry were also identified. (1) In the Sahel region (and most likely other semiarid regions), the disappearance of perennial grasses, the replacement of annual species with long growth cycles by short cycled annuals (Breman and De Ridder, 1991) and the massive mortality of woody plants (Breman and Kessler, 1995) created a reduction of water use by vegetation. Whereas at the start of the growing season during the seventies (Penning de Vries and Djit~ye, 1982) water availability was limited to the first 50 cm of the soil profile, the current situation at the Niono ranch (SoumarE, 1996; Breman and Sissoko, 1997) shows a water profile exceeding 250 cm soil depth. This implies potentials for tree planting and reforestation. (2) In Faidherbia albida agroforestry parklands nitrogen availability for crops is improved by at most 5 kg/ha without fertiliser use, corresponding to about 25% of the average natural annual N availability (Penning de Vries and Djit~,ye, 1982). In addition,
the improved organic matter status will improve the feasibility of using fertiliser inputs. Using woody plants that shed their foliage during the dry season, the yield of associated crops (e.g. vegetables) during the dry season, even when irrigated, can be stimulated by benefiting from this added value of the trees in terms of improved nutrient availability and organic matter status, in a similar way as for Faidherbia albida during the rainy season.
References Production Soudano-SahElienne, 1991. Production SoudanoSahElienne (PSS). Exploitation optimale des ElEments nutritifs en Elevage. Projet de cooperation scientifique. IER, Bamako et AB-DLO, Wageningen, 38 pp. Breman, H. and Kessler, J.-J., 1995. Woody plants in agro-ecosysterns of semi-arid regions (with an emphasis on the Sahelian countries). Advanced Series in Agricultural Sciences, Vol. 23. Springer-Verlag, Berlin, 340 pp. Breman, H. and De Ridder, N., 1991. Manuel sur les p~iturages des pays sahEliens. ACCT-CTA-KARTHALA, Paris, 485 pp. Breman, H., Ketelaars, J.J.M.H. and TraorE, N., 1990. Un remade contre le manque de terre? Bilan des ElEments nutritifs, production primaire et 61evage au Sahel, SEcheresse, 2: 109-117. Breman, H. and Sissoko, K., 1997. L'intensification de L'agriculture SahElienne. KARTHALA, Paris (in press). De Ridder, N. and Van Keulen, H., 1990. Some aspects of the role of organic matter in sustainable intensified arable farming systems in West-African semi-arid tropics. Fert. Res., 26: 299310. Groot, J.J.R., KonE, D. and de WiUigen, P., 1997. L'engrais chimique pour une intensification durable en zone soudanosahElienne. In: H. Breman and K. Sissoko (Editors), L'intensification de L'agriculture SahElienne. KARTHALA, Pads (in press). Hiernaux, P., CissE, M.I., Diarra, L. and de Leeuw, P., 1992. Fluctuations saisonni~re de la feuiaison des arbres et des buissons sahEliens, amelioration de l'Evolution des ressources fourrag~res des parcours sahEliens. CIPEA, Document de travail no. 002/92. ILCA, Addis Ababa. KonE, D. and Groot, J.J.R., 1996. EfficacitE d'utilisation de phosphore et de l'azote par Stylosanthes hamata, Vigna unguiculata, Andropogon gayanus et Pennisetum pedicellatum en zone soudano-sahElienne du Mali. Rapports PSS no. 21, IER, Bamako et AB-DLO, Wageningen, 45 pp. Penning de Vries, F.W.T. and Djit~ye, M.A., 1982. La ProductivitE des Paturages SahEliens. Une l~tude des Sols, des VEgEtations et de L'exploitation de Cette Ressource Naturelle. PUDOC, Wageningen, 525 pp. Radersma, S., 1996. Influence des arbres agroforestiers sur le sol et la strate herbacEe du Sud du Sahel. Rapports PSS no. 26, IER, Bamako et AB-DLO, Wageningen, 106 pp.
47
Soumar6, A., 1996. Utilisation des 616ments nutritifs par deux arbres du Sahel: Acacia seyal et Sclerocarya birrea. Rapport PSS no. 22. IER, Bamako et AB-DLO, Wageningen, 120 pp. T6m6, B., Breman, H. and Sissoko, K., 1996. Intensification agricole au sahel: mythe ou r6alit6. Rapport de synth6se des travaux du Colloque International, Bamako 28 novembre-2 d6cembre 1995. IER, Bamako et AB-DLO, Wageningen, 28 p. Van Keulen, H. and Breman, H., 1990. Agricultural development in the West African Sahelian region: a cure against land hunger? Agric. Ecosyst. Environ., 32:177-197.
Van Keulen, H. and Wolf, J., 1986. Modeling of Agricultural Production: Weather, Soils and Crops. Pudoc, Wageningen, 478 pp. Winterbottom, R. and Hazelwood, P.T., 1987. Agroforestry and sustainable development: making the connection. Ambio, 16: 100-110. Wooning, A., 1992. Les prix du b~tail, de la viande, des produits laitiers et des engrais dans les pays sah~liens. Rapport PSS no. 1. IER, Bamako et AB-DLO, Wageningen, 82 pp. Young, A., 1989. Agroforestry for soil conservation. ICRAF, Nairobi, 276 pp.
This Page Intentionally Left Blank
© 1997 Elsevier Science B. V. All rights reserved Perspectives for Agronomy - Adopting Ecological Principles and Managing Resource Use M.K. van lttersum and S.C. van de Geijn (Editors)
49
Chemical crop protection research and development in Europe R. N e u m a n n
*
Research and Development Novartis Crop Protection AG, R-1004.8.15, CH-4002 Basel, Switzerland
Abstract The following five theses are discussed: 1. The majority of the crop protection research and development money is spent by European companies. These efforts are targeted at a global market. The relative importance of the European market strongly varies with the crops and the different pests. 2. Crop protection research and development is a very costly process which will accentuate the concentration process in the crop protection industry. 3. Crop yields will have to increase significantly, if we want to meet the growing demand for food without claiming more land from nature. High yielding varieties and new agrochemicals with further improved characteristics will be important pillars. 4. Risks and benefits must be seriously judged and compared. Overemphasising the risks reduces the benefits. Insufficient crop protection is a waste of resources. 5. Crop protection has tremendously improved in the last 50 years and will continue to do so. Chemical synthesis will remain the most important source for new active ingredients. Natural products will not per se play a dominant role in the near future. Keywords: Crop protection; Research and development; New chemical products; Risk and benefit
1. Introduction
2.
Do we really need research and development in crop protection? Don't we have more than enough herbicides, insecticides, and fungicides? Many people would argue that we have too many. But this is wrong. In the past, we had a steady flow of new active compounds from research to the markets and this must continue. We need more and better products for resistance management, for integrated pest control, and for higher yields. Otherwise, we will not be able to satisfy the needs of a growing population without destroying more rainforests and other important natural habitats. This paper is structured around five theses with the following topics"
3.
1.
Europe as part of a global market
* Corresponding author. Phone +41 (61) 697 34 73; Fax +41 (61) 697 84 57.
4. 5.
The increasing expenses in crop protection R and D The need for increasing yields in crop production Risks and benefits in crop protection New developments in crop protection R and D
2. 1st Thesis The majority of the crop protection research and development money is spent by European chemical companies. These efforts are targeted at a global market. The relative importance of the European market strongly varies with the crops and the different pests. In 1994 the 20 biggest crop protection companies invested around 2.5 billion US$ in research and development (Wood Mackenzie, 1995a). Out of these 20 companies 7 have their headquarters in Europe, 6 are based in the US and 7 in Japan. This could in-
50
dicate a more or less equal importance of the three regions for research and development in crop protection. However, the seven European companies spent 1.4 billion US$, corresponding to 58% of the grand total, the six US based companies spent little over half that amount, namely 756 million US$ or further 31%, and the R and D effort of the seven Japanese companies add up to 282 million US$ i.e. the remaining 11%. In relative terms, the industry spent 9.7% of sales for R and D which shows that these companies invest intensely into the future of crop protection. Today, these R and D efforts must be targeted at the global market. Home markets are too small to justify the ever increasing costs of R and D. The size of the respective world markets is a major factor for the setting of priorities in R and D. Due to the globalization, this is valid for all big companies. Smaller companies, on the other hand, look for allies in order to market their new products globally. Home markets are of little special importance for most European crop protection companies. Specific European needs are taken into consideration as far as it can be justified by the size of the respective markets. In Germany, this has led to a situation where more than 40 pest problems are without a permitted pesticide. Products exist for these problems, but the size of the respective markets can-
not justify the cost of their registration (Meinert, 1991). A few figures shall help to sense the weight of Europe in world-wide crop protection. •
•
•
•
Europe represents a quarter of the global market of about 28 billion US$. North America - which is mainly the US - lies at 30%, and Japan accounts for 17%. After the US and Japan, France is number three in a country ranking, followed by Brazil, and Germany (Wood Mackenzie, 1995b). Together, these five most important countries add up to 60% of the global market. Looking at West Europe on its own" France takes a third, Germany 15%, followed by Italy with 14%, the UK with 10% and Spain with 9%. All other countries use the remaining 20% (Wood Mackenzie, 1995b). Almost half of the pesticides used are herbicides (45% in Europe and 47% world-wide); due to the different climatic conditions, fungicides are more important in Europe (29% and 20%) and insecticides more in the rest of the world (18% and 29%) (Wood Mackenzie, 1995c). Europe's importance in the world-wide herbicide markets varies with the crops. It repre-
Table 1 Major companies" the 75% Club 1980
1985
1990
1994
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 1 I. 12. 13. 14. 15. 16. 17. 18.
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 1 I. 12. 13. 14. 15.
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
1. Ciba 2. AgrEvo 3. DuPont 4. Monsanto 5. Zeneca 6. Rh6ne-Poulenc 7. Bayer 8. DowElanco 9. American Home 10. BASF 11. Sandoz
Bayer Ciba-Geigy Monsanto Shell ICI Rh6ne-Poulenc BASF Eli Lilly DuPont H6chst Stauffer Dow Union Carbide Am. Cyanamid FMC Rohm&Haas FBC Kumiai
Bayer Ciba-Geigy Monsanto ICI Shell Rh6ne-Poulenc BASF H fchst Dow DuPont Schering Eli Lilly Am. Cyanamid Stauffer FMC
Ciba-Geigy ICI Bayer Rh6ne-Poulenc DuPont DowElanco Monsanto H6chst BASF Schering Sandoz Am. Cyanamid
51
In 1980, 75% of total world-wide sales were made by 18 companies, in 1985 there were 15, in 1990 only 12. Today 11 companies hold 75% of the market and by the year 2000 it should be expected that there will be fewer than 8 research based companies in the 75% club. With Ciba and Sandoz merging to Novartis it already is down to 10. One of the major driving forces behind this concentration process is the need to keep or obtain 'critical mass' as a basis for successful research and development.
sents 43% in cereals, 30% in fruits and vegetables, 16% in maize, and below 5% in rice and soybeans (Wood Mackenzie, 1995c).
3. 2nd Thesis Crop protection research and development is a very costly process which will accentuate the concentration process in the crop protection industry.
4. 3rd Thesis
A total of 148 new active ingredients were introduced between 1983 and 1993, or about 13.5 per year. From this the R and D expenses for a new active ingredient can be roughly calculated to be in the vicinity of 200 million US$. 20 years ago, this figure would have been close to 80 million, or 2.5 times less, after correcting for the inflation. Modem chemical crop protection started after World-War II. Before that time almost no adequate solutions were available. A large number of compounds have since been discovered. The Pesticide Manual (1994) lists 1285 active ingredients that at some time were used for crop protection, out of which 725 still are in use today. This number will decrease with time. Due to the increasing costs of development, registration and re-registration, more products will be withdrawn from the market than newly introduced. As another consequence, the concentration process in crop protection companies will continue (Table 1).
Wheat
Barley
Maize
Rice
World Germany
Crop yields will have to increase significantly, if we want to meet the growing demand for food without claiming more land from nature. High yielding varieties and new agrochemicais with further improved characteristics will be important pillars. Today, almost nobody contests the need for further increases in the production of food and fibres for a growing and prosperous population. FAO and others agree that we will at least have to double if not triple crop production in the next 50 years. We have not only to keep up with the human population increase, but also to provide the feed for the increasing demand for meat. This is easy to say, but will be hard to achieve. Since we cannot afford to take more land from nature, the only way out is to increase the yields. The last 25 years have seen increases in the world-
I+
55 % }+62%
[ . + 18 %
World Germany
1+41%
World Germany
1+70% 1+35%
World
!.... 1
2
3 yield:
I +53% 4
5
t/ha
Fig. 1. Increase of yield 1969/71 - 1994.
52
wide yields, of 55% for wheat, of 18% for barley, even 70% for maize, and 53% for rice (Fig. 1). As e.g. the European countries have shown, further substantial increases are possible (Oerke et al., 1994). The same publication gives an estimate of the contribution made by world-wide crop protection to the production of the eight principal food and cash crops. Of the theoretically attainable production still about 40% is actually lost due to weeds, animal pests, and diseases. The 60% harvested would further be cut in half without crop protection. This alone clearly shows that today's pesticides and the way they are used must be improved. We cannot afford to loose such a high percentage of the attainable production. The increasing importance of integrated pest management (IPM) and the inevitable spread of resistance are further reasons for the need for new products. There are many definitions of IPM around. Most of them give 'minimizing the use of chemicals' as the main goal. Extreme views equate IPM with pesticidefree farming. The definition which Ciba supports reads: Integrated Pest Management is the farmer's best combination of cultural biological and chemical measures that yield the most cost-effective, environmental sound and socially acceptable insect, disease and weed management for crops in a given situation. High yielding varieties and new agrochemicals with further improved characteristics will be important pillars; certainly among other things like e.g. soil fertility or a sufficient supply of water. The main driving forces for better products are in the areas of safety and performance as discussed in the next thesis.
5. 4th Thesis Risks and benefits must be seriously judged and compared. Overemphasising the risks reduces the benefits. Insufficient crop protection is a waste of resources.
Crop protection products have to be very safe" safe for the consumer, safe for the environment, safe for the farmer and safe for the industry worker.
About 40% of the R&D expenses are spent for the safety evaluation of the products, i.e. about 80 million US$ for each successful new active ingredient. Product safety is very important and has always been the major hurdle to receive the official registration. But on the other hand, the benefits from the crop protection chemicals have to be considered, too. They are now included in the guidelines from the EU. Risks and benefits must be seriously judged and compared. They have to be discussed case by case, separately for each use of a specific product. Region specific facts like climate, soil conditions, availability of other solutions, level of education of the farmers, pest pressure, resistance status, just to name of few, also need to be taken into account. Overemphasising the risks reduces the benefits. Insufficient crop protection is a waste of resources. As has been said before, we must increase production. If we cannot do it through increased yields, we will have to claim more land from nature by reducing the area for forests or wildlife. Now-a-days, this is bad politics everywhere, also in a wealthy Europe. Safer products will come from new chemicals with better intrinsic properties like lower use rates, higher selectivity, and lower persistence. And they will also come from innovations in formulation, packaging, application equipment, and the timing of applications. In the area of packaging, a new development from Ciba by the name of LinkPak could be~cited, which is a closed filling system that uses re-tillable containers. So the farmer has no problem with contamination or the disposal of empty packages. Beside safety, performance is the other major driving force: Better performance for the farmer means lower application rates or fewer applications, means better suitability for system approaches like the integrated pest management (IPM) or resistance management, and means - more generally- better reliability, versatility, and ease of handling. In addition, changes in cropping methods, like no-till cropping or rice that is sown and not transplanted, will generate the need for new products with different properties. Better performance for the farmer translates into cheaper products for the consumer: This is not so important for many of us in Europe, but very important for a lot of people as the hunger in the world is directly related to poverty. Lower costs of food
53
means fewer people starving. This is something to remember for those people who extrapolate from our European surpluses to world-wide agriculture. And, last but not least, better performing products will certainly also help to secure the profitability of the crop protection industry and thus continuation of their research activities.
6. 5th Thesis Crop protection has tremendously improved in the last 50 years and will continue to do so. Chemical synthesis will remain the most important source for new active ingredients. Natural products will not per se play a dominant role in the near future.
Research adapted to the new needs and developments by improving the discovery process. Although serendipity still is the main avenue to new and better products, two major changes have become apparent. The first change is a considerable increase in the number of compounds tested, thereby enhancing the chances to find new and better products by luck, and the other exploits new biological targets by making use of the fast growing knowledge about biochemical and physiological processes in important crop plants and in pests and beneficial organisms. Random screening where a large number of molecules is evaluated in selected biological tests has been the standard method for many years. It certainly was improved over time, but will now be drastically expanded. 20 years ago, block screening tests needed gram quantities of chemicals. Therefore, relatively few compounds only were available. Today many more compounds can be screened as 100 mg is sufficient. The aim is to increase the access to new chemistry further by decreasing the amount of material needed. As a consequence, the focus is now on high-speed synthesis, high throughput screening systems and specialised tests. The standard block screen in Ciba tests in stage 1A about 12,000 compounds a year which come in regular amounts of 100 to 200 mg. A much larger number of structures about 100,000 will be evaluated in a 'high throughput screen' (HTS). Such a new test system has to be automated and work with specialised biological and biochemical mini tests that do not
need more than 1 mg. Promising structures will then have to be synthesised in larger amounts for the standard block screen. Chemical synthesis will remain the most important source for new active ingredients. Robots will help to increase the efficiency. Natural products will become more important, especially those from plants and micro-organisms. However, it does not seem probable that they per se will play a dominant role in the near future. Nevertheless, the structural richness of the natural products make them interesting research objects and a valuable source for new ideas for synthesis chemists. Fludioxonil (Gehmann et al., 1990) and pymetrozinc (Fliickiger et al., 1992) are two examples to illustrate this type of approach: The phenylpyrroles are a successful example of a new class of fungicides closely related to a natural bioactive compound. Pyrrolnitrin is a secondary metabolite produced by Pseudomonas and some Myxobacterales. It served as the lead structure for the phenylpyrroles fenpiclonil and fludioxonil. Pyrrolnitrin has a high activity against various plant pathogenic fungi but has the disadvantage of being very photolabile. The new phenylpyrrols are at least as active as pyrrolnitrin and are significantly more stable in light. From the main areas of improvement, as mentioned above, two points should be highlighted: environmental safety and resistance management. Coating of the seeds results in optimal targeting of this fungicide and, as a consequence, in a very low rate of application per hectare, thereby keeping the environmental burden to a minimum. The new mode of action makes fludioxonil a very valuable tool in resistance management. The second example is pymetrozine which represents a novel class of insect control agents with again a new mode of action. The compound affects the behaviour of homopterous insects and causes them to stop feeding which eventually leads to their death. It has an excellent activity against susceptible and resistant aphids, whiteflies and hoppers. Thanks to its selectivity it is an ideal IPM product and, again, its new mode of action makes it a very valuable tool in resistance management programs. More generally, in the field of insect control, new chemicals like pymetrozine or the Bayer product imidacloprid, with new modes of action have already
54 been or will soon be introduced. They reduce the dependence on the small number of chemical classes that presently dominate the markets, e.g. the organophosphates, carbamates, and pyrethroids. And, among others, they will improve our choices in integrated pest management and in resistance management. The second approach for the discovery of new solutions exploits new biological targets by making use of the fast growing knowledge about biochemical and physiological processes in important crop plants and in pests and beneficial organisms. Two examples are given to illustrate this approach. The first is a new method to fight fungal and microbial diseases. In the course of evolution plants have developed natural defence mechanisms that help to protect them from damage by pathogens. One important self defence mechanism is a phenomenon called Systemic Activated Resistance (SAR). When attacked by fungi, bacteria or viruses, plants react with the development of local necrosis which aim to block further development of the pathogen. In addition, plants produce a signal molecule which activates the defence systems throughout the plant - very similar to an immunisation achieved by vaccination and, thereby, helps the plant to help itself. New biological and biochemical test systems made it possible to specifically search for compounds that could imitate this biological process and, thereby, activate the corresponding resistance mechanism. The project already was successful. A molecule (CGA 245'704) was found that stimulates the plant defence mechanisms and, as a consequence, provides effective and long-lasting protection against a wide range of pathogens without being a fungicide, i.e. without having a direct effect against plant pathogens (Kessmann et al., 1996). The other example of using biological knowledge stands under the title: Attract and Kill. This method combines the attracting power of pheromones with insecticides. Male codling moths are attracted to drops of the female pheromone placed on the leaves. If the drop contains also some insecticide, e.g. cypermethrin, the male insect will die instead of enjoying the expected female companionship. By this method, populations are kept low by eliminating the male insects. In field trials in Egyptian cotton against the
pink bollworm Attract and Kill gave much better results than conventional insecticide sprays (Hofer et al., 1995). Manual application of the droplets had to be done which is certainly more laborious than aerial sprays. On the other hand, the high specificity and the extremely low residues have to be put into the cost and benefit evaluation.
7. Concluding remarks To improve crop protection means to optimise in many different directions. It ~is a difficult task with many contradicting objectives and needs of farmers, the industry and the society. It is the task of industry to discover and develop today those products that the farmer will need in 10 to 20 years from now and that will then be accepted by the local society. Crop protection is of public interest and has to be discussed publicly. Industry has to accept its role in these discussions and has to acquire the relevant capabilities. The case of 'Brent Spar' between Shell and Greenpeace has neatly shown how important it is today, not only to know the right arguments but also to communicate them professionally. The last 50 years brought chemical crop protection to a very high standard. The next years will bring various changes, many will be linked to biotechnology. To improve our crop protection methods will remain a basic need for us and our children. No prejudice should impede the implementation of the best solutions. We have to integrate all our knowledge, methods and products to optimise agriculture, be they chemical, biological or biotechnological. There is no reasonable alternative.
References Fliickiger, C.R., Kristinsson, H., Senn, R., Rindlisbacher, A.S., Buholzer, H. and Voss, G., 1992. PYMETROZINE- A novel agent to control aphids and whiteflies. Brighton Crop Protection Conference- Pests and Diseases l: 43-50. Gehmann, K., Nyfeler, R., Leadbeater, A.J., Nevill, D. and Sozzi, D., 1990. CGA 173'506: A new phenylpyrrole fungicide for broad-spectrum disease control. 1990 Brighton Crop Protection Conference- Pests and Diseases 2: 399-406. Hofer, D. and Angst, M., 1995. Control of Pink Boll Worm in Cotton with SIRENE, a Novel Sprayable Attract and Kill
55
Formulation. Proceedings Beltwide Cotton Conferences, San Antonio, TX: 949-952. Kessmann, H., Oostendorp, M., Ruess, W., Staub, T., Kunz, W. and Ryals, J., 1996. Systemic Activated Resistance- A new Technology for Plant Disease Control. Pesticide Outlook, June 1996: 10-13. Meinert, G., 1991. Minor uses - a difficult proplem. Gesunde Pflanzen, 43: 238-240. Oerke, E-C., Dehne, H-W., Sch6nbeck, F. and Weber, A., 1994. Crop Production and Crop Protection. Elsevier, Amsterdam, 808 pp.
The Pesticide Manual, 1994. C. Tomlin (Editor). British Crop Protection Council, Surrey and The Royal Society of Chemistry, Cambridge, UK, 1341 pp. Wood Mackenzie, 1995a. Agrochernical Service, Update of the Companies Section, 307 p. Wood Mackenzie, 1995b. Agrochemical Service, Update of the Countries Section, 167 p. Wood Mackenzie, 1995c. Agrochemical Service, Update of the Products Section, 198 p.
This Page Intentionally Left Blank
© 1997 Elsevier Science B. V. All rights reserved Perspectives for Agronomy - Adopting Ecological Principles and Managing Resource Use M.K. van lttersum and S.C. van de Geo'n (Editors)
57
Emissions of CO2, CH4 and N20 from pasture on drained peat soils in the Netherlands C.A. Langeveld a'*, R. Segers a, B.O.M. Dirks a, A. van den Pol-van Dasselaar b, G.L. Velthof c, A. Hensen d aDepartment of Theoretical Production Ecology, Wageningen Agricultural University, P.O. Box 430, 6700 AK Wageningen, The Netherlands bDepartment of Soil Science and Plant Nutrition, Wageningen Agricultural University, P.O. Box 8005, 6700 EC Wageningen, The Netherlands CNMI, Department of Soil Science and Plant Nutrition, Wageningen Agricultural University, P.O. Box 8005, 6700 EC Wageningen, The Netherlands dNetherlands Energy Research Foundation ECN, P.O. Box 1, 1755 ZG Petten, The Netherlands Accepted 8 April 1997
Abstract Soils have an important role in the global budgets of the greenhouse gases carbon dioxide ( C O 2 ) , methane (CH4) and nitrous oxide (N20). In particular, peat soils are considered to exhibit relatively high emission rates. The purpose of this study is to make an integrated assessment of the emissions of CO2, CH4 and N20 from pasture on drained peat soils in the Netherlands (almost 10% of the total land area). The study is based on monitoring studies, described or to be described in more detail elsewhere, at the experimental farm 'R.O.C. Zegveld'. Besides emissions from the soil-plant part of the system, CO2 emission from cattle and their excreta, and emissions of CH4 and N20 from excreta in grazed pastures were included. The annual fluxes of CO2, CH4 and N20 were estimated. The estimated net CO2 emission was (l I + 3) x 103 kg/ha per year, CH4 emissions ranged from --0.3 + 0.1 to 0.1 + 0.1 kg/ha per year and N20 emissions from 14:1:1 to 61 + 4 kg/ha per year. By extrapolation we estimated the contribution of the investigated sources on drained peat pastures to the overall national greenhouse gas emission of 244 x 109 kg CO2 equivalents/year, assuming that (l) the emissions at the monitored site were representative for the Netherlands, and (2) the uncertainties found in the monitoring studies were the sole sources of uncertainty. The percentage contributions to the overall national greenhouse gas emission were estimated at 1.3 _.+0.3% for CO2, 0.0% for CH4 and 0.9 + 0.1% for N20. © 1997 Elsevier Science B.V. Keywords: Carbon dioxide; Enhanced greenhouse effect; Groundwater table; Methane; Nitrous oxide; Pasture; Peat
1. Introduction The enhanced greenhouse effect refers to a rise of the equilibrium temperature at the earth's surface due *Corresponding author. Tel.: +31 317 482140; fax: +31 317 484892; email:
[email protected]
to anthropogenic emissions of greenhouse gases into the atmosphere. Carbon dioxide, methane and nitrous oxide are major gases responsible for this enhanced greenhouse effect (Table 1). Soils have an important role in the global budgets of these gases. Changes in tropical land use probably constitute about 10% of the global CO2 sources
Reprinted from the European Journal of Agronomy 7 (1997) 35-42
58
(IPCC, 1995). Wetlands and rice paddies may contribute about 30% to the global gross emission of CH4 (Prather et al., 1995). Probably at least 60% of the global gross N20 emission evolves from soils, in particular wet tropical forest soils and cultivated soils (Prather et al., 1995). The main processes underlying exchange of greenhouse gases between soils and the atmosphere are of a biological and physical nature. CO2 is produced by respiration of various organisms and consumed in assimilation by plants. CH4 is produced in methanogenic breakdown of organic compounds and consumed in methane oxidation, both by micro-organisms. Denitrification and nitrification by microorganisms can produce N20; N20 is also further reduced or consumed during denitrification. The presence of organic material drives respiration, methanogenesis and denitrification. Therefore, peat soils which have a high organic matter content could constitute important sources. Drainage probably affects the balance between emissions of CO2, CH4 and N20, and the overall contribution to the enhanced greenhouse effect. Nyk~inen et al. (1995), for example, found higher CO2 and N20 emissions and a lower CH4 emission at a site drained for agriculture than at a corresponding non-drained 'virgin' site, resulting in a higher overall contribution to the enhanced greenhouse effect at the drained site. Besides drainage the fertilisation and grazing regimes affect the emissions of greenhouse gases from drained peat soils used for dairy farming. In the Netherlands, peat soils cover approximately 0.28 × 10 6 ha (F. de Vries, Winand Staring Centre, Wageningen, pets. commun.), almost 10% of the total land area. Most soils are drained, fertilised and used as pasture land. About 30% of the pastures in the Netherlands is situated on drained peat. Thus far, no estimates based on measurements in the Netherlands have been made of CO2 and CH4 emissions from drained peat pasture soils. Kroeze (1994) estimated N20 emissions from this source, but did not account for differences in groundwater tables. This study aimed at integrating the field monitored data available in the Netherlands on emissions of CO2, CH4 and N20 from drained peat pasture soil and at assessing the importance of emissions from this soil type. We focused on the soil-plant-atmosphere part of the system, but included CO2 emissions from
cattle and their excreta, and CH4 and N20 emissions from excreta in the case of grazing. Methane emission by cattle was omitted, however.
2. Materials and methods
2.1. Site description Emissions of CO2, CH4 and N20 were monitored at the experimental farm 'R.O.C. Zegveld', The Netherlands. The soil is a Terric Histosol (FAO classification). Within the refined Dutch soil classification system (De Bakker and Schelling, 1989) the soil is a 'koop' peat soil. The top layer of about 0.2 m has a relatively high clay content of about 300 g/kg soil (Velthof and Oenema, 1995b) and covers a 6-7-m thick layer of wood sedge peat. With respect to greenhouse gas emissions, we considered the site an adequate representative of the peat pasture soils used for intensive dairy farming in the Netherlands for two reasons. 'Koop' peat soil is the most common of the nine peat soil types found in the Netherlands (Hendriks, 1992), while the management like imposed groundwater tables, fertilisation regime, mowing/ grazing regime, composition and stocking density of the cattle, reflects the common farming practice on these soils well. Since 1969 the soil area at the farm has been divided into two sections, one where a high groundwater table was maintained (high GWT; median at 0.3 m depth), and one with a low groundwater table (low GWT, median at 0.5 m depth). The average carbon content in the 0-0.2 m layer of the high GWT plots was 156 g C/kg dry soil at a C/N ratio of 9.6, the corresponding values at the low GWT plots were 223 g C/kg dry soil and 12.0 (Velthof and Oenema, 1995b). N fertiliser was applied in split applications, as calcium ammonium nitrate (CAN). Applications ranged from 161 to 464 kg N/ha per year (Velthof et al., 1996).
2.2. Monitoring studies Instantaneous CO2 fluxes were measured throughout April, May, June, August, September and October, 1994 (Hensen et al., 1995), by an eddy correlation method having a fetch of approximately 500 m. The resolution of the monitor was 0.1 p.p.m, by volume
59
with a reproducibility of the concentration measurements, i.e. a coefficient of variation for repeated measurements of a standard concentration, of 0.03%. The error in a 30-min averaged flux value was about 10% due to the combination of errors in measurements of the vertical wind velocity and CO2. The set-up of the flux measurements implied that treatments differing in GWT, grazing/mowing regime and N fertilisation regime were spatially integrated, and that CO2 emitted by cattle was largely included in the measurements. Besides the CO2 fluxes air temperature (T) and global radiation (R) were recorded during the monitoring period. Local instantaneous CH4 and N20 fluxes at different treatments were monitored with vented closed flux chambers (Van den Pol-van Dasselaar and Oenema, 1997; Velthof and Oenema, 1995a,b), sampling 0.10.2 m 2 weekly or bi-weekly. The specifications for the CH4 monitoring study were: reproducibility of the concentration measurements 0.08%; lower limit detectable fluxes about 0.03 mg CHn/m2 per day. The specifications for the N20 study were: reproducibility of the concentration measurements 5%; lower limit detectable fluxes about 0.4 mg N20/m 2 per day. 2.3. Estimation of annual emissions The instantaneous CO2 flux measurements were used to estimate the annual net CO2 emission. We estimated the instantaneous total CO2 emission by cattle by evaluating their C balances. The C uptake was calculated using the tabulated energy requirement of the cattle (IKC, 1993), ignoring growth. It should be noted that the calculated CO2 emission by cattle is partly derived from C in feed concentrates (about 30% of the C taken up annually by dairy cows in the Netherlands comes from feed concentrates; E.A. Lantinga, Dept. Theoretical Production Ecology, Wageningen, pers. commun.). The pasture's (upward) respiratory CO2 fluxes (Fc.r) were calculated by subtracting the cattle's CO2 emission from the measured night-time CO2 fluxes. Subsequently, an exponential relationship was fitted between T and Fc,r. The instantaneous respiratory CO2 fluxes from pasture and cattle were subtracted from the measured instantaneous day-time CO2 fluxes to obtain the (downward) assimilatory CO2 fluxes (Fc,a). A hyperbolic relationship was fitted
between R and Fc, a. Diurnal patterns of radiation (R) and temperature (T) during 1994 were reconstructed (Goudriaan and Van Laar, 1994) using daily R and the minimum and maximum T for Wageningen, 50 km east of the site. To estimate the annual net CO2 emission, the two obtained relationships - between T and Fc,r, and between R and Fc,a - were applied to the full year's diurnal pattems of T and R. The uncertainty in the estimated annual net CO2 emission associated with the statistical error in the two obtained relationships was calculated using the principles of Monte Carlo simulation: 1. The relationships between T and the coefficient of variation of Fc,~, and between R and the coefficient of variation of Fc,a, were determined considering classes of 100 successive (T, F~,r) and (R, F~,~) pairs. 2. Drawings were done from the normal distribution to obtain stochastic realisations of F~,r and F~,~for each half hour of the year. We assumed that any drawing applied to both Fc,~and F~,a(i.e. the deviations from their fitted values were correlated). 3. Net CO2 emissions for a year were calculated from 17520 (=365 x 48) realised values of Fc.r and Fc,a. 4. The estimated annual net CO2 emission and its uncertainty were obtained as the stabilised average and standard deviation of repeated calculations of the annual emissions, respectively. The annual CH4 and N20 emissions were obtained by trapezoidal integration of the mean fluxes over time (Van den Pol-van Dasselaar and Oenema, 1997; Velthof et al., 1996). An upper limit for the uncertainty of the annual CH4 emission was estimated by applying an error propagation analysis on the trapezoidal integration method. We assumed that (1) the maximum of 0.2 mg CH4/m2 per day in the typical range of standard deviations found in replicated measurements (Van den Pol-van Dasselaar and Oenema, 1997) was relevant during the whole observation period, and that (2) bi-weekly monitoring took place. For N20, we assumed that the differences between the two observation years were an adequate measure for the variability of the annual emissions. The total emissions from peat pasture soils in the Netherlands were estimated by extrapolating the Zegveld data in a straightforward linear way, discussed more extensively below.
60 Table 1 Characteristics of CO2, CH4 and N20 as greenhouse gases Gas
Pre-industrial atmospheric concentrationa (ppmv)
1992-Atmospheric GWPb (kg CO2 concentrationa equivalents/kg) (ppmv)
Radiativeforcing of emissions 1850presenta'c (Wm 2)
Global sum of sources° (109 kg/ year)
Sum of sources in the Netherlandse (109 kg/year)
CO2 CH4 N20
280 0.700 0.275
355 1.714 0.311
1.5 0.5 0.2
26000 + 4000 550 + 90 25
189 1.0 0.06
1 24.5 320
ppmv, parts per million by volume. aIPCC (1995). bDirect Global Warming Potential in 1995 following addition of 1 kg of each gas to the atmosphere, relative to CO2 for a 100-year time horizon (Albritton et al., 1995). CTotal direct radiative forcing 1850-present: 2.4 W/m2 (IPCC, 1995). dFrom budgets for CO2 (IPCC, 1995), and CH4 and N20 (Prather et al., 1995). eEstimates for 1994 (RIVM, 1995).
3. Results
(54.5 + 0.6) x 103 kg CO2/ha -I per year and a respiratory flux of the pasture of (55.8 + 2.5) x 103 kg CO2/ ha -1 per year. Including the respiration by the cattle
3.1. Carbon dioxide
We estimated the instantaneous total CO2 emission by cattle at 27.5 kg CO2/ha per day (Table 2). The obtained regression equation for the pasture's respiratory CO2 flux (in mg CO2/m2 per s) was Fc,r = 0.33 x 2.20 ~r-20)/10 (n = 1009~ r 2 = 0.12; T is temperature, in °C). The regression equation for the assimilatory CO2 flux (in mg C O y m 2 per s) was Fc,a = - 1 . 1 1 x R/(335 + R)+0.03 (n = 2182; r 2 = 0.67; R is global radiation, in W/m2). Using these equations and applying the described m e t h o d to estimate uncertainties we arrived at a cumulative assimilatory flux of the pasture of
of (10.0 + 1.6) x 103 kg C O J h a per year (Table 2), we arrived at an estimate for the total annual net CO2 emission in 1994 of (11 + 3) × 103 kg CO2/ha per year (Table 3). Since the cattle largely feeds on the grass and subsequently respires it, it was considered to be an integral part of the pasture ecosystem. As a consequence, a relevant estimate for the CO2 emission by the pasture sec could not be given. The open nature of intensive dairy farming ecosystems - in terms of flows of matter and energy makes any system boundary and associated CO2 emission an arbitrary choice. Moreover, annual variations in pasture productivity and respiration
Table 2 Calculation of daily CO2 emissions from cattle at Zegveld using C balances Cattle type
C uptakea (kg C/head per day)
C not excreted as manureb (kg C/ head per day)
Milkc (kg C/ head per day)
CO2 emission (kg C/head per day)
Stockingdensityd (number of heads/ha)
CO2 emission (kg CO2/ha per day)
Female adult Female young (>1 year) Calves (<1 year) Total
6.56 + 0.92 3.2 + 0.4
4.92 + 0.77 2.4 + 0.4
1.0 + 0.1
3.92 + 0.78 2.4 + 0.4
1.48 0.41
21.3 + 4.2 3.6 + 0.6
1.9 + 0.2
1.4 + 0.2
1.4 + 0.2
0.50
2.6 + 0.4 27.5 + 4.3
aUsing IKC (1993) and Van Dijk et al. (1995); assumption: C content dry matter 40%. Uncertainties from estimated uncertainties in underlying dairy index numbers. bSome of the herbage taken up (75 + 5%) is digested (Meijs, 1981). CRegional data (Van Dijk et al., 1995). Assumptions: milk dry matter content 13.2% (IKC, 1993), milk dry matter C content 40%. dZegveld data (Van Houwelingen, R.O.C. Zegveld, pers. commun.).
61
Table 3 Estimates of annual emissions of CO2, CH4 and N20, and their uncertainties, for different treatments at drained peat pasture soils in Zegveld, The Netherlands, and annual contribution to the enhanced greenhouse effect Gas
Treatment
Annual emission Annual contribu(kg/ha year) tion enhanced greenhouse effect (kg CO2 equivalents/ha per year)
CO~ Integrated (11 +3) x 103 CH4b HighGWT, grazed 0.1 + 0.1 High GWT, mown -0.2 + 0.1 Low GWT, grazed -0.3 + 0.1 Low GWT, mown -0.3 + 0.1 N20c High GWT, grazed 23 + 4 High GWT, mown 14 +_ 1 Low GWT, grazed 61 + 4 Low GWT, mown 28 + 3
(11 +3) x 103 2+2 -5 + 2 -7 + 2 -7 + 2 (7 -I- 1) × 103 (4.5:1: 0.3) x 103 (20 + 1) x 103 (9 :t: 1) x 103
GWT, groundwater table. a1994; CO2 emitted from cattle and their excreta included; uncertainties estimated via Monte Carlo simulation. bl994; CI-I4 from cattle not included; uncertainties estimated via error propagation analysis. CAverages + absolute deviations from these averages, for the periods March 1992-March 1993 and March 1993-March 1994 (Velthof et al., 1996).
found during the growing season, 1-3 weeks after fertiliser application. During dry periods in the summer and during the winter, fluxes were low. We also conclude that there was a clear difference between the treatments. Higher fluxes in the grazed treatments may have resulted from the stimulation of N20 production by cattle excreta. Higher N20 fluxes at lower groundwater tables might be explained by the combination of increased availability of C and N compounds, enhanced nitrification and an increased NEO/N 2 end product ratio in denitrification.
4. Discussion
4.1. Comparison of Zegveld with another drained peat pasture site
further complicate the establishment of robust emission figures.
We found substantial emissions of CO2 and N20, while those of CH4 were negligible or slightly negative (Table 3). Using the different Global Warming Potentials for the three gases, we conclude that the contributions of the CO2 and N20 emissions to the enhanced greenhouse effect had the same order of magnitude, while the contribution of CH4 emissions was negligible. Nyk~inen et al. (1995) also carded out a study on the emission of C O 2 , CH4 and N20 from a drained peat
3.2. Methane
Table 4
Measured CH4 emissions were negligible or slightly negative (Table 3). This suggests that relatively aerobic soil conditions suppressed methanogenic processes in favour of methanotrophic processes. Besides, the abundance of other electron acceptors than oxygen, like nitrate, probably suppressed methanogenesis during wet periods in winter.
Ranges of estimated average annual emissions of CO2, E l 4 and N20, and annual contribution to the enhanced greenhouseeffect of a hypothetical fen in the Netherlands, based on literature data Gas
Annual emission (kg/ha per year)
Annual contribution to the enhanced greenhouse effect (kg CO2 equivalents/ha per year)
CO2 a
1.5 x 103 to -0.5 x 103 100 to 160 0.07 to 0.7
- 1.5 x 103 to --0.5 x 103
-
3.3. Nitrous oxide
CH4b N20c Total
2.5 x 103 to 3.9 x 103 0.02 x 103 to 0.2 x 103 1 x 103 to 4 x 103
Compared with N20 emissions from mineral soils (Velthof et al., 1996), the N20 emissions in Table 3 were relatively high. The high availability of carbon (C) and nitrogen (N) compounds used in denitrification, combined with high denitrification capacities and locally low partial oxygen pressures, probably promoted N20 fluxes. In general, high fluxes were
aUsing typical average values on C accumulation: 0.1 x 103 to 0.4 X 103 kg C/ha per year (Gotham, 1991; Korhola et al., 1995; Armentano and Menges, 1986). bAssumption: average flux of 87:1:18 mg CH4/m2 per day during an emission season of 150 days, zero fluxes during the rest of the year (Bartlett and Harriss, 1993). CUsing regression lines of N20 flux with water table of Moore (1996) and assuming an average groundwater table at 0.1 m depth.
62 soil. However, grazing cattle was absent at their site in Finland. They observed a high COz emission of 22 x 103 kg/ha per year. This might be explained by their removal of the green vegetation throughout the monitoring period. The N20 emission of 13 kg/ha per year they found is lower than we found in any treatment. This was probably associated with (i) the lower N fertilisation rate, 80 kg N/ha per year, and (ii) the absence of grazing cattle. Compared with our negligible or slightly negative CH4 emissions, they found a small positive CH4 emission.
4.2. Comparison with wetlands Implementation of policy decisions for conversion of the peat soils in the Netherlands to natural wetlands could result in the generation of fens. This would have implications for agriculture (less area for dairy farming) and environment ('nature restoration', change of emissions). To assess the impact of this type of land use change on greenhouse gas emissions, we estimated the emissions of CO2, CH4 and N20 from a hypothetical fen using the typical ranges of data found in literature (Table 4). Compared with the data of Table 3, the N20 and CO2 emissions for fens were much lower and the CH4 emissions much higher. The total contribution to the enhanced greenhouse effect per hectare of pasture on drained peat in the Netherlands is probably larger than the contribution per hectare of fens. The main causes for the differences are the increased aerobicity, the presence of cattle, and N fertilisation at the drained site. To put the CH4 emission from the hypothetical fen in perspective, we estimated the CH4 emission from cattle at Zegveld. Using the specific emissions by cattle under Dutch dairy farming conditions (grazing regime, constitution of diets) of Van Amstel et al. (1993), we estimated this emission at 200 kg CHa/ha per year, which is of the same order of magnitude as the CH4 emission from the fen.
4.3. Emissions from drained peat pastures compared to the total national emissions To assess the importance of emissions from drained peat pastures in the contribution of the Netherlands to the enhanced greenhouse effect we extra-
polated the data from Zegveld. With this we did a number of assumptions that seemed reasonable, taken into account the methodological marginal notes with the estimation procedure of the emissions, discussed below, and the still limited quantitative knowledge on the dependence of the emissions on environmental factors. Further studies on the possibly limited validity of the assumptions could be used to refine to extrapolation procedure. The assumptions were:
1. In the monitoring studies reliable estimates for the instantaneous, average fluxes from the different treatments at Zegveld were obtained. The method for monitoring CO2 implied that relatively large areas were sampled, resulting in representative measurements for these areas. However, as a result, spatial integration over plots differing in management took place and we were not able to identify effects of N fertilisation or groundwater tables in our data set. In the CH4 and N20 monitoring studies, the sampling was confined to small areas. The estimation of fluxes for larger areas from these samples is methodologically difficult, because of the high spatial variability of the fluxes. 2. The calculated annual emissions estimates were representative, despite the temporal confinement of the observations. All fluxes are known to exhibit temporal variability. The use of regression equations for relating the CO2 fluxes to temperature and radiation to estimate the total annual emission is a relatively reliable method, because of the critical dependence of the relevant biological processes on temperature and radiation. Annual CH4 and N20 emissions were estimated by trapezoidal integration of flux estimates obtained from weekly or bi-weekly measurements during day-time. This method did not account for possible effects of (i) a strongly non-linear temporal behaviour of fluxes at a smaller time scale, and (ii) a diurnal pattern of the fluxes. 3. The high and low GWT plots in Zegveld represent pastures on drained peat in the Netherlands in classes II/II* (69% of the total peat pasture area of 0.28 x 106 ha) and III/III* (20%) of the soil map of the Netherlands (F. de Vries, Winand Staring Centre, Wageningen, pers. commun.; classes
63
described in Van der Sluijs, 1992), respectively. The on average lowest GW"r is between 0.5 and 0.8 m depth for class II/II*, and between 0.8 and 1.2 m depth for class III/III*. By this classification, the high GWT plot at Zegveld was classified as II/II*, and the low GWT plot as III/III*. We neglected differences between the types of peat soil and duration of the drainage. 4. The emission per hectare from the remaining pastures on drained peat (11%) equals the average of the emission per hectare in class II/II* and in class III/III*, using the relative areas as weight factors. Quantitative information on the relation between GWT and emissions is scarce. This assumption was introduced as an obvious possibility. We expect that possible errors due to this assumption were small, because of the relative small area of these soils. 5. The effect of Nfertilisation on other peat pastures is comparable to that of the N fertilisation at Zegveld. Most pastures on drained peat are N fertilised. CAN, which was used at Zegveld, is the most important N fertiliser in the Netherlands (LEI, 1996). The CAN application rates at Zegveld corresponded with recommended total N application rates in the Netherlands amounting to 200-400 kg N/ha per year (Unwin and Vellinga, 1994). In the fanning practice in the Netherlands, a part of the N is applied via cattle slurry. The effect of cattle slurry on greenhouse gas emissions is not clear; unpublished exploring studies suggest lower N20 and higher CH4 emissions than for CAN. 6. Effectively half of the drained peat pasture area is used for grazing, the other half for mowing. We assumed that half of the grass production is directly consumed by grazing cattle during the grazing season of about 180 days. The other half is mown and supplied as silage during the rest of the year. This assumption is reasonable for dairy farms where fresh and conserved grass are the only farm-grown roughage components. Using the data in Table 3, we estimated the respective total emissions from pastures on drained peat soils in the Netherlands at (3.1 + 0.8) x 109 kg CO2/year, (-0.03 + 0.02)x 106 kg CHa/year and (6.9 + 0.5) x 106 kg N20/year. The last figure is in
the lower half of the range of 2.6 x 106-13.2 x 106 kg N20/year that can be derived from Kroeze (1994) for N20 emissions from drained peat pasture soils. By implicitly assuming that high and low GWT plots were equally important, Kroeze (1994) probably slightly overestimated N20 emissions from drained peat pasture soils. The emissions we found corresponded to (3.1 + 0.8)x 109 kg CO2 equivalents/ year for CO2, (-0.7 + 0.5)x 106 kg CO2 equivalents/year for CH4 and (2.2 + 0.2)x 109 kg CO2 equivalents/year for N20. We compared these figures with the estimated total greenhouse gas emissions in the Netherlands. In 1994, the emissions of CO2, CH4 and N20 corresponded with 189 x 109, 25 x 109 and 19x 109 kg CO2 equivalents/year, respectively (RIVM, 1995). Including 11 x 109 kg CO2 equivalents/year for other greenhouse gases (Van Amstel et al., 1994), the estimated overall emission of greenhouse gases in the Netherlands in 1994 was 244 x 109 kg CO 2 equivalents/year. Thus, CO2, CH4 and N20 emitted from pastures on drained peat soils could constitute about 1.3 + 0.3, 0.0 and 0.9 __ 0.1%, respectively, of this overall emission. In these figures 02 emitted from cattle and their excreta was included, while CH4 emitted from cattle was not included. It should be noted that the given uncertainties only reflect the uncertainties associated with the respective monitoring studies. An attempt to quantify the uncertainties associated with the assumptions made for the extrapolation does not seem relevant yet. This because neither sufficient knowledge of the processes underlying emissions, nor sufficient field measurements are available to do so in a reliable way.
Acknowledgements Funding for this study was received from the National Research Programme on Global Air Pollution and Climate Change, and the C.T. de Wit Graduate School Production Ecology. We thank J. Goudriaan, M.H.H. Hermsen, M. Hoosbeek, K. van Houwelingen, L. Klemedtsson, C. Kroeze, E.A. Lantinga, P.A. Leffelaar, H. Nyk/inen, O. Oenema, R. Rabbinge and F. de Vries for various contributions.
64
References Albritton, D.L., Derwent, R.G., Isaksen, I.S.A., Lal, M. and Wuebbles, D.J., 1995. Trace gas radiative forcing indices. In: J.T. Houghton, L.G. Meira Filho, J. Bruce, Hoesung Lee, B.A. Callander, E. Haites and K. Maskell (Editors), Climate Change 1994. Cambridge University Press, Cambridge, pp. 205-231. Armentano, T.V. and Menges, E.S., 1986. Patterns of change in the carbon balance of organic soil - wetlands of the temperate zone. J. Ecol., 74: 755-774. Bartlett, K.B. and Harriss, R.C., 1993. Review and assessment of methane emissions from wetlands. Chemosphere, 26: 261320. De Bakker, H. and Schelling, J., 1989. Systeem van Bodemclassificatie voor Nededand: de Hogere Niveaus. PUDOC, Wageningen, 209 pp. (in Dutch). Gorham, E., 1991. Northern Peatlands: role in the carbon cycle and probable responses to climate warming. Ecol. Appl., 1: 182195. Goudriaan, J. and Van Laar, H.H., 1994. Modelling Potential Crop Growth Processes. Kluwer, Dordrecht, 238 pp. Hendriks, R.F.A., 1992. Afbraak en Mineralisatie van Veen. Rapport 199. DLO-Staring Centrum, Wageningen, 152 pp. (in Dutch). Hensen, A., Kieskamp, W.M., Vermeulen, A.T., Van den Bulk, W.C.M., Bakker, D.F., Beemsterboer, B., M61s, J.J., Veltkamp, A.C. and Wyers, G.P., 1995. Determination of the Relative Importance of Sources and Sinks of Carbon Dioxide. Report ECN-C-95-035. ECN, Petten, 68 pp. IKC, 1993. Handboek voor de Rundveehouderij. IKC, Lelystad, 629 pp. (in Dutch). IPCC (Working Group 1), 1995. Summary for policy makers: Radiative forcing of climate change. In: J.T. Houghton, L.G. Meira Filho, J. Bruce, Hoesung Lee, B.A. Callander, E. Haites and K. Maskell (Editors), Climate Change 1994. Cambridge University Press, Cambridge, pp. 7-34. Korhola, A., Tolonen, K., Turunen, J. and Jungner, H., 1995. Estimating long-term carbon accumulation rates in boreal peatlands by radiocarbon dating. Radiocarbon, 37: 575-584. Kroeze, C., 1994. Nitrous Oxide (N20). Emission Inventory and Options for Control in the Netherlands. RIVM, Bilthoven, 163 PP. LEI (Landbouw-Economisch Instituut), 1996. Landbouwcijfers 1995. LEI-DLO/CBS, The Hague. (in Dutch). Meijs, J.A.C., 1981. Herbage Intake by Grazing Dairy Cows. PhD Thesis. PUDOC, Wageningen, 264 pp.
Moore, T., 1996. Trace gas fluxes from northern peatlands. In: R. Laiho, J. Laine and H. Vasander (Editors), Northern Peatlands in Global Climatic Change. Oy Edita Ab, Helsinki, pp. 115-121. Nyk/inen, H., Aim, J., L~ng, K., Silvola, J. and Martikainen, P.J., 1995. Emissions of Cl-h, N20 and CO2 from a virgin fen and a fen drained for grassland in Finland. J. Biogeogr., 22: 351-357. Prather, M., Derwent, R., Ehhalt, D., Fraser, P., Sanhueza, E. and Zhou, X., 1995. Other trace gases and atmospheric chemistry. In: J.T. Houghton, L.G. Meira Filho, J. Bruce, Hoesung Lee, B.A. Callander, E. Haites and K. Maskell (Editors), Climate Change 1994. Cambridge University Press, Cambridge, pp. 73-126. RIVM, 1995. Achtergronden bij Milieubalans 95. RIVM, Bilthoven. (in Dutch). Unwin, R.J. and Vellinga, Th.J., 1994. Fertilizer recommendations for intensively managed grassland. In: L. 't Mannetje and J. Frame (Editors), Grasslands and Society. Proceedings of the 15th General Meeting of the European Grassland Federation. Wageningen Pers, Wageningen, pp. 590-602. Van Amstel, A.R., Albers, R.A.W., Kroeze, C., Matthijsen, A.J.C.M., Olivier, J.G.J. and Spakman, J., 1994. Greenhouse Gas Emissions in the Netherlands 1990, 1991, 1992 and Projections for 1990-2010. RIVM, Bilthoven, 93 pp. Van Amstel, A.R., Swart, R.J., Krol, M.S., Beck, J.P., Bouwman, A.F. and Van der Hoek, K.W., 1993. Methane - The other Greenhouse Gas. RIVM, Bilthoven, 108 pp. Van den Pol-van Dasselaar, A. and Oenema, O., 1997. Effects of grassland management on the emission of methane from intensively managed grasslands on peat soil. Plant Soil. (in press). Van der Sluijs, P., 1992. Hoofdstuk 11: Grondwatertrappen. In: W.P. Locher and H. de Bakker (Editors), Bodemkunde van Nederland. Deel 1. Algemene Bodemkunde. 2nd edn. Malmberg, Den Bosch, pp. 167-180. (in Dutch). Van Dijk, J.P.M., Douma, B.E. and Van Vliet, A.L.J., 1995. Bedrijfsuitkomsten in de Landbouw (BUL). Boekjaren 1990/91 t/m 1993/1994. Landbouw-Economisch Instituut, The Haag. (in Dutch). Velthof, G.L. and Oenema, O., 1995a. Nitrous oxide fluxes from grassland in the Netherlands: I. Statistical analysis of flux-chamber measurements. Eur. J. Soil Sci., 46: 533-540. Velthof, G.L. and Oenema, O., 1995b. Nitrous oxide fluxes from grassland in the Netherlands: II. Effects of soil type, nitrogen fertilizer application and grazing. Eur. J. Soil Sci., 46: 541549. Velthof, G.L., Brader, A.B. and Oenema, O., 1996. Seasonal variations in nitrous oxide losses from managed grasslands in the Netherlands. Plant Soil, 181: 263-274.
Section 3 CROP PHYSIOLOGY AND IDEOTYPING Effects of CO2 and temperature on growth and yield of crops of winter wheat over four seasons G.R. Batts, J.I.L. Morison, R.H. Ellis, P. Hadley and T.R. Wheeler ........................................................... 67 Reprinted from the European Journal of Agronomy 7 (1997) 43-52 Use of in-field measurements of green leaf area and incident radiation to estimate the effects of yellow rust epidemics on the yield of winter wheat R.J. Bryson, N.D. Paveley, W.S. Clark, R. Sylvester-Bradleyand R.K. Scott .................................................... 77 Reprinted from the European Journal of Agronomy 7 (1997) 53-62 Simulating light regime and intercrop yields in coconut based farming systems J. Dauzat and M.N. Eroy ................................................................................................................................. 87 Reprinted from the European Journal of Agronomy 7 (1997) 63-74 Improving wheat simulation capabilities in Australia from a cropping systems perspective: water and nitrogen effects on spring wheat in a semi-arid environment H. Meinke, G.L. Hammer, H. van Keulen, R. Rabbinge and B.A. Keating ..................................................... 99 Reprinted from the European Journal of Agronomy 7 (1997) 75-88 Comparison of CropSyst performance for water management in southwestern France using submodels of different levels of complexity C O. Stockle, M. Cabelguenne and P. Debaecke ......................................................................................... 113 Reprinted from the European Journal of Agronomy 7 (1997) 89-98 Root growth of three onion cultivars A.D. Bosch Serra, M. Bonet Torrens, F. Domingo Oliv~ and M.A. Melines Pages ...................................... 123 Interspecific variability of plant water status and leaf morphogenesis in temperate forage grasses under summer water deficit J.-L. Durand, F. Gastal, S. Etchebest, A.-C. Bonnet and M. Ghesqui&e .................................................... 135 Reprinted from the European Journal of Agronomy 7 (1997) 99-107 Evaluation of sunflower (Helianthus annuus, L.) genotypes differing in early vigour using a simulation model F. Ag~iera, F.J. Villalobos and F. Orgaz...................................................................................................... 145 Reprinted from the European Journal of Agronomy 7 (1997) 109-118 Options of breeding for greater maize yields in the tropics A. Elings, J.W. White and G.O. Edmeades .................................................................................................. 155 Reprinted from the European Journal of Agronomy 7 (1997) 119-132
This Page Intentionally Left Blank
1997 Elsevier Science B. V. All rights reserved Perspectives for Agronomy - Adopting Ecological Principles and Managing Resource Use M.K. van lttersum and S.C. van de Geijn (Editors)
67
Effects of CO2 and temperature on growth and yield of crops of winter wheat over four seasons G.R. Batts a'b'c, J.I.L. Morison d'*, R.H. Ellis b, P. Hadley c, T.R. Wheeler b aDepartment of Meteorology, University of Reading, PO Box 239, Reading, RG6 6AU, UK bDepartment of Agriculture, University of Reading, PO Box 236, Reading, RG6 6AT, UK CDepartment of Horticulture and Landscape, University of Reading, PO Box 221, Reading, RG6 6AS, UK dDepartment of Biological Sciences, John Tabor Laboratories, University of Essex, Wivenhoe Park, Colchester, C04 3SQ, UK
Accepted 1 July 1997
Abstract Crops of winter wheat (Triticum aestivum L. cv. Hereward) were grown in the field in four consecutive seasons from 1991/ 1992 to 1994/1995 at Reading, UK, within polyethylene-covered tunnels along which a temperature gradient was superimposed on the ambient temperature variation at normal atmospheric (ca. 370) or an increased [CO2] (ca. 700 #mol CO2 mol-t air), producing many environments from one sowing date in each season at one location. Mean seasonal temperatures varied by up to 4°C along the temperature gradient. Increased [CO2] had no effect on crop duration, or on the rate of reproductive development, which had the same temperature sensitivity across all years. A 2°C warming, on the 4-year ambient mean temperature (10°C), reduced crop duration by 42 days (from 254), and reduced the reproductive phase by 16 days (from 130). Crop biomass generally declined with increase in mean temperature, and was greater at increased [CO2], with the effect of increased [CO2] varying with temperature and between years (6-34% range in relative stimulation by increased [CO2]). Grain yield was substantially reduced by warmer temperatures, and increased by doubling [CO2], but the effect varied greatly between years and with temperature (7-168% range). There were both positive and negative interactions of temperature and increased [CO2] on biomass and grain yield. In all 4 years, the increase in grain yield from doubling [CO2] was negated by an increase in mean seasonal temperature of only 1.0-2.0°C. Year-to-year variation in the responses of biomass and grain yield to [CO2] and temperature resulted from differences in environmental conditions, influencing biomass partitioning and altering the role of different yield components. © 1997 Elsevier Science B.V. Keywords: Wheat; Growth; Yield; Interannual variation; CO2; Temperature
I. Introduction The potential impacts of an increasing atmospheric CO2 concentration ([CO2]) and warmer global temperatures on crop production are substantial (Watson et al., 1996). Increased [CO2] is expected to increase * Corresponding author. Fax: +44 1206 873416; e-mail:
[email protected]
primary biomass production in temperate crops through a stimulation of net photosynthetic rate and/ or a decrease in water use, whereas global warming per se reduces crop duration in determinate crops such as wheat (e.g. Slafer and Rawson, 1994; Batts et al., 1996; Morison, 1996) resulting in lower biomass production and reduced grain yields (Mitchell et al., 1993, 1995; Wheeler et al., 1996a; Batts et al., 1998), at least where only one crop is grown per
Reprinted from the European Journal of Agronomy 7 (1997) 43-52
68
year. Therefore, any increase in yield due to increased [CO2] may be partly or wholly offset by warmer conditions. However, many papers and several observations suggest that the stimulating effect of increased [CO2] on biomass accumulation is larger at higher temperatures (see Rawson, 1992, 1995), so that the final outcome on yield of these two key aspects of global environmental change, temperature and CO2 needs clarification and quantification. Changes in yield of small grain cereals such as wheat can occur through changes in tiller numbers per area and numbers of grain per ear, or through changes in grain size. Wheat plants grown in an increased [CO2] have frequently shown more tillers (e.g., Gifford and Morison, 1993; Mitchell et al., 1993; Rawson, 1995; Batts et al., 1996, 1998) with more grains per unit area from more ears (e.g., Rawson, 1995; Batts et al., 1998; see also review by Lawlor and Mitchell, 1991). Increased temperature can result in reduced smaller grains, fewer grains per ear, and fewer ears m -2 (e.g., Mitchell et al., 1993,
1995; Rawson, 1995), in addition to any effects of high temperatures during particular critical periods on grain fertility (e.g., Mitchell et al., 1993). Clearly, the potential interactions between temperature and [CO2] on wheat growth and yield are complex, and this paper reports a 4-year investigation of these interactions for winter wheat crops (cv. Hereward) grown in our field-based temperature gradient facility (Hadley et al., 1995; Batts et al., 1996). Our previous papers have reported the detailed results of individual years, but the aim here is to compare the responses to temperature and [CO2] between years, and to highlight the variation in responses observed.
2. M a t e r i a l s a n d m e t h o d s
Four field experiments were conducted at University of Reading School of Plant Sciences field unit (51o27 ' N, 0056 ' W). Certified seed of winter wheat cv. Hereward was sown in rows spaced 0.12 m apart
Table 1 Crop husbandry of winter wheat cv. Hereward grown in 4 years in double-wall temperature-gradient tunnels
Sowing date Mean establishment (plants m-2) Soil analysisa Previous crop Pre-emergence herbicide Fungicide fumigantsb Fertilisers (kg ha-I) Seedbed Top dressing 1 Top dressing 2 Top dressing 3 Pesticides Soil applications
Fumigantsb
1991/1992
1992/1993
1993/1994
1994/1995
8 Jan 1992 278
24 Nov 1992 298
8 Nov 1993 296
30 Nov 1994 250
4; 3; pH 6.5 Grass ley Tribunil at 2.25 kg ha-I, 9 Jan Fungaflor
2; 2; pH 6.7 Winter wheat Tribunil, 7 Dec
3; 2; pH 6.8 Maize Tribunil, 12 Nov
3; 2; pH 6.9 Winter wheat Tribunil, 5 Dec
Fungaflor
Fungaflor
Fungaflor
30 N lO0 N double ridge; 31 Mar 40 N late stem extension; 7 May
25 N; 70 P2Os; 110 K20 50 N pre-double ridge; 16 Feb 50 N pre-terminal spikelet; 6 Apr 45 N late stem extension; 4 May
30 N; 70 P205; 110 K20 50 N double ridge; l Mar 50 N terminal spikelet; 24 Mar 40 N mainstem flag leaf; I l May
30 N; 70 P2 05; 110 K20 50 N double ridge; 17 Mar 50 N pre-terminal spikelet; 27 Mar 40 N mainstem flag leaf; l0 May
Gamma-Col,9 Nov
Gamma-Col,8 Nov
Gamma-Col,1 Dec
Fumite 7000
Birlane 24 at 4.2 1 ha-I" 12 Nov Fumite 7000
Fumite 7000
-
Gamma-Col at 1.4 1 hal, 6 Jan
Fumite 7000
ap and K values, respectively, expressed as ADAS scil indices. bUsed following canopy closure until maturity.
69
on 8 January and 24 November 1992, 8 November 1993 and 30 November 1994 into a sandy loam soil to provide uniform stands of 250-300 plants m -2 (Table 1). Four 20 m by 1.75 m areas were sown in each experiment, and a double-wall polyethylenecovered (replaced at each sowing) temperature gradient tunnel was constructed over each crop area (see Batts et al., 1996, 1998). One pair of tunnels was controlled to a target concentration of ca. 360 ttmol CO2 mol -~ air ('normal'; with seasonal mean daily concentrations of 390, 380, 365 and 371 for years 1-4, respectively) and the other pair to a target of 700 /zmol CO2 mol -I air ('increased'; seasonal means of 713, 684, 698 and 691 for years 1-4). The precision of control of the high [CO2] treatment gave concentrations in the tunnel between 630 and 770 /xmol mol -~ for 60%, 86%, 78% and 94% of the duration of the experiment in years 1-4, respectively. The temperature gradient along each tunnel was imposed, and enrichment with CO2 began soon after sowing and maintained until harvest maturity. The temperature gradient facility was relocated ca. 40 m away for the experiments in 1993/1994 and 1994/1995. The crop along each tunnel was divided into five (1991/1992) or six (1992/1993, 1993/1994 and 1994/1995) plots, thereby providing plots grown at different temperatures. Vertical mixing fans between the plot increased the coupling of plants to the air, although these were only installed late in the first experiment. Standard agronomic practices (Table 1) were followed in all 4 years and crops were irrigated to maintain the soil at near field capacity through a network of porous pipe
placed on the soil surface at intervals of 0.32 m. In each year, crops were also grown in ambient field plots near the tunnels. At crop maturity, a 1 m 2 (1991/1992), 0.18 m 2 (1992/1993), 0.19 m 2 (1993/1994) or 0.36 m 2 (1994/ 1995) area was separately harvested from two locations within each plot. Above-ground biomass only was harvested. The number of ears were counted, then separated into straw and seed fractions. The grains were counted and mass recorded. The moisture content of a 5 g sample of grain, and the dry mass of the straw were determined after drying at 80°C for at least 72 h. In summary, these experiments provided a total of 20 (1991/1992) or 24 (1992/1993 to 1994/1995) plots comprising a combination of two [CO2], five or six temperature regimes and two replicates. The effects of mean temperature (between sowing and maturity) and [CO2] on the dependent variables were examined using comparison of regressions in order to determine the significance of differences between the CO2 treatments and whether or not there was any interaction between the effects of [CO2] and temperature.
3. Results 3.1. Growth conditions The mean temperature from sowing to harvest maturity of the ambient plots outside the tunnels ranged from 11.0°C in the late sown 1991/1992 season to
Table 2 Mean temperatures (°C) in ambient plots and in plots at the warm and cool ends of temperature gradient tunnels during different stages of development (SW, sowing; DRS, double ridges formation; MAT, harvest maturity)
Increased Normal Ambient Seasonal ambient mean
SW-DRS DRS-MAT SW-DRS DRS-MAT SW-DRS DRS-MAT
1991/ 1992
1992/1993
1993• 1994
1994• 1995
8.3-7.7 19.5-13.1 7.9-7.8 18.1-15.0 6.5 14.7
6.5-6.3 14.7-11.6 6.6--6.4 14.2-12.4 6.5 14.2
7.0-6. I 13.9-11.0 6.5-5.8 14.2-11.2 5.1 13.2
8.3--6.4 15.4-12.4 8.0-6.4 15.0-12.4 6.1 14.2
I 1.0
9.7
9.2
10.4
Averages of pairs of tunnels maintained at normal (ca. 376 #mol mol-I) or increased (ca. 697/zmol mol-I) CO2 concentration during field experiments with winter wheat in 4 years.
70
9.2°C in the earlier sown and colder 1993/1994 season (Table 2). The temperature gradient in the tunnels varied diurnally and during the season, being typically small in the winter because of low insolation, and the requirement of the air conditioning equipment to deice, and increased markedly from spring onwards. During the whole vegetative development period (from sowing to double ridge formation) the mean daily difference between end plots within the tunnels in each season was <0.5°C except in the last year, when improvements in the system gave larger differences (1.4-1.9°C, Table 2). During the reproductive phase the mean daily difference was typically 3-4°C (Table 2). It should be noted that the effectiveness of the temperature gradient system is larger than suggested by the figures in Table 2 because the temperatures shown are those calculated over a particular phase, not a particular calendar time. As there is a general rising temperature trend during the season, plants developing more slowly tend to 'catch up' in mean temperatures over a particular phase. The 4-year mean [CO2] were 376 and 697 #mol CO2 mol-! in the 'normal' and 'increased' treatments. Mean insolation was similar during reproductive development, with a 4-year mean of 16.6 MJ m -2 per day (Table 3), but varied substantially between years during vegetative development from 2.9 (1993/1994) to 7.4 MJ m -2 per day (1994/1995). A seasonal mean of ca. 65% of this insolation was incident on the crop canopies within the tunnel growing area; although the autumn-spring average was closer to ca. 75%. While there are no spectral changes in the tunnel, except for reduction of UV transmission, the solar radiation inside the tunnels is much more diffuse (see Hadley et al., 1995 for further details of the temperature gradient system).
Table 3 Mean solar radiation flux (MJ m -2 d-j) incident on ambient plots during development between emergence (EM) and double ridges formation (DRS) and between DRS and harvest maturity (MAT) during field experiments with winter wheat in 4 ears
EM-DRS Duration DRS-MAT Duration
1991/1992
1992/1993
1993/1994
1994/1995
5.9 72 16.0 77
3.5 110 15.3 126
2.9 108 16.6 136
7.4 105 18.3 128
Approximate duration of periods (days) are also shown.
320 a
300-= :3 0
•~
280 -
240 -
E
220 -
¢: 0
200
-
180
-
= "1o
o o
260 -
160
8
~
0.012
I
I
9
10
,, .
I
I
I
I
11
12
13
14
.,
c o
15
b ......-'"" "''"'"
0.010
..-" ,.
"10 A • ,c--x .z
.......'"
0.008
"0 O.
""O"
.-'"" "
0..'""" ....'"
0.0O6
0.004
I
10
'
I
12
'
I
14
' "
I
16
'
I
18
'"v
20
Temperature (oc) Fig. 1. Relations between (a) crop duration from sowing to maturity (Yl), and (b) the rate of reproductive development (Y2) and mean temperature (t) during each phase for winter wheat crops cv. Hereward grown at increased (solid symbols) or normal (open symbols) [CO2] during 1991/1992 (i--1), 1992/1993 (A), 1993/1994 (O), and 1994/1995 (O). Results for crops grown in field plots in ambient conditions outside the tunnels are shown for comparison (shaded symbols). The fitted line in (a) is: 1/yl = 3.931 x 10-4t (SE = 1.223 x l0 -6) (re = 0.918, 96 DF) in (b) is: Y2 = 5.492 x 10-4t (SE = 3.00 x l0 -6) (r E = 0.889, 96 DF), with 95% confidence intervals for the prediction (dashed lines).
3.2. Crop development The temperature differences imposed between the warmest and coolest plots within the tunnels altered the durations of crops by as much as 6 weeks over the 4 years. A single relation between the crop duration (days from sowing to maturity) and temperature described all the data for the four seasons;(P < 0.001; Fig. l a), indicating a base temperature of 0°C and a thermal time requirement of 2544°C.d (SE _ 8°C.d). Similarly, the rate of progress (the reciprocal of duration) through the reproductive development phase (double ridges formation to harvest maturity) was positively and linearly related to temperature (P < 0.001; Fig. lb) with a base temperature of 0°C and a
71
1 55 0000
.....
1992/1993 season which had a limited range of temperature (P > 0.1; Fig. 2b). The stimulating effect of increased [CO2] was significant in all 4 years (all P < 0.001). There was a larger effect of increased [CO21 at warmer temperatures in 1991/1992 (P < 0.05; Fig. 2a); no interaction in either 1992/1993 or 1993/1994 (Fig. 2b,c); and much larger increases of biomass with increased [CO2] at cooler than warmer temperatures in 1994/1995 (P < 0.05; Fig. 2d). Therefore, the range in the relative effect of increased [CO2] at different temperatures on biomass varied between years (Table 4).
e
000
50O 0
~
4000
3000
0
~
1992/93
b"
•
4000
~
1500
"
. . . . .
1000
........
1993/94
C
E>, 1 5 0 0 "0
=
~
30oo
2000
3.4. Grain yield
l ooo
o~=
500
1000 0
~
~
i
4000
~
~9a4~5
3000
,
0
*
I
h
1000
,
2000
500
1000 0
"t
1500
d
!
8
! -
10
i
I
I
12
14
16 Temperature
0
! 8
10
i - i 12
14
16
(°C)
Fig. 2. Relations between biomass (left column) and grain dry mass (right column) and the mean temperature from sowing to harvest maturity in 1991/1992 (a,e), 1992/1993 (b,f), 1993/1994 (c,g) and 1994/1995 (d,h) for crops grown at increased (solid symbols) or normal [CO2] (open symbols). All regressions are significant (P < 0.05), and where separate regressions are shown for the two CO2 treatments they are significantly different (P < 0.05).
thermal time requirement of 1821 °C.d (SE + 11 °C.d). There was no significant effect of [CO2] treatment on either of these relations. For both phases of development, the durations of crops grown outside the tunnels were comparable to those grown inside, except in 1991/1992 (Fig. la,b), which is probably because of the poorer coupling of the plants to the air stream in that year prior to the installation of the vertical mixing fans.
3.3. Crop biomass Crop above-ground biomass at maturity declined linearly with increasing seasonal mean temperature (P < 0.001; Fig. 2a,c,d), with the exception of the
In 1991/1992, grain dry mass per unit area declined as a curvilinear function of temperature (P < 0.001; Fig. 2e). Increased [CO2] increased grain yield (P < 0.001) with more effect at warmer temperatures. In all other years, grain yield was a negative linear function of mean temperature (1992/1993 and 1994/1995 P < 0.001, 1993/1994 P < 0.01; Fig. 2fh). In 1992/1993, increased [CO2] increased grain yield (all P < 0.001) with no interaction with temperature (P > 0.05; Fig. 2f). In 1993/1994 and again in 1994/1995, however, this interaction was significant (both P < 0.05); with grain yield less sensitive to temperature at normal [CO2] (Fig. 2g,h), hence, the relative effect of increased [CO2] was greater at cooler than at warmer temperatures, with increases in grain yield of up to 3.0 and 4.5 t ha -~ in the coolest plots, in the respective years. In the last 3 years the grain yield at 10°C varied by only 103 g m -2 in normal [CO2], but by 554 g m -2 at increased [CO2] so that the relative effect of increased [CO2] ranged from 27 to 89% in the three different years (see Table 4). In order to understand the differing patterns of grain yield response to temperature and [CO2] between years the relations between grain yield and biomass or yield components were examined (Table 5). As Fig. 2 indicates, the pattern of grain yield variation with temperature and CO2 was often similar to that of biomass so that biomass changes accounted for a large part of the effect of temperature and [CO2] on grain yield (Table 5). However, harvest index was not constant, showing a large decline with increasing temperature in the first 2 years and showing increases with increased [CO2] in all except the first year,
72
Table 4 Response of above ground biomass and yield to increased CO2 concentration (approximately a doubling) in winter wheat crops grown in temperature gradient tunnels in the field in four seasons Year
1991/1992 1992/1993 1993/1994 1994/1995
Range
Grain yield at 10°C
Biomass (%)
Grain yield (%)
Normal[CO2]
Increased[CO2]
Grainyield (%)
Warming(°C)
6-31 34 8-17 33-17
7-44 72-168 46-7 58-31
b 683 580 623
b 1292 738 994
b 89 27 60
1.0 1.8 2.0 1.2
The range shown is the range in relative effect of increased CO2 between the coolest and warmest plots on biomass and yield. The absolute and relative effect of increased CO2 on yield (g m-2) at 10°C are also showna together with the warming above the ambient seasonal mean which negates the increase of grain yield by increased CO2. a4-year ambient mean temperature. bprediction beyond 1991/1992 temperature range inadvisable. with particularly large effects in the second year (Fig. 3). The yield variation across temperature and CO2 treatment attributable to each of the three primary components (the n u m b e r of ears m -a, number of grains per ear, and individual grain mass) differed substantially between years, and in several cases the increased [CO2] was not a significant factor. This is notable in the contribution of variation in ear number m -2 to variation in grain yield, which was negligible in the second year and very high in the last, but with no effect of increased [CO2] in either case. In 1992/1993, small contributions of variation in ear number m -2 to variation in grain yield were linked to large contributions from numbers of grain per ear, and in 1994/1995 the opposite occurred. In contrast in years 1 and 3, there was a large [CO2] effect on ear number m -2. Individual grain mass was particularly important in determining yield in the warm 1991/1992 season, where total grain mass declined drastically with
increasing temperature (Fig. 2e) and individual grain weight declined from ca. 35 to 15 mg grain -1 over the temperature range (see also Wheeler et al., 1996a;Wheeler et al., 1996b). Similarly in 1994/ 1995 cooler plots had markedly heavier total (Fig. 2h), and individual grain mass, which declined from ca. 33 to 25 mg grain -1 over the 3°C temperature range at normal [CO2] and from 46 to 25 mg grain -I at increased [CO2] (see Batts et al., 1998). The combined effect of grain number per ear and ear number per area resulted in a high correlation of grain yield with number of grains per area in all years.
4. Discussion While some studies have shown effects of increased [CO2] on apical development in wheat and other cereals during particular phases of development (e.g.,
Table 5 Percentage variation accounteda for by linear regression analysis of grain yield (g m-2) on biomass or components of grain yield for all treatments in 4 years Year
Biomass(g m-2)
Ears (m-2)
Grains (ear-~)
Grain mass (mg grain-I)
Grains (m-2)
1991/92 1992/93 1993/94 1994/95
45.1 66.6 (77.3) 49.4 (68.5) 86.1 (87.4)
1.4 (65.6) 6.2 26.6 (69.8) 81.0
79.4 (82.5) 62.4 (71.5) 3.7 (46.4) 0.0 (15.8)
81.9 8.6 (66.0) 1.4 (48.5) 89.5 (91.0)
86.7 80.5 (82.8) 63.4 (78.4) 77.4
Values in parentheses are the correlations when [CO2] is included as a factor in the regression. aPercentage variation accounted for = [(Total MS - Residual MS)/Total MS] × 100. Degrees of freedom for 1991/1992 to 1994/1995 were 38, 42, 46 and 46, respectively. Single figures are presented where correlation values are not significantly different (P > 0.05) from those when [CO2] was included as a factor.
73
Marc and Gifford, 1984; Batts et al., 1996), these are generally small, and were not detectable here over the whole duration from sowing to maturity or over the reproductive period alone (Fig. l a,b). Similar lack of developmental response to increased [CO2] has also been shown in winter wheat by Slafer and Rawson (1994) and Mitchell et al. (1995), (1996). The relations in Fig. 1 can be used to calculate that the effects of a 2 and 4°C warming (in the range of current climate change scenarios for Europe over the next 50100 years) added to the 10°C 4-year ambient mean would result in reductions of 42 and 72 days, respectively, for total duration, and added to the 4-year mean for the reproductive phase (ca. 14°C) would result in reductions of 16 and 29 days, respectively. However, using the AFRCWheat development model, Butterfield and Morison (1992) predicted that a mean warming of 2 and 4°C would reduce the total duration of a winter crop by 20 and 35 days in the southern UK, and the reproductive phase would be reduced by 8 and 20 days, respectively. Part of the discrepancy could be because the simulation used a model parameterised for a different, slower developing cultivar (Hustler) and for a much earlier sowing date, which combined to reduce the mean seasonal temperature in the control simulation to 9°C, and give considerably longer growing seasons. Another reason could be that the seasonal warming in the model was throughout the season, where temperature effects particularly during early development can be affected by photoperiod and vernalisation responses, whereas the characteristic of the temperature gradient tunnels is that the warming is more pronounced in the spring and summer, than in the winter. Therefore, assessing the temperature treatment by the mean season-long value as used in Fig. 1 could be overemphasising the temperature sensitivity. However, until we have more firm information on the likely seasonal pattern of temperature change expressing crop durations as a function of seasonal mean temperature seems the simplest approach. The major objective of the experimental approach developed for these studies was to detect and quantify any interaction between the effects of [CO2] and temperature on grain yield. The negative effect of temperature and the positive effect of [CO2] on grain yield resulted in interactive effects in three out of the four years (Fig. 2e,g,h). The relative effects of increased [CO2] on biomass and grain yield of these crops varied
greatly within and between years (Table 4). For example, there were large differences in the relative effect of increased [CO2] at 10°C (Table 4), indicating that the response of grain yield to [CO2] was influenced by environmental conditions, of which the main variables were probably the temperature and the insolation as the crops were maintained well-watered. While the sowing date in the first year was late, due to technical problems, which may have influenced the results, the other years had similar sowing dates yet showed differences in yield response. For example, the relative effect of increased [CO2] on grain yield (at 10°C) was 89% in 1992/1993 but only 27% in 1993/1994 despite similar mean whole season, vegetative and reproductive phase temperatures (Table 2). However, changes in the exact seasonal time course of temperature and radiation can have a more substantial influence on mean yield and yield stability than changes in the mean conditions (e.g., Nonhebel, 1994). Clearly, each year has a unique seasonal time course of temperature and insolation, and plants in the -
13
40-
1992/93
1991/92
c
3020~X 10-
&
"0
•~- 0
m
!
W
1993•94
~ 60-
b
1994195
d
e-
50-
o•
40-
000° ; ' ~ ; -
-
30208
I
I
t
10
12
14
16
8
I
!
I
10
12
14
16
Temperature (°C)
Fig. 3. Relations between harvest index (grain mass/above ground dry mass) and the mean temperature from sowing to harvest maturity in 1991/1992 (a), 1993/1994 (b) 1992/1993 (c), and 1994/1995 (d) for crops grown at increased (solid symbols) or normal [CO2] (open symbols). All regressions are significant (P < 0.05) except in (b) where the lines indicate the mean values, and where separate regressions are shown for the two CO, treatments they are significantly different (P < 0.05).
74 different temperature plots develop at different rates in these experiments and are therefore subjected to a unique coincidence of weather conditions with ontogenetic stage. There is inevitably a confounding effect in any such experiments using the normal seasonal time course of insolation, such that plants in higher temperatures mature earlier and therefore 'miss' the generally increasing solar radiation (see also Mitchell et al., 1993). Moreover, the coincidence of stress events during sensitive developmental phases, particularly high temperatures following anthesis (Stone and Nicolas, 1995) significantly contributes to seasonal variation in grain yields via influence on the rate and duration of grain growth (Wardlaw and Moncur, 1995; Wheeler et al., 1996b). Therefore, the different magnitudes of response to both increased [CO2] and temperature in each year is not surprising, and it emphasises the difficulty of making yield predictions for future climate and atmospheric conditions, without very detailed scenarios. Seasonal warming of only ca. 1-2°C was sufficient to negate the grain yield stimulation due to increased [CO2] (Table 4). Indeed, in 1993/1994 and 1994/1995 the benefit of doubling [CO2] to grain yield was particularly small at the warmer temperatures as a result of the negative interaction between [CO2] and temperature, so much so that only limited extrapolation of the relations (see Fig. 2g,h) would be required before the increase in grain yield from doubling [CO2] was lost. Thus, in some seasons, and at some temperatures, the response to increased [CO2] was small given the large [CO2] treatment; large compared to scenarios of only approaching twice the pre-industrial [CO2] (viz. 540 ~mol CO2 mo1-1) by the end of next century, (IS92d scenario, IPCC, 1996). The limited warming of only 1-2°C to offset the grain yield from doubled [CO2] is considerably smaller than the 4.5°C estimated to almost negate yield gains in a modelling study by Goudriaan and Unsworth (1990), indicating a greater sensitivity of yield to temperature than characterised by the model. However, the same comments apply here as in the above discussion over predictions of changes in duration that as most warming in these experiments was in the spring and summer expressing the warming expressed on a whole season basis may be overestimating the sensitivity. If Figs. 2 and 3 were shown using the mean temperature during grain filling (from anthesis to maturity) the data would have a span
of ca. 6°C for years 1 and 2 and ca. 4°C in years 3 and 4, rather than the spans shown of between 2 and 3°C (see also Wheeler et al., 1996a,Wheeler et al., 1996b). Using this temperature basis would increase these experimentally derived estimates of the temperature required to negate the CO2 effect to 2-3°C, which still represents a substantially higher temperature sensitivity than the model used by Goudriaan and Unsworth (1990). While such comparisons between the relative effects of mean seasonal temperature and [CO2] increase are useful, the detailed pattern of temperature and other climate variables during the season is important, as shown here and in modelling work elsewhere (Grashoff et al., 1995; Nonhebel, 1994). While above ground biomass and grain yield showed a broadly similar pattern of response to temperature and [CO2] in all years (Fig. 2), there were substantial changes of harvest index with temperature and with increased [CO2] in most cases (Fig. 3). The marked declines in harvest index that occurred in the first 2 years were when yields were affected by high temperature episodes during grain filling (Wheeler et al., 1996a,b). Analyses of which yield components accounted for the observed variation of grain yield between treatments, showed that for all years the number of grain per unit area was a major determinant, and this arose either through variation in ear number m-2 or grain number per ear (Table 5). Grain size and its response to temperature and [CO2] differed between years, and in normal [CO2] was largest in years 2 and 3 (ca. 45 mg at cooler temperatures), compared to ca. 32 mg in equivalent temperatures in years 1 and 4 (see Batts et al., 1998; Wheeler et al., 1996a) clearly indicating different source-sink relations between years. The role of the ear number in determining yield varied between years, and the relative effect of increased [CO2] on ear number at maturity varied (9%, 0%, 12% and 20% increases, in years 1-4, respectively). Many other reports have found substantial stimulation of tillering in wheat with increased [CO2] (e.g., Gifford, 1977; Mitchell et al., 1993, 1995). The propensity of winter wheat to tiller during the period around double ridge to terminal spikelet formation is linearly related to main stem dry mass at the same stage (Batts and Baylis, 1985; Batts et al., 1996), which depends on the cultivar (Batts et al., 1998) and on the environmental conditions from sowing. For example, Mitchell et al.
75
(1996) found that shading a wheat crop near terminal spikelet formation had a substantial effect on reducing the number of ears per unit area, compared with two alternative shading periods between terminal spikelet and the start of grain-fill. The lack of [CO2] promotion of tillering in 1992/1993 (Batts et al., 1996) may have been caused by the low solar radiation flux during the vegetative development period in this year (Table 3), and this effect may have been exacerbated by the reduction in solar radiation within the temperature gradient tunnels. However, there was still a pronounced increase in yield in response to increased [CO2] in this year because of more grain per ear, and it should be noted that the grain filling rate of the cooler treatments at normal [CO2] was the same as that outside (Wheeler et al., 1996b) indicating that the radiation regime in the tunnels was not too low. Clearly, wheat plants have substantial flexibility in the partitioning of the increased photosynthate supply in increased [CO2], and therefore increased tillering may not be a prerequisite for effects on grain yield. Unravelling these complex effects is difficult experimentally in the field. There is a need to explore further the influence of climatic variation within and between years on crop performance and yield, with both experiments and modelling studies (e.g., Semenov and Porter, 1995; Grashoff et al., 1995). Detailed field datasets, from the same location, over several unique environmental patterns such as summarised here are valuable in calibrating and validating crop growth models that are to be used to predict the impacts of climate change on crop production. The variation in response to increased [CO2] we observed here has important implications to any attempt to provide climate change impact predictions, either for changes in the mean climate, or for the equally important possible changes in climate variability (e.g., Kattenberg et al., 1996).
Acknowledgements We thank X. Liu, Pam Nkemka, Marelia Puche, G. Ruffles and M. Cantwell for technical assistance. This research was funded by the UK AFRC Wheat-CO2 Minitopic (GRB, TRW) and the UK BBSRC BAGEC programmes (GRB), The Royal Society and The University of Reading Research Endowment
Fund. This