Soil Responses to Climate Change

NATO ASI Series Advanced Science Institutes Series A series presenting the results of activities sponsored by the NA TO Science Committee, which aims at the dissemination of advanced scientific and technological knowledge, with a view to strengthening links between scientific communities. The Series is published by an international board of publishers in conjunction with the NATO Scientific Affairs Division A Life Sciences B Physics

Plenum Publishing Corporation London and New York

C Mathematical and Physical Sciences D Behavioural and Social Sciences E Applied Sciences

Kluwer Academic Publishers Dordrecht, Boston and London

F Computer and Systems Sciences G Ecological Sciences H Cell Biology I Global Environmental Change

Springer-Verlag Berlin Heidelberg New York London Paris Tokyo Hong Kong Barcelona Budapest

NATo-pea DATABASE The electronic index to the NATO ASI Series provides full bibliographical references (with keywords and/or abstracts) to more than 30000 contributions from international scientists published in all sections of the NATO ASI Series. Access to the NATO-PCO DATABASE compiled by the NATO Publication Coordination Office is possible in two ways: - via online FILE 128 (NATO-PCO DATABASE) hosted by ESRIN, Via Galileo Galilei, 1-00044 Frascati, Italy. - via CD-ROM "NATO Science & Technology Disk" with user-friendly retrieval software in English, French and German (© WTV GmbH and DATAWARE Technologies Inc. 1992). The CD-ROM can be ordered through any member of the Board of Publishers or through NATO-PCO, Overijse, Belgium.

Series I: Global Environmental Change, Vol. 23

Soil Responses to Climate Change

Edited by

Mark D. A. Rounsevell Peter J. Loveland Soil Survey and Land Research Centre Cranfield University Silsoe, Bedfordshire MK45 4DT, UK

Springer -Verlag Berlin Heidelberg New York London Paris Tokyo Hong Kong Barcelona Budapest Published in cooperation with NATO Scientific Affairs Division

Proceedings of the NATO Advanced Research Workshop on soil responses to climate change, held at Silsoe, Bedfordshire, UK, September 20-24, 1993

ISBN-13:978-3-642-79220-5 e-ISBN-13: 978-3-642-79218-2 DO I: 10.1007/978-3-642-79218-2

CIP data applied for This work is subject to copyright. All rights are reserved. whether the whole or part of the material is concerned. specifically the rights of translation. reprinting. reuse of illustrations. recitation. broadcasting. reproduction on microfilm or in any other way. and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9. 1965. in its current version. and permission for use must always be obtained from Springer-Verlag. Violations are liable for prosecution under the German Copyright Law. © Springer-Verlag Berlin Heidelberg 1994

Soft cover reprint of the hardcover 1st edititon 1994

Typesetting: Camera ready by authors SPIN 10100886 31/3130 - 543210 - Printed on acid-free paper

PREFACE

This volume has resulted from the papers and discussions of a NATO Advanced Research Workshop held at the Soil Survey and Land Research Centre, Cranfield University, 20-24 September 1993. The principal aim of the workshop was to address the central role soils will play in mediating both the overall responses of ecosystems to predicted climate change, and the rates at which these responses will occur. This aim arose from the clear need to highlight the importance of soils and soil processes in environmental change research and their contribution to the proper understanding of ecosystem behaviour. In doing so, the Workshop attempted to emphasise the dynamic nature of soils and to encourage treatment of soils as ecosystems in their own right. This collection of papers is intended to cover the wide range of scientific disciplines covered by Soil Science including the chemistry, physics and biology of soils, and further emphasises the inter-disciplinary nature of ecosystem studies. A feature of the Workshop was the high quality of the poster exhibitions, and so we felt it appropriate to publish extended summaries of these presentations. Some of these summaries represent research at an early stage of development, and others outline techniques or methodologies which are appropriate to the study of soils and global environmental change. We would like to extend our gratitude to NATO for sponsoring the Workshop and the publication of this volume, Dr. Jean-Paul Legros and Prof. Diego de la Rosa for their

contributions to the Workshop Organising Committee, Mrs. Elaine Cousins and Dr. Thomas Mayr for assistance during the organisation of the Workshop, and to Mrs. Rosemary Foskett and Me Steve Lawson for help in compiling the manuscript for publication.

Mark D.A. Rounsevell and Peter J. Loveland Soil Survey and Land Research Centre, Cranfield University, Silsoe, Beds. MK45 4DT,

UK.

WORKSHOP PARTICIPANTS

From left to right: C.S. Kosmas, M.DA Rounsevell, P.J. Loveland, S.P. Theocharopolous, S. Frolking, C. Ramos, D.S. Powlson, D. Bachelet, E. Fernandez, T.R. Mayr, M.J. Mausbach, J. Crompvoets, P. Ineson, P. Botterweg, J.J. Lee, J.R. Kramer, M.R. Carter, D. Sauerbeck, G. Wagner, F. Robinet, D. Favis-Mortlock, G. Varallyay, F. Nachtergaele, D. Magaldi, C. Simota, D.G. Zvyagintsev, L. Trombino, U. Diny, J-P. Legros, M. Voltz, A.C. Armstrong, D. de la Rosa, D.L. Phillips, E.A. Davidson, P.B. Tinker, R.W. Arnold, N. Santos, C. Nys, P. Lorenzoni, M. da Conceiyao Gonyalves, P. Loiseau, J.U. Smith.

CONTENTS

Preface Workshop participants

Plenary papers Soils and global change: an overview

3

P.B. Tinker and J.S.I. Ingram Relevance of understanding landscape evolution in relation to

13

climate-induced soil behaviour R.w. Arnold Climate change, desertification and the Mediterranean region

25

C.S. Kosmas and N.G. Danalatos Climate change, soil salinity and alkalinity

39

G. Varallyay Spatial modelling approaches to evaluate the effects of climate

55

change on future crop potential and land management T.R. Mayr, M.D.A. Rounsevell and D. de la Rosa Crop models: principles and adaptations to the problem of

71

climate change J-P. Legros, C. Baldy, N. Fromin and D. Bellivier The effects of climate change on irrigated soils: water resources and solute leaching C. Ramos, A.L. Lid6n and A. Rodrigo

99

VIII

Modelling the effects of climate change on the hydrology and

113

water quality of structured soils A.C. Armstrong, A.M. Matthews, A.M. Portwood, T.M. Addiscott and P.B. Leeds-Harrison The potential impact of global environmental change on nitrogen

137

dynamics in arable systems N.J. Bradbury and D.S. Powlson Climate change and soil microbial processes: secondary effects

155

are hypothesised from better known interacting primary effects E.A. Davidson Old sediment carbon in global budgets

169

J.R. Kramer

Poster papers Global climate change and the necessity to 'scale-down

187

P. Botterweg The agricultural management effects on carbon sequestration in

193

eastern Canada M.R. Carter, D.A. Angers, E.G. Gregorich, R.G. Donald, C.M. Monreal and R.P. Voroney Demonstration of the Rothamsted carbon model

197

K. Coleman and D.S. Jenkinson An expert evaluation system to assess agricultural soil erosion

199

vulnerability J. Crompvoets, F. Mayol and D. de la Rosa MicroLEIS 3.2: a set of computer programs, statistical models and expert systems for land evaluation D. de la Rosa

205

IX

Modelling soil erosion on UK agricultural land under a changed

211

climate D. Favis-Mortlock The development of pedotransfer functions for the hydraulic

217

properties of Portuguese soils M. da Conceiyao Gonyalves The use of EPIC in a statistical framework for regional analysis

221

of soil responses to climate and management J.J. Lee, D.L. Phillips and

v.w. Benson

Effect of climatic changes (C0 2 , temperature) on grassland ecosystems: first five months' experimental results

223

P. Loisseau, J-F. Soussana and E. Casella Significance of two soil components of the pedosphere as

229

carbon sinks M.J. Mausbach and L.D. Spivey The role of site characteristics, species and soil horizon on the

231

evaluation of carbon contents of forest soils C. Nys, S. Didier and J.L. Hubert Statistical study of soil respiration: calculation of present day

237

rates and anticipation for a double CO 2 world F. Robinet Demonstration of SUNDIAL: simulation of nitrogen dynamics in

243

arable land J.U. Smith and N.J. Bradbury Seasonal climatic variability and upward nitrate movement in Greek soils S.P. Theocharopoulos, M. Karayianni-Christou, P. Gatzogianni and S. Aggelides

245

x Standard operation procedures for sampling and sample treatment of soils for environmental specimen banking G. Wagner and J. Sprengart

249

Summary paper Soils and climate change - where next? J-P. Legros, P.J. Loveland and M.D.A. Rounsevell

257

Bibliography

267

Index

303

PLENARY PAPERS

SOILS AND GLOBAL CHANGE - AN OVERVIEW

P.B. Tinker and J.S.I. Ingram GCTE Office, Department of Plant Sciences, Oxford University, South Parks Road, Oxford OX1 3RB,

u.K.

BACKGROUND There have been a number of meetings and publications on soils and global change in the last few years (Anderson, 1992; Scharpenseel et aI., 1990; Bouwman, 1990; Arnold et al., 1990). In continuing further with this work, we ought to bear in mind that the study of soil has been changing rapidly over the last decade or so. Soil science used to be seen almost entirely as a support for agriculture and forestry, and the justification for its study was the increase in productivity which it could bring. Recently this focus has widened enormously.

Soil science is now a major

component of any environmental science course. Soil biota are an important part of world biodiversity, and soil has a critical part to play in several essential elemental cycles.

Soil pollution is as important as, and often far more persistent than,

atmospheric and aquatic pollution (Eijsackers and Hamer, 1993). When we consider the impact of global change on soils, we do so from a far broader viewpoint than we would have done only a couple of decades ago. However, despite this massive interest in newer fields, we must not forget the agricultural imperative. The main economic purpose for which soil is used is still agriculture, and amid the potential catastrophes of global change, famine must surely rank as one of the most serious. The problems of global change and soils have covered a wide range of changes and processes, because the effects of human activity are becoming pervasive on a global scale. By the word "global", we mean effects that are obviously transmitted around the world, such as changes in the methane concentration of the atmosphere. In relation to soil we normally mean something rather different: local effects that occur so frequently, in so many different ecosystems, that they can be considered to have global significance. Soil erosion is one of the most obvious examples. It is NATO ASI Series. Vol. I 23 Soil Responses to Climate Change Edited by M. D. A. Rounsevell and P. J. Loveland © Springer-Verlag Berlin Heidelberg 1994

4 Tinker and Ingram

necessary to stress this point, because so many researchers in the atmospheric and marine fields tend to think in terms of the first type of effects only. Location-specific factors are usually the more difficult to investigate, because of the difficulty of obtaining results of wide application.

GLOBAL CHANGE DRIVERS The International Geosphere-Biosphere Programme (IGBP) (1990) has focused its work on three main driving variables.

These are changes in atmospheric

composition, climate and land use. The last of these three is wide, complex and brings in many social and economic issues, and this meeting will not deal explicitly with it.

However, we must all bear in mind that these complex socioeconomic

changes (Sanchez et al., 1990) are occurring all the time, and will interact strongly with changes in the first two, the focus of this paper. This is particularly so because we sometimes ignore the impact, adaptation and response aspects of climate change; these will not be of a wholly technical nature, as is sometimes implied. Even when the response needed is technical, such as a change of farming system, rotation or cultivar, there are important socioeconomic questions; advice and skills must be made available, and there will almost certainly be reduction in yield during the change-over, as new techniques have to be learnt. If the climatic changes are more extreme, or the agricultural system is less developed, there may be insufficient flexibility to cope with the new conditions, and the agricultural system may break down. This is in effect what happened during the great drought of the 1930's in North America, when a large fraction of the population simply left the land. It is essential to remember that effects on soil will be part of this complex response; the most important effect on a soil may simply be that it is no longer farmed, or that its land use is switched from arable agriculture to grassland. It is essential to emphasise that both soils and vegetation must be addressed, for two reasons. Firstly, vegetation will be very strongly influenced by climate change (Woodward, 1992), and to a currently unknown extent by elevated atmospheric C02. Secondly, climate and vegetation are two closely linked soil forming factors, and in this context they cannot be separated. Climate change impacts on soil will thus be both through changes in vegetation and through direct effects (e.g. changed rainfall patterns leading to changed erosion). On the other hand, elevated C02 is only likely to have a significant impact through its effect on vegetation, because the soil atmosphere C02 concentration is always very high relative to that in the atmosphere. Thus, elevated C02 will induce changes in below-ground allocation of

5 Soils and global change

photosynthate, and change the quantity and composition of root exudates.

The

impacts of changes in climate and atmospheric composition on vegetation cover will thus have rapid and important effects on soil biology, organic matter dynamics, nutrient ion relations and water relations. If climate change leads to a change into arable agriculture, tillage will have immediate and drastic effects on surface morphology, soil carbon contents and structure, which will then alter the stability of the soil to erosion.

THE CLIMATIC SCENARIO The future processes controlling change in atmospheric composition are reasonably well defined - the main uncertainties concern the effectiveness of the political response to the threat of climate change, and whether the emission of greenhouse gases will start to taper off because of this. The most recent and authoritative study (Houghton et a/., 1992) has on the whole supported the conclusions of the original IPCC Report, except that methane is now increasing more slowly than formerly believed.

However, it is not able to make any conclusion on the likelihood of

successful political action in this sphere, and the possible future rates of emission still cover a very wide range. There are still major uncertainties in both sources and sinks, and in the relation between emission rates and atmospheric concentrations. Similarly, the predictions of climatic effect are not greatly altered, but neither are they much more certain. We are still expecting a global temperature increase of 1.5° to 4.5°C and it has not been possible to predict local variations, weather variability or precipitation with much greater accuracy. Improvements in modelling of various processes have been incorporated, including terrestrial processes, but the broad picture has remained much as before - a tribute to the quality of the work done by IPCC in 1990. We therefore still do not know with any certainty what local climate will be like, which is particularly frustrating for soil scientists because of the essentially location-specific nature of many soil impacts. The changes in rainfall are also still very uncertain. Regional scenarios for climate are now being produced, but this is still on a rather large scale for our purposes (Goodess and Palutikof, 1992).

SOILS EQUILIBRIA AND TRANSIENT STATES We tend to assume at present that most of our soil-vegetation systems are in approximate equilibrium, as long as they have not been perturbed by human action, because the world's climate has been moderately stable for a few thousand years. There is now considerable evidence for periods of very rapid change before and

6 Tinker and Ingram

during the Ice Ages, and the more detail that is obtained about past climates from ice cores, the more variability is discovered (White, 1993). Most previous changes in climate have been very slow compared to what is predicted by the Global Circulation Models for the next century or so (Houghton et al., 1992), but the knowledge that there have been very rapid changes in the past (Dansgaard et al., 1993) tends to support the possibility that our perturbation of the atmosphere may cause a similarly rapid climatic 'flicker' now. The soils and climates are therefore likely to be out of equilibrium over the whole period of climate change, and the same also applies to the vegetation (Huntley, 1992). This is especially true since climate change is not expected to be a single, one-off occurrence, but a progressive, lengthy, and irregular development. Different soil properties have very different characteristic rates of change under external forces (Yaalon, 1971; Arnold et al., 1990). Indeed many soil properties are not in equilibrium with the environment over large parts of the worlds surface at present, but the rate of change for these properties is sufficiently slow to ignore. It is the relatively unstable properties that will be affected by climate change, and these are the subject of our meeting. If this rapid change occurs, there may well be, over the next century, combinations of climate, vegetation and soils' properties that are quite new to our experience. The information from palaeobotany suggests that vegetation associations or ecosystems do not move coherently from one place to another as the climate changes. Species appear to move independently, mingling with different species at different times, so that a variety of associations may be produced.

MAIN CHANGES The most important processes appear likely to be the following: 1. Change in soil type

The overall relationship of soil type with temperature, rainfall and evapotranspiration is well known. Only fugitive properties may be altered in the short term as climatic factors alter, such as salt content. However, if the soil classification depends upon this - e.g. a solonchak - then the soil type will alter. There are also some situations where soil types are known to be in an unstable equilibrium, and may be changed sharply by a change in climatic factor and/or vegetation (Sombroek, 1990). On the basis of Global Circulation Models' predictions it is expected that temperature

7

Soils and global change

changes will be largest at high latitudes, and the loss of organic matter in tundra soils and peats could be rapid, and could alter soils quite markedly. There are also known cases where a difference in the tree or shrub species that is established leads to the fairly rapid development of different soil types (Miles, 1985; Moffat and Boswell, 1990); the predicted changes in vegetation may therefore have some quite rapid consequences. 2. Erosion

While major changes in rainfall and wind could cause great changes in erosion (Wicherek, 1993), most serious erosion is currently a consequence of interference with the vegetation cover, either artificially or through drought or fire. Under climate change it is predicted that rainfall will increase on average, and extreme events may increase more than proportionately.

The converse does not necessarily hold,

because Valentin (personal communication) found in the Sahel that the number of extreme rainfall events did not decrease during the drought, and that their impact on erosion was even more serious than before. In brief, there are likely to be large areas where changes in rainfall and in vegetation leads to greatly increased erosion danger.

A similar argument can be made for wind erosion in semi-arid areas,

especially when they are under grazing pressure (Dregne, 1990). 3. Element cycles and soil organic matter (SOM)

A great deal of attention has focused on the effect of climate change on the C, Nand P fluxes and pools in soils (Tinker and Ineson, 1990; Anderson, 1992).

The

pervasive and manifold effects of organic matter on soil behaviour and fertility justify this importance. Basically, two approaches have been taken to the problem of predicting organic matter levels. Firstly, one may proceed by analogy with existing soils, and on the basis of current data. The best known examples are in the work of Zinke et al. (1984) in defining the amount of carbon in soils in relation to climate. In more detailed work, Jansson and Berg (1985) determined an empirical relation between the rate of decomposition of litter and evapotranspiration across a European transect. Also, the net primary productivity is reasonably well related to actual evapotranspiration (Lieth, 1975). However, these broad generalisations may cover wide divergences in particular situations (Anderson, 1992). The second general approach is to model carbon or nitrogen dynamics from a knowledge of the processes in soils, using empirical relationships where necessary. These models, such as CENTURY (Parton et al., 1987), are now being linked with

8 Tinker and Ingram

models of vegetation to include most of the changes in soil organic matter that would be driven by a changed climate. The effects of increased C02 in the atmosphere are very important here. There are still a number of facts to be confirmed, but the results are expected to show more rapid vegetation growth, improved water use efficiency, more resistant litter of lower N content, changes in the root-shoot ratio and the root exudation rate. These effects will vary with environmental condition and with plant species. There are still uncertainties in the basic processes of carbon input to soils, that will need to be clarified. quantify.

Thus, the input of carbon below ground is very difficult to

C02 may be produced either from root respiration or from microbial

metabolism of exuded compounds or sloughed-off residues, and it is extremely difficult to distinguish between these sources. Only the latter process will contribute to soil organic matter. A considerable amount of work both in growth chambers and in the field (Whipps, 1990) suggests that whereas seedlings tend to transfer around

50% of fixed carbon below-ground, this can rise to around 70% for adult trees. Anderson (1992) has suggested that because of the stabilisation mechanisms for SOM in soil, organic matter injected below ground may be converted to SOM with particularly high efficiency. The Free Air C02 experiments that are being set up should start to give more reliable data on total carbon export into the soil from roots, because the C02 supplied to the plants is normally from fossil carbon, and hence carries an isotopic label. Our understanding of the amounts and the behaviour of carbon that is injected below ground level, either from or as part of root systems, is therefore still incomplete.

It is difficult to see how soil carbon dynamics can be

modelled accurately so long as this remains true.

4. Greenhouse Gases It is now well established that soils and vegetation have in the past been the major sources of additional C02 in the atmosphere, which drove the increase in the atmosphere until the early part of this century. It is still a major source, even if it is now much less than the supply from fossil fuels.

There are many sources of

methane, but again those associated with the soil are very important, though difficult to quantify.

Perhaps even more important is the sink function of the soil for

methane. The main source of the nitrous oxide is certainly in the soil, and emissions may be increased by the use of nitrogen fertiliser. These are difficult research areas

9 Soils and global change

because the fluxes often depend upon the precise level of soil conditions such as aeration, and the measurement of fluxes over large areas is very difficult. AGRICULTURAL IMPLICATIONS

The practical implications of these issues are now becoming extremely important (Parry et al., 1988; Scharpenseel et al., 1990). However, much work tends to focus on the circumstances "after climate change". In practical terms climate change is likely to be an irregular and continuing process, so that all states are transients. Farmers and agricultural researchers are therefore likely to be faced with a continuing series of judgements on probability and risk. Their ability to cope with changing conditions, and with the risks, varies widely in different societies, and is probably inadequate in most. The impact of the droughts in central USA in the 1930's, and in the Sahel in the 1970's, indicates the problems that can arise. The best contribution that scientists can make to coping with these problems is to ensure that we have dependable basic information about processes, and robust models that can guide applied research programmes and advisory support. The management of soil organic matter and of erosion hazards will be of particular importance. THE WORK OF THE INTERNATIONAL GEOSPHERE·BIOSPHERE PROGRAMME (IGBP)

One of the IGBP Core Projects is Global Change and Terrestrial Ecosystems (GCTE). This has 4 Foci, and one of these (Focus 3) deals with agriculture, forestry and soils (Steffen et al., 1992). The outline of the soils research programme is in Table 1. The soils' programmes are being designed in detail at workshops planned for 1994; one in Nairobi in conjunction with the International Society of Soil Science (ISSS) co-sponsored Soil Organic Matter conference is particularly important. This will develop a network for the comparison and development of organic matter models. This is one of the basic modes of working, the best example being the GCTE Wheat Network. This now contains 23 groups, with 14 models (or variants) and 33 datasets from a wide range of locations around the world. Following a great deal of hard preparatory work, the models will be tested on common datasets, from well distributed points in climate space, the aim being to identify a model, or to develop one, that is reliable under climate change conditions. This basic concept is intended to be applied to other crops and processes.

10

Tinker and Ingram

It is very important that an initiative is now in progress to develop an integrated World Soils' Database, driven by IGBP's Data and Information System. This was started because IGBP-DIS discovered that the highest priority for global modellers was a unified soil database; it is encouraging that the importance of soil has been recognised in this way, and gives an example of how the new subjects into which soil science is spreading can produce developments that formerly were seen as desirable but unattainable. Table 1 The Structure of the GCTE Focus 3 Soils Activity (Activity 3). Task 3.3.1 Global Change Impact on Soil Organic Matter Objective: To determine the impacts of global change, as expressed at the plant physiological, vegetation and ecosystem levels, on soil organic matter dynamics.

Task 3.3.2 Soil Degradation under Global Change Objective 1: To develop the capability to predict soil degradation by water erosion caused by interactive changes in land use and climate. Objective 2: To develop the capability to predict soil degradation by wind erosion, and subsequent deposition, caused by interactive changes in land use and climate.

Task 3.3.3 Greenhouse Gas Emissions from Agricultural Soils Task under revision.

SOILS AND POLLUTION Whereas at one time soil had little importance for anyone other than agriculturalists, foresters and ecologists, it is now much more prominent in pollution studies. There is a particular interest in how pollutants are held within soil, and whether there are any conditions in which they would be released. All heavy metal pollutants are strongly bound by soil organic matter, and the pollutant equilibria in the soil will be markedly altered if soil organic matter content is decreased by temperature or

11 Soils and global change

rainfall change (Eijsackers and Hamers, 1993).

There is consequently some

concern at the long-term implications of climate change for heavy metals stored in the soil, especially those with low levels of organic matter. CONCLUSION

The most difficult aspect of these changes is the complexity of the soil/vegetation system, and the variety of the changes that will flow from the interactive drivers of global change. Many of the models that are presently being developed to predict what will happen under global change focus on one or a small number of issues. There is little point in applying soil organic matter models unless the likely vegetation changes are known, and there is little point in predicting this until the reaction of the agricultural and forestry industries is clear.

These issues become steadily more

complex as the social and economic dimension is included, as the Human Dimensions Programme is now doing.

We can however be certain that the

behaviour and properties of soils are now seen as fundamentally important in this most complex situation, and this is focusing attention on soils in a genuinely new way.

RELEVANCE OF UNDERSTANDING LANDSCAPE EVOLUTION IN RELATION TO CLIMATE-INDUCED SOIL BEHAVIOUR

R.W.Arnold

Director of the Soil Survey, Soil Conservation Service, U.S. Department of Agriculture, Washington, D.C., U.S.A.

ABSTRACT I want to convey to you the concept of doing experiments with soils and not just doing them to soils. I would like to reinforce the idea that soils as natural bodies of the pedosphere only have relevance within their own environmental setting. This setting is larger than an 'on-site' because one has to see a bigger picture to obtain a proper, or at least a healthy, perspective of the complexity of soils in their landscapes. Trends of observations in time and space must exceed the inherent noise or background variability to be useful.

I therefore, want to emphasise the

value of local variability.

WHAT NEEDS TO BE DONE? Data collection

Almost everything about a landscape depends on making observations. They need to be careful, patient, objective ones. Samples are to be taken and measurements made to answer questions related to hypotheses that you or others have proposed. Even baseline characterisation should satisfy part of a specific hypothesis. Remember that we seldom start with 'no knowledge' about an object. Experiments with soils vary tremendously. Imagine the clearing of the rain forest in Indonesia. Once cleared and the logs burned or hauled away, the surface is very uneven. What will be the impacts of removing varying thicknesses of the surface soil? One can imagine that there will be major impacts on inherent productivity. It is a question needing an answer for today, and for tomorrow. NATO ASI Series. Vol. I23 Soil Responses to Climate Change Edited by M. D. A. Roonsevell and P. J. Loveland © SpringerNerlag Berlin Heidelberg 1994

14

Arnold

Many field experiments can be linked with existing ones at experimental stations to take advantage of accumulated databases and known relationships. The synergism of building new answers by utilising several sources of available information is just good common sense. Biogeochemical processes currently receive more attention in research than the history of a landscape, and rightfully so, because shOrter term questions and answers relate to current processes rather than to the results of millennia. Decadal processes are also important in studying global climate change (Arnold et al., 1990). They include soil organic matter, nutrients, contaminants, and even some features used in soil classification schemes.

Representative sites An important consideration is the selection of representative sites that are repeatable in larger landscapes. When one stands and looks out over the broad landscape there is so much variability that one wonders how to start. Start with geomorphic relationships within landscapes - they are wonderful stratifiers of variability. Geomorphology and pedology are strongly linked especially at scales of about 1:20000 (Arnold and Schargel, 1978). Assume that a number of soil profiles have been sampled perpendicular to the slope and particle. sizes determined at several depths in each profile. A plot of the geometric mean of the particle size distribution against the distance from the summit position commonly shows definite patterns of change related to hill slope position (Daugherty et al., 1975). This is very common on many slopes because the surficial materials move faster downslope due to gravity than the underlying subsoil and substrata.

Another interesting way to illustrate hill slope sorting is to plot the

average textures as isograms, which is a stratigraphic correlation technique. If you are not aware of such sorting relationships on hill slopes, some of your interpretations of other kinds of research may be misleading.

Acceptable variations Consider how you can develop criteria for variations that are acceptable for your purposes. That is, what are you willing to accept as a meaningful difference, or as

15 Climate-induced soil behaviour

being so similar that they are the same. With careful measurements and with comparative analyses between sites it is thought that the effects of local variability can be minimised. Erosion plots in Sumatra on a recently cleared jungle site have so much local variability that meaningful results may be very difficult. In Hawaii, boulders of weathering basalt in the subsoil may influence the results of leaching potential studies. In southem Ontario, Canada there are tongues of organic enriched, acidic areas, within a matrix of calcareous sands, which cannot be detected nor anticipated from the surface. Extrapolating results It is important to learn about those properties and features that are distributed throughout a region so that you are able to extrapolate results to other areas of interest.

The examples cited above are local in nature and do not enter into

extrapolations except by accident or lack of knowledge of their occurrence.

PROBLEMS NEEOING ATTENTION For purposes of discussion I assume that field research is undertaken to be able to determine soil responses to climate change. understand a soil site in its landscape.

Overall the basic problem is to

Such understanding is linked with

geomorphology, stratigraphy, and hydrology (Daniels and Hammer, 1992). Geomorphology Geomorphology begins with surfaces such as the depositional surfaces of river terraces or the surface of a mantle of volcanic ash. There are erosional surfaces and associated depositional surfaces which occur side by side on most slopes. They are easier to see or visualise in cultivated cropland. On such slopes there is an intermingling of erosional-depositional surfaces on a microscale. It is very important if you are studying small surface features changes and fail to recognise these integral associated phenomena. An erosional surface is one on which removal is dominant over additions; that is, erosion is greater than deposition. It would be rare to have a surface that is 100% erosional or 100% depositional. Reality is usually somewhere in between.

16 Arnold

Surfaces are only skin deep in the sense that they are two-dimensional. They are recognised, however, by the nature of the sediments that underlie them. In some steep sloping areas of Indonesia the indigenous people have placed rice terraces on the former depositional surface areas and left the erosional upland slopes under trees or grasses. Thus, it does have practical significance to recognise erosional and depositional geomorphic surfaces. Geomorphology and stratigraphy go hand in hand in most landscapes. In upstate New York we studied segments of a river valley (Scully and Arnold, 1981). The region had been glaciated and water from the melting glaciers deposited high level outwash material which was later dissected by streams that were underfit for the existing valley. The river meandered back and forth across the valley, cutting out some of the older alluvium and depositing accretion materials behind. There were periods of non-deposition when surface soil organic matter accumulated only to be buried by additional accretion sediments. These events could be interpreted from plots of texture variations throughout the area and by the organic matter profiles coupled with radiocarbon dating. Such unravelling of geomorphic events leads to an improved understanding of dynamics of landscape evolution and their association with changing environmental conditions, especially the sensitivities to changes of climate and their effect on vegetation.

Stratigraphy Stratigraphy is necessary to unravel a stream's history as mentioned above. Uplands are complex also. In southern Iowa there is Wisconsin-age loess overlying Kansan-age glacial till (Ruhe et al., 1967). The mantle of loess is truncated by the current erosion surface. Glacial till was deposited, soil formation occurred, erosion stripped off most of the soil, loess was deposited, soil formation occurred, erosion has stripped some of these soils and removed part of the loess on the hill slopes, soil formation is currently taking place. In a gently sloping upland in Ecuador the landscape looks uniform enough to layout some field experiments.

But perhaps a look beneath the surface would be

instructive. Two black layers of volcanic ash are separated by a white tuff layer. Remember that landscape surfaces often hide important aspects of underlying stratigraphy that are the keys to understanding.

17 Climate-induced soil behaviour

Two-storied materials in landscapes are common; more common than most people would believe (Arnold, 1968).

In the Venezuelan tropics there are examples of

ironstone gravels resting on a truncated layer of hardened plinthite. What appears at first glance to be a simple roadcut in southern Iowa has a thin band or layer of pebbles that separate till from the overlying loess-like materials. This is a classic stone line of an erosional surface which has been covered with sediments derived from upslope (Ruhe, 1959). The surface is a pediment and the overlying materials are pedisediments. This is a world-wide phenomenon once your eyes and minds are alert to the features related to multiple sediments in landscapes. Stone lines that mark lithologic discontinuities are global. When the erosional surfaces and the overlying sediments are not marked by pebbles or stones it is helpful to plot depth distributions of soil properties such as structure, consistence, and colour.

In many profiles depth functions of a sand

fraction index and bulk density can be interpreted as markers of important spatialtemporal boundaries (Wang and Arnold, 1973). Soil hydrology

Water in landscapes has time and space implications. Two or more soil moisture regimes occur in many landscapes.

In the semi-arid regions of southern

Kazakhistan there are more humid soils in the depressions and drainage-ways than on the associated hill slopes and summits. To understand biogeochemical cycles in this type of landscape, the moisture regimes must be integrated. Hydrology in a landscape deals with the movement of both surface and subsurface water in and through soils as well as their effects on soil processes and properties. In hummocky gilgai of Vertisols there is an interesting micro-hydrology imposed on a larger regional pattern of water recharge and discharge. The convoluted horizons observed in many Vertisols attest to complicated internal hydrology and associated processes (Bouma and Loveday, 1988). A spatial pattern of organic carbon in a Vertisol profile suggests that the processes of carbon sequestration are not the same in all soils. Near Rossbaden, Austria a mountain stream empties into a small alluvial fan and then moistens a meadow further down slope.

Hydrology is a major factor in

18 Arnold

understanding soils in landscapes. One should imagine how water moves in all kinds of landscapes. If you do not, or have trouble - ask someone for help. On this high elevation alluvial fan the soil is a loamy skeletal, micaceous Aquic Cryochrept. The classification and all of its implications is another part of the story of soils in lal1dscapes.

The geomorphology of a fan an a high bench, the stratigraphy of

sediments in a mountain alluvial fan, and the hydrology of such a setting gave rise to this soil and its current set of properties. As we walk over the landscape we often overlook the little zones or areas that have a few more stones.

They commonly seem to be concentrated in ephemeral

drainage-ways that are nearly perpendicular to the slope. A roadcut in the Bitterroot mountains in Montana reveals a small upland drain filled with stones. This serves as a drain for the hillside. This feature is a relict periglacial stone stripe that markedly influences the hydrology of the landscape today.

Integrating information Geomorphology, stratigraphy and hydrology are ingredients of understanding soils in their landscapes.

In eastern Venezuela, early Pleistocene sediments have been

reworked and the area uplifted to form the 'Mesa' (Arnold and Schargel, 1978). Maps of geomorphology and of pedology (soils) throughout this region have many similarities of boundaries. There are many similar patterns at scales near 1:25000. An open pipeline trench in Ontario, Canada exposed glacial lacustrine sediments which have been eroded and with some reworked materials deposited along the drainage-ways. The continuum of soil reveals the changing properties with landscape position indicating the relevance of hydrology in interpreting the features in such a soil landscape. Soil surveys have traditionally worked hard to separate the recognisable soils in landscapes. Now the soil scientists are working to explain the integration of such differences within catchments.

It is the linkages that help us understand the

changes that are occurring throughout ecosystems.

Soil variability It has been common for soil scientists to make soil maps where and when the variations are observed to be systematic. Where the variability is systematic, either

19

Climate-induced soil behaviour

in space or in time, it is predictable and its location can be noted and plotted on maps. In a similar way it has been noted that random variability in space or in time cannot be mapped even though it can be described. This concept of systematic versus random has been the guideline for whether or not differences and similarities could be delineated in map form. I find it interesting that many soil features that seem to be random in space or in time are 'biological'. For example, termite mounds, gopher mounds, crayfish chimneys, tree throw mounds and pits, tongues of organic enriched soil in subsoils, and haystacks in fields. These biologically produced entities are components in soil map units rather than being large enough to be map units themselves, mainly because of scale but also because of non-predictable location within the delineated areas (Fridland, 1976). Components occur in all map units. They have sets of properties, their extent can be estimated, but their exact location is unknown. Wetland spots, or organic soil inclusions, in some map units contribute most of the organic carbon in an area thus these almost random entities are important for certain 'extrapolations'. The French school of Pedology explodes the soil profile and each horizon is treated as a stratigraphic layer useful for tracing out the spatial distribution of selected sets of properties in a landscape (Ruellan and Oosso, 1993). It is powerful and useful, and perhaps a necessary way to understand a landscape. As they say "try it - you might like it". If one starts with soil descriptions of pedons taken from pits as controlling sections of soil bodies, it is easy to use an auger to check along transects and map out variations of horizons, textures, and other features. Questions often asked are: How many transects should be made? How many observations are needed? How far apart should the observations be made (Steers and HajeK, 1979)? There are no well defined answers because local variability influences the decisions.

As with all

statistics, it is best to obtain some preliminary results and see where you want to go from there. In some sampling studies we found that cumulative information from transects commonly under- and over-estimated the real proportions of the components in a landscape unit (Dos Santos, 1978). It was impossible to know when or where the results differed from the real proportion. However, the standard error of the

20 Arnold

estimates narrowed as samples increased. Of course, the cost or effort to increase precision may be high and consequently it is important to consider what will be acceptable variations for your purposes. What measures are you willing to use and what range of variation is necessary to support or deny your hypotheses?

WHERE TO GO FROM HERE? Where does one go from here? Assuming that an appropriate hypothesis has been devised or selected, there are three things about the landscape that should be of assistance. A good site description, a working model of the landscape's evolution, and knowledge about the kinds of current processes taking place in the soil landscape. A careful description of the research site in the field allows you to examine unusual relationships that are not predictable and to have a reliable baseline for evaluating further results.

The local setting relies on the area geomorphology and the

hydrology that influences the vegetation and acts on the parent materials. It is an old model of soil forming factors that give rise to processes which interact and result in soil bodies in space and time (Jenny, 1941). It is useful to know how to use the soil science model to your advantage in designing appropriate research. Once the site has been examined and information about the materials, stratigraphy, soil properties, and hydrology obtained it is useful to develop descriptive models of how the soils might have formed. This opens up possibilities of explanations of current phenomena. For example, in southern Ontario we studied a soil profile that had a stone-free mantle overlying a thin stone line at the top of calcareous basal till (Wang, 1965). One explanation of this profile is that it is two soils in two materials. A profile developed in stable time followed by erosion stripping off the upper horizons and concentrating a few coarse fragments at the surface. With continued landscape erosion the stone line was covered with sediment and another soil developed in the overlying materials. The observed profile would then be a relict soil overlain by a newer one. Two other hypotheses suggest that there is one soil developed in two materials. Soon after the till was laid down the upper mantle was deposited and soil horizons developed through the discontinuity. This is similar to the early periglacial events in Germany (Kleber, 1992). A second explanation is that ablation drift was deposited from stagnating glacial ice over the basal till. Such a phenomenon is common throughout New England in the northeastern U.S.A. (Wang and Arnold, 1973).

21

Climate-induced soil behaviour

In considering research at a site it is relevant to keep in mind the kinds of processes that have already taken place, especially where the soils have well expressed properties. Micro-time and micro-scale soil processes eventually give a whole new character to a larger landscape. Remember, that a millisecond per hour is a long time over ten million years. Contours and slopes perpendicular to the contours can be grouped into nine configurations, with convex, straight, and concave classes of each (Ruhe, 1975). Consider how each of these nine would react with surface runoff. Join them together in different arrangements and rethink how surficial water would behave. This kind of exercise can be the expert system that is the starting point of hydrology models of landscapes and watersheds. Think about the possibilities, you might enjoy them. Landscapes are a part of a continuum that we barely comprehend. We should never forget the connection - search for the linkages. The end result of good research is not just information - it is the ability to extrapolate, to generalise, to make significant contributions for uncharted areas. What links the similarity of patterns that exist at different scales?

Is there a self-similarity of such patterns from one scale to

another? What is a fractal dimension for us in soil science? A fractal dimension, D, can be a group of polygons whose perimeters and areas have been measured (Arnold, 1990). The slope of a plot of the log perimeters versus the log of the areas is one-half of D, the fractal dimension. With geographic information systems and digitised detailed soil surveys it will be relatively easy to determine the fractal dimensions of many areas of soil polygons. We do know that fractal dimensions of maps are between 1 and 2,

but other than that,

we simply are not yet

knowledgeable. SOME WORDS OF ENCOURAGEMENT

The principles of stratigraphy are deceptively simple.

The law of superposition

unlocks numerous details in landscapes (Daniels and Hammer, 1992). Reality is commonly much more complex than our models of how landscapes are put together. Know what to sample, and sample it. Take careful measurements, use calibrated standards, and do good scientific work. The burden is yours, not your neighbours. Be aware of expected variability and realise that most experimental work must overcome the thresholds set by inherited local variability. Plan for it, plan how to use it, and plan to overcome it.

Estimates and best guesses often help us visualise

22 Arnold

important implications of change. Leam about some characteristics of the climate for your research sites. There are models that estimate soil temperature and soil moisture regimes from weather station data (Eswaran and Reich,

1993).

Modifications of such a model also estimate growing seasons and various indicators of stress. You, or your research group, must decide what is, or may be acceptable levels of variability - for just about everything. Confidence limits narrow around the mean as more observations are obtained, but the cost goes up accordingly. When is enough enough, and at what cost (Aronoff, 1985)? Where should you, and where do you, draw the line for your proposals?

Time, money, preCision, accuracy - what

determines your value system of science? We should also consider the scientific limits for risk taking. When we want to know the 'at least' situation, the lower confidence limit suggests where, or how much (Ginevan, 1979). It is associated with the user's risk; it is commonly the politician's crutch: a yardstick of not how much, rather how little. Be conscious of such things and learn the value of the 'user's risk' for your information. It may not be as important to others as you want to believe it is.

SOME CONCLUDING REMARKS Many interesting things happen because landscapes are three dimensional, such as water movement, genetic processes, erosion of sediments, nutritional losses and productivity decline (Larson and Foster, 1988). Landscapes can serve as a basis for improved understanding. For some years I have made an interesting observation about people.

When

showing slides I have shown a close-up view of the rice paddies of Banuae, Philippines some of which were constructed over two thousand years ago. I explain all of this and then tell the audience that I have misled them.

I show them the

landscape as they would see it in the field and explain that the previous slide was upside down. Why do this? I do this to remind us that there are many possible illusions and that we sometimes see what others, or even ourselves, want to see. We are often confronted with illusions. Some soil properties are likely to change over a ten-year period (Arnold et al., 1990). Others are more likely to change over a hundred-year period, and some take

23 Climate-induced soil behaviour

thousands of years to detect change. It is not so important that you believe in these lists of properties or the placement of a particular feature in a specific time period, as it is that you consider that there are meaningful differences in the modification and alteration of soil features that are observable or measurable in soil profiles. Design your climate-change experiments based, in part, on a reasonable knowledge of the kinds of soils and their distribution in your research area. Proper grouping of results minimises the local variations of soils and maximises differences you are searching for. Let soil information strengthen your research, not be the weak link. Do not hide soil variability; learn to use it to reinforce your results. Global change research involves many disciplines. As soil scientists we have an unique opportunity to provide valuable information to the work of others. We have many kinds of information that are supportive of integrated systems. If we give it away - we may receive more in return. This requires teamwork, mutual trust, and understanding. You must be good, you must have integrity, you need humility and a good sense of humour to balance all of these aspects. At each stage of development, a civilisation utilises the very best information available. The belief in the Gods of yesterday was just as consistent a human behaviour as is our current belief in the truth of scientific investigation and in the belief that governments can, and will, make informed judgements for the good of our global habitat.

Thank you for not forgetting the power of knowing soils in their

landscapes and for realising that extrapolations can take advantage of the beauty of nature. It is our wonderful Pedosphere.

CLIMATE CHANGE, DESERTIFICATION AND THE MEDITERRANEAN REGION

c.s. Kosmas and N.G. Dana/atos Agricultural University of Athens, Department of Soils and Agricultural Chemistry, lera Odos 75, Botanikos 11855, Athens, Greece.

ABSTRACT

Land desertification, a series of natural processes leading to gradual environmental degradation, is now considered as a serious threat to the semi-arid areas of the Mediterranean, and particularly to the marginal hilly lands of the region. Soil erosion comprises the dominant process of land deterioration and desertification. Adverse climatic conditions, irregular terrain with steep slopes, geology and long periods of land misuse are the main factors responsible for desertification in the Mediterranean. The climate of the region is characterised by strong seasonal and spatial variations in rainfall and large oscillations between minimum and maximum daily temperatures. Moreover, higher temperatures and more pronounced aridity are predicted to prevail during the next decades. The extensive deforestation and intensive cultivation of sloping lands since ancient times has already led to soil erosion and degradation through the progressive inability of the vegetation and soils to regenerate themselves.

Hilly soils developed on Tertiary and Quaternary formations usually

have a restricted effective rooting depth for plant growth. Under hot and dry climatic conditions, the tolerance of these soils to erosion is low, and rainfed vegetation can no longer be supported. Many areas on limestone formations are already desertified with the soil mantle eroded and the vegetation cover completely removed.

For

example, soils formed on marl deposits, despite their considerable depth and high productivity in normal and wet years, are very susceptible to desertification in dry years. Intensive human interference in hilly areas vulnerable to desertification has severely damaged or totally destroyed the productivity of these lands due to the loss of soil volume beyond a critical point.

NATO ASI Series. Vol. I 23 Soil Responses to Climate Change

Edited by M.D.A. RounseveII and P. J. Loveland © Springer·Verlag Berlin Heidelberg 1994

26 Kosmas and Dana/atos

INTRODUCTION There are indications that the earth is undergoing a long time-scale warming as a result of the increased concentrations of carbon dioxide, water vapour and certain other gases; a process referred to as the greenhouse effect. Greenhouse warming might have undesirable consequences for the Mediterranean region including greater soil loss and water shortages, diminishing fertility and ecological well-being. The climate of the Mediterranean region is characterised by strong seasonal and spatial variations in air temperature and rainfall. In addition, greenhouse warming may be accompanied by greater climatic extremes such as prolonged droughts or heavy rains as well as the possibility of increased frequency of such extreme events. Such changes in weather conditions will cause even greater seasonal variations in soil moisture and thus in plant growth and productivity. In addition to global climatic changes, rapid shifts in agricultural patterns can result from economic pressures and urbanisation, which are characteristic of extensive areas of the Mediterranean. Such shifts are closely related to changes in vegetation cover, the ability of soil to recover after abandonment, and the capacity of soils to withstand extreme events. The reduction andlor damage of the vegetation cover may enhance land degradation and ultimately land desertification. Information on desertification processes is modest. Recent data from the United Nations indicates that as much as 30 million hectares may be subject to desertification in the Mediterranean region. Since ancient times, agriculture, animal grazing and other human activities such as fires and timber extraction have radically altered the natural vegetation to such a degree that in a few cases only a little remains of the original plant cover, the rest having been altered or completely destroyed.

As a result, soil erosion represents a serious hazard for southern

Europe, bringing about large reductions in vegetation growth, siltation of water courses and reservoirs, loss of fertilisers and finally land degradation and desertification.

Desertification is now considered a serious problem requiring

immediate and urgent attention especially in the marginal lands of the Mediterranean region. For example, under the current European Community research program EPOCH, major research projects have been initiated dealing with desertification and land surface properties: the MEDALUS project (Mediterranean Desertification and Land

27 Desertification and the Mediterranean region

Use) and EFEDA (ECHIVAL Field Experiment in a Desertification-threatened Area). The objective of the first project is to investigate the ecological and soil processes involved in desertification through detailed site studies and mathematical modelling which will then develop into a policy framework. The second project is focusing on the climatological-hydrological interactions between vegetation, atmosphere and land surface properties.

FACTORS AFFECTING LAND DESERTIFICATION IN THE MEDITERRANEAN REGION Knowledge of desertification is generally limited because of a lack of basic information on many important sub-processes. Information on desertification is currently being collected at individual sites (Thornes, 1991). Until now this research was not multidisciplinary, focusing mostly on specific ecological or erosion problems. The complex interactions between climate, vegetation, soils and hydrology that are important for desertification have only recently been identified as specific research goals (Thornes, 1991). The rate of soil degradation is dependent upon the rate of vegetation degradation, which in turn is influenced by both the climate and the intensity of land use (Grainger, 1992). Natural and semi-natural ecosystems with evergreen sclerophyllic species are well adapted to the Mediterranean climate and landscape conditions. Disequilibrium of these systems can be caused by extreme climatic conditions and human manipulation. The vegetation cover may be altered radically by Man within a short time, but physical and biological changes within the soil, affecting erosion rates, may take longer.

The main factors affecting desertification in the Mediterranean region are the following: climate, topography, geology, and human activity (Yassoglou and Kosmas, 1990).

Climate The contemporary climate of the Mediterranean region is characterised by strongly seasonal and spatially highly variable rains with sudden changes from cold to warm months and large oscillations between minimum and maximum daily temperatures.

28 Kosmas and Danalatos

The total mean annual rainfall is less than 830 mm and the mean monthly temperature is greater than 9°e most of the time (8 to 12 months) (Palutikof et al., 1992). The most characteristic feature of the Mediterranean climate is the long, dry summer and the winter rainfall type (also known as Mediterranean) with the occurrence of most of the precipitation during the 8 or 9 cooler months. A recent analysis of the global temperature record is given in Figure 1. The yearly values are the global-mean temperatures derived from a common reference period (1950 to 1979) with each value representing over one million observations on both land and sea (Warrick et al., 1990). A warming trend is evident up to about 1940, followed by several decades of little change. From about the mid-1970s the warming resumed at a more rapid rate. The six warmest years in the series occurred in the 1980s. Overall, the world has warmed by about o.soe since the late 19th century.

0.0

-0.5

1860

1880

1900

1920

1940

1960

1980

2000

Figure 1 The global temperature record 1860 to 1988 (After Jones et al., 1988).

Maheras (1988) studied the annual and seasonal precipitation for the period 1891 to 1985 and showed that two major wet periods occurred in the western Mediterranean (1901 to 1921 and 1930 to 1940) followed by two drier periods (1958 to 1963 and 1975 to 1979).

Furthermore, two major dry periods were recorded in the same

intervals (1922 to 1929 and 1942 to 1954), and a decrease in rainfall since 1980 with the most recent years also being dry. Palmieri et al. (1991) studied the annual and monthly rainfall at 27 Italian meteorological stations for the period 1870 to 1970 and demonstrated that a negative trend existed for only six of these stations.

Rainfall

data from Gibraltar, one of the longest European rainfall series (1791 to 1990), show four major dry periods centered on 1822, 1873, 1945 and 1979 (Wheeler and MartinVide, 1992).

29 Desertification and the Mediterranean region

General Circulation Models (GCM's) predict greater aridity and higher summer temperatures in the Mediterranean region, winters in southern and central Europe becoming warmer by 4 to 6°C (Warrick et a/., 1990). Moreover, the average model scenarios for winter show an increase in precipitation in the Mediterranean region from 0.2 mm day-1 to 0.6 mm day-1. These changes represent an increase by 10 to 30% of the average winter precipitation, but the credibility of this scenario is low. Rainfall amount and distribution are the major determinants of biomass production on hilly lands under Mediterranean conditions. Decreasing rainfall combined with increased rates of evapotranspiration markedly reduce the soil moisture available for plant growth.

Reduced biomass production, in turn, directly affects the organic

matter content of the soil and the aggregation and stability of the surface horizon to erosion. Kosmas et a/. (1993c) studied the effect of diminishing soil moisture on soil properties and biomass production of rainfed wheat in a two year rainfall exclusion experiment.

This work showed that the total above ground biomass production

(TAGBP, in kg m-2 ) was reduced proportionally with the amount of rainfall excluded (RE) according to the relationship TAGBP = -0.46 + 1.64 * RE.

Reductions in

biomass of 90%, 71.4%, and 53.4% were measured in the experimental plots in which rainfall was reduced by 65%, 50% and 30%, respectively (total amount of rain falling in the open field during the growing period R = 361 mm). As Figure 2 shows the leaf area index (LAI) of the crop was also greatly affected throughout the growing period. The maximum values of LAI measured in the plots of 100%, 70%, 50% and 35% rain interception were 5.2, 3.7, 2.9 and 1.6, respectively.

Diminishing soil

moisture affected the organic matter content and the aggregate stability. The mean organic matter content was decreased to 0.9%, 1.1 %, and 1.2% from an initial value of 1.4% in plots with 35%, 50% and 70% rain interception in a two year period, respectively. Large areas of the Mediterranean region cultivated with rainfed cereals are confined to hilly lands with shallow soils and thus are very sensitive to erosion. In a number of years, the prevailing weather conditions during the growing period of these crops may be so adverse that the soils remain bare, creating favourable conditions for overland flow and erosion.

Any loss of soil volume from these marginal lands,

markedly reduces the potential for plant growth, ultimately leading to desertification and abandonment.

30 Kosmas and Dana/alos

o

35%

tJ.

50%

•

•

70%

100%

6 ,.... ...... "A

i

\

/

\ \

,.

4

\

I ,~

if' \

;;:

f/ A

..J

/ './ r,

/:/>

2

/./,

¥' /

r

....~...:::.::.:t

~..."? / /~ q-6'

\\

\... lI."

\

•

\. l!j

'\

\.

\

I

\

'

\ ., \ \ \ \ \

~

\

0~~/~7~:;_-~-~A=--~~~_~A~~==~~~~~'~\~\~I~~~ 300

350 400 450 500 550 Nov I Dec I Jan I Feb I Mar I Apr I May I June

Figure 2 The change in LAI of fainfed wheat grown in plots with 35%, 50%, 70% and 100% rain interception, Topography

Lands of high quality are scarce around the Mediterranean due to physical and historical factors such as irregular terrain, steep slopes, soil limitations, high rainfall variation and long periods of misuse (CORINE, 1990). The particular nature of Mediterranean relief, with slopes subject to extensive deforestation and intensive cultivation since ancient times, has led to soil erosion and the formation of skeletal soils through the progressive inability of the vegetation and soils to regenerate themselves. As an example, hilly areas intensively eroded with high slope instability represent 17% of Italian soils, with the lowland areas representing 23% of the total surface area.

In the Basilicata basin, southern Italy, sediment deposition in the

valleys reaches several thousands of tons per square kilometer per year (EEC, 1990).

31

Desertification and the Mediterranean region

Apart from slope grade, slope aspect is also considered an important factor for land degradation processes.

Cottle (1932) found that soils on south-facing slopes in

south-western Texas were 5 to 11°C warmer at 5 cm depth, had 5 to 15% less moisture and 24 to 44% greater evaporation than soils on the north-facing slopes. The situation is similar in the Mediterranean region where lands with southern aspects are drier than those with northern aspects and have slower recovery of vegetation.

According to the results of the CORINE project (1990), the largest area in southern Europe characterised by 'high actual erosion risk' (defined on the basis of topography, soil properties, climate and vegetation cover) occurs in Spain, covering 29% of the country or 145000 km 2 . In Portugal, areas with high erosion risk cover almost one-third of the country (26878 km 2). Similar areas occupy about 20% of the rural land surface in Greece (24712 km 2) and as much as 10% of the total extent of Italy. Such data are not available for north Africa and the Middle East.

Geology Soils derived from different parent materials react differently to soil erosion, vegetation and desertification. Limestones produce shallow soils with a relatively dry moisture regime.

These soils are characterised by high erodibility and slow

vegetation recovery. As Plate 1 shows, several areas on limestone formations in the Mediterranean region are already desertified with the soil mantle eroded, and the vegetation cover completely removed. Under Mediterranean conditions, regeneration of soils and vegetation is impossible, and desertification is irreversible. Similarly, acid igneous parent materials produce shallow soils with high erodibility and high desertification risk. Extensive areas on hilly agricultural lands in the semi-arid zone of Greece and more generally in the Mediterranean region are mainly cultivated with rainfed cereals. The production of cereals is largely dependent on the amount and distribution of precipitation during the growing period. The soils of these areas are formed mainly on Tertiary and Quaternary marls, conglomerates and shale-sandstones. Soils on conglomerates and shale-sandstones usually have a restricted effective rooting depth resulting from erosion and limiting subsurface layers such as a petrocalcic horizon or bedrock at shallow depth. The tolerance of these soils to erosion is low,

32

Kosmas and Dana/atos

and under hot and dry climatic conditions and severe soil erosion, rainfed vegetation can no longer be supported leading to desertification.

Plate 1 Severely eroded and deserlified land with hard limestone as geological

substratum (Thiva, Greece, photo by C. Kosmas, June 1993). Kosmas et al. (1993a) demonstrated the impacts of parent material and landscape position on drought and biomass production of rainfed wheat along hillslope catenas under semi-arid Mediterranean conditions. These catenas are located in hilly areas with rolling topography and soils formed on marl, conglomerates and shalesandstones.

Total above ground biomass production was measured at specific

hillslope positions (shoulder, backslope and foots lope) in two successive growing periods and was related to soil properties, landscape position and climatic data. Based on the long term average climatic data for the area, the first growing period was extremely dry (total amount of rainfall R = 95 mm against R = 370 mm in an average year) while the following growing period was exceptionally wet (R = 663 mm). As Figure 3 illustrates, under dry conditions, areas on marl were almost bare. The total above ground biomass production was 960 and 1355 kg ha- 1 for shoulders and footslopes, respectively, whereas averages of 4620 and 11110 kg ha-1 were measured on the shoulders and footslopes of the shale-sandstones, respectively. In

33 Desertification and the Mediterranean region

the following wet growing period, the opposite occurred with soils on marl more productive than soils formed on conglomerate and shale-sandstone formations (Figure 3).

Shale-sand.

I$i;;~~;~ Marl

20000

15000

ro1:

Ol

~

(/) (/)

ro E 0

1989-90

10000

5000

0

20000

en

1990-91 15000

10000

5000

0

Shoulder

backslope

footslope

Sampling sites

Figure 3 Effect of parent material and landscape position on biomass production of rainfed wheat under exceptionaffy dry (1989 to 1990) and exceptionaffy wet (1990 to 1991) weather conditions in the semi-arid zone of Greece (After Kosmas et al., 1993b).

34 Kosmas and Dana/atos

The study demonstrated that in dry years gravel and stones are extremely important for the prevention of desertification by conserving appreciable amounts of soil water from evaporation through surface mulching.

Stony soils along slope catenas of

conglomerates and shale-sandstones, despite their normally low productivity, may supply appreciable amounts of previously stored water to the stressed plants and ensure an adequate biomass production in dry years (Kosmas et al., 1993b). As Table 1 illustrates, the biomass production of wheat growing under water-limiting conditions was reduced by 10 to 30% in plots in which the rock fragments were removed from the soil surface during cultivation, as compared with the stony plots of the same soils along hills lope catenas. The biomass production was also affected by the landscape position, soil depth and texture (Kosmas et al., unpublished data). Soils formed on marl are free of coarse fragments and despite their considerable depth and high productivity in normal and wet years, they are susceptible to desertification in particularly dry years. In such dry years, they are unable to support any vegetation due to adverse soil hydraulic properties and the absence of gravel and stone mulching. Table 1 Effect of removal of rock fragments (RF) from the soil surface on total above ground biomass production (TAGBP) of rain fed wheat along catenas with soils formed on conglomerates.

Site

Cover(%)

TAGBP (kg m-2 ) measured in plots with RF

NS 2 NF 7

9.0

0.326

28.6

NS 9 NB11

Reduction (%)

without RF

0.794

0.291 0.616

10.7 22.4

19.0

0.554

0.446

19.5

13.2

0.554

0.476

14.1

NB12

4.6

0.875

0.788

9.9

NB14

19.0

0.960

0.818

14.8

NB15

52.2

0.640

0.449

29.8

NB17

51.4

0.426

0.337

20.9

35 Desertification and the Mediterranean region

Human activity In recent decades, favourable soil and climatic conditions and the availability of ground or surface water has resulted in intensive farming of the lowlands of the Mediterranean region.

The development of high input agriculture in the plains

provided much higher net outputs than those obtained from terrace agriculture. The result was the continuing abandonment of the agricultural lands on sloping terrains followed by the collapse of the traditional agro-pastoral forms of management. Fires and overgrazing ensued, the relatively deep soils were eroded, and the land was degraded. Land abandonment had the opposite effect in non-terraced areas. The growth of natural vegetation slowed down desertification, enabling some less degraded areas to recover fairly rapidly. In many cases, rapid development resulted in the over-exploitation of the aquifer system for a variety of uses (agricultural, industrial and drinking water) causing the gradual intrusion of sea water into aquifers. Irrigation using such water with high salt contents increased the salinity of the soil, yielding unproductive desertified land. Soil salinization is a potential land desertification threat in the soils located in climatic zones characterised by high xerothermic indices.

Moreover, as tourism

grows continuously, water allocation will be shifted towards domestic consumption. It has already been documented that water consumption will increase following the increase in tourism (Grenon and Batisse, 1989), and combined with the substantial increase in water requirements associated with high input agriculture, a significant water allocation problem is likely to occur with further degradation of the plains through salinization. However, the main mechanism of land desertification in the Mediterranean region is the loss of soil volume capable of supporting a soil protective vegetative cover. Intensive human interference in hilly areas vulnerable to desertification has severely damaged or totally destroyed the productivity of these lands due to the loss of soil volume beyond a critical point. The widespread occurrence of Mollisols in an extensive, undulating plain was recognised long ago in Thessaly, the greatest lowland formation of Greece. Some 40 years ago, the organic matter content of the topsoils was commonly in excess of 5%, but since the beginning of this century, human activity disturbed the long established equilibrium in the area. Large scale deforestation and intensive cultivation of the land was initiated in the 1920's (Danalatos, 1993). This resulted in a drastic decrease in organic matter content to

36 Kosmas and Dana/atos

its present level of less than 2.5% because of oxidation, water and wind erosion and frequent burning of the crop residues.

In Mediterranean environments, the

aggregation of the surface horizon is strongly determined by the vegetation dynamics which provide the primary organic carbon for developing soil structure. It is well known that aggregation of the surface soil layer plays an important role in determining the hydrological characteristics including water movement and retention capability (Tisdall and Oades, 1982).

The intensive cultivation of the Thessaly

sloping lands accompanied by accelerated erosion finally created adverse conditions for crop productivity resulting, in some cases, in the abandonment of the land. The soils in these areas are classified as Xerochrepts or Xerorthents according to the Soil Taxonomy (Soil Survey Staff, 1975). Desertification comprises a series of natural processes leading to gradual environmental degradation.

These processes can be observed, but only after a

sufficiently long time period, when the accumulated effects of change in plant cover become apparent. Kosmas (1993) demonstrated the impact of land use change on soil conditions and erosion and land degradation along a hillslope catena occurring over a long period of time. The major land use of this area for the past 160 years has been olive trees under semi-natural conditions. In part of the area, the whole vegetation has been removed and vines have been grown since 1979. The area was described in detail before clearing of the vegetation in 1979 and after 12 years, in 1991. The data obtained suggested that the change in land use from olives to vines had a large degrading effect on soil properties related to soil erodibility. The size and stability of the soil aggregates under vines decreased by about 10 fold compared with the soils under olives (Table 2). The organic carbon content was reduced by 33% while the depth of the A-horizon was altered as a consequence of the 12 years of cultivation with vines. It was estimated that at least 247 kg m-2 of soil material was eroded from the upper part of the study catena. In contrast, the depth of the A-horizon under olive trees remained almost unaltered for the same period. The total run-off measured in a period of 21 months (from March 1991 to November 1992) ranged from 4.9 to 12.3 mm under vines, whereas under olives much lower values were observed (0.3 to 2.6 mm). The total average sediment measured during the same period were 0.4 g m-2 in the olive grove, and an extreme of 441.9 g m-2 was measured in the vineyard despite the lower slope gradients on this transect. This information reveals the great impact of land use change on environmental conditions particularly on land degradation in hilly areas under Mediterranean

37 Desertification and the Mediterranean region

conditions, and should be considered seriously in future land use planning. Vines create favourable conditions for overland flow, erosion and desertification. Table 2 Some chemical and physical characteristics of the studied soils along two

transects cultivated with olive trees and vines. Property

Transect under olives site 1

site 2

site 3

Transect under vines site 4

site 5

site 6

site 7

134.50

139.20

153.20

133.50

129.08

127.00

137.50

Slope (%)

17.02

18.00

22.30

12.40

9.10

7.00

16.70

Clay (%)

21.40

28.40

36.50

29.10

30.10

34.40

14.80

(23.40)

(27.90)

(37.50)

(28.60)

(27.10)

(35.90)

(19.7)

29.00

20.90

34.20

19.00

9.60

10.90

5.20

Altitude

Gravel (%) Organic C (%)

Total N (%) Aggregate

1.22

1.57

1.82

1.06

0.87

0.92

0.50

(1.28)

(1.66)

(1.75)

(1.22)

(1.10)

(1.24)

(1.7)

0.11

0.13

0.15

0.09

0.08

0.09

0.04

(0.12)

(0.14)

(0.16)

(0.12)

(0.10)

(0.11 )

(0.15)

3.20

5.80

7.10

0.60

0.10

0.90

0.10

2.30

2.60

0.30

12.30

4.90

8.90

67.10

83.80

size (mm) Runoff (mm) Sediment

0.80

0.50

0.10

441.90

(g m-2) Data in brackets represent measurements in 1979.

Soils on stable surfaces and under good conditions of plant cover may improve with time by accumulating organic material, increasing floral and faunal activity, enchancing aggregate stability, increasing infiltration capacity, and decreasing their erosional potential (Trimble, 1990). Soeiro de Brito et al. (1993) demonstrated that erosion in hilly areas of Portugal is markedly higher in areas cultivated with wheat (even in plots under fallow) than in abandoned fields.

Furthermore, natural

vegetation such as Cistus maquis is more protective of the soil than wheat. However, the dominance of this species over other annuals is responsible for the absence of low bushes leaving large patches of bare soil.

38

Kosmas and Danalatos

CONCLUSION Desertification represents a serious problem for the dry areas of the Mediterranean region. Climatic changes such as increasing temperature and aridity tend to reduce the biomass production potential, enchance erosion and land degradation, increase land albedo and thus further increase air temperature and aridity. This is followed by a net decrease in carbon dioxide assimilation by the plants and hence greatly contributes to climate change at the local and regional scale and thus to desertification in a broad sense.

Desertification comprises a series of natural

processes in which our knowledge is still limited. Multidisciplinary research on the complex interactions between climate, vegetation, soils and hydrology is absolutely necessary to improve future policy and land management strategy decisions with respect to environmental protection against desertification.

CLIMATE CHANGE, SOIL SALINITY AND ALKALINITY

G. Varal/yay Research Institute for Soil Science and Agricultural Chemistry (RISSAC) of the Hungarian Academy of Sciences, Herman Ott6 ut. 15, H-1022 Budapest, Hungary.

INTRODUCTION

Human activities are leading to changes in the global environment at virtually unprecedented rates, with potentially severe consequences for our future.

The

study and solution of the problems of global environmental change require urgent and efficient action. This crucial task formulates a challenge for science: "to describe and understand the interactive physical, chemical and biological processes that regulate the total Earth System, the unique environment for life" (Toward, 1988). Due to natural processes and human activities, such as increasing energy consumption, industrialisation, intensive agriculture, urban and rural development, considerable changes have taken place in the gas composition of the atmosphere. The CO 2 concentration, which was at about 180 to 200 ppm after the last glaciation and 270 ppm in early industrial times rose to the present 320 to 350 ppm, an approximately 25 to 30% increase, during the last 100 to 200 years. Similar tendencies were registered for other major greenhouse gases (CH 4 , NOx, carbohalogenides) (Scharpenseel, 1990).

This may lead to a rise of global

temperature at a rate of 0.1 to 0.8°C per decade. The spatial and temporal patterns of temperature increases will be heterogeneous and are expected to be greatest in the northern mid-continental region of North America and Eurasia.

The various

Global Circulation Models (GCM's) predict an increasing rate of temperature rise from the Equatorial to the Polar regions, with relatively higher temperature increases during summer and lower increases during winter periods in both hemispheres. These predictions are still rather uncertain, because in addition to solar radiation the

NATO AS! Series. Vol.! 23 Soil Responses to Climate Change Edited by M. D. A. Roonsevell and P. J. Loveland © Springer-Verlag Berlin Heidelberg 1994

40 Varal/yay

influences of circulation, changes in vapour content, cloudiness, albedo and surface roughness have to be better evaluated quantitatively. The changing temperature regime will be accompanied by changes in precipitation characteristics: quantity of rain and snow, their spatial and temporal distribution pattern, rain intensity, etc.

Forecasting precipitation changes is even more

uncertain. Within the World Climate Programme, WMO makes serious efforts to improve the complex GCM's and quantify at least some basic, plausible, hypothetical climate change scenarios.

From the state-of-the-art of these GCM's it can be

concluded that in the first period of the forecast warming, the average global precipitation will decrease, with high spatial variability and considerable regional redistribution (Manabe and Holloway, 1975; Manabe and Wetherald, 1986). Later, these tendencies will be counterbalanced by increasing evaporation from water surfaces, firstly from the world's oceans. The evaporation rate from the sea surface is about 44000 km 3 year1. If the present average temperature of 11.5°C increases by 1°C, evaporation will increase by 20% or 10000 km 3 year1.

This increasing

evaporation will lead to higher air humidity and (probably) more precipitation, again in a more uneven spatial and temporal distribution, and with increasing frequency of heavy rainfall and extreme weather events (droughts, storms, floods, etc.) (Climate Change, 1992). A changing climate will cause considerable changes in the natural vegetation and in land use practices. The great vegetation zones will move in the direction of the poles, at a predicted rate of 25 to 200 km per 100 years. Land use practices will follow or modify these changes; depending on environmental and socio-economic conditions (Climate Change, 1992; Toward, 1988). Changes in temperature, precipitation, natural vegetation and land use practices will result in significant changes in soil formation processes and soil properties. These impacts and their relationships are summarised in Figure 1. It can be concluded from Figure 1 that the quantitative evaluation of any climatic change on soil conditions is rather difficult. The uncertainty of long-term global temperature and precipitation forecasts are combined here with the complex, integrated influences of changing vegetation

and

land

use

patterns

and

the

changing

hydrological

cycle.

Consequently, the global soil change prognosis can only be a rough, qualitative estimate. Such a rough estimation is summarised in Figure 2, showing the influence of four main climatic scenarios on the major soil degradation processes, and

41 Soil salinity and alkalinity

indicating their natural and anthropogenic causative factors (Varallyay, 1990b, 1990c).

5

E

evapo transtorage ration

air

h~at

r.

r.

groundwater

surface runoff

soil erosion

Figure 1 The influence of climate change on soil formation and soil processes.

42 Varal/yay

Climatic scenarios Soil degradation. processes

Cold Cold Hot Hot Sign and and and and lOry Wet Dry Wet

mm ~ "'3'" m m 4

Soil erosion by wind

4 .... , , 0 :-:3'" 4

Acidification

A

4

Salini'Zation / Alkalization

S

Soil erosion by water

E

~ · .

4

~ , 3" ~~ 4

I~

p

Extreme mOisture regime(water logging

M

Biological degradation Unfavourable nutrient regime

8 .3.. · .... N ·::'3':"

Soil pollution (toxicity)

T

.'"

Strong

4

Natural

. " ..

4

~~

~

rn ~ .. .:3"..: "'3"

4 . .. '

"

2-

Antrop

1,2,3

9,10,11,12

3

9,10,11,12

2,4

13,15

5,6,8

14

"

Physical degradatfon

m

4

Causa1ive factors

-

10,12

5,6,7

1t12)4

-

11,16

(2,6 )

13

-

16

. 4

~ Medium V~d Slight ~ No or negligible Causative factors:

Natural

Anthropogeneous

1. Undulating surfaces 2. Parent rock 3. Lack of penn anent and dense vegetation 4. Litter decomposition 5. Low-lying lands 6. Improper drainage 7. High water table (non-saline) 8. High water table (saline)

9. Deforestation 10. Overgrazing II. Irrational land use 12. Improper tillage practices 13. Irrational fertilizer application 14. Improper irrigation IS. Acid deposition 16. Chemical soil pollution

Figure 2 The influence of four main climatic scenarios on the main soil degradation processes, and their natural and anthropogenic causative factors.

43 Soil salinity and alkalinity

FACTORS IN SALT ACCUMULATION, DEVELOPMENT OF SALINITYI ALKALINITY AND FORMATION OF SALT AFFECTED SOILS The basic preconditions for salt accumulation are as follows: 1.

Salt sources: (i)

local weathering (dissolve and mobilise the soluble compounds of rocks,

primary and secondary minerals and other geological deposits);

(ii) surface waters (world oceans, seas, rivers and smaller waterways, lakes); (iii) subsurface waters (groundwater, saline seep); (iv)

human

activities

(irrigation

water;

fertilisers

and

other

applied

agrochemicals; sewage waters and sludges, wastes of various origin).

2.

Transporting agents (water, wind), transporting salts by horizontal or vertical flow: (i) from a large area (e.g. from an extensive watershed) to a small(er) accumulation territory; (ii) from the whole soil - geological strata profile to a specific accumulation horizon.

3.

Driving force for solution movement: relief (for surface runoff); hydraulic gradient (for saturated flow as groundwater movement, saline seep, leaching); suction gradient (for unsaturated flow as capillary transport); concentration gradient (for diffusion and hydrodynamic dispersion).

4.

Negative water balance: evapotranspiration (ET) is higher than precipitation (P) plus irrigation water (I): ET> P + I

(1)

and equilibrated by horizontal flow (surface runoff, R, plus infiltration (saline seep) in the unsaturated zone, F, plus groundwater flow, G): P + I + Rj + F j + G j '" ET + Ro + Fo + Go

(2)

and, Rj + Fj + G j > Ro + F0 + Go

(3)

and (at least periodical) drying.

5.

Rj , F j , G j = inflow

(4)

Ro' F0' Go = outflow Limited drainage conditions:

(5)

(i) poor vertical drainage of the soil profile (low infiltration rate, permeability and hydraulic conductivity); (ii) poor horizontal drainage of the area (low-lying territory with imperfect outflow).

44 Varal/yay

For a comprehensive impact analysis of climate change on salinization/alkalization processes a quantitative assessment and evaluation of these factors is required. Salt sources There are different, sometimes contradicting, opinions on the impact of predicted (and rather uncertain) climate change on the rate and type of weathering processes. If we accept the three main characteristics of predicted climate change: • global warming (Scharpenseel, 1990; Scharpenseel et a/., 1990); • increasing spatial and temporal variability of rainfall distribution and soil moisture (Fisher, 1990; Manabe and Holloway,

1975; Manabe and

Wetherald, 1986; Sombroek, 1990; Varallyay, 1990a, 1990c); and • increasing frequency of extreme weather events, including droughts (Climate Change, 1992), we may come to the conclusion that the weathering rate, the dissolution, mobilisation and redistribution of soluble compounds of the geological deposits (parent material) will increase in the future because both higher temperature (Yu, 1985), and the periodical wetting and drying help these processes, thus leading to the formation and accumulation of water (or soil solution) soluble salts. These salts will appear and may accumulate in the water and sediments of lakes, small waterways, rivers, seas and world oceans, as well as in the various subsurface waters. Because a warmer and (at least potentially) drier climate will increase aridity and drought-hazard, the extension of irrigation will be necessary over large areas for efficient agricultural production to satisfy the increasing food and fibre demand of the sharply growing world population. According to expert assessments about 25% of the present irrigated lands face damage by salinity/alkalinity and each day many thousands of hectares will be affected by salts due to poor-quality irrigation water, improper irrigation practices and misguided soil management (Climate Change, 1992; Szabolcs, 1979, 1989, 1990, 1991; Szabolcsand Redly, 1989). The reduction of this tendency is particularly difficult, because most of the 'no problem' fields are under agricultural utilisation, and the extension of irrigated agriculture is limited only to the marginal lands, where biomass production faces more and more difficulties,

e.g. increasing hazard of salinization/alkalization (Climate Change, 1992; Stigliani and Salomons, 1992; Scharpenseel et a/., 1990). These lands are classified by

45 Soil salinity and alkalinity

Szabolcs (1979, 1989, 1990, 1991) and others (Kovda and Szabolcs, 1979; Szabolcs and Redly, 1989) as 'potentially salt-affected areas', where any extension of irrigated agriculture without adequate salinity/alkalinity control will result in unfavourable environmental side-effects such as waterlogging and secondary salinization/alkalization. The increasing rate of fertiliser application and the (socially-driven) disposal of sewage waters and sludge's, and wastes of various origin (consequently with different chemical composition and hazard for the environment) result in an additional water-soluble salt load to agricultural lands (Stigliani and Salomons, 1992). Eustatic rise of sea-level

Because of a 0.6°C temperature increase over the last 100 years a 6 to 25 cm (0 to 22 cm, 10.5 cm, 1 mm year 1) eustatic rise of sea-level was recorded by various authors, and they forecast a 5 to 44 cm (17 to 26 cm, 9 to 29 cm) rise by 2030, and a 0.5 to 1 m rise in the next 100 years (Brammer and Brinkman, 1990; Brinkman and Brammer, 1990; Climate Change, 1992; Scharpenseel et al., 1990). Table 1 The rate and causes of sea level rise.

The main causes of sea-level rise

Rate of sea-level rise (in cm) in the last 100 years

Thermal expansion of sea water Melting of large mountain glaciers Melting of inland ice caps (e.g. in Greenland) Melting of Antarctic ice sheet Total

2 to 6 1.5 to 7

(4)

1985 to 2030

1 to 4

(4) (2.5)

6.8 to 14.9 2.3 to 10.3 0.5 to 3.7

(10.1) (7.0) (1.8)

-5 to 5

(0)

-0.8 to 0

(-0.6)

-0.5 to 22

(10.5)

8.8 to 28.9

(18.2)

The 50 to 100 cm sea-level rise expected in the next 100 years is somewhat analogous to the rise which occurred at the end of the last ice age: 80 to 120 mover an interval of 14000 to 18000 years (Toward, 1988; Climate Change, 1992). However, even a 50 cm eustatic sea-level rise is a serious environmental threat for many millions (perhaps a billion) of people with the following harmful or undesirable

46 Varal/yay

ecological, economical and social consequences (Ahmad, 1990; Brammer and Brinkman, 1990; Brinkman and Brammer, 1990; Climate Change, 1992; Day and Templet, 1989; Hekstra, 1989; Jelgersma, 1988; Pirazzoli, 1989; Sombroek, 1990; Titus, 1987): 1. Flooding (inundation) of coastal lowland plains by saline sea water; flat, low-lying coastal fringes; marshlands and swamps; deltas and estuaries of big rivers such as the Nile, Niger, Ganges, Yangtse, Mekong, Irriwadi, Indus, Amazon, Parana, Mississippi, Volga, Danube, etc.; Pacific island states, Latin American and Caribbean countries. 2. Increase in storm tides (which could effect extensive areas up to the present 5 m contour, and the penetration of saline or brackish water by tidal flooding further inland. Inundated and tide-affected coastal lowlands, and tidal estuarine flood plains potentially at risk represent one third of the world's total cropland (with a population of about 1 billion); this area might be doubled within the next 100 years. 3. Increasing frequency and intensity of large-scale flood disasters which considerably hamper agricultural production in highly productive alluvial soils. For such reasons Bangladesh may loose 20%, and Egypt 15%, of their arable land. 4. Rapid erosion of coastlines. About 350000 km of coastline requires protection, sometimes by expensive engineering works. Without enormous empoldering, the eroded coastlines represent a permanent environmental risk for the coastal regions. 5. Intrusion of saline sea water or brackish tidewater into estuaries and groundwater bodies (freshwater aquifers and water lenses) near coastlines. The consequences are the limited applicability of these waters for irrigation and human consumption (e.g. the Delaware river in the USA for the water-supply of Philadelphia; Pacific atolls, etc.); the upstream movement of the saltwater wedge; seasonal salinization of delta areas particularly in dry periods. 6. Unfavourable changes in the sediment transport of rivers and canals: greater silting up -+ reduction of their drainage function -+ impeded soil drainage in coastal plains -+ rising water table + salinization of groundwater resources and increasing possibility of pollution.

These factors amplify changes in moisture

regime (e.g. shift in monsoons or snow melt) and lead to territorial extension of salt-affected soils (Szabolcs, 1990, 1991; Szabolcs and Redly, 1989).

47 Soil salinity and alkalinity

7. Flooded or groundwater-connected landfills represent a particular environmental hazard: creating favourable conditions for the accumulation of pollutants in the sediments of deltas and estuaries of highly polluted rivers, and for the mobilisation of potentially toxic elements in soils and ecosystems (chemical timebomb effect, CTB) (Hekstra, 1989; 8tigliani and 8alomons, 1992).

Field water balance, soil moisture regime Most climate-induced soil changes are related to the field water balance and to the soil moisture regime. Their components are summarised in Figure 3. The general conclusion drawn from Figure 3, is that an increase in the average annual precipitation will be followed by an increase in: • surface runoff (R) in hilly lands with undulating surfaces and without permanent and dense vegetation, if the infiltration rate, permeability and water storage capacity of soil is limited; • infiltration (I) and water storage (8) within the soil, if these are not limited, such as in flat lands; • groundwater recharge (G) if the soil profile has good vertical drainage, and permeability is not limited, especially in low lying areas; • evaporation (E) if infiltration is limited; • transpiration (T) in the case of well-developed plant canopies. A decrease in precipitation results in adverse changes. Increasing temperature will result in: • increases in the potential E and T, if the plant canopy is not suffering from limited water supply due to climatic or soil induced drought, e.g. low precipitation or limited water content • decreases in R, I, 8 and G, especially if accompanied by low precipitation; • decreases in the intensity (depth) of permafrost, which will modify its geographical boundaries, opening possibilities for increased water storage and water movement, biological activity and soil formation processes within the unfrozen part of the soil.

48 Varallyay

p

0 RiRo FiFo

;y

zone (in and out)

GiGo

Or

Gs I S

V

P

R G

I I

5 E T F

Gr GS

Cald, '""

I I i

I i

=groundwater flow (in and out) =rise of water table =lowering of the water table = infiltration into the soils =infiltration into the groundwater =storage within the soil (recharge) =filtration to the plant roots, uptake by plants

~

~:I

=precipitation =irrigation water =surface runoff (in and out) =filtration in the unsaturated

T E C

Cald, dry

I

. =. transpiration =evaporation =capillII}' transpon from groundwater

Hal,""

HOI, dry

d,D

I I

d

'j

d

I

0

(i)

0 0 0

0

0

D

I 0

d

(I)

0

E

0

E

E I

I I

-

-

-

1

i

-

-

-

-

I

(I)

-

Figure 3 Components of the field water balance and soil moisture regime and the influence of 4 potential climatic scenarios on these factors. d and D: slight and strong decrease, i and I: slight and strong increase, E: no change (equilibrium).

49 Soil salinity and alkalinity

Decreasing temperature will result in adverse changes. These general influences are modified by the impact of vegetation characteristics (type, density, dynamics, species composition, biomass production, litter and root characteristics), and depend greatly on the type, intensity and spatial and temporal distribution of atmospheric precipitation. Man's influence is still more complex. Land use, cropping patterns, amelioration (including water and wind erosion control, chemical reclamation, irrigation and drainage) and other activities sometimes radically modify the field water balance and its components. In regions under agricultural use the influence of global climatic change on the soil moisture regime is partly affected through these human actions (Ahmad, 1990; Hekstra, 1989; Sombroek, 1989; Scharpenseel et al., 1990). Based on General Circulation Models (GCM's) and hypothetical scenarios several prognoses have been elaborated for the assessment of the influence of predicted climatic change on the soil moisture regimes.

Manabe and Holloway (1975) and

Manabe and Wetherald (1986) indicated a substantial drying in spring and summer in the middle-to-upper mid-latitudes; and significant high-latitude increase in precipitation, and hence runoff. They showed increases in continental soil moisture between 30 0 S and 60 0 N, with greatest increases in the winter and spring. Decreased soil moisture is found for most seasons in tropical latitudes and in the southern hemisphere. Manabe and Holloway (1975) prepared a map indicating the estimated global distribution of the annual mean rate of runoff. In addition to the global-scale approaches numerous authors have analysed the regional hydrological impacts of global climatic change. The subtropical regions may become drier with greenhouse-effect-induced warming which could interact with anthropogenic desertification: overgrazing, irrigation-induced salinization, accumulation oftoxic elements (Toward, 1988). Stigliani and Salomons (1992) published some predictive data on the percentage changes in soil moisture in Europe in the summer assuming a doubling of atmospheric CO2 to 600 ppm. According to their prediction, northern Europe would experience a reduction of between 20 and 30%, western Europe 30 to 50%, central Europe 20 to 40%, most of the Mediterranean region 20 to 30%, and the Iberian peninsula by as much as 70%. In northern latitudes the reduction results from the earlier occurrence of the snow melt season followed by a period of intensive evaporation.

In southern Europe it is mainly due to decreased precipitation and

50 Varallyay

increased potential evaporation during the summer months. The increase in soil moisture in winter would be greatest in northem Europe and least in southern Europe. In northern Europe much warmer winters are predicted with increased soil wetness, and somewhat warmer summers with increased drying. In southern Europe winter and summer would be considerably warmer with a slight increase in soil moisture in winter and significant drying in summer. In westem and central Europe the changes will be less extreme (Stigliani and Salomons, 1992). Salt transport and salt accumulation processes Predicted global warming may lead to two plausible climatic scenarios from four potential ones: (a) warm and dry (before the considerable evaporation increase from the world's oceans); and (b) warm and wet (if the increasing evaporation from the water surfaces leads to both greater cloudiness and higher precipitation over the continents). Before attempting to summarise the impacts and consequences of these two climatic scenarios on salt transport and salt accumulation processes, it is necessary to emphasise the different character of coastal and continental (inland) salinity. Coastal salinity depends primarily on sea-level and its tidal, seasonal and long-term fluctuations; temperature, concentration and chemical composition of sea-water; the geology, geomorphology and relief of the coastal area including the river deltas and estuaries, and; the climate (temperature, precipitation, rate of evaporation and their spatial and temporal variability). The assessment and prediction of coastal salinity and salt accumulation requires comprehensive information on the above-mentioned factors and on the hydrological, chemical and ecological consequences of a potential rise in the eustatic sea-level, summarised earlier. The simplified general conclusion is that any rise in the eustatic sea level will result in the territorial extension of coastal salinity under the direct and indirect influences of saline sea water. The conceptual impact analysis of climatic scenarios on the continental salt transport and salt accumulation processes needs a much more complex approach. Here we summarise only the most important potential consequences of the selected climatic scenarios using the symbols in Figure 3.

51

Soil salinity and alkalinity

Scenario (a): warm and dry. The basic processes are: increasing temperature (during summer) higher aridity

~

~

higher E + T

~

higher concentration of the soil solution + higher capillary transport

from the groundwater to the overlying horizons (in the case of a shallow water-table) ~ increasing salinity. The schematic diagram of this process is shown in Figure 4 indicating also the changes in the chemical composition of the migrating soil solution (Varallyay, 1968). Potential additional processes are: •

high capillary transport

~

lowering of water table (in the case of stagnant

groundwater with slow horizontal flow) •

~

~

increasing salinity;

seepage from unlined canals and reservoirs and/or filtration losses from overirrigated or imperfectly irrigated fields ~ horizontal groundwater inflow and/or saline seep from the surroundings transport

•

reduction of capillary transport

prevention of salt accumulation or decreasing salinity (Figure 4); high capillary transport ~ lowering of the water table is prevented by horizontal inflow (F j or G j)

•

~

~

rise of water table

~

~

higher capillary

increasing salinity (Figure 4);

drainage operations

~

vertical and horizontal outflow lower water table

reduced or prevented capillary transport

~

~

prevention of salt accumulation.

If the temperature increase occurs in winter its impact on salinity is much less: e.g. changing rain/snow ratio and/or shorter frozen period ~ better infiltration ~ leaching or winter rise of water table

~

negligible or no capillary transport

~

no salt

accumulation.

Scenario (b); warm and wet. The basis process is: increasing temperature + increasing precipitation ~ lower aridity ~ lower concentration of the soil solution + downward flow in the soil profile (at least during most of the year) ~ leaching (decreasing salinity). Potential additional processes are: •

rise of water table (in the case of stagnant groundwater with slow horizontal flow) overwetting of the soil ~ difficulties in mechanical and tillage operations;

~

aeration problems; water-logging hazard; •

periodical drying ~ capillary transport from shallow or rising groundwater to the overlying horizons ~ temporal or seasonal salt accumulation.

52 Varal/yay

(e

97

I (hi

-----Ijl

NaIM glHCn_--

--(kl

III

Figure 4 Schematic diagram of salt accumulation in the Hungarian Danube Valley a) change in the chemical composition of water during concentration; b) chemical composition of the soil solution; c) sequence of the precipitation of salts; d) carbonate accumulation horizon; e) height above sea level; f) intense salt accumulation on the surface (solonchak soils); g) salt accumulation near the surface (solonchak-solonetz soils); h) leaching (meadow and meadow alluvial soils); i) soil surface; j) equilibrium (critical) level of water table; k) actual level of water table; i) hydrostatic pressure. The direction of the arrow indicates the direction of water and salt movement. The width of the white arrows is proportional to the intensity of water movement, while that of the black arrows is proportional to the salt content of the water.

53 Soil salinity and alkalinity

These two main summary scenarios can be combined with each other and may appear in various combinations in nature. Many further variations are caused by the physiographical variability of the whole region (agroecological unit, water catchment area, etc.) and the various human activities. Among these anthropogenic influences the most significant are agricultural water management practices (irrigation, drainage, water conservation). Climate change, particularly the 'warm and dry' scenario, sometimes make the extension of irrigation necessary (and more and more frequently) into marginal lands. Here irrigation is faced with ever increasing constraints: ecological, economical and sometimes socio-economic limitations. Some of the most important potential impacts of irrigation on salinity are as follows: accumulation of salts from irrigation water; uneven distribution of water ~ local over-irrigation water table •

~

increasing capillary transport

~

~

filtration losses

~

rising

increasing salinity (Figure 4).

horizontal and vertical redistribution of salts within an area and in a soil profile, respectively; leaching of salts from the soil profile or from the root-zone; production of soluble salts from local weathering processes.

Consequences of salt accumulation

According to the dynamic equilibrium of the solid-liquid interface phenomena (cation exchange, miscible displacement, etc.) the chemical characteristics (concentration, ionic composition) of the soil solution governs the exchangeable ion composition of the soil absorption complex and the properties of solid soil constituents. Further, these influence the mineralogical status: degradation, destruction and formation of clay minerals, orientation of clay particles, etc.; rate of hydration and dispersion; swelling-shrinkage-cracking phenomena; arrangement of primary particles, shape, size and stability of soil micro- and macro-aggregates, structural elements and consequently pore-size distribution; factors of soil moisture regime: spatial (vertical and horizontal) and temporal distribution of water, moisture potential and moisture movement.

The interactions between heavy-textured swelling clays and alkaline

sodium-salt solutions are particularly significant, and result in radical changes in the water economy of the soil, and consequently in mass and energy transport and transformation, thus creating a specific and potentially extreme soil ecological

54

Varal/yay

environment for native vegetation or agricultural crops. The various impacts of soil salinity/alkalinity on agroecological potential are summarised in Figure 5 .

... .

~

.~~

~51------l

.lI U

Figure 5 potential.

The main factors of salinity-alkalinity and their influence on ecological

SPATIAL MODELLING APPROACHES TO EVALUATE THE EFFECTS OF CLIMATE CHANGE ON FUTURE CROP POTENTIAL AND LAND MANAGEMENT

T.R. Mayr1, M.D.A. Rounsevell1 and D. de la Rosa2 1 Soil Survey and Land Research Centre

Cranfield University, Silsoe, Bedford MK45 4DT, U.K. 2Consejo Superior de Investigaciones Cientificas, Instituto de Recursos Naturales y Agrobiologia, P.O. Box 1052, 41080 Sevilla, Spain.

INTRODUCTION

Air temperatures are likely to rise as a consequence of increases in the tropospheric carbon dioxide (C02) concentration and other radiatively active trace gases. These elevated temperatures are expected to influence other climatic factors such as precipitation, cloudiness, humidity and windiness. Resulting changes in global and regional weather patterns could have profound implications for agricultural production and thereby the food resources of an ever-increasing world population. Climate impact assessment studies have two closely-linked objectives: firstly, to study the physical interactions of climate and the environment with their likely socioeconomic consequences on mankind and, secondly, to provide policy makers with the best possible information to formulate appropriate responses. A great deal of scientific research has been undertaken in the agricultural sector to assess the likely consequences of climate change, covering issues from specific crop phenological effects to broad-scale continental changes in agricultural suitability. There have, however, been very few attempts at climate change impact assessment for regional agricultural potential within the European Union countries.

NATO ASI Series. Vol. I23 Soil Responses to Climate Change Edited by M. D. A. Rounsevell and P. J. Loveland © Springer·Verlag Berlin Heidelberg 1994

56 Mayretal.

This paper discusses the merits and drawbacks of different modelling techniques available for climate change assessment and introduces a new approach to regional crop suitability modelling.

In doing so, the paper addresses some of the wider

problems encountered in climate change impact research. Current climate change predictions and their possible implications for the soil-plant-climate continuum are also discussed.

MODELLING CLIMATE CHANGE IMPACTS ON LAND-USE SYSTEMS There have been numerous attempts to establish reliable relationships between climate and agriculture. Monteith (1981) provided a detailed review of the effects of important climatic variables on plant growth and development.

In more general

terms, Roy and Hough (1978) discussed the effects of climate on British agriculture. Biophysical models can be used to evaluate these interactions and there are two main model types: empirical-statistical models and mechanistic simulation models. Carter et al. (1992) discussed these two types of models in detail and their comments are paraphrased below. Empirical-statistical models are based on the statistical relationship between climate and the crop (e.g. Brignall and Rounsevell, 1994). They are usually developed on the basis of present-day climatic variations. Thus, one of their major weaknesses in considering future climate change is their limited ability to predict effects of climatic events that lie outside the range of present-day variability. They are also limited by being based on statistical relationships between factors rather than on an understanding of the important causal mechanisms. However, where models are founded on a good knowledge of the determining factors and where there are good grounds for extrapolation, they can still be useful predictive tools in climate impact assessment.

Empirical-statistical models are often simple to apply, and less

demanding of input data than simulation models. Mechanistic simulation models make use of established physical laws and theories to express the dynamics of the interactions between climate and a crop.

In this

sense, they attempt to represent processes that can be applied universally to similar systems in different circumstances. Usually additional model calibration is required to account for features of the local environment that are not modelled explicitly, and this procedure is generally based on empirical data. Nevertheless, there are often firmer grounds for conducting predictive studies with these process-based models

57 Modelling approaches to evaluate the effects of climate change

than with empirical-statistical models.

The major problem with most simulation

models is that generally they have demanding requirements for input data, both for model testing and for simulating future impacts. This tends to restrict the use of such models to only a few points in geographical space where the relevant data are available. In addition, theoretically-based models are seldom able to predict system responses successfully without considerable efforts to calibrate them for actual conditions. An empirical approach The following suitability model is an example of an empirical approach for climate change impact assessment on crop potential.

This work originated in the land

suitability classification for particular crops and cropping systems developed by Thomasson (1982), and was later modified by Jones and Thomasson (1987) for use at the national scale using the Soil Survey and Land Research Centre's Land Information System (LandiS). The work was then extended to model crop potential and climate change by Rounsevell and Brignall (1993) and Brignall and Rounsevell (1994). The approach relies on the calculation of a range of agroclimatic indices expressing land qualities in a land suitability model. The indices include: Accumulated temperature

Accumulated temperature (AT), the integrated excess of temperature above a fixed base value or threshold over an extended period such as a month or year, is used as a reasonable guide to energy input to a crop since it correlates with crop potential and vegetative growth. The average AT above O°C for January to June (AT> 0), and the average AT above 5.6°C for the whole year (AT> 5.6) are used as a measure of the heat energy available when assessing crop suitability (Jones and Thomasson, 1987) Potential soil moisture deficit

The Potential Soil Moisture Deficit (PSMD, mm) is calculated on a monthly basis from the cumulative difference between rainfall (R) and potential evapotranspiration (PET) during the period of the year when a deficit occurs i.e. during the summer and autumn months (Jones and Thomasson, 1985). PSMD ==

L (PET - R)

(1 )

58 Mayretal.

PSMD reflects the climatic contribution of water to a growing plant.

Figure 1a

illustrates PSMD for the baseline climate (1961-1975) for England and Wales and Figure 1b approximates to the current best estimate (+2°C relative to the baseline) of climate change for UK summers (Climate Change Impacts Review Group, 1991). Crop adjusted moisture deficit

The crop specific Moisture Deficit (MDcrop) can be derived by adjusting PSMD according to specific crop types. This accounts for the difference in ground-cover during the growing season between grass and an arable crop. For example, winter wheat in England and Wales attains full ground-cover by April, but requires no water from the soil after mid-July (Thomasson, 1979; Jones and Thomasson, 1985), thus; MD (winter wheat) = PSMD (mid-July) - PSMD (April/3)

(2)

Profile available water

In addition to the climatic water balance, the water available to a growing crop from the soil profile (AP) has to be taken into consideration. This is calculated from the soil water contents between specific limits of matric potential, adjusted according to the rooting habits of different crops.

zr

AP=IFC-PWP

o

(3)

where FC is the field capacity, PWP the permanent wilting point and zr is the depth of rooting (Hall et al., 1977). Droughtiness

A measure of droughtiness can be derived from the difference between the soil available water (AP) and the crop adjusted moisture deficit MD(crop) (Thomasson, 1979); droughtiness = AP - MD (crop)

(4)

The numerical value of droughtiness indicates the shortfall in crop water demand. If yield reductions are to be avoided then irrigation must compensate for this deficit. Figures 2a and b show the change in droughtiness for winter wheat for an increase in temperature of 4°C relative to the baseline.

59 Modelling approaches to evaluate the effects of climate change

(SOmm SO- 100mm 100-150 mm 150-200mm 200-250mm >250mm

Figure 1a PSMD for the baseline climate, 1961-1975, for England and Wales at a spatial resolution of 5 km x 5 km.

250mm

Figure 1b PSMD for England and Wales at a resolution of 5 km x 5 km when the temperature is increased by 2°C relative to the baseline.

60

Mayretal.

Vf!TY c!roUl:llty Mod ...... tely drouglllJ' Sl;,IlUy drougllty Non drougbly Urb ..n 200 Itm

Figure 2a Droughtiness for winter wheat at the baseline climate for England and Wales (after Brignall and Rounsevell, 1994).

200 I0 to 20

after 1 September

Over 40

>20 to 40

Over 80

Well

Well

>50 to 80

Well

Well

Moderate

Marginal

>20 to 50

Moderate

Moderate

Moderate

Marginal

20 and less

Marginal

Marginal

Marginal

-20 and below

Well

Moderate

Marginal

Results

The model is normally used to give an average response of the soil based on climatic patterns derived from long-term meteorological datasets, i.e. to predict soil status and crop suitability in 6 years out of 10. However, the model also can be used to simulate a single growing season, using data for a specific year. The output of the model is the classification of a particular soil unit in relation to a particular

62 Mayretal.

crop. A soil map can then be classified in terms of suitability. The model is based on empirical relationships of restrictions to crop production and yield, and so is strictly valid only under the conditions in which these relationships were derived because crop physiological processes are not taken into account. A hybrid approach Empirical-statistical models do not sufficiently define the mechanisms of a crop system and so cannot be extrapolated confidently to a future climatic scenario. Because of large input data demands, simulation models cannot easily provide information on the broad-scale ability of a region to sustain adequate crop production, which is of most relevance to local communities or policy makers. A hybrid approach, in which simulation modelling is combined with more conventional land evaluation procedures (FAO, 1978) to estimate crop suitability ratings, may provide an appropriate way forward. The ACCESS model (Agroclimatic Change and European Soil Suitability) is currently being developed to estimate the suitability of soils within the European Union for a range of European Land Utilisation Types (LUT's) (Loveland et al., in press). The model is being developed for test regions from central England, southern France and southern Spain so that the diverse farming and socio-economic systems found in Europe can be incorporated. ACCESS will run at two levels of complexity, where a simpler regional model (ACCESS I) is supported and validated by a detailed, sitespecific model (ACCESS II). The difference in complexity between the two models is reflected in the time-steps used for simulation as well as in the number of input parameters required to run the two models. The following sections provide an overview of the work undertaken for ACCESS I and in so doing outlining some of the difficulties encountered in modelling climate change impacts. Aspects of ACCESS II are discussed in another paper in this volume (Legros et al., 1994). Soil-Water-Crop model

The central component of soil-crop-climate modelling is the successful simulation of the soil-water balance. Considerable achievements have been made in modelling soil hydrological processes and their influence on crop yield (Childs et aI., 1977; Feddes et al., 1978; van Keulen, 1982; Diericks et al., 1986; Simota et al., 1986). Unfortunately, most simulation models require daily time-steps of weather data and many input parameters to characterise soils and crops. This renders these models

63 Modelling approaches to evaluate the effects of climate change

impracticable for land suitability assessment on a regional, national or European scale (Simota et al., 1993). In order to work with minimum data inputs, ACCESS I runs on monthly time-steps for weather data, limited soil information and very simple crop information. The model makes extensive use of pedotransfer functions to establish soil water retention curves and matric potentials for individual soil map units. Potential evapotranspiration (PET)

Estimates of PET are central to any soil water balance and crop growth modelling, but Global Circulation Models currently do not provide these estimates. With an increase in temperature, a corresponding increase in PET could be expected and Budyko (1989) estimated this as 4% per °C (Carter et al., 1992).

PET is most

sensitive to changes in humidity on an annual basis, but solar radiation and wind speed are also influential. Vegetation changes such as those that might occur in response to the direct effects of rising CO 2 concentrations can also be important (Institute of Hydrology, 1992; McKenny and Rosenberg, 1993). Both ACCESS I and ACCESS II require PET estimates to be provided. However, because PET data are not always available for the spatial application of ACCESS I, PET can be calculated using Thornthwaite's technique. The technique was chosen on the basis of minimum data requirements and a monthly temporal resolution. In addition, the formula does not require any variables which are not provided by GCM's. However, less accurate PET estimation must be expected with Thornthwaite's technique compared with Penman's formula: a comparison of Thornthwaite with Penman shows that the former tends to under-estimate PET in southern Europe, but over-estimates it in northern Europe (Kenny and Harrison, 1992). The differences between Penman and Thornthwaite arise because Thornthwaite ignores changes in humidity and radiation (Linacre, 1963). This is confirmed by Ward (1963), Smith (1964) and Jones et al. (1990). The shift between the two curves in Figure 3 is due to the phase lag of temperature on radiation conditions (Linacre, 1963). A non-linear response to temperature is apparent with the Thornthwaite method;

the rate of

change of PET is greater at higher temperatures, particularly during the warmer part of the year and at warmer locations. Thornthwaite predictions of the rate of change of PET with temperature greatly exceed those of the other methods (McKenny and

64

Mayretal.

Rosenberg, 1993). However, temperature is unlikely to be the only climatic element affected by greenhouse warming, and simulations have shown that changes in other climatic elements can offset or intensify the effects of rising temperature on PET (Institute of Hydrology, 1992; McKenny and Rosenberg, 1993).

120.00 100.00 80.00

E

- - - Penman

.§. 60.00

...w 0

--0-

Thornth .

40.00 20.00 0.00

c

...,--i:-::---'~-:'---11 Succion .

Interception •

tI

5011 water

I

I I

I I I

8 Figure 1 The principles of crop models (solid arrows indicate mass or energy

transfer; dashed arrows indicate cause and effect). Nevertheless, if the model operates only on the basis described so far, growth could be infinite! In reality however, biomass production is restricted in different ways: energy resources, C02, water and minerals are not unlimited.

Further, different

specific constraints appear: temperature too low to guarantee normal vegetative growth, soil too shallow, too wet, etc. All these factors limiting to growth can produce what it is called 'stress'. Finally, the plant cannot grow beyond its genetic potential.

75 Crop models: problem of climate change

The plant's genetic make-up regulates the appearance of several phases of development (crop phenophases): emergence, anthesis, maturity, etc. It is important to note that one should not confuse growth (quantitative, cf. mass) and development (qualitative, cf. floral initiation). The overall effect of the qualitative and genotype aspects is difficult to measure since the phenomena concerned involve complex biochemical concepts.

Fortunately, from field experience or from controlled

laboratory experiments it has been found that, in the majority of cases, the different stages of plant development can be related to temperature.

More precisely, it is

seen empirically, and without necessarily much-scientific justification, that each stage appears when a value is reached of the cumulative mean daily temperature since sowing. For example, sunflower requires about 150 degree-days from sowing to emergence (Mollet, 1991). Similarly, it is possible to define a number of degrees indicating the appearance of physiological maturity or the onset of senescence, this being characterised by the cessation of growth and the onset of start of leaf decline. In other words, most models convert day-calendars to thermal-time related to the time-temperature integral. The prinCiples of simulation. Simulations are organised on the basis of some general principles, not always identical from one model to another. The essential points are reviewed in turn. Origin of leaf and root biomass

We have seen above the essential role played by the leaves and roots in exploiting resources of matter and energy. It is convenient, therefore, to examine the model principles which control the growth of these organs. The production of total biomass can be calculated as a function of solar energy received, then divided into the effective biomass of roots, leaves and other parts of the plant. It is on this basis that, for example, the CORNGRO model functions for maize growth (Childs et al., 1977). However, the conversion of energy to biomass and even more its partitioning between the different plant organs, assumes the introduction, in the equations, of coefficients of variable value, which are not easy to fit (Bonhomme and Ruget, 1991). This is why many modellers, at risk of some loss of coherence, do not consider the growth of all the organs starting from conversion of energy to biomass followed by redistribution between the organs. Instead they infer leaf or root growth directly starting from an empirical relationship based on

76 Legros et al.

thermal time. This is partly justified by the relationship seen, by different authors working with different plants, between air temperature and the number of, or extension of, leaves. Thus, EPIC calculates the leaf area as a function of thermal time and the root mass mainly as a function of energy. Similarly, in models concerning the growth of sorghum, SORGF (Maas and Arkin, 1978) and SORKAM (Rosenthal et al., 1989), leaf growth is separated from the growth of other organs. There are also intermediate cases, for example CERES-MAIZE (Jones and Kiniry, 1986). In this model, leaf biomass is indeed part of the total biomass, but the number of leaves is related to the thermal calendar. In some models total biomass is assumed to be a function of the amount of water transpired since the beginning of the crop cycle (Wilson and Jamieson, 1985). This approach implies that a plant cannot increase its production without further transpiration. This remains empirical and cannot be used in all cases. Time steps Incident solar energy and temperature vary within the same day. Ideally, models

should thus be working at hourly time steps. This is effectively the case with the BACROS model (Goudriaan et al., 1985), which is concerned mainly with photosynthesis. Unfortunately however, the necessary climate data are not always available. One solution is to reconstruct them as was done for the water balance models developed at METEO-FRANCE (Choisnel and Jacquart, 1991). Temperature, air humidity and solar energy were reconstructed at hourly intervals. The SORGF model (Maas and Arkin, 1978) works at hourly steps; many others (e.g. SOYGRO) at daily steps. This reduces the calculations and there are no particular difficulties in relating plant growth and recorded climate data at this time scale. Generally, the models use minimum and maximum temperature, rainfall, and solar energy (Ritchie and Otter, 1984). Models operating at weekly or ten-day time steps are rare. Stages of development

Obviously, stages of development depend on the crop considered. The principal stages (crop phenophases) are not the same for cotton, wheat or sunflower. Thus, crop models are adapted to the relevant crop. Beyond the vegetative differences, one can identify two concepts within models. For some, the different phases must be described carefully within the model; not only to signal their appearance or disappearance, but also to understand the consequences of modification of the

77

Crop models: problem of climate change

plant's functions.

For example, the daily sharing of the assimilates between the

roots and the leaves is not the same at the beginning and the end of the vegetative cycle. Thus, good management, within the model, of the development phases allows better representation of the true functioning of the plant. However, this functioning is controlled, for each of the phases, by parameters for which the values must be calibrated. The complexity of the modelling is thus increased. This explains why other modeliers prefer to limit themselves to a minimal description of the vegetative cycle. For example, in the basic version of the EPIC model, only two phases are distinguished: growth and senescence. It is true that many more phases have since been introduced into the model so as to better simulate the consequences of different stresses on final yield taking into account the period in which the stress appears (Quinones and Cabelguenne, 1990). In a feasibility study on grain maize at the European scale, an intermediate position was taken i.e. 3 phenophases: sowing to flowering, flowering to onset of senescence, onset of senescence to maturity (Begon et al., 1990). Establishing the thermal time

In all models the accumulation of measured average daily temperatures, or estimated values ((Tmin + Tmax) / 2), is used in conjunction with threshold temperatures (above or below which growth is halted) to establish thermal time. The following form of expression is used: HUi = ~((Tmin + Tmax) / 2 - Tb)

(1 )

where HUi represents accumulated heat units up to day i and Tb (0C) is the cropspecific base temperature. Besides the lower threshold, Tb, it is possible to introduce an upper limit, Th, expressing the fact that above a definite temperature the functioning of the crop is prevented. From thereon, modellers have different choices between two extremes: 1. either, use the heat units to calculate, with the aid of adequate phenological data, the number of days corresponding to each stage of plant development; in this case each stage is characterised, from the preceding one, by a certain quantity HU, or 2. use thermal time concerned only with the end of crop development i.e. to maturity; in which case one can calculate a heat unit index (HUI).

To obtain HUI, HUli

(corresponding with day i) is divided by PHU (potential heat units), which is

78 Legros et 81.

calculated in the same manner and corresponds to the thermal time necessary to achieve the theoretical maturity. HUI

=HUIi / PHU

(2)

Thus one knows, for each day, one's position on a relative time scale from HUla (sowing) to HUI 1 (maturity). This approach is used in EPIC. However, one should note that the growth of vegetative organs is not directly proportional to the heat units HU and thus cannot be directly proportional to the time HUI. In effect, in certain stages of the cycle the effect of heat is more evident than in others. This is why, in EPIC, supplementary quantities are introduced; HUF (heat unit factor for driving leafarea-index

development)

and

HUFH

(factor for estimating

harvest

index

development), calculated simply from HUI. These factors express how leaf growth or the reserves leading to final harvest will vary over time relative to HUI. In other words, the idea of a phenological calendar is not dismissed, but presented differently.

Moreover, intermediate approaches are possible.

For example, for

wheat, different models distinguish greater or fewer numbers of stages. Stress and limiting factors

In most models, the possible stresses experienced are estimated, for each day, by coefficients lying between 0 and 1. These coefficients affect growth (e.g. EPIC) or development (Brisson and Delecolle, 1991). For example, in EPIC, the increase in leaf area index (~LAI) is controlled by a multiplication factor known as the crop growth regulating factor (REG). If REG becomes zero, ~LAI is also zero for the day in question. If REG equals one, growth is unaffected.

This approach has the

advantage of closely representing the action of stress, which can only reduce growth or development in a well-defined manner. The number of stresses taken into account depends on the objectives and purpose of the model.

Lack of water is

almost always considered. Some models also deal with excess of water, nitrogen deficiency, salinity, etc. When several stresses occur at the same time, it is necessary to consider how to combine their actions. One could think of multiplying the coefficients corresponding to each stress. However, the most logical method, a priori, is to compare all the stresses and to consider only the most important. For example, suppose that one had, at the same time, a water deficit and a nitrogen deficit, capable of reducing growth to 40% and 60% of the optimum respectively. In such a case, it is more

79 Crop models: problem of climate change

logical to assume that growth is reduced to 40% (taking into account only one stress maximum) rather than 24% (product of the two stresses). In effect, nitrogen which allows 60% of the maximum growth, should not be deficient if the real growth is 40% of the same maximum. It is on this basis, corresponding to the law of limiting factors, that the EPIC model works. However, it is important to note that this law cannot be applied when CO 2 is one of the causative factors. An increase in the amount of CO 2 increases photosynthesis, even if the plant has suffered additional different stresses (Warrick et al., 1986). In some cases, the daily stresses are added. If their values are aj and if the stress period is n days then k

= :E(1

- aj) representing, roughly, k

days of zero growth.

PRINCIPAL MODEL FUNCTIONS Energy interception Many authors have produced evidence, for many plants, of an approximately linear relationship between solar energy absorbed and dry-matter produced, over a crop growing season (Biscoe and Gallagher, 1977; Monteith, 1977; Charles-Edwards, 1982; Gosse et al., 1986; Rosenthal et al., 1989; Russel, 1993). Models make use of this observation, but with some differences in detail, including:

Incident solar energy, or more exactly global radiation (GR), is measured or else calculated using simplified or precise models (Durand and Legros, 1981).

Photosynthetic active radiation (PAR) is about half of the global radiation (VarletGrancher et al., 1989).

Theoretically it requires only 8.4 photons to reduce a

molecule of C02. The actual output of the reaction is close to 17 x 10-9 kg C02 joule- 1 (Goudriaan et al., 1985; Whisler et aI., 1986).

Maximum gross photosynthesis (Mgp) is usually derived from PAR e.g. fro,mthe following formula in the SOYGRO model (Wilkerson etal., 1985); Mgp

= 1.67 * PAR - 0.0173 * PAR2

(3)

where Mgp is g CH 20 m-2 day-1, and PAR is MJ m-2 day-1. This relationship is not linear, but is valid on a daily basis.

80 Legros et al.

Net photosynthesis (Pn) corresponds to the amount of carbohydrate formed after

deduction of the biomass consumed in maintenance-respiration and in the respiration involved in growth.

Maintenance-respiration can be regarded as

proportional to the mass of the plant and also a function of temperature. Growthrespiration is the amount of CH 20 utilised in producing the energy to maintain an enzyme pool compatible with growth (Brisson, 1987) and is proportional to the allowable biomass production. In practice, respiration is difficult to model (Moorby, 1985) and so is often neglected or, more exactly, it is assumed to be globally proportional to the biomass. In this case, it cannot feature in models and instead is taken into account indirectly by reducing the efficiency of the conversion of energy into biomass. In other words, one passes from PAR to Pn, thus (Jones and Kiniry, 1986); Pn = EFP * PAR

(4)

where EFP is the photosynthetic efficiency. This efficiency is of the order of 2.5 to 5 g of dry matter per MJ of PAR absorbed. Its exact value is rather difficult to specify for various reasons (Ruget et al., 1991): (i)

We have just seen that the idea of efficiency, as it is presented, implies some

(ii)

approximation of its impact on respiration. In other words, it varies with growth respiration, that is with the level of biosynthesis, or again, with the date and stage of development; The efficiency depends on the amount of energy received at the leaf surface

(iii)

which can be at saturation especially at midday. The saturation threshold, beyond which energy is squandered as heat, is not the same for all plants; Efficiency varies when going from leaf to crop canopy. During crop growth, the surface covered by leaves increases and the average illumination of each leaf decreases. Leaves saturated by light are thus less numerous and consequently the average efficiency increases. However, a threshold can be reached beyond which leaves are too numerous to be sufficiently illuminated and the efficiency of the canopy decreases;

(iv)

All calculations of efficiency are made on the above ground parts without taking roots into account, whereas the root biomass/above-ground biomass

(v)

ratio varies in the course of the cycle; Experimental determinations are difficult to make. They must show very precisely the conditions under which they have been made, taking into

81

Crop models: problem of climate change

account, or not, reflected radiation, length of measurement period, etc. (Gosse et al., 1986). Energy interception by vegetation is only partial, especially if there is poor soil cover

by the crop, and this must be taken into account. Generally, light interception is estimated as a function of leaf area, which is commonly expressed as Leaf Area Index (LAI). LAI represents the relationship between the total leaf surface and the corresponding soil surface and can reach 4 or 5 or more. It is not always easy to calculate (Sinoquet and Andrieu, 1993). Application of the BEER-LAMBERT law gives, for example (Jones and Kiriny, 1986);

v = VO * (1

(5)

- exp(-b * LAI))

where VO is the incident value, defined below, V is the intercepted value, and b is a coefficient dependent on the spatial arrangement of leaves; it reaches, for example, 0.65 for cereals (Varlet-Grancher et al., 1989) and 0.90 for sunflower (Quinones et al., 1990). This equation can be inserted at different levels within the calculation

sequence and V can represent different things. For example, it is used to pass from 'PAR' to 'absorbed PAR' or from 'maximum photosynthesis' to 'gross photosynthesis'. Figure 2 shows which options have been used in different models. Even if the basic concepts are unchanged, the intermediate results of the calculations can have different definitions and significance. Confusion abounds. Use of the BEER-LAMBERT law implies a continuous cover. In practice, for the law to be verified, it is at least necessary that the leaves of the different plants begin to touch (closed canopy). Leaf growth

There are at least two ways of calculating leaf area. The first method is to determine the fraction of biomass reserved for leaf growth, but this is difficult for several reasons. Firstly, the division of the assimilates between roots, stems, leaves and fruits varies over time.

It is thus necessary to consider the phenological stage.

Further, this division varies also as a function of stresses and roots suffer much less water stress than leaves.

82 Legros et al.

Place of LAI impact in the process of simulation

Main steps'

e

Simplification

GLOBAL RADIATION

ARKIN an~ RITCHIE 1973

I

EPIC WILLIAMS et al. 1989

l-

rL ~

~ CERES.M~ize

CERES. Maize

............... p1R

PHOTOSYNTHETIC ACTIVE RADIATION

PAR

'y6_____

JONES and KINIRY

I

absorbed:...... 11 _ _ _ _ _ _ _ _--,

...1

I

e I

FEDDES 1985

SOYGRO

I

GPh max GROSS PHOTOSYNTHESIS GPh

I,------,t

I

CERES. Maize

~ RESPIRATION

NET PHOTOSYNTHESIS

Figure 2 Energy to biomass conversion.

SORKAM ROSENTHAL et al. 1989

I

83 Crop models: problem of climate change

Finally, having determined the leaf biomass, it is necessary to estimate leaf area which depends, in principle, on the bulk density and the average thickness of the leaves. In fact, this leads to the introduction of a supplementary parameter in the model. For example, in CORNGRO, one uses: ilLAI

= ilBIOmF * k

(6)

where ilLAI is the increase in leaf surface area, ilBIOmF the increase in leaf biomass and k the area per unit mass (dm2) of leaf per gram of dry matter. The calculation, made over periods of 24 hours, removes variations in the assimilate contents which occur during the course of the day and can represent up to 20% of the leaf dry-matter for wheat (Baldy, 1973). The SWACROP model, adapted for grassland (de Jong and Kabat, 1990), works on the basis of a similar approximation. The second method for calculating leaf surface area and obtaining it directly and empirically starts from heat units. These have been discussed above. We will see in the second part of this paper that this procedure can be justified. There are also mixed methods, because in the calculation of ilLAI in equation 6, the coefficient k can be a function of the temperature sum rC days) (O'Leary et al., 1985). Root growth Root growth in annual plants is not well understood and models simulate it in a rather arbitrary manner.

In principle, the following factors need to be taken into

account: (i)

environmental constraints (the presence of an obstacle to downward

(ii)

genetic factors (the rooting depth which the plant could attain unaffected by

(iii)

conditions which influence growth (the amount of assimilates provided by

penetration, or of a water table, a low pH, etc.); external constraints; general form of the root system); photosynthesis and temperature). In this way, relationships are established between total root length and accumulated energy (PAR), or between total root length and thermal time in DC days (Vincent and Gregory, 1989). Water is equally involved, at least under definite conditions. For example, for wheat, the rate of root growth is probably directly related to the soil water potential (Bouaziz and Bruckler, 1989).

84 Legros

et al.

To account for root growth, the values of two parameters are often provided: 1.

Depth reached by the roots by day i; this is often calculated as a function of

the thermal time in degree days, but at times it depends only on the number of days (Chopart and Vauclin, 1990), and

2.

Root density in each horizon; the sum of root densities in all horizons takes

the value '1 '.

This parameter does not always have an obvious physical

meaning. Models do not generally take note of root diameter; all are assumed identical.

In this hypothesis, root density is mixed with the mass or the

surface area of the roots in each horizon, expressed as a relative value. As an example of the calculation of root depth, the approach of Borg and Grimes (1986) is summarised by the following expression: RD = ROm * [0.5 + 0.5 * Sin(3.03 * DAP / DTM - 1.47)]

(7)

where RD is the current rooting depth, ROm the maximum rooting depth, DAP the current day after planting, DTM the days to maturity and Sin is in radians.

Root

density can be calculated in at least three ways: (i)

The genes controlling growth are dominant. The root system is supposed to increase in a pre-determined manner (sometimes homothetically i.e. in keeping the same form), until a maximum size is reached after a fixed time, expressed in days or heat units. Root density is thus given as a simple mathematical function taking into account the depth reached by day i. This approach is adopted in the MOBIDIC model (Leenhardt, 1991) and in CORNGRO. Thus, there is no coupling between the root dynamics and the biomass reserved for their growth (Voltz and Rambal, 1987);

(ii)

The strategy of the plant is dominant. The plant roots develop as a function of the available biomass and priority is given to the exploitation of the most favourable horizons (e.g. the wetter horizons). Further, in the EPIC model the roots explore the horizons as a function of the amount of water in them. In SOYGRO, growth is inhibited if the horizons are dry.

In such models the

distribution of the roots can be distinctly different from that which is usually characteristic of the chosen plant;

85 Crop models: problem of climate change

(iii)

Root growth is related linearly to leaf growth by a simple proportionality coefficient. This is the concept of 'morphological homeostasis' (Fernandez and McCree, 1991).

A third type of parameter is added in some models, to express the degree of exploration of horizons by roots. In CORNGRO this is the inter-root distance, while in others it is root penetration into the structural elements of the soil. Soil water extraction

It is not our intention to reproduce here the detail of soil water behaviour in the soilplant-atmosphere system, because several reviews of the literature have been devoted to this (Brisson and Voltz, 1991; Leenhardt et al., in preparation). Below we raise some pertinent points. During its cycle a plant will evaporate an amount of water 300 or 400 times its dry weight. The water forming part of the plant is negligible and the plant can be thought of as an evaporating machine.

As a result, the plant water requirements are

functions of two main factors, on the one hand, potential evapotranspiration (PET), which represents climatic water demand, and on the other, the LAI, which indicates what the plant can or should evaporate as water without any physiological or pathological limitation (no disease, soil well provided with water, no root disturbance, PET not abnormally high). Formulae of the type below are used: Tp = PET * (1 - exp(-B * LAI))

(8)

where Tp is the potential transpiration i.e. the demand for water by the atmosphere; PET is the potential evapotranspiration and the B coefficient depends on leaf geometry, and thus on the plant. On the other hand, the availability of water also depends on: (i)

the amount of water present in each horizon;

(ii)

the ease of access to this water

ct.

the root density function examined above,

and also the coefficient of root penetration in the PERFECT model (Littleboy et al., 1989), or the coefficient of the inter-root distance in the CORNGRO

model;

86 Legros et al.

(iii)

the ease with which this water can be mobilised in relation to the soil water

(iv)

the capacity of the roots to extract this water by exerting a suction capable of

potential and factors controlling the rate of water transfer; exceeding the suction of the soil and overcoming the forces of friction (the concept of leaf potential). Thus, one can examine to what extent the demand for water can be satisfied by in

situ sources on functioning.

a daily balance, which clearly is a rather large simplification of plant

For the models ARFEJ, (Rambal and Cornet, 1982), MOBIDIC

(Leenhardt, 1991) and ACCESS II (Kling, 1993), one proceeds as follows. Firstly, the flux of maximum plant transpiration is calculated on the basis of van Bavel's formula, which takes into account the partial closure of stomata: Ta1

=Tp / (1 + ((psil / b)n)

(9)

where Ta1 is the actual transpiration, psil is the leaf water potential, b- is a coefficient with a value of about 6000 and n is a coefficient with a fixed value of 5. Secondly, the water supply is determined as the water that can actually be evaporated. This depends on the soil water status and can be written as the difference between the soil (psis) and leaf (psil) water potentials: Ta2 = (psis - psil) * RD / res

(10)

where Ta2 is the actual transpiration, RD is the root density and res is the overall plant resistance to the water flux. The value of psil is unknown, a priori, but it can be calculated by combining two expressions which give Ta. It is necessary to proceed by iteration, taking note that the resulting equation has a complex form, since several horizons are involved, each having a potential and a root density. In practice, one must solve: (11 ) In certain models soil evaporation and plant transpiration are not separated. Maximum evapotranspiration is derived from PET by means of a crop coefficient which depends on the phenophase (Jacquart and Choisnel, 1985).

87 Crop models: problem of climate change

Yield The most commonly used method for estimating yield is to deduce it from the increase in biomass. For example, in EPIC the following formula is used (Williams et a/., 1989):

YLD = HI * AGB

(12)

where YLD is the harvest, HI is the harvest index as given by Monteith (1977) and AGB is the above ground biomass. The harvest index is calculated, day-by-day, on the basis of the modified thermal time, as was shown above, in order to take into account the non-linearity of the relationships between heat and bioaccumulation. The interest in calculating the potential yield throughout the vegetative cycle, and not only at maturity is justified for most plants: grass, forest crops, etc. However, the yield is sometimes estimated differently to avoid the conversion from energy to biomass. For example, it can be related directly to the water behaviour of the plant, as is the case in the YIELD model (Terjung et a/., 1984): Ya = Ym - Ym * (Ky * (1 - AEP / PET)

(13)

where Ya is the yield, Ym is the maximum possible yield, Ky is the yield response factor (from 0 to 1), AEP is the actual evapotranspiration and PET is the potential evapotranspiration, calculated from Penman's formula, adjusted for different crops and their different phenological stages (ct. DTM). The value of Ym corresponding to the formula above can be calculated on the basis of local experience. However, it can also be calculated by an empirical formula based on the crop life-cycle, temperature and the average vapour pressure gradient. Other formulae involve water efficiency in the calculation of yield by establishing the relationship between the amount of water transpired and the amount of dry matter produced over the same period. Finally, there are more sophisticated solutions. For example, grain growth depends at the same time on both the biomass allocated to it, and on that from a reserve pool (Weir et a/., 1985).

88 Legrosetal.

Water stress In some cases, it is possible to relate total above ground dry matter linearly to water use (Innes and Thomasson, 1983). The decrease in water or water stress plays, therefore, a particularly important role in crop models. It is interesting to see how it is calculated and used. Water stress calculation

The method most widely used in the calculation of water stress assumes that the stress is a function of the daily relationship between actual transpiration and maximum or potential transpiration (non-limiting provision of water to the plant): WSI=1-Ta/Tm

(14)

where WSI is the Water Stress Index, Ta is the actual transpiration and Tm is the maximum transpiration. It should be noted here that the stress is maximum when it is equal to 1 for Ta = O. This kind of equation, proposed by Nix and Fitzpatrick (1969) is used by Feddes et al. (1978), EPIC, CERES-MAIZE, and PERFECT (Littleboy et al., 1989), etc. Water stress is occasionally assessed more simply as a direct result of a lack of water in the soil, thus: WSI = 1 - W / AWC

(15)

where W is the actual water content and AWC is the available water capacity. Finally, it is possible to relate water stress directly to leaf potential. However, this potential is very variable over a single day, and its proper estimation by calculation is particularly difficult. Models based on this principle e.g. CORNGRO, must be very well calibrated and validated for these measurements. Water stress action

Water stress limits the increase in leaf size, and eventually the number of leaves, if the plant is in a growth phase. Leaf growth is more sensitive to water stress than photosynthesis. This is why in the model McSTRESS, where the water stress is moderately important, leaf growth stops but not photosynthesis leading to a storage pool for assimilates corresponding to filling of the existing leaves. This pool can be mobilised, if the stress disappears, by making new leaves (McCree and Fernandez, 1989). However, if the water stress is marked, the lifespan of the leaves can be

89 Crop models: problem of climate change

affected. Naturally, once they begin to fall they are no longer sensitive to the effect of a water deficit. The yield of a crop can be affected by water stress. Several authors (Robelin, 1984; Feddes, 1985) found evidence of a relationship between dry matter production and satisfaction of the water requirements of the plant. This can be shown in the form: (16)

(DOM / DOMmax) = a * (AET / MET) - b

where DOM is the organic matter (dry weight), DOMmax is the DOM maximum (optimum yield), AET is the actual evapotranspiration and MET is the maximum evapotranspiration.

Photosynthesis and grain yield are limited in the case of the

CERES-maize model. The effects of water stress on root growth are less obvious. Some authors consider the roots not to be affected because they are closer to the water resource and receive some kind of priority (McCree and Fernandez, 1989).

Moreover, in a

situation of extreme but progressive drought, root development can increase. As the above-ground parts will be more sensitive to water stress, roots seem likely to be relatively favoured. Conversely, some authors take account of the influence of soil water content on root growth.

In the SOYGRO model, dealing with soybean

(Wilkerson et al., 1985; Brisson, 1989), a stress factor is calculated if the water content of the horizon falls below 25% of the available water reserve.

As shown

above, this factor takes a value between 0 and 1: SWr = 4 * (Wi - Wp) / AWC

(17)

where SWr is the water stress on roots, Wi is the water content in horizon i, Wp is the water content corresponding to the wilting point and AWC is the available water capacity. SWr is equal to 1 (no stress) if the water content of the horizon under consideration represents 25% of the maximum reserves. It is equal to 0 (maximum stress) if there is no available water in the horizon. However, it is necessary above all else to note that in the case of a water deficit, even if the roots continue to increase, the root-soil interface decreases. The fraction of effective root area is thus sometimes related to the soil water potential (Fernandez and McCree, 1991). Figure 3 shows the working principles of the EPIC model, which illustrates some aspects of the above discussion. We have used EPIC ourselves (Bellivier, 1993).

90 Legros et al.

I

HU P- - - - - - - - - -HUI------,---...,

(-----------!

HUFH

HUF

.--------------------1----------- ---.-

---------------- ---------.-

---YIELD

Figure 3 Principle of the EPIC crop growth model.

91 Crop models: problem of climate change

ADAPTATION OF CROP GROWTH MODELS TO CLIMATE CHANGE Generalities Climate change will have an effect on crop growth in two ways: i) increasing carbon dioxide

concentrations,

and

ii)

changing

rainfall,

temperature

and

evapotranspiration. In the first case, future developments are quite well understood, even if the equilibrium of atmospheric CO 2 with dissolved carbonates and bicarbonates of the oceans is still poorly known.

Most experts, on the basis of

analysis of trapped bubbles of air in the Antarctic ice, consider that we have entered a period of rapid enhancement of atmospheric carbon dioxide.

The atmospheric

level was about 280 ppmv prior to the industrial revolution and is now about 345 ppmv. It increases annually by about 1 or 1.5 ppmv (Dahlman, 1985). The most pessimistic scenarios predict a doubling of atmospheric CO2 before 2050! However, even if future increases in CO 2 are understood, it will be difficult to incorporate this into crop growth models.

In essence, we will see further that the physiological

consequences of a change in content CO2 concentrations are both numerous and also poorly understood. Furthermore, it should be remembered that the question of changing CO 2 levels has not previously been considered: several crop models have been conceived on the basis that the CO 2 level is unvarying. For other climate characteristics the situation is reversed: crop models have rainfall, temperature and PET as input variables, and so are perfectly adaptable to climate change studies. However, it will be necessary to clearly perceive future evolution of these variables, whi.ch currently is far from being the case. We will see further which alternative solutions have been proposed for making predictions i.e. to make the crop models work with more realistic data (as much as possible) for future climates.

The problem of CO 2 Literature review

The following review owes much to the excellent overview by Warrick et al., (1986). Plants are referred to as either C3 or C4 with C3 plants the most numerous. In these, photosynthesis produces assimilates with 3 carbon atoms. The causative enzyme is mainly ribulose bisphosphate carboxylase-oxygenase usually called RuBP. In the presence of light, this catalyses reactions with CO 2 and 02, the two gases entering

92 Legros et al.

into competition for the same adsorption sites. Oxygenation of this enzyme leads to expulsion of CO2 - this is photorespiration. C4 plants are less numerous and include for example, maize, sorghum, sugar cane and millet.

Photosynthesis in these plants produces assimilates with 4 carbon

atoms. The causative photosynthetic enzyme is 'PEP Carboxylase' resulting in dark respiration rather than photorespiration. PEP has a strong affinity for CO2 by which it is practically permanently saturated. Therefore, an increase in atmospheric CO2 is linked with an increase in plant CO 2 , but without changing the state of this enzyme, and the photosynthesis of C4 plants (except in a dense canopy where the CO2 content can be insufficient to ensure unrestricted functioning of the enzyme system). In contrast, in C3 plants containing RuBP, CO 2 can be fixed in very large amounts which can both increase photosynthesis and decrease oxygen fixation i.e. photorespiration. Under normal climatic conditions, C4 plants are more effective in the conversion of energy to biomass than C3 plants (Gosse et al., 1986) and the growth of C4 plants is more rapid. Their production potential, owing to highly efficient use of intercepted radiation, has been shown to be 45 to 55 t ha-1 yr 1 in the Paris Basin, compared with 32 to 38 t ha- 1 yr 1 for C3 plants. CO2 and plant physiology On the basis of numerous laboratory investigations, the probable influence of an

increase in CO 2 on plant physiology can be summarised as follows: Photosynthesis and biomass.

We have just seen that there is an increase in

photosynthesis in C3 plants. An experimental doubling of the CO 2 content showed that a rise in biomass production, in the absence of limiting factors, can vary by more than 25% in trees to more than 125% in certain crops (Kimball, 1983, adapted by Warrick, 1986). Yields also increase, but to a lesser degree (17% to 14%). In the case of C4 plants, photosynthetic efficiency is not particularly affected by an increase in C02'

Thus, their total biomass does not change much; an average

increase of 14% has been estimated. However, rather curiously, some C4 plants, maize and sorghum in particular, show a net increase in leaf surface area because of CO 2 , This shows that an increase in leaf surface area is not directly related to an overall increase in biomass and justifies, to some extent, the approach of some

93 Crop models: problem of climate change

models which separate leaf growth from that of overall biomass e.g. EPIC - see above. Photorespiration. For C3 plants an increase in CO2 decreases photorespiration and thereby gives a better return on the conversion of CO2 into carbohydrates, but

respiration increases with increasing air temperature. C4 plants are not involved in photorespi ration. Dark respiration. The effect of an increase in CO2 is rather variable depending on the species; respiration decreases in wheat but increases in soya. It also depends

on the variation of temperature between noon and dawn. Transpiration. For all plants, transpiration decreases when the amount of available

CO2 increases. Diffusion of water vapour at the stomatal level is necessary for the opening of the stomata to allow entry of CO2. Thus, increases in atmospheric CO2 cause plants to conserve water or, more exactly, to increase the efficiency of water use. For example, in sorghum, a doubling of CO2 decreases stomatal conductance by about 40%. However, transpiration decreases it by only about 15% because of the reduction in evaporative cooling. This results in heating of the leaves by 1 to 3°C, a phenomenon which itself favours an increase in evaporation through the difference in vapour pressure between leaf and air.

However, the biochemical

mechanism of the stomatal response to CO2 is unknown. It is important to note the role of crassulacean acid metabolism (CAM). This is relevant to the Crassulaceae and the Bromeliaceae. These plants open their stomata at night and absorb and store CO2 in their vacuoles. This CO2 is then utilised during the day, when the stomata are closed and the acidity of the cells increases. This mechanism limits transpiration considerably, and is an adaptation to drought. Pineapples and bananas are examples. Other effects of C02 increase. Several other effects can be identified. In particular,

it is possible to have, at least in some plants, a reduction of the period of development by advancing the time of flowering, and accelerating senescence. In sunflower, this phenomenon can be related to an increase in the production of ethylene. When associated with a general increase in temperature, doubling of C02 represents a gain of 4 to 14 days in the period necessary for the maturation of spring wheat in Scandinavia or Canada (Williams et al., 1988, cited by Parry and Carter,

94 Legros et al.

1989). In the USA, simulations show that it is possible to gain a similar amount of time with winter wheat (Meams et al., 1992).

Discussion C02 stimulates the functioning of a plant and this is verified even when environmental conditions are not very good for growth.

For example, water

efficiency is increased by CO2, even if the plant is in conditions of relative drought, and growth also increases at low light levels. CO2 also enhances the efficient use of nitrogen. Thus, C02 is involved in the functioning of a plant in many ways, and it is difficult to come up with general rules for this behaviour. If the amount of C02 doubles in the course of the next century, plant growth is going to be stimulated. In principle, the demand for water should also increase greatly at the same time as increases in the efficiency of water use. It is not known, therefore, if more irrigation water will be needed than at present, to obtain the same yield.- Further, one asks if, in the long term, there could be adaptation by plants to new climatic conditions which negate these effects. Further, in the understanding of the functioning of plants in the future, it is necessary to take into account the overall changes of climatic parameters: C02 but also cloudiness and rainfall, temperature, and potential evapotranspiration. With such complex situations the importance of crop simulation models cannot be underestimated.

Modelling The developments anticipated for EPIC to account for the problem of C02 (Stockle

et al., 1990) are given here as an example. The role of C02 is encompassed by several equations. Firstly the acceleration of photosynthesis is taken into account by:

(18) where RUE is the radiation use efficiency, C02 is the atmospheriC C02 concentration (ppmv) and b 1 and b2 are coefficients for which values are given or deduced from an understanding of at least two sets of values of RUE and C02. However, a correction must be introduced to allow for the influence of the vapour pressure deficit (VPD) on RUE. This is expressed as:

95

Crop models: problem of climate change

RUEc = RUE - S * (VPD - 1)

for VPD > 1 kPa

(19)

or better RUEc = RUE

if mean daily VPD < 1 kPa

(20)

where RUEc is the adjusted value of RUE, VPD is the vapour pressure deficit (kPa) and S is the radiation-use efficiency sensitivity to VPD (kg m-2 ha-1 MJ-1 kPa- 1); the value of S being plant dependent and occurring between 6 and 8 as a first approximation. Evapotranspiration also needs to be considered. Penman's formula assumes that the crop surface offers no resistance to evaporation. It is thus not suitable, in its initial form, for the study of the effects of C02, which precisely reduces stomatal conductance and evapotranspiration (Perrier,

1977).

The Penman-Monteith

equation is preferable (Monteith, 1975). In this equation, the canopy resistance (Rc) is estimated from the leaf resistance (RI) and the LAI: Rc = RI / (0.5 * LAI)

(21)

Leaf resistance or better, its inverse, leaf conductance (GI = I / RI), should be calculated by taking into account the effect of the vapour pressure deficit because this plays an important role in the regulation of stomatal functioning (Massman, 1992).

However, this should be done on a very fine timescale (minutes, hours),

which poses problems because most crop models work on a daily basis.

This

introduces an equation of the type: Glh(VPD) = Glmax * fh

(22)

where Glh(VPD) is the hourly leaf conductance corrected for VPD, Glmax is the maximum stomatal conductance measured or found from tables (Korner et al., 1979) and fh is a coefficient «1) easily calculated from VPD. Further, it is necessary to correct the effects of the concentration of C02, on leaf conductance.

This is

obtained as an average, from the following equation: Glh(VPD, C02) = Glh(VPD) * (1.4 - 0.4 * (C02 / 330))

(23)

96 Legros et at.

where Glh(VPD, C02) is the corrected hourly conductance in terms of VPD and C02, Glh(VPD) was calculated above (Eqn. 22), and C02 is the atmospheric C02 concentration.

Hourly leaf conductance is the inverse of resistance.

Finally, a

summation is needed to obtain an average daily resistance of the measurements because the crop model under study operates at this timescale. This is done by introducing a weighting factor for each hour depending on net radiation, which gives particular importance to the middle hours of the day.

Consequences of evapotranspiration.

changes

in

rainfall,

temperature

and

potential

It is not our objective to examine here how rainfall, temperature and PET will vary because of the greenhouse effect. Expert opinion suggests that the temperature will increase by between 1.5 to 3.0°C per century. Rainfall is difficult to predict because of the large differences between regions.

Evapotranspiration will be affected in

different ways as a function of climates and ecosystems.

Anticipated variations

range from -20% to +40% (Rosenberg et al., 1989). Crop models can be used to determine the agronomic or ecological consequences of these predicted changes.

The method generally used consists of subjecting the

current climate data to transformations which serve to simUlate the future situation. These transformations are of two types. Firstly, one can perturb an average value of the parameter of interest. For example, let us assume that the daily temperature is increased by 2°C. Alternatively, one can assume that the variability in the climate is increased. For example, it is believed that the temperature of the Nordic regions will increase more strongly in winter than in summer (Choisnel, 1991). The equation which follows allows the construction of simUlations on this basis (Mearns et al., 1992):

x =a * m + b * (xo - m)

(24)

where x is the new value of the affected climate parameter, m is the average current value of this parameter, a is a coefficient which modifies the mean, b is a coefficient with modifies the variance and Xc is the initial value of the parameter. Coefficient b can be fitted, such that the initial variance is doubled or tripled. However, this equation sometimes leads to negative values, which must be reset to zero. e.g. for rainfall. These changes are then imposed on the mean. This type of approach gives

97 Crop models: problem of climate change

a definite problem: the modification of one parameter, without affecting the others, is not necessarily valid. It is hardly feasible that an increase in global radiation will not affect other climatic characteristics. Simulations organised in this way form a kind of simple sensitivity analysis of the model, but do not give very accurate results. Some climatic changes, however minor, also have important agronomic consequences as threshold effects. For example, the winter kill of seeds or plantlets can disappear if the temperature is raised slightly.

Simulation on the basis of 20 or 30 years of

measurement, serves to determine, statistically, the reduction in risk.

GENERAL CONCLUSIONS This review of crop growth models is still incomplete and not all aspects have been covered. An example is vernalization, which is the need for a cold period to ensure germination of particular seeds.

Also not discussed in detail is the effect of

photoperiodism i.e. the consideration of day length on growth (Oelecolle et a/., 1985; Bonhomme et a/., 1991).

Finally, we have assumed that plants grow in an

environment without insects, disease, weeds, etc. However, this analysis has led to some final thoughts.

There are numerous

interesting models, perhaps more than 30, corresponding to as many research teams around the world. These models, with their different options and detail, have a common objective; to reproduce the functioning of a plant.

Moreover, the

modellers often integrate, in their simulation programme, submodels which have been tested and which have been internationally accepted (Feddes, 1978, 1985; Ritchie, 1991). Each element of the working of a plant system is necessarily reproduced in a very simplified form. The equations proposed can appear similarly simplistic in some cases. However, we should not draw a negative conclusion from this. It is necessary to take into account the objectives of the simulation, the quality of the input data, the relative precision of the different compartments of the model, the results of sensitivity analysis and equally importantly, the accuracy of the end results. It is impossible to cover all these questions here. Moreover, the existence of such models implies that many field and laboratory studies have been conducted, which are necessary to define the weights of the factors affecting growth and development in different plants.

98 Legros et al.

ACKNOWLEDGEMENTS The authors would like to thank Mrs. Nadine Brisson, Mr. Maurice Cabelguenne, Mr. Serge Rambal, Mr. Marc Voltz and Mr. Jacques Wery for their assistance in bibliographic references and the editors for help with the text. The work reported here was supported by the European Union under contract EV5V-CT92-0129.

THE EFFECTS OF CLIMATE CHANGE ON IRRIGATED SOILS: WATER RESOURCES AND SOLUTE LEACHING

c. Ramos, A.L. Lid6n and A. Rodrigo Departamento de Ec%gfa, Instituto Valenciano de Investigaciones Agrarias, Apdo. Oficial, 46113 Moncada, Spain.

INTRODUCTION The probable effects of climate change on soils were summarised by Wild (1993): a)

The rise in temperature will result in an acceleration of the rate of soil processes such as mineral weathering, oxidation of organic matter and water evaporation;

b)

The increase in atmospheric C02 will produce greater crop yields that in turn will increase the rate of return of plant residues to soil and this will tend to counterbalance the faster mineralisation rates of soil organic matter;

c)

The regional patterns of rainfall and evaporation might change, but the effects of these changes on soils are difficult to predict at this time;

d)

The sea level is expected to rise at a rate of about 0.6 cm year 1 during the next century, and this will have serious effects on soils in coastal areas.

Most reviews on the effects of climate change on soils have considered only natural systems and dryland agriculture (Bouwman, 1990; Scharpenseel et al., 1990; Mooney et al., 1991; Anderson, 1992), probably because vegetation and crop performance in these soils depends heavily on climate.

In irrigated soils this

dependence is lessened because the water supply is regulated through an artificial system of water resources (surface and groundwater). Dregne (1990), in a study on the impact of climatic warming on arid region soils, suggested that climate change should have little effect on irrigated soils, although he considered a slight increase in organic matter decomposition rate to be of some importance. This idea is probably based on the fact that irrigated soils are typical of intensive agriculture where manmade differences in soils can mask those resulting from small climatic changes. NATO ASI Series, Vol. 123 Soil Responses to Climate Change Edited by M. D. A. Roonsevell and P. J. Loveland © Springer-Verlag Berlin Heidelberg 1994

tOO Ramos etal.

In an attempt to assess the effect of temperature on soils, Dregne (1990) studied

soils in the Texas High Plains, in areas with the same physiography and similar parent materials and rainfall, but with a difference in the annual mean temperature of a few degrees centigrade. The soils between Big Spring and Amarillo, about 400 km apart, matched these requirements, and several of them belong to the same series and textural class. Unfortunately, almost all land in this area is cultivated and the variability imposed by the cropping practices masked the effect of differences in temperature.

Unger and Pringle (1981), in a comparison of soils near Amarillo

(Texas), found the organic matter content of five soils in adjacent counties to range from 1.4% to 2.2%. These results illustrate the difficulty of assessing the influence of climatic effects in isolation, since we cannot know whether the observed variation in organic matter is natural or the result of different managements. Irrigated soils depend on the available water resources and, since climate change will have an important impact on the global and regional water cycle, we think it is appropriate to examine some of the likely effects of the atmospheric C02 increase and global warming on these soils. Table 1 illustrates the importance of irrigated land.

Moreover, in arid and semi-arid countries there is a trend to increase the

surface area of irrigated land because of their higher yields per unit area compared with rainfed soils. Thus, the significance of irrigated soils for food supply is often greater than that indicated by their surface area alone. Table 1 Cultivated and irrigated land in some Mediterranean countries (FAD, 1988). Cultivated land {106 hal

Irrigated land(1) {%}

Egypt

2.6

100

Morocco

8.5

15

Turkey Spain (2)

27.9

8

20.3

17

France

19.5

6

3.9

29

12.2

25

Country

Greece Ital~

(1) relative to cultivated land (2) data for Spain are for 1989

Irrigated soils are also important because they allow more stable yields than rainfed soils. Stan hill (1986) reported that inter-annual variations of irrigated wheat yield in

101

Climate change on irrigated soils

Egypt had a variation coefficient of 7%, similar to that found in England, whereas in Israel, the rainfed wheat yields in different years had a variation coefficient of 26%. In this paper we will review the possible effects of climate change on irrigated soils by looking at impacts on water resources for irrigation, on the salt balance, and on nitrate leaching. We will not consider potential changes in land use resulting from an increase in yield due to C02 enrichment, not that resulting from new geographic thermal and moisture limits for growing crops. These latter effects may be more important than the changes in the pedogenic processes, but their quantitative prediction is difficult (Anderson, 1992; Parry, 1992). CLIMATE CHANGE SCENARIOS

On a global scale, a mean warming of 1.8°C is expected by the year 2030 when atmospheric C02 will be approximately doubled (Mitchell et al., 1990), assuming there are no changes in the global rate of emission of CO 2 ,

In relation to

precipitation, changes of about +7% to +15% have been reported, although confidence in these predictions is lower (Schneider, 1992). For southern Europe, where irrigated soils are common, Mitchell et al., (1990) predicted changes in climate using three models (Table 2). All three models predicted similar increases for mean temperature in the warm and cold months, but the simulated changes in precipitation varied between models. However, there was a tendency to predict a slight increase in precipitation in winter months and a decrease of about 10% during the summer months. In a study on climate change in Europe, Lough et al. (1983) predicted that global warming will result in a decrease in inter-annual variability of precipitation in winter for southern European countries. In summer this variability will decrease in some countries (Portugal, Spain and most of France) and increase in others (most of Italy and Greece). An additional expected effect of global warming is a rise in sea level estimated to be 20 ± 10 cm above the present level by the year 2030, and 30 ± 15 cm by the year 2050 (Warrick and Oerlemans, 1990).

102 Ramosetal.

Table 2 Changes in temperature and precipitation from pre-industrial time to the year 2030 in southem Europe (35 to 50 oN; 100W to 45°E) estimated by three climate models (after Mitchell et al.,1990). Model(1)

A Precipitation (%) Months

CCC GDFL

UKMO

DJF

JJA

DJF

JJA

2 2 2

2 2 3

5

-15

10

o

-5

-15

(1) models are from: the Canadian Climate Centre (CCC), the Geographic Fluid Dynamics

Laboratory (GFDL), and the UK Meteorological Office (UKMO). EFFECTS OF CLIMATE CHANGE ON IRRIGATED AGRICULTURE Before considering the impact of climate change on soils we will discuss its effects on irrigated agriculture, since the two are related. It has been proposed that, with well-irrigated crops, a doubling in the atmospheric C02 could produce an increase in yield of about 34 ± 6% (mean ±95% confidence interval) in C3 plants and about 14 ± 11 % in C4 plants (Kimball and Idso, 1983). The effects of increased C02 interact with those of temperature (Idso et al., 1987), but since plants experience a wide range of temperatures during a growing season, the effect of increasing the C02 will depend on whether the plants are growing near their optimum temperature (Rosenberg et al., 1990). In a review of 18 studies, Kimball (1985) observed that the effect of rising C02 on plant growth (expressed as dry-matter) was greater for plants under water-stress than for those which were irrigated: a 43% increase in growth in the well-irrigated plants against 76% for the water-stressed ones.

Thus, the beneficial effects of

increasing C02 on plant growth and yield would be relatively greater for dryland agriculture than for irrigated crops. However, some of these increases in plant growth might be counteracted by the losses resulting from pests because of more favourable conditions for their development (Cammell and Knight, 1992). Another important factor is the effect of C02 on water use efficiency, defined as the amount of transpired water required to produce a unit mass of plant dry-matter.

In a review of this topic, Allen (1991)

concluded that doubling C02 will result in an increase in photosynthetic rates,

103 Climate change on irrigated soils

growth, yield and water-use efficiency for C3 crop plants; he also estimated that a reduction in precipitation in irrigated areas might reduce the land surface available for irrigation, and that the competition for water between agriculture and the urban and industrial sectors would increase.

Other authors have suggested from the

available evidence that doubling of atmospheric C02 will also double water use efficiency (Fleming, 1991). We should also consider the effects of climate change on the surface extent of irrigated soils. Adams et al. (1990) estimated the effect of C02 doubling on the surface extent of irrigated land in the U.S.A. using two climate models: the Goddard Institute of Space Studies (GISS) model and the Princeton Geophysical Fluid Dynamics Laboratory (GFDL) model.

The increase in the area of irrigated land

predicted by the GISS model was 1.4 million ha and that by the GFDL model was 3.8 million ha. These increases represented increments of 8% and 21 %, respectively, relative to the base irrigated surface extent. In Japan, it has been estimated that an increase in mean temperature of 3SC on the island of Hokkaido would allow the surface area of land for growing irrigated rice to be more than doubled because of the disappearance of the severe risk of crop losses due to frost (Yoshino et al., 1988).

In a recent study on the possible effects of climate change on irrigated

agriculture in the Murray-Darling basin in Australia, Pigram et al. (1992) concluded that: 1) Increased temperatures may lead to a shift of crops in the south to higher elevations with cooler temperatures; 2) Irrigated rice (a crop with large water consumption) may be reduced or moved to areas where an increase in water resources may occur; 3) Irrigated cotton could move further south if the growing season is extended there as a result of an increase in temperature; 4) A decrease in rainfall in the southern parts of the basin could cause a decline in dairy production from irrigated pasture; 5) Pests and diseases may increase if the climate becomes warmer and wetter. Effects on water resources

Changes in water resources for irrigation will undoubtedly be one of the main impacts of climate change on irrigated soils, due not only to the variation in total water resources in a given region, but also resulting from increased demands in the urban and industrial sectors.

However, the impact of climate change on

104 Ramoseta/.

precipitation, runoff and evapotranspiration, which control water resources, are quite difficult to predict at a regional scale (Schneider, 1992). In a review of methods for evaluating the hydrologic impacts of global climate change at a regional scale, Gleick (1986) proposed the application of watershed water balance models using climatic input data from predictions by General Circulation Models (GCM). An assessment of the effect of climate change on water resources for irrigation is important because agriculture is the main consumer of water resources.

For

example, da Cunha (1989) reported that in southern European Union countries, 70 to 80% of water consumption was by agriculture, and Waggoner and Schetter (1990) gave a similar figure (81%) for the U.S.A.

Peterson and Keller (1990) made an

estimate of the effect of climate change on water resources for irrigation in the U.S.A. They predicted that the greatest impact of the warmer climate would be in the western states where farmers would find it difficult to maintain present levels of irrigation.

In contrast, in the eastern states farmers would profit from a modest

expansion of irrigated land. In many cases the competition for available water between agriculture, industry and the urban sector will increase irrespective of whether water resources increase or decrease. This situation will stimulate the use of available irrigation technology and new developments in this field to improve application efficiency (the ratio of field applied water to evapotranspiration) and the uniformity of water application. There are considerable possibilities in this area (Table 3). Irrigation efficiency depends not only on the irrigation system, but also on soil type, crop and factors such as irrigation rates and scheduling that constitute good irrigation management.

An example of the effect of crop type on field irrigation

efficiency was given by Oster et al. (1986) who, in a study of many irrigated fields in the Imperial Valley (California), found that irrigation efficiency ranged from 20% to 79% and that the variation between fields was less than that between crops. recent review of factors related to irrigation efficiency is given by Wolters (1992).

A

105 Climate change on irrigated soils

Table 3 Estimated application efficiency ranges for various irrigation systems (after Keller et al., 1980).

Irrigation system

Seasonal average efficiency range (%)

Surface Furrow

40 to 50

Border strip

50 to 80

Sprinkler Periodic move lateral

60 to 75

Fixed lateral

60 to 85

Center pivot

75 to 85

Lateral move

80 to 88

Trickle Point source

65 to 92

Line source

60 to 85

Effects on leaching of salts

A favourable salt balance in irrigated soils is critical for long term agricultural productivity. However, large amounts of salt can enter the soil in irrigation water. For example, in semi-arid areas where irrigation is frequent, the salt content of irrigation water can be about 0.4 to 0.8 kg m- 3 , assuming a normal irrigation rate of 7000 to 12000 m-3 ha-1 yea(1. This results in salt inputs of 2.8 to 9.6 t ha-1 yea(1 of which only a minor fraction is taken up by the crop.

Leaching of these salts is

required to prevent salt accumulation from becoming harmful to plant growth. The effect of climate change on the salt balance in irrigated soils can be produced by: 1) a reduction in available water resources and consequently in the applied irrigation water, which would result in an insufficient leaching fraction (the proportion of applied water that flows past the bottom of the root zone); 2) a change in the amount of rainfall in the wet season when leaching is likely to occur. With respect to the first point, we have already discussed the possible effects of climate change on water resources; leaching requirements for different crops and irrigation water qualities have been summarised by Rhoades and Loveday (1990). In relation to the second point, in irrigated semi-arid regions it is common for the

106 Ramos etat.

soluble salts accumulated in the soil profile during the irrigation season to be partially or totally removed by leaching in the rainy season when evapotranspiration is low. An example of this is shown in Figure 1 which presents the soil chloride levels throughout a period of a year and a half in an irrigated citrus orchard; chloride content was found to be a good indicator of soil salinity since they were both highly correlated. It is apparent that from May to December in both years there was an accumulation of chloride during the irrigation season (which in these unusual years extended until late autumn) and then a decrease in the winter months.

El El

.... Ii

1;

...'"

80 60 40

80

a

....2 ....'..."

l1li

.~

60 40

20 0 500 400 300

I

t .)

200 100

~ M J J A SON 0 E F M A M J J A SON 0

months Figure 1 Changes in chloride content in the 0 to 100 cm soil layer with time in a

citrus orchard in the Valencia area, Spain. Open and closed symbols refer to different nitrogen fertiliser treatments.

107

Climate change on irrigated soils

If climate change affects rainfall distribution during the winter months and extreme values increase, this would favour runoff and therefore a decrease in effective rainfall and leaching. Rainfall distribution is important to leaching since it has been shown that, for a given amount of applied water, small applications are more effective than continuous ponding (Hoffman et al., 1981). A review of soil salinity problems in different countries was made by Szabolcs (1986), who also assessed the impact of climate change on the development of soil salinization and alkalinization in Europe (Szabolcs, 1990). Effects on nitrate leaching

Nitrate leaching is an important process· in irrigated soils because of large applications of N fertilisers. In addition, leaching is strongly influenced by the deep percolation term in the soil water balance and changes in climate can affect many of the N cycle components in these soils.

Although it is difficult to generalise, we

present here an example of a simulation exercise to assess the effects of climate change on nitrate leaching. The NLEAP model (Shaffer et al., 1991) has been used because it does not have excessive data requirements. This model allows three types of use: screening, monthly, and an event-by-event analysis.

We used the

monthly budget mode. Our objective was to study the effects of an increase of 3°C in the monthly mean temperature on nitrate leaching using the climatic data for average, dry and wet years obtained from the agroclimatic station at the Instituto Valenciano de Investigaciones Agrarias (these data can be considered typical of the Mediterranean climate). The soil was a loam with 2% organic matter and a mineral N content before planting of 25 kg ha- 1 in the top 30 cm and 60 kg ha- 1 in the 30 to 150 cm layer. The cropping sequence was potato (January to June) and fallow the rest of the year. The average climatic data were taken from the period 1980 to 1992 and the dry and wet years from this period were 1981 and 1989 respectively. These data are summarised in Table 4 and the main crop management data are shown in Table 5. Three simulations were made for each type of year: 1) using the given climate data; 2) increasing the mean temperature by 3°C with no change in Class A evaporation (Eo); and,

108

Ramosetal.

3) increasing the temperature by 3°C and assuming an increase in Eo proportional to the increment in potential evapotranspiration estimated from temperature data (Hargreaves and Samani, 1982). In the three types of year, the irrigation amounts and dates were based on farmer practice in the area: during January and February, a maximum of two irrigations depending on the rainfall; in each of the following months (March, April and May) a weekly irrigation unless a rainfall event ~ 30 mm occurred in the previous week. The simulation results are presented in Figure 2. In each type of year, the imposed changes in temperature and evaporation resulted in negligible variations in the monthly and annual nitrate leaching. A striking result was that more nitrate was leached during the growing cycle in the dry year than in the average and wet years. This probably resulted from greater applications of irrigation water in the dry year, which increased percolation losses. The result would have been different if, during the dry year, fewer irrigations had been applied because of water restrictions. Another point to note is that nitrate leaching in the autumn months increased from the dry to the wet year, indicating the importance of rainfall in this part of the year. In these simulations, temperature and evaporation were varied but not rainfall; it is likely that changes in rainfall amount and seasonality arising from climate change would affect nitrate leaching losses in a way similar to the results shown here for different types of year.

Other effects The sea level rise predicted by climate change studies could have an important impact on irrigated soils in coastal areas by creating anoxic conditions, and by raising the water table near the soil surface, thus establishing a net upward flux of water from the phreatic layer causing soil salinization. Groundwater quality in these areas would also become poorer as a result of sea intrusion, and irrigation with these waters might cause soil salinization and sodification. As an example of the importance of sea level rise in some cases, it has been estimated that a rise of 1.5 m would inundate about 20% of all farmland in Bangladesh. In Egypt, this rise would result in the loss of 20% of all farmland, especially of the most productive land (UNEP, 1989).

109 Climate change on irrigated soils

Table 4 Climate data used in the NLEAP model simulations. Average

T

Rainfall

(0C) January

9

February

11 12 13 17 21 23 24 22 18 13 11 16

March April May June July August September October November December Mean

T Rainfall Eo (OC) {mm} {mm}

Eo

{mm} {mm}

36 38 23 38 25 23 10 20 46 64 76 33

53 66 109 127 157 185 213 188 137 99 61 56

9 8 14 14 16 21 21 22 21 18 14 12 15

358 1332

Total

Wet

D~

3 8 10 81 8 10 0 43 20 15 0 5

T (0C)

46 48 99 112 155 175 208 165 130 94 48 94

8 11 13 14 17 21 24 26 21 19 15 13 15

182 1363

Rainfall

Eo

{mm} {mm}

13 61 43 33 30 18 5 25 170 23 229 48

30 97 114 137 145 196 229 196 104 117 64 66

689 1254

Eo: Class A tank evaporation

Table 5 Crop management data. Month

Fertiliser

Irrigation Average year Dry year Wet year

January February March April

Mal'

Amount Type(1)

{mm}

{mm}

{mm}

kg N ha- 1

36 36 79 89 104

36 71 114 104 104

36 0 36 71 104

0 176 167 167 0

A B B

(1) A: ammonium sulphate; B: ammonium nitrate.

Potato harvest was on June 10th and thereafter it was left fallow. Some of the predicted changes in climate, such as frequency of drought and heat stress, should have less impact on irrigated soils than on dryland and natural soils because of the buffering effect of the applied irrigation water on the moisture and

110

Ramosetal.

temperature regime. Other effects, such as changes in flooding frequency (Rind et al., 1989) might affect irrigated soils more severely since in many cases irrigation is located on river terraces.

These effects, although considered important for

agricultural soils, are very difficult to predict quantitatively (Parry, 1992).

«I

200 r-------------------------------------~

.c:

175

::

150

-ti II)

125

.........

..1:1 ~

100

~

75

Z I

'" ~ ~ ';;;; ~

WET

CJ

350

_

335

~ 337

Total N-Ieached (kg N/ha)

SI)(UL. 1 SI)(UL. 2 SI)(UL. 3

50 25

o

175 150

r-~_UUL~~~~~~_ _~~WU~~~UL~~

AVERAGE

Total N-Ieached (kg N/ha)

CJ ~

_

285 282 258

125 100 75

Z

50

'"oZ

25

I

o

~~~~~~~~UL~~=-~~~~~~~~~

175

DRY

Total M-Ieached (ka M/ha)

150

CJ

331

_

322

~ 329

125 100 75 Z

50

'"o

25

I

Z

o

~

__ E

_W~~LLUL~~

F

M A

M

____

J

~~wu~am~

J

A

____

SON

~

0

month Figure 2 Monthly values of nitrate leaching estimated with the NLEAP model for three types of years and three climate change scenarios (simulations 1,2,3).

111

Climate change on irrigated soils

CONCLUSIONS As pointed out by Brinkman (1990), human activity will continue to have a greater effect on soils than potential changes arising from future climate change. However, since human influence on irrig?ted soils is dependent on the availability of water resources, climate change can have important effects on these soils, although quantification of this effect is difficult because of economic

~nd

social factors.

Sombroek (1990) suggested that climate change predictions apply to very large areas, and that at a smaller scale the individual characteristics of each soil and its management should be considered. To do this, local data are necessary. Many uncertainties still remain about the effects of climate change on the degree and variability of rainfall at a regional scale.

Thus, the establishment of soil,

irrigation and water resource policies should carefully consider irrigation practices and the development of large irrigation projects or expensive works for water transfer between regions. These have potential long term effects when the likely impact of climate change on water resources and demand in different regions is taken into account. Parry (1992) suggested that adaptations to counteract the undesirable effects of climate change on crop production will have to be investigated at two levels; these are also applicable to the effects on irrigation: a) the enterprise level, and b) the regional and national policy level. In the first case, the people involved would be farmers and irrigation engineers. At the policy level the response would be at a larger scale, cover longer time spans, and would involve regional and national governments in matters such as new irrigation areas or water transfer between hydrological basins.

ACKNOWLEDGEMENTS We thank A. Sahuquillo, J. Mataix, G. Cruz and A. Yeves for providing some references used in this work, and Elena Lorenzo for making the climate data available.

MODELLING THE EFFECTS OF CLIMATE CHANGE ON THE HYDROLOGY AND WATER QUALITY OF STRUCTURED SOILS

A.C. Armstrong1, A.M. Matthews1, A.M. Portwood1, T.M. Addiscott2 and P.B. Leeds-Harrison3 1ADAS

Soil and Water Research Centre,

Anstey Hall, Maris Lane, Trumpington, Cambridge CB22LF, UK. 2AFRC Institute of Arable Crops Research, Rothamsted Experimental Station, Harpenden, Herts. AL52JQ, UK. 3Cranfield University, Si/soe Campus, Si/soe, Bedford MK45 4DT, UK.

ABSTRACT

The behaviour of cracking clay soils is characterised by the interaction between the development of soil macroporosity and the soil moisture status. If climate change alters the magnitude of the soil moisture deficit, that change may alter the importance of macroporosity in soil hydrological behaviour. This interaction is illustrated by models of three differing degrees of complexity: a simple partition model, a redistribution model, and a relatively complex model. These models can also be used to indicate some of the impacts of possible changes in the climate on the components of the hydrological cycle.

One of these models, when used to

predict the leaching of nitrate suggested that, despite higher nitrate concentrations in response to greater mineralisation, the nitrate leaching loads may be lowered in a drier climate, because of reduced through-drainage. Lastly, a simple water balance model is used to suggest the impacts of climate change on the hydrology of a managed wetland. Here, the greatest impact was on the recharge volumes needed to maintain wetland status.

NATO ASI Series, Vol. I 23 Soil Responses to Climate Change Edited by M. D. A. Rounsevell and P. 1. Loveland © Springer-Verlag Berlin Heidelberg 1994

114

Armstrong et al.

INTRODUCTION

Changes in the global climate due to an increased CO2 content of the atmosphere will also change the local climate.

Agriculture will be immediately affected, and

studies, for example by Bennett (1989) and Parry et al. (1988), have contemplated the way in which patterns of production will alter.

Equally, climatic changes will

affect the hydrological cycle (e.g. Climate Change Impacts Review Group, 1991). This paper is concerned primarily with identifying the impacts of possible changes to the climate on the hydrology of soils, and in particular with structured soils, which are primarily, though not exclusively, clay soils.

In the UK these are nationally

important, forming approximately 60% of the agricultural soils.

Their major

differentiating characteristic is the presence of seasonally variable macropores formed largely by shrink-swell. Hence, the degree of macroporosity is related to the moisture content of the soil. A major interaction between soil hydrological behaviour and climate change will be the degree of development of the macroporosity: if such soils become drier, then macroporosity may become more important; if soils become wetter, then macroporosity may become less important.

We have the intriguing

prediction in the UK of wetter winters and drier summers, in which wetting/drying cycles may become more pronounced, so profoundly affecting the hydrological behaviour of clay soils. The hydrological behaviour of clay soils will depend largely on the degree of soil structural development and the associated flow in the macropores.

It has been

observed that impermeable clay soils generate runoff very rapidly through macropore flow, particularly when drained for agricultural production (see e.g. Armstrong and Garwood, 1991; Harris et al., 1993). Macropores are thus responsible for both the rapid transmission of water from the soil surface to the drainage outlets (Hallard and Armstrong, 1992), and for the rapid movement of solute and chemical loads. In modelling climate change impacts on the hydrology of macroporous soils we can no longer assume that the properties of the soil will remain unchanged from their current condition. Not only will structure vary dynamically during the year, as at present, but it may also change in response to the changed climate regime. There is thus the possibility of major interaction between climate and soil, in which the dominant soil hydrological property also varies in response to climatic events.

115 Hydrology of structured soils

This paper reports some first attempts at modelling the hydrological behaviour of soils as they respond to climatic sequences. It is a report of work in progress, so different components are at different stages of development.

Nevertheless, an

underlying thesis is that we must handle the problems in a variety of ways, and at a variety of degrees of complexity, depending on the kind of problem we are attempting to address, and the kind of resources that are appropriate (both data for inputs and computer time for deriving the model results). Thus, when our models are to be included in spatially distributed systems, they need to be relatively rapid in operation and require easily available parameters, whereas when models are developed to understand the basic physics, they can be made as complicated as we deem necessary and may demand many parameters.

This paper thus seeks to

demonstrate some of the different ways of modelling the hydrology of clay soils, and to identify ways in which these different models can be used.

MODELLING WATER FLOW IN CRACKING CLAY SOILS

Modelling of water flow within cracking soils requires a conceptual subdivision of the soil into two components: the peds, and the intervening cracks and structural components generally termed macropores. It is then possible to identify the types of flow corresponding to these two components of the soil: matrix flow through the peds, and macropore flow through the cracks and fissures. We have so far identified and investigated three models which adopt such a representation, and which exhibit three different levels of complexity: •

a simple partitioning model (SWAP) which attempts to split the incoming rainfall

•

a simplified model of soil and solute movement (SLIM) which includes an element

into the two components of matrix and macropore flow.; of vertical redistribution of the incoming water down the profile; •

a complex physically based model of water and solute movement (CRACK) which attempts a complete description of the movement of water in cracked clay soils.

These three models have been applied to a common dataset, based on a single site and a common year, so that comparisons can be made between their relative performances.

116 Armstrong et al.

Site details: the Tippler's catchment. The modelling studies have been undertaken in the context of a small catchment, the Tippler's sub-catChment which covers 56 ha at the southern end of the Swavesey study site approximately 15 km northwest of Cambridge (Harris and Parish, 1992). The catchment (Figure 1) comprises six fields in intensive arable production. For the year considered in this study (1991/92) five of the fields grew winter wheat with one in winter barley. The soils of the catchment are a uniform pelo-stagnogley of the Denchworth series, derived from thin clayey drift over Jurassic clay. They are highly structured with large cracks developing during dry summers and seasonal waterlogging occurring due to the slowly permeable subsoil.

The site is

underdrained with mole drains at 2 m spacing linking into an intensive pipe drainage system. The outfall from the catchment is instrumented with a pre-formed flume, associated head recorder and water sampler. Data available for model validation included patterns of water and nitrate loss, crop performance and soil mineral nitrogen data.

Meteorological data were available from the ADAS Experimental

Husbandry Farm at Boxworth, approximately 5 km from the site.

The climate inverse problem A major problem in such studies is the derivation of an acceptable set of climatic inputs to represent changed conditions. Changes to the mean climate cannot immediately be translated into daily weather values suitable for modelling, and so some scheme must be used to create these values; this is the climate inverse problem (Kim et al., 1984). The approach adopted throughout this paper has been to adjust current weather data to meet altered means. Daily temperatures are all altered by the same absolute amount, and daily rainfall by the same percentage amount. This has the advantage of being simple, but carries with it the assumption that the structures within the data are similar.

Thus, to produce, for example, a

greater rainfall amount, it assumes more rainfall each rain day, not more rain days. A common scenario was also adopted in which temperature was assumed to increase by 3°C and rainfall by 10% except in the summer when rainfall was assumed to decrease by 10%. This scenario is within the range of climate change predictions considered by the Climate Change Impacts Review Group (1991). Evapotranspiration was recalculated for each day from a series of monthly coefficients derived from climatic data.

These gave a relationship between

117 Hydrology of structured soils

temperature and evapotranspiration, and were then used to identify the change in potential evapotranspiration with change in temperature which could be calculated without reference to the 'Penman' variables.

Ii.!!DIMs Catchment

200 1

6001

meler'

Figure 1 The Tipplers catchment.

10001

118

Armstrong et al.

Simple partition model (SWAP) An initial model (SWAP: Soil WAter Partitioning), not previously described, is concerned with calculating the soil moisture balance without attempting to identify the redistribution of water within the profile. In practice, this is the least developed of our models. Results are still preliminary and should be considered indicative of the potential of such a simple model. The model considers the soil as a single store, and partitions the incoming daily rainfall amounts into three components: •

Infiltration

•

Macropore flow

•

Soil surface runoff

The partitioning of rainfall into these components depends on the rainfall intensity. This must, however, first be estimated where, as is usual, only daily rainfall data are available.

The total rainfall amount is then replaced by a synthetic triangular

hyetograph (rainfall distribution graph), in which the total rainfall amount is distributed around an estimated peak rate (Figure 2).

r Infiltration

T

T

Figure 2 Partition of rainfall into soil water components. Statistical analysis of hourly rainfall data from several sites in the UK has indicated that it is possible to estimate this peak rainfall intensity from the daily rainfall total. Simple regression analysis of rainfall totals greater than 10 mm day-1 has suggested that peak rainfall intenSity is about 28 to 30% of the daily rainfall rate. Although the regression relationships are weak (r = 0.5 to 0.6), and should in the long run be

119 Hydrology of structured soils

replaced by a more sophisticated analysis, they do offer a way forward in the short term. The resulting hourly hyetograph may then be compared to various thresholds for the partition of the rainfall amount into its various components. This may exceed one of two thresholds: the matrix infiltration, and the macropore infiltration capacity. If the rainfall peak is less than the matrix infiltration capacity, then all the rainfall infiltrates. When rainfall exceeds that capacity, the excess 'overflows' and becomes macropore flow. If the maximum rate of acceptance for macropore flow is exceeded, then the resultant excess becomes surface runoff. However, these two thresholds are not in themselves constant soil properties, but are functions of the soil state, and in particular the degree of soil structural development.

For example, Harris et al.

(1993) present evidence that different cultivation practices lead to dramatic changes in the behaviour of clay soils, which are a reflection of the different degrees of macro pore development. We must then estimate the two critical thresholds from the soil state. Both infiltration capacities, the matrix infiltration capacity, lx, and the macropore infiltration capacity, 1m , can be modelled as a function of the soil moisture content, e. Initially we assume a simple linear function: Ix = a + ~e

FC > e > er

(1a)

FC > e > er

(1 b)

where FC is field capacity, a,

~,

a' and

P' are constants and 8r is a residual

moisture

content below which the parameters do not change. For matrix infiltration we can identify the parameters by reference to the normal infiltration equations.

However, for the macropore component, the physical

significance of the parameters may be clear, but their measurement is problematic. In practice, the degree of macropore development can be related to the moisture content using the Coefficient of Linear Extensibility (which defines the shrinkage of the peds) and the hydraulic behaviour of the cracks. The residual value, 8p may thus be related to the point at which the soil ceases to contract and the structure becomes stable.

Although such links might be developed from detailed physical

120 Armstrong et al.

investigation, for the present we use a linear relationship with empirical parameters and p'.

~

The SWAP model thus partitions incoming rainfall between the three components: recharge to the soil moisture store, macropore flow, and surface runoff (Figure 2). Rainfall, up to and including the maximum infiltration capacity, becomes matrix infiltration, any excess above it up to the maximum macropore infiltration rate becomes macropore flow, and any further excess above that becomes direct surface runoff. The model then calculates a soil water balance for the soil considered as a single uniform store, with matrix infiltration and evapotranspiration being added to, and removed from, the store, respectively. Macropore flow, however, is assumed to move directly to the drainage systems, and is therefore unavailable to the soil storage. Because it considers the soil as one store only, this model is inappropriate for soils whose hydrological properties vary strongly with depth. The model has been applied to the data from the Tippler's catchment and shows moderately good representation of the observed hydrology (Figure 3). We also used the model to predict the impacts of possible climate change, using the standard climate change scenario (an increase in mean temperature of 3°C and rainfall increased by 10% in winter and decreased by 10% in summer). The model was run for a period of 10 years, and the mean contributions to each of the flow components calculated for both current and changed climates (Figure 4). This shows that as a result of the change in'climate, the amount of actual evapotranspiration will increase. However, because of the greater soil moisture deficit in summer and the increased intensities of rainfall, the amount of water entering the soil and passing as drainflow is reduced. Hence the contribution of through drainage relative to macropore flow is decreased (Table 1). The SWAP model predicts that a major impact of climate change will be the reduction of drainflow, but no major increase in the macropore or surface flow. It thus suggests that a consequence of climate change will be an increase in the contributions of flows generated near the surface of the soil, with consequent implications for water quality issues. The important advantage of this simple model, however, is that its Simplicity renders it very fast in operation. It is thus possible to embed this sort of model within a GIS for spatially distributed modelling.

121

Hydrology of structured soils

!!i

! o

(0

~

""

o

"" ""

.. ..

,

..

,

·· ·.'

z o

.,

.

a. oSl -'

~ ~

"'iI

"'x z ..

"00

oen

::::!~

.,en u

_

ex 0.-

o~

U

0::

:t VI

~~ :.'.'

o

~~ 1>0 -om

:t --

.'

f.'.'

(0

!!!

':

f

Q: ~«

~~ o

\

~

o

e

I/')

ww

0

N

NlY~

II')

ww

0

I/')

0

"i " ~ 13Y WUJ oriS

~

I/')

N

'I

0

N

0

~

0

0

I

ww MOU

~I

OO$?~ ~ I I ww MOl.:!

III

«

0

XI

Figure 3 Results of the SWAP partition model for the Tippler's catchment under current conditions.

The diagram reports in order (from top to bottom): the rainfall

inputs (RAIN), the actual evapotranspiration (AET), the soil moisture deficit (SMD), and the partition into surface runoff, macropore, infiltration and drainage; and the two moisture-dependent controlling parameters, maximum matrix infiltration (1m) and maximum macropore infiltration (Ix).

122 Armstrong et al. Curr.nt

Chang.d

Figure 4 Partition model (SWAP): components of the hydrological cycle. Results for mean of 10 years for current and changed climates (increase in mean temperature of 3°C and rainfall increased by 10% in winter and decreased by 10% in summer).

Table 1 Components of flow (mm) modelled by the SWAP model for current and changed climate.

Component

Current Climate

Changed climate

575 387

600

Actual ET Drainage

Rainfall

458

100

61

Macropore flow

44

39

Surface

42

41

123

Hydrology of structured soils

A simplified model of soil water movement (SLIM)

The simple partition model, SWAP, considers the whole soil as a store, and does nothing to consider the distribution of water within the soil profile. More complex models consider this redistribution of water up and down tj1e profile subject to the twin processes of downward gravitational drainage and upwards evapotranspiration. There are many models that would serve this purpose, but in the work reported here we have used the SUM model of Addiscott and Whitmore (1991) because there is an extended version of the model, SACFARM, which includes N-mineralisation and crop uptake, so that changes in soil mineral nitrogen and losses of nitrate by leaching can be computed on a daily basis. These are described by Addiscott and Whitmore (1987).

The model has simple parameter requirements and should be

applicable to a range of soil types. The water balance components of the model are very simple. The model divides the soil into layers and the water within each layer is divided into mobile and immobile categories, wm and wr' which are estimated from the soil moisture characteristic as follows: (2a)

except in the top layer where (2b)

Immobile water: (2c)

where 9s, 90 .33 and 915 are the volumetric water contents at saturation, 0.33 and 15 bar respectively and d is the layer thickness, usually 50 mm. Downward movement of water is controlled by a rate parameter

(x,

which is the

proportion of the mobile water in a layer which moves down to the layer below each day. The value of

(x,

which is estimated from a regression on the percentage of clay

in the profile, acts as a surrogate for the hydraulic conductivity of the soil, having the properties of a velocity (Addiscott and Whitmore, 1991). Upward movement of water

124

Armstrong et al.

and solute is possible in both phases in response to evaporation from the soil surface.

Rapid movement of water to the drains can take place if the top layer Water reaching the surface in excess of its water holding

becomes saturated.

capacity thus becomes macropore flow. Nitrate in the top layer is leached via this mechanism, but there is assumed to be no interaction with soil layers further down the profile. The model also includes a crop growth model which simulates dry-matter production, N-uptake and root development of winter wheat. Degree-days of soil temperature in the topsoil is the driving variable behind the crop development. The root distribution is used to determine the fraction of water and solute in each layer which can be taken up by the crop. This fraction decreases exponentially with depth. The model was run against a single field in the Tippler's catchment at Swavesey for the year 1991/92.

The soil physical parameters required by the model include

percentage clay and bulk density in both topsoil and subsoil and the water content at three points on the soil moisture characteristic curve. The parameters used were provided by the model for a clay soil, with the exception of bulk density for which measured values were available. The clay content of the soil indicated a value of 0.54 for the rate parameter a.. Rainfall data used for this and for subsequent runs are shown in Figure 5.

35 30

E 25

.s

-

'iij 20 c:

'i!!

15

>-

~ 10

o

5 O~~~~

09/10/91

__

09/11/91

~~~~~-L~~u--u~~.u-L~~~~~

10/12191

10/01/92

10/02192

12103192

Figure 5 Daily rainfall patterns at ADAS Boxworth 1991/92.

12104/92

13105/92

125

Hydrology of structured soils

The hydrological component of the results of SACFARM (Figure 6) show a number of drainflow peaks where rainfall has exceeded the infiltration capacity of the top layer and the excess water has passed as bypass flow to the drains. This peaky response is typical of cracking clay soils in which the major component of flow is rapid movement through the macropores. In addition, the model also predicts two periods of steady flow produced by water moving slowly downwards to the drains through the soil matrix. This behaviour is not generally observed in soils of high clay content where the water tends to be retained in the fine pores of the soil matrix. The model predictions could be made more realistic by reducing the value of the rate parameter a below the modelled value. This would have the effect of reducing the amount of water moving to the drains via the soil matrix and increasing the likelihood of rapid bypass flow occurring. Figure 6 also shows the SACFARM results for 1991/92 using the changed climate dataset. The pattern of drainflow is similar to that obtained in the original run, but with one additional peak in March. The peaks are generally larger with the changed climate data, reflecting the increased rainfall, with the exception of the two peaks on 30th March and 14th April which are smaller than the original run because of the increased evapotranspiration causing greater drying of the topsoil. The amount of water passing to the drains as a steady flow via the soil matrix is also reduced for the changed climate model run again because of the increased drying of the profile by evapotranspiration. Totals of rainfall, PET and predicted drainflow for the modelling period are given in Table 2.

~ (01

10

~8

8

6

~

4

~ .;

2

'-'

0

r.. ~

0 0911(),)1

- - - Current - - - - Changed

09/11191

10112191

10101192

10l02I92

12103192

Figure 6 Daily drainflows predicted by the SACFARM model.

12104192

13105192

126

Armstrong et at.

Table 2

Yearly totals of rainfall, potential evapotranspiration and predicted drain

outflow for SACFARM simulations of Tippler's catchment 1991/92.

Current

Total rainfall

Total PET

Total drainflow

(mm)

(mm)

(mm)

298.4

190.9

52.9

325.8

260.1

42.5

climate Changed climate These results show a change very similar in direction to those predicted by the SWAP model (Table 1), although they cannot be directly compared because they refer to part of one year only. They show that despite an increase in total rainfall, this is more than offset by the increase in evapotranspiration, with a consequent decrease in the quantity of drainflow. One important effect predicted by the model, which results from the increase in air temperature (and consequently soil temperature), is the more rapid development of the winter wheat crop. This is a direct consequence of the use of degree-days to drive the crop growth. Table 3 shows the crop dry-weight and root depth at 31st March for the two runs. The increased root development predicted for the changed climate run results in an increased depth of drying of the soil profile which contributes to the reduction in drainflow over the period from March onwards. The crop growth component of the model does not, however, consider the reduction in crop performance as a consequence of summer soil moisture deficit stress. The interactions between climate change and crop performance is explored by Legros et al. (1994) and between climate change and soil nutrient dynamics by Bradbury and Powlson (1994) in this volume. Table 3 Crop development up to 31st March 1992.

Crop dry weight (t ha-1) Depth containing 63% of roots (cm) Normal climate Changed climate

8.9

37

10.6

41

127 Hydrology of structured soils

Complex physical models

A further level of complication in the conceptualisation of the processes is given by the CRACK model described by Jarvis and Leeds-Harrison (1987).

This was

developed as a hydrologic model for structured cracking clay soils to provide simulations for single rainfall events, but has been subsequently modified to model multiple wetting and drying events over a complete annual cycle. The hydrology describing the model is illustrated schematically in Figure 7. Water reaching the soil surface may infiltrate, but when the rainfall intensity (1) exceeds the infiltration capacity of the peds at the soil surface (2), it begins to enter the cracks (3). The amount of water flowing in the cracks (5) may be reduced by sorption losses to peds within the soil (4).

If a water table in the cracks rises above drain depth, drain

outflow occurs (6).

~--.'-~_;-;;-r_l

"~I:. J,: 5

?::.:r,.::::::......:.----.,

,.

.~ :--,{

Figure 7 Components of the hydrological cycle identified and defined in the CRACK model. In the CRACK model, soil structural status is dependent on the moisture status of the soil. The volume of cracks is a function of the difference between the initial moisture

128 Armstrong et al.

content in each layer and the 'field capacity' moisture content. It is assumed that the vertical soil movement is small compared to swelling and shrinking in the horizontal dimension. Water flow in cracks is initiated when the uptake rate into peds at the soil surface is exceeded by the rainfall intensity. As soon as crack water flow starts, peds take up water through internally wetted surfaces as well as through the soil surface. The amount of wetted surface area increases with time, so that the normal decay in infiltration rate below a wetted surface is compensated for by the generation of new wetted surfaces. The height of the water table in the cracks is controlled by the balance between the input at the soil surface and the loss due to ped water uptake within the soil and drain outflow. The rate of rise or fall is also a function of the crack volume in the layer in which the water table is located. If the water table rises to the surface, any further rainfall input is immediately lost as surface run-off. Water infiltrating into peds in any given layer is assumed to be stored entirely within that layer. If the layer reaches field capacity, any excess is transmitted to the layer below. If the bottom layer reaches its field capacity moisture content, any further excess is assumed to be lost as deep seepage. The CRACK model was run for a single field in the Tippler's catchment over the 1991/92 season for the current climate and then for the changed climate. The total rainfall, PET and drainflows are given in Table 4 for both climates.

Table 4 Summary of rainfall, PET and drainflow for the current climate and the changed climate. Total (mm) Rainfall PET Drainflow

Current climate

Changed climate

289.1 190.1

261.1

34.7

23.3

316.2

Table 4 can be compared with the results predicted from SACFARM (Table 2) and from the long term averages predicted from the SWAP model (Table 1). They show a similar pattern to that of SACFARM, in which the increase in winter rainfall is more than offset by the increase in evapotranspiration, with a subsequent decrease in drainflow. The model predicted daily drainflow (Figure 8) and the soil moisture deficit (Figure 9) for the period. Using current weather data, the start of the simulation was very dry,

129 Hydrology of structured soils

so there was no response in the drainflow until the winter. The initial soil moisture deficit (SMD) was high (over 100 mm), but decreased steadily with the small rainfall events in the autumn until stabilising to a more constant level for nearly two months. The large rainfall event in January (41 mm) produced a peak drainflow of 5.2 mm day-1 with water flowing down the cracks. The SMD rapidly decreased by a third to 29 .2 mm.

There was a much smaller drainflow two weeks later and the SMD

decreased slowly thereafter, but later rose again in the dry latter half of the winter. There was a prolonged period of flow in March and April which included six peak events ranging from 0.1 mm to 8.1 mm. In this period, the SMD decreased from 14.6 mm to 5.9 mm as the soil wetted . The total drainflow for the period was 34.7 mm and no runoff or seepage was predicted by the model. The sequence of hydrographs predicted by the CRACK model are more characteristic of the rapid response to rainfall found in cracking clay soils, and in particular do not show the period of prolonged flow predicted by SLIM. It is thus considered that the CRACK model gave a better representation of site hydrology than the SLIM model.

9 8 7

6 5

Current

- - - - Changed

4

3 2 1

o

09flOf9!

A

09fllf91

10112191

1010 1192

J 0102192

At

12J03f92

~ 12I04f92

13105f92

Figure 8 Daily drainflows predicted by the CRACK model. The

changed

climate,

with

rainfall

increased

by

10%,

and

potential

evapotranspiration (PET) increased by the relationships described earlier, resulted in a smaller drainflow and a larger SMD.

This result is in agreement with those

predicted by the simple partition model. The total PET increased from 190 to 261 mm in this simulation period .

The large rainfall event in January produced a

drainflow of 5.2 mm in April, which is just smaller than that for the current climate, and the SMD was slightly larger because of the increased rainfall (45.1 mm).

130

Annstrong et al.

09110191

.... ·u

1.=

41---.

.-~ ~ e

10112191

10/0 1192

10/02192

12/03192

12/04192

13/05192

20

41

"C

0

09111191

0

'-'

·8

Ul

40 60

- - -

80 100 120

Current

- - - - Changed

Figure 9 Soil moisture deficit predicted by the CRACK model.

The prolonged period of flow in March and April included only three peak events of 3.2 mm, 6.7 mm and 5.7 mm, which were all smaller than for the current climate. The reason for the decrease in flow events was that more water was taken up by evapotranspiration. The SMD decreased from 28.2 mm in the middle of March to 15.2 mm, which is a much smaller fall compared with the current climate. The total drainflow for the period was 23.3 mm, which is a third less than for the current climate. Modelling soil chemical dynamics The three models discussed here are representative of different levels of complexity when looking at soil water movement. Moving from simple to complex models, we improve the conceptual model, but at the cost of increased parameterisation. The models are however consistent in describing a decrease in drainflow as a response to an increase in evapotranspiration. As well as predicting the soil water balance, the models can be used to assess the quality of water leaving a site, and so begin to estimate the impacts of climate change on water quality. The SAC FARM model was developed primarily as a solute leaching model, and so contains routines for N-transformations and leaching, and so can be used to speculate on the possible impacts of climate change on nitrate leaching patterns in a structured soil. Figures 10 and 11 show the nitrate leaching results for the current and changed climates. The patterns of leaching are similar, the difference being in the amounts and concentrations of nitrate-N leached. Mineralisation of soil organic-N to ammonia

131

Hydrology of structured soils

and the subsequent nitrification to nitrate-N are driven by soil temperature. The increased soil temperatures in the changed climate therefore result in more nitrate-N being available for leaching. This is reflected in the higher predicted concentrations, suggesting a potential increase in short-term pollutant loadings. However, because of the lower drainflow rates (see section 4.2) the total nitrate-N load leached is lower for the climate change scenario compared with the current climate (Table 5) .

....... 800

~7oo

5600

- - - Current

§

- - - - Changed

~

v

Z

500 400 300

~

o z

I

:1

III 200 I I 100 oL--------o~~~~~==~~LJ~~~~=wUL_L 09/10191 09/11191 10/12191 10101/92 10/02192 12103/92 12104192 13105/92

Figure 10 Nitrate-N concentrations in drainage water predicted by the SACFARM model.

~

2.5

~ .:.I:

2

'-'

Z

~

o

Z

1.5

-

- - Current

- - - - Changed

1

0.5 0~--------------~~~~~----~·~~~~1

09/10/91 09111/91 10/12191 10/01/92 10/02/92 12103/92 12/04/92 13/05/92

Figure 11 Nitrate-N loads predicted by the SACFARM model. An important component of running these models is the crop management decisions, which are required as external inputs to the models.

However, we cannot easily

predict how crop management decisions will be affected by climate change. Thus, there is an important interaction missing in the current models, concerned with relating farmer actions to changed weather patterns and soil conditions.

132 Armstrong et al.

Table 5 Totals of predicted drain outflow and nitrate-N leached for SACFARM simulations of Tippler's catchment 1991/92. Total drainflow (mm)

Total N0 3-N leached (kg ha-1)

Current climate

52.9

16.0

Changed climate

42.5

9.3

Impacts on managed wetlands As a last example, pertinent to the theme of the workshop, we present some ideas about the impact of climate change on managed wetland ecosystems.

The

management of wetlands for ecological aims is being increasingly considered, both to maintain wetland status and to recreate wetlands that have been lost. One component of studies of recreated wetlands has been a soil water balance model, which includes an explicit consideration of the interaction between the field water table levels required to achieve the ecological aims and the ditch water levels imposed to effect these ends. Armstrong (1993) has shown how a simple model (DITCH: Ditch InTeraction with Channel Hydrology) can be used to examine the degree to which the desired effect can be achieved simply by setting ditch levels. He showed that where soils are conductive, it is possible to create high water tables in the field centre, but that this requires large volumes of water in the summer to feed the evaporative demand. Where however, the soil has a low conductivity, it is not possible to feed water from ditches to the field centre, and so the field dries out even though the ditch levels are maintained at a high level. Armstrong et al. (1993) were then able to show that the results from the model compared well with observations from the Broads Environmentally Sensitive Area (ESA) in eastern England (MAFF, 1989). To do this, it was necessary to extend the model to consider soils with a hydraulic conductivity varying with depth, using the analysis originally proposed by Youngs (1965), based on the Girinsky seepage potential. Where the logarithm of hydraulic conductivity, K(z), varies linearly with height above the drain, z: (3)

133

Hydrology of structured soils

where

Ko

is the hydraulic conductivity of the saturated soil at z=O, and

P is

a

constant, the flux through the drains is given by:

(4) where Hm is the water table height at mid-drain spacing, Hw is the height of water in the ditches and D is half the distance between the ditches. It is worth noting that this model represents variations in soil structure with depth, by modelling empirically the effect on the hydraulic conductivity. The DITCH model was able to predict the observed behaviour of. the soil water regime in the Broads ESA, where the water level was being maintained close to the ground level.

The model can also be used to compare the impacts of climate

change, by applying the same procedures as previously described to generate a synthetic weather sequence (Figure 12). The model identifies the need for a supply of water in the summer months to offset the evaporative demand from the vegetation. Results were calculated for a situation where the water level in the ditches was maintained at 45 cm below the field level to maintain field wetness throughout. The summary of fluxes from a 12 year run (Table 6) shows that the impact of climate change is to increase this evaporative demand considerably, and so increase the amount of recharge dramatically.

The conclusion is that the economic cost of

maintaining wetlands in a warmer climate is likely to be considerable. Table 6

Mean annual fluxes for the ditch water table model under the current

climate and a simulated changed climate. Rain

Actual ET

Recharge

Drainage

Current

594.0

614.0

278.0

254.0

Changed

624.0

781.0

394.0

211.0

5.1

27.5

41.5

-16.9

% Change

CONCLUSIONS The studies presented in this paper represent a variety of ways of predicting the hydrological behaviour of soils, and consequent solute movement, under conditions of a changed climate. Several of these studies are still in progress, and the results reported here are preliminary. They do however, show there is no single way that is likely to be applicable to all situations.

Depending on the amount of information

134

Armstrong et a/.

available for input, and required from the output, either a very simple model or else a very complex model may be appropriate. Modelling, as always, is a matter not just of getting the equations together in a computer program, but a matter of skill, experience and judgement on behalf of the modeller to match the right degree of complexity to the issues addressed. It is suggested that where water resource issues are of major interest, and where multiple model runs are required for spatial analysis, then the simple partition model is an adequate representation. The parameters however are semi-empirical, and may be difficult to identify except by the use of pedo-transfer functions. By contrast, the detailed simulation of soil moisture behaviour by the CRACK model gives mechanistic predictions of the impact of climate change, but requires a level of parameterisation that can only bEi undertaken for tightly controlled experimental sites. In addition to its computing resource need, the parameter need prevents the use of CRACK for anything except single point research studies. An intermediate level of complexity is offered by the SLiM/SACFARM model, using readily obtainable parameters, which may then be of widespread applicability. However, these models are not developed specifically for cracking soils, and thus require further refinement before they can be applied with confidence.

ACKNOWLEDGEMENTS We are grateful to our various collaborators who have provided models, data, or ideas for the various components, in particular to Mr J. Cochran, of the Meteorological Office, for suggesting the scheme for revised PET estimates. We are grateful to our colleagues, notably Graham Harris, Steve Rose, and Adrian Muscutt of ADAS SWRC for making data from the Swavesey catchment available. Financial support for this work from the UK Ministry of Agriculture, Fisheries and Food (MAFF),

The

European

Commission

(contract

EV5V-CT92-0129),

and

the

AFRC/NERC joint initiative on solute transport in soils and rock, is gratefully acknowledged.

•

c

I

"'.-j

0

LI\

0

0 0

0

0

0

:x:

•

.

\.,

J

I

f I"

M

r- A T

M -

r J -1

J A

SON

0I

The diagram reports (in order from top to bottom) the rainfall

table, and the drainage fluxes.

;;: Ii

8:

Orslnage flux

- - Current ••••••. Changed

Ditch level

inputs, the actual evapotranspiration, the percentage flooded area, the ground water

revised climates.

eM

"tl "tl

Ground level

o

Figure 12 Results of the DITCH model applied to the Broads ESA for current and

I

LI\

gJ

1 iA ;:jA'"M~~= "_

.s=en

•> ~ ..•

ao

E

LIJ

...

~

.....

"tl

0:

~

&

~

Q.

~ Cil

(f)

0~ Q.

~

~

THE POTENTIAL IMPACT OF GLOBAL ENVIRONMENTAL CHANGE ON NITROGEN DYNAMICS IN ARABLE SYSTEMS

N.J. Bradbury and D.S. Powlson AFRC Institute of Arable Crops Research, Rothamsted Experimental Station, Harpenden, Herts. AL5 2JQ,

u.K.

ABSTRACT The response of nitrogen cycle processes to anticipated changes in climate will depend both on the direct effects of these changes (e.g. increased temperature, changes in rainfall amount and distribution, increased atmospheric C02) and on indirect effects such as modifications of land use or cropping pattems prompted by the warmer environment. This paper outlines how climate change will influence different N-cycle processes and focuses on the complexity of interactions between soil, plant and atmosphere. A model of the nitrogen cycle is used to investigate the impact of some climate change scenarios, for the UK, on the organic nitrogen content of soil under two different cropping sequences over a run of eighty years, and to examine how pattems of mineralisation, leaching and denitrification change over the course of the run. The results indicate that, should C02 fertilisation of crops in the field lead to more carbon being returned to the soil, this additional input of C would, for some time, maintain the organic nitrogen (and carbon) content by compensating for the increased rate of decomposition induced by higher temperatures.

Eventually, assuming a O.6°C rise in temperature every

twenty years, the rate of turnover would be sufficient to cause a decline in organic nitrogen content. The greater mineralisation so caused would potentially increase N losses. A change in cropping sequence from continuous winter cereals to an altemating winter and spring rotation led to a greater decline in soil organic N.

NATO ASI Series, Vol. 123 Soil Responses to Climate Change Edited by M. D. A. Rounsevell and P. J. Loveland © Springer-Verlag Berlin Heidelberg 1994

138 Bradbury and Pow/son

INTRODUCTION The activity of microbial populations and their demand for organic carbon substrates as a source of energy and for nutrients, notably nitrogen, phosphorus and sulphur is a major driving force in the cycling of nutrients in the soil. Mineral nitrogen released by the microbial decomposition of organic matter may be taken up by plants, lost to ground and surface waters by leaching of nitrate, transformed to gaseous products by denitrification or volatilisation and emitted to the atmosphere, or re-absorbed by soil micro-organisms. Nitrogen is returned to the soil in organic form in plant residues and animal manures and, in inorganic form, as fertiliser and by wet and dry deposition.

It can also be transferred from

atmosphere to soil and plants through biological nitrogen fixation. Changes in the cycling of nitrogen and other nutrients under the influence of climate change will depend not only on the direct effects of increased temperature, but also on the vast array of interactions which will characterise the response of soils and plants to a changing climate. Some of these interactions are considered below. Superimposed upon these will be the effects of the changes in land use prompted by a warmer environment and also, perhaps, by economic pressures or policy decisions. The latest assessments using coupled ocean-atmosphere models suggest a rate of global warming of 0.3°C per decade (IPCC, 1992). It is thought that global warming will bring a wetter climate to the northern mid-latitudes, but considerable uncertainty is attached to different model predictions of rainfall distribution in specific regions. For the UK, it is suggested that by the years 2010, 2030 and 2050 respectively, summer rainfall may change by 0 ± 5%, 0 ± 11 %, 0 ± 16%, and winter rainfall by 3 ± 3%, 5 ± 5% 8 ± 8% (Climate Change Impacts Review Group, 1991). Similar changes are likely to occur over north-western Europe.

In this

paper, we first consider the possible effects of climate change on individual Ncycle processes and then discuss the results of some simulations from a nitrogen cycle model using different climate and land-use inputs.

139

Nitrogen dynamics in arable systems

N-CYCLE PROCESSES Decomposition of organic matter

Many studies have demonstrated the increase in the rate of decomposition of organic matter with temperature (Ayanaba and Jenkinson, 1990), and a decrease in the rate as the soil dries (Stanford and Epstein, 1974; Orchard and Cook, 1983). Under the anticipated changes in climate, soil moisture content over the course of a season will be altered as a result of changes to the terms in the balance between water added in precipitation and lost through drainage, evaporation or crop extraction. Both the amount of rainfall received and its distribution over time are likely to change, and a change in soil organic matter content could in itself lead to structural changes which will alter drainage patterns. The capacity of the air to hold water vapour increases by 5 or 6% per °C, so a warmer climate will promote greater evaporation (Rosenberg et al., 1989). However, any temperature-driven increase will be modified by changes to other mediating factors in the energy balance such as radiation, humidity and wind conditions, and the magnitude of changes in these factors under global warming is uncertain. Water loss from soil through plant transpiration is additionally controlled by canopy and stomatal resistance and will be influenced by changes in the immediate environment, (such as stomatal closure in response to reduced humidity), and by crop response to C02, for example, in increased rates of canopy expansion.

In the possible

scenario for the UK of warmer, drier summers, the combination of higher temperatures and less moisture could result in no net change in decomposition. However, warm and wet weather in autumn and winter would promote mineralisation and could lead to increased leaching of nitrate. The sudden rewetting of a dry soil commonly leads to a flush of mineralisation as the microbial population undergoes a rapid increase (Powlson and Jenkinson, 1976; Birch, 1960). This effect is likely to occur in the late summer and early autumn, and would further increase the amount of nitrate at risk to leaching. In addition to the direct effects of temperature and moisture, decomposition will also be influenced by the quantity and composition of plant residues being returned to the soil.

Both above-ground production and root growth appear to

respond to an increase in atmospheric C02 concentration (see for example, Lawlor and Mitchell, 1991, for a review of field experiments under such conditions). There is thus the potential for the amount of carbon input to soil to increase if elevated

140

Bradbury and Pow/son

atmospheric C02 concentrations lead to increased dry matter production by crops in the field. An analysis of yields from the Park Grass experiment at Rothamsted, (which has been in grass for several centuries and under experiment since 1856), was unable to identify any upward trend in yield due to C02 fertilisation after the year to year variation in yields due to climatic factors was eliminated (Jenkinson et al., 1994). It may be that background variation was too large and obscured the existence of a small trend, but the study also indicates that if other factors, such as

water stress and nutrient availability, are limiting growth, the response to C02 may be reduced. An increased input of carbon to the soil from roots and other plant residues

would

offset

the

increased

decomposition

induced

by

higher

temperatures and so the organic matter content of the soil would tend to be maintained. However, a rise in temperature will in itself tend to counter increased C fixation by increasing the developmental rate of crops, thereby decreasing the time available for growth and lowering yield (Mitchell et al., 1993). For central England, this decline has been estimated to be of the order of a 6% decrease in yield per one °C rise (Climate Change Impacts Review Group, 1991). On balance, one would expect a slight increase in C inputs if the crop is not limited by other factors, notably water stress. Here again, there is some evidence that restrictions in root growth caused by drought are compensated for by the additional root production induced by elevated C02 (Chaudhuri et al., 1990). Along with a change in the quantity of plant residues, the composition of the material will also be subject to change. Nitrogen is known to be used more efficiently by some C3 species of plants under elevated C02 (Hocking and Meyer, 1991). With carbon fixation either remaining constant or possibly increasing, the C:N ratio of plant residues will widen so that additional soil mineral N may be immobilised during the course of microbial decomposition. Thus, a little less N will be available for crop uptake or loss. In the course of time, re-mineralisation would release this immobilised N and make it available for crop uptake, leaching and other processes causing loss. There are both long-term and short-term questions to address here. Over several decades, what will be the effect of the continued incorporation of wide C:N ratio material on soil organic matter dynamics and how can these be detected? In long-term experiments where cereal straw of wide C:N ratio has been incorporated for a number of years, large increases in biomass carbon and nitrogen have been observed before significant changes in total soil organic matter content could be detected accurately. At one such trial in Ronhave in Denmark the incorporation of barley straw for 18 years led to increases of 5%

141

Nitrogen dynamics in arable systems

and 8% in total organic carbon and nitrogen respectively compared to plots on which the straw was burned, whilst biomass C increased by 37% and biomass N by 46%. So whilst an increase in biomass C of 101 kg C ha-1 from 273 to 374 kg C ha- 1 was significant and could be detected without difficulty, the increase in

1500 kg C ha- 1 in total C was not significant and had to be measured against a background of 28000 kg C ha-1 already present (Powlson et al. , 1987). Data from such experiments also provide an analogy for testing models under current conditions. In the shorter timescale of the three or four years of a crop rotation, will changes in crop residue quality be significant enough to alter patterns of immobilisation and mineralisation, and how will this affect crop nutrient requirements and N loss processes? Denitrification The atmospheric concentration of nitrous oxide is 310 ppbv and is increasing by

0.2 to 0.3% per year, giving a global atmospheric increase of 3 to 4.5 Tg N per year (IPCC, 1992). Current estimates of emissions from cropped and fertilised soils range from 2.4 to 3.7 Tg N20-N per year out of total emissions from all sources of 11 to 17 Tg N (Bouwman and Sombroek, 1990). Emissions from soil are mainly the result of the reduction of nitrate to nitrous oxide by denitrifying bacteria under anaerobic conditions.

As the oxygen content of soil declines,

further reduction of N20 to N2 occurs.

The denitrification process is strongly

related to the moisture content of the soil and also to the availability of carbon substrate, to the nitrate supply and to temperature. Under global warming, any changes in the amount or the timing of denitrification losses will be mediated by changes to the whole soil-plant system. Higher temperatures in themselves will promote denitrification, the more so if allied to higher rainfall giving wetter soils. Such conditions are expected to occur in the autumn and spring in the UK. The loss of fertiliser N by denitrification has been linked to the rainfall received in the three weeks following its application (Powlson et al., 1986). As fertiliser is generally applied between February and May, it appears that the risk of loss is likely to increase in this time period. The presence of plant roots has been shown to increase denitrification by providing readily decomposable carbon substrates in the form of root exudates and sloughed off root material (Smith and Tiedje, 1979). The increase in root growth under elevated C02 could therefore promote denitrification through greater

142

Bradbury and Pow/son

substrate availability. An opposing effect would occur if the bigger root system extracted more water, thereby drying the soil and reducing the number of anaerobic sites in which denitrification could take place. However, elevated CO 2 has also been found to increase water use efficiency so, despite a greater root system, total water use by the crop could remain unchanged.

To evaluate the

overall effect of global warming scenarios on denitrification losses there is clearly a case for using detailed mechanistic models which encompass all the interactions between the opposing influences outlined here. These, in turn, rely on a sufficient mechanistic,

and quantitative,

understanding of the processes and their

interactions; it is by no means clear that this degree of understanding currently exists. As a result of climatic change, and the altered cropping systems which may follow, changes in organic matter content may produce important changes to the water holding capacity and the structure of the soil. These factors in part determine soil aeration and so will influence both the timing and amount of denitrification, and the ratio of N20 to N2 produced. A decline in organic matter content will generally serve to decrease the water-holding capacity of the soil and reduce soil aggregation, giving more frequent occurrences of anaerobic conditions. This may increase the total amount of nitrogen lost through denitrification, but would perhaps tend to increase the proportion of nitrogen lost as N2 compared to N20. Nitrous oxide can also be produced in semi-aerobic conditions during nitrification (Klemendtson et al., 1988), so increasing wetness when nitrification is occurring will tend to increase loss of nitrous oxide. Leaching In current UK climatic conditions, most soils return to field capacity in late autumn, and nitrate leaching ~ainly occurs over the period of winter to early spring. Nitrate can also be lost from the profile by bypass flow through wide pores and cracks which develop as the soil dries. The importance of this mechanism depends on soil type and is more prevalent in soils of high clay content. In addition to clay content, shrinking, swelling and crack formation are also influenced by the bulk density, cation exchange capacity and organic matter content of the soil (Reeve

al., 1980).

et

Conditions leading to greater soil moisture deficits will make soil

cracking potentially more severe particularly if higher temperatures have led to a decrease in organic matter content.

Although mineralisation of N may be slow

143

Nitrogen dynamics in arable systems

under dry conditions, applied fertiliser N would be at greater risk to loss through cracks so the timing of N application to minimise losses may become more critical particularly on heavier textured soils. Table 1 shows the effect of rainfall received after fertiliser application on the recovery of 15N-labelled fertiliser in two contrasting seasons. In 1981, the soil, (a sandy clay loam overlying a heavy clay subsoil with mole drains) was cracked prior to fertiliser application, and rainfall amounting to 113 mm in the three weeks following led to recovery of labelled N in the crop and soil of only 67% at harvest. Presumably, much of the unrecovered fertiliser N was lost by bypass flow through the cracks to the drains. The much lower rainfall of 15 mm in the same period in 1982, when the soil was not cracked, did not have the same effect and 89% of the fertiliser was recovered in that year (Powlson et al., 1992).

Table 1 Effect of rainfall on recovery of 15N-labelled fertiliser at Saxmundham. Year

15N_ labelled

Labelled Labelled N in crop N in soil

3 weeks

after application

N applied at harvest

(0 to 23cm) at harvest

recovered in crop

kg N ha-1

kg N ha-1

kg N ha-1

1981

142

1982

143

kg N ha-1 65.7 100

Rainfall in

Total Recovery labelled N

~Ius

soil %

mm

22.2

87.9

62

113

27.5

127.5

89

15

Warmer, wetter conditions at other times of the year will favour mineralisation. Much of the additional nitrate will be formed at periods when crop uptake is small so a greater quantity will be at risk to leaching. This would be offset by a later return to field capacity in the autumn and the incorporation of wide C:N ratio plant residues which will tend to immobilise soil mineral N. With drier summers, soil cracks may persist longer into the autumn. In this case, iNater leaching through cracks may contain relatively low concentrations of nitrate as, at this time, soil mineral N will be contained within micropores and will not be exposed to water flowing through channels. In the longer term, higher temperatures may increase decomposition sufficiently to lead to a decline in the organic matter content of soil. Eventually, a new equilibrium level may be reached as microbial populations adapt to the new environmental conditions.

In the transition period, the increased

mineralisation of N along with a decline in water-holding capacity (which would

144

Bradbury and Pow/son

accompany a decrease in organic matter), will tend to increase the loss of nitrate by leaching. Volatilisation

Volatilisation occurs through the dissociation of ammonium ions to ammonia gas which diffuses from soil solution to soil atmosphere and finally escapes into the air from the soil surface. Thus, for volatilisation to proceed, there must be a source of ammonium ions near the soil surface. This source may be the excreta of grazing animals, applied animal manures and ammoniacal fertilisers, or NH4+-N released from decomposing soil organic matter.

The extent of volatilisation losses are

dependent on a number of factors including soil pH, soil moisture, temperature and atmospheric conditions.

Volatilisation increases with temperature because the

constants governing the equilibria between NH4+ and aqueous NH3, and between NH3 in solution and in gaseous form are temperature-dependent and promote the flux of gaseous ammonia from the soil to the atmosphere. Volatilisation also increases if the soil is drying as the loss of water from the soil increases the concentration of bicarbonate ions at the drying surface, allowing the release of ammonia (Rachhpal-Singh and Nye, 1986).

Volatilisation is also affected by

windspeed, the transport of NH3 away from the ground surface decreasing its concentration and serving to increase the flux from the soil atmosphere. The factors to which volatilisation is sensitive are subject to considerable spatial variation so changes to the amount of NH3 lost by this process will be very sitespecific. In agricultural terms, the greatest risk of loss comes following fertiliser or manure application. On soils where there is known to be a high risk of NH3 losses, for example, those of high pH, the timing of fertiliser applications may become critical in a warmer climate. Ammonia may also be lost from the crop as the canopy senesces, although estimates of quantities lost in this way vary widely (Schjorring et a/., 1989; Parton et a/., 1988). Nevertheless, higher temperatures will both speed the development

of crops, giving a longer senescent phase, and may also prompt the onset of senescence when crop tissue still contains a relatively high concentration of nitrogen. The potential for NH3 loss from canopies would appear to be increased in a warmer environment.

145 Nitrogen dynamics in arable systems

SIMULATING THE IMPACT OF SELECTED CLIMATE CHANGE SCENARIOS ON N-CYCLE PROCESSES

The SUNDIAL model (SimUlation of Nitrogen Dynamics in Arable Land) has been developed to simulate the major processes in the nitrogen cycle (Bradbury et al., 1993). It considers the addition of N to the soil from inputs of fertiliser, organic manures, crop residues and atmospheric deposition; the transformation of N between organic and inorganic forms by mineralisation and immobilisation, and the removal of N from the system through crop uptake, denitrification, leaching and volatilisation. In addition, the fate of applications of 15N-labelled fertiliser can be followed through soil and crop. In the model, organic matter (carbon and nitrogen) is held in three pools: fresh plant residues (including dead root material), biomass and humus. Mineralisation or immobilisation of nitrogen occurs depending on the relative C:N ratio of each pool. The model was parameterised using data from experiments carried out at Rothamsted using 15N-labelled fertiliser.

By taking

account of a field's soil properties, its previous cropping history and by using appropriate weather data, the model can simulate the amount of organic and mineral N in the soil throughout the growing season and accounts for N inputs, transformations and losses on a weekly basis.

The model can be run over a

number of years using the simulated soil conditions and the crop residues produced at harvest as the inputs to the next year's run. In this way, it can be used to examine how different cropping sequences affect N dynamics. The model was run nominally from 1990 until 2070. Four climate change scenarios were selected from those described by the Climate Change Impacts Review Group (1991) and also outlined in the introduction to this paper. Details of the selected scenarios are summarised in Tables 2 and 3. Scenario A was run with no change in climate. In Scenarios B, C, and D, changes to rainfall and Mean temperature was temperature were introduced every twenty years. increased by 0.6°C, 1.2°C and 1.8°C in 2010, 2030 and 2050. Changes to rainfall were made on a seasonal basis using four three-monthly seasons, (winter December, January, February; spring - March, April, May; summer - June, July, August; autumn - September, October, November). Changes made to rainfall were the same for winter, spring and autumn, but differed in the summer. So, Scenario B gave no change in summer rainfall, but slightly wetter conditions in winter, spring and autumn; Scenario C gave drier summers and much more rainfall in the other seasons; Scenario D gave wetter conditions all year round.

146

Bradbury and Pow/son

Table 2 Climate change scenarios: changes in rainfall. Rainfall (% change) and Year

Scenario 1990 to

2011 to

2031 to

2051 to

2010

2030

2050

2070

Sep-May

0

0

0

0

Jun-Aug

0

0

0

0

Sep-May

0

+3

+5

+8

Jun-Aug

0

0

0

0

Sep-May

0

+6

+10

+16

A

B

C

Jun-Aug

0

-5

-8

-16

Sep-May

0

+6

+10

+16

Jun-Aug

0

+5

+8

+16

D

Table 3 Climate change scenarios: changes in carbon and nitrogen inputs, and in temperature.

Scenarios B,C,D

C input to

N input to

Harvest date

Temperature

soil

soil

(weeks)

(0C)

{% change)

{% change)

1990

0

0

0

+ 0.0

2010

+15

0

-1

+ 0.6

2030

+15

0

-2

+ 1.2

2050

+15

0

-3

+ 1.8

The model runs on a weekly time-step and uses total weekly rainfall and evapotranspiration, and mean weekly air temperature as inputs. Meteorological data encompassing the changes to rainfall and temperature in Scenarios B, C and D were generated in the following way.

From a 30-year run of Rothamsted

meteorological data for the years 1961 to 1990, the distributions of mean weekly rainfall, evaporation over grass and temperature in each of the four seasons were found to be normally distributed, and the standard deviation of each 30-year set of seasonal data was determined.

147 Nitrogen dynamics in arable systems

Using an appropriate incremental step (e.g. 0.1 mm of rainfall or evaporation, or 0.1°C of temperature), the probability of particular amounts of rainfall, evaporation or temperature occurring over the recorded range were calculated.

Weekly

meteorological data were then generated for a run of years by randomly selecting values from the seasonal distribution, given their particular probabilities of occurrence. Every twenty years in the model run, these distributions were shifted, keeping variability constant, by the amounts shown in Tables 2 and 3. No explicit changes in evapotranspiration were made due to the difficulty in interpreting how climate change will affect its controlling variables, but the scenarios chosen covered a range of rainfall distributions which would generate wetter or drier soil conditions and would to some extent mimic the effects of reduced or increased evapotranspiration. Carbon inputs from crop residues were increased by 15% to allow for possible C02 fertilisation, but N use was kept at current amounts to reflect the increase in efficiency of use. The harvest date was brought forward by one week every twenty years to allow for the faster development of crops in higher temperatures. RESULTS

Model runs were made with appropriate inputs for a continuous winter wheat cropping sequence and with a rotation alternating a winter and spring cereal. Figure 1 shows how soil organic N content changes over 80 years in each of the four scenarios. With no change in climate (Scenario A), the soil is in a steady state and there is no decline in organic matter because of the inputs chosen for this particular example. In the other scenarios, a rise in temperature and increase in C inputs to soil is introduced in 2010. Initially, there is a rise in organic N as more is immobilised but after 60 years, the rise in temperature is sufficient to bring an increase in decomposition which more than compensates for the additional carbon, and organic N declines below the steady-state level. By 2070, Scenario D, (wet winters and summers), has led to a 5% decrease in soil organic N content. Figure 2 shows how organic matter might decline in the same soil if the winter cropping sequence were changed to an alternating spring and winter crop. With no change in climate, soil organic N content has declined by 12% and with Scenario D by 16% by the year 2070. Again, in earlier years of the run, additional C inputs are sufficient to maintain organic matter above those in the unchanged

148

Bradbury and Pow/son

climate. Thus, the change in the cropping sequence to include spring-sown crops leads to a decrease in soil organic N over and above that caused by climate change alone.

5.4

51

5.0

z:

4.4

41

4.0

1900

2010

2040

2020

2050

2000

2070

Yea Figure 1 Simulated change in soil organic N over eight decades in four climatic scenarios - continuous winter wheat: Scenario A{ _ _}; Scenario B {..... }; Scenario C L __}; Scenario 0 C-._}.

149

Nitrogen dynamics in arable systems

5.4

51

5.0

z:

4.2

4.0

1990

2010

2050

2000

2070

Year

Figure 2 Simulated change in soil organic N over eight decades in four climatic

scenarios - alternate spring and winter cereal: Scenario A (_ _); Scenario B (..... ); Scenario C

C __}; Scenario D C_._}.

150

Bradbury and Pow/son

The model can also be used to examine the effects on N losses. Figure 3 shows the mean annual amounts of mineralisation, denitrification and leaching in each 20-year period and in each of the four scenarios.

With climate change,

mineralisation increases by about 12% by 2070 and this additional mineral N appears to cause an increase in leaching rather than denitrification. The model does not take account of all possible interactions between the various N-cycle processes, so the partitioning of losses between leaching and denitrification may be questionable. ,However it does seem that larger overall losses of N will occur. It is also useful to examine the losses over the course of the year.

Denitrification

Leaching

Mineralization

Figure 3

Mean annual mineralisation, leaching and denitrification under winter

wheat in four twenty-year periods and four climate scenarios.

151

Nitrogen dynamics in arable systems

Figure 4 shows the cumulative amounts of mineralisation, leaching and denitrification in the four 20 year periods of the model run, using Scenario B with a winter-cropping sequence.

The values shown are the mean amounts of N

mineralised, leached or denitrified in each week of the year in a twenty-year period cumulated over the course of the year.

As harvest and incorporation of crop

residues is brought forward, immobilisation of N increases but subsequent remineralisation gives rise to earlier leaching losses. The soil is thus depleted of available mineral N over winter and leaching losses are reduced in early spring. Denitrification increases over the summer as the climate is warmer giving an increase in microbial activity.

CONCLUSIONS

The results obtained here give some indication of the ways in which changes to N cycle processes may affect arable cropping systems in a warmer climate. Higher temperatures will tend to cause a decrease in organic matter content of soil, but changes in land use or cropping sequence may have a greater influence. In the southern part of the UK, a warmer climate may encourage the adoption of more spring-sown crop species such as linseed, sunflower and maize which are grown only on a small scale at present. Because of their shorter growing season, carbon inputs to soil from these crops may be less than from a winter crop and, in addition, the soil may be left bare over winter before sowing. Both of these factors will lead to a decline in soil organic matter. The model simulations demonstrate that the onset of such a decline will be sensitive to the quantity of carbon returned to the soil and may be delayed for some time should crops respond to C02 fertilisation and increase carbon fixation. Even under present conditions, there is a shortage of quantitative data on total carbon inputs to soil under different cropping systems.

At one scale, inputs can be measured from plants grown in an

atmosphere containing 14C-labelled C02 in controlled environments (Dinwoodie and Juma, 1988). Measurements of root production in the field give an indication of inputs, but are incomplete because of the difficulty in measuring fine roots and exudates (Barraclough and Leigh, 1984). At the ecosystem scale, models can be used to estimate the carbon inputs required to maintain a given stock of soil organic matter (Jenkinson et al., 1992).

There is a need to 'bridge the gap'

between these approaches in order to understand how the responses of processes at the level of the microsite accrue to the field or catchment scale response and beyond. In this way, the most dominant factors may be identified and strategies to

152

Bradbury and Pow/son

mitigate adverse effects targeted on components of the system where they will be most effective. So, for example, in this study it was found that the higher rates of mineralisation induced by higher temperatures led to greater losses of nitrogen and, in particular, losses of nitrate earlier in the season than in the current climate. These losses could be lessened by minimising the time the soil is bare by bringing forward the sowing date of the following crop. Similarly, the risk of fertiliser N losses by denitrification, (in warmer, wetter soils), or by bypass flow, (through persistent cracks), may increase, but the timing of applications could be managed to minimise these risks.

FUTURE RESEARCH DIRECTIONS The following issues require further attention from soil scientists:

1)

The role of soil organic matter in carbon release and sequestration.

How much extra C might be released as a direct result of climate change or, indirectly, due to changes in land use? What is the scope for sequestering significant quantities of C in soil organic matter, e.g. through afforestation? 2)

The role of plant residues. How will climate change influence the quantity and composition of C

deposited in soil by plants and how will these influence soil microbial activities and nutrient cycles? 3)

Use of models. Few process-based models encompass all possible interactions between soil, crop and climate. There is a need to integrate models of plant growth,

soil carbon and nitrogen cycling and soil hydrology with climate models and use them to identify the processes which will dominate the response of soil to climate change.

How can such integrated models be scaled up and

linked to spatial data to indicate trends at the ecosystem or global scale?

4)

Soil physical properties.

To what extent will a decline in soil organic matter content induced by climate or land use change have a detrimental effect on soil physical properties such as aggregate formation and stability, water-holding capacity, aeration and ease of tillage?

How will changes in physical

properties have feedback effects on other nutrient cycling processes? Can susceptible soils be identified, and how should management of these soils be changed in response to a decline in soil organic matter?

153 Nitrogen dynamics in arable systems

Mineralization

100

nI

K

>M >K

>0 >L >G >1

>E High

MQ 2.1: Cultivation "ractices, soil A 8 C 0 E F

G

H

I J K L M N 0 P

Tillage Type Tillage Depth Conservation TechniCS Conservation Technics Conservation Technics Conservation Technics Conservation Technics Artificial Drainage Artificial Drainage Artificial Drainage Artificial Drainage Artificial Drainage Artificial Drainage lnigation lnigation lnigation

>8 >C >H >1 >L >K >J >N Moderate >0 >P >0 Very low Moderate Low Very low

>E >0 >1 >J >K >M >K High >N >N >0 Moderate >P High Moderate Low

MQ 2.2: Cultivation "ractices, "Iant A 8 C 0 E F

G

H

I

J K L M N 0 P Q R S T U V W X

Plant density Row spacing Row spacing Row spacing Row spacing Crop rotation Crop rotation Crop rotation Crop rotation Crop rotation Crop rotation Herbicides Herbicides Herbicides Herbicides Herbicides Herbicides Residues treatment Residues treatment Residues treatment Residues treatment Residues treatment Residues treatment Residues treatment

>8

>J >R >L >M >S >N >Q High >S Low >X >W >V High High Low Moderate Moderate Low Mod

>C >F >H >K >T >V Very Low >W >R Moderate >W Low >T Low High Mod Very Low Moderate Low Low

>0 >X

>S >R >U >U >P >N

Moderate Moderate Very Low Low Low Very Low Low

NOTE: Under each class the symbol> followed by a letter (8 to X) is used to direct the user to the next step of the decision trees. The pathways of the decision trees are followed untill a severity level of the MQ is encountered.

203 Assessing agricultural soil erosion vulnerability

Categories of classification

The attainable soil erosion risk submodel is formed by a combination of three LO's and divides into four risk classes: 1) very small, 2) small, 3) moderate, and 4) high; with corresponding subclasses according to the respective LO: relief (r), soil erodibility (e) and rainfall erosivity (f). The management soil erosion risk submodel is formed by a combination of the two MO's and divides into four risk classes: 1) very low, 2) low, 3) moderate and 4) high, with corresponding subclasses according to the respective MO's: crop properties (c) and cultivation practices (p). The final actual erosion vulnerability classes (total of sixteen) are a combination of the attainable and management soil erosion risk classes. Automated data processing

The core of the evaluation modules (decision trees) were initially developed within the ALES framework (Rossiter, 1990), and then translated into the Microsoft™ C language. An important part of this module is a CLIPPERTM program with menus and explanatory screens to aid generation (preparation and editing) of input data for the models, and another program to display (in tabular and/or graphical mode) the results of the evaluation. Thus, it is easy to modify parameters, create hypothetical scenarios, run the evaluation and observe their effects.

The software is a

compressed package of the CLIPPERTM and C programs, which runs on an IBM PC or compatible with MS-DOS 3.0 or later operating system; 640 Kb RAM memory; video, VGA, EGA or CGA cards; floppy drive; hard disk. Application and validation

The attainable soil erosion submodel has been applied to 62 Andalucian benchmark areas (de la Rosa, 1984). In addition, 42 European (CEC, 1985) and 44 English (SSLRC, 1988) land units were evaluated.

The correlation (r2) between the four

classes of erosion according to the soil profile descriptions of the 62 Andalucian soils and the four modelled attainable erosion risk classes was 0.83. The correlation between the risk classes of the 'erosion risk class map of England' (SSLRC, 1988) and the four modelled classes was 0.85. The first step in applying the actual erosion vulnerability submodel was to develop a management knowledge base (MKB) for the Andalucia region. This knowledge base

204 Crompvoets et al.

contains 25 field utilisation types (FUT's), which correspond to the most representative annual and perennial 'crops' of Andalucia. The application of this part of the model is based on the created FUT's, published literature, expert knowledge and farmer contacts.

It is also partly based on the 62 benchmark areas. The

correlation (r2) between the four classes of erosion according to the soil descriptions of the 62 typical Andalucian soils and the four modelled classes of the actual erosion (from field description) was 0.97.

Acknowledgement The financial support of the Commission of the European Communities (DGXII) ECproject EV5V-CT92-0129 'ACCESS: Agro-climatic Change and European Soil Suitability' is gratefully acknowledged.

MICROLEIS 3.2: A SET OF COMPUTER PROGRAMS, STATISTICAL MODELS AND EXPERT SYSTEMS FOR LAND EVALUATION

D. de la Rosa Consejo Superior de Investigaciones Cientificas, Instituto de Recursos Naturales y Agrobiologia,

P.o. Box 1052, 41080 Sevilla, Spain.

SUMMARY Introduction Although increasing consideration is being given to agricultural diversification and to lower input agriculture, it is still important to identify optimum land use systems for resource sustainability and environmental quality. Land evaluation makes it possible to use land according to its potential.

During the last few years, increasing

application of information technology to land evaluation procedures has led to the development of land evaluation information systems.

For these computerised

applications, the microcomputer (PC platform) has become an essential tool. Since 1975, several land evaluation projects have been developed by the Instituto de Recursos Naturales y Agrobiologia, Sevilla (formerly, Centro de Edafologia y Biologia Aplicada del Cuarto), and the Agencia de Medio Ambiente, Junta de Andalucia, Spain. MicroLEIS is based on the results of these projects. The principal objective of MicroLEIS was to establish an interactive and user-friendly system for the optimal allocation of land use and management systems under Mediterranean agroforestry conditions. The MicroLEIS system must be considered a tool for land use planning rather than an accurate predictive model.

NATO ASI Series, Vol. 123 Soil Responses to Climate Change Edited by M. D. A. RoonseveIl and P. J. Loveland © Springer-Verlag Berlin Heidelberg 1994

206

delaRosa

The basic framework The framework of MicroLEIS is in accordance with the FAD Framework for Land Evaluation with adaptations established for the European Community, and integrates land evaluation methods previously developed. Several land capability, suitability, yield prediction and vulnerability methods may be automatically applied. MicroLEIS addresses land evaluation at various scales: reconnaissance, semi-detailed and detailed, in an interrelated manner. Biophysical land evaluation procedures are combined using corresponding scale-appropriate models, which vary from purely qualitative through semi-quantitative to quantitative empirical models. Economic attributes were not considered.

The ultimate output of the system is the

classification of a specific soil in relation to a particular agricultural or forestry use, as well as the environmental impact assessment for sustainable land management in the Mediterranean. General land capability The general land capability module (named Cervatana) represents a qualitative land evaluation method as a first stage to screen land units as suitable or not suitable for agricultural use. The module rates capability according to limitations imposed by: a) site factors: slope; b) soil factors: useful depth, texture, stoniness, drainage and salinity; c) erosion risk: slope, soil erodibility, rainfall erosivity and vegetation density; d) bioclimatic deficiency: moisture degree and frost risks. Matching-tables were used to express inferences and define, by the maximum limitation method, four capability classes: Class S1-excellent; Class S2-good; Class S3-moderate; and Class N-marginal and unsuitable.

Four subclasses are also

defined according to site (t), soil (I), erosion risk (r) and bioclimatic deficiency (b) limitations. Forestry land suitability The forestry land suitability module (named Sierra) describes the land requirements of 22 representative tree species. The land use requirements were estimated as the

207 Expert systems for land evaluation

minimum conditions necessary for the successful and sustained growth of a given species, according to limitations imposed by: a) site factors: latitude, altitude, physiographic position; b) soil factors: useful depth, texture, drainage, pH; c) climate factors: minimum and maximum temperature, precipitation. These land requirements were structured so that a land suitability classification could be used to indicate whether a land unit was suitable (Order S) or not suitable (Order N) for the tree species under consideration. Maximum limitation procedures were followed to establish the physical suitability method for forest use. Agricultural soil suitability

The soil suitability module (named Almagra) was based on an analysis of edaphic factors which influence the productive growth of twelve traditional crops: wheat, corn (maize), melon, potato, soybean, cotton, sunflower, sugar-beet, alfalfa, peach, citrus and olive. Effective depth (p), texture (t), drainage (d), carbonate content (c), salinity (s), sodium saturation (a) and soil profile development (g) are the soil characteristics considered as diagnostic criteria.

For each soil characteristic, a gradation matrix

was established which relates the soil characteristic value to the corresponding soil crop requirements.

Following the maximum limitation procedure, five relative

suitability classes are determined: Class S1-very high, Class S2-high, Class S3moderate, Class S4-low and Class S5-very Low. The subclasses are indicated by the letters corresponding to the main limiting soil criteria. Crop yield prediction

The crop yield prediction module (named Albero) was based on the use of statistical modelling to formulate and calibrate multiple regression equations to predict yields of wheat, corn (maize) and cotton crops.

These agroecosystem models were

formulated, calibrated and validated over a particular range of management practices, climate, soils and time scales. A high level of management, the general characteristics of a Mediterranean climate, the best agricultural soils and the estimated average yields obtained in recent years, are the experimental parameters which define the selected benchmark Sevilla zone.

208

de/a Rosa

The following were considered as diagnostic criteria or x variables: useful depth (x1), clay content (x2), depth to hydromorphic features (x3) , carbonate content (x4) , salinity (x5), sodium saturation (x6) , and cation exchange capacity (x7). According to the statistical test, the independent variables and their interactions accounted for a large part of the variation in yield.

Agricultural field vulnerability The field vulnerability module (named Arenal) is a knowledge-based model that allows one to predict the relative vulnerability of fields to agrochemical compounds, in terms of soil and groundwater contamination.

The following field factors are

combined: a) land factors: precipitation, physiography, water table depth, soil texture, salinity, pH and CEC; b) management factors: farming system, artificial drainage, water extraction. Expert knowledge was captured into the ALES system shell (Cornell University, USA), through computer-based decision trees. The mobility of agricultural pollutants (fertilisers and pesticides) by soil infiltration into groundwater was especially taken into consideration. Four field vulnerability classes: S1-none, S2-slight, S3-moderate, S4-severe, were chosen and defined. This expert system can be used to estimate the environmental impact of agricultural activities, with reference to chemical degradation of soil and water resources.

Dataset and toolkit The dataset module within MicroLEIS includes the major Andalucia datasets used to develop the MicroLEIS system, and makes application of the other evaluation modules within the system easier. The datasets comprise several standard data files from 62 benchmark sites of the Andalucia region of monthly climate data, of landscape and soil morphological and analytical data, and of crop and management data for specific land use systems. The toolkit includes a group of simple tools to estimate soil water balance, to assess erosion and agrochemical transport, along with several pedotransfer functions. Bioclimatic classification uses the method of Thornthwaite, rainfall erosivity estimation by Fournier index and precipitation leaching degree by the Arkley method. Soil texture classes according to several

209 Expert systems for land evaluation

systems of classification, and their relation to soil physical and chemical properties, are included as qualitative pedotransfer functions. The computing environment

MicroLEIS was designed and constructed to be applied as a sequential and userfriendly set of tools. Input data are entered from the computer keyboard for each soil, land or field unit to be evaluated, following a menu system and explanatory screen mode. The executable files, when called from the 'Main Menu', will apply the corresponding evaluation modules.

Several documentation files give ample

information on MicroLEIS by means of an 'Electronic Manual'. System utilities such as: 'Dataset and Toolkit' chapters derive and prepare input data; an 'Evaluation Results' system allows the user to view, edit, print, copy and delete output data files; and 'Presentation Language' section to change from the Spanish to English languages. The software is a compressed package of compiled BASIC programmes, which runs on IBM PC, XT, AT, or compatible machines with MS-DOS 2.0 or later operating system; 640Kb RAM memory; video, VGA, EGA or CGA cards; floppy drive: 3.5" (1.44Mb or 720Kb) or 5.25" (1.2Mb); and the printer assigned to port LPT1. MicroLEIS 3.2 requires 2.8Mb of hard disk space; and the software is distributed on 5.25" or 3.5" floppy disks. Materials needed to use MicroLEIS

Although the dataset and toolkit module can make the application of MicroLEIS easier, to use this system effectively, the following information will be needed: a) the relevant soil survey report, in order to identify the soil types; b) tabular climate data statistics on a monthly basis from the relevant meteorological stations; c) the last report on labour operations that identify the farming system of the field types.

MODELLING SOIL EROSION ON UK AGRICULTURAL LAND UNDER A CHANGED CLIMATE

D. Favis-Mortlock Environmental Change Unit, University of Oxford,

1a Mansfield Road, Oxford OX1 3TB,

u.K.

SUMMARY Introduction

Agricultural land in the UK has experienced an increase in soil erosion by water during the last two decades. This increase is due to continued intensification of farming methods and to the widespread adoption of autumn-planted cereals. It is certain that agricultural activity will continue to be a major influence on erosion. Nonetheless, the rate and extent of future erosion is also likely to be strongly determined by any change in the amount, frequency or intensity of precipitation which might occur under a changed climate. Methodology

In order to assess the impact of such climate change, a modelling approach has been adopted. The EPIC (Erosion-Productivity Impact Calculator) model has been used in a number of studies to estimate erosion rates at a test site on the UK South Downs for changed climatic conditions (Boardman et al., 1990; Favis-Mortlock et al., 1991; Boardman and Favis-Mortlock, 1993a; Favis-Mortlock and Boardman, submitted).

Due to the uncertainty surrounding projections of future climate -

precipitation in particular - a number of possible climate scenarios were constructed. From these, sequences of synthetic weather were generated and used to drive the EPIC model. Both equilibrium and transient simulations were undertaken.

NATO ASI Series. Vol. 123 Soil Responses to Climate Change Edited by M. D. A. Rounsevell and P. J. Loveland © Springer-Verlag Berlin Heidelberg 1994

212 Favis-Mortlock

Results

Increase in temperature proved to have little direct influence on erosion rate. Change in rainfall amount, however, affected erosion rate strongly. A 27% increase in mean annual erosion rate for a 15% increase in rainfall was indicated by the equilibrium simulations (Figure 1). An attempt was made to extend these results to other areas of the UK (Table 1). Results from the transient runs also suggested an increased year-to-year variability of erosion rate, despite rainfall variability being held unchanged from present values (Figure 2). Erosion rates for wet years in particular were predicted to increase by up to 35% for an 8% increase in rainfall amount.

.M ~e ~a~n~an ~n~ u~ ru~e~ro ~s~iO~n~(~m~3~~~a~)__________________________,

4.0 ""T

•...

3.6

3.6

3.4

3.2

3.0

+--------,--------,-------1

1990

2030

2010

2050

Year

Figure 1 Results from the equilibrium runs: simulated mean annual soil loss for a site on the South Downs. The solid line is the erosion rate for a 'best-guess' climate scenario; the other lines are for higher and lower estimates (after Boardman and Favis-Mortlock, 1993a).

213 Modelling soil erosion under a changed climate

Table 1 Median erosion rates (m 3 ha- 1 yr 1) for sample areas of England and Wales (after Boardman et al., 1990).

Soil texture type

Location

mean

annual

+5

+10

+15

0.43

0.49

0.52

Ratio to South Change in Downs rainfall (%) erosion rates

Shropshire

4.0 2.9

Staffordshire

3.1

all areas

3.5

clayey

Dorset

3.5

"

Cambs./8eds.

light

Norfolk

0.6 1.8

0 0.39 3.63 2.89 3.08 1.75 1.56 1.13 1.21 1.37 1.37 0.23 0.70

fine silt

GwentlHereford

1.7

0.66

0.73

0.83

0.88

silt loam

Weald

1.2

0.47

0.52

0.59

0.62

silty

sandy

South Downs

1.0

Kent

9.3

Somerset

7.4

all areas

7.9 4.5

Isle of Wight Nottinghamshire

4.00

4.56

4.84

3.18

3.63

3.85

3.40 1.94 1.72

3.87 2.21

4.11 2.34

1.96

2.08

1.25

1.42 1.52

1.51 1.61

1.72

1.82

1.51

1.72

1.82

0.26

0.29

0.31

0.77

0.88

0.94

1.33 1.51

loam

The simulations indicate a broadly linear relationship between annual soil loss and annual rainfall (Figure 3). However the form of this relationship apparently contrasts with that obtained from ten years' monitoring of erosion in an area of the South Downs (Figure 4). Even allowing for the limited size of the dataset, the non-linear and/or discontinuous nature of this pattern implies that the simulations may nevertheless be underestimating the sensitivity of the erosion-climate system.

214

Favis-Mort/ock

10 ~------------------------------------------~

9

-

CONTROL2

....... Residual

- - MID2

8 7

~

6

.~

4

~ 5 .Q

(ij

~

3 2

2020

Figure 2

2040

2030

Year

2050

Results from a transient run: simulated annual soil loss for the South

Downs site. CONTROL2 is a stationary present-day weather sequence, MID2 a nonstationary 'best-guess' sequence (after Favis-Mortlock and Boardman, submitted).

10 9

8

~

.s

7

.. " . ..." ."" ~

C')

I/) I/)

~

6

~

.Q

5

I/)

4

'5 (ij

:>

c

.'l

~"

~

~

.....

.. ....

:...

.~

~

3

~

2

o Figure 3

100 200 300 400

500 600 700 800 900 1000 1100 1200 1300 Annual rainfall (mm)

Annual soil loss and rainfall for the transient simulations (after Favis-

Mortlock and Boardman, submitted).

215 Modelling soil erosion under a changed climate

6.----------------------------------------------, 87-88

5

or

90-91 or

82-83 89-90 91-92 88-89

85-86 T

86-87

.. 83-84

O+----,,----.----,-----r----.-----,----.---~

o

100

200

300 400 500 600 'Erosion season' rainfall (mm)

700

800

Figure 4 Median soil/oss and rainfall in the 'erosion season' (September to March) a monitored area of the South Downs. See Boardman (1993) for further details

for

(redrawn from data in Boardman and Favis-Mort/ock, 1993b).

THE DEVELOPMENT OF PEDOTRANSFER FUNCTIONS FOR THE HYDRAULIC PROPERTIES OF PORTUGUESE SOILS

M. da Concei9iio Gon9alves Esta9ao Agronomica Nacional, 2780 Oeiras, Portugal.

SUMMARY One of the major difficulties in applying models to describe the water movement in a soil profile root zone is the lack of data on the hydraulic properties of different soil types, namely the soil water characteristic curve, h(a), and the hydraulic conductivity curve, K(h). These properties can be measured in the laboratory or in situ and the data used to derive empirical relationships relating the hydraulic properties with other basic soil information (pedotransfer functions). In Portugal, the determination and the availability, in a continuous form, of the curves h(a) and K(h) for different soils is a priority task. The crust method of Bouma et al. (1971) and the hot air method of Arya et al. (1975) were used in the laboratory to measure points of the hydraulic conductivity curve. A sand table, kaolin table and pressure membrane apparatus were used to measure points of the soil water characteristic curve. The two hydraulic properties were also measured in the field with the instantaneous drainage profile and the zero-flux-plane methods. For the different layers of the soil profile studied the points of the curves h (a) and K(h) obtained in the laboratory closely matched follow those obtained with the field methods. A nonlinear least-squares parameter estimation technique was used to fit two models to the observed data: Gardner's (1958) model for hydraulic conductivity (K =

K sat

l+(bh)n

)

and, van Genuchten's (1980) model, with four parameters, for the soil

. . (a - Sr

water c haractenstlc - - =

as -a r

1 ).

l+(ah)n

NATO ASI Series, Vol. I 23 Soil Responses to Climate Change Edited by M. D. A. Rounsevell and P. 1. Loveland © Springer-Verlag Berlin Heidelberg 1994

218

A multiple regression analysis was used to obtain the pedotransfer functions which relate the fitted parameters for the different soils to more easily measured basic soil properties such as soil texture, bulk density, organic matter content and the measured saturated hydraulic conductivity.

The following pedotransfer functions

were obtained: 1. Parameters of the van Genuchten model (178 observations fitted); 9r = 0.0570 - 0.00058 + 0.0015C + 0.0003G80 - 0.03330M (r2=0.50) 9s = 1.0007 + 0.00118 + 0.0015C - 0.2316GPO - 0.3724p - 0.0005Z (r2=0.6614) log a = - 0.1119 - 0.01268 - 0.023881 + 2.247GPO - 0.0013G80 - 0.19030M - 0.9609p - 0.0064Z

(r2=0.29)

log n = - 0.1153 - 0.00448 - 0.G127C + 0.9792GPO + 0.0012G80 - 0.03030M + +0.0007Z (r2=0.40) 2. Parameters of the Gardner model (99 observations fitted); (r2=0.9717) log b = - 0.8413 + 0.01098 + 0.009781- 2.4434GPO + 0.709110gK (r2=0.7010)

log Ksat = - 0.002349 + 0.1130GPO + 0.0005G80 + 0.976110gK

log n = - 0.0048 - 0.00128 - 0.001681 + 0.0008C + 0.3239GPO + 0.00890M + + 0.0582p + 0.017410gK (r2=0.46) with 9r the residual moisture content (cm 3 cm- 3 ); 9s the saturated moisture content (cm 3 cm- 3 ); a the inverse of the air entry value (cm- 1); n the pore distribution index; Ksat the saturated hydraulic conductivity considered as a parameter of the Gardner'S model (cm day-1); K the measured saturated hydraulic conductivity (cm day-1); band n soil dependent parameters; S, 81 and C respectively the coarse sand, silt and clay content (%) (International Classification); p the bulk density (g cm-3); OM the organic matter content (%); GPO the geometric mean particle diameter; G80 the geometric standard deviation and K the measured saturated hydraulic conductivity (cm day-1). In the case of the parameters nand b of Gardner's model, the correlation could be improved by dividing the soils according to texture. For the moment, the results for the prediction of the shape parameters (a and n) of the van Genuchten model from the basic soil properties are not good.

The

introduction of the saturated hydraulic conductivity as an independent variable will

219 Hydraulic properties of Portugese soils

be tried, as well as the cross products of this variable with the organic matter and textural information in order to represent some kind of structure effect. For the parameters of Gardner's model, the introduction of the measured saturated hydraulic conductivity as an independent variable significantly increases the amount of variance explained. The b parameter seems also to be greatly influenced by the saturated hydraulic conductivity. However, care should be taken when using this variable as a predictor because of its high spatial variability and measurement dependency.

The n parameter (slope of the curve) seems to be particularly

influenced by the geometric mean particle diameter.

THE USE OF EPIC IN A STATISTICAL FRAMEWORK FOR REGIONAL ANALYSIS OF SOIL RESPONSES TO CLIMATE AND MANAGEMENT

J.J. Lee 1, D.L. Phillips1 and V. W. Benson2 1U. S. Environmental Protection Agency,

Environmental Research Laboratory, Corvallis, OR, U.S.A. 2USDA. Soil Conservation Service, Grassland, Soil and Water Research Laboratory, Temple, TX, U.S.A.

SUMMARY Soil processes are strongly influenced by soil properties and condition, land management, and weather. These factors can change greatly over short distances and times. For example, the susceptibility of a field to water erosion can change dramatically during the time it takes for ploughing (i.e. a few hours).

Similarly,

conditions on either side of a boundary line can be very different if farmers employ different management practices (e.g. no-till versus conventional tillage). This variability is in addition to the inherent variability of soil properties. Furthermore, variations in weather can occur over times from several minutes to years. This variability suggests that, optimally, models of soil processes should operate at fine spatial and temporal scales.

Frequently however, characterisation of the

response of soils to climate and management changes is most useful if done for large areas and long time periods. One solution to this dilemma is to use a site simulation model within a statistical framework, which can provide a representative sample of sites within a heterogeneous environment (Lee and Lammers, 1990). For example, the EPIC model has been used in a statistical framework to characterise soil processes, as influenced by tillage practices, over the next 100 years in Illinois (Phillips et al., 1993) and the U.S. corn belt (Lee et al., 1993). Currently, we are applying this approach to the effects on the carbon budget of the South Central NATO ASI Serid. Vol. 123 Soil Responses to Climate Change Edited by M. D. A. Rounsevell and P. 1. Loveland © Springer-Verlag Berlin Heidelberg 1994

222 Lee etal.

U.S.A., to the effect of climate change on crop yield, soil erosion, and soil carbon in the U.S. corn belt, and to the effects of conversion of crop land to forests. The EPIC model was developed by the U.S. Department of Agriculture for regional analyses of crop yield loss due to soil erosion.

It simulates soil erosion, plant

growth, weather, hydrology, nutrient cycling, tillage, and soil temperature on a daily time step (Sharpley and Williams, 1990). The model simulates a small drainage area « 1 ha), with up to 10 soil layers defined. A stochastic weather generator simulates daily temperature, precipitation, relative humidity, solar radiation, and wind values. Water erosion processes are modelled on an individual storm basis by a choice of three erosion equations.

Crop growth is simulated by a generic crop

growth model capable of simulating more than 30 crops. The tillage component simulates mixing of crop residue and nutrients in the plough layer, soil bulk density changes, and conversion of standing residue to flat residue by management operations and weather.

Hydrologic and nutrient cycling components simulate a

number of nutrient fluxes (e.g. Nand P) important to surface water and groundwater quality. Soil organic carbon is also modelled. More detailed explanation of EPIC model components is given by Williams and Renard (1985) and Sharpley and Williams (1990). Within the U.S.A., the 1987 National Resources Inventory (NRI) provides a basis for statistical selection of a representative sample of sites. The 1987 NRI is the latest in a series of inventories conducted by the Soil Conservation Service (SCS) to provide information on the status, condition, and trends for soil, water and land use. It contains information on 335878 sample points on non-Federal land in the United States.

Data of interest for each site include: crops grown for 1984 to 1987;

Universal Soil Loss Equation (USLE) (Wischmeier and Smith, 1978) estimates of sheet and rill erosion; all USLE factors; tillage practices; and 10 numbers for the SCS Soils-5 Soil Interpretation Record (SIR) database. In a typical application, we select 100 sites randomly from a well-defined subset of the NRI sites (e.g. corn (maize)/soybean sites in the com belt). Data from the NRI, from the SIR database, and from a climate data base provide the input required to run EPIC at each site. On a UNIX workstation, a 100 year simulation takes about 1.5 minutes for each site, or about 2 hours for 100 sites.

Thus, it is practical to repeat the simulation for

numerous assumed climate and/or management scenarios. Regional averages are obtained by applying the statistical weights (from the NRI) for individual sites.

EFFECT OF CLIMATIC CHANGES (C02, TEMPERATURE) ON GRASSLAND ECOSYSTEMS: FIRST FIVE MONTHS' EXPERIMENTAL RESULTS

P. Loiseau, J-F. Soussana and E. Casella Laboratoire Fonctionnement et Gestion des Prairies, Station d'Agronomie, 12 Avenue du Brezet, F-63039 Clermont-Ferrand, France.

SUMMARY The C and N cycles were examined in an experiment concerning the impact of climate change (C02 , temperature) on perennial grassland ecosystems (Lolium perenne cultivated on a loamy soil), at two levels of inorganic N application. The various experimental techniques are summarised here, as well as the first experimental results of the effects of climate change on the utilisation of water resources and dry matter accumulation in roots, stubble, litter and the coarse OM fraction. At a similar, limited water supply, the beneficial effect on water consumption, water use efficiency and harvestable production of increasing atmospheric CO2 to 350 ppm disappears when the temperature is also increased by 3°C. This is because of drought. Further analysis will show how N interception by roots is involved in dry matter production and increases under doubled atmospheric CO2 , This results in a lowering of the applied N productivity. Organic accumulation at the soil level is increased by 21 % (N+) to 31% (N-) in the case of CO 2 doubling, but does not occur with supplemental temperature elevation.

Objectives and methods An experiment was started in March 1993 concerning the impact of climate change (C02 " temperature) on perennial grassland ecosystems.

The sward consisted of

Lolium perenne sown 18 months before exposure to climatic changes, in containers (0.5 m2; 45 cm deep) filled with loamy soil (15% clay, 41 % fine loam, pH 6.9, 1.29%

C). The climatic conditions were produced in 3 ventilated tunnels (44 containers in NATO ASI Series. Vol. 123 Soil Responses to Climate Change Edited by M. D. A. Roonsevell and P. J. Loveland © Springer-Verlag Berlin Heidelberg 1994

224 Loiseau et al.

each): control with outdoor temperature and CO2 level (C); CO2 level increased by 350 ppm with outdoor temperature (C02); increased CO2 and temperature (C02, +3°C). The same water supplies were given in all climatic conditions, according to the limiting conditions in summer in the control. P and K supplies did not limit growth, but two levels of mineral N fertiliser were compared: 150 and 500 kg N ha-1 year 1. The containers were all cut together and all harvested materials were removed. The different measurements focused on harvestable production, organic matter in roots and soil particle-size fractions, water utilisation per regrowth period, daily C assimilation, respiration and transpiration, C allocation in swards by pulse labelling with excess 13C, C turnover in soils by long-term labelling with industrial CO2 (.1 13C40 pm), N mineralisation/immobilisation turnover by 15N tracer, microbial biomass, microbial activities and metabolisable C and N (R. Chaussod, Laboratory of Soil Microbiology, Dijon, pers. comm.).

Water utilisation The results were accumulated for the period from March to July.

For the high N

supply pots, the maximum evapotranspiration was reduced by 14% in the case of C02 doubling and increased by 61% in the case of a 3°C elevation of temperature (C0 2 , +3°C), but at a similar limiting water supply during the period, the water consumption was decreased only by 7% (C02) and increased by 7% (C02, +3°C) (Figure la). Drought severity was reduced by 8% (C02) and increased by 24% (C02 , +3°C). Harvestable potential dry matter production was increased by 34% (C02) and 36% (C02, +3°C), but at limiting water supply, the harvestable production was only increased by 25% (C0 2) or was unchanged (C02, +3°C) (Figure Ic). Under these conditions, the water use efficiency (WUE) was increased by 35% (C02) and decreased by 6% (C02, +3°C) (Figure Ib), respectively. The actual evapotranspiration and severity of drought were increased by the level of N fertiliser, shown by an N effect on the recovery of the leaf area after cutting. The supplemental applied N increased the actual dry-matter production by 218% (C), 155% (C02) and 106% (C02, +3°C). The productivity of N fertiliser was respectively 23.4, 25.9 and 14.4 kg DM kg N-1. The supplemental N fertilisation increased the water efficiencies by 210% for the control, 137% for CO2 and 80% for CO2, +3°C (Figure fb).

225 Effect of climatic changes on grassland ecosystems

750

E E

W-,W+: Water limiting or not for growth N-,N+: Nitrogen limiting or not for growth

w+ N+

600

c: o

r-

w+

~ L.

450

0.



N+

~

.c:

L.

j1 w+

w+ W-

12000

~ 10000

C02 +3dC

C02

O+-----J-----~

'"

N-

w-

1000

I

w-

N-

w-

w-

t..-

'--

N-

4000 2000

we!:!.:. Control

C02

N-

C02 +3dC

Q)

~

0

1-------L------L__

~L______L__~L______L__

I

Figure 1 Water utilisation under 3 climatic and atmospheric conditions: influence of water and N supply. la: evapotranspiration; Ib: Water Use Efficiency; Ic: harvested dry matter.

226 Loiseau et a/.

Organic matter at the soil level

At the higher N supply, the roots and stubble phytomass (0 to 15 cm) was increased by 47% (C02) and 4% (C0 2 , +3°C). The amount of coarse OM fraction (200 to 1000 was increased by 44% (C02) and 53% (C0 2 , +3°C), but the climatic changes did

~m)

not affect litter mass (coarse debris> 1 mm). As a whole, the supplemental organic accumulation of soil plant material and litter (> 200 ~m) amounted to 25% of the control in the case of CO2 increase, but was not significant with simultaneous temperature elevation (Figure 2).

Supplemental applied N did not significantly

increase the root or the total soil dry matter accumulation: it only increased the mass of stubble and usually decreased the mass of coarse OM (C, CO 2 ),

12000

N+

T"'"

I

C'O .I:: O>

10000

~

~

0

D

N+

N-

8000 6000

Q)

N-

D

L

L

N+

D

N-

L

(/)

I-

C'O

0

u

0

4000 2000

P P

P

C/)

0 Control

Figure 2

CO2

CO2 +3dC

Coarse soil organic matter in 3 climatic and atmospheric situations:

influence of N supply in water limited condition for the control (same water supply in each climatic situation). P: root and stubble phytomass; L: litter (> 1 mm); D: coarse OM fraction (200 to 1000 ~m).

227 Effect of climatic changes on grass/and ecosystems

Discussion

In the case of an atmospheric CO2 increase alone, the harvestable dry matter production was enhanced by immediate increases in CO 2 fixation and a delayed consumption of water resources, brought about by stomatal closure and simultaneous increase in WUE. The harvested material was increased even at the low level of N input, suggesting nitrogen dilution in the harvest or better root interception of existing mineral N rather than faster N mineralisation. Despite the increase in harvest at the low N level, the higher growth potential resulted in unchanged supplied N productivity. The increase in non-harvestable material and coarse OM was more pronounced than the increase in harvestable production, indicating more C allocation to the below ground parts of the sward. The N availability seemed to be responsible for less accumulation of fine debris through better C mineralisation. In the case of CO 2 and temperature increases, the potential for carbon fixation remained the same as for CO 2 increases alone, but at similar limiting water supplies, the beneficial effect of CO 2 enrichment disappeared because of earlier growth, increase in water consumption and subsequent drought effects. The root and stubble phytomass became similar to the control. The increase in harvest at the low level of N inputs, compared with the control, cannot be attributed to better N interception, but could be due to an effect of temperature on soil N mineralisation.

The earlier

limitation of water resources lowers the productivity of mineral N fertiliser. Despite similar non-harvested phytomass, the total dead material > 200 11m decreased compared with the control.

So, temperature elevation limits the effects of CO 2

increase on carbon accumulation in soils at first by limitation of growth through water shortage and then by increased evolution rates of the new residues and old organic matter. In the first 5 months, CO 2 doubling increased the organic accumulation of dead material by 21% (N+) and 31% (N-), and the total dry matter harvested and accumulated by 23% (N+) and 38% (N-). Supplemental temperature elevation caused no supplemental soil accumulation.

Under the present experimental

conditions, the percentage accumulated in soil represents 50% of the total dry matter (harvested + accumulated) for all climates in the case of no limiting N. This ratio became 76% (control), 72% (C0 2 ) and 65% (C0 2 , +3°C) in the case of limiting N, underlining the impact of climatic changes on the C cycle through their N effects.

228 Loiseau et al.

Conclusion The direct effects of climatic changes on water economy, C fixation and soil N mineralisation imply other indirect effects on water reserves, N resources, and finally new regrowth, including allocation between harvested and returned parts of the sward. Eventual organic matter accumulation in soils will result not only from increased organic returns but also from further interactions between the C and N cycles. The level and manner of opening the N cycle (N fertilisation, harvest and losses) seem to be of major importance for modelling C processes and manipulating C resources. The experiment allows climatic effects on these C N cycles to be examined in the longer term (3 years).

Acknowledgement This work is part of INRA's AGROTECH/EFFET program, of the French contribution to IGBP and the EC 'Crop Change' program.

SIGNIFICANCE OF TWO SOIL COMPONENTS OF THE PEDOSPHERE AS CARBON SINKS

M.J. Mausbach and L.D. Spivey Soil Conservation Service, U.S. Department of Agriculture, Washington, DC, USA.

SUMMARY In this paper we explore the distribution of wet soils (organic and mineral) in the conterminous United States with respect to land use and carbon sequestration. We used information from the National Resource Inventory of the U.S. Department of Agriculture (Soil Conservation Service, 1987), and from the literature, to estimate potential carbon storage in wet soils. Two classes of soils in the pedosphere, organic soils and wet mineral soils, were known to be sinks for a large portion of pedosphere carbon (Armentano and Menges, 1986; Armentano, 1980; Bramyd, 1980; Moore and Bellamy, 1974). Eswaran et al. (1993) in their analysis of carbon distribution in soils of the world showed that Histosols (organic soils) make up about 1% of the land area but contain about 25% of the carbon stored in the pedosphere. The area of wet soils in the conterminous U.S.A. is mainly concentrated in the south east (coastal and riparian areas), the mid-western Great Lakes area and the north eastern U.S.A. Soils in the Taxonomic (Soil Survey Staff, 1975) classes of Aquolls and Aqualfs make up the largest area of the U.S.A., followed by Aquepts, Aquents and Histosols. Wet Mollisols (Aquolls and Albolls) and Aqualfs make up the majority of the cultivated wet soils in the U.S.A. The cropped area of Histosols is small in comparison with most of the wet mineral soils. Most of the cropped Histosols are in the south east and southern parts of the mid-west (warmer parts of the country). Armentano and Menges (1986) showed that in warm climates, soils can sequester about an order of magnitude more carbon than in cool climates and conversely when NATO ASI Series, Yol. I 23 Soil Responses to Climate Change Edited by M. D. A. RounseveIl and P. 1. Loveland © Springer·Yeriag Berlin Heidelberg 1994

230 Mausbach and Spivey

drained these same organic soils will release an order of magnitude more carbon dioxide than the cool climate organic soils. Under the scenario of returning drained organic soils to their original undrained conditions, the total potential carbon storage in organic soils of the conterminous U.S.A. is about 7.8 Gg C yr1. This is a relatively small amount of C yr 1 compared with mineral soils, but is an annual sequestration of carbon. Mineral soils, unlike organic soils, reach equilibrium or steady state levels of carbon (Jenny, 1980; Schlesinger, 1985; Jenkinson et al., 1987). Wet mineral soils have organic carbon levels that range from 20 to 60% higher than well-drained soils (Franzmeier et al., 1985). When cultivated, mineral soils lose an average of 30% of the carbon from native conditions (Schlesinger, 1985).

In this study, we adjusted

carbon in wet mineral soils to reflect native organic carbon contents. The potential for storage of organic carbon in wet mineral soils if returned to their natural undrained state is an additional 7 Pg C or a total of 17 Pg C in wet mineral soils of the conterminous U.S. We realise that returning presently cultivated wet soils to their natural condition is not totally feasible or even desirable.

However, through programs that encourage

residue management and incorporation of residue into the soil, we may be able to attain 50% of the potential sequestration of 7 Pg C in wet mineral soils. This paper demonstrates the potential for managing the effects of agriculture on global carbon fluxes through possible changes in agricultural policy.

THE ROLE OF SITE CHARACTERISTICS, SPECIES AND SOIL HORIZON ON THE EVALUATION OF CARBON CONTENTS OF FOREST SOILS

C. Nys, S. Didier and J.L. Hubert INRA,

Recherches ForestiiHes Cycles Biogeochimiques, 54280 Champenoux, France.

SUMMARY Forest soils can immobilise organic carbon.

One of the characteristics of this

reservoir is its variable nature during the sylvicultural life of the forest stand. The way the carbon reserve cycle functions depends on the site (climate, soil, mineralisation capacity), the species (broad-leaved or coniferous), and the soil horizon (humic or mineral).

In· this summary we present evaluations of carbon

reserves in neighbouring forest stands where the species was the only variable factor. Materials and methods

The sampling was carried out in 23 forest sites in France, on a range of acid soils, from a Typic Eutrochrept to a Typic Haplorthod - the pH in the A horizon varying from 3.5 to 5.5. The important aspect of this study is that 3 or more stands were present at each site being differentiated only by the tree species, all other ecological conditions were similar. The species selected were Oak, Beech, Spruce, Douglas fir, Larch, Scots pine and Fir. Four layers were sampled in the soil profile: the humus, divided into 01 and Of + Oh levels, using a quadrant 0.1 m2 and the mineral layers 0 to 5 cm and 5 to 15 cm, using 3 cylinders. The profile was replicated 10 times in each stand. Dry matter and

NATO AS[ Series, Vol. [23 Soil Responses to Climate Change Edited by M. D. A. Rounseve" and P. J. Love[and © Springer-Verlag Berlin Heidelberg 1994

232 Nys etal.

soil bulk density allowed us to calculate the carbon reserves in each layer, after carbon analyses of the dry samples. Carbon analyses The carbon content of the samples was determined using two different methods on a subsample of each of the 630 samples from 4 of the 23 chosen sites. The aim of this work was to verify whether the 'Ioss-on-ignition' method was adequate to evaluate carbon contents in soil. In this study, the soil variation is due to organic matter content, all the soils were acid with no mineral carbonate content, and the clay content and type allowed us to neglect any error related to the clay water content.

The reference carbon method (C-chn) used was the furnace technique

coupled with an IR cell for carbon dioxide detection.

12

Model

8

C = 0.4702 LOI

4

00





c9 0

0

0

0

-4

§

0



-8



C -chn

-12 0

10

20 •

L

•



40

30

F+H

o At

A

A(B)

50%

I

Figure 1 Representation of the residual between the measured and the calculated values. The relationships between these two methods were studied using a linear regression model.

The global model shows a high value of the adjusted r2 (0.99) and an

apparently good dispersion around the model curve: C

=

0.4702 LOI.

An

examination of the residual of this global model (Figure 1) and the parameter of variability of carbon concentration showed that 'horizon' is a determinant factor and

233 Evaluation of carbon contents of forest soils

should be considered in the model. The equation of this modified model is C = 0.45365 LOI + Hor, where 'Hor' depends on the horizon and the values of these coefficients are: Hor ( L)

= 0.90237

Hor (A 1)

= - 0.24502

Hor (Of + Oh)

= 2.42508

Hor (A[B])

= - 1.11559

The adjusted r2 value was 0.98. The statistical tests showed that this model may be improved if one includes soil characteristic variables.

The evaluation of carbon

content in the selected sites was calculated using this model as an initial estimate of soil carbon reserves. Carbon reserves in forest soils of France

The variable 'horizon' is the most important factor of the variability relative to the 'species' and 'site index' factors as shown in Table 1 and Figure 2. Table 1 Test of variance analyses on the carbon contents in the soils. Factor

OF

Sum of

Mean Square

F Value

Pr > F

1.49

0.86

0.4619

Squares

4.46

Site

3

Species

5

7.20

1.44

0.83

0.5260

Horizon

3

38.65

12.88

7.46

0.0001

Site x Species

7

5.16

0.74

0.43

0.8857

Hor. x Site x S~e.

7

3.45

0.49

0.29

0.9598

The results of the evaluation of the carbon reserves in different soil types (USA classification) are shown in Table 2. In this table the standard error is not quoted, but calculation shows that the coefficients of variation are between 50 to 100% in the humus layer and from 5 to 15% in the organo-mineral horizon, depending on the soil type.

234 Nys eta/.

Table 2 Evaluation, by tree species, of the carbon reserves (t ha- 1) in different soil types. Soil type

Horizon

Oak

Beech

Spruce Douglas

Parent material

Larch S. Pine

Fir

Typic Dystrochrept

L

6.19

6.46

5.94

12.53

4.52

Loess

0-5cm

19.21

21.84

23.55

15.41

18.44

5-15 cm

19.87

16.09

15.96

15.16

14.44

2.59

Of+Oh

Typic Hapludalf

Fir

4.43

2.59

2.68

2.49

L

1.97

1.85

2.31

2.33

Of+Oh

2.75

8.00

6.07

6.30

0-5cm

23.49

15.56

25.25

18.74

5 -15 cm

16.54

15.61

18.56

14.45

Typic Dystrochrept

L

3.82

4.39

5.56

5.23

Of+Oh

0.00

0.00

0.00

0.00

granite

0-5cm

13.58

14.74

17.48

15.22

5 -15 cm

11.18

13.04

14.36

11.66

Loess

Umbric Dystrochrept

granite

Typic Eutrochrept

L

1.31

4.04

3.24

2.45

3.41

3.34

Of+Oh

3.36

7.28

7.14

6.08

11.05

10.36

0-5cm

26.82

30.81

27.19

28.24

28.66

25.83

5 -15 cm

37.45

35.12

32.58

40.30

37.33

31.28

L

2.48

2.70

2.59

Of+Oh

7.03

4.14

5.66

0-5cm

14.18

15.98

11.23

5 -15 cm

20.47

22.81

16.68

Aquic Hapludalf

L

2.11

2.38

3.05

Of+Oh

13.65

7.12

6.46

6.39

loess

0-5cm

22.15

29.46

27.77

22.42

5 -15 cm

47.60

37.08

25.62

23.85

L

2.39

1.75

5.49

5.90

Of+Oh

11.36

21.63

44.48

10.84

loess

Typic Haplorthod

3.04

0-5cm

20.26

16.35

13.78

18.60

5 -15 cm

13.80

13.27

12.33

10.67

Typic Haplorthod

L

1.64

1.66

4.98

Of+Oh

6.99

11.31

27.07

3.67

25.03

sandstone

0-5cm

19.15

21.89

25.39

15.62

27.63

5 -15 cm

14.56

10.75

12.90

12.75

10.21

sand

4.42

3.58

235 Evaluation of carbon contents of forest soils

40 30

20 5-15cm 0-5cm

10

Spruce

Douglas

Figure 2 Histogram of the mean carbon content in

Gubon 5

20yerus C=25kg1n1Z

a range of forest soils.

J)Oyerus C=4kg1m2

+

80

J)O

4

3 2

1

0 0

20

40

60

120

Figure 3 Changes in carbon reserves (kg m-2) during one rotation of a Sitka spruce stand (from page 1968, Commonwealth Forest Rev., 47).

236 Evaluation of carbon contents of forest soils

Conclusion The carbon reserves in forest soils are related to the soil type and to the species growing on them and species behaviour is dependent on the soil activity. In this initial approach we did not study the changes in the carbon reserves with time because of: i) the seasonal changes in mineralisation activities during the year, or ii) the stage of development of the stand during rotation as shown by Page (1968) for a Sitka spruce stand (Figure 3). The data collected in this project (INRA - AGROTECH) will allow us to produce a database, which could be used to validate agricultural models for forests.

With

reference to global climate change and its consequences for species distribution in the landscape, we should be able to predict the change in the carbon sink capacity of forest ecosystems by modelling the modifications in the carbon reserves in forest soils due to species change.

STATISTICAL STUDY OF SOIL RESPIRATION: CALCULATION OF PRESENT DAY RATES AND ANTICIPATION FOR A DOUBLE CO2 WORLD

F. Robinet Laboratoire de Physique AtmospMrique et Planetaire, Institut d'Astrophysique, Universite de Liege, Avenue de Cointe 5, B-4000 Liege, Belgium.

SUMMARY On a global scale, soil respiration rates are positively correlated with mean annual air temperature and mean annual precipitation.

In this work, we present and test

different regressions of soil types. We use these regressions to calculate present day rates of respiration and those expected for a double CO 2 world. The global CO 2 flux is estimated to be 60 Gt C yr1. Using climatic fields calculated with a GCM for a double CO 2 world, the flux is higher by 20%. This increase can produce a positive feedback to the greenhouse effect. Introduction

Terrestrial ecosystems play an important role in the global carbon cycle (Tans et al., 1990).

Reliable estimates of carbon dioxide exchange in various terrestrial

ecosystems are needed for an improved understanding of biospheric-atmospheric interactions (Kim et al., 1992). Within the soil, carbon dioxide is produced mainly by the respiration of soil biota and plant roots. Soil microbial activity is essential for the mineralisation of organic matter, which provides nutrients for plant growth and root respiration, and depends on the vitality of the plant (Mogensen, 1977). The soil surface carbon dioxide flux is likely to depend on soil temperature (Kucera and Kirklam, 1971; Sharkov, 1987), soil water content (Grammerer, 1989), the occurrence of recent precipitation (Sharkov, 1987)

NATO ASI Series. Vol. I 23 Soil Responses to Climate Change Edited by M. D. A. Rounsevell and P. J. Loveland © Springer-Verlag Berlin Heidelberg 1994

238 Rabinet

and photosynthetic rate within the plant as well as substrates available for decomposition within the soil (Blet-Charaudeau et al., 1990). Measurement Sites

Soil respiration (n = 162) measurements have been performed in some terrestrial ecosystems although relatively few measurements exist for arid, semi-arid, and tropical regions and no measurements are given for warm deserts e.g. Sahara (Raich and Schlesinger, 1992).

Europe and North America are well covered, but Africa,

North Asia and South America are not.

Figure 1 Measurement Sites. Statistical analysis

A statistical analysis was performed to correlate soil respiration data with climate data. The corresponding mean annual climate conditions were taken from a global database (1 0 x 1°; Leemans et al., 1991). In each pixel of this dataset, soil respiration data are averaged, so we were able to obtain 70 values of the respiration rate in the climate. Ecosystems were separated into ligneous (woody) and herbaceous classes. In the herbaceous case, soil respiration was better correlated with preCipitation than with temperature, while in the woody case, the correlation with temperature was better. The reason for this is that woody (perennial) ecosystems are absent from low precipitation areas such as deserts. The linear contribution of both precipitation and temperature improves the correlation with soil respiration in these ecosystems. The improvement is not so marked with the use of parameterisation correlating the lowest of the two rates calculated from temperature and from precipitation.

239 Statistical study of soil respiration

Table 1

Monthly mean soil respiration rates, parameterisation for 32 data of

herbaceous ecosystems. Variable(s) used

Correlation

for correlation

Parameterisation R (9 C m-2 month-1}

Temperature (T)

Rr/lerb = 1.87 T + 21.8

0.62

Precipitation (P D)

Rpoherb = 0.62 PD- 0.85

0.79

Temp (T), Prec (P)

RT,pherb = 0.8 T + 0.5 P -1.2

0.82

Min (T, PD)

RAilerb = 1.63 min (RT' RpD) - 8.6

0.65

coefficient

Table 2 Monthly mean soil respiration rates, parameterisation for 38 data of woody ecosystems. Variable(s) used

Parameterisation

Correlation

for correlation

R (9 C m-2 month-1)

coefficient

Temperature (T)

R~ood

= 2.5 T + 27.1

0.82

Precipitation (PD)

RpDwood = 0.44 PD + 15.4

0.70

Temp (T), Prec (P)

RT,pWood = 1.6 T + 0.24 P +18.3

0.80

Min (T, PD)

RA,fWood = 1.48min (RT' RpD)- 6.3

0.81

Extension to the global scale The fraction of ligneous and non-ligneous material in each 10 x 10 pixel is based on the distribution of ecosystems published by Wilson and Henderson-Sellers (1985). The global distribution of soil respiration rates was obtained by application of the four previous regressions to the global climate database. Only the distribution obtained by correlating the minimum temperature and precipitation functions is consistent with both warm deserts and cold-humid climates, despite the relatively poor correlation

240 Rabinet

coefficient of its regression on data (selection of ecosystems). The global value of the annual mean soil respiration rate is 60 Gt C yr1, a figure comparable to previous estimates, which are in the range of 50 to 75 Gt C yr 1 (Houghton and Woodwell,

1989; Schlesinger, 1977). Rxtot = RxWood 0'

* 0' + Rjlerb * (1

- O')

= woody contribution to grid vegetation x = T, Pp' T, P or M

Figure 2 Monthly mean soil respiration

o

500

1000

1500

Figure 3 Map of annual soil respiration rate (g C m-2 yr 1).

Global soil respiration in a double CO 2 world Based on the climatology of the UK Meteorological Office General Circulation Model, we predict that the global soil respiration rate would be increased by about 20% if the atmospheric CO 2 level were doubled. The response to the CO2 increase varies from one ecosystem to another and is small near the equator and large in boreal regions. Boreal soils have high contents of carbon, which will be released into the atmosphere in the future.

241

Statistical study of soil respiration

Table 3 Annual global mean soil respiration rate (Gt C yr 1).

Monthly mean temperature and precipitation from

Annual global mean soil respiration (Gt C yr1) Variable(s) used for

Observations

GCM 1 X CO2

GCM2xC02

(IIASA)

(UK Met)

(UK Met)

Temperature (T)

82

78

95 (+22%)

Precipitation (P)

61

74

82(+11%)

Temp (T), Prec (Pol

69

76

91 (+20%)

Min (T, Pol

60

72

85 (+18%)

correlation

Discussion and conclusions Arid, semi-arid and tropical regions should be priorities for soil respiration measurements.

These measurements must include climate variables at different

sites. Other important parameters include litter fall, the soil content and the chemical composition or organic carbon, the soil texture and moisture. The seasonality of the phenomenon is very important in order to validate the biospheric models by the atmospheric signal of CO 2 , The residence time of various organic compounds must be known by 14C dating: climate change can influence residence times and soil contents of carbon.

Another problem is the contribution of root respiration to the

release of CO 2 into the atmosphere from soils. Root death is an invisible litter 'fall', and a source of soil carbon. The root biomass, the fraction of NPP allocated to root growth and the mass of roots dying annually are not well known. At equilibrium, the rate of soil respiration minus root respiration is equal to the net primary production. A better knowledge of soil respiration processes is necessary to develop global models of the continental biosphere.

Acknowledgements This study was supported by the EC Environment Program (contract no. EV5V-CT920119), the Global Carbon Cycle and its Perturbation by Man and Climate II, part B. Biosphere, ESCOBA project: European Study of Carbon in the Ocean, Biosphere and Atmosphere.

DEMONSTRATION OF SUNDIAL: SIMULATION OF NITROGEN DYNAMICS IN ARABLE LAND

J.U. Smith and N.J. Bradbury AFRC Institute of Arable Crops Research, Rothamsted Experimental Station, Harpenden, Herts. AL5 2JO, U.K.

SUMMARY SUNDIAL is a user-friendly PC-based version of the Rothamsted Nitrogen Turnover model (Bradbury et al., 1993).

It is a dynamic model, describing the turnover of

nitrogen and carbon in the soil crop system: inputs from fertilisers, organic manures, the atmosphere and crop seed; transformations by mineralisation, immobilisation, nitrification and crop uptake; and outputs by denitrification, leaching, volatilisation, senescence and harvest. A great strength of SUNDIAL is that it is based on simple equations, using only a small number of parameters, and so is robust and nonspecific in its execution.

The model was originally developed for cereals under

funding from the UK Home Grown Cereals Authority and so has been extensively tested for nitrogen turnover under cereal crops. It is currently being extended to include further crops and soil types with funding from the U.K. Ministry of Agriculture, Fisheries and Food. By providing SUNDIAL with details of the soil, weekly weather, organic manure applications and cropping history, the turnover of nitrogen under a sequence of crops may be simulated. This information may be entered either interactively through a detailed menu system or by loading a previously saved simulation set-up file.

To

illustrate the use of SUNDAL in the investigation of the effects of climate change, four different weather scenarios were created: Normal weather data: ten years' average weekly air temperature, weekly

rainfall and weekly evapotranspiration for grass at Rothamsted, 1979 to 1989;

NATO ASI Series, Vol. 1 23 Soil Responses to Climate Change Edited by M. D. A. Rounsevell and P. J. Loveland © Springer.Verlag Berlin Heidelberg 1994

244 Smith and Bradbury

Hot generated from the normal weather data with a mean temperature increase of 2.5°C, the weekly values randomly selected from a normal distribution.

Hot and Dry: generated from the normal weather data with a mean temperature increase of 2.5°C and a mean decrease of 25% in the rainfall selected as described above.

Hot and Wet generated from the normal weather data with a mean temperature increase of 2.5°C and a mean increase of 25% in the rainfall selected as described above. Winter wheat was grown continuously for ten years, with sowing in the second week of October and harvesting in the second week of August in the following year. Fertiliser containing 25% nitrate, 25% ammonium sulphate and 50% urea was applied in the third week of February at a rate of 150 kg ha-1. The expected yield from each crop was 9 t ha-1 and straw was not incorporated. The results can be viewed as balance sheets of inputs and outputs, graphical plots, pie charts of the fate of applied fertiliser and flowcharts of the changes in the nitrogen status of the soil.

Overlaid graphical plots showed a decline in the total organic

nitrogen content of the soil with a temperature increase of 2.5°C. The rate of decline in total organic nitrogen increased with a 25% increase in rainfall and decreased with a 25% decrease in rainfall. Similar trends were observed in the biomass and humus nitrogen contents of the soil. A general decrease in nitrogen losses by leaching and by denitrification was observed with a 2.5°C increase in temperature. This effect was enhanced if the rainfall also decreased by 25%, but was reversed if the rainfall increased by 25%. This short demonstration serves to illustrate the potential of SUNDIAL for use in climate change investigations. The menu-system makes SUNDIAL accessible to the novice user. The explicit and simple input requirements generate consistent results. The graphical display of results facilitates the quick interpretation of possible climate change scenarios.

SEASONAL CLIMATIC VARIABILITY AND UPWARD NITRATE MOVEMENT IN GREEK SOILS

S.P. Theocharopoulos 1, M. Karayianni-Christou 1, P. Gatzogianni 1, and S. Aggelides2 1N.AG.RE.F., Soil Science Institute of Athens,

14123 Lykovrissi, Greece.

2Agricultural University of Athens, Laboratory of Soil Hydraulics, 11855 Botanikos, Athens, Greece.

SUMMARY Nitrate concentrations in the groundwater and rivers of Greece have gradually increased, although this exhibits seasonal variation. It is necessary to take action to avoid further increases in the future. The main purpose of this study was to measure nitrogen

leaching,

under a wheat

crop

in

the

Kopais

area

of

Greece

(Theocharopoulos et al., 1993), under very changeable climatic conditions during the cropping period and for different nitrogen fertilisation treatments. We did this by investigating the pattern of nitrogen distribution in the soil profile of two soil map units during the cropping period and quantifying losses from one of the main crops of the area, winter wheat. We tried to detect the dominant factors throughout the year causing leaching for each soil type in relation to seasonal climatic variability and nitrogen applied, since wheat is sown in October-November and is harvested in June-JUly. The movement of nitrates in relation to the soil water balance (potential evapotranspiration minus rainfall) were studied, for two years, in two experimental fields (EF) established in heavy calcareous soils (Haplaquents) in the Viotia area of Greece.

The wheat variety MEXICALI was grown while nitrogen fertiliser was

applied in four replicates at five different levels 0,100,200,400, and 600 kg N ha- 1 . Evapotranspiration was estimated, for ten day periods using the modified Penman equation (Doorenbos and Kassam, 1979) while the movement of nitrates was NATO ASI Series. Vol. 123 Soil Responses to Climate Change Edited by M. D. A. RounseveII and P. J. Loveland © Springer-Verlag Berlin Heidelberg 1994

246 Theocharopoulos st al.

inferred from sequential soil sampling at depths down to 150 cm. Concentration of nitrates increased in all soil sampling depths with the nitrogen fertiliser applied, for both fields and cropping seasons. When rainfall exceeded evapotranspiration the concentration of N03-N decreased in the first experimental field (EF), for both years,

o to 30 cm and 30 to 60 cm depths, while at the lower depths N03-N increased.

In

the second EF this decrease extended to 150 cm in the soil which was probably due to differences in the soil physical properties. Table 1a Average concentrations of soil nitrate-N (ppm dry weight) at different depths and sampling time in Experimental Field 1. Depth

Date

(cm} 3/11/87

9/2188

14/4/88

13/7/88

1/12188

27/2189

19/4/89

19/7/89

36.6

26.0

72.5

34.3

34.0

68.2

133.1

0-30

44.7

30-60

66.0

63.4

55.0

64.0

59.5

55.4

69.2

62.4

60-120

22.6

36.0

40.1

27.9

21.5

35.9

25.9

21.8

120-135

8.9

12.8

17.4

12.8

11.3

15.7

14.5

14.8

135-150

8.8

8.5

13.9

13.9

10.8

15.2

13.5

14.6

Table 1b Average concentrations of soil nitrate-N (ppm dry weight) at different depths and sampling time in Experimental Field 2. Date

Depth

(cm} 14/4/88

13/7/88

1/12188

27/2189

19/4/89

19/7/89

0-30

12.4

20.0

5.8

11.3

9

26.6

38.9

55.5

30-60

18.7

18.8

16.8

14.8

16.8

10.2

15.7

19.6

60-95

18.3

20.4

18.8

14.5

15.2

12.0

9.5

12.9

95-150

5.2

8.1

6.5

3.1

6.4

5.7

5.8

6.6

3/11/87

9/2188

When evapotranspiration exceeds rainfall the concentration of N03-N increases in the first EF in the top 60 cm in the first year and in the top 30 em in the second year. In the second EF the increase in N03 -N was only in the top 30 cm in the first year and at all depths the second year. The groundwater from the different treatments in both EF showed that the concentration of nitrate increased with the time and with the

247 Upward nitrate movement in Greek soils

dosage of applied nitrogen fertiliser, and at the first sampling after fertilisation, since rainfall was higher than evapotranspiration at the time of fertilisation. Table 2a Concentration of nitrate (mg N03 ( 1) in the groundwater of Experimental Fields (EF) 1 and 2 against time. EF1

EF2

S.E.

S.E.

Mean

Date

Mean

11/11/87

155.09

47.47

14.08

3.67

13/01/88

148.75

34.33

37.58

32.76

10/02/88

127.27

26.44

18.67

8.01

24/04/88

196.03

84.57

25.04

16.04

16/06/88

181.87

66.19

10.42

2.46

17/08/88

150.64

37.16

5.09

1.04

28/11/88

180.02

72.18

27.01

30.50

12/01/89

201.13

94.28

23.89

7.89

01/03/89

185.33

70.48

13.27

5.79

03/04/89

238.24

120.97

38.54

24.55

09/05/89

175.12

72.40

20.70

6.72

05/06/89

108.03

27.94

10.52

1.64

5.50

1.36

17/07/89

Table 2b Concentration of nitrate (mg N03 1- 1) in the groundwater of Experimental Fields (EF) 1 and 2 in relation to nitrogen fertiliser applied Applied N kg ha-1

0-0 200-200 400-400 600-600 EF1 EF2

122.00 126.63 18.65

12.48

143.38

236.74

15.92

32.72

Conclusions Climatic variability is the main determinant of nitrate movement in soil profiles under the bioclimatic conditions of Greece. This is highlighted by the increase of soil N0 3N in the topsoil when evapotranspiration exceeds rainfall. When rainfall exceeds

248 Theocharopoulos et al.

evapotranspiration a decrease is found in the concentration of N03 -N in the topsoil and an increase in N03 -N in the groundwater, which is also affected by the nitrogen fertiliser applied. Acknowledgements

We would like to express our sincere thanks to the General Directorate of Research for financing this project and to the Soil Science Institute of Athens, especially Mr. J. Fragiadakis, Mr. C. Demopoulos, Dr. P. Papadopoulos and Dr. C. Paschal ides.

STANDARD OPERATION PROCEDURES FOR SAMPLING AND SAMPLE TREATMENT OF SOILS FOR ENVIRONMENTAL SPECIMEN BANKING

G. Wagner and J. Sprengart Environmental Research Centre, Institute of Biogeography, University of the Saarland, PF 151150, 0-66041 SaarbrDcken, Germany.

SUMMARY What is Environmental Specimen Banking? Environmental Specimen Banking (ESB) is a new instrument for long-term environmental monitoring, assessment and research. It depends on the systematic and long-term storage of selected environmental materials for deferred analysis and evaluation (Lewis and Klein, 1990). Banked specimens such as soils, sediments, plants, animals, and human tissues offer the chance to analyse retrospectively the concentrations and effects of potentially dangerous substances in the environment which are hitherto unknown, undetectable or thought to be innocuous.

More than 100000 different chemical substances are produced worldwide and may affect human health and the environment.

For most of them we lack sufficient

information about their effects on man, animals and plants and about their further reaction and fate in the environment. Hitherto unknown substances may tomorrow cause danger for our environment and our lives (see Figure 1). Plants, animals and human beings show impacts of harmful substances. Chemical analysis of representative environmental specimens give reference to possible causes and sources of chemical pollution. Banked samples of selected specimens can also be analysed retrospectively. Thus, they can help to recognise and defuse such problems at an early stage. NATO ASl Series. Vol. l23

Soil Responses to Climate Change Edited by M. D. A. Rounsevell and P. J. Loveland © Springer-Verlag Berlin Heidelberg 1994

250

Wagner and Sprengart

Deficits in Assessment and Control of Environmental Chemicals

Out of > 100 000 substances less than10% are actually monitored

Knowledge deficiencies about toxicological properties, metabolism, behaviour and remain in the environment

Restrictions In availability, senslvity, and reliability of analytical methods

of knowledge I...,.. Lack and evidence

•

Figure 1 Deficits in assessments and control of environmental chemicals. Specimen Banking - a Guide for an International Standard

"A Specimen Bank is the systematic collection, characterisation, and secure storage of samples for deferred examination (analysis or other evaluation). These samples must be stored under conditions which will minimise the occurrence of any changes in the sample (i.e., physical, chemical, or biological).

The examination of the

samples is generally deferred for a period of years, even decades, following collection, so that samples can be studied in relation to newer knowledge, new substances of interest, and new and improved analytical techniques ... " (BMU/U.S. EPA,1989).

251 Sampling and sample treatment of soils

The German Environmental Specimen Banking Programme The ESB Programme of the Federal Republic of Germany was started in 1985. The concept was based on the results of an extensive pilot phase (BMFT, 1988; Lewis, 1987) and a series of international workshops (Berlin et al., 1979; Luepke, 1979; Lewis et al., 1984). There is also international cooperation with ESB's in the USA, Canada, Japan, Denmark, Sweden, and emergent ESB Programmes in many other countries (see Wise et al., 1988; Rossbach et al., 1992; Stoeppler and Zeisle, 1993).

Selection of research areas Sampling areas have been chosen to form a national Network of Ecological Assessment Parks coordinating Environmental Specimen Banking with long-term ecological research and environmental monitoring. The list of at present fourteen areas comprises the major ecosystems and habitat types that occur within the Federal-Republic of Germany including two main rivers, two urban industrial ecosystems, and three coastal areas.

Selection of specimen types As a (minimal) set of specimen types to represent main functions in tliese ecosystems the following species have been selected: atmospheric essence (filters); wet deposition; soil (litter/humus-layer and topsoil); terrestrial plants (Norway spruce Picea abies, Scots pine Pinus sylvestris, beech Fagus sylvatica, Lombardy poplar

terrestrial animals (earthworms Lumbricus terrestrisiAporrectodea longa; urban pigeons (eggs Columba /ivia f. domestica); roedeer (liver Capreolus capreolus); from freshwater ecosystems ~ake and river Populus

nigra

Italica);

sediments, zebra mussel Dreissena polymorpha, bream Abramis brama); from the Wadden Sea (brown algae Fucus vesiculosus, common mussel Myti/us edu/is, lugworm Arenicola marina, eel pout Zoarces viviparus, herring gull Larus argentatus (eggs)); human specimens (whole blood, urine, hair, adipose and other tissues).

Standard operating procedures The Standard Operating Procedures (SOP) are the basis for the comparability, reliability, and repeatability of the banked samples.

They contain detailed

instructions for: the selection of sampling sites and specimens; sampling; providing

252 Wagner and Sprengart

cover for repeatability of sampling; area and sample characterisation; sample treatment and long-term storage; documentation of sampling and storage conditions; chemical analysis; data processing and evaluation and quality assurance. Function and importance of soils in ESe Soils are one of the most essential compartments of ecosystems for environmental specimen banking, monitoring and assessment.

All the environmental chemicals

released into the atmosphere are deposited directly or indirectly on the soil surface. Depending on the specific substance properties, climatic influences and soil quality, deposited substances in the soil are accumulated, translocated, transformed, fixed or released.

The status and changes of soil properties and concentrations of

environmental chemicals in different soil horizons are, together with plants and animals, valuable indicators of environmental quality. The challenge is to obtain spatially and temporally representative, comparable and reproducible soil samples for ESB without any changes in their chemical state. Numerous existing guidelines have been studied and followed (e.g. ISO/CD 1038101.2, 1992; Arbeitsgruppe Bodenkunde, 1982; Hodgson, 1985). However, none of these could be used directly because ESB demands the following special requirements: samples must not be contaminated or changed in any way during sampling, sample treatment and long-term storage; samples must be representative for the selected areas and ecosystems; analytical results must be related to mass, surface and volume units respectively; quantity and number of samples is restricted; drying and sieving in the traditional way is not adequate because of possible substance losses and metabolisation during' this process. The specific sampling strategy and plan developed for soils in ESB contains detailed instructions for all steps of selection, sampling and sample treatment as briefly summarised in the following. The full version of the SOP will soon be published (Sprengart and Wagner, 1994). 1. Selection and delimitation of sampling areas and sites

Selection of a well delimited ecosystem typical for the area by making full use of existing maps of all kinds, aerial photographs etc., mapping in the field by soil core boring, visible land marks and vegetation, elimination of disturbed places, definition

253 Sampling and sample treatment of soils

of a soil type representative for the area, delimitation of a homogeneous sampling area of about 2500 m2 with uniform land use, vegetation and history. 2. Preliminary examinations to develop a site-specific sampling plan and protocol

Pedological and analytical characterisation of sites and samples, calculation of the number of individual samples necessary to reach the demanded sample mass. The following measures are necessary to guarantee long-term repeatability: Installation of a soil profile pit as well as interpretation of soil cores and cleaning of augers must be done outside the sampling area, avoiding any disturbing material deposition or land use for other purposes, long-term demarcation of the borders and successive displacement of the sampling grid.

3. Sampling Establishing a regular sampling grid to guarantee spatially representative random sampling (Monte Carlo sampling method using special tools and methods for surface and volume related sampling: sampling of humus layers with a steel frame (20 x 20 cm), removing the humus layer carefully with tweezers and spatula, sampling topsoil with a split-tube-sampler (diameter 5 cm, 10 cm deep) from the prepared spots (cleaned of humus layer), quantitative collection and manual mixing of the individual samples in steel tubs, manual separation of coarse soil fractions and plants for separate analysis, deep-freezing of the samples on the spot using liquid nitrogen, transportation, homogenisation, subsampling and long-term storage in the gas phase above liquid nitrogen, subsequent detection and consideration of soil texture and moisture content.

4. Sample treatment in the laboratory Avoidance of any substance losses and contamination by milling in a titanium mill cooled with liquid nitrogen, subsampling in a cold nitrogen atmosphere, storage in 10 g portions ready for analysis < -140°C (Schladot and Backhaus, 1988).

5. Pedological site characterisation Parameters to identify the main soil profile by field methods, using drill cores (e.g. 'Pyrkhauer drill') and a soil profile pit: depth and thickness of the humus layer and different soil horizons, soil colour and colour distribution, humus content and distribution, soil texture and stoniness, soil structure and consistency, root distribution, bulk density, soil moisture, content of carbonates, content, distribution and form of iron and manganese (hydromorphological markers), clay transfer. For

254 Wagner and Sprengart

additional laboratory examinations samples are taken from a representative soil profile pit and analysed analogous to (6). 6. Pedological sample characterisation

The following parameters must be measured using representative subsamples separated before milling: actual water content, carbonate content, humus content (organic matter), particle size distribution and content of coarse material (> 2 mm), soil reaction (pH in water and in salt solutions, 0.01 M CaCI 2 , and 0.1 M KCI), relation between carbon and nitrogen content (C:N), exchange capacity of bases (BNK), effective and potential ion exchange capacity (Ake and Akpot).

7. Analytical sample characterisation The following substance groups are analysed routinely for sample characterisation in the sense of analytical finger-prints: trace and bulk elements, toxic metals, polycyclic aromatic hydrocarbons (PAH), chlorinated hydrocarbons.

The banked

samples are held ready for deferred examination on any chemical substance or property.

8. Data processing and evaluation, documentation and quality assurance A specially developed ESB information system is fed with all the data to allow fast answers on questions such as: from which areas and dates are banked samples available and what is already known about them? Are there spatial differenciations and temporal trends of the analysed chemical substances and properties? Are there correlation's between chemical substance concentrations in soils and associated biota and possibly related effects in biota and ecosystems? Further application fields for specimen banking (ESS)

As a new instrument for science, administration and management, ESB can support analytical and environmental research and monitoring generally in many ways and make it more effective and reliable, e.g. supply of reference materials for environmental analysis; preservation of authentic records as an archive for long-term comparisons of environmental change; securing and perpetuation of evidence in biology, medicine, forensic medicine, biotechnology, deposition or conversion of problematic wastes; environmental planning and risk assessment.

SUMMARY PAPER

SOILS AND CLIMATE CHANGE - WHERE NEXT?

J-P. Legros 1, P.J. Loveland2 and M.D.A. Rounsevell2 1INRA, Unites de Science du Sol et Laboratoire d'Ecophysiologie des Plantes sous Stress Environnementaux, Centre de Recherches de Montpellier, Place Viala, 34060 Montpellier Cedex, France. 2Soil Survey and Land Research Centre, Cranfield University, Silsoe Campus, Silsoe, Bedford MK45 4DT, UK.

INTRODUCTION

The objective of the Workshop was to address the impact of climate change on soils and the role that soils will play in mediating the overall responses of ecosystems to predicted climate change.

Because of the geographic distribution of NATO

countries, and therefore the origins of the delegates, attention focused mostly on temperate and Mediterranean parts of the globe.

We are aware that there is

vigorous research into similar questions in tropical and semi-tropical areas. The fact that these areas were under-represented at the Workshop does not mean that we regard their problems or research as unimportant. Many of the problems and their potential solutions are universal. This short paper summarises both the points made during the formal presentations at the Workshop, the comments - informal and otherwise - made during the various sessions, and in follow-up correspondence after the Workshop.

The purpose of

much of the discussion was to highlight those areas where soil science could, in association with other disciplines, further climate change impacts research.

The

paper concludes with suggested research topics, and time scales over which they could be attempted.

NATO ASI Series. Vol. 123 Soil Responses to Climate Change Edited by M. D. A. Rounsevell and P. J. Loveland © Springer-Verlag Berlin Heidelberg 1994

258 Loveland et al.

OVERALL REACTIONS

At the heart of the discussion lay the recognition that soil science is one of many important aspects, though neglected by some, of a complex inter-disciplinary research activity. First and foremost, the thrust of climate change impacts research has to be directed towards understanding and predicting the kinds and the magnitudes of the effects of climate change on man's economic activities and the environment, and hence well-being. The research itself has to be carried out with the knowledge that the changes in climate are expected to be rapid. Thus, some of the answers (or informed estimates) are required in the near rather than distant future, because policies to deal with the consequences of climate change need to be set in motion quickly. That said, there was a clear view that the longer term aspects should not be entirely forgotten, and that part of the research effort should be directed to a better understanding of the time-factor in soil and soil-mediated processes. While soils are an extremely important ecosystem in their own right, they are only part of the total ecosystem. Changes to some soil components, as a result of climate change, might be trivial over a short time-scale, compared with the potential changes in such things as vegetation and water supply. For example, most of the current mineral components of soils will change hardly at all over the next century, given a rise in temperature of about 3°C and an overall change in precipitation of 10% or so. It is thus important to focus priorities firstly on those soil components expected to change most rapidly under anticipated climate change, and secondly on changes in soil conditions or states. Examples of the first are changes in organic matter content, and hence the size and nature of the carbon and nitrogen pools, and of the second, changes in the amount of water stored in, or storable by, soils. Changes in either of these factors can affect the kinds of vegetation which soils can support and hence change in cropping potential, change in erosion potential, changes in salt content and so on. Similarly, there is a pressing need to establish the spatial nature of these potential changes; what holds for the US corn belt will not hold for Mediterranean agriculture or boreal forests.

This is implicit in the phrase 'global

climate change', but the understanding of the spatial distribution of these changes is far from satisfactory, and reflects in part, the uncertainties in the predictions derived from global climate modelling. Better estimation of these uncertainties is, in itself, a valid research objective.

259 Where Next?

These are questions which soil scientists cannot answer alone. They involve chemists, agronomists, crop physiologists, ecologists, etc., etc.

However, it

needs to be stressed that scientists from these other disciplines are rarely sufficiently knowledgeable to undertake a full appraisal of the ways in which changes in soil properties or behaviour will affect 'their' part of the ecosystem, without considering the response of the soil to climate change in some detail. It must be an inter-disciplinary team effort. SPECIFIC ISSUES

Data Very large amounts of data exist about the natural environment - soils, river systems, climates (both local and regional), cropping systems and so on. The tools also exist to accumulate more data for specific ecosystems by a variety of means e.g. remote sensing, international experimental sites, monitoring networks, etc. Much effort has been put in over very many years to develop models of various kinds which make use of these data e.g. 'weather' modelling, crop growth modelling, ecosystem modelling, and so on. A lot of scientists from different disciplines have collaborated in these exercises. The Global Climate Change debate has extended awareness within the scientific and policy communities that these databases and tools exist, and that global issues can be addressed through them. There was general consensus that the use of such large datasets and of models has been successful, and will continue to be so. However, there is still a lot to be desired in the way models and databases are used. An immediate issue is that the detailed knowledge of the existence and content of such databases is often fragmentary, nationalistic, bedevilled by questions of ownership and intellectual property rights, and by pragmatic problems such as computer codes, formats, units of measurement, scales of observation and many others. The same also applies to models. For ecosystem modelling and prediction, the priority was seen increasingly to be large-scale work i.e. at high resolution, whilst low resolution i.e. small scale 'global' modelling was accorded a lower priority. This reflected the feeling that many landscapes are too fragmented in ecosystem terms to be modelled at a resolution of tens or hundreds of kilometres, if sensible policy advice is to be given.

260 Loveland et al.

A major problem which affects many soil-crop-climate models is that very few have been validated outside the situation/dataset from which they were developed; calibration is the norm once a model is transported. These are not new problems, but they require urgent attention if endless regurgitation of similar exercises is to be avoided, in which the only distinguishing feature is to be differences between models, and in which differences in output may be due solely to model structure and conceptual differences.

There appears to be no comprehensive, readily-available source of information on models and databases; people rely on word of mouth and personal contact, and this inevitably tends to parochialism.

The best models need to be

identified and tested, their data requirements thoroughly aired, the nature of the output well understood and publicised, and resources concentrated on their improvement. We would encourage the expansion of model 'shoot-out' workshops, which should have both components - database and models. These could be run under the umbrella of one or more of the International Global Climate Change organisations, such as IGBP is doing for wheat models. Spatial questions Discussion of models and data led to numerous questions about resolution on the ground. Whilst the contribution of the climate community, through General Circulation Modelling (GCM) w.as recognised, there is a serious discrepancy between the scale, and possibly the conceptual level, at which soil-plant scientists and climatologists operate. As said earlier, ecosystem scientists are tending to operate at high resolution both spatially and temporally i.e. a few km and weekly or monthly time steps. They are, in effect, interested in weather rather than climate. Climatologists have tended to operate at the scale of GCM output i.e. annual means for several thousand square kilometres. There is also a real danger that large scale science is grossly and unnecessarily oversimplifying some issues. Examples were quoted of climatologists expressing the view that one value of one parameter was sufficient to describe the moisture status of the 'soil' (sic) over the whole of Europe for climate modelling purposes. This highlights a serious divide, which needs to be crossed much more quickly than seems to be happening at present. It is recognised that the resolution of GCM output is unlikely to improve significantly in the next decade, so other measures are needed. A possible tool is the weather generator. This can use existing runs of long-term climate data to produce surrogate climates

261 Where Next?

which match the general predictions of GCM modelling.

Realistic seasonal

perturbations can be included, as can weather patterns based on different frequencies of extreme events. There is a clear need to agree on the use of a small number of robust weather generators, derived from good, long-term baseline climate data for an agreed range of climate zones over the globe. The output from such weather generators should be made widely available to the terrestrial science community and, where appropriate, access to the computer code and datasets should also be possible.

Further, climatologists and ecosystem modellers

should be encouraged to interact much more closely so as to define very clearly what one can expect from the other. The question of spatial resolution is equally valid for modelling changes in plant-soil communities. Much vegetation modelling has been derived from very site-specific research, and is data intensive. Although the same plant(s) might have been considered at more than one site, the geographical dimensions of the modelling are likely to have been small. For example, the well-known ARCWHEAT crop model was developed at two principal sites in southern England, and has not been widely tested outside cool, humid north-west Europe. Extension of crop modelling to larger geographic areas requires the development of simpler models. These can be validated against more complex, site-specific models so that the increase (if any) in the uncertainty of the modelling can be established. The extension of such modelling to larger geographic areas requires robust methods for the interpolation of data and/or the derivation of data from more widely available data. A good example is the calculation of a soil water release curve (slow and expensive to determine; datasets very limited) from particle size data (rapid, cheap; one of the most widely measured soil properties). The most immediately visible effects of climate change will almost certainly be on plant communities, especially their composition and distribution. These changes will be mediated partly by changes in soils, but the latter will themselves undergo other changes because of plant community change. The system is dynamic and requires dynamic, interactive modelling. This topiC is seen as one of the most important targets for future research.

262 Loveland et al.

Carbon, nitrogen and nutrients Although the soil-vegetation carbon pool is relatively small compared with that of the oceans, it is much more potentially labile in the short term. Vegetation in this context includes plant remains found predominantly in the bogs, peats, tundras and boreal soils of the colder, wetter parts of the northern hemisphere, and in the litter of forest floors in warmer regions. The likely fate of this material is extremely important in terms of the near-future global atmosphere greenhouse gas inventory.

Several

uncertainties were identified:

1. The size of this carbon pool (and its entrapped greenhouse gas pool) is still relatively poorly known. 2. The chemical forms of organic carbon in these materials is not well understood. 3. The reaction mechanisms, including kinetics, and the influence of the soil microbial population and other soil biota (see below), on the breakdown of these organiC materials, is poorly understood. 4. The sensitivity of this pool of material to climate change is uncertain. Is it more sensitive to slight changes in temperature, or in soil wetness or, most likely, both; but in what proportion. 5. What will be the effect of a changed vegetation cover on these organic materials . .6. Can the pool of organic carbon in soils be manipulated in meaningful terms e.g. by deliberately changing the vegetation. It is widely advocated that planting trees is beneficial in terms of carbon storage. Long term benefit is likely to accrue, however, only if the below ground storage of carbon can be substantially changed. Little is known about how this might be done, which plants are most likely to be most effective, and which combinations of soiVclimate/hydrology should be targeted. 7. Can the soil carbon pool be manipulated by adjusting the soil nitrogen cycle? If so, how, and in which soil-crop systems is it likely to be most effective, and by what order of magnitude. There is less certainty about other nutrients such as phosphorus and sulphur. Perhaps this reflects neglect of their place in research into soils and climate change. In general, it was believed at the Workshop that release of nutrients from soils other than nitrogen associated with organic matter (and possibly phosphorus) - is

263 Where Next?

unlikely to change significantly in the time-scale over which marked climate change is expected to occur. However, this assumption is by no means fully tested, and some background research into nutrient release through different weathering rates is justified. This would be possible through a study of nutrient status of analogue soils

i.e. those soils with similar parent materials and land-use history, but differing from each other by slight differences in climate. Land quality, vegetation systems and other biota

Large parts of the land surface soils and associated vegetation systems are fragile, whether in terms of cultivated crops, or of more natural systems. Good examples of fragile landscapes can be found around the Mediterranean, in much of Africa, parts of the Indian sub-continent and the Americas, although this Workshop considered only the first of these. Such systems are acutely sensitive to small absolute, but large relative, changes in climate - especially in precipitation (amount and distribution). Such landscapes can be particularly affected by extreme events such as droughts or especially heavy storms; even more so if these extreme events occur in succession. A major consequence of extreme events is often soil erosion, with its resultant permanent loss of soil quality. Already major increases in the amount of erosion in some parts of the world threatens the ability of the soils to produce sufficient food.

How will climate change affect this? Will currently stable soils

become unstable? There is a pressing need to extend research into soil management factors i.e. water conservation practices, tillage practices and cropping patterns, so as to better protect fragile ecosystems against the more extreme events which may accompany climate change.

This is particularly true for dryland arable

cropping systems, afforestation strategies, and the regions of shallow, sloping soils susceptible to erosion.

There is often a rather facile assumption that the solution to soil-vegetation problems in drier areas is to increase the amount of irrigation. This begs the question as to whether there are enough water resources to do this, but also ignores the likely risk of soil salinization. chemistry and tillage.

Much is known about salt-affected soils in terms of physics,

264 Loveland et al.

However, a concerted effort is required to develop a model or predictive tool which will allow the estimation of salinization risk against a given cropping system/irrigation policy/climate change combination, for particular soils and landscapes. Finally, there is also the question of plant community structure and soil pattern, Biologists know that plant expressed in terms of landscape fragmentation. communities are not fixed combinations; they very often represent some kind of short-term equilibrium or quasi-equilibrium grouping, corresponding to a particular soil-climate system. This of itself reinforces the need to study such systems at high spatial resolution as mentioned earlier. Changes in any of the factors which make up the set of properties controlling the current assemblage of plants in a particular area will cause that community to change at different rates for different species. Very little attention has been given to this kind of 'disequilibrium' scenario, and the effects it might have on the landscape. Some of the plants involved in community break-up, re-formation and migration will be of considerable economic importance. There is also the related question of habitat occupation. Are all possible habitats occupied by the appropriate plant community? If not, why not? What is the role of soil in this equation? Can habitat occupation be predicted spatially, and what are the implications for future landscapes and biodiversity? There is an urgent requirement that multidisciplinary studies of this kind should be targeted at major plant-soil associations in selected areas of the globe e.g. humid arable agriculture, savannah/grassland, lowland and montane forest, dryland agriculture and so on. Relatively little attention is being paid to the role of soil biota in relation to soil physical and chemical conditions or properties. The importance of the soil microbial population has already been mentioned, but studies are also required of other soil fauna and flora. Some pathogens are known to be soil-borne, and others will follow their host plants, which in turn will affect the distribution of many plant diseases and their vectors. Given the interaction between soil properties and plant communities, there is clearly room for further predictive work in the changes which might occur in major crop pests and diseases.

265 Where Next?

CONCLUSIONS This paper identifies areas in which soil scientists could cooperate with others to improve our ability to predict the effects of climate change on the land surface of the earth and its vegetation cover. The purpose of this research is to minimise, where possible, the consequences which some of these changes could have for many societies. In terms of priorities, the tasks seem to us to be: •

Better integration of available data sources, and clear identification of the gaps where more information is required. This needs to be, and could be, achieved over the next five years.

The improvement of databases means that soil

scientists should be working much more closely with other members of the climate change community in deciding how those databases might be used. •

A parallel exercise needs to identify which models are worth supporting from the point of view of predicting the nature and effect of climate change on the soils and vegetation cover. Parallel research should be encouraged to identify those mathematical techniques which will allow robust interpolation of data between collection points, and the derivation of necessary data from limited measurements. Emphasis in this modelling approach should be given to food crops and their water requirements, with a special regard for fragile land use systems. The realities of irrigation as a modifier of climate change effects of soils should be put into perspective. A programme of this nature could be undertaken within the next decade to the point where good policy guidance is possible.

•

We need to consider urgently how the soil-vegetation carbon- storage and release system is likely to respond to small changes in climate over the next 50 to 100 years. Little is known about below ground carbon storage, turnover rates of carbon of different degrees of reactivity (and how these relate to decomposition processes), and the role played by the soil microbial biomass in these systems.

An integrated research programme over the next ten years

should address many of these questions. •

There is insufficient research into the way in which current, seemingly coherent, plant communities will fragment, migrate and re-form under the influence of climate change, and how habitats occupy landscapes within the constraints imposed by the soil pattern.

A research programme to understand these

266 Loveland et al.

processes better should be targeted at several of the world's major plant-soil systems over the next 15 to 20 years. This applies equally to studies of potential changes in the distribution of pathogens, pests and disease vectors for major plants.

Such studies could be run in parallel with those of changes in plant

community structure. •

There should be some long-term research into the possible effects of climate change on basic soil properties e.g. how changed weathering rates might affect soils in temperate regions, what happens to buried carbon, will significant inorganic nutrient release occur from mineral sources? Such a programme could well be undertaken in parallel with the construction of better databases, and with better understanding of the processes involved.

•

The climate change impacts research community should assist in the development of practical systems of agriculture for the more fragile soil/cropping systems of the world, by showing as clearly as possible what changes are likely to occur where. This is more a ~hange of emphasis and explanation than research per se. The development of improved cultivation systems, and/or their introduction into parts of the world where they might be regarded a new or novel departures from existing practice, should involve careful study of the soils' context into which they will be introduced. sustainable development.

•

This is at the heart of future

Much of what has been outlined could be achieved by bringing people together who otherwise might remain in ignorance of each others skills and data. Such interactions are being encouraged through such Intergovernmental Organisations as the IPee and IGBP, but it was felt that more recognition needs to be taken within the national research programmes, that there could be an international dimension to the work.

Research into the effects of climate change impacts on the land surface of the globe is a multi-disciplinary activity. Soil scientists can contribute in several areas, as summarised above, but must not operate in isolation. They should, however, identify clearly the contribution they expect to make, and offer their contribution loudly and clearly. They should not wait to be asked, or someone else will do the work instead and probably less well.

BIBLIOGRAPHY

Abelson, P.H. (1978). Organic matter in the earth's crust. Annual Review of Earlh Planet Science, 6, 325-351.

Adams, R.M., Rosenzweig, C., Peart, R.M., Ritchie, J.T., McCarl, B.A., Glyer, J.D., Curry, R.B., Jones, J.w., Boote, K.J. and Allen, L.H. (1990). Global climate change and U.S. agriculture. Nature, 345, 219-224. Addiscott, T.M. (1977). A simple computer model for leaching in structured soils. Journal of Soil Science, 28, 554-563.

Addiscott, T.M., Heys, P.J. and Whitmore, A.P. (1986). Application of simple leaching models in heterogeneous soils. Geoderma, 38, 185-194. Addiscott, T.M. and Whitmore, A.P. (1987). Computer simulation of changes in soil mineral nitrogen and crop nitrogen during autumn, winter and spring. Journal of Agricultural Science, Cambridge, 109, 141-157.

Addiscott, T.M. and Whitmore, A.P. (1991). Simulation of solute leaching in soils of differing permeabilities. Soil Use and Management, 7, 94-102. Ahmad, N. (1990). The agricultural environment in Latin America and the Caribbean and the greenhouse effect. In: (H.W. Scharpenseel, M. Schomaker and A. Ayoub, eds). Soils on a warmer Earlh, Developments in Soil Science, 20, 249265. Allen, L.H. Jr. (1991). Effects of increasing carbon dioxide levels and climate change on plant growth, evapotranspiration, and water resources. In: Managing water resources in the West under conditions of climate uncerlainty. National

Academy Press, Washington D.C. pp.101-147. Anderson, J.M. (1992). Responses of soils to climate change. In: (F.1. Woodward, ed.). The ecological consequences of global climate change. Advances in Ecological Research, 22, 163-210.

Arbeitsgruppe

Bodenkunde

(1982).

Bodenkundliche

Kartieranleitung.

Schweizerbart'sche Buchhandlung Hannover, pp. 331. Armentano, TV. (1980). Drainage of organic soils as a factor in world carbon cycle. Bioscience, 30, 825-830.

Armentano, T.V. and Menges, E.S. (1986). Patterns of change in carbon balance of organic soil-wetlands of the temperate zone. Journal of Ecology, 74, 755-774. Armstrong, A.C. (1993). Modelling the response of in-field water tables to ditch levels imposed for ecological aims, a theoretical analysis. Agriculture, Ecosystems and Environment, 43, 345-351.

268

Armstrong, A.C. and Garwood, E.A. (1991). Hydrological consequences of artificial drainage of grassland. Hydrological Processes, 5,157-174. Armstrong, AC., Portwood, A.M. and Castle, D.A. (1993). Simple models to predict field soil water regimes in the presence of ditch water levels managed for environmental aims. In: (E. Lorre, ed.). IC/o Workshop on Subsurface Drainage Simulation Models, The Hague, September 1993, pp.147-157.

Arnold, R.W. (1968). Pedological significance of lithologic discontinuities. Transactions of the 9th International Congress of Soil Science, Adelaide, 4, 595-603. Arnold, R.W. (1990). Fractal dimensions of some soil map units. Transactions of the 14th Congress of the International Soil Science Society, Kyoto, 5, 92-97.

Arnold, R.w. and Schargel, R. (1978). Importance of geographic soil variability at scales of about 1:25,000 - Venezuelan examples. In: Diversity of soils in the tropics. ASA Special Publication 34, ASA, CSSA, SSSA, Madison, WI, pp. 13-

27. Arnold R.W., Szabolcs I. and Targulian V.O. (eds) (1990). Global soil change. IIASA, Laxenburg, Austria. Aronoff, S. (1985). The minimum accuracy value as an index of classification accuracy. Photogrammetric Engineering and Remote Sensing, 51, 99-111. Arya, L.M., Farrell, D.A. and Blake, G.R. (1975). A field study of soil water depletion patterns in presence of growing soy bean roots. I. Determination of hydraulic properties of the soil. Soil Science Society of America Journal, 45, 1023-1030. Axelson, D.E. (1985). Solid State Nuclear Magnetic Resonance of Fossil Fuels, Canada Government Publishing Centre, Supply and Services, Ottawa, 7. Ayanaba, A and Jenkinson, D.S. (1990). Decomposition of carbon-14 labelled ryegrass and maize under tropical conditions. Soil Science Society of America Journal, 54, 112-115.

Baldy, C.H. (1973). Etude de peuplements de ble tendre (Triticum aestivum L. em. Thie/~:

croissance, repartition de la lumiere et assimilation nette. These Doc.

Ing. Universite Paris Sud, Orsay, pp. 160. Barraclough, P.B. and Leigh, R.A (1984). The growth and activity of winter wheat roots in the field: the effect of sowing date and soil type on root growth of highyielding crops. Journal of Agricultural Science, Cambridge, 103, 59-74. Barrow, E. (1993). Future scenarios of climate change in Europe. In: (G.J. Kenny, P.A. Harrison and M.L. Parry, eds). The effect of climate change on agricultural and horticultural potential in Europe. Research Report No.2,

Environmental Change Unit, Oxford University, pp. 11-40.

269

Begon, J.C. and 15 other authors. (1990). Evaluation des contraintes sol-climat sur la faisabilite des cultures. Application au ma'is-grain. Rapport de fin de travaux, CCE, Bruxelles, No. 5011, pp. 120. Bellivier, D. (1993). Etude du modele de culture "EPIC". Integration dans un programme de simulation du bilan hydrique. DAA, INRA-INAPG, Paris, pp. 40. Bennett, R.M. (ed.) (1989). The 'Greenhouse Effect' and UK agriculture, CAS Paper 19, Centre for Agricultural Strategy, Reading. Berlin, A., Wolff, A.H. and Hasegawa, Y. (eds) (1979). The Use of Biological Specimens for the Assessment of Human Exposure to Environmental Pollutants. Martinus Nijhof Publishers, The Hague. pp. 368.

Berner, RA (1974). Kinetic models for the early diagenesis of nitrogen, sulphur, phosphorus and silicon in anoxic marine sediments, In: (E.D. Goldberg, ed.). Wiley, NY, The Sea, 5, 427-450. Berner, RA Lasagna, A.C. and Garrels, R.M. (1983). The carbonate-silicate cycle and its effect on atmospheric carbon dioxide over the past 100 million years. American Journal of Science, 283, 641-683.

Billings, W.o. (1987). Carbon balance of Alaskan tundra and taiga ecosystems: past, present, and future. Quaternary Science Reviews, 6,165-177. Birch, H.F. (1960). Nitrification in soils after different periods of dryness. Plant and Soil, 12, 81-96.

Biscoe, P.V. and Gallagher, J.N., (1977). Weather, dry matter production and yield. In: (Landsberg, J.J. and Cutting, C.v. eds). Environmental effects on crop physiology. Academic Press, New York, pp. 75-100.

Blet-Charaudeau, C., Muller, J. and Laudelout, H. (1990). Kinetics of carbon dioxide evolution in relation to microbial biomass and temperature. Soil Science Society of America Journal, 54, 1324-1328. BMFT.

(ed.)

(1988).

Bundesministerium

fOr

Forschung

und

Technologie.

Umweltprobenbank, bericht und bewertung der pilotphase. Springer-Verlag,

Berlin Heidelberg. pp. 158. BMU/U.S.EPA (1989). Specimen banking - a guide for an international standard, Ottawa. pp. 8. (unpublished). Boardman, J., Evans, R., Favis-Mortlock, D.T. and Harris, T.M. (1990). Climate change and soil erosion on agricultural land in England and Wales. Land Degradation and Rehabilitation, 2(2), 95-106.

Boardman, J. (1993). The sensitivity of Downland arable land to erosion by water. In: (D.S.G. Thomas and R.J. Allison, eds). Landscape Sensitivity. Wiley, Chichester, pp. 211-228.

270

Boardman, J. and Favis-Mortlock, D.T. (1993a). Climate change and soil erosion in Britain. Geographical Journal, 159(2), 179-183. Boardman, J. and Favis-Mortlock, D.T. (1993b). Simple methods of characterizing erosive rainfall with reference to the South Downs, southern England. In: (S. Wicherek, ed.). Farmland Erosion in Temperate Plains, Environments and

Hills. Elsevier Science Publishers B.V., Amsterdam, pp. 17-29. Bonhomme, R. and Ruget, F. (1991). Modelisation du fonctionnement d'une culture de ma"is: cas de CORNGRO et de CERES - MAIZE. In: Physiologie et production du mars. Colloque INRA. Universite Paris Sud XI. AGPM Pau, pp. 365-372. Bonhomme, R., Ruget, F. and Derieux, M. (1991). Prise en compte de la photoperiode dans CERES-MAIZE. In: Physiologie et production du mars. Colloque INRA. Universite Paris Sud XI. AGPM Pau, pp. 411-418. Borg, H. and Grimes, D.W. (1986). Depth development of roots with time: an empirical description. American Society of Agricultural Engineers, 0001-2351, pp.194-197. Bormann, B.T., Bormann, F.H., Bowden, W.B., Pierce, R.S., Hamburg, S.P., Wang, D., Snyder, M.C., Li, C.Y. and Ingersoll, R.C. (1993). Rapid N2 fixation in pines, alder, and locust: evidence from the sandbox ecosystem study. Ecology, 74, 583-598.

Bormann, F.H. and Likens, G.E. (1979). Pattem and Process in a Forested Ecosystem. Springer-Verlag, New York. Botterweg, P.F. (1990). The effect of frozen soil on erosion - a model approach. In: (K.R. Cooley, ed.). Proceedings of the International Symposium: Frozen soil impacts on agricultural, range, and forest lands, March 21-22, 1990, Spokane, Washington. CRREL Special Report, 90-1, pp. 135-144. Botterweg, P. (In press). Modelling the effects of climate change on runoff and erosion in the central-southern part of Norway. Soil and Technology, Special Issue. Bouaziz, A. and Bruckler, L. (1989). Modeling wheat seedling growth and emergence: seedling growth affected by soil water potential. Soil Science Society of America Joumal, 53, 1832-1838.

Boudreau, B.P. and Ruddick, B.A. (1991). On a reactive continuum representation of organic matter diagenesis. American Journal of Science, 291, 507-538. Bouma, J., Hillel, D., Hole, F.D. and Amerman, C.R. (1971). Field measurement of unsaturated hydraulic conductivity by infiltration through artificial crusts. Soil Science Society of America Proceedings, 35, 362-364.

271

Bouma, J. and Loveday, J. (1988). Characterizing soil water regimes in swelling clay soils. In: (L.P. Wilding and R. Puentes, eds). Vertisols: their distribution, properties, classification, and management.

Texas A & M Press, College

Station, TX, pp. 83-96. Bouwman, A.F. and Sombroek, W. (1990). Inputs to climate change by soil and agriculture related activities. In: (E.W. Scharpenseel, M. Schomaker, and A Ayoub, eds). Soils on a warmer Earth. Developments in Soil Science, 20, pp. 15-30. Bouwman, AF. (1990). Soils and the greenhouse effect. John Wiley, Chichester. Bradbury, N.J. and Powlson, D.S. (1994). The potential impact of global environmental change on nitrogen dynamics in arable systems. (This volume). Bradbury N.J., Whitmore A.P., Hart P.B.S. and Jenkinson, D.S. (1993). Modelling the fate of nitrogen in crop and soil in the years following application of 15 N_ labelled fertilizer to winter wheat. Journal of Agricultural Science, Cambridge, 121,363-379. Brammer H. and Brinkman R. (1990). Changes in soil resources in response to a gradually riSing sea-level. In: (E.W. Scharpenseel, M. Schomaker, and A. Ayoub, eds). Soils on a warmer Earth. Developments in Soil Science 20, 145156. Bramyd, T. (1980). The role of peatlands for the global carbon dioxide balance. Proceedings of the 6th International Peat Congress, Duluth, MN. pp. 9-11.

Brignall, AP. and Rounsevell, M.D.A. (1994). The effects offuture climate change on crop potential and soil tillage opportunities in

England and Wales.

Environmental Change Unit, University of Oxford, and the Soil Survey & Land Research Centre, Silsoe, Joint Research Report. ISBN 1-874370-04-4, pp. 94. Brinkman,

R.

(1990).

Resilience

against

climate

change?

Soil

minerals,

transformations and surface properties, Eh, pH. In: (H.W. Scharpenseel, M. Shomaker and A Ayoub, eds). Soils on a warmer Earth. Developments in Soil Science, 20, pp 51-60. Brinkman, R. and Brammer, H. (1990). The influence of changing climate on soil properties. Transactions of the 14th International Society of Soil Science Congress, Kyoto, pp. 283-288.

Brisson, N., (1987). SOYGRO. Un modele de croissance et de rendement du soja, son fonctionnement et son utilisation pour definir des potentialites. CIRAME, pp.51. Brisson, N. (1989). Modele de simulation de la culture du soja et de son fonctionnement hydrique. These. INRA Paris-Grignon, pp. 90.

272

Brisson, N. and Delecolle, R. (1991). Developpement et modeles de simulation de cultures. Agronomie, 12, 253-263. Brisson, N. and Voltz, M. (1991). Le comportement hydrique des modeles de simulation de la culture du ma'is. Analyse comparative de trois modeles existants. In: Physiologie du mais et production du mars. Colloque INRA. pp. 393-402. Budyko, M.1. (1989). Empirical estimates of imminent climatic changes. Soviet Meterorology and Hydrology, 10, 1-8. Bultot, F., Coppens, A., Dupriez, G.L., Gellens, D. and Meulenberghs, F. (1988). Repercussions of a CO2 doubling on the water cycle and on the water balance - a case study for Belgium. Journal of Hydrology, 99, 319-347. Cammell, M.E. and Knight, J.D. (1992). Effects of climatic change on the population dynamics of crop pests.

In:

(F.1.

Woodward,

ed.).

The

ecological

consequences of global climate change. Advances in Ecological Research, 22,

117-162. Carter, T.R., Parry, M.L., Nishioka, S. and Harasawa, H. (1992). Preliminary guidelines for assessing impacts of climate change. Environmental Change

Unit, University of Oxford, UK and the Centre for Global Environmental Research, Tsukuba, Japan. pp. 28. Carter, M.R., Parton, W.J., Rowland, I.C., Schultz, J.E. and Steed, G.R. (1993). Simulation of soil organic carbon and nitrogen changes in cereal and pasture systems of southern Australia. Australian Journal of Soil Research, 31, 481491. CEC. (1985). Soil map of the European Communities, 1/1,000,000. Commission of the European Communities, DGVI (Agriculture), Coordinator of Agricultural Research. Brussels. Cess, R.D., and 19 authors. (1989). Interpretation of cloud-climate feedback as produced by 14 atmospheric general circulation models. Science, 245, 513516. Charles-Edwards, D.A. (1982). PhYSiological determinants of crop growth. Academic Press, Sydney. Chaudhuri, U.N., Kirkham, M.B. and Kanemasu, E.T. (1990). Root growth of winter wheat under elevated carbon dioxide and drought. Crop Science, 30, 853-857. Childs, S.W., Gilley, J.R. and Splinter, W.E. (1977). A simplified model of corn growth under moisture stress. Transactions of the American Society of Agricultural Engineers, 20(5), 858-865.

273

Choisnel, E.K. (1991). Changements climatiques et impacts sur les forets et I'agriculture. La Jaune et la Rouge, pp. 43-46. Choisnel, E.K. and Jacquart, C. (1991). Les methodes de suivi du bilan hydrique du sol utilisees

a la

Meteorologie Nationale. In: Milieux poreux et transferts

hydriques. Bulletin du G.F.H.N., 28, 11-33.

Chopart, J.L. and Vauclin, M. (1990). Water balance estimation model: field test and sensitivity analysis. Soil Science Society of America Journal, 54, 1377-1384. Climate Change. (1992). Climate change: a threat to global development. Acting now to safeguard the future. Economica Verlag - Verlag C.F. MOiler, Bonn-

Karlsruhe. Climate Change Impacts Review Group. (1991). The potential effects of climate change in the United Kingdom. HMSO, London. Conte, M. (1993). Climatological investigations: regional scale short term climatology. In: Mediterranean Desertification and Land Use - MEDALUS I, Final Report. Commission of the European Communities, DGX/I (Environment): EPOCHCT90-0014-(SMA), pp. 1-16.

CORINE. (1990). Soil erosion risk and important land resources. Commission of the European Communities, DGXII (Environment, Consumer Protection and Nuclear Safety), Brussels, pp. 97.

Cottle, H.J. (1932). Vegetation on north and south slopes of mountains in S.E. Texas. Ecology, 13,121-134.

da Cunha, L.V. (1989). Water resources situation and management in the EEC. Hydrogeologie, 2, 57-69.

Curtis, P.S., Balduman, L.M., Drake, B.G. and Whigham, D.F. (1990). Elevated atmospheric CO 2 effects on belowground processes in C3 and C4 estuarine marsh communities. Ecology, 71, 2001-2006. Dahlman, A.C. (1985). Modeling needs for predicting responses to CO2 enrichment: plants, communities and ecosystems. Ecological Modelling, 29, 77-106. Danalatos, N.G. (1993). Quantified analysis of selected land use systems in the Larissa region, Greece. Ph.D. Thesis, Agricultural University of Wageningen. Wageningen, pp. 370. Daniels, A.B. and Hammer, R.D. (1992). Soil geomorphology. John Wiley & Sons, New York. Dansgaard, W., Johnsen, S.J., Clausen, H.B., Dahl-Jensen, D., Gundestrup, N.S., Hammer, C.U., Hvidberg, C.S., Steffensen, J.P., Sveinbj6rnsdottir, A.E., Jouzel, J. and Bond, G. (1993). Evidence for general instability of past climate from a 250-kyr ice-core record. Nature, 364, 218-220.

274

Daugherty, L.A., Hanna, W.E. and Arnold, R.W. (1975). Geomorphology of a glaciated first-order valley in south-central New York. Soil Science Society of America Proceedings, 39, 710-716.

Davidson, E.A. (1991). Fluxes of nitrous oxide and nitric oxide from terrestrial ecosystems. In: (J.E. Rogers and W.B. Whitman, eds). Microbial Production and Consumption of Greenhouse Gases: Methane, Nitrogen Oxides, and Halomethanes. American Society for Microbiology, Washington, pp. 219-235.

Davidson, E.A. (1992). Pulses of nitric oxide and nitrous oxide flux following wetting of dry soil: an assessment of probable sources and importance relative to annual fluxes. Ecological Bulletins, 42, 149-155. Davidson, E.A. (1993). Soil water content and the ratio of nitrous oxide to nitric oxide emitted from soil. In: (R.S. Oremland, ed.). The Biogeochemistry of Global Change. Chapman and Hall, New York, pp. 369-386.

Davidson, E.A. and Ackerman, I.L. (1993). Changes in soil carbon inventories following cultivation of previously untilled soils. Biogeochemistry, 20,161-193. Davidson, E.A., Hart, S.C. and Firestone, M.K. (1992). Internal cycling of nitrate in soils of a mature coniferous forest. Ecology, 73, 1148-1156. Davidson, E.A. and Swank, W.T. (1990). Nitrous oxide dissolved in soil solution: an insignificant pathway of N loss from a southeastern hardwood forest. Water Resources Research, 26,1687-1690.

Day, J.W. and Templet, P.H. (1989). Consequences of sea level rise: implications from the Mississippi delta. Coastal Management, 17, 241-257. Degens, E.T. and Hunt, J.M. (1964). Thermal stability of amino compounds in recent and ancient sediments, humic acids and kerogen concentrates. International Meeting. on Organic Processes in Geochemistry, Paris, pp. 28-30. Delecolle, R., Couvreur, F., Pluchard, F. and Varlet-Grancher, C. (1985). About the leaf day length model under French conditions. In: (W. Day and R.K. Atkin, eds). Wheat growth and modelling. NATO ASI Series A, Vol. 86, pp. 25-31. Plenum, New York. Dereppe, J., Boudou, J.P., Moreaux, G. and Durand, B. (1983). Structural evolution of a sedimentologically homogenous coal series as a function of carbon content by solid state 13C NMR. Fuel, 62, 575-579. Detweiler, R.P. and Hall, CAS. (1988). Tropical forests and the global carbon cycle. Science, 239, 42-47. Diericks, J.C., Belmans, C. and Pauwels, P. (1986). SWATRER: a computer package for modelling the field water balance. Laboratory of Soil and Water Engineering, K.U. Leuven, Reference Manual 2, pp. 141.

275

Dinwoodie, GoO. and Juma, N.G. (1988). Factors affecting the distribution and dynamics of 14C in two soils cropped to barley. Plant and Soil, 110, 111-121 Doorenbos, J. and Kassam, AH. (1979). Yield response to water. FAO Irrigation and Drainage Paper, No. 33. Rome.

Dos Santos, H. (1978). Some strategies of quality control for reconnaissance soil survey. Unpublished MSc thesis., Cornell University, Ithaca, NY. Dregne, D.H. (1990). Impact of climate warming on arid region soils. In: (H.w. Scharpenseel, M. Shomaker and A. Ayoub, eds). Soils on a warmer Earth. Developments in Soil Science, 20, 177-184.

Driessen, P.M. and Konijn, N.T. (1992). Land use systems analysis. CIP-Data, Koninklijke Bibliotheek, The Hague. Dunfield, P., Knowles, R., Dumont, R. and Moore, T.R. (1993). Methane production and consumption in temperate and subarctic peat soils: response to temperature and pH. Soil Biology and Biochemistry, 25(3), 321-326. Durand, R. and Legros, J.P. (1981). Cartographie automatique de I'energie solaire en fonction du relief. Annales Agronomiques, 1(1), 31-39. Eckhardt, F.E.W. (1984). Solubilization, transport and deposition of mineral cations by micro-organisms - efficient rock weathering agents. In: (J.1. Drever, ed.). The Chemistry of Weathering. NATO ASI Series C: Mathematical and Physical Science, 149,161-173.

EEC (1990). Environment Project 1990-94: Specific Program in the Field of Environment. Official Journal No. L 192, pp. 16-91. Eijsackers, J.P. and Hamers, T. (1993). Integrated soil and sediment research: a basis for proper protection. Kluwer Academic Publishers, Dordrecht.

Esser, G. (1984). The significance of biospheric carbon pools and fluxes for the atmospheric CO2 : a proposed model structure. In: (H. Lieth, R. Fantechi and H. Schnitzler, eds). Interaction between Climate and Biosphere. Proceedings of a Commission of the European Communities Symposium, Osnabruck, Germany. Esser, G. (1989). Global land-use changes from 1860 to 1980 and future projections to 2500. Ecological Modelling, 44, 307-316. Eswaran, H., Van Den Berg, E. and Reich, P. (1993). Organic carbon in soils of the world. Soil Science SOCiety of America Journal, 57, 192-194. Eswaran, H. and Reich, P. (1993). Variability in soil and climatic conditions in IBSRAM's vertisol network sites in Africa. In: (E. Pushparajah and C. Elliot, eds). Report on the 1992 Annual Meeting on: Management of Vertisols in Africa. IBSRAM, Bangkok. pp. 141-166.

276

FAO. (1978). Report of the agro-ecological zones project 1. Methodology and results for Africa. World Soils Research Report, 48, Rome. FAO. (1983). Guidelines: land evaluation for rainfed agriculture. FAO Soils Bulletin, 52. FAO, Rome. FAO. (1988). FAO production yearbook 1987. FAO, Rome. Favis-Mortlock, D.T., Evans, A., Boardman, J. and Harris, T.M. (1991). Climate change, winter wheat yield and soil erosion on the English South Downs. Agricultural Systems, 37(4), 415-433. Favis-Mortlock, D.T. and Boardman, J. (submitted). Soil erosion under a changing climate. Proceedings of the Memorial Symposium to Prof. J. de Ploey: Experimental Geomorphology and Landscape Ecosystem Changes. Leuven, Belgium, March 1993, pp. 1990-2050. Feddes, A.A., Kowalik, P.J. and Zaradny, H. (1978). Simulation of field water use and crop yield. Simulation Monographs PUDOC. Wageningen. pp. 189. Feddes, A.A. (1985). Crop, water use and dry matter production: state of the art. Colloque INRA. Les besoins en eau des cultures. pp. 222-234. Fernandez, C.J. and McCree, K.J. (1991). Simulation model for studying dynamics of water flow and water status in plants. Crop Science, 31, 391-398. Fisher, W.A. (1990). Influence of climatic change on development of problem soils, especially in the alluvial domains. In: (H.W. Scharpenseel, M. Shomaker and A. Ayoub, eds). Soils on a warmer Earth. Developments in Soil Science, 20, 89-99. Flaig, W. (1972). Biochemical factors in coal formation. In: (H.A. von Gaertner and H. Wehner, eds). Advances in Organic Geochemistry. Pergamon, Elmsford, N.Y. pp.197-232. Fleming, P.M. (1991). Climate change and climate variability. An examination of its impact in water resource planning. CSIRO, Canberra. Franzmeier, D. P., Lemme, G.D. and Miles, A. J. (1985). Organic carbon in soils of north central United States. Soil Science Society of America Journal, 49, 702708. Freeman, C., Lock, M.A. and Reynolds, B. (1993). Fluxes of CO2 , CH 4 and N20 from Welsh peatland following simulation of water table draw-down: potential feedback to climate change. Biogeochemistry, 19,51-60. Freidli, H., Lotscher,'H., Oeschger, H. and Siegenthaler, U. (1986). Ice core record of the 13C/12C ratio of atmospheric CO 2 in the past two Centuries. Nature, 324, 237-238.

277

Fridland, V.M. (1976). The soil cover pattern: problems and methods of investigation. In: (Fridland, V.M., ed.). Soil combinations and their genesis. Amerind Publishing Co., New Delhi, pp. 3-31. Fromin, N. (1992). Revue des modeles de culture

a des fins

d'integration dans un

modele de bilan hydrique. Rapport de stage INRA-ISIM Chambery, pp. 50. Fyfe, W.S. (1984). The geochemical cycle. In: Environmental Geochemistry, Mineral Association of Canada, short course handbook, 10. Royal Ontario Museum, Toronto, pp. 1-14. Gardner, W. R. (1958). Some steady state solutions of the unsaturated moisture flow equation with application to evaporation from a water table. Soil Science, 85, 228-232. Garrels, R.M. and Perry, EA Jr. (1974). Cycling of carbon, sulphur and oxygen throughout geological time. In: (E.D. Goldberg, ed.). The Sea, 5, 303-336. van Genuchten, M. Th. (1980). A closed-form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Science Society America Journal, 44, 892-898. Gifford, R.M. (1992). Interaction of carbon dioxide with growth-limiting environmental factors in vegetation productivity: implications for the global carbon cycle. Advances in Bioclimatology, 1, 24-58. Ginevan, M.E. (1979). Testing land-use accuracy - another look. Photogrammetric Engineering and Remote Sensing, 45,1371-1377. Gleick, P. (1986). Methods for evaluating the regional hydrologic impacts of global climatic changes. Journal of Hydrology, 88, 97-116. Goodess, C.M. and Palutikof, J.P. (1992). The development of regional climate scenarios and the ecological impact of greenhouse gas warming. Advances in Ecological Research, 22, 33-62. Goodroad, L.L. and Keeney, D.R. (1984). Nitrous oxide emissions from forest, marsh, and prairie ecosystems. Journal of Environmental Quality, 13, 448-452. Gosse, G., Varlet-Granchet, C., Bonhomme, R., Chartier, M., AIIirand, J.M. and Lemaire, G. (1986). Production maximale de matiere seche et rayonnement solaire intercepte par un couvert vegetal. Agronomie, 6(1), 47-56. Goudriaan, J. and Ketner, P. (1984). A simulation study for the global carbon cycle, including man's impact on the biosphere. Climatic Change, 6, 167-192. Goudriaan, J., van Laar, H.H., van Keulen, H. and Louwerse, W. (1985). Photosynthesis, CO 2 and plant production. In: (W. Day and R.K. Atkin, eds). Wheat growth and modelling. NATO ASI Series A, Vol. 86, pp. 107-122. Plenum, New York.

278

Grainger, A. (1992). Characterization and assessment of desertification processes. In:

(G.P.

Chapman,

ed.).

Desertified grasslands:

their biology and

management. Academic Press, London. pp. 17-33.

Grammerer, K., (1989) Respiration of soil and vegetation in grassland, M.Sc. Thesis, Department of Agronomy, University of Nebraska, Lincoln.

Greenwood, D.J., Nye, P.H. and Walker, A. (1990). Quantitative theory in soil productivity

and

environmental

pollution,

a

discussion.

Philosophical

Transactions of the Royal Society of London, B, 29, 307-428.

Gregorich, E.G., Ellert, B. H., Angers, DA, Carter, M.R., Monreal, C. M. and Donald, R.G. (1993a). Impact of management on the quantity and composition of organic matter in soils of Eastern Canada. Research Report, Canadian Society of Soil Science Annual Meeting, St. John's, Newfoundland, Canada.

Gregorich, E.G., Carter, M.A., Angers, D.A., Monreal, C. M. and Ellert, B. H. (1993b). Towards a minimum data set to assess soil organic matter quality in agricultural soils. Canadian Joumal Soil Science (In press). Grenon, M. and Batisse, M. (1989). Futures for the Mediterranean Basin: The Blue Plan. Oxford University Press, pp. 279.

Groffman, P.M. and Tiedje, J.M. (1989). Denitrification in north temperate forest soils: spatial and temporal patterns at the landscape and seasonal scales. Soil Biology and Biochemistry, 21, 613-620. Hall, D.G.M., Reeve, M.J., Thomasson, A.J. and Wright, V.F. (1977). Water retention, porosity and density of field soils. Soil Survey Technical Monograph No.9., Soil Survey of England and Wales, Harpenden. Hallard, M. and Armstrong, A.C. (1992). Observations of water movement to and within mole drainage channels. Journal of Agricultural Engineering Research, 52,309-315. Hammond, T. E., Cory, D.G., Ritchey, W.M. and Morita, H. (1985). High resolution solid state 13C NMR of Canadian peats. Fuel, 64,1687-1695. Harden, J.W., Sundquist, E.T., Stallard, R.F. and Mark, R.K. (1992). Dynamics of soil carbon during deglaciation of the Laurentide ice sheet. Science, 258, 19211924. Hargreaves, G.H. and Samani, Z.A. (1982). Estimating potential evapotranspiration. Journal of Irrigation and Drainage Engineering, 108, 223-230.

Harris, G.L., Howse, K.R. and Pepper, T.J. (1993). Effects of moling and cultivation on soil-water and runoff from a drained clay soil. Agricultural Water Management, 23,161-180.

279

Harris, G.L. and Parish, T. (1992). Influence of farm management and drainage on leaching of nitrate from former floodplains in a lowland clay catchment. In: (PA

Carling

and

G.E.

Petts,

eds).

Lowland

Floodplain

Rivers:

Geomorphological Perspectives. Wiley, pp. 203-216. Hedges, J.I., Cowie, G.L., Ertel, J.R., Barbour, R.J. and Hatcher, P.G. (1985).

Degradation of carbohydrates and lignins in buried woods. Geochimica et Cosmochimica Acta, 49, 710-711.

Hekstra, G.P. (1989). Global warming and rising sea levels: the policy implications. Ecologist, 19,4-15.

Henderson, G.S., Swank, W.T., Waide, J.B. and Grier, C.C. (1978). Nutrient budgets of Appalachian and Cascade region watersheds: a comparison. Forest Science, 24, 385-397. Hocking, P.J. and Meyer, C.P. (1991). Carbon dioxide enrichment decreases critical nitrate and nitrogen concentration in wheat. Journal of Plant Nutrition, 14(6),

571-584. Hodgson, J.M. (1985). Describing and Sampling Soil Profiles. Soil Survey Field Handbook, Soil Survey Technical Monograph No.5. Soil Survey of England

and Wales, Harpenden, pp. 99. Hoffman, G.J., Ayers, R.S., Doering, E.J. and McNeal, B.L. (1981). Salinity in irrigated agriculture. In: (M. Jensen, ed.). Design and operation of farm irrigation systems. American Society of Agricultural Engineers, St Joseph,

Michigan, pp. 145-185. Houghton, R.A., Boone, R.D., Fruci, J.R., Hobbie, J.E., Melillo, J.M., Palm, CA, Peterson, B.J., Shaver, G.R. and Woodwell, G.M. (1987). The flux of carbon from terrestrial ecosystems to the atmosphere in 1980 due to changes in land use: geographic distribution of the global flux. Tel/us, 39B, 122-139. Houghton, RA and Woodwell, G.M., (1989). Global climatic change. Scientific American, 260, 36-44. Houghton, J.T., Callendar, B.A, and Vamey, S.K. {1992}. Climate Change 1992Supplementary Report to the IPCC Scientific Assessment. Cambridge University Press, Cambridge. Houghton,

J.T.

(1990).

Scientific

assessment

of climatic

change.

World

Meteorological Organisation, Geneva. Houghton, R.A. (1991). Tropical deforestation and atmospheric carbon dioxide. Climatic Change, 19, 99-118. Hunt, J.M. (1972). Distribution of carbon in the crust of the earth. American Association of Petroleum Geologists Bul/etin, 56, 2273-2277.

280

Hunt, J.M. (1977). Distribution of carbon as hydrocarbons and asphaltic compounds in sedimentary rocks. American Association of Petroleum Geologists Bulletin, 61,100-116. Hunt, H.W., Trlica, M.J., Redente, E.F., Moore, J.C., Detling, J.K., Kittel, T.G.F., Walter, D.E., Fowler, M.C., Klein, DA and Elliott, E.T. (1991). Simulation model for the effects of climate change on temperate grassland ecosystems. Ecological Modelling, 53, 205-246.

Huntley, B. (1992). How ecosystems respond to environmental change: evidence from late-Quatemary paleoecological records. In: (A. Teller, P. Mathy and J.N.A. Jeffers, eds). Responses of forest ecosystems to environmental changes. Elsevier Applied Science, London, pp. 45-54.

Idso, S.B., Kimball, BA, Anderson, M.G. and Mauney, J.R. (1987). Effects of atmospheric C02 enrichment on plant growth: the interactive role of air temperature. Agriculture, Ecosystems and Envimnment, 20, 1-10. IGBP. (1988): The International Geosphere-Biosphere Program: a study of global change. A plan for action. Report No.4: Stockholm, pp. 200.

IGBP. (1990). Intemational Geosphere-Biosphere Programme: the initial core projects. Report No. 12, Stockholm.

Innes, P. and Thomasson, A.J. (1983). Soil water relationships and their effects on cereal yields. In: The Yield of Cereals. International Seminar. Royal Agricultural Society of England. Monograph Series no.1. pp. 49-57. Institute of Hydrology. (1992). Implications of climate change for the National Rivers Authority. R&D Project 358/2/T. Institute of Hydrology, Wallingford, pp. 152. IPCC. (1992). Climatic Change. The Supplementary Report to the IPCC Scientific Assessment. (J.T. Houghton, B.A. Callander, and S.K. Varney, eds).

Cambridge University Press. ISO/CD 10381-1.2 (1992). Soil quality - Sampling Part 1: Guidance on the design of

sampling programmes. International Standards Organisation, Geneva, pp. 34.

Jacquart, C. and Choisnel, E. (1985). Les besoins en eau du mais. In: Les besoins en eau des cultures. Colloque INRA, Paris. pp. 451-470.

Jansen, A.E. (1991). Importance of ectomycorrhiza for forest ecosystems. In: (A. Teller, P. Mathy, J.N.A. Jeffers, eds). Responses of Forest Ecosystems to Environmental Changes. Elsevier Applied Science, London. pp. 456-461.

Jansson, P-E. and Berg, B. (1985). Temporal variation of litter decomposition in relation to simulated soil climate. Long-term decomposition in a Scots pine forest. V. Canadian Journal of Botany, 63, 1008-1016.

281

Janssen, B.H. (1984). A simple method for calculating decomposition and accumulation of 'young' soil organic matter. Soil and Plant, 76, 297-304. Jarvis, N.J. and Leeds-Harrison, P.B. (1987). Modelling water movement in drained clay soil. Journal of Soil Science, 38, 499-509. Jelgersma S. (1988). A future sea-level rise: its impacts on coastal lowlands. In: Geology and urban development: atlas of urban geology. UN-ESCAP. I, 61-81.

Jenkinson, D.S. and Rayner, J.H. (1977). The turnover of soil organic matter in some of the Rothamsted Classical Experiments. Soil Science, 123, 298-305. Jenkinson, D.S., Hart, P.B.S., Rayner, J.H. and Parry, L.C. (1987). Modelling the turnover of organic matter in long-term experiments at Rothamsted. INTECOL Bulletin, 15, 1-8.

Jenkinson, D.S., Adams, D.E. and Wild, A. (1991). Model estimates of C02 emissions from soil in response to global warming. Nature, 351, 304-306. Jenkinson, D.S., Harkness, D.o., Vance, E.D., Adams, D.E. and Harrison, A.F. (1992). Calculating net primary production and annual input of organic matter from the amount and radiocarbon content of soil organic matter. Soil Biology and Biochemistry, 24, 295-308.

Jenkinson, D.S., Potts, J.M., Perry, J.N., Barnett, V., Coleman, K. and Johnston, A.E. (1994). Trends in herbage yields over the last century on the Rothamsted long-term continuous hay experiment. Journal of Agricultural Science, Cambridge, 122,365-374.

Jenny, H., (1980). The soil resource: origin and behaviour. Ecological Studies, 37. Springer-Verlag, New York. Jenny, H. (1941). Factors of soil formation. John Wiley and Sons, New York. Johnson, OW., Cole, OW., Gessel, S.P., Singer, M.J. and Minden, R.V. (1977). Carbonic acid leaching in a tropical, temperate, subalpine, and northern forest soil. Arctic and Alpine Research, 9, 329-343. Jonasson, S., Havstrom, M., Jensen, M. and Callaghan, T.V. (1993). In situ mineralization of nitrogen and phosphorus of arctic soils after perturbations simulating climate change. Oecologia, 95, 179-186. Jones, CA and Kiniry, J.R. (1986). CERES-MAIZE, a simulation model of maize growth and development. Texas A & M University Press, College Station, TX, pp.194. Jones, CA (1990). Use of crop simulation models in dryland agriculture. Advances in Soil Science, 13, 333-345.

282

Jones, P.O., Wigley, T.M.L., Folland, C.K., Parker, D.E., Angell, J.K., Lebedeff, S. and Hansen, J.E. (1988). Evidence for global warming in the past decade. Nature, 332, 790. Jones,

P.O.,

Marsh,

R.

and

Farmer,

G.

(1990).

Calculation

of potential

evapotranspiration (PET) over the European Community Countries. Climatic Research Unit Report to the European Commission. Contract No. 3578-88-12 ED ISP GB. pp. 29. Jones, R.J.A. and Thomasson, A.J. (1985). An agroclimatic databank for England and Wales. Soil Survey Technical Monograph No. 16. Soil Survey of England and Wales, Harpenden. Jones, R.J.A. and Thomasson, AJ. (1987). Land suitability classification for temperate arable crops. In: (K.J. Beek, P.A. Burrough, and D.E. McCormack, eds) Quantified Land Evaluation Procedures. ITC Publication, NO.6. pp.29-35. de Jong, R. and Kabat, P. (1990). Modelling water balance and grass production.

Soil Science Society of America Journal, 54, 1725-1732. Jorgensen, S.E. and Mejer, H. (1976). Modelling the global cycle of carbon, nitrogen and phosphorus and their influence on the global heat Balance. Ecological Modelling, 2, 19-31. Kasting, J.F. (1984). Comments on the BLAG model: the carbonate - silicate geochemical cycle and its effect on atmospheric carbon dioxide over the past 100 million years. American Journal of Science, 284, 1175-1182. Katz, R.W. and Brown, B.G. (1.992). Extreme events in a changing climate: variability is more important than averages. Climatic Change, 21, 289-302. Keeling, C.D. and Whorf, T.P. (1990). Atmospheric CO2 Concentrations - Mauna Loa Observatory, Hayvaii, 1958-1990. Carbon Dioxide Information Analysis Center (CDIAC), NDP-001, Oak Ridge National Laboratory, Institute for Energy Analysis, Oak Ridge, TN. Keller, J., Corey, F., Walker, W.R. and Varva, M.E. (1980). Evaluation of irrigation systems. Irrigation challenges for the 1980s. Proceedings of the 2nd National

Irrigation Symposium. American Society of Agricultural Engineers, pp. 95-105. Keller, M., Veldkamp, E., Weitz, A.M. and Reiners, W.A. (1993). Effects of pasture age on soil trace gas emissions from a deforested area of Costa Rica. Nature, 365, 244-246. Kenny, G.J. and Harrison, P.A. (1992). Thermal and moisture limits of grain maize in Europe: model testing and sensitivity to climatic change. Climate Research, 2, 113-129.

283

van Keulen, H. (1982). Crop production under semi-arid conditions as determined by moisture availability. In: (FW.T. Penning de Vries and H.H. van Laar, eds). Simulation of plant growth and crop production. Pudoc, Wageningen. pp. 159-

174. Kim, J-W., Chang, J-T., Baker, N.L., Wilks, D.S. and Gates, W.L. (1984). The statistical problem of climate inversion: determination of the relationship between local and large-scale climate. Monthly Weather Review, 112, 2069-2077. Kim J., Verma, S.B. and Clement, R.J. (1992). Carbon dioxide budget in a temperate grassland ecosystem. Journal of Geophysics Research, 97, 6057-6063. Kimball, BA (1983). Carbon dioxide and agricultural yield. An assemblage and analysis of 770 prior observations. WCL Report 14. Water Conservation

Laboratory. Agricultural Research Service. U.S. Department of Agriculture. Phoenix. AZ. Kimball, BA and Idso, S.B. (1983). Increasing atmospheric C02: effects on crop yield, water use, and climate. Agricultural Water Management, 7, 55-72. Kimball, BA (1985). Adaptation of vegetation and management practices to a higher carbon dioxide world. In: (B.R. Strain and J.D. Cure, eds). Direct effects of increasing carbon dioxide on vegetation. DOE/ER-0238 Carbon Dioxide

Research Division, U.S. Department of Energy, Washington, DC, pp. 185-204. Kleber, A. (1992). Periglacial slope deposits and their pedogenic implications in Germany. Paleogeography, Paleoclimatology, Paleoecology 99, 361-371. Klemendtson, I., Svensson, B.H. and Rosswall, T. (1988). Relationships between soil moisture content and nitrous oxide production during nitrification and denitrification. Biology and Fertility of Soils, 6, 106-111. Kling, L. (1993). Programmation, calibration et validation d'un modele de bilan hydrique. DAA IENSA/ INRA Montpellier, France. pp. 40. Korner, C., Scheel, J.A., Bauer, H. (1979). Maximum leaf diffusive conductance in vascular plants. Photosynthetica, 13(1), 45-82. Korner, C. and Arnone, JA III (1992). Responses to elevated carbon dioxide in artificial tropical ecosystems. Science, 257, 1672-1675. Kosmas, C.S. (1993). Field sites: Spata-Athens, Greece. In: Mediterranean Desertification and Land Use - MEDALUS I, Final Report. Commission of the European Communities, DGXII (Environment), Brussels, EPOCH-CT90-0014(SMA), pp. 581-607.

Kosmas, C.S., Danalatos, N.G., Moustakas, N., Tsatiris, 8., Kallianou, Ch. and Yassoglou, N. (1993a). The impacts of parent material and landscape position

284

on drought and biomass production of wheat under semi-arid conditions. Soil

Technology, 6, (in press). Kosmas, C.S., Moustakas, N., Danalatos, N.G. and Yassoglou, N. (1993b). The effect of rock fragments on wheat biomass production under highly variable moisture conditions in Mediterranean environments. Catena, Supplement 24, (in press). Kosmas, C.S., Moustakas, N., Danalatos, N.G. and Yassoglou, N. (1993c). The effect of diminishing soil moisture on soil properties and biomass production of rainfed wheat. In: Mediterranean Desertification and Land Use - MEDALUS I,

Final

Report.

Commission

of

the

European

Communities,

DGXII

(Environment), Brussels, EPOCH-CT90-0014-(SMA). Kovda, VA and Szabolcs, I. (1979). Modelling of soil salinization and alkalization.

Agrokemia

es Talajtan, 28, 1-208. (Supplementum).

Kucera, C.L. and Kirklam, D.R. (1971). Soil respiration studies in tallgrass prairie in Missouri. Ecology, 52, 912-915. Larson, W.E. and Foster, G.A. (1988). Erosion and sediment deposition. In: (H.R. Finney, ed.). Proceedings of the International Interactive Workshop on Soil

Resources: Their Inventory, Analysis and Interpretation for Use in the 1990's. 22-24 March 1988, Minneapolis, Minnesota Extension Service, St. Paul, MN. pp. 222-230. Lawlor, D.W. and Mitchell, RAC. (1991). The effects of increasing CO2 on crop photosynthesis and productivity: a review of field studies. Plant, Cell and

Environment, 14, 807-818. Lee, J.J. and Lammers. D.A. (1990). An approach to the regional evaluation of the responses of soils to global climate change. In: (A.F. Bouwman, ed.). Soils and

the Greenhouse Effect. J. Wiley & Sons, England. pp. 395-399. Lee, J.J., Phillips, D.L. and Liu, R. (1993). The effects of trends in tillage practices on erosion and carbon contents of soils in the US corn belt. Water, Air, and Soil

Pollution, 70, 389-401. Leemans, A. and Cramer, W.P. (1991). The IIASA database for mean monthly values of temperature, precipitation and cloudiness on a global terrestrial grid. IIASA Research Report RR-91-18, Laxenburg, Austria. ISBN: 3-7045-0113-1. pp.62. Leenhardt, D. (1991). Spatialisation du bilan hydrique. Propagation des erreurs d'estimation des caracteristiques du sol au travers des modeles de bilan hydrique. Cas du ble d'hiver. These. ENSAIINRA, Montpellier, France. pp. 129.

285

Leenhardt, D., Voltz, M. and Rambal, S. (Submitted). A critical view of soil water balance models with reference to their spatial applications. Geoderma. Legros, J-P., Baldy, C., Fromin, N. and Bellivier, D. (1994). Crop models: principles and adaptations to the problem of climate change (this volume). Lewis, RA, Stein, N. and Lewis, C.w. (1984). Environmental specimen banking and monitoring as related to banking. Proceedings of the International Workshop, SaarbrOcken, 10-15 May. Martinus Nijhof Publishers, The Hague. pp. 358. Lewis, RA (1987). Guidelines for environmental specimen banking with special reference to the Federal Republic of Germany: ecological and managerial aspects. U.S. Department of the Interior, National Park Service. U.S. MAB Report No., 12. pp. 182. Lewis, R.A. and Klein, B. (1990). A brief history of specimen banking: storage, institutions and applications. Toxicological and Environmental Chemistry, 27, 251-266. Lieth, H. (1975). Modelling the primary productivity of the world. In: (H. Lieth and P.H. Whittaker, eds). Primary productivity of the biosphere. Springer-Verlag, New York, pp. 237-263. Linacre, E.T. (1963). Determining evapotranspiration rates. Journal of the Australian Institute of Agricultural Science, 29, 165-181. Littleboy, M., Silburn, D.M., Freebairn, D.M., Woodruff, D.R. and Hammer, G.L. (1989). A computer simulation model of productivity/erosion runoff functions to evaluate conservation techniques.

Queensland

Department of Primary

Industries, Brisbane. pp. 119. Logan, J. (1983). Nitrogen oxides in the troposphere: global and regional budgets. Journal of Geophysics Research, 93, 1593-1599. Lough, J.M., Wisley, T.M.L. and Palutikof, J.P. (1983). Climate impact scenarios for Europe in a warmer world. Journal of Climatology and Applied Meteorology, 22, 1673-1684. Loveland, P.J., Rounsevell, M.D.A., Mayr, T.M., Legros, J.P., Voltz, M., de la Rosa, D. and Armstrong, A. (In press). ACCESS: agroclimatic change and European soil suitability - a spatially distributed model. In: (I. Troen ed.). Climate Change. Proceedings of a European Community Workshop, September 1993, Copenhagen.

Commission

of

the

European

Communities,

DGXII

(Environment), Brussels. Luepke, N.P. (ed.) (1979). Monitoring environmental materials and specimen banking. Martinus Nijhof Publishers, The Hague. pp. 591.

286

MAFF. (1976). The Agricultural Climate of England and Wales. Technical Bulletin No. 35. HMSO, London, pp. 147. MAFF. (1989). Environmentally Sensitive Areas (first report). HMSO, London, pp. 50. Maheras,

P.

(1988).

Changes

in

precipitation

conditions

in

the

western

Mediterranean over the last century. Joumal of Climatology, 8, 179-189. Manabe, S. and Holloway, J.R. (1975). The seasonal variation of the hydrologic cycle as simulated by a global model of the atmosphere. Journal of Geophysics Research, 80,1617-1649.

Manabe, S. and Wetherald, R.T. (1986). Reduction in summer soil wetness induced by an increase in atmospheric carbon dioxide. Science, 232, 626-628. Marshman, N.A. and Marshall, K.C. (1981). Bacterial growth on proteins in the presence of clay minerals. Biology and Biochemistry, 13,127-134. Mass, S.J. and Arkin, G.F. (1978). User's guide to SORGF: A dynamic grain Sorghum growth model with feed back capacity. Blackland Research Centre,

Temple, TX. pp. 100. Massman, W.J. (1992). A surface energy balance method for partitioning evapotranspiration data into plant and soil components for a surface with partial canopy cover. Water Resources Research, 28(6), 1723-1732. McCree, K.J. and Fernandez, C.J. (1989). Simulation model for studying physiological water stress responses of whole plants. Crop Science, 29, 353360. McKenny, M.S. and Rosenberg, N.J. (1993). Sensitivity of some potential evapotranspiration estimation methods to climate change. Agricultural and Forest Meteorology, 64, 81-110. McLaren, A.D. and Barshad, I. (1976). Clays, protein and nonsense. Soil Science, 121,188. Meams, L., Rosenzweig, C. and Goldberg, R. (1992). Effects of changes in interannual climatic variability on CERES-WHEAT yields. Agricultural and Forest Meteorology, 62,159-189.

Melillo, J.M., McGuire, A.D., Kicklighter, D.w., Moore, B. III, Vorosmarty, C.J. and Schloss, A.L. (1993). Global climate change and terrestrial net primary production. Nature, 363, 234-239. Middelburg, J.J. (1989). A simple rate Model for organic matter decomposition in marine sediments. Geochimica et Cosmochimica Acta, 53, 1577-1581. Miles, J. (1985). The pedogenic effects of different species and vegetation types and the implications of succession. Joumal of Soil Science, 36, 571-584.

287

Mitchell, J.F.B., Manabe, S., Meleshko, V. and Tokioka, T. (1990). Equilibrium climate change and its implications for the future. In: (J.T. Houghton, G.J. Jenkins,

J.J.

Ephraums,

eds).

Climate

change.

The

IPCC scientific

assessment. WMO & UNEP, Cambridge University Press, Cambridge, pp 131-

172. Mitchell, RAC., Mitchell, V.J., Driscoll, S.P., Franklin, J. & Lawlor, D.w. (1993). Effects of increased CO 2 concentration and temperature on growth and yield of winter wheat at two levels of nitrogen application. Plant, Cell and Environment, 16, 521-529.

Moffat, A.J. and Boswell, R.C. (1990). Effect of tree species and species mixtures on soil properties at Gisburn Forest, Yorkshire. Soil Use and Management, 6, 4651. Mogensen, V.O. (1977). Field measurements of dark respiration rates of roots and aerial parts in Italian ryegrass and barley. Journal of Applied Ecology, 14,243252. Mollet, J.M. (1991). Aptitude

a la

diversification en grandes cultures. DAAI ENSA

/INRA Montpellier, France. pp. 73 (and annexes). Monteith, J.L. (1977). Climate and the efficiency of crop production in Britain. Philosophical Transactions of the Royal Society London, B, 281, 277-294.

Monteith, J.L. (1981). Climatic variation and the growth of crops. Quarterly Journal of the Royal Meteorological Society, 107, 749-774.

Mooney, HA, Drake, B.G., Luxmoore, R.J., Oechel, W.C. and Pitelka, L.F. (1991). Predicting ecosystem responses to elevated C02 concentrations. Bioscience, 41,96-104. Mooney, HA and Koch, G.W. (1993). The impact of rising CO 2 concentrations on the terrestrial biosphere. Global Change Newsletter, 13, 19. Moorby, J. (1985). Wheat growth and modelling: conclusions. In: (W. Day and R.K. Atkin, eds). Wheat growth and modelling. NATO ASI Series A, Vol. 86,pp. 381-385. Plenum, New York. Moore, P.O. and Bellamy, D.J. (1974). Peatlands. Elek Science, London. pp. 221. Moore, T., Roulet, N. and Knowles, R. (1990). Spatial and temporal variations of methane flux from subarctic/northern boreal fens. Global Biogeochemical Cycles, 4, 29-46.

Moore, T.R. and Knowles, R. (1989). The influence of water table levels on methane and carbon dioxide emissions from peatland soils. Canadian Journal of Soil Science, 69, 33-38.

288

Morgan, R.P.C. (1979). Soil erosion: topics in applied geography. Technical Report. Silsoe, UK. pp. 113. Mueller, E.L. and Kramer, J.R. (1994). Computer simulations of terrestrial carbon and atmospheric interactions. Environmental Pollution, 83, 113-120. Nix, HA and Fitzpatrick, E.A. (1969). An index of crop water stress related to wheat and grain sorghum yields. Agricultural Meteorology, 6, 321-337. Norby, R.J., Gunderson, CA, Wullschleger, S.D., O'Neill, E.G. and McCracken, M.K. (1992). Productivity and compensatory responses of yellow-popular trees in elevated CO 2 . Nature, 357, 322-324. Norton, J.M., Smith, J.L., Firestone, MK (1990). Carbon flow in the rhizosphere of Ponderosa Pine seedlings. Soil Biology and Biochemistry, 22, 449-455. O'Leary, G.J., Connor, D.J. and White D.H. (1985). A simulation model of the development, growth and yield of the wheat crop. Agricultural Systems, 17, 126. Oades, J.M. (1988). The retention of organic matter in soils. Biogeochemistry,S, 3570. Oberbauer, S.F., Gillespie, C.T., Cheng, W., Gebauer, R., Sala Serra, A. and Tenhunen, J.D. (1992). Environmental effects on CO 2 efflux from riparian tundra in the northern foothills of the Brooks Range, Alaska, USA. Oecologia, 92, 568-577. Oechel, W.C., Hastings, S.J., Vourlitis, G., Jenkins, M., Riechers, G. and Grulke, N. (1993). Recent change of Arctic tundra ecosystems from a net carbon dioxide sink to a source. Nature, 361, 520-523. Orchard, VA and Cook, F.J. (1983). Relationship between soil respiration and soil moisture. Soil Biology and Biochemistry, 15, 447-453. Oster, J.D., Meyer, J.L., Hermsmeier, l. and Kaddah, M. (1986). Field studies of irrigation efficiency in the Imperial Valley. Hilgardia, 54(7),1-15. Palmieri, S., Siani, A.M. and Agostino, A. (1991). Climate fluctuations and trends in Italy within the last 100 years. Annales Geophysicae, 9, 769-776. Palutikof, J.P., Guo, X., Wigley, T.M.L. and Gregory, J.M. (1992). Regional changes in climate in the Mediterranean basin due to global greenhouse gas warming. UNEP Mediterranean Action Plan Technical Reports Series No 66, UNEP.

Athens, pp. 172. Parry, M.L. and Carter, T.R. (1989). An assessment of the effects of climatic change on agriculture. Climatic Change, 15, 95-116.

289

Parry, M.L., Carter, T.R. and Konijn, N.T. (eds). (1988). The impact of climatic variations on Agriculture, Volume 1: assessments in cool temperate and cold regions. Kluwer Academic, Dordrecht. Parry, M. (1992). The potential effect of climate changes on agriculture and land use. In: (F. I. Woodward, ed.). The ecological consequences of global climate change. Advances in Ecological Research, 22, pp. 63-91. Parton, W.J., Morgan, JA, Altenhofen, J.M. and Harper, L.A. (1988). Ammonia volatilization from spring wheat plants. Agronomy Journal, 50, 419-425. Parton, W.J., Schimel, D.S., Cole, C.V. and Ojima, D.S. (1987). Analysis of factors controlling soil organic matter levels in Great Plains grasslands. Soil Science Society America Journal, 51, 1173-1179. Parton, W.J., Stewart, J.W.B. and Cole, C.V. (1988). Dynamics of C, N, P and S in grassland soils: a model. Biogeochemistry, 5, 109-131. Paul, E.A. (1984). Dynamics of organic matter in soils. Plant and Soil, 76, 275-285. Paustian, K., Parton, W.J. and Persson, J. (1992). Organic amendments and Nfertilization in long-term plots: model analyses of soil organic matter dynamics. Soil Science Society America Journal, 56, 476-488. Perrier, A. (1977). Projet de definitions concernant I'evapotranspiration en fonction de considerations theoriques et pratiques. Meteorologie, 11, 7-16. Peterson, D.F. and Keller, AA (1990). Irrigation. In: (P.E. Waggoner, ed.). Climate Change and U.S. Water Resources. John Wiley & Sons, New York. Phillips, D.L., Hardin, P.D., Benson, V.w. and Baglio, J.V. (1993). Non-point source pollution impacts of alternative agricultural management practices in Illinois: a simulation study. Journal of Soil and Water Conservation, 48, 449-457. Philp, R.P. and Calvin, M. (1976). Possible origin for the insoluble organic (kerogen) debris in sediments from insoluble cell-wall materials of algae and bacteria. Nature, 262,134-136. Pigram, J.J., Shaw, K.L. and Coelli, M.L. (1992). Climate change and irrigated agriculture. Implications for the Murray-Darling basin. Centre for Water Policy Research, University of New England, Armidale, Australia. Pi razzoli , P.A. (1989). Present and near future global sea-level changes. Global and Planetary Change. 75(4}, 241-258. Porter, J.R. (1985). Approaches to modelling canopy development in wheat. In: (W. Day and R.K. Atkin, eds). Wheat growth and modelling. NATO ASI Series A, Vol. 86, pp. 69-81. Plenum, New York.

290

Powlson, D.S. and Jenkinson, D.S. (1976). The effects of biocidal treatments on metabolism in soil II. Gamma irradiation, autoclaving, air-drying and fumigation. Soil Biology and Biochemistry, 8, 9-188. Powlson, D.S., Pruden, G., Johnson, A.E. and Jenkinson, D.S. (1986). The nitrogen cycle in the Broadbalk Wheat Experiment: recovery and losses of 15N-labelled fertilizer applied in spring and inputs of nitrogen from the atmosphere. Journal

of Agricultural Science, Cambridge, 107,591-609. Powlson, D.S., Brooks, P.C. and Christensen, B.T. (1987). Measurement of soil microbial biomass provides an early indication of changes in total soil organic matter due to straw incorporation. Soil Biology and Biochemistry, 19(2), 159-

164. Powlson, D.S., Hart, P.B.S., Poulton, P.R., Johnston, A.E. and Jenkinson, D.S.

(1992). Influence of soil type, crop management and weather on the recovery of 15N-labelled fertilizer applied to winter wheat in spring. Journal of Agricultural Science, Cambridge, 118,83-100. Preston, C.M. and Ripmeester, J.A. (1982). Application of. solution and solid state 13C NMR to organic soils, their humic acids, fulvic acids, humins and hydrolysis residues. Canadian Journal of Spectroscopy, 27, 99-105. Preston, C.M., Ripmeester, J.A., Mathur, S.P. and Levesque, M. (1986). Application of solution and solid-state multinuclear NMR to a peat-based composting system for fish and crab scrap. Canadian Journal of Spectroscopy, 31, 63-69. Preston, C.M., Shipital, S.E., Dudley, R.L., Fyfe, C.A., Mathur, S.P. and Levesque, M. (1987). Comparison of l3C CPMAS NMR and chemical techniques for measuring the degree of decomposition in virgin and cultivated peat profiles. Canadian Joumal of Soil Science, 67, 187-198.. Quinones, H. and Cabelguenne, M. (1990). Use of EPIC to study cropping systems. II. Improved simulation of the water use, growth and harvest index in corn. Agriculture Mediterraneenne, 120, 241-248. Quinones, H., Texier, V., Cabelguenne, M. and Blanchet, R. (1990). Simulation du comportement hydrique du toumesol et de ses repercutions sur la croissance et la production. Les Points Scientifique du CETIOM. Le tournesol et I'eau. pp. 56-71. Rachhpal-Singh and Nye, P.H. (1986). A model of ammonia volatilization from applied urea. III. Sensitivity analysis, mechanisms and applications. Journal of Soil Science, 37, 31-40. Raich, J.W. and Schlesinger, W.H. (1992). The global carbon dioxide flux in soil respiration and its relationship to climate. Tel/us, 448, 81-99.

291

Rambal, S. and Cornet, A. (1982). Simulation de I'utilisation de I'eau et de la production

vegetale d'une phytocenose sahelier'lne du

Senegal.

Acta

Oecologia (Planta), 3(17), 381-397.

Rastetter, E.B., McKane, R.B., Shaver, G.R. and Melillo, J.M. (1992). Changes in C storage by a terrestrial ecosystem: how C-N interactions restrict responses to CO 2 and temperature. Water, Air, and Soil Pollution, 64, 327-344. Reeve, M.J., Hall, D.G.M. and Bullock, P. (1980). The effect of soil composition and environmental factors on the shrinkage of some clayey British soils. Journal of Soil Science, 31, 429-442.

Rhoades, J.D. and Loveday, J. (1990). Salinity in irrigated agriculture. In: (B.A. Stewart and D.R. Nielsen, eds). Irrigation of agricultural crops. American Society of Agronomy. Madison, WI, pp. 1089 -1142.

Rind, D., Goldberg, R. and Ruedy, R. (1989). Change in climatic variability in the 21st century. Climatic Change, 14,5-37. Ritchie, J.T. (1991). Specifications of the ideal model for predicting crop yields. In: Climatic Risk in Crop Production. CAB International, Wallingford, pp. 97-122.

Ritchie, J.T. and Otter, S. (1984). Description and performance of CERES-WHEAT: a user oriented wheat yield model. USDA, Temple. pp. 27. Robelin, M. (1984). Fonctionnement hydrique et adaptation

a la

secheresse. In:

Physiologie du MaTs. INRA - Paris. pp. 445-476.

Roberson, E.B., Sarig, S. and Firestone, M.K. (1991). Crop cover management of polysaccharide-mediated aggregation in an orchard soil. Soil Science Society of America Journal, 55, 734-739.

Rodhe, H. (1990). A comparison of the contribution of various gases to the greenhouse effect. Science, 248, 1217-1219. Ronov, A.B. (1983). The Earth's Sedimentary Shell. American Geological Institute Reprint Series: V. Plenum Press, N.Y. de la Rosa, D., Cardona, F. and Paneque, G. (1977). Soil evaluation for several agricultural uses: a system developed for Mediterranean regions. Anales Edafologia y Agrobiologia, 36, 1100-1112.

de la Rosa, D. (Coordinator) (1984). Catalogo de Suelos de Andalucia. AMA-CSIC, Junta de Andalucia, Sevilla. de la Rosa, D., Mudarra, J.L., Romero, R. and Martin-Aranda, J. (1984). Characterization and evaluation of agricultural benchmark soils from Sevilla, Spain. Soil Science Society America Joumal, 48, 358-362. de la Rosa, D., Cardona, F. and Almorza, J. (1981). Crop yield predictions based on properties of soils in Sevilla, Spain. Geoderma, 25, 267-274.

292

de la Rosa, D. and Moreira, J.M. (Coordinators) (1987). Evaluacion Ecologica de Recursos Naturales de Andalucia. AMA, Junta de Andalucia, Sevilla.

de la Rosa, D., Moreno, J.A., Garcia, l.V. and Almorza, J. (1992). MicroLEIS: a microcomputer-based Mediterranean Land Evaluation Information System. Soil Use and Management, 8, 89-96.

de la Rosa, D., Moreno, J.A. and Garcia, l.V. (1993). Expert evaluation system for assessing field vulnerability to agrochemical compounds in the Mediterranean region. Joumal of Agricultural Engineering Research, 56, 153-164. Rosenberg, N.J., McKenney, M.S. and Martin, Ph. (1989). Evapotranspiration in a greenhouse warmed world: a review and a simulation. Agricultural and Forest Meteorology, 47, 303-320. Rosenberg, N.J., Kimball, B.A., Martin, P. and Cooper, C.F. (1990). From climate and C02 enrichment to evapotranspiration. In: (P.E. Waggoner, ed.). Climate change and U.S. Water Resources. John Wiley & Sons, New York, pp. 151-

175. Rosenthal, W.D., Vanderlip, RL, Jackson, B.S. and Arkin, G.F. (1989). SORKAM: a grain sorghum crop growth model. Texas Agricultural Experiment Station, College Station, TX. pp. 205. Rossbach, M., Schladot, J.D. and Ostapczuk, P. (eds) (1992). Specimen Banking. Springer-Verlag. pp. 242. Rossiter, D. (1990). ALES: A framework for land evaluation by using a microcomputer. Soil Use and Management, 6, 7-20. Rotty, A.M. and Marland, G. (1984). Carbon dioxide production from fossil fuels and cement, 1960-1982. Carbon Dioxide Information Analysis Center (CDIAC), NDP-006, Oak Ridge National Laboratories, Institute for Energy Analysis, Oak Ridge, TN. Rounsevell, M.D.A. (1993). A review of soil workability models and their limitations in temperate regions. Soil Use and Management, 9,15-21. Rounsevell, M.D.A. and Brignall, A.P. (1993). The climatic sensitivity of land suitability for strategic crops in Europe. In: (G. Liangzhi, W. Lianhai, Z. Dawei and H. Xiangling, eds). Proceedings of the Intemational Symposium on Climate Change, Natural Disasters and Agricultural Strategies. Chinese

Meteorological Press, Beijing, pp. 14-18. Rounsevell, M.D.A. and Jones, A.J.A. (1993). A soil and agroclimatic model for estimating machinery work-days: the basic model and climatic sensitivity. Soil and Til/age Research, 26,179-191.

293

Rounsevell, M.D.A. and Loveland, P.J. (1992). An overview of hydrologicallycontrolled soil responses to climate change in temperate regions. SEESOIL, 8, 69-78. Roy, M.G. and Hough, M.N. (1978). Effects of weather and climate change on modem British farming. Journal of the Royal Agricultural Society of England, 139,43-53. Ruellan, A. and Dosso, M. (1993). Regards sur Ie sol. Foucher, Paris. Ruget, F., Bonhomme, R. and Varlet-Grancher, C. (1991). Analyse de la fonction photosynthese dans CERES-MAIZE. Colloque INRA. Physiologie et production du mars. pp. 445-453.

Ruhe, R.V. (1959). Stone lines in soils. Soil Science, 87, 223-231. Ruhe, R.V. (1975). Geomorphology. Houghton Mifflin Co., Boston. Ruhe, R.V., Daniels, R.B. and Cady, J.G. (1967). Landscape evolution and soil formation in southwestem Iowa. USDA Technical Bulletin 1349. Russel, G. (1993). Absorbed radiation and crop growth. Colloque Science Update. (INRA Paris, ed.). pp. 459-470. Sadras, V.O. and Hall, A.J. (1989). Patterns of water availabfity for sunflower crops in semi-arid central Argentina: a simulation based evaluation of their interactions with cropping strategies and cultivar traits. Agricultural Systems, 31, 221-238. Sanchez, P.A., Palm, C.A. and Smyth, T.J. (1990) Approaches to mitigate tropical deforestation

by

sustainable

soil

management

practices.

In:

(H.W.

Scharpenseel, M. Schomaker and A. Ayoub, eds). Soils on a warmer earth. Developments in Soil Science, 20, pp. 217-220.

Scharpenseel, HW., Schomaker, M. and Ayoub, A. (eds) (1990). Soils on a warmer earth. Developments in Soil Science, 20, Elsevier, Amsterdam. Scharpenseel, HW. (1990). Global soil changes and their dynamics in a changing environment. Transactions of the 14th ISSS Congress, Kyoto, Vol. V, 255-288. Schjorring, J.K., Nielsen, N.E., Jensen, H.E. and Gottschau, A. (1989). Nitrogen losses from field-grown spring barley plants as affected by rate of nitrogen application. Plant and Soil, 116,167-175. Schladot, J.D. and Backhaus, F. (1988). Preparation of sample material for environmental specimen banking purposes - milling and homogenization at cryogenic temperatures. In: (S.A. Wise, R. Zeisler and G.M. Goldstein, eds). Progress in Environmental Specimen Banking, National Bureau of Standards,

Washington, D.C., NBS Special Publication, 740,184-193. Schlesinger, W. H. (1985). Changes in soil carbon storage and associated properties with disturbance and recovery. In: (J.R. Traflka, and D.E. Reichle, eds). The

294

changing carbon cycle: a global analysis. Springer-Verlag, New York. pp. 194220. Schlesinger, W.H., (1977). Carbon balance in terrrestrial detritus. Annual Review of Ecological Systems, 8, 51-81. Schlesinger, W.H. (1991). Biogeochemistry: an analysis of global change. Academic Press, New York. pp. 443. Schneider, S.H. (1992). The climatic response to greenhouse gases. In: (F.I. Woodward, ed.). The ecological consequences of global climate change. Advances in Ecological Research, 22, 1-32. Scully, R.W. and Arnold, R.W. (1981). Holocene alluvial stratigraphy in the Upper Susquehanna River basin, New York. Quatemary Research, 15,327-344. Semenov, MA, Porter, J.R. and Delecolle, R. (1993). Simulation of the effects of climatic change on growth and development of wheat in the UK and France. In: (G.J. Kenny, PA Harrison and M.L. Parry, eds). The effect of climate change on agricultural and horticultural potential in Europe. Environmental Change Unit, Oxford University,Research Report No.2, pp.121-136. Shaffer, M.J., Halvorson, A.D. and Pierce, F.J. (1991). Nitrate leaching and economic analysis package (NLEAP): model description and application. In: (R.F. Follett,

D.R. Keeney and R.M. Cruse, eds). Managing nitrogen for

groundwater quality and farm profitability. Soil Science Society of America, Madison, WI, pp 285-322. Sharkov, LN. (1987). An improved absorption method for measuring C02 evolution from soil under field conditions. Soviet Soil Science (Pochvovedeniye, English Translation), 1, 127-133. Sharpley, AN. and Williams, J.R. (1990). EPIC - Erosion Productivity Impact Calculator: 1. Model Documentation. USDA Technical Bulletin No. 1768, Temple, Texas, pp. 235. Shaver, G.R., Billings, W.D., Chapin, F.S. 111., Giblin, AE., Nadelhoffer, K.J., Oechel, W.C. and Rastetter, E.B. (1992). Global change and the carbon balance of Arctic ecosystems. Bioscience, 42, 433-441. Simota, C., Canarache, A, Colibas, I. and Colibas, M. (1986). A simulation model for field water use and crop yields as applied to soil ripping and compaction studies. Transactions of the 13th ISSS Congress, Hamburg, 2, 391-396. Simota, C., Mayr, T. and Loveland, P. (1993). A spreadsheet oriented model for simulation of crop yield formation using monthly weather data and soil survey information. Soil Survey and Land Research Centre Internal Paper. (unpublished)

295

Sinclair, T.R. (1992). Mineral nutrition and plant growth response to climate change. Journal of Experimental Botany, 43, 1141-1146.

Singh, J.S. and Grupta, S.R. (1977). Plant decomposition and soil respiration in terrestrial ecosystems. Botanical Reviews, 43, 449-528. Sinoquet, H. and Andrieu, B. (1993). The geometrical structure of plant canopies: characterization and direct measurement methods. In: (INRA, Paris, ed.). Structure and light microclimate. Colloque Science update, pp. 131-158.

Skejemstad, J.O., Frost, R.L. and Barron, P.F. (1983). Structural units in humic acids from

south-eastern

Queensland

soils

as

determined

by

13C

NMR

spectroscopy. Australian Journal of Soil Research, 21,539-547. Smith, J.L. and Paul, E.A. (1990). The significance of soil microbial biomass estimations. In: (J. Bollag and G. Stotzky, eds). Soil Biochemistry 6, 357-395. Smith, K. (1964). A long-period assessment of the Penman and Thornthwaite potential evapotranspiration formulae. Journal of Hydrology, 2, 277-290. Smith, M.S. and Tiedje, J.M. (1979). The effects of roots on soil denitrification. Soil Science Society of America Journal, 43, 951-955.

Soeiro de Brito, R., Roxo, M.J. and Casimiro, P.C. (1993). Field investigations: Lower Alentejo, Beja and Mertola, Portugal. In: Mediterranean Desertification and Land Use - MEDALUS I, Final Report. Commission of the European Communities, DGXII (Environment), Brussels, EPOCH-CT90-0014-(SMA), pp.

406-432. Soil Conservation Service. (1987). Basic statistics 1982, National Resources Inventory. USDA, Soil Conservation Service and Iowa State University

Statistical Laboratory. Statistical Bulletin Number 756. Soil Conservation Service, Washington, DC. Soil Survey Staff. (1975). Soil Taxomony - a basic system of soil classification for making and interpreting soil surveys. USDA Soil Conservation Service, Agriculture Handbook No. 436. U.S. Government Printing Office, Washington, D.C. pp. 754. Sombroek, W.G. (1990). Soils on a warmer earth: tropical and subtropical regions. In: (HW. Scharpenseel, M. Schomaker and A. Ayoub, eds). Soils on a warmer earth. Developments in Soil Science, 20,157-174. S0renson, L.H. (1981). Carbon-nitrogen relationships during the humification of cellulose in soils containing different amounts of clay. Biochemistry, 13, 313-321.

Biology and

296

SOVEUR. (1991). Mapping of soil and terrain vulnerability to specified chemical compounds in Europe at a scale of 1:5 M. (N.J. Batjes, ed.). IS RIC,

Wageningen. Spedding, P.J. (1988). Peat. Fuel, 67, 883-900. Sprengart, J. and Wagner, G. (1994). Umweltprobenbank des Bundes: Richtlinie zur Probenahme

und

Probenbearbeitung

fOr

Boden.

Ecomed-Verlag

(in

preparation). SSLRC. (1988). Water erosion in England and Wales 1982-1984. Soil Survey and Land Research Centre. Silsoe, UK. pp. 57. Stach, E. (1975). Coal Petrology, Gebruder Borntraeger, Stuttgart, pp. 428. Stanford, G. and Epstein, E. (1974). Nitrogen mineralisation - water relations in soils. Soil Science Society of America Proceedings, 38, 103-106.

Stanhill, G. (1986). Irrigation in arid lands. Philosophical Transactions of the Royal Society of London, A, 316, 261-273.

Steers, C.A. and Hajek, B.F. (1979). Determination of map unit composition by a random selection of transects. Soil Science Society of America Journal, 40, 156-160. Steffen, W.L., Walker, B.H., Ingram, J.S.I. and Koch, G.w. (eds). (1992). Global Change and Terrestrial Ecosystems: The Operational Plan. IGBP Report No. 21, Stockholm. pp. 95. Stevenson, F.J. (1985). Geochemistry of soil humic substances. In: (G.R. Aiken, D.M. McKnight, R.L. Wershaw. and P. MacCarthy, eds). Humic Substances in Soil, Sediment and Water. Wiley-Intersceince, NY, pp. 13-52. Stewart, J.w.B., Anderson, D.W., Elliott, E.T. and Cole, C.V. (1990). The use of models of soil pedogenic processes in understanding changing land use and climatic change. In: (H.w. Scharpenseel, M. Schomaker and A. Ayoub, eds). Soils on a Warmer Earth. Developments in Soil Science, 20, 121-131. Stigliani, B. and Salomons, W. (1992). Pollutants and some not impossible environmental problems caused by climate change. IIASA, Laxenburg, Austria. pp.1-15. Stockle, C.O., Williams, J.R. Rosenberg, N.J. and Jones, C.A. (1990). Estimating the effect of carbon dioxide - induced climate change on growth and yield of crops: modification of the EPIC model for climate change analysis. Internal Note. USDA-ARS, Temple, Texas. Draft 4. pp. 22. Stoeppler, M. and Zeisler, R. (eds) (1993). Biological environmental specimen banking. The science ofthe total environment, 1391140 (BESB special issue:

297

Proceedings of the 1st International Symposium on Biological Environmental Specimen Banking, Vienna, Austria, 22-25 September 1991. pp. 553. Sundquist, E.T. (1985). Geological perspectives on carbon dioxide and the carbon cycle. In: The Carbon Cycle and Atmospheric CO2 : Natural Variations Archean to Present, Geophysical Monographs (American Geophysical Union), 32, 5-

60. Sundquist, E.T. (1993). The global carbon dioxide budget. Science, 259, 934-941. Szabolcs, I. (1979). Review of research on salt-affected soils. UNESCO, Paris. Szabolcs, I. (1986). Agronomical and ecological impact of irrigation on soil and water salinity. Advances in Soil Science, 4,189-218. Szabolcs, I. (1989). Salt affected soils. CRC Press Inc., Boca Raton, Florida. Szabolcs, I. and Redly, M. (1989). State and prospects of soil salinity in Europe. Agrokemia

es Talajtan, 38, 537-558.

Szabolcs, I. (1990). Impact of climatic change on soil attributes: influence on salinization and alkalinization. In: (H.W. Scharpenseel, M. Shomaker and A. Ayoub, eds). Soils on a warmer Earth. Developments in Soil Science, 20, 6169. Szabolcs, I. (1991). Salinization potential of European soils. In: (F.M. Brouwer, E.J. Thomas and M.J. Chadwick, eds). Land use changes in Europe: processes of change,

environmental transformations and future patterns. Geojournal

Library, Vol. 18, 293-315. Kluwer Academic, Dordrecht. Tans, P.P., Fung, I.Y. and Takahashi, T. (1990). Observational constraints on the global atmospheric CO 2 budget. Science, 247,1431-1438. Tarutis, W.J. (1993). On the equivalence of the power and reactive continuum models of organic matter diagenesis. Geochimica et Cosmochimica Acta, 57, 1349-1350. Tegelaar, E.W., de Leeuw, J.W., Derenne, S. and Largeau, C. (1989). A reappraisal of kerogen formation. Geochimica et Cosmochimica Acta, 53, 3103-3106. Terjung, W.H., Liverman, D.M. and Hayes, J.T. (1984). Climatic change and water requirement for grain corn in the North American Great Plains. Climatic Change, 6, 193-220.

Theocharopoulos, S.P., Karayianni, M., Gatzogianni, P., Afentaki, A. and Aggelides, S. (1993). Nitrogen leaching from soils in the Kopais area of Greece. Soil Use and Management, 9, 76-84.

Thomasson, A.J. (1979). Assessment of soil droughtiness. In: (M.G. Jarvis and D. Mackney, eds). Soil Survey Applications. Soil Survey Technical Monograph No. 13, Soil Survey of England and Wales, Harpenden.

298

Thomasson, A.J. (1982). Soil and climate aspects of workability and trafficability. In: Proceedings of the 9th Conference of the International Soil Tillage Research Organisation, Osijek, Yugoslavia. pp. 551-557.

Thomasson, A.J. and Jones, R.J.A. (1991). An empirical approach to crop modelling and assessment of land productivity. Agricultural Systems, 37, 351-367. Thomes, J. (1991). Mediterranean Desertification and Land Use - MEDALUS I. Commission of the European Communities, DGXI/ (Science, Research and Development), Brussels, EPOCH-CT90-0014-(SMA}.

Tiessen, H. and Stewart, J.W.B. (1988). Light and electron microscopy of stained microaggregates: the role of organic matter and microbes in soil aggregation. Biogeochemistry, 5, 312-322.

Tinker, P.8. and Ineson, P. (1990). Soil organic matter and soil biology in relation to climate change. In: (H.w. Scharpenseel, M. Schomaker and A. Ayoub, eds). Soils on a warmer Earth. Developments in Soil Science, 20, 71-88.

Tisdall, J.M. and Oades, J.M. (1982). Erosion-induced changes in the soil properties of a fragipan soil. Soil Science Society of America Journal, 54, 223-228. Titus, J.G. (1987). Causes and effects of a sea level rise. In: Preparing for climate change. Proceedings of the First North American Conference on preparing for climate change. (October 27-29, 1987), Government Institutes Inc., Rockville,

MD. pp. 125-139. Toward. (1988). Toward an understanding of global change. National Academy Press, Washington, D.C. Toy, T.J. (1983). Accelerated erosion: processes, problems and progress. Geology, 10,524-529. Trimble, S.W. (1990). Geomorphic effects of vegetation cover and management: some time and space considerations in prediction of erosion and sediment yield. In: (J.B. Thomes, ed.). Vegetation and Erosion, Processes and Environments. John Wiley and Sons, London, pp. 55-65.

UNEP. (1989). Criteria for assessing vulnerability to sea level rise: a global inventory to high risk areas. UN Environmental Programme and the Government of The Netherlands, Draft Report.

Unger, P.w. and Pringle, 8. (1981). Pullman soils: distribution, importance, variability and management. Texas Agricultural Experimental Station, 8-1372, College Station, Texas, USA. Varallyay, G. (1990a). Influence of climatic change on soil moisture regime texture, structure and erosion. In: (H.W. Scharpenseel, M. Schomaker and A. Ayoub, eds). Soils on a warmer Earth. Developments in Soil Science, 20, 39-49.

299

Varallyay, G. (1990b). Consequence of climate induced changes in soil degradation processes. Transactions of the 14th ISSS Congress, Kyoto, Vol. V, 265-270. Varallyay, G. (1990c). Potential impacts of global climatic changes on soil moisture regime and soil degradation processes. Proceedings of the UNESCO/MAB International Seminar: Future Research Trends in MAB,

University of

Fisheries, Tokyo. pp. 256-267. Varallyay, G. (1968). Salt accumulation processes in the Hungarian Danube Valley.

Transaction of the 9th ISSS Congress, Adelaide, 1, 371-380. Varlet-Grancher, C., Gosse, G., Chartier, M., Sinoquet, H., Bonhomme, R. and AIIirand, J.M. (1989). Mise au point: rayonnement solaire absorbe ou intercepte par un couvert vegetal. Agronomie, 9, 419-439. Velben, MA (1985). Geochemical mass balances and weathering rates in forested watersheds of the southern Blue Ridge. American Journal of Science, 285, 904-930. Vincent, C.D. and Gregory, P.J. (1989). Effects of temperature on the development and growth of winter wheat roots II. Fields of temperature, nitrogen and irradiance. Plant and Soil, 119, 99-110. Vitousek, P.M. and Howarth, R.w. (1991). Nitrogen limitation on land and in the sea: how can it occur? Biogeochemistry 13, 87-115. Vitousek, P.M. and Matson, PA (1993). Agriculture, the global nitrogen cycle, and trace gas flux. In: (R.S. Oremland, ed.). The Biogeochemistry of Global

Change. Chapman and Hall, New York, pp. 193-208. Vogt, KA, Publicover, DA and Vogt, D.J. (1991). A critique of the role of ectomycorrhizas in forest ecology. Agriculture, Ecosystems and Environment, 35,171-190. Voltz, M. and Rambal, S. (1987). Evaluation de CORNGRO: un modele de simulation des transferts hydriques et de la croissance d'une culture de mai's. Agronomie, 7(9),727-738. Voroney, R.P. and Angers, DA (1993). Analysis of the short-term effects of management on soil organic matter using the CENTURY model. In: World

Soils and Greenhouse Gas Emissions. Advances in Soil Science. Lewis Publishers, CRC Press, Boca Raton, Florida. (In press). Waggoner, P.E. and Schetter, J. (1990). Future water use in the present climate. In: (P.E. Waggoner, ed.). Climate change and U.S. Water Resources. John Wiley and Sons, New York, pp. 19-39.

300

Wahlen, M., Allen, D. and Deck, B. (1991). Initial measurements of CO 2 concentrations (1530 to 1940 AD) in air occluded in the GISP-2 ice core from central Greenland. Geophysics Review Letters, 18, 1457-1460. Wang, C. (19?5). Occurrence of lithologic discontinuities in two Guelph soils. Unpublished MSc Thesis, University of Guelph, Guelph, Ontario. pp. 43.

Wang, C. and Arnold, R.W. (1973). Quantifying pedogenesis for soils with discontinuities. Soil Science Society of America Proceedings, 37, 271-278. Ward, R.C. (1963). Observations of potential evapotranspiration (PE) on the Thames floodplain. Journal of Hydrology, 1, 183-194. Wardle, DA (1992). A comparative assessment of factors which influence microbial biomass carbon and nitrogen levels in soil. Biological Review, 67, 321-358. Warrick, RA, Barrow, E.M. and Wigley, T.M.L. (1990). The greenhouse effect and its implications for the European Community. Commission of the European

Communities, DGXII (Science, Research and Development), EUR 12707, pp. 30. Warrick, R.A., Gifford. R.M. and Parry, M.L. (1986). CO 2 ,climatic change and agriculture. In: The greenhouse effect, climatic change and ecosystems. John Wiley and Sons, Chichester, pp. 393-473. Warrick, RA and Oerlemans, J. (1990). Sea level rise. In: (J.T. Houghton, G.J. Jenkins, J.J. Ephraums, eds). Climate change. The IPCC Scientific assessment. WMO & UNEP, Cambridge University Press, Cambridge, pp.

257-282. Weir, A.H., Bragg, P.L., Porter, J.R. and Rayner, J.H. (1984). A winter wheat model without water or nutrient limitations. Journal of Agricultural Science, Cambridge, 102, 371-383. Weir, A.H., Day, W. and Sastry, T.G. (1985). Using a whole crop model. In: (W. Day and R.K. Arkin, eds). Wheat growth and modelling. NATO ASI Series A, Vol. 86, pp. 339-355. Plenum, New York. Wheeler, D. and Martin-Vide, J. (1992). Rainfall characteristics of mainland Europe's most southerly stations. International Journal of Climatology, 11, 1-25. Whipps, J.M. (1990). Carbon economy. In: (J.M. Lynch, ed.). The Rhizosphere, Wiley, Chichester, pp. 59-98. Whisler, F.D., Acock, B., Baker, D.N., Fye, R.E., Hodges, H.F., Lambert, J.R., Lemmon, H.E., McKinion, J.M. and Reddy, V.R. (1986). Crop simulation models in agronomic systems. Advances in Agronomy, 40, 141-208. White, JW.C. (1993). Don't touch that dial. Nature, 364, 186.

301

Wicherek, S. (1993). Farm land erosion in temperate plains environment and hills. Elsevier, Amsterdam. Wild, A. (1993). Soils and the environment: an introduction. Cambridge University Press, Cambridge. Wilkerson, G.G. et al. (1985). SOYGRO V.5.D: Soya bean crop growth and yield model. Technical documentation (Internal Report), Gainesville University,

Florida. pp. 220. Williams, J.R., Jones, CA, Kiniry, J.R. and Spanel, D.A. (1989). The EPIC crop growth model. Transactions of the American Society of Agricultural Engineers, 32(2),497-511. Williams, J.R. and Renard, K.G. (1985). Assessment of soil erosion and crop productivity with process models (EPIC). In: (R.F. Follett and SA Stewart, eds). Soil Erosion and Crop Productivity. American Society of Agronomy; Crop Science Society of America; Soil Science Society of America, Madison, WI, pp.67-103. Williams, E.J., Guenther, A. and Fehsenfeld, F.C. (1992). An inventory of nitric oxide emissions from soils in the United States. Journal of Geophysics Research, 97,7511-7519. Wilson, MA (1987). NMR Techniques and Applications in Geochemistry and Soil Chemistry. Chapters 7-11. Pergamon Press, New York.

Wilson, MA, Barron, P.F. and Goh, K.M. (1981). Cross polarisation 13C-NMR spectroscopy of some genetically related New Zealand soils. Journal of Soil Science, 32, 419-425.

Wilson, D.R. and Jamieson, p.o. (1985). Models of growth and water use of wheat in New Zealand. In: (W. Day and R.K. Atkin, eds). Wheat growth and modelling. NATO ASI Series A, Vol. 86, pp. 211-216. Plenum, New York. Wilson, M.F. and Henderson-Sellers, A. (1985). A global archive of land cover and soils data for use in general circulation models, Journal of Climatology, 5, 119143. Wischmeier, W.H., Johnson, C.B. and Cross, B.V. (1971). A soil erodibility nomograph for farmland and construction sites. Journal of Soil and Water Conservation, 26, 189-193. Wischmeier, W.H. and Smith, D.o. (1978). Predicting rainfall erosion losses: a guide to conservation planning. USDA Agriculture. Handbook No. 537, Washington, D.C. pp. 58. Wise, SA, Zeisler, R. and Goldstein, G.M. (eds) (1988). Progress in environmental specimen banking. National Bureau of Standards, Special Publication 740,

302

U.S. Department of Commerce, U.S. Govemment Printing Office, Washington, D.C. pp. 202. Wolters, W. (1992). Influences on the efficiency of irrigation water use. ILRI Publication 51, Wageningen.

Woodward, F.1. (1992). Predicting plant responses to global environmental change. New Phytologist, 122, 239-251.

Yaalon, D.H. (ed.). (1971). Paleopedology: origin, nature and dating of paleosols. Israel University Press, Jerusalem. Yariv, S. and Cross, H. (1979). Geochemistry of Colloid Systems. Springer-Verlag, New York. Yassoglou, N. and Kosmas, C.S. (1990). Potential erosion and desertification risks in Greece. Balkan Scientific Conference on Environmental Protection. Bulgarian Academy of Science.

Yoshino, M.M., Horie, T., Seino, H., Tsujii, H., Uchijima, T. and Uchijima, Z. (1988). The effect of climatic variations on agriculture in Japan. In: (M.L. Parry, T.R. Carter and N.T. Konijn, eds). The impact of climatic variations on agriculture: Vol. 1: Assessments in cool, temperate and cold regions. Kluwer, Dordrecht,

pp. 723-868. Youngs, E.G. (1965). Horizontal seepage through unconfined aquifers with hydraulic conductivity varying with depth. Journal of Hydrology, 3, 283-296. Yu Tiam-ren. (1985). Physical chemistry of paddy soils. Science Press, Beijing. Zak, D.R., Pregitzer, K.S., Curtis, P.S., Teeri, J.A., Fogel, R. and Randlett, D.1. (1993). Elevated atmospheric CO2 and feedback between carbon and nitrogen cycles. Plant and Soil, 151,105-117. Zinke, P.J., Strangenberger, A.G., Post, W.M., Emanuel, W.R. and Olson, J.S. (1984).

Worldwide

organic

soil

carbon

and

ORNLfTM8857, Oak Ridge National Laboratory, TN.

nitrogen

data.

Report:

INDEX Absorption, 53 Accumulated temperature, 57 Accumulation, 43, 44, 47, 49, 50-53, 73,77,105,106,156,159,175,194,223,226-228 Acid deposition, 157 Aeration, 9, 51,142,152,160 Aerobic, 142,164-166,168,175 Aggregation, 29, 36,142,160,167 Agrochemicals, 43 Agroclimatic zones, 65 Agroforestry, 205 Agronomy, 72 Albite, 170 Albolls, 229 Alfalfa, 207 Alginites, 175 Aliphatics, 177 Alkalinity, 39, 43-45, 54,170,174,180 Allophane, 176 Alluvial fan, 17, 18 Amino acids, 176, 178 Ammonia, 130, 144 Ammonium, 109, 144,244 Anaerobic, 141, 142, 164, 166, 168 Anoxic conditions, 108 Anthesis, 75 Anthropogenic activities, 171 Aqualfs, 229 Aquents, 229 Aquepts, 229 Aquic Cryochrept, 18 Aquolls,229 Arctic, 163, 165, 169, 174 Aridity, 25, 29, 38, 44, 51 Arkley method, 208 Aromatics, 177 Assimilates, 77, 81, 83, 88, 91, 92 Asynchrony, 164, 168 Atolls, 46 Available water, 58, 68, 72, 73, 88, 89, 100, 104,105 Barley, 116, 140 Basalt, 15 Beech,231,234,251 Benzene rings, 178 Bicarbonate, 144, 161 Biodiversity, 3, 264 biogeochemical cycles, 17 Biogeochemical cycling, 169 Biogeochemical processes, 14

304

Biological activity, 47, 178 Biome, 181 Biopolymers, 175 Biosphere, 4, 9,155-158,187,241 Biosynthesis, 80 Bitumen, 174, 175 Boreal, 164, 165, 181, 194,240,258,262 Brunisols, 194 Buffering, 109, 199 bulk density, 17, 124, 142,218,222,232,253 Bypass flow, 125, 142, 143, 152 C:N ratio, 140, 143, 145, 162 C3,91-93, 102, 103, 140 C4, 91-93, 102 Calcareous, 15, 20, 245 Canopy, 47,80, 81,92,95, 139, 144 Capillary, 43, 51,53 Capillary transport, 43, 51,53 Carbohydrate, 80, 177 Carbon, 5, 7, 8,17,19,26,36,38,55,68,73,91,92,137-141,145-147, 151, 152, 155, 157-159, 169-183, 193-195, 197, 198,221,222,227,229,230-237,240,241,243,254, 258,262,265,266 carbon sequestration, 17, 193, 229 Carbonato, 170, 180 Carotenoids, 176 Catastrophe, 190 Catchments, 18 Cation exchange, 53, 142,208 Cation leaching, 155 Cations, 161 Cellulose, 176 CH 4 ,39, 155, 156, 158, 162, 165-168 Chloride, 106 Citrus, 106, 207 Clay minerals, 53, 183 Clays, 53, 176 Climate scenarios, 150, 211 Cloudiness, 40, 50, 55, 68, 94, 188 CO 2 , 4, 8, 39, 49, 55, 63, 68, 71, 74, 79, 91-96, 99,100-103,114,137,139,140-142,147, 151,155,156,158-162,165-171,174,177,180,181,183, 193, 197, 198, 22~ 224, 226, 227,237,240,241 CO 2 fertilisation, 137, 140, 147, 151, 155, 158, 160 Coal, 175, 177, 178 Coastline, 46 Coefficient of Linear Extensibility, 119 Conductivity, 43, 123, 132, 133, 217-219 Cotton, 65, 72, 76,103,207 Cracking clay soils, 113, 115, 125, 127, 129 Crop potential, 55, 57, 67 Crop residue, 141, 160, 222 Crop yield, 61, 62, 207, 222

305

Cultivar,4 Dairy production, 103 Dark respiration, 92, 93 Decomposition, 7, 99,137-140,143, 147, 160, 162-165, 168-170, 173, 178-180, 182, 197, 238,265 Degradation, 10,25-27,31,35,36,38,40,42,53,66,169-171,175-178, 180,201,208 Degree-days, 75, 124, 126 Delta areas, 46 Denitrification, 137, 138, 141, 142, 145, 150, 151-153, 155, 157, 159-161, 163, 164, 167, 243,244 Denitrifying bacteria, 141 Deposition, 10, 15, 16,30,138,145,157,172,174,176,179,199,251,253,254 Depositional surfaces, 15 Depth to water table, 162, 165, 166, 168 Diffusion, 43, 93, 180 Diseases, 64, 103, 264 Douglas fir, 231, 234 Drainage, 17, 18,43,46,47,49,51,53,65,113,114,116,120-123,131, 133, 135, 139, 165,166,201,206-208,217,222 Drainfiow, 120, 125, 126, 128-132 Drought, 4, 7,32,44,47,67, 73,89,93,94,109,140,165,223,224,227 Droughtiness, 58, 60, 61, 67 Dry-matter, 68, 79, 83, 102, 124, 224 Dryland, 99,102,109,263,264 Dystrochrept, 234 Ecosystems, 3, 6, 9, 18,27,47,96,132,155,156,158-160,162-164,187,188, 223, 235, 237-240,251,252,254,257,259,263 Emergence, 72, 75 Emissions, 8,10,141,155,156,162,165-169,174,180,182,198 Environmental Specimen Banking, 249, 251, 252 Enzyme, 80, 91, 92, 160 Ethylene, 93 European Community, 26, 206 Eustatic rise, 45 Eutrochrept, 231,234 Evaluation, 40, 44, 62, 66, 72, 178, 187, 189, 190, 199, 201, 203, 205, 206, 208, 209, 231, 233,234,249,250,252,254 Evaporation, 31, 34, 40, 47, 49, 50, 73, 86, 93, 95, 99,107-109,124,139,146,147,197 Evapotranspiration, 6, 7, 29, 43, 57, 63, 64, 85-87, 89, 91, 94-96, 104, 106, 108, 116, 117, 120,121,123,125,126,128-130,135, 146, 147,224,225,243,245-248 Expert systems, 66, 205 Extreme, 4, 7, 26, 27, 36, 40, 44, 50, 53, 68, 89, 107, 170, 182, 188-190,261,263 Exudates, 5, 141, 151, 159 Fallow, 37, 107, 109 Farmyard manure, 197 Fatty acids, 176 Feedback, 152, 158, 162, 168, 174,237 Fertiliser application, 45, 143 Fertility, 7, 26, 68, 161

306

Field capacity, 58,119,128,142,143 Fine roots, 151, 159, 162, 167 Fir, 231, 234 Flood,46 Flooding, 46, 110 Flow, 11,29,37,43,51,114,115,118-120,122,124,125, 128-130, 142, 143, 152, 159, 167 Flux, 86, 108, 133, 144, 157, 165, 171, 180, 182,217,237 Forests, 160, 163, 164, 222, 235, 258 Fournier index, 208 Frost, 103, 188,206 Fulvic acids, 175 Gardner's model, 218, 219 Genotype, 75 Geomorphology, 14-16, 18,20,50 Geopolymers, 175, 176 Germination, 66, 97 Gilgai,17 Glaciers, 16, 45 Gleysols, 194 Global Circulation Models, 6, 39, 63, 188 Global radiation, 79, 97 Granite, 234 Grass, 58,87, 140, 146,243 Grassland ecosystems, 223 Greenhouse gases,S, 8, 39,168,187 Haplaquents, 245 Haplorthod, 231, 234 Hapludalf, 234 Harvest index, 78, 87 Harvestable dry matter, 227 Heat-trapping, 156, 162, 165 Herbaceous, 238, 239 Humic, 169, 175, 177, 178, 182, 183,231 Humin, 175 Huminites, 175 Humus, 145, 174, 175, 180, 181, 197,231,233,244,251,253,254 Hydraulic conductivity, 43,123,132,133,217-219 Hydrographs, 129 Hydrology, 15, 17, 18,20,21,27,38,63,64, 113-115, 120, 127,129,132,152,222,262 Hyetograph, 118, 119 Hyphae, 160 Ice caps, 45 IGBP, 4,9, 10, 187,228,260,266 Immobile, 123 Immobilisation, 141, 145, 151, 157,159,163,224,243 Impermeable, 114 Industrial, 35, 39, 91,102,103,157,173,224,251 Infiltration rate, 47, 120, 128

307

Intrusion, 35, 46, 108 lon, 5, 53, 254 IPCC, 5,138,141,266 Ironstone, 17 Irrigation efficiency, 104 Irrigation water, 43, 44, 53, 73, 94, 105,108,109 Kaolinite, 176 Kerogen, 169, 174-178, 182, 183 Kinetics, 179, 262 Knowledge base, 66, 203 Land capability, 206 Land Characteristics, 201 Land evaluation, 62, 66, 205, 206 Land management, 38, 55, 64, 65, 66, 199, 201, 206, 221 Land Qualities, 57, 199, 200 Land use, 4, 10,27,36,37,40,49,64,66,69,101,137,138,151,152, 155, 156, 158, 180,193,194,199,205,206,208,222,229,253,265 Land Utilisation Types, 62 Land-use systems, 56 Landfills, 47, 156 Landscape, 13-22, 27, 32-34, 208,235, 264 Larch, 231, 234 Leaching, 15,43,51-53,99, 101, 105-108, 110, 113, 123, 130, 131, 137-140, 142-145, 150,151,153,155,161,163,208,243-245 Leaching potential, 15 Leaf area index, 29, 78,81,85,95 Leaf conductance, 95, 96 Leaf resistance, 95 Light, 73, 80, 81, 91, 94, 156, 158, 194, 213 Ligneous, 238,239 Lignin, 175 Lignite, 175 Linseed, 151 Lipids, 176 Litter, 7, 8, 49, 162,223,226,241,251,262 Loam, 107, 143,213,223 Loess, 16, 17,234 Lolium perenne, 223 Luvisols, 194 Machinery work days, 61 Maize, 75-77, 88, 89, 92, 151,207,222 Managed wetlands, 132 Management practices, 53, 65-67,193,194,207,221 Management Qualities, 200 Map units, 19, 63, 245 Maps, 18, 19,21,252 Marginal lands, 26, 29, 44, 53 Matric potential, 58' Matric potentials, 63

308

Matrix flow, 115 Maturity, 75,77,78,84,87,183 Melon, 207 Mesozoic, 174 Methane, 3, 5, 8 Methanotrophic bacteria, 165 Micro-organisms, 138 Microbial, 8, 65,138,139,140,143,151,152,155-168,171,175-178, 182, 183, 194,224,

237,262,264,265 Microbial activity, 65,151,159,164,171,177,178,182,237 Millet, 92 Mineralisation, 99, 113,123,130, 137, 139, 140-143, 145, 150-153, 155, 157, 162-164, 194,224,227,228,231,235,237,243 Minerals, 43, 53, 74,170,176,183 Mobile, 123 Models, 6-9, 11,20-22,29,39,49,56,57, 62, 63, 67, 69, 71-73, 75-81,83-86,88, 91, 9397,101-104,113,115,118,123,127,130,131,134,138, 141, 142, 151, 152, 165, 169171,187, 188, 195,203,205-207,217,221,235,241,259-261,265 Moisture, 17,22,26,29,31,44,46,47,49,50,53,57,58,61,66,101, 109, 113, 114, 118121,123,124,126-130,134,139,141,142,144,155,158, 162, 164, 167, 168, 197,206, 218,241,253,260 Mole drains, 116, 143 Mollisols, 35, 194, 229 Monsoons, 46 Montane forest, 264 Montmorillonite, 176 Mycorrhizae, 159, 167 N fixation, 155, 157, 159, 160, 167, 168 N mineralisation, 155, 163, 164,224,227,228 N20, 141, 142, 155, 156, 158, 162, 167, 168 Net primary production, 180, 197,241 Net radiation, 96 Nitrate, 101, 107-110, 113, 116, 123, 124, 130-132, 138, 139, 141-144, 152, 244-247 Nitrates, 72, 245, 246 Nitrification, 131, 142, 164,167,243 Nitrogen, 7, 8, 78, 79, 94,106,116,123,137,138,140-142,144-146,152,155, 157, 175,

193,194,227,243-248,253,254,258,262

Nitrogen dynamics, 7,137,145,243 Nitrous oxide, 8, 141, 142 NMR, 177 NO, 4,13,44,48,51,58,76,84,85,89,95,101,107,114,120, 124, 129, 133, 139, 142,

145,147,155,156,158,162,163,167-170,178,179,181, 182, 199,221,227,232,238, 241,260,263 Nutrients, 14, 138, 158, 159, 164, 171, 174, 182,222,237,262 Oak, 231, 234 Old sediment carbon, 169 Organic carbon, 17, 19,36,68,138,141,169,170,172-180,182,183,193-195, 198,222,

230,231,241,262

309

Organic matter, 5, 7-11,14,16,29,35,68,89,99,100,107,138-140, 142-145, 147, 151, 152,155,157,163,178,179,193-195,197,201,218,219, 224, 226, 227, 228, 232, 237, 254,258,262 Organic N, 137, 147-149 Organo-clay complex, 176 Oxic horizons, 165 Palaeobotany, 6 Particle size distribution, 14,201,254 Partition, 113, 118, 119, 121-123, 129, 134 Pasture, 103 Peach,207 Peat, 165, 167, 169, 172, 174, 175, 177, 178 pediment, 17 pedisediments, 17 Pedogenesis, 72 Pedology, 14, 18, 19 Pedons, 19 Pedosphere, 13,23,229 Pedotransfer functions, 63, 208, 209 Peds, 115, 119, 127, 128 Pelo-stagnogley, 116 Penman, 63,64,69, 87,95, 117,245 Peptides, 176 Percolation, 107, 108 Permafrost, 47, 165 Pests, 64, 102, 103, 264, 266 Petroleum, 175, 177 pH,83,144,207,208,223,231,254 Phenological, 55, 66, 71, 77, 78, 81,87 Phenols, 176, 178 Phenophases, 75-77 Phosphorus, 138,262 Photoperiodism, 97 Photorespiration, 92 Photosynthesis, 71, 73, 76, 79-81, 83, 88, 89, 91, 92, 94, 158, 170 Phreatic layer, 108 Phytomass, 169, 174, 180,226,227 Plant Physiology, 72, 92 Plinthite,17 P02,160 Podzols, 194 Policy, 27, 38, 55, 62,111,138,230,259,264,265 Pollutant, 10, 131 Pollutants, 10,47,208 Pollution, 3, 10, 46, 71,72, 155, 158,249 Polysaccharides, 160, 176, 178 Pore-size, 53 Potato, 107, 109,207 Precipitation patterns, 67, 164 Protein hydrolysis, 176

310

Q10effects, 158, 168 Radiation, 39, 63, 68, 79, 81, 92, 94-97,139,222 Radiocarbon, 16, 175, 198 Rate, 6-8, 27, 28, 39, 40, 43-45, 47, 49, 50, 53, 63, 68, 83, 86, 99, 101,105, 118-120, 123125,128,137-140,160-162,169,171-173, 176, 178-182, 197,211,212,238,240,241, 244 Recalcitrant material, 174 Recharge, 17,47, 113, 120, 133 Resins, 176 Resolution, 59, 63, 67, 187,259-261,264 Respiration, 8, 71, 73, 80, 92, 93,155-158,160-162,164,166,168,170,224,237-241 Rice, 16, 22, 103, 156 Rock,34, 134, 169, 171, 176-178, 180-183 Root density, 84-86 Root exudates, 5, 141, 159 Root growth, 73, 75, 83-85, 89,139-141,241 Root-zone, 53 Runoff, 21, 43, 47, 49,104,107,114,118-121,129,157,188-190 Sahel, 7, 9 Salt, 6, 35, 43, 44-46, 50-53, 101, 105, 254, 258, 263 Salt accumulation, 43, 50-53,105 Salt affected soils, 43 Sand,17,217,218,234 Sandstone, 33, 234 Savannah,264 Scots pine, 231, 251 Sea level rise, 45, 108 Sedimentary rocks, 170, 171, 173-176, 178, 179, 182 Seed-bed, 66 Seepage, 51, 128, 129, 132 Semi-arid, 7,17,25,31-33,100,105,238,241 Senescence, 75, 77,93, 144,243 Sequestration, 17, 152, 166, 193, 194,229,230 Sewage,43,45,156 Shrink-swell, 114 Simulation, 56, 57, 62, 66, 67, 69, 75, 94, 97, 107, 108, 128, 129, 134, 145, 190,221,222, 243 SimuSolv, 180 Sink,8, 160, 164, 169, 171, 174, 182, 183, 193,235 Sitka spruce, 235, 236 Smectites, 177 Snow,40,49,51,189 Snowmelt, 46 Socio-economic, 40, 53, 55, 62, 66, 69 Socio-economics, 69 Sodification, 108 Sodium saturation, 201,207,208 Soil aeration, 142, 160 Soil aggregates, 36,159,160 Soil biota, 3, 237, 262, 264

311

Soil classification, 6, 14,72 Soil database, 10 Soil erosion, 3, 25, 26, 30-32, 188, 199,201,203,211,222,263 Soil formation, 16, 40, 41, 47 Soil hydrology, 17, 152 Soil loss, 26,190,212-215,222 Soil macroposity, 113 Soil microorganisms, 155, 156, 158, 167 Soil moisture, 17,22,26,29,44,47,49,50,53,57,66,113,118-121, 123, 124,126, 128130,134,139,142,144,155,162,164,167,168,197,253 Soil moisture regimes, 17,22,49 Soil organic matter, 7-11,14,16,99,139,140,144,151,152,155,157,163,179,194, 195,197,226 Soil pollution, 3 Soil profile, 19,20,43,47,51,53,58,72,106,123,126,165,203,207, 217, 231, 245, 253,254 Soil surveys, 18, 21, 201 Soil temperatures, 131, 155, 156, 162 Soil type, 6, 104, 142,233-235,245,253 Soil water characteristic curve, 217 Soil water retention, 63 Soil water status, 66, 86 Solar energy, 71, 75, 76, 79 Solar radiation, 39, 63, 222 Solonchak, 6, 52 Solute, 99, 114, 115, 124, 130, 133, 134 Sorghum, 76, 92, 93 Spatial distribution, 19, 258 Spatially distributed systems, 115 Sprinkler, 65, 105 Spruce, 231, 234-236, 251 Stomata, 86, 93 Storm tides, 46 Stratigraphy, 15, 16, 18, 20, 21 Stress, 4, 22, 71, 74, 77-79, 81, 88, 89,102,109,126,140,257 Structured soils, 113, 114 Stubble, 223, 226, 227 Sugars, 176, 178 Suitability, 55-57, 61-63, 66, 68, 204, 206, 207 Sulphur, 138,262 Sunflower, 65, 75, 76, 81, 93, 151, 207 Surface waters, 43, 138 Temperature, 76, 146, 181, 189, 239 Temporal variability, 44, 50 Thermokarst, 165 Thornthwaite, 63, 64, 208 Till, 16, 17,20,221 Tillage, 5, 51, 65, 66, 152, 201, 221, 222, 263 Transects, 19,37 Transpiration, 47, 71, 73,76,85,86,88,93,139,224 Transport, 43, 46, 50, 51, 53,134,144,158,199,208

312

Trickle, 105 Tropospheric, 55 Tuff, 16 Tundra, 7 Typhoons, 188 Uptake, 71, 73,123,124,128,140,143,145,157,164,174,182,183,243 Urban, 39, 103, 104, 180,251 Urea, 65,244 Vapour pressure, 87, 93, 94, 95 Vernalization, 97 Vertisols, 17 Very old carbon, 171 Vitrinites, 175 Volatilisation, 243 Volcanic ash, 15, 16 Volcanism, 173 Vulnerability, 66, 199-201, 203, 206, 208 Water, 17,62,83,88,99, 103, 113, 123, 124, 127, 128, 139,221,222,224,225 Water quality, 113, 120, 130 Water Resources, 67, 99, 101, 103-105, 111,208,223,227,263 Water stress, 81, 88, 89, 140 Water table, 46, 51-53, 83, 108, 127, 128, 132, 133, 135, 162, 165, 166, 168,208 Water use efficiency, 8, 102, 103, 142, 223-225 Waterlogging, 45, 116 Watershed, 43, 104, 164 Waxes, 176 Weathering, 15,43,44,53,99,159,161,167,170,182,263,266 Wetland, 19, 113, 132, 168 Wetlands, 132, 133, 166, 180 Wetting/drying cycles, 114 Wilting point, 58, 89 Wind, 7,10,36,43,49,63,139,188,199,222 Windspeed, 68, 144 Winter, 28, 29, 39, 49, 50, 51, 58, 60, 61, 65, 94, 96, 97,101,106,107,116,120,122, 124,126,128,129,137-139,142,145,147-151,153, 164, 189,244,245 Workability, 61 Yield, 4, 58, 61, 62, 65, 77, 87, 89, 94, 100-103, 140, 206-208, 222, 244

The ASI Series Books Published as a Result of Activities of the Special Programme on Global Environmental Change This book contains the proceedings of a NATO Advanced Research Workshop held within the activities of the NATO Special Programme on Global Environmental Change, which started in 1991 under the auspices of the NATO Science Committee. The volumes published as a result of the activities of the Special Programme are: Vol. 1: Global Environmental Change. Edited by R. W. Corell and P. A. Anderson. 1991. Vol. 2: The Last Deglaciation: Absolute and Radiocarbon Chronologies. Edited by E. Bard and W. S. Broecker. 1992. Vol. 3: Start of a Glacial. Edited by G. J. Kukla and E. Went. 1992. Vol. 4: Interactions of C, N, P and S Biogeochemical Cycles and Global Change. Edited by R. Wollast, F. T. Mackenzie and L. Chou. 1993. Vol. 5: Energy and Water Cycles in the Climate System. Edited by E. Raschke and D. Jacob. 1993. Vol. 6: Prediction of Interannual Climate Variations. Edited by J. Shukla. 1993. Vol. 7: The Tropospheric Chemistry of Ozone in the Polar Regions. Edited by H. Niki and K. H. Becker. 1993. Vol. 8: The Role of the Stratosphere in Global Change. Edited by M.-L. Chanin. 1993. Vol. 9: High Spectral Resolution Infrared Remote Sensing for Earth's Weather and Climate Studies. Edited by A. Chedin, M.T. Chahine and NA Scott. 1993. Vol. 10: Towards a Model of Ocean Biogeochemical Processes. Edited by G. T. Evans and M.J. R. Fasham. 1993. Vol. 11: Modelling Oceanic Climate Interactions. Edited byJ. Willebrand and D.L.T. Anderson. 1993. Vol. 12: Ice in the Climate System. Edited by W. Richard Peltier. 1993. Vol. 13: Atmospheric Methane: Sources, Sinks, and Role in Global Change. Edited by M. A. K. Khalil. 1993. Vol. 14: The Role of Regional Organizations in the Context of Climate Change. Edited by M. H. Glantz. 1993. Vol. 15: The Global Carbon Cycle. Edited by M. Heimann. 1993. Vol. 16: Interacting Stresses on Plants in a Changing Climate. Edited by M. B. Jackson and C. R. Black. 1993. Vol. 17: Carbon Cycling in the Glacial Ocean: Constraints on the Ocean's Role in Global Change. Edited by R. Zahn, T. F. Pedersen, M. A. Kaminski and L. Labeyrie. 1994. Vol. 18: Stratospheric Ozone DepletionlUV-B Radiation in the Biosphere. Edited by R. H. Biggs and M. E. B. Joyner. 1994. Vol. 19: Data Assimilation: Tools for Modelling the Ocean in a Global Change Perspective. Edited by P. O. Brasseur and J. Nihoul. 1994.

Vol. 20: Biodiversity, Temperate Ecosystems, and Global Change. Edited by T. J. B. Boyle and C. E. B. Boyle. 1994. Vol. 21: Low-Temperature Chemistry of the Atmosphere. Edited by G. K. Moortgat, A. J. Barnes, G. Le Bras and J. R. Sodeau. 1994. Vol. 22: Long-Term Climatic Variations - Data and Modelling. Edited by J.-C. Duplessy and M.-T. Spyridakis. 1994. Vol. 23: Soil Responses to Climate Change. Edited by M. D. A. Rounsevell and P. J. Loveland. 1994. Vol. 24: Remote Sensing and Global Climate Change. Edited by R.A Vaughan and A. P. Cracknell. 1994. Vol. 25: The Solar Engine and Its Influence on Terrestrial Atmosphere and Climate. Edited by E. Nesme-Ribes. 1994. Vol. 26: Global Precipitations and Climate Change. Edited by M. Desbois and F. Dssalmand. 1994. Vol. 27: Cenozoic Plants and Climates of the Arctic. Edited by M. C. Boulter and H. C. Fisher. 1994. Vol. 28: Evaluating and Monitoring the Health of Large-Scale Ecosystems. Edited by D. J. Rapport, C. L. Gaudet and P. Calow. 1995.