BERNARD TINKER together with JAN GOUDRIAAN, PAUL TENG, MIKE SWIFT, SUNE
LINDER, JOHN INGRAM AND SIEBE VAN DE GEIJN
Department of Plant Sciences, University of Oxford, UK
Overview of the GCTE and focus 3 structures
Effects of global change on major food crops (a component of activity 3.1)
Improved pastures and rangelands (a component of activity 3.1)
Effects of global change on multi-species cropping systems (activity 3.4)
Experimental studies on the relationship between plant-based diversity and complexity and system sustainability (task 3.4.1)
Modelling complex agricultural systems (task 3.4.2)
Long-term agricultural experiments and databases as a resource for global change research (task 3.4.3)
Effects of global change on managed forests (activity 3.5)
Changes in pests, diseases and weeds (activity 3.2)
Effects of global change on soils (activity 3.3)
Conclusion
References
The concept of global change is now well understood and to a large extent accepted amongst scientists, agencies and the informed public. It is usually also accepted that the driving forces for changes are: (i) change in atmospheric composition; (ii) climate change (which arises from the first); and (iii) land-use change (driven by both socio-economic factors and by climate change). A number of associated topics, such as ozone/UV-B, acid rain and other forms of pollution are sometimes included, but are not central to the discussion in this paper. Of these factors, change in atmospheric composition is well documented, and a reasonably good prediction of its future progress can be given. Climate change may be occurring, but it is still possible that what is being observed is a normal fluctuation (IPCC, 1992). Land-use change cannot yet, therefore, be associated with climate change in the global sense, but it is occurring very rapidly because of socio-economic driving forces.
The potential impacts of these three factors are agreed to be massive, and could well be catastrophic in some cases, while beneficial in others. However, the progress and impact of climate change is difficult to predict, atmospheric composition may be controlled to an as yet unknown extent by human action, and the progress of land-use change depends upon demographic, social and economic factors which cannot be predicted with accuracy over long periods. The details of the impacts are therefore extremely uncertain, especially for rainfall.
The International Geosphere-Biosphere Programme (IGBP) was set up, under the auspices of the International Council of Scientific Unions, to conduct research into these matters. Of the Core Projects of the IGBP, Global Change and Terrestrial Ecosystems (GCTE) (Steffen et al., 1992) has the most obvious and direct involvement with the land surface. Within GCTE, Focus 3 has the specific responsibility for agriculture, forestry and soils (Table 12.1). GCTE has responsibility for work on both impacts of global change, and also on feedbacks which further control the extent of global change. For example, the change and extent of different types of vegetation has important implications for the regional climate. GCTE's main responsibility is, however, for studies of impacts.
Table 12.1. Structure of GCTE Focus 3
|
Activity 3.1: |
Effects of Global Change on Key Agricultural Systems |
|
Task 3.1.1: |
Experiments on Key Crops with Changed Atmospheric Composition and Climate, on Different Soils |
|
Task 3.1.2: |
Modelling Growth of Key Crops Under Changed Atmospheric Composition and Climate |
|
Task 3.1.3: |
Global Change Impact on Pastures and Rangelands and Associated Animal Production |
|
Activity 3.2: |
Changes in Pests, Diseases and Weeds |
|
Task 3.2.1: |
Global Monitoring Network and Data Sets for Pests, Diseases and Weeds |
|
Task 3.2.2: |
Distributions, Dynamics and Abundance of Pests and Diseases Under Global Change |
|
Task 3.2.3: |
Weed Distribution, Dynamics and Abundance Under Global Change |
|
Activity 3.3: |
Effects of Global Change on Soils |
|
Task 3.3.1 |
Global Change Impact on Soil Organic Matter |
|
Task 3.3.2 |
Soil Degradation Under Global Change |
|
Task 3.3.3 |
Global Change and Soil Biology |
|
Activity 3 4: |
Effects of Global Change on Multi-species Agro-ecosystems |
|
Task 3.4.1 |
Experimental Studies on the Relationship between Plant Species Number and Function in Agricultural Systems |
|
Task 3.4.2 |
Modelling Complex Agricultural Systems |
|
Task 3.4.3 |
Long-term Agricultural Experimental Network |
|
Activity 3.5: |
Effects of Global Change on Managed Forests |
|
Task 3.5.1: |
Experimental and Observational Studies of Managed Forests |
|
Task 3.5.2: |
Modelling Global Change Impact on Function, Structure and Productive Capacity of Managed Forests |
At a series of international 'open' workshops, GCTE Focus 3 has considered carefully what research can and should be done, in a situation where the climatic predictions are so uncertain. Given this uncertainty on a worldwide scale, let alone on a site-specific basis, the research priorities have aimed at improved system understanding leading to more robust predictive models. Because of the global scale of the problem it has been essential to base the research programmes on international collaboration, so that the results have generic value. National programmes will subsequently fill out the more local details. The work undertaken should therefore be of a precautionary and preparative type, so that it simplifies and clarifies the tasks which must be undertaken as and when it becomes clear that climate change is occurring, and the extent and direction of the changes. This is one reason why systems analysis and the preparation of better predictive models occur so frequently in our programme.
It seems likely that when climate change becomes apparent, there will be a period of increasing turbulence within agriculture. Farmers and advisers are familiar with the possibility of weather variation in a single year or over a short run of years, as in many regional or local droughts. Two questions which they are always called upon to decide are, firstly whether these new conditions will hold in the following year, because they have to take immediate decisions on cropping; and secondly whether this change may last for several years, because they then may have to adjust their farming system. A long-term trend, possibly with greater annual variability, is however something with which farmers are not accustomed to cope.
Much will depend upon the degree of reliability of local climate predictions. If climate cannot in fact be predicted in detail at the local scale, the essential characteristic must be flexibility and rapid response. This may involve change of cultivar, change of agronomic variable such as sowing date, change to a different crop, or change to a completely different agricultural system, e.g., from arable to grassland. In some cases, it may mean abandonment of farming, as happened so often during the drought of the 1930s in the American Midwest (Worster, 1979). If the predictability of climate change on a local basis becomes better, then farmers can plan these changes in a more rational way. In either case, the best tool for guiding their response is likely to be a set of the best possible predictive crop models, tested over widely different climates, photoperiods and soil types.
IGBP has no specific funding, other than a limited amount of money for its own internal administrative operation. Its work proceeds in two main ways: (i) by the launching of major collaborative projects which directly address specific aspects of the GCTE Operational Plan (Steffen et al., 1992), and which gain national or international funding; (ii) by the coordination and gathering together of many projects (usually nationally funded) which have similar aims. The latter strategy aims to build a global picture by synthesizing what is usually locally based research. The total programme cannot therefore be a tightly structured and organized one, such as is possible where there is a large funding stream to determine the research to be done. We are to some extent guided by Lord Rutherford's statement 'We have no money, so we have to think!'. However, that thinking has produced a number of important initiatives, some already in being, some still at the planning stage, which we explain below.
The GCTE Core Project encompasses a very wide range of terrestrial biomes and scientific disciplines. Given this wide scope, the early planning phase of GCTE recognized the need for a highly structured programme, to ensure both that appropriate collaboration at the international scale is established and that important areas (either geographically or thematically) were not overlooked or duplicated. GCTE has thus been designed around four major Foci: ecosystem function; ecosystem structure and composition; global change impact on agriculture, forestry and soils; and ecological complexity.
GCTE is hierarchical in design, each Focus being divided into Activities, each of which are in turn subdivided into Tasks. This arrangement allows Tasks of a similar nature to be clustered and necessary collaboration at the Task level to be maximized. There are, however, numerous instances where a particular scientific topic spans Activities or indeed Foci. To minimize duplication of effort while maximizing the benefits of interdisciplinary study, inter-Foci and/or inter-Activity aspects have been identified and are strengthened as much as possible. Good examples of such collaboration are, respectively, the system response to elevated CO2 and the pests and weeds component within the GCTE Crop Networks; both are described later in this paper.
Within Focus 3, four major types of production system have been identified: monocrop agriculture, pasture and grazing systems, multi-species cropping (including agroforestry and rotational systems), and forestry. These have been grouped into three Activities:
1. Key agricultural systems. This includes a representative selection of monoculture food crops, together with improved pastures and rangelands; the emphasis in this Activity is on harvestable product, i.e., the harvested portion of the crop or animal.2. Multi-species agro-ecosystems. These are the norm in much of the world, be they spatially or temporally mixed. Modelling spatially complex agricultural systems is in its infancy, relative to modelling monocrop production or even rotation farming.
3. Managed forests. These encompass the spectrum from intensively managed plantation crops through more natural forests which are either selectively logged or put to some other use with socio-economic implications.
Two further Activities cross-cut these three major areas: one addresses the effect of global change on pests, diseases and weeds and the other addresses the effect of global change on soils. The full list of the Focus 3 Activities and their associated Tasks is given in Table 12.1, and their rationale and work plans are discussed below. (As GCTE planning has progressed, new Activities have been assigned the next sequential number. While this avoids the possibility of confusion brought about by renumbering Activities and Tasks, it must be stressed that the number does not imply importance.)
The need for within-GCTE collaboration has already been highlighted. There is however an equal need to develop appropriate collaboration with other international programmes, be they other IGBP Core Projects (notably IGAC 1, BAHC 2 and LUCC 3) or non-IGBP initiatives; GCTE's role is often one of international coordination and engendering collaboration, linking new initiatives and synthesizing results within a common, internationally agreed framework. This aspect will be especially necessary for linking to the human dimension, particularly apposite when dealing with ecosystems from which a 'harvest' is taken; global change impact on agriculture and forestry is more than just impact on biology - it is impact on food, wealth, development and people.
1 International Global Atmospheric Chemistry project
2 Biospheric Aspects of the Hydrological Cycle project
3 Land Use and Land Cover Change project
Anticipated changes in global rainfall and temperature patterns together with the established increase in atmospheric CO2 will affect the production of crops throughout the world. Of the many crops which might form the subject of a research programme in global change, GCTE selected a shortlist of six for initial studies: wheat, rice, potatoes, cassava, maize and groundnuts. The crops were selected both in recognition of their importance in global food production, and as representatives of a diverse range of crop 'functional types'. GCTE has also recently included sorghum, a major crop of marginal tropical regions.
GCTE is launching a series of Crop Networks, one for each of these crops. These worldwide research networks are designed to promote integrated experimental and modelling research to determine the interactive effect of global change factors on these key agronomic species. Each will be led by a small working group of modellers, experimentalists, and pests and disease specialists. The Networks will build on the many crop experiments already established and on models already developed. Existing models, however, are generally constructed and validated using data from one locality, so the identification of modelling approaches that yield generic capability is needed. The worldwide network approach will greatly facilitate the required cross-comparison of models with appropriate data sets, and both modellers and experimentalists are essential to the given Network's success. The GCTE research will lead to the development of crop models that are robust under a wide range of changed environmental conditions.
The Crop Networks are designed to maximize the understanding of global change impact on the cropping system. This means on the crops per se, and on the crops' pests, weeds and soils (key aspects which are discussed elsewhere in this paper) - i.e., on the whole agro-ecosystem. This approach requires the consideration of not only all components of the system, but also the interactive effects of global change drivers. This integrative approach makes GCTE uniquely innovative. Furthermore, by taking the systems approach to crop modelling, these Networks cover many aspects of GCTE, and provide a forum for the global change community to undertake and discuss collaborative modelling research.
The Networks also strongly promote international collaboration in experimentation. One of the key experimental techniques, of equal interest to Focus 1 and Focus 3, is Free Air CO2 Enrichment (FACE); FACE systems are now established in crops (wheat and cotton) and in pastures. (Methodological problems are encountered when using the technique in higher-stature systems, i.e., forests, but as the technique is so valuable, these are being tackled by various groups around the world.) The real, and unique value of FACE is that the site's microclimate is not disturbed; in addition, pests and diseases are not excluded from, or trapped in, the system. Another main advantage is size; being some 300-400 m2 in area, subtreatments (e.g., nitrogen or water) can be incorporated and, even so, the experiment is not seriously disrupted by the destructive sampling of soils and vegetation from all plots. The question of whether Open-Top Chambers (OTCs) can be 'calibrated' against FACE is important, as OTCs are being widely used. The issue is currently being researched by GCTE collaborators.
GCTE Crop Networks started with wheat, because models are already well developed and its physiology is well understood. An international workshop (Ingram, 1992) launched the GCTE Wheat Network, and initiated the evaluation of current wheat production models using a minimum set of standard model initialization parameters. A subsequent workshop (Goudriaan et al., 1993) undertook a highly refined analysis of the Network's wheat production models, using a small number of agreed data sets. A summary of this most recent workshop's design and results is given below, to illustrate the effectiveness of working in this truly global, collaborative way.
The workshop (and preparative exercises) covered three main areas, and relied upon a series of pre-circulated datasets:
THE HIGHLY REFINED SENSITIVITY ANALYSIS (EXERCISE 1)
This required models to be run with common baseline data, but with artificially perturbed temperature (from -6 °C to +6 °C, at 1 °C increments) and CO2 concentration (seven levels). Two sets of 30-year mean meteorological data were distributed to modellers in advance of the workshop, one set being for a winter wheat growing season (European data), the other being for spring wheat (North American data). Modellers were asked to fit genetic coefficients which resulted in standard phenological development dates for the current (zero °C offset) case. Output data included, inter alia, phenological development, leaf area index and grain yield.
Contrary to expectations, the variance of the models' results was unexpectedly large, even though the dates of emergence, of anthesis and of maturity had been prescribed (see Figure 12.1). Models, however, contain many feedbacks, making it difficult to trace the precise causes of deviation. (A good example is the major feedback route occurring through growth of leaf area - once leaf area in early growth is overestimated, its high value will tend to be amplified though a higher rate of photosynthesis.) For this reason, it was decided to 'force' the models in line by imposing the same prescribed time course of leaf area index on all models. This artificial move enabled exclusion of morphological model differences (specific leaf area, leaf area formation, tillering, leaf appearance), leaving only functional model differences (light interception, photosynthesis and respiration). The results of this forcing exercise are presented in Figure 12.2, from which it can be seen that the variance was only marginally reduced.
The models included in this exercise differ greatly in complexity, ranging from one with only 90 lines of code in GWBASIC to one requiring a Cray computer to run. There was no apparent relationship between model output values and model complexity.
Figure 12.1. Scatter plot of simulated yield against simulated
biomass for the North American site 
and for the European
site 
Figure 12.2. Scatter plot of simulated yield, as in Figure 12.1, but with the same imposed time course for leaf area index in all models
REAL DATA SETS AND MODEL ASSESSMENT (EXERCISE 2)
Together with the 30-year mean meteorological data for Exercise 1, modellers were sent meteorological, soil, crop management and phenology data for crops at the European (winter wheat) and North American (spring wheat) sites. Modellers were asked to bring their output data, including grain yield, for the cases provided. This exercise was designed to highlight the strengths and limitations of each model, and was not intended to be a 'horse race'; the models differ in objective and design, in operational scale and in geographic, climatic and phenological aspects.
The full results from Exercises 1 and 2 will be reported and discussed in two multi-authored scientific papers. One will discuss the sensitivity analysis, and the other will analyse the precise backgrounds of the model differences. It must be stressed, however, that the workshop did not aim to tell which model does the best job; rather, it set out to improve understanding of why the models behave as they did, and which components of the models were critical.
COMBINING NETWORK DATA SETS WITH NETWORK MODELS (EXERCISE 3)
This exercise aimed to promote the Network's data/model dialogue by using the Network's diverse experimental data sets for further model validation and comparison. The methodology needed to run each model for each available data set was established. The intention of the exercise was to set the scene for Network efforts that will identify which combinations of model and data set are most likely to improve understanding and to develop interesting research questions.
This 'workshop series' approach has proved a very productive way of working, and the networks for the other crops will be similarly designed. In addition to workshops, Network activities include the distribution of collated metadata sets to all formal collaborators and other interested parties. These metadata describe the nature of the Network's data sets and models; they do not list the data or model code per se. These metadata compilations are published in GCTE Report No. 2, and will be regularly updated. Formal Network collaborators share actual data and model code according to an agreed code of conduct; the IBSNAT data format was chosen and is now the GCTE standard for all the GCTE Crop Networks, although the conversion of all data sets to this standard is not yet complete; a GCTE Networks' Officer position has been funded by the Dutch Global Change Programme (NOP) to assist with this important task.
The GCTE Rice Network was launched with an international workshop held at IRRI in March 1994 (Ingrain, 1994a), and network activities will be planned in detail at a workshop in 1995, co-funded by FAO. Similar Networks for potato and cassava are in advanced planning, and initial Network membership will be established in 1995.
These GCTE Networks represent a major step forward in international, collaborative agronomic science. They add value to the large investments in crop modelling that are being made in many countries, and being dynamic in both scientific content and participation, and are very much open to further scientists.
Uncertainties surrounding the possible effect of global change on food production are not limited to crops. Therefore, to complement the Focus 3 work in this area, an important Task in Activity 3.1 (Task 3.1.3) deals with the potential impact of global change on livestock production. Global change will have far-reaching consequences for dairy, meat and wool production, though mainly via impacts on grass and range productivity; direct impacts on the animal itself are not expected to be great. This concept underpins GCTE's work plan in this area, as manifested in the Task's title: 'To predict the effects of global change on pasture and range composition and production, and the consequent effects on livestock production'.
Grasslands and rangelands form a continuum. At one end lie improved pastures, i.e., areas where the original vegetation has been removed and grassland is maintained by regular grazing and/or mowing, and fertilizer and/or herbicide applications. This definition obviously encompasses intensively managed grasslands (e.g., in the Netherlands), but it also includes the areas of sown grassland which are used less intensively for livestock production. Improved pastures can therefore range from simple monocultures to complex mixtures of species.
At the other end of the continuum lie rangelands. These occur in areas where either climate or soil factors are so limiting that intensive primary production is not viable. While rangelands may contain introduced species, this end of the continuum is characterized by systems where (i) the pasture has a substantial component of native species; (ii) the pasture is composed of more than three or four species; and (iii) there is little or no addition of fertilizer or irrigation.
Improved pastures are invariably restricted to commercial operations, whereas rangelands are used by commercial and subsistence farmers. Vegetation productivity and animal growth models will be formulated for each type and use, and ultimately a generic model will be developed that could be applied to intermediate types.
The improved pasture and rangelands components share two general objectives. These are:
(i) to predict the effects of changes in atmospheric composition and climate on forage production, and the consequential impact on livestock production at the patch and landscape scales; and(ii) to predict the feedback effects to the atmosphere-climate system of changes in management induced by changes in atmospheric composition and climate.
Improved pastures and rangelands of course have specific individual objectives in addition to these. Systems with significant management intervention via, for instance, forage and animal germplasm improvement or irrigation, will provide opportunities for adaptation to meet changing environmental conditions; research is needed to prepare adaptive strategies to cope with the interactive effects of changes in such conditions. On the other hand, in more extensive systems (where adaptation is hard to implement due to the scale of operation), the research strategy is one of improving the ability to predict the effects of interactive changes in intrinsic and extrinsic forces.
The GCTE approach is to build on points of commonalty between the types of systems, rather than highlight their differences. This helps link to other areas of GCTE interested in, for instance, rangelands - Focus 2 includes an Activity dealing with patch-scale modelling, an important aspect when developing decision support systems for range managers.
GCTE also has an important role in linking natural and social sciences in this area; the rangelands component of this Task is currently leading Focus 3 's links to the Human Dimensions Programme of global environmental change (HDP), although this will also become a major aspect of the developing research programme of Activity 3.4 'Effects of global change on multi-species agroecosystems' (see below). The essential point of contact is the proposed joint IGBP-HDP Core Project, Land Use/Cover Change (LUCC). The Task ultimately aims to predict the interactions between the biophysical driving factors and the extrinsic driving factors, the focus of LUCC. The socio-economic understanding of contrasting systems, on a region-by-region basis, will have immediate regional value for land-use planning. (Further details about this GCTE research programme can be found in Stafford Smith and Campbell, 1994.)
GCTE Focus 3 launched its operational phase with research on monocropped systems, and substantial progress has been reported above. GCTE is now turning its attention to cropping systems designed around mixtures of species, and detailed planning is well under way, launched with an international workshop in Kenya in 1994.
Global change will impact on all agriculture systems to some extent. The effects of global change may however first be observed in multi-species agro-ecosystems, because the competitive advantage of the different species in the system may change. The resulting overall effect may be more pronounced than in monocropped systems. There are also two further, powerful reasons for highlighting multi-species agricultural systems in a programme of research on global change. The first is socio-economic; a major proportion of the world's population depend on such systems for their food and livelihood. The second is ecological; study of such systems can give insight into the relationship between complexity, stability, diversity and an ecosystem's capacity to respond to change.
In addition to spatially mixed systems, temporally mixed (i.e., rotation) agriculture offers numerous, well-established benefits, and it has long been a principal agricultural management technique. As with other aspects of this 'multi-species' component of GCTE research, the consequences of global change on rotational systems will be undertaken in collaboration with other GCTE Foci, Activities and Tasks, and with other international programmes; the residual effects on soil organic matter and nutrient dynamics, or the pest, disease and weeds management are obvious examples where maximum collaboration is needed.
Standard classifications recognize about six or eight major types of cultivation system. They vary greatly in the diversity of the plant and livestock components from shifting and recurrent fallow-based cultivation through permanent mixed and rotational crop cultivation to annual and perennial monocrops. Livestock production similarly ranges from pastoral nomadism utilizing natural savannas to intensive ranching on species-regulated grasslands. Furthermore, livestock and arable farming may also be combined in a range of mixed farming systems (Frissel, 1978; Grigg, 1974; Ruthenbag, 1980).
The major advances in world sufficiency in food production have come from intensive monocultures of improved varieties of a small range of cereals. These crops are often grown in rotation, and receive high inputs of manufactured fertilizers and energy-subsidized management of soil, water, pests and weeds. This type of agriculture is characteristic of the industrialized regions of the northern hemisphere and has also made considerable impact in a number of countries of the developing world by means of the so-called 'Green Revolution'. Most of the world's farmers, however, still depend for their food and income on multi-species systems of one type or other. In comparison with monocultures these systems are less researched, more complex to model and there are more options to consider. Their behaviour and productivity under global change is therefore less easy to predict.
Nonetheless, in recent years cropping systems of 'intermediate' complexity, such as those based on intercropping of a few plant species including those incorporating trees, have replaced intensive monocultures as the target for much of modern agricultural research, particularly in the tropical regions. This change of strategy is largely based on the hypothesis that multi-species systems are more sustainable and more environmentally conservative than monocrops. Research with intercropping (e.g., Francis, 1986) or alley cropping (e.g., Kang et al., 1990) has shown that these two- or three-species systems may indeed gain comparable or even higher returns as intensive monocrops, although results are not conclusive. There have also been many claims for the greater yield stability of intercropping systems (e.g., Rao and Willey, 1980) although Vandermeer and Schultz (1990) could find no theoretical justification as to why this should necessarily be so.
The GCTE Activity dealing with multi-species systems under global change will build as much as possible on this existing research. There is, however, a clear need to increase basic understanding of the interactions occurring in temporally and spatially mixed crops, which will of course be of great benefit in the short term for improving yield and yield stability, and for longer-term global change studies.
The Activity has three major components, to be managed as individual, but mutually dependent Tasks:
As outlined above there is a considerable but inconclusive body of evidence that cropping systems with more than one species of plant (e.g., intercropping or agroforestry) have both greater productivity and greater resilience in the face of disturbance than mono-cultural systems. These effects may be enhanced when the system is structurally and chemically complex as well as species diverse. This could be due, for example, to the plants forming complex microenvironments through multiple canopy layers, and providing a wide range of allelochemicals and other substances which influence the function of other members of the community. Research will rigorously evaluate the evidence for this hypothesis, particularly in the context of global change. The research programme will be developed in close collaboration with GCTE's Focus 4, 'Ecological Complexity'.
The influence of the plant system on the herbivore and decomposer subsystems will be important components of the studies. The below-ground community is essential to the maintenance of ecosystem function through its role in soil fertility by means of nutrient cycles regulation, control of soil organic matter dynamics and modification of soil structure. This is of particular significance to agricultural systems where the soil subsystem is constantly disturbed by management practices. The focus in this research will be on the role in system productivity and sustainability of key groups of the fauna and microflora such as earthworms, termites, nematodes, nitrogen-fixers and mycorrhiza. This is an area where improved system understanding will be particularly beneficial for improving current productivity and reducing yield variation; it is also the essential prerequisite to building improved predictive models, robust to global change.
The time scale of change relevant to sustainability and global change is much longer than that of a realistic experimental programme and also extends beyond the usual planning horizon of the farmer. As stated elsewhere in this paper, the most rigorous way of extrapolating predictions of change beyond the short term is by means of simulation models. Whilst there is now a comparatively effective (but by no means adequate) suite of such models for individual crops such as wheat, maize or rice, the modelling of more complex systems such as intercrops and agroforestry (with the inclusion of competitive and/or synergistic effects), is in its infancy. More rapid advancement will come from an interactive effort between modellers tackling the issues of simulating complex cropping systems than if they operate in isolation. Further strength would be gained moreover from linkages with the monocrop and ecological modelling activities also being promoted in GCTE.
A further dimension to the modelling activities comes from the need to assess the economic and social implications of effects on agro-ecosystem performance, including modifications in agricultural practice, resulting from the impact of global change. GCTE scientists in Activity 3.4 will collaborate with social scientists from the Human Dimensions Programme and the CGIAR to develop economic-ecological models and/or decision support systems that link the biological and socio-economic dimensions of agroecosystem change.
The only unequivocal way of determining the relationship between external factors and system response over time is by means of long-term experiments. The data from such monitoring can also be used to validate the predictions of simulation models. A large range of such experiments were initiated at various times over the last century, many of them in developing countries. Only a minority are still in existence and the data from many have never been published in the open literature. The need for publication of data, rehabilitation of selected experiments and establishment of experiments in new sites was discussed at a conference at Rothamsted Experimental Station, UK (Swift et al., 1995). This discussion provided a strong foundation for developing this Task; this will be done in collaboration with the Task dealing with soil organic matter (see below), where long-term data sets are crucial for model development.
Forests cover more than one-third of the land surface of the earth and are almost equally divided between the temperate and the tropical regions. While the intensity of management and use of the forests varies both within and between regions, forest and other wooded land constitutes the largest component of current and future land use in terms of area, and plays an important role in the global carbon balance. Global change is having, and will continue to have, a major impact on forest cover.
From a socio-economic point of view the objective of securing long-term wood supply to timber and pulp industries is an obvious imperative. Biomass production in less intensively managed forest ecosystems may, however, be equally important since it provides fuel, fodder and other utilities for a large proportion of the world's population. Hence, the term 'managed forests' is used here in a broad sense to include forest ecosystems managed and used for purposes other than industrial production. Managed forests thus cover the range from plantations (where management includes all silvicultural activities), through natural forest managed (in the real sense of the word) for wood production, to natural forests simply exploited for timber and other products.
Forests are long-lived communities; the rapidly changing atmospheric CO2 concentrations and the rising temperatures associated with climate change are likely to have significant impacts not only on the forests of the future but also on forests already in the ground. The current changes in forest cover brought about by the socio-economic drivers of global change will be compounded by changes in these biophysical drivers. To provide the information needed to predict the interactive effects of global change on forests, and the knowledge required to provide the basis for appropriate management programmes, we need focused, coordinated forest research programmes throughout the world.
RESEARCH ON FOREST ECOSYSTEMS WITHIN GCTE
Studies of forest ecosystems occur in all four GCTE Foci, in other IGBP Core Projects and in many other international and national research programmes. There is, however, one Activity specially designated to study the 'Impacts of global change on managed forests', which lies within Focus 3. The primary aim of this Activity is to understand and evaluate potential effects of global change on future structure, biomass production, and yield of managed forests; and hence to identify resource management strategies for sustainable forestry under changed climatic conditions. This will require the prediction of both short-term forest responses to altered climate, disturbance and silvicultural practices, and effects on long-term sustainable site productivity and biodiversity; clearly close cooperation among all GCTE Foci is needed. The role of managed forests as carbon sinks or sources should be evaluated as well as the possibility of increasing carbon sequestering in forest ecosystems by means of silvicultural practices. This long-term emphasis will require conceptual advances in capacity to characterize ecosystem sustainability and resilience in response to altered rates of input, loss and cycling of carbon, water and mineral nutrients.
The GCTE Managed Forests Implementation Plan has been recently published jointly with the International Union of Forestry Research Organisations (IUFRO) (Landsberg et al., 1995). This document, which functions in support of the GCTE Operational Plan, provides a framework for planning and implementing scientific research concerned with forests. Emphasis has been placed on ensuring consistency in terms of concepts, procedures and data recording. These will lead to results that can be compared and used to test and validate models, and assess the impacts of global change on forest growth and production.
The research strategy underlying this plan is to sample a representative range of forest types, growing under different conditions in different parts of the world. Within these the establishment of identical, or very similar, experiments will allow evaluation of the variation of, and constraints on, the productivity and biological diversity of forests. Networks analogous to those coordinated in other Focus 3 Activities (e.g., the GCTE Wheat Network, discussed above) will be established. As more projects addressing suitably similar goals, or using similar techniques, 'come on-line', they will be encouraged to join the most appropriate network or other GCTE structure. This will maximize collaborative effort, and help synthesize results from around the world.
To ensure rapid progress in information and understanding, an agreed, standardized primary data recording system will be developed, which would be available for exchange and comparison. The use of Geographical Information Systems (GIS) with models, to deal with the heterogeneity of large areas of forest, and as a framework for the recording and analysis of remotely sensed information, is strongly advocated.
This Activity is designed around two closely related Tasks, one dealing with experimental and observational studies (Task 3.5.1), and the other with modelling (Task 3.5.2). While the latter Task will not be geographically specific, the former will be split between boreal and temperate, and tropical and semi-arid regions. A further Task specifically addressing plantation forestry is being considered, and would probably be split into tropical and temperate components.
STUDY SITES AND TRANSECTS
For assessing the impact of environmental conditions, and hence global change, on the growth and performance of forests it is important that study sites be located (i) in areas where significant change is expected; (ii) where there is reason to believe that the forests are likely to be susceptible to change; and (iii) where the forest ecosystems are economically important. Based on these criteria, the following systems have initial priority:
· boreal forests, which contain much of the world's available softwood and where considerable warming has been predicted, though physiological responses to CO2 enhancement may be limited by low temperature and infertile soils;· northern hemisphere temperate coniferous and mixed coniferous-deciduous forests which form the basis for most of the present timber and pulp industry;
· softwood and eucalypt plantations in semi-arid and subtropical regions where changes in precipitation may have considerable effects;
· tropical forests, where warm conditions may lead to a large CO2 fertilization effect.
EXPERIMENTAL TREATMENTS, MEASUREMENTS AND OBSERVATIONS
Appropriate models should provide a framework for organizing and focusing measurements and experiments, and modelling should therefore be done a priori rather than a posteriori. It is not the primary role of models to 'integrate' the observations made in experiments; parameters can generally be adjusted to obtain a good 'fit' to observation without learning about the system or the model. Models also provide the only means of evaluating the likely consequences of various management actions, particularly in relation to the uncertainties arising from global change. It should be noted that regression-based models (which include most of the models developed as part of traditional forest research and management) are unable to predict the effects on growth of changing climatic conditions; they are derived from historical observations, without a mechanistic basis. There is therefore an urgent need to link process-based models to tree population models and hence to produce estimates of the harvestable product, which is the information that forest managers need.
Experimental work should be aimed not only at providing empirical information, but also at understanding the physiological control mechanisms and key processes underlying forest response to environmental factors, at both the tree and stand levels. The basic minimum requirement, at each experimental or observational site, is a series of baseline measurements and observations (Level I), preferably carried out over a long period. Where resources exist these baseline observations should be supplemented by more detailed studies on community dynamics (Level II) and, at some sites, measurements of the physiological processes governing forest growth and productivity (Level III).
Level I includes the normal measurements that can be made with minimal infrastructure but which provide the essential baseline information about the state of forests and their long-term growth patterns. Such measurements accumulated over long periods at many well-characterized sites will provide an extremely valuable database from which considerable information about forest growth and performance, and the effects of weather and climate, can be obtained.
Level II measurements are aimed at providing information about the factors that cause changes in species composition. Such studies are important for all natural stands since species composition, and hence biodiversity, are likely to become significantly affected by many aspects of global change.
Level III measurements are concerned with the physiological processes underlying and driving the growth and production of trees. These studies should be performed on specific sites, when experimental treatments increase water availability and mineral nutrients similar to field experiments under way in Australia, Sweden and the United States. Where possible, experiments should also incorporate elevated CO2 and soil warming treatments. The sites should incorporate the same experimental treatments, and should follow the same protocols, as far as possible.
DATABASES AND GEOGRAPHICAL INFORMATION SYSTEMS (GISs)
The long-term objective must be to develop databases for each forest type which provide accurate (geo)references for site location, long-term climatic data, properly documented site descriptions and standardized records of research data and observations. Such records, over time and for a number of sites/transects, will allow the refinement and testing of a complete range of models as well as providing invaluable empirical information about the productivity and growth patterns of the world's forests. This will be a major GCTE product, also invaluable for monitoring forest cover change.
The technology has great potential: the expenditure of considerable resources on accurate documentation (in GIS) of forest biomass, species composition and a range of other parameters, is justified; in fact it is arguable that this should have a higher priority for funding than a great deal of other research, since without good information about forest resources, and their state, detailed information about processes is of limited value for policy and management purposes. Given these data in GIS, in compatible formats, for many regions of the world, it will be possible to compare productivity and evaluate the consequences of global change at sites around the world.
One of the first noticeable effects of global change may be changes in agronomic pests and weeds ('pests' here refers to insect pests and microbial pathogens). This is because global change will potentially affect the pest/weed-host relationship in one (or more) of three ways: by affecting the pest/weed population; by affecting the host population; and by affecting the pest/weed-host interaction. It is hypothesized that the net effect will manifest in one of several ways: (i) pests/weeds currently of minor significance may become key thereby causing serious losses; (ii) the distribution and intensity of current key pests/weeds may be affected, leading to changed effects on yield and also on mitigation techniques such as pesticides and integrated pest management; and (iii) the competitive abilities in weed-plant interactions may be affected through changes in ecophysiology. The goal of this Activity is therefore to determine, through worldwide networks for research, data sharing and modelling, the potential impact of global change on pest and weed distribution and dynamics; and from this, the socio-economic cost of such impact. A conceptual framework for generating specific outputs (assessments of impact) is presented in Figure 12.3.
All pest effects should be determined in the context of the crop model(s) that are being used to estimate global change effects on crops under pest-free environments. This Activity is therefore being developed hand-in-hand with GCTE's Activity 3.1 (Effects of global change on key agricultural systems). The data sets used for crop modelling and pest modelling will ultimately be fully integrated, although the interpretation of pest effects should be done only when the effects of global change on crops are established. This will allow tangible targets and outputs to be generated for an assessment of impact. Table 12.2 lists crop-pest combinations which will receive initial emphasis in the GCTE research programme.
GLOBAL MONITORING NETWORK AND DATA SETS FOR PESTS, DISEASES AND WEEDS (TASK 3.2.1)
This Task aims to establish appropriate data sets in support of the experimental and modelling work of Tasks 3.2.2 and 3.2.3, which are discussed below. GCTE will not attempt to include all pests, diseases and weeds (clearly a vast undertaking), but will concentrate initially on those pests, diseases and weeds of the six priority crops in Activity 3.1. This work will be linked to several other initiatives on global data networks, such as (i) a US initiative to develop a global database on crop losses; (ii) a USDA-ARS initiative to develop a global database on pest distribution and intensities; (iii) the former IBSNAT (International Benchmark Sites Network for Agrotechnology Transfer) Project Office at the University of Hawaii, with its minimum data sets (MDS) for crop and pest models; and (iv) the University of Hannover's global database on pest-induced crop losses for the major food crops.
Figure 12.3. Conceptual framework for generating specific outputs from GCTE Activity 3.2
Table 12.2 Crop-pest combinations for assessment of global change impact
|
Agronomic species |
Pest |
Domain |
|
Wheat |
Aphids |
Europe, North America |
|
Leaf rust |
Europe, Americas, Oceania |
|
|
Septoria tritici |
Europe, N America |
|
|
(To be refined after consultation with CIMMYT) |
|
|
|
Rice |
Brown planthopper |
Asia |
|
Blast |
Asia |
|
|
Weeds |
Asia |
|
|
(To be refined after consultation with IRRI) |
|
|
|
Cassava |
Mealy bug |
West Africa |
|
Mosaic virus |
West Africa |
|
|
(To be refined after consultation with IITA) |
|
|
|
Maize |
To be determined after consultation with CIMMYT |
|
|
Groundnut |
Leaf spots |
N America, West Africa |
|
(To be refined after consultation with IITA) |
|
|
|
Potato |
Colorado P. Beetle |
N America |
|
Late blight |
Americas, Europe |
|
|
(To be refined after consultation with CIP) |
|
Note: The term 'Pest' is used here to include insects, pathogens and weeds.
DISTRIBUTIONS, DYNAMICS AND ABUNDANCE OF PESTS AND DISEASES UNDER GLOBAL CHANGE (TASK 3.2.2)
There is already much literature on global change-pest interactions, although many of these are conflicting. Table 12.3 (from Manning and von Tiedemann, 1993) shows one interpretation of the literature. GCTE will initially aim towards making an objective assessment of such literature to arrive at a common, accepted statement on global change effects on selected pests. Pest-crop combinations to be initially researched are shown in Table 12.2. Research will not, however, be restrictive in the number of combinations and GCTE welcomes further suggestions. Experimental and modelling studies will be developed in close collaboration with the Crop Networks, and representatives from Activity 3.2 are included in the Crop Networks' Working Groups.
Table 12.3. Effect of selected global change parameters on pathogens (from Manning and von Tiedemann. 1993)
|
GC factor |
Type (it pathogen |
No. of reports |
Nature of impact |
|
Elevated CO2 |
Bacteria |
0 |
- |
|
Necrotrophic fungi |
9 |
4(+), 4(0), 1(-) |
|
|
Biotrophic fungi |
7 |
6(+), 0(0), 1(-) |
|
|
UV-B |
Bacteria |
0 |
|
|
Necrotrophic fungi; |
13 |
9(+), 2(0), 2(-) |
|
|
Biotrophic fungi |
3 |
3(-) |
|
|
Ozone |
Bacteria |
4 |
0(+), 1(0), 3(-) |
|
Necrotrophic fungi |
25 |
16(+), 6(0), 3(-) |
|
|
Biotrophic fungi |
14 |
6(+), 2(0), 6(-) |
WEED DISTRIBUTION. DYNAMICS AND ABUNDANCE UNDER GLOBAL CHANGE (TASK 3.2.3)
Changing environmental conditions will alter the competitive advantage of one species over another in a given system. This will potentially affect ecosystem structure and composition, and has major implications not only for agricultural systems (where farmers strive to control weeds), but also for less managed systems (where invasive species may significantly alter the species composition). While the latter aspect is primarily of interest to Focus 2. the former clearly concerns focus 3.
Within Focus 3. a two-track approach is being planned. One component will concentrate on those weeds currently important for the relatively narrow range of crop species covered by Activity 3.1. The other will address weeds in a broader sense, and will try to determine what will make an otherwise benign species an economically important one under global change: obviously individual species will be hard to spot a priori, but the improved understanding of what triggers a shift in species composition (the Focus 2 aspect) will he a valuable contribution.
An international workshop in May 1995 (sponsored by the Australian Government) will consider the effects of global change on the nation's pests and diseases. and GCTE will sponsor a symposium on 'Assessment of Global Change Effects on Pests' during the XII International Plant Protection Congress, to be held in June 1995 in the Netherlands. Meanwhile, the current international effort on the Activity's topics are being documented, and will comprise three parts: (i) a list of long-term data sets for weather, global change factors, crops and pests (part of Task 3.2.1): (ii) a list of scientists working on global change-pest effects estimation for the six crops (part of Task 3.2.2): and (iii) a list of scientists working on global change-weeds effects (part of Task 3.2.3).
Soil science forms an obvious cross-cutting issue in GCTE. and it will he dealt with in all the Foci. as is necessary for their research. However, a number of research issues are specific to soils as such. and for this reason Activity 3.3 deals with soil science,
Soil properties vary from the transitory and rapidly variable, such as nitrate content. to the virtually permanent, such as texture. The number of properties that will be altered directly by changes in temperature, rainfall or CO2 concentration is fairly small, though a few soils are morphologically poised, and may alter rather rapidly (Sombroek, 1990). Changes in climate will affect soils in a terrain, in the sense that their erosion potential will alter with rainfall, plant cover and cultivation: and changes in atmospheric CO2 concentration may lead to changed soil organic matter quantity and type. via its impact on vegetation. The topics listed as Tasks below are those where global change is likely to have a direct or indirect impact on soils, and where this impact has important practical consequences.
In almost any activity concerning soil science, there will he a need for spatially referenced information. This is not a direct task of GCTE. but a geo-referenced World Soil Database is being assembled by a collaborative project involving FAO, ISRIC and USDA. with coordination being provided by the IGBP Data and Information System (IGBP-DIS): GCTE strongly supports this initiative.
GLOBAL CHANGE IMPACT ON SOIL ORGANIC MATTER (TASK 3.3.1)
Soil organic matter (SOM) is probably the single most important soil variable, and its level and properties contribute to both structure and fertility. Increased temperature and altered water status may change organic matter, but these effects can largely he modelled already. The potential effects of enhanced CO2 in changing the root/shoot ratio of plants, root exudation and the chemical composition of plant tissue requires much more research, but could be very important for SOM dynamics. SOM models will need to be altered accordingly, for different types of vegetation and climates. This Task therefore aims to determine the impacts of global change, as expressed at the plant physiological, vegetation and ecosystem levels, on SOM dynamics.
Most of these subjects are of course already being researched heavily, especially given that global change in the sense of land-use change is of course the most important agent of loss of SOM at present (see Task 3.3.21. GCTE will add to and help to coordinate this work. In particular, it will do its best to ensure that all major enhanced-CO2 experiments - particularly in EACH experiments - include studies on soil organic matter. GCTE is establishing a worldwide network to compare existing models under widely varying conditions of soil climate and vegetation, and to validate the models against the world's major long-term trials data sets.
GCTE has formed particular links with two international programmes. 'Alternatives to Slash and Bum' and the Tropical Soil Biology and Fertility Programme (Swift. 1991). because both focus, in different ways. upon the vegetation-soil-biota-SOM system. Both these are based in the tropics, but there are also important questions about the behaviour of SOM and related materials in higher latitudes during climate and land-use change.
SOIL DEGRADATION UNDER GLOBAL CHANGE (TASK 3.3.2)
Soil degradation is a wide-ranging and emotive term. GCTE has to focus, and it is doing so, on enhanced water erosion in the humid tropics, and wind erosion in semi-arid regions. Land-use change, or land cover change caused by climatic change, are the likely causes. The Task's objective, therefore, is to develop the capability to predict soil degradation by erosion caused by interactive changes in land use and climate. GCTE has established a worldwide network of erosion scientists (Ingram, 1994b), including modellers, experimentalists and scientists coordinating long-term monitoring studies (essential for validation erosion models). A detailed meeting held in early 1995 (sponsored by US EPA and USDA) planned in detail a model comparison workshop to be held in 1995, supported by NATO; the meeting also planned a major field campaign to be launched in West Africa in 1996. GCTE will endeavour to undertake genuinely strategic work, and to avoid site-specific studies. To ensure that research is in progress in the most-at-risk areas, the work will be conducted closely with bodies already involved in this area, such as ISRIC and FAO.
GLOBAL CHANGE AND SOIL BIOLOGY (TASK 3.3.3)
Originally this Task related to the production of greenhouse gases only. It was gradually perceived that its scope should be widened, because the behaviour of soil biota will be so critical in many global change situations, such as litter breakdown or new soil-borne diseases. The general questions in ecology will also apply to soil biota - speed of migration, invasion of new areas, speed of adaptation to new climates, and the classification of functional types. Soil microfauna and symbiotic organisms may be particularly interesting. It is absolutely essential, however, that the GCTE emphasis shall be on the processes mediated by soil biota, rather than on counting of species.
Work will be conducted in close collaboration with GCTE Focus 4 (Ecological Complexity) and with the IGBP Core Project 'International Global Atmospheric Chemistry' (IGAC).
The work outlined here is only one part of GCTE, i.e., Focus 3. There are strong links and contacts between this work, essentially designed for production, and the work dealing with the less managed ecosystems of the world. Cross-linking activities, such as the effect of CO2 on both wild and managed plants, or the importance of changes in soil and soil organic matter, are established throughout. Focus 2, dealing with structures of ecosystems, has great relevance for mixed ecosystems and landscapes, where some land is agricultural and some is wild. In general, it has much to contribute towards the understanding and prediction of land use. GCTE is fully aware that it needs collaboration with the socio-economic approach in the Human Dimensions Programme, which stands in parallel to the IGBP. Furthermore, it will need the involvement of scientists from all aspects of the global research community, be they from national systems (industrialized and developing nations), the international agricultural research centres, UN organizations and programmes, or other bodies.
Focus 3 is very committed to the idea that its work should be useful. With the possibility, indeed the likelihood, of global change occurring, it is sensible to carry out preparatory work for the amelioration which will have to be undertaken. This falls squarely under the tasks of the IPCC Working Groups 2 and 3. Focus 3 cannot give exact prescriptions, because the future is too uncertain, but we believe that we can clear some of the ground, so that better decisions can be taken when they are needed.
Francis, C.A. 1986. Multiple Cropping Systems. Macmillan, New York.
Frissel, M.J. (ed.). 1978. Cycling of mineral nutrients in agricultural ecosystems. Developments in Agricultural and Managed-Forest Ecology 3. 355 p.
Goudriaan, J., van de Geijn, S.C. and Ingram, J.S.I. 1993. Report of the GCTE Wheat Modelling and Experimental Data Comparison Workshop, November 1993. GCTE Focus 3 Office, Oxford.
Grigg, D.S. 1974. The Agricultural Systems of the World: An Introductory Approach. Cambridge University Press, Cambridge.
Ingram, J.S.I. 1992. Report of the GCTE Focus 3 Meeting: Effects of Global Change on the Wheat Ecosystem., July 1992. GCTE Focus 3 Office, Oxford.
Ingram, J.S.I. 1994a. Report of the GCTE Rice Ecosystems Workshop, March 1994. GCTE Focus 3 Office, Oxford.
Ingram, J.S.I. 1994b. Report of the GCTE Workshop Soil Erosion under Global Change, Paris, March 1994. GCTE Focus 3 Office, Oxford.
IPCC. 1992. Climate Change 1992 - The Supplementary Report to the IPCC Scientific Assessment. J.T. Houghton, B.A. Callander and S.K. Varney (eds.). Cambridge University Press, Cambridge.
Kang, B.T., Reynolds, L. and Attah-Krah, A.N. 1990. Alley farming. Advances in Agronomy 43: 315-359.
Landsberg, J.J., Under, S. and McMurtrie, R.E. 1995. GCTE Activity 3.5: Effects of Global Change on Managed Forests - Implementation Plan. GCTE Report No 41 IUFRO Occasional Paper No 1.
Manning, W.J. and von Tiedemann, A. 1993. Global climate change: potential impacts on plant-pathogen interactions. Paper presented at symposium on Global Climate Change: Implications in Plant Pathology, Sixth International Congress of Plant Pathology, 28 July-6 August 1993, Montreal, Canada. Paper S 18.2, Abstracts of ICPP. 20 p.
Rao, M.R. and Willey, R.W. 1980. Evaluation of yield stability in intercropping: studies with sorghum/pigeon pea. Experimental Agriculture 16: 105-116.
Ruthenbag, H. 1980. Farming Systems in the Tropics. 3rd edn. Clarendon Press, Oxford. 424 p.
Sombroek, W.G. 1990. Soils on a warmer earth: tropical and subtropical regions. In: Soils on a Warmer Earth. H.W. Scharpenseel, M. Schomaker and A. Ayoub (eds.). Elsevier, Amsterdam. pp. 157-174.
Stafford Smith, M. and Campbell, B. 1994. GCTE Task 3.1.3 Implementation Plan. GCTE Report 3. GCTE, Canberra.
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 21, 95 pp.
Swift, M.J. (ed.). 1991. Soil fertility and global change: the role of TSBF studies in the IGBP. Biology International, Special Issue No. 25. 24 p.
Swift, M.J., Seward, P.D., Frost, P.G.H., Qureshi, J.M. and Muchena, F.N. 1995. Long-term experiments in Africa: developing a database for sustainable land-use under global change. In: Long-term Experiments in Agricultural and Ecological Science. R.A. Leigh and A.E. Johnston (eds). CAB International, Wallingford. pp 229-251.
Vandermeer, J.H. and Schultz, B. 1990. Variability, stability and risk in intercropping: some theoretical considerations. In: Agroecology: Researching the Ecological Basis for Sustainable Agriculture. S.R. Gliessman (ed.). Springer-Verlag, New York. pp. 205-232.
Worster, D. (1979). Dust Bowl: The Southern Plains in the 1930s. Oxford University Press, New York.
