Previous research and education programs associated with academic institutions, government agencies, and non-government organizations have provided a strong foundation for continued progress towards the goal of agricultural sustainability. Modern initiatives in conservation tillage, integrated pest management, soil testing, and farming systems research and extension offer examples of programs that deal with both short term productivity and long term sustainability. This paper presents a brief description of the newly emerging science of agroecology and offers an example of how research and educational institutions can continue to respond to public concerns about the impact of agricultural production on environmental and social systems.

    The emerging science of agroecology provides an appropriate framework in which to develop new programs that encourage both public service and scholarly research and education. Much of the confusion over sustainability, profitability, and productivity disappears when we employ ecological principles and measures to study agricultural systems. Further, agroecology is the framework which will allow us to scientifically address multiple primary objectives and understand complex systems. Implicit in agroecological research and education is the idea that knowledge of ecosystem relationships will allow farmers to manage inputs and processes in agricultural production systems and thereby optimize for productivity, sustainability, stability and social equity.

    In order to understand agroecosystems, boundaries must first be placed on the system under study. An agroecosystem may be thought of as a complex of air, water, soil, plants, and animals in a defined area that people have modified for the purpose of agricultural production. It exists in an environmental setting which defines resources available. Farmers and consumers exist in a social setting which conditions how they interact with each other and with the environment. An agroecosystem may be a field, a farm, or a larger region such as a river valley. It is important to recognize that the boundaries of an agroecosystem exist in space, as well as in time.

    While agroecosystems are simpler in some ways than natural ecosystems, they are complicated by the human component. People are part of agroecosystem analysis and social equitability is a co-objective, right alongside of productivity and sustainability. Today, few individual scientists have the training to scientifically address these multiple objectives. Nevertheless, cooperative programs among farmers, production agriculturists, sociologists, economists, and ecologists can and should be encouraged to improve our understanding and subsequent management of agricultural ecosystems.

    A systematic research method for preliminary agroecosystem analysis was described by Gordon Conway, of the Centre for Environmental Technology at the Imperial College of Science and Technology in London (1). The procedure is structured as a 7 day workshop which results in a set of questions for research. These questions are formed into hypotheses which are then tested using traditional scientific methods. Over time a better understanding of the agroecosystem provides the basis for development of improved production systems.

    Agroecosystem analysis is founded on the first principle of ecology that all things are interconnected, and is based on ecological principles that govern relationships among biotic (living) communities and the abiotic environment. Biotic relationships may be described by principles of predator/prey interactions, competition for food sources and habitats, cooperation and commensalism etc. Abiotic relationships are described by nutrient cycling, carbon cycles, energy cycles, etc. over space and time.

    Conway suggests that four system properties may be used to understand the dynamics of an agro-ecosystem. They are; productivity, stability, sustainability, and equitability. These may be used as neutral descriptors of system behavior for the purpose of understanding, or they may be used as indicators of performance (co-objectives). In the latter case a value judgment is made, and humans determine which of the performance indicators are most important to the human community.

    Productivity is the quantity of product or output from an agroecosystem per unit of some specified input. For an agroecologist, output may include a marketable product such as bushels of corn, as well as negative products of a system like water runoff, pesticides lost, or soil lost. Of course, tons per hectare is a standard measure of productivity. But productivity can also be expressed in other units of output per unit of input. Inputs may be measured in tons of fertilizer, monetary value of pesticide, or kilocalories needed to deliver irrigation water.

Production relationships might be expressed as tons of grain produced per unit of soil loss due to erosion. While this may have little short term economic significance, it can be used to help us better understand the dynamics of the agroecosystem.

    Stability is consistency of production in spite of short term upsetting influences such as uneven rainfall, pest explosions, price variability, etc. Annual variations in productivity indicate a lack of stability.

    Sustainability is the ability to maintain a desired level of production over time, in spite of long term destabilizing influences. Examples of these are; increasing soil salinity, off-site effects of soil erosion, declining market prices, or accumulation of biotoxins in the environment. Systems which rely on heavy inputs of non-renewable and rapidly diminishing resources are not considered to be sustainable. Sustainability is a measure of persistence or long-term resilience of a production system.

    Equitability can be used as an indicator of agroecosystem performance which incorporates the social dimension. Social equity is a measure of the degree in which resources and products of a production system are equally distributed throughout the human population. This implies that equality of product availability (output) and equality of resource availability (inputs) are the preferred norm.

    Irrigation capability can be used to help understand how these system properties may be used as performance indicators by agroecologists. Irrigation may improve productivity because yields per acre increase. Stability is generally improved since farmers are no longer susceptible to unreliable rainfall. However, the system is only sustainable if increased irrigation doesn't result in other problems such as soil salinity or depletion of limited water resources. In addition, social equity might be higher when all farmers were subject to natural rainfall. Once the system dynamics are adequately understood, individual components can be manipulated to improve one or more of these variables. With a better understanding of the impact of a specific technology, such as irrigation, on productivity, sustainability and social equity, improved system performance can follow.

    The objectives of agroecosystem analysis are to identify research priorities and propose tentative guidelines for improving agroecosystem performance. Conway proposes the following procedure:

    1. Define the of specific objectives and boundaries of the agroecosystem. This might be to increase productivity on a specific farm or improve equitability of resource distribution within a farming region. Boundaries are defined in space and time, with biological, physical and social components.

    2. Describe the agroecosystem under study. Conway calls this pattern analysis. It includes a physical description the agroecosystem using a series of maps and transects. A pattern analysis over time uses graphs of seasonal and annual changes. These might include indicators of credit and labor availability, rainfall, prices, cropping sequences and product supply. Another pattern analysis describes the flow and transformation of energy, basic materials, money, and information through the system. This is done using flow diagrams, graphs and charts. The last pattern analysis describes methods of decision making. It includes a list of the important choices that must be made and who influences these decisions.

    3. Describe key relationships as they impact the four basic system properties of productivity, stability, sustainability and equitability. Conway provides some examples of key relationships that impact these properties for a rice agroecosystem in Northeast Thailand. They are:

        a. productivity

            - product demand by world markets

            - government rice and fertilizer policies

            - water resource development

        b. stability

            - rainfall, especially floods and droughts

            - rice production

            - human migration

        c. sustainability

            - increasing salinity

            - increasing indebtedness

            - deterioration of communal help arrangements

        d. equitability

            - subsistence rice crop

            - government information sources

            - credit availability

    4. Finally, once the system is described through pattern analysis and a discussion of system properties, research questions are identified. These may address gaps in the knowledge base, such as questions about meteorological conditions, cropping sequences, or amounts of fertilizer applied. Or they may be broader questions about relationships and system properties. The most powerful questions, according to Conway, address system properties and, in particular, the actual potential tradeoffs between them. For example, "to what extent are the gains in productivity and stability from water resource management and irrigation capability likely to be offset by a decline in sustainability and equitability?" Or "if large land holdings were broken up into smaller farms, to what extent would the gain in equitability offset potential declines in productivity? - or would there indeed be declines in productivity?"

    Questions must be developed into testable hypotheses before research is initiated. Likely research areas will be agronomic as well as social. The research that follows will be similar to projects that arise from independent initiatives. However, they will differ in that the individual research projects will result directly from a multidisciplinary systems analysis. Results of individual projects will then feed back into, and modify that analysis. Conway maintains that the likelihood of effective multidisciplinary programs is enhanced following this approach. Further, the research findings which result from this procedure are more likely to be acted upon and manipulations of the system based on those findings are more likely to be successful.

    A research and/or educational institution that employs agroecosystem analysis will probably conduct more studies on complex relationships among the various components of a agroecosystem that go beyond simple cause and effect. There will be more research on inputs and outputs that are measured in currencies other than monetary units, such as carbon, nitrogen, and calories. This will provide the basis for better nutrient management recommendations. There will be more studies on relationships among pest populations, predator populations, host populations (both agricultural crop and non-crop species), and environmental influences on these. This will provide the basis for better pest management recommendations. There will be more studies on interspecific relationships among crop plants and the effect of these on pest populations. This will provide the basis for innovative multiple cropping systems.

    This work will be similar to the research we conduct today, in that they will be well-conducted scientific studies. They will be similar in many ways to the component research, that is being criticized today as too specialized or too reductionist. But the answers to the questions we ask; and the understanding of the hypotheses we pose, will be thoroughly embedded in a holistic understanding of the systems we study. There is great power in complementing the study of multiple relationships among components of a system, with the study of the components themselves. The reductionist thinking, for which we have been so criticized of late, will be better appreciated as the power of systems analysis enhances the impact of our specialized studies.

    If we are doing everything right, nothing will change. If we are not, ecological principles will provide an appropriate framework, in which to better understand and manage the impact of agricultural systems on food supply, the environment and people. I believe that ecological approaches to agriculture can be the common ground upon which we address the needs of society, while employing the tools of scholarly research and teaching, for which public institutions are best equipped.

1. Conway, G. 1985. Agroecosystem Analysis. Agr. Admin. Vol. 20:31-55.

This paper was presented by John Gerber, former Assistant Director in the Illinois Agricultural Experiment Station, as part of a U.S.A.I.D. sponsored workshop on sustainable agriculture, Lusaka, Zambia, September 18-21, 1990.