The papers in this issue of the American Journal of Alternative Agriculture are devoted to the proceedings of a symposium, “The Role of Ecology in Lower-Input, Sustainable Agriculture.” This symposium was sponsored by the Ecological Society of America as a part of its annual meetings with the American Institute of Biological Sciences at The Ohio State University, Columbus, August 10, 1987. The symposium was organized for the purpose of relating ecological approaches and concepts to sustainable agriculture. Additionally, it was our intention to indicate research opportunities for ecologists in agricultural systems.

Conventional agriculture has caused economic problems associated with overproduction of crops, increased costs of energy-based inputs and lessened farm incomes. It has also produced ecological problems such as poor ecological diversity, soil erosion, and soil and water pollution. The adoption of integrated systems of agricultural production involving lower inputs of fertilizer, pesticides and cultivations can alleviate these economic and ecological problems. Such systems are dependent upon a good understanding of the nature of interactions between the four main components of such systems, which are fertilizers, pesticides, cultivations and rotations and how these interactions influence crop yields and farm income. Alternatives to energy-based inputs include legume rotations, use of waste organic matter as well as that from animals and crops, integrated pest management, pest and disease forecasting, biological and cultural pest control, living mulches and mechanical weed control, conservation tillage methods, and specialized innovative cultural techniques, including intercropping, strip cropping, undersowing, trap crops, double-row cropping. Two commercial farms in Europe operated on integrated lower input systems have increased farm incomes by 5–20 percent. Further research on such techniques is needed, as well as computer-aided management systems, appropriate extension aids and farmer education.

This paper has three goals: (1) to convince ecologists and evolutionary biologists to study the evolution in insects of the ability to overcome crop protection measures, (2) to provide insights into the kinds of data needed to develop methods for retarding the evolution of such traits, and (3) to suggest that the study of these phenomena can further our understanding of evolution. The evolution of resistance to chemical insecticides often results in higher application rates and constant development of new classes of these potentially environmentally degrading toxicants. Moreover, it is important to understand resistance phenomena related to alternative crop protection measures involving plant genetic resistance and biological insecticides so that these less environmentally damaging control measures can be maintained. Insecticide resistance evolves in insect populations in response to selection by chemical compounds. Similarly, selection by host-plant defenses of resistant crops leads to the evolution of virulence to those varieties. The evolution of these traits constitutes an important subject of applied evolutionary biology. In the context of single-gene evolutionary models, this article reviews the most common strategies that have been suggested for retarding the evolution of insecticide resistance. These models are also used to illustrate the effects of ecological factors and genetical properties of insect populations on the evolution of resistance. Where appropriate, the relevance of these models to the evolution of virulence to resistant crop varieties is also described. Durability in the insecticidal effectiveness of a plant protective chemical is not incompatible with the requirement for health safety in the same material.

Low external-input agriculture seeks to optimize interactions between species that permit processes at the ecological community level to shift inputs away from nonrenewable or synthetic sources. This goal can be reached through the study of the ecological basis of an array of plant-plant and plant-insect interactions, especially in intercropping systems where particular plant mixtures are of benefit to the entire crop community. These polyculture systems can be managed for nutrient cycling efficiency and pest and disease regulation using knowledge of multi-trophic level interactions and application of recent developments in mutualism and competition theory. A mechanistic model of additive and removal reactions on the environment is proposed as a means of studying species interactions.

Developing reduced-input agroecosystems is a challenge to researchers and practitioners in a number of disciplines. In particular, because of the complexity of managerial inputs, understanding the effects of decreases in levels of one or several inputs presents a difficult experimental problem. Moderate reductionism is proposed as an appropriate hierarchical analysis, which first characterizes system level behavior and then explains it in terms of system components and their interactions. An example is given using reduced nitrogen fertilizer inputs; higher nitrogen use efficiency at lower levels of input is explained by increased nutrient recycling between soil and plant subsystems. Transition from conventional to reduced-input management involves changes in both economic and ecological properties of agroecosystems. Biological, chemical and physical changes, including increased species diversity, tighter nutrient recycling, and longer-term perennial effects, may parallel ecological succession. Hypothetical trends for several such changes are discussed.

The conversion from high to low input agricultural systems has been studied at the field and farm scale but analysis of the effects of conversion at the landscape or regional scale has not previously been attempted. In this paper, we apply existing historical data bases on changing inputs and changing land uses to project possible effects of conversion on productivity and environmental quality on the landscape scale. We chose two areas for the analysis of changing input effects on landscape structure and function. One area is a region defined as the southern half of the lower peninsula of Michigan (Southern Lower Michigan or SLM). The second area is a small watershed, Watershed K of the Little River Watershed, in the Gulf-Atlantic Coastal Plain of Georgia. In SLM, land use data suggest that lower input use will lower per hectare productivity, and that regional production can only be maintained by bringing marginal lands into production. The prime lands not already in crop production in SLM are mostly poorly drained or sandy soils which will require either drainage modifications or supplemental irrigation for crop production. Intensification of input use in SLM during the last 20 years has led to increases in aggregate production, but may have contributed to declining water quality in the Grand River, the major drainage of that region. On Watershed K, corn grain yield was linearly related to N fertilizer application. A hypothetical yield response curve with a yield of 1500 kg/ha of corn grain without nitrogen fertilizer was selected for use in the analysis of changing inputs. Land use data on Watershed K showed that even with no nitrogen input to corn, the aggregate production of corn could be maintained without encroachment on riparian areas which are critical to maintenance of water quality. Our analyses, which are limited to historical data on the relationships among input, yield, and environmental quality, raise many questions about the landscape scale effects of conversion from high to low input systems. Relationships between inputs and yields are poorly understood at the field scale and relationships among inputs, productivity, and environmental quality have not been established at the landscape scale. Our analyses also raise questions about the need to maintain aggregate landscape production in the light of crop surpluses and declining environmental quality.

Modern, conventional agriculture has generalized the technology of farming, minimizing the significance of the relationship between farmers and their particular land. This generalization undermines farm culture and impedes its transmission from generation to generation. To avoid making the same mistake, ecologists designing the new, lower input agriculture should: 1) reject any black box presentations of their systems that the farmer cannot understand and that interfere with the relationship between the farmer and the land; 2) make systematic efforts to rediscover traditional farm wisdom and incorporate it into the new system; 3) utilize academics in a major program of local, adult education for farmers; and 4) reorganize extension services to facilitate the lateral transfer of information from farmer to farmer, and develop methods of incorporating farmers into the agricultural research process at the planning stage.