In the last decade cortccrrts for sustainability have replaced the maximisation of productivity as the target for agricultural development. This has generated increased interest in "agroecosystem design", a more holistic concept than the "commodity-led technology development" paradigm which has dominated the post-world-war period of agricultural development. The fundamental features of this sustainability agenda, are that productivity should meet the aspirations of the farmers and society, whilst at the same time conserving resources and environments for the future. It has been hypothesised that the inclusion of biodiversity is a key feature of such sustainable agroecosystem design (izac and Swift 1994). Agroecosystem design should thus draw on scientific information derived from the study of "complex agroecosystems" rather than simply on reductionist information drawn from the study of crop plants in isolation. Fundamental to this information base are the principles associated with the trade-offs between productivity and competition described in previous sections.
An appropriate metaphor to illustrate the differences between agroecosystem design and commodity-led technology development is that of the engineer planning a road. One option is an autoroute designed by drawing a straight line from place to placc on a map and leading to the construction of a straight and level road regardless of the physical impediments. The second option is to build the roadway along contour lines utilising the heterogeneity of the environment rather than dispensing with it. Designing ecosystems is similar in that an ecosystem already exists where the farm will be established. In terms of sustainability, it makes little sense to begin the design project from the bottom up, so to speak (drawing a straight line on the map). Rather, understanding the relationships between structure and function in natural ecosystems provides valuable guidelines about what might be planted, what might be the problems to be encountered, and what goals are reasonable. The relationships between the physical contours and limits of the system, the variety and complexity of biological components, and the functional profile are all important components of this understanding.
Developing this analogy further, we can recognize three general philosophies associated with agroecosystem design which in their turn influence the process of agricultural intensification. First is that of incremental change, which is characteristic of early stages in intensification of traditional agricul tural practices. This practice is based on the concept of introduction of small but significant changes in biotic composition or agricultural practice in response to local necessities or modified goals. At the other extreme, associated with a more regulated and "planned" approach to agricultural development, is the conversion of the natural ecosystem into one that contains only those biological and chemical elements that the planner desires, almost irrespective of the background ecological conditions, what we have called above the "autoroute path" to agricultural development. This is the approach which has been adopted most vigorously in the development of modern intensive agriculture in Europe and North America in the years following the Second World War and subsequently in the Green Revolution in many tropical countries. The effect of this approach is totally to reconstruct the landscape. Intermediate to these two approaches, representing a substantial modification of the landscape from a natural vegetation to a site of intensive agricultural production, is the "contour pathway" approach. This approach recognises the need to reconstruct the ecosystem in response to the pressures of agricultural development, but to do so by utilising biological diversity and complexity rather than rejecting it.
How each of these pathways relates to biodiversity is partially a function of how agroecosystem goals are formulated for each. In the autoroute metaphor, for example, it is usually the case that narrow production targets are pursued, and all that is asked of the background ecosystem is basic information about production potential. In the incremental change pathway, by contrast, production targets are only one of a set of interrelated goals that include cultural factors and sustainability requirements. The contour pathway seeks to acknowledge and work with the ecological forces that provide the base on which the system must be built, as well as to acknowledge the cultural, economic and social requirements of the farming communities which will run the agroecosystem.
An example of the intermediate, contour approach is the Sloping Agricultural Land Technology (SALT) developed by the Mindanao Baptist Rural Life Centre in the southern part of the Phillipines (Tacio 1993). It is based on the planting of field and perennial crops in 3-5 m bands between double rows of nitrogen-fixing trees and shrubs planted on contours for soil conservation. The crop species include rice, maize, tomatoes and beans, while the perennials are cocoa, coffee, banana, citrus and other fruit trees. The contour lines are planted with Leucaena leucocephala or Flemingia macro-phylla and Desmodium rensonii.
It is worthwhile to examine the major ingredients of SALT in terms of resource-capture principles. The first objective of SALT is to establish a stable ecosystem using several soil conservation measures and involving a range of legumes, cereals, vegetables and trees. The crops provide a continuous supply of food and vegetative cover, while the legumes and perennial crops (eg. cocoa, coffee and fruit trees) ameliorate the chemical and physical properties of the soil. Over a 6-year period the SALT system reduced soil erosion from 1160 to 20 t ha 1 in this high-rainfall region of the Philippines. Even more remarkable was the dramatic increase in the income of the farmers, which was about seven times greater than the traditional systems over a 10-year period.
In terms of ecological suitability, SALT is also applicable to 50% of the hillside farmers in the region, and is thus being tested in Indonesia, Thailand, Malaysia, Nepal, India, Bangladesh, Sri Lanka, Cambodia and Vietnam. However, initial reaction to the extension of SALT to the sloping uplands outside the initial site area was disappointing for two reasons: first, farmers with uncertain land tenure were unable to accept the technology, a common problem with tree planting in many parts of the tropics; second, few farmers were able to afford the heavy initial investment in labour and subsequent investment for pruning.
11.5 CONCLUSIONS: BIODIVERSITY, AGROECOSYSTEMS AND LANDSCAPES
The evidence reviewed above clearly shows that biodiversity, and the system complexity associated with it, plays a very important role in fulfilling the functions of agroecosystems when considered across the full spectrum of intensification. The relationships with specific biological functions remain enigmatic, however.
On the one hand, we have cited examples where particular ecological functions are not particularly dependent on. or enhanced by, increased biodiversity. For example, the evidence from the biological control literature suggests that the regulation of an individual pest is effected quite well by a single parasitoid, multiple parasitoid releases adding little to the outcome (Hawkins 1993); the function of water-use efficiency in an agroforestry system can be readily optimized with one or a few species of trees (Ong 1995). On the other hand, we have also cited many examples where biodiversity clearly serves an agronomic function. The repeated emergence of secondary pests in cotton, for example, was a presumed consequence of the elimination of a diverse assemblage of natural enemies; the utilisation of a range of genotypes at the cultivar or specific level assists disease control; greater structural and chemical complexity in the plant system stabilises soil processes. These cases from the agricultural literature lend to support the evidence now emerging from ecological experiments that increase in species richness, particularly from very low to intermediate levels of diversity, has significant effects on ecosystem function (Ewel et al. 1991; Vitousck and Hooper 1993; Naeem et al. 1994; Tilman and Downing 1994).
Yet for the generality of agroecosystems it is not really clear what the ecological function of biodiversity might be. For example, despite the fact that intercrops generally yield better than their monocultural components, surveys from intercropping experiments show that in a significant number of eases the monocultures actually would be better (Trenbath 1976; Willey 1981), and most frequently not enough information is accumulated in intercropping studies to evaluate the question properly (Vandermeer 1989). In the final analysis we have very few case studies that can unequivocally relate biodiversity to function. Often the studies that do exist are restricted to narrow production goals, whereas many farmers and farming communities have a diverse set of goals, only one of which may be production (such as minimizing risk, attaining a minimum production, preserving cultural traditions, preserving the sustainabilitv of the system). While we can speculate on the role of biodiversity in serving these various agroecosystem functions, for the most part hard data are difficult to come by. The resolution of this is a critically important question for agroecosystem design, for attaining the goal of combining productivity with sustainability, and also because agroecosystems. because of their relative simplicity, probably offer the best opportunity for testing hypotheses linking diversity and function.
Most emphasis in biodiversity conservation has been on the preservation of a few charismatic and conspicuous organisms, or of pristine environments within national parks and reserves. In fact such organisms are a very small fraction of threatened biodiversity and such habitats represent only a small percentage of total land area (Western and Pearl 1989; Pimentcl et al. 1992). Concern about biodiversity loss during agricultural intensification has usually emphasized the transformation from natural forest to agriculture, while the transformation from low-intensity forms of management to high-intensity ones, which is today the main feature of change, has been largely ignored. It is generally accepted that the greatest biodiversity per unit area exists in tropica! forests (Wilson 1988), and since these forests are being destroyed at such a rapid rate (World Resources Institute 1990), the bulk of the world's efforts at cataloguing and conserving biodiversity are justifiably aimed at these disappearing ecosystems.
Agroecosystems are also biologically diverse in two important senses: first, depending on the management system, farmers and farming communities purposefully manage the biodiversity, sometimes at very high levels (e.g. home gardens) and other times at very low levels (modern cereal production); second, as an indirect consequence of these management practices the incidental, or associated, biodiversity also varies very considerably, and in many agroecosystems is often surprisingly large. The preceding review, for instance, provides evidence that the total biodiversity is frequently very high in traditional agroecosystems, and significant even in substantially intensified intermediate systems, but that transformation into "modernism" (very high intensity management) results in a substantial loss of biodiversity. The design of agroecosystems and agricultural landscapes thus becomes a legitimate mechanism for biodiversity conservation. Agroecosystems have functions that are clearly not found in unmanaged ecosystems, as a result of the human values associated with them. These functions are often directed at. or utilise, biodiversity and have to be reconciled with the desire to conserve biodiversity per se.
It is possible to construct two contrasting landscape models for biodiversity conservation (Vandermeer and Perfecto 1994). At one extreme we might see islands of pristine unmanaged ecosystems, set in a sea of intensive large-scale agroecosystems. Such a landscape often carries with it a substantial cost in terms of social disruption and inequity. For instance, the existence of protected conservation areas may exacerbate social tensions by emphasising the uneven distribution of resources between owners and "visitors" on the one hand, and "workers" on the other. At the other extreme we see a mosaic of unmanaged ecosystems, casually managed ecosystems, traditional agroforestry, abandoned agricultural fields, home gardens and other agricultural systems perhaps designed with structural features resembling a forest (Ewel 1986). As to which area truly conserves the most biological diversity there can be no certain answer (Simberloff and Abele 1976), but the second strategy seems more likely to support biological diversity than the first. A diverse landscape model creates a range of micro-habitats, thus providing more opportunities for various assemblages of spccies to invade and take hold: the "meta-community" effect may give rise to far greater diversity than simply the sum of the individual community patches.
Whatever and wherever the agroecosystem design, we must acknowledge that it is narrow-minded to think of it as a single plot in a single year with the purpose (function) of producing as much biomass as possible. Agroecosystems are inevitably extended over space, over time and over "values". As Mahatma Gandhi, on a number of occasions, pleaded, we must pursue conservation through austerity and proper amalgamation of ecology, economics and ethics. His prescription of development in harmony with nature anticipated what the world now recognizes - that conservation and sustainable development are two sides of the same coin, closely interlinked in that one cannot be achieved at the expense of the other. Such an integrated approach demands satisfying basic human needs in an equitable manner and sustaining and indeed promoting social, cultural and biological diversity, along with the maintenance of the ecological integrity of the system.
Was this article helpful?