From the biodiversity viewpoint, there is a difference in agroecosystems between the planned biodiversity and the unplanned, or associated, biodiversity. The latter are all those organisms, above- and belowground, that have found niches to fill among the planted trees and crops. The extent to which unplanned biodiversity occurs in different agroecosystems is not well known or understood, although studies have started to address the effects of a broad spectrum of agricultural practices on wildlife populations (McLaughlin and Mineau, 1995; Perfecto and Snelling, 1995). Swift et al. (1996) have, however, drawn four very different scenarios for the relationships between agricultural intensification and biodiversity (Figure 3), although these may be very scale dependent (e.g., from farm to landscape) and probably also vary depending on the level of biodiversity at the time of planting. Thus, the biodiversity associated with an agroforest planted on recently cleared land at the forest margin, as an alternative to slash-and-burn agriculture, would almost certainly be very different from the same tree/crop mixture planted to rehabilitate already degraded land. There is a need for controlled experiments to determine these relationships between intensification and biodiversity. In addition there is a need to determine the patterns of diversity in different agroforestry systems and their implications for ecological functioning at different scales. There is currently unresolved debate about the functional role of species diversity in ecosystems (Johnson et al., 1996), which makes it difficult to suggest best practices. However, it seems that planned biodiversity should aim at maximizing ecosystem processes (nutrient cycling, production of different products, light requirements, etc.) and structural complexity, rather than increasing the number of species per se. However, species numbers may also impact on function. The few studies in which the diversity of agricultural ecosystems has been manipulated suggest that increases in diversity from 0 to 10 plant species alters ecosystem function, but that there is little effect beyond that point (Schulze and Mooney, 1993).
Species also vary in their importance in different food webs, where, once again, scale is important. For example, for a fruit tree species to support a population of monkeys will require a large habitat. A small area of fruit trees may therefore avoid monkey damage and so be preferable for production, but without a large area there may not be a market for the crop. Thus, ecological and economic factors affecting
Unman aged Shifting Low intensity \íiddle intiwsity High intensity Intensive system cultivation management management management monoculture
Figure 3 Hypothetical relationships between agricultural intensification and agroecosystem biodiversity. (From Swift, M. J. et al., 1996. Functional Roles of Biodiversity: A Global Perspective, Mooney, H. A. et al., Eds., SCOPE 55, John Wiley & Sons, New York. With permission.)
the viability of production systems can be antagonistic. Typically, predators require larger range than their prey; thus for the sustainable production of these fruits the area may need to be big enough to support leopards. Top predators are, however, often unwelcome in farmland while their absence creates pest management issues for farmers. In this context, a land-use mosaic with corridors between forest patches may be the most appropriate means of acquiring the scale needed to achieve some level of ecological equilibrium; however, the likely benefits of corridors between forest patches are not fully understood (Hobbs, 1992). This concept thus returns to issues of scale, land-use design, and the "integrate vs. segregate" debate (van Noordwijk et al., 1995b) in developing optimal land-use strategies.
FOREST PATCHES, BIOGEOGRAPHICAL ISLANDS, AND AGROFORESTRY
In the tropics, conservationists have focused their attention on the protection of natural forests and woodlands and, until recently (Schelas and Greenberg, 1996), have not given much attention to the widely dispersed forest patches throughout human-occupied landscapes. These patches are often critical components of a farmer's environment, being a source of products and environmental services of importance to the farmer's livelihood and welfare. As biogeographical islands, their role in maintaining biological diversity is also crucial. Issues of scale are central to this role, and thus, in a landscape mosaic, forest patches and areas of agroforestry are potentially complementary, especially when considering the need for ecological equilibrium and population size vis a vis genetic diversity. There is thus a need for ecologists and geneticists to become more involved in agroforestry research and the process of farmer reforestation. Biological and genetic diversity are, for example, becoming an issue in the densely populated, high-potential areas of Kenya. These are areas where the situation of "more people, less erosion" (Tiffen et al., 1994) is also becoming "more people, more trees" (Holmgren et al., 1994), as tree biomass is increasing at 4.7%/year, while the high human population is expanding at 3.0%. However, of concern is the fact that about half the trees planted on a farm are of a single species (Grevillia robusta), while the other half consists of about 10 to 20 species. This raises the question at what point the tree population reaches a size and distribution such that the controls that prevent isolated outbreaks of pests and disease break down. Not enough is known about the patch dynamics of such outbreaks relative to species numbers and configurations to steer agroforesters in the development of risk-averse strategies. The use of only a few tree species in agroforestry to provide corridors between forest patches could obviously allow their pests and diseases the opportunity to spread, while reducing the effectiveness of corridors for the ecological and genetic stability of other species. This reaffirms the need for both diversity and the need to consider the ecological implications of landscape design. In the case of G. robusta, which was introduced into East Africa from Australia, the risks of its dominance in the landscape could be exacerbated by its limited genetic base. Fortunately, new and more diverse accessions of G. robusta have now been introduced in East Africa, but it will take some time for these to be disseminated around the countryside. This example illustrates the need to ensure that agroforestry promotes the importance of diversity and not just the planting of trees by farmers.
Strictly, biogeographical islands are completely isolated populations, in which island size is positively related to biodiversity. The concept has, however, also been applied to shrinking habitats such as woodlands in farmland. Although Simberloff (1988) concluded that, in a wildlife conservation context, treating woodlands as islands in this way provided contradictory or inconclusive evidence, there does seem to be some merit in considering (1) how agroforestry could perhaps benefit biodiversity through increasing island sizes and (2) what might be the possible impacts of such action. It is not known, however, if the species-area relationships apply when habitats are being expanded. Can agroforestry, as a form of restoration ecology, reverse the 18 factors (e.g., low population density, inbreeding, catastrophe, competition, habitat destruction, disease, etc.) associated with species extinction listed by Soule (1983)? Agroforestry can probably reverse some of them, but which of the 18 factors are the most crucial and which should these be the prime targets for the development of a healthy agroecosystem? At present, it is only genetic diversity that is getting any attention. The decrease in size of biogeograph-ical islands lowers genetic diversity, increases the likelihood of inbreeding, and lowers the chances of future allopatric speciation (Soule, 1986). Agroforestry, because of its production orientation, is promoting the retention and introduction of genetic diversity among the tree species planted by farmers. This aspect of tree domestication at least is one positive step toward a healthier and more productive environment. It is hoped that, as agroforestry is increasingly seen as applied ecology, other such steps will follow, such as the geographic expansion of the multistrata agroforestry systems already described.
Was this article helpful?