Diversity, complexity, stability and pest control.
It should be possible to design agroecosystems to reduce the severity of insect pest problems. Ecological theory about stability should be relevant whenever pest control relies heavily on the action of insect enemies (predators and parasites). To pursue such a control strategy the crop system should contain enough plant species to maintain continuity of the enemy species (and perhaps sometimes of the pest species). However, current evidence from agriculture, a comparison of different natural ecosystems, laboratory communities and mathematical models does not support the more general hypothesis that species diversity per se enhances stability. Thus, unless additional species contribute to the action of natural enemies, the addition of diversity, qua diversity, is not likely to be a useful strategy in agriculture. It is suggested that the marked instability of agroecosystems (and other artificial communities), in contrast to the stability of natural communities, results from the frequent disruption of crops by human intervention and from the lack in crop systems of co-evolutionary links between the interacting species. This second feature of crops is caused by the haphazardness of the collection of species on any given field crop, the changing selective regime imposed by man, and the fact that crops have lost many species that were present in the previously existing natural communities. Thus, it is concluded that natural communities, whose stability results from their complement of co-evolved species provide a poor model for the design of crops systems. By the same token, laboratory and mathematical models, like the agricultural systems, are relatively poor in co-evolved peculiarities, and such models may therefore be useful in yielding insights about agricultural systems. Recently formulated mathematical models of predator-prey systems suggest that physical complexity, especially a patchy distribution in space, enhances population stability. This result is supported by some meagre evidence from the laboratory and field. A major difficulty in the application of this principle to agriculture hinges on the scale on which patchiness should be incorporated in fields of crops, but field experiments could examine this question. Stability is assumed to be desirable in crop systems, being translatable into reduced pest fluctuations and pest damage. However, little attention has been given to the question of how to obtain both stability and an acceptably low average density of pests. Some of the tactics suggested here achieve density dependence at the cost of lowering the overall predation rate at low pest densities. This trade-off will be worthwhile if the predators can thus keep the pest at low densities.