Community structure and dynamics in desert ecosystems: potential implications for insecticide risk assessment.

Abstract

Insecticides are increasingly being used in hot arid ecosystems. The evaluation of the ecological risk these insecticides may pose, however, is based largely on data derived from temperate organisms and ecosystems. The major differences in the composition, structure, and functioning of desert animal communities when compared to temperate terrestrial ecosystems are discussed. Desert communities are characterized by a high fraction of ectotherms (both vertebrates and invertebrates);rodents and insectivores appear to dominate the mammalian fauna; and detritivores make up a very large part of the arthropod fauna. Presently available toxicity data cover these groups only to a very limited extent. It is not known if the ranges of insecticide susceptibility observed in temperate species are representative of those in arid organisms. Thus, it is not certain that ecotoxicological assessments based on such data sets adequately protect desert animal communities. It is shown that food web connectance is higher in desert ecosystems than in temperate grasslands or forests. This may to a large extent be due to the high degree of omnivory among desert organisms. Population regulation between predators and prey appears to be weaker in deserts. The same is often, though not always, the case for competition among desert organisms. It is argued that such characteristics will reduce the chance that strong indirect effects of insecticide perturbations will occur. In spite of the fact that many desert organisms are well adapted to cope with high temporal and spatial environmental variability, there is no reason to believe that they will always recover more rapidly from population perturbations caused by insecticides. The relatively large physiological and life-history plasticity encountered in many desert animals may increase tolerance to insecticide stress. Food chains are longer in deserts than in temperate grasslands and forests. The implications of this observation for the risk of biomagnification of contaminants are discussed.