ABSTRACT
Recent workshops have focused on the need for and use of experimental ecosystems (agricultural microcosms and mesocosms) to assess the environmental risk of pesticides. The accuracy with which scientists can predict contaminant effects in natural systems from such experimental field studies, with coordinated laboratory testing, is often questioned. Successful extrapolation of response data collected from mesocosms to surface waters depends on physical and biological characteristics of the mesocosms (morphometry, trophic status, nature of resident species) and the functional redundancy and assimilative capacity of these experimental ecosystems. Regulatory interpretation often relies upon univariate measures of key single species indicators. A review of published studies indicates that often, direct responses of fish can be predicted by laboratory toxicity responses. However, for various reasons, indirect responses among aquatic biota cannot be predicted solely from laboratory data. How well test results on experimental ecosystems (for example, mesocosms or agricultural microcosms) protect natural aquatic environments furthermore depends upon the physical/chemical properties of the chemical, and the accuracy with which organism exposure is measured. Measurement of biotic responses in micro- or mesocosms needs to match the appropriate scale of chemical perturbation to life history characteristics of the affected species. Interpreting the response endpoints hinges on an awareness of the species in the community: their trophic level status and manner of obtaining food, micro-habitat requirements, and size and reproductive potential. Extrapolating mesocosm results to natural ecosystems calls for conservative estimates of hazard because of the limits on predictability in natural ecosystems and the complicating effect of exposure to multiple chemicals.
