Incorporating results of avian toxicity tests into a model of annual reproductive success.

Modeling the effects of pesticide exposure on avian populations requires knowledge of how the pesticide changes survival and fecundity rates for the population. Although avian reproduction tests are the primary source of information on reproductive effects in the pesticide risk assessment process, current tests cannot provide a direct estimate of the effects of a pesticide on fecundity rates. We present a mathematical model that integrates information on specific types of effects from reproduction tests with information on avian life history parameters, the timing of pesticide applications, and the temporal pattern of pesticide exposure levels to estimate pesticide effects on annual reproductive success. The model demonstration follows nesting success of females in no-pesticide or pesticide-exposed populations through a breeding season to estimate the mean number of successful broods per female. We demonstrate the model by simulating populations of a songbird exposed to 1 of 2 hypothetical pesticides during a breeding season. Finally, we discuss several issues for improving the quantitative estimation of annual reproductive success.