Retrospective estimation of population-level effect of pollutants based on local adaptation and fitness cost of tolerance.

We present a novel framework for estimating site-specific effects of pollutants on natural populations. Our method is based on fitness optimization and uses observed differences in tolerance (sensitivity) to a particular pollutant between populations at contaminated and uncontaminated sites (i.e., target and reference populations). In addition, the method uses laboratory estimates of the fitness cost of tolerance, that is, the reduction of population growth rate (fitness) of a target population compared to that of a reference population when both are maintained in uncontaminated conditions. As a case study, we applied this framework to analyze observed genetic differentiation in tolerance to the pyrethroid insecticide fenvalerate between Daphnia galeata populations in Lake Kasumigaura and an adjacent agricultural pond. The estimated exposure level at the contaminated site was about 0.015 μg/L, and the population-level risk corresponded to about a 24 % reduction of the intrinsic rate of natural increase.