Abstract
Pesticides are often toxic to nontarget organisms, especially to those living in rivers that drain agricultural land. The brown trout (Salmo trutta) is a keystone species in many such rivers, and natural populations have hence been chronically exposed to pesticides over multiple generations. The introduction of pesticides decades ago could have induced evolutionary responses within these populations. Such a response would be predicted to reduce the toxicity over time but also deplete any additive genetic variance for the tolerance to the pesticides. If so, populations are now expected to differ in their susceptibility and in the variance for the tolerance depending on the pesticides they have been exposed to. We sampled breeders from seven natural populations that differ in their habitats and that show significant genetic differentiation. We stripped them for their gametes and produced 118 families by in vitro fertilization. We then raised 20 embryos per family singly in experimentally controlled conditions and exposed them to one of two ecologically relevant concentrations of either the herbicide S-metolachlor or the insecticide diazinon. Both pesticides affected embryo and larval development at all concentrations. We found no statistically significant additive genetic variance for tolerance to these stressors within or between populations. Tolerance to the pesticides could also not be linked to variation in carotenoid content of the eggs. However, pesticide tolerance was linked to egg size, with smaller eggs being more tolerant to the pesticides than larger eggs. We conclude that an evolutionary response to these pesticides is currently unlikely and that (a) continuous selection in the past has either depleted genetic variance in all the populations we studied or (b) that exposure to the pesticides never induced an evolutionary response. The observed toxicity selects against large eggs that are typically spawned by larger and older females.