Can pollinators choose between pesticide-laced and pesticide-free foods? And is choice a good thing in terms of conservation? Paola Olaya‐Arenas and Ian Kaplan discuss their new research. It is becoming increasingly clear that non-target insects such as pollinators and natural enemies are forced to develop on food sources (e.g. pollen, leaves) containing pesticide residues intended for crop fields. This is often harmful to their survival, growth, and all-around performance. The assumption however, is that insects cannot tell the difference between pesticide-laced vs. pesticide-free foods and thus readily feed without displaying aversion behaviours. It is unclear whether to expect insects to reject pesticide-contaminated foods, particularly those at sublethal concentrations, which impose a lower selection pressure. In co-evolved systems, the presence of evolutionarily novel substances like pesticides could be readily detected by specialist insects that are attuned to even minor chemical differences in their environment. To date, pesticide-mediated feeding preferences have been mostly observed in bees, but these outcomes are complicated by the complex foraging regulations and feedbacks in social insects. Also, all of this work is with managed honey or bumble bees (Apis mellifera, Bombus sp.), which are non-native and/or extreme generalists. The process of the study. Illustration: Tom Kronewitter In this new study, we tested the role of pesticide presence and concentration on larval feeding and adult oviposition preferences in the monarch butterfly, Danaus plexippus (Linnaeus, 1758). In North America, eastern migratory monarchs are undergoing a long-term population decline and pesticide exposure is one of several factors implicated. We had previously shown that leaves of their main host-plant, the common milkweed Asclepias syriaca (L.), contains traces of several agricultural pesticides. This is not surprising, as milkweeds are largely relegated to field-edges and mostly exist in close proximity to corn or soybean fields in the Midwestern United States. Because monarchs are specialists on milkweed, the quality of their host-plant is a critical factor in the management, ecology, and population dynamics of this iconic species. Overall, we have found that mated female butterflies place ca. 30% fewer eggs on milkweeds treated with a cocktail of pesticides, simulating the concentrations measured from field plants. Similarly, newly hatched caterpillars show feeding aversion for four of the six pesticides tested at mean and maximum concentrations. These data from both life stages suggest that butterflies and their offspring commonly prefer milkweeds that are free of pesticide residues when given the choice. Interestingly, we observed this, not only with insecticides, but also some fungicides and herbicides, whose developmental effects on insects are poorly documented. Photo: Paola Olaya-Arenas From a conservation perspective, is choice a good or bad thing? It depends. On the one hand, if monarchs can behaviorally regulate pesticide exposure risk by simply avoiding plants with high pesticide loads, this should be beneficial to their persistence in agricultural landscapes. However, this possibility is limited by the fact that we know little about the range of pesticide values across milkweed plants or patches experienced by monarchs on a given foraging bout. We know even less about within-plant variation (i.e., among or within leaves) that would allow caterpillars to self-select optimal food. Alternatively, displaying choice could be a bad thing from a population perspective. Milkweeds have dramatically declined in agricultural areas in recent decades and monarchs may not be well-positioned to reject plants. Individuals could die while seeking out ‘better’ options or pesticide presence could, in some cases, interfere with the co-evolutionary arms race between monarchs and milkweeds, encouraging them to place eggs on plants express