Taste evolution in an herbivorous drosophilid.

Plant secondary metabolites pose a challenge for generalist herbivorous insects because they are not only potentially toxic, they also may trigger aversion. On the contrary, some highly specialized herbivorous insects evolved to use these same compounds as 'token stimui' for unambiguous determination of their host pants. Two questions that emerge from these observations are how recently derived herbivores evolve to overcome this aversion to plant secondary metabolites and the extent to which they evolve increased attraction to these same compounds. In this study, we addressed these questions by focusing on the evolution of bitter taste preferences in the herbivorous drosophilid , which is phylogenetically nested deep in the paraphyletic . We measured behavioral and neural responses of and a set of non-herbivorous species representing a phylogenetic gradient (, , and ) towards host- and non-host derived bitter plant compounds. We observed that evolved a shift in bitter detection, rather than a narrow shift towards glucosinolates, the precursors of mustard-specific defense compounds. In a dye-based consumption assay, exhibited shifts in aversion toward the non-mustard bitter, plant-produced alkaloids caffeine and lobeline, and reduced aversion towards glucosinolates, whereas the non-herbivorous species each showed strong aversion to all bitter compounds tested. We then examined whether these changes in bitter preferences of could be explained by changes in sensitivity in the peripheral nervous system and compared electrophysiological responses from the labellar sensilla of , , and . Using scanning electron microscopy, we also created a map of labellar sensilla in and We assigned each sensillum to a functional sensilla class based on their morphology and initial response profiles to bitter and sweet compounds. Despite a high degree of conservation in the morphology and spatial placement of sensilla between , electrophysiological studies revealed that had reduced sensitivity to glucosinolates to varying degrees. We found this reduction only in I type sensilla. Finally, we speculate on the potential role that evolutionary genetic changes in gustatory receptors between and may play in driving these patterns. Specifically, we hypothesize that the evolution of bitter receptors expressed in I type sensilla may have driven the reduced sensitivity observed in , and ultimately, its reduced bitter aversion. The system showcases the importance of reduced aversion to bitter defense compounds in relatively young herbivorous lineages, and how this may be achieved at the molecular and physiological level.