Abstract

Changes in the way animals perceive and respond to the environment are key to adaptation and speciation. However, we still know little of the genetic mechanisms underlying shifts in behavior. Identifying these genetic changes would provide an important route towards understanding how behavior is generated, both during development and across evolutionary time. To begin to fill this gap, in this thesis I investigate the genetics of behavioral evolution in Heliconius butterflies. In particular, I investigate the genetics of visual mate preferences as well as broader visual adaptations across Heliconius species. Heliconius butterflies display a striking diversity of warning patterns, which they also use as mating cues to recognize conspecifics. Preferences for conspecific warning patterns have a strong genetic component, but unlike the warning pattern cues, the exact genes responsible remain unknown. In chapter 1, I analyse a causative genomic region for such divergent visual behaviors in two Heliconius species: red H. melpomene and white H. cydno. I couple population genomic and gene expression analyses of neural tissue of these species and their hybrids across development, to identify five genes that are strongly associated with divergent visual preferences. The functions of these candidate genes suggest shifts in behavior involve changes in visual integration or processing, which would allow mate preferences to evolve without altering perception of the wider environment. In chapter 2, I expand my analyses to include another species: red H. timareta, a co-mimic of H. melpomene. There is substantial evidence that H. timareta acquired its red wing pattern coloration through hybridization (adaptive introgression) with H. melpomene. In this chapter, together with my colleagues, we test the hypothesis that H. timareta also acquired alleles for visual mate preference from H. melpomene. We first show that the same causative region associated with the divergent visual preferences of H. melpomene and H. cydno, also controls visual divergence between H. timareta and H. cydno. I then find genomic signatures of adaptive introgression at the level of candidate behavioral genes identified in chapter 1. One of these candidates, regucalcin1, also shows patterns of gene expression strongly linked to visual preference across Heliconius species and their hybrids. Overall, I find evidence that visual preference alleles have crossed species barriers to facilitate adaptive shifts in behavior in Heliconius. Finally, chapter 3 goes beyond divergence in mate preferences to investigate broad-scale neural divergence associated with speciation in the H. melpomene/H. cydno group. Species within this group are separated across an ecological gradient of open to closed forest. We find evidence that species have adapted to this ecological transition at the neural level, through heritable, volumetric expansion of visual processing regions of the brain. We find that these same visual structures show intermediate morphologies in F1 hybrids, which likely disrupt their behavioural function. I then show that these brain volumetric changes are mirrored by adaptive divergence in gene expression level in the neural tissue of these species. Finally, I find evidence for selection against the introgression of alleles (with distinct neural expression level) between species, further indicating that neural divergence contributes to reproductive isolation. Overall, we show that broad-scale sensory/neural adaptations to the visual environment, at both morphological and gene expression level, contribute substantially to behavioral isolation and speciation across Heliconius.