The evolution of herbivory in insect lineages has generated one of the most diverse groups of organisms on Earth. Adapting to an herbivorous lifestyle, however, comes with significant challenges for insects, as their host plants mount substantial defenses. Herbivorous insects must contend with the plant’s first line of defense – a battery of physical barriers (i.e., spines, trichomes, waxy and thickened cuticles) and their second line of defense – the mass production of chemical cocktails aimed at killing and/or repelling them. The mechanisms by which insects evade and overcome these plant defenses have captivated the attention of researchers for over a century. But only recently have we begun to study the genetic and molecular mechanisms underlying these adaptations. In my dissertation, I investigate the evolution and genetic basis of key adaptations that facilitate the penetration of tough leaf surfaces, and adaptions that enable changes in chemosensory reception towards and metabolism of plant toxins.

My work focuses on the herbivorous lineage within Scaptomyza, a genus that is sister to Hawaiian Drosophila and nested within the paraphyletic genus of Drosophila. While close relatives feed on decaying plant tissues and microbes, most species of herbivorous Scaptomyza feed on plants of the order Brassicales (mustards and allies). Scaptomyza have become models for studying herbivory because of their recent evolution of herbivorous feeding (ca. 10-15 million years ago), in addition to their close evolutionary proximity to Drosophila melanogaster, which affords a wealth of genetic and genomic tools.

In Chapter 1, I investigate the evolution and genomic architecture of the plant-penetrating ovipositor in herbivorous Scaptomyza. Females use their dentate ovipositors to gain access to physically-defended leaves for feeding and egg-laying. This morphological trait has evolved several times in the Drosophilidae, in most cases involving an increase in hardened bristles along the margins. I tested the hypothesis that increased ovipositor bristle number was associated with herbivory in a comparative phylogenetic framework and found a significant increase coinciding with the evolution of herbivory in Scaptomyza. We then conducted a genome wide association study on ovipositor bristle number in S. flava. We found that variation in ovipositor bristle number was most notably associated with genetic variants that were located near or in transcriptional repressors affecting bristle development. These findings suggests that increased ovipositor bristle number, a key trait for the cutting ovipositor, may have been driven by regulatory evolution on conserved developmental genes.

In Chapter 2, I examine the molecular evolution of major detoxification and chemosensory gene families. Because these genes directly mediate interactions with plant chemical defenses, I aimed to test the hypothesis that herbivory was associated with dramatic gene losses, duplications, and significant protein evolution, and to identify specific gene candidates involved in this dietary transition within Scaptomyza. We annotated these genes in three herbivorous and nine non-herbivorous drosophilid species. I then generated phylogenetic trees for each gene family and estimated evolutionary rates of gene loss and duplication. Indeed, I found that the branch leading to all three herbivorous Scaptomyza experienced significantly higher rates of gene loss in almost all chemosensory and detoxification gene families, and significantly higher rates of gene duplication in more than half. I then assessed changes in selective pressure, employing several codon-based models of protein evolution. I found that both changes in gene copy number and changes in selection pressure most strongly impacted chemosensory genes involved in the detection of yeast volatiles and bitter phytotoxins. These results add to the growing literature that indicates that herbivory is enabled by the losing the ability to detect attractive ancestral cues (e.g., yeast volatiles) and repellent host plant cues. Among detoxification genes, gene losses, duplications and positive selection were notably found in genes involved in oxidative stress responses, insecticide resistance, and phytotoxin metabolism. This study serves as a critical first step in identifying specific gene candidates that underlie behavioral adaptations that drive herbivorous insects towards appropriate food sources, and physiological adaptations that enable them to metabolize host plant toxins.

In Chapter 3, I investigate how the taste system of S. flava has evolved to deal with plant-derived bitter chemical defenses. Generalist insects are often repelled by or averse to bitter plant toxins. Indeed, the results from Chapter 2 indicated that herbivorous Scaptomyza may have lost their ability to detect bitter toxins, through the loss of gustatory receptors involved in bitter detection. This would enable them to feed undeterred on their new hosts. However, many other specialist herbivores use these toxins to identify their host plants. I therefore hypothesized that S. flava may have also evolved gustatory attraction towards mustard-specific defense compounds (specifically glucosinolates), while losing perception of bitter compounds derived from non-mustard plants. In feeding assays, I found that female S. flava, contrary to expectations, were less averse not only to a non-mustard compound (caffeine) but also to mustard-derived glucosinolates. This contrasted with non-herbivorous Scaptomyza, which were strongly averse to all bitter compounds. I then performed electrophysiological recordings on the labellar taste sensilla to test whether these behavioral changes in bitter preferences could be explained by changes in sensitivity in the peripheral nervous system. I found that one class of sensilla in S. flava (I type) had reduced sensitivity to glucosinolates, compared to homologous sensilla in the non-herbivore S. pallida. Using existing knowledge of the expression patterns of gustatory receptors in D. melanogaster, coupled with the genetic changes in gustatory receptors identified in Chapter 2, these results suggest a critical role for losses of candidate gustatory receptors (paralogs of Gr39aA and Gr28bA) in driving overall changes in bitter detection. While it remains unclear the extent to which young herbivores evolve gustatory attraction towards plant defense compounds, the S. flava system showcases the importance of reduced aversion to bitter defense compounds, and how this may be achieved.