Identifying the macro- and micro-evolutionary factors that create and maintain biodiversity is a central goal of evolutionary biology. It is clear that interactions between plant hosts and insect herbivores drive increased diversification rates in insects. Yet, there are also challenges to herbivory as a life history strategy. My dissertation research has focused on the evolution of herbivory and the adaptive leaps that must occur to survive in this trophic niche including behavioral, morphological, nutritional and the ability to detoxify a well-defended plant diet. This work is largely focused on adaptations in Scaptomyza flava, a representative of a lineage nested within the Drosophila that transitioned to herbivory ~10-15 million years ago.

First, I characterized the trophic niche of S. flava. We used natural history studies to determine if S. flava is a true herbivore or a cryptic microbe-feeder, given that the ancestral character state for the family Drosophilidae is likely microbe-feeding. We quantified oviposition substrate choice and larval viability across food types, including yeast media, decaying plants and fresh plants. We found that S. flava had a strong preference for fresh plants and did not lay eggs in alternate food options. Next, we described morphological traits related to feeding across putatively herbivorous and non-herbivorous drosophilids. We found an overall enlargement of the mouth hook of herbivores, with a reduction in the number of teeth along the hook. We also found no difference in gut length among Scaptomyza species studied. Next, we compared nitrogen isotope and sterol profiles to understand how nutrients were being integrated into the body. We found a reduction in the N15 isotope ratio in S. flava compared to non-herbivores, which is expected in a species with fewer trophic levels between itself and its food. Through these multiple lines of evidence, we confirmed that S. flava is an obligate herbivore of living mustard plants.

Next, I explored the relationship between S. flava and microbes that live in their gut. I tested the hypothesis that the gut microbiota of nascent herbivores can facilitate the initial switch to herbivory by providing detoxification services of plant defensive chemicals. This is called the gut microbial facilitation hypothesis. We first characterized the gut microbiome of S. flava. We found that the life-stage of the fly influenced the microbial community in the gut the most, with species also playing a role. We then focused on identifying bacteria that are able to detoxify the common plant defensive chemical phenethyl isothiocynate (PEITC) using the SaxA enzyme. We isolated bacterial strains from the S. flava gut that are both resistant to and can detoxify PEITCs using SaxA. Lastly, we tested whether bacteria with saxA conferred benefits to non-herbivores when challenged with PEITC compared to bacteria without saxA. We found that the functional contribution of detoxification by one bacterial species was not enough to confer an advantage. However, future directions look toward testing the gut microbial facilitation hypothesis in this powerful tripartite system in other contexts and with a representative gut community.

Finally, I focused on testing the pre-adaptation hypothesis, which posits that adaptations of herbivorous flies that enable them to detoxify plant chemicals may also enable them to detoxify other xenobiotic chemicals, like pesticides. We have taken a comparative look at cyp6g1 across related herbivorous and non-herbivorous drosophild species. Cyp6g1 encodes a Cytochrome P450 monooxygenase that has been implicated in both insecticide resistance and phytochemical detoxification. We used a branch-site test to uncover signatures of positive selection along the branch leading to a subset of the cyp6g1 copies in herbivores. We then focused on the seven functional cyp6g1 copies that S. flava has in its genome. We also discovered a P-transposon inserted in the intergenic region between cyp6g1b and cyp6g1c. Transposable element insertions cis of detoxification genes tend to be related to an increase in the ability to detoxify. We determined whether this transposon was present in several pooled datasets and found that it was present in all three of the focal populations which spanned two sites across two years. This finding indicates that at the scale that we investigated, the P-transposon is being maintained in this genomic location. Increases in expression of genes cis of transposon insertions can be one way detoxification increases are obtained, so we measured differential expression of cyp6g1 copies in response to plant defensive chemicals. We found an increase in expression of one copy (cyp6g1h), though not the copy directly cis of the P-transposon, indicating it is unlikely to have influenced this increase. Finally, we developed a homologous expression assay to measure the affinity of each Cyp6g1 copy across the clade to detoxify either indole glucosinolate derived plant defensive compounds or insecticides, with further results of this assay forthcoming.

Overall, I have worked to understand the major axes of adaptation enabling the evolution of herbivory (morphological, behavioral, nutritional and metabolism of plant toxins) with a main focus on detoxification either through endogenous adaptations or through interactions with gut bacteria.