Microbial Mediation of Leafcutter Bee-Plant Interactions

Abstract

Nest building is an important component of reproduction for many animals. Features of nests, including location, architecture, and construction materials, are expected to be critical to fitness and therefore under strong selection. However, other than in birds, the factors that affect animals’ decisions during nest building are rarely studied. These decisions may have consequences not only for the individual building the nest but also for other members of the community, especially when nests are built from material taken from living organisms. Like other antagonistic interactions, interactions between nest-building animals and the species they use as resources for nest material may be affected by biotic context, including the presence of microbes. Indeed, microbes may be particularly important in these interactions, given evidence that nest microbial communities can impact animals’ fitness. In my dissertation, I studied the factors that affect choices about plant-derived nest material made by leafcutter bees (Megachile spp., Megachilidae), which line their brood cells with discs cut from living leaves. Leafcutter bees have been observed to cut particular plant individuals and leaves more heavily than others. Leafcutter bees’ use of leaves as nest material makes them antagonists of plants, and brings leaf-associated microbes into prolonged contact with their developing offspring. My dissertation research explores how leafcutter bees make decisions about which leaves to cut, including how leaf chemistry and microbiome affect these decisions, as well as how leafcutter bee cutting affects plant-pathogen interactions. Using cultivated rose plants (Rosa × hybrida, Rosaceae), I first tested the hypothesis that leafcutter bees take into account previous decisions of other beeswhen deciding to cut a leaf, with the predictions that they would behave differently on cut and non-cut leaves and that experimentally cutting leaves would lead to significantly more cuts. I found no evidence consistent with either of these predictions, suggesting that clustering of leafcutter bee cuts on particular leaves is not sufficiently explained by positive feedback. Rather, it may be driven by characteristics of the leaves. I next tested whether leaf surface microbial communities and/or chemistry could explain patterns of leafcutter bee cuts. I first predicted that if either of these factors affected cutting, they would differ between cut and non-cut leaves. I surveyed the composition of the metabolome and the bacterial and fungal communities associated with cut and non-cut rose leaflets. I found evidence consistent with this prediction for microbial community composition, but not for chemistry. To test whether differences in microbial communities cause differential cutting, I inoculated leaves with several common leaf-associated microbes, and compared cut accumulation between inoculated and sham-inoculated leaflets. Fungi in the genus Aspergillus tended to increase cutting when added to leaves. I also tested one hypothesis for this possible preference for leaves hosting Aspergillus: that it protects leafcutter bee larvae from pathogens in the nest. I grew Aspergillus with several insect pathogens. Consistent with my hypothesis, I found that it was able to inhibit their growth. Finally, I explored how leafcutter bee cutting affects plant-pathogen interactions. I found that leaflets of Clitoria mariana (Fabaceae) cut by leafcutter bees had more pathogen damage than non-cut leaflets. To test whether cutting causes increased pathogen damage, and how both wounding and introduction of microbes from bees’ bodies affect pathogen damage, I experimentally cut and/or inoculated C. mariana leaflets. While experimentally cutting leaflets increased pathogen damage relative to controls, experimentally cutting and inoculating leaflets with microbes from bees’ bodies did not, suggesting that these two components of leafcutter bee cutting have opposite effects on pathogen damage. Overall, my dissertation research shows that microbes, including plant pathogens, bee pathogens, and non-pathogenic leaf-associated fungi, contribute to the mediation of interactions between leafcutter bees and the plants they use for nest material. Understanding the role of microbes in the species interactions involved in nest building allows us to more accurately predict how this critical element of reproduction will evolve or respond to perturbations.