Integrating Leaf-Microbial Interactions and Leaf Water Content to Understand Leaf Selection by Leafcutter Bees

Abstract

Introduction: Leafcutter bees (Megachile spp.) are a cosmopolitan group of solitary bees that are important pollinators for agricultural crops and wild plants. These bees make brood nests lined with semi-circular pieces cut from leaves selected from various plant species. The physical and chemical characteristics of leaves selected for nesting should be important in supporting the survival and health of larvae. Surprisingly, little is known about leaf selection: why are certain leaves chosen by leafcutter bees while others on the same plant are not? This study examined whether antifungal chemical properties, water content, or the presence of endophytic symbionts may be factors why leafcutter bees choose leaves for nests. I hypothesized that leaves cut by Megachilids show more antifungal activity than uncut leaves to provide better protection of larvae from fungal infection. Along with this hypothesis, it was further predicted that leaf water content of leaves selected by Megachilids would have lower water content compared to leaves not selected and that leaves selected would have fewer culturable endophytes. Methods: The study was conducted in Tucson, Arizona, in the Sonoran Desert biome of the southwestern USA. My study included three introduced plant species that are used by leafcutter bees: Nicotiana glauca (Solanaceae), Punica granatum (Lythraceae), and Rosa sp. (Rosaceae). I collected leaves from each species, including those cut by bees and those on the same plants that were not cut by bees in the same timeframe. I evaluated in vitro the growth of representative fungi in response to leaf discs from these cut and uncut leaves. I measured water content of these leaves by comparing fresh mass to dry mass. To test for the presence of endophytic symbionts, discs from cut and uncut leaves were surface-sterilized, placed on a standard growth medium, and observed over two weeks. Results: I found no meaningful differences in water content between cut and uncut leaves. Culturable endophytes were not common in these leaves. However, I observed a significant difference in antifungal activity: contrary to expectations, more enhanced growth of focal fungi was observed to and on leaves that were cut when compared to uncut leaves. Conclusion: Leaves cut by leafcutter bees had more enhanced growth to and on fungi in vitro than uncut leaves. This result was unexpected, as I anticipated that leaves cut by bees would have more antifungal activity than uncut leaves, potentially protecting larvae from fungal infection during development. My finding raises the possibility that cut leaves may attract potentially beneficial fungi, including the leaf-symbiotic fungi used in my assays, but exclude potentially harmful fungi not evaluated here. In future work I will examine the prediction that the fungi attracted to cut leaves have a greater capacity to inhibit insect-pathogenic fungi than fungi that are not attracted to cut leaves. It is also possible that fungal infection is less a threat to leafcutter bees than kleptoparasitism, such that the affinity of cut vs. uncut leaves to parasitic insects should be investigated in future studies. Finally, it is plausible that microbes not considered here -- such as epiphyllous microbes -- could influence bees’ selection of leaves, as illustrated in ongoing work inspired in part by this thesis (Pryor et al., 2019).