Bumble bees are important pollinators that face threats from multiple sources, including agrochemical application. Declining bumble bee populations have been linked to fungicide application, which could directly affect the fungi often found in the stored food and gastrointestinal (GI) tract of healthy bumble bees. We test the hypothesis that fungicides impact bee health by disrupting bee–fungi interactions. We examined the interactive effects of the fungicide propiconazole and fungal supplementation on the survival, reproduction and microbiome composition of microcolonies (queenless colonies) using two species, Bombus vosnesenskii and B. impatiens. We found that in B. vosnesenskii, fungicide exposure decreased survival, while fungal supplementation mitigated fungicide effects. For B. impatiens, fungicide application had no effect, but fungal supplementation improved survival and offspring production. Fungicides reduced fungal abundance in B. vosnesenskii microcolonies, but not in B. impatiens, where instead fungal addition decreased fungal abundance (ITS copy number). Fungal composition varied between treatments but differently between bee species. In B. impatiens, fungal addition increased microbiome diversity. In B. vosnesenskii, the abundance of the pathogen Ascosphaera was negatively associated with survival, while the yeast Zygosaccharomyces was positively associated with survival. Our results highlight that bumble bee species differ in response to fungicides and in the nature of bee‐fungi associations. Fungicides can alter bee–fungi interactions with consequences for bee survival and reproduction, and exploring the mechanisms of such interactions, including interactions among fungi in the bee GI tract, may offer insights into bumble bee biology and conservation strategies.

For full methodology please see the manuscript associated with these data. Microcolonies (queen-less colonies) of two species of bumble bee (Bombus vosnesenskii and B. impatiens) were exposed to either nectar containing the fungicide propiconazole or control nectar for one week. After this, they were either given a collection of bee-associated fungi in their nectar, or given control nectar for the following 2-3 weeks. Microcolony survival over time was measured, as was consumption of both fungicide-treated nectar and fungi-treated nectar. For B. impatiens, offspring production was quantified. Following microcolony termination, workers from each microcolony were dissected to remove the gut, which was separated into two sections (crop and midgut+hindgut). Three samples were pooled for each microcolony and DNA was then extracted from each sample. Samples from colonies that microcolonies were sourced from were included in processing and analysis, as were pollen samples. These samples were sent for amplicon sequencing of the 16S and ITS regions to determine bacterial and fungal gut community composition (sequences available on the NCBI SRA - BioProject: PRJNA759617). Total fungal gut abundance was quantified using qPCR. Sequences were processed using DADA2, and were then arranged into phyloseq objects for community analyses.

Refer to the ReadMe files for information on the included data and script files. microcolony_ReadMe.txt provides details on the microcolony response datasets. sequences_ReadMe.txt provides information on how to access raw sequences and on sample data associated with 16S and ITS phyloseq objects. Script_ReadMe.txt provides information on the R scripts used to analyze data associated with this project.