Primers, site data, and virus abundance in honey bees and Andrena spp.

Bees are focal pollinators, essential for maintaining biodiversity and crop production. Thus, reports of high annual honey bee colony losses and population declines among many wild bees in different parts of the world are of major concern. The spread of viruses is highlighted as a potential threat to bee communities. Viruses infect a wide range of bee species, and can be transmitted interspecifically through shared floral resources. Therefore, the role of flowers as hubs of bee virus transmission requires a community ecology perspective. Here, we investigate local and landscape-scale characteristics of floral communities potentially associated with the spread of viruses in the solitary Andrena spp. (mining bees). We surveyed 14 sites in a Mediterranean agroecosystem with varying local densities of honey bee (Apis mellifera) foragers and diversity of flowering species, and assessed the prevalence of four common Hymenoptera-associated viruses [deformed wing virus (DWV), black queen cell virus (BQCV), sacbrood virus (SBV) and Lake Sinai virus-2 (LSV2)] in co-foraging honey bees and mining bees. We found that the probability of virus presence in mining bees was generally associated with the diversity and composition of the local (site level) floral community, and with floral resource availability at the landscape scale (up to 1000 m range). In addition, SBV and DWV prevalence in mining bees was positively related to the density of SBV-infected, and total honey bee foragers, respectively. These findings demonstrate the focal role that the floral community at multiple spatial scales, and co-foraging pollinator species, may play in virus spread and, potentially, pollinator health.

Methods
We conducted a field survey during spring 2018, which included a total of fourteen sites. Each site comprised a 625 m2 plot located at the center of a larger wild bloom patch (mostly >3000m2). Field work was conducted in mid-March, corresponding to peak spring bloom, and lasted two weeks. Each site was sampled once between 8:00 and 16:00 when weather conditions were favorable for bee activity (temperature >17 °C, wind velocity <3 m/s, and clear or partially clear skies). Foraging bee sampling was carried out for a total of 40 person-minutes (excluding handling time of captured bees) and conducted by slowly walking throughout the plot and hand-netting any bee that was observed landing on a flower, while keeping records of the visited flower species. Each captured bee was identified in-situ to the lowest taxonomic level possible, immediately placed in a vial, and kept on dry ice. We collected additional foraging bees as needed to attain a sample of at least 11 Andrena and 11 A. mellifera individuals per site for viral screening. At the end of each sampling day, all the collected specimens were brought to the lab and stored at –80 °C.  The sample of mining bees that was used for viral screening consisted of the three most dominant Andrena species (collectively comprising > 70% of the observed Andrena individuals in the study): A. ocraceohirta, A. aerinifrons levantina, A. urfanella, and a closely related morphospecies from the subgenus Truncandrena. From each site, virus identification efforts were focused on individual A. mellifera and Andrena (median=11). High throughput sequencing identified regional virus genome sequence variations for BQCV and LSV-2, which required utilization of new primer sequences (Daughenbaugh et al., 2021). The presence and abundance of four common hymenopteran viruses, namely BQCV, DWV, SBV, and LSV-2, was assessed using the procedure of RNA extraction followed by cDNA preparation, polymerase chain reaction (PCR) and qPCR, as described in Daughenbaugh et al. (2021).