Charles C. Carey

Bees are important plant pollinators in agricultural and natural ecosystems. High average annual losses of honey bee (Apis mellifera) colonies in some parts of the world, and regional population declines of some mining bee species (Andrena spp.), are attributed to multiple factors including habitat loss, lack of quality forage, insecticide exposure, and pathogens, including viruses. While research has primarily focused on viruses in honey bees, many of these viruses have a broad host range. It is therefore important to apply a community level approach in studying the epidemiology of bee viruses. We utilized high-throughput sequencing to evaluate viral diversity and viral sharing in sympatric, co-foraging bees in the context of habitat type. Variants of four common viruses (i.e., black queen cell virus, deformed wing virus, Lake Sinai virus 2, and Lake Sinai virus NE) were identified in honey bee and mining bee samples, and the high degree of nucleotide identity in the virus consensus sequences obtained from both taxa indicates virus sharing. We discovered a unique bipartite + ssRNA Tombo-like virus, Andrena-associated bee virus-1 (AnBV-1). AnBV-1 infects mining bees, honey bees, and primary honey bee pupal cells maintained in culture. AnBV-1 prevalence and abundance was greater in mining bees than in honey bees. Statistical modeling that examined the roles of ecological factors, including floral diversity and abundance, indicated that AnBV-1 infection prevalence in honey bees was greater in habitats with low floral diversity and abundance, and that interspecific virus transmission is strongly modulated by the floral community in the habitat. These results suggest that land management strategies that aim to enhance floral diversity and abundance may reduce AnBV-1 spread between co-foraging bees.

Common genetic variants buffered by Hsp90 are candidates for human diseases of signaling such as cancer. Like cancer, morphological abnormalities buffered by Hsp90 are discrete threshold traits with a continuous underlying basis of liability determining their probability of occurrence. QTL and deletion maps for one of the most frequent Hsp90-dependent abnormalities in Drosophila, deformed eye ( dfe ), were replicated across three genetically related artificial selection lines using strategies dependent on proximity to the dfe threshold and the direction of genetic and environmental effects. Up to 17 dfe loci (QTL) linked by 7 interactions were detected based on the ability of small recombinant regions of an unaffected and completely homozygous control genotype to dominantly suppress or enhance dfe penetrance at its threshold in groups of isogenic recombinant flies, and over 20 deletions increased dfe penetrance from a low expected value in one or more line, identifying a complex network of genes responsible for the dfe phenotype. Replicated comparisons of these whole-genome mapping approaches identified several QTL regions narrowly defined by deletions and 4 candidate genes, with additional uncorrelated QTL and deletions highlighting differences between the approaches and the need for caution in attributing the effect of deletions directly to QTL genes.