Affiliations: 1) University of Virginia, Charlottesville, VA; 2) University of Richmond, Richmond, VA
Keywords: h. selection; m. adaptation
Vinegar flies (Drosophila melanogaster) living in temperate regions experience strong fluctuations in the strength and direction of selection due to seasonality and evolve rapidly to track the changing fitness landscape. Recent work has shown the existence of genomic loci whose seasonal signal is replicated across continents. Despite these advances in identifying the presence of fluctuating selection and adaptation between seasons, we still lack a detailed understanding of the functional genetic architecture of seasonal adaptation. This gap in knowledge is further compounded by the fact that wild Drosophila populations experience demographic “boom-and-bust” cycles concomitant with seasonality, a process which may alter levels of standing genetic variation. We characterized both the genomic consequences of boom-and-bust cycles as well as the targets of seasonally varying selection. We accomplished this by leveraging temperature data combined with genomic sequences from densely sampled orchard populations collected within (June-November) and across growing seasons over multiple populations on two continents. Principal component analysis reveals temporal structure, and using simulation, we show that this temporal structure is consistent with cyclical winter population collapse. This temporal signal is observed genome-wide except for chromosome 2L, where allele frequencies of the cosmopolitan inversion Inv(2L)t drive genetic structure. We further investigated these patterns at Inv(2L)t by regressing the mean temperature, 30 days prior to sample collection, onto allele frequencies of each SNP using a generalized linear model. This analysis reveals an enrichment of loci responding to temperature clustered within Inv(2L)t. Linkage analysis suggest that the loci associated with temperature changes lie within the inversion but is not the inversion per se. Additional analyses show multiple mutations in high linkage at distances ~0.1 – 0.5 Mb, suggesting long range associations among seasonal loci inside the inversion. Enrichment analysis show an excess of non-synonymous mutations among the Inv(2L)t seasonal outliers. Overall, our data provides two insights about Drosophila seasonality. First, we show that severe demographic contractions and expansions are an important, and measurable, component of the seasonal dynamics of temperate Drosophilids. Second, our data supports the general hypothesis that supergenes are important drivers of adaptation in highly fluctuating environments.