A direct and highly informative way to study evolution and evolutionary adaptation is to quantify the process in action, and within the context of a natural environment. In collaboration with Paul Schmidt’s group at the University of Pennsylvania, we combine wild population sampling, highly-replicated field mesocosm experiments, and time-series phenotyping and genome sequencing to quantify patterns of rapid adaptation in Drosophila melanogaster.

Populations of D. melanogaster in the Northeastern United States expand exponentially from spring through summer as resources become abundant, decline during fall as abiotic conditions deteriorate and resources become scarce, before entering a period of arrested development during the winter. This process is associated with rapid and repeatable evolution of life-history, reproductive, and stress tolerance traits. Our first effort to characterize the genome-wide patterns of variation associated with this process in wild populations suggested that alleles at hundreds of independent loci fluctuate in frequency in accordance with seasonal patterns of abiotic, demographic, and phenotypic change (Bergland et al. 2014). We have since demonstrated the ubiquity of these genomic signatures, identifying parallel patterns of allele frequency shifts across geographically wide-spread populations of the species (Machado et al. 2021; part of the Drosophila Evolution in Space and Time Consortia, DestBio).

To study this process in a more controlled, yet still natural, manner, Paul Schmidt created an experimental system of highly-replicated, field mesocosms at the University of Pennsylvania. Through this, we are able to preclude confounding variables such as migration, and powerfully infer parallel and deterministic allele frequency change (taking advantage of our lab’s tools for high-accuracy, haplotype-informed allele frequency estimates from pooled DNA samples; Tilk et al. 2109). In a recent paper, we showed that patterns of adaptation occur on monthly, sub-seasonal intervals, and are underpinned by fluctuations in the selective environment whereby alleles favored during population expansion become selected against during population decline (Rudman et al. 2022). The most recent study showed that adaptation proceeds of (sub)generational timescales and shows patterns of pervasive, strong, and rapidly fluctuating selection at tens to hundreds of loci (Bitter, et al, 2023).

We have many ongoing projects using this system – exploring the pervasiveness of fluctuating selection, linking observed patterns of genomic and phenotypic change, and exploring how perturbations (e.g. via the addition of pesticides) alter adaptive dynamics

We anticipate the principles discovered in this system to transcend our focal species and seasonal environments, and be broadly applicable to patterns of adaptation in large populations undergoing demographic boom and bust dynamics.