Rapid and transient evolution of local adaptation to seasonal host fruits in an invasive pest fly

The data and R scripts for our analyses are available at: https://github.com/nrode/NatPop2021 and in the Dryad Repository: https://doi.org/10.5061/dryad.2z34tmpqh.International audienceBoth local adaptation and adaptive phenotypic plasticity can influence the match between phenotypic traits and local environmental conditions. Theory predicts that environments stable for multiple generations promote local adaptation, whereas highly heterogeneous environments favor adaptive phenotypic plasticity. However, when environments have periods of stability mixed with heterogeneity, the relative importance of local adaptation and adaptive phenotypic plasticity is unclear. Here, we used Drosophila suzukii as a model system to evaluate the relative influence of genetic and plastic effects on the match of populations to environments with periods of stability from three to four generations. This invasive pest insect can develop within different fruits, and persists throughout the year in a given location on a succession of distinct host fruits, each one being available for only a few generations. Using reciprocal common environment experiments of natural D. suzukii populations collected from cherry, strawberry, and blackberry, we found that both oviposition preference and offspring performance were higher on medium made with the fruit from which the population originated than on media made with alternative fruits. This pattern, which remained after two generations in the laboratory, was analyzed using a statistical method we developed to quantify the contributions of local adaptation and adaptive plasticity in determining fitness. Altogether, we found that genetic effects (local adaptation) dominate over plastic effects (adaptive phenotypic plasticity). Our study demonstrates that spatially and temporally variable selection does not prevent the rapid evolution of local adaptation in natural populations. The speed and strength of adaptation may be facilitated by several mechanisms including a large effective population size and strong selective pressures imposed by host plants.Impact SummaryNatural populations often exhibit good “fit” to the environment they are in. However, environments change over time and space, and following change, the fit between a population and its environment may be poor. A question of long-standing interest is how do populations track changing environments to maintain performance? Two main mechanisms are known: (i) genetic shifts in the form of local adaptation, in which traits evolve over time through differences in performance of individuals harboring different genetic variants, and (ii) plastic shifts, or adaptive phenotypic plasticity, in which traits immediately change in response to environmental change. Adaptation is common when environments change over multiple generations, whereas plasticity is common when environments change over an individual's lifetime. However, it remains unclear whether plasticity or adaptation is more vital to maintaining performance when environments change at an intermediate pace. Drosophila suzukii is well-suited to evaluating the relative importance of plasticity and adaptation in response to an intermediate pace of environmental change. This invasive pest species experiences an environment that shifts every one to four generations as host fruits change over time and space. Here, we studied natural populations of D. suzukii collected from different hosts. Using reciprocal common environment experiments, we evaluated their performance on their source and alternative hosts. Drosophila suzukii populations were most fit on their source host, successfully tracking an intermediate pace of environmental change. We developed a statistical method to quantify the contributions of adaptive plasticity and local adaptation in determining performance. We found that performance was mainly maintained through local adaptation to each new host in succession. This study highlights that spatially and temporally variable selection does not prevent local adaptation and, on the contrary, illustrates how rapid the adaptive process can be. It also provides a novel statistical tool that can be applied to other systems