Estimating the timing, rate, and magnitude of maladaptation under changing climates based on landscape genetics and realised niche limits

Joanne Bentley, University of Cape Town, Cape Town, South Africa, Alex L Pigot, Genetics, Evolution and Environment, UCL, London, United Kingdom, Cory Merow, Department of Ecology and Evolutionary Biology, University of Connecticut, Storss, CT and Christopher Trisos, African Climate and Development Initiative, University of Cape Town, Cape Town, MD, South Africa

Background/Question/Methods
Many recent studies have used landscape genetics approaches to identify populations vulnerable to future climate change, typically identifying a range of genes that show an adaptive response to one or more climate variables. However, most assessments have made comparisons of genetic change between two, often remote, time slices, usually between current conditions and conditions near the end of the century. While this approach demonstrates the magnitude of genetic change required to maintain current gene-environment optima, it fails to indicate the rate of change in genetic offset necessary in different locations across a species’ range or the amount of time available until genetic offset is predicted to begin. To address this, we develop a framework for modelling future genetic change in a more continuous way—the ‘climate genetic offset profile’—which combines generalised dissimilarity modelling of climate-associated genes with annual projections of future climate conditions across the geographic distribution of a species. We illustrate this approach in 81 populations of the forest tree species balsam poplar (Populus balsamifera), showing how projected time series of genetic offset can be used to explore the timing, rate, and magnitude of genetic change in populations across the species’ range.
Results/Conclusions
We found that the shape of the projected genetic offset profiles varied substantially across populations. Notably, among populations experiencing similar magnitudes of future genetic offset, we found remarkable differences in the rates of expected genetic offset, with some populations predicted to experience far more rapid genetic turnover than others. Similarly, there was also high interpopulation variability regarding the time until the initiation of genetic offset, which could be used to inform conservation prioritization. These results have implications for understanding the capacity for adaptation under climate change, as those populations rapidly perturbed from their adaptive optima are those that are less likely to respond in time, resulting in maladaptation. More generally, our results highlight the value of moving beyond projections of genetic offset based on single climate time slices to understand the dynamics of genetic responses to climate change in a more continuous way.