On climate change and genetic evolution in Arabidopsis thaliana

Global climate change is already impacting Earth’s biodiversity, but we are still struggling to understand which species will perish and which will thrive. As many species will not tolerate a rapidly-changing climate nor migrate fast enough to escape it, survival will depend on whether populations are able to genetically adapt. Some species, however, seem to rapidly adapt and spread in the new status quo of human-dominated ecosystems. We are just beginning to understand the genomic footprints of past adaptation to climates and how this has prepared populations for future rapid adaptation, but many questions still need to be answered. Furthermore, evolution and adaptation knowledge is rarely integrated into predictive biodiversity models, even though that would increase the accuracy of predictions and help design better conservation strategies. Here I aim to tackle those challenges using the mustard-related plant Arabidopsis thaliana, for which there are public genomic sequences, geographic information, and seed collections of thousands of individuals. Chapter One was my first approach to understand how populations of the same species might respond to climate change. I examined survival of 220 natural Arabidopsis thaliana lines whose genomes are known to a simulated extreme drought in the greenhouse. Severe droughts are being forecast as some of the most drastic threats for plant communities as a consequence of global change. Extending the use of environmental niche models in combination with genome-wide association techniques, I found the hotspots of adaptive variants are primarily at the North and South margins of the species’ distribution range. The populations at those areas, that live in more extreme environments, will perhaps become reservoirs of adaptive variation under future, more hostile climates. In Chapter Two, I carried out a large-scale field experiment to directly quantify climate-driven selection in natural conditions. We planted a global panel of 517 natural A. thaliana lines in rainfall-manipulated common gardens both in a region with a moderate climate, in Central Europe, and in a region with a more extreme environment, the Mediterranean. Using image analysis to estimate reproductive success, I generated close to 25,000 fitness measurements. Combining fitness and genomic data, I could infer massive changes in genome-wide allele frequencies within one generation, especially under hot temperatures and reduced precipitation where many Central European genotypes died. Integrating the theory of local adaptation with machine learning tools, I showed that a significant portion of natural selection is predictable from the climate at the geographic areas where genetic variants are found. Following a decrease in rainfall in the future, I then predicted that the intensity of natural selection will increase the most in transition areas from the Mediterranean to Central Europe, putting populations at evolutionary risk. This is in stark contrast to the generally accepted notion that marginal “warming” populations are at higher risk of extinction than populations at the center of the geographic distribution. Chapter Three, in contrast to the previous chapters that studied the adaptive value of pre-existent variants to future climate change, focuses on how novel mutations could directly contribute to adaptation. Using herbarium samples as genetic snapshots in time, I studied a 400-year-old lineage of A. thaliana that was isolated in North America. I was able to identify over 5,000 new mutations, some of which generated novel morphological differences likely related to adaptation to the newly colonized continent. I concluded that even large organisms such as plants might evolve and adapt from new mutations in contemporary timescales. This work advances our knowledge on how and whether different populations of a species will genetically adapt to the changing climate. Some of the insights generated here include (1) that adaptation to climate occurs thanks to hundreds of genetic variants (polygenic adaptation), (2) that new mutations occur often enough that they could contribute to rapid adaptation in colonizing populations, and (3) that statistical models that learn the relationship between current climates and genetic variants can be used to predict whether populations will have the appropriate genetic makeup to adapt to climate change or whether they will be at evolutionary risk. All in all, these studies move us one step closer to address ecological challenges using the genetic theory of evolution