Abstract: Rapidly increasing temperatures threaten global ecosystems. Marine environments are some of the most vulnerable to future climate change and have particularly pronounced effects on habitat-forming species such as corals, seagrass, mangroves, oysters and kelp. While biodiversity loss across taxa is predicted to intensify under climate change, the potential effects of eco-evolutionary processes and environmental characteristics on large-scale species persistence and community diversity remain underexplored. In this study, we used an eco-evolutionary model to track the relative abundance and mean thermal optimum of 25 hypothetical metapopulations or ‘species’ defined by different combinations of thermal tolerance breadth and genetic variation. We simulated the response of each of these metapopulations to sea surface temperatures (SST) from 2010 to 2100 based on SST ranges and rates of projected SST increase in 44 large marine ecosystems (LMEs). We then used generalized linear models to quantify the relative association of metapopulation and environmental characteristics to abundance and species richness at 2100.
Across LMEs, the number of persisting metapopulations at 2100 ranged from 0 to all 25 species tested. Metapopulations with the narrowest thermal tolerance breadth consistently led to the highest final abundances across LMEs, likely because these species experienced strong selection that facilitated a rapid evolutionary response to rising temperatures. In contrast, the level of genetic variation that led to the highest abundances varied from intermediate to high values across regions, suggesting that populations can persist even at middling levels of genetic diversity depending on environmental conditions. We also found that the rate of SST increase was the strongest predictor of metapopulation abundance and species richness, and was inversely associated with those response variables under both low and intermediate emissions scenarios. A wider range of temperatures within the metapopulation was positively associated with final abundance, and may have been due to the dispersal of individuals from warmer to cooler patches facilitating evolutionary rescue in the latter. Overall, we conclude that LMEs with low rates of temperature increase and broad environmental gradients will likely support the greatest number of species. Furthermore, species with a narrow thermal tolerance breadth and intermediate to high levels of genetic diversity will be most likely to maintain abundance across different environmental conditions under increasing temperatures.
