Understanding the geographic distribution of life and its connection to physiological thermal limits is crucial for predicting the impact of climate change on biodiversity. Species often retain ancestral climatic affinities (niche conservatism), potentially limiting their ability to adapt to rapidly changing climates. The ability to tolerate cold temperatures varies greatly across species, clades, and geographic locations, while heat tolerance shows striking invariance across latitudes, elevations, and phylogeny. This counter-intuitive pattern, considering the generally left-skewed thermal fitness/performance curves, suggests underlying mechanisms shaping thermal tolerance evolution. Three non-mutually exclusive mechanisms are proposed: 1) "deep-time climate legacies," where ancestral climatic affinities influence current thermal limits; 2) "physiological boundaries," where physiological limitations constrain adaptation beyond certain temperatures; and 3) "adaptation to current climatic extremes," where current selective pressures shape thermal limits. This study uses a large dataset of thermal limits to investigate the relative roles of these mechanisms in shaping global variation in thermal physiological limits, acknowledging that species distributions can shift through time to remain within a thermal niche.

Previous research has highlighted the variability in cold tolerance and the relative invariance in heat tolerance across various taxa and geographical locations. Studies have suggested the role of niche conservatism, where species retain their ancestral climatic preferences, limiting their adaptation to new climates. Other studies have explored the possibility of physiological boundaries limiting the extent of thermal tolerance adaptation. However, a comprehensive analysis integrating these factors and considering the tempo and mode of thermal limit evolution across a large dataset of diverse taxa was lacking before this study.

This study utilized the largest existing dataset of thermal limits, encompassing over 2000 terrestrial and aquatic species. The researchers investigated three hypotheses: deep-time climate legacies, physiological boundaries, and adaptation to current climatic extremes. To investigate deep-time climate legacies, they determined the thermal ancestry of each species based on paleoclimatic conditions at the time of its order's origin. Evolutionary age of each species' order was determined to assess evolutionary constraints (physiological boundaries hypothesis). Current thermal regimes experienced across species' ranges were used as proxies for adaptation to current climatic extremes. The researchers used phylogenetic analyses to investigate evolutionary patterns in upper and lower thermal limits, comparing the fit of Ornstein-Uhlenbeck (OU), Brownian Motion (BM), and White Noise (WN) models. Random forests were used to compare the relative importance of thermal ancestry, evolutionary age, and current climate in determining thermal tolerance variation.

The study's key findings include: 1) Adaptation to current climatic extremes is the strongest determinant of species' thermal tolerance limits. Both lower and upper thermal limits increased with current environmental temperatures across all taxonomic groups, though the effect was stronger in ectotherms and plants than in endotherms. 2) A deep-time climatic legacy is evident in the cold tolerance of ectotherms, where species from orders originating in glacial periods showed lower cold tolerance than those from warm periods. This pattern was less apparent in endotherms and aquatic plants. 3) Lower thermal limits evolved faster than upper limits across taxa, supporting the existence of physiological constraints on the tempo of evolution of upper thermal limits. This asymmetric rate was more pronounced in endotherms than in ectotherms or plants. 4) The evolution of thermal limits better fits an OU model than BM or WN models, suggesting directional selection towards attractor values, particularly for upper thermal limits in endotherms and terrestrial plants. 5) Random forest analysis indicated that current environmental temperatures, along with clade age, were significant predictors of thermal tolerance limits, with paleoclimate playing a minor role, primarily affecting cold tolerance in ectotherms. The OU model fits suggest either stabilizing selection or phylogenetic niche conservatism, but also possibly indicate directional selection acting together with physiological barriers constraining the evolution of thermal tolerances.

The findings highlight the interplay between adaptation to current climate and evolutionary constraints in shaping thermal tolerance. The strong influence of current climate suggests that species are adapting to contemporary temperature extremes, but the slower evolution of upper thermal limits and the OU model fits indicate physiological boundaries or an optimum, constraining adaptation to higher temperatures. The deep-time climate legacy effect in ectotherms highlights the long-term influence of past climates on current thermal tolerance. The faster evolution of cold tolerance compared to heat tolerance suggests that adaptation to colder temperatures might be more readily achieved, possibly due to fewer biophysical constraints.

This study demonstrates the complex interplay between current climate adaptation and evolutionary constraints in determining species' thermal tolerance. The limited adaptive potential for upper thermal limits, coupled with the rapid pace of climate change, raises concerns about the ability of many species to survive future warming. Future research could explore the specific physiological mechanisms limiting heat tolerance adaptation and investigate the potential for evolutionary rescue in different taxa and environments.

The study's limitations include potential biases in data availability, with some taxa being better represented than others, which may affect the generalizability of findings. The phylogenetic analyses rely on certain assumptions about the evolutionary processes, and alternative interpretations of the OU model fits are possible. The study focused on critical thermal limits and did not address other aspects of thermal physiology that might affect species' responses to climate change.