Rapid climate change results in long-lasting spatial homogenization of phylogenetic diversity

Climate stability is a key factor influencing biodiversity. Stable climates allow for species evolution and persistence, resulting in higher biodiversity and the presence of endemic species with limited ranges. In contrast, rapid climate change dramatically alters biodiversity patterns, favoring generalist species with high dispersal capabilities and leading to the loss of endemic specialists. Current rapid climate change forces species to track suitable habitats, creating high climate change velocity which significantly impacts future biogeography. While previous studies focused on temporal changes in phylogenetic diversity, less is known about the spatial impact of rapid climate change on phylogenetic turnover (phylo-β) across large regions. This study aims to understand how past rapid climate changes, specifically during the Quaternary, have affected spatial phylo-β in European seed plants. Europe serves as an ideal case study due to the significant impact of glacial cycles and climate oscillations on species distributions. The authors hypothesized that lower phylo-β would be observed in regions with higher past climate change velocity, increased distance to refugia, and larger mean species range sizes. Understanding these historical impacts is critical for anticipating the effects of ongoing climate change on species distributions and extinctions.

The study draws upon established knowledge linking climate stability to biodiversity. Studies have shown that climatically stable regions harbor high biodiversity with endemic and small-ranged species, leading to high species and phylogenetic turnover along environmental and geographic gradients. Conversely, rapid climate changes negatively impact biodiversity by favoring generalist species with larger ranges and strong dispersal capabilities, leading to the loss of specialists. The impact of climate change velocity on species endemism has been previously documented. Many studies project significant loss of evolutionary heritage due to future climate change using phylogenetic diversity, emphasizing the need for conservation efforts. However, few studies have examined the spatial patterns of phylogenetic diversity in relation to past rapid climate change. The authors draw on existing work highlighting the role of ecological and evolutionary processes (immigration, competition, speciation, extinction) in shaping biodiversity patterns, emphasizing the lineage-specific nature of these processes and how factors like environmental heterogeneity and landscape complexity impact spatial phylo-β. Previous research has shown that phylo-β increases with geographic and environmental distance and richness differences among sites.

The study uses seed plant distribution data from the Atlas Florae Europaeae (AFE), encompassing approximately 25% of European vascular plants. The data consists of presence/absence records for each species within 50 x 50 km grid cells across Europe. Phylogenetic data was obtained from the species-level plant megaphylogeny (PhytoPhylo) of Qian and Jin (2016), with analyses also conducted using a second phylogeny (Smith and Brown 2018) for robustness. The study area includes the entire European subcontinent excluding some smaller islands and eastern regions due to data limitations. Data cleaning involved standardizing species names according to The Plant List and handling synonyms and subspecies. The data was then divided into angiosperms and gymnosperms. Environmental data included 19 bioclim variables from Worldclim, spatially aggregated to 50 km resolution. Environmental distances between grid cells were calculated using PCA on standardized bioclim variables. Climate change velocity (Vocc) was calculated using a gradient-based approach from paleoclimate data providing 1000-year time steps since the Last Glacial Maximum (LGM). The Vocc data incorporated both the rate of climate change and elevation change. Climate stability (ClimStab) was also calculated based on the same paleoclimate data, using the standard deviation of climate difference between time slices and inversely relating it to stability. Potential LGM refugial areas were inferred for each species using the KISSMig model by identifying cells that were climatically suitable during the LGM and also resulted in meaningful predictions of current species distributions based on simulated migrations. Distance to refugia (DistRef) for each grid cell was determined as the mean shortest distance among all species present to their closest reconstructed LGM refugial cell. Mean species range size was calculated as the average number of occupied AFE points across Europe for all species in a grid cell. Phylogenetic beta diversity (phylo-β) was calculated using the Simpson's pairwise dissimilarity index between each grid cell and its 24 nearest neighbors. To remove local geographic and environmental distance effects, a generalized linear model was used, and the residuals were interpreted as the larger-scale variation in phylo-β. Finally, standardized linear regression models were used to examine the relationship between the residual phylo-β and the explanatory variables (range size, distance to LGM refugia, velocity of climate change, climate stability). Variable importance was assessed using the lmg method.

The study revealed a clear spatial structure in phylogenetic turnover, independent of regional environmental or topographic heterogeneity. Phylo-β was generally higher in Southern Europe and lower in Northern Europe, with some peaks and troughs in Central Europe. For angiosperms, a strong model fit (R² = 0.49) was observed, primarily explained by distance to LGM refugia (68%), velocity of climate change (18%), and species range size (14%). For gymnosperms, the model fit was slightly lower (R² = 0.36) with the contribution of distance to LGM refugia (43%), velocity of climate change (25%), and species range size (32%). Southern Europe exhibited higher species diversity and acted as a major refugium during the LGM. A negative relationship between phylo-β and distance to refugia was observed, indicating lower diversity further from the refugia. Similarly, lower phylo-β was found in areas with higher past climate change velocity. Conversely, climate stability and larger species range sizes were positively associated with higher phylo-β. Notably, some regions in Central and Northern Europe showed higher phylo-β, potentially indicating the presence of previously hypothesized northern refugia. The results were consistent across different phylogenetic trees and for both angiosperms and gymnosperms, suggesting robustness and broad applicability across clades. Removing local geographic and environmental distance effects revealed strong large-scale patterns linked to historical climate changes.

The findings strongly support the hypothesis that past rapid climate change has significantly homogenized phylogenetic diversity across large landscapes in Europe. The observed patterns are best explained by the interplay of climate change velocity, distance to refugia, and species range size. The homogenization in Northern Europe is likely due to incomplete range filling by species following the LGM, the dominance of generalist species with large ranges, and the extinction of specialists unable to migrate quickly enough. The negative correlation between phylo-β and distance to refugia highlights the influence of historical processes. The identification of regions with unexpectedly high phylo-β in Central and Northern Europe, consistent with the locations of proposed northern refugia, provides further evidence for the existence of such refugia. The findings confirm the long-lasting impact of Quaternary climate oscillations on current biodiversity patterns. The spatial homogenization observed in this study suggests that future climate change may result in an even more dramatic loss of regional phylogenetic diversity than previously projected by studies focusing solely on local temporal changes.

This study demonstrates the significant and long-lasting impact of Quaternary climate oscillations on the spatial patterns of phylogenetic diversity in European seed plants. The homogenization observed in northern and central Europe highlights the importance of past climate change velocity and distance to refugia in shaping current biodiversity. The results strongly suggest that future, even more intense climate change will likely lead to an even greater loss of phylogenetic diversity through spatial homogenization. This underscores the critical need for conservation efforts focusing on preserving both local and regional diversity, especially in areas currently exhibiting high phylogenetic turnover.

The study relies on distribution data from the AFE, which represents only about 25% of European angiosperms. While analyses using alternative phylogenetic trees and consideration of both angiosperms and gymnosperms demonstrated robustness, future research could benefit from the inclusion of a more comprehensive dataset. Furthermore, the study focuses on seed plants, limiting the generalizability to other taxonomic groups. While the authors accounted for local geographic and environmental distance effects, other factors not explicitly considered here, such as biotic interactions, might also influence phylogenetic turnover. The use of coarse-resolution distribution data could mask finer-scale patterns. Finally, modeling the future impact of climate change was outside the scope of this study and represents a key future research direction.