Climate-induced range overlap among closely related species

Contemporary climate change is causing large shifts in biotic distributions1, which has the potential to bring previously isolated, closely related species into contact2. This has led to concern that hybridization and competition could threaten species persistence3. Here, we use bioclimatic models to show that future range overlap by the end of the century is predicted for only 6.4% of isolated, congeneric species pairs of New World birds, mammals and amphibians. Projected rates of climate-induced overlap are higher for birds (11.6%) than for mammals (4.4%) or amphibians (3.6%). As many species will havedi culty tracking shifting climates4, actual ratesof future overlap are likely to be far lower, suggesting that hybridization and competition impacts may be relatively modest. Widespread changes in species distributions due to climate change are documented for diverse taxa and are expected to become more pronounced over the coming century as rates of warming increase1. One expected outcome of climate change-induced range shifts is the establishment of geographic range overlap among previously isolated taxa, leading to novel species interactions and assemblages5,6. The potential for climate change to result in new interactions among closely related species has given rise to conservation concern, as these may have negative consequences for species persistence. Climate-induced range contact between ecologically similar species may introduce high levels of interspecific competition to populations already stressed by changing climatic conditions7,8. In addition, recently diverged species with incomplete reproductive barriers may hybridize, reducing population fitness through genetic admixture or leading to species extinctions through asymmetric hybridization9,10. Although few studies have empirically documented climate-induced contact among closely related species2, many have expressed concern that it could lead to a significant loss of biodiversity3,11. Despite potential for negative impacts, no attempt has yet been made to estimate future rates of climate-induced geographic overlap among previously isolated, closely related species. We used bioclimatic models to predict potential end-of-century (2071–2100) areas of climatic suitability for 9,577 congeneric species pairs, including NewWorld birds (n=3,858), mammals (n=1,661) and amphibians (n= 4,058). From this data set, we calculated the number of nonoverlapping (that is, allopatric), congeneric species pairs with ranges projected to come into contact (that is, sympatry) in the coming century. We accounted for variability among estimates by including in our results only species pairs projected to come into contact under a majority (>5) of 10 general circulation models (GCMs). We found that 6.4% of 4,796 allopatric species pairs are projected to come into geographic contact by the end of the century (Fig. 1). Rates of future contact for species pairs were significantly greater for birds than mammals or amphibians (generalized linear mixed model, F1,4781=8.54, P<0.0002), for tropical than temperate species (F1,4781= 5.21,P < 0.0055), and increased with current geographic range size (F1,478=11.55, P<0.0007). Our finding that future range overlap is more than twice as common for bird species pairs (11.6%) than mammals (4.4%) or amphibians (3.6%) cannot be explained by higher dispersal rates in birds, as our models assume equal dispersal capacities across taxa. Rather, it is best explained by the larger range sizes of birds (analysis of variance (ANOVA), F2,402= 9.897, P < 0.001) and the positive relationship between current range size and rate of future overlap (Fig. 2). For most newly overlapping species, projected areas of future overlap span a relatively small (<25%) percentage of their future bioclimatic envelope (Fig. 3). However, this percentage is negatively correlated with the size of a species’ future bioclimatic envelope (R2=0.358, d.f.=1,614, P<0.001; Fig. 4), which is itself highly correlated with current range size (R2=0.543, d.f.=1,614, P<0.001). Thus, species with smaller ranges, now and in the future, may have the greatest potential for negative impacts, because higher proportions of their future ranges are expected to overlap with that of a congener, and because smaller-ranged species are particularly vulnerable to the negative effects of overlapwith closely related species9. Our finding that most areas of future overlap are in the tropics (85.8%) can be explained by current species distributions (Fig. 1): 86.2% of species pairs include a tropical species. Additionally, future bioclimatic envelopes of pairs that include tropical species are projected to have greater overlap (ANOVA, F2,305=3.231, P=0.041; Table 1) and size asymmetry (ANOVA, F2,305= 27.220, P < 0.001; Table 1) than pairs with only temperate species, both of which may exacerbate the potential for negative impacts9. Thus, both the occurrence and impact of climate-induced range overlap between closely related species may be greatest in the tropics. As newly overlapping species do not overlap in current climate space, future areas of overlap must be interpreted as having no-analogue climates. Caution must be used when extrapolating future species distributions from no-analogue future climates because it is difficult to predict how species will respond to new environments5,12. Thus, although our results may estimate the proportion of species pairs that are likely to come into geographic contact (that is, parapatry), whether species are able to occupy