Elevational shift in seed plant distributions in China's mountains over the last 70 years

Aim: Significant changes in species elevational ranges in mountains have been repeatedly documented, yet the direction, magnitude and drivers of these shifts remain controversial. Presently, there is still lacking evidence about the general nature of species elevational range shifts in eastern Eurasia in response to anthropogenic climate change. By using historical specimen records and recent field observations for 735 seed plant species across 29 China's mountains, we assessed changes in species' elevational centroids and their drivers.Location: China.Time Period: 1950-2018.Major Taxa Studied: Seed plant species.Methods: The elevation records of all sampled occurrences in each mountain during the two time periods were estimated, and the null models were developed to test the sampling bias. Ecological niche models (ENMs) were used to evaluate the relative importance of climate factors in constraining each species distribution. Generalized linear models (GLMs) to test the relationships between the centroid elevational range shifts of species and different divers.Results: We found that 54% of the species shifting upward and 46% downhill. However, species' elevational shifts significantly differed among species and mountains. Herbaceous and lowland species moved upward faster than woody and high-elevation species. Species in temperate mountains and in mountains with taller elevational gradients moved upward, while species in subtropical mountains and in mountains with shorter elevational gradients moved downward. Precipitation changes experienced by species, species' climatic adaptations, several species' functional traits and mountain size all contributed to explain the magnitude of species' centroid elevational range shifts.Main Conclusions: Our results highlight complex biodiversity redistribution of seed plants across Chinese mountains, not necessarily conforming to the trend of species upward shifts in elevation. Changes in precipitation regimes may blur the simplistic assumption of isotherm tracking. This study fills an important geographic shortfall for our understanding of biodiversity redistribution under anthropogenic climate change.