Vascular plant extinctions outpace colonizations under climate change

Abstract

Climate change is expected to shift plant distributions poleward and upslope, yet requires global empirical evidence. Using historical range shifts from global herbarium specimens for 109,242 vascular plant species over the past 28 years as baseline, we fit 1.8 million species distribution models on a global dataset of 184,545 vascular plant species to project future range shifts over an equivalent period, the next 20 years (2021-2040). We find that plant species have already shifted their distribution centroids at a median velocity of 5.6 km per year globally, exceeding projected future velocities of 3.27 km yr per year. Projected shifts will vary widely in direction and magnitude, with varied velocities, colonization and extinction rates. Shifts will vary across clades with Superasteridae, Gymnosperms and Monocots moving at slower velocities, and Ferns, Chloranthales, and Magnoliidae with faster velocities. Although most clades will shift polewards, velocities will be highest in temperate and boreal regions, while shifts will be slower in regions of high topographic variation in the tropics. High latitude species will trend poleward and upslope, while desert shrubs, and tropical conifer forests will show slower and directional shifts. Extinctions will outpace colonizations, confirming that climate change will be a driver of biodiversity loss. Colonization will be frequent in protected areas, higher elevations, and among threatened species whereas extinctions will cluster near species historical ranges and at lower elevations. Our findings show the complexity of climate-driven range shifts and emphasize the importance of considering multiple interacting processes beyond simple latitudinal or elevational movements.