In the past decades, studies related to the implications that a changing climate will have on biodiversity have increased. However, the effects of climate change at community level in tropical and subtropical latitudes are still poorly addressed. We analyzed the impact of climate change on local tree richness, and on the appearance of novel or disappearance of known tree community assemblages in Mexico, a megadiverse country with high levels of species richness and endemism. Community models were based on stacking single species distribution models for 1974 tree species, the 96% of the total tree community reported on the Mexican National Forest and Soil Inventory. For species distributions predictions in year 2070, we considered a ‘business as usual’ climate change scenario (RCP 8.5). To calculate the number of novel or disappearing community assemblages in the future we performed dissimilarity analyses.
Results/Conclusions
Our study predicted a general decrease on local tree species richness for 2070. Forests (temperate, tropical dry, and rainforest) are predicted to lose species in more than 60% of its territories. Deserts, however, will gain species on half of their areas. Regarding community composition, we found that 9.2% of Mexico is predicted to undergo disappearance, or extreme rarefaction, of known assemblages, while 5.4% of its territory will have novel assemblages in the future. Mediterranean California will be the most affected ecoregion, losing current assemblages in 86.9% of its area and having novel communities on 23.2% of its extension. Even when forecasted to lose species richness, forests showed the most stable community composition, with less than 5% of their territories losing current assemblages or having novel combinations of species.The local loss of species in the forests and gain of species in deserts points towards a savannization of the ecoregions as a response of trees to the climate change. Additionally, our results forecast the reshuffling of Mexican tree communities driven by change in climate on the next 50 years, a fact that could potentially affect associated animal communities through the modification of the ecosystems functioning. Our predictions should be taken with extreme caution, given that they do not take into account relevant processes such as biotic interactions, and the role of microclimates or of climatic extremes. However, they can be useful as they might indicate general trends at large scales and serve as a null model of community changes exclusively based on species’ climatic preferences.