Modeling future tree species distributions under climate change to guide restoration planning: Application to the Brazilian Amazon

Tiago de Oliveira Gonzaga Teixeira, Manon Villa, Rebecca Montemagni Almeida, Lucas José Mazzei de Freitas, Emira Cherif

Addressing climate change and biodiversity loss requires innovative approaches to ecosystem restoration. This study aims to (1) develop a statistical tool to predict species distribution shifts under future climate scenarios and (2) apply it to 30 key tree species in the Brazilian Amazon, a biodiversity hotspot increasingly threatened by deforestation and climate change.

Using MaxEnt, we modeled species distributions under three climate scenarios (optimistic, medium, and pessimistic) for 2040, 2070, and 2100, integrating bioclimatic and soil variables. The tool generates interpretable maps highlighting areas of stability, expansion, and contraction for each species.

Results reveal a dual effect of climate change: while some species may initially expand their ranges, long-term survival is uncertain, particularly for those with low ecological plasticity. Functional trait analysis identified three species clusters, emphasizing the role of wood density, phenology, and plasticity in species selection for restoration.

Our findings highlight the need to integrate predictive modeling with ecological knowledge to guide species selection and enhance restoration success. By prioritizing climate-resilient species, this approach supports ecosystem stability and long-term conservation goals. Strengthening collaboration between scientists and practitioners is essential to refining and scaling these decision-support tools for effective restoration efforts.