Climate change facilitates fungal pathogen expansion while driving endemic host range contractions in a tropical biodiversity hotspot

Non-native plant pathogens are reshaping ecosystems globally, yet their virulence and spread potential under future climate and land-use change remains underexplored, particularly in biodiversity hotspots where endemic hosts may be highly vulnerable. In Madagascar, a vascular wilt pathogen (Leptographium calophylli, formerly Verticillium) has been increasingly observed infecting native forest trees. We modelled the future spread of this pathogen under changing climate conditions and directly compared projections to those of a susceptible endemic host tree, Calophyllum paniculatum, to assess overlap, divergence, and implications for extinction risk. Using an ensemble species distribution modelling approach (Random Forest, Boosted Regression Trees, MaxEnt), we forecasted potential distributional ranges for both pathogen and host using the Coupled Model Intercomparison Project (CMIP6) Global Climate Model (GFDL-ESM4) under three shared socioeconomic pathways (SSP1–2.6, SSP3–7.0, SSP5–8.5) and three time periods (2011–2040, 2041–2070, 2071–2100). Ensembles showed high predictive performance (AUC > 0.97), with precipitation seasonality and moisture availability in the driest month as key drivers of wilt distribution. The pathogen was predicted to retain 68.5–91.7% of the current projected distribution by 2100, with westward expansion into humid and sub-humid ecoregions, consistent with climate-facilitated invasion dynamics. In contrast, C. paniculatum was forecast to contract severely (65.9–97.0% range loss by 2100) and shift south-eastward, leading to a partial spatial decoupling. However, overlap with pathogen distributions persisted, indicating sustained mortality risk. Our results demonstrate that climate change may facilitate fungal pathogen expansion while simultaneously reducing the range of endemic hosts, an asymmetric dynamic that could intensify biodiversity loss in island ecosystems. Even without human mediated, long-distance dispersal, biological invasions can expose native species to prolonged biotic pressure. By explicitly modelling future host–pathogen interactions, this study highlights the importance of incorporating disease threats into biodiversity forecasts and conservation strategies, particularly in tropical systems facing rapid environmental change.