Climate-Informed Species Distribution Modelling and Conservation Planning for Threatened Endemic Plants of the Jos Plateau, Nigeria

The Jos Plateau montane ecosystem harbours a significant assemblage of endemic and threatened plant species whose distributions and conservation status have not been assessed in the context of projected climate change. This study develops climate-informed species distribution models for twenty-three endemic and near-endemic vascular plant species of the Jos Plateau, using an ensemble modelling approach incorporating eight modelling algorithms and climate projections from five global climate models under RCP 4.5 and RCP 8.5 scenarios for 2050 and 2080. Occurrence data were compiled from herbarium records, recent field surveys, and community ecological knowledge elicitation. Model performance was rigorously evaluated using spatially blocked cross-validation and tested for the influence of sampling bias through target-group background filtering. Projected habitat suitability changes were overlaid with existing protected area boundaries and land use maps to identify projected climate refugia, expansion zones, and extinction-prone areas. Results project net habitat loss averaging 52 percent across the species ensemble under RCP 8.5 by 2080, with seven species projected to lose more than 80 percent of current climatically suitable area. Highlands above 1,400 metres elevation emerge as the most critical climate refugia for the majority of endemic species, yet currently receive limited formal protection. Three endemic species are projected to gain suitable area in adjacent highland regions of Benue and Plateau States, suggesting opportunities for assisted range expansion. The study provides actionable prioritisation data for conservation planning on the Jos Plateau and recommends expansion of protected area coverage to encompass projected climate refugia identified by the ensemble models. Keywords: species distribution modelling, climate change, endemic plants, Jos Plateau, conservation planning.