Combining ensemble and Bayesian distribution models to guide conservation planning for a narrow‐range endemic skink on ultramafic substrates in New Caledonia
Year: 2026
Type: article
Abstract: Conservation planning for squamates is frequently constrained by their cryptic ecology and by the rare species conservation paradox, whereby limited distribution data restricts the development of effective management strategies. We focus on Lacertoides pardalis , a narrow endemic skink restricted to ultramafic substrates in New Caledonia, where suitable habitats are increasingly fragmented by nickel mining and recurrent wildfires. To overcome data limitations, we combined ensembles of small models with Bayesian additive regression trees to model habitat suitability using a parsimonious set of bioclimatic variables and a remotely sensed vegetation index (NDVI). We then spatially overlaid predicted suitable habitats with mining concessions, fire history, and protected areas to assess exposure to major anthropogenic pressures. Our results reveal a highly restricted and fragmented potential distribution, with a substantial proportion of suitable habitats overlapping active mining areas and fire‐prone landscapes, while remaining poorly represented within protected areas. By explicitly linking species distribution modeling with spatial threat analysis, this study provides decision‐relevant information to support conservation planning and land‐use management for highly range‐restricted taxa in complex landscapes.
DOI: 10.1111/csp2.70466
Institutions: Centre National de la Recherche Scientifique, Aix-Marseille Université, Institut de Recherche pour le Développement, Institut Méditerranéen de Biodiversité et d'Ecologie Marine et Continentale, Institut de Recherche pour le Développement, University of New Caledonia, Ifremer, University of Reunion Island, Écologie Marine Tropicale des Océans Pacifique et Indien
Language: English
Subfield: Ecological Modeling
Field: Environmental Science
Domain: Physical Sciences
Keywords: skinks, ultramafic substrates, New Caledonia, conservation planning, species distribution modeling, ensemble learning, Bayesian additive regression trees, habitat suitability, habitat fragmentation, nickel mining, forest fires, protected areas, NDVI, bioclimatic variables, land use management
SDG: Life on land
Open Access status: gold
Estimated APC paid: $2,240
OpenAlex ID: W7232301966
