Separating Fundamental from Realized Habitat: A Transferable Dual-Model Framework for Quantifying Anthropogenic Constraints on Species Distributions in Human-Dominated Landscapes

Résumé fourni par la source

Conservation planning for large-bodied species in human-dominated landscapes is undermined by a persistent error in species distribution models (SDMs): the conflation of fundamental environmental suitability with realized habitat availability constrained by anthropogenic barriers. This leads to systematic overestimation of accessible habitat, misallocation of restoration priorities, and ineffective targeting of mitigation efforts. A transferable dual-model framework is developed to resolve this limitation by pairing an environment-only SDM (Ideal Model, representing the fundamental niche) with a model incorporating anthropogenic predictors (Full Model, approximating the realized niche). The pixel-wise difference between these outputs yields a spatially explicit anthropogenic constraint surface with direct management relevance. Applied to Asian elephants across a 2,607 km² landscape in south-central Bhutan, the Full Model significantly outperformed the Ideal Model (AUC = 0.896 ± 0.014 vs. 0.848 ± 0.017), demonstrating that anthropogenic variables capture critical distributional constraints absent in environmental predictors alone. Anthropogenic pressures reduced highly suitable habitat by approximately 57 km² (~10%) and degraded an additional ~410 km². Human–elephant conflict was concentrated in high-quality habitat (mean suitability = 0.733 vs. 0.197 landscape-wide), indicating that conflict arises from overlap in productive landscapes rather than displacement into marginal areas. The framework produces a three-tier conservation triage: restoration targets (high Ideal, low Full), mitigation priorities (high both, with conflict), and low-priority areas (low both). This approach is broadly applicable to wide-ranging species with available occurrence and anthropogenic data. Separating environmental suitability from anthropogenic feasibility is essential for effective, evidence-based conservation planning in increasingly human-modified ecosystems.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.