Habitat-partitioning improves regional distribution models in multi-habitat species: a case study with the European bilberry

摘要: Modelling the spatial distribution of multi-habitat species is challenging since they show multi-dimensional environmental responses that may vary sharply through habitats. Hence, for these species, achievement realistic models useful in conservation planning depend on appropriate consideration habitat information model calibration. We aimed to evaluate role different types predictors, along with habitat-partitioning, improve inference, detect non-stationary across habitats and simulate impact sampling bias predictions. As a case study, we modelled occurrence plant bilberry (Vaccinium myrtillus) Cantabrian Mountains (NW Spain), where it represents basic trophic resource threatened brown bear capercaillie. used MaxEnt compare baseline approach calibrated topo-climatic variables against three alternative approaches using explicit based vegetation maps remote sensing data. For each approach, ran non-partitioned (all together) habitat-partitioned (one per habitat) evaluated performance, overfitting extrapolation. The highest performance was including predictors. lowest model, at cost achieving predicted fractional area. extrapolation success low, approach. Our results highlight are habitats, habitat-biased data resulting weak When modelling regional scale, recommend either or data, realism outputs its applicability planning.