Incorporating landscape context into species distribution models improves predictions for migratory shorebirds

Migratory shorebirds are declining worldwide due to coastal habitat loss, making the identification of important habitats a conservation priority. While species distribution models are widely used to map potential habitats, they typically rely on environmental conditions measured at individual locations and underuse the surrounding landscape context. Focusing on six species across the coastal zone of the East Asian-Australasian Flyway during the non-breeding season, we compare landscape-scale models with conventional models based on pixel-scale and distance predictors. Here we show that incorporating neighborhood variables describing habitat composition and configuration (area proportions, patch densities, and land cover diversity) improves predictive performance for most species. Key predictors include distances to tidal flats and inland seasonal wetlands, tidal flat area proportion and patch density, land cover diversity, and elevation. Predicted suitability shows weak positive correlations with observed abundance in most species, suggesting landscape context captures habitat attributes relevant to local abundance. These results indicate that landscape context improves shorebird habitat predictions. Incorporating landscape-context variables can improve habitat predictions for migratory shorebirds, based on landscape-scale species distribution models of six species in the East Asian-Australasian Flyway.