Study on mapping and assessment of pollinator habitats in rural areas, using an optimized Maxent model

Introduction Pollinators are vital for rural ecosystems and agriculture, yet their diversity is declining, threatening biodiversity and ecosystem stability. Method This study evaluates pollinator habitats in Pukou District, Nanjing, using an optimized Maximum Entropy (Maxent) model and landscape pattern analysis to map spatial distribution and assess village-level habitat suitability. By combining field survey data with multi-source database data, we optimized Maxent parameters and applied entropy-based weighting with K-means clustering to evaluate 66 villages. Results Results reveal a three-tiered gradient pattern (mountainforest core, agricultural transition, urban fringe), with 19 high-, 28 moderate-, and 19 low-suitability villages. Natural–artificial edges form synergistic hotspots, while roads support 69%–71% of habitats, and settlements suppress 33%–47%. High-suitability villages benefit from low patch density and high heterogeneity, unlike low-suitability villages with intensive agriculture. We propose core protection, ecological corridors, and community-based conservation, offering a framework for rural biodiversity management in urbanizing regions.