Environmental DNA-derived indicators reveal species turnover and connectivity-related ecological patterns across patch-corridor-matrix framework

Habitat fragmentation alters biodiversity patterns and ecological connectivity across agricultural landscapes. Here, we investigated how Patch-Corridor-Matrix (PCM) landscape components influence animal community structure within a vineyard-dominated landscape in Southern Italy. Biodiversity was assessed using leaf-surface environmental DNA (eDNA) metabarcoding, integrating arthropod and vertebrate detections with multivariate analyses, beta-diversity partitioning, indicator species analysis, and co-occurrence networks. A total of 27 sampling units (9 per PCM element) were surveyed using a stratified grid-based design. Patches supported significantly higher observed taxon richness than corridors and matrix habitats. Community composition was significantly structured by both elevation and landscape element, whereas precipitation and temperature did not show significant effects. The dbRDA model explained approximately 24.3% of the total variation, and landscape element remained significantly associated with community composition after controlling for elevationand spatial structure. Beta-diversity partitioning revealed that differences among PCM elements were overwhelmingly driven by species turnover (βJTU = 0.962) rather than nestedness (βJNE = 0.009), indicating strong species replacement and limited compositional redundancy among landscape components. Corridors shared taxa with both patches and matrix habitats, suggesting a potential dual role as connective elements and distinct habitat units. Co-occurrence networks exhibited clear modularity, reflecting habitat filtering and landscape structure rather than direct ecological interactions. Overall, our findings demonstrate that eDNA-derived metrics, including species turnover, modularity, and indicator taxa, can serve as ecological indicators of habitat differentiation in complex landscapes.