Urban green spaces rely on bird–plant interactions that support biodiversity and ecosystem functioning, yet it remains unclear how urban landscape configuration shapes bird–plant interaction networks. We examined how landscape configuration influences the structure and beta diversity of bird–plant multilayer networks in 13 Tokyo urban parks, and which bird traits, landscape variables, and spatial scales best explain among‐park variation. We sampled bird-dispersed seeds and feces along transects and identified birds and plant genera using DNA and morphology. For each park, we constructed a multilayer network linking foraging and seed-dispersal interactions and related network metrics to bird traits and landscape variables within 500–1500 m buffers. Multilayer networks contained relatively few bird species but many plant genera. Longer bills, semi-open habitat use, and ground-dwelling or perching behaviors were associated with higher multilayer centrality, whereas body mass showed no clear effect. Within 1500 m buffers, higher green-space proportion increased network density and centrality and reduced average path length, whereas greater mean nearest-neighbor distance among patches was associated with higher network modularity. Park area, vegetation quality, and surrounding landscape attributes jointly explained among-park variation in interaction beta diversity, with most dissimilarity driven by interaction rewiring rather than species turnover. Overall, urban landscape configuration appears to shape bird–plant multilayer networks by filtering bird functional traits and reorganizing interactions. Maintaining well-connected green spaces within ~1500 m of parks, together with high within-park vegetation quality, may help sustain these networks and their associated ecological functions. • Bird–plant multilayer networks were constructed for 13 urban parks in Tokyo using DNA metabarcoding. • Birds with longer bills, a preference for semi-open habitats, and ground-dwelling or perching lifestyles were central species in the networks. • Green-space proportion within 500–1500 m increased network density and centrality and decreased average path length. • Patch isolation at 1500 m was positively associated with network modularity and overall network dissimilarity. • Urban green-space planning should prioritize landscape configuration within ~1500 m of parks, together with high within-park vegetation quality, to sustain interaction networks.
Tang et al. (Sun,) studied this question.