Dynamic changes in seed germination trait networks correlate with shifts in successional gradients of alpine meadow on the eastern Tibet Plateau

BACKGROUND AND AIMS: Seed germination traits regulate population regeneration and community assembly. The germination trait network, integrating multiple traits as nodes and quantifying pairwise associations, reveals complex interactions and identifies key hubs. However, a research gap remains regarding how seed germination trait network undergoes dynamic shifts along with community succession, which constrains a comprehensive understanding of the underlying succession mechanisms. METHODS: We constructed and visualized seed germination trait networks for meadows across different successional stages by integrating field survey data with germination experiment results, with ten plots established in ten grasslands for each successional stage. The relationships between these networks and community diversity were also analyzed. KEY RESULTS: With the progression of succession, the seed germination trait networks exhibited enhanced connectivity, shortened paths, and weakened modularity (i.e., significant increases in edge density and average clustering coefficient, and significant decreases in average path length, diameter, and modularity). In all successional stages, germination percentage-related traits functioned as primary hub, while for early-successional communities also did germination rate-related traits. Species richness, Shannon-Wiener, and Simpson's indeces were significantly positively correlated with edge density and average clustering coefficient, and significantly negatively correlated with average path length, diameter, and modularity, while Pielou's evenness index showed no significant association with any of the network parameters, suggesting dynamic trait changes are related to community assembly. CONCLUSIONS: This study provides the first evidence of systematic shifts in seed germination trait network architecture across succession, linking trait integration patterns directly to community assembly processes and offering a novel network-based perspective for understanding plant regeneration strategies.