Interactions between biotic and abiotic factors shape phylogenetic community assembly in southern range-margin Picea jezoensis forests

Identifying factors that control community formation is crucial for understanding biodiversity patterns. Although numerous studies utilize either within- or among-community indicators to explain community formation processes, few have assessed the additional insights gained by integrating the two. Accordingly, we examined Picea jezoensis forests in South Korea and northeastern China using complementary within- and among-community approaches and discuss the implications of interpreting these metrics together: (1) we identified the abiotic (e.g., topography and climate) and biotic (e.g., stand structural diversity and community-weighted proportion values of dormancy and dispersal forms) controls of phylogenetic community structure (net relatedness index [NRI] and nearest taxon index [NTI]) across forest strata (whole strata, upperstory, and understory) using piecewise structural equation modeling (pSEM); and (2) we quantified among-community phylogenetic differentiation by decomposing phylogenetic beta diversity (PBD) into turnover and nestedness and testing their relationships with climatic distance, geographic distance, and pairwise elevational difference. (3) We explored how within-community phylogenetic structure (NRI/NTI) relates to among-community phylogenetic differentiation (PBD components). Our results showed that the primary drivers of phylogenetic structure varied by forest strata and metric. Specifically, elevation, mean annual temperature (MAT), and community-weighted proportion (CWP)-based dormancy forms were the dominant predictors for NRI in both the whole strata and understory. In contrast, NTI were primarily shaped by abiotic factors (elevation and MAT) in the upperstory, while biotic constraints—specifically dormancy forms—showed a greater contribution in the understory. Furthermore, among PBD components, turnover was the primary contributor and was strongly associated with climatic distance. This climatic influence remained dominant even after controlling for geographic distance. In contrast, nestedness was best explained by pairwise elevational differences. By bridging within- and among-community metrics, this study helps contextualize how local-level assembly signals may scale up into regional phylogenetic patterns, offering a more unified understanding of forest community formation across spatial scales.