Divergent soil filtering of leaf functional traits of Pinus yunnanensis in non-karst and karst habitats: from pH screening to stoichiometric control

Introduction: by examining leaf trait shifts and coordination patterns. Methods: We compared key leaf functional traits between paired karst and non-karst sites. An integrated analytical framework was applied, including correlation analysis, network analysis, redundancy analysis (RDA), and partial least squares structural equation modeling (PLS-SEM) to identify trait variation patterns and their environmental drivers. Results: in karst habitats exhibited a resource-conservative strategy, with significantly lower specific leaf area and leaf phosphorus concentration, but higher leaf thickness, leaf dry matter content, and leaf nitrogen to phosphorus ratio. Network analysis revealed a fundamental reorganization of trait relationships: karst habitats formed a tightly connected network centered on stoichiometric traits, whereas non-karst habitats displayed an SLA-centered architecture. RDA identified soil phosphorus availability and stoichiometric ratios as primary drivers in karst habitats, while soil pH and nitrogen availability predominated in non-karst habitats. PLS-SEM indicated that karst environments attenuated the trade-off between leaf morphological and chemical traits. Conclusion: Karst environments function as potent biogeochemical filters that not only select for conservative trait values but also fundamentally reconfigure phenotypic integration. The architectural shift from a resource-acquisitive to a persistence-oriented phenotype demonstrates how extreme environments reshape plant adaptation through altered trait coordination networks.