Scale Dependence in the Relationship between Forest Structural Diversity and Vascular Plant Diversity across Ecoclimatic Gradients

Abstract:

Studies have used the empirically-observed positive covariance between forest structural diversity (FSD) and vascular plant diversity to predict and map the distribution of biodiversity over landscapes. However, the role of scale dependency across eco-climatic gradients in constraining this relationship remains to be robustly quantified. If the FSD ~ plant diversity relationship is premised on a complex, yet deterministic series of processes whereby structure constrains diversity by mediating local resource acquisition and competitive dynamics, we hypothesize that structure's influence on diversity will exhibit a distinct distance decay profile, with its peak dependent on biotic and abiotic conditions. In this study, we employ in situ diversity and airborne LiDAR data from all NEON plots to assess distance decay in the FSD ~ plant diversity relationship, and how it varies by growth form and ecoclimatic condition. Decay functions and hypothesized abiotic drivers are inferred from a combination of ensemble machine learning and Random Effects Bayesian GLMMs. Results indicate that while plant diversity is generally best predicted by an FSD zone of influence subsuming a 20-40m buffer around each spatially-nested (1-400m²) plot, tree diversity (plants >10cm DBH) peaks at equivalent, not larger, scales. This finding suggests that while plant diversity is influenced by larger-scale niche and assembly processes including recruitment and competition mediated by incident light from canopy gaps outside of plot boundaries, tree diversity is best predicted by FSD from coincident geographic footprints, and thus reflects the pronounced role of sampling effects. Climate exhibits a distinct influence on the distance decay of FSD's zone of influence. In general, environmental filtering in climatically stressful conditions is associated with larger (~10,000m²) FSD zones of influence compared with more climatically benign regions, where small-scale local competition is the prevailing constraint on diversity. These findings provide a novel perspective on the scale-dependent role of forest structure in mediating the micro-site conditions that ultimately constrain the assembly of biodiversity, that has distinct implications for the optimization of biodiversity models based on air- and space-borne LiDAR.