Direct effects of selection on aboveground biomass contrast with indirect structure-mediated effects of complementarity in a subtropical forest

Understanding the multiple biotic and abiotic controls of aboveground biomass (AGB) is important for projecting the consequences of global change and to effectively manage carbon storage. Although large-scale studies have identified the major environmental and biological controls of AGB, drivers of local-scale variation are less well known. Additionally, involvement of multiple causal paths and scale dependence in effect sizes potentially confounds comparisons among studies differing in methodology and sampling grain. We tested for scale dependence in evidence supporting selection, complementarity and environmental factors as the main determinants of AGB variation over a 50 ha study extent in subtropical China, modelling this at four sampling grains (0.01, 0.04, 0.25 and 1 ha). At each grain, we used piecewise structural equation models to quantify the direct and indirect effects of environmental (topographic and edaphic properties) and forest attributes (structure, diversity and functional traits) on AGB, while controlling for spatial autocorrelation. Direct scale-invariant effects on AGB were evident for structure and community-mean traits, supporting dominance of selection effects. However, diversity had strong indirect effects on AGB via forest structure, particularly at larger sampling grains (≥ 0.25 ha), while direct effects only emerged at the smallest grain size (0.01 ha). The direct and indirect effects of edaphic and topographic factors were also important for explaining both forest attributes and AGB across all scales. Although selection effects appeared to be more influential on ecosystem function, ignoring indirect causal pathways for diversity via structural attributes risks overlooking the importance of complementarity on ecosystem functioning, particularly as sampling grain increases.

Diego Ismael Rodríguez-Hernández, David C. Deane, Weitao Wang, Yongfa Chen, Buhang Li, Wenqi Luo, & Chengjin Chu