Multivariate relationships between litter productivity and its drivers in a tropical karst seasonal rainforest

Litter is a vital component of ecosystem functionality, which is inherently linked to net primary productivity and global biogeochemical cycles. To better understand the drivers that modify net primary productivity, we need to disentangle the processes that underly changes in the generation of litter. We employed structural equation models (SEMs) to test for the direct and indirect effects of biodiversity, the stand structural attributes, and abiotic drivers on annual litter production in a tropical karst seasonal rainforest. For this study, annual litter production was evaluated over a five-year period in 90 permanent sample litter traps in a dynamic 15-ha long-term forest plot in Southern China. For each circular plot stand (with central litter traps) we quantified biodiversity, structural variations, basal area, and growth rate, which are important for net primary productivity. The four best-fit SEMs with different metrics of biodiversity for the paths between biodiversity, structural attributes, abiotic drivers, and annual litter production had a similar goodness of fit to the data. They accounted for 42.7–47.8% of the variation in annual litter production, 4.7–16.8% of the variation in stand-level growth rate, 29.3% of the variation in stand-level basal area, 8.2–13.2% of the variation in stand structural variation, and 4.4–6.5% of the variation in biodiversity. Although biodiversity had a positive effect on annual litter production through direct or indirect effects via the stand-level growth rate, abiotic drivers had the strongest impacts on annual litter production. We also found that the structural attributes of the stands had positive effects on the annual litter production; however, only the stand-level growth rate had significant consistently positive effects on annual litter production in the four best-fit SEMs. These results implied that annual litter production is most strongly driven by abiotic factors, followed by stand structural attributes and biodiversity. Theoretically, our study supports that niche differentiation and facilitation might enhance productivity; however, this enhanced capacity is primarily contingent on how abiotic variations are controlled in the sampling of natural karst forest systems. Biodiversity and the conservation of specific habitats should therefore be a key component in the management of forest resources, forest conservation, and sustainability.