Cumulative effects of multiple biodiversity attributes and abiotic factors on ecosystem multifunctionality in the Jinsha River valley of southwestern China

Various biodiversity attributes contribute to ecosystem multifunctionality (EMF) in different ways in terrestrial ecosystems. The relative influence of biodiversity and abiotic factors in driving EMF simultaneously remains relatively unclear. We evaluated the effects of woody species richness along with phylogenetic, functional, fungal, and bacterial diversity as well as abiotic factors on EMF. The analysis also included a review of function indices (including biomass production, water regulation, carbon stock, and nutrient cycling) along the elevation gradient in the Jinsha River valley, southwestern China, based on the bivariate relationship, random forest model, structural equation model (SEM), randomization tests, and redundancy analysis. The results showed that EMF increased with increases in woody species richness, phylogenetic diversity, elevation, and soil water content, while it decreased with increases in bacterial diversity and soil pH. The random forest model showed that elevation, soil pH, soil water content, and woody species richness made the main contribution to EMF and function indices. Further analysis using SEM confirmed these results. Indeed, woody species richness played a relatively important role in driving EMF, biomass production, carbon stock, and nutrient cycling compared to other diversity metrics. Dominant species contributed differently to the EMF and functions. However, phylogenetic diversity also appeared to have a positive effect on water regulation. Simultaneously, as a belowground trophic group, fungal diversity played a vital role in shaping EMF, biomass production, and nutrient cycling, which may be assigned to a change of arbuscular mycorrhizal fungi and ectomycorrhizal fungi along the elevation gradient. In addition, abiotic factors regulated the relationship between biodiversity and EMF. Elevation, soil water content, and soil pH appeared to have direct and indirect effects on the EMF and function indices. Woody species richness, fungal diversity, soil water content, soil pH, and elevation jointly mediated EMF in the Jinsha River valley, whereas woody species diversity was the main driver of EMF and other function indices following the complex environmental conditions. Our result should be useful for land managers involved in ecological restoration in the Jinsha River valley, where biodiversity and soil quality will determine the supply of multiple ecosystem functions.