Loss of plant diversity impairs ecosystem functioning. Compared to other well-studied processes, we know little about whether and how plant diversity affects root decomposition, which is limiting our ability to predict biodiversity-carbon cycling relationships in the soil. Plant diversity potentially affects root decomposition via two non-exclusive mechanisms: by providing roots of different substrate quality and/or by altering the soil decomposition environment. Within the frame work of the Jena experiment, one of the longest-lasting grassland diversity experiments in Europe, we conducted three decomposition experiments using litter bags to disentangle these mechanisms. We studied the root substrate quality effect by decomposing community roots from all experimental plots in a single plot, the decomposition environment effect by decomposing standard roots in all experimental plots, and the overall plant diversity effect by decomposing community roots in situ. In addition, we applied structural equation modeling (SEM) on data from the in-situ decomposition experiment and integrated morphological and chemical traits of community roots as well as soil physical and chemical conditions and abundance of soil biota to understand the individual pathways in details.
Results/Conclusions
Results from the three decomposition experiments and the SEM all showed that with increasing plant species richness, root decomposition decreased. Both root substrate quality and soil decomposition environment contributed to the negative relationship between plant species richness and root decomposition. Functional group composition played an important role, with the presence of grasses reducing root decomposition while the presence of legumes enhanced it. In addition, SEM revealed important drivers underlying the diversity and composition effects on root decomposition. Plant species richness positively affected root decomposition via increased root K concentration and negatively affected root decomposition via increased abundance of oribatid mites in the soil. Functional group composition, in contrast, mainly affected root decomposition via affecting root chemistry. Presence of grasses negatively affected root decomposition via increased root lignin concentration, increased oribatid mite abundance and decreased root K concentration, while presence of legumes positively affected root decomposition only via reduced root lignin concentration. By unraveling the causes of a negative diversity-root decomposition relationship, our study provides novel insights into the mechanisms underlying the strong diversity effect on soil C storage in grasslands.