Soil Nutrients, Forest Structure and Species Traits Drive Aboveground Carbon Dynamics in an Old-Growth Temperate Forest.

Huiying Cai,Fengri Li,Guangze Jin

DOI: https://doi.org/10.1016/j.scitotenv.2019.135874

2020-01-01

Abstract:Forests store a substantial amount of terrestrial carbon (C), but the drivers of forest C dynamics remain poorly understood, especially in old-growth forests. Here, we evaluate how aboveground C dynamics (i.e., net C change and its demographic processes: C gain from the growth of surviving trees (∆C-surv), C gain from the growth of recruited trees (∆C-recr) and C loss by tree mortality (∆C-mort)) are driven by vegetation attributes (diversity, trait composition and forest structure) and habitat conditions (soil properties and light environment), as well as how ∆C-surv, ∆C-recr and ∆C-mort contribute to net C change. Using 10-year interval demographic data from a 9-ha permanent plot in an old-growth temperate forest in northeastern China, we performed structural equation model to relate the C dynamics to the vegetation attributes and habitat conditions. The net C change is most strongly determined by ∆C-mort. High soil phosphorus concentrations increased ∆C-surv, soil moisture increased ∆C-recr, and leaf area index increased both ∆C-surv and ∆C-recr. Diversity (i.e., structural diversity) had a positive relationship with ∆C-surv but was not related to ∆C-recr or ∆C-mort. Trait composition was significantly related to all three demographic processes. Forest structure was the best predictor of ∆C-surv and ∆C-recr. The net C change increased with higher soil phosphorus concentrations and basal area and in communities dominated by conservative traits (i.e., high wood density). This study highlights that soil nutrients, forest structure and trait composition are important drivers of net C change in old-growth temperate forests. Better insights into C storage and productivity can be gained by simultaneously evaluating the vegetation attributes and habitat conditions of C dynamics in natural ecosystems.