Drivers of subtropical forest dynamics: The role of functional traits, forest structure and soil variables

Abstract

AbstractQuestionHow do biotic forest conditions, in terms of plant functional traits and tree basal area, and soil variables influence old‐growth forest dynamics?LocationA subtropical old‐growth forest in Southern Brazil.MethodsForest inventories were conducted in 1999 and 2017, in 24 permanent plots of 500 m2. All trees with a circumference at breast height of ≥ 30 cm were measured, identified to the species level, and tagged. We fitted linear models to evaluate the effect of the biotic variables of initial forest communities (tree basal area, trait community‐weighted mean [CWM] and both species and functional diversity metrics) and of soil variables on forest dynamics. These models included biomass dynamics (annual rate of biomass increment and mortality) and demographic rates (recruitment and mortality rates). Furthermore, linear models were used to evaluate the effect of tree survival, recruitment and mortality and biomass increment on the net biomass change.ResultsWe observed a positive effect of species diversity and a negative effect of CWM leaf phosphorus content on the biomass increment. Biomass loss was positively related to basal area and negatively related to the CWM of leaf nitrogen content. Overall, net biomass change was negatively influenced by biomass loss due to mortality. Recruitment rates were negatively affected by soil potassium content and were positively affected by soil pH and functional evenness of the community. Mortality rates were positively predicted by soil clay content and negatively predicted by species diversity.ConclusionsCommunity functional traits, basal area and soil variables predicted growth, mortality and recruitment dynamics of this subtropical old‐growth forest. This work provides important information about the biomass and demographic dynamics of subtropical forests, showing that initial community characteristics influence the gain and loss of biomass and individuals over a period of almost two decades.