Context‐dependent changes in the functional composition of tree communities along successional gradients after land‐use change

Summary Land-use change can modify the functional composition of tree communities, which is an essential determinant of the ecosystem functions. The lack of consensus about the functional responses of tree communities to land-use change is a major uncertainty in the assessments of human impacts on terrestrial ecosystem functions. In this study, we applied a machine-learning method to a large data set consisting of 2574 tree communities across Japan to examine changes in the functional composition of tree communities after land-use change while considering contexts including successional trajectories, forest types and the presence of gymnosperms. Specifically, we hypothesized that functional changes along successional gradients after land-use change can be different in different contexts. Effects of two successional variables (stand age and basal area) on functional composition were highly significant throughout the study region. Changes in functional composition with changes in the two successional variables differed greatly, and the effects of basal area often outweighed those of stand age. Tree communities with small basal area were generally characterized by functional traits related to the resource-acquisitive strategy, that is short adult stature, low leaf mass per unit area, small seeds, low wood density and large leaves, especially when gymnosperms were excluded from the analysis. Decreasing basal area but not decreasing stand age often led to a considerable loss in functional diversity. Despite these general trends, functional changes along successional gradients after land-use change were not necessarily parallel and the opposite patterns were sometimes observed among forest types, traits and taxonomic groups. Synthesis. As a whole, our analyses demonstrate that the functional changes in tree communities after land-use change are highly evident in a given context but can be different under different contexts. These changes in functional composition can trigger variable changes in ecosystem functions such as carbon and nutrient cycling that depend on the context.