Effects of climate, soil, structure and management on functional traits in forest understory

Many studies explored how the observed pattern of plant functional traits (PFTs) may be influenced by environmental variables. However, studies on forest ecosystems including also stand structure and management are lacking. A first attempt to test the relative effect of variables related to the latter groups, together with climate and soil gradients, on the community weighted mean (CWM) values of PFTs was performed on forest understory in Italy. The Level I biodiversity dataset (extensive CONECOFOR network) has been used, based on a probabilistic sampling design, by 201 sites on a representative 16 x 16 km systematic grid. Following a harmonized protocol (ICP Forests, BioSoil-Biodiversity project) 29 explanatory variables were recorded and four plots 10x10 m have been surveyed for vascular specific cover, on each site. Variance partitioning was used to identify the relative role of climatic, soil, structural and management variables on the CWM values of specific leaf area (SLA), plant height (H) and seed mass (SM). Redundancy analysis was used to assess the relation between traits and variables. The combination of the selected variables explained the variation of H (34.3%) better than SLA (14.9%) and SM (11.1%). Climate alone, and in combination with other variables, demonstrated to explain the largest proportion of the variation for H (29.5%) and SM (9.3%); however, also structure and soil showed a relevant role. Forest management (9.9%) and structure (5.4%) were the main drivers for SLA. Considering a gradient of increasing temperature, aridity and nutrient availability, we detected plant understory communities with higher mean values of H and SM. High-SLA communities appeared in forests characterized by a larger amount of deadwood.In forest understory vegetation, the PFTs pattern is linked to a complex combination of variables. Not only climate and soil, but also forest structure and management played a role, suggesting the importance of taking into account such parameters in future research, at larger spatial scale including different Country-level policies. The variation of SLA, H and SM is controlled by different variables, making no obvious any attempt to predict the effects of climate and land-use changes on understory functional signature.