Trait similarity among dominant highly-competitive species rather than diversity increases productivity in semi-arid Mediterranean forests

Abstract Biodiversity plays a pivotal role in forest productivity and, through it, on the provision of ecosystem services. Consequently, understanding the mechanisms by which biodiversity drives productivity is a challenge for sustainable forest management. It has been proposed that functional dissimilarity can increase forest productivity, particularly in water-limited ecosystems (as Mediterranean forests), where water stress might increase complementarity resource use among species (hypothesis of niche complementarity). Alternatively, strongly water-limited forests may favor the dominance of woody species with slow-growth and conservative resource-use trait values, which in turn, might increase plant community productivity proportionally to their biomass (hypothesis of mass-ratio). The effects of such mechanisms on productivity is yet understudied in water-limited forests. We assessed the effect of both previously described mechanisms on productivity in a semi-arid Mediterranean-type climate matorral of Chile through a trait-based approach. We hypothesized that: (1) higher dissimilarity of functional trait values of crowns, roots and phenology among woody species increases community productivity. Alternatively, (2) higher dominance of woody species of conservative trait values increases forest productivity. We installed twenty-three 25 × 25 m plots in the matorral of the National Reserve of “Robleria de Los Cobres de Loncha” located in the Coastal Range. Forest productivity was estimated as above-ground biomass changes between 2010 and 2017. The mean species pairwise dissimilarity (MPD) and the community-weighted mean trait values (CWM) were calculated for each single trait as surrogates of the hypotheses of niche complementarity and mass-ratio, respectively. Environmental variables (e.g. altitude, slope, precipitation) were included as potential drivers influencing productivity. We performed a selection procedure of multiple linear regression models based on the Akaike Information Criterion. Our findings show that higher forest productivity was related with higher trait similarity of plant maximum height and onset of growth among species. Besides, productivity increased with CWM values that reflect greater dominance of species exhibiting acquisitive traits (i.e. greater maximum plant height, lower wood density and earlier onset of growth) rather than conservative traits. Moreover, forest productivity increased with increasing altitude, precipitation and decreasing temperatures from 300 to 900 m. Therefore, in contrast with our hypotheses, productivity increased by greater trait similarity among dominant highly-competitive species - which might be favored by mesic conditions at middle altitudes of the Coastal Range. The sustainable management of semi-arid Mediterranean forests with the main focus on productivity should promote higher dominance of highly-competitive species with similar functional trait values in mesic conditions, and conservative species with greater trait functional dissimilarity to increase drought resistance in stressful environments.