Amazon forest biomass: intra- and interspecific variability in wood density drive divergences in Brazil's far north

Wood density (WD) is an important functional trait of tree species. Understanding spatial WD variability as a function of environmental determinants improves our ability to estimate carbon stocks in the woody biomass of tropical forests. However, the role of each environmental variable affecting the intra- and interspecific variability of WD is not entirely clear for most forest ecosystems. In Amazonia there are recurrent uncertainties in estimates of regional woody biomass. The aim of the study was to investigate the effects of environmental conditions on the intra- and interspecific variability of WD for tree assemblages in forests of the northern Brazilian Amazon. A single sample was extracted from each of 680 individuals (108 species, 82 genera, 38 fami lies; stem diameter >= 10 cm) dispersed among 129 plots distributed along a hydro-edaphic gradient. General community-averaged WD (0.703 +/- 0.133 g cm(-3); range: 0.203 to 1.102 g cm(-3)) was high in relation to other Amazonian areas because 62% of the species and 69% of the sampled individuals had high WD values (>0.650 g cm(-3)). Altitude (a proxy for drainage), clay and soil micronutrient content explained 23% of the spatial variation in WD. Partitioning WD variation into species-substitution (turnover) and intraspecific-variation components slightly increased the explanatory power to 26%. The analysis of interspecific variability showed that forests occurring in seasonally flooded areas are characterized by tree assemblages with species tolerant to P-poor soils, where mean WD (0.742 g cm(-3)) is about 4% higher than the mean (0.713 g cm(-3)) for tree assemblages on unflooded uplands where soils have less limitations from nutrient poverty. Our results represent an improvement in the estimates of biomass because they promote adjustments (1.4%-16.3%) to the previous estimates of woody biomass in the northern Brazilian Amazon forests considering different environmental conditions.