The importance of stand structure and tree allometry for local-scale variation in aboveground biomass

Aboveground biomass (AGB) plays a critical role in determining the long-term dynamics of carbon in tropical forests. Consequently, understanding what factors are important in controlling AGB in tropical forests has major implications for projecting the terrestrial carbon stocks, in the context of an increasingly uncertain future. In this study, we aimed to explore the local-scale AGB variation in two forest sites in northern Congo, representative of two contrasted forest types under the same climate but growing on vastly different soils and parent material (quartzite substrate for CIB and sandstone substrate for Mokabi). Tree diameter was measured in 36 permanent forest plots of 1-ha in each site, and tree allometry (total tree height, height of the first branch and crown dimensions) was measured on a subsample of 18 plots of 1-ha in each site. Allometric data were available for a total of 2202 trees (1040 for CIB and 1162 for Mokabi) covering a large range of diameters (10 – 200 cm). We first developed site-specific allometric models that were then used to estimate AGB at plot level. We then explore the determinants of AGB variation at plot level using multiple regressions and mixed linear models. For a given diameter, trees tended to be taller and to have deeper crown in the Celtis forest of the CIB (rich soil) while they tended to have larger crown in the Manilkara forest of the Mokabi (sandy soil). Similar trends were reported within species for the sixteen species shared by both forest types, suggesting an environmental control of tree allometry. We found that AGB strongly varied between the two forest sites, with greater AGB per hectare in the Celtis forest of the CIB. Within sites AGB variation was positively related to basal area, though between-site allometric attributes (total height and crown radius) were important determinants of AGB variation. These results have strong implications for forest biomass and carbon monitoring.