Stocks of soil organic carbon in forest ecosystems of the Eastern Usambara Mountains, Tanzania

An inventory of soil organic carbon (SOC) stocks to 100-cm depth was carried out at four pedologically comparable forest sites in the Amani Nature Reserve (Eastem Usambara Mountains, NE Tanzania). Three plots at primary tropical submontane forests and one plot at a secondary forest derived from selective cutting were sampled. The aim was to quantify the SOC stocks including the spatial variation, and to put the findings into a pedological context. At each site, a combined approach was applied with auger samples taken to 40-cm depth (topsoil) at eight points (10 cm depth intervals) surrounding a central soil pit. The subsoil was sampled at 40-100 cm depth in the soil pit, accompanied by a detailed soil description and classification according to FAO and WRB. In addition, soil analysis comprised a set of chemical and physical parameters. According to WRB, the soils belonged to the reference groups Acrisols and Alisols, depending on the method used for estimating clay-related potential cation exchange capacity. The SOC stocks to 100-cm depth ranged between 16.9 and 22.4 kg C m(-2) (mean 19.7 kg C m(-2)). The SOC stock at the secondary forest (20.2 kg C m(-2)) did not differ from those of the primary forest plots. In all profiles the subsoil contributed to the SOC stocks to 100-cm depth, whereas the contribution of the forest floor was small. The variation of SOC stocks between the forest sites may be explained by 'old' land-use and/or differences in soil chemical parameters controlling SOC stabilization. The SOC stocks appear to be related to the amounts of poorly crystallized Fe- and Al-oxides. The mechanisms and potentials for SOC sequestration are, however, unclear and merit further detailed study. Our inventory of SOC under forest represents a suitable baseline for the assessment of SOC stock changes after land-use conversion to agriculture on typical soils of the humid tropics. (C) 2014 Elsevier B.V. All rights reserved.