Assessing Soil Organic Carbon Dynamics At Different Spatial Scales and Land Uses in East Africa.

Soil organic carbon is an important indicator of soil health, yet drivers of carbon dynamics across different land uses in East Africa are poorly understood. The Land Degradation Surveillance Framework is a methodology that assesses the variability of ecological metrics at different spatial scales. In this analysis, soil organic carbon dynamics were assessed within and between different land-use systems (forest, woodland, shrubland, grassland and cropland) in Tanzania. Seven 100-km² sentinel sites were sampled in 2010, each site consisting of 16-1 km² clusters and 160 plots (plot area=1000 m²). Composite soil samples were collected at each plot at two depths (0-20 cm and 20-50cm) and cumulative soil mass samples were collected at the center subplot to 50 cm. Soil samples were scanned using mid-infrared spectroscopy and calibration models were developed using soil organic carbon values from 15% of the samples and a Global Spectral Prediction Tool developed at the World Agroforestry Centre.  A total of 1,993 plots were included in the current study. Woodland and cropland had the highest frequency  (560 and 475 plots, respectively). Average topsoil organic carbon was 10.9 g C kg^-1 (n=1058) and average subsoil organic carbon was 8.3 g C kg^-1 (n=935) for the seven sites.  Forested plots had the highest mean topsoil organic carbon concentrations 14 g C kg^-1 followed by cropland 13 g C kg^-1. Soil mass at 20 cm varied more between sites than between vegetation types, with average soil mass at 20 cm of ~1012 g. These data will be used to calculate carbon stocks for the different land uses. Systematic sampling allows for robust comparisons of soil organic carbon values and soil mass within and between land uses and at different spatial scales.