Soil C is a function of C inputs, from net primary production, and C outputs, through decomposition, both of which are spatially variable. As a result, soil C is highly variable, with coefficients of variation as high as 20 percent even in flat fields that have been under long term uniform crop rotation and tillage management. Quantifying C sequestration in response to changes in agricultural management is, thus, confounded by substantial variability inherent in soils. Further, even at high rates of C sequestration following changes in management, net changes in soil C are likely to be small relative to the amount of C in soil. Here we evaluate implications for detecting changes in soil C and for farm- level, regional, and national assessments. Soil C variability implications for regional and national assessments were evaluated using information from the USDA/NRCS pedon database to assess soil C sample variability at national-, state-, and county-levels. Farm- level variability was investigated on pastures in Virginia that were intensively sampled using a method designed to maximize ability to detect changes over time following implementation of new management. Six soil samples were collected in two by five meter microplots, with three microplots per field. Total C analyses were performed on individual cores and samples bulked by microsite. These data were used (1) to determine variability in pasture and forest sites, (2) to evaluate the utility of this method for sampling and re- sampling the same plots, and (3) to investigate variability in fields or forests and implications for detecting changes in soil C over time.
