Whole-Profile Soil Carbon Stocks: The Danger of Assuming Too Much from Analyses of Too Little
- A. N. Kravchenko *a and
- G. P. Robertsonb
Whole soil profile assessments of soil C stocks are enormously important for evaluating a soil's potential for sequestering C. Carbon stock measurements are often highly variable, which makes the detection of statistically significant differences among different land uses and management systems difficult. A common mistake is to interpret a lack of statistical significance as evidence for the absence of differences. This mistake is costly: failure to identify practices and habitats that accumulate C can affect our understanding of soil C cycling and hide possible sequestration potentials. In this study, we investigated inadequate replication as a major cause of a lack of statistical significance. Power analysis is a well-established statistical tool that allows researchers to infer whether the lack of statistical significance is due to the absence of meaningful differences or due to insufficient replication. We used data from two published studies to illustrate the value of power analysis. In particular, we showed that in these studies the chances of detecting even substantial (e.g., 50%) differences among experimental treatments can be as low as 20 to 60% due to the natural variability in soil C concentrations and stocks. Carbon variability in different soil horizons suggests that statistical analyses of soil C stocks should be conducted separately for each sampled horizon and that conclusions about the profile as a whole should be made based on the individual horizon results. Devising an optimal sampling strategy that identifies the appropriate number of replicates needed to detect significant C change by horizon is essential.Please view the pdf by using the Full Text (PDF) link under 'View' to the left.
Copyright © 2011.