Accurate laboratory assessments of soil carbon content require the use of carbon (combustion) analyzers. Analytical methods based on mass loss on ignition or wet digestion (see Nelson and Sommers, 1982; Tiessen and Moir, 1993) are prone to biases. Although these biases can be calibrated (Kalembasa and Jenkinson, 1973), they add another level of uncertainty to the estimate. Methods for sample preparation and analysis for SOC and SIC are well documented elsewhere (Lal et al., 2000). The carbon stock per unit area of land can be calculated from the percentage of carbon and the soil bulk density. For transparency and verifiability purposes, soil samples can be geo-referenced and archived in a dried state in a dark, cool place for future carbon analysis, and original data can be retained. Some archived samples should be re-analyzed at the same time as repeat samples to confirm that there has been no drift in analytical precision. Methods also exist to measure different components of SOC (see Section 4.2).
Modern dry combustion procedures have an analytical precision of �2-3 percent. The error associated with assessments at the regional and global scales also depends on errors introduced through spatial variability and sampling error and therefore will be much higher than 3 percent. Establishing quality protocols for all methods is important. The cost of analysis varies from region to region, by number of samples, and by efficiency. On a global scale, the per-sample cost of carbon analysis may range from US$3 (where labor costs are very low) to about US$20, with significant proportions of the total cost accounted for by sampling (high labor cost), sample preparation, bulk density measurement, inorganic carbon analysis (or pre-acidification of samples to remove carbonate), archival of samples, and quality control efforts. This cost range is broadly consistent with that reported for a Canadian study by Izaurralde et al. (1997).
Special consideration is necessary for peatlands and other wetland soils-collectively known as Histosols-in which the layer of undecomposed organic matter may be several meters in depth. These soils are vulnerable to large losses of organic matter upon changes in regional drainage or soil temperature. Particularly in areas that have been drained, accurate measurement of changes in carbon stock in peat soils requires reliable measurement of carbon loss through decay of organic material. Accurate measurement of changes in carbon in peat soils probably requires fixed reference points for measuring subsidence from decay. When subsidence is being measured, corrections should be made for depth changes resulting from changes in soil moisture content. Estimates of changes in the storage of carbon in peatlands must avoid biases that may result from changes in bulk density because of compaction of surface layers. Often, the age of peat at a given depth may be determined by 14C or 210Pb dating, allowing workers to assess additions or losses of organic carbon above a depth of known age.
Other reports in this collection