The soil contains two major types of carbon: soil organic carbon and soil inorganic carbon (which occurs primarily as carbonates). Although SOC constitutes the majority of soil carbon in most soils globally (Batjes, 1996), soil carbon in arid and semi-arid regions may be dominated by SIC (Schlesinger, 1982). In fire-prone ecosystems, there may be significant amounts of charcoal present.
SOC held in the various layers, or horizons, of the mineral soil largely occurs as humus. Much of the humus is adsorbed to the soil mineral fraction, particularly to clay minerals (Richardson and Edmonds, 1987). A small amount of the total content of soil organic matter consists of dead roots, which become part of the soil organic matter upon their demise. In most cases, however, the amount of carbon associated with such roots is <5 percent of the total carbon in the soil profile (Ruark and Zarnoch, 1992).
There is a wide range of laboratory procedures available to measure SOC stocks, the most accurate of which is dry combustion (usually performed using a carbon analyzer). Changes in SOC stocks over a 5-year commitment period may be difficult to measure in some soils because although they are potentially large in absolute terms, they may be small compared with background levels (Batjes, 1999). Despite these difficulties, the rate of change in SOC stock during a commitment period can be measured (Lal et al., 2000); because of high spatial variability, however, many sub-samples may be required to obtain a mean with an acceptable standard error (Izaurralde et al., 1997; Garten and Wullschleger, 1999; Post et al., 1999). Figure 2-9 (from Garten and Wullschleger, 1999), for example, shows the calculated minimum detectable difference in SOC as a function of variance and sample size for SOC changes after 5 years under a herbaceous bioenergy crop.
|Figure 2-9: Soil carbon inventories under a bioenergy crop (Garten and Wullschleger, 1999).|
In some cases, the cost of demonstrating the change in stocks to the required level of accuracy and precision may exceed the benefits that accrue from the increase in stocks. As the size of the stock change decreases, there will be a point when the cost of demonstrating the change with sufficient confidence exceeds the value of the carbon offset provided. The cost of demonstrating a change in SOC stock could be decreased by developing and verifying locally calibrated models that can use more easily collected data (Section 2.4.5).
Other reports in this collection