Coarse and fine woody debris on the forest floor often accounts for about 20 percent as much carbon as living biomass. The amount of woody debris may increase markedly immediately following forest harvest, which typically leaves a large amount of organic debris on the soil surface (Covington, 1981). These residues will decompose within a few years following harvest, although there is little information on the carbon content of different types of litter. Sampling the volume of woody debris is as straightforward as it is for live trees, but assessing its mass is more problematic because decay affects wood density dramatically.
The litter layer-also known as the L and O horizons-is the layer of dead plant material that lies on top of the mineral soil. During forest regrowth, the litter layer may accumulate rapidly, so changes in its carbon content are an important component of a total carbon inventory in ecosystems (Richter and Markewitz, 1996). During a cycle of forest harvest followed immediately by regrowth, however, there is usually little overall change in carbon storage in the forest floor (Johnson, 1992).
Litter quantities change dramatically with the seasons. As a result, remeasurements must be made at the same time of the year as initial measurements. If litter were to be sampled together with the top layer of mineral soil, temporal variability in the standing crop of litter could confound detection of changes in soil carbon. Typically, dividing litter from mineral soil when sampling is not difficult.
The number of samples necessary to estimate the accumulation of surface litter will differ greatly between ecosystems. In some forests, as few as 10-15 randomly located samples may provide an accurate estimate of changes in the mass of the forest floor over large areas (e.g., Schiffman and Johnson, 1988); in shrub deserts, where the spatial variability of soils is enormous, it is often necessary to take a larger number of random samples in locations that are stratified by plant cover (e.g., Conant et al., 1998). In any particular location, some preliminary sampling is probably required to determine the number of forest floor samples that is necessary to estimate the total mass to the desired level of accuracy.
Other reports in this collection