Appendix 1. Methods To Estimate Carbon Storage and Accumulation

Carbon Storage in Trees

Estimates of growing-stock volume in a forest area were converted to estimates of carbon storage in trees in a twostage process. First, growing-stock volume was converted to total forest tree volume by multiplying by a ratio to account for the additional tree volume excluded from estimates of growing-stock volume: tops and branches, foliage, rough and rotten trees, small trees (less than 5.0 in. dbh), standing dead trees, stump sections, roots, and bark (table 1.1 Separate ratios were computed for softwoods and her woods to account for differences in the average propc total volume to growing-stock volume. Ratios were d~ from two principal sources: a new nationwide biomasc prepared by the USDA Forest Service containing the I estimates of above-ground biomass by tree compone (Cost and others 1990), and a special report preparec Koch (1989) containing estimates of the proportion of volume that is below the ground. Separate ratios wer derived for each of the regions to account for differen tree form and to be consistent with the data used to e growing-stock volume. The validity of this method res the assumption that the ratio of total above-ground bi to merchantable biomass (estimated in dry weight uni equivalent to the ratio of total above-ground volume to growing-stock volume.

There is considerable variation in the ratios of total vo fume to growing-stock volume among regions and species groups (table 1.1). For the United States as a whole, the average ratio of total volume to growing-stock volume is 1.91 for softwoods and 2.44 for hardwoods.

The second step involved converting total tree volum' cubic feet to carbon in pounds. Separate factors wer developed for major forest types, for softwoods and h woods within each forest type, and for broad geograp regions (table 1.2). The volume-to-carbon conversior was computed in two steps. First volume in cubic feet was converted to biomass in dry pounds by multiplying the number of cubic feet times the mean specific gravity times the weight of a cubic foot of water (62.4 Ibs.). A weighted mean specific gravity for softwoods or hardwoods was estimated from the relative frequency of the three predominant hardwood and softwood species in each forest type and region. The second step was to multiply the biomass pounds by a factor to account for the average carbon content of the tree. Estimates of the carbon content used in past studies have generally ranged from 45 to 50 percent (Houghton and others 1985); however, Koch (1989) found that, for the United States as a whole, the avererage percent carbon for softwoods was 52.1 and for hardwoods was 49.1, with some slight regional variations. The final factors used to convert volume to carbon ranged from 11.41 to 17.76 for softwoods, and from 11.76 to to 19.82 for hardwoods (table 1.2).

Carbon Storage In the Soil

To estimate carbon storage in forest soils, a regressic . model was developed to relate soil carbon in relative!' undisturbed, secondary forests to temperature and pr tion. The method was an extension of the model dev, by Burke and others (1989) for soil organic carbon in cropland and pasture in the Central Plains grassland and adjacent areas.

The data in Post and others (1982) were used to estimate regression coefficients for forest lands. They used published sources of data to estimate mean soil carbon density for all of the life zone groups of the Holdridge life zone system (Holdridge 1967). To estimate regression coefficients for forest lands, the mean soil carbon densities were associated with the average precipitation and biotemperature for each of the life zone groups as read from the Holdridge life zone chart.

To apply the regression equations to the United States, temperature and precipitation averages for timberland and other forest land within each State were estimated from published weather records (Ruffner and Bair 1987). Separate estimates for timberland and other forest land allowed some sensitivity to the wide variation within some large States. For example, eastern Texas is largely covered with timberland and has a climate different from that of western Texas, where forests are primarily classified as other forest land. State-level estimates of temperature and precipitation were aggregated to the regional level by weighting the individual State estimates by the areas of timberland and other forest land (table 1.3).

For developing estimates of soil carbon for regional aggregates of forests with different age classes, it was necessary to make some assumptions about when forests reached the level of development represented in the data by Post and others (1982). It was assumed that these levels would be reached at age 50 in the South and at age 55 elsewhere. Then the average per-acre estimate of soil carbon for a State or a region was adjusted to reflect the actual age structure of the forests. This was accomplished by first determining the average age distribution by age classes, and then converting the distribution to percent and computing a weighting factor by comparing the age distribution with a model of soil carbon changes over time. On average, eastern forests are younger than the reference age of 50 or 55, and western forests are older than the reference age. The weighting factor was multiplied by the initial estimate of soil carbon for a State or region to obtain the final estimate used in the tables.

Carbon Storage on the Forest Floor

Estimates of the amount of carbon or organic matter on the forest floor are available for very broad forest classifications (Schlesinger 1977, Vogt and others 1986) and for very specific ecological types. These sources were used to estimate carbon in the forest floor for reference age classes. For State or regional carbon yields, the estimates of Vogt and others (1986) for broad forest ecosystems were applied to the broad forest types common in the area (table 1.4). These reference estimates were assumed to be representative of relatively undisturbed secondary forests.

As was done in estimating soil carbon, a weighting procedure was used to account for the general composition and relative age structure of the State or regional forests. First, area estimates were compiled for hardwood timberland, softwood timberland, reserved timberland, and other forest land for each State. Then the estimated carbon on the forest floor from Vogt and others (1986) was used for timberland , and other sources were used for other forest land. The weighted average for all forest land was computed for each State. The weighted average was multiplied by a factor to account for the actual age distribution of forests within the State. The factor was derived in the same way as the age factor for soil carbon.

Carbon Storage in Understory Vegetation

The understory has such a small percentage of the tc carbon stock in forests that it is often ignored or adde trees in estimates of all live vegetation. Estimates of understory biomass are generally available only from published results of ecological studies of specific forest ecosystems (e.g., Messina and others 1983, Ohmann 1984, Switzer and Nelson 1972, Turner and Long 1975).

It was assumed that there was no carbon in the understory at age 0, and that understory biomass peaked at age 5 for all regions and forest types. It was assumed that understory biomass declined to a reference level by age 50 in the South and age 55 elsewhere. Reference levels were defined as 2 percent of the carbon in the overstory, except for Douglas-fir and red pine, for which a value of 1 percent was used. The distribution of values by age class was compared with the actual age-class distributions of forest land by forest type to estimate a weighted average value for carbon in the under story vegetation in each State.

Table 1.1 - Ratio of total volume to merchantable volume

Table 1.2 - Factors to convert tree volume (cubic feet) to carbon (pounds)

Table 1.3 - Estimates of organic soil carbon in relatively undisturbed secondary forests in the United States, by region

Table 1.4 - Estimates of organic matter and carbon on the forest floor by region and forest type

[ Table of Contents | Next ]