Coleman, Mark (USDA-Forest Service, Savannah River, PO Box 700, New Ellenton, SC, 29809; Phone: 803-725-0513; Fax: 803-725-0311; Email: coleman.m@earthlink.net)

Carbon Sequestration Through Belowground Carbon Allocation: Forest Stand Development vs. Soil Resources Availability

M. D. Coleman *

Substantial amounts of carbon can be sequestered in belowground biomass of forest plantations. However it is difficult to verify the accuracy of belowground biomass estimates on a case-by-case basis. Stand-level growth process models operating with monthly or annual time steps can be used to predict belowground carbon allocation, if they are properly validated. Many of these models assume belowground allocation is controlled by soil resource availability. Favorable soil moisture and nutrient conditions are predicted to minimize belowground allocation. As soil resource availability declines, increasing belowground allocation is predicted. Such models are scaled to regional and continental levels for carbon accounting purposes, because of limited parameter requirements and accurate predictions of aboveground production. Yet, there is strong evidence that stand development, not soil resources, predominately controls belowground allocation. Limited opportunity to validate model assumptions regarding belowground production causes concerns that errors may multiply upon scaling. To understand the relative importance of stand development and soil resource availability, we monitored above and belowground production in plots of four southern forest tree species. Biomass of stem, branch, stump, and coarse and fine roots was measured annually in control, irrigated, fertilized, and irrigated + fertilized plots. Relative root biomass consistently declined with both age and resource availability, demonstrating the need to simultaneously determine which factor predominately controls belowground allocation. When aboveground vs. belowground tissues were plotted on log-log plots to control for tree size, there were few slope differences among treatments despite more than a three-fold difference in growth. These results indicate that the relationships between above and belowground biomass is controlled mainly by development rather than soil resource availability. Furthermore, all species tended to fall on a single line. Data from literature with a variety of species, conditions and developmental stages also show very consistent relationships between above and belowground biomass. These relationships can be used to reliably predict belowground biomass as a constant fraction of aboveground biomass. Use of soil resource availability to control allocation to roots in process models introduces errors that are not consistent with a wide range of belowground data. Research that reports declining root production with increased soil resources must account for development. Prediction of belowground biomass as a proportion of stem biomass is a quantitatively accurate method of accounting for belowground carbon sequestration.