Analysis of alternative methods for estimating carbon stock in young tropical plantations

Estimates of carbon stock in forest plantations are generally based on allometric equations relating either carbon or biomass to diameter at breast height (DBH). These equations are usually based on measurement of the fresh mass of each tree with subsamples taken to determine moisture content to convert to dry weight. However, drying time and the number of sub-samples varies between studies. Furthermore, the carbon concentration of different tree parts is rarely measured directly, but generally assumed to be 50% of dry weight. This study analyzed those assumptions and determined their effect on regression equations and on species-specific stand level estimates of carbon stock for Anacardium excelsum and Dipteryx panamensis growing in 7-year-old mixed-species plantations in Panama. Four methods were used to develop aboveground carbon estimates for the same sample of trees. Results indicated that the drying time, the number of sub-samples taken, and whether or not carbon was measured directly had only a small effect on the estimate of carbon stock for the entire cohort of trees. None of the methods developed using the same sample of Panamanian trees gave stand level estimates of carbon stock that differed by more than 10% from the best estimate for either species. Another sample of slightly larger D. panamensis trees growing in 5and 6-year-old mixed-species plantations in Costa Rica [J. Trop. For. Sci. 13 (3) (2001) 450] was used to develop a second set of regression equations. We hypothesized that a regression equation would give a more accurate estimate of carbon stock if the range of tree sizes used to produce the regression more closely matched the range of sizes that the regression was being applied to. When the Costa Rican equation developed using the full range of trees was compared to a Panamanian equation developed using the full range of tree diameters that we sampled, the estimates of carbon stock for the Panamanian plantation differed by 10.2%. However, when two additional regression equations were created using the range of tree diameters that overlapped, the estimates of carbon stock for the Panamanian plantation differed by only 5.2%, supporting our hypothesis. # 2003 Elsevier Science B.V. All rights reserved.