Longleaf pine forests were once an important forest ecosystem in the southeastern United States and their restoration has been suggested as pathway to increase the resilience of forests to changing climate. Little is known about the potential role of longleaf pine restoration to the southern landscape in mitigating greenhouse gas emissions. We quantified ecosystem carbon (C) stocks in 20 longleaf pine stands ranging in age from 5 to 118 years located across the species range in Georgia, Louisiana, North Carolina and Florida. We sampled aboveground biomass of 127 trees and below-stump biomass (including the entire tap root) of 86 trees to develop robust allometric equations for longleaf pine and quantified C density in live tree biomass, live understory biomass, standing dead trees, down dead wood, the forest floor and soil to a 1 m depth. Ground penetrating radar (GPR) was used to quantify coarse root biomass outside excavation pits. Our specific objectives were to: (1) quantify C density (Mg C ha-1) in longleaf pine forests representing a range in age, forest structure, management, and site quality, (2) derive above and belowground C accumulation trajectories for longleaf pine, and (3) determine what forest structural variables were related to ecosystem C stocks.
Aboveground live tree C ranged from < 1 to 78 Mg C ha-1 and increased with stand age following a saturation function. Predicted live tree C accumulation at 100 years was 60 Mg C ha-1. Down dead wood, understory vegetation and forest floor C pools were relatively small components of total ecosystem C. Ecosystem C, which included all above and belowground pools except soil C, was dominated by C in live trees and increased with stand age following a saturation function, with a predicted asymptotic maximum of 119 Mg C ha-1. Residuals were linearly and positively related to basal area but not tree density. Total root C increased with stand age and was as high as 25 Mg C ha-1; whereas the root to shoot (R/S) decreased with stand age. Excluding the 5-year-old grass stage stand, mean R/S was 0.26 if only caorse roots in excavation pits were considered and increased to 0.52 with inclusion of GPR detected roots and fine roots. It is unclear if high R/S values were a result of more intensive sampling or proportionally greater allocation to far-reaching lateral roots. In either case, these results indicate a larger-than-suspected root C pool in longleaf pine ecosystems.