Changes in the net carbon balance following a shelterwood harvest at Howland Forest in central Maine seven years after harvest

As CO2 emissions continue to increase, policy-makers are considering various ways to help slow the rise in atmospheric CO2 concentrations. Forests exchange significant quantities of carbon with the atmosphere, so any measures that increase carbon storage in forests could help mitigate rising CO2 emissions. Some proposed C trading markets include payments for enhanced C storage due to changes in forest management, but others exclude management of existing forests due to large uncertainties in sequestration rates, validation, and leakage. Ideally, forest management practices could be designed to provide multiple benefits to society, including provision of wood and paper products, creating economic returns from natural resources, and sequestering C from the atmosphere. To evaluate the impact of a forest management practice on C storage, it is important to quantify both on-site and off-site C fluxes. We began studying changes in C sequestration following a shelterwood harvest at the Howland Forest in central Maine in 2000. Shelterwood harvesting removed about 30% of live aboveground biomass from the forest (15 Mg C ha-1), reduced leaf area by about 40%, and created detrital carbon pools of about 10.5 Mg C ha-1. Net ecosystem carbon storage (NEE), measured using eddy covariance, went from about 1.9 Mg C ha-1y-1 to almost zero in both 2003 and 2004. Live trees, however, stored about 1.5 Mg C ha-1y-1 in 2003 - this was only slightly lower than C storage in live vegetation in the control (unharvested) stand. In 2005, NEE increased to about 1.5 Mg C ha-1y-1 and tree growth increased to about 2.2 Mg C ha-1y-1 in spite of the fact that leaf-area index (LAI) remained about 25% lower in the harvested stand. Soil respiration was significantly lower in the harvested stand, but only in areas impacted heavily by harvest. This is likely due to decreased root respiration as a result of tree removal. When accounting for both on- and off-site carbon pools, this forest returned to being a net C sink in 2005 and has remained a net C sink since that time. In 2007, the combination of C storage in live biomass and wood products put the net C balance at about 0.8 Mg C ha-1y-1. Our results demonstrate that rates of tree growth recovered rapidly after the harvest, returning the system to a net C sink within about four years after harvest. Further research will investigate the mechanisms behind the rapidly increasing tree growth rates after the harvest.