Summary
To estimate carbon emissions from selective logging in Central Africa, we employed the reduced-impact logging for carbon emissions reductions (RIL-C) protocol to quantify baseline carbon emissions from legal timber harvests by source (i.e., hauling, skidding, and felling). We modeled the relationships between emissions and biophysical conditions, logging practices, and forest policies and then used these models to estimate potential emission reductions from full implementation of RIL-C practices. We applied the method in 8 forest management enterprises (FMEs; i.e., concessions) in the Democratic Republic of Congo (DRC), 9 in Gabon, and 6 in the Republic of Congo (RoC). Committed logging emissions expressed per cubic meter of timber harvested (to control for differences in logging intensities) ranged from 0.63 Mg C m(-3) in a FME in RoC to 4.8 Mg C m(-3) in a FME in Gabon, with an overall average of 2.1 Mg C m(-3). Logging emissions were dominated by damage caused by road and log landing construction (i.e., hauling; 50%) and felling (43%; includes carbon in extracted logs). Total emissions represented only about 9% of unlogged forest biomass carbon stocks. Average emissions were highest in Gabon (2.65 Mg C m(-3)) followed by DRC (1.84 Mg C m(-3)) and RoC (1.54 Mg C m(-3)). Emissions from concessions certified by the Forest Stewardship Council (FSC, N = 6) and those that were not certified (N = 17) did not differ. Nearly half (51%) of logging emissions could be avoided without reducing timber yields if all best examples of RIL-C logging practices observed were applied in the same FME. At the country level, if all FMEs were to utilize these practices, emissions reductions would be 34% in RoC, 45% in DRC, and 62% in Gabon. When combined with country-level logging statistics, emissions from selective logging as currently practiced in the six countries of the Congo Basin are equivalent to 40% of the region's total emissions from deforestation.