Long-term carbon loss and recovery following selective logging in Amazon forests
Article first published online: 30 SEP 2010
Copyright 2010 by the American Geophysical Union.
Global Biogeochemical Cycles
Volume 24, Issue 3, September 2010
How to Cite
2010), Long-term carbon loss and recovery following selective logging in Amazon forests, Global Biogeochem. Cycles, 24, GB3028, doi:10.1029/2009GB003727., and (
- Issue published online: 30 SEP 2010
- Article first published online: 30 SEP 2010
- Manuscript Accepted: 1 JUN 2010
- Manuscript Revised: 12 MAY 2010
- Manuscript Received: 16 NOV 2009
- carbon dynamics;
- selective logging;
 Amazon deforestation contributes significantly to global carbon (C) emissions. In comparison, the contribution from selective logging to atmospheric CO2 emissions, and its impact on regional C dynamics, is highly uncertain. Using a new geographically based modeling approach in combination with high resolution remote sensing data from 1999 to 2002, we estimate that C emissions were 0.04–0.05 Pg C yr−1 due to selective logging from a ∼2,664,960 km2 region of the Brazilian Amazon. Selective logging was responsible for 15–19% higher carbon emissions than reported from deforestation (clear-cutting) alone. Our simulations indicated that forest carbon lost via selective logging lasts two to three decades following harvest, and that the original live biomass takes up to a century to recover, if the forests are not subsequently cleared. The two- to three-decade loss of carbon results from the biomass damaged by logging activities, including leaves, wood, and roots, estimated to be 89.1 Tg C yr−1 from 1999 to 2002 over the study region, leaving 70.0 Tg C yr−1 and 7.9 Tg C yr−1 to accumulate as coarse woody debris and soil C, respectively. While avoided deforestation is central to crediting rain forest nations for reduced carbon emissions, the extent and intensity of selective logging are also critical to determining carbon emissions in the context of Reduced Emissions from Deforestation and Forest Degradation (REDD). We show that a combination of automated high-resolution satellite monitoring and detailed forest C modeling can yield spatially explicit estimates of harvest-related C losses and subsequent recovery in support of REDD and other international carbon market mechanisms.