1. As the role of forestry-based options for reducing or mitigating greenhouse gas emissions is debated by policymakers, there is a need to inform the debate by synthesizing existing information on carbon dynamics in tropical forest systems and by applying this information to a range of possible interventions in forestry.
2. To investigate the consequences of reductions in logging damage for ecosystem carbon storage, we constructed a model to simulate changes in biomass and carbon pools following logging of primary dipterocarp forests in south-east Asia. We adapted a physiologically driven tree-based model of natural forest gap dynamics (FORMIX) to simulate forest recovery following logging.
3. Following selective logging, simulated ecosystem carbon storage declined from prelogging levels (213 Mg C ha−1) to a low of 97 Mg C ha−1, 7 years after logging. Carbon storage in biomass approached prelogging levels about 120 years after logging.
4. The relationship between fatal stand damage and ecosystem carbon storage was not linear, with biomass recovery following logging severely limited by 50–60% stand damage.
5. Results from simulations suggest that when 20–50% of the stand is killed during logging, replacing persistent forest species with pioneer tree species can reduce the site's potential for carbon storage by 15–26% over 40–60 years.
6. Reducing fatal damage from 40% to 20% of the residual stand, as was the case with a pilot project in Malaysia, was associated with an increase of 36 Mg C ha−1 in mean carbon storage over 60 years.
7. Efforts to monitor and verify the benefits, either through carbon sequestration in new growth or carbon retention in existing biomass, of offset projects involving tropical forests and natural forest management should focus on above-ground biomass, particularly the large trees. Selection of the most appropriate allometric equations for a site and species is important because of their influence on biomass estimates.