Luyssaert et al.1 reported that unmanaged, old-growth forests continue to sequester atmospheric carbon (C) at a rate of 2.4 ± 0.8 Mg C ha–1 yr–1 for stand ages exceeding 200 years. This claim was based on an analysis of net ecosystem productivity (NEP) and other C flux data from temperate or boreal forest plots compiled from published studies and databases. Their conclusions and quantitative estimates are widely cited and essential in the debate on the role of forests in climate mitigation. Thus, it is important to discuss the validity of these estimates.

Their conclusion was surprising given that such forests were thought to be C neutral2. Using data from several different viewpoints (for example, global net land flux, soil C accumulation rate, N demand to sustain C accumulation) and a re-analysis of the data provided by Luyssaert et al.1, we find evidence that the forest NEP values reported in the paper are markedly overestimated. Our analysis suggests that NEP is only 1.6 ± 0.6 Mg C ha–1 yr–1 for forests more than 200 years old, or about one-third less than the value reported by Luyssaert et al. Moreover, comparisons of net primary production (NPP) and heterotrophic respiration (Rh) suggest that some old unmanaged forests might serve as C sources. These inconsistencies bring into question the value of the NEP flux data given in Luyssaert et al.1 and their conclusion that unmanaged old-growth forests are important C sinks.