Comparison of carbon balances of a Douglas-fir forest from pre- to post-harvest

Stand-replacing disturbances, such as harvesting, have a major impact on the CO₂ exchange between forested land and the atmosphere. Former forest CO₂ sinks become net CO₂ sources due to significantly reduced photosynthetic uptake and continued respiratory losses. Chronosequence studies have been widely used to quantify the effect of harvesting on the carbon (C) exchange of forested stands; however, they assume that all sites differ only in stand age and their abiotic and biotic components have had the same history. This is often not the case so the relationship between C exchange and age obtained from a chronosequence study is not necessarily representative of a particular site.

This study used data from a Fluxnet-Canada Douglas-fir chronosequence on Vancouver Island, where the most mature site (62-year-old; DF49) was fertilized four years prior to harvesting and commercially harvested in 2011 creating a 77-ha clearcut (HDF11). CO₂ exchange was measured using the eddy-covariance technique at DF49 for more than a decade pre-harvest and for two years at HDF11. Automated non-steady-state chambers were used to measure soil respiration (R_s) before and after harvesting. These measurements allowed a direct comparison of pre-harvest and post-harvest C balances, thus eliminating the uncertainty resulting from using chronosequence data to estimate C exchange after harvesting. The weather conditions during the years used in this analysis were similar and within the regional climate normals. Half-hourly energy balance closure was greater than 80% for both sites.

DF49 transitioned from being a moderate C sink (560 g C m^-2 yr^-1) before harvesting to being a strong C source (1000 g C m^-2 yr^-1) in the first year after harvesting. Gross ecosystem photosynthesis (GEP) decreased to almost zero in the first year after harvesting, while ecosystem respiration (R_e) decreased by only 23% from 1460 to 1130 g C m^-2 yr^-1. This could be attributed to decreased autotrophic respiration (R_a) due to the loss of respiring roots, boles, branches and foliage, and increased heterotrophic respiration (R_h) from the decomposition of soil organic matter and logging residue. Pre-harvest at DF49, R_s accounted for 62% of R_e annually, while post-harvest at HDF11 it was 88%. The post-harvest value was similar to the pre-harvest winter value (86%), when R_a was lowest during the year.

C balance results from HDF11 were also compared with those from a previously harvested stand (HDF00) in the same Douglas-fir chronosequence. HDF00 was harvested in 2000 and was located 3 km from HDF11. Large differences in net ecosystem productivity (NEP), GEP and R_e were measured in the two clearcuts in the first year following harvesting, with HDF11 being a much stronger source of C (1000 g C m^-2 yr^-1) than HDF00 (620 g C m^-2 yr^-1). This was partly because recovery of vegetation (planted seedlings, shrubs and herbs) following harvesting was faster at HDF00 than at HDF11. In the second year the annual source strengths were less at both sites but more similar to one another (700 vs. 520 g C m^-2 yr^-1, respectively). R_e was greater in the first year at HDF11 due to a greater amount of decomposing soil organic matter and logging residue, but it became virtually the same as at HDF00 in the second year. These results show that caution is necessary when estimating the effects of harvesting on the C balance at a forest site from CO₂ flux measurements in different-aged stands of a chronosequence study.