Short-term effect of the harvesting method on ecosystem carbon budget in hemiboreal Scots pine forest: Shelterwood cutting versus clear-cut

Shelterwood cutting (SC) has been highlighted as an alternative method to clear-cut (CC) -based even -aged forest management. However, compared to CC, the effect of SC on stand carbon (C) balance is still poorly understood at the ecosystem level. We examined the prompt effect of SC versus CC on ecosystem net primary production (NEP) on a short-term scale, using the C budgeting method combined with eddy covariance (EC) measurements in hemiboreal mature Scots pine stands. The early effect of SC on annual C budget after removing 30 - 40 % of the growing stock revealed diverse patterns in the studied stands; NEP varied from 0.62 to -1.3 t C ha -1 yr - 1 for the C sink and the C source, respectively. However, C loss decreased already in the second post -thinning year and levelled out attaining -0.41 t C ha -1 yr - 1 , which is close to balance. Furthermore, C loss declined in the second post -harvesting year at the CC sites as a result of the increased production of the ground vegetation and the decreased soil heterotrophic respiration (Rh) flux. However, C loss from dry mesotrophic clear -cuts was almost -1.8 t C ha -1 yr - 1 for both study sites. Since estimation of the individual C fluxes for C budgeting is associated with variability -induced errors, the estimated values of NEP contain uncertainties of various levels. The estimated annual cumulative Rh flux was significantly higher in the SC versus CC areas and the mean annual Rh values across the study sites were 3.57 +/- 0.25 and 2.59 +/- 0.09 t C ha -1 yr - 1 , respectively. Annual estimated NEE after SC was 1.8 +/- 0.52 t C ha -1 yr - 1 , gross primary ecosystem production was 9.28 +/- 0.97 and total ecosystem respiration was 7.47 +/- 0.28 t C ha -1 yr - 1 . The discrepancy between the estimated values of NEE and NEP was 1.2 t C ha -1 yr - 1 . In the short term SC demonstrated some advantage over CC from the perspective of the C cycle, but the difference in NEP values between the SC and CC treatments was not convincingly overwhelming. Hence, SC allows to maintain the forest cover for a longer period and to avoid drastic changes in the landscape; still, after SC, C budget varied between the C sink and the C source.