Managing harvest residues: Implications for terrestrial forest carbon stocks in eastern Canadian forests

Harvest residue management influences forest carbon (C) dynamics, but its long-term, pool-specific effects remain poorly quantified. At eight experimental sites in Québec, Canada — two intensively monitored (Montmorency Forest, FOM; La Tuque, VLT) additionally including an extreme residue removal treatment and an unharvested reference forest — we measured all major ecosystem C pools (living biomass, detrital mass, forest floor, mineral soil) and forest floor and mineral soil nitrogen (N) over 10 years after clearcut harvesting, in coniferous and mixed stands. We compared residue retention, standard removal, and extreme removal. Coarse (CWD), small (SWD), and fine (FWD) woody debris, litter and moss, stumps, and the forest floor coarse fraction (>6 mm) declined across treatments over the decade. Woody debris C showed the strongest responses to removal: for most detrital pools, the difference between retention and standard removal was no longer significant by Year 10, whereas CWD C remained significantly lower under extreme removal throughout. Mineral soil C showed a small but significant reduction under removal while no significant differences were found at the ecosystem scale. Forest floor and mineral soil N stocks were unaffected, but declining soil C:N ratios indicate that the soil N cycling is still changing 10 years after harvesting. Total ecosystem C was slightly but persistently lower under removal, reflecting additive small differences across pools; recovery was faster in mixed than coniferous stands. Field-based woody debris decay rates exceeded CBM-CFS3 default values, suggesting that distinguishing decay by forest type and size class would improve estimates of residue residence time and forest bioenergy carbon parity.