CO2 Fluxes Of A Boreal Black Spruce Chronosequence In Eastern North America

Forest harvest and subsequent stand development can have major effects on the carbon cycle of boreal stands. Carbon dioxide (CO₂) fluxes of a three-point black spruce harvest chronosequence located in the boreal forest of eastern North America were measured over a one-year period at the ecosystem scale with the eddy covariance technique and CO₂ efflux from soils was measured with a portable infrared gas analyzer. The three sites (pre-harvest, recently harvested, and juvenile) were 105-, 8- and 33-years old, respectively. On an annual basis, the pre-harvest site (EOBS) was a weak carbon sink (6 ± 4 g C m⁻² yr⁻¹), the recently harvested site (HBS00) a source (−87 ± 3 g C m⁻² yr⁻¹) and the juvenile site (HBS75) a moderate to strong sink (143 ± 35 g C m⁻² yr⁻¹). Annual gross ecosystem production (GEP) at the pre-harvest site was only 28% greater than at the recently harvested site (646 ± 6 versus 504 ± 5 g C m⁻² yr⁻¹), while GEP at the juvenile site (1107 ± 32 g C m⁻² yr⁻¹) was more than double that at the recently harvested site, suggesting significant physiological constraints to photosynthesis at the pre-harvest site. Annual ecosystem respiration (R_e) followed the same pattern, but intersite differences were somewhat less (640 ± 8 to 591 ± 6 to 964 ± 50 g C m⁻² yr⁻¹). Annual soil respiration (R_s) decreased following harvest from 593 to 500 g C m⁻² yr⁻¹ and increased with further stand development to 644 g C m⁻² yr⁻¹, although the changes were less than for GEP and R_e. Q₁₀ and R₁₀ of R_s for the snow-free period varied between sites, were lowest for the recently harvested site, and appeared to be related to GEP via substrate supply. The annual ratio of R_s to R_e was lower for the juvenile site (67%) than for the pre-harvest and recently harvested sites (93 and 85%, respectively). These results quantify how some of the major physiological processes that influence the carbon cycle of boreal black spruce stands evolve following harvest and should be useful for better incorporating stand-age effects into regional and global scale models.