Ecological factors affecting the abundance of advance regeneration in Quebec's southwestern boreal forest

Boreal forest disturbance regimes have changed during the past century, in turn changing regeneration and stand dynamics of these forests. This is occurring at the same time that the forest industry is emphasizing operations that take advantage of pre-established natural regeneration. This study has therefore investigated the effect of various ecological factors on the abundance of Abies balsamea (L.) Mill., Picea glauca (Moench) Voss, Picea mariana (Mill.) BSP, Thuja occidentalis L., Betula papyrifera Marsh., and Populus tremuloides Michx. natural regeneration in some boreal stands in the Abitibi region of Quebec. Abiotic ecological site classification variables were found to be poorly correlated with seedling densities for most species. Parent trees, as a seed source, and stand type, for its influence on the quality of the seedbed, were strongly correlated with abundant conifer regeneration. As expected, time since fire was positively correlated with seedling abundance for late successional species, whereas it was not related to the abundance of early successional species. Post-fire disturbances, such as those due to spruce budworm (Choristoneura fumiferana (Clem.)) outbreaks, accounted for the variations in seedling densities observed with respect to stand age. The presence of budworm-caused gaps was positively correlated with regeneration abundance, whereas shrub competition, which was greatest in large gaps, was negatively correlated with advance regeneration density. Forest management based on abundance of pre-established regeneration should focus mainly on mixed stands because seedling density is very low in hardwood stands and because of the strong competition found in large gaps of resinous forests. The low number of seedlings observed, especially in coniferous stands, may limit the effectiveness of operations that take advantage of advance regeneration. The large variability observed among site and stand types limits their operational use in predicting seedling densities.