Ecological drivers of post-fire regeneration in a recently managed boreal forest landscape of eastern Canada

Clearcutting practices combined with the predicted increase in fire activity may induce post-fire regeneration failure in boreal forest landscapes. This study aims (1) to evaluate if recently managed landscape by clear cut logging is susceptible to be affected by post-fire regeneration failure; and (2) to explore the ecological drivers of black spruce (Picea mariana (Mill.) BSP) post-fire regeneration. In 2014, we surveyed the regeneration of 36 stands in northwestern Quebec that had burned in a major fire in 2005. Fire severity was evaluated for each site with the differenced Normalized Burn Ratio. Using linear models, we explored the relationship between environmental variables (fire severity, pre-fire stand maturity, nature of the seedbed, and physiographic variables) and black spruce post-fire regeneration. Black spruce post fire seedling density was highly variable (range: 25-16 000 seedlings/ha; mean +/- standard deviation: 4549 seedlings/ha +/- 4752) within the studied fire, but did not significantly differ between stands that had been logged 50 years prior to fire and those that were mature prior to the 2005 fire. However, post-fire regeneration failure (defined as <40% stocking that corresponds in our study region to a regeneration density <1750 seedlings/ha) was observed in 48% of the stands that had been logged, but only in 29% of the stands that were mature prior to the fire. The presence of residual trees left after clearcutting may explain why regeneration level was relatively good (>50%) in stands affected by past logging activities. Our study illustrates how biological legacies, environmental conditions and fire severity determine post-fire recovery and resilience of black spruce-dominated ecosystems of eastern Canada. By identifying the drivers of post-fire regeneration success, our study will help forest managers allocating resources where restoration of productive forest are truly needed. (C) 2017 Elsevier B.V. All rights reserved.