Complex ecological pathways drive boreal forest successional dynamics

Pervasive shifts in forest successional dynamics are accelerating in response to global environmental change. However, the relative importance of variables driving successional transitions and how these are altered by climate change remain unclear, in part because the ecological pathways among variables can be complex. To untangle the web of interactions driving successional transitions between consecutive catastrophic disturbances, we utilized a long-term forest inventory database repeatedly measured over 43 years, encompassing 3465 boreal forest plots in central Canada and covering stand ages 1–261 years. We developed a hybrid analytical approach that combines boosted regression tree (BRT) and structural equation modelling (SEM). The BRT assessed the relative importance of variables among 37 potentially influential variables. The SEM examined multiple causal pathways of 14 top-ranking (relative importance > 1 %) drivers. Overall, we found an average 4.6 % probability of transitioning to a different forest type over consecutive censuses with a mean interval of 6.8 years. By ranking the relative importance of variables in BRT and SEM, we show that multiple, simultaneously occurring within-community dynamics, rather than climate variations or site and soil conditions, primarily drive successional transitions. In particular, the compositional proportion of the most dominant species was the most influential driver. As it increased from a minimum of 0.24 to monoculture (1.00) in the plot, the likelihood of transition decreased from 41.1 % to 0.2 %, emphasizing slow successional transitions in the mono-species dominated boreal forest. Our empirical findings, spanning the course of secondary succession, suggest that the widely predicted climate-driven transitions in boreal forests may be context-dependent and highly variable than previously thought.