A dendrochronological reconstruction of sugar maple growth and mortality dynamics in partially cut northern hardwood forests

Abstract

We investigated the growth and mortality dynamics of 22 northern hardwood stands with well-known disturbance histories in southern Quebec, Canada. The main objective of this study was to evaluate the effect of short- and long-term growth trends on both the probability of mortality and the radial growth response of trees following partial cutting. This was first achieved by identifying growth release episodes and characterizing growth patterns prior to tree death from entire tree-ring series of 86 live and 26 dead sugar maple (Acer saccharum Marsh.) trees. Using mixed-effects logistic regression, we quantified the effects of short- and long-term growth trends, as well as external variables such as neighbourhood competition and tree vigour status on the probability of mortality and growth release events. Results showed that long periods of continuous growth decline or long periods of slow radial growth after a severe decrease occurred in 88% of dead trees. These growth patterns were associated with a progressive decline in tree vigour that was mostly initiated by drought and insect defoliations during the 1980s, thereby supporting the relevance of studying long-term tree-ring chronologies to deduce causes of post-harvest mortality. Our model showed that predictions of post-cut mortality events could be significantly improved by considering long-term radial growth trends (25 years before cutting) in individual trees. In terms of growth response, 24% of live trees had a release event occurring after the selection cut, which is slightly lower than the response induced by natural disturbance in previous decades. None of the tested long-term growth variables were significantly related to the occurrence of growth release events. Consequently, suppressed sugar maple trees appear to maintain their capacity to respond positively to gap creation, regardless of the length of the suppression period. We conclude that while decreasing trends in long-term radial growth increase mortality probabilities, the growth release potential of trees is negatively impacted by the presence of defects affecting tree vigour (e.g. fungal infections and cambial necroses) and positively related to the removal of neighbouring trees, rather than to prior growth trends.