“Mortality or not mortality, that is a question …” How to treat removals in tree survival analysis of managed forests

Abstract Tree mortality is an objective and easily definable criterion for forest health and is particularly suitable for long-term and large-scale studies of forest condition. However, it is not possible to accurately determine actual tree mortality in managed forests that are subject to silvicultural and sanitary, periodic or continuous tree removal. In this case, the only way to approximate actual tree mortality is to define the range in which it occurs by determining minimum and maximum mortality thresholds. For this purpose, we performed a survival analysis and considered the removals as either censored or complete observations. The results obtained showed significant differences, indicating the importance of how removals are classified in the analysis. An attempt to determine the similarity of removed trees to alive or dead trees in terms of defoliation, severity of damage, DBH and age of the trees revealed inconsistencies between species and the year for which this analysis was also performed. Removed trees from species in good health (pine) resembled alive, while removed trees from species in poor health (spruce) resembled dead trees more. This result suggests that in managed stands of healthy species the actual mortality is closer to the minimum, while in species in poor health it is closer to the maximum mortality.