Resin pocket occurrence in Norway spruce depending on tree and climate variables

Abstract Resin or pitch pockets cause a significant reduction in timber quality of Norway spruce ( Picea abies (L.) Karst). Their occurrence is very variable within the stem and between trees and stands. Researchers have suggested several causes of resin pocket formation with a majority arguing in favour of storms as the main causing factor. The objective of this study was to examine the influential factors and to model the probability of resin pockets in the stem predicted by single tree descriptors and climatic variables. For this reason 29 trees from a long-term thinning experiment were felled at the age of 37 years. The samples were chosen along a stand density gradient. A tree ring analysis was carried out on six disks per stem, including an exact dating of the resin pockets. The tree dimensions including tree growth were reconstructed for each year. These variables were correlated with wind and precipitation variables during that growth period. Statistical tests and correlation analysis were performed in order to identify the influencing factors and a two step generalized linear mixed regression model (GLMM) was used to model the resin pocket occurrence within a tree ring. The first step was a logistic GLMM that was fitted to determine the probability of a resin pocket occurrence. In a second step we fitted a Poisson GLMM to the data sets with resin pocket numbers greater than zero in order to model the number of resin pockets per year ring. The results suggest significant influences of tree ring width, height at the stem and the crown length as well as an influence of average wind speed and drought. Conversely, there was no influence of storm incidents, which is contrary to the prevailing storms hypothesis. In contrast, our results support the drought hypothesis.