TLidar-based crown shape indicates tree ring pattern in Norway spruce (Picea abies (L.) H. Karst) trees across competition gradients. A modeling and methodological approach

Tree crowns and growth rings are physiologically and functionally connected through supporting and resource sharing. Management interventions may strongly influence tree growth by altering this linkage. However, conventional approaches have limited ability to characterize crown shape precisely, thus hindering our under-standing of the relationship between crown shape and tree ring patterns. We, thus, aimed to test three hy-potheses: (HI) Crown shape (regularity vs. irregularity) and ring patterns (regularity or irregularity) are significantly correlated and (HII) vary across density gradients; if so, (HIII) internal ring patterns could be predicted from external crown shape metrics. We, therefore, employed terrestrial laser scan-based crown shape and coring-based tree ring width metrics for Norway spruce (Picea abies (L.) H. Karst.) trees covering a range of density gradients to assess temporal changes of crown shape and tree ring patterns. We found a significant and positive influence of crown shape quantifying metrics on ring patterns, indicating crown regularity or irregularity strongly reflects tree ring regularity or irregularity (p < 0.05). Crown shape and ring patterns always showed comparable patterns across density gradients (e.g., trees from lower-density stands produced transgressive crown and ring growth) and significantly varied across competition level. Trees grown in lower-density stands are more likely to produce upper-reaching crowns (maximum crown radius expansion shifted to the mid-to upper-crown) than trees grown in competitive conditions, which result in lower-reaching crowns (maximum crown radius shifted to the crown base) with reduced crown shape and ring pattern parameters. Crown irregularities increased as density decreased through competition reduction, resulting in more regular ring patterns (stable growth). Since both crown shape and ring patterns are simultaneously impacted by stand density or competition, the relationship between crown shape and ring patterns is competition-neutral. When viewed separately, both patterns had a strong relationship with the competition index. Finally, our comparative model predictions showed that approaches ranging from simple linear models to complex machine learning techniques (e.g., random forest, neural network, support vector machine, etc.) were effective in predicting ring patterns using external TLidar-crown shape, indicating a potential method to evaluate the crown shape and ring pattern link. The relationship between the crown and growth ring and their synchronous patterns across competition gradi-ents suggests that internal growth can be assessed from the external appearances of trees and recommends further consideration in forest modeling.