Properties of boundary-line release criteria in North American tree species

Boundary line release criteria are increasingly applied to evaluate forest disturbance histories from tree-ring data. However, a number of important properties central to the technique have not been evaluated, including: (i) the ability of boundary line release criteria to standardize releases across various sites, species, and tree life stages (ii) the minimum sample sizes necessary for developing boundary lines, and (iii) the degree to which the criteria can resolve the degree of crown exposure following a disturbance event. In an analysis of eleven North American tree species, boundary line release criteria do not fully compensate for declines in release response a tree experiences with increasing age and size, with the exception Tsuga canadensis. A bootstrapping analysis indicates that approximately 50 000 ring width measurements are necessary to develop boundary line release criteria for a given species. In a Quercus prinus stand, boundary line release criteria better predict the degree of crown exposure following a disturbance than an earlier running mean technique. Despite certain limitations, boundary line release criteria have the potential to standardize release calculation across most life stages of a tree, and possibly among sites and species.