Comparative analysis of dendroecological methods used to assess disturbance events

Abstract

The use of dendroecological methods to identify and describe disturbance histories has become quite common, and numerous techniques have been developed. These methods simultaneously use accurately dated and measured tree rings and objective criteria to identify release events; i. e., sustained increases in radial growth. The goals of this investigation were to perform a literature review to identify the various methods used to identify release events, critically compare the methods, and use an appropriate method to describe the disturbance history of a mixed mesophytic old-growth forest of Ohio, U. S. A. Application of 30 different methods resulted in substantial differentes in the number of disturbance events and the disturbance return interval. Analysis of individual growth segments revealed a large portion to be non-normally distributed. Furthermore, non-normality was especially high during identified disturbance events. We subsequently modified an existing disturbance identification method by changing the measure of central tendency from the mean to the median growth rate since medians are more robust estimators of central tendency than means. Application of this running median method (> 25 or 50 % increase in radial growth when medians of adjacent 10-yr growth segments are compared) to tree-ring series obtained from the study site led to the identification of numerous synchronous and asynchronous radial-growth increases among the individual tress. Based an the observed release patterns, we hypothesize that the disturbance regime of the old-growth stand is best categorized by gap-phase dynamics.