Testing for Change in Structural Elements of Forest Inventories

摘要

In this article we develop a methodology to test for changes in the underlying relationships between measures of forest productivity (structural elements) and site characteristics, herein referred to as structural changes, using standard forest inventories. Changes in measures of forest growing stock volume and number of trees for both hardwood and softwood on forestland in North Carolina are evaluated using plot-level data aggregated at both the state and survey unit level from the last three available completed Forest Inventory and Analysis surveys using exploratory data analysis and nonparametric statistics. When the survey data are aggregated at the state level, we accept the null hypothesis of no discernible between-survey differences in the means of the forest productivity measures for at least 90% of the plots in each of the four models. We also accept the null hypothesis of no discernible between-survey differences in the variance or higher moments of these forest productivity measures for at least 82% of comparisons. At a finer scale, our results show that structural stability is questionable in the Coastal Plain units of North Carolina. Overall, results provide evidence of some structural change in the forests of North Carolina but do not address the causes of such changes. The systematic comparison of forest inventories conducted in this article constitutes a new approach to testing for structural changes in forest relationships, one that can be implemented as a monitoring protocol within standard repeated forest inventories. [PUBLICATION ABSTRACT]