Especially in closed-canopy stands, annual growth rings are simultaneously affected by various environmental factors: not only regional forcings such as climate fluctuations but also gap-scale, ecological dynamics such as competition and stochastic disturbance. Although these local effects are often discarded by dendrochronologists as a noise, the dendroecological signals are valuable information of past forest dynamics. Since dendroecological trends arisen from local stand dynamics often have medium frequencies in persistence ranging from several years to a few decades, it is usually difficult to separate the local, gap-scale forcings from regional, medium-frequency forcings such as El Nino Southern Oscillation, North Atlantic Oscillation, and solar cycles. Moreover, conventional dendroecological practices failed to analyze the continuously changing medium frequency trends. In this study, a method to obtain a continuous index of medium-frequency dendrochronological trends was developed, by generalizing previous analytical methods that evaluate relative changes in moving averages. The new method was then tested against a tree ring dataset from a site with known forest history of release and suppression due to a disturbance. To quantify the effects of local gap dynamics against the regional, often climatic effects, increments cores of black spruce (Picea mariana) were sampled from boreal forests in Saskatchewan, Canada, using a stratified sampling design.

Results/Conclusions

Assuming that regional forcings affect trees in the given stand homogeneously and local gap-scale forcings have heterogeneity within stand, the relative effect of stochastic heterogeneity was quantified using the proportion of explained variance in stratified time-series ANOVA. In closed-canopy stands, stochastic local effects explained 12.9-35.4% of variance in tree ring widths, because interactions between neighbor trees were likely to be intense. In open-canopy stands, on the other hand, the proportion of explained variance was 1.4-10.2%, reflecting the less-intense local tree interactions in sparsely wooded stands. Our results were generally consistent to the conventional observations in dendrochronology; in closed, high-density forests, local forest dynamics often played a significant role in medium-frequency patterns in tree rings. The extracted medium-frequency trends were significantly different from raw-data based trends, because the latter were largely dominated by distinct high-frequency signals. These advancements in statistical analysis and study design will help both ecologists and paleo-climatologists to objectively evaluate the forcings of climate fluctuations and ecological interactions, relative to each other.