Modeling tree growth in a selectively logged temperate forest

Background/Question/Methods

Single-tree selection logging is similar to “background” tree mortality of individuals in unmanaged hardwood forests, and is believed to be suitable for shade-tolerant species in smaller areas of woodland. However, few studies have specifically evaluated this method for improving quality and controlling stocking of the residual stand necessary for sustained yields. In my dissertation study, I used likelihood methods and model selection to quantify the effect of competition on tree growth. Specifically, I asked the following questions: (1) Are different species of neighboring trees functionally equivalent competitors? And (2) how does the magnitude of neighborhood competition vary with the size and spatial distribution of trees in the neighborhood?

My study system is a secondary oak-mixed forest in New York’s Catskill Mountains, USA. I set out three 40 x 120 m plots, located in the region where single-tree selection treatment was implemented in 2001- 2002. At each plot, I recorded the species, diameter at breast height (1.3 m), and mapped the location of each tree. Increment cores were extracted using an increment borer at a height of 1.0 m for the five major study species. I used a spatially explicit, maximum-likelihood, neighborhood model approach to examine tree competition on radial growth.

Results/Conclusions

In the tree growth results, models that treated neighbors as equivalent in their competitive effects were the most parsimonious for the four species studied due to insufficient neighbor numbers of any given species. One exception was eastern white pine, whose best model estimated separate competition coefficients for all five species of neighbors. I found most species had strong size-dependency in potential growth in the absence of competition. Furthermore, I found strong asymmetries in competitive interaction. Red maple and eastern white pine had strong crowding effects on three oak species, while oak species had little effect on the former.

Overall, my findings highlight the predicted change of residual composition within stands after removal of different species and tree sizes. It is recommended that both eastern white pine and red maple be removed gradually to stimulate growth of residual oak stock. Based on research at Camp Deerpark Forest, it is recommended that other private landowners in the Catskills apply the single-tree selection method for optimal woodland management.