10 20 30 40 50 60 70 



TOTAL HEIGHT IN 1974 



Figure 5 — Influence of tree height on western spruce budworm defolia- 

 tion of Douglas-fir, subalpine fir, and Engelman spruce. 



Defoliation 



Defoliation in 1973 by western spruce budworm was 

 significantly related to the natural log and the coefficient 

 of variation (cv) of initial height, most strongly for sub- 

 alpine fir (table 7). Taller trees sustained more defolia- 

 tion than smaller ones. Higher cv of Douglas-fir height 

 resulted in slightly increased defoliation, whereas the 

 opposite was true for subalpine fir. The only important 

 independent variable influencing defoliation of Engel- 

 mann spruce was natural log of initial height. The coeffi- 

 cient of determination was highest for subalpine fir (0.42) 

 and lowest for Douglas-fir (0.20). Percentage of defolia- 

 tion is shown as a function of tree height for the three 

 species (fig. 5). 



DISCUSSION 



Height and/or diameter growth of all species increased 

 following treatment and were directly related (P ^ 0.05) to 

 at least one of the variables representing competition — 

 degree of overstory removal or change in plot basal area. 

 Change in plot basal area was calculated as the difference 

 between pretreatment and posttreatment basal area; 

 therefore larger values in basal area change represent 

 decreasing competition. Degree of overstory removal also 

 represents decreasing competition. Some of the overstory 

 basal area was included in the calculation of basal area 

 change when an overstory tree occurred on a plot. Thus, 

 the separation of degree of competition due to plot basal 

 area change or degree of overstory removal was not al- 

 ways clear. So both variables need to be considered as 

 measures of stand competition. 



In this respect, the results are sensible. Height growth 

 of Douglas-fir increased with increasing change in plot 

 basal area; height growth of subalpine fir increased with 

 change in basal area and degree of overstory removal 

 (table 3). Engelmann spruce height growth did not re- 

 spond to either measure of competition, however. 

 McCaughey and Schmidt (1982) showed that subalpine 



fir tended to release more than spruce, although in their 

 study spruce released only minimally. Diameter growth 

 of all three species increased with decreasing competition. 

 Douglas-fir and Engelmann spruce grew faster as degree 

 of overstory removal increased, whereas subalpine fir 

 responded to increasing change in plot basal area 

 (table 4). 



Crown ratio was a significant predictor of postrelease 

 d.b.h. growth of subalpine fir. But the sign of the coeffi- 

 cient was opposite of the expected positive relation; better 

 crown ratios were associated with poorer growth. This 

 anomaly remains unexplained. In spite of the statistical 

 significance, this variable added little to the prediction of 

 height growth. A 50 percent change in crown ratio re- 

 sulted in only about 1.2 feet of height growth over 11 

 years. 



There was little physical difference between the com- 

 plete overstory removal on "large" units and complete 

 removal on "small" units. The former were 3 to 5 acres, 

 the latter were about 2 acres and in reality should have 

 been considered identical treatments. This lack of real 

 difference is reflected in the regression models for height 

 and diameter growth. Where degree of overstory removal 

 was statistically important, the coefficients representing 

 both categories of complete removal were similar in size to 

 each other but were quite different from the coefficient 

 representing the partial removal. But because the origi- 

 nal study had broader objectives than reported here 

 (Barger 1980), we chose to keep the treatments separate. 



Notwithstanding the above discussion on the effect of 

 reduced stand competition on tree growth, the variable 

 exerting the most influence on posttreatment height and 

 diameter growth was radial growth prior to treatment 

 (figs. 1, 4). Trees that were vigorous prior to treatment, 

 for whatever reasons, simply outperformed other trees 

 following treatment. This variable was significant in 

 height and diameter models for all three species. 



Initial height was not an important predictor in any of 

 our models. Tall trees responded about as well as short 

 ones. Other researchers obtained somewhat different 



9 



