MATERIALS AND METHODS 



Four clearcut areas, supporting natural regeneration of western hemlock and Sitka 

 spruce less than 20 years old, were sampled. Areas were selected on the basis of 

 comparable age, stand origin, macrocl imate , and documented occurrence of S. stvobvlvnus . 

 Clearcuts were from 0.2 to 0.5 mi apart and all on recent glacial outwash. One of the 

 four areas had been thinned; the other three were unthinned. 



The three unthinned areas had been clearcut in 1959-1962, so the oldest regenera- 

 tion was 16 years. One of these areas had received 464 kg Agricultural Grade urea/ 

 hectare in May 1969; the two remaining areas were untreated. A line of 4 plots, 100 m 

 apart and 4 m 2 area, was established in each of the three areas. Total number of (1) 

 trees, (2) infected trees, (3) dead trees, and (4) potential crop trees (dominants or 

 codominants) were recorded for each plot. 



Five potential crop trees were selected from each plot for second-order sampling. 

 These trees were cut at ground line and removed from the plot. Total height, total 

 number of branches, and year of dead terminal leader (6 years accumulation) were 

 recorded for each tree. Six primary branches (two lower, two middle, and two upper 

 crown) were removed from each tree and used for third-order sampling. Total length of 

 primary, secondary, tertiary, etc., branching and total length of branching with 

 symptoms of Sivococous infection were recorded for each primary branch sampled. The 

 ratio of infected branch length: total branch length for the tree was calculated and 

 used as a disease index. An analysis of variance was conducted for the disease index 

 data from the three unthinned areas, using the following model for an unbalanced 

 design (Snedecor 1959) : 



Y = y + y • + a.,.. + E.. 7 



I JO) 



where 



Y = dependent variable 

 y = mean 



Y • = treatment 



t 



a.,., = plot nested in treatment 

 Z . = error 



The fourth area sampled was a 0.4 ha plot that was mechanically thinned in 1975 to 

 a 2.4 x 2.4 m spacing. Maximum age of the regeneration was 20 years. Twelve western 

 hemlock trees, comparable in size, were selected from the plot and sampled in the same 

 manner as those from the previous three areas. The disease index data were not included 

 in the analysis of variance because the severely infected western hemlock was removed 

 during the recent thinning and Sitka spruce was favored in selection of crop trees. 



3 



