Data are presented in terms of milacre stocking computed by dividing the number of 

 plots with one or more trees by the total number of plots sampled. 



From file records and field measurements, I obtained information about the topog- 

 raphy, soils, seedbed, shade, competing vegetation, habitat type, distance from timber 

 edge, and history of each area since cutting. 



I used the Daubenmires' key (1968) to name habitat types throughout northern Idaho 

 and eastern Washington. Where a habitat classification system was not yet developed, 1 

 used the tree species that appeared to be climax to identify the habitat of our cutting 

 areas and made no attempt to subdivide further on the basis of understory vegetation. 

 Accordingly, data from all areas are summarized primarily by climax tree species, which 

 are used to approximate the Daubenmires' habitat type series. 



RESULTS 



Because the northern part of the study area has a larger proportion of its forested 

 areas in moist environments than does the south, 1 expected success of stand regenera- 

 tion to decrease from north to south. Survey data show this to be generally the case. 

 The percent of stocked milacre plots in our sampled areas decreased from about 26 in 

 the north to 10 in the south for plantations and from about 47 to 20 percent on areas 

 where natural regeneration was the goal. 



Of even more interest than the geographical relationship was the apparent corres- 

 pondence betvsfeen habitat type and regeneration success. As I was sampling an area on 

 tlie Payette National Forest about midway through the survey, my transects crossed a 

 narrow ecotone between a subalpine fir [Abies lasiocarpa) type and a grand fir (Abies 

 grandis) type. A Douglas-fir plantation was established and growing on the subalpine 

 fir site, but on the grand fir site Douglas-fir survival was near zero. This same 

 phenomenon was observed in other areas. 



It became obvious that summarizing data for other factors without regard to habitat 

 type could be misleading. I lacked a habitat type classification system for most of 

 the area and compromised by ignoring understory vegetation and using the climax tree 

 species to classify the habitat of each plot. 



When summarized in this manner, the data in table 1 suggest that (1] greater 

 stocking of natural Douglas-fir seedlings occur on the western hemlock habitat than on 

 other habitats, and (2) Douglas-fir plantation success on the western redcedar and 

 western hemlock habitats approximately doubles that of v\?armer and drier habitats. We 

 sampled few direct-seeded areas, but those sampled had results similar to those of 

 naturally seeded plots. 



8 



