3. Same as 2 with commercial thinning at 50 years to 

 100 trees per acre, such that the average d.b.h. before 

 thinning is the same as after thinning (PCI). 



4. Same as 3 but with commercial thinning from above 

 at 50 years to 100 trees per acre (PCA). 



5. Same as 3 but with commercial thinning from below 

 at 50 years to 100 trees per acre (PCB). 



Yields for natural stands are based on the clearcutting 

 method of regeneration, with site preparation consisting of 

 64 percent of area burned and an additional 11 percent 

 mechanically disturbed. No thinnings are scheduled. 



METHODS 



Growth and yield projections for 150 years were made 

 using version 5.2 of the Stand Prognosis Model. 



Geographic, topographic, and ecologic data to be used in 

 the simulations were obtained from inventory plots that 

 were a random sample of the forest. These plots represent 

 the range and joint occurrence of site variables throughout 

 the Inland Empire. The inventory plots were used to de- 

 fine the distribution of five major site variables influencing 

 tree growth: geographic location (defined by National For- 

 est), habitat type (Daubenmire and Daubenmire 1968), 

 slope percentage, aspect (degrees from north), and 

 elevation (hundreds of feet). 



Combinations of these five variables constituted a "cell," 

 each cell corresponding to a unique set of model coeffi- 

 cients in the Prognosis Model. The values used for the 

 slope and elevation variables were the mean ±1 standard 

 deviation within the geographic and habitat type classes. 

 Aspect variation is represented by two classes correspond- 

 ing to the maximum and minimum of the aspect effect on 

 increment for that species. The appendix contains detailed 

 information on the combinations of these variables and the 

 nominal values used for the three planted species and 

 naturally regenerated stands. 



Initial stand conditions were defined by the Regenera- 

 tion Establishment Model (Ferguson and others 1986) 

 component of the Prognosis Model. 



Site index was calculated for each projection by monitor- 

 ing the simulated height growth for each tree record. 

 Height at 50 years (rings at 4.5 feet) was calculated for 

 each tree, then site index calculated as the average of the 

 20 trees per acre having the greater 50-year heights. 



However, to prevent management treatment from influ- 

 encing site calculations, each cell was assigned the site 

 index value derived for the planting-no thinning manage- 

 ment regime for that species. 



Yields from the simulation were then sorted by 10-foot 

 site index classes and management treatment and a 

 weighted average yield value was calculated. Weights for 

 each cell were proportional to the frequency with which 

 inventory locations fell within the cell. 



Although the site variables define a wide range of site 

 productivity, there is still substantial variability within a 

 cell. To represent this variability, multipliers of basal area 

 increment for trees larger than 3 inches d.b.h., and height 

 increment for smaller trees (<3 inches d.b.h.), were as- 

 signed to each cell. The multipliers were entered with the 

 BAIMULT and REGHMULT keywords (Wykoff'and others 

 1982). Logarithm of BAIMULT was selected randomly 



from a normal distribution having a mean of zero and 

 standard deviation of 0.3. The choice of standard deviation 

 was based on estimates of residual variation in growth 

 histories of 102 permanent sample plots in northern Idaho 

 (Stage 1987). 



DEFINITIONS OF VARIABLES 



Site index. Average height in feet of the tallest 20 trees 

 per acre when the trees have 50 rings at 4.5 feet above 

 ground level. 



Age. Number of years from planting or from date of har- 

 vest or site preparation for natural-stand yields. 



Top Height. Average height in feet of the 40 largest di- 

 ameter (d.b.h.) trees per acre. 



Volumes. Reported for three categories of size limits: 



Total cubic feet includes all stems, from 1-foot stump to 

 tip of main stem. 



Merchantable cubic feet includes all stems greater than 

 7.0 inches d.b.h. from 1-foot stump to a top diameter in- 

 side bark of 4.5 inches. 



Board feet Scribner (Kemp 1956) includes all stems 

 greater than 7.0 inches d.b.h. from 1-foot stump to a top 

 diameter inside bark of 4.5 inches. 



Accretion (ACC in yield table headings). Periodic mean 

 annual change in total cubic foot volume of trees surviv- 

 ing through the period,the length of which is indicated 

 by column headed PRD yrs. 



Mortality (MOR in yield table headings). Total cubic foot 

 volume per year of trees dying in the period. 



Total Yield. Standing volume plus volume removed by 

 thinning, if any. 



YIELD TABLES FOR NATURAL 

 REGENERATION 



Table 1 and figures 1, 2, and 3 represent the yields for 

 the mix of species typically found in naturally regenerated 

 stands of the grand fir-cedar-hemlock ecosystem. Grand 

 fir is the most prevalent species followed by Douglas-fir 

 and western white pine (Pinus monticola). 



Entries for ages 5 and 10 represent only the established 

 regeneration that would have been inventoried at age 5 

 years. From age 15 years on, the trees are those that 

 would have been inventoried at age 15. Any subsequent 

 regeneration is ignored. 



Trends of tree d.b.h. with age and site index are dis- 

 played in figure 4. 



The distribution of site indices represented in these 

 simulated yields (fig. 5) was determined by the distribution 

 of site variables in the inventory and by the range of the 

 random variation of the growth multipliers. 



Individual simulations showed surprising range in yield 

 for the same site index. These variations are attributed to 

 factors influencing initial stocking and time to reach 4.5 

 feet height. Figure 6, for example, shows a range in yield 

 of approximately 10,000 cubic feet for site index class 60 at 

 120 years. This variation is nearly as large as the range of 

 variation of the mean yields over the full range of sites 

 represented in the jrield tables. 



2 



