Table Z . --Location and least square mean height of populations according to origin 



Popul a- 



tion Location Latitude Longitude Elevation 16-year height 



Number 





Degrees 



Degrees 



Meters 



cm 



1 



Anderson 



44.55 



116.20 



1520 



292 



2 



Mill Creek 



44.45 



116.30 



1370 



294 



5 



Camp Creek 



44.62 



115.65 



1830 



286 



6 



Crawford 



44.53 



115.97 



1460 



296 



9 



Warm Springs 













Ridge 



43.85 



115.88 



1520 



296 



11 



Lester Creek 



43.62 



115.38 



1830 



281 



14 



Sagehen 



44.32 



116.23 



1580 



308 



15 



Third Fork 



44.37 



116.33 



1250 



309 



16 



Powellson Creek 



44.78 



115.87 



1770 



293 



17 



Circle Creek 



45.05 



116.27 



1280 



311 



18 



Bear Creek 



43.95 



115.47 



1830 



• 275 



19 



Idaho City 



43.72 



115.72 



1460 



301 



22 



South Fork 













Payette River 



44.07 



115.50 



1220 



299 



23 



Mill Creek 



44.67 



116.87 



1250 



309 



24 



Pine Creek 



44.70 



116.83 



1160 



309 



25 



Babbit Crgek 



43.80 



115.72 



1950 



283 



26 



Mica Creek 



44.60 



116.27 



1520 



285 



27 



Cabin Creek 



44.63 



116.27 



1370 



331 



28 



Mud Creek 



45.05 



116.38 



1520 



308 



29 



Rough Creek 



44.90 



116.48 



975 



314 



35 



Dutch Creek A 



43.80 



115.37 



1520 



289 



38 



Circle "C" 



45.05 



116.27 



1220 



318 



39 



Meadow Valley 



44.98 



116.22 



1220 



318 



41 



Dutch Creek B 



43.73 



] 15.45 



1980 



288 



42 



Trail Creek 



43.60 



115.75 



1190 



315 



46 



West Fork Creek 



44.30 



115.87 



1370 



286 



47 



Garden Valley 



44.07 



115.93 



1040 



315 



50 



Bumgartner 



43.62 



115.08 



1310 



271 



53 



Camp Creek 



44.88 



115.70 



1220 



320 



54 



Zena Creek 



45.07 



115.75 



1520 



271 



2 OA 



Idaho City 



43.77 



115.72 



1280 



296 



20B 



Idaho City 



43.87 



115.83 



1280 



282 



21A 



Hazard Creek 



45.18 



116.30 



1070 



321 



21B 



Scriver Creek 



44. 17 



116.00 



1310 



318 



43A 



Rice Creek 



45.75 



116.33 



1430 



301 



43B 



Fleming Creek 



44.03 



116. 10 



1310 



296 



In the multiple regression analysis, elevation, latitude and longitude of the seed source 

 accounted for 51 percent of the variation in mean performance of populations (table 4) . 

 Visual examination of the residuals from regression suggested that transformations of the 

 independent variables would not necessarily improve the fit. The regression statistics show 

 that for an elevational gradient of 1 000 m, mean height of populations decreases by about 10 

 percent; and across a geographic interval of one degree of latitude or longitude, mean height 

 changes by 3 percent and 2 percent, respectively. 



Since the simple correlations (r) of population means at age 16 with those at age 11 were 

 0.78, 0.89, 0.93, and 0.91 at Jack's Creek, Holcomb, Boulder Creek, and Idaho City, respectively, 

 it is not surprising that multiple regression models that fit mean performance at each site to 

 physiognomic variables essentially duplicated results of similar models for height at age 11 

 (Rehfeldt 1980). Moreover, models that involved mean values of 16-year height across all 

 sites (table 4) accounted for more variance than those models involving 11- or 16-year height 

 at a single site. Undoubtedly, analyses that combined observations from all sites provided 

 better estimates of population means and greater statistical power than previous analyses that 

 were made for each site separately. 



4 



