For periodic gross annual board-foot volume growth per acre (Y3) 

 the regression solution to Model 5 was: 



Y3 = 



—381.1 + 7.437 (P) + 6103.202 (-?-) + 9.199 (_§LjE_BA_) _ ^^s (-^UlML) 



A ^ lOO 1000 



where: Coefficient of determination (R2) = .403 

 Standard error = 23.49% 



The solution to the periodic gross annual board-foot volume growth 

 per acre equation is shown in Table 8 and is shown graphically in 

 Figure 3. 



For periodic net annual board-foot volume growth per acre (Y4) 

 the regression solution to Model 5 was: 



Y4 = 



—255.605 + 5.904 (P) + 4222.069 i-h + 9.487 (-^I^LM.) _ .162 ( ^^ "l^^' ) 



A 100 1000 



where: Coefficient of determination (R-) = .273 

 Standard error — 27.39% 



The solution to the periodic net annual board-foot volume growth 

 per acre equation is shown in Table 9 and is shown graphically in 

 Figure 4. 



Although site index by basal area squared interaction ( ) 



and age variables in both equations were not significant they were re- 

 tained in the model. All other variables in these equations were sig- 

 nificant (5-percent level). Since the percent basal area in sawtimber 

 is fixed at 70 percent, Figures 3 and 4 are apt to be a bit misleading 

 because at very high stand basal areas the percent sawtimber would 

 likely be less than 70. 



16 



