Table 3 . --Equations for live crown weight of trees less than 2 inches d.b.h. 



Spe- 

 cies 



MSR 



1/ 



Range 

 in ht 



Equations 



Weight at 

 5 £t^ 10 ft 



- -Ft- 



Lh 



DOMINANTS 



DF 



0. 83 



1 . 



156 



11 



3 



7-11 . 



4 



w = 



EXP[-4.212 + 2.7168(lnh)] 



1 



2 



7 



7 



s 



. 94 



7 



290 



12 



1 



9-10 



4 



w = 



EXP[-3.932 + 2.571 (lnh)]^ 



1 



2 



7 



3 



AF 



. 90 



1 



343 



13 



2 



5- 9 



9 



w = 



EXP[-3.335 + 2.303(lnh)] 



1 



4 



7 



2 



c 



. 90 



1 



908 



12 



1 



8-10 



1 



w = 



0. 04833 (h^) 



1 



.2 



4 



8 



GF 



. 79 



3 



138 



12 



3 



1-14 







w = 



0.4284 (h) 



2 



. 1 



4 



3 



PP 



.81 



1 



235 



12 



2 



3-10 







w = 



0.3451(h) 



1 



7 



3 



4 



WP 



.97 







190 



13 



2 



8-11 



5 



w ~ 



0. 3292(h) 



1 



7 



3 



3 



LP 



.96 





220 



12 



1 



6-13 



1 



w = 



0.03111(h2) 







.8 



3 



1 



WBP 



.93 





085 



10 



2 



5-10 







w = 



0.070 + 0.02446(h2) 





7 



2 



5 



WH 



.91 



2 



221 



12 



3 



6-13 



6 



w = 



2/ 



EXP[-5.126 + 2.563(lnh)]^ 





4 



2 



2 



L 



. 80 



1 



230 



12 



2 



8-18 







w = 



0.1128(h) + 0.00813(h2) 





8 



1 



9 



INTERMEDIATES 



GF 



.85 



.374 



9 



3 



7- 



9 



5 



w = 



0.0538(h2) 



1 



.3 



5.4 



C 



.96 



. 142 



11 



3 



7- 



10 



4 



w = 



0.0307(h2) 





.8 



3.1 



DF 



.66 



.985 



10 



3 



6- 



15 



6 



w = 



EXP [-2. 8065 + 



1 .4802(lnh)]^ 



.6 



1.8 



PP 



. 35 



.268 



10 



3 



9- 



14 



6 



w = 



EXP[-2.7297 + 



2/ 



1.1707(lnh)]-^ 



.4 



1.0 



2-/ MSR indicates mean square residuals. For logarithmic functions, MSR was 

 calculated as Z(P-0)'^/df, where P and are predicted and observed values trans- 

 formed to arithmetic units and df is the residual degrees of freedom. 



2J 



These equations are of the form Iny = a + blnX + (mean square error/2). 

 The latter term corrects for bias in transforming logs and is included in the 

 intercept term in the equations. The intercept term was adjusted by (mean square/2) 

 when the summation of predicted minus observed values in arithmetic units showed 

 less bias with the correction term than without it. 



14 



