Table 23 . --Definition of independent variables and abbreviations 



Abbreviation Independent variable 



D Diameter class size (e.g., for the diameter class 8.6 to 9.5, D = 9) 



S Site index 



GRF Average annual 5 -year growth period rainfall 



ARF Average annual 5-year rainfall 



BA Total stand basal area per acre 



CCF Crown competition factor 



T Total number of trees per acre 



PCT Percentile in the basal area distribution 



DCP Diameter class size divided by the diameter of mean basal area 



LBA| 



Total 



basal 



area 



under 



a 



given diameter class 







MBAi 



Total 



basal 



area 



in a 



giv 



en diameter class 







UBAi 



Total 



basal 



area 



over 



a g 



iven diameter class 







LBA2 



Total 



basal 



area 



under 



(a 



given diameter class 





1) 



MBA2 



Total 



basal 



area 



in (a 



gi 



ven diameter class ± 



1) 





UBA2 



Total 



basal 



area 



over 



(a 



given diameter class 



+ 



1) 



LBA3 



Total 



basal 



area 



under 



(a 



given diameter class 





2) 



MBA3 



Total 



basal 



area 



in (a 



given diameter class ± 



2) 





UBA3 



Total 



basal 



area 



over 



(a 



given diameter class 



+ 



2) 



TIME 



Number of 5- 



-year 



periods 



since last cutting 







The first set of screening runs was made on the separate plots (61, 62, 71, 72, and 

 Taylor Woods separately) with the purpose of identifying those independent variables most 

 highly correlated with the log of basal area growth, and eliminating those independent variables 

 that proved to be poor(er) predictors. To do this, the independent variables were classified 

 into sets, and then each set was placed in a group (table 24). A set is defined as one or 

 more independent variables such that, when chosen by combinatorial rules, all independent 

 variables in the set are included (or excluded when not chosen) . The only sets containing 

 more than one independent variable were those formed with the three independent variables 

 LBA^, MBA^ , and UBA^, or their squares. 



56 



