and latitude make interpretation difficult, cone yield 

 per tree and locality appears weakly associated with 

 tree age and tree elevation. Significance of the in- 

 dicated correlation between cone yield of localities 

 and locality latitude is doubtful because values for 

 Boulder Creek (highest latitude and high elevation) 

 exert enough influence to inflate the correlation 

 coefficient over the 5 percent level of significance. 



Yield of Seeds Per Cone 



The average wind-pollinated cone yielded approxi- 

 mately 104 filled seeds; thus, each tree produced 

 nearly 3,000 filled seed per year. However, yields 

 of filled seeds per cone were positively associated 

 with the yield of cones (r = .79, significant at the 

 1 percent level of probability). Mean yield of filled 

 seeds per cone thus varied according to seed years 

 (table 4) , localities (table 5) , and trees within 

 localities (table 5). In fact, yields of seeds per 

 cone ranged from 25 to 149 for two trees in different 

 localities and in different seed years. 



Exogenous variables cannot entirely account for 

 the tremendous variation in total number of filled seed 

 per tree. Tree 212 near Cedars Guard Station and tree 

 58 in the Crystal Creek area (see figure 1) outwardly 

 were similar in terms of age, size, dominance, and 

 local stand density (table 6) . Although the cone yield 

 of tree 212 was about one-half that of tree 58, the 

 seed yield of 212 was less than one-fourth that of 58. 

 Possible genetic control of these traits requires 

 elucidation; their importance in seed orchards might 

 be great. 



Yields From Wind-, Cross-, and 

 Self-Pollinated Cones 



Statistical analyses (paired "t" tests) of balanced 

 samples of filled seeds per cone (table 7) indicate 

 that yields derived from self-pollinations were less 

 (1 percent level of probability) than those of wind- 

 or cross -pollinations ; no differences were detected 



6 



