404 W. A. BECKER AND M. A. MARSDEN 



Table 3. Analysis of variance of adjusted and transformed data 









Elevation 







Source of 



Low 







Mid 





High 



variation 



d.f. 



MS 



d.f. 



MS 



d.f. 



MS 



Replications 



9 



12,694 



9 



11,237 



9 



4,209 



Testers 



3 



2,134 



3 



2,433 



3 



3,023 



Candidates 



50 



877 



20 



714 



16 



1,144 



Tester X Candidate 



150 



361 



60 



374 



48 



413 



Tester- Candidate 















combinations 















X Replications 



1,827 



276 



747 



316 



603 



305 



Table 4. Means of healthy seedlings and variance components 













Elevation 





Component 



Low 







Mid 



High 



Mean (%) candidates 





56.67 







44.86 



54.23 



controls 



33.56 













°\ 





3.48 ± 



2 



.65 



9.81 ± 7.33 



15.36 ± 11.26 



° Z c 





12.90 ± 



4 



40 



8.50 ± 5.64 



18.28 ± 9.75 



° 2 TC 





8.43 ± 



4 



24 



5.75 ± 6.91 



10.73 ± 9.59 



° 2 -i" 2 ., 



e m a 





178.57 







217.07 



174.32 



1 2 



m b 





97.80 







99.43 



131.16 



Standard error of variance components were obtained according to 

 Anderson and Bancroft (1952) . 



Table 5. Heritabilities (%) for different selection methods 

 and elevational zones, when selecting for healthy seedlings 



Selection 







Elevation 





method 



Low 



Mid 



High 



Individual 



Progeny test 

 (= selection 



unit) 



5.0 

 59.3 



3.2 

 38.6 



7.0 

 66.0 



