HERITABILITY AND RUST RESISTANCE IN WHITE PINE 401 



Table 1. Model for analysis of variance (testers and candidates 

 assumed random; fixed replications) 



Me an ~~ 



a b 



Source of variation d.f. squares Expectation of mean squares 



Replications K-\ MS p 



Testers 7-1 MS T o 2 g ♦ i o 2 fc - i a 2 d ♦ &*„ - KJo^ 



Candidates J-l MS r a 2 +io 2 - - i o 2 , + Ko 2 „ r + Kla 2 r 



Tester X Candidate (J-l) (J-l] MS mn a 2 +ic 2 . - -^ o 2 + Zo 2 



TC e m b m d TC 



Tester -Candidate 



combinations 



X Replications (IJ-1) (K-l) MS„„ D a 2 +ia 2 , -ia 2 , 



r TCi? e m b m d 



1, J, and K = total numbers of testers, candidates, and replica^ 

 tions, respectively. 



Formulas for estimating individual variance components are as 

 follows: 



:,'5_ - ::s 



rp L + k ~'rpp 



— ^77 a 2 = variance due to testers 



MS Q - MS TC 

 C KI " C 



c 2 = — „r — a 2 = variance due to candidates 



MS TC - MS 



Q mf = y ° rr = variance due t0 interaction 



of testers and candidates 



c 2 - -=r0 2 , = MS mm --=-0 2 , a 2 = variance due to effect of 



e m w d l ^TCR m u b 



e plot 



a 2 , = variance due to effect of 

 individuals 



a 2 -, =821 a 2 -, = variance due to effect of 



UK 



binomial sampling 



m = harmonic mean number of 



1 — seedlings 



m . .-, fe 



