HERITABILITY AND RUST RESISTANCE IN WHITE PINE 403 



AG = S 



a 2 r = cov„, , (candidate half sibs) 



where 



d *2 C . yr^2 rc/7 + b 2 tcr/k 



a 2 = cov r . (candidate half sibs) 

 o i: ^ l^J 



o o 1 o - 1 



a TOT = a % + 1T a 2, ¥ 



But, because pine trees are monoecious the terms combine and 



RESULTS 



Because of the large rumber of individual seedlings involved in each 

 test, the statistical analyses of stage B were done on an electronic com- 

 puter. A special set of computer programs was written to edit the data, 

 compute the statistical analysis, and produce maps (printer plots) of 

 infection locality and intensity (for these maps see McCluskey, these 

 proceedings) . 



Progenies from the three elevation zones were analyzed separately. 

 Included in the program output were: an analysis of variance table; a 

 table of variance components; and a table of the percent healthy seedlings 

 for each of the tester-candidate combinations. This last table was sorted 

 on the average percent healthy for each candidate over all testers. 



The degrees of freedom (df) and mean squares (MS) of the analysis of 

 variance of adjusted and transformed data are given in Table 3. The 

 variance components and their standard errors are given in Table 4. The 

 harmonic mean number of seedlings (m) for each elevation was: 8.39 low, 

 8.26 mid, and 6.26 high. These seedlings had been exposed to blister rust 

 during the fall of their second growing season and, as expected, the means 

 of healthy seedlings were higher (56.67%, 44.86%, and 54.23%) than had 

 previously been reported (Bingham et at. , 1969) for seedlings exposed to 

 the pathogen during their first growing season. 



The estimates of 1/4 additive genetic variance provided by o 2 j, and 

 o 2 q ranged from 3.48 to 18.28. The interaction variance o 2 j>c which esti- 

 mates 1/4 dominance variance and about 1/8 of epistatic variance was high 

 in relation to the estimates of additive genetic variance in each eleva- 

 tional zone. 



The heritabilities were calculated for individual and progeny tests 

 for each elevation in accordance with the procedures given previously. 

 Table 5 presents these heritabilities by elevational zones and method of 

 selection. 



To illustrate the calculations required for estimating heritability 

 the mid elevational zone data are taken from Table 4. 



