General and specific combining ability (GCA and SCA, respectively) wer
e examined in 36, 6-parent disconnected partial diallels across 4 diff
erent experimental series in coastal Douglas-fir (Pseudotsuga menziesi
i var. menziesii DOUGL.) to examine the ratios of the 2 genetic varian
ces, the distribution of GCA and SCA effects, and estimates of genetic
gain from GCA and SCA for 3 growth traits. Height at age-7 and height
and volume at age-12 were measured on approximately 150 trees per ful
l-sib family in each diallel, across 11 different test sites within ea
ch series. The average percentage ratio of SCA variance to GCA varianc
e was 36%, across all series and the 3 growth traits, with a range of
19% to 65%. GCA and SCA variances did not appreciably change for heigh
t growth from age 7 to age 12. Diallel set effects were generally negl
igible. From theoretical considerations assumed for the diallel model,
clear separations of additive and dominance effects (vis-a-vis the as
sumptions of selecting on GCA and SCA variances) are likely not possib
le: the effects are subject to degrees of dominance, epistasis and Lin
kage in the population. However, these genetic details did not manifes
t themselves in any noticeable pattern or correlation among GCA and SC
A effects. While these results confirm current strategies in Douglas-f
ir breeding to select primarily on GCA, there are opportunities to uti
lize SCA variance in the production population. Controlled matings for
elite production populations are now common in coastal Douglas-fir im
provement programs, and controlled crossing with specified parents cou
ld i) elevate gains by as much as 3.0% in 12-year volume (assuming com
petition effects have not biased volume estimates), and ii) access add
itional sets of parents that would otherwise be disregarded. Gains for
height growth from utilizing SCA are lower, which reflect either, i)
lower SCA variances associated with height, or ii) competition has bia
sed upward volume SCA variance by age 12 (relative to height).