Abstract:
Eight genotypes of rapeseed (Brassica napus L.), including six cultivars and two breeding
lines, were used to determine the genetic parameters for number of pods per main
axis (NPM), number of pods per plant (NPP), length of pod (LOP), number of seeds per
pods (NSP), 1000-seed weight , seed yield, and total glucosinolate. Analysis of variance
based on Hayman’s method revealed significant general (a) and specific (b) combining
ability mean squares for all traits except for 1000-seed weight, which indicated the importance
of additive and non-additive genetic effects. For 1000-seed weight, only the general
combining ability mean square was statistically significant. Significant ratios of a to b
mean squares and high narrow-sense heritability estimates were observed for 1000-seed
weight and total glucosinolate, which indicated the importance of additive genetic effects
for these traits. Therefore, the efficiency of selection for improving these traits will be
high. The significant b1 (mean deviation of F1’s from their mid-parental values) mean
squares for all of the studied traits except LOP and 1000-seed weight, exhibited directional
dominance and subsequently significant average heterosis. Significant maternal (c)
mean squares were observed for all of traits, except LOP. Among yield components, NPM
and NPP had a significant correlation with seed yield and can therefore be used as good
criteria for improving seed yield. The correlation between seed yield and total glucosinolate
was not significant, and so reducing this antiquality trait without any considerable
changes in seed yield is possible. In general, the parents PF7045/91 and BL1 were good
combiners and the crosses BL1×PF7045/91, BL1×BL2 and PF7045/91×Shiralee were good
combinations for improving seed yield and total glucosinolate, simultaneously.
