Abstract
The present investigation involved four experiments which were carried out during 2014-2015,
2015-2016, 2016-2017 and 2017-2018 at Regional Agricultural Research Station (RARS), Bangladesh
Agricultural Research Institute (BARI), Jamalpur with the aim to investigate genetic diversity
among rapeseed germplasm, to generate information on genetic variability, selection of
prospective genotypes, estimate heterosis as well as heritability, degree of interrelationship and
combining ability (general, specific and reciprocal), nature and magnitude of gene action for 18
qualitative and 10 quantitative characters. Thirty diverse germplasm were evaluated for the
genetic diversity analysis using multivariate approach. High phenotypic coefficient of variation
(PCV) and genotypic coefficient of variation (GCV) were found for seed yield/plant, number of
secondary branches/plant and number of siliquae/plant; while, they were found as low for days
to first flowering, days of flower completion, days to maturity and plant height. Principal
component analysis based on correlation matrix of 16 quantitative characters showed that the first
seven components were accounted about 88.28% of total variation among the genotypes. Cluster
analysis grouped the 30 germplasm into five clusters. The result on intra and inter-cluster
distances exhibited that the germplasm were highly diverse from each other. Nine potential and
diversified genotypes; Tori-7, TS-72, BC-0828-1, BC-0830-1, BC-0835-1, ВС-11046, BARI Sarisha-9,
BARI Sarisha-12 and BARI Sarisha-14 were identified based on short duration coupled with yield
for diallel crossing. Selected nine genotypes were crossed in all possible direction following a full
diallel fashion. The performance of the parental lines and 72 F₁ hybrids were assessed upon 18
characters. ANOVA confirmed the variances among parents, Fi hybrids for general combining
ability (GCA), specific combining ability (SCA) and reciprocal combining ability (RCA) across all
the studied characters, suggesting the presence of additive and non-additive gene actions for
regulating those characters. Deviation from dominance (VD) was much higher than the
corresponding additive genetic variance (VA) for almost all the characters except days to first
flowering and siliqua length. According to GCA, SCA, RCA and heterosis analyses, the parental
lines BC-0828-1, ВC-0830-1, BC-0835-1 and BARI Sarisha-14 and the F₁s BC-0828-1 × BC-11046,
Tori-7 x BARI-12, Tori-7 × BC-0830-1, Tori-7 × BC-0828-1, BARI-9 × TS-72, BARI-9 × BC-0835-1,
BC-0830-1 × BARI-14, BC-0835-1 × BARI-14, BARI-14 × BARI-12 and BARI-14 × BC-11046 were
identified as short duration coupled with high yield and yield components. Twelve crosses were
used to assess gene actions through scaling and joint scaling tests, which indicated the presence of
non-allelic gene interaction for most of the characters. The result showed that magnitude of
dominance gene effects was higher than additive gene effect. Three epistatic components additive
x additive (i) and dominance x dominance (1) were found as significant for almost all the
characters in all the crosses, but additive x dominance (j) was found as significant for some of the
characters in most of the crosses. Duplicate epistasis played a vital role in controlling majority of
characters for all the crosses. The interspecific hybrids of B. napus and B. rapa, viz. Nap-0717-2 x
BC-0830-1, Nap-0837 × Tori-7, BARI-8 × BC-0828-1, Nap-0876 × BC-0835-1, and Nap-0876 × BC0828-1 were identified as desirable toward selection for early-flowering and maturity coupled with
high yield potential. In the summary, it could be recommended that the crosses Tori-7 × BC-0828-
1, BARI-9 × TS-72, BC-0828-1 × BС-11046, ВC-0830-1 × BARI-14 and BARI-14 × BARI-12 could be
selected for consecutive selfing and selection for short duration desired genotypes for better
adopting in rice-rice cropping pattern in Bangladesh.