EVALUATION OF BRASSICA JUNCEA GERMPLASM THROUGH HERITABILITY ESTIMATION AND CORRELATION ANALYSIS

The experiment was conducted to estimate genetic parameters of heritability and correlation for yield traits of indigenous Brassica juncea genotypes. RCB design with three replications was used to sow 20 Brassica genotypes. Highly significant genetic variations were obtained among studied parameters. Days to half flowering revealed positive associations with primary branches per plant (r= 0.337**). Primary branches per plant was positively correlated with pod length (r= 0.407**), seed per pod (r= 0.418**) seed yield per plant (r= 0.479**). Similarly, pod length showed positive correlation with seed per pod (r= 0.324*) and seed yield per plant (r= 0.331**), while seed per pod was positively correlated with seed yield per plant (r= 0.878**). Highest broad-sense heritability was recorded for seed yield per plant (0.97), while lowest heritability was estimated for primary branches per plant (0.46).


INTRODUCTION
Edible oil is on the top in the imported bill of Pakistan mainly due to low local production.Brassica oilseed is the third edible oil source after Palm and Soybean world widely [1].This high dependency on imported oil can be reasonably decreased by the development of oilseed genotypes having better adaptability into the cropping system of Pakistan [2].Seed yield can be improved by improving parameters having high genetic divergence, heritability and selection response [3].To run an effective breeding program, high heritability along with high selection response is desirable because a good understanding of selection response is helpful in the selection process [4].It must be noted that high heritability with high selection response is an indication of additive genetic operation, while parameters with high heritability but low selection response is due to operations other than additive genes [5].
Evaluation of the variability among the genetic resources of the material is the top priority for the improvement because it generates necessary information which can speed up the breeding program [6,7].The necessary tools for the genetic improvement of crop plants in a breeding program are divergence among the material and parameters having high heritability with high selection response [8].Selection response is a very useful tool which shows the extent of increment of a parameter under study and can be easily calculated with the help of heritability estimates.It helps the breeder to proceed in a right way [9].Correlation analysis is very helpful in understanding relationship between different sets of parameters [10].

MATERIALS AND METHODS
The planned study was conducted at the University of Agriculture, Peshawar during 2011-12 to estimate genetic parameters of heritability and correlation analysis among 20 Brassica genotypes.The research material was planted in RCB design with three replicates.Data were recorded on six agro-morphological and maturity parameters.Fifteen plants were selected for data recording in a single replication.

Statistical analysis
LSD test was used for means separation.Steel and Torrie [11] approach was used for the computation of analysis of variance.Broad-sense heritability estimates were recorded as a ratio of genotypic variance to the total phenotypic variance [12].Selection response was calculated using Fehr [13] approach having selection intensity of 20%.differences among 20 genotypes (table 1).Our results are strongly supported by the previous findings [10,14,15].Days to half flowering ranged from 58.00 to 89.00 d.Maximum days to half flowering (89.00 d) were observed in genotype J.600, whereas minimum days (58.00 d) were found in genotype J.617 (table 2).Genetic variance (40.02) was greater than environmental variance (37.53) for days to half flowering.Moderate broad-sense heritability (0.52) and high selection response (6.54) were recorded for the said trait (table 3).Our current findings are in contradiction with Mahto and Haider [16], who reported high broad-sense heritability which discourages our results but their findings agree with our results of obtaining moderate to high selection response for days to flowering.Highly significant and positive correlation were recorded between days to half flowering and primary branches per plant (r = 0.337**) (table 4).Positive but non-significant associations between days to half flowering and primary branches per plant were noted by Nasim et al. [17] and strengthen our findings.

Primary branches plant -1
For primary branches plant -1 , highly significant (P≤0.01)differences were computed from analysis of variance (table 1).Previous studies reported significant variation for the said trait and encourage our findings [18,19,20].Primary branches plant -1 ranged from 8.3 to 15.7.Maximum primary branch plant -1 (15.7) was observed in genotype J.600, while genotype J.617 exhibited minimum number of primary branches (8.3) (table 2).Genetic variance (2.97) was lower in magnitude than environmental variance (3.53).Moderate broad-sense heritability (0.46) and low selection response (1.68) were recorded for the said trait (table 3).Our results are contradicted by the findings of Iqbal et al. [21] who observed high broad-sense heritability but are in agreement by reporting low selection response value.Low heritability estimates were obtained by Zhang and Zhou [1], for the said parameter.Primary branches plant - 1 showed significant and positive associations with pod length (r = 0.407**), seed pod -1 (r = 0.418**) and seed yield (r = 0.479**) (table 4).Positive and significant associations among primary branches plant -1 with pod length, seed pod -1 and seed yield were reported by Ali et al. [2] which confirmed our finding.

Main raceme length
Analysis of variance indicated significant (P≤0.01)differences for the said parameter (table 1).Zada et al. [7] observed significant differences in Ethiopian mustard.Our findings are also in similarity with the previous results [22,14].Main raceme length ranged from 40.1 to 77.8 cm.Maximum main raceme length (77.8 cm) was observed in genotype J.616, whereas minimum raceme length (40.1 cm) was found in genotype J.600 (table 2).Genetic variance (73.26) was greater in amount than environmental variance (53.86) for main raceme length.Moderate broad-sense heritability (0.58) and high selection response (9.36) were recorded for studied trait (table 3).Muhammad et al. [23] recorded moderate broad-sense heritability and extend our findings, while Nasim et al. [17] manifested low heritability estimates for the said parameter in Brassica napus.Main raceme length was negatively correlated with all the studied parameters.Our current results are discouraged by Choudhary et al. [24] who reported positive associations for seed yield plant -1 with the said parameter (table 4).
Seed pod -1   There was significant (P≤0.01)variations for seed pod -1 (table 1).Samad and Khaleque, [25] reported highly significant variation in B. campestris for the said parameter.However Bilgili et al. [26] observed nonsignificant differences for seed pod -1 and discourage our findings.Seed pod -1 ranged from 8 to 19.Maximum seed pod -1 (19) was observed in genotype J.606, whereas minimum seed value (8) was found in genotype J.603.Genetic variance (10.28) was higher in magnitude than environmental variance (1.36) (table 2).High broad-sense heritability (0.88) and moderate selection response (4.34) were recorded for seed pod -1 in current study, which suggested that the inheritance and improvement is under genetic control (table 3).High heritability with moderate selection response was resulted by Chaghakaboodi et al. [27] and confirm our result for the said trait.

Pod length
Significant (P≤0.01)differences were exhibited for Pod length (table 1).Previous reports also showed significant values for pod length and encourage our results [10,20,22].Pod length ranged from 1.9 to 4.3 cm.Genotype J.612 showed maximum pod length (4.3 cm), while minimum length (1.9 cm) was recorded in genotype J.616 (table 2).Genetic variance (0.34) was good enough in magnitude than environmental variance (0.11) for the said trait.Broad-sense heritability (0.76) was higher in amount however, low selection response (0.73) were recorded for pod length (table 3).Similar findings were obtained by previous studies [29,30,17].
Pod length showed significant and positive associations with primary branches per plant (r = 0.407**), seed per pod (r = 0.324*) and seed yield (r = 0.331**) (table 4).Positive and significant associations between pod length with primary branches plant -1 were reported by Ali et al. [2] which confirmed our finding.