Estimation of Heritability for Seed Cotton Yield in Cotton Based on Regression Approach

Main focus of this study was to calculate the heritability for seed cotton yield based on regression approach by utilising the ten F5 and ten F6 lines of two populations derived from a heterotic box. Ten F6 lines of RSG group derived from (DSMR-10 X DSG-35) cross were regressed over same F5 lines for seed cotton yield. Similar procedure was done for lines of RGR group (DRGR-32-100 X DRGR-24-178) cross. By utilising the regression value heritability of seed cotton yield was calculated as suggested by Smith and Kinman, 1965. The narrow sense heritability for RSG group and RGR group lines was 24.90 and 21.21 per cent. Broad sense heritability was also calculated based on RBD analysis separately for RSG and RGR group lines.

Cotton is an important commercial crop grown for its fibre.Seed cotton yield determines the commercial value of cotton and the heritability value plays a key role in determining selection strategy for improvement.Apart from establishing half sib and full sib relations for determining components of variance and determining heritability of the trait, it is also possible to use information on consecutive generations to work out regression value for a trait and estimate heritability for a trait (Smith and Kinman, 1965 and Salimath and Patil, 1990).There are very limited studies focusing on determining heritability of yield per se based on regression approach.In this study F 5 and F 6 populations of a heterotic box representing opposite heterotic groups subjected to reciprocal selection were utilized.

MATERIALS AND METHODS
For determining heritability of yield per se F 5 and F 6 lines developed through reciprocal selection in cotton of a heterotic box involving elite lines of robust/stay green group (RSG) and RGR (high relative growth rate) were used.This heterotic box comprises of DSMR-10 line (of stay green group), DSG-3-5 line (of robust group) and two DRGR-32-100 and DRGR-24-178 lines (of RGR group).These lines were crossed (DSMR-10 X DSG-3-5) (DRGR-32-100 X DRGR-24-178) two give two F 1 .Resulting F 1 s were advanced to the F 4 and F 5 generation where recombinational variability for combining ability was evaluated.Here, regression of ten parental lines of F 6 generation over F 5 generation was carried out to determine the heritability of yield per se.

Regression Approach
The seed cotton yield values of F 5 and F 6 lines were utilized for determining regression values TANTUWAY et al.: STUDY OF HERITABILITY FOR SEED COTTON YIELD i.e., (b F6F5 ).Heritability (h 2 NS ) of yield per se was calculated based on regression approach given by Smith and Kinman, (1965).h 2 = (b / 2r XY ) where, h 2 = Narrow sense heritability b = Regression coefficient r XY = Coefficient of parentage [which works out to be (31/32) for this situation] The mean seed cotton yield of lines was used for regression of F 6 lines over F 5 lines, finally giving the regression value (b F6:F5 ).The regression value (b F6:F5 ) was divided with the coefficient of parentage (31/32) depending upon the generations of the lines used in the analysis.
The set of F 5 and F 6 lines were evaluated in replicated block design and MSS values in the ANOVA for genotypes and error component were utilised in determining broad sense heritability (h 2 = V g /V p ).

RESULTS AND DISCUSSION
Regression of seed cotton yield of F 6 lines over F 5 lines were carried out for RSG and RGR group, ANOVA of regression coefficient was presented in table 1 (a) and 1(b) respectively.

Table 2 .
Regression Regression of per se yield of F 6 lines over F 5 lines (b) = 0.48 Heritability (h 2 NS ) of per se yield = 24.90%

Table 4 .
Comparison of broad sense heritability obtained by RBD analysis and narrow sense heritability obtained by Regression approach