Suitability of Pedigree Method for Improving Seed Cotton Yield and Fiber Quality Traits

Background: Cotton is grown around the globe for its ber, which consists of unicellular seed trichome. Converging decent ber quality and good ber yield in cotton varieties is crucial for textile industry of any country. Pedigree method is mostly used for developing cotton varieties in Pakistan. Suitability of pedigree method for developing cotton variety is accessed in an experiment. Genotypic variance, phenotypic variance, heritability, co-heritability, genetic advance, mean comparison and correlation analysis was carried out of nine breeding lines of in fth lial generation . Gene action was thus determined to target the traits for further yield and quality improvement of cotton ber. Results: Substantial genetic variability existed in F 5 lines. Genotypic and phenotypic variances had less differences and phenotypic variances were higher than genotypic variance indicating inuence of environment on the nal expression of the traits. All traits had medium to high heritability. Seed cotton yield per plant came up with high genetic advance and high heritability indicating additive gene action and can be improved by selection. Mean comparison indicated more variation for GOT% compared to other traits. Correlation analysis indicated selecting more sympodial branches for improving seed cotton yield per plant and selecting more monopodial branches for improving ber neness and ber length. However, co-heritability had high value indicating that all the traits are in balance for improvement. Conclusion: Pedigree method is suitable for improving seed cotton yields and ber quality. However, statistical check points are recommended with each proceeding generation to apply suitable selection pressure and monitor the gene action for each trait in each generation.


Background
Cotton (Gossypium hirsutum L.) is a cash crop of Pakistan contributing 0.8% in GDP and 4.5% in value addition of agriculture. Cotton has emerged as life line of Pakistan's economy but its production has been decreased by 17.5% in 2018-19 from last year. The decrease in cotton production was due to decrease in area for cotton production, unfavorable weather conditions and insect /pest attack (Economic Adviser's Wing 2019). Cotton ber is largely produced natural ber. Unlike the other natural bers, cotton ber is unicellular seed trichome (Boopathi et al. 2011). The quality of ber is equally important to ber yield of cotton crop to support the economy of the country. Cotton is mostly bread in Pakistan using pedigree method, where selection is performed in F 2 and continued till F 7 to get homozygous and homogenous population of newly developed cultivar. The plants of F 5 becomes 93.8% homozygous and plant breeders select best performing lines from F 5 (Acquaah, 2009).For genetic improvement, it is important to estimate the extent of genetic variation. The genetic information on broad sense (b. s.) heritability and genetic advance are very important to predict possible improvement in breeding material.
Heritability is the proportion of variation in progeny that is inherited, while genetic advance is difference in mean phenotypic value of parents and offspring (Acquaah, 2009). Broad sense heritability and predicted genetic advance guide the plant breeder to apply an appropriate selection pressure for variety development. High genetic advance and high broad sense heritability estimates offers a most effective response to selection (Larik et al. 1997). In addition, correlation of seed cotton yield per plant with other morphological traits is critical for breeder to tackle of low yield of cotton crop. The correlation analysis also provides a good catalogue to predict the parallel change in one character at the expanse of the proportional change in the other character (Marwat, 2002;Ahmad et al. 2008).
An experiment was designed to predict heritability and genetic advance with different selection intensities in F 5 generation of nine breeding lines. These breeding lines were established to develop cotton variety with high seed cotton yield and reasonable ber quality. Correlation analysis was further carried out to predict the percent change in one trait by selecting for other trait. Mean comparison was also done to check the performance of all nine breeding lines. The study was aimed to determine the suitability of pedigree method for improving seed cotton yield per plant and ber quality simultaneously.  Data were recorded for traits including seed cotton yield per plant (g), GOT (%), ber neness (mic), ber length (mm), ber strength (g/tex), number of bolls plant -1 , number of nodes to rst fruiting branch, plant height (cm), total number of monopodial and number of sympodial branches.
Statistical analysis: The data was analyzed as per Steel et al. (1996) for analysis of variance. Heritability (broad sense) estimate was calculated from ANOVA table as the ratio of genotypic variance to total variance. Genetic advance was also calculated in percent at 10% and 5% selection intensity. Co-heritability was estimated to estimate the inheritance of two traits simultaneously (Acquaah, 2009). Tuckey test was also done to check the variability in the breeding lines. These analyses were done following Soomro et al. (2010). Correlation analysis was further carried out to check the association of traits with each other (Kwon and Torrie, 1964). STATISTIX 8.1 was used to do the statistical analysis.

Assessment of variability:
There exist considerable variance among genotypes for all traits under consideration. These lines had highly signi cant differences for number of sympodial branches plant -1 , seed cotton yield, number of bolls plant -1 and GOT. While, for some traits signi cant differences were observed in germplasm under consideration. These traits include ber length, ber neness, ber strength, number of monopodial branches plant -1 , number of nodes to rst fruiting branch and plant height. All the traits had higher phenotypic coe cient of variance compared to genotypic coe cient of variance ( Table 2). The difference between phenotypic and genotypic coe cient of variance were high for all the traits except for GOT indicating the variation in these lines are due to environmental in uence except for GOT which is due to genetic effect.

Mean Comparison:
Tuckey test separated the genotypes in three groups for number of bolls per plant, ber strength, number of monopodial and sympodial branches, ber neness, plant height and seed cotton yield per plant. The lines were divided into four diverse groups for GOT, while, number of nodes and ber length showed no diversity (Table 3).
Nature of gene action: All the traits showed medium to high broad sense heritability estimates. Fiber length, ber neness, ber strength and GOT had low genetic advance combined with high heritability indicating non-additive gene action of these traits. Thus, delayed selection from these breeding lines or using these lines in another round of hybridization to get transgressive segregants is recommended to improve ber quality traits and GOT. Plant height, number of nodes to rst fruiting branch, monopodial branches plant -1 and number of bolls plant -1 had medium genetic advance and moderate heritability showing non-additive gene action controlling these traits. High heritability indicates the role of genes in nal phenotypic expression but low and medium genetic advance indicates less gain by selection for these traits. Thus, late selection in advance breeding lines for nal variety approval can improve plant height, number of monopodial branches plant -1 and number of nodes to rst fruiting branch. Sympodial branches and seed cotton yield plant -1 had high genetic advance and high heritability showing additive gene action of these traits and pure line selection for these traits would be rewarding. Co-heritability values for all the traits are also high indicating possibility of simultaneous improvement of yield and quality traits. Genetic advance seems more promising at 5% selection intensity compared to 10% selection intensity (Table 2). where RMS is replication mean sum of sequare, TMS is treatment mean sum of sequare, PCV is phenotyic coe cient of varience, GCV is genotypic coe cient of variance, H 2 is broad sense heritability, GA (5%) is expected genetic advance at 5% selection intensity, GA (10%) is expected genetic advance at 10% selection intensity, CO-H 2 is coheritability, * is signi cant difference at p=0.5 and ** is signi cant at p=0.01 Correlation analysis (Table 4)   FF is ber neness, SCY is seed cotton yield per plant, "*" is signi cant at p = 0.05 and "**" is highly signi cant at p = 0.01.

Discussion
Genetic diversity is prerequisite for variety development. Selection after hybridization is done to get genetic purity of superior genetic combinations in progeny lines. Progeny lines become 93.8% pure in F 5 generation and where selection of superior lines is done more rigorously to get advance breeding lines. Genetic diversity in F 5 lines can lead to the multiple variety development lately. The genetic variability seems to be su cient in F 5 . Earlier studies of cotton also indicated signi cant differences in F 2 generation (Eswari et al. 2011), in F 2 and F 3 generations (Soomro et al. 2010) and in varieties and F 1 hybrids (Ahmed et al. 2006;Khalid et al. 2018).
High phenotypic coe cient of variance of all traits than genotypic coe cient of variance indicates environmental in uence on the extent of variation present in F 5 . However less differences in these variances indicates close association of genotypic and phenotypic expressions. Previous studies in cotton also indicated high phenotypic coe cient of variance and less differences between genotypic and phenotypic coe cient of variation (Gnanasekaran et al. 2018;Eswari et al. 2017) Fiber length and number of nodes to rst fruiting branch showed no grouping in mean comparison, while for GOT % , genotypes were divided in four groups. Earlier study in cotton also grouped genotypes of F 1 , F 2 , F 3 and parents into groups on basis of mean comparison (Ahmed et al. 2006;Soomro et al. 2010). Seed cotton yield plant -1 and other traits grouped genotypes in F 5 lines into three groups. The difference in the variation of traits in earlier studies is due to different generations under consideration.
Fiber length, ber neness, ber strength and GOT had high heritability and low genetic advance indicating non additive gene action of these traits. Plant height, number of nodes, monopodial branches and number of bolls per plant showed nonadditive and over dominance types of gene action due to medium genetic advance and high heritability. High heritability indicates the role of genes in nal phenotypic expression but low and medium genetic advance indicates less gain by selection for these traits. Only sympodial branches and seed cotton yield plant -1 showed additive gene action and continued pure line selection can result in improved seed cotton yield and more sympodial branches. Earlier studies of cotton also indicated additive gene action of seed cotton yield plant -1 and estimated the gene action of other morphological and ber quality traits (Ahmed et al. 2006;Gnanasekaran et al. 2018;Eswari et al. 2011;Soomro et al. 2010;Kumar et al. 2019). Thus for improving staple length, ber neness, ber strength and GOT another round of hybridization is needed to shu e the genetic makeup of these traits. Delayed and rigorous selection till F 7 and F 8 can also give desired results.
Seed cotton yield plant -1 came up with positive correlation with number of sympodial branches plant -1 indicating that more sympodial branches in variety would yield more seed cotton yield. While, GOT showed positive correlation with monopodial branches and number of initial nodes plant -1 . This indicates that more ber yield can be gain by giving more importance to monopodial branches and the ber in this case would be ner and long. Selection should proceed with less number of initial nodes to improve ber quality. Earlier studies also indicated positive association of seed cotton yield with number of sympodial branches per plant (Rahman et al. 2013;Khalid et al. 2018). While positive connotation of GOT with ber neness are contradictory with earlier studies (Khalid et al. 2018) may be due to difference in genetic material.

Conclusion
Genetic variability in F 5 is suitable for variety development. Seed cotton yield has additive gene action while all other traits came up with non-additive type of gene action. Correlation studies indicated selecting more of monopodial branches to improve ber quality while more number of sympodial branches plant -1 to improve seed cotton yield plant -1 . Increased selection pressure of up to 5% is more promising for more genetic gain. Thus, pedigree method practiced in cotton breeding is likely to improve ber quality and seed cotton yield. However, continued statistical analysis are recommended for determining the selection pressure and gene action for improving targeted traits.