Introduction
Scientists have described 50 species in genus Gossypium that have wide range to live on arid to semi-arid regions of world’s tropic and subtropics region. The four species have extra-ordinary morphological variation, mainly herbaceous or small tree type with diverse vegetative and reproductive characteristics that make it suitable to live in many environments. Each of four species of cotton has its own diversification, utilization and domestication. Cotton is being cultivated in whole world, and according to estimates of United States Department of Agriculture, its production was 117.1 million bales during the year 2013-2014 (USDA, 2014).

Gossypium hirsutum is known as upland cotton and is grown on 99% of the total cotton area in Pakistan and contributes 78% domestic oil production. Cotton is playing an important role in the agriculture sector of the country by contributing 1.5% to GDP and 7.0% in value addition. Cotton is cultivated on 2.879 million hectares with the production of 13.02 million bales (Economic Survey of Pakistan, 2013-14).

As Pakistan is an agriculture economy based country, cotton has a major role in it. Cotton seed oil is fulfilling 70% requirement of our vegetable oil. Cotton seed cake is major dairy food having a lot of protein quantity Cotton is also a source of foreign exchange earnings in raw form and also in finished product form. Cotton fibre is raw material for textile industry, ginning factories and oil expellers. Cotton fibre is basically cellulose that originates from the epidermis of seed.

As Pakistan has a huge cotton industry, we need high yield and good fibre quality to fuel up the economy. That’s why we need to know yield association with other traits. Fibre fineness and staple length are important for textile industry. Fibre strength is quality character and important for spinners. High tensile strength is needed for good spinning.

The knowledge of association of traits influencing yield in any crop has utmost importance for carrying out effective breeding program. A lot of improvement in production of high yielding variety has been made. As yield is the prime importance of breeder whereas textile industrialists give preference to fibre related traits i.e., fibre strength, fibre length and fibre fineness. Iqbal et al. (2006) emphasized that the seed cotton yield is positively correlated with number of boll weight, number of monopodial branches and number of bolls, whereas, negatively correlated with GOT. Rasheed et al. (2009) concluded that yield is positively correlated with number of bolls, boll weight and GOT, whereas, negatively correlated with fibre strength and plant height. There are many factors that effect on seed cotton yield such as number of bolls, plant height, number of fruiting branches, ginning out turn percentage, boll weight and seed index (Salahuddin et al., 2010). It also reflects the response of one trait with its counterpart, although, correlation of agronomic characters will definitely help to enhance yield in cotton and characterization in order to screen the promising lines. Aim of this study is to find out the extant of association, heritability, direct and indirect effects of polygenic traits on genotypes.

Results and Discussion
The means of ten polygenic traits, correlation, path coefficients and heritability is illustrated in Table 1~ Table 4, respectively. The data is subjected to the analysis of variance (ANOVA) prescribed by Steel and Torrie 1987. All the genotypes/crosses have significant differences among them. The mean values are shown in Table I along with their units in which they measured.

Table 1 Mean of the parents/crosses for the following traits

Table 2 Correlation coefficient of yield and fibre traits of upland cotton

Table 3 Direct and Indirect effects of yield and fibre traits of upland cotton

Table 4 Broadsense heritability (h2¬BS), Genetic Advance (G.A) along with variances

Correlation Analysis
Seed cotton yield had positive correlation with 100 seed weight, plant height, number of sympodial branches, fibre length and fibre strength. Whereas, seed cotton yield showed strong negative correlation with number of monopodial branches, ginning-out-turn and fibre fineness. Iqbal et al. (2006) proved the same results. The results of Karademir et al. (2010) were contrary to the present research. Joshi et al. (2006) found negative correlation between seed cotton yield and GOT whereas positive correlation with plant height.

Number of monopodial branches illustrated strongly negative association with weight of 100 seeds, plant height, fibre length and seed cotton yield. Whereas, monopodial branches were moderate negative associated with fibre strength and number of sympodial branches. Monopodial branches had positive correlation with GOT and fibre fineness. Ginning- out-turn and monopodial branches had significant negative association with yield per plant. Study conducted by Tariq et al. (1992) reported the same results. Ahuja et al. (2006) reported negative correlation of number of monopodial branches to plant height, fibre strength and fibre length. Kazerani et al. (2012) and Hatamee and Latifi (2004) reported negative association of plant height and number of monopodial branches.

Ginning-out-turn percentage had strong negative association with seed cotton yield, weigh of 100 seeds, plant height, fibre length and fibre strength. Whereas, GOT had positive association with monopodial branches, sympodial branches and fibre fineness. So in case of selection process we have to be very careful to focus toward both fibre yield and fibre fineness.

Weight of 100 seeds had positive correlation with seed cotton yield and plant height whereas strongly negative association with monopodial branches, GOT, number of bolls and fibre fineness. Weight of 100 seed showed moderately negative association with sympodial branches and non-significance association with fibre strength and fibre length. Karademir et al. (2009) found negative association of weight of 100 seeds and GOT, and further enunciated that 100 seed weight had no any effect on yield and fibre traits.

Plant height is the one of the most promising traits which contributes in yield. Plant height had strong positive correlation with seed cotton yield, weight of 100 seeds, sympodial branches and fibre length, whereas, negative correlation with monopodial branches, GOT, fiber fineness and fibre strength. From these results, we can conclude that if we want good fibre quality than we must go for short heighted varieties. Taohua & Haipeng (2006) and Khan et al. (2009) reported that plant height positively correlated with yield and number of bolls. Murthy (1999) reported the same results and noted that plant height contributed 70% in seed cotton yield. Camlho et al. (1994) reported negative association of plant height and fibre strength. Number of bolls proved to have strong positive association with seed cotton yield, fibre strength and sympodial branches. Karademir et al. (2009) authenticated that number of bolls had positive association with sympodial branches and weight of seed cotton.Bibi et al. (2011) reported negative association of number of bolls and sympodail branches. 

Number of sympodial branches showed negative association with monopodial branches and 100 seed weight but had positive correlation with seed cotton yield, GOT, plant height, number of  bolls, fibre length and fibre strength. It clearly depicts that number of sympodial branches is yield contribution trait not only for seed cotton yield but also for fibre length and strength. The sufficient increase in yield had been proved by the research findings of Badr (2003), Iqbal et al. (2006), Wang et al. (2004) and Salahuddin et al. (2010). Hussain et al. (2010) and Natera et al. (2012) found positive association of sympodial branches and number of bolls.

Fibre fineness is the promising trait which is much bothered by textile industry. Fibre fineness had positive correlation with monopodial branches and 100 seed weight. But showed strong negative correlation with seed cotton yield, 100 seed weight, plant height and fibre length. Whereas, fibre fineness and fibre strength had negative correlation between them. Increasing trend of fibre fineness will lead toward decreasing trend of seed cotton yield.

Fibre length had strong positive correlation with seed cotton yield, plant height and fibre strength whereas negative correlation with monopodial branches, GOT and fibre fineness. It is clear that if we are planned to enhance fibre length, strength and fineness than we must have to be very careful in breeding program. Fibre strength showed positive association with seed cotton yield, fibre strength and number of bolls. Fibre strength showed negative association with monopodial branches, plant height, GOT and fibre fineness. Ahuja et al. (2006) found negative association between seed cotton yield and fibre strength.

Path Coefficient Analysis
Purpose of path coefficient analysis is to find out the influence of dependent variables on seed cotton yield. These effects are in shape of direct or indirect effects. The results showed that seed cotton yield had experience positive effects form number of monopodial branches (0.060), GOT (0.145), plant height (1.175), number of sympodial branches (0.014) and fibre strength (0.0.534), directly. But only plant height, fibre strength and GOT showed some good effects. (Baloch et al., 2009) and (Kerademir et al., 2009) authenticated same results, whereas, Rasheed et al., (2009) were in favour of GOT and fibre strength. Natera et al., (2012) also proved the significant positive effect of fibre strength on seed cotton yield. Magadum et al. (2010), Ahuja and Dhyal (2007), Salahuddin et al. (2010) and Afiah and Ghoneim (2000) authenticate strong positive effect of sympodial branches on yield, but here in this research its values are just confine to small percentage, this may be due to small genotypic correlation between seed cotton yield and sympodial branches.

Weight of 100 seeds, fibre fineness and fibre length had negative direct effect on seed cotton yield i.e., -0.692, -0.936 and -0.232, respectively. Same results were authenticated by Natera et al. (2012) and Rasheed et al. (2009). Whereas, research study of Magadum et al. (2010) and Rasheed et al. (2009) was in favour of direct positive effect of fibre fineness on seed cotton yield.

Heritability
Heritability was highest for fibre strength (99.3) followed by number of sympodial branches (99.757), plant height (97.616), fibre fineness (96.738) and GOT (94.86). Whereas, lowest heritability value was of number of monopodial branches (70.354). Same results of h2bs was in the finding of Hussain et al. (2010). Rasheed et al. (2009) proved high h2bs value for fibre strength, fibre fineness and GOT in his studies.

Conclusion
Overall there was good performance of hybrids and high heritability indicates that selection would be easy. Yield and fibre traits cannot be enhanced simultaneously. We can maximize yield by focusing on plant height, fibre strength and GOT. These traits are very responsive toward selection. 

Materials and Methods
The study was conducted in the research area of the Department of Plant Breeding and Genetics, University of Agriculture, Faisalabad, Pakistan. Five parent cultivars were grown in glasshouse under controlled condition in October, 2012. Temperature in glasshouse was maintained approximately at 35℃. Parent cultivars used in the study are listed in the following:
B557
FH 153
CRS 2007
KZ 191
S 12

At the time of flowering direct crosses were made. Emasculation was done at evening and pollination was performed next morning at 9 to 10 am. In June 2013, F0 seed of 10 hybrids and 5 parents were planted in the field in randomized complete block design with three replications. In each replication there was single row for each hybrid and parent. Row-to-row and plant-to-plant distance was maintained 75cm and 30cm, respectively. Proper agronomic practices were adopted from sowing till harvesting of crop. Five plants were selected from each parent for recording of data. Data was recorded for following polygenic traits: plant height, number of sympodial branches, number of monopodial branches, number of bolls, fibre length, fibre strength, fibre fineness, ginning-out-turn, weight of 100 seeds and seed cotton yield.

After harvesting, ginning was performed with help of ginner, in department of Plant Breeding and Genetics. Weight of randomly selected 100 seeds was measured on electrical balance. Ginning out trun (GOT) was calculated by formula:


Fibre related characters - fibre length, fibre strength and fibre fineness- were calculed by help of Zellwegar Uster HVI-900 in the Department of Fibre Technolog, University of Agriculture, Faisalabad. Data of fibre length, fibre strength and fibre fineness was recorded in millimeter, gram per tax and microgram per inch.

Analysis of Data:
Data was subjected to analysis of variance by procedure given by Steel and Torrie (1987) for determining significant differences among the genotypes under study. Correlation analysis-phenotypic and genotypic correlation- and Path coefficient analysis was done by procedure outlined by Dewey and Lu (1959) and path coefficient analysis was done by using genotypic correlation. Broad sense heritability was determined according to formula of Mahmud and Kramer (1951) as follows:


Genetic advance calculated according to prescribed formula of Allard (1960).
G.A = σ ρ× h2× i

References
Ahuja S.L., L.S. Dhayal, and R. Prakash, 2006, A Correlation and path coefficient analysis of components in G. hirsutum L. hybrids by usual and fibre quality grouping. Turk. J. agric. 30: 317-324

Ahuja S.L. and Dhayal L.S., 2007, Comparative characteristics and gene action in three petal-spotted mutants of Gossypium hirsutum L.J. Genet. 86: 81-84
http://dx.doi.org/10.1007/s12041-007-0012-0

Allard R.W, 1960, Principles of Plant Breeding. pp: 96. John Willey and Sons, Inc. New York.

Badr S.S.M, 2003, Evaluation of some Egyptian cotton varieties by the yield and seven methods of Earliness of crop maturity measurements. Egyp. J. Agric. Res., 81: 671-688

Baloch M.A., A.R. Lakho, H. Bhutto, and M.Y. Solangi, 2001, Path coefficient analysis for assessing direct and indirect effects of yield components on seed cotton yield in Gossypium hirsutum L. Online J. of Biol. Sci. 1(5):354-355
http://dx.doi.org/10.3923/jbs.2001.354.355

Bibi M., N.U. Khan, F. Mohammad, R. Gul, A. A. Khakwani. and O.U. Sayal, 2011, Genetic divergence and association among polygenic characters in Gossypium hirsutum L. Pak. J. Bot., 43(6): 2751-2758

Camlho H., R. Steave and S.T. Roberts, 1994, Population analysis, correlation and regression analysis of some agronomic and technological characteristics of upland cotton (Gossypium hirsutum L.) using diallel analysis in two different zones. Online J. Biol. Sci. 1(3): 120-123

Dewey D.R. and K.H. Lu., 1959, A correlation and path coefficient analysis of components of crested wheat grass and seed production. Agron. J., 51: 515-7
http://dx.doi.org/10.2134/agronj1959.00021962005100090002x

Economic Survey of Pakistan, 2012-13, Economic advisors wing, Finance Division, Government of Pakistan, Islamabad.

Hatamee H., and N.Latifi , 2004, The  effect of planting arrangement on quantitative and qualitative  properties of cotton under intercropping with berseem clover. Gorgan, J. Agric. Sci. Nat. Resour., 11: 95-106.

Hussain S., N. N. Nawab, M. A. Ali, A. Hussain, M. A. Nawaz and T. A. Malik, 2010, Evaluation of performance, genetic divergence and character association of some polygenic traits of upland cotton. J. Agric. Social Sci. 4-79-82

Iqbal M., K. Hayat, R.S.A. Khan, A. Sadiq and Noor-ul-Islam, 2006, Correlation and path coefficient analysis for earliness and yield traits in cotton (G. hirsutum L.). Asian J. Pl. Sci. 5: 341-344
http://dx.doi.org/10.3923/ajps.2006.341.344

Joshi H.J., P.K. Chvoatia, and D.R. Mehta, 2006, Genetic variability, character association and component analysis in upland cotton. Ind. J. Agric. Res. 40: 302-306

Karademir E., C. Karademir, R. Ekinci, and O. Gencer, 2010, Relationship between yield, fibre length and other fibre related traits in advanced cotton strains. Not. Bot, Hort. Agrobot. Cluj. 38: 111-116

Karademir C., E. Karademir, R. Ekinci and O. Gencer, 2009, Correlation and path coefficient analysis between leaf chlorophyll content, yield and yield components in cotton (Gossypium hirsutum L.) under drought stress condition. Not. Bot. Hort. Agrobot. Cluj 37: 241-244

Kazerani B., 2012, Determination of the best cotton cultivars and selection criteria to improve yield in Gorgan climatic region. African J. Agric. Res. Vol. 7: 2004-2011
http://dx.doi.org/10.5897/ajar11.1602

Khan N.U., G. Hassan, K.B. Marwat, Farhatullah, S. Batool, K. Makhdoom, I. Khan, I.A. Khan, and W. Ahmad, 2009, Genetic variability and heritability in upland cotton. Pak. J. Bot. 41(4): 1695-1705

Magadum S., U. Banerjee, R. Ravikesavan, K. Thiyagu, N.M Boopthi, and S. Rajarathinam, 2012, Association analysis of yield and fibre quality characters in interspecific population of cotton (Gossypium Spp.). J. Crop Sci. Biotech. 2012. 15(3): 239-243
http://dx.doi.org/10.1007/s12892-012-0027-9

Mahmud I. and H. H. Kramer, 1951, Segregation for yield, height and maturity following a soybean cross. Agron. J. 43: 605-609
http://dx.doi.org/10.2134/agronj1951.00021962004300120005x

Manzoor A. and F. M. Azhar, 2000, Genetic correlation and path coefficient analysis of oil and protein contents and other quantitative characters in F2 generation of G. hirsutum L. Pak J. of Biol. Sci. 3(6):1049-1051
http://dx.doi.org/10.3923/pjbs.2000.1049.1051

Murthy  J.S.V.S, 1999, Character association and component analysis in upland cotton. Mad. Agric. J. 86: 39-42
Natera M.J.R., A. Rondon, J. Hernandez, and J.F.M. Pinto, 2012, Genetic studies in upland cotton. III. Genetic parameters, correlation and path analysis. SABRO J. Breed. Genet. 44: 112-128

Rasheed A., W. Malik, A. A. Khan, N. Murtaza, A. Qayyum, and E. Noor, 2009, Genetic evaluation of fibre yield and yield components in fifteen cotton (Gossypium hirsutum) genotypes. Int. J. Agric. Biol. 5: 581-585

Salahuddin S., S. Abro, M.M. Kandhro, L. Salahuddin, and S. Laghari, 2010, Correlation and path analysis of yield component of upland cotton (Gossypium hirsutum L.). World Appl. Sci. J. 8: 71-75

Steel R.G.D and J.H. Torrie, 1997, Principal and procedure of statistics: A biometrical approach. 3rd edition, McGraw Hill Inc., New York.

Taohua Z. and Z. Haipeng, 2006, Comparative study on yield and main agri-characters of five hybrids colored cotton varieties. J. Anhui Agric. Univ., 33: 533-536

Tariq M., M. A. Khan, and G. Idrees, 1992, Correlation and path coefficient analysis in upland cotton. Sarhad J. Agric. 8 (3): 335-341

USDA, 2014, Monthly economic letter. Cotton Incorporated. Available at: www.cottoninc.com/corporate/.../World-Cotton-Production-Exp.pdf

Wang C., A. Isoda, and P. Wang, 2004, Growth and yield performance of some cotton cultivars in Xinjiang, China, an arid area with short growing period. J. Agron. & Crop Sci., 190: 177-183
http://dx.doi.org/10.1111/j.1439-037X.2004.00090.x