Cotton (Gossypium hirsutum L.) is a major cash crop, often cross-pollinated, long day plant belonging to the family Malvacae. It is a sixth largest source of vegetable oil in the world and can boost up the economy of any country by producing edible oil for human consumption, feed (seed cake) for animals and local consumption in textile industries. Worldwide, most common cultivated species of cotton is Gossypium hirsutum L., a tetraploid, also called as upland cotton and provides 90% fiber production while Gossypium barbadense called as Egyptian cotton produces only 3% fiber. These both species are also known as new world cotton  .
Heterosis (over better parent) is a best source for improvement of F1 and F2 hybrids for production. It has ability which can lead breeders for commercial utilization of valuable hybrid combinations in breeding program. It can also play a vital role for selection of potential parents with desired vigor, maternal effects and genetic variance. Hybrid cotton has successfully attracted the concentration of cotton breeders for commercial utilization of hybrid progenies  . However, hybrid cotton has been produced successfully on primary basis since 1960s in countries like China and India due to availability of cheaper labor.
Inbreeding depression is also related to high heterosis in F1 hybrids and has to search for moderate type of heterosis which has some stability to have less inbreeding depression at F2 level  . Allelic and non-allelic interactions of genes in presence of specific environmental effects will lead us to successful heterosis results. Superiority of hybrids over commercial cultivars and genotypes is known as useful heterosis  . It is a well-known fact that without proper combination of parents, heterosis does not occur. Heterosis can be used for enhancement of cotton production by exploitation of heterozygosity and to get such cotton hybrids which are superior to best parents. The comparison of performance of the best hybrids with standard cultivars will result in determination of economic heterosis.
The F2 lines revealed larger range of adaptation regarding to their parents and F1 hybrids due to presence of genetic variation and larger heterogenity   . F2s can produce better combinations of fiber quality and yield than their parents and F2s yield was highly correlated with F1s and parental lines. It is expected that F2 populations express only 50% of the economic heterosis shown by F1 hybrids, and even less when heterosis is defined in terms of the higher yielding parent. Nonetheless, F2 lines with lower inbreeding depression in yield and superior performance than well-adapted cultivars have been found. The existence of such lines lends credibility to the use of F2s in hybrid cotton production. Previous findings are also of view about the F2 hybrids heterosis in cotton  . On the other hand, a group of physical properties can play a vital role in determination of textile use and economic value of cotton. The cotton hybrid can be produced through heterobeltiosis by crossing two cotton cultivars to enhance seed production in combination with better fiber quality  .
A successful cotton breeding program depends upon the choice and use of best parental lines for crossing and selection of favorable gene combinations. Improvements in cotton yields can be made by applying dependable tools to cotton breeders regarding to heterotic studies and inbreeding depression. For enhancement of production per unit area by proper availability of environment, identification and selection of genotypes with best genetic potential is a regular requirement. In view of economic importance of hybrid cotton cultivation and importance of selected hybrids in segregating generations, the current research was undertaken to study the genetic potential, heterotic effects and inbreeding depression for yield related traits in 6 × 6 F2 diallel populations of upland cotton.
2. Materials and Methods
Genetic material: Breeding material used during the study of genetic potential, heterotic effects and inbreeding depression for yield related traits in 6 × 6 F2 populations of upland cotton were consisted of six different Gossypium hirsutum L. genotypes i.e. CIM-446, CIM-496, CIM-499, CIM-506, CIM-554 and CIM-707 (having broad genetic base and varied by pedigree, date of release, seed cotton yield and fiber quality traits) and their 30 F2 cross combinations in a complete diallel fashion. Some of the characteristics of used varieties/genotypes during the study have been mentioned below.
Experimental design and field procedures: The mean performance, heterosis
and inbreeding depression in parental genotypes and their F2 populations of upland cotton for polygenic traits were conducted at Khyber Pakhtunkhwa Agricultural University, Peshawar during 2010. The hand sown method was used for parental genotypes and their F2s in a randomized complete block (RCB) design with three replications. Each treatment was consisted of two rows having four meter length with 30 cm plant to plant and 75 cm row to row distance. All the recommended cultural practices and inputs were applied for all the entries from time of sowing till the harvesting including fertilizer, hoeing, irrigation and pest control and the crop was grown in uniform conditions to reduce the environmental variability. Boll picking was done for two times on single plant basis and ginning was made with eight saw-gins.
Statistical analysis: Data for all the variables were analyzed through analysis of variance (ANOVA)  . After getting the significant differences, the means were further compared and separated with least significant difference (LSD) test.
Heterosis: The F2 heterosis over mid parents was calculated in terms of percent increase (+) or decrease (−) of F2 hybrids against its mid parent values according to the following formula  .
The F2 heterobeltiosis (heterosis over better parents) was formulated in terms of percent increase/decrease of F2 hybrid over its better parent according to  by using following formula.
The “t” test was used to determine whether the mid and better parents F2 heterosis was significant or not. The “t” values were computed by using the following formula according to  .
= Mean of the ijth cross.
= Mid parent value for the ijth cross.
= Better parent value for the ijth cross.
EMS = Error mean square.
For assessment of economic heterosis, the cultivar CIM-473 was used as check cultivar and was also grown with other cultivars but was not included in 6 × 6 diallel hybrids. The check cultivar was compared with mean values of other parental cultivars and F2 hybrids and economic heterosis was formulated.
Inbreeding depression: Inbreeding depression in F2 hybrids was calculated as percent decrease of F2 hybrids when compared with F1 hybrid means as outlined by  .
3. Results and Discussion
Regarding analysis of variance, the mean values for thirty six cotton genotypes including six parents and their 30 F2 hybrids revealed highly significant differences (p ≤ 0.01) for seeds per locule, seeds per bolls, lint % and seed cotton yield per plant, respectively (Table 1). The traits wise results about genetic potential, heterosis over mid and better parents and inbreeding depression in F2 populations in light of previous review are discussed as follows.
Seeds per locule: Seeds per locule varied from 5.26 (CIM-554) to 7.12 (CIM-446) among parents and 4.43 (CIM-506 × CIM-496) to 6.60 (CIM-496 × CIM-499 & CIM-554 × CIM-707) in F2 population (Table 2). Seeds per locule play a vital role in determining of seed cotton yield and is highly associated with production of boll number means high boll number will lead to more number of locules and ultimately produce more seed cotton yield, before ginning. Out of 30 F2 cross combinations, 20 showed positive value of mid parents heterosis ranged from 0.00% (CIM-446 × CIM-554) to 22.22% (CIM-554 × CIM-707), while 14 F2 hybrids revealed superior performances over their parents ranged from 0.17% (CIM-496 × CIM-707) to 19.13% (CIM-554 × CIM-707). Remaining, F2 population showed negative heterotic performance for mid parent and better parent. Collectively, 22 F2 hybrids showed significant heterosis for mid parent and better parent, respectively     . Moreover, 23 F2 hybrids revealed 76.66% economic heterosis ranged from 1.99% (CIM-554 × CIM-506) to 19.13% (CIM-554 × CIM-707). It is expected that F2 populations express 50% of the economic heterosis shown by F1 hybrids, and even less when heterosis is defined in terms of the higher yielding parent. Overall, 66.66%, 46.66% and 76.66% heterotic performance was recorded for mid parent, better parent and for economic heterosis, respectively (Table 3). In case of inbreeding depression, F2 population performed better and showed negative inbreeding depression except the cross CIM-554 × CIM-499 which only performed (0.00%) positively means F2 population have increased the number of seeds per locule which is desirable
Table 1. Mean squares and CV% of various Morpho-yield traits of upland cotton.
**, Significant at p ≤ 0.01.
Table 2. Mean performance of parental cultivars and F2s for Morpho-yield traits of upland cotton.
Table 3. Heterosis (MP, BP & Eco.) and Inbreeding Depression for Seeds locule−1 and Seeds boll −1 in F2s of upland cotton.
Check: CIM-473 = 5.54 Check: CIM-473 = 24.10. *: Significant.
to enhance the seed cotton yield at end. Highest negative inbreeding depression was noted in CIM-506 × CIM-496 while lowest negative inbreeding depression was observed in CIM-707 × CIM-506 having values −41.86% and −3.64%, respectively (Table 3). It has also been suggested that little inbreeding depression exists for F2 and F3 generations and it is possible to screen and select high yielding F2 hybrids  . Indication of high inbreeding depression even by having superior heterotic performances was also revealed that high performing hybrids had showed high inbreeding depression. It has elaborated that F2 populations can be used as hybrid cotton if have better performance over their superior parents because F2 crop can easily be managed with increased amount of seed produced by F1 plants  . Therefore, in cotton the F2 populations could be used for hybrid cotton production.
Seeds per boll: Seeds per boll varied from 21.10 (CIM-554) to 28.03 (CIM-446) among parental cultivars and 20.40 (CIM-496 × CIM-506) to 28.50 (CIM-506 × CIM-707) among F2 population (Table 2). Seeds per boll play a vital role in contributing seed cotton yield. Path coefficient analysis showed that seeds per boll had positive effect on yield. Out of 30 F2 hybrids, 18 showed positive mid parent heterosis ranged from 0.00% (CIM-707 × CIM-496) to 23.32% (CIM-506 × CIM-707), while 09 F2 hybrids performed superior over their parents and showed positive heterobeltiosis ranged from 1.22% (CIM-506 × CIM-554) to 18.60% (CIM-506 × CIM-707). Rest of F2 population revealed negative heterotic performances for the respective trait. Collectively, 06 F2 hybrids showed significant heterosis for both mid and better parent, respectively. F2 hybrid heterosis in cotton has been reported by several workers and F2s can express at least 50% of the economic heterosis shown by F1 hybrids, which can lead to cultivar improvement    . Moreover, half number of F2 hybrids showed 50% positive economic heterosis ranged from 0.41% (CIM-496 × CIM-707) to 18.26% (CIM-506 × CIM-707). Overall, 60%, 30% and 50% of mid parent, better parent and economic heterosis has been recorded for the said trait (Table 3). It was suggested that little inbreeding depression exists for F2 and F3 generations and it is possible to screen and select high yielding F2 hybrids as already mentioned above. In case of inbreeding depression, only one F2 hybrid (CIM-506 × CIM-707) performed positively and showed (2.41%) positive inbreeding depression while rest of F2 population performed well by showing stability even after segregation and revealed negative inbreeding depression ranged from −5.71% (CIM-499 × CIM-446) to −37.15% (CIM-506 × CIM-554) (Table 3). It was also showed that high performing hybrids showed high inbreeding depression. High inbreeding depression was performed by high performing hybrids. The existence of superior lines in F2 reveals that superior alleles for the trait may have accumulated in the same line. The seeds per ball is a quantitative trait so additive gene action may be involved and those have contributed in the better performance. Results supported the idea that F2 populations could work as a hybrid crop if properly managed and if parents selected on basis of F2 performance, because F1 hybrids cannot clarify the stability of F2 populations  . Therefore, such F2 populations would be desirable to use as hybrid cotton to enhance the boll number and eventually seed cotton yield.
Lint% (GOT): The lint % (GOT) varied from 32.20 (CIM-707) to 34.80 (CIM-506) among parents and 32.10 (CIM-446 × CIM-496) to 35.05 (CIM-496 × CIM-446) among F2 population (Table 2). Lint% after ginning of seed cotton becomes a chief output for cotton breeders because cotton is basically grown for purpose of obtaining fibers. We extract edible oil from cotton seed which serves as a byproduct for people used. Out of 30 F2 cross combinations, 13 cross combinations showed positive mid parent heterosis ranged from 0.32% (CIM-506 × CIM-496) to 6.27% (CIM-496 × CIM-446). Further, 09 F2 hybrids revealed positive heterobeltiosis ranged from 0.36% (CIM-707 × CIM-554) to 4.23% (CIM- 554 × CIM-446), while remaining all F2 population performed heterotically negative for both mid and better parent. Moreover, only 01 F2 hybrid (CIM-496 × CIM-446) revealed (0.68%) positive economic heterosis while rest of F2 population showed negative economic heterosis. Overall, 43.33%, 30% and 3.33% of the F2 population showed positive mid parent, better parent and economic heterosis, respectively for lint% (Table 4). The current findings are accorded with the findings that revealed low lint% values for heterosis after staple length  . In case of inbreeding depression, more than half number of F2 hybrids revealed negative inbreeding depression ranged from −1.90% (CIM-499 × CIM-707) to −12.07% (CIM-554 × CIM-707) means F2 population has shown stability and performed well after segregation in comparison with the F1 generation. The F2 hybrids including CIM-446 × CIM-506, CIM-506 × CIM-707 and CIM-496 × CIM-446 performed positively by having values of 15.55%, 9.76% and 8.54%, respectively and revealed positive inbreeding depression for the said trait (Table 4). Lower magnitude of inbreeding depression for lint % and staple length indicated that additive genes were responsible for the expression of these traits  .
Seed cotton yield per plant: Seed cotton yield varied from 55.74 (CIM-554) to 85.47 (CIM-496) among parental cultivars and 45.94 (CIM-446 × CIM-499) to 92.04 (CIM-506 × CIM-496) among F2 population (Table 2). Seed cotton yield per plant is a major and very important trait in growing cotton besides lint yield. As seed cotton yield was highly depended on boll plant−1 due to presence of close relationship between them  . Out of 30 cross combinations, 11 showed positive mid parent heterosis ranged from 0.53% (CIM-446 × CIM-496) to 30.71% (CIM-554 × CIM-499) while only 05 F2 hybrids showed positive heterobeltiosis ranged from 2.23% (CIM-707 × CIM-499) to 15.41% (CIM-446 × CIM-554). Collectively, 06 F2 hybrids showed significant heterosis for both mid and better parent, respectively. It was observed that significant mid parent heterosis in 13 hybrids and heterobeltiosis in 11 hybrids for seed cotton yield in upland cotton. Seed cotton yield was found with greater variation mostly affected by different yield contributing traits in different previous studies  . Moreover, 06 F2 hybrids revealed positive economic heterosis ranged from 0.22% (CIM-496 ×
Table 4. Heterosis (MP, BP & Eco.) and Inbreeding Depression for lint% and Seed cotton yield plant −1 in F2s of upland cotton.
Check: CIM-473 = 34.81 Check: CIM-473 = 76.72. *: Significant.
CIM-554) to 19.97% (CIM-506 × CIM-496). In case of inbreeding depression, all F2 populations showed negative inbreeding depression except three F2 hybrids. Overall, 36.66%, 16.66% and 20% heterosis was noted for mid parent, better parent and for economic heterosis (Table 4). Out of 30 cross combinations, 27 cross combination revealed negative inbreeding depression means F2 has more stability to perform well after segregation and have great potential to explore it in future to come out with a fruitful results. Maximum inbreeding depression was observed in the cross CIM-554 × CIM-446 (−68.89%) while minimum inbreeding depression were recorded in CIM-506 × CIM-707 (−0.48%). High inbreeding depression has also been observed for the seed cotton yield. Moreover, only three F2 hybrids i.e., CIM-499 × CIM-707, CIM-496 × CIM-499 and CIM-707 × CIM-506 performed positively by having values of 37.36%, 7.04% and 1.97%, respectively (Table 4). F2 hybrids with lower inbreeding depression in yield expressed superior performance over well-adapted cultivars as already mentioned by Meredith. Even after inbreeding depression in F2s, presence of some promising population can reveal better performance hence positive selection can provide better opportunity for further improvement.
The F2 hybrids manifested a remarkable percentage and ranges of mid parent, better parent and economic heterosis having negligible effects of inbreeding depression for majority of the characters of upland cotton. The remarkable heterotic performance for mid parent and better parent was revealed by the F2 population including, CIM-554 × CIM-707, CIM-499 × CIM-506, CIM-499 × CIM-707, CIM-506 × CIM-707, CIM-499 × CIM-554, CIM-554 × CIM-499, CIM-496 × CIM-446 and CIM-446 × CIM-554, respectively while CIM-496 × CIM-499, CIM-506 × CIM-707, CIM-496 × CIM-446, CIM-506 × CIM-496 and CIM-554 × CIM-499 were the F2 hybrids which showed the highest positive economic heterosis. Collectively, CIM-499 × CIM-554, CIM-554 × CIM-499, CIM-554 × CIM-707 and CIM-506 × CIM-707 showed superiorly with significant heterosis for yield and yield contributing traits for all parameters except lint%. Majority of the F2 population did not reveal inbreeding depression meant having superiority over F1s even after segregation which is more desirable for selection of best hybrid cotton. Overall, these parental cultivars and F2 hybrids can be used and studied for further betterment in advanced generations for improvement of seed cotton as they have the capabilities to produce excess yield and production of cotton in future breeding program.
 Pareetha, S. and Raveendren, T.S. (2008) Genetic Appraisal of Yield and Fiber Quality Traits in Cotton Using Interspecific F2, F3 and F4 Population. Journal of Integrative Plant Biology, 3, 136-142.
 Soomro, A.R. and Kalhoro, A.D. (2000) Hybrid Vigor (F1) and Inbreeding Depression (F2) for Some Economic Traits in Crosses between Glandless and Glanded Cotton. Pakistan Journal of Biological Sciences, 3, 2013-2015.
 Wu, Y.T., Yin, J.M., Guo, W.Z., Zhu, X.P. and Zhang, T.Z. (2004) Heterosis Performance of Yield and Fiber Quality in F1 and F2 Hybrids in Upland Cotton. Plant Breed, 123, 285-289.
 Tang, B., Jenkins, J.N., McCarty, J.C. and Watson, C.E. (1993) F2 Hybrids of Host Plant Germplasm and Cotton Cultivars. 1. Heterosis and Combining Ability for Lint Yield and Yield Components. Crop Science, 33, 700-705.
 Wynne, J.C., Emery, D.A. and Rice, P.M. (1970) I. Combining Ability Estimates in Arachis hypogeae L. 11. Field Performance of F1 Hybrids. Crop Science, 10, 713-715.
 Percy, R.G., Cantrell, R.G. and Zhang, J. (2006) Genetic Variation for Agronomic and Fiber Properties in an Introgressed Recombinant Inbred Population of Cotton. Crop Science, 46, 1311-1317.
 Batool, S., Khan, N.U., Makhdoom, K., Bibi, Z., Hassan, G., Marwat, K.B., Farhatullah, Mohammad, F., Raziuddin and Khan, I.A. (2010) Heritability and Genetic Potential of Upland Cotton Genotypes for Morpho-Yield Traits. Pakistan Journal of Botany, 42, 1057-1064.
 Yuan, Y.L., Zhang, T.Z., Guo, W.Z., Pan, J.J. and Kohel, R.J. (2001) Heterosis and Gene Action of Boll Weight and Lint Percentage in High Quality Fiber Property Varieties in Upland Cotton. Acta Agronomica Sinica, 28, 196-202.
 Khan, N.U., Abro, H.K., Kumbhhar, M.B., Hassan, G. and Khan, M. (1999) Exploitation of Heterosis Can Combat Cotton Leaf Curl Virus (CLCuV) Incidence in Cotton (G. hirsutum L). The Pakistan Cottons, 43, 21-33.
 Xing, C.Z., Yu, S.X., Guo, L.P., Miao, C.D., Feng, W.J., Wang, H.L. and Zhao, Y.L. (2000) Heterosis Performance and Correlation Analysis on Economic Traits of Upland Cotton Hybrid in Different Ecological Environments. Cotton Science, 19, 3-7.
 Wei, X.C., Li, Q.Z., Pang, J.Q., Zhang, J., Zhao, J.H. and Wang, L.G. (2002) Heterosis of Pre-Forest Lint Yield of Hybrid between Cultivars or Lines within Upland Cotton (G. hirsutum L.). Cotton Science, 14, 269-272.
 Galanopoulou-Sendouca, S. and Roupalias, G. (1999) Performance of Cotton F1 Hybrids and Its Relation to the Mean Yield of Advanced Bulk Generations. European Journal of Agronomy, 11, 53-62.
 Khan, N.U., Hassan, G., Kumbhar, M.B., Kang, S., Khan, I., Parveen, A., Aiman, U. and Saeed, M. (2007) Heterosis and Inbreeding Depression and Mean Performance in Segregating Generations in Upland Cotton. European Journal of Scientific Research, 17, 531-546.
 Khan, N.U., Hassan, G., Marwat, K.B., Farhatullah, Batool, S., Makhdoom, K., Khan, I., Khan, I.A. and Ahmed, W. (2009) Genetics Variability and Heritability in Upland Cotton. Pakistan Journal of Botany, 41, 1695-1705.
 Babar, S.B., Soomro, A.R., Anjum, R. and Kalawar, M.S. (2001) Estimation of Heterosis, Heterobeltiosis and Economic Heterosis in Upland Cotton (G. hirsutum L). Pakistan Journal of Biological Sciences, 4, 518-520.