The genetic makeup of an attribute comprises additive, dominance and epistasis variance. The information of the relative degree of genetic variance components is fundamental in the selection of appropriate breeding methodologies. The diallele analysis is the quickest method to enable predictions to be made based on early generations, which could greatly increase the efficiency of breeding programs. Component analysis revealed that additive-dominance model fitted only for number of sympodial branches per plant in both the generations under all the three environments. The present study indicates the preponderance of dominance components for inheritance of this character. Heritability values in the narrow sense for this character varied from low (17.50 % in E₂ F₁) to moderate (46.58% in E₂ F₂). Since both additive and non-additive types of gene actions were present in different proportions, reciprocal recurrent selection is suggested for the improvement of the character.

The information of the relative degree of genetic variance components is fundamental in the selection of appropriate breeding methodologies. A number of studies have shown the prevalence of different types of gene actions. Hence, the knowledge of the prevalence of the type of gene actions in the present set of material is also required for formulating an effective breeding program for the improvement of the trait under study. This shows the importance of the testing of genetic material in more than one environment in order to obtain unbiased estimates of various genetic components. Therefore, the present investigation was undertaken to bring out the various aspects of gene action for various characters through a study of diallele crosses with 10 parents in upland cotton.

Ten genetically diverse lines of G. hirsutum L., namely, F 1867, F 1861, C-2602-WIR-6109, PIL 8-5, LH 1836, LH 1861, LH 1896, Pusa 101, H 1123 and RS 2115, were crossed manually in all possible combinations in half diallele fashion during Kharif 2000. All 45 F₁'s were grown during kharif 2001 to get respective F₂'s. All 10 parents, their 45 F₁'s and 45 F₂'s were raised in randomized block design replicated thrice in three environments (E₁, E₂ and E₃) at the Agricultural Research Station, Sriganganagar, during Kharif 2002. The environments were created by applying three sowing dates, i.e., early-sowing on May 11 (E₁), timely-sowing on May 26 (E₂) and late-sowing on June 10, 2002 (E₃). The temperature during the week of sowing ranged from 27.80 °C to 46.01 °C for E₁; 26.30 °C to 44.17 °C for E₂; and 29.71 °C to 44.37 °C for E₃ (Appendix). Each entry consisted of 6 meter long, two rows for non-segregating generations and four rows for segregating population (F₂'s). The inter-row and intra-row spacings were kept 67.5 cm and 60 cm, respectively. Two border rows were grown around the experiment to avoid border effects. All the recommended cultural and management practices were followed to raise a good crop. The sample size consisted of 10 competitive randomly selected plants for non-segregating material and 30 plants for segregating material per replication. Observation on days to 50% flowering, 2.5% span length (mm) and fiber fineness (micronaire value) were recorded on plot basis, while for plant height (cm), number of sympodia per plant, number of monopodia per plant, number of bolls per plant, number of seeds per boll, ginning percentage (%), seed index (g), lint index (g), and seed cotton yield were recorded on competitive randomly selected plants per replication. The mean values over replications of all the characters for parents, F₁'s and F₂'s were subjected to diallele cross analysis. Various components of genetic variance were computed by the diallele cross method by Hayman (1954). For those characters where the genotype ´ environment interaction was significant, the components of genetic variance were estimated from the data of individual environment, and where genotype ´ environment interaction was nonsignificant, the pooled data were considered for the above estimation.

Genetics & Evolution Journal, Component Analysis, Breeding Methodologies, Breeding Programs, Environmental Components, Multilocation Tests, Dominant Alleles, Asymmetrical Distributions, Recessive Alleles, Dominant Genes, Recessive Genes.