Characterisation and diallel analysis of commercially planted cotton (Gossypium hirsutum L.) germplasm in Tanzania

[摘要] English: 1. Thirty cotton varieties were studied. Seven parents generated 21 F1 diallel progenies.The main objective was to evaluate genotypes through oil content, fatty acids,morphological and molecular characteristics. Statistical procedures were performed withAgrobase, NCSS and NTYSS computer programmes. Correlation, heterosis, heritability,G x E interaction and stability values were obtained. Stability across four environmentswas analysed by five stability measures, that of Shukla, Wricke, Lin and Binns, Eberhartand Russell and the AMMI analysis.2. Significant differences were observed for oil and fatty acids content. Oil content resultsrevealed an average of 20.23%. The ratio for polyunsaturated:saturated fatty acid was2:1 and for unsaturated:saturated 3:1. The predominant fatty acids were linoleic,palmitic, oleic, stearic, myristic, palmitoleic, arachidic and behenic. Palmitic fatty acidwas high while oleic and stearic fatty acids were low for almost all varieties studied. Oiland fatty acid content were significantly influenced by the environment. CIM 70 andCyto 12/74 can be selected as parents for increased oleic and stearic fatty acid andStoneville 506 as a parent for reduced palmitic fatty acid, increased oleic, stearic andlinolenic fatty acids in hybridisation.3. ANOVA results indicated significant differences between varieties for all agronomicalcharacteristics studied, where NTA 93-21 and NTA 93-15 were the best performers almostfor all characteristics. Genetic distances for morphological (qualitative characteristics)analysis ranged from 0.18 to 0.80. Two major groups, one including many varieties fromAfrica and the other including varieties from the USA, were revealed.4. AFLP analysis showed high genetic similarities (0.85 to 0.98). Varieties weredistinguished according to pedigree and origin. The increased genetic similarityobserved, indicates the need for the introduction of more diverse cotton material.Comparison of agronomical and AFLP markers for determining genetic diversity indicated that genetic similarities between varieties were low for agronomical comparedto AFLP markers. The correlation coefficient between AFLP analysis and agronomicalgenetic similarities was not significant,5. Highly significant differences between genotypes over environments were observed andNTA 93-21 crosses outperformed the seven parents for almost all characteristics. Theparent Okra leaf performed poorly for many characteristics. Trial sites Ukirigurufollowed by Kanziga were high yielding environments.6. Combining ability analysis results indicated significant variation in GCA estimatesamong parents for different characteristics. Genotypes having high positive GCAestimates for yield components exhibited negative GCA estimates for fibre quality,excluding NTA 93-21. Some combinations indicated positive and significant SCAeffects for seedcotton, boll weight, GOT and lint yield. The GCA/SCA ratios for most ofthe studied characteristics were larger than one.7. Correlation among characteristics indicated highly significant positive and negativecorrelations. The observed high and significant GCA and SCA effects for thosecharacteristics are promising for cotton population improvement. Positive heterosis forboll weight was observed for more than 90% of the combinations. Hybrids with highSCA values presented high values of heterosis. Heritability values for the studiedcharacteristics were moderate to high.8. The combined ANOVA over four environments indicated significant differencesbetween genotypes and locations for all characteristics. The percentage contribution oflocation to total variation was high compared to genotype variation except for a fewcharacteristics. Highly significant differences of G x E for lint yield and fibre strengthwere observed.9. Three stability procedures (Wricke, Shukla and ASV) identified similar stablegenotypes. Eberhart and Russell's procedure, was in harmony with the above threemethods but with some divergence. Cultivar superiority measure deviated from the otherfour measures. Therefore the procedures of Wricke, Shukla and AMMI can be appliedfor cotton stability studies.

[发布机构] University of the Free State