GENETIC VARIABILITY, INTER-RELATIONSHIP AND PERFORMANCE OF GROUNDNUT (Arachis hypogea L.) VARIETIES FOR YIELD AND OIL TRAITS IN ETHIOPIA

Abstract

Genetic analysis for traits of interest in crop varieties is necessary in designinig appropriate breeding methods. Several varieties of groundnut have been recommended for cultivation in Ethiopia; however, little genetic information has been generated so far. Therefore, this research was conducted with the objectives of assessing genetic variation and association of traits in groundnut varieties; estimate genetic progresses made through decades of breeding efforts and determine effects of genotype x environment interaction (GEI) on yield, oil content and oil quality traits. The varieties were evaluated for 12 agromorphological and 17 oil traits and quality parameters. The ANOVA computed across locations revealed the prevalence of significant difference among the varieties for all traits. High to moderate GCV, high heritability and expected genetic advance were observed for stearic acid, eicosenoic acid, palmitic acid, arachidic acid, behenic acid, lignoceric acid and O/L ratio. These suggested genotypes can be selected based on their mean values for all traits since additive gene action is prevalent in such traits. However, direct selection for oil content was not rewarding due to the low heritability of the trait. The path analysis based on genetic correlation has shown that selection for oil content trait is effective through OY, GY, NMP and NSPOD. The path analysis has also shown that breeding for high O/L ratio can be conducted through selection for AGBP, NSPOD, oil content, TPUS/TS and oleic acid traits likewise breeding for oil yield (OY) can be conducted via selection for AGBP, NBP and TPUS/TS. The cluster mean analysis, for agromorphological traits, has shown that genotypes in clusters 1, 2, & 4 can be used as parents in hybridization program for grain yield improvement. Similarly, cluster mean analysis for biochemical analysis of oil traits has shown genotypes in cluster-1 as promising for improvement of oil quality trait while genotypes in cluster-2 including Tole-1, Tole-2 and Behagudo can be used for improvement of most important traits in oil crops including grain yield, oil content, oil yield and oil quality traits. The PCA analysis has shown xxi the three principal components (PC) accounted for 83% of the variation of cultivars for 12 agromorphological traits of which high positive first component loadings were from NBP, AGBP, NMP, PWP, SWP as well as GY and found to associate with NC-343, Bahajidu, Lote, Manipeter, Roba, Werer-962, Tole-1, Tole-2 and Oldhale varieties. The PC2 had high positive component loading from 100SW, PWP and GY that correspond to Behagudo, Fetene, Manipeter, Werer-962, and Werer-961 varieties. The GY has shown positive loading in all the first three components; however, it had highest positive loading in PC2 indicating the cultivars associated in this PC2 had highest grain yield. The PCA analysis based on Euclidean distance matrix (D) for 17 oil traits and quality parameters has shown that the first principal component had high positive component loading from monounsaturated and saturated fatty acid traits/parameters: oleic, TMUS, O/L, OY, GY, TUS, lignoceric and eicosenoic acids. The corresponding varieties were Oldhale, Lote, Beha jidu, Werer-962, and NC-343. On the other hand, the first component has got high negative loading from polyunsaturated fatty acid traits/parameters including palmitic, linoleic, TPUS/TS, TS and IV. The corresponding loads were Fetene, Sedi, Werer-961 and Werer-963 varieties. Thus, the most distinct varieties were being Oldhale/Lote/Werer-962 and Fetene/Werer-961. Similarly, PCA2 had high positive component loading from arachidic acid, grain yield, behenic acid, stearic acid, TS, oil content and O/L ratio. The corresponding genotypes were Baha gudo, Tole-2 and Tole-1 whilst negative PC2 loads from IV, eicosenoic acid, TPUS/TS and TUS. The corresponding varieties were Werer-963, Sedi, Roba, NC-343 and Bulki. Thus, the most distinct varieties in the second component were being Beha gudo and Werer-963. The relative genetic gain for grain yield during the last 36 (1976 to 2012) years' breeding effort was 1.08% and the varieties recommended for cultivation since 1976 had increased grain yield, 100 seed weight, plant height and harvest index since these traits showed significant correlation with year of release of varieties. The genetic gain in groundnut oil content and oil quality across four locations in Ethiopia using aggregative data from 1976 to 2012 was shown that the overall decrease in mean of oil content, oil yield and also oil quality traits. The absolute gain for oil content, iodine value, palmitic acid, oleic acid, linoleic acid, eicosenoic acid, lignoceric acid, TMUS, TPUS, TPUS/TS and O/L ratio for the time period of 1976 to 2012 did not differ significantly from zero, thus no genetic gain can be estimated for these traits. However, the absolute gain for oil yield, stearic acid, arachidic acid, behenic xxii acid and TS was positive and significant indicating considerable progress has been made with repect to these traits. The estimates of genetic gain showed that substantial efforts in breeding activities have been made to increase the grain yield of groundnut but not for oil content and quality traits in Ethiopia which needs due attention in the future breeding activities, to meet the ever growing consumers' and industrial demand of groundnut. The ANOVA across location has shown significant mean squares for genotype x location suggested further analysis of GEI and stability for grain yield, oil yield, oil content and oleic to linoleic acid ratio (as a major oil quality parameter. The GGE biplot has shown Guba for grain yield and oil quality, Mechara for oil content were the most favorable environments that could be used to test cultivars for the respective traits, while Bahagudo for grain yield, oil yield and also oil quality; Tole-2 for oil content were identified as ideal varieties that are highest performing and stable across environments. Generally, the results suggested the importance of genotype with the interaction of environment for the expression of these traits and indicating the importance of testing cultivars across environments to identify high performing and stable cultivars. Further studies should be conducted by considering multiyears (across seasons).