ABSTRACT

Genetic diversity is essential for effective and efficient breeding programme. Planned plant introduction becomes inevitable when there is narrow genetic base for desirable traits. To know the status of genetic diversity in a collection, evaluation and characterization are essential. Therefore, two experiments were carried out in two locations (Umudike, Abia State, and Otobi, Benue State) to evaluate and characterize 18 genotypes collected from Ibadan, and 3 checks from the National Root Crops Research Institute, Umudike for their agronomic and morphological attributes. The trials were established in a randomized complete block design with 3 replications. 37 descriptor traits were characterized and 17 agronornic and .morphological traits were evaluated. Broad descriptor variations were observed among the genotypes for most traits, while little or no variations were observed in colour of stigma and style, number of sepals, and secondary roots skin colour. Minor differences were observed in the degree of expression of some descriptor traits in some genotypes in the two locations. In general, Otobi supported the expression of metric traits more than Umudike. For agronomic traits, only vine length showed significant (P<O.05) difference among the 21 genotypes at Otobi, while no trait showed any significant variation among the genotypes in Umudike. However, Digitate recorded the highest yield (total roots weight) across the two locations (18.75 tiha). All traits except stand count at harvest were positively and significantly (P<0.005) correlated with yield (total roots weight). Forward selection multiple regression analysis identified weight of saleable and unsaleable roots, and total number of roots as the most important yield components. The first three principal component axes explained 76% of the observed variations among the genotypes. The three axes identified weight of saleable and total roots, stand count at harvest, number of unsaleable roots, number of branches, incidence of roots Cylas and length of smallest roots as traits that characterized the genotypes. Across locations, environmental variances were much higher than genetic variances. This resulted into moderate broad sense heritability estimates for number of saleable roots (30.00%), weight of saleable and total roots (39.20 and 40.76% respectively), and incidence of roots Cylas (35.35%). Others recorded low to negative heritability estimates. This resulted to high phenotypic coefficients of variation (PCV) and low genotypic coefficients of variation (GCV) in all the traits. Genetic advance(%) was however high for weight of total and saleable roots (81.62 and 84.3 6% respectively), and for incidence of roots Cylas (75.42%). Therefore, genetic diversity among the 21 genotypes was low. No duplicate genotypes were however identified among the genotypes characterized. Weight of marketable and unmarketable roots, and total number of roots had been identified as important yield components for sweet potato crop improvement. Digitate, Caroline Lee and TIS 4400-2 are recommended for possible inclusion in a sweet potato crop improvement where better lines are not available.