Roselle (Hibiscus sabdariffa L. var. altissima Wester) is a bast fiber crop of global economic importance. Although West Africa is considered the centre of diversity, roselle research and utilization is widely ignored. The awakening of industrialization in Ghana presents roselle as a candidate crop for exploration, however, information on genotypes of economic importance is lacking. Our objective was to map roselle population hotspots in northern Ghana and examine genetic variability therein. Thirty-six roselle accessions collected from five regions in Ghana were planted in field trials using a 6 × 6 lattice square design in three replications and evaluated for seven qualitative and four quantitative morphological traits covering plant type, leaf and stem characteristics, and growth habit. Data were analysed by Shannon-Weiner Diversity Index (SDI) and analysis of variance. A large variability was identified between the accessions. The mean SDI values in the 18 districts ranged from 0.53 to 0.73 with Savelugu-Nanton district having the largest diversity and having accessions with the highest mean plant height of 308.27±48.91 cm, highest branching point at 107.19±64.66 cm, and few branches not exceeding 5.0 in number. Majority of the accessions exhibited low branching points. The most variable trait was branch number with SDI of 0.83±0.12. Accessions HA-07, HA-11, HA-12, HA-21, and HA-33 ranked highest with respect to plant height with few branches at high branching points, and large basal diameter. The ample diversity in roselle and identification of genotypes of economic importance await their exploitation for genetic improvement, particularly for fiber yield.

Roselle (Hibiscus sabdariffa var. altissima), a bast fibre crop adapted to the warm climate of Northern Ghana, offers a great economic potential not yet explored for lack of information on its distribution, collection, and genetic diversity. Little variability is reported in exotic genotypes to merit trait improvement. The objective of this study is to investigate distribution and diversity in roselle of Northern Ghana. Twenty-five accessions collected from seven districts were field evaluated in a 5×5 lattice square design in three replications at twelve qualitative and five quantitative morphological traits. Data were analysed for within-and between-population variability and multivariate analysis. Large within-population variability of SDI 0.72 to 0.87 was identified in accessions of Kassena-Nankana East district. The most variable traits, plant height and branch number, varied from 184 cm to 284 cm with six accessions HA-44, HA-47, HA-43, HA-38, HA-52, and HA-42 having the tallest plants and least basal branching of four. Mean flowering time was between 96 and 104 days. Mean Euclidean distance of 3.03 ± 0.90 ranged from 0.41 to 5.17. Based on means across pairwise distances of 2.22 and 3.94, three accessions were divergent, namely, HA-61 (3.94), HA-57 (3.66) and HA-59 (3.63). Clustering and principal components analyses delineated three distinct groups. The first three PCs explained 100% of the variance. The ample diversity in roselle awaits exploitation for genetic improvement, particularly for fibre yield.