African Journal of Biotechnology Vol. 3 (4), pp. 220-225, April 2004            

Full Length Research Paper

Analysis of genetic diversity in bambara groundnut [Vigna subterranea (L.) Verdc] landraces using amplified fragment length polymorphism (AFLP) markers

Wazael H. Ntundu^{1, 2*}, Inga C. Bach¹, Jørgen L. Christiansen¹ and Sven B. Andersen¹

¹Department of Agricultural Sciences, The Royal Veterinary and Agricultural University, Thorvaldsensvej 40, DK-1871 Frederiksberg C, Denmark.

²Tropical Pesticides Research Institute, National Plant Genetic Resources Centre, P O Box 3024, Arusha, Tanzania.

*Corresponding Author. Phone: +255 27 2509674. Fax: +255 27 2509674. E-mail: wntundu@yahoo.com.

Accepted 28 December 2003

Amplified fragment length polymorphism (AFLP) was used to assess genetic diversity among 100 selected bambara groundnut [Vigna subterranea (L.) Verdc] landraces from a diverse geographic area of Tanzania. Eleven informative AFLP primer combinations generated a total of 49 scorable polymorphic amplification fragments across the bambara groundnut accessions. Genetic distances between all accessions based on Jaccard’s variability index ranged from 0.1 to 0.68, with a total average of 0.3. The results showed that bambara groundnut landraces from Tanzania form a genetically diverse population, and AFLP markers can be effectively employed to assess genetic diversity and to measure genetic relationship among accessions. Cluster analysis revealed that bambara groundnut from Tanzania constitute two major groups in line with their putative geographic origins, one genetically distinct group from the Southern agro-ecological zone and a mixed group with accessions from Central, Lake Victoria and Western agro-ecological zones. The clustering of accessions compared relatively well to clustering based on phenotypic characters. However, correlation of the AFLP marker distances with phenotypic distances showed r-values of 0.4 only.

Key words: AFLP markers, genetic diversity, landraces, Tanzania, Vigna subterranean.