Genetic diversity and population differentiation of traditional fonio millet (Digitaria spp.) landraces from different agro-ecological zones of West Africa.

Fonio millets (Digitaria exilis Stapf, D. iburua Stapf) are valuable indigenous staple food crops in West Africa. In order to investigate the genetic diversity and population differentiation in these millets, a total of 122 accessions from five countries (Benin, Burkina Faso, Guinea, Mali and Togo) were analysed by Amplified Fragment Length Polymorphisms (AFLPs). Genetic distance-based UPGMA clustering and principal coordinate analysis revealed a clear-cut differentiation between the two species and a clustering of D. exilis accessions in three major genetic groups fitting to their geographical origins. Shannon's diversity index detected in D. iburua was low (H = 0.02). In D. exilis, the most widespread cultivated species, moderate levels of genetic diversity (Shannon's diversity H = 0.267; Nei's gene diversity H' = 0.355) were detected. This genetic diversity is unequally distributed with the essential part observed in the Upper Niger River basin while a very low diversity is present in the Atacora mountain zone. Analysis of molecular variance (AMOVA) revealed that a large part of the genetic variation resides among the genetic groups (70%) and the country of origin (56%), indicating a clear genetic differentiation within D. exilis. Influence of mating system (inbreeding or apomixis), agricultural selection and ecological adaptations as well as founding effects in the genetic make-up of the landraces were visible and seemed to jointly contribute to the genetic structure detected in this species. The genetic variability found between the analysed accessions was weakly correlated with their phenotypic attributes. However, the genetic groups identified differed significantly in their mean performance for some agro-morphologic traits. The results obtained are relevant for fonio millets breeding, conservation and management of their genetic resources in West Africa.