Diversity of durum wheat (Triticum durum Desf.) at in situ conservation sites in North Shewa and Bale, Ethiopia

Biological resources used by local people for a long period of time preserve historical records and scientific competence attained while they develop, conserve, maintain and utilize the landrace varieties of those crops sustainably, ensuring their descent to successive generations. Ethiopian farmers and the traditional farming systems they still exercise have a tremendous positive impact on maintaining wheat genetic diversity with the associated indigenous knowledge. Twenty-one durum wheat landraces, 11 of which were collected from North Shewa and the remaining 10 from Bale were examined for five qualitative morphological characters. The long beak character was monomorphic at lower altitude (< 2100 m a.s.l.). The phenotypic frequencies for characters showed that the long and intermediate beak types were found in both sites, with the long one being the predominant type in most of the strategic sites (localities). The short beak type was rare in most of the strategic sites (localities) in Bale whereas in North Shewa its frequency reached a maximum of 51%. Black glume was in rare frequency in most of the strategic sites (localities) and districts (depending on size, a district contains several localities). The three seed colour groups (white, brown and purple) were found in a similar proportion across districts, and in situ conservation zones (farmer-based on-farm crop conservation zones, and depending on size a zone contains several districts) with considerable differences among strategic sites (localities). While white glume colour was predominant at lower altitude (< 2300 m a.s.l.), brown glume was most frequent at higher altitude (> 2301 m a.s.l.). The frequency of the purple seed colour increases with increasing altitude. Intra-population polymorphism was common for most traits. The overall estimate of diversity in in situ crop conservation sites is very high (H' = 0.81 +/- 0.01). The first three canonical functions accounted for 85.5% (48.5, 19.7 and 17.4% respectively) of the variation between varieties. The hierarchical cluster analysis revealed two major clusters and eight subclusters. In most cases, the cluster analysis based on the qualitative traits did not necessarily cluster varieties on the basis of their collection sites. The two major clusters contained varieties from Bale and North Shewa in situ crop conservation zones.