Effects of sexual reproduction and its management on the genetic diversity of vegetatively propagated crops in traditional farming systems : the case of Oxalis tuberosa Mol

Oca (Oxalis tuberosa Mol.) is a tuber crop that is traditionally cultivated in the Andes where it plays an important role for food security. It is vegetatively propagated by the farmers but presents a fully functional, complex and poorly studied sexual reproduction system. Both reproductive systems act antagonistically on genetic diversity. On one hand, vegetative propagation allows farmers to fix and easily propagate agronomically valuable genotypes. On the other hand, sexual reproduction allows to recombine parental genotypes into new genotypes. Sexual reproduction of oca has never been observed in natural conditions and its use by farmers has never been reported. Modern plant breeding has never been attempted for oca. Oca is currently under the threat of genetic erosion. Obtaining a better understanding of the dynamics of its genetic diversity is crucial to the development of effective conservation strategies. The development of plant breeding programs would permit to renew the interest of farmers for this crop, which suffers from the competition of modern potato varieties. The present thesis investigates the effect of the sexual reproduction of oca and its management (both traditional and modern) on the genetic diversity of this crop trough a multidisciplinary approach combining genetics, biology and ethnobotany. In this thesis, the sexual reproduction of oca was shown to be functional under natural conditions. All the sexually-produced individuals showed unique genotypes, distinct from the parental landraces. Following interviews with farmers, an unconscious incorporation of sexually-produced new genotypes into the landraces traditionally cultivated by farmers was evidenced to be highly probable. SSR molecular markers studies showed that farmers grow complex landraces, showing important intra-landrace diversity and very distinct genotypes arising from different sexual reproduction events. A lot of seed exchanges are undertaken by farmers, which manage very differently self-consumed and cash landraces. Therefore, the introduced new genotypes can be propagated to farmers of other communities and regions very fast. Outbreeding depression was observed for the first time for vegetatively propagated crops. By measuring genetic distances between parental genotypes, plant breeders will be able to avoid this phenomenon. A pathway for effective plant breeding programs was proposed. It should involve participatory techniques that will allow a sustainable use of genetic resources. In conclusion, this thesis fills an important gap in the understanding of the use of sexual reproduction of minor vegetatively propagated crops. Perspectives concerning plant breeding and genetic resources conservation of these crops are formulated.