Structuration de la diversité génétique du genre Milicia: taxonomie, phylogéographie, dynamique des populations

Characterizing population genetic structure using phylogeographic approach contributes to understanding of evolutionary processes that affect plant populations. The present study investigated how past environmental variations can shape the spatial genetic diversity of an African timber tree species, Milicia excelsa (Welw) C.C. Berg. Given that plant responses to disturbances are closely linked to life history traits, this research also aimed to describe relationships between the reproductive traits of M. excelsa and its population genetic structure. Phylogeographic analyses were performed throughout the range of M. excelsa using nuclear and chloroplastic markers. Ecological characterization was conducted for a single population located in south-East Cameroon.
Since M. excelsa is similar to the other species of its genus, M. regia (A. Chev) C.C. Berg, we first performed a morphogenetic comparison of these two taxa. The results confirmed an important morphological similarity between the two species with very rare discriminating criteria. In fact Bayesian assignment methods and reproductive isolation patterns suggest the occurrence of interspecific hybridization, which appeared to be unidirectional: hybrids displayed the M. regia morphotype. Because M. regia is known to be more adapted to humid tropical forests, one may expect it to occur in the Congo basin zone. However, its range is restricted to the western part of West Africa while M. excelsa occurs from Guinea-Conakry to Zimbabwe. Interestingly, in the case of M. excelsa, Central and West African specimens had no shared haplotype. This result was unexpected because long gene dispersal distances were found in Cameroonian populations, as expected from the behaviours of the main dispersal animals we identified, the bat Eidolon helvum and the parrots Psittacus erithacus and Agapornis swindernianus. In general, despite the likelihood for long distance seed dispersal, distinct genetic clusters were identified over West and Central Africa and their location may support the forest refugia hypothesis. In addition a phylogeographical signal was detected in chloroplast DNA indicating a prolonged separation between West and Central Africa populations with an accumulation of new mutations. At a smaller spatial scale, patterns of genetic structure also seemed to be weakly influenced by divergent reproduction times among sympatric individuals. The level of genetic diversity in the study populations falls within the range of values calculated for other tropical tree species.
If phylogeography helps to infer population history, it also requires ecological or biological information to confirm some uncertainties. Such detailed data were often lacking to completely interpret numerous of our results. The present study raised new questions and led to new perspectives for further investigation both in phylogeography and botany.