Cocoa pods, Venezuela. Photo: C.Lanaud ©CIRAD
Arabica coffee, Ethiopia. Photo: ©Jean-Pierre Labouisse
Yams in Benin. Photo: J-L Pham ©IRD
Rice harvest, Guinea. Photo: J-L Pham ©IRD
Maize corn. Photo: ©Brigitte Gouesnard

Rationale

Domestication strongly impacted phenotypic and genomic evolution in crop species. Crop species typically exhibit lower genetic diversity than their wild ancestors, and may show dramatic phenotypic changes in their morphology, phenology and metabolism. Understanding the domestication process is thus a key to crop breeding but also a unique opportunity to study rapid evolutionary processes on a short time scale. Differences among genomic patterns are still not adequately explained. For instance, in many crops, how many genes - and which - are involved in domestication and artificial selection is still not clear. Comparing the domestication process in a range of species, varying from ancient domesticated species (Vitis, Sorghum) to more recently cultivated species (Coffea) should provide key information on the dynamics of adaptation and the correlated evolution of polymorphism patterns.

It is also especially important to compare molecular evolutionary patterns among species with contrasted life-history or ecological traits. Life-history or ecological traits may influence genome evolution through their effect on key population genetic parameters (effective size, recombination rates, and mutation rates). Genomic patterns may also vary among phylogenetically distant species because of specific molecular mechanisms such as recombination and repair mechanisms.

Knowing the molecular functions that are targeted by selection is also of interest to increase our understanding of adaptation. Thus, studying the evolution of gene families and relating it to expression data across lineages may help to identify which molecular functions play a key role in adaptation.

Today, a comparative population genomic approach among many species is both indispensable and possible thanks to massive sequencing technologies.