Abstract
Description. Coffee cultivation plays an important economic and social role in Côte d'Ivoire. Among the dozens of Coffea species, only Coffea canephora, also known as robusta, is grown in the country. The genetic improvement of this species has been the subject of a selection program set up by the National Center for Agronomic Research (CNRA). This program involves two basic genetic groups, Guinean and Congolese. These groups have been determined using enzymatic markers which are known to be less informative. So little is known about the reliability of the genetic structuring of the parental populations. Objectives. The objective of this work is to refine the genetic structuring of the parental populations used to set up the C. canephora breeding program in Côte d'Ivoire using SNP markers and to relate this structure to test values in the genotypes concerned. Method. Sixty-six individuals representing parental genotypes of the two populations were analyzed using 200 SNPs markers. The parameters of genetic diversity were calculated and then the genetic structure was determined. The yield over four years of production, susceptibility to rust, and bean size have been evaluated. Results of these agronomic traits have been included in this analysis. The Guinean genotypes were crossed with a Congolese tester named 464 and the Congolese individuals were crossed with a Guinean tester 410, for assessing their combining ability. Results. The analysis of molecular variance (AMOVA) revealed that 75% of the total variation was due to variation between populations. Bayesian classification and Principal Coordinates Analysis (PCoA) structure the 66 genotypes into two main groups of Guineans and Congoleses. While the Guinean group appears to be homogeneous, the Congolese group is split into four subgroups of SG1, SG2, B and C. Maximum test values are found for the SG1 and SG2 subgroups. The Guinean group showed lower performance. It is important in this group to select the best progenitors. Conclusions. SNP markers were effective in genetic structuring of the population. This study also allows the identification of a genotype that was not determined. The use of this set of markers will be useful for the control of future progenitors. The same genetic progress can be achieved by crossing SG1 or SG2 with a good Guinean male. It seems important to make a selection among the Guineans before crossing them with a Congolese tester.