Résumé
Maize streak virus and maize mosaic virus diseases can both cause significant grain yield reduction in maize. In order to have a better knowledge of the genetic control of the resistance to each disease and thus to improve the selection schemes in current use, an RFLP mapping study was performed. D211 and L61, two completely resistant lines to MSV from La Réunion island, were each crossed to B73, completely susceptible, to form two F2 segregating populations and their derived F2:4 families (163 and 190 families respectively). These families were tested for MSV resistance under artificial infestation in two different environments: Harare, Zimbabwe, and Saint Pierre de la Réunion. MMV resistance was mapped in an F2 population derived from the cross between Hi40, resistant, and B73, susceptible. 105 F2:4 families were scored under insect proof tunnels after artificial infestation in St. Pierre. In all the trials, 6 disease assessments were made on 15 to 21 plants per plot, two replications. Between 100 and 150 RFLP markers were mapped depending on the population. Composite interval mapping was used on the AUDPC (Area Under the Disease Progress Curve) variable in Harare for MSV resistance and in la Réunion for MMV resistance. Resistance to MSV was controlled by one major QTL (or cluster of QTL) on chromosome 1 (QTLS1), explaining 69% of the variation, and at least two minor ones on chromosome 10. QTLS1 had a predominantly additive effect. Analyses of other available data will allow us to test the stability of these QTL across environments and germplasm. Two QTL for MMV resistance were detected, one on chromosome 1 (QTLM1), and the other on chromosome 3 (QTLM3) which may be a cluster of 2 or 3 loci. They explained 6% and 68% of the phenotypic variance, respectively. The susceptible parent contributed the resistance allele at QTLM1. The genotypic effect at QTLM3 was predominantly additive, which should simplify selection for this genomic region. (Texte intégral)