Résumé
Small citrus is an increasing component of the world citrus industry. New high quality, parthenocarpic, sterile mandarin varieties will play a pivotal role for its sustainable development. To produce seedless varieties, the CIRAD breeding program is focused on triploid hybrid selection. The first method to develop triploid progenies exploits 2n gametes that are naturally produced by diploid cultivars to obtain triploids in 2x X 2x crosses. The second method is based on interploid crossings (2x X 4x and 4x X 2x). For the last strategy the parental tetraploid gene pool has been diversified by selecting spontaneous tetraploids in apomictic cultivars, generating doubled diploids by colchicine treatments and production of allotetraploids by somatic hybridization. Several thousand triploid hybrids have been created by CIRAD using these strategies with the support of embryo rescue and ploidy evaluation by flow cytometry. This breeding program is supported by basic and methodological research performed in collaboration with IVIA (Spain), DAK (Morocco) and INRA (France) in the following topics: (1) citrus germplasm management and characterization, (2) studies of polyploidization mechanisms, tetraploid meiosis and its implications on the genetic and phenotypic structure of triploid progenies, and (3) studies of the implication of polyploidy on genomic and phenotypic expression. (Texte intégral)