The breeding system of three Paspalum species with outstanding forage potential

Abstract

Knowledge of the breeding system is a fundamental step in any program dealing with genetic improvement of wild grasses for forage purposes. This information is essential for Paspalum, a large genus with a variety of reproductive methods including apomixis. We conducted cytoembryological studies in 3 species of the group Plicatula (Paspalum limbatum, Paspalum lenticulare and Paspalum guenoarum) in order to understand the breeding systems of 9 accessions which are currently under selection and being bred for forage purposes in our institution. Both accessions of P. limbatum were diploid (2n = 2x = 20), showed bivalent chromosome pairing at meiosis and were allogamous due to self-incompatibility. Five accessions of P. lenticulare and 2 of P. guenoarum were tetraploid (2n = 4x = 40) with mainly bivalent and quadrivalent chromosome pairing at meiosis. The high frequency of quadrivalents suggests autoploidy as the most probable origin for these tetraploids. All these tetraploid accessions showed complete degeneration of the megaspores. The embryo sacs, one to several per ovule, always developed from somatic nucellar cells. These nonreduced embryo sacs produced embryos through parthenogenesis, and were hence genetically identical with the mother plant. Notwithstanding, pollination was required for the development of the endosperm. Only one accession of P. lenticulare has some reduced potential for sexual reproduction. Our results indicate that P. limbatum is a sexual allogamous diploid species while P. lenticulare and P. guenoarum reproduce by means of aposporous obligate apomixis.