Defining heterotic groups in alfalfa: A molecular marker perspective

E. Charles Brummer
Iowa State University, Ames, IA 50011


Capturing heterosis remains a desirable but elusive goal for alfalfa (Medicago sativa L.) breeders. Virtually all commercially marketed alfalfa cultivars are synthetics, developed by successively intermating selected plants and increasing seed through 2-3 generations. The inbreeding present in such a system generally counterbalances any heterosis that may have been expressed in the initial progeny population. Despite repeated efforts along several lines, hybrid cultivars cannot be produced feasibly with current technology. However, the potential for capturing a partial heterotic yield gain has been generally overlooked: some population hybrids have been shown to express high levels of heterosis and could be developed using current seed production processes (Busbice and Rawlings, 1974, Sriwatanapongse and Wilsie, 1968). Population hybrids are actually "semi-hybrids;" half of the progeny will be hybrid (i.e. between populations); the other half will be crosses within each of the parental populations. Due to competition-induced seedling death during establishment, semi-hybrids may yield as much as pure hybrids. In order to express heterosis, two factors are required: dominance at loci controlling the trait of interest, and differences in allele frequencies between the parents at those loci. Most data in maize (Hallauer et al., 1988) and alfalfa (Bingham et al., 1994) show partial to complete dominance to be the genetic basis for yield, and population allele frequency differences have been reported (e.g. Brummer et al., 1991; Kidwell et al., 1994) indicating that alfalfa breeders should be able to realize heterosis. The key to making successful population crosses is to identify or develop heterotic groups and to keep them distinct through breeding, only bringing them together for testing and seed production. The resulting hybrid populations should not be recycled for further breeding to avoid mixing alleles (or linkage blocks) and breaking apart beneficial complementarities between populations. Molecular markers offer several opportunities to capitalize on heterosis: assigning populations to heterotic groups, identifying loci important for the expression of heterosis, selecting parents for multiple or complementary alleles, and identifying particularly useful alleles. Simulations show striking differences in the proportion of multi allelic individuals (i.e. having multiple linkage blocks) resulting from different population matings. Discussions focusing on the inbreeding coefficient, F, cannot easily compare the number of individuals containing multiple complementary linkage blocks (e.g. two-allele and four allele reference populations would each have F=0, even though the former has no individuals with three or four different linkage blocks at a particular locus). Best case crosses are those in which the populations contain contrasting sets of alleles; marker-assisted selection greatly increases the number of individuals with multiple linkage groups. Finally, a method to develop pure population hybrids will be discussed.

References

Bingham, E.T. et al. 1994. Crop Sci. 34:823-829.
Brummer, E.C., et al. 1991. Theor. Appl. Gen. 83:89-96.
Busbice, T.H. and J.O. Rawlings. 1974. Euphytica 23:86-94.
Hallauer, A.R. et al. 1988. In: Corn and Corn Improvement, ASA, Madison, WI.
Kidwell, K.K., et al. 1994. Crop Sci. 34:230-236.
Sriwatanapongse, S. and C.P. Wilsie. 1968. Crop Sci. 8:465-466.

Previous Page