Use of single cross hybrids to measure heterosis for yield in diverse lucerne genotypes growing in a subtropical environment

Abstract

Yield stagnation is a worldwide issue for lucerne breeding, and reasons for the yield plateau include emphasis on disease and pest resistance and not yield per se, and the broad-based synthetic approach to lucerne breeding which is generally used. In this study, an incomplete diallel was made between 50 lucerne clones with representatives from the 3 hypothetical heterotic groups, Medicago sativa subsp. falcata, dormant subsp. sativa, and non-dormant subsp. sativa. Male sterile clones were also included among the dormant group. The single crosses were compared in a subtropical environment at Gatton, Queensland, for yield and other relevant agronomic traits against the adapted synthetics Sequel (dormancy group 9), UQL-1 (group 7), and a highly non-dormant experimental synthetic (line B) derived by introgression of highly non-dormant Arabian germplasm into Sequel. The trial was conducted in a known low-disease-pressure site for Phytophthora root rot, and anthracnose was managed by regular application of prophylactic treatments. The best single cross outyielded Sequel and line B by 13% and 8%, respectively. In this environment, yield was very much influenced by the dormancy group of the test material, with group 9 material significantly outyielding more dormant material. General combining ability (GCA) effects were more important determinants of cumulative yields than specific combining ability (SCA) effects, with these effects being significantly greater than zero for only 4 of the 236 crosses tested over the 15-month period. Similarly, GCA effects were more important for determining autumn height and persistence. The research did identify a small number of clones with good GCA for yield per se, and it would appear that future work should focus on developing more narrow-based synthetics with 4–8 parents which have been selected on the basis of their GCA for yield per se. DNA markers would appear to have a role in selecting clones carrying multiple resistances, and in establishing marker pedigrees for high-yielding parental clones such as we have identified, which can be traced through subsequent generations of recurrent selection in cultivar improvement.