The Breeding Behavior of Yield and Related Variables in Alfalfa. III. General and Specific Combining Ability 1

T basic problems in alfalfa breeding were reviewed in 1942 by Tysdal et al. (15). These investigators noted that the highest yielding of 28 single crosses exceeded the average yield of the check varieties by 39%. The average of the top ten hybrids was 15% above the yield of the commercial variety checks. Furthermore, when the hybrids were compared in the field, certain ones showed much more hybrid vigor than others, indicating good combining ability for some lines and relatively poor combining ability for others. Bo]ton (1) also noted differences in combining ability for forage yield in both inbred and open-pollinated groups in alfalfa. Open-pollinated sources of material appeared to be as productive of good hybrids as were inbred lines. He suggested that open-pollinated sources would be preferred over the inbred sources, because of their greater ability to produce seed, an important attribute of parental clones. Tysdal and Kiesselbach (14) and Tysdal and Crandall (13) showed that alfalfa synthetics, based on general combining ability measured from polycross progeny testing, were highly productive. Wilsie and Skory (17) suggested that for maximum improvement specific combining ability should be determined also. A positive correlation between general and specific combining ability was shown (17). The yield of polycross progenies of parental clones of alfalfa, all high in general combining ability, showed little relationship with average synthetic-1 through synthetic-4 yields of 18 derived synthetics (11). Davis (3), however, tested three 4-clone alfalfa combinations and concluded that yield of such synthetics could be predicted from polycross progeny tests. Wilsie (16) reported an F, single cross that yielded 81% above the higher-yielding inbred parent. Davis and Panton (4) noted that the yield of the best of six single crosses was 36% higher than the expected yield of the synthetic of the four parental clones. Reports such as these emphasize the need for breeding methods that will make more effective use of hybrid vigor in the development of new alfalfa varieties. Morley et al. (12) noted that alfalfa strains differed combining ability for winter and summer growth rates. Differences between strains for combining ability were usually more evident in winter than in summer growth. However, a Canadian creeping-rooted strain had a high combining ability for summer growth but was poor in combining ability for winter growth. Heinrichs (9) has developed strongly creeping-rooted lines of alfalfa in a breeding program based on the combining ability for the creepingrooted character. In respect to estimates of gene action, Kohel a found that