RECURRENT SELECTION FOR FORAGE AND SEED YIELD IN ZIGZAG CLOVER

Phenotypic recurrent selection on an annual basis was conducted with zigzag clover (Trifolium medium L.) to determine the efficiency of selection for improved seed and forage yield and the relationships among these characters. Seedlings were exposed to winter temperatures in a cold frame before they were transplanted in a field in the spring. The undesirable types were mowed prior to flowering and the selected plants were allowed to cross. Heads and seeds were harvested from each plant, and after seed threshing and counting, plants were further selected on the basis of numbers of seeds and heads. The effectiveness of four cycles of selection was evaluated over a 2-yr period in an experiment initiated with remnant seed and with management similar to that imposed during the selection process. Recurrent selection was effective in increasing vigor, seeds per head, heads per plant, and seeds per plant. A significant linear increase over cycles was found for each character. Vigor was not correlated with seeds per head, but was correlated with the other characters. The more vigorous plants also exhibited less stand reduction. Data from the two evaluation years were correlated suggesting that materials selected on an annual basis performed similarly to those grown in a 2-yr stand. Although recurrent selection was effective for improving all traits, the rate of change, particularly for seeds per head, was limited. Many more cycles of such selection would be required to change zigzag clover into a useful forage species. It was suggested that the phenotypic recurrent selection program should be continued on an annual basis, but modified if possible to select against the vernalization requirement for flowering which may be associated with poor seedling growth and low aftermath vigor.Key words: Trifolium medium, recurrent selection, vigor, persistance, heads and seeds per plant