Along with the increasing competition in its primary production area the Brazilian dairy searchs for production systems that result higher productivity and lower costs. Among the adopted strategies are the herd and forage improvement. Alfalfa, known for its excellent nutritional quality and palatability, becomes an option. However, to obtain cultivars well adapted to tropical conditions still requires a lot of time and work. Theoretical information regarding forage value characters are required in order to orient improvement programs. So, the objective of this work was to estimate the repeatability coefficient of alfalfa dry matter productivity in Brazil southeast region and to determinate the ideal evaluation number necessary to select higher genotypes. Four methodologies were used. Dry matter productivity data from four alfalfa evaluation experiments carried in different areas of Minas Gerais State were used. The repeatability coefficient (r) for each experiment was estimated using variance analysis (Anova) of the main components based on the covariance (CP-1) and the correlation (CP-2) matrix and by the structural analysis based on the correlation matrix. A good response among the four methodologies was observed in all experiments. The average estimate for repeatability of alfalfa dry matter production for all experiments was 0,59, with the determination coefficient average of 0,96. What was also noted was that four cuts, in average, were sufficient to shown with 85% confidence the real genotipic value of the evaluated cultivars.
Medicago sativa L.; yield; breeding; forage
