Repeatibility coefficients were estimated for dry matter production (DMP), plant height at cutting time (PHC), disease tolerance (DT) and height at flowering (HF). The characteristics were evaluated in 92 genotypes in the years 2004 and 2005. Five cuttings were carried out in the rain period (November to March) and six cuttings in the dry period (April to September). The experiment was arranged in randomized block design, with two replications. The estimates of the repeatability coefficient were obtained by different methods and, then, the study of genotype stabilization was performed. The statistical analyses were done using the GENES software. During the dry period, only two cuttings were enough to achieve 85 % accuracy in the estimation of the real DMP value, whereas in the rain period, five cuttings were required to achieve the same reliability. Overall, in the rain period a larger number of evaluations was required to obtain the same data reliability. For genotype stabilization, there was increase in reliability by leaving out some cuttings. The variables DMP and PHC, during the rain period, showed larger estimates for the repeatability coefficient than the variables DT and HF.
Agronomic traits; genotype's stabilization; Medicago sativa L.; repeatability; rain period; dry period
