Estimates of repeatability coefficients and optimum number of measures for genetic selection of Cynodon spp

The genus Cynodon includes species that have huge potential to produce high-quality fodder, for grazing and hay production. In perennial forage breeding, selection should be performed based on repeated measurements in the same individual over time, which maximizes the selective efficiency. Thus, the objectives of this study were to estimate the coefficients of repeatability and optimal number of measures for the genetic selection of Cynodon spp. clones. We evaluated 202 clones, being five checks, during four harvests, in augmented block design. A mixed model methodology (REML/BLUP) was used to estimate the variance components and predict the genotypic values. The repeatability coefficient for plant height, green weight, dry matter, and plant vigor were equal to 0.50, 0.67, 0.25, and 0.73, respectively. The accuracies obtained by performing m repeated measures revealed that 5, 2, 13, and 2 measurements are required to attain an accuracy ≥ 90% in genetic gain from the selection for plant height, green weight, dry matter, and plant vigor, respectively. Furthermore, the efficiency of performing four measurements compared with that of performing only one measurement was 27%, 15%, 51%, and 12% for plant height, green weight, dry matter, and vigor, respectively. In addition, the additive index, adopting a 10% selection intensity, was used in simultaneous selection, leading to desirable genetic gains for plant height, green weight, and plant vigor. In conclusion, the use of the repeatability coefficient suggests performing three measurements to obtain a high accuracy and good efficiency on the selection of superior genotypes of Cynodon spp.