The objective of this work was to estimate the repeatability of agronomic traits and to determine the appropriate number of cuts for selectig Urochloa ruziziensis. One hundred and eighteen progenies of half-sib U. ruziziensis and four checks, the cultivars Basilisk (U. decumbens), Marandu (U. brizantha), Comum (U. ruziziensis) and one access Urochloa sp. were evaluated. A randomized complete block with two replications and plots in a 3.0-m line with 1.0x0.5-m spacing was used. Seven cuts were made at intervals of 60 and 90 days in wet and dry seasons, respectively. The evaluated characteristics were: plant height, dry matter weight (DM), fresh matter weight (FM) and percentage of dry matter (PDM). The repeatability was estimated by analysis of variance, principal components of the covariance matrix and of the correlation matrix, and structural analysis using the correlation matrix. Repeatability estimates varied from 0.31-0.38 for plant height, 0.31-0.43 for FM, 0.16-0.50 for PDM and 0.23-0.43 for DM, with coefficients of determination between 57-87%. The number of cuts necessary to assess the actual value of the characteristics ranged between 7-8 cuts for FM and plant height, and 10-14 for DM and PDM, with coefficients of determination greater than 80%.
Brachiaria ruziziensis; Urochloa spp.; efficiency of selection; forage breeding; forage crops
