-

The diallel cross constitutes an informative genetic design for choosing genitors and crosses in breeding programmes since it provides estimates of general (GCA) and specific (SCA) combining abilities. Because the breeding programme for Urochloa humidicola ([Rendle] Morrone & Zuloaga; syn. Brachiaria humidicola [Rendle] Schweick) at Embrapa Beef Cattle is recent, these estimates are still unknown. Thus, the aim of this study was to obtain the estimates of GCA and SCA from the partial diallel cross design between sexual and apomictic parents. The crosses involved nine sexual and ten apomictic parents. Seventy‐one full‐sib progenies were obtained and evaluated in an incomplete block design, using all parents and the cv. BRS Tupi as a check. The agronomic and forage quality traits were evaluated using seven harvests. Statistical analysis was performed using the mixed model approach. Significant variations were associated with GCA among sexual parents for biomass yield and forage quality traits. However, this cannot be observed for apomictic parents for GCA. The SCA effect was not significant for the assessed traits. These results showed the predominance of additive effects. Sexual parents SEX3 and SEX4 presented the highest potential for contributing favourable alleles in hybrid combinations. The absence of GCA among apomictic parents and SCA suggests the need to increase genetic variability for the generation of breeding populations of U. humidicola.

No abstract

ABSTRACT. Sweet sorghum has considerable potential for ethanol and energy production. The crop is adaptable and can be grown under a wide range of cultivation conditions in marginal areas; however, studies of phenotypic stability are lacking under tropical conditions. Various methods can be used to assess the stability of the crop. Some of these methods generate the same basic information, whereas others provide additional information on genotype x environment (G x E) interactions and/or a description of the genotypes and environments. In this study, we evaluated the complementarity of two methods, GGEBiplot and Toler, with the aim of achieving more detailed information on G x E interactions and their implications for selection of sweet sorghum genotypes. We used data from 25 sorghum genotypes grown in different environments and evaluated the following traits: flowering (FLOW), green mass yield (GMY), total soluble solids (TSS), and tons of Brix per hectare (TBH). Significant G x E interactions were found for all traits. The most stable genotypes identified with the GGEBiplot method were CMSXS643 for FLOW, CMSXS644 and CMSXS647 for GMY, CMSXS646 and CMSXS637 for TSS, and BRS511 and CMSXSS647 for TBH. Especially for TBH, the genotype BRS511 was classified as doubly desirable by the Toler method; however, unlike the result of the GGEBiplot method, the genotype CMSXS647 was also found to be doubly undesirable. The two analytical methods were complementary and enabled a more reliable identification of adapted and stable genotypes.

No abstract