Heterose em sorgo sacarino em diferentes épocas de colheita

Abstract

The sweet sorghum [Sorghum bicolor (L.) Moench] is an alternative feedstock for ethanol production of first generation, due to juicy stalks with directly fermentable sugars, such as sugar cane. The sweet sorghum breeding programs aim to obtain hybrids with high ethanol yield per hectare, which means selecting genotypes that associate favorable phenotypes for the several traits, such as fresh biomass production, juice yield and total soluble solids content. The phenotypic variation in these traits over to the development stage of the crop has influence on the recommendation of cultivars and agro-industrial planning by plants. The aim of this study was to evaluate the potential per se of male sterile and restorer lines of sweet sorghum, as well as to detail the heterosis effect manifested on the average performance, accumulation rate and predictability for the traits related to directly and/or indirectly to the ethanol yield during different times of harvesting. We evaluated 20 genotypes, being four fertility restorer lines (R), three male sterile lines (A), 12 hybrids resulting from the partial diallel cross between A and R, and the commercial hybrid CV 198 in four harvesting time (97, 111, 125 and 139 days after sowing - DAS). The experiments were conducted in the municipalities of Lavras and Sete Lagoas, both in the state of Minas Gerais, following the triple rectangular lattice design 4 x 5. The treatments were arranged in a split plot scheme 20x4, where the genotypes were allocated in plots and harvesting times in subplots. The measured traits were plant height (AP, m), fresh biomass production (PMV, t ha -1 ), juice extraction (EXT%), total soluble solids (TSS, °brix) and tons of brix per hectare (TBH). In addition, they were estimated the parameters b 1 and r 2 by plot, corresponding to the adjustment of the simple linear regression between the traits and times of harvest. The phenotypic data and estimates of b 1 and r 2 were submitted to analysis of variance and diallel analysis. We used the Miranda Filho and Geraldi (1984), adapted of Gardner and Eberhart (1966) diallel model. The experimental precision was assessed by experimental coefficient of variation. There was significant difference among genotypes for all traits. The male-sterile line A1 and restorer lines R1 and R2 had better potential per se when considering the average performance, the accumulation rate and the predictability of behavior of the traits over the harvests. The heterosis effect was significant for all traits. Regarding the accumulation rate also there was heterosis, except for the juice extraction. The hybrids H3, H4 and H6 showed higher performance per se and higher heterosis.