Controle genético de caracteres associados ao enchimento de grãos em feijão

Abstract

One of bean breeders’s challenge is to obtain a line with the dry grain period possible associated with a high daily dry matter rate (RATE), without changing the plant cycle. the main alternative is to obtain lines with the lowest number of days to flowering (NDF). The objective of this research was to provide information of the dry grain period and the dry matter accumulation rate genetic control; to verify if it is possible to select lines with early flowering, but with a longer period of effective grain filling, and show if it is feasible to use these traits as a strategy for an indirect selection aiming grain yield. For this, a diallel of ten parents involving early and late flowering lines, differing in grain size and belonging to two gene pools, Andean and Mesoamerican, was conducted at the Federal University of Lavras. Of the 45 possible hybrid combinations, 39 hybrid combinations were obtained. These hybrids with the parents were evaluated from the F2 generation, in the randomized block design in three harvests. The first sowing was in November 2014 with three replicates. The second one in February 2015 and the third one in July 2015, the latter two with four replicates. The evaluated characters were three: the date of beginning of flowering; Dry matter of the grains (MSG) and Grain yield (YIELD). MSG was obtained seven days after the date of flowering and subsequently every five days until harvest. From the MSG data, a quadratic regression equation was fitted between the independent variable X, number of days of sampling and the dependent variable Y, dry matter per grain. With this equation, the maximum point was estimated. From this estimate the number of days for physiological maturity (NDFM) was obtained. To estimate the dry matter accumulation rate, a linear equation was obtained between the independent variable X, date until the physiological maturity and the dependent variable Y, dry matter. From the NDF, NDFM, RATE and YIELD estimates, analyzes of variance and diallel analyzes according to Griffing's model II (1956) were performed. It was verified that: for all evaluated characters variability was found; for NDF, NDFM and RATE the overall combining ability explained most of the variation. In grain yield, both CGC and CEC were important. For this character, the heterosis was high, but in the general average it is negative, since many hybrid combinations involved parents of Andino x Mesoamerican gene pool; None of the segregating populations evaluated showed that it is feasible to obtain new lines that are early flowering, with a longer period of accumulation of dry matter in the grains and high RATE.