The genetic progress estimative is an important aspect in breeding programs, in order to evaluate the breeding selection methods efficiency for the characteristics of interest in the populations under selection. In this research, the gains obtained from the selection of agronomic traits were evaluated in a carrot population through successive generations. The experiment was performed at Embrapa Hortaliças, Brasília, Brazil. A carrot population was selected every summer during six consecutive generations from 1999 to 2004. In the summer of 2005, samples of seeds harvested in the previous years were sown in plots of 1 m2 in randomized blocks design with five replications. The plants were evaluated for severity of leaf blight at 60 and 90 days after sowing and harvested at 90 days after sowing when each plot was evaluated for the characters of number and mass of waste roots, number and mass of marketable roots with green shoulder, number and mass of marketable roots without green shoulder, number, mass and average mass of total roots and number, mass and average mass of marketable roots. The data were submitted to analyses of variance, and comparisons of treatment means by Scott-Knott, and the gain obtained with selection was calculated. Superior gains occurred only in the first cycles of selection towards the decrease of refused roots and increase of resistance to leaf blight. The selection for leaf blight resistance was efficient only when performed at 90 days after sowing, because the disease severity was low at 60 days. The amount of roots classified as marketable with and without green shoulder increased during all the selection cycles, although the gains were higher for the roots without green shoulder.
Daucus carota L.; recurrent selection; genetic progress
