The estimate of the optimal number of plants per plot for the evaluation of carrot populations is important in order to reduce the experimental error and to enhance the accuracy of the information obtained in breeding experiments. Additionally, it contributes to reduce the research costs. The objective of the present work was to estimate the minimum number of plants per plot for evaluation of root characters in carrot populations using the modified maximum curvature and the stabilization of the average and genetic variance methods. The essays were conducted in the summer of 2004 at three representative areas for carrot production in Brazil, respectively: Brasília, DF; São Gotardo, MG and Lapão, BA, in randomized blocks design with four replications, and plots with 2 m². Harvesting took place 90 days after sowing. In each plot of nine populations, 15 competitive plants were harvest and roots evaluated for mass, length, green shoulder length, tip shape, shoulder shape, diameter, xylem diameter, relationship among xylem diameter / root diameter and the L* a* b* color caracters for the xylem and phloem. The procedures of re-sampling and the modified maximum curvature method were used to estimate the minimum plant number to represent each plot. The results obtained for each procedure indicated that a sample of 13 plants per plot is enough to guarantee an appropriate evaluation of carrot populations for the group of studied characters.
Daucus carota L.; modified maximum curvature; stabilization of the average and genetic variance; sampling
