The objective of this study was to determine the number of harvests necessary to predict the performance of tomato (Lycopersicon esculentum) genotypes under greenhouse conditions. The genotypes Empire, Fanny, Floradade and Olympo were transplanted on October 3, 2002 and submitted to 11 harvests. The genotypes Empire, Rodas, Floradade and Olympo were transplanted on February 28, 2003 and submitted to 14 harvests. Harvests were carried through in three intervals of four days. The mass and the number of fruits and the ratio mass/number were evaluated for commercial and discarded fruits. Repeatability estimates were obtained through variance, structural, and principal component analysis. Seven harvests are sufficient to identify superior genotypes for all characteristics, with 90% accuracy of the real value, in both grown conditions.
Lycopersicon esculentum; repeatability; selection; protecting culture
