ESTIMATED AND OBSERVED GENETIC ADVANCES OF THREE SELECTION METHODS IN HUSK TOMATO (Physalis ixocarpa Brot.)

Abstract

Mass selection in husk tomato (Physalis ixocarpa Brot.) has been efficient to increase yield, but it is becoming harder to recognize the best plants in a selection plot; it is then convenient to prove another selection methods to choose the most efficient and practical one. In this work it was estimated the expected genetic advances (EGA) accomplished by mass selection (MS), half-sib families selection (HSFS) and combined half-sib selection (CHSS) in the CHF1-Chapingo variety, and to compare them with their respective observed genetic advances (OGA). The EGA was estimated based on a random sample of 200 maternal half-sib families evaluated in three environments during the Spring-Summer cycle of 1995 and 1996. The first selection cycle of each method, the CHF1-Chapingo variety, and six more populations, were evaluated in three environments during the SpringSummer cycle of 1996. The results showed that the first harvest is more important than the second one in relation to the genetic advance. Based on the EGA, the HSFS method was the most efficient and the MS was the least efficient, being HSS and CHSS better than MS in 34 and 18 %, respectively, in the first harvest yield. MS and HSFS were equally efficients in relation to OGA for increasing total yield (2.8 and 2.5 %, respectively), while CHSS was inefficient. There was a high (> 90 %) concordance between EGA and OGA in MS, but it was low (> 30 %) in CHSS.