Scientia Horticulturae 220 (2017) 317–325 Contents lists available at ScienceDirect Scientia Horticulturae journal h om epage: www.elsevier.com/locate/scihorti Exploring the breeding potential of Mexican tomato landraces Esaú de los Ángeles Martínez-Vázquez a , Aurelio Hernández-Bautista a , Ricardo Lobato-Ortiz a,∗ , J. Jesús García-Zavala a , Delfino Reyes-López b a Postgrado de Recursos Genéticos y Productividad-Genética, Colegio de Postgraduados, Campus Montecillo, Km 36.5 Carretera México-Texcoco, 56230, Montecillo, Texcoco City, State of México, Mexico b Facultad de Ingeniería Agrohidráulica, Benemérita Universidad Autónoma de Puebla, Av. Universidad s/n, 73965, San Juan Acateno, Teziutlán, Puebla, Mexico a r t i c l e i n f o Article history: Received 1 December 2016 Received in revised form 8 March 2017 Accepted 21 March 2017 Available online 23 April 2017 Keywords: Solanum lycopersicum Crosses Mexican landraces Hybridization Yield a b s t r a c t General (GCA) and specific (SCA) combining ability are parameters that provide insight into the genetic value of lines utilized in plant breeding programs based on hybridization. The aim of this study was to estimate the GCA for yield of ten S 5 lines derived from Mexican tomato landraces and four testers S 5 derived from commercial varieties, as well as to estimate the SCA of forty hybrids obtained under the line × tester mating design. The traits evaluated were fruit yield and six of its yield components. The lines with the greater GCA effects for yield were LOR82, LOR91 and LOR111, with values of 240, 208, and 99, respectively, being higher than those obtained by most of the testers, which suggested that those native lines have important alleles. In the highest-yielding crosses participated at least one line with high positive GCA effects. In contrast, in crosses with the lowest yield, at least one parent or both exhibited negative GCA. All this results indicated that the best or worst hybrids are not necessary obtained from parents having a high or low GCA, respectively. Valuable germplasm with high yield potential was identified to be used in the generation of hybrid or open-pollinated varieties in tomato breeding programs. © 2017 Elsevier B.V. All rights reserved. 1. Introduction For any breeding program, the selection of parental germplasm is one of the most important decisions to be taken by breeders, and knowing the combining ability of parents improves the efficiency of the breeding program (Gutiérrez et al., 2004). General (GCA) and specific combining ability (SCA) values let us estimate the breeding potential of a set of studied lines under hybrid combinations, since the per se evaluation of the lines is not a reliable parameter. General combining ability is defined as the mean performance of one line under hybrid combinations, whereas specific combining ability are those cases in which certain specific hybrid combinations express good performances in rela- tion to that observed on their parents (Sprague and Tatum, 1942). Based on these definitions, Cockerham (1963) related the GCA with the genetic additive effects and the SCA with the dominance and epistatic effects. SCA and GCA were introduced by Sprague and Tatum (1942), and such concepts were used as a base by Griffing ∗ Corresponding author at: Colegio de Postgraduados, Postgrado de Recursos Genéticos y Productividad-Genética, Km. 36.5 Carretera México, Texcoco, 56230, Texcoco City, State of México, Mexico. E-mail address: rlobato@colpos.mx (R. Lobato-Ortiz). (1956) to develop four diallel designs, which are frequently used for the estimation of GCA and SCA (De la Cruz et al., 2010). To date, there are many genetic designs to estimate combin- ing ability; however, the most common genetic designs are the diallel mating and line × tester designs (Lobato-Ortiz et al., 2010). Davis (1927) and Lobato-Ortiz et al. (2010) indicated that since 1920–1930, the estimation of GCA on maize inbred lines consisted on the evaluation of all possible n(n–1)/2 direct crosses; however, this procedure require too many human and material resources when the number of parental lines is high. In 1932, this situation allowed Davis to develop and introduce the line × tester design (Hallauer et al., 2010). Basically, this mating design is an exten- sion of the topcross design and involves the hybridization of n lines with more than one tester (Singh and Chaudhary, 1985). The tester utilized could be a population with broad or narrow genetic base as well as a low-yielding line (Hallauer and López, 1979). This eval- uation allows selecting the best parents in a series of crosses and identifies specific combinations with above-average performance of the lines, which could help to develop breeding programs based on reciprocal recurrent selection or hybridization (De la Cruz et al., 2010). A recent research has demonstrated that tomato domestication was a two step-process; a first domestication in South America http://dx.doi.org/10.1016/j.scienta.2017.03.031 0304-4238/© 2017 Elsevier B.V. All rights reserved.