Search for a command to run...
As part of ongoing studies regarding the genetic basis of quantitative variation in phenotype, we have determined the chromosomal locations of quantitative trait loci (QTLs) affecting fruit size, soluble solids concentration, and pH, in a cross between the domestic tomato (Lycopersicon esculentum Mill.) and a closely-related wild species, L. cheesmanii. Using a RFLP map of the tomato genome, we compared the inheritance patterns of polymorphisms in 350 F2 individuals with phenotypes scored in three different ways: (1) from the F2 progeny themselves, grown near Davis, California; (2) from F3 families obtained by selfing each F2 individual, grown near Gilroy, California (F3-CA); and (3) from equivalent F3 families grown near Rehovot, Israel (F3-IS). Maximum likelihood methods were used to estimate the approximate chromosomal locations, phenotypic effects (both additive effects and dominance deviations), and gene action of QTLs underlying phenotypic variation in each of these three environments. A total of 29 putative QTLs were detected in the three environments. These QTLs were distributed over 11 of the 12 chromosomes, accounted for 4.7-42.0% of the phenotypic variance in a trait, and showed different types of gene action. Among these 29 QTLs, 4 were detected in all three environments, 10 in two environments, and 15 in only a single environment. The two California environments were most similar, sharing 11/25 (44%) QTLs, while the Israel environment was quite different, sharing 7/20 (35%) and 5/26 (19%) QTLs with the respective California environments. One major goal of QTL mapping is to predict, with maximum accuracy, which individuals will produce progeny showing particular phenotypes. Traditionally, the phenotype of an individual alone has been used to predict the phenotype of its progeny. Our results suggested that, for a trait with low heritability (soluble solids), the phenotype of F3 progeny could be predicted more accurately from the genotype of the F2 parent at QTLs than from the phenotype of the F2 individual. For a trait with intermediate heritability (fruit pH), QTL genotype and observed phenotype were about equally effective at predicting progeny phenotype. For a trait with high heritability (mass per fruit), knowing the QTL genotype of an individual added little if any predictive value, to simply knowing the phenotype. The QTLs mapped in the L. esculentum X L. cheesmanii F2 appear to be at similar locations to many of those mapped in a previous cross with a different wild tomato (L. chmielewskii).(ABSTRACT TRUNCATED AT 400 WORDS)