Abstract Quantitative Trait Locus (QTL) allelic varia- tion was studied by analyzing near-isogenic lines (NILs) carrying homologous introgressions on chromosome 4 from three green-fruited wild tomato species. The NILs affect agronomic (yield, brix, fruit weight) and fruit (fruit shape, color, epidermal reticulation) traits in a sim- ilar manner. However, significant differences were de- tected in the magnitudes of the effects, the dominance deviations and epistatic interactions, indicating that those species carry different alleles for the QTL. As the QTL did not show any interaction across environments, gene- tic backgrounds or other QTLs, it can be used to intro- duce novel genetic variation into a broad range of culti- vars. Analysis of new recombinant NILs showed that fruit traits are controlled by several linked genetic loci, whereas multiple genetic loci control the agronomic traits within the original introgression. The hypothesis that QTLs may be composed of multiple linked genes can not be rejected prior to implement projects for QTL isolation and cloning. Loci involved in color enhance- ment could not be related to any known gene involved in the carotenoid biosynthesis pathway, therefore it is hy- pothesized that the function of those loci must be related to the genetic regulation of the carotenoid biosynthetic pathway. Keywords QTL · Fine-mapping · Epistasis · Pleiotropy · Breeding Introduction The use of wild germplasm as a source of novel quantita- tive trait locus (QTL) allelic variation is one of the more important issues for crop improvement and breeders to- day. From a breeder’s viewpoint, it would be most useful to identify new favorable QTL alleles that display low QTL × environment (QTL × E) and QTL × genotype (QTL × G) interactions. Also desirable would be the identification of combinations of QTLs’ alleles at differ- ent loci that interact in an additive or complementary manner. QTL effects and chromosomal position are usually esti- mated in early segregating populations, such as the F 2 or BC 1, where QTLs can be detected using a relative small population size. However, QTL analysis in these popula- tions is problematic for several reasons: (1) estimates of QTL effects and positions generally are biased (Van Ooijen 1992; Jiang and Zeng 1995; Melchinger et al. 1998); (2) QTL-by-environment interactions (QTL × E) are difficult to detect (Stuber et al. 1992; Ragot et al. 1995; Austin and Lee 1998; Melchinger et al. 1998); (3) the distinction be- tween pleiotropy or close linkage when a chromosomal re- gion shows an effect on several traits is not always possible (Lebreton et al. 1998). Additionally, epistatic and QTL-by- genetic background (QTL × G) interactions are difficult to estimate (Yu et al. 1997). Advanced backcross QTL (AB-QTL) analysis has been proposed and tested as a method to mitigate these problems and facilitate utilization of the exotic germ- plasm (Tanksley and Nelson 1996; Tanksley et al. 1996; Fulton et al. 1997; Bernacchi et al. 1998a; Xiao et al. 1998). From an AB population, near-isogenic lines (NILs) carrying small introgressions from the donor par- ent, can be easily isolated by marker-assisted selection (MAS). NILs can then be used to obtain better estimates of the magnitudes of QTL × E, QTL × G and QTL × QTL interactions (Eshed and Zamir 1995, 1996), fine-map QTLs, eliminate undesirable affects caused by linkage drag and, eventually, perform positional cloning of the QTLs (Alpert and Tanksley 1996). Communicated by G. Wenzel A.J. Monforte · S.D. Tanksley ( ✉ ) Department of Plant Breeding and Department of Plant Biology, 252 Emerson Hall, Cornell University, Ithaca, NY 14853-1902, USA e-mail: sdt4@cornell.edu E. Friedman · D. Zamir Department of Agriculture, Vegetables and Genetics, Faculty of Agriculture, Hebrew University of Jerusalem, Kennedy Lee Building, Room 222, Box 12, Rehovot 76-100, Israel Theor Appl Genet (2001) 102:572–590 © Springer-Verlag 2001 ORIGINAL PAPER A.J. Monforte · E. Friedman · D. Zamir S.D. Tanksley Comparison of a set of allelic QTL-NILs for chromosome 4 of tomato: Deductions about natural variation and implications for germplasm utilization Received: 14 April 2000 / Accepted: 12 May 2000