Quantitative Trait Loci and Breeding


Plant breeding methodologies have changed greatly over the past century as new techniques and tools become available to breeders. The introduction of computer technologies, new statistical methods and molecular markers have greatly enhanced the rates of genetic gain that can be achieved in breeding programs. Many important traits for crop improvement, such as yield, quality and quantitative or durable disease resistances, are regarded as quantitative traits and are controlled by multiple quantitative trait loci (QTL). Recent advances in molecular marker systems have created new opportunities and strategies to select for quantitative traits. Strategies for deploying QTL in breeding programs vary from monitoring specific loci through to the deployment of molecular markers flanking the target QTL to the use of whole‐genome marker scans to identify individual plants that will offer the greatest opportunity for genetic gain.

Key Concepts:

  • Yield, many aspects of quality and durable disease resistance are all examples of quantitative traits.

  • Interactions between the genotype of the plants and the growing environment can be critical in determining the expression of quantitative traits.

  • Breeding programs usually select for quantitative traits at multiple stages in the breeding process and using replicated trials across multiple locations.

  • The regions of the genome controlling quantitative traits, QTL, can be mapped by measuring the association between molecular markers and the phenotype across multiple environments in populations segregating for the trait.

  • The stability of QTL across environments and the proportion of the variation accounted for by the QTL are key criteria in determining the value of the QTL to a breeding program.

  • Genomic selection, based on whole‐genome scan with markers, offers an opportunity to select for large numbers of minor QTL without prior knowledge of the location of the QTL.

  • Effective delivery of molecular breeding technologies to breeding programs, particularly in poor countries, is dependent on providing access to marker screening platforms resources, and training.

Keywords: quantitative trait; breeding; marker‐assisted selection; genomics selection; yield; genotype; phenotype

Figure 1.

Example of QTL mapping of the average yield at 12 field trials across Australia (purple), 10 field trials in South Australia (orange) and 6 field trials under drought conditions in South Australia using simple interval mapping (a). The confidence interval of the QTL controlling the average yield in 6 dry environments corresponds to the maximum LOD minus 1, which is 23.1 cM. The LOD score for the average yield across 12 field trials increased from 2 to 3.9 by using composite interval mapping model (b). The threshold of significant LOD score was calculated by performing 1000 permutations of the data (c) and showed that the QTL identified in (b) was significant at p<0.05.

Figure 2.

Example of MABC process for a single locus. Plants are screened for the linked markers at each generation to increase the chances of recombination close to the target region, using 2–4 well spread polymorphic markers per chromosome for background selection and 2–3 flanking markers on each side of the locus to introgress. For a QTL, the whole confidence interval needs to be spanned (4–5 markers). (a) In the BC1 generation, the focus is on finding the closest possible recombination events on one side of the target trait (besides ensuring that the proper alleles on the other side are still present). Enough plants are selected at this stage to still allow for background selection. (b) In the BC2 generation, the same takes place for the other side of the target trait. (c) Selfing will then be needed to fix the introgressed region. That will be done at the end of the background selection process, which may take an additional generation.

Figure 3.

Example of MARS process for F3‐derived populations. Progenies are advanced to the F3 generation through single‐seed descent and single F3 plants are selfed to generate F3:4 or F3:5 progenies. DNA samples are obtained directly from the F3 plants, or from bulked F4 progenies from each F3. These samples are genotyped at the polymorphic loci identifies from the parental screening. Once a set of key QTLs has been identified, a few sets of F3‐derived progenies are chosen based on their complementarity for the presence of favourable alleles and on their overall phenotypic performance. Several individual plants (F4 or F5) of each progeny are grown and genotyped (nearest marker to the QTL peak, or flanking markers) to identify the best individual plants to use in the recombination crosses. An example would be to cross four pairs of progenies (8 lines), then the two pairs of resulting F1s in the second cycle, and then the final two F1s in the final cycle. At each stage, the F1s are genotyped and the best ones are used again for the next cycle of recombination. At the end of the process, the resulting lines are selfed a few times for fixation.

Figure 4.

Genomic selection. Modified from Heffner et al. . There are two key components to a genomic selection strategy; the active breeding program and the training population. The training population is both genotyped and phenotyped and forms the base for developing predictive models and genomic‐estimated breeding values (GEBV). The training population must be based on germplasm relevant to the breeding program and should be regularly updated with new lines that come through the breeding process. Within the breeding program, lines are genotyped only and the genotypic data is combined with the predictive models (from the training population) to derive GEBVs. Lines from within the breeding program that show the highest GEBVs are selected for subsequent breeding cycles and also to feed into the training populations for further development of the predictive models.



Bernardo R and Yu JM (2007) Prospects for genomewide selection for quantitative traits in maize. Crop Science 47: 1082–1090.

Bouchez A, Hospital F, Causse M, Gallais A and Charcosset A (2002) Marker‐assisted introgression of favorable alleles at quantitative trait loci between maize elite lines. Genetics 162: 1945–1959.

Burnett RJ and Larkins BA (1999) Opaque2 modifiers alter transcription of the 27‐kDa gamma‐zein genes in maize. Molecular General Genetics 261: 908–916.

Cavanagh C, Morell M, Mackay I and Powell W (2008) From mutations to MAGIC: resources for gene discovery, validation and delivery in crop plants. Current Opinion in Plant Biology 11: 215–221.

Charmet G, Robert N, Perretant MR et al. (1999) Marker‐assisted recurrent selection for cumulating additive and interactive QTLs in recombinant inbred lines. Theoretical and Applied Genetics 99: 1143–1148.

Charmet G, Robert N, Perretant MR et al. (2001) Marker assisted recurrent selection for cumulating QTLs for bread‐making related traits. Euphytica 119: 89–93.

Heffner EL, Sorrells ME and Jannink JL (2009) Genomic selection for crop improvement. Crop Science 49: 1–12.

Hospital F (2001) The reduction of linkage drag around introgressions in marker‐assisted backcross programs. Quantitative Genetics and Breeding Methods: The Way Ahead 96: 199–204.

Kuchel H, Fox R, Reinheimer J et al. (2007) The successful application of a marker‐assisted wheat breeding strategy. Molecular Breeding 20: 295–308.

Lander ES and Botstein D (1989) Mapping Mendelian factors underlying quantitative traits using RFLP linkage maps. Genetics 121: 185–199.

Magalhaes JV, Liu J, Guimaraes CT et al. (2007) A gene in the multidrug and toxic compound extrusion (MATE) family confers aluminum tolerance in sorghum. Nature Genetics 39: 1156–1161.

Mayor PJ and Bernardo R (2009) Genomewide selection and marker‐assisted recurrent selection in doubled haploid versus F‐2 populations. Crop Science 49: 1719–1725.

Meuwissen TH, Hayes BJ and Goddard ME (2001) Prediction of total genetic value using genome‐wide dense marker maps. Genetics 157: 1819–1829.

Nadolska‐Orczyk A, Gasparis S and Orczyk W (2009) The determinants of grain texture in cereals. Journal of Applied Genetics 50: 185–197.

Taylor J and Verbyla A (2011) R package wgaim: QTL analysis in bi‐parental populations using linear mixed models. Journal of Statistical Software 40: 1–18.

Thomson MJ, de Ocampo M, Egdane J et al. (2010) Characterizing the Saltol quantitative trait locus for salinity tolerance in rice. Rice 3: 148–160.

Wang C, Rutledge J and Gianola D (1994) Bayesian analysis of mixed linear models via Gibbs sampling with an application to litter size in Iberian pigs. Genetic Selection Evolution 26: 91–115.

Yu J, Holland JB, McMullen MD and Buckler ES (2008) Genetic design and statistical power of nested association mapping in maize. Genetics 178: 539–551.

Further Reading

Delannay X, McLaren CG and Ribaut J‐M (2011) Fostering molecular breeding in developing countries. Molecular Breeding 29: 857–873. doi: 10.1007/s11032‐011‐9611‐9.

Hospital F and Charcosset A (1997) Marker‐assisted introgression of quantitative trait loci. Genetics 147: 1469–1485.

Langridge P and Fleury D (2011) Making the most of ‘omics’ for crop breeding. Trends in Biotechnology 29: 33–40.

Liu B (1998) Statistical Genomics. CRC Press: New York.

Lynch M and Walsh B (eds) (1998) Genetics and Analysis of Quantitative Traits. Sunderland: Sinauer Associates Inc.

Tester M and Langridge P (2010) Breeding technologies to increase crop production in a changing world. Science 327: 818–822.

Varshney RK, Glaszmann JC, Leung H and Ribaut J‐M (2010) More genomic resources for less studied crops. Trends in Biotechnology 28: 452–460.

Contact Editor close
Submit a note to the editor about this article by filling in the form below.

* Required Field

How to Cite close
Fleury, Delphine, Delannay, Xavier, and Langridge, Peter(May 2012) Quantitative Trait Loci and Breeding. In: eLS. John Wiley & Sons Ltd, Chichester. http://www.els.net [doi: 10.1002/9780470015902.a0023712]