marker-assisted selection in wheat

Chapter 19 – Technical, economic and policy considerations on marker-assisted selection in crops 387Xanthomonas oryzae and is one of the mostimportant diseases of rice. Several genesthat confer resistance to bacterial blighthave been tagged with molecular markers.Huang et al. (1997) and Hittalmani et al.(2000) developed strategies for combiningfour resistance genes, namely Xa-4, Xa-5,Xa-13 and Xa-21, in a single cultivar usingpairwise combinations of the genes. Dueto the co-dominant nature of the markersused, the authors were able to select fromF 2 generations without having to performprogeny testing. The derived lines containingpyramided genes showed higherlevel of resistance and/or a wide spectrumof resistance compared with the parentalmaterial. Another gene pyramiding exampleusing MAS involves stacking of the resistancegenes rym4, rym5, rym9 and rym11for the barley yellow mosaic virus complexusing molecular markers and doubledhaploids (Werner, Friedt and Ordon, 2005).Other examples include pyramiding forbarley stripe rust resistance (Castro et al.,2003), and powdery mildew resistance inwheat (Liu et al., 2000) and, in MAS applicationsat CIMMYT, crosses have beenmade to combine two genes for cereal cystnematode resistance and three differentgenes for stem rust resistance (Sr24, Sr26and Sr25) in targeted wheat germplasm.Manipulation of quantitative traitsQuantitatively inherited traits are geneticallycomplex, are conditioned by a numberof genes each having relatively small effects,and their expression often depends on interactionsamong different genetic components(epistasis). The environment also has a highdegree of influence on the expression of thetrait, which confounds the interpretation ofQTL identification and often renders theresults obtained from QTL studies crossspecific.When it is necessary to manipulateseveral genomic regions simultaneously,each having different effects on the sametrait of interest, MAS-based approachesbecome more complicated and presentformidable challenges. Mapping studies conductedat CIMMYT identified five genomicregions associated with the anthesis-silkinginterval which is a parameter associatedwith drought tolerance in maize (Ribautet al., 1996, 1997). The drought tolerantparent was used in MAS experiments asthe donor parent to transfer the five QTLto CML 247, an elite inbred line with goodcombining ability that was drought-susceptiblebut high-yielding under favourableconditions. Markers were used to generate70 BC 2 F 3 lines containing the favourablealleles from the drought-resistant parentafter two backcrosses and two self pollinations.These lines were crossed withtwo testers for field evaluation. Field testsindicated that under severe drought stressconditions, the 70 MAS-derived lines weresignificantly better yielding than the controls.The differences were less prominentunder reduced drought stress (Ribaut andRagot, 2007).Other CIMMYT experiments aimed atcomparing MAS with phenotypic selectionhave been conducted for stem borers intropical maize (Willcox et al., 2002). In thecase of maize stem borer resistance, threeQTL identified through mapping experimentswere used in MAS. Three BC 2 S 2families that carried all three target QTLfrom the donor parent in homozygous statewere developed. Comparative studies withMAS and traditional phenotypic selectiondid not establish a clear advantage for MAS,but both approaches yielded significantgenetic gains in reducing leaf damage. MASis not being used currently on a routinebasis at CIMMYT for drought and stemborer resistance.