marker-assisted selection in wheat - ictsd

Chapter 16 – Possibilities for marker-assisted selection in aquaculture breeding schemes 315wide diversity of experimental populationsbeing used in such studies. Double haploids(see below) have been used for QTLmapping in salmonids (by androgenesis;Robison et al., 2001; Zimmerman et al.,2005) and tilapias (by gynogenesis; Paltiet al., 2002). Backcross populations havebeen set up where strains with large phenotypicdifferences in the trait of interestare crossed, e.g. for temperature tolerancein rainbow trout (Danzmann, Jackson andFerguson, 1999). The strains can come fromone species, but crosses between speciesalso have been used (Streelman, Albertsonand Kocher, 2003 for tilapia; Rodriguez etal., 2005 for salmonids). Finally, familiesderived from a breeding nucleus have beenused for QTL mapping. (e.g. Moen et al.,2006 for Atlantic salmon).Most QTL mapping is based on markerassociation studies (e.g. Sakamoto et al.,1999) or on a marker association studyfollowed by interval mapping (Moen etal., 2006). Combined linkage/linkage disequilibriummethods (e.g. Meuwissen etal., 2002) have high precision for mappingQTL in outbred populations, but have notbeen used for QTL mapping in aquaculturespecies. This lack could be explained by thesparcity of genetic marker maps for the speciesunder study and by many of the studieshaving used special crosses as mentionedabove, where linkage is the main source ofinformation for mapping the QTL.Various reproductive manipulationsmay be applied to aquaculture species.One interesting reproductive manipulationtechnique for outbred populations formarker and QTL mapping is gyno- andandrogenetic double haploids (Chourrout,1984). In gynogenesis, the sperm’s chromosomesare inactivated by irradiation andfollowing fertilization, diploidy is restoredby applying a temperature or hydrostaticpressure shock. The result is an individualthat is double haploid with only the female’schromosomes. Depending on when diploidyis restored, two types of doublehaploid individuals can be produced. Ifan early shock is applied, extrusion ofthe second polar body is suppressed and,when the two maternal chromosome setsfuse, some heterozygosity is retained. Ifa late shock is applied, the first mitoticcleavage of the zygote is suppressed and,when the two maternal chromosome setsfuse, the resulting individual is virtuallyhomozygous. In androgenesis, the egg isirradiated and, after “fertilization”, the eggis shocked to suppress the first mitosis(Parsons and Thorgaard, 1984). The resultis an individual that is a double haploidwith only the male’s chromosomes and thatis virtually homozygous.The power of an experiment to detectQTL depends on the effect of QTL alleles,the recombination rate among the markerand QTL loci, and the sample size of themapping population. The effect of QTLgenotypes is higher for double haploid thanfor full- or half-sib family designs in anoutbred population because the QTL genotypesoccur only in a homozygous form indouble haploids (i.e. in their most extremeform). The relative advantage of a populationof mitotic double haploids, where thetwo chromosome sets are copies of eachother, is largest when the QTL has a smalleffect (Martinez, Hill and Knott, 2002). Inmeiotic haploid individuals, the two chromosomesets in an individual have beenrecombined. The power of QTL detectionin these meiotic double haploid populationsis therefore expected to resemble thatof selfed populations (Martinez, Hill andKnott, 2002). Double haploids have beenused for genetic marker and QTL mapping,as noted above.