267wide use of Mega-varieties or varieties covering wide extensions and used mas-
sively as parents for future varieties may have created a genetic bottleneck for crop
diversity.
The use of landraces and other lines non-adapted to the target region for any crop
in a breeding program to increase genetic diversity has, however, some inconve-
niences. The direct use of a landrace in the crossing block will result in a progeny
with several unsuitable alleles to select against, increasing the number of F 2 plants
needed to identify the ones with acceptable agronomical types. Wheat landraces
usually carry alleles for tall straw or photoperiod sensitivity (Worland and Snape
2001 ) unsuitable for wheat productivity and adaptation while in modern varieties
these traits have been already fixed. Especially, adaptation to modern culture or
bread-making techniques and processes are significantly different than the ones
used when these landraces were selected by farmers. Also, the introduction of for-
eign material may alter suitable co-adapted gene complexes fixed in the elite mate-
rial. Normally, backcrosses to elite material will help in overcoming the linkage
drag. Even after 20 years of conventional breeding, a single gene transferred by
backcrossing may still be linked to a DNA region carrying more than 100 poten-
tially undesirable genes (Young and Tanksley 1989 ). However, marker assisted
backcross, when closely linked markers are available, can greatly simplify this task.
The use of wild relatives, as it was previously stated, can help in improving mod-
ern wheat diversity beyond the limits of the landraces germplasm. However, besides
the laborious and sometimes expensive process of making synthetic wheat, other
complications may reduce the selection efficiency when they are used in a crossing
block. Thus, normally primary SHW lines (directly developed from the crosses of
wheat with wild relatives normally a tetraploid and Ae. tauschii) carry too many
undesirable alleles to produce useful F 2 population. Breeders usually prefer to back-
cross the primary synthetic to increase the chances of obtaining suitable BCXF 2
plants to select superior SDW. Unfortunately, this implies that two full breeding
cycles – the breeding cycle to obtain a suitable synthetic derivative and the breeding
cycle to obtain the final variety after the cross of the derivative with a modern
adapted variety (SHW>SDW>variety released) – are needed to obtain a variety with
superior performance to be released.
Additional complications may occur when crossing SHW and modern varieties:
Often the F 1 plants from such crosses have a high frequency of hybrid necrosis, usu-
ally lethal or semi lethal, resulting in gradual death or loss of productivity (Tomar
et al. 1991 ; Tomar and Singh 1998 ). Hybrid necrosis is controlled by two loci, Ne1
and Ne2 (Pulkhalky et al. 2000 ). The presence of at least one copy of these two loci
dominant alleles, located in the BB genome (Nishikawa et al. 1974 ), results in
hybrid necrosis. Modern bread and durum wheat varieties vary in the frequency of
these alleles (Trethowan and Van Ginkel 2009 ) and, therefore, the cross between
SHW and modern varieties often results in the combination of both dominant alleles
and a necrotic F 1. Most of the research in this area has been based on phenotypic
characterizations of modern varieties according to the alleles they carry at these loci
Breeding and Genetic Enhancement of Dryland Crops