Agriculture Reference
In-Depth Information
single-gene IMI tolerance is largely a recessive
trait, so that after spraying, only plants that are
homozygous for herbicide tolerance remain, thus
ridding the population of susceptible homozy-
gous or heterozygous plants. In the latter case,
spring growth habit is typically dominant to
winter growth habit (elaborated in Chapter 3), so
that if the population segregating for winter and
spring growth habit is planted in a spring envi-
ronment, the winter types do not vernalize and
never fl ower, and thus are removed from the
population. However, the heterozygous plants do
remain. If the population segregating for winter
and spring growth habit is planted in a winter
environment and the winter is suffi ciently cold,
the spring types are killed by the winter condi-
tions and thus are removed from the population.
Retention of desirable types that meet the
breeding objective would be typical of positive
mass selection. Positive mass selection increases
the gene frequency of desirable types, and gener-
ally a small fraction of the population is retained
and advanced to the next generation with an
aggregate seed sample. An example would be
retention of F
2
plants that express a high degree
of resistance to a complex of diseases, such as
rusts (
Puccinia
spp.), bunts (
Tilletia
spp. and
Ustilago
spp.), and
Fusarium
spp., and that meet
the target expression of plant stature and archi-
tecture. Only 5% of an F
2
population might be
retained.
Molecular markers may be applied at some
stage to aid in either positive or negative selection.
Molecular markers can have a major impact on
shifting gene frequency, especially in a topcross
F
1
or backcross F
1
population (Knox and Clarke
2007).
One challenge with bulk breeding is to opti-
mize selection for desirable types while minimiz-
ing unwanted plant-to-plant competition in the
bulk (Harlan and Martini 1938). For example,
with the advent of semidwarf wheat cultivars,
numerous studies showed that tall wheat plants
shaded the desirable and shorter semidwarf plants
(Khalifa and Qualset 1974, 1975). The loss of
semidwarf lines in both the mechanical mixtures
and bulk populations occurred despite the semi-
dwarf line in the mixture or the semidwarf segre-
gants in the bulk being higher yielding. Of course,
competition in a bulk could be reduced by chang-
ing the bulk planting system to one that was less
competitive, such as using thinly spaced bulks
(Khalifa and Qualset 1975) or by keeping the lines
in the bulk for fewer generations so that the
impact of competitive effects would be less.
As mentioned previously, breeders would
ideally like to compare populations on the basis of
their line means and standard deviations. Busch
et al. (1974) showed that the grain yield of F
4
and
F
5
bulks was correlated with the yield of F
2:5
lines
when the bulks and lines were grown in the same
environment, thus reducing the G × E effect.
Interestingly, the highest yielding line came from
a cross between high-yielding and low-yielding
parents, but the highest frequency of high-
yielding lines came from crosses between strictly
high-yielding parents. Hence they concluded that
bulks from crosses between high-yielding (elite)
parents could be evaluated for grain yield and
would be indicative of progeny lines developed
from the bulk populations. Later Cregan and
Busch (1977) found that by using bulks derived
from adapted material, the yield of the F
2
bulks
was positively correlated with the yield of their
derived F
5
lines and that the amount of variation
among the F
5
lines did not seem to affect the yield
of the highest yielding derived lines. This research
is also noteworthy because the process of han-
dling the bulks to develop the lines included an
initial screen for many disease resistance traits,
illustrating how early-generation breeder selec-
tion and bulk breeding can be integrated.
From a practical standpoint, G × E interactions
would affect the applicability of the results from
both of these studies because (i) early-generation
bulk yield trials normally occur in years preceding
progeny-line evaluation, and (ii) often the bulks
are evaluated in one or a few locations whereas the
progeny lines are tested for adaptation in many
more locations. Also, bulk populations are rarely
grown in replicated trials to precisely estimate
their grain yield. However, the ease of planting
and harvesting bulks often leads to bulks being
planted at more than one location. This allows the
breeder to assess their performance in different
environments and to avoid generation loss in areas
