Agriculture Reference
In-Depth Information
were preferred to elite-by-nonelite crosses, with
the idea that the nonelite lines might have rare
or unused benefi cial alleles (Busch et al., 1974;
reviewed by Baenziger and Peterson 1992). Virtu-
ally every program now relies on elite-by-elite
crosses for most of their breeding effort. Simply,
these crosses have the highest likelihood of pro-
ducing new cultivars. As such, most breeders
heavily intermate their own material, followed by
crossing to elite material in their region. In this
effort, most breeders can use single or three-way
crosses because the parents have equivalent or
nearly equivalent value. Only when new diseases
or pests or new traits are identifi ed and where the
resident breeding program is unlikely to have
useful genetic diversity for the trait will the
breeder use less-adapted germplasm. This could
include elite germplasm from another region or
even a wild or alien species.
Once the breeder decides to cross to nonelite
material, rarely are single crosses used except as
an interim cross to a three-way cross or backcross.
The value of a three-way cross is that nonelite
germplasm will contribute only 25% of the genes
in the segregating population, including those
which are truly targeted. If the germplasm source
has very few useful genes, a backcross can be used
to further reduce the amount of the nonelite
germplasm genes in the subsequent progeny.
Most progeny lines that are developed from a
cross are discarded, a few are used as parents, and
very rarely is a progeny line released as a cultivar.
Hence germplasm that has few useful genes might
be used as a parent to develop lines (without back-
crossing), which in turn may become parents for
future crosses (effectively an extended cycle and
“backcross”).
Despite the research on which crosses yield the
best progeny (reviewed by Baenziger and Peter-
son 1992), crossing remains a relatively simple
process in wheat, while evaluating parents and
using specifi c mating systems to determine paren-
tal value are cumbersome. Thus most breeders
make a large number of crosses knowing that most
will fail to produce a cultivar. Often their progeny
will become new parents for future cultivars or
again parents. All breeding programs are funda-
mentally recurrent selection programs as previ-
ously alluded (Acquaah 2007), with differing
periods of time for inbreeding, evaluation, and
intermating.
Once the objective is decided and the cross is
made, the breeder must decide which breeding
method(s) to use. In our preceding discussion on
breeding methods, each method was presented in
a pure form. In practice, every breeder modifi es
the method to fi t the available resources. Also,
during the 7 to 12 years it takes to release a culti-
var, different opportunities arise for selection,
which require a highly fl exible approach.
For example, many breeding programs may
start as a bulk breeding program to allow natural
selection (perhaps winter survival) or, in the case
of IMI-tolerant wheat, artifi cial selection (herbi-
cide application to remove the susceptible plants)
to effi ciently remove unwanted types. Often only
one or two generations of selection are needed to
remove the undesired types, so thereafter the
breeding method might change. For example, in
the later generations when it is more diffi cult to
select lines phenotypically, perhaps single-seed
descent will be used to rapidly advance
generations.
Alternatively if key traits can be selected in the
fi eld during the fi rst two generations of inbreed-
ing, pedigree breeding may be used to identify the
best F
2
or F
3
plants for creating progeny rows.
This procedure is an example of allele enrich-
ment, or moving the population to having more
desired types genetically and phenotypically as
further described later.
Most breeding programs are so large that their
selection procedures have to be effi cient and
repeatable. In this area, great care needs to be
taken as to where the early generations are grown,
because they are often nonreplicated and if they
are lost will the work and resources used to
develop the plant material will be lost as well.
Loss of a key nursery often impacts the genetic
progress by more than a year by disrupting the
acquisition of data upon which to make selections
and by losing the genetic materials themselves. At
the same time, natural selection can be advanta-
geous for moving populations to the desired phe-
notype (Baenziger and Peterson 1992; Baenziger
et al., 2006b). Ideally selection nurseries, or nurs-
