Agriculture Reference
In-Depth Information
information. At harvest, only the lines which have
met all selection criteria are retained and others are
discarded.
Visual assessment is often very subjective and differ-
ent people have been shown to give differing emphasis in
screening, depending on individual preference. Overall,
however, different evaluations based on visual ratings
can be remarkably similar in the lines that are chosen.
Despite the problems with visual selection it can be car-
ried out relatively quickly so many lines can be evaluated
and at low cost. The biggest failing of visual selection
alone is that characters can only be assessed if they can
be seen. Therefore it is not usually possible to use visual
evaluation to screen, for example, for quality charac-
ters. If more objective selection is to be achieved than
it may be necessary to actually record information on
yield, disease rating or quality.
Even when there is no replication it is possible to
obtain some indication of error in visual assessment
if the assessment operation is repeated. This will not
provide environmental error estimation but can often
be useful in determining the repeatability of visually
assessed characters.
Visual assessment of yield and yield components has
been examined and visual evaluation of yield from single
rows or small plots has proved unreliable in predict-
ing actual yield in subsequent generations. The highest
yielding progeny bulks, derived from F
2
and F
3
single
plants, do not necessarily produce the highest yield-
ing segregants. Visual selection for yield on individual
plants in cereals results in only a random reduction in
population size with little or no effect in increasing yield.
Even when the actual yield of an early generation of a
cereal pedigree bulk breeding scheme (say F
2
or F
3
)
was measured and it was found to be significantly cor-
related with yield in later generations (say F
5
or F
6
)
the
association found between segregating populations was
usually so poor that it was questionable whether selec-
tion at the early stages (along with the expense that this
would incur) would be justified.
Selection for yield
per se
in the early segregating gen-
erations of other inbreeding species has also been shown
to produce an effect which is no better than random.
Examples from past research include chickpea, cotton,
soybean and rice. In addition selection for yield com-
ponents such as seed size in chickpea and grains per ear
in spring barley was shown to be slightly more effec-
tive in the early generations than selection for yield
itself.
The large numbers and small plots used in the early
generations dictates that selection is only carried out
for characters which are highly heritable. Often these
only include single gene traits. As might be expected,
selection for qualitative disease resistance in the early
generations has been found to be more effective than
selection for quantitatively inherited resistance or other
polygenic characters.
The efficiency of selection in a pedigree bulk breed-
ing scheme has been related to the heterozygosity of
the bulks under selection. As homozygosity increases,
selection becomes more effective. Homozygosity can
be accelerated by single seed descent. However, care is
needed to ensure that in single seed descent there is not
a non-random loss of genetic material. Homozygosity
can also be accelerated by various doubled haploid tech-
niques. Again however, care must be taken in the use of
these procedures as there is evidence of non-random
success and a strong genotypic response to
in vitro
regeneration.
Most research into the efficiency of selection on
clonally reproduced crops has been on potatoes and
Mass selection
It is often possible to use mass selection in the early
generations. Examples of this would include selection
for short plants by cutting tall ears from popula-
tions in wheat. Mass selection can also be used to
select larger seeds, higher specific gravity in tubers and
morphological traits such as fruit colour.
Mass selection in the early generations can be
achieved by growing the early population under specific
environmental conditions. The more adapted lines will
be more productive and the frequency of less adapted
genotypes will reduce. Bulk selection has been shown
to be effective in increasing the frequency of drought,
heat, salt and other stress tolerances.
Efficiency of early generation selection
The efficiency of early generation selection has been
examined in a number of different crops. When breed-
ing an autogamous species, for example wheat or barley,
selection will be influenced by the highly heterozygous
nature of the breeding lines in the early generations.
Segregation effects can be avoided by advancing towards
homozygosity prior to selection but has not been com-
mon in the past because of time restraints or cost factors.
