Agriculture Reference
In-Depth Information
where breeding nurseries can be lost to hail, win-
terkilling, drought, or other unforeseen disasters.
Even with these concerns, information on early-
generation bulks can be valuable to wheat breed-
ers as they decide on which populations to focus
their efforts (Busch et al., 1974).
generation. If any phenotypic selection is prac-
ticed, it is based on the individual plant (hence
must be for highly heritable traits) without regard
for progeny performance.
Knott and Kumar (1975) compared single-seed
descent with early-generation yield testing and
found that, in general, early-generation yield
testing identifi ed higher yielding lines on average.
The lower yielding lines were removed from the
population, indicating that in early generations it
is often easier to cull poor-performing lines than
to select high-performing lines. However, there
were very few differences among the higher yield-
ing lines; hence the authors felt the early-genera-
tion testing was not worth the extra effort. Because
plant breeders are most interested in the high-
performing lines, they concluded that single-seed
descent would have considerable value. However,
the authors did not consider the lost opportuni-
ties from single-seed descent. From Table 13.3 it
is noted that the highest frequency of all 20 desir-
able alleles, from a cross segregating for 20 target
loci, will be present primarily in a heterozygous
condition. Because single-seed descent only
samples one seed per individual, the probability
of selecting individuals with the maximum
number of desirable alleles is greatly reduced.
Probably the greatest concern with single-seed
descent is that it should not be used, if without
selection, in crosses expected to produce greater
genetic variation. As can be seen in Table 13.1,
with as few as fi ve segregating loci, only 3 of 100
single-seed descent lines (or random inbred lines)
would be expected to have the desired allele at all
5 targeted loci. With 10 segregating loci, only 1
plant of 1,000 would have the desired allele at all
10 loci. Hence most plant breeders restrict single-
seed descent to their elite × elite crosses, in which
many of the favorable alleles are already fi xed;
hence segregation for important traits is less.
Also, in these crosses, poor segregants occur at
much lower frequency, making the need to cull in
early generations much less.
In addition to being used in applied breeding
programs, single-seed descent is often used to
create mapping populations (Marza et al., 2006),
where a random sample of lines needs to be
developed which best represents the genetic
Single-seed descent
Single-seed descent is a method to achieve homo-
zygosity while often practicing minimal selection
(Goulden 1939). Brim (1966) referred to the pro-
cedure as a modifi ed pedigree selection method,
most likely because he suggested including selec-
tion in the early generations. In its current usage,
the objective is usually to develop a random
sample of the 2
n
possible homozygous individuals
from a parental mating. Though often done
without selection, single-seed descent does not
preclude selection during generation advance
(Brim 1966). Selection and evaluation often are
delayed until inbred lines have been produced,
when selection can be based on data from repli-
cated fi eld trials for agronomic performance,
biotic and abiotic stress nurseries, and end-use
quality testing. The method consists of selfi ng a
random sample of F
2
-derived plants in each gen-
eration and advancing only one seed per plant.
This process attempts to represent each genotype
present in the F
2
regardless of merit. Techniques
to accelerate generation advancement may be
used to grow two to three generations per year.
This makes single-seed descent ideally suited for
spring wheat, which does not require vernaliza-
tion, but single-seed descent is also used in winter
wheat breeding.
The rationale for practicing single-seed descent
is to reduce the time to achieve near-homozygos-
ity, to reduce operational costs and record keeping
to achieve near-homozygosity, and to maintain
maximum genetic variation while inbreeding.
Natural selection does not infl uence gene fre-
quency in this procedure, unless genotypes differ
in their ability to produce viable seed under the
growing conditions used. Genotypes which do
not produce seed will be lost, and genotypes
which are very or less prolifi c would be repre-
sented equally by a single seed in the following
