Agriculture Reference
In-Depth Information
Then
isolation principle or progeny test. He proposed that
the only means to determine the value of an individual
plant (or genotype) was to grow and evaluate its progeny.
Ever since, of course, the progeny test has become well
established and is frequently used by plant breeders to
determine the genetic potential of parental lines. The
diallel cross is simply a more sophisticated application
to Vilmorin's progeny test.
The term
diallel cross
has been attributed to a Danish
geneticist (J. Schmidt) who first used the design in
animal breeding. The term and design came to plant
breeding and began to be used by plant scientists in the
mid 1950s.
The diallel cross was then described as
all possible
crosses amongst a group of parent lines
. With
n
par-
ents there would be
n
2
families. The
n
2
families or
progeny are called a
complete diallel cross
. If the recipro-
cal crosses are not made, making
n
Statistic
Female
parent
Male
parent
Mid-parent
value
Regression
0.476
0.468
0.813
slope (
b
)
se(
b
)
0.1632
0.1895
0.1269
t
8df
2.898
2.259
6.407
From the regression of offspring on one parent:
2 families, the
result is called a
half diallel
.A
modified diallel
is one in
which all possible cross combinations are included but
the parental selfs (diagonal elements) are excluded. This
type of diallel will include
n
2
[
n
−
1
]
/
male
h
n
=
2
×
b
=
2
×
0.473
=
0.946
female
h
n
=
×
=
×
=
2
b
2
0.468
0.936
−
n
families. In a
partial
[
−
]
/
diallel
fewer than the
n
2 cross combinations
are completed. However, the crosses that are included
are arranged in such a way that valid statistical analysis
and interpretation can be carried out.
Initially, only inbred homozygous lines were used
as parents in diallel crossing designs. Techniques
that allow for parents to be non-inbred genotypes
(i.e. heterozygous) are now available.
According to some critics of the designs '
the diallel
mating design has been used and abused more extensively
than any other
n
1
From the regression of offspring onto the average
phenotype of both parents (mid-parent) we have:
h
n
=
0.813
You will notice that the heritability using only one
parent is larger than that from both parents. It should
be noted that the estimation based on both parents will
be more accurate. Despite the difference, however, it is
obvious that there is a high degree of additive genetic
variance for this character.
Finally, always remember that a heritability estimate,
no matter which method is used to obtain it, is only
valid for that population, at that time, in that environ-
ment! Change the environment, carry out (or allow)
selection to occur, add more genotypes, sample another
population and the heritability will be different! This
should be clear from the descriptions and methods of
calculating heritability but you will find many exam-
ples in the literature where the basic limitations of the
concept are forgotten.
...
'. Whether this statement is true or
otherwise, there is little doubt that if the theory of dial-
lel analysis is adhered to and if interpretation can be
carried out in a logical manner, then the use of diallel
crossing designs can be of great benefit to plant breeders
in aiding understanding of qualitative inheritance and
providing invaluable information regarding the genetic
potential of parental lines in cross combinations. The
limitation of the design arises in terms of the sample of
parental genotypes that can be handled, which is always
somewhat restricted.
It will not be possible to cover the whole spec-
trum of information or even indeed the types of diallel
crossing schemes that are available or to investigate
the interpretation of many examples within the space
available. There are therefore several approaches to
the analysis and interpretation of diallel cross data
DIALLEL CROSSING DESIGNS
It has been over 130 years now since the publication
by Louis de Vilmorin that became known as Vilmorin's
