ANIMAL GENETICS & BREEDING

UNIT – II
Principles of Animal & Population Genetics
Lecture – 8

Repeatability

Dr K G Mandal
Department of Animal Genetics & Breeding
Bihar Veterinary College, Patna
Bihar Animal Sciences University, Patna
 Importance of partitioning the variance
• To understand the relative importance of each
and every component of variance attributable to
the different sources of variation
• To estimate the relative importance of various
determinants of the phenotype, particularly the
role of heredity and environment.
• To determine genetic determination, VG/VP.
• To estimate heritability of the character, VA/VP.
• To determine the degree of resemblance between
relatives.
• To determine the repeatability of a character.
 Repeatability
• Repeatability measures the correlation
between repeated measurements of a
character on the same individuals.
• When more than one measurement of a
character is made on each individual, the
phenotypic variance can be partitioned into
two components, variance within individual
and variance between individuals.
• The ratio of between individual variance
component to the total phenotypic variance
is called repeatability.
• Examples:
The following characters shows repeated
measurements on each individual:
(i) Lactation Milk Yield – it can be measured
in subsequent lactations on the same animal.
(ii) Litter size – it can be obtained in
successive pregnancy of litter bearing
animals.
(iii) Wool production in subsequent shearing
of sheep.
Several measurements can be obtained from
each individual.
• Repetition of a character:
• There are two ways by which repetition of a
character may provide multiple measurements :
(i) by temporal repetition
(ii) by spatial repetition
Temporal repetition:
• The variance of milk yield per lactation or number
of young per litter can be partitioned into a
component within individuals and a component
between individuals.
• The within individual variance component
measures the differences between the
performances of the same individual.
• The between individual variance component
measures the permanent differences between
individuals.
• The within individual variance component is entirely
environmental in origin and caused by temporary
differences of environment between successive
performances.
• The between individual variance component is
partly environmental and partly genetic, the
environment part is caused by circumstances that
affect the individuals permanently.
• By this analysis, the variance due to temporary
environmental circumstances is separated from
the rest and can be measured.
Spatial repetition:
• The character repeated in space are chiefly
structural or anatomical, and are mostly found
in plants than in animals.
Example:
• Plants that bear more than one fruit, provide
more than one measurement of any character
of fruit such as fruit shape and no. seed content.
• Spatial repetition in animals is chiefly found in
characters that can be measured on the two
sides of body or on serially repeated segments
of body parts, such as number of bristles on the
abdominal segments of Drosophila.
• In spatially repeated characters the within
individual variance component is again
environmental in origin but represents the
‘developmental’ variation arising from localized
circumstances operating during development.
• In both temporal and spatial repetition, the term
special environmental variance component will
be used to refer to the within individual variance
component arising from temporary or localized
circumstances and it will be denoted as VEs.
• And general environmental variance component,
VEg, will refer to the environmental variance
contributing to the between individual variance
component (VG + VEg) and arising from
permanent or non-localised circumstances.
• The ratio of between individual variance component to
the total phenotypic variance is intra-class correlation
‘r’. It is the correlation between repeated
measurements of the same individual and is known as
repeatability.
Repeatability, r = (VG + VEg)/VP
or, rVP = VG + VEg
• Therefore, repeatability express the proportion of the
variance of single measurements that is due to
permanent or non localized differences between
individuals to the total phenotypic variance.
• The special environmental variance component, VEs,
as a proportion of total is given by
VEs / VP = 1 – r
or VEs = (1 – r)VP
 Repeatability
 Repeatability measures the correlati
repeated measurementsof the traits o
individual.
Example – milk yield per lactation
- Litter size in successive partur
- Wool production in successive
- Egg production in subsequent month
• The phenotype variance (VP) of such ch
be partitioned into variance within ind
and variance between individuals (VG +
r=
• It ranges from 0 to 1
 Use of Repeatability
• To show how much is gained from
repeated measurements.
• To sets upper limit of genetic
determination (VG/VP) and
heritability (VA/VP).
• To predict future performance from
the past records.
1. To sets upper limit of genetic
determination and heritability:
Repeatability, r = (VG + VEg)/VP
Genetic determination = VG/VP
Heritability, h2 = VA/VP
Since, (VG + Veg)/VP > VG/VP > VA/VP
Hence, r > Genetic determination > h2
2. Gain from repeated
measurements:

 If repeatability is high, little will be gained; if

repeatability is low, more will be gained.
 Only special environmental variance
component, VEs, can be reduced by
increasing the number of measurements.
 Then, VP(n) = VG + VEg + 1/n(VEs)
 This reduction in phenotypic variance
represents the gain in accuracy.
• The variance of mean of n measurements as a
proportion of the variance of one measurement can
be expressed in terms of repeatability, as follows:
Repeatability, r = (VG + VEg)/VP
or, rVP = VG + Veg
VEs / VP = 1 – r or, VEs = (1 – r)VP
VP(n) = VG + VEg + 1/n(VEs)
= rVP + 1/n(1 – r)VP
= VP[r + 1/n(1 – r)]
VP(n)/VP =( nr + 1 – r)/n
= [1 + (n – 1)r]/n
3.Prediction of future
performance:
 If X and Y are first and second
performances, x and y are the
population means, and ‘b’ is
regression coefficient of y on x, then
prediction is given by ( Y – y) = b(X –
x).
 The relationship between
regression and correlation coefficient
is
b = r(6y/6x), where 6x and 6y
Use of repeatability
• To slow how much is gained from
measurements.
VP = VG + VEg + VEs
VP(n) = VG + VEg + 1/n VEs
=
• To set upper limits of
determination (VG/VP) and heritab
/ VP).
(VG + Veg)/ VP > VG/VP>VA/VP
• To predict future performance f
past record.
 Conclusion

 When repeatability is high, multiple

measurements give little gain in accuracy

 When repeatability is low, much will be

gained in accuracy from multiple
measurements

 The gain in accuracy falls rapidly as the

number of measurements increases

 It is seldom worthwhile to make more

than two or three measurements.
THANK YOU