Methods of Selection
Methods of Selection
AMITOSH
METHODS OF SELECTION
• From the practical standpoint, the net value of an animal depends upon several
traits that may not be of equal economic value
• Dairy cow will be more economical to maintain if she produces more milk with
higher fat % for a longer lactation period and remained dry for a shorter time
between successive calving
• On the other hand if a cow produces more milk daily with low fat % with shorter
lactation length, it will not be economical to maintain
• In this method selection is practised for only one trait at a time until
improvement has been made in the trait
• Selection efforts for this trait are then relaxed and efforts are directed towards
the improvement of a second trait, then a third and so on
• If the traits are negatively correlated then when trait ‘A’ improves, the trait ‘B’
becomes worse (See-saw effect)
Merits
(1) It is simple to practice
(2) Intensity of selection is more
(3) If there is a desirable genetic correlation with other traits, it produces
progress in the related traits
Demerits
(1) Only one character is considered at a time
(2) Efficiency of this method is low
(3) It takes very long period of time to achieve desirable efficiency. In the mean
time the breeder may change his goal
(4) If there is undesirable correlation between traits that will also hinder the
progress. i.e. genetic improvement obtained in previous generations will be
lost
• In this method, selection may be practiced for two or more traits at a time
• In this method, for each trait a minimum standard is set so that each animal must
meet those standards in order to be selected
• Failure to meet the minimum standard for any one trait makes the animal to be
rejected
• Effectiveness of this method depends on the level kept for each of the traits
• In case of keeping low levels, very few animals are culled and it affects the genetic
progress. On the other hand, if the standards kept are high, number of animals
selected will be very small
• Generally this method is used in selection of animals for show purposes where the
animals are selected for type, colour and body conformation traits ignoring its
performance for economic traits
• In this method the animal is scored for its merit in each of the traits included in
selection
• An index is a single numerical value, which is the total of scores given for each trait
considered in the selection
• Each trait is weighted by a score and the individual scores are summed to a total
score / index value for each animal which is the selection criterion
• By this method, superiority in some traits can be made up by mediocrity in others
• Selection index is a total score that includes all the advantages and disadvantages of
an animal
• Animals with highest score / index are selected for breeding
• Amount of weight given to each trait depends on the relative economic value,
heritability of characters and genetic correlation among characters
• Selection index method is the most efficient of the three because it results in more
genetic improvement
• The index is the best estimate of animal’s breeding value
• Principle of selection index theory for animals was developed by Hazel (1943)
AKS/AGB Unit-3/Selection Methods 7
3. SELECTION INDEX OR TOTAL SCORE METHOD
Merits
(1) Most efficient method of selection among the three methods
(2) It allows the individuals, which are superior in some traits to be selected
regardless of their deficiency in other traits
(3) All the characters are given proper weightage for optimum selection
(4) Advantage of index method increases with the number of traits
(5) It requires less time to bring about an overall improvement
Demerits
(1) Selection index is difficult to construct because of complex computations
involved
(2) Relative economic value, genetic parameters are likely to change from time
to time and from population to population. Therefore, an index selected for
a particular generation and for particular population cannot be valid
subsequently
(3) Efficiency of the index is reduced unless the genetic parameters are
estimated with good precision from large data
If there are ‘n’ traits under selection, the selection index is constructed by multiple
regression analysis as:
X1, X2……. Xn are the phenotypic value of the 1, 2 ……. n traits of the individual
b1, b2…….. bn are the corresponding weighing factors of the trait (multiple
regression coefficient) by which the each measurements / phenotypic value to
be weighed
The b’s are estimated by solving set of normal simultaneous equations whose
number is equal to the number of traits under selection
• Different traits may not have equal heritability estimates but some may have
low heritability
• Accuracy of selection for such traits can be increased by using the information
from other sources (relatives’ records of individual) for such traits
• It is obvious that due to selection there will be change / response in each of the
trait included in the index and even in other traits not included in the index
• Generally it is followed for traits which have negative genetic correlations. For
e.g. in poultry, the situation arises when one wants to increase the egg
production without affecting the egg size, which is negatively correlated
3. Multistage selection
• There are two stages of selection. The selection of the first stage is based on
physical attributes of phenotypic performance and the selection at the second
stage is based on the breeding value of the animals, which is the final selection
• The genetic gain expected from this combined selection is always higher than
those obtained from either individual selection or any other selection aids
alone