Comparison of Various Mexhods for the Estimation of Breeding Values in Gir Cattle 1674
Loading...
Files
Date
2012-11
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
JAU,JUNAGADH
Abstract
A total of 1257 lactation records of 345 Gir cows sired by
52 Gir bulls maintained at Cattle Breeding Farm, JAU,
Junagadh from 1987-2010, were analyzed to compare various
sire and cow evaluation methods for lactation milk yield (LMY)
(standardized to 305-2X-ME basis and also to SLPY), lactation
length (LL) and calving interval (CI). Earlier records (1965-
1985) were also collected to be partly used in some of the
methods. Sire evaluation was done using conventional indices
(viz. Modified Simple Daughter Average index, Equiparent
index, Norton index, Rice index, Tomar index, Corrected
Daughter Average index. Contemporary Comparison method
and Corrected Contemporary Comparison method), Least
Squares method (LS), Best Linear Unbiased Prediction (BLUP)
method and Average Information Restricted Maximum
Likelihood (AIREML) method. Cow evaluation was done using
LS, BLUP and AIREML methods.
BLUP and AIREML models were constructed using
univariate models, multivariate models (based on first lactation
records) and multivariate repeatability models (based on all
lactation records) using the three traits in all possible
combinations. Criteria for comparison of all the models
involved error variances to measure the efficiency of models.
Additionally, coefficients of determination (%) i.e. R2 values and
coefficients of variation (CV%) were used to measure the
accuracy and stability of models, respectively for linear fixed
and mixed models (LS, BLUP and AIREML).
Among the conventional indices. Modified Simple
Daughter Average index (I-l) was the most efficient method for
sire evaluation for 305-DMY and LL followed by Corrected
Daughter Average index (1-6). Sire rank correlations of breeding
values for 305-DMY and LL by I-l and 1-6 were highly
significant (P<0.01). For CI, 1-6 was the most efficient method
followed by I-1.
Among the linear fixed and mixed models, all sire and cow
rank correlations of breeding values for the three traits were
highly significant (P<0.01) except those with AIREML model (A-
13) which involved 305-DMY and CI. Repeatability AIREML
model (A-S) which incorporated SLPY, LL and CI was the most
accurate method of evaluation of animals (sire and cow) for
LMY followed by AIREML model A-12 which involved 305-DMY
and LL. A-12 was found to be most accurate method of
evaluation of animals for LL. A-13 was the most accurate
method for evaluation of animals for CI. LS was the most
stable method for all the three traits. A-12 was the most
efficient method for evaluation of animals for LMY and LL. A-S
had relatively higher error variance than A-12. Most efficient
method of evaluation of animals for CI was A-13. Sire rank
correlations of breeding values for the three traits obtained by
first lactation models with their respective all lactation models
were highly significant (P<0.01) except those with A-13. Use of
all lactation records instead of first lactation records lead to
33.17%, 40.48% and 21.67% gain in accuracies; 36.73%,
28.78% and 18.26% gain in stabilities and 21.38%, 37.64%
and 26.50% gain in efficiencies for LMY, LL and Cl,
respectively.
Thus, AIREML repeatability model incorporating 305-DMY
and LL is recommended for evaluation of animals for the
production traits viz. lactation milk yield and lactation length,
and AIREML repeatability model incorporating 305-DMY and
CI is recommended for evaluation of animals for calving
interval, a reproduction trait.