Multiple Breed Evaluation
Can MBE enhance crossbreeding?
John Pollak
Cornell University
Director, NBCEC
Yes
Can MBE enhance crossbreeding?
Outline:
1. Brief overview of motivation for MB models.
2. On the models themselves.
3. Justification for the answer “yes.”
4. Where are we today in MBE.
National Genetic Program
Objective: Design a national genetic program for profitable production of beef.
What would be desirable feature of that program?
The Beef Industry
Commercial
Feeders
Seedstock
Genetics
Calves
ProcessorsFinishedcattle
ConsumersAnimalProtein
Invests in information.
Wealth of opportunity for useful information for genetic improvement down stream.
Adaptability
Feed Performance
Carcass Health
• Integrated system of data capture across all segments of the industry on sire identified animals from birth to slaughter or maturity.
Non-traditional sources like commercial ranches and feedlots. Typically crossbred cattle.
Barriers: Animal ID and Sire ID
National Genetic Program
• Commercial industry that:– Maximizes the expression of heterosis
through crossbreeding and developing composite populations.
– Utilizes the “best” genetics from the seedstock industry regardless of breed.
National Genetic Program
• Seedstock industry that:– Focuses selection decisions on economically
relevant traits for the commercial industry.
– Provides genetic assessments that are comparable across breeds.
National Genetic Program
Motivation
All require some method of assessing animals of different breeds or breed composition.
That is, all aspects of that national program could benefit from multibreed evaluation technology.
Two programs for weight traits:
ASA/Cornell 1997Several breed associations participate.
Georgia 2002Single breed association evaluations with crossbreds
Current MBE Programs
Both use external EPDs.
MB system for carcass traits/ultrasound at Cornell.
Motivation
Built the ASA MB-ICE for several reasons:
• Better handle the evaluation of the composite, Simbrah.
5/8 Simmental + 3/8 Brahman
We wanted to appropriately account for the influence of different Simmental founders on the Simbrah breed.
Motivation
Built the ASA MB-ICE for several reasons:
• Better handle the evaluation of Simbrah.
• To provide evaluations for F1 seedstock.
Growing number of Simmental members were producing F1 bulls for sale as seedstock and we wanted to better
describe the genetics of these animals.
MotivationBuilt the ASA MB-ICE for several reasons:
• Better handle the evaluation of Simbrah.
• To provide evaluations for F1 seedstock.
• To allow for evaluations of other composite cattle that were being reported to the ASA database.
MotivationBuilt the ASA MB-ICE for several reasons:
• Better handle the evaluation of Simbrah.
• To provide evaluations for F1 seedstock.
• To allow for evaluations of other composite cattle that were being reported to the ASA database.
• If other breeds joined in we would have EPDs that were directly comparable across those breeds to enhance selection decisions of commercial cattlemen.
Motivation
Notice:
All the motivating factors for building the system relate directly to the theme of this presentation and to the theme of this BB series.
MBE ModelsArnold et al. (JAS, 1992) presented the basic animal model for
multibreed evaluation.
Observation = Contemporary group effects +
other effects (e.g. age of dam) +
Breed effects + Breeding values +
Non-additive effects+
Residual
Maternal effect: include breed effects, breeding values and non-additive effects for the dam.
Foundation for both the AS/Cornell and UG multiple breed models.
MBE Models
The estimated breeding value is:
EBV = Combination of breed effects +
random genetic effect of that animal
The nonadditive effect is to account for the effect of heterosis on
performance.
Multibreed Models
e.g. compare the performance of:
Simmental calf to a F1 Simm-Angus calf in the same CG
Account for the contribution to performance from the expected heterosis in the F1 calf
Heterosis
10 Effects in the ASA/Cornell MB-ICE
Continental ... Cont, Brit, Zebu, Other
British ... Brit, Zebu, Other
Zebu ... Zebu, Other
Other … Other
Could include breed specific heterosis values
(e.g. Angus with Hereford)
Heterosis
Probability alleles originate from different breeds
CharolaisCharolais x x AngusAngus = = 1.001.00
CharolaisCharolais xx (1/2)A(1/2)A (1/2)C (1/2)C = = 0.500.50
CharolaisCharolais xx (1/4)A(1/4)A (3/4)C (3/4)C = = 0.250.25
This cross would get the full contribution of This cross would get the full contribution of the Continental by British heterosis.the Continental by British heterosis.
Heterosis
Charolais Charolais x x (1/2)Angus(1/2)Angus (1/2)Simmental (1/2)Simmental
This cross would get:This cross would get:
50% of the Continental by British heterosis 50% of the Continental by British heterosis
++
50% of the Continental by Continental heterosis50% of the Continental by Continental heterosis
1. The motivation for developing the MB evaluations are consistent with the enhancements they would bring to crossbreeding programs.
2. An MB system allows for the opportunity to expand the scope of genetic evaluations.
3. The model for MB allows for crossbred (composite) data.
Summary
Previous comments:
The quality of the crossbred (or composite) program is a function of the quality of the parents selected as foundation animals.
Why “Yes”
EPDs, in general, allow for characterizations of foundation animals and for setting bench marks
for future selections.
In cases where several breeds are joined in a multiple breed evaluation, direct comparison of the genetic
merit of animals from those breeds is possible.
Why “Yes”
F1 bulls have achieved a significant level of market share.
Why “Yes”
Multiple breed EPDs allow for characterizations of the F1 animals in comparison to their purebred
counterparts.
Previous comments:
One concern with composites is the lack of access to genetic evaluations.
Why “Yes”
Multiple breed EPDs allow for evaluations of composites and accounts for the differences in
breed composition within the composites in both the initial stages of composite development and in
some cases, at “stabilization.”
Caveat
Is our EPD an EPD?
EPD = Expected Progeny Difference
The difference is technically in the additive genetic merit of the two calves.
However, all else equal, this is ALSO a prediction of the difference in calf performance.
Caveat
Is our EPD an EPD?
Sire 1 Dam Sire 2
Calf 1 Calf 2
EPD = Expected Progeny Difference
If all are the same breed then the EPD predicts both the differences in the additive genetic merit of the calves as well as the expected
difference in their performance.
Predicting performance differences as a result of selection is important in the economic assessment
of a selection decision.
Economic Assessment
Is our EPD an EPD?
Simmental Dam Angus
Calf 1 Calf 2
Simm-Angus
In this example, the contribution of heterosis is the same regardless of choice of sire so the prediction of the performance difference is
still the differences in the calves EPDs.
Is our EPD an EPD?
Simmental Dam Angus
Calf 1 Calf 2
Simmental
In this case, the difference in performance is influenced by the difference in expected heterosis which is not a part of the EPD.
With the use of MBE EPDs we need to make clear in educational programs that the EPD has a
specific definition.
Prediction of performance differences in calves is not a by product of the EPDs in certain cases.
Decision support tools for phenotypic prediction. NBCEC project at CSU
NBCEC MB Prototype
We are currently in the process of researching the use of the MB model for a greatly expanded
number of breeds.
NBCEC MB Prototype
• Developing a national pedigree file
• Developing a national database of weight records
• Using the ASA/Cornell model for this investigation
• Use the Georgia model for application
Current prototype run included:
ASA data (Simmental, Chianina, Maine Anjou)
Red Angus
Limousin
Gelbvieh
Charolais
Brangus
Currently loading South Devon and Saler data
NBCEC MB Prototype
Process:
“Load and run”
Each time we load a breed into the national pedigree and data files we run an evaluation.
Report “anomalies” to breed associations for review and, if necessary, correction for the next
run.
NBCEC MB Prototype
Rank correlations between the within breed runs and the NBCEC prototype run are >0.9 for
high accuracy bulls.
NBCEC MB Prototype
Nice check of all three systems!
Issues:
Identification across registries
Contemporary groups definitions
Adjustments to records
Data filters and edits
Breed Specific Parameters
Commercial information
Independent composite programs (e.g. Bell Ranch)
NBCEC MB Prototype
Evolution of MB programs in the NBCEC:
NBCEC MB
Cornell will run the MB weight trait prototype and take ultimate responsibilities for the national database.
Georgia is preparing a multiple breed carcass/ultrasound evaluation and random regression model for weights.
CSU is developing the web based selection decision support programs to utilize the MB EPDs.
Other Traits?NBCEC MB
Obviously, to provide complete profiles of potential parents we will needed to evolve all trait evaluation programs to MB.
For the cow-calf decision support program need:
Calving easeStayability
Cow Maintenance requirements
For the “finishing” decision support program need:
Days-on-feedRFI?