Logo British Limousin Cattle Society

technical

bbc weather

market prices

projects : technical : european genetic evaluation

European Genetic Evaluation

Interbeef, the story so far

Eric VENOT, Research Engineer, INRA-SGQA (UR337), F-78352 Jouy-en-Josas, FRANCE
eric.venot@jouy.inra.fr

Thanks to the increasing use of artificial insemination these last four decades, beef cattle breeds have been developing outside their historical breed cradles. Breeders can now make their choice in an international panel and are seeking to compare domestic to foreign breeding animals. However, genetic evaluation of seed stock is usually performed within country. Genetic links between countries created through these exchanges can be used to compare seed stock between countries.

I. Several common genetic evaluations already in place

Since the 90's, several countries have developed common genetic evaluation for their beef cattle breeds. This is usually made of a leader country with a large population and associated countries with smaller populations. These countries generally use the same rules for performance recording and have similar environment conditions and farming systems. Examples can be found in Oceania (Autralia and New Zeland) with Breedplan (Reverter et al., 2002), North America (United States and Canada) (Benysheck, 1998; Bullock et al., 2003), and in Europe (France, Italy and Luxemburg) with the IBOVAL system (Laloë and Menissier, 1990; Menissier et al 1996). These joint genetic evaluations suppose nevertheless no interaction between Genotype and Country, which means that all breeding animals are considered as coming from the same country and the animals ranking is the same in each participating country.

This simplifying assumption should be verified by preliminary studies: several works have shown for instance that weaning weights can be evaluated as one character for Angus Breed in Australia and New Zealand (Meyer, 1995), for Hereford breed in USA, Canada and Uruguay (de Mattos et al., 2000 and Lee and Bertrand, 2002) or for Charolais breed in Australia, New Zealand and USA (Donoghue, 2004) (if heteroscedasticity between countries is taking into account in the model). On the other hand, Genotype by Country interaction has been detected for these countries for birth weight or after weaning growth. In this case, genetic evaluation should be run with more sophisticated models.

II- The EUropean project for international BEef EVALuation (EUBEEVAL project)

A- A collaborative project

In 1999, Irish Cattle Breeding Federation (ICBF) in association with the Meat Livestock Commission (MLC) from the United Kingdom (UK) and the Institut de l’Elevage (IE) from France took the initiative in establishing a research project with a double objective:
1)  develop prototype software to compute breeding values with data from different European countries taking into account the heterogeneity of production systems,
and
2)  study the best way for comparing breeding values obtained in different systems.

The first item was developed by the Institut National de la Recherche Agronomique team (INRA) from France on the Charolais breed and the second by the Animal Genetics and Breeding Unit (AGBU) based at the University of New England in Armidale (Australia) on the Limousin breed. These were funded by the Irish, French and UK participants in the collaboration.

B- Scientific results

1- It is feasible!

At the term of the project in 2004, INRA showed:
-       the feasibility of a joint genetic evaluation for beef cattle taking into account this heterogeneity,
-       the heterogeneity of the environment effects on weaning weight,
-       the potential benefit for the selection intensity.
This preliminary study provided a specific ranking of the French, UK and Irish bulls in each country for Charolais pure bred weaning weights. But it also pointed out data quality problems: better genetic links were needed to compute more reliable genetic parameters in the following studies (see practical results).

2- The model of choice

The AGBU team, with the support of Florence Phocas from INRA, compared different models to apply on beef cattle data:
-       a single trait animal model with heterogeneous variances between countries,
-       a sire model on deregressed EBVs such as for the international dairy cattle genetic evaluation of Interbull,
-       an animal model using raw data and accounting for across country interactions.
The last one was the model of choice to compare beef cattle seed stock from different countries (Phocas and Donoghue, 2004, Phocas et al., 2004). This model presented more precisely below can take into account each country's farming specificities through different fixed effects models but also the different variances heterogeneities and genetic correlations between countries.

C- Practical results

1- Cross reference file: key point to trace back the genetic links between the countries

Associated with these scientific results, this study underlined the necessity of a clean and complete cross-reference table which establish the correspondence between a unique international identification number, a national number (both at the Interbull format) and a national number (at the national format) for every exchanged animals (Quintanilla et al, 2002; Renand 2004): this file is the key point of the joint genetic evaluation and allows the organism responsible for the international genetic evaluation to trace back all the genetic links between the countries. The accuracy of the genetic evaluation depends dramatically on the quality of the genetic links between the countries. Therefore, a country importing an animal from another country should keep the original identification number of this animal in its database.

Following this idea, international identifications of many Charolais and Limousin animals that had not been previously identified as coming from a foreign country were corrected in the Irish and UK databases (worked out by Antunes (2004a, 2004b), Pabiou and Journaux thereafter): the quality of the cross reference table between Ireland, United Kingdom and France have been considerably improved for the two breeds. Based on these new links, the INRA team computed a new set of genetic parameters between France and Ireland for Limousin breed (Venot, 2005a 2005b). Estimation of the genetic parameters between France and UK is presented in details in the last part of this paper (Part IV).

download full report pdf (448k)