Introduction

Beef genetic evaluations have usually been developed within country. Increasingly, breeders are seeking to compare domestic and foreign bulls in order to enlarge the choice of animals that best fit their selection needs. As more and more genetic material is traded around the world, the need for accurate comparisons of the genetic merit of bulls from different countries has become increasingly important. A general survey carried out in 2006 by the International Committee for Animal Recording (ICAR), showed the need of international genetic evaluation for beef cattle to be run by the Interbull Center in Sweden. Prototype results prepared by INRA (France) and ICBF (Ireland) showing the feasibility of this type of evaluation (Venot et al., 2007, 2008, 2009), ICAR decided to establish, in 2008, a new service called Interbeef. Over the past seven years, the different tools and infrastructure have been built to make Interbeef genetic evaluation a reality.

The objective of this paper is to present the results of the 2014 Interbeef test run for weaning weights performed at Interbull Center and to show the different benefits that can be expected by breeding organisations of the different countries members of Interbeef.

Materials and Methods

Data: Weaning weight was the first trait considered for Charolais (CHA) and Limousin (LIM) pure bred animals in Interbeef joint genetic evaluation. Ten countries are currently members of Interbeef: i) for CHA and LIM evaluations: Czech Republic (CZE), Denmark (DNK), Finland (FIN), France (FRA), Ireland (IRL), Sweden (SWE), Germany (DEU), Switzerland (CHE) ii) for LIM evaluation only: same countries as for CHA plus Spain (ESP) and United Kingdom (GBR). Different trait definitions occurred: several countries used adjusted weaning weight at a specific age (200 or 210 days; e.g., FRA, GBR) based on several weights, while other countries considered weight measured around weaning and correct for age at weighing in the model (e.g., IRL, DNK). Weights sent by SWE were calculated as the difference between adjusted weaning weight and birth weight. Phenotypic records and pedigree data were extracted from the Interbull IDEA database based on the animalís unique international identification number in September 2014. Each member country had uploaded phenotypic performances edited to their respective national genetic evaluation standards. Therefore, data editing at the Interbeef level was thus limited. The final performance dataset edited for genetic evaluation contained 2,672,286 animals from ten countries for Limousin, and 4,224,114 animals from eight countries for Charolais. Back-ancestry, traced back to founders, was also extracted from the IDEA database and comprised of 2,973,893 and 4,963,270 lines for Limousin and Charolais, respectively.

Genetic evaluation model: The model chosen for the Interbeef joint genetic evaluation is a multiple trait animal model based on raw data and considering each country as a separated trait (Phocas et al, 2005). Along with country specific fixed effects, the model included direct and maternal genetic effects as well as permanent environment effects. Across-country genetic parameters were estimated in 2014 by Pabiou et al. (2014). The MiX99 software was used by Interbull Centre for breeding value prediction (Lidauer et al., 2011) while reliability computation used MTEDC5 package (Sullivan, 2011). The genetic evaluation was run in September 2014. Interbeef breeding values can be published at country level pending the following publication rules: breeding values can be published only i) for animals with Interbeef reliability (reliability=accuracy ²) greater than 0.50 and with more than 25 progenies with records taken into account in Interbeef evaluation or ii) for animals already publishable in the country.

Analysis: In order to assess the benefits of the Interbeef collaboration, pseudo national genetic evaluations were mimicked at Interbeef level by setting between-country correlations to zero in the Interbeef model. That way, only national performances were taken into account in each country breeding value prediction. The comparison of sire ranking and reliability between Interbeef and pseudo national evaluations were used to assess the different impacts of a joint genetic evaluation at country level. The software used to run the analysis was SAS v9.1 (SAS, Cary, NC, USA, 2003).

Results and Discussion

The IDEA web-based database was launched in 2010 by Interbull center and allowed member countries to upload and automatically check for potential error pedigree or performance files. This automatic process ensured that a maximum number of correct records entered the InterBeef genetic evaluation by minimizing errors due to mis-identification or duplicates. As expected for Charolais and Limousin breeds, a large part of the data comes from France (Table 1). The United Kingdom and Sweden had the second largest populations of Limousin and Charolais, respectively. The amount of data sent by Ireland represented only 15% of the data used nationally as the majority of performances are collected on commercial animals.

By taking into account all information available on animals and relatives across all member countries, Interbeef can provide more accurate genetic evaluation results (compared to national evaluation) to breeders.

Larger choice of breeding animals: The Interbeef joint genetic evaluation provides breeders of the different member countries with a larger spectrum of breeding animals (Table 2). Every country was offered a wider panel of publishable bulls. On one hand, the country with the narrowest panel of bulls in their national evaluation will see the greatest increase in the size of their publishable bull list (e.g., SWE, CHE). On the other hand, FRA which is holding most of the purebred data in Limousin had a modest input (14%) of foreign sires in its publication list.

Reliability increase: Using all available sources of performances outside the national supply is hugely beneficial in terms of reliability (Figure 1).

Across all country, the general median reliability increase was +0.44 and general maximum increase of +0.89. This reliability increase depends on the country and the type of bulls considered.

Figure 1 shows that reliability increase concerns mainly foreign bulls but can also occur for local bulls via the genetic linkage between countries.

For example, the average increase of reliability for Limousin bulls with at least 10 progenies was +0.47 for GBR and +0.06 for FRA.

Interbeef ranking: Table 3 gives, for Limousin top 1000 best bulls in each country scale, the origin of these bulls associated with the best rank (only for bulls with at least 10 progenies and associated reliability higher than 0.30).

In all country scales, a large majority of the top 1000 best bulls comes from France (between 72% and 96%) followed by United Kingdom (4% to 25%).

However, bulls from all participating countries can be found in all scales (except in French scale).

A direct benefit of Interbeef is to allow breeders to have access to a much larger panel of new favorable bulls selected on EBV specific to their country scale.

More than 90% of the bulls presented in Table 3 originated from FRA across all scales, and the best ranked bull was also FRA across all scales. However, GBR bulls were ranked in the top 10 in CZE, ESP, GBR, and IRL scales, and has a bull ranked 97 in the FRA scale.

Since genetic correlations between countries are not equal to unity (Pabiou et al., 2014), re-ranking between countries will occur. Rank correlations between Interbeef and pseudo national evaluation showed that including data from other countries had a great impact on bull ranking in each country (Table 4). As expected, rank correlation in FRA scale is the largest (0.96) allowing for limited re-ranking of bulls in the French scale. Accounting for all available performances across countries in a single genetic evaluation has had the largest impact on IRL and CZE.

Conclusion

Over the past years Interbeef has met and overcome challenges, and successfully answered several technical and research questions. Interbeef is a relatively new entity offering across country genetic evaluation to its member country and beef breeds. By uploading pedigree and performance files into the Interbeef centralised database housed at the Interbull Center, the participating countries enjoy full benefit of data sharing through pedigree consolidation and an access to a large database of phenotypic records.

The 2014 genetic evaluation showed large benefits of Interbeef genetic evaluation collaboration in terms of individual animal reliability increase but also by providing member country with a bull ranking accounting for country specificities.

Interbeef will help the breeders to better choose their breeding animals in their own country but also abroad on the same scale. It deals at the moment only with pure bred animals and weaning weights. However, research is ongoing to include cross-bred animals and expand Interbeef trait list to calving, slaughter and female fertility traits.

References

IDEA Interbull (2013). https://idea.interbull.org/.Lidauer, M., Matilainen, K., Mäntysaari, E. and Strandén, I. (2011). Mixed model equations solver MiX99 – Release VI/2011. MTT Agrifood Research, Finland. http://www.mtt.fi/BGE/Software/MiX99.

Pabiou, T., Nilforooshan, M., Laloë, D., Hjerpe, E. and Venot, E. (2014). Across-country genetic parameters in beef cattle for Interbeef weaning weight genetic evaluation. Proc 10th WCGALP.

Phocas, F., Donoghue, K. and Graser, H.U. (2005). Investigation of three strategies for an international genetic evaluation of beef cattle weaning weight. Gen. Sel. Evol. 37: 361–380.

Sullivan, P. (2011). ftp://cdn.ca/pub/outgoing/general/mtedc_5e.zip

Venot, E., Fouilloux, M.N. and Laloë, D. (2008). Interbull Bulletin, No. 38, 41–48.

Venot, E., Fouilloux, M.N., Forabosco, F., Foghs, A., Pabiou, T., Coffey, M., Eriksson, J.-A., Renand, G. and Laloë, D. (2009). Interbull Bulletin, No. 40, 61–67.