To EBV or not to EBV

By Jack Allen, Technical Director,ABRI

Why do performance recording?
Comparison of animals has been used in breeding programs for centuries. Whether it is done in the show ring, the paddock or over the scales – animals have been selected on their perceived advantages. The problem encountered by many breeders is how to allow for different management, feed quality and quantity, animal ages, etc in assessing if one animal is better suited to their breeding program than another. It gets even more difficult when you are presented with lots of animals from different and diverse farms.

Performance recording is just measuring some of the more commercially orientated aspects on the animal and then comparing them with other animals within the herd. Over the years, more sophisticated ways of analyzing the information have been developed, allowing animals to be compared across herds as well as within herd.

Calculating Estimated Breeding Values (EBVs) necessarily involves using sophisticated statistical routines over the information supplied by breeders to the Society. These EBVs are aimed at estimating the genetic potential of the animals – what is passed on from one generation to the next, rather than how good the paddock was where the animal was raised.

EBVs allow: You to compare animals within your herd; You to compare animals from outside your herd with those that you have; Others to compare your animals to theirs (should they buy your animals - or your neigbours?).

How much is a good bull worth? What is the cost of a bad bull?
The sires you use in your herd are there to produce calves – not show rosettes, not EBVs - these are just by-products. Show rosettes may tell you how the animal compares to other animals at this point in time – rosettes do not indicate how good the progeny will be. EBVs try to predict what the progeny might be like as well as an emphasis on the current animal for the traits recorded.

From a stud herd perspective, a “good” bull will not only produce good calves for market, but also good replacement breeding animals – either for your herd or others. Hence the legacy of a “good” bull lasts for generations.

Similarly, the legacy of a “bad” bull can also last for generations. EBVs are designed to help you, and others, select the “good” bulls. You still need to consider the animal, his structure and the purpose you want to use him for. EBVs do not select the bull – EBVs help rank animals for commercial traits as an aid to help you to select the bull.

An interesting point to note is that at the recent Perth Bull Sales in November, 36 bulls sold that were in the top 10% of the Breedplan Terminal Sire Index sold to an average of £5400. 39 Bulls that were outside the top 10%sold to an average of £2700 and 8 out of the 39 made more than the sale average.

Performance data – it’s your call
While we run “sophisticated statistical routines” tocalculate the EBVs, the requirement for the analysis is simply good comparisons between animals in your herd. All you need to do is manage your herd to get good comparisons between animals - you don’t need to worry about the complicated statistical stuff.

Both small and large herds need to get good comparisons between animals. It is usually easier to get good comparisons in larger herds, although they can still have problems.

No matter how smart the computer is, the system needs to know which animals have been managed on farm together and what their pedigree and performance is. The whole system relies on you. Not only do we need you to tell the system that calf “A” weighed 20kg heavier than calf “B”, but also when the calves were born, who their sires and dams are, etc and whether you think it is “fair” to compare these calves’ performance. The system is not necessarily interested in how well you feed the animals, rather how well the animals performed for the feed available. Hence it is “unfair” to compare sick calves with healthy ones. It is “unfair” to compare animals given special preparation for a show or sale with those that were not.

Why are good comparisons between progeny necessary?
Generally speaking, the more “fair” comparisons there are between progeny in your herd, the better the analysis can rank these animals in your herd. Just as importantly, the performance of the progeny is also a reflection on their parents. Just as we rank the progeny within “fair” comparisons, we also can rank the parents of those calves based on the calf performance. The analysis combines the rankings of the sires and dams from lots of “fair” comparisons to better estimate the breeding values of the calves, their parents, grand-parents, etc within your herd and potentially across the whole data base.

What can I do?
Consider the following ideas to help improve the comparisons of animals within your herd:

At mating Try to have just one mating per year – or at least as few as possible. The more mating periods you have in a year, the harder it is to get good comparisons between animals born around the same time.

Restrict the mating period – the tighter the mating period, the more calves that are born around the same time which can be directly compared. Animals are compared when born within 60 day windows – the more comparison windows in your herd the bigger the likely “pane” in the butt for the analysis.

Use at least 2 sires – direct comparison of progeny of different sires is important. Consider using AI, particularly in smaller herds where it may not be economical to have more than one bull. Not only will this improve the comparisons within your herd, but it will also improve the comparison to other herds and while giving you access to better genetics not generally affordable in a herd bull.

At birth
Record sire, dam, date of birth, sex, single/twin If the calf is by embryo transfer, flag it as such and record the recipient dam (tag, breed and year the recipient was born).

Record calving ease and birth weights whenever possible.

For performance data
Manage groups of animals together wherever possible and for as long as possible.

Measure the whole group of animals on the same day.

Measure all the animals in the group – particularly for scanning and scrotal (not just the sale bulls).

Try to get at least 2 weights plus scanning and scrotal measures on the whole group of animals prior to selecting show/sale animals.

If you need to sub group animals, use calf sex as the first criteria (males are not directly compared to females in the analysis anyway).

If you do treat animals differently, then record this as a management group difference. Remember we want fair comparisons within a group.

If you are giving extra feed to sale animals, etc, then you must subgroup these from the other animals not being fed. However, try to ensure you feed more than one animal and preferably from a range of sires. Remember that the progeny comparisons are also a reflection on the parents.

You control the analysis of your data
The information you supply to the Society determines how that information is analysed. By following a few of the basic principles outlined above, you can help maximize the usefulness of the data in the analysis and get better value for your performance recording efforts.

In particular, reduce your mating period, manage animals in as large a group as possible and record all the traits on all the animals.

DON’T DELAY JOIN BREEDPLAN TODAY GREAT FIGURES ATTRACT BUYERS

For further information on Breedplan, how to join, costs etc please contact Barbara Webster on Telephone: 01738 622478 or email barbara@breedplan.co.uk or contact the Society office Telephone: 02476 696513/ Fax: 02476 696724 or email information@britishsimmental.co.uk