Making EPDs Work in Your Herd

 by Brad Wright

Fall bull sale season is in full swing and makes this a good time for a refresher course in utilizing EPDs to help you pick your next bulls.  For many of you, this will be pretty redundant to things you have heard and read in the past, but it sure doesn’t hurt to make sure you are using the best breeding predictor available correctly.

What is an EPD?  EPD stands for Expected Progeny Difference.  An EPD is a value that represents an animal’s genetic breeding value for a given trait.  As the name implies, an EPD estimates the difference between the averages of 2 animals’ progeny for any given trait due to the genetic differences in the 2 animals.  An EPD is a comprehensive statistical tool that combines actual performance and contemporary group data with pedigree data, related animal data, and in some breeds, DNA markers.  An EPD is the best predictor of breeding value available to us and should be used for selection over actual data of any animal.

How is an EPD used?  In defining EPD, the word “difference” was used several times.  This word is key to understanding how to use an EPD.  An EPD is a means of comparison to evaluate the difference between an animal and either a) another animal or b) breed average.  To calculate a difference, you must have two values.  I bring this up only to drive the point that a set of EPDs on an animal with no other knowledge is useless.  An EPD cannot be used by itself and it does NOT relate to any value of actual performance (i.e. a BW EPD of 1.8 does NOT mean the calves will be 75 lbs).  An EPD can only be used to estimate the difference of the average of two animals’ progeny.  When you hear someone say something about “good” EPDs or “bad” EPDs on an animal, they are generally comparing to breed average.  Breed average can be found on most breed association websites or printed in most sale catalogs.  The important thing to remember about Breed Averages is they are most often NOT ZERO.  This one simple fact has disappointed many unknowing producers when the realized their +15 WW bull was actually 9 lbs below breed average for WW (IBBA – Average WW 24).  Also, EPDs are calculated within breeds and cannot be compared across different breeds.

BULL A
WW EPD – 24

BULL B
WW EPD – 34

Based on the above information, we would expect the average weaning weight of the calves sired by BULL B to be 10 lbs greater (34-24=10) than the average weaning weight of the calves sired by BULL A, given the same environment and comparable genetic makeup of the dams.

What do you need to help your herd?  Back on this discussion of “good” and “bad” EPDs, this cannot be based solely on the bulls ranking within the breed, and will not be the same for every breeder.  Know your herd and know your environment and utilize the EPDs to best fit your needs to reach your end product goals.  A “good” Milk EPD for a producer in Northern Georgia with 50 inches of annual rainfall may be ABOVE +20.  A “good” Milk EPD for a producer in Southern Arizona with 6 inches of annual rainfall may be BELOW +11.  Even though EPDs are represented with + and – and percentile rankings, doesn’t mean that + is good and – is bad, and 5th percentile isn’t necessarily better than 50th percentile if it doesn’t fit your herd, your environment and your end product goals.

How accurate is an EPD?  Ever wondered what that little number was printed below and EPD?  The accuracy of an EPD, sometimes labeled ACC, is often times printed below the EPD.  The accuracy will be a number between 0 and 1 and is the statistical representation of the confidence in the EPD.  The closer the accuracy is to 1, the more reliable the EPD is and the less likely the EPD will move or shift.  Accuracy increases as progeny are recorded.  A common problem when buying bulls is that most will be unproven bulls with a large percentage having never sired a calf.  This does add a degree of difficulty when evaluating young bulls.  With non-parent bulls, the highest accuracies available will be .25 to .35 and these accuracies will only be available on bulls with data collected in large, high quality contemporary groups with at least some proven AI sires represented in the group.  This magnifies the need to buy bulls from breeders who are diligent in collecting and reporting proper data and contemporary groups.

Do EPDs work?  Figure 1 shows the genetic trend for the growth traits in the Brangus breed since 1970.

EPD Trend Lines for Growth Traits

EPD Trend Lines for Growth Traits

These trend lines are similar to those found in other breeds as well.

 

With the use of EPDs, breeders have been able to increase Weaning Weight and Yearling Weight while keeping Birth Weight essentially the same.  EPDs do work when used appropriately.  They can even work too well if only selecting for one trait.  There will be constant updates to the EPD analysis in the future, whether it is the addition of DNA data to increase the accuracy of these unproven animals, or the creation of new EPDs to help producers better select for economically important traits like fertility, age of puberty or feed efficiency; but quality progeny data will always be the backbone of EPDs because “proof is in the pudding” (or in the cattle industry, “proof is in the progeny”).

Importance of Records, Individual Selection and Breed Choices in Crossbreeding Systems

Joe C. Paschal

Last month I ended my article with some comments on whether or not crossbreeding was still economically relevant in today’s beef market based on an article written by Dr. Nevil C. Speer of Western Kentucky University. One interpretation was that it is not, that in today’s value based marketing system (read by some as “marbling score driven”), the value of carcasses with high marbling scores somehow outweighs the overall increase in lifelong performance of crossbred cows and their crossbred fed offspring. I doubt that that is case but still crossbreeding is not free. There is a cost to produce the purebreds that form the basis of the cross but this cost is usually passed on to the producer who purchases these females. If both male and female crosses are equally desirable as feeders or replacement females then the additional cost can be recouped equally. If there is a significant discrimination between the crosses based on gender then the more favorable gender will have to pick up more of the cost of production.

A well known example is the production of British x Brahman F1 steers and heifers. The steers will bring a lower price per pound than some other crosses (but less so today than a few years ago) but the heifers will command a premium as replacements. Discounts for crossbred bulls and steers depend on the demand for specific types. In the early years of the Brangus breed a good 3/4 Brahman x 1/4 Angus bull could be worth quite a bit of money to a breeder with Angus cows wanting to produce Brangus cows. On the Gulf Coast of the US there are places where 1/2 and sometimes 3/4 Brahman bulls are very useful, but these represent only a small fraction of the beef cattle produced in the US, but you get my point.

Most of the advantages that are pointed out in different crossbreeding systems come from the increased productivity of the crossbred cow, mainly attributed to heterosis. In animal breeding we define heterosis or hybrid vigor as the difference in level of performance in a trait when comparing the crossbred offspring to the average of the parents. There are parental breeds that are high performing in one trait or another (adaptation, growth, longevity, etc) but no breeds that I know of excel in ALL traits. This is why selection of breeds and individuals within breeds to make the cross is very important. It is equally as important as selection of individuals in a straightbred or purebred herd. In most discussions of why crossbreeding and hybrid vigor is beneficial (or is not) to commercial producers, the selection of breeds (and what they bring as “breed effects” to the cross) and the importance of individual selection are often glossed over or ignored in the larger discussion. I am not going to discuss breeds here, I think most of us have a good understanding of what different breeds have to offer, either as purebreds with some Ear influence (or not) or in a crossbreeding program with Ear influence (or not).

The purpose of selection is to increase, reduce or to maintain performance of the traits that you as a breeder or a producer have an interest. These traits may increase your income, reduce your costs or just lower your stress. These traits may be relatively easy to select for (or against) or not depending on the heritability of the traits, the amount of variation in the breeds being used, and the source and quality (accuracy) of the records that are available. The more information that you can bring to bear on your selection decision, the more likely your breeding or crossbreeding program will be successful. There are a number of efforts underway in the different Eared breed associations to encourage their members to increase the quantity AND quality of the production records that are being collected and reported to them. In this age of genomics and genetic markers we tend to forget that these are not “silver bullets” but aids to selection. The use of these new tools will improve the prediction of an animal’s genetic value, more so for young animals without progeny records. In the future these genomic tools will account for more genetic variation of these traits and others of significant interest (such as longevity, reproduction and health) but right now their best use is to improve the accuracy of the EPDs for animals with no (or few) offspring. It might change their EPD for the trait but the greatest impact will be on the accuracy or reliability of their EPD. The problem is that many of the genetic tests are for carcass merit traits, not for traits earlier in life affecting productivity and economics at the cow calf level.

Collecting Data at JDH. Picture Courtesy of Dr. Paschal.

Collecting Data at JDH. Picture Courtesy of Dr. Paschal.

Now, as in the past, the Eared breeds and their breeders have quickly and roundly embraced new programs and technology. They were among the first to utilize (and continue to do so) ultrasound for carcass merit, to participate in feedyard and carcass merit (including tenderness) data collection programs, and the use of genetic markers and other genomic tools. But these technologies should be used IN ADDITION TO the time consuming and less cutting edge documentation and reporting of calving, weaning and yearling data on all animals in your herd. I am familiar with the concept and the drudgery of record keeping and reporting but it is very important to collect to ensure that the accuracy of your animal’s actual and adjusted within-herd performance records and your breed’s genetic predictions or EPDs.

Collecting ultrasound data at JDH.  Picture courtesy of Dr. Paschal.

Collecting ultrasound data at JDH. Picture courtesy of Dr. Paschal.

Eared cattle breeds are raised in a wide range of environments (not just climatic conditions) which affects their performance. If complete records are not collected, this nongenetic environmental influence may not be completely accounted for in genetic evaluations. In addition, there can be considerable “selection bias” injected into the records as only the records of the more/most desirable animals are sent. This bias tends to affect bulls more than heifers since fewer bulls are kept and registered. Selection bias will make the remaining animals look better than they really are because the lower performing (or less desirable) animals have been culled and their performance records were not collected. We all spend a lot of time and money breeding and raising good cattle for our customers, feeder cattle as well as replacements, and we need to collect and report as much data as we can, sometimes even if it is a little onerous. Just do it!

 

Dr. Paschal is a livestock specialist for the Texas A&M AgriLife Extension Service and is based in Corpus Christi, Texas. He can be reached at (361) 265-9203 or j-paschal@tamu.edu