• If you are having problems logging in please use the Contact Us in the lower right hand corner of the forum page for assistance.

EPD Basics

M

Mike

Well-known member
Joined
Feb 10, 2005
Messages
28,480
Location
Montgomery, Al
The Basics of EPD's


It's hard to believe in this day and time there still exist a good deal of mistrust and misunderstanding when it comes to Expected Progeny Differences (EPDs). As a purebred breeder, it is critical that you not only understand the application of EPDs, but also how they are calculated. I don't mean the specific statistical methodology, but rather how the data you submit affects the EPDs of your cattle. If you read no further, remember this, EPDs work, but only with complete data submission and proper contemporary grouping.



EPDs are a prediction of the genetic transmitting ability of an animal. They tell you in a given environment how one animal will compare to another; they do not estimate actual weights, only weight differences. An animal's genetic value does not change over time, but the amount of information we have to estimate that value does. The difference in that animal's true genetic value and the predicted value is known as accuracy. EPDs with a higher accuracy have less chance of changing far from their predicted value. It is also important to realize that the EPD value is reflective of the average of an animal's calves and could represent a wide range.



The calculation of EPDs incorporates the relationship between all animals in the analysis, records submitted on all related animals, and the value of correlated traits (the manner in which traits affect one another). An animal's own records also play a role in their EPD. This individual performance is all based on their ratio within their contemporary group.



Contemporary grouping is how environmental differences are removed from the genetic equation. A given sire's calves will not weight the same at birth in Texas versus North Dakota and it would not be fair to compare the weight of a calf on feed with one that isn't. A contemporary group is a group of animals raised together under the same conditions. When you report weight information it is critical to properly define contemporary groups. Only calves managed similarly should be assigned the same contemporary group code. The system will further break contemporary groups by sex and age.



Maintaining large contemporary groups is vital to the effectiveness of the genetic evaluation system. Ratios are calculated within contemporary groups and this is how the genetic differences are determined. Single-animal, or single-sire contemporary groups are of little value in determining genetic differences. The best way to ensure an accurate genetic evaluation is to keep contemporary groups large and properly identify them using management codes and try to use a mixture of both high-accuracy reference sires and young sires in each contemporary group. EPDs are by far the best prediction of performance we have, but they require complete data submission and proper contemporary grouping to maximize their effectiveness.
 
Am I to understand, that EPD does not stand for Ear Potential Development? :?

And now I can't even remember who 'splained to me, that's what it meant, so I can't even yell at them!!!!!!!!

This sounds like a good fair entry for Al, for his Cattle Breeding Project this year.........we may have to do a bit more research on it.

Was it MCG that did one on EPDs last year for their fair, or do I have my wires crossed again? :roll:

Interesting topic, anyway, Mike! :wink:

As HayMaker would say.............Good Luck!

:wink: :lol: 8)
 
Andy said:
EPD's are only as good as the people that report the numbers. That is why I ONLY work with honest, trustworthy breeders.
Click to expand...

Andy, It is ALWAYS good to deal with honest people. Always.

But can you explain to me how I could fudge the numbers on my calves' data and somehow skew my EPD's to my advantage?
 
Inaccurate reporting of wieghts and or ages. I have even heard of breeders selling a bad cow and then keep reporting bad numbers on her just to make the rest of the calves in the contempary group look better. If you don't sell to other reg breeders you can skew EPDs big time. There are people out there that will srew you over just to make a buck, some of them will even shot themselves in the foot to do it.
 
Andy said:
Inaccurate reporting of wieghts and or ages. I have even heard of breeders selling a bad cow and then keep reporting bad numbers on her just to make the rest of the calves in the contempary group look better. If you don't sell to other reg breeders you can skew EPDs big time. There are people out there that will srew you over just to make a buck, some of them will even shot themselves in the foot to do it.
Click to expand...

Turning in a high weight on one animal will bring his ratio up and the ratios down for the rest. What good is that doing you? Would it make sense to sacrifice the whole contemporary group for one animal? Hardly.

EPD manipulation is a myth.
 
all the numbers in the world can be put in front of me, and I am still gonna buy with eye appeal.
 
jigs said:
all the numbers in the world can be put in front of me, and I am still gonna buy with eye appeal.
Click to expand...

As you should. EPD's are the only genetic predictor we have though. They are to be used only in conjunction with other attributes and assessments.
 
Back to the original thread:

Expected Progeny Differences Are Comparable to Realized Progeny Differences



University of Kentucky and University of Florida researchers conducted a summary of many previous studies that compared expected progeny differences (EPDs) with actual realized progeny differences for various beef cattle traits. The summary involved data from six breeds: Angus, Brangus, Charolais, Limousin, Polled Hereford, and Simmental. Traits were: birth wt. (BWT), weaning wt. (WWT), yearling wt. (YWT), marbling (MAB), carcass wt. (CWT), fat thickness (FAT), loin eye area (LEA), percent lean yield (% LY), milk (MLK), maternal (MAT), and scrotal circumference (SC). Following is a summary: • Realized progeny differences agreed well with EPDs for BW and WW, but for YW, realized tended to be greater than EPD, especially when YW was the primary sire selection criterion.


Relative to sires with low EPDs for MAB, CWT, FAT, LEA, and % LY, sires with high EPDs sired progeny with higher MAB scores and greater CWT, FAT, LEA, and % LY. Sires with high EPDs for MLK and MAT sired daughters that produced more milk and weaned heavier calves than sires with low EPDs. Sires with high EPDs for SC sired daughters that reached puberty earlier.



The authors noted that the similarity between expected and realized progeny differences should be greater when high EPD accuracy sires are used, but when a small number of low accuracy yearling bulls are used, expected results may not be realized (F.A. Thrift and T.A. Thrift. 2006. Prof. Anim. Sci. 22:413).



Source: Dr. Rick Rasby, Professor of Animal Science, University of Nebraska - Lincoln, Lincoln, NE
 
MikeC:

Who wrote the first article you posted?

This line is incorrect, "Ratios are calculated within contemporary groups and this is how the genetic differences are determined."

EPD are not calculated on ratios. They are calculated with the adjusted weights. Not a big mistake, but lots of folks don't realize that EPD have nothing to do with ratios.

Badlands
 
Hey Mike, as an "old fashioned" breeder, one of the things that has never been explained to me in adequate detail is how EPDs can cross contemporary groupings and result in any usuable information. A fine example is the Texas vs North Dakota birthweights or fed calf vs roughed calf growth.

Lets say in Texas, a bull ends up with a +2 BW in a large contemporary group. When his semen is used in North Dakota in a large comtemporary group, he ends up at +4 BW. How do the EPDs reflect those differences?

What if a sire (and his offspring) is used almost exclusively in the south for many years in large groups, resulting in fairly accurate EPDs? How do you know that when that bulls semen moves north what you'll be getting?

Rod
 
References sires, DiamondS.

There are enough sires used in many different groups that they tie everyone together, unless you are distinctly disjoint with a completely closed herd.

If the reference sires' calves weigh 500 pounds in group 1 and sire B's calves weigh 525, then in group 2 the references sires calves weigh 525 and sire C's calves weigh 500, we know that Sire B produces more weight than sire C and the reference sire.

I won't bog you down with the heritability of the trait which ends up making the EPD less that the 25 pound differences.

The point is the reference sires tie the groups together across the wide ranging groups.

EPD aren't perfect, and there can be some interaction between sires and environments, but for the most part, they aren't that significant. They don't cause the animals to re-rank too much.

Badlands
 
EPD are not calculated on ratios. They are calculated with the adjusted weights. Not a big mistake, but lots of folks don't realize that EPD have nothing to do with ratios.

Badlands[/quote]

Dear Badlands what about this case. There are two 3/4 brothers. The sires are the same and the dam's sire is the same. Here are there EPD's

bull one1
3.5 birth
45 ww
84 yw
17 milk


0.05 marb
0.44 rib
0.00 fat


bull two
5.4 birth
46 ww
89 yw
14 milk

0.03 marb
0.32 rib
0.00 fat

Bull one has an actual bw of 80 at a ratio of 97. His dam has an actual bw of 80 and a ratio of 106. Her bw epd is a 3.5.

Bull two has the same sire.
His actual bw epd is 97 at a 115 ratio. His dam has an actual bw of 81 at a 105 ratio. Her bw epd is a 2.3.

I called the RA office and asked why they were so different and they said it was the birth weight ratio that made bull two's bw epd so much higher.

I may add that bull one has calves on the ground and bull two is a yearling.

The RA office said that weaning weight is required and all other weights are optional.

have a cold one

lazy ace
 
The actual weight differences means the ratio would be higher or lower on the groups of calves.

The amount the ratio is higher or lower is not what is measured, it is the actual pounds.

With enough information, EPD's will always correct, however they can be skewed in the short term.

The famous (infamous) case of a breeder that recorded mulitple calves from a popular A.I sire from cows that were dead or no longer used to produce purebred offspring. He "weighed" the phantom calves and they all were bigger at birth but smaller at weaning and yearling than his calves from his homegrown bull.

Of course when the bull was used in other herds, and against real offspring of the popular bull, the real EPD's would prevail. In the meantime hundreds of thousands of dollars could have been made from syndication and semen sales from his phony numbers.

How about the bull that scanned 7% imf at a year, generating a huge EPD, but as of yet none of his offspring have scanned above average for marbling, and his EPD has dropped like a stone in the river. That initial EPD helped sell the bull for 70K. How would you manipulate his scan? He must have actually scanned 7%, but if he was fed better than the bulls in his contemporary group, the higher fat he scanned compared to nutritionally less fat of the other bulls would boost the EPD.

Those examples are the reason if you want an EPD that has value, look for accuracy, and track the parentage EPD's as well.

One other thing, just because an EPD is high or low for a certian trait doesn't mean all offspring of that animal will be high or low in that trait. EPD's for an animal are on a bell curve. The majority of the offspring will fall in the relatively narrow range of the center of the bell, some will be significantly higher or lower. The steeper the bell the more consistant the animal.
 
lazy ace:

Jason hit it right here, "The actual weight differences means the ratio would be higher or lower on the groups of calves.

The amount the ratio is higher or lower is not what is measured, it is the actual pounds."

We talk about ratios because that is how we have done things traditionally, we do it by convention. What BLUP procedures use however, it the adjusted weights.

In the case of your animals, it considers the performance of the dams too. Plus all the other relationships. It is not something you can do in your head, trust me. Maybe for one year, but you can't dynamically handle all the effects over time of having different sires and dams in different groups over several years. Now, that isn't and insult at all. Nobody does this sort of math in their heads, that's all.

Your numbers might be looking a little odd to you, but part of how the EPD end up where they do is from the "competition" in the contemporary group. In other words, what other sires were in there?

I can see that in the case of the dams they did not originate in the same contemporary group, so they had different sire and dam competition within those groups resulting in differing EPD.

ONe thing to remember too, is that we never get enough progeny on most cows to get much accuracy, so they are pretty much stuck with their Interim EPD for life. If they were born with a high BW, but transmit low BW, they EPD will seldom change enough to truly reflect that.

Some call it a flaw in the system, but it's simply a numbers game, and cows don't raise enough calves to move their interims too much. It is odd to hear people argue for cows having EPD that should be able to change, while saying bulls EPD change too much. On one hand they aregue for more volatility, on the other hand they argue for less, forgetting that the EPD are all pooped out by the same system.

Know your cows, use high accuracy bulls.

Badlands
 
Jason said:
The actual weight differences means the ratio would be higher or lower on the groups of calves.

The amount the ratio is higher or lower is not what is measured, it is the actual pounds.

With enough information, EPD's will always correct, however they can be skewed in the short term.

The famous (infamous) case of a breeder that recorded mulitple calves from a popular A.I sire from cows that were dead or no longer used to produce purebred offspring. He "weighed" the phantom calves and they all were bigger at birth but smaller at weaning and yearling than his calves from his homegrown bull.



Of course when the bull was used in other herds, and against real offspring of the popular bull, the real EPD's would prevail. In the meantime hundreds of thousands of dollars could have been made from syndication and semen sales from his phony numbers.

How about the bull that scanned 7% imf at a year, generating a huge EPD, but as of yet none of his offspring have scanned above average for marbling, and his EPD has dropped like a stone in the river. That initial EPD helped sell the bull for 70K. How would you manipulate his scan? He must have actually scanned 7%, but if he was fed better than the bulls in his contemporary group, the higher fat he scanned compared to nutritionally less fat of the other bulls would boost the EPD.

Those examples are the reason if you want an EPD that has value, look for accuracy, and track the parentage EPD's as well.

One other thing, just because an EPD is high or low for a certian trait doesn't mean all offspring of that animal will be high or low in that trait. EPD's for an animal are on a bell curve. The majority of the offspring will fall in the relatively narrow range of the center of the bell, some will be significantly higher or lower. The steeper the bell the more consistant the animal.
Click to expand...

Jason I have a heifer that scanned Actual Pct Imf 6.21, Adj Pct Imf 5.90. This heifer was not on any grain just grass. She weighed 473 at 238 days of age. She is a registered Angus and I have some high hopes for her offspring with respect to carcass to grade well but in reality my opinion is they may or may not, even when bred to a high accuracy high Imf bull. What do you think?? Have you used any bulls that seem to pass on consistently high Imf numbers to their progeny? I also might add although it seems I am chasing one epd here, I'm not, as I am watching the others as well . Any suggestions?
 
How much of an affect does creep feeding bulls have on WW epd's. When weining weights are submitted they don't add that info do they. When looking at sires to use for A.I., I see those high WW Epd's it leaves alot of questions in my mind. Just how common is creep feeding with registered breeders? (angus)
 
TSR what was her backfat scan?

The imf scan is pretty good, but heifers will lay down more marbling. Steers will be slightly less (I don't know the %) under the same conditions and steers are 30% fatter, and about 1% more imf than bulls under the same management.

The other thing is when an animal is smaller the imf numbers will appear greater, but if that animal was fed more, and the ribeye was bigger, would the animal have enough marbling in the bigger area.

That is the reason implants hurt marbling scores, they tend to make ribeyes bigger and to get the same % imf, it takes more fat.

Shrinking frame size has the effect of increasing imf % wise.

Horseless, creepfeeding will affect milk epd's if not reported. I am not sure how prevalent it is but I never use it, I opted for early weaning when I was having my sale. I know some of the sales where calves are 1100 pounds plus in Nov/Dec don't lack any nutrition.
 

Latest posts

Back
Top