Expected Progeny Differences are the most beneficial and accurate tool we have in estimating the true genetic variation among cattle of the same breed. This tool separates genetics and environment giving an estimate based on additive gene effects. They are the most proven way at this point to go beyond phenotypic observation and gauge the animal’s true potential in passing on genes. Expected Progeny Differences are calculated via matrix algebra and the use of complex, statistical models. They take into account animal relationships, environment, and population size.
Expected progeny differences (EPD’s) provide estimates of the genetic value of an animal as a parent. Specifically, differences in EPD’s between two individuals of the same breed predict differences in performance between their future offspring when each animal is mated to animals of the same average genetic merit.
Hereford EPDs are calculated by the Agriculture Business Research Institute in Armidale, Australia using complex statistical models and equations. Best-Linear Unbiased Prediction (BLUP) is the method in which this is achieved. These statistical models use all known information on the animal to calculate its EPD. This information includes performance data (i.e., weight records) on the animal itself, information from its ancestors (sire and dam, grandsire, great grandsire, maternal grandsire, etc.), collateral relatives (brothers and sisters), and progeny (including progeny that are parents themselves). These records are adjusted for factors such as age, sex, and dam age to allow fair analysis. Additionally, genetic merit of mates is accounted in evaluating progeny information. The statistical analysis used for EPD calculation also accounts for the effects of environment (nutrition, climate, geographical location, etc.) that exists between herds. These environmental effects can be estimated via the widespread use of A.I. to establish “connectedness” between herds. A.I. sires create genetic links between herds with different environments and serve as the foundation of performance data and index calculation across herds
Accuracy values are published for EPD values reported for any animal. Accuracy can be defined as the relationship between the estimated EPD and the “true” EPD of the animal. This relationship is expressed as a number between 0 and 1. Accuracy is primarily a function of the amount of information available to calculate an EPD for any given trait. As the volume and quantity of records used in the estimation of an EPD increases, so does the confidence we have that the EPD is estimated correctly.
Accuracy is a function of the number of producers who submit performance records and the number of animals in each herd that have their records submitted. To achieve high accuracy values data needs to be submitted on replacements, culls, high quality and low quality animals alike. EPD’s do not allow the best to rise higher because they stand out in your herd. In fact they are penalized because the animals they outperformed were left out of the analysis, therefore making their data closer to average than would be the case otherwise.
A few things to think about:
a. The only way to increase consistency in a set of calves is through the use of inbreeding. Higher accuracy EPD’s have or more favorable breeding values have no bearing on this.
b. The EPD of an animal for birth weight or any trait is more reliable that the actual performance measurement on the animal.
c. Bulls will have higher accuracy values than females for the simple reason that they pass genes onto a far greater number of progeny. This fact can’t even be overcome with the use of embryo transfer in most cases either.
d. Older sires obviously have higher accuracy values because of the time they have had to affect more progeny. This increased accuracy comes at the cost of generation interval or turnover in the subsequent progeny.
e. Indexes are only useful on a within herd index because the index does not account for genetic differences between herds.
Many people are still skeptical of EPD’s, yet with the upcoming molecular genetic technology coming on stream an understanding of these values being integrated with molecular markers will be more important than ever. At this point I don’t think there is enough information coming from genetic tests of marbling, tenderness or any other trait to give a comprehensive look into an animal’s genotype. Several of these tests are being marketed via one gene that is connected to a certain trait, when in reality tens or hundreds of other genes may also be affecting the same trait. Molecular DNA testing and marker assisted selection has long-term promise, but right now I don’t think it is an accurate let alone profitable predictor of animal performance.