M.T. See, Ph.D.

Department of Animal Science

North Carolina State University


Today's pork producer is confronted by a dynamic swine genetics market. The prevalence of carcass-merit buying programs has fueled a need for improved genetics that is being answered by an ever increasing number of seedstock sources. A confusing problem for producers is interpreting, understanding and comparing the various types of performance information that are offered as genetic fact.

A performance record is a measurement of an animals phenotype and is made up of two components: genetic and environmental. Genetic improvement through selection operates only on the genetic component of the animals record. The environmental component is not passed from parent to progeny and, therefore, needs to be accounted for when determining the genetic value of an animal. Some of the environmental factors such as parity of sow or sex can be accounted for mathematically. However, other factors such as health, management and feed cannot be accounted for as easily. These are referred to as unknown sources of environmental variation.

This presentation will:

1) Illustrate the different types of genetic information available.

2) Explain how well each separates the genetic and environmental components.

3) Explain when and how the genetic information can be used for comparisons between animals.


When making genetic decisions economically important traits should be emphasized, but it is also important to understand the effect that selection has on different traits. Table 1 provides heritability estimates that tell us the strength of inheritance for each trait. Heritability is the percent of the variation in performance due to genetic effects. For example, backfat has about a 40% heritability. Thus, about 40% of the variation (the phenotypic differences between animals raised in the same group) in backfat is due to gene effects while the remaining variation is due to environment. Selection will be less effective for lowly heritable traits like pigs born alive because they are affected by the environment to a greater extent. Most litter trait have a low heritability , while production and carcass traits have higher values as shown in Table 1.

Table 1. Heritability estimates of some economically important traits for swine.


Heritability, %

Pigs born alive


Pigs weaned


Adjusted 21-day litter weight


Birth weight


Feed efficiency


Days to 230 pounds


Average daily gain


Backfat probe



Actual and Adjusted Data

To make any genetic comparisons, performance records must first be collected. All seedstock should be tested in some manner. Universities, companies, breed groups and testing associations, have programs designed for on-farm testing. The actual data collected gives information about a "specific" pig. However, when records are collected on animals there is an opportunity to account for the unknown environmental effects by creating contemporary groups. A contemporary group is a set of animals of the same breed, or cross, managed together under the same conditions. A contemporary group might consist of boars of the same breed farrowed during a one week period and fed a common ration in the same building. These pigs are similar in age, identical in sex and breed and therefore, we expect them to respond to unknown environmental effects similarly. They are also raised together with the same effects of feed, weather, health and management. How the data is collected and type of measurement equipment used is also important. Contemporary groups help to standardize non-genetic effects. Limited comparisons can be made of animals in the same contemporary group. When ranking the five boars in Table 2 on actual backfat, boar A or B is selected first, followed by D, C and E. However, there are other environmental factors that need to be adjusted for, like weight at probing.

Table 2. Unadjusted and adjusted backfat on five boars.

Boar ID

Actual Backfat

Weight at Probing

Adjusted1,2 backfat

at 230 lbs.





















1 The contemporary group average on 40 boars was .75 inch.

2 Adj. backfat = act. backfat +(230 - act. wt.)x[act. backfat/(act. wt. -25)]


Adjusting actual data for known sources of variation puts all of the animals in a contemporary group on a more comparable basis. For example, backfat thickness is affected by the pigs weight when the probe is taken; to reduce this effect, pigs should be weighed close to the target weight and backfat values adjusted to a constant weight (see Table 2). After the actual data for the five boars is adjusted to a constant weight off-test of 230 pounds the boars would be ranked; C, A, B, D and E. Boar C would move for fourth to first in the rankings. Adjustment procedures for most traits can be found in "Guidelines for Uniform Swine Improvement Programs" (NSIF, 1987). When using adjusted data to make comparisons within a contemporary group the group average for each trait should be computed so customers will know which animals are above average. Commercial producers should select animals that perform better than the group average.

Actual and adjusted data should only be used for comparisons within a single contemporary group and will have low accuracy. Actual or adjusted data alone is not a predictor of genetic potential.

Ratios and Deviations

Ratios and deviations are used to compare the adjusted data of each individual to the average of it's contemporaries, giving a more precise estimate of genetic merit. When calculating either ratios or deviations, actual data must first be collected in structured contemporary groups and adjusted for known sources of environmental variation.

A ratio is calculated by dividing the pigs adjusted record by the average, adjusted performance of its contemporaries and multiplying that result by 100. The average ratio is always 100 and a ratio of 125 means the pig is 25% grater than the contemporaries for that trait.

A deviation is the difference between the adjusted record of an individual and the average, adjusted performance of its contemporaries. Deviations are expressed in the units of measure being used for each trait, but one must know the average value of the trait and the variation of the deviations from the average value to properly interpret the comparison.

Ratios and deviations are only to be used when comparing animals in a single contemporary group. Different contemporary groups may be subjected to different environmental conditions and the genetic ability of the pigs in two different contemporary groups is probably not identical. The example in Table 3 shows that the rankings of the five boars did not change by using either ratios or deviations when compared to adjusted data. The ratios and deviations allow for easier comparison among the boars and are slightly more accurate when making selections because the group average is accounted for.

Estimated Breeding Values and Expected Progeny Differences

The genetic merit of an individual is its breeding value. A pigs true breeding value is never actually known but by applying genetic theory to adjusted data, Estimated Breeding Values (EBVs) can be obtained. The genetic merit that is passed on to an individuals progeny relative to other parents is one-half of the EBV, known as the Expected Progeny Difference (EPD).

A simple EBV using only an individuals own record can be calculated as follows: Simple EBV = h2 x Deviation. In Table 3 simple EBVs are shown for each of the five boars and it can be seen that the ranking does not change. These simple EBVs give precise genetic difference between the boars for backfat in inches but are only accurate within the contemporary group.

Table 3. Deviations, ratios and simple EBVs on five boars for backfat.

Boar ID



Simple EBV






















After actual data has been collected on the individuals in a contemporary group and adjusted for all known sources of variation that information can be combined with all other available records in the population to estimate across-contemporary group or across-herd EPDs. The total set of performance records and pedigree information is combined in a statistical procedure known as Best Linear Unbiased Prediction (BLUP) by an "animal model" that describes gene flow over time and the biology of the trait. The pedigree information allows the performance records of progeny, cousins, sibs, parents and grandparents to help predict the genetic merit of the individual. Comparisons can be made across herds where genetic ties are present and when sires have been used in multiple herds through artificial insemination.

The BLUP procedure can also utilize information from correlated traits in multiple trait evaluations. A multiple trait evaluation is one that estimates EPDs simultaneously for more than one trait. If two traits are correlated (either positively or negatively), information on one trait is useful to predict the EPD for the other trait.

Table 4. EPDs and accuracies for five boars for backfat and days to 230 pounds.

Boar ID


EPD Days

Acc. Days


Acc. Fat
































The accuracy value associated with an EBV or EPD gives an idea of the reliability of the estimate. Accuracy values will range from 0 to 1. The closer the value is to 1, the higher the accuracy. Low accuracy values are caused by very few pieces of information on the animal being evaluated or by low heritability for a given trait. As more information is used in the genetic prediction the accuracy value will increase. If an animal has EPDs that meet a producer's selection goals it should be used regardless of the accuracy value. A producer may wish to limit the use of an animal with low accuracy, whereas a boar with many progeny and hence a higher accuracy may be used more extensively. Accuracy values are most effective as a risk management tool because, regardless of accuracies, EPDs are the best estimates of genetic value available. Figure 1 shows that the genetic changes in progeny are greatest when using EPDs when compared to selection on adjusted data.

When using EPDs it is important to know how much information was used in the calculation. EPDs can be calculated and used for comparing individuals in a contemporary group, a farm, or a breed. At present, EPDs cannot compare animals across breeds. Using the information in Table 4 the following example will describe how EPDs are used. Example: Comparison of B and E for backfat - If boars B and E are mated to a large number of sows randomly picked, the expected difference in their progeny for backfat is determined as follows: B has an EPD for backfat of -.10 while E has an EPD for backfat of .05 (Table 5). Subtract one EPD from the other to get the difference between the Two i.e. [ -.10 - .05 = -.15]. The -.15 obtained say that the progeny of boar B are expected to have, on the average, .15 in. less backfat at 230 lbs than those of boar E.

Multiple Traits

Overall profitability is influenced by many factors. Several traits may need to receive emphasis in a well-designed breeding program. The difficulty is determining the appropriate emphasis to place on each trait. Traits are measured in different units (number of pigs, pounds per day, inches, etc.), are not of equal economic importance, and are not genetically influenced to the same degree (different heritabilities). These factors make it a difficult problem to determine the appropriate emphasis to apply to each trait in a breeding or selection program. Selection indexes are used to assign emphasis to each trait and provide a single value to use when comparing animals. Most producers will be familiar with the Sow Productivity Index; SPI = 100 + 6.5(L) + 1.0(W). Terminal indexes for the five boars in Table 5 show that when selecting for both growth and fat boar C rises in the ranking. Indexes that are calculated with anything other than across-contemporary group and across-herd EPDs can only be used for comparisons within a single contemporary group.

Table 5. Terminal Indexes for five boars.

Boar ID

Terminal Index











Multiple Breeds

None of the procedures discussed presently account for genetic differences across breeds or lines of hogs. However, research is currently underway, developing procedures for multi-breed genetic evaluations. Limited information is available from research trials and progeny tests that have been conducted. Also, the National Pork Producers is currently conducting a National Genetic Evaluation Program. This program will provide comparative information on terminal lines and the first results are expected to be available in 1995.


Genetics and understanding the numbers is important today, it affects your bottom line. NO genetic information can be compared across breeds of swine and most types of genetic information can only be used when making selections within a contemporary group. To make sound selection decisions you must understand how the measuring is done, what equipment is being used, and how the data is being reported. If you wish to make accurate genetic comparisons across different herds within a breed you must use EPDs or EBVs.

Commercial producers should demand that seedstock suppliers follow a sound genetic improvement program and use genetic information in seedstock purchasing along with acceptable health and structure.

Suggested Reading

Belonsky, G.M. and B.W. Kennedy. 1988. Selection on individual phenotype and best linear unbiased predictor of breeding value in a closed swine herd. J. Anim. Sci. 66:1124.

Mabry, J.W. and M.T. See. 1990. Selection with the animal model versus selection within contemporary groups for swine. J. Dairy Sci. 73:2657.

NSIF 1987. Guidelines for uniform swine improvement programs. USDA Program Aid #1157

Pollak, E.J. 1992. Interpretation of EPDs and use for single trait improvement. In: Proceedings of Symposium on: Expected progeny differences to livestock improvement. 84th Annual Meeting of the Amer. Soc. of Anim. Sci. Pittsburgh, PA.

See M.T. 1993. Using expected progeny differences for swine selection. Animal Science Facts, ANS 93-801S.

Return to Home Page!Return to MARK Home Page

Questions, comments, or for additional information contact

Dr. Todd See or (919)515-8797!

Notes: Information presented here represents the views of the Authors, and may not represent the views of North Carolina State University. The information provided herein is intended to be used for educational purposes and may not be reproduced without the consent of the Authors!