Table of Contents

Common Acronyms Used in this FAQ

What is NSIP?

NSIP -- the National Sheep Improvement Program -- is a computerized, performance-based program for genetic selection. NSIP is designed to help purebred sheep producers identify the best genetic stock for their breeding programs. NSIP also gives breeders reliable information that they can use to advertise and sell their breeding stock. NSIP uses the most modern, scientifically-proven technology to measure genetic performance. This technology -- called EPDs -- has been used extensively in the dairy, beef cattle, and swine industries for many years, and is only now being implemented in the sheep industry.

NSIP works through the breed associations, and in certain situations groups of producers, to deliver across-flock EPDs to purebred producers. Breeders use these EPDs to guide them in their selection and genetic improvement programs. A producer who is a member of NSIP receives reports on the genetic values for every animal in a flock, based on the performances of those animals and all the animals that are genetically related to them, over many years and management systems. By using EPDs, a breeder can make genetic improvements efficiently and reliably. EPDs allow a breeder to rank all the animals by genetic value, identify high-producing replacements, and cull poor-producing animals.

What are EPDs?

"EPD" is short for "Expected Progeny Difference." An EPD is an estimate of the genetic merit of an animal for a single trait. Specifically, the EPD of an animal is the expected difference between the performance of that animal's progeny and the average progeny performance of all the animals in the breed, for that trait.

How are EPDs calculated?

First, purebred producers record the performance values for their animals (weights, numbers of lambs born, wool characteristics, etc) and enter all this information into electronic data entry forms. They then send these forms to the breed association offices, where the data is compiled and checked and then sent to the NSIP computer. For each breed, NSIP collects these performance records from purebred flocks across the country, breed by breed. This data comes from sheep reared under many different management systems, year after year after year.

The NSIP computer then identifies the genetic linkages between these flocks and across years -- like when rams are sold or traded, or when progeny are distributed into many flocks -- and puts this data into one massive calculation for each breed. The NSIP dataset for a breed also includes all the data from previous years, for all the relatives, across generations. The EPD calculations even include data from related traits, because an animal's performance in any trait gives information on how it will perform in a similar trait (for example, a good preweaning weight for a fast-growing lamb suggests it will also have a good postweaning weight). These calculations produce EPD values on every trait for every ram, ewe, and lamb in the system. And these EPDs are recalculated annually (or more often for accelerated flocks), after the performance records from each new production cycle are entered into the computer.

Definition of Prediction Error   

The 2001 NSIP genetic evaluations introduce a new measure of the accuracy of EPDs. The measure is known as "prediction error" and is designed to avoid some of the confusion and misunderstanding that has been associated with the accuracy measures used in the past. The prediction error directly reflects the amount of future change in EPD that can be anticipated as more data accumulates on an animal, its relatives, and, most importantly, its progeny. Prediction error is expressed in the same units as the trait being evaluated (pounds for weight traits, microns for fleece grade, etc.). Thus each animal has a prediction error associated with the EPD for each trait.

Recall that the EPD is an estimate of its genetic merit based on accumulated performance records. As more information accumulates, the EPD can change. When little information is initially available, future changes can be relatively large. But once a substantial amount of performance data accumulates, the EPD becomes increasingly stable. The prediction error is a measure of the anticipated stability of an EPD. It differs from the accuracy values used before in that it directly addresses the magnitude of possible future change in the EPD whereas the accuracy gave only a relative measure of stability of the EPD. The prediction error was used for a time in the beef industry, where it was known as the "possible change", but was eventually discarded in favor of accuracy. In NSIP, where accuracy values are lower than in the beef industry, we believe that prediction error is a more useful measure of the stability of the EPD. 

An EPD for an animal can be though of as an estimate surrounded by error. The prediction error quantifies the magnitude of that error. The properties of prediction error can be summarized relatively easily: 

1. There is about one chance in three (a probability of about .33) that an animal’s EPD for a given trait will change (either increase or decrease) by more than the amount of the prediction error. The probability that the EPD will go down by an amount greater than the prediction error is thus about one chance in six. The corresponding probability that an EPD will go up by an amount greater than the prediction error is likewise about one chance in six.

2. There is only about one chance in 20 that an EPD will change by more than two times the prediction error.

3. An EPD is unlikely to change by more than three times the prediction error (about one chance in 385).

Figure 1 shows the probability that the true EPD will lie within the interval defined by the reported EPD plus or minus some multiple of the prediction error. For example, the true EPD is expected to be within the interval defined by the reported EPD, plus or minus .5 times the prediction error, about 40% of the time. As discussed above, the true EPD is expected to be within the interval defined by the reported EPD, plus or minus 1.0 times the prediction error about 67% of the time. The true EPD is expected to be within the interval defined by the reported EPD, plus or minus 1.5 times the prediction error over 80% of the time.

It is important to realize that the reported EPD is the best estimate we have of the true EPD, and the most likely value of the true EPD. When possible change is large, future changes in EPD may be relatively large. However, it is also important to recognize that the direction of these changes is not predictable. An animal with a large positive EPD and high possible change value could show a significant future drop in its EPD. Or its EPD could go up substantially. Either result is equally likely, which is why breeders should focus on the reported EPDs, and largely ignore accuracy and prediction error, when attempting to determine relative genetic merit. Prediction error should be used only as a risk management tool. If two rams being considered for use have similar EPDs but differ substantially in prediction error, the ram with the smaller prediction error is the less risky choice, since his EPDs are more stable. Prediction error alone is of no value; a ram with mediocre EPDs and low prediction errors is simply an animal that you can be confident will be mediocre. 

In order to assist with comparisons between the accuracy and the prediction error, Figure 2 shows the relationship between the two measures of reliability of the EPD for each trait. The maximum value for prediction error should not exceed the value shown for an accuracy value of zero. The only exception to this rule if for inbred animals. If an animal is inbred, its prediction errors will be somewhat larger than those for noninbred animals.

Figure 1. Probability the true value of the EPD will lie within the interval defined by the reported EPD, plus or minus the indicated multiple of the prediction error

 Examples

Table 1 shows examples of EPDs, prediction errors, and accuracy values for several types of     animal drawn from this year’s genetic evaluation. The animals chosen include: 

 A) A trait-leader for 120-day weight

 B) The ram with the highest accuracy (lowest possible change) for weaning weight

 C) A ram lamb in the top 5% for weaning weight EPD

 D) A 5-year-old ewe in the top 5% for percent lamb crop EPD

 E) A yearling ewe in the top 5% for maternal milk

 F) A ewe lamb in the top 5% for percent lamb crop EPD

Note that there is a direct correspondence between the values for accuracy and prediction error: within a trait, high values for accuracy are associated with smaller prediction errors. 

Animal A is a high-growth sire with substantial numbers of progeny. For this sire, accuracies are relatively high for growth traits and prediction errors are substantially below EPDs for these traits, indicating that the genetic superiority suggested by the EPDs is real and unlikely to be lost as more information accumulates. For animal A, the weaning weight and 120-day weight EPDs are 2.2 and 2.0 times their prediction error. These values suggest that the probability that this animal's true EPDs are actually at or below breed average for these traits would be only .014 for weaning weight and .023 for 120-day weight. Even though accuracies are high for animal A for growth traits, accuracies remain low, and prediction errors are correspondingly high, for maternal traits. This result reflects the low heritabilities of these traits and also suggests that this sire may have few daughters with records in NSIP flocks. Still, the maternal milk EPD of .8 for this sire suggests that his daughters do milk well.

 EPDs for animal B demonstrate that high accuracies are not necessarily associated with genetic superiority. This ram has the highest accuracy for weaning weight in the breed, but all that means is that we can be relatively confident that he is an inferior sire. Of course, this is valuable information, but we would be wrong to consider using him just because his accuracies are high. 

The situation for animal C, the ram lamb, is somewhat different from that observed for the older rams. This animal has good EPDs for growth traits, but prediction errors are larger than those for animals A and B, reflecting the fact that animal C has yet to produce progeny. Still, in many cases, the EPDs approach or exceed the prediction errors, strongly suggesting that this is a superior individual for growth traits. Although accuracies are low, his positive EPD for maternal milk suggest that he will produce good daughters, while the percent lamb crop EPD of -0.3 ± 9.0 indicates that we should expect him to be close to breed average for daughters' litter size. 

EPDs and prediction errors for ewes (animals D, E, and F) demonstrate how difficult it is to achieve high levels of accuracy in evaluating individual ewes, although our level of confidence in evaluating the older ewe (animal D) is very similar to that attained for a ram lamb (animal C). Thus genetic improvement in the ewe flock relies more on identifying groups of females with high average EPD values or on keeping sets of daughters of high EPD rams. Only a few older ewes will have enough progeny to achieve small prediction errors. Still, EPDs for maternal milk and percent lamb crop for animal D; weaning and 120-day weight for animal E; and weaning weight for animal F exceed or closely approach their prediction errors, suggesting that these individuals can be used to improve these traits with only modest risk. 

Table 1. EPDs, accuracies, and prediction errors for selected animals from the 2000 Suffolk national genetic evaluation.  PE = Prediction Error

 Using Prediction Error

The main use of the prediction error comes when you are forced to choose between a young ram with very high EPDs but also with relatively high prediction errors and an older progeny-tested ram with good, but lower, EPDs and prediction errors. Again, the issue is risk aversion. If your goal is the leap to an elite position in the breed and if you are willing (and have the resources) to take some chances to get there, the younger ram will be the best choice. Over time, it is always better to go with the higher EPD animals, but when prediction errors are large, you may have to weather a few disappointments along the way. On the other hand, if consistency and reliability of production are key to you, you may pay more attention to prediction error, preferring to use proven rams with less risk of future changes in EPD. But overall genetic progress in the flock may be slower.

 A rough guideline to assessing reliability of an EPD is that if the EPD exceeds the possible change value, an animal with a positive EPD is unlikely (one chance in six) to drop below zero in the future, and an animal with a negative EPD is not very likely to move above the average. Also, small differences in possible change are not worth worrying about. The issues of importance come when making choices between ram lambs and progeny-tested sires or between adult ewes and ewe lambs. Differences in prediction errors among ewes in the breeding flock are almost never large enough to be important. Focus on EPDs in selecting and culling breeding ewes. Don’t worry about small differences in possible change!

 Finally, realize that without widespread AI, the sheep industry will not have the large numbers of proven sires found in dairy cattle. In most cases, our objective is not to find a few exceptional rams, although when such animals do emerge, they will, of course, be welcome. Our goal is to select groups of replacement ewes and rams that will provide consistent genetic improvement. Thus flocks of reasonable size need to focus on the average genetic merit of the rams purchased or the ewe lambs retained each year.  

The concept of prediction error can be extended directly to groups of animals. If a breeder goes out to buy four ram lambs, each with EPDs for weaning weight of about +1.5 pound, the prediction error for these rams will typically be around 1.7 pounds. For the group of four, the average EPD will still be +1.5, but the prediction error of the group average will now be only about .85 pounds. Thus, your new rams, as a group, can be expected to reliably enhance weaning weights in your flock. Similarly, if you purchase a group of 10 ewe lambs, each with EPD and prediction error of .5 ± 1.2 pounds for maternal milk, the mean EPD of the group is still .5 but the prediction error of the group mean reduces to .4 pounds, giving you more confidence in the genetic merit of the set of ewe lambs. These examples demonstrate why larger flocks can pay less attention to prediction errors while relying on difference among individual animals to average out future changes in EPDs. In contrast, small, single-ram flocks will have to decide for themselves how much risk is acceptable when they only buy a new ram every year or two.

How are EPDs reported?

An EPD is reported in the normal units of a trait, such as +0.5 pounds (for weights) or -0.3 microns (for wool diameter). It's important to note that an EPD value is not a ratio or an index. EPDs are expressed as deviations (+ or -) from the average population value, which is considered to be zero. Therefore, EPDs always have a positive (+) or negative (-) sign in front of them.

The positive and negative symbols don't always mean better or worse -- it depends on the trait. For example, an Weaning Weight EPD of +0.5 pounds is good (i.e. more weight of lamb at weaning), but an Fiber Diameter EPD of -0.3 microns can also be good (i.e. smaller diameter fiber, which is more valuable to fine wool producers). Not only can we compare sheep with positive and negative EPDs, but we can also use EPD values to compare animals who both have positive EPDs. For example, a ram with a Weaning Weight EPD of +1.0 is good, but a different ram with a Weaning Weight EPD of +2.0 is better. EPDs may take a little getting used to, but once you get the hang of them, they give the most objective and reliable estimation of genetic value possible.

Can Rams Be Compared Under Different Management Conditions?

Yes. across-flock EPDs are designed to allow this comparison. The calculation of EPDs uses data from many different flocks, and this procedure is mathematically valid across flocks. This means that a range operation in dry country can use rams from a Midwestern corn-soy crop farm, and that a Midwest farm can identify top-quality range rams reared on sagebrush and rattlesnakes. Of course, on each farm, the groceries and health still have to be good enough to permit good performance, and in particularly stressful environments (such as desert range) there may be unique genetic adaptations that affect performance. But at least EPDs give a producer a clear and reliable report about an animal's genetic potential.

Can Animals Be Compared Between Different Breeds?

No. An EPD in one breed cannot be compared to an EPD from another breed. Across-flock EPDs are calculated only within a breed. Each breed database is independent from all other breed databases, and the numbers are not commingled.

What about Commercial Producers?

EPDs are only calculated on purebred animals. EPDs are not calculated for commercial flocks. Commercial producers do not join NSIP directly. Commercial producers, however, can really benefit from NSIP because they can purchase rams (and ewes) from NSIP purebred flocks that have precisely the improved traits that they need.

Because EPDs are provided on a trait-by-trait basis, commercial producers can decide what traits they need for their operation and then use NSIP to find rams and ewes that excel in those specific traits. The Breed Associations publish "Sire Summaries" -- in printed form and on their websites -- which are genetic catalogs that list all the NSIP sires in that breed, trait by trait. These sire summaries often include lists of "trait leaders," which is convenient for quickly identifying top genetics. Commercial producers can study these sire summaries and easily find the best sires and dams which carry the improved traits for their own operations

What are FEPDs?

FEPDs are the same as EPDs except that all the data is derived from a single flock. The "F" stands for "Flock." FEPDs do not use across-flock data. For over 10 years, NSIP calculated FEPDs for producers.

When NSIP first began, across-flock analyses for sheep were not feasible because there was not enough good information on identifiable genetic linkages between flocks. NSIP calculated FEPDs to serve producers and provide them with the best genetic information that was available at that time. For ten years NSIP calculated FEPDs as part of its program to collect data for the development of across-flock EPDs.

In contrast, across-flock EPDs provide far more information about genetic value than FEPDs because they are derived from many flocks and over different management systems.

NSIP currently calculates only across-flock EPDs for producers.

What Traits does NSIP Evaluate?

Maternal Traits:
NSIP evaluates all individual animals within a flock for three very important maternal traits: (1) number of lambs born per ewe lambing, and (2) maternal milk, and (3) Milk+Growth. To obtain an accurate evaluation of genetic merit for each of these traits, producers record information on all ewes exposed for breeding and all lambs born in each production cycle.

Growth Traits:
NSIP evaluates growth for three possible weights: weaning weight, postweaning weight, and yearling weight. Farm flocks and range flocks are analyzed differently because their weighing schedules are so different. Farm flocks receive 60-day weaning weights and 120-day postweaning weights. For farm flocks, the cutoff point between weaning weight and postweaning weight is 90 days. Range flocks receive 120-day weaning weights and yearling weights. Some range flocks also chose to take 60-day preweaning weights, and those weights are used in their genetic analysis. NSIP accepts generous time windows around each age to weigh lambs, so that any flock can arrange convenient weigh dates to fit its management schedule.

Wool Traits:
NSIP calculates EPDs on three wool traits: grease fleece weight, fiber diameter, and fiber length. These measurements only need to be taken once during an animal's lifetime, usually at a year of age. These measurements must be taken on a full year's growth of wool. A producer can also record codes for face cover and skin folds, although no EPDs are calculated on these traits.

Carcass Traits:
Carcass traits are still under development but will be incorporated into NSIP very soon. These traits will be fat thickness, ribeye area, and an index trait called the "Carcass Value Trait", which will be calculated from the age of a lamb, the weight of the lamb, ribeye area, and fat depth between the 12^th and 13^th ribs. Producers will record values for these traits either from direct measurement of the carcass or from ultrasound measurements on the live animal.

Accelerated Traits:
For those breeds using accelerated lambing systems, NSIP is working on two important traits: date of first lambing, and lambing interval. Currently, implementation of these two traits are still under development.

Does NSIP accept electronic data entry?

Yes. In fact, NSIP only uses electronic data entry. All producers enter data onto specially-designed, easy-to-use spreadsheets. These spreadsheets look very much like the traditional paper forms -- i.e. they have rows and columns. However, these computer forms also do automatic error-checking, so that the spreadsheet catches typos and other data errors when they are first entered, which makes these errors very easy to correct. If a producer doesn't have a computer or prefers to avoid doing data entry, they can have a friend/relative/neighbor do it or make arrangements with the personnel at the breed association office to do this for them. After producers enter their data, they send the file via the Internet (or just mail a diskette) to the Breed Association office.

How does NSIP work?

NSIP works hand-in-hand with the breed associations. Performance data flows from farms and ranches to the breed association offices and then to Virginia Tech University, where the actual EPD calculations take place. The EPD results then flow back to the breed association offices and then back to the individual farms and ranches. This means that there is no central do-it-all NSIP office. (This style of organization is similar to the well-known DHIA system that has been functioning successfully in the dairy industry for many years).

Specifically: Purebred breeders collect performance data on their farm or ranch and enter that data into data-entry spreadsheets on their own computer, or have someone else enter it for them. These spreadsheets have been developed by NSIP and are supplied to all the breed association offices and breeders. Producers then send their completed spreadsheet to their respective breed association. People at the breed association office combine all the files into a single large spreadsheet file. They do additional checking of the data (especially registration numbers) and then send the breed data on to Virginia Tech for genetic evaluation. At Virginia Tech, geneticists run the complex software to calculate EPDs. They also archive all the NSIP databases in a secure mode. The specialized EPD software (BLUP = Best Linear Unbiased Prediction) calculates the genetic values for each animal and each trait. After these EPDs are calculated, the results are sent back to the breed associations. The breed office then uses these results to publish the breed sire summary and other genetic documents for the breed. It also sends genetic reports back to the individual breeders. Most of this work is done over the Internet. NSIP oversees and coordinates the entire process. NSIP also develops new procedures and traits; it tests improved data collection and data evaluation techniques; it maintains the databases for security and archival purposes, and it coordinates genetic research with the data.

The NSIP main office is located at the ASI headquarters in Englewood, Colorado. NSIP clerical and bookkeeping tasks are performed there.

What role does a breed association play in NSIP?

Breed Associations play a pivotal role in the NSIP genetic evaluation process. The association office tracks all breed members who are in NSIP and acts as the focal point for the final data checking and assembly. It sends NSIP producers the data-entry spreadsheets. It collects the completed spreadsheets, compiles them into a single breed spreadsheet, does additional error-checking on the data, and works directly with NSIP and Virginia Tech to assure accuracy and integrity of the data. After the EPDs are calculated, the breed association processes the results and sends the genetic reports back to the individual breeders. The Breed Association office publishes the breed Sire Summary, and also publishes other genetic documents that are derived from the EPD results, including lists of trait leaders. The Breed Associations also collect fees from producers for joining NSIP.

How can a producer join NSIP?

Simple. A producer submits an Enrollment Form to the NSIP office. This Enrollment Form is a very easy-to-fill-out form that asks a few questions about contact information, breed, and flock size. Send this form to the NSIP office along with your payment.

Enrollment forms can be obtained from this website, the NSIP office and breed association offices,

All purebred producers with registered animals can join NSIP. Calculation of across-flock EPDs, however, is dependent on the establishment of good across-flock genetic linkages. NSIP is currently working closely with ten breeds to calculate across-flock EPDs. These breeds are Targhee, Suffolk, Polypay, Dorsets, Hampshires, and Columbia. Producers in other breeds will receive across-flock EPDs if enough flocks join NSIP so that good genetic linkages can be established. NSIP will help facilitate this. If the breed association office cannot act as a collection point for data, then the group of breeders needs to find someone else to do this role. Again, NSIP will facilitate this. There are people in the sheep industry already doing these tasks

How much does NSIP cost?

The NSIP fee structure is very simple and reasonable. Annual fees are based on two things: (1) a flock charge plus (2) a charge per each breeding animal in the flock.

NSIP defines a breeding animal as an adult ewe or a ewe lamb that will be part of the breeding flock or a ram that will be used to sire lambs. On the Enrollment Form, a producer counts all animals, male or female, that were used in breeding during the past year.

Producers enroll only purebred animals with registration numbers. They also enroll any ewe lambs or ram lambs used for breeding that will be registered but have not yet been assigned registration numbers.

The producer would send a check with the completed Enrollment Form to the NSIP office.

When are fees determined?

At enrollment. On the Enrollment Form, a producer lists the number of breeding animals and pays the sum of the ewe and flock charges.

Who pays whom?

Breeders pay their NSIP fees directly to their Data Coordinator.  The Data Coordinator will mail producers a receipt for their payment.

Does NSIP have a website?

Yes. NSIP

Where is the NSIP office?

NSIP
6911 South Yosemite Street, Suite 200
Englewood, CO 80112-1414

Phone: 303-771-5717
Fax: 303-771-8200
Email: NSIP Office

Where can I get more information?

Purebred producers: Contact your Breed Association or t