BREED VALUE and the GSD - The SV Zuchtwert Program Part 2

by Fred Lanting


© Copyright June, 2000 - 2005

ZW Mandatory As Of July 1999

Wide implementation of the method of Breed Value Assessment has been effective in the “commercial animal” breeds (food & textile). This includes, in a computer record, all available data of the ancestors and any available progeny as well as of the personal phenotype of the animal concerned, and can work authentically and exceedingly effectively in the purebred dog. In cooperation with the Institute for Breeding Animals and Animal Genetics at the University of Gießen, the SV developed a breeding program that became obligatory as of July 1, 1999. Before that, starting in September 1998, the process was voluntary. Each breeder currently and henceforth knows about the inheritance (HD-production) risk of his bitch through this HD breed value. The breed value numbers of all stud dogs are likewise disclosed. From both of these sides comes a greater knowledge of HD risk for their progeny. The official position of the SV is that considerable breeding progress in the area of declining HD incidence can be had by implementing this measure.

The HD Program of the SV

Success speaks for itself. The SV has had a protocol for HD control since 1966, which was recognized at an international technical convention in l974 as "the best in the world." Nothing has changed about it until recently.
Through mid-1999, over 260,000 German Shepherd Dogs had been X-rayed in the SV’s HD scheme, and about 240,000 dogs received that coveted "a" stamp, which denotes official permission for the dog to be bred. Approximately 13,000 new "a" stamps are awarded annually. The percentage of dogs with medium and severe HD was drastically reduced by initially 27%, until today there are only around 6% as identified in the old diagnostic technique. The SV has indeed made greater strides in lowering the number of severe-HD cases and preventing severely dysplastic dogs from breeding, than have U.S. or many other counterparts, and this is to be admired. However, because of the failings in their “standard” diagnostic method (see my articles in the Internet on the PennHIP and OFA methods) the SV has not made as much progress as they should have in reducing the number of “carriers” and actual dysplastic dogs that slip through the net designed to catch them. This is why the SV has decided to use Breed Value. What they really ought to do is use both ZW and PennHIP. I am sure you noticed the proximity of those numbers of total dogs radiographed and those “passed”: this should create even more questions about how stringent the “a”-stamp program is. It is time to raise the bar, since the SV has reached a plateau just as the OFA-type evaluation has, although with the ZW, that SV plateau will prove to be at a slightly higher altitude.

Decisively involved with the success of the battle in Germany are the SV-certified HD radiologists and HD veterinarians. Nearly 3,000 veterinarians have become authorized by the SV as “acknowledged HD veterinarians”. Contrast this with the most well-known American system, where any vet can take a picture and send it to OFA or just give his own reading. And with the PennHIP one, in which only P-H-certified and trained vets may participate — in the former, there is a statistical bias if not all films are sent in; in the latter, the vet would lose privileges if he were to withhold any from the database. In this respect, and in the amount of useful information gathered, the SV’s total system is intermediate in value or usefulness between OFA and PennHIP. Toward the end of this article, I condense the instructions the SV gives to their listed vets.

Dr. Beuing, the administrator of the ZW project for the SV, answers the following most frequent questions on the SV website and mailings to certain members. However, he does not really answer satisfactorily, partly because geneticists use such complicated formulas for BV. My (Fred’s) comments in brackets.

1. How is the Breed Value HD worked out for my dog? What is the formula for it?

There is no precise formula in these calculations. The ZW of every animal [whose parents were not already calculated] was considered “unknown" at the beginning of the calculations. Then an equation is put up for every animal, with the formula:

(ZW for HD) = the breed average + Breed Value of this animal + variation of sex + other influences.
In this equation the breed average, breed value and the effect of the sex are formulated as the "cause" of HD. In the
SV, 450,000 equations were formulated for 450,000 animals, with the 450,000 unknown Breed Values. Then some
supplementary conditions (stipulations or prerequisites) of the laws of inheritance and heredity are mathematically
formulated. Following this, the computer has to solve the 450,000 equations with the 450,000 unknown factors. The
answers are the assessed (estimated) Breed Values. This means that there is no formula, in which a dog's breed value
is worked out on its own.

2. What role does its own HD score play in my dog's HD Breed Value?

The dog's own HD grade modifies that which was already known about the sire and dam. Siblings may differ because of their own HD status. If an animal has progeny, its own HD status [“a”-stamp grade] recedes more and more into the background, becomes less of a factor. With 30 - 40 progeny, it is practically of no consequence. [We see here that the ZW program uses that old reliable “progeny testing” approach, which makes it far better than what we use in the U.S. Hinted in Beuing’s answer here is the tacit acknowledgement that the “a”-stamp for hips is not all that reliable, something he dares not say unequivocally. Even though the “Normal” is given more weight than “Fast Normal or Noch Zugelassen, the unreliability of these assessments casts some doubt on the reliability of the ZW itself. The greater the number of radiographed dogs that contribute to a parent’s ZW, of course, the more these discrepancies are minimized.]

3. How is a foreign “a” [or equivalent] evaluated for the ZW Breed Value Assessment?

At this time, dogs from abroad with the “a”, OFA, etc. are evaluated in the formulations as though they were not X-rayed. Their ZW is gained only through their X-rayed relatives. As these foreign animals cannot be clearly termed as “HD-free” or “still acceptable” [noch zugelassen], an assumption in this case, would be unfair.

4. How can a dog that was not X-rayed have a ZW for HD?

In the system of equations (see Answer 1), there is a lateral clause, that the father and the mother each contribute half of their genes to the progeny. Therefore the Breed Value of the progeny can be assessed or estimated, if nothing is known about the animal (not X-rayed), as well as if the breed value of the sire and dam are known. In the same way, a parent can be assessed if the progeny is known, because progeny have one half of their genes from each parent. Siblings on the one hand help to characterize the parents, the knowledge about the breed value of the parents helps on the other hand, to assess the breed value of the untested siblings.

5. How accurate can the Breed Value Assessment be, if no HD information is available for more than half of the whelps?

In the ZW Estimate, the percentage of the pups does not matter, but the total number. The last Sieger has his heritage well demonstrated by 71 X-rayed progeny as of early 1999. [Each month I get from the SV the news of breedings by German dogs, and can watch the ZW numbers change. Sieger Rikkor, mentioned here, had as high as 118, meaning that he has produced an uncomfortably high number of dysplastic dogs. As his owner requires (or is limited to) lower ZW numbers on prospective partners, his own ZW slowly comes down a little. This has been seen in the past year and more, with a ZW of about 110 as of mid-2000. The ZW of an unproven dog is an average of both parents, and the “requirement” now is that this average must not exceed 100 (the breed “average”); that such a mating will not be allowed/registered. However, even though the SV says that, they allow plenty of exceptions, as the monthly stud service/litter announcement publication shows.]

6. My bitch has passed on [the ability to get] the “a”-stamp well up to now. What happens if I mate her to a dog that has a HD Breed Value over 100?

If the bitch has produced well, this must be seen in her ZW. Good progeny, however, can also be the result of good matings. If the bitch, for instance, is mated to a dog with a value of 100 [or higher], then the risk for the puppies is higher. The bitch is only “accredited negatively” if the pups are worse than could be expected with that stud dog. Generally it might well be that the stud dog owner would want to accept “poor” bitches, because that way his dog can show plainly that he “improves”. [(in regard to hips) Unfortunately, this is not the best way to improve the breed as a whole, and can be considered a very selfish and deceptive tactic, as well as very risky, because more people will look at his ZW than calculate how much better the progeny’s hips are than their dams’ were]. With HD, where there is no better rating than “HD-free”, it is difficult to prove a positive heritability of the top bitches.

7. How do dogs that have no HD findings flow into the HD ZW calculations?

At this stage, non-radiographed animals are “neutral” for the ZW; they are not taken into consideration.

8. What does the mean Breed Value of 100, and the stated deviation of the standard of ten points mentioned in the breeding plan mean?

The ZW of 100 means that the animal's hip heredity is typical or average for the breed. At this time the breed mean is 1.71 (that is, between HD-free and mild HD). What is passed on can be better or worse than 100. The average deviation above or below this mean will be set at 10 points.

9. What does a hip Breed Value higher than 100 mean for my dog?

This ZW for such a dog means that he passes on poorly [in regard to sound hips]. As the breed average itself is still unsatisfactory, matings should be striven for that bring the average under 100. Still, “you must not throw out the baby with the bath water.” The first priorities are working ability, temperament, and so on. If one finds a stud dog who brings the expectations for the pups (the average of father and mother) below the 100 value, that is sufficient.

10. Conversely, what does a breed value below 100 mean for my dog?

A ZW below 100 means that there is a greater freedom in the choice of dogs [to breed with] who are firm in character, strong in performance, and of good body construction, even if their breed value is around 100. With a bitch value of 83, one can even accept breeding to a dog of 117!

11. Can a Breed Value be worked out for criteria other than HD and, if so, for which ones?

A ZW estimate can be obtained for other characteristics. For this, it is important that the particular characteristic is accurately defined.

Dr. Reiner Beuing, Institute for Breeding Animals and Animal Genetics, University of Gießen, Germany.

Some Recommendations for Breeders: The Next Step in Progress

Dr. Beuing says: The most important aspect is that we care for the health of the dog. It is also a question of humane animal protection: the animal protection law in Germany, section 11-b, forbids breeding animals if, in the procedure, pain or grief is expected because of a faulty part or organ. [Dr. Beuing, with the official blessing of the SV, has admitted what his colleagues have complained about, which is the need for better orthopedic soundness.] Breeders have invested 30+ years in the HD fight, with an annual cost to the breeder communities between one and two million Deutsche Marks (up to about a million dollars). Of course, although some long-term improvement can be shown, the amount of success is nevertheless unsatisfactory. In addition, in the last several years the progress has stagnated, so that more effective methods of breeding are long overdue. [That is the plateau I mentioned earlier.] First, each of our breeders should select more strictly. One should exclude not only the animals with medium and severe HD, but also stay away from breeding the dogs with mild and borderline cases of HD.

Oh, yes, objections will immediately come up here! We should not base our selection only on what the individuals themselves are like but, since it deals with the selection of breeding animals, we must pay attention to how the animals produce. Therefore it must also be possible to weed out from the breed some supposedly “HD-free” individuals if they produce badly, and we must perhaps give good-producers from good lines a chance, even if their own radiographic evaluations have shown HD in a borderline or mild case. [Remember, there are inaccuracies in the standard view of hips. Using the combination of low ZW numbers and the best diagnostic method, this weeding and improving can be accomplished most rapidly.] The concept that is really new and lacking from previous education in SV thinking is that now the inherited traits, that is to say, the puppies’ quality, moves into the forefront. We breeders should no longer concentrate on the individual breeding animal, but on the pairing, the combination! Both partners in the breeding pair jointly determine the hereditary factors (genes) in the puppies. [It is interesting to note, as I did earlier, that the SV is trailing by many years the cattle industry and The Seeing Eye, Inc., both of which have long been relying on Breed Value numbers for guidance in such pairings.]

Comments on the Individual Items

1. General

[The GSD is certainly not the only breed in which hip dysplasia appears, but as mentioned elsewhere, HD has a very high profile in the breed. A hereditary predisposition is responsible for it. Therefore, measures were subsequently formulated to serve the genetic improvement of the breed. In addition, the SV continues to give advice for the breed’s suitable nourishment and necessary care in infancy. According to the SV, scientific investigations have shown that for HD, as diagnosed with the old leg-extended view, around 20% of the (phenotype) variation from one animal to another depends on differences in the genes. This low heritability shows that other factors (in the individual) must also be considered, and therein lies the SV’s mistaken or undue amount of emphasis on “upbringing”. Especially on optimized nourishment and expert, cautious raising and training. Therefore they say that a breeding program can be concerned with and solve only a part of the problem. Beyond that, it is the duty of the SV, they say, to wage the HD fight via nourishment advice and enlightened training and upbringing/maintenance practices.]

2. Determination of genetic risks

The SV employs an estimate, in its attention to genes, by means of the Zuchtwert evaluation. Presently the statistical process is determined by MMP and MME as the best available procedure, which includes information on all relationships. The breed worthiness is expressed as relative breed values with the median value being 100 (representing a breed cross-section) and having a standard deviation of 10 points. In the breeding program it must be determined how best to make use of the Zuchtwert numbers and information. Since it is a complex mathematical process, what is presented to the public is only a summary of this scientifically studied and accepted method. The door is left open for changing the process for a more effective one if it became available. Presently, however, it is the technology of the Mixed Linear Models that is the optimal process:
• MME: Mixed Model Estimates for the determination of any gross false clutter or interference with the evaluation, and
• MMP: Mixed Model Prediction for the forecasting of the inherited transmission by the animals). The process some perhaps know as “BLUP” belongs to this methodology.

3. Information

The HD classifications serve as information for the Zuchtwert evaluation according to the guidelines of the FCI. Further knowledge (from an earlier radiograph, for example) is handled according to its relevance and importance. If one wants to calculate breed worthiness from currently available information, it must be regulated and defined just what information that would be. It should be clear that current radiography remains unchanged at present. It was determined that the current HD classifications [Normal, Fast-normal, and Noch Zugelassen, plus the obvious and disqualifying grades of HD] suffice to define the inherited distinctions. The disadvantages of a shift in policy, direction, and judgement of the classification would lead to too large a change, and would result in a loss of all current knowledge, [being replaced by something that would be no longer comparable]. Nevertheless, the situation anticipates that further knowledge can be included. We (the SV) will work on such possibilities. [Thus, they close the door to PennHIP adoption yet say they leave it open a crack at the same time.]

4. Present calculation and information responsibility

The current Zuchtwert numbers are made available in a suitable manner at least at the beginning of each quarter to members, breeders, and local clubs. In addition, the Zuchtwert is updated on “information-software” CD-ROMs (“SV Genetics”), which are obtainable from the SV headquarters. These numbers are required to be published in the mating-notice bulletins (Nachrichten des SV-Zuchtbuchamtes, — stud service or mating news — etc.). In Germany, Landesgruppen (regional groups) delegates with access to the Internet ( are responsible to the members and breeders for making available this information to those not on the ‘net. This obligation is resolved by the EDP (electronic data) department at the SV headquarters through the news media, through PC (computer) support, and through the Internet, but naturally also by direct inquiries to the headquarters (for a fee).

5. Progress

Animals with medium and severe HD have already been excluded from the gene pool — forbidden to breed —according to Section 4, paragraph 1.3 of the breeding rules of the VDH. Nothing is lost if animals that are themselves severely affected by HD are left with no potential partners with far-enough below-average ZW numbers to breed to, anyhow… their own Zuchtwert numbers are too high. Dogs that in regard to other features are allowed to breed according to the Zuchtordnung (the breeding rules of the SV), may be so included only in regard to their partners’ scores; which is to say, if the HD risk for the puppy itself as averaged from the ZW numbers of the parents does not exceed a certain limit. At this time, a limit of 100 is determined to be an upper boundary. It is recommended that you strive for the lower values. The respective breed value published in the quarter in which the date of mating falls, is the figure that counts as the Zuchtwert of the breeding pair. If the mating should occur within a maximum of 2 weeks after publication of the Zuchtwert numbers, you could also use the numbers given in the prior quarter as the bases for the average.

[Without such restrictions and/or selection, no further progress could be expected. The peculiarity here, however, lies in the distinctive form of the restrictions. To a certain extent, it is contradictory to the idea of the master plan, if these breeding restrictions were to be based on the owner's judgement of the animals and he wanted to loosen them. However, the new rules now take effect. In the breed scene the breeder now has the duty to look for such studs for his bitch who prove to be the good genetic producers, those that are expected to produce puppies with below-average risk. According to the SV, it is important that a bitch remains in the gene pool as long as the breeder is successful in finding dogs that are good enough for that bitch.]

Let us play through another example: A bitch has a breed value (ZW) number of 102, i.e., she ranks slightly higher than the medium risk. The breeder must be sure his first-choice dogs (for performance, beauty, character, type, etc.) have ZWs of 98 or lower. Let us assume he chooses a dog with the breed value of 96; then the initial expectation for the puppies is 99, which is acceptable for the breeding program. No one prevents him from taking her to perhaps “an 89 dog”. Both studs are acceptable, but the conclusion is that the 98 is not as good as the 89!
Now, a ZW assessment is no guarantee. It certainly can happen that bad results can arise from this combination. If the father already had many x-rayed progeny and one knows his breed value with a high degree of confidence, this mostly affects the bitch. Let us say her breed value amounts to 107 and that of the dog is 97 (average 102), so a repeat mating would be not permitted. [However, I have already noted that this “rule” is being circumvented all the time!] The breeder must choose for the next breeding a dog with a maximum of 93 (average of 107 and 93 is 100). The breed progress regulations thus allow for a range that permits much freedom in the choice of dogs, if the bitch produces well. This always puts greater limits on choices, and forces the use of good genetic producers, in proportion to how badly the bitch produces. Also, in this standard range young animals from good producing lines have the best chances. The principle of this Zuchtwert-oriented breeding is called “strategic pairing”.

Now, there was still the problem in the practical routine that after a cutoff date, breed values might have lost their validity and new numbers put in their place. A mating planned and arranged today could have become inadmissible tomorrow. In the discussion leading to the establishment of this breeding scheme, this failing was revealed and disarmed: in the first 14 days after validating a new Zuchtwert figure, the numbers of the preceding quarter can also be put on the application. If the new is better than the old, then the new of course is valid. If the new number no longer meets the averaging requirements, but all is arranged, the mating still is possible, insofar as the two-week time limit is observed.

6. Matings in foreign countries

Dogs from the foreign countries are welcomed and may be used only if:

  • they are found in the database of the SV with at least 3 generations,

  • an HD evaluation comparable with the German method exists, and

  • calculated breed value was used in the breeding.

Proof of the "a"-stamp for HD was also necessary. In the passage of the new breeding program, it was anticipated that the owners of dogs certified in the foreign country may get a similar registration by applying to the SV and submitting the HD evaluation result. After that first step, it is put "into the German market". If the dog is initially integrated into the EDP of the SV, and the SV has a studbook (SZ) number for him, then he is automatically included in the breed value procedure. His progeny in Germany then determine his future image. Without this regulation we would have bitches for which the suitable partner is simply difficult to find in Germany, but is licensed in foreign countries. However, that cannot be in the best interest of the breed.



Fred Lanting is an internationally respected show judge, approved by many registries as an all-breed judge, has judged numerous countries’ Sieger Shows and Landesgruppen events, and has many years experience with SV. He presents seminars and consults worldwide on such topics as Gait-&-Structure, HD and Other Orthopedic Disorders, Anatomy, Training Techniques, and The GSD. He conducts annual non-profit sightseeing tours of Europe, centered on the Sieger Show (biggest breed show in the world) and BSP. Check out his website: