Genetic management of a breed with very low numbers

One scenario, already mentioned, is not to be taken into consideration here: where a breed had effectively disappeared and been recreated by out-crosses. However agreeable such a method might appear, it has nothing to do with saving a breed, except perhaps keeping the  population in memory.

There is no inevitability linked to numbers that are too low, provided there is still a little variability in the animals. The technique consists of using as many males as possible in reproduction, ideally almost all (obviously those that express defects must be eliminated). Selection proceeds to the second level and it is therefore necessary to accept subjects which, in the normal course of events, would not have been authorised for reproduction. Several examples are known in farm animals, of isolated populations that have ended up achieving high numbers starting from a foundation stock of only a few dozen animals. We know personally the example of the Mouton des Landes in Brittany which, in 15 years, rose from 60 to 600 head of stock: in the initial remaining flock, each year's male lambs were left to run with the ewes until their departure to the abattoir. One might say that 60 breeding stock was already a not insignificant group, which is true, but it should not be forgotten that ewes are not such prolific breeders as bitches. It is also important to know that specialised geneticists consider that with anything under 100 head of stock, a population is lost. This is not true.

As soon as numbers of the breed to be saved have risen sufficiently, it will then be possible to  progressively reduce the number of males for use in reproduction, without however claiming to have returned to a normal situation for a considerable length of time, and to have carried out a minimum of selection.

If the residual variability in the re-founding animals is too low, any success of the operation will however be compromised and the introduction of new blood is virtually inevitable. Obviously this should be done with a related breed and limiting the extent so that the genotypes that one wants to safeguard are “diluted” as little as possible. What is termed an introduction of “half-blood”, limited to a single generation, should be enough to reintroduce sufficient heterozygosity.

Sperm banks provide a very interesting tool. One sometimes tends – when it involves saving populations – to think of them as a last resort. This can effectively be the case but there is of course an advantage to anticipating and not waiting for the almost total disappearance of a breed before freezing what remains of its genetic heritage. The problem is that financing will no longer be found on the side of any breeders because there are virtually none remaining; the  situation is therefore even more difficult than in the case where it involves, for a breed with “normal” numbers, freezing reserves of variability in the long term interests of all members. For breeds with very low numbers, there is only the national Kennel Club which might possibly be in a position to do something. It should be noted, on this point, that the French Kennel Club is interested in this possibility: we can only be thankful for this.

WHAT CAN BE DONE TO PREVENT HYPER-TYPES AND OTHER GENETIC DRIFTS?

The drift towards hyper-types

The term "hyper-type" is currently used to describe animals that exaggerate the expression of their breed's morphological type overall, or else a particular feature which is looked for. An entire breed may be hyper-typed – classic examples of this are the Bulldog and the Sharpei – and nonetheless include some animals even more hyper-typed than others: they are deliberately used by advertising, which contributes to the fashion in their favour. However, the most frequent case is where one sees the appearance from time to time of hyper-typed animals within a breed which nonetheless still remains “normal”; for example, excessive weights, features that initiate – or intensify – a flattening or a lengthening, more skin folds than usual, short backs that become excessively short, hair that grows too long, etc.

The drift towards hyper-types constitutes one of the "diseases" of today's dog world. It is unfortunately encouraged by dog owners, who are easily attracted by an animal that is "different", for example because its morphological type, overall or on a particular point, is significantly pronounced. It is then encouraged by judges, who have no hesitation in giving top awards to hyper-typed dogs. To cap it all, such animals very often do not comply with their breed standard! The French Kennel Club's zoo-technical commission emphasised, more than ten years ago, that "hyper-type" should be assimilated to "lack of type", which, quite logically, bars from confirmation. This means that certain champion dogs are not eligible for confirmation!

One sometimes hears it said that breeders need hyper-typed dogs so that, with judiciously planned matings, they can return to the middle range which is tending to disappear. We challenge the word "need" since there are other solutions, more progressive ones, in the context of a properly conducted breeding selection; in any case, how did we manage at a time when there were few, if any, hyper-types? Even if true, it would be best at least to keep such animals in the kennels and not “display” them at dog shows.

If hyper-type was only to give shape to a new direction for the breed, there would be no more than two questions to be asked: 
- wonder about the image that is being given to the breed, at a time when "animal r" is taking on more and more importance, sometimes too much. It is concerned about the well-being of animals as well as the respect that is shown to them: and it is fairly obvious that a hyper-typed dog is, in a way, a "toy"... , 
- what can we do to prevent this new direction from eliminating all the others? This brings us back to the question of how we manage variability, as has already been amply discussed.

In reality, often, hyper-typed dogs are weakened with respect to certain ailments and their life-expectancy tends to be shortened. Their well-being is therefore compromised and the show-dog world, because of this, offers arguments to its detractors, especially when these belong to the radical fringe of the animal rights movement. They are now actively proposing, via European legislation, to get a ban imposed on the breeding of a large number of breeds that it describes as "tortured".

In order to prevent the excesses of animal rights, we should not simply reject all its arguments, refusing to enter into the debate, in brief clothing ourselves in our “dignity” but, to the contrary, make a point of calmly identifying the real deviations which are harmful to the animals and acknowledging that it is necessary to deal with them. The debate is not necessarily simple in scientific terms but it needs to be tackled. Its the collection of multiple observations and of discussion that will reveal the objectivity.

Convincing the various sides

We must first of all convince the national kennel clubs which, in absolute terms, is not necessarily easy. The European Convention on the welfare of pet animals has however aroused fear, and even undermined certainties. Even the British Kennel Club has been roused to action and has endeavoured to distinguish between the truth and exaggeration in the discussion of its principles. The British have agreed to recognise that six breeds possessed a standard where certain elements constituted a risk for the dog's health: Bloodhound, Bulldog, Clumber Spaniel, Pekingese, Sharpei, St Bernard (in alphabetical order); corrections have apparently already been made for the Bulldog and Pekingese and are in preparation for the four other breeds. As far as the remainder is concerned, the British Kennel Club, basing itself on statistics provided by insurance companies, challenges much of what is advanced (even by veterinarians) on the relationships that exist between morphological or anatomical features and specific diseases. As we heard at the last meeting of the FCI's Scientific Commission, it is important not to focus on the blockages as demonstrated by the Kennel Club, but rather to be happy that it has recognised the existence of a problem and taken steps, even small ones. It was noted at this same meeting that the question of hyper-types had been asked for the first time at the World Small Animal Veterinary Association (WSAVA) congress in Paris, in 1969!

Convincing the national kennel clubs is one thing, convincing the judges is altogether different. Firstly, the problem is less in possible modifications to be made to the standard than in the use that is made of the document itself. We are unable to resist reporting a recent fact, that we heard from Raymond TRIQUET, who was asking an English judge invited to a show in France whether he knew of the modifications made to the Bulldog standard by the Kennel Club. His reply was: "Judges do not read the standard, they know the breed" ! That being said, it is important not to lose sight of the fact that the standard must all the same be corrected when it  is an indirect incentive towards straight hocks, microphtalmia, luxation of the eyeball, etc., while still remembering that this will probably not be sufficient since, as we have seen, hyper-typed animals are often outside of the standard. This means to say that a strict application of standards (except on the few points that are disputed) should make it possible to take a big step forwards on the road towards preventing hyper-types. The judges therefore need to be kept informed and, what is more difficult, convinced. Perhaps experiments such as that conducted this year by the French Kennel Club at the Paris Agricultural Show might be of help here: breed champions were examined by a panel which included a scientist. This was an opportunity for interesting discussions.

At breed club level, it is important for discussion to take place regularly on trends that are observed in breeding selection, and for this to incorporate the opinions of a veterinarian who is himself aware of the issue. For this purpose, a critical reading of the standard(s) needs to be carried out and it is necessary to ask oneself questions about the dogs given awards in the show-ring. Results of the discussion are obviously to be communicated to the breeders by means of the club's newsletter, as well as to the judges, trying to explain diplomatically to those who tend to prefer the hyper-types why this choice is not thought desirable by those responsible for the breed.

The role of breeders is fundamental, but how can they be asked to adopt selection criteria that go against their interests, at least inasmuch as the situation has not changed in the preceding stages? Economic aspects take precedence, and this is perfectly normal. Any action at breeder level must therefore be in conjunction with information given to the general public. The media are an ideal means for this but unfortunately today they are used for everything – to circulate extremist tracts on animal welfare as well as to  encourage the popularity of hyper-types, in particular via advertising. When one knows the role television plays on fashion in favour of one breed or another, one can easily imagine the effect it can have on fashion in favour of just one type of dog. That is to say that hyper-types should be banned on television, but how can one do that? It is obviously the same for advertising on posters or in the press.

Other genetic drifts

What other genetic drift can be mentioned, other than that towards hyper-types? For the record, since the issue has been indirectly raised in the first section, there is a tendency towards a steep reduction of intra-breed genetic variability, linked to the excessive use, for reproduction, of a small number of stud dogs, then of their offspring. Let us remember the definition of a non-inbred animal - its parents have no ancestors in common over five generations – and ask ourselves how many kennel club registered dogs are non-inbred. Let us point out, which is not necessarily an excuse, that the situation is the same in farm animals. Everything goes on as if selection in itself, even if it has refrained from doing so, had not been capable of creating genetic progress without resorting to extensive inbreeding. In dairy cattle breeds, subject to artificial insemination, the reduction in what is known as the “genetic population” is considerable.

Another genetic drift is represented by the widespread expansion of certain hereditary anomalies or ailments. This is very similar to the previous problematic since the expression of recessive genes, often incriminated in medical genetics, is favoured by inbreeding. It is also linked to other factors, such as the absence of precautions prior to the large-scale use of a stud dog – here we will add: especially when it comes from abroad – the ease with which uncertainties relating to the ailment's genetic determinism lead to a conclusion that heredity is not involved, and the difficulty for those in charge of a club to know what is the exact situation of the breed with regard to the anomaly. Is it necessary to point out that the effort to prevent a hereditary ailment is in the interest of all breeders but involves the transparency and adherence of all parties to the programme for eradication?

We will come to an end, even though it is only indirectly that this is a genetic drift, with the eternal opposition between selection for "show-ring" and selection for "work". Such opposition has no reason for existence: even if the working dog no longer has the same overall importance as before, even if the nature of work required of the dog is changing, there is no justification for separating selection for "show-ring" from selection for "work": a dog must, by definition, be beautiful and good at the same time, with its "goodness" being appreciated through a function that is capable of evolving, or even simplifying (agility is a "minimum" function).

CONCLUSION

One does not breed dogs without passion, nor without finding a certain pleasure in it. It is nonetheless important not to underestimate the responsibility that this engages:

- with regard to the people, whether talking about the buyers who have made a big psychological investment in the purchase of a dog, or future generations of breeders who will in their turn be involved in managing the breeds,

- with regard to the dogs themselves, in respect of which ethical concerns are taking on more and more importance, which is fortunate provided one does not fall into extremist patterns.

Assuming one's responsibilities presupposes not necessarily letting oneself fall into habits nor being swayed by the prevailing fashion, but rather exercising a critical observation of one's own activities as a breeder, encouraging other breeders to do the same and exchanging experiences and thoughts at club level. Criticisms, experiences and studies will be all the better conducted and interpreted if an effort has been made to understand and, to a certain extent, accept the message from the scientists on the genetic drifts which are threatening the world of dog breeding.

Article resulting from the seminar of the Société Française de Cynotechnie de Nantes, held on 12 & 13 December 2003.

- English Translation by Susan Bamford -



 

 

 

 

REFLECTIONS ON THE GENETICS OF PUREBRED DOGS AND THE ESSENTIAL MAINTENANCE OF GENETIC VARIABILITY
Bernard DENIS
Professor emeritus at the Ecole vétérinaire de Nantes
Chairman of the Société d'Ethnozootechnie
The organisers of this colloquium have suggested a very well-defined framework by asking us to respond to three questions concerning: 
- registration without known parents, 
- safeguarding of breeds that are low in numbers, 
- fight against hyper-types.
We shall obviously not be limiting ourselves to a simple reply, which might be fast enough, but will use the opportunity that is given to us to expand on certain points in relation to the subject.
REGISTRATIONS WITHOUT KNOWN PARENTS – MUST THEY BE ENCOURAGED?
The question is always recurring: we know that the spontaneous tendency in the dog showing world, is towards closing the stud book, while the Ministry of Agriculture refuses to do so. 
Before giving a direct response, we should maybe remind ourselves of some of the elements relating to the concept of breed.
When does a breed start to exist?
For many people, a breed exists as soon as breeders are in agreement on the definition of a standard and have opened a stud book. This does not mean to say that, subsequently, animals without papers cannot be identified as being "of that breed" but that the starting point of the breed corresponds to the two events that we have just mentioned. These two events can be pinpointed in time, most often occurring in the second half of the 19th century and the first half of the 20th century.
This concept, which is the accepted way, has the merit of establishing a precise point of reference for the origin of breeds, but also the disadvantage of artificiality since one comes to believe that they did not exist before receiving what might almost be seen as their "birth certificate".
In reality, across all animal species, breeds are the result of a long process of regional  differentiation, that out-crosses have continually come along to modulate, but certainly not to overwhelm. For farm animals, up until the 19th century, it is important not to exaggerate the importance of out-crosses, the effects of which end up being diluted in the shared genetic pool of the regional population concerned. They were of course far more  important in dogs, although this species does not however get away from the standard description of the situation as it may have existed before the definition of standards and the opening up of stud books: heterogeneous populations existed in all regions, from within which sometimes emerged a particular type, which was recognised as that of a future breed. If many breeds have taken on the name of a region, this is precisely because they existed before their official “birth” and breeders had been able to reach agreement without difficulty in order to identify them. Any belief that, across all species, the animal populations were in virtually disorderly variation and that breeds owe their existence to the entirely artificial intentions of mankind at the end of the 19th century, is erroneous.
At the time of a breed's “official creation”, man sorts through and identifies animals that conform most closely to the chosen standard and, because this selection is by phenotype, will be forced to continue in this manner over successive generations: it is in fact quite possible that an animal may have an interesting phenotype more or less by chance, but in reality being the result of a cross-breeding which may distance it from the genotype of the regional population and produce offspring that are far too disparate.
Much later, can it be considered that all animals of the breed are derived from what we call the  "foundation stock", i.e. those resulting from the initial selection? This must have been rare. It is probable that, in one way or another, rightly or wrongly, genes coming from unregistered animals even though belonging to the same population or appearing to do so, were introduced on a regular basis.
At its beginnings, one can therefore assume that a breed, in a given region, includes a proportion of foundation animals – the most beautiful in accordance with the newly established standard – and a common population, which is considered not in conformity with the desired model but nonetheless resembling it. Naturally, only the former are entitled to the breed name. As and when selection increases homogeneity in the group of registered animals, the breed makes itself better known and registered males are used to mate “common” females. The overall heterogeneity of all the animals in the old regional base is somewhat reduced but obviously does not entirely disappear.
NB: this situation corresponds to the beginnings of the dog showing world. In the 1950s, when some breeds were re-created, it was still possible to find farmyard dogs expressing the old regional type, but it would no doubt be far more difficult today.
Is a breed only made up of registered animals?
One can understand that the law will reply in the affirmative. One can also understand that kennel club breeders will adhere to that idea, with unregistered dogs being described as of "breed appearance" or, pejoratively, as "without papers". Scientifically, however, this is wrong. The "papers" provide official recognition or guarantee, but the breed is recognised by the "eyes", even if they can be mistaken.
In farm stock, the percentage of animals registered in the stud book has always been in the minority. The French Law on Breeding in 1966 has introduced a new category of animals: those which, while not being registered with UPRA (organisation holding the stud book), are nonetheless subject to declarations of birth and control of performances (check on milk production, on growth, etc... ) and constitute the breed book. Animals that are not with UPRA, nor in the breed books, are simply identified. The three categories include “breed” animals but certification is only actual for those registered with UPRA.
This situation is not without interest since, while there are many more chances that common animals benefit from the efforts of selectors (resorting from time to time to a male registered for reproduction) than the reverse, the very existence of a population without specific references constitutes an important reservoir of variability, within which can be found interesting examples. If necessary, the means of using them legally will always be found.
If selectors registered with UPRA are aware of owning the best animals (which, in fact, is not always the case), it would never occur to them to use such expressions as “breed appearance” or “without papers” pejoratively, to designate animals that are not registered but of the same breed.
Once again, if registration in the stud book certifies that an animal belongs to a breed, it is only the eye that confirms it, with some risks of mistake, it is true. The assimilation that "pure-bred animal = animal with papers" is only made in Dogs and Cats, usually before being decided by law, for various reasons.
Are animals of the same breed very similar?
For them to be similar is obvious, otherwise they would not belong to the same breed. But must they be very close, one to another? The answer is no. The greatest possible homogeneity is the prerogative of inbred lineages. It is normal and even desirable to have the coexistence of different types in a breed, in terms of morphology, behaviour, etc... Naturally, the extent of such differences must not be too wide and should be specified within each club.
The reasons why a breed must not become too homogeneous are twofold: 
- if tastes change and in the future one looks for a slightly different style of dogs, it must be possible to get the breed to evolve under its own steam, 
- homogeneity is most often attained at the price of an insidious increase in the coefficient of inbreeding, for which there is no shortage of unfavourable consequences.
Consciously or unconsciously, it always seems to have been towards “greater homogeneity” that selection for breeding has moved. In order to facilitate this process, in breeds where there are several varieties, each variety is made to reproduce independently of the others, which can lead to an accentuation of the differences and for them to be considered as separate breeds, which was not the case to start with. It cannot be overemphasised that the varieties must be crossed with each other from time to time, since they form a sort of reserve supply for the introduction of new blood from one variety to another!
Obviously, buyers often have specific ideas on the dog that they want – the one they saw on television for example – and, at the end of the line, they will not accept an animal that they consider is too different from what they expect. It is a question of educating the potential buyer: logically, a breeder cannot guarantee a specific type of dog except through inbreeding; the buyer must be informed of this and, of course, agree to paying for this guarantee.
The question is then to know just how far to go in this reasoning. Since we have been holding to it, we have heard remarks such as: "You are suggesting that we go backwards and undo what we have done". A German colleague, also a zoo-technician, one day even said to us approximately as follows: "At the end of the day, you are against selection... And yet, the interest for us as zoo-technicians, is to achieve evolution in populations. When there is no more variation, we move on to something else. If the breed has become inbred and looks like disappearing, there will always be plenty of others to keep us occupied" (!)
We are obviously not against selection although we do assert that this needs to be implemented in a context of breed management, which involves: 
- certainly getting to evolve in the desired direction for the current circumstances, 
- but also to preserve enough variability in order to possibly move on in a different direction at a later date.
It's all a question of proportion. This year we had the opportunity of seeing a gathering of some thirty dogs of the Corsican "breed", Ù Cursinù, and around the same number of Savoy Sheepdogs and (or) Alpine Sheepdogs. It is clear that current heterogeneity is too high in these two populations and that selection is going to have to reduce it, although this is nothing new: all dog breeds have started in this way. It will be necessary to know how far to go and, above all, at a certain point, to avoid favouring the offspring of any particular stud dog to the point of changing the characteristics of the whole breed!
It is therefore quite clear that  animals of any particular breed must obviously express sufficient characteristics that are shared but should not in any way be carbon copies of each other.
How to maintain intra-breed variability?
The issue is difficult if one tackles it from a practical point of view but it cannot be ignored.
No doubt it is the responsibility of the clubs, at a given moment, to draw up a sort of “inventory” identifying what exists in terms of foundation stock, families, bloodlines, etc... in the breed, then to ensure, over subsequent generations, that none of the identified individuals disappears from reproduction. If the risk of this happening for any one of them, consideration must be given to the sperm bank. Indeed, what is the point of freezing sperm from outstanding studs that have already been much used for reproduction? To the contrary, it would be better to keep in reserve the characteristics of studs that are not much sought after but which maybe one day will offer an interesting introduction of new blood. All this is only valid, of course, if one is looking at it from the point of view of breed management, in which case the general interest should prevail; the financing involved can therefore only be assumed collectively, which is not easy to put in place.
It is good that dogs winning at shows should not always be the same style. There is therefore, here again, a study to be conducted within the club, and suggestions to be formulated with judges. The judge's role is primordial, as much for the expression of a “broadness of outlook” as for the educational skills with which they will need to explain their judgements.
In breeds where there are a number of varieties, cross-breeding between the varieties must be at least tolerated: while certain matings may run the risk of poor results (we are thinking of certain colours or different ear carriages), breeders can be strongly advised against such practices, with explanation of the risks being incurred, but they should, in all logic, be prohibited.
Limiting the number of litters allowed for the most popular stud dogs is a good step to take. We will not suggest a figure since the response obviously depends on the breed. It is more a matter of the percentage of matings performed in a given generation, than their quantity. Identification of stud dogs who have reproduced in the previous generation and the number of matings that they carried out, is a good starting point for analysis, with common sense being more important than mathematical calculations.
A practical method should be encouraged: that of a regional breeding programme,   as an incentive for breeders to agree amongst themselves in order to conduct a joint breeding programme, with animals that they know very well. This is probably not the best way to produce champions but it is a very good way of avoiding the introduction of hereditary problems in their own line (sometimes, by champions... ), provided of course that there is total transparency between breeders. For what interests us here, this method of doing things consists of setting up founder lines, which must essentially differ from one another, at least to a certain extent, and form reservoirs of variability.
It is also desirable from time to time to seek out other bloodlines, thereby in a way introducing new blood from within the breed. These can first of all be looked for abroad, assuming such lines do exist. One might also resort to using dogs without a pedigree (defining with those responsible for the breed the means of doing so legally), which brings us back to the question asked at the start of this first study of  intra-breed genetic variability.
The use of registrations without known parents
We have seen, up until the present day, that: 
- at the beginning, breeds probably drew for a certain time from the "reservoir" of unregistered animals, so as to genetically enhance the descendants of the initial founders, 
- from a strictly scientific point of view, animals that have the appearance of a breed are part of that breed, since it is the eye that signs – in all probability – for an animal belonging to a breed and not the "papers", 
- even if the tendency, in the dog world, is to have breeds that are always becoming more homogeneous, it is still essential for them to conserve a variability that will, possibly, enable them to evolve of themselves in another direction.
One can easily imagine therefore that dogs without papers are worthy of attention, that they are likely to reinforce a chosen orientation or quite simply to provide an interesting introduction of new blood. Obviously, therefore, it is desirable to proceed with complete openness for their use and, as a result, to register them as without known parents.
That being said, there are three types of dogs without papers: 
- dogs from kennel club registered breeding stock that, for one reason or another, have left the system: these are not true "unregistered" stock and they do not provide an introduction of new blood in the proper sense of the term; 
- dogs that have had a studbook origin, but a long time ago, or even that may have animals in their ancestry from the old regional base stock of which we have spoken. These are particularly interesting, even if their morphological type deviates from the model currently preferred among kennel club breeders; 
- subjects that may express an interesting phenotype but have an input of blood that is foreign to the breed during previous generations. If introduced into the selected breeding stock, these are likely to cause damage. 
The problem is that in the current state of affairs, it is very difficult to know which of the three categories pertains to any given dog that may appear interesting.
This means to say that registration without known parents involves paying very critical attention with regard to the offspring of a newly registered animal. It will be very quickly apparent if it belongs to the first or third category, due to the appearance of subjects identical to those from French kennel club registered breeding or, to the contrary, litters that are very heterogeneous. If it belongs to the second category, resulting in an intermediary situation for offspring, this is perceived differently by those in charge of the breed, certain of whom will admit to being very disappointed by the results and announce themselves from that point on, as being against such registration without known parents. This is probably because they had in mind the model looked for under French kennel club registration and had forgotten that the role of dogs under registration without known parents is not necessarily to pass this on... Others however will be happy to rediscover types of dog that had been forgotten within the French studbook and will make every effort to reintroduce them.
Registration without known parents presupposes that the clubs allowing this system understand completely that it is done in a context of openness, for the purposes of genetic management within the breed. However it is always uncertain. Such uncertainty is acceptable provided all precautions are taken for being able to block the way for a product that may show itself to be a bad influence. We do not believe that this possibility exists in the current state of affairs under French kennel club rules. It may be appropriate to have a review of the methods of registration without known parents by bringing it closer to a “waiting register”.
It is important to put such registration without known parents in its rightful place, that of a method making it possible to make the link between dogs with kennel club registration and those without papers, so as to return to a genetic diversity that, often, the official selection has been unable to conserve. Must this be encouraged? Maybe not, although, to the contrary, it should not be rejected, as is often the case with the chairmen of clubs seeking to close the studbook. They need to recollect that the breed for which they are responsible is not intended to become a completely inbred homogeneous population, and that registration without known parents requires cautious implementation.
BREEDS WITH VERY LOW NUMBERS – WHAT CAN BE DONE TO PREVENT THEIR DISAPPEARANCE?
In farm animals, there is a new awareness at international level that many breeds with low numbers run the risk of disappearing, leading to the implementation of measures for their protection with greater or lesser efficiency depending on the countries and areas involved. This is just one aspect of the preservation of biodiversity, so badly served over recent decades. Even the Dog seems to have been overtaken in its turn by this concern relating to its heritage – one can only rejoice about this.
Three topics deserve to be covered under this heading: ensuring that the populations that are disappearing are in fact original, investigating the causes of their disappearance, implementing suitable measures of genetic management.
Ascertaining the reality of genetic losses
First of all, it is a matter of asking oneself if the breeds that are tending to disappear were in fact real breeds, in which case there is indeed a genetic loss, or not. The obvious example for the veterinary college in Nantes is that of the Levesque, created just 300 yards away as the crow flies from the college, at the château de la Poterie, by two brothers Donatien and Rogatien LEVESQUE, which disappeared with them. Although recognised as a breed, it was  certainly not a breed, but rather a variety or branch of Gascon-Saintongeois. The Dupuy Pointer was probably not a breed either (we should point out, however, that we have not really studied the question). The disappearance of an individual branch is obviously regrettable but does not bear prejudice to the future of the breed to which it belongs. The disappearance of the branch or, rather, inbred line known as Laverack, did not prevent the English Setter from becoming what it is.
Next, for a breed with low numbers and an uncertain future, it is important to know whether the surviving animals really belong to the old population. There are relatively recent examples of breeds that have been completely recreated: it would seem pointless trying to preserve a genotype produced in reality from cross-breeding.
Looking into the causes of a breed's disappearance
When a breed is moving towards its disappearance, it is of interest to look into the reasons for such a state of affairs, with a view to possibly being able to remedy the situation.
There is one situation where not much can be done – where the breed was linked to a particular function, with little likelihood of redeployment. A recent example is that of the Berger de Crau, which used to accompany the flocks of sheep in the South-East when they were moved, on foot, to different pastures. Some do still exist – a study is in progress - but, for the summer pastures, other breeds are now preferred. As for the possibility of converting such a rustic breed, with its long and felted coat, into a house pet, this is not likely. More generally, there are also utility breeds that find it difficult to adapt to life as a companion animal either because they are unable to tolerate the new living conditions offered to them or because their upkeep (coat, cleanness, etc... ) is considered a problem.
In the case of hunting dogs, it is sometimes relevant to make a difference between the wishes of the hunter himself, who may prefer a certain type of hunting and a breed of hound well suited to it, and the orientation given by field trials. The problem goes far beyond the scope of breeds with low numbers – it is widely accepted that Field-Trials, in their current form, have even gone so far as to greatly favour English breeds to the detriment of the continental breeds, some of which have even gone so far as to become "anglicised", overtly or covertly. However, we have always been told that there were still enthusiasts – and apparently more and more of them – for continental hunting! Under such conditions, a major study should be conducted – if one is promoting the preservation of genetic diversity, competitions that tend to reduce it are poor competitions and should therefore be made more flexible or be conducted in a different manner. The problem is exactly the same in the case of herding dogs, where it appears to us to be necessary to design competitions reserved to continental dogs. To return to hunting and to breeds with low numbers, it must not be forgotten that, in the picture that is given of this activity to the general public, ever greater emphasis is placed on the contact with  nature and the pleasure of being accompanied by one's dog and seeing it work. Even if this dog belongs to a breed that doesn't “perform” particularly well, one wonders why it is necessary to encourage its owner to change it, since he is happy with it and is aware of taking part in the conservation of a regional breed. Improving the qualities of such a breed, not necessarily based on the dominant model, is a question of selection and time, from the moment that the numbers have increased sufficiently to make this possible.
It is possible to imagine that a breed can see its numbers dwindle, because of becoming a problem to breed, due to the high frequency of one or more hereditary ailments, a marked reduction in life expectancy, etc... This scenario is theoretical since, while it might have the merit of providing objective explanations for the breed's situation and indicate the means of recovery, experience shows that the factors evoked rarely have a dissuasive effect, either through lack of information, or through a taste for risk or a search for the exceptional.
The human factor is often important in breeding. It happens that a breed ends up being forgotten because of a loss of dynamism, even motivation on the part of a few breeders who remain faithful to the breed. When numbers have grown scarce and any presence in the show-ring has become exceptional, one can understand that the dog showing fraternity believes that it has disappeared.
Lastly, fashion can obviously come into play, a trend against which it is very difficult to argue. When acting negatively against a population which was already numerically challenged, the consequences can be disastrous.
In the five situations which we have just envisaged – there are of course others – we know, theoretically at least, on what we need to take action in order to reverse the situation; putting this into practice is often more complicated but at least the way is outlined.
There is another aspect in the issue of saving breeds with very low numbers, which is the manner of carrying out reproduction in order to conserve sufficient genetic variability.