Thursday, 29 November 2012

Inbreeding is quite safe

Click to enlarge/read

A fabulous bit of archive from 50 years ago in this week's Dog World... dismissing concerns about inbreeding in response to what the author calls "the sheer volume of uninformed criticism which has lately appeared in the lay Press and been given public utterance by those who should know better than to indulge in such scientifically unsound generalities".

Those who should know better were, in fact, senior vets - not, of course, that they are necessarily experts in genetics. But then neither was Dr E Fitch Daglish.

Daglish's article was in response to the first high-profile challenge in the early 1960s to the way dogs are bred under kennel club rules. This included an article published in The Field on 8th Feburary 1962 and, again, in December 1962, in the Canadian Journal of Comparative Medicine and Veterinary Science. (When A Vet Should Speak Out, Can J Comp Med Vet Sci, 26 (12), 1962). Despite the title, the author chose to remain anonymous - an indicator, perhaps, of the pressure vets felt even then about not alienating their best customers.

The anonymous vet wrote: "Breeding from defective animals not only destroys bloodlines, but discredits a breed both at home and overseas. Unfortunately, this warning sometimes falls on deaf ears. It is a great temptation to the owners of a champion to disregard or deny any defects that are repeated in its offspring. Their natural inclination is to keep quiet about the matter. There is no suggestion that dog breeders and exhibitors are deliberately dishonest. Their worst faults are ignorance and a disinclination to face up to unpleasant facts."

The author went on to refer to an article published in 1961 in the Journal of Small Animal Practice (JSAP) in which Dr S Hodgman stated that the Executive Committee of the British Small Animal Veterinary Association had become 'very concerned' at the serious increase in the number of pedigree dogs suffering from defects and abnormalities.

"Dr Hodgman's central point was that there is strong evidence that many of these conditions are hereditary. The present situation may have arisen because of the ignorance of many dog breeders of elementary genetics. Irrespective of the causes, the alarming fact remains that the percentage of unsound stock is on the increase."

Dr S F Hodgman was a founding member of the BSAVA and a senior research vet at what is now the Animal Health Trust in Newmarket.  Hodgman wrote several articles on the issues at this time, including:

Hodgman S, ‘The duties of veterinary surgeons in attendance at championship dog shows’, Brit Vet J, 117(5), 1961*

Hodgman S, ‘Abnormalities in pedigree dogs: Their surgical correction, related to Kennel Club Rules and Regulations’, J Small Anim Pract, 2 (1-4), 1961*

Hodgman S, ‘Abnormalities of possible hereditary origin in dogs’, Vet Rec, 74(46), 1962*


This third paper, presented to the BVA Congress in September 1962 and picked up by the newspapers of the time, is the chief reason for claims of an "attack" in the above article by Daglish. But in fact, it is clear from the history books that Hodgman worked quite closely with the Kennel Club. His articles are measured and, although critical of some breeders, he was not unsupportive of the KC - maintaining that it was not their job to tell breeders how to breed their dogs.  In fact, the Kennel Club collaborated with a survey - largely prompted by Hodgman -  into inherited problems. 

The following year, in 1963, came this article:

Hodgman S, ‘Abnormalities and defects in pedigree dogs: I, An investigation into the existence of abnormalities in pedigree dogs in the British Isles’, J Small Anim Pract, 4(6) 1963 *


In this, Dr Hodgman wrote“The preliminary results of an investigation into the existence of certain deleterious conditions that are hereditary are presented. Thirteen conditions found to be of major concern and of these five were considered to be of importance and needing immediate consideration are hip dysplasia, patella luxation, entropion, retinal atrophy, and prolonged soft palate. The second priority group consists of abnormal temperament, skin fold dermatitis, uterine inertia, elbow dysplasia, ectropion, trichiasis, and deafness”. 

That same year, during his opening address to the 6th BSAVA Congress, president Dr Oliver Graham-Jones announced: "We have recently been to the House of Commons on your behalf and met many members of both Houses. We told them of our tremendous interest in the abnormalities of some of the dogs that we are called upon to treat; and explained that our concern is that dogs are being bred and born into this world to suffer throughout their lives from certain conditions which probably could be prevented." 

The issues were also discussed during a symposium at that year's Congress entitled Abnormalities and defects in pedigree dogs and six accompanying papers were published in JSAP - the last one by Hodgman referenced above, plus:

Hein H, ‘Abnormalities and Defects in Pedigree Dogs-II. Hereditary Aspects of Hip Dysplasia’*

Knight G, ‘Abnormalities and Defects in Pedigree Dogs–III. Tibio-Femoral Joint Deformity & Patella Luxation’*

Barnett K, ‘Abnormalities and Defects in Pedigree Dogs–IV. Progressive Retinal Atrophy’*

Willis M, ‘Abnormalities and Defects in Pedigree Dogs—V. Cryptorchidism’*

Frankling E, ‘Abnormalities and Defects in Pedigree Dogs—VI. The Breeders' Point of View'*

In fact, some action did follow. The fuss was directly responsible for the founding of the BVA/KC Hip Scheme (largely thanks to strong lobbying by geneticist Malcolm Willis) and the concern about crypotchordism prompted the Kennel Club to bring in the rule that all dogs must have "two fully descended testicles" - hence why all show judges today fondle them in the ring. 

But the inbreeding message got lost - until, pretty much, Pedigree Dogs Exposed in 2008 - and inbetween there has been almost 50 years of breeding that has led to a loss of genetic diversity that may well prove cataclysmic in some breeds, as is now more generally acknowledged.

Daglish (a naturalist, engraver, dog show judge and author of several books on dogs) writes above: "It has for long been commonplace to ascribe such undesirable traits as loss of size, lack of resistance to disease, infertility or the incident of physical deformities or abnormalities, to say nothing of mental aberrations, the practice of inbreeding; wholly disregarding the fact that if such defects come to light as a result of inbreeding it is the fault, not of inbreeding as such, but of faulty selection of the partners to a mating."

Of course some breeders still maintain this (and there were some very strong editorials in the dog press defending  inbreeding following Pedigree Dogs Exposed). But we now know that they are wrong. Well, mostly.

We now know that inbreeding in and of itself does indeed lead to the very problems Daglish denies above, irrespective of specific, disease-causing deleterious recessives. This is because it leads to not just the fixing of traits that are good and desirable, but homozygosity in other areas of the genome, reducing  fertility, fecundity and our dogs' ability to respond to life's onslaughts. This is why every conservation programme of wild species (and indeed domesticated rare breeds) centres on trying to maintain as much genetic diversity as possible. 

Sameness is the spice of life? Doesn't have quite the same ring to it, does it?

Good selection - particularly that which mimics natural selection in ensuring that only the fittest get to breed - can mitigate these effects for a while. This is why we do see some fit wild populations that are very inbred. It may even prove possible for one or two dog breeds of the many to come out the other side of a lot of inbreeding, having purged all deleterious recessives. But the odds are stacked against it, hence the increasing focus on the importance of maintaining genetic diversity in dog-breeding.

*I'm afraid none of the above-cited articles are freely available online, hence no links - but if you would like to see them, please email me privately.

Many thanks to bulldog historian Stuart Thomson for some of the history above (see here).

58 comments:

  1. "We now know that inbreeding in and of itself does indeed lead to the very problems Daglish denies above, irrespective of specific, disease-causing deleterious recessives."

    Dagliesh includes physical deformities or abnormalities ....

    So you're saying that if a dog and bitch are genetically clear of a certain condition, but are very closely related, their offspring are at risk of having that condition?

    ReplyDelete
    Replies
    1. No, Anonymouse - by the way, there´s no risk involved in putting your name to a post! - No, Jemima is not saying that if a dog and a bitch are genetically proven clear of condition X, then because they are very close relatives, if they are mated condition X will (miraculously) appear in their offspring. Jemima and anybody else with a basic understanding of genetics is saying that condition X is the one thing they will not have!
      Unfortunately, the probability of a great number of other conditions, many of them not possible to foresee, will be greater.

      Delete
    2. Exactly, Bodil. Inbreeding cannot create hereditary genetic conditions that aren't already carried by the individuals; it highlights them, making them known and therefore easier to avoid in future, rather than masking them the way outcrossing does.

      Delete
    3. To add; there is less need for 'test mating' of close relations now that there are DNA tests for a number of conditions to determine whether or not they are carriers, but there is still a way to go. And of course if you can produce individuals homozygotically clear of a hereditary condition (such as one hopes will be produced by breeding together some of the few heterozygotic LUA dalmatians) how much more quickly a remedy can be spread through a population. If that means close inbreeding it should be considered as a means to an end.

      Delete
    4. There are 20,000 genes in the dog. The DNA tests for single-gene recessive disorders - which by no means can be considered a health test; they're just tests for a handful of horrible inherited conditions - measure a tiny, tiny part of all that goes into the genetics of ensuring a dog is fit and healthy.

      There are no plans to make the LUA Dals homozygous for the LUA gene - indeed the opposite, because of the concern that throwing away the HUA gene might have unintended consequences. Again, genes are very rarely discrete entities, only coding for one trait. Even "bad" ones may be exerting a desirable effect in other respects. (EG in humans, the gene for sickle-cell anaemia confers resistance to malaria).

      Delete
    5. Mary, like Jemina says: if a dog breed had TWO GENES ONLY, and one of them carried a recessive harmful variety, whereas the other one commanded all the other myriads of functions on the mammalian body and IF that other one was known to be intact and healthy - yes, then indeed your aergument for inbreeding (it brings out and demonstrates exactly where we find the harmful receseeive gene) would be valid.
      Ufortunately, that is not the case. Following your recipe, we would certainly find out rather quickly - in a few generations - just where the mutated allele was and just what it caused. We would also, over rather more generations, discover a number of other genes out of the 20 000 that were out of working order, as shown when when doubled... Which seems to be exactly what has happened to a number of breeds.

      Delete
  2. No, of course not. The genetic disposition has to be there. The ills I'm referring to are the more generalised ones such as less vigorous puppies, reduced fertility/fecundity, a higher incidence of immune-mediated problems. In humans, inbreeding is known to reduce intelligence - and increase the number of birth defects. These are signs of inbreeding depression and well-documented.


    ReplyDelete
  3. So a one-off closely inbred mating, followed by a few generations of more distantly-related matings, need not do any harm at all but can indeed 'strengthen' beneficial genes (as Dagliesh says, by choosing the breeding stock wisely).

    ReplyDelete
    Replies
    1. It NEED not do any harm, but it very well might. It increases the gamble of deleterious alleles pairing up and then influencing the gene pool down the way, especially if the inbreeding is behind a popular sire.

      As an example, how do you believe the very RARE condition - Episodic Falling Syndrome - now has an estimated 20% carrier rate in the Cavalier King Charles (a reported 30% amongst whole colors). It is a condition so rare it is not heard of in other breeds. That does not mean, however, it is not lurking in the odd individual and if you, by luck, decide to inbreed on THAT individual you would create a problem - especially as it is incompletely penetrant and an affected can be sitting in front of you looking perfectly healthy.

      There are 20,000 genes on the canine genome. Do we really want to go trolling for more of these types of conditions?

      Kary

      Delete
  4. That depends on a host of things - including how close and how distant. But I'd agree that a one-off close mating in an otherwise outbred pedigree is less harmful than generations of closer breeding.

    ReplyDelete
    Replies
    1. That's right; repeated outcrossing eventually loses breed 'type' and can increase the level of recessives, resulting in some very nasty and unexpected surprises, and repeated inbreeding loses diversity whilst exposing recessives, for good or evil. The wise breeder opts for something in between the two.

      Delete
    2. What are you speaking of when you say "increase the level of recessives"? If you are speaking to frequency of specific recessive, outcrossing does NOT increase frequency, in fact quite the opposite. Inbreeding certainly does increase frequency unless they are recognized and culled, and with the case of many late onset and incompletely penetrant conditions that commonly does not happen.

      Kary

      Delete
    3. What utter nonsense.

      Delete
  5. As I see it, breed type is maintained by selection, hence selective breeding. Then inbreeding/linebreeding was found to be a shortcut to "improve" type but, as time went by, it turned out to be one with considerable risks.
    People are free to inbred/linebred their dogs (until some degree) within the system yet selective outbreeding is practicaly forbiden when it could be another useful tool.

    ReplyDelete
  6. "Those who should know better were, in fact, senior vets - not, of course, that they are necessarily experts in genetics. But then neither was Dr E Fitch Daglish." just remind us of your qualification in genetics Ms Harrison? Degree level, A level, O level ..............Brownies Badge level?

    ReplyDelete
    Replies
    1. The idea that close inbreeding is acceptable is not upheld by the majority of research. Research that has been done by very qualified people. Do you disagree that inbreeding is a bad idea, or are you trying to draw attention away from an important issue by mocking people who have good intentions?

      Delete
    2. Anon is pointing out the irony of pots and kettles, Jordan.

      Science has never been able to prove that occasional close inbreeding of healthy individuals is damaging; it's continual repetition that can become harmful.

      Delete
    3. I don't see any irony here. Ms. Harrison uses research that has been done by very qualified scientists. Not having a degree doesn't mean you can't understand what scientists tell you, it just means you probably shouldn't do the research yourself.

      Isn't the problem here that there IS repetitive inbreeding? With most breeds descending from just a few dogs, some inbreeding is inevitable, but the amount that currently goes on is unacceptable. Yes, two perfectly healthy dogs will have healthy offspring, but genetic diversity is still reduced, and the risk of an unknown disease being passed on is greater.

      Delete
  7. I see nowhere where out crossing is "forbidden" or even cross breeding.. NOTHING keeps people from doing whatever they want to in their breeding programs..want to cross a poodle with a Great Dane.. no one will stop you.. want to cross breed two already cross breds.. no one will stop you and yet the KC stops people from registering certain breedings based on what?/ nothing but hype and comparisons to humans.. but since most people here bash the KC's from their respective countries I do not see why they just don't do their own breedings and mind their own business.. don;t like Pekes..?? who cares.. breed your own version of the "Peke" want a Pug that has a long nose.. breed one..
    what was they said about that mousetrap?

    If a man has good corn or wood, or boards, or pigs, to sell, or can make better chairs or knives, crucibles or church organs, than anybody else, you will find a broad hard-beaten road to his house, though it be in the woods.
    —Ralph Waldo Emerson,

    ReplyDelete
    Replies
    1. Why people bother? Well... because dogs are living creatures who suffers because of the monstrously bad and unscientific breeding programs that many purebred breeders use.

      Delete
    2. So put your money's where you mouths are and start breeding better pugs and bulldogs then come back and tell us all about it.

      Delete
    3. Anno: That was a bad argument. I hope you understand that you don't need to breed dogs at all to see the bad breeding used? Dogs are not unique when it comes to population genetics, but i prefer to read scientific litterature about the subject and base my opinions on this rather than guess like some people.

      Delete
  8. The simple fact is we do not fully understand the dog genome, we can spot the known problems but a host of unkown variables remain, which is why inbreeding is generally a bad idea

    ReplyDelete
  9. Not quite sure whether the software ate my previous comment, so here goes...

    Inbreeding increases the percentage of homozygous dogs. If there are recessive bad alleles in the population, inbreeding therefore tends to bring these out, and the result is called inbreeding depression. However, this also means that if an inbred population is subject to selection for fitness, this selection will eliminate these bad alleles much more efficiently than in a non-inbred population, as fewer of them survive selection in their heterozygous form. This is called “purging”, and it renders the purged population much less susceptible to inbreeding depression.

    This is more than a theoretical consideration, given that many dog breeds were subject to strong bottlenecks in combination with strong selection for fitness during WW2, which was a time when veterinary medicine was far, far from its current capabilities that have a habit of reducing selection pressure. One example where there is some scientific evidence for purging is the Irish Wolfhound – a breed that went to two severe bottlenecks associated with WW2, and where the oldest dog on record lived to be 16.5 years old and had an inbreeding coefficient of over 39% over five generations.

    Now, imagine such a purged dog breed with fairly low numbers whose gene pool is threatened by a popular sire that is known to carry a recessive disease. Reducing short-term inbreeding depression in such a breed will result in this sire being behind every dog within a couple generations, which is obviously not a desirable outcome. On the other hand, keeping several distinct more inbred lines will preserve more genetic variability and result in a continued base for selection outside of the popular sire. This sweeping one-size-fits-all approach to inbreeding depression is therefore not applicable to every breed and may very well make the situation worse in some breeds. The road to hell is paved with good intentions, as they say.

    Interestingly, there is virtually no scientific work that looks at inbreeding depression in individual breeds. This should warn us about applying general “recipes” to each and every dog breed, when in fact the genetics of these breeds may vary considerably and would have to be studied in detail before a course of action that would in fact stand a chance of improving the situation could be decided on.

    On a non-dog note, Chillingham Cattle are a pretty interesting example of purging (http://en.wikipedia.org/wiki/Chillingham_Cattle): They have been highly inbred for over 300 years and do not seem to mind. Of course, they are a semi-feral breed, so the point about strong selection for fitness does apply.

    ReplyDelete
    Replies
    1. Anonymous wrote: "Inbreeding increases the percentage of homozygous dogs. If there are recessive bad alleles in the population, inbreeding therefore tends to bring these out, and the result is called inbreeding depression."

      It's also called "lots of affected offspring". If we're talking dogs, "eating one's mistakes" isn't an option as it is in livestock...

      And then there's the MHC...

      Anonymous wrote: "However, this also means that if an inbred population is subject to selection for fitness, this selection will eliminate these bad alleles much more efficiently than in a non-inbred population, as fewer of them survive selection in their heterozygous form. This is called “purging”, and it renders the purged population much less susceptible to inbreeding depression."

      Riiiight! Ask the Basenji folks how well that worked for them. They tried to eliminate one recessive trait, found they couldn't, and discovered other nastiness they'd never known was there.

      And take a look at the bearded breeds that have been created from landrace breeds with both bearded and smooth-faced varieties. (The Tibetan breeds come first to my mind, of course, but there are many others.) Even after many, many generations of selection entirely for the dominant trait (bearded), smooth-faced pups *still* show up. And it doesn't take expensive genetic tests to reveal the lack of a bearded face. And a "smoothie" can't be passed off as a "clear" or "carrier" by a deceptive breeder.

      (I wonder if there is *any* bearded breed that doesn't produce the occasional "smoothie"...)

      Anonymous wrote: "This is more than a theoretical consideration, given that many dog breeds were subject to strong bottlenecks in combination with strong selection for fitness during WW2, which was a time when veterinary medicine was far, far from its current capabilities that have a habit of reducing selection pressure."

      And how much secret crossbreeding has gone on since then?

      (to be continued...)

      Delete
    2. (continued)

      Anonymous wrote: "One example where there is some scientific evidence for purging is the Irish Wolfhound – a breed that went to two severe bottlenecks associated with WW2, and where the oldest dog on record lived to be 16.5 years old and had an inbreeding coefficient of over 39% over five generations."

      As we all know, one example doesn't prove *anything* except that that one dog was very lucky. Even if that entire litter had been long-lived, the N would be too small to be meaningful.

      Anonymous wrote: "Now, imagine such a purged dog breed with fairly low numbers whose gene pool is threatened by a popular sire that is known to carry a recessive disease. Reducing short-term inbreeding depression in such a breed will result in this sire being behind every dog within a couple generations, which is obviously not a desirable outcome. On the other hand, keeping several distinct more inbred lines will preserve more genetic variability and result in a continued base for selection outside of the popular sire. This sweeping one-size-fits-all approach to inbreeding depression is therefore not applicable to every breed and may very well make the situation worse in some breeds. The road to hell is paved with good intentions, as they say."

      I assume that "purging" is the "one-size-fits-all" approach being referenced? I'd say it would make the situation worse in virtually any breed.

      Anonymous wrote: "Interestingly, there is virtually no scientific work that looks at inbreeding depression in individual breeds. This should warn us about applying general “recipes” to each and every dog breed, when in fact the genetics of these breeds may vary considerably and would have to be studied in detail before a course of action that would in fact stand a chance of improving the situation could be decided on."

      "More study". "Not now". I.e., wait until it's too late and crossbreeding is the only possible way to salvage a "breed".

      Anonymous wrote: "On a non-dog note, Chillingham Cattle are a pretty interesting example of purging (http://en.wikipedia.org/wiki/Chillingham_Cattle): They have been highly inbred for over 300 years and do not seem to mind. Of course, they are a semi-feral breed, so the point about strong selection for fitness does apply."

      "The Chillingham cattle herd are not domesticated in any way, and are wild animals," says the Wikipedia article. If they are, indeed, wild and their breeding is not interfered with by humans, no doubt the cows do all that *they* can to avoid inbreeding...

      Delete
    3. Judy said ""The Chillingham cattle herd are not domesticated in any way, and are wild animals," says the Wikipedia article. If they are, indeed, wild and their breeding is not interfered with by humans, no doubt the cows do all that *they* can to avoid inbreeding..."

      There's nothing they can do to avoid inbreeding, other than never to mate. They are all so closely related as to be considered genetically identical.

      Delete
    4. For information:

      http://www.chillinghamwildcattle.com/beasts

      Delete
  10. There is limited evidence for purging in inbred populations (http://link.springer.com/article/10.1023%2FA%3A1012299230482?LI=true#page-1) and what we see in dogs is unlikely to be purging in this sense, anyway.

    There were, indeed, bottlenecks as a result of WW2, but it is not true that the selection of the remaining dogs was on the basis of fitness. Sure, these would be dogs that had withstood the early infections that today vaccination guards against, but in many instances it was a case of use whatever there was in order to increase numbers - often very related individuals - and, as ever, many dogs were bred from young before their health and fitness was proven.

    "Now, imagine such a purged dog breed with fairly low numbers whose gene pool is threatened by a popular sire that is known to carry a recessive disease. Reducing short-term inbreeding depression in such a breed will result in this sire being behind every dog within a couple generations, which is obviously not a desirable outcome."

    Well, either the population is purged or it's not and how would using this dog (one of the inbred small population) used so profligately reduce inbreeding depression?

    "On the other hand, keeping several distinct more inbred lines will preserve more genetic variability and result in a continued base for selection outside of the popular sire. "

    Population modelling does not show that this is the case. Inbreeding results in the loss of genetic variation within a species/breed. That's why population geneticists advise that the best way to maintain as much genetic variation as possible is to breed from as large a number of animals every generation as possible - and from the least-related individuals.

    "Interestingly, there is virtually no scientific work that looks at inbreeding depression in individual breeds. This should warn us about applying general “recipes” to each and every dog breed, when in fact the genetics of these breeds may vary considerably and would have to be studied in detail before a course of action that would in fact stand a chance of improving the situation could be decided on."

    Yes, you're right - and there needs to be more studies.. But there is a large body of literature documenting inbreeding depression in other species and there is absolutely no reason to assume that dogs are immune to it. Genetics is genetics and the principles apply across the board. So in the meantime, that "general recipe" is the safest route.

    Finally, you cannot compare dogs to Chillingham Cattle. As you say, natural selection applies ("These unique beasts fight for the chance to mate. Any weak herd members are gored to death") and they are recognised as an anomaly: "The animals are regarded as a scientific marvel; inbreeding throughout history is well known to lead to extinction because of the small gene pool that the animals share. DNA samples from the cattle show that the cattle are all genetically identical; however the Chillingham Wild Cattle have managed to survive in spite of this, and the herd continues to grow."

    There will always be exceptions to the rule. To assume every breed of dog will be one of them would be completely foolhardy.

    ReplyDelete
    Replies
    1. (Same anon you were replying to, part 1)

      The evidence for purging is actually more significant than what one insect paper says. I recommend e.g. "The Natural History of Inbreeding and Outbreeding" by Nancy Thornhill, University of Chicago Press.

      Why should I not be able to compare dogs to Chillingham Cattle? The work on inbreeding and inbreeding depression that you apply to dogs has been mostly based on plants and insects. Surely, another mammal could be expected to be a more suitable model than these?

      My point though is that the earlier bottlenecks that many dog breeds went through do in fact mimic the situation in these cattle. Considering the standard of dog care during WW2, the situation was rather more grim than it is today. First, food rationing meant that dogs, especially large breeds that did not have a war use, were fed a rather inadequate diet. Second, veterinary care standards were nowhere near where they are today. When you consider that there were not even antibiotics available, you might understand that the risk of dying from what today would be a banal infection was quite real. Add to this the fact that breeders of large, non-useful dogs were forced to resort to substantial culling of their breeding stock because they were simply unable to find food for them, and you might start realising that strong selection for fitness in combination with strong inbreeding was very much a reality during these days.

      Delete
    2. (Part 2)

      As for my population model, I seem to have fallen victim to a typo – apparently, I am so used to typing “inbreeding depression” that my cerebellum must have played a trick on me here. What I meant to type was:

      "Now, imagine such a purged dog breed with fairly low numbers whose gene pool is threatened by a popular sire that is known to carry a recessive disease. Reducing short-term inbreeding COEFFICIENTS [not inbreeding depression, sorry] in such a breed will result in this sire being behind every dog within a couple generations, which is obviously not a desirable outcome."

      Whether the detrimental allele survived the purging scenario or whether the mutation occurred after purging is fairly immaterial. Obviously, there are thousands of alleles, some of which will survive even a purging event after which inbreeding depression becomes unmeasurable – that is irrelevant, as long as these alleles do not become sufficiently widespread to have an overall effect. This popular sire scenario, however, will cause them to do exactly that – an unintended consequence of reducing inbreeding coefficients.

      Why would this occur? Again, assume we have a population of inbred, but genetically distinct lines. Then the popular sire is used by some people, his get started winning shows, and the dog and his offspring are used more and more. Because he at first is an outcross for most breeders, lines on which he was used will initially have a lower inbreeding coefficient than those where he was not. If we declare reducing inbreeding coefficients the primary goal and breed the least related individuals to each other, we will be then spreading this dog and the associated detrimental allele through all the population within a short time. After a couple more generations, the lack of alternatives will then result in the population essentially becoming one single generic inbred line, though this time this line will be susceptible to inbreeding depression, as this practice allowed a detrimental allele to become widespread in the population again. In that situation, it is clearly better to keep separate purged lines with a higher inbreeding coefficient than reduce this coefficient in the short term while rendering the whole breed susceptible to the detrimental effects of inbreeding depression once again.

      I am not saying most breeds will be immune to inbreeding depression. Large ones with low numbers that were of no use in a war effort seem to be candidates for such a scenario, and in some of these, your well-meant one-size-fits-all approach to inbreeding reduction may very well prove to be a larger challenge to their survival than current breeding practices are.

      Delete
    3. You are making assumptions here that I don't think would apply in real life. If a breed recognises that bringing in a less related dog is desirable, why would they then turn the dog into a popular sire? That makes no sense.

      Now that's not to say that it hasn't happened... I suspect it has. But with today's knowledge, it shouldn't. Trying to avoid popular sires is an integral part of diversity breeding.

      Delete
  11. "There will always be exceptions to the rule. To assume every breed of dog will be one of them would be completely foolhardy."

    Equally it would be foolish and irrational to assume that ANY breed of dog wouldn't ...

    ReplyDelete
  12. The general argument is scientifically sound in relation to selecting against physical defects EXCEPT that it ignores the problem that homozygosity in the Master Histocompatibility Complex (MHC) is a major component of inbreeding depression. This is the difference between old style genetics and post HIV research genetics. HIV pushed us to learn a lot about the genetics of the immune system and what we learned is that the more different genes in the MHC, the better your ability to fight infectious disease, recognize cancers early and destroy them, and avoid having severe allergies and autoimmune diseases.

    Certainly selecting against bad genes that produce physical deformities is important but it does not help to avoid auto immune problems, disease susceptibility, and severe allergies.

    I suspect the reason the many large breed dogs came out of the WWI and Great Depression + WWII bottlenecks showing little inbreeding depression in the first several generations after the bottlenecks was that many breeders delayed breeding during the tough times and the dogs that were the subsequent parents were middle age or older by the time they were bred so overall health had been selected for. I would not be surprised if these older dogs had very low homozygosity in the MHC. Also the initial recovery litters may well have been from relatively unrelated dogs.

    It takes a number of generations for inbreeding depression to start to appear. It is generally recognized in relation to shortened lifespans, lowered fertility and lowered disease resistance when looked at from a population based level.

    I have been told that in England after WWII some breeds (Affenpinschers and Scottish Deerhounds) were actually out crossed to related breeds because the population had become so small. This can be checked in the studboooks.

    Another matter that cannot be ignored in relation to this claim is correctness of published stud book records. It is certainly very possible that as fertility drops, interbreed crosses occur (by accident) and go undetected or are done deliberately. Some interbreed crosses will be practically invisible in the f1 generation. For example a Lab crossed with an Irish Setter - the F1's will be pretty lab like, black, short coat. They might be considered a bit fine boned - but there is variation in a breed.

    The F2's bred back to the setter will produce a few long coated red pups. I the breeder culls the short coats and blacks then no one would challenge the litter as Setter although the dogs might lack elegance.

    I can think of quite a few crosses of related breeds that could be made to breed pure for one breed's breed specific traits in two generations.

    The other component is of course severe selection. If you are dealing with a bottleneck then only breed from middle aged or older dogs that are completely healthy. This may well mean ignoring certain fine points of beauty conformation for example.

    There is actually quite a bit of research on the problem of inbreeding in small populations of mammals. The people concerned with protecting endangered species are very concerned with this problem.

    here is one book that looks at it:

    The Natural History of Inbreeding and Outbreeding: Theoretical and Empirical Perspectives [Paperback]

    Also you might want to look at the essays collected at:

    http://dogdimension.org/DiversitySite/index.shtml

    Bonnie Dalzell, MA

    Freelance anatomist, vertebrate paleontologist, writer, illustrator, dog
    breeder, computer nerd & iconoclast... Borzoi info at www.borzois.com.

    ReplyDelete
    Replies
    1. (same anon Jemima was replying to)

      The alleles determining MHC are of course an explanation why inbreeding depression does occur. The considerations on Deerhounds, Setters etc. are equally interesting to read. However, all of this somewhat misses the point.

      Inbreeding depression is a reduction of fitness that occurs in relation to an increase in inbreeding coefficients. The commonly accepted effects of inbreeding depression are e.g. a reduced lifespan, a reduced resistance to disease, and reduced fertility.

      What we are sorely lacking are studies that look at a breed or even at dogs in general (!) and test whether such a correlation between COI and parameters of inbreeding depression exists in them. Virtually all of it is conjecture based on other models. The theoretical mechanisms that cause it are interesting to consider, but only if inbreeding depression is actually found in practice.

      Delete
    2. There are a few studies of dogs (eg John Armstrong showed that Poodles with very high COIs live less long than Poodles with the lowest COIs). More are needed. And I would agree that every breed needs proper analysis. But to not act - at least to do the research as a matter of some urgency - on the basis that the studies are lacking is stupid given the sheer weight of evidence that inbreeding depression is an issue in inbred populations (and *of course* in mammals - wild and domesticated - as well as plants and insects...)

      It's enough for the Animal Health Trust, btw, who issue advice regarding inbreeding depression (and of course the whole existence of Mate Select is due to the acknowledgement that high rates of inbreeding are a problem). You might also be interested in this, from Tom Lewis at the AHT - his presentation on inbreeding and outcrossing (you'll have to copy and paste all of the following... still haven't figured how to do a hyperlink...)

      http://www.google.co.uk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=4&ved=0CEwQFjAD&url=http%3A%2F%2Fdanecouncil.danemoor.com%2Fimages%2FTalk_5_-_Inbreeding_and_outcrossing_-_Dr_Tom_Lewis.pdf&ei=rmO6UNv2BsyY1AWVgYGACw&usg=AFQjCNGkAyKANaqQw-ENn19vlDC9-Saxxg&sig2=U4r-vzOOFAaKt515IngpSg

      Delete
    3. I am aware of the poodle study, though that one was not actually published in the peer-reviewed literature. One peer-reviewed study that demonstrated ID in a breed that I recall was a German study on Dachshunds that showed a negative correlation between litter size, the percentage of stillborn pups and COI (http://europepmc.org/abstract/MED/15803761/reload=0;jsessionid=4JOXMKYoOE9d2QY78RkG.0 ), as one would expect in a breed subject to ID. Also, Kosovska et al. showed that non-inbred Dachshunds live longer than inbred ones (http://www.ejpau.media.pl/volume8/issue4/art-71.html). No surprise so far: Dachshunds never went through an important bottleneck in their modern history as far as I am aware.

      On the other hand, Rehfeld found no correlation between inbreeding and fertility in Beagles (Am. J. Vet.Res. 31:723-732). Then there is another one that found no correlation between COI and lifespan in Wolfhounds (http://scholar.google.com/scholar?cluster=2187023651852669798&hl=en&as_sdt=0,48 ) that was later confirmed and also showed no correlation between inbreeding and litter size(http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3060614/ ). As I wrote above, this is a breed that was subject to severe bottleneck events and strong selection for fitness around WW1 and WW2.

      Basically, the whole correlation between inbreeding and fitness in dogs is a whole lot less clear-cut than what you or the Animal Health Trust seem to believe. You both mean well, of course, but neither of you are acting based on a level of evidence that resembles anything that would be acceptable to, say, the Cochrane collaboration.

      I suggest focusing on cases where exaggerated anatomy is clearly detrimental to a dog. Nobody in their right mind has the slightest scientific base to argue that exaggerated skin folds, ectropion, brachycephalic syndrome, exaggerated chondrodysplasia and all the other conformation-related and/or inherited diseases that you have pointed out are not actively harming the dogs involved. As opposed to inbreeding, these are clear-cut cases where broad action is entirely justified and no unintended detrimental consequences are obvious.

      I applaud your stance against exaggerated anatomy and the detrimental consequences it causes for the affected dogs, which are clearly a concern from an animal welfare point of view. What sets it apart from your stance on inbreeding depression is that you are perfectly aware that breed-specific paths of action are required for appropriate actions against these problems (surely, you would agree that selection against brachcyephalic syndrome would not be particularly appropriate for German Shepherds). Given what we know, why should the same approach to inbreeding e.g. be appropriate for both Dachshunds and Beagles?

      Delete
    4. A great many breeds went through genetic bottlenecks because of the Second World War and so far there is evidence that one - the wolfhound - may have been lucky.

      The Poodle too has undergone a severe bottleneck - Wycliffe rather than the war.

      Rehfeld may not have found a correlation between inbreeding and fertility in his inbred beagle colony, but he did find other effects - including an increase in neonatal death.

      Inbreeding effects have been found *in every species that has been studied". It is illogical to assume that dogs would be immune.

      So the default - in my view - should be that inbreeding fx are likely.

      But I completely agree (and have never argued otherwise) that this needs to be studied breed by breed. Inevitably, it will vary.

      Delete
    5. Nobody is arguing that inbreeding depression does not occur in all species with sexual reproduction that have been studied so far. However, that is besides the point: We are not talking about a species here. We are talking about hundreds of quite different, highly inbred, genetically isolated subpopulations of a subspecies of the wolf.

      Lack of inbreeding depression in highly inbred subpopulations of a species has been demonstrated to occur in quite a few such subpopulations of various species. I am not entirely certain that such subpopulations have been found in every species where this has been studied, but I certainly cannot think of any where it was not the case. Barring evidence to the contrary, it is therefore sensible to assume that susceptibility to ID will not be present in some breeds of dog, though their percentage cannot be reliably estimated based on our present knowledge.

      So yes, I entirely agree that this needs to be studied breed by breed, and results will likely vary. The problem is that COI's will, in principle, always be reducible in the breeds that turn out to be subject to ID, using outcrosses if necessary – but once advantageous purging effects are lost in a breed, they will be virtually impossible to re-attain:

      Why? Because purging requires both a bottleneck and strong selection for fitness. However, consciously provoking a bottleneck in a breed that is subject to ID is not exactly a sensible idea, and as for then telling breeders that they must no longer use dogs for breeding that would be dead without antibiotics or surgery... you can imagine the rest ;-).

      Delete
    6. OK, so I think we can agree that there are *some* highly-inbred subpopulations that are OK so - theoretically - it's possible for individual dog breeds to be OK.

      In the main, though, those apparently robust, highly-inbred subpopulations have been subjected to strong natural selection - something dog breeds do not enjoy - although there may have been *some* purging either by such a bottleneck as WW2 (and possibly in a limited way by some breeder selection).

      Of course most dog breeds *are* bottlenecked - and the high level of disease in many is evidence that bottlenecking has not had a purging effect.

      We can also agree that we need to look at each breed as the impact of inbreeding is likely to vary - from minimal to serious. Consequently, the response would need to be as variable.

      And I think we can also agree that consciously provoking a bottleneck and exerting selection that mimics natural selection is probably not that great an idea, for many reasons.

      So that leaves us with needing to do the research into individual breeds - longevity/litter sizes/neonatal mortality etc.

      Delete
  13. Stuart Thomson's website. A useful bibliography of more good reading

    http://www.gaiaresearch.co.za/bbgaiabulldogreport.html

    ReplyDelete
  14. What seems to be forgotten is that after the Bateson Report the Kennel Club rejected the idea of limiting the use of popular sires. Instead it was going to introduce Mate Select and Estimated Breed Values for all breeds.

    Cavaliers were going to be the first, the EBVs were initially going to select for Syringomyelia and Mitral Valve Disease. Now there are new DNA tests for two unique cavalier problems, Episodic Falling Syndrome and Dry Eye/Curly Coat, that could presumably be added.

    Four years on what progress has been made to provide these breeding tools for cavaliers and the other two hundred plus KC registered breeds?

    The gathering of reliable Cavalier health information proved difficult, needing two official health screening schemes to be set up.

    There is now a BVA/KC CMSM Scheme but it is being boycotted by many cavalier breeders and the KC seems to be content to let that happen. The scheme has not even been made a requirement for ABS breeders( nor have the two new DNA tests )

    There seems to be no realistic prospect of an official heart scheme for years to come.

    So is anyone asking whether the KC has any other plans to control loss of genetic diversity through the popular sire syndrome? Some questions need to be asked, because four years have gone by and their widely promoted method to tackle the problem, throughout all KC registered breeds, has just not materialised.

    ReplyDelete
  15. Wright's formula, the common measure of inbreeding, is based on known common ancestry. According to that, if a highly inbred dog of breed A mates with a highly inbred bitch of breed B, the offspring will have CoI zero; inbreeding disappears in one generation. With one bound we are free!

    Now I've heard of experiments with lab rats which ended with the offspring of many generations' inbreeding being completely infertile. Has anyone ever mated the depressed products of two separate inbreeding programmes, to see if inbreeding depression actually does vanish in a generation, as predicted by Wright's formula? If so, then outcrossing is the answer, but it can't really be that easy, can it?

    ReplyDelete
  16. Nice, we're still in the 1960's.

    ReplyDelete
  17. I was watching a science program on human sexuality, and it noted that the primary reason for kin repulsion (people mostly not being attracted to relatives, even if they are never told they are relatives) is genetic immunity; mating to closely related individuals leads to selecting for offspring whose immune systems are similar to both parents. It is genetically healthier to mate with individuals with different genes, thereby increasing the odds that the offspring may be immune to new pathogens. Since pathogens continue to evolve at a fast rate (short life-spans, rapid reproduction, genetic variability), a population (NOT an individual) stands the best chance of surviving if there are widely varied immune systems. Indeed, there are suspected points in human history where there were population crashes when a new disease came out and only certain groups had a certain immunity to it, and so survived. Sometimes these genes also carry deleterious effects, but you can't have everything. We are likely seeing something similar now with bats in the Eastern US and White Nose Syndrome. If all the bats were genetically similar, they ALL would have been wiped out, but pockets are surviving and they believe these may be individuals who happened, by sheer chance, to have some immunity.

    Reduce the gene pool and you reduce those chances.

    I have mixed views about health testing (as opposed to breeding for fitness and longevity), because health testing just leads to one more "popular sire" rising up if he's free from known genetic diseases and also typey.

    I know I've mentioned DM and Pembroke Welsh Corgis before. There is reason to be concerned that breeding rapidly away from a gene with high saturation but low rates of expression might lead to much more serious problems down the road in an otherwise fairly healthy and long-lived breed (early onset cancer, for instance). Thankfully most breeders seem to be aware of this and are not publishing "DM Status" but there is a small and vocal minority who makes claims like "We will never breed another DM positive dog again" against all recommendations from geneticists. Hopefully that movement won't lead to pressure on other breeders; so far it seems not to have (going by what I see from the outside looking in).

    I think that many breeders want to do the right thing, and if the message can get out there that occasional line-breeding to cement type is ok, but MOST of the time in MOST of your pedigrees you should see more out-crossing, then we might make headway. The problem is, to get true genetic diversity you sometimes have to accept less-than-perfect individuals into your gene pool. Are we willing to accept dogs with not quite the right size or ears or tailset? Are we willing to occasionally breed from working border collies who don't have eye, or working retrievers who have slightly lower tolerance for cold water, or the occasional dog with an auto-immune disease, in order to preserve true genetic diversity? I think to a certain degree, perfection in health or type or temper is by definition counter to achieving a wide gene pool.

    ReplyDelete
  18. "I have mixed views about health testing (as opposed to breeding for fitness and longevity), because health testing just leads to one more "popular sire" rising up if he's free from known genetic diseases and also typey."

    Health testing is very useful in that it provides knowledge. It shouldn't be considered as a way of removing animals from the gene pool but rather a way of making sure that they're only bred to compatible mates. Breeding from carriers of conditions is fine as long as they're not bred to another carrier. Even breeding from 'affected' animals might not be as cataclysmic as it sounds so long as its mates are all 'clear', so the offspring will be carriers but not affected.

    ReplyDelete
    Replies
    1. Well, that sounds great until you run into something like DM in Corgis, where only about 10% of all Pems are clear, and the rest either carriers or at-risk, yet only one or two percent of dogs come down with the disease.

      As testing gets more common, we must realize that many diseases are poly-genic, with other genes determining expression (age of onset, severity, environmental triggers, etc).

      So for something like vWD, which is a straight recessive with 100% expression on homozygous individuals, of course that's what you do.

      But I doubt cancer works that way. You may find that a certain set of genes trebles the risk of cancer. If cancer is common, that triple risk is too high. If the cancer is rare, say impacting one-tenth of one-percent of individuals, than a gene that causes the risk to increase to three-tenths of one percent of individuals might not be deleterious.

      If there is a gene for a disabling disease with age of onset at 5 we want to avoid it, but what if average age of onset is 14 and average lifespan is only 13?

      What if a gene that increases risk of hip dysplasia and also decreases risk of IVDD?

      What if we find a gene increases risk of cancer, but ONLY in dogs that are exposed to, say, ivermectin? Do we breed away from this or use it to know which dogs should avoid ivermectin?

      What if we have a high risk of cancer in a breed, and we find that most of the dogs who have lower risk according to a new genetic test ALSO carry a very mild form of vWD?

      And so on. Genetic testing can be straightforward, or oh-so-complicated.

      Delete
    2. I should add that the fact that the gene for DM is so ubiquitous in Corgis means that there is a strong chance that it confers some benefits elsewhere. Otherwise why would it have become so prevalent?

      Delete
    3. The benefit could be a a human one - a trait wanted by breeders - rather than one that is of any particular benefit to the dog.

      Delete
    4. Jemima,

      That may well be true. It may also be true that they carry a benefit--say, reduced incidence of IVDD: in Corgis, IVDD can strike young dogs, while in this breed DM tends to strike older dogs; average age of onset is 11 (while age of onset is much younger in German Shepherds). Or it may be that it was just present in the founding population, or it just happened to be present in a few popular sires, and so it got to be widespread. But when you are looking at 90% presence in the gene pool, chances are it either directly impacts something else, OR it rides along closely with another gene that has been selected for. We don't know enough to know for sure.

      Since over half of Corgis seem to be "at risk" and only a few percent (single digits) get the disease, there must be either modifier genes or environmental factors that play a role.

      We also know there are certain genes that turn other genes off and on, so in some cases where breeds have recessive genes that cause disease, it may be possible to add other genes into the mix, rather than just subtracting out. The gene for DM is an autosomal recessive with incomplete penetrance. That sort of information leaves breeders with hands more or less tied until more information about what determines expression comes to light. I doubt this is the only genetic disease with that sort of fingerprint.

      My fear is we know just enough about genetics to be dangerous, which is why in many cases I think it's still better to look at the big picture (actual health and longevity of the family tree, minimizing breedings of close relatives to maintain diversity) rather than just genes. In the example I give of DM, what if we breed aggressively away from the disease and end up with a host of other problems, so instead of having a small percent of dogs develop a disease with average age of onset at 11 years, we now have a breed with very little DM, but a high rate of cancer and average lifespan plummets to, say, 10? If that were to happen, you wouldn't have to worry much about DM even if it were around because most dogs would be dead before they were old enough to get it...

      My (very long) point is that maintaining genetic diversity and breeding away from every known disease to impact dogs are by definition mutually exclusive. We must be careful in what we breed away from, and how, because when it comes to recessive genes, once we get enough tests it will likely be impossible to find mates where no deleterious recessives at all match up, and not worth while to even try.

      We should be breeding for genetic diversity and the overall health of the breed, but healthy breeds will at times produce individual pups with genetic health problems. Indeed, humans have mated freely for millennia and there are huge numbers of people impacted by diseases with a strong genetic component. There are sub-populations that are more prone to certain diseases as well.

      Delete
    5. Beth, humans may have mated freely for millenia and still produced unhealthy progeny, but despite the current resistance, I think DNA tests will become widespread.

      I think once DNA tests become more widely available, it will be deemed unethical not to use them. I doubt any child would thank his/her parents for being born with a serious disease or disability, that could easily have been prevented through the use of a DNA test.

      Or maybe I just presume humans are more rational than they actually are?

      Delete
    6. Many mothers who know their foetus will have a disease choose to carry to term anyways. Many people who have such diseases that are detectable in utero wonder that someone who learned of their disease before birth would not have given them a chance to live simply for being imperfect. When it comes to humans, this is a sticky issue, even though many people agree it is easier for everyone - from parent, to child, to family and community - simply to not have a disabled child.

      Delete
  19. While I'm still thinking about it I wanted to give a straightforward example of why we need to breed dogs and not just genetic tests:

    Let's say we have the case of DM above, where it's a recessive with incomplete penetrance (meaning not all dogs who are homozygous for the recessive gene will go on to show symptoms).

    Now let's say I have a potential stud dog who is a nice dog from a very healthy family. He's got nice temperament, he's got good healthy conformation, he's typey enough, he's not heavily line-bred.

    But I test him for DM, and he's homozygous for the recessive gene that causes the disease. Despite the fact that I don't know of any DM dogs in his family tree, I decide not to breed him because I'm trying to be responsible and breed away from DM.

    But what if, paired with the coding for DM, is another gene that turns off uncontrolled cell reproduction? And this gene works if there is just a single copy (dominant gene?)

    Now I have just removed a cancer-beating dog from a gene pool.

    And therein lies the danger. With genes that are recessives with complete penetrance (every dog with a double recessive gets the condition) the decision is straightforward, if the resulting disease significantly impacts the well-being or lifespan of the resulting pups. But many, many genes have incomplete penetrance, and whole series of genes act as switches and modifiers, and we haven't done enough mapping to begin to know which are which.

    ReplyDelete
  20. I havent been able to get into the blog for several days, so havent commented. I see there are several references to the white Chillingham Cattle as an example of a breed who have miraculously survived in isolation for centuries in spite of inbreeding. In fact if one starts to dig deeper, things are not so good with the Chillingham cattle. There are only about sixty in the herd , with another smaller reserve herd of 12 or so in NE Scotland. In the very bad winter of 1947 only 17 of the cattle survived , and it has taken them 65 years to reach a total of about sixty, so still not back to the prewar herd size. Pretty low fertility. It would take a very long time to get the herd back to the numbers in the nineteenth century. Genetically the herd are almost identical and it is thought they would have very little resistance to disease so they have to be kept in isolation from other livestock for their survival. They have problems with their teeth due to anomalies of dentition. What one is seeing is a very small herd which survives due to a high degree of protection and isolation, being kept alive partly as a tourist attrction, partly out of scientific interest. If this herd were turned loose on the hills to fend for themselves, it isnt very likely they would survive for long! Not a good analogy for domestic dogs, unless one is willing to preserve breed by isolating them from any possible infection and accepting low fertility as the norm

    ReplyDelete
  21. Just out of interest, why did you only use the first part of the original title, and not the whole? Surely not for sensationalism? After all, context is very important, and it appears you deliberately tried to mislead the reader.

    ReplyDelete
    Replies
    1. Maybe she chose to assume that her readers don't all have the attention span of goldfish and can actually read past an article title. The title isn't even misleading.

      Delete
    2. The title of the original article makes the position clear; Jemima's proven track record on the subject (and indeed the article) shows she disagrees with the writer rather than learning about moderation.

      Delete