NORWEGIAN ELKHOUND FORUM

Clarifications of misconceptions by Mike Tempest

Created: 20/08/2013
One of the most difficult aspects of being a scientist is to get over scientific knowledge in a simple, easily understood way. In my experience the canine community – and that includes breeders, commentators and kennel clubs – are one of the most difficult sectors with which to achieve this understanding – but we should not blame them, because failure to understand is a failure on the part of ‘educators’, and I include myself in that. If you cannot get a message across it is never the fault of those trying to receive the message, but is always the fault of those trying to communicate the message!
I did say right back in my very first article that I would not desist from prodding any of the constituent groups that make up the dog world, especially those whose heads are still buried in the sand, even if they are breeders. Although there are some breeders who try very hard to understand genetic principles, they are the exception. As a broad generalisation, breeders demonstrate a continued lack of progress in accepting genetics. Maybe this is because there is no certainty of outcome when breeding and therefore breeders are dismissive of genetics; but maybe also because scientists are hellbent on theoretical issues that do not relate to what is practical for breeders. I will defend science to the hilt – it is the very backbone of our existence, but to be useful it has to have practical relevance to our lives.
It is clear from the dog press and internet sites that there are still many myths and misconceptions held both by breeders and others. In order to ‘explode’ the myths, I will be quoting them, but to save embarrassment I will not attribute them to the individuals who quoted them.


Inbreeding

Inbreeding is probably the subject on which there is greatest misunderstanding, and is also the subject on which I receive the greatest number of enquiries in my mail. Here are a couple of quotes from breeders writing to the UK dog papers: "Inbreeding is now recognised as a cause of genetic problems” and "We all know that inbreeding causes health problems” – no we don’t all know this because it is incorrect. Inbreeding does not create problems; it only brings them to the surface if the population/breed has the problem on its genes in the first place. This is a subtle distinction but a very important one to understand.
Most defects are caused by mutations, which is a change in the genetic material in the cell, changing one allele of a gene into a different form. Mutations are unpredictable and occur very rarely, and can be induced by external factors such as radiation and chemicals. If a mutation has some biological advantage to the dog or does not cause its death it may be perpetuated, but it can skip many generations undetected (if it is a recessive mutation) until it surfaces through a chance combination with the same mutation carried by its mate.
This sudden appearance does not mean there has been a gene mutation there and then in a particular dog, the mutation probably happened several generations ago, and had no effect on the dogs that carried it, but it has only come to light now by chance. Inbreeding, which is the mating of a dog and a bitch which have a common ancestor, increases the chance of two mutated alleles coming together in the offspring from the common ancestor that carried them – ‘common ancestor’ meaning that it is on both the sire’s and dam’s side of its pedigree.


Coefficient

The chance of a dog acquiring the same form of a gene from a common ancestor can be worked out mathematically. This is known as the coefficient of inbreeding, otherwise the COI. Going right back to basics, you will recall that in an early article I explained that a male dog only passes half of his chromosomes into a sperm, and that a bitch only passes half of her chromosomes into an egg, so that when the sperm fertilises the egg the full complement of chromosomes are restored. The segregation of the chromosomes and recombination of them takes place at random, but in every generation coming forward from a common ancestor the halving process takes place. Therefore the further back in the pedigree the common ancestor is, the less is the chance that his/her particular chromosomes (and therefore genes) will end up in the present subject dog. The level of inbreeding, and therefore the coefficient, depends on which generation the common ancestor is in, and indeed on how many times the common ancestor is present.
It is therefore not too difficult to understand that you cannot have the same common ancestor in the first generation, because in that parental generation one ancestor is male and one is female – so it is impossible for them to be one and the same ancestor! If it was possible then there would be a one in two chance of getting the same version of a gene (ie a COI of 50 per cent). The closest we can get with a common ancestor is when a father mates his daughter, the father being the sire is in the first generation of the sire’s side of the pedigree, the daughter is in the first generation on the dam’s side with her father in the second generation. This is known as inbreeding 1:2 on the sire, gives a COI of 25 per cent and a one in four chance of the same version of a gene going into the offspring (the same is true of mother to son matings). If the common ancestor was in the second generation on each side of the pedigree, ie 2:2, then the COI would reduce to 12.5 per cent, and the chance of getting the same form of a gene from the common ancestor would be one in eight.
This is all very straightforward when there is only one common ancestor, but what is the situation when there is more than one common ancestor? Generally in the case of several common ancestors the COIs to each ancestor are added up to give a total, but this can give rise to a very misleading interpretation. We need to illustrate this.
Table 1 leaves out some dogs that are superfluous to the issue and simply illustrates that Bruce has mated his granddaughter Adelaide. Bruce is in the first generation on the sire’s side and the third generation on the dam’s side, so we say that the offspring of this combination are inbred 1:3 to Bruce. This gives a COI of 12.5 per cent.
Table 2 shows inbreeding to four common ancestors all in the third generation. The offspring of Mikey x Babs are thus inbred 3:3 to John (which is a COI of 3.125 per cent), 3:3 to Margaret (another COI of 3.125 per cent), 3:3 to Tommy (another COI of 3.125 per cent) and 3:3 to Annie (yet another COI of 3.125 per cent). The total COI is thus 12.5 per cent, but it is a very different 12.5 per cent to Example 1.
The Kennel Club’s Mate Select programme offers an explanation of what the COI value means. It says that a COI of 12.5 per cent means that there is a one in eight chance (derived from 12.5 per cent which is 12.5 chances in 100) that a dog will inherit the same version of a gene from the same dog that appears in both the sire’s and dam’s pedigree. This is only correct when there is only one common ancestor as in my Example 1, but this interpretation is scientifically and statistically incorrect for the 12.5 per cent in my Example 2. In this example there is a 1 in 32 chance (from 3.125 chances in 100) that a pup will inherit the same version of a gene from John, a 1 in 32 chance that a pup will inherit the same version of a gene from Margaret (not necessarily the same gene as with John), a 1 in 32 chance that a pup will inherit the same version of a gene from Tommy (not necessarily the same gene as with John or Margaret) and a 1 in 32 chance that a pup will inherit the same version of a gene from Annie (not necessarily the same gene as with John, Margaret or Tommy). So although the total COI is 12.5 per cent, the influence of each of the four common ancestors is limited, and if the four common ancestors were all different in type or construction, inbreeding would be moving in several different directions at once which would be a pointless exercise. Indeed, to call it ‘inbreeding’ would be a contradiction in terms.
We therefore need the KC’s Mate Select programme to show, when there is more than one common ancestor, the COI to each common ancestor. The Finnish Kennel Club’s programme does this and therefore recognises the correct science. Why does our own KC not do likewise? I can only think that this was to produce a quick response to pander to the pressure groups, but in doing so it has put in place a system that is scientifically incorrect and way behind what other KCs already have!
A further question is why does Mate Select calculate COIs ad infinitum beyond the horizon? Even just going to the sixth generation, there are 64 dogs in that generation, 32 on the sire’s side and 32 on the dam’s side, but the current pups only acquire 39 chromosomes from the sire’s side and 39 from the dam’s side. There will be some dogs in the sixth generation that do not contribute one chromosome (and therefore genes) to the present generation. Yes, some dogs may be duplicated, but COIs are being calculated to dogs that are probably not making any genetic contribution to the present. Going beyond five generations to calculate COIs is just a theoretical exercise that may be euphoric for a scientist, but it has no practical relevance to the present day breeder. Please Mate Select, just calculate COIs over five generations – there is no point in going to the million ancestors in the 20th generation, we all know that this represents more dogs than there has ever been in most breeds and some dogs will occur several times, but being so far back they are not adding anything significant to current inbreeding levels.
I have just bred a litter and the straight five-generation COI was 2.72 per cent. When I factored in the five-generation COIs of the common ancestors themselves it went up to 2.81 per cent. Round it up to the nearest per cent and let’s say it is 3 per cent. Mate Select went all the way back to 12 generations to get it up to 4 per cent. So what! And this litter had no other breeding in it than breeding that went back to the original foundation stock of the breed in the UK. The 2.81 per cent was the total of five common ancestors – the highest was 1.21 per cent, then 0.61 per cent, 0.41 per cent, 0.39 per cent and 0.19 per cent. This means that for the common ancestor with the highest COI there is a 1 in 83 chance that a pup will inherit the same version of a gene from it, and for the common ancestor with the lowest COI there is a 1 in 526 chance of a pup getting the same version of a gene from it. Come on – let’s get inbreeding into perspective!
While on the subject of inbreeding, it gives me the opportunity to correct more misconceptions expressed in the press by breeders. One correspondent wrote: "Wright’s COI does have some limitations in that it only takes into account ancestors that are common on both sides of a pedigree” – but that’s what inbreeding is, the amount of relationship between the sire and the dam; it has nothing to do with whether the sire and dam are themselves inbred. At least another correspondent got it right when saying, "If there are no common ancestors, even though the sire and dam might be very much inbred themselves, the resulting COI for their offspring will be 0 per cent”. Well done for understanding correctly. Yes – if the sire and dam are unrelated, being themselves inbred to totally different dogs, then mating them together is not inbreeding – it is outbreeding – the mating together of two totally unrelated lines.
I do acknowledge the tremendous progress made by the KC with Mate Select and MyKC, tools that breeders could only dream of a few years ago, but both need fine tuning and both need simple and scientifically correct explanations for breeders and the dog buying public.