Genetic diversity – what about it? by Mike Tempest

Created: 23/05/2012

Two columns ago (DW, March 23) I said we would have to come back to this topic because of the suggestion that certain breeds should be crossbred to other breeds, and now we have further topicality to the subject with another suggestion that bitches should be mated to two stud dogs at the same season. It is claimed that all of this is in order to increase genetic diversity, so I thought we had better have a look at these issues before they become carved in stone as the way to go!

Diverse species

The dog species is probably the most genetically diverse of all the animal species. It varies enormously in size and shape, from the small Chihuahua weighing 1.5kg (height approximately 15cm) to the Irish Wolfhound weighing 60kg (almost 90cm tall). Such diversity is man-made. However the current issue is not about the species as a whole but is concerned with how genetically diverse individual breeds are, and should be. This is a fundamental question for us all to think about.
There have been many public expressions of various viewpoints, largely unqualified, that such and such a breed has "a rather limited gene pool” without knowing what this means. Often the reason given for this view is that it must be so because most of our present day dogs descend from a very small number that founded a breed. Such a rationale is without scientific foundation.
I have previously described the segregation of the pairs of chromosome into single chromosomes when sperm and eggs are being formed, and the recombination of single chromosomes back into pairs when fertilisation of an egg by a sperm takes place. This segregation and recombination takes place totally at random, and with 39 pairs of chromosomes in the dog the number of possible segregations and re-combinations is vast (39 multiplied by itself 39 times is the number of possible gene combinations in either the sperm or the eggs). When different breeders all around the world have approached the choice of sire and dam mating combinations, and then selection of pups, in their own individual and unique ways, it is not difficult to come to the conclusion that there is an enormous amount of genetic diversity in any breed (the exception will be very rare breeds) – and to be precise what we mean is a heck of lot of different combinations of alleles, which are the alternative forms of any one gene. Every breed originally descends from a limited number of dogs, but that does not mean that they are not genetically diverse now.
So even if only viewed from a parochial UK position it is generally not correct to say there is a low level of genetic diversity in a breed, and it is even less correct (if that is possible) if the world population of a breed is taken into account. The exception to this is for breeds that are highly inbred, mainly but not always a consequence of being very rare breeds that have extremely low numbers, which inevitably means that it will not be possible to avoid mating dogs and bitches with common ancestors and therefore there will be high levels of inbreeding. Generally speaking a high level of inbreeding will mean low genetic diversity and the gene pool will be relatively small; a low level of inbreeding will mean much genetic diversity and reflects a relatively large gene pool. Thus an examination of inbreeding levels in a breed will indicate the amount of genetic diversity in a breed and therefore the relative size of the gene pool.
A large gene pool is not simply a reflection of how many breeding animals there are in a breed. Individual dogs often occur several times in the background (pedigree) of any dog and this has to be taken into account to give us a measure of the number of dogs that are contributing genetically. This measure is known as the ‘effective breeding population’. Scientists at the English Kennel Club’s Genetics Centre at the Animal Health Trust have developed a formula to work this out based on the average inbreeding coefficient of a breed and the rate at which inbreeding is changing over time and per generation, so the ‘effective breeding population’ can be worked out for every breed (see DW March 2, p2 New research on genetic diversity could ‘save breeds at risk from extinction’). The scientists have stated that the size of ‘effective breeding population’ that they think has enough genetic diversity to sustain a breed is 100. This means that those breeds that have an effective breeding population below 100 need more genetic diversity to sustain them. What we always need to be aware of is that this analysis is based on UK registrations only, and the effective breeding populations calculated are the effective breeding populations in the UK only.
Almost immediately after the AHT’s announcement of its new research findings on genetic diversity Prof Sheila Crispin’s Dog Advisory Council announced its ‘view’ that a long term strategy for out-crossing and back-crossing is needed to resolve major problems in ‘problematic’ breeds (DW March 9, p2 Outcrossing strategy needed to resolve breed problems, says DAC). The AHT’s research had shown that five breeds have effective breeding populations of between 20 and 28 and the KC said it will be talking to breed clubs about introducing out-crossing strategies. It is wrong, in my view, to jump into this as the first and immediate response. It smacks of yet another knee-jerk reaction or fashionable PR stunt by the KC to say look how good we are. True crossbreeding might be needed if the global population of a breed is so low as to threaten its future, or to introduce a gene that would resolve health issues such as high uric acid in Dalmatians, but there are other more acceptable alternatives.

Breeding bottleneck

I often use examples from my own breed Tibetan Terriers, and I will now do so again. The breed went through a breeding bottleneck which resulted from using two popular stud dogs, one of which was the grandson of the other. ‘Effective breeding population’ had not been invented at the time, but we are now told by the AHT scientists that their calculations show that in the early 1980s, the effective breeding population of TTs went as low as 13, much lower than the five breeds mentioned above. Breeders at the time didn’t need ‘effective breeding population’ to inform them that the gene pool was restricted – the senior breeders knew it themselves; nor did they need the KC to talk to them about crossbreeding - they knew that new bloodlines were needed. So several responsible breeders imported dogs from abroad and are still doing it – from Germany, Denmark, Finland, Sweden, America, France, Holland, Belgium, the Czech Republic – mostly new breeding but on some occasions a combination with an export that had been sent out. So successful was this that the effective breeding population of TTs is now 126 and expanding. 
The above example shows that the first step for increasing genetic diversity in a breed should not be to crossbreed, but to import new bloodlines, and it is much easier to do this now than it was 40 years ago. The example also demonstrates the mark of responsible breeding, but like all good things it never made any headlines, only the bad things do that, but I am happy to use it to confirm my belief that the best of dog breeders are supremely responsible in their breeding activities.
And by this we mean mating a bitch to two dogs at the same season. Why on earth would breeders want to do this? We are told it is to reduce the influence of popular sires, and thereby increase genetic diversity. My understanding is that popular sires become popular sires because they produce quality offspring, so what is being suggested is that breeders use a second and therefore inferior stud dog which will produce inferior specimens of a breed just for the sake of increasing genetic diversity. This is crazy thinking! Experienced responsible breeders generally have a pretty good hunch about which stud dog will suit their bitch, and evidence of the quality produced over time would back this up.
But let us examine a little closer the impact on genetic diversity. Let us suppose that Stud Dog A and Stud Dog B each separately mate 10 bitches, thus 20 bitches are mated. If the average litter size is eight, 160 pups will have been produced in total, 80 by Stud Dog A out of bitches 1-10 and 80 by Stud Dog B out bitches 11-20. The alternative proposal is that the 20 bitches are mated by both Stud Dogs. On average it would be expected from mixed matings done at the right time that half the offspring will be by each dog. Thus Stud Dog A produces 80 pups out of bitches 1-20 and Stud Dog B produces 80 pups out of bitches 1-20. So what has been gained by using mixed matings with two stud dogs compared to separate matings? Stud Dog A has gained 40 offspring out of bitches 11-20, but ‘lost’ 40 offspring out of his own original bitches 1-10. Stud Dog B has gained 40 offspring out of bitches 1-10, but ‘lost’ 40 offspring out of his original bitches 11-20. All the pups are still by the same two stud dogs. The increase in genetic diversity is miniscule, and it is not going to "increase the gene pool twice in one litter” as claimed in various postings.
 But – and this is a big but – in order to increase genetic diversity the offspring have themselves to enter the ‘effective breeding population’ and be bred from. This will not happen unless breeders using mixed matings retain offspring from each sire, or find homes for the offspring with other breeders. The tendency as always will be for the breeder to keep the best pup and place the others in pet homes. Result – no effect on genetic diversity!
The assessment of whether we have genetic diversity in our breeds is based on the KC registration database. It is therefore an assessment of genetic diversity within the UK. In order to create more genetic diversity in the UK we have to end this parochial, inward looking ‘tiny little island’ approach and ‘British is best’ mentality, as if we are fenced off from the rest of the world. The first solution should be to import new bloodlines, either by bringing in dogs or semen. Forget about crossbreeding and mixed matings, except if the threat of breed extinction is a real threat because of low numbers and no alternative bloodlines.