Feline Blood Groups & their Implications for Breeders

by Judith Picknell

The problem of neonatal mortality.

For very many years cat breeders have suffered problems with unexplained neonatal mortality of kittens, which seems to affect some breeds more than others. At one time it was considered that these problems, dubbed ‘fading kitten syndrome’ were probably caused mostly by latent FeLV or FIV infection, and many of us spent sleepless nights as we awaited the results of yet another re-test from the virology lab at Glasgow (no in-vitro tests done in minutes at the surgery in those days!). Eventually, to our relief, most tests showed no evidence of infection, but failed to explain why a healthy queen, with plenty of milk, had lost most or all of an apparently healthy litter of kittens within a few days off birth.

Sometimes mating to a different stud had a more successful outcome, and certain matings were simply dismissed as ‘incompatible’ without any real understanding of why they had failed to produce viable kittens. Often, however, good quality queens were neutered, written off as ‘poor breeders’ or ‘poor mothers’, because of their consistent inability to rear their kittens.

However work done by Professor Urs Giger and his team at the University of Pennsylvania in the early 1990’s, revealed another reason for neonatal death: incompatibility of blood group between the queen and her kittens.

Basics of Feline Blood Groups.

In humans there are several different series of factors involved in blood type, including the ‘rhesus factor’. The most commonly known and understood series are the primary Blood Group factors: A, B, and O. (To simplify a complex matter, type ‘A’ individuals produce antibodies which react if they come into contact with antigens in ‘alien’ type ‘B’ blood, and type ‘B’ individuals produce similar antibodies against type ‘A’ blood. Individuals with the relatively uncommon type ‘AB’ can cope with blood of either type, as their body ‘recognises’ both factors. Type ‘O’ individuals react against all other types of blood, but their blood can safely be given to any other group, as it doesn’t contain any antagonistic factors).

Neither type ‘A’ nor ‘B’ is dominant over the other (an individual who inherits ‘A’ from one parent and ‘B’ from the other will be type ‘AB’), but both are dominant to ‘O’.

However, blood groups in cats work rather differently; no ‘O’ group exists. The two main types are ‘A’ and ‘B’, with ‘A’ being a simple dominant over ‘B’; so a cat may be type ‘B’ (homozygous recessive), type ’A’ but carrying ‘B’ (heterozygous), or ‘pure A’ (homozygous dominant).

There is a third type, ‘AB’, which is very uncommon, and until recently, poorly understood. It is not the same as AB in humans (if a cat receives ‘A’ from one parent and ‘B’ from the other, it will not be type ‘AB’, but type ‘A’ carrying ‘B’). Recent research suggests that type ‘AB’ operates as part of the same series as types ‘A’ and ‘B’, although it appears to be inherited separately; it is recessive to ‘A’ but dominant to ‘B’. Because it is so uncommon, for practical purposes type ‘AB’ will be ignored for most of this article, as it is not something, which most breeders are likely to encounter.

In cats, type ‘A’ individuals (the dominant type) produce no antibodies, or only very weak ones, to type ‘B’ blood, and type AB cats produce no antibodies against either of the other blood types. However type ‘B’ individuals produce powerful antibodies to type ‘A’ blood, and this is where, in a breeding situation, problems may arise.

Problems with incompatibility.

There are two main areas in which blood group incompatibility may have serious consequences. The first lies in the situation where a cat receives a transfusion of blood of a different type, giving rise to a serious incompatibility reaction, usually with fatal consequences. The second, which is more likely to be a problem and concern for most breeders, is in the breeding situation, where incompatibility of blood type exists between a queen and her kittens.

The main problem in this situation is the neonatal mortality of type ‘A’ kittens born to a type ‘B’ queen. As mentioned above, type ‘A’ cats produce only very low levels of antibodies to type ‘B’ blood, but type ‘B’ cats produce powerful anti-‘A’ antibodies; these antibodies are present in high concentrations in the colostrum of a type ‘B’ queen. When newborn kittens suckle, and ingest the colostrum, the antibodies pass across the lining of the gut into the kittens’ bloodstreams. If the kittens have type ‘A’ blood, the antibodies react to the surface proteins of their red blood cells, and destroy them; this process is called isoerythrolysis. This can cause acute anaemia, and usually produces noticeable signs of jaundice as the kittens’ immature livers struggle to clear them of the dead blood cells. The destruction of the oxygen-carrying red cells and the resulting anaemia may cause necrotic damage to the kittens’ vital internal organs, and/or necrosis of their extremities, such as the tips of ears or tails. Typically, affected kittens will pass characteristically dark brown or red-coloured urine, due to the excretion of the dead blood cells. In cases where isoerethrolysis is a risk or is suspected, the kitten may be stimulated to urinate onto a pad of white cotton wool or tissue to check for the characteristic discolouration.

For reasons that are not fully understood, the severity of the disease is variable, with some kittens being more badly affected than others; many are severely affected and die, but a small number which should theoretically be at risk appear to be unaffected. Symptoms may include jaundice and death within the first two days of life; the kitten may survive but the tail tip may become necrotic and fall off at 10 to 14 days of life; or in rare cases, no signs of disease may be evident at all. It is also possible that where damage has occurred to a kitten’s internal organs it may fail to thrive, and die at several weeks of age.

Fortunately, the susceptibility of kittens to the effects of the maternal antibodies only lasts for approximately the first 16-24 hours of life. After this initial period the kitten’s gut lining becomes impermeable to them, so they are unable to pass across it into the blood stream and cause damage. For this reason it is safe for type ‘A’ kittens to be returned to feed from a type ‘B’ mother after this initial critical period has passed. (Some older papers on this subject state that it is necessary to avoid type ‘A’ kittens feeding from a type ‘B’ queen for as much as the first 48 hours, but some more recent research suggests that the critical period may be as little as 16 hours. Caution suggests allowing at least 24 hours, on the basis of ‘better safe than sorry’).

There is anecdotal evidence (largely involving British Shorthairs, which have a high proportion of blood type ‘B’) that some type ‘B’ queens may fail to conceive when mated to a type ‘A’ stud, or may resorb their litters early in the pregnancy. It also seems that some queens fail to carry their litters to full term and spontaneously abort at around 6-8 weeks gestation. At present there has been little scientific research to explain the mechanisms involved in this pre-natal loss, but it appears possible that in some cases maternal antibodies may be crossing the placenta and causing the intrauterine death of kittens of incompatible blood type. However this is clearly not the general rule, and the exact reasons for these occurrences remain uncertain at this point.

Practical implications.

In practice, when dealing with the potential of blood group incompatibility in a mating, knowledge is the key to avoiding problems!

The most important thing is to know the blood group of the queen to be mated; if she is blood type ‘A’ there should not be a risk of neonatal mortality through isoerethrolysis, irrespective of the blood group of the stud, as she will not produce any antibodies against the blood in type ’A’ kittens, and either extremely low levels or none at all against that of type ‘B’ kittens. However, when the queen is type ‘B’ the potential for problems exists.

When planning to breed from a type ‘B’ queen, the blood group of the stud is of utmost significance. Mated to a group ‘B’ stud, only type ‘B’ kittens will be produced, therefore there is no risk of incompatibility. However with a type ‘A’ stud, the potential of the queen giving birth to type ‘A’ kittens exists, and all of these will be at risk of developing isoerethrolysis if allowed to suckle from their dam in the first 24 hours of life. (See table 1).

The most straightforward option may seem to be only ever to use type ’B’ studs when breeding from a type ‘B’ queen. This may be feasible in breeds such as British Shorthair, where there is a fairly large genepool and a substantial majority of the population (59% in UK) is type ‘B’, but even there breeders may be denying themselves the positive benefits of using a type ‘A’ stud which is most suitable in other respects, either because of superior conformation or because of his pedigree.

In breeds such as the Devon Rex, where cats with type ‘B’ blood comprise a substantial minority of the breed and the genepool is already limited, further problems exist, which will be discussed later.

Some stud owners may, for whatever reason, decide not to blood-type their boy, possibly because they have never had neonatal mortality problems themselves, or are working entirely with type ‘A’ queens, and therefore do not perceive blood-group incompatibility as a problem.

Suffice to say, for now, that there will be occasions where it will be desirable to mate a ‘B’ queen to an ‘A’ stud, or to a stud whose blood type is unknown. In these circumstances it is still possible to avoid isoerythrolysis and subsequent kitten mortality, providing that care is taken.

The most important thing to remember is that newborn type ‘A’ kittens will only be at risk if they ingest ‘anti-A’ antibodies from their mother’s colostrum during the first 16-24 hours after birth. After that time the antibodies cannot cross the gut wall into the blood stream to destroy the kittens’ red blood cells. The really difficult thing is to make sure that you are there (and awake!) when the queen gives birth, so that you can remove the kittens from her before they have the opportunity to suckle. Assuming that you manage this without mishap there are several possible courses of action to tide you all over the critical 24-hour period.

First, and possibly most simple, is to hand feed all the babies with Cimicat or a similar feline milk replacement for the initial period, returning them to mum once they have ceased to be at risk. The disadvantage of this option is that a mother who has all her kittens taken away may become very distressed, but it is obviously preferable to losing some or all of the kittens. However some breeders successfully manage this situation by fitting the queen with a long section of stocking or lightweight tubular elastic bandage, to cover the whole length of the body from shoulders to haunches, ensuring that all the nipples are well covered to prevent the kittens from suckling (see picture, right). This allows the kittens to be returned to the queen between feeds to be kept warm, washed and mothered, allowing bonding and benefiting both queen and kittens.

Sexyrexy Babbetjie, owned by Laraine Lister of Rexitude Devon Rex, wearing a home-made 'leotard', made from part of a pair of thick tights, to prevent her newborn kittens from suckling.

Secondly it is now possible to order and obtain blood-typing kits through veterinary surgeries, with which it is possible to type the individual kittens at birth using blood from the umbilical cord. However breeders who have used these kits report that they can be fiddly to use, and that it is not always easy to obtain enough blood from the cord to perform an accurate test. The advantage of this approach is that once the kittens have been tested, any type ’B’ kittens, which are not at risk, can be returned to the queen for her to feed and care for in the normal way, which should prevent her from fretting, and the type ‘A’ kittens can be handfed and returned to her when it is safe to do so. There is always the chance that ALL the kittens will be type ‘A’ and that they will all therefore need to be handfed in any case; but if all test as type ‘B’ then the problem is over!

Possibly the optimum solution, if you are in a position to arrange it, is to plan for a type ‘A’ queen to kitten a week or so before the type ‘B’ queen is due to give birth, and then to cross-foster the kittens for at least the first 24 hours. The older kittens will be unaffected by the type ‘B’ queen’s anti-A antibodies, and the newborns will be safe to suckle from a type ‘A’ queen.

Obviously it is of paramount importance to know the blood type of queens in those breeds where both blood groups are known to occur with any frequency. Even in those breeds where type ‘B’ is known to occur only very rarely it may be useful to type queens when either their pedigree or a previous history of neonatal or pre-natal loss suggests that there might be a problem. It should also be considered before outcrossing a queen from a breed where type ‘B’ is known to occur in the population (eg Cornish Rex), to a stud from a breed known to be predominantly or exclusively type ‘A’ (eg Siamese/Oriental or Russian Blue). However it is not critical to know the blood group of the queen if the stud to which she is mated has already been confirmed as type ‘B’, as no incompatibility problems arise where the male is of this type.

Ideally all cats used for breeding, which belong to those breeds where both blood groups are known to occur, should be blood-typed, but this will not always be practicable, or seen as necessary by all breeders.

The blood typing of newborns, in situations where a type ‘B’ queen is likely to produce a mix of type ‘A’ and type ‘B’ kittens, can be a useful tool in avoiding neonatal mortality due to isoerethrolysis; however it will not necessary or practical to do this for kittens from all matings, the majority of which are not at risk.

In the rare instances of cats with blood type ‘AB’, they should for practical purposes be treated as though they were type ‘A’. A type ‘B’ queen mated to a type ‘AB’ stud should be treated as if she had been mated to a type ‘A’ stud, and any type ‘AB’ kittens from a type ‘B’ queen should similarly be treated as if they were type ‘A’ kittens, and should not be allowed to feed from their mother for the first 24 hours, to avoid the risk of isoerethrolysis. However a type ‘AB’ queen in theory should be safe to mate to studs of any blood type, as she will not produce antibodies against the blood of kittens of her own or either of the other types.

Distribution of blood groups.

The blood group that constitutes the greatest proportion of cats over all breeds and in non-pedigrees is the dominant type ‘A’.

The proportion of cats with type ‘B’ blood varies significantly between breeds, with the greatest numbers occurring among British Shorthairs, Persian-type longhairs, Exotics, and Devon and Cornish Rex.

The overall proportion of cats displaying type ‘AB’ is very tiny, and the Bengal breed seems to be disproportionately heavily represented within this group. This may be reflect the use of hybrids, produced by crosses with the Asian Leopard Cat, in the foundation stock of this breed, as the genetic makeup in terms of blood groups is significantly different in some wild species to that of domestic cats.

The most extensive study to date of feline blood-type distribution across different breeds has been carried out by Professor Giger and his team at the University of Pennsylvania, which shows 33% of Cornish and 45% of Devons are type ‘B’.

Limited testing done in the UK, in a study based at the Royal Edinburgh Veterinary School, shows a somewhat different picture to the overall results in the international study; however as no Cornish and only 2 Devons were included in this study it does not provide useful statistics on blood types in these breeds in the UK. Results of tests recorded by the joint Rex BAC suggest that distribution of blood types in Devons broadly mirrors the international picture. On the basis of anecdotal evidence, the vast majority of Cornish in the UK now seem to be blood type ‘A’, although the BAC now accepts that the presence of type ‘B’ within the UK breeding pool, and suggests that all breeding cats should be bloodtyped.

Overall, differences between the American and UK studies emerged, especially in respect of the unexpectedly high proportion of type ‘AB’ cats in the UK study. This particular difference may reflect the proportion of Bengals typed in this study, but may also be influenced by the fact that a new testing method was also under trial.

The explanation for the differences between the UK figures and the wider international picture may in fact lie, at least in part, in the relatively isolated position of the UK in respect of the import of cats from abroad. Its geographic status as an island means that the native non-pedigree cat population may have developed largely separate from that even in neighbouring parts of mainland Europe; a slightly higher proportion of group ‘B’ blood in the original population may have been exaggerated and consolidated by ongoing interbreeding within this (relatively) restricted gene pool. During the 20th century the imposition of quarantine regulations, introduced in order to protect the UK’s rabies-free status, will undoubtedly have increased the isolation in the indigenous non-pedigree population caused by basic geographical factors.

This has been of even greater importance in respect of the pedigree cat fancy. It has always been possible to export cats, and indeed both Cornish and Devon Rex breeds have spread around the world from their origins in South West England. However, importing cats has been severely restricted by both the bureaucratic complexity and the sheer financial expense of putting an animal through six months quarantine in addition to purchase and shipping costs.

Until very recently, therefore, the feline gene pool in the UK has had fairly low levels of input from abroad, compared with those countries in which the cross-border movement of animals has been unrestricted, either for sale or for stud service. However, the introduction of the Pet Travel Scheme (‘Pet Passports’) may in time have a significant impact upon the gene pool of pedigree cats in the UK. Not only does it make the import of breeding animals either from or via other EU countries much easier and less expensive, but also opens up the possibility in the future of taking queens to stud in mainland Western Europe.

Differences in breeding and showing policies under the GCCF as compared with FIFe, CFA and other international cat fancy organisations may also account for some differences in the gene pools of specific breeds in the UK, compared to their equivalents elsewhere in the world. These policies, especially in respect of permitting outcrosses, have in many cases been influenced by the difficulty of importing cats and by the consequent restriction of the numbers of breeding stock available within certain individual breeds.

It is not unreasonable to suppose that overall differences in the gene pool are reflected by the differences in the prevalence of blood groups. When introducing overseas bloodlines to their breeding programs, UK breeders need to be aware where differences in blood group distribution profiles exist, and where appropriate to type and to take appropriate action to avoid running into incompatibility problems, either in the F1 or subsequent generations.

It is interesting that Australia, which is also a rabies-free country and operates a quarantine system for domestic cats, also shows significant differences from the norm in this respect, with a much higher percentage of type ‘B’ cats than that in the overall international figures. It may be significant that Australia has no indigenous felines, and that the majority of the cats that formed the foundation of the species in this continent probably originated from the UK.

Noticeable variations in blood-type distribution of non-pedigree domestic cats also occurs regionally within the USA, with blood type ‘B’ occurring at less than 1% in the North East and Midwestern states, and at up to 6% on the West coast. No theories are available at present to account for these differences.

Seeking to eliminate an ‘unwanted’ gene.

In theory, eliminating a dominant gene is straightforward, and simply requires that all cats of the dominant phenotype (in this case blood type ‘A’) should be neutered. This is only likely to be considered in British Shorthairs, as this is the only breed in which group ‘B’ is the majority blood group.

In practice, this is less easy to carry out. If a particularly good specimen of a breed is type ‘A’, the owner is likely to be reluctant to neuter the cat and to lose the positive benefits that it could give to their bloodlines, and, indeed, to the breed as a whole. People with good type ‘A’ queens will not suffer problems with neonatal erethrolysis in kittens, and are unlikely to see any benefit in having them spayed.

A possible working compromise under these circumstances would be to neuter all type ‘A’ males, using only type ‘B’ studs within the breed, so that problems with neonatal mortality are avoided irrespective of the blood group of the queen to which they are mated. This in turn is likely in the longer term to further reduce the incidence of blood group ‘A’ within the breed.

To eliminate a recessive gene from breeding stock is a far more difficult proposition than eliminating a dominant. Even if the recessive phenotype (blood type ‘B’) occurs at fairly low frequency and all cats of that blood type are neutered, there is likely to remain a surprisingly high number of cats that are type ‘A’ but carrying the recessive gene for type ‘B’. For example, in a breed with only 10% type ‘B’ individuals, over 43% could be heterozygous ‘A’ (carrying ‘B’). (See Table 2).

The GCCF Cat Welfare Trust is currently supporting a project proposed by Dr Matthew Binns of the Animal Health Trust, to produce a feline genome map. When this project is complete it will be possible to genetically screen cats to check their carrier status for a whole range of recessive genetic diseases, as has already been done for the recessive hereditary disease gangliosidosis in the Korat breed. It has recently become possible to test whether cats with type 'A' blood carry the recessive genes for type 'B' or 'AB' blood types, although at present this test is only being carried out by a limited number of laboratories.

Where a test to detect carrier status is not available, the only available method of identification is test mating. This involves mating the cat to be tested either to cats of the recessive genotype, or to proven carriers of the gene. If a type ‘A’ cat produces a total of 11 kittens, none of which are type ‘B’, when mated to type ‘B’ cats, or 19 kittens (again all type ‘A’) when mated to a proven type ‘B’ carrier, there is only a 0.1% probability (ie one chance in a thousand) that the tested cat is a carrier of the recessive type ‘B’. However a single type ‘B’ kitten is sufficient to prove that the cat is a carrier!

Clearly a large number of kittens would need to be produced, all of which would be either possible or known carriers of the ‘unwanted’ recessive gene, and all of which would have to be blood-typed. Because of the numbers of kittens involved, test-mating queens is certainly unlikely to be feasible, and even test-mating studs could prove extremely expensive. Also it should be remembered that this only gives a probability that the tested cat is not a carrier – there are always the ones that defy the statistics, and even with testing to this high level of probability there is always a tiny chance of the recessive gene ‘slipping through the net’. It has been said that recessives, like diamonds, are forever! Realistically, it is doubtful whether it is worthwhile to go to these lengths in respect of a non-lethal gene.

Table 2
Probable distribution of Blood Types from known percentages of Type ‘B’

Some breeders who work with breeds where both ‘A’ and ‘B’ blood groups occur may feel that their best course of action would be to eliminate the minority group from their own breeding stock, and it may indeed be feasible for individual breeders to choose to test their cats, and to breed only from cats of one blood group. However one possible scenario is that of some breeders choosing to work only with type ‘A’, and others only with type ‘B’. In the long term this could lead to a situation where a single breed, in terms of registration and standard of points, effectively divides into two parallel breeds with separate, reduced genepools, split along the lines of blood group.

If co-ordinated efforts were made to eliminate the minority blood group throughout a breed, the genetic implications could be serious, especially in a breed that already has a limited genepool and relatively small numbers of breeding cats, as this would inevitably cause significant diminution of the existing genepool and exacerbate the problems of inbreeding which already exist.

The Historical Background in Cornish & Devon Rexes.

The presence of type ‘B’ blood in both Cornish and Devon breeds should not be surprising, given the origins of both breeds in the non-pedigree cats of the South West of England; it is quite probable that the gene has been there since the beginnings of both breeds. The indigenous cat population of the British Isles contained a higher than average proportion of type ‘B’ individuals, and is the same basic genetic base that gave rise to the British Shorthair, with its very high incidence of blood type ‘B’. In the early days of both breeds, the most common breeds used to expand the genepools were non-pedigree domestic shorthairs and British Shorthairs. It is probable that many of the early exports from the UK to the USA were either group ‘B’ or carriers of the gene.

In North America, the CFA closed both breeds to outcrossing several years ago, which will to a large extent have had the effect of ‘freezing’ the genepools. In the UK this has not been the case, and the breeds used as outcrosses at different times may have caused variations in the blood group distribution over a period of time.

In the early years of the Devon Rex, before it was established as a different genetic mutation from the Cornish, several crosses were made between the two breeds, and the straight-coated hybrids were used in the progression of both. Consequently it is probable that all Devon lines share a common genetic inheritance with the Cornish Rex, and indeed Devons carrying the Cornish gene, from crosses made almost forty years ago, are still occasionally exposed by accidental matings with Cornish! Similarly, very few UK Cornish lines are without any trace of Devon in the pedigree, although Du-Bu, Amaska, Senty-Twix and Zureiqa Cornish Rex were all scrupulous not to include any cats descended from the Devon hybrid lines in their breeding programs.

With Devons, probably the most common outcross has been to the British Shorthair. Early matings were also made to Persian-type longhair breeds in an effort to improve coat density, although the current breeding policy now prohibits such outcrosses due to the undesirability of perpetuating the longhair gene in the breed (the incidence of type ‘B’ blood in these breeds currently stands at around 15%). Both British and Persians may have either increased or consolidated the high incidence of group ‘B’ blood within the Devon genepool.

More recently a number of Burmese have been used as outcrosses; with their almost exclusively type ‘A’ blood, these cats may have caused a movement in the balance of blood type distribution within the Devon breed, but this is almost certainly counterbalanced by the continued popularity of British Shorthairs as an outcross.

With Cornish Rex the picture is rather more complex. British Shorthairs were heavily used in the early development of the breed within the UK; however this caused a noticeable change in type away from the original moderate foreign type of Kallibunker. Several moves were made to counteract this.

Alison Ashford imported Riovista Kismet, a blue Cornish male, from Canada, and he is behind the pedigrees of many of today’s cats, especially those descended from Alison’s ‘Annelida’ lines. It is quite possible that Kismet was himself either type ‘B’ or a carrier of the gene, as blood typing of Cornish in recent years has shown that an unusually high proportion of cats in Canada, and those bred from Canadian lines, are of type ‘B’.

Other moves to refine the type and to eliminate the cobbiness brought in by the British Shorthairs involved a number of outcrosses to Siamese. This is likely to have increased the proportion of type ‘A’ within the breed, as this breed is exclusively blood type ‘A’. The downside of these outcrosses was a significant decline in coat density, and in an effort to counteract this without again losing type, Hetty Hamilton of Zureiqa cats outcrossed to Arctic Piatrovitch, a Russian Blue, another exclusively type ‘A’ breed.

During the past 20 years or so British Shorthairs have not been a popular outcross for Cornish, although they remain an approved outcross breed in the breeding policy; the vast majority of outcrosses have been to Siamese, Siamese-type Orientals, and occasionally to Russians. This has almost certainly had a significant effect in terms of changing the overall balance of blood type prevalence within the breed within the UK, and is likely to be a major reason for the variance between the distribution in the UK compared with the broader international picture.

Over the years, a lot of type ‘B’ queens have probably been spayed, dismissed as ‘poor breeders’ due to heavy neonatal or pre-natal loss of kittens. It might be thought that this would, in time, also have the effect of reducing the proportion of type ‘B’ cats within a breed, as their genes would not be passed on. However it should be remembered that a type ‘B’ stud could be perceived as a potent and successful stud, as no problems associated with incompatibility reactions would occur in queens mated to a stud of this type, irrespective of their own blood group; therefore a type ‘B’ stud could strongly perpetuate this blood type within a breed. Simple reproductive biology dictates that a single stud may have an impact on the genepool of his breed in excess of that of very many queens, and the overall effect within a closed genepool may well tend towards equilibrium.

The presence of different blood groups within a breed can present serious problems to cat breeders, including the heartbreak of neonatal loss if incompatibility reactions occur. However, with knowledge and care it is a situation which can be worked with in order to minimise or avoid problems.

This is arguably less damaging to individual breeds in the long term than taking steps which would diminish an already limited gene pool.

Acknowledgements:

I would like to thank the following people for their assistance in the preparation of this article:

Professor Urs Giger, University of Pennsylvania, for supplying information and references.

Professor Michael Day, Langford, Bristol University Veterinary School, for supplying information and copies of articles.

Dr. Sandy Robertson, University College London, for calculating the statistical information in Table 2.

Caroline Prout, Fergan British Shorthairs, for information on blood group incompatibility in British Shorthairs.