I've been looking up pedigrees of various dogs in my search for a Miniature Schnauzer breeder. I've noticed that some breeders will breed Grandfather to Granddaughter and Uncle to niece. The average breeding coefficient for schnauzers is 4.6% but these matings bump it up to 13 - 17%.

I thought the idea was to get as close as possible to the average, so why do some breeders do it, and are there additional health risks to puppies from these litters? Is this normal across breeds ? I can understand it happening in breeds where breed numbers are low but this isn't the case with minis.

> I thought the idea was to get as close as possible to the average

I would say this is a rule of thumb, which can be departed from when justified. Not only in low numbers breeds but any breeds, depending on what the breeder is seeking to achieve. Linebreeding is very complex requiring vast knowledge and experience of the breed and its genetics in order to be done correctly.

COI is one of the most important things (or inbreeding generally) to look at.  Every time we breed above average, we put another bullet in the head of our breeds. 

We can't do that and say we are breeding to 'better' the breed - we're not, we're breeding to kill it.  It's only a matter of time before stud books will HAVE to be opened but we can hold that day off by using what stock we have to preserve genetic diversity within breeds.

It's a fact that most (if not all) very successful show dogs will have higher than average COIs  More diversity, lower COIs also creates more diversity in the way the dogs look. Successful breeders have walked the tightrope of line breeding/inbreeding for generations, having to go out occasionally but then more often than not finding this detracts from what they were trying to achieve. or at the very least creating more differences within a litter. So they go back in to fix the type they like again.

I'm not saying I agree with it, just that it's a fact. My 1st champion borzoi had a COI of 32%, he was born in 1988. Anyone would know what lines he was from by looking at him.

In my current breed silken windhounds breeders are much more aware of COI and the need for genetic diversity, whilst I love that they are such a healthy breed I would have to admit that there are huge differences from dog to dog, including huge differences within a litter. I personally have one silken who is 47cm and one who is 60.5cm, one with rose ears, two with prick ears and one semi prick.

So, the balancing act is lower COI versus keeping dogs that look the way you want them to look.

After a number of basically outcrossed litters (living overseas with imported hounds), I did two matings, one of uncle to niece, and the other aunt to nephew.  Just because I felt I needed to see where I was in my breeding programme.    The uncle to niece was good.  The aunt to nephew only produced one puppy, so hardly put me any further forward re knowing where I was with my programme.   I didn't repeat that mating because her first litter was only 2 so she clearly wasn't meant to be a brood bitch much as of the two litter sisters I'd kept from that mating, I always felt she'd be the more prolific bitch.   'Uncle' and 'aunt' were brother and sister and 'Niece' and 'nephew', brother and sister.

Those two matings were the closest I'd done, although I did line-breed as much as I could, once back in the UK, going back to what was behind my UK-bred foundation bitch, used when overseas.  Not easy to do given the passing years.    My location situation apart -  which most breeders don't have to cope with, breeders do line-breed to fix type, but there does come a point when an outcross mating is needed as eventually the line starts to loose vigour, size and so on.  Such is the art of breeding.   However, when a breeder hits a good combination via line-breeding, it's very hard not to carry on perhaps beyond the point they should outcross.    But it becomes inevitable!   The trouble with outcrossing, even if you go for type, is you are likely to get the best of the best, or the opposite. It's a bit of a lottery, especially with two lines that have never been bred together!   Many years down the road now, I'm always interested to see what comes out of a breeding involving my earlier imports, used by other breeders.   And more than happy when the result is good.

In breeding & line breeding any animals takes a good degree of understanding & knowledge.

If you want to fix type when developing a new breed(which Miniatures are relatively to the original Schnauzers) you have to in breed at the start to get the size & conformation. Once you have this you can go to line breeding to the known good producing animals. This is how various species of domestic animals have been developed over the centuries.

COI isn't an exact science & is one of many tools now avaiable to breeders.

I thought the idea was to get as close as possible to the average,

Because that's what the media would like you to believe, and have hyped up. But if this was the way to a healthy dog, then all first crosses would be healthier than pedigrees, wouldn't they? And we know they're not. Also I don't believe for a moment that the published averages shows a true picture. In my main breed, we now sadly have something like 80 % irresponsible breeders vs 20 % responsible. The irresponsible breeders outcross. They do that because it's what they've heard they should do, and it's easier to sell puppies that way. As a result they also very often do no health testing at all, they think outcrossing means all pups will be healthy. So the average COI is low, but it's based mainly on matings carried out by less than responsible breeders, who very often don't know what they are doing at all.

Linebreeding carries risks, but so do outcrossing. With linebreeding you will quickly bring to the front what potential problems there may be in the line. You will also fix looks and temperament/character. With outcrossing, you could unwittingly bring in a new problem which lies hidden for years and then suddenly appears.

My mentor in my main breed, who was breeding for many decades, always said the rule of thumb was "Twice in, once out". That's what the big, successful breeders used to do in years gone by - before people started breeding by numbers.

Don't get me wrong, like I said both ways carry risks. And I have done both (everything from 0 % COI up to 16 %, and my best dog ever, bought in, was 20 %) and seen the results of both. Good and bad. The most important thing is knowing what's in your pedigrees, and by that I mean everything, not just names - health, longevity, looks, temperament etc. If you do, then you should technically end up with healthy pups of good temperament. But with the total outcross, you will not get a litter looking all the same. Some will take after one side, some the other, and some both.

Thank you everyone, didn't realize it would be so interesting.

From my point of view, I'm never going to be breeding any dogs I own. When choosing a pup am I correct in thinking one close relative crossing in the pedigree shouldn't ring alarm bells and is ok, but if the same dog / bitch shows up several times in the lines then that could be dodgy ?

>The aunt to nephew only produced one puppy, so hardly put me any further forward re knowing where I was with my programme.   I didn't repeat that mating because her first litter was only 2 so she clearly wasn't meant to be a brood bitch much as of the two litter sisters I'd kept from that mating,

Inbreeding results in reduced fertility and smaller litter sizes ON AVERAGE.  (Putting that bit in capitals because there will now be lots of people giving evidence of their own inbreedings which saw average sized litters!  This is why anecdotal evidence gets us no where.)

>When choosing a pup am I correct in thinking one close relative crossing in the pedigree shouldn't ring alarm bells and is ok, but if the same dog / bitch shows up several times in the lines then that could be dodgy ?

No, it's more about the COI.  However, even the COI doesn't give a full picture - really we want to be moving towards DNA testing dogs so we can then compare genetic material.  Because not all puppies in a litter will inherit the same genes from their parents - otherwise they would look and be identical.  Even within one litter, one pup might inherit a lot more genetic material from one ancestor whilst another pup inherits a lot from another ancestor - so although the whole litter will have the same COI, they will not have the same genetic material.  You can only test what genetic material they have through DNA testing. 

Which is all to say - COI is a crude tool which can only give a vague sense of relatedness and inbreeding - but for the vast majority of breeds and dogs, it's all we have at the moment in the UK.  Slowly, as more dogs are DNA tested and added to databases, we will be able to accurately determine exactly how much inbreeding there is in each individual dog.

As for fixing traits:  As well as fixing desirable traits, if you inbreed and reduce the gene pool, you are also fixing undesirable traits and increasing the number of recessive genes which come together to create affected puppies - in diseases which we don't even know about to start with.  We only see them, when we start to see affected puppies appearing.  And then we can't do anything about it, because we've lost the other genes in the breed - the genes that weren't affected.  This situation has occurred in multiple breeds at the moment and is threatening their very existence.

Inbreeding results in a higher risk of cancer, autoimmune disease and the inheritance of recessive genes for such diseases we don't know about.  Inbreeding results in decreased litter size and affects fertility.  To inbreed is to kill a breed and to breed towards its death. 

Those who like a particular appearance or performance need to be breeding to genetically unrelated dogs which have appearance or performance they like... These exist in my breed.  They may not exist in every breed.

And science (rather than anecdotal evidence) tells us that the effects of inbreeding begin at 5%. :  https://www.instituteofcaninebiology.org/blog/coi-faqs-understanding-the-coefficient-of-inbreeding

You have to look at the pedigree as a whole. I know with working Border Collies, breeders look to line breed to known good producing dogs/bitches usually on 4th /5th generations via good producing offspring.

In breeding is breeding close relatives ie Grand parent to Grandchild, uncle/aunt to niece/ nephew etc

They are both important & if done with knowledge neither is bad, but you do have to be careful & not just presume that good dogs will be the result.

My very best ever dog was from a pedigree with 1 dog 4 times in the 4th & 5th generations of his pedigree(non were parent/child or sibling to sibling breeding)

>> The aunt to nephew only produced one puppy, so hardly put me any further forward re knowing where I was with my programme.   I didn't repeat that mating because her first litter was only 2 so she clearly wasn't meant to be a brood bitch

Just to say her first litter, of only 2, was very much an outcross mating and I chose the dog for his phenotype, not genotype.   And I knew the dogs behind that dog.   This, going for type rather than genetics, was how I had to proceed with my imported (exported from here) foundation stock when living abroad.

Just to add, unless changed, the KC didn't accept the result of very close breeding for registration, and even with half brother to half sister, which is done quite often, with the two unrelated parents not closely bred, they recommend breeders take the advice of knowledgeable fellow breeders before going ahead.

> However, even the COI doesn't give a full picture - really we want to be moving towards DNA testing dogs so we can then compare genetic material.

My breed has reasontly been done threw UCDavis diversity testing. Better bred database that uses that data gives potential maitings grades based on how genetically related that are, 1 being the equivalent of identical twins to 10 being genetically unrelated. It's been interesting looking at these compared to pedigree COI. There are dogs who share some of the same dogs in the first 5 gens yet come out as grade 10 maitings. Also vice versa, low COI but genetically more alike.
A good tool for breeds looking to maintain as much of the diversity their breed currently has, also would be good for outcrossing projects to ensure the more diverse dogs are put back in.

> Better bred database

It's fascinating isn't it, we're doing the embark testing which gives a similar result as well as some 170 health tests and various other things, like identifying genes for low shedding, size modifiers etc and research towards finding the genes for things like kinks in the tail or mon/cryptorchid. It's interesting to see the dogs that come out as high or low diversity.

It is hard to know which of these databases to go with, how do we decide?  Is it simply whichever has more of our breed?

> It is hard to know which of these databases to go with, how do we decide?  Is it simply whichever has more of our breed?

Guess it depends on owner, what each offer and the differences. Mine are done with mydogdna and ucdavis, from what I've read the ucdavis and mydogdna do the diversity differently and have been told uc is more accurate picture. But I'll still continue to use both.

> It's fascinating isn't it, we're doing the embark testing which gives a similar result as well as some 170 health tests and various other things, like identifying genes for low shedding, size modifiers etc and research towards finding the genes for things like kinks in the tail or mon/cryptorchid. It's interesting to see the dogs that come out as high or low diversity.

It is isn't it, I had fun with doing all sorts of test breeding. I've had mine also done with mydogdna, same sort of idea as embark but done by a different lab.

If you go onto Genescoper's MydogDNA tools and you will see each dog tested goes on a graph showing it's diversity.

Here is one of mine: https://www.mydogdna.com/crm/index.html#en/animals/870023219845467/pass/diversity

Hers shows as 35.1% (our breed median is 30.9%, All purebred dogs 33.9%, and Mixed breed median 43.2%). Her COI is 0.4% based on 17 generations with 4 complete.

I have had no need to test her daughter as she is hereditary clear for the tests needed and 199 Eueos is a bit steep just to know her genetic diversity, but her sire is an import and an outcross as was her mother's sire.

She has a COI of 0.4% based on 18 generations and 4 complete. Her sire has a COI of 0.1% based on 10 generatiosn with 3 complete.

FB replies:

Gemma Brown says:  It's can be called line breeding and it can be bad if you wont know what your doing it can also be called select breeding and it can improve lines and help to breed out some bad things but should only be don't if the breeder is very experienced and know how to do it x x

Sue Parr says:  IMO thats modern 'PC' ideas....one of the best litters I had was uncle to niece. Some top breeders I know used to say 2 in 1 out and they produced some fab dogs

Julie Wright says:  I'd imagine after such a mating the next generation would be crossed to a bloodline with no shared ancestor, therefore reducing the average breeding coeff. When they were creating all these breeds a couple of centuries or so ago, this is how they'd do it. It's called line-breeding as opposed to in-breeding.

Ren Melo says:  Too bad for the breed, for any breed actually. Genetics can't be messed up

Thanks Barbara, I think I'm going to go with MyDogDNA too, to start with.  Maybe do Embark next...

Did you need to get a vet to do the cheek swab or sign any paperwork?  Only it says that a vet is supposed to do it and verify the microchip number so I'm just wondering if we have to get a vet involved or if we can do it ourselves...?

> Did you need to get a vet to do the cheek swab or sign any paperwork?

No at the moment the KC doesn't require it, but in some countries they do, hence the bit on the form for verification.

Great, thanks

I got told many years ago now, that to maintain type, which includes temperament, conformation and ability, you line breed, and to maintain diversity, you out cross. It’s not rocket science to say if you continually line breed, or inbreed, then you lose diversity at the expense of type, and vice a versa.

Not really, it is possible to breed choosing phenotypically desirable and similar mates which are also genetically different in many breeds.

We can see from the staggeringly high COIs in the majority of dog breeds what's happening over time, and it's a bit pointless to have evolved the most amazing 'type' (based on whatever is most in trend in the ring) whilst the dog is susceptible to all kinds of recessive diseases and weakened immune system due to being excessively inbred.  It's great to suggest that balance is the goal - but balance isn't what's happening. 

To avoid health effects, 5% - just 5% - is the COI above which research suggests we should not go.  Many breeds of dog are currently around 25%....