Animal breeding, genetics abuse, and Powerpoint pedigrees

I’m about to tell a little story, here, and be aware that there are far worse and far more egregious cases of what I’m going to call “genetics abuse” than the situation I am about to describe.

After dog training the other day, as we were walking Amos (our purebred Rottweiler, for those who don’t know) back to our car, a woman stepped out of our car and went “Oh, he’s beautiful!” (thank you, we know our dog is gorgeous).

“What line is he from?”

This question, well-intentioned as it is, makes me twitch. While we did glance at Amos’s pedigree (in my case, and as I’ll explain below, it was to calculate his inbreeding coefficient) and look at his parents’ records for eye examinations and elbow and hip scores, we don’t care what line he is from. We’re not breeders. We won’t breed Amos for a number of reasons. We got Amos from Oscelly because his parents were healthy and good-natured, and the breeders were supportive.

“Oscelly,” Husband replied politely. “In Kyneton.”

“I used to breed Rottweilers, in Queensland,” the woman explained. “But I can’t have another one since my last one died.”

I flooded with sympathy. “Oh, that’s awful,” I said.

“Yes, and there was no sign! I line-bred him as close as I could-”

At this point the conflict between “It’s so sad to lose a dog” and “you fucking idiot” caused a blood vessel to pop somewhere in the tact-zone of my brain. I actually inhaled my own spit when I went, “Ah… ha?”

“-and I’d been wrestling with him the night before, and he was fine, and I went out to the garage in the morning and he was dead. He was only seven.”

A healthy Rottweiler has a good chance of making it to ten years old and beyond. Quietly I observed, “Sounds like a heart problem?”

“Yes, I thought so.”

Then she admired Amos some more while he wriggled in sociable happiness, particularly his beautifully shaped head (he does have a nice head), and we parted ways. As in all truly frustrating situations, it actually took a few hours for me to get angry enough to start composing this post in my head.

Let Auntie Kate explain line-breeding to you, as well as some basic genetics, and let’s get cracking on what happened to this poor dog and his owner.

Line breeding is where a particular breeder/stud decides that they want their dogs (or horses, or cats, or alpacas) to have a particular look and tendency. They basically want a genetic stamp that says “These are our dogs. You have a dog from Stud X!” They also want all the traits that will allow them to win more dog shows, which will get them more stud fees. In order to achieve that, they try to concentrate the particular desired traits in their dogs.

They do this by inbreeding. Let’s be clear here: line breeding is just another way of saying “inbreeding.” There is no fucking around. It is the same thing, and it is very, very bad for any population to have this happen.

I know a good deal about population genetics (the thesis is coming along nicely) and next to nothing about animal husbandry. All that I know about animal husbandry is based on breeders of any kind of animal delivering bone-headed pronouncements that make me want to smack them upside the head with a population genetics text book (heavy enough to concuss). Having said that, animal husbandry is one of the long arms of human agriculture. Humanity has been breeding animals to conform to their specifications for millennia, and we’ve become pretty good at it. Without an understanding of DNA, or complex inheritance patterns, we managed to work out that breeding too close was a bad idea.

We worked this out from a number of signs: deformed offspring is one obvious sign. Less obvious signs are infertility or reduced fertility. If, for example, you want to buy a puppy from a breeder, and they announce that the bitch only had one pup in that litter, do not buy that pup. That pup has a good chance of breaking your heart. If you do cave, and buy the pup that almost certainly carries a number of recessive defects, do not breed the pup. Ever. If you do, you’re part of the problem.

So some animal husbandry has limits as to how close they are willing to breed their animals. They draw these limits far, far closer than they should, and then cite some pretty random research to say that this is okay. This research does not say what they think it says, and I’m going to explain why.

In every cell of your body*, you have two copies of your master genome** – the 23 (usually) chromosomes that are tightly wound strings of DNA. You have one copy from your biological mother, and one from your biological father. In the simplest scenario, when you produce gametes (eggs and sperm), each gamete contains only one copy. They might have your paternal copy, or your maternal copy (they might have a slight mix of the two as your chromosomes cross over and recombine, which is how new combinations arise). The successful gamete will pass that copy on to the next generation. Odds of having a mostly paternal versus a mostly maternal copy are 50:50.

So we have two copies of every gene and every gene region. When these copies are different in any way, we refer to the different versions as “alleles”. For example, I have red hair. Red hair is what we call a recessive trait, which means that I have to have inherited alleles relating to red hair from both my mother and my father (hair colour is actually a polygenic trait, so it’s not that simple, but both parents have to be involved for my hair colour to express the way it does). This also means that, since I have two copies of the relevant alleles, I am guaranteed to pass on one redhead allele copy to any kidlets I one day have.

So we’ve got the background for simple recessive inheritance. Now we get to the scary part.


No, really, you are. DNA is self-replicating, and it even has proofreading systems to make sure it copies everything perfectly. It still screws up, and leads to mutations. We all have these mutations, every single one of us. Most of them are probably relatively benign on their own – after all, you’re reading this, right? – but in many cases if you had two copies, the result would be lethal. In fact, most of the time, if you have two copies that were flawed in the same way, you wouldn’t develop all the way to viability. You would be miscarried, maybe as a zygote, or an embryo, or a fetus.

Here’s the thing: because these mutations are so sparsely distributed throughout the genome, and so individual, and because there is so much diversity in the human genome (we have what is called a large effective population size, more on that later), it’s astoundingly rare that you will get two copies of a flawed allele (with the exception of some named recessive disorders that have persisted in the population – and there are some nasty ones out there).

Of course, if you’re closely related to someone, it’s much more likely that you both have that allele. I am now going to illustrate why cousins shouldn’t marry and have kids****, only in my example, they are DOG cousins.

(****early footnote to prevent offense also pasted in here: Of course, cousins are perfectly allowed to marry. It’s their right and their choice. It’s just that most of them seek genetic counselling to make sure that they are not going to pass on some deleterious recessive trait. In most cases, it’s perfectly fine.)

We begin with a basic pedigree.


Meet Lord Doggington and Princess Jemima. They are both perfectly healthy, show line dogs of the breed I just made up called Zimbabwean Slothhounds. They win awards for how amazingly pretty they look. Depending on the breeder, they may also be valued for their robust strength and good nature, or the breeder might just be interested in appearance. Let’s give them the benefit of the doubt: they are, after all, particularly good at chasing down sloths.

Like all dogs – and all DNA-bearing organisms – Lord Doggington is carrying a nasty secret in his genome. It’s okay, because it arose with him due to a point mutation in one of his parents’ gametes, so it doesn’t exist anywhere else in the Zimbabwean Slothhound population. We should be safe from anyone ending up with two copies of this allele.

Alas, Lord Doggington’s owners think he is super awesome, and they want their whole stud to be like him. They want to concentrate his traits.

Lord Doggington and Lady Jemima have a bunch of puppies (ideally around eight). The stud sells six, and keeps two, Daisy and Fido, whom they like the look of. Daisy and Fido each have a 50% chance of inheriting Lord Doggington’s crap-arse allele. However, they will only ever have one copy, because Princess Jemima does not have the crap-arse allele. She probably has her own deleterious alleles, but we’ll focus on the other one here.


Let’s imagine that both Daisy and Fido inherit the crap-arse allele. There is a 25% chance that this will happen (0.5 x 0.5 = 0.25, for those who forgot probability as soon as they graduated high school). Daisy and Fido are each bred to outstanding Slothhounds from other studs, and produce, respectively, Butcho and Miss Dogface, who are adorable puppies.


There is a further 50% chance each that Miss Dogface and Butcho respectively will inherit the crap-arse allele from Fido and Daisy. If we know that Fido and Daisy have each inherited the allele, then, again, there is a 25% chance that both Miss Dogface and Butcho will as well. If we don’t know the status of Fido and Daisy – i.e., we’re just guessing, which is the fun part of recessive trait pedigrees – then there is a 25% chance that each pup will have the allele, and a 6.25% chance that both of them will. That’s a more than 1 in 20 chance; it’s actually quite high.

Let’s say that this untoward event happens, and the breeders decide to breed Dogface to Butcho (some breeders won’t do this. Some will. It varies) to create the pure strain of awesome that descended from Lord Doggington, thus elevating the Zimbabwean Slothhound from relative obscurity as a breed to glorious renown.

If we know that Dogface and Butcho have the crap-arse allele, there’s a 50% chance that their pup will have one copy, a 25% chance that they will have no copies, and a whopping remaining 25% chance the pup will have two copies. If it is still unknown – if, in fact, all we know is that Lord Doggington is on both sides of the pedigree and that Butcho and Dogface are first cousins – then we have a 1.56% chance that their offspring will have two copies of any allele from Lord Doggington. This is what we call the probability of identity by descent (IBD).*****


In this last figure, I’ve now changed the colour to indicate that this unlikely event has happened, and we can see the carriers of the allele. The offspring of Miss Dogface and Butcho inherits both copies (something that is a 1.56% chance from the word go), and either aborts, or is born with a severe defect that may even require euthanasia. This is why you want to avoid dogs with tiny litter sizes – unless they’ve been bred for reduced ovulation, small litter sizes generally represent resorbed pups that did not develop or even useless gametes.

Lest you think 1.56% is vanishingly small, in a population-level analysis, it is huge. It is a matter of some concern, and this shit happens all the time. In livestock animals, offspring may often be bred back to their parents – now that you’ve got the basics, I’ll leave you to draw out the pedigree of all the shit that comes out of that.

When Amos’s breeder showed me his pedigree, he brought out a photo of his very own Lord Doggington, explained how fantastic that dog was, and that he was on both sides of Amos’s pedigree. This was stated to me as a good thing, and if I were not a geneticist, I would have been convinced by his superior knowledge. That is why I am writing this post and using too much bold typeface.

I gently pointed out that I was a population geneticist doing a PhD and that this was of some concern to me. The breeder told me about some research some fellow in the UK had done on plants where he experimented and showed that inbreeding wasn’t so bad.

I haven’t managed to find whatever research he was quoting, but here’s the question I want to ask: what sort of plant? Was it a native population? Because this, I think, is where animal husbandry and population geneticists part ways. Animal breeders know what I’ve just been explaining. These are not stupid people – they have a science all of their own. They’re just not updating it, and they are missing a crucial piece of the puzzle: the piece that comes from population genetics.

They are missing effective population size.

Effective population size is basically a way of describing the background genetic diversity in the population and what is passed on to the next generation. For example, you might have an enormous population of corals on a reef, but if the millions are all reproducing by cloning themselves, and if they are all descended from the same clone, the corals will have a very, very small population size (i.e.: 1). However, since they are clonal, they already have their two genome copies, and we already know that they are functional. There are advantages to clonality.

You might have a very large herd of deer, but if all the young males keep getting their arses handed to them by the boss male, he’s going to do all the mating. Because you have a large number of different females, the effective population size will be larger than 1, but it’s not going to be enormous. There’s going to be a lot of deer wandering around with that paternal genome copy.

The basic message is this: when you breed to concentrate physical traits in a population, you are removing variation from the genome, because you are not breeding the animals that don’t fit your requirements. These animals do not contribute to the next generation; therefore they are not included in your effective population size. You are concentrating the physical traits you desire, but you are also concentrating the invisible, deleterious alleles. You are increasing the likelihood that matings will result in double copies of the same rare, lethal allele. You are reducing your effective population size.

You are drying up your gene pool #nerdpun.

So inbreeding is tolerable and manageable when you have a large effective population size, although it is itself reducing your population diversity. It is a horrible thing to do when you have a small effective population size.

This just in: purebred dogs have a very small effective population size. We’re not talking clonal corals, but we’re definitely not talking human-grade levels either. It’s a closed system; generally they only breed purebreds of a breed to others. That’s the whole point. I decided to get a purebred dog despite knowing all this, because I grew up with and love the breed, and also because this way there is a measure of predictability regarding temperament (I like large dogs. Temperament is important) and disease (I want to know what to look for).

I decided to look at Amos’s pedigree myself. All up, his probability of any allele being identical by descent was less than 0.06%. Since I’d been told that the odds of finding a breeder who does not practice line breeding are slim to none, I decided that was an acceptable risk. Amos has an undershot jaw, so it would be irresponsible to breed him (it’s not a huge deal for him personally), but otherwise he is perfectly healthy.

I think. I’m not sure, because here’s the rub: Rottweilers became very fashionable at some point in the 1980s. Before that, they were known for the elbow and hip problems, and possibly the minor eye issues – and that was it. These are also common to many breeds of dog.

When a dog breed becomes fashionable, you get a lot of what is termed “backyard breeding” – people either breeding out of ignorance (because they like puppies), or greed (because purebred dogs are worth squagloads) or both. This means people didn’t necessarily check that their dogs were healthy before they bred them.

Rottweilers are now known for heart problems. This is a new thing. I now have to worry about my dog’s heart, because even though he has a 0.06% chance of identity by descent, the odds of these crappy heart alleles floating around are much higher (there’s also cancer. Don’t get me started).

And this woman I met in the parking lot – to come back to the point after a long and circuitous journey – line-bred her dog as close as she could, and she did it because she honestly believed that it would produce a healthier dog. She believed this because all the dogs she bred were outwardly healthy, with no consideration for or knowledge of what was hanging around unexpressed in their DNA. This is the problem. Because they don’t consider effective population size, because they don’t consider recessive traits, and because they do not acknowledge how common these mutations are, they only breed from the appearance and behaviour of the animal, i.e., from the phenotype, not the genotype (this is not universal: a number of recessive traits are very well understood and bred for, but this doesn’t connect with effective population size).

It’s possible that if she outbred her dogs with other studs – and you could even do this and keep a dog purebred – the resulting pup wouldn’t have had a congenital heart defect.

It’s just possible she wouldn’t have gone out to the garage and found her best mate dead on the floor at seven years old.

So, even in purebreeding systems, you can outbreed. Do that. Avoid genetic abuse. Reduce animal suffering. Avoid getting your heart broken.

*Except red blood cells. They need extra room for hauling around oxygen, so they don’t have a nucleus.

** We’re not talking about the mitochondrial genome. While that is important in its own way, it is only maternally inherited, it doesn’t recombine, and because cellular respiration is so crucial and you only have one copy rather than two, deleterious mutations don’t tend to survive.***

***Having said that, there are rare cases of species or individuals where a mitochondrion may be bi-parentally inherited and/or undergo recombination, and there are rare human diseases that are caused by mitochondrial mutations.

****Of course, cousins are perfectly allowed to marry. It’s their right and their choice. It’s just that most of them seek genetic counselling to make sure that they are not going to pass on some deleterious recessive trait. In most cases, it’s perfectly fine.

*****IBD! Identity by descent! Not to be confused with Isolation By Distance, or any other number of concepts for which IBD is an acronym…

(Backdated entry: Originally put together November 2012)


4 thoughts on “Animal breeding, genetics abuse, and Powerpoint pedigrees

  1. THIS IS WHY MUTTS ARE SO DANG HEALTHY in comparison to purebreds! I’d rather have a mutt or a non-“popular” cross (like Labradoodles – as soon as a cross becomes A Thing they can basically become as sick as purebreds). I’m not even a geneticist and i’m liable to rant at anyone who thinks purebreds are hot stuff (i know enough about mendel squares to be dangerous…) Of course, i say all this while owned by a greyhound – but i went into this relationship knowing i’m probably going to have my heart broke. Bone cancer, amputations, all kinds of wacky and horrible diseases – greys are subject to them all. But the racers get the bum end of the deal, and if my girl had to put up with crap food and no personal attention for the first years of her life, she is by golly going to spend the rest of her days on her back on the couch with her feet in the air. Drooling, of course.

    • One of the problems as I see it is that, if it weren’t for the inbreeding issue, it would still be difficult to convince breeders that certain traits are not worth breeding to extremes. In a sense my rottie boy gets off lightly here, because they are a working breed, but even then the show lines get shafted a bit on conformation; and also, German Shepherds are a working breed, and they are bred for the most ludicrous conformation. This doesn’t even get into the little toy dogs with the eye problems, ear problems and breathing problems because someone decided a certain configuration of features was “cute” (and possibly lethal). While I will probably always own rotties (because <3), you won't be likely to hear me saying that purebreds are totally awesome, because of all these issues. All dogs have their own personalities, and how a dog is raised and socialised is an enormous factor when it comes to temperament, but there's a large genetic component – and as a large dog person that is one of the reasons I try to go for something standardised, while minimising the other problems.

      I recently learned that for the working breeds like rotties and shepherds, there are "working lines", where breeders breed them for robust health rather than looks. Apparently there are some adorably ugly purebreds out there, but they will be healthsome. I'm thinking for my next pup I will look out that way (although I am not ruling out a rescue rottie. Like other supposedly "tough macho" dogs they often end up abandoned because they respond very poorly to neglect)."

      Wow that turned into a screed… Anyways, in summary, I completely agree with you. Although with crossbreeds you can get some very unexpected results, they are only occasionally undesirable (i.e., normal sized dog with titchy legs, results in joint issues, etc. etc. but then might have had joint issues anyway so…).

      Also your grey girl is a very lucky lass. I was quite charmed by her 🙂

  2. Was it perchance inbred octoploid strawberries?

  3. […] take a puppy home to treasure and train. Many breeders are lovely, responsible people even if they don’t know enough about population genetics to prevent inbreeding. While some breeders may view their studs as assembly lines, many do genuinely love their charges […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s