Recently, dog shelters here in the Bay Area have been putting out the red alert that they are overwhelmed with good boys and girls that need homes. I do not want a second dog, but I do have a house with a big fenced-in yard and I wanted to help, so I signed up to be a foster. Within days I got the call to come pick up my new temporary dog: Bug.
Bug is a 9-month old mutt, or mixed breed dog. However, she has almond eyes, a velvety coat, strong neck and jowls, and a tapered tail, so the popular perception is that she is a “pit bull,” a type of dog that isn’t really a recognized breed so much as a collection of breeds that share those physical traits: American Pit Bull Terriers, American Staffordshire Terriers, American Bullies, Staffordshire Bull Terriers and others.
And because she has those physical traits, there are some cities and towns, and thousands of apartment buildings, where she isn’t even allowed to live. Those traits also mean that when I first posted a cute photo of her on Twitter, I was almost immediately hit with a completely unhinged reply from a stranger reading “New York has kept record of every reported dog bite since 2015. Pitvulls (sic) were massively over represented.” Wow, thanks Milo Thatch from Wish, what a helpful response.
I blocked him and forgot about it because I wasn’t interested in getting in a “nature vs. nurture” fight with an anti-pit bull activist because, obviously, they are weird, obsessive, and already have their minds made up based on an overly simplistic understanding of genetics. But I decided to talk about it here today because of a new study I’ve seen pop up in various places that specifically looked into the genetic basis for the behavior of various dog breeds.
I first saw it when someone in my Discord sent along this summary in Ars Technica, which is headlined “Genetics goes to the dogs, finds there’s not much to breed behavior.” Later I saw the New York Times also picked it up and titled their piece “They’re All Good Dogs, and It Has Nothing to Do With Their Breed.” So I read both of those and then I read the study: as usual, here are the correct facts (as I understand them) up front: those headlines (especially the New York Times) are incorrect, as many headlines are. The actual study found that SOME aspects of a dog’s behavior DO have some basis in their genetics that corresponds to their breed’s stereotypes, but the vast majority of their behavior is due to other factors, mostly how they are raised. The authors conclude that “breed offers only modest value for predicting the behavior of individual dogs.”
That’s just the tl;dr though, so let’s get into the actual study, which is really interesting and well done. Ancestry-inclusive dog genomics challenges popular breed stereotypes was published last month in the journal Science and is available in full to read online.
One of my favorite aspects of this study is that it relies upon a citizen science initiative: Darwin’s Ark is a huge, open-source project that enlists pet owners to aid in scientific research done on dogs, ticks, and soon, cats. It’s run in part by Elinor K. Karlsson, Director of Vertebrate Genomics at the Broad Institute of Harvard and MIT, along with a bunch of veterinarians, computational biologists, and scientists like Kathleen Morrill, who led this study. It’s super cool and I’m so glad this study tipped me off to its existence as now Indy is officially a citizen scientist and his DNA kit is on the way.
This study took the data of more than 18,000 dogs on Darwin’s Ark (about half purebred), including about 2,000 mutts who had their DNA sequenced. They then had the dogs’ humans fill out 12 short and easy surveys about their dogs’ behavior. Self-reported data is always tricky, and the authors acknowledge that: some owners may be more likely to notice behavior that is in line with what they breed they think their dog is, or they may see their dog’s behavior in a rosier light than a more objective observer. For instance, the owners of purebred golden retrievers were more likely to say their dogs are friendly to strangers, while the owners of mutts who unknowingly had golden retriever genes weren’t.
But unfortunately, owner surveys are the best and most accurate tool available as the only other option would be researchers personally observing 18,000 dogs for an extended period of time under a wide variety of conditions. We work with what we’ve got!
When they examined this data, they found that about 9% of a dog’s behaviors seemed to be tied to their genetics. Some behaviors are more heritable than others – “biddability,” which is a readiness to listen and obey commands, is very heritable and so purebred dogs with biddability as part of their breed standard were more likely to demonstrate that behavior. On the other hand, agonistic threshold, which is how easily a dog can be provoked or freaked out, is not very heritable so breed doesn’t tell us much about it.
Fun fact: the strongest association between a behavior and a specific gene was “getting stuck behind objects” and gene SNX29, which they point out is associated with cognitive performance in human genome studies. They also found a significant association between “howling” and, as they write, “an intergenic region between SLC38A11 and SCN3A, a voltage-gated sodium channel involved in the development of speech and language (in humans). Neat!
The study includes some really helpful infographics, I assume because the whole thing is based on citizen science and so they wrote it to be helpful to the general public, which is so awesome. Check out this graphic, which shows eight major behavioral propensities and whether they’re affected by genetics, whether they correspond to breed, whether there’s a significant difference between breeds, and other factors, all of which are shown relative to one physical trait: the size of the dog (the bottom line of dots).
As you can see, human sociability is very tied to genetics (big dot) and agonistic threshold really isn’t at all (the tiniest possible dot). And under “does breed matter,” you can see that no trait gets a bigger dot than 25% of the “body size” control.
Why doesn’t breed have more impact on behaviors? Well, the researchers point out several factors: first of all, dogs have evolved alongside humans for thousands and thousands of years, while the modern breeds we have today have only been bred for a tiny blip of a few hundred years – not necessarily long enough to really dial in behaviors.
Second, with the exception of carefully bred working dogs, most breeds that live in our homes as pets today aren’t bred for behavior. They’re bred for aesthetics! It’s the same reason why this apple LOOKS delicious, is even CALLED “delicious,” and yet it tastes like fucking TRASH. It’s a trash apple. I’d rather eat a raw potato. Because the Red Delicious apple was selectively bred to be marketable to the average supermarket shopper: brighter reds, more uniform color, bigger fruit, and a thicker skin to allow it to be shipped around the country, all of which came along with a corresponding loss of flavor, making it the worst fucking thing in the produce aisle. It’s like if sadness was edible.
In the same way, we have golden retriever puppies that have never been more golden but might not necessarily like the people who have mistakenly chosen to purchase them instead of going to the shelter and getting a, frankly, way better dog at a fraction of the price. And consider that these days, responsible breeders are also concerned with making sure their inbred products don’t end up with one of the hundreds of genetic disorders common in purebred dogs, like hip dysplasia or cardiomyopathy. And yes, by the way, this study did confirm that purebred dogs are super inbred.
It’s also not particularly surprising considering what we know about how genetics affect human behavior: not a whole lot! Our genes definitely influence us in many ways, but they aren’t simply the program that runs the computer. Why would we expect dogs to be different?
All of which brings me back to Bug, here. You might call her a pit bull, and assume she has the behaviors you associate with pit bulls, and now you know which behaviors those are more likely to be: it’s true that 150 years or so in the past, pit bulls were bred to fight bulls and bears, and when that was outlawed they fought rats, and then other dogs, just like other fighting dog breeds like bulldogs, akita inus, and sharpeis. But they weren’t simply bred to be mean – at the time, the ideal pit bull’s behaviors were tenacity (they didn’t want their dog to give up a fight) and most importantly an incredibly strong resistance to hurting a human. When the dogs were fighting, humans would be in the ring with them. If a dog gets all hyped up in a fight and then turns on the ref, that’s the fastest way to lose a fight. Permanently, in fact. A dog that bit a person would likely get the Ol’ Yeller treatment. So pit bulls were selected for the ability to fight other animals but never humans.
When we look at the heritability of behaviors from this study, we see that the strongest genetic factor, by far, is human sociability. Agonistic threshold, which again is how easily a dog can be provoked, is at zero. So from what we know about how pit bulls were bred, and from what this study shows us, we could expect that if you took the average pit bull bred in 1854, any genetic influence it had would mostly just cause the dog to be very affectionate and gentle with humans, which is probably why in the mid-20th century they had a reputation for being “nanny” dogs for children. And I don’t know about you, but I’m not even adopting pit bulls specifically bred to fight dogs or bears or bulls in 1854. I’m just fostering this little velvet hippo who landed in a packed shelter here in 2022.
And as I pointed out earlier, “pit bull” isn’t really even a standardized breed. In fact, here are six of the dogs whose DNA was sequenced for this study. Which one is the American pit bull terrier? I’ll give you a few seconds to make your guess.
The researchers also had people guess which one was the pit bull, and here are the percentage of people who guessed “pit bull” for each dog. But all six of these dogs have almost exactly the same amount of genetic ancestry detected from American pit bull terriers: 25-30%. In fact, “American pit bull terrier” was the most common breed found in this study’s mutts at nearly 10%, with the next closest being lab retrievers at 6%. This is a self-selected group of dog owners, of course, but if 10% of the mutts out there in the world are walking around with a significant portion of pit bull genetics, then you may want to rethink your stereotypes.
So why do pit bulls seem to dominate our bite statistics? Well, based on this study and others I have some ideas: for a start, people tend to call any dog a “pit bull” if it looks a certain way, regardless of if it is even in that collection of breeds that the term tends to encompass. Even shelter staff do this, according to a study from 2015 that found that out of 62 dogs identified by staff as pit bulls, genetic tests showed only 25 were pit bulls. In 2018, another study showed that among dogs that DO have pit bull genes, the majority had less than 50% of those genes.
So if pit bulls are so commonly over-identified in general, would it surprise you if they’re also over-identified when a dog behaves in an aggressive manner? It wouldn’t surprise me!
And if many of these dogs don’t even have a “pure” pit bull DNA, isn’t it much more possible that many of these dogs are biting because they’re one of the most mistreated dogs in the Western world, bred and abused and thrown in shelters and put down at rates far worse than any other kind of dog? That wouldn’t surprise me, either.
Does that mean I think pit bulls are no more dangerous than any other breed? Absolutely not. Look, pit bulls AND all the other dogs who are mistaken for pit bulls have traits that make them more likely to cause damage when something goes wrong. Dogs are dogs and sometimes things happen. They panic, they play too hard, they lash out due to mistreatment, and when that happens in an 8-pound chihuahua it may not be a big deal. But when it happens with a dog that weighs 50 pounds, it may be a big deal. And when it happens with a dog with particularly muscular jaws, it may be a very big deal. Every dog should be raised with love and boundaries and effective training, but it’s especially important for dogs that can cause more damage.
That’s why I’m not just providing Bug with a place to live for a few weeks, but I’m also teaching her how to be a polite member of society. She doesn’t know it, but she’s got a lot of human hatred to overcome if she’s going to have a good life.