Today’s pet peeve column is inspired by a friend who sent me some screenshots of a debate that was occurring on a FB page about the pros and cons of ear cropping.
On one side, Person A argued that ear cropping was a good thing because it mimicked natural selection. Dogs with droopy ears get more ear infections (citation needed), so when you crop ears, you’re mimicking natural selection by making them less prone to disease.
On the other side, Person B argued that if drooped ears caused an evolutionary disadvantage, then they wouldn’t have evolved in domestic dogs at all. Furthermore, Person B suggested that drooped ears are actually healthier anyway because they developed alongside the docile temperament that evolved in dogs after they were domesticated, citing the Russian fox experiments.
Ear cropping aside, both Person A and Person B are using really poor arguments to support their conclusions. And here’s why.
In a context where artificial selection is the primary deciding factor for an animal’s reproductive ability, natural selection is more or less irrelevant.
We breed dogs to hunt, to herd sheep, and to be cuddly companions. We breed them to match our lifestyles. We breed them to match our cultural notions about what certain types of dogs should look like (breed standards). We do NOT breed them to match the wild type domestic dog, the pariah type or street dog (pictured above); they can do this all by themselves.
Natural selection and artificial selection are two very different things, and only the latter is really relevant in dog breeding.
In any context where we are breeding toward a specific goal, we are using artificial selection to achieve that goal.
If you breed dogs low to the ground so they run into holes after badgers and vermin, that’s artificial selection. If you breed dogs to look like sheep and fool livestock predators, that’s artificial selection. If you breed dogs to look like toads because somehow you think that makes them look more badass, that’s [really stupid] artificial selection.
Any way you slice it, selective breeding is artificial selection.
Artificial selection does not work nearly the same way as natural selection because humans are dictating what does and does not contribute to the gene pool and how big of an impact specific dogs have upon the collective genome of a breed.
This should be really obvious. If not, think about it a little harder: artificial selection means selection is controlled artificially (via human interference), as opposed to natural selection where the shape of the gene pool is derived naturally through survival of the fittest and natural factors like geography as well as random events like epidemics or natural disasters. Of course humans control the contents of the gene pool in a population shaped via artificial selection!
For example, in Italian Greyhounds, the American dog Ch. Dasa’s King of the Mountain was used so prolifically at stud that 99% of American IGs have him at least once in a 10 generation pedigree and his genes make up about 18% of the modern American IG genome. In a wild, free-breeding population, this would never ever happen outside of an extreme situation where a tiny, geographically isolated population experienced an extreme genetic bottleneck where only one or two males existed at one point in that population’s history and the population remained geographically isolated long after this bottleneck had occurred. The phenomenon of popular sires completely goes against anything that would occur via natural selection in wild populations, just like a lot of other things which happen fairly often in the context of artificial selection.
Artificial selection can easily distribute a large number of deleterious genes throughout a population and saturate the population with these genes, where natural selection naturally works against this. This may happen either intentionally or unintentionally.
In reality, domestication is the perfect setting for the proliferation of deleterious genes and traits. Here’s a recent study which confirms this.
The exact reasons why this is the case will be explored in this article. For starters, we need to examine how genes move differently through wild, free-breeding populations than they do through selectively bred populations like dog breeds.
The ways in which genes flow through a selectively bred population are very different from the ways genes flow through wild populations.
This should be another “DUH” thing especially considering the popular sire example I previously referenced. The following paragraphs describe other examples of this basic principle.
Wild animal populations, including free-ranging street dogs, naturally tend toward higher levels of heterozygosity (meaning they have a higher level of genetic diversity) than selectively bred populations. This is because a high level of genetic diversity is protective against the negative effects of inbreeding depression which can have a huge impact on reproductive fitness, so animals have evolved in ways which promote maintaining high levels of genetic diversity whenever possible.
One of the simplest means of looking at the genetic background and diversity of a dog population is looking at the diversity and distribution of dog leukocyte antigen (DLA) haplotypes. DLA haplotypes are blocks of genes that encode for part of the genome called the major histocompability complex (MHC) which governs the immune system. The MHC is one part of the genome that especially benefits from a genetic principle calledoverdominance, whereby loci that are heterozygous for different alleles have an advantage in reproductive fitness over loci that are homozygous, which essentially means that higher levels of genetic diversity is inherently superior from a survival of the fittest standpoint. A population that has a large number of DLA haplotypes which are fairly evenly distributed will have superior immune function than a population which has a fewer number of unevenly distributed DLA haplotypes, meaning lower rates of immune-mediated diseases and better resistance to infection and disease. Street dogs undergo natural selection and have the ability to free-breed like wild animal populations, so their DLA haplotype diversity and distribution is essentially a blueprint for a hardy canine immune system.
I previously wrote briefly on the subject of DLA haplotypes in street dogs versus different dog breeds. Here’s a quote from page 106 of The Genetics of the Dog ed. Ostrander & Ruvinsky:
There is clearly much diversity to be found in semi-tame and feral street dogs, as has already been demonstrated in Bali street dogs (Runstadler et al.,2006). […] These dog populations are more outbred than most domestic dog breeds, and this is demonstrated when we assess the number of different haplotypes found in each group (where n = 50-100 dogs) and compare the frequency of the most common haplotypes. The highest haplotype frequency is around 12%, and there are, on average, 15 haplotypes with frequencies of 2-10%, with a further 30 or more haplotypes at lower frequencies.
I then compared the genetic diversity of the street dog to that of the Italian Greyhound:
In stark contrast to the healthy diversity of the street dog is the genetic homogeneity of the Italian Greyhound, which suffers from the highest number of immune-mediated diseases of any known breed as a result.According to the research done by UC Davis, the breed is split into two genetically distinct subpopulations of American and European dogs with slightly different DLA haplotypes and allele frequencies. Together, the two subpopulations have 18 DLA haplotypes, with 5 of these being unique to the US and 4 unique to Europe.
Of the two groups, the American dogs are slightly more diverse and have somewhat more evenly distributed allele frequencies. Of their 14 DLA haplotypes, 2 have a frequency >20%, 2 have a frequency of 10-19%, 5 have a frequency of 2-9%, and 5 have a frequency of <2%. Their most common haplotype has a frequency of 21.5%, and the second most common a frequency of 20.1%. All of the 5 haplotypes that are unique to the US subpopulation are very rare, with a frequency of <2%.
The European IG population has 13 known DLA haplotypes with a notably uneven distribution. 2 of these have a frequency of >20%, 1 has a frequency of 10-19%, 4 have a frequency of 2-9%, and 6 have a frequency of <2%. Their most common haplotype has a frequency of 31.5%, and the second most common a frequency of 24%. All of the 4 haplotypes that are unique to the European subpopulation are exceedingly rare, with a frequency of <1%.
It should be obvious that selective breeding has greatly affected the genetic make-up of the MHC in IGs, even though no IG breeders were actively selecting for DLA haplotypes prior to the creation of UC Davis’ IG genetic diversity test, which was just released last year.
So, if no breeders were actively selecting for these differences, how did they come to arise? The simplest answer to this is a genetic principle called linkage disequilibrium.
Linkage disequilibrium refers to the state of two or more genes being typically inherited together when under normal, natural circumstances, this would not be the case. When a lot of a population shares many of the same genes, which is very common in purebred dogs, linkage disequilibrium increases as there’s a much greater chance of genes being inherited together when there’s little allelic variation to randomize the assortment and inheritance of specific genes.
For example, in purebred Dalmatians, the population is genetically fixed for a single gene that causes high uric acid production and thereby significantly accelerates the formation of kidney stones. There is no aspect of the Dalmatian’s physical appearance or phenotype that necessitates this to be the case (e.g. their spotted coat does not inherently cause high uric acid production). However, over the years, Dalmatian breeders continued to select for genes which were commonly found in the same dogs which carried the gene encoding for high uric acid production. After enough time, these genes entered a state of linkage disequilibrium to the point where the Dalmatian as a breed became fixed for HUA. (For more information on this gene and what some breeders are doing to ameliorate these health issues, check out the site for the Low Uric Acid Dalmatian project.)
Several factors can accelerate linkage disequilibrium by increasing the number of genes which are inherited as blocks, or haplotypes. Inbreeding, small effective population size, and low levels of heterozygosity are a few of the biggest factors involved. All of these come into play much more frequently in selectively bred dogs, especially purebred dog breeds, than in wild animal populations.
The way in which genes ebb and flow within a population controlled by artificial selection are very different from that patterns we observe in wild populations. Referring to these terms as if they were interchangeable is wildly incorrect.
Some breeders intentionally select for traits which have a negative impact on reproductive fitness, which would never happen in the context of natural selection.
Some of these traits, like the extreme brachycephaly of the Pugs and several other breeds which causes a long list of health problems, are obvious. Others, not so much.
For example, the natural bob-tail gene, which is found in several breeds including the Australian Shepherd, Pembroke Welsh Corgi, and Swedish Vallhund among others, is a lethal semi-dominant trait which almost always leads to the death of all homozygous NBT puppies in utero. In at least one known case study, it also leads to extreme, fatal spinal defects of homozygous puppies which do survive to birth.
In the context of natural selection, this trait would die out very quickly and almost certainly never become highly prevalent throughout a wild population. An up to 25% loss of fertility is HUGE; dominant deleterious genes which result in even a 1% loss of reproductive fitness will be very quickly lost from a wild population. And yet, many dogs of the aforementioned breeds do carry this NBT gene, because breeders consider natural bob-tail to be a desirable trait that deserves to be reproduced, so the respective gene quickly saturates the breed population.
A list of similar traits (not exhaustive) is as follows:
All of these traits would never become common in the wild, but all of them have become common within specific breed populations (or in the case of the last trait in the list, is shunned by the breed community in general but is still bred for by certain breeders who find the trait appealing). To breeders who select for these traits, the appeal of the traits outweigh the drawbacks associated with them. The selective criteria is far different than that which would occur via natural selection.
Pet dogs live in a protective bubble of human care and assistance which partly negates many of the harmful traits a dog may have. Wild animals produced via natural selection have no such protection.
This is one of the biggest reasons why comparing selectively bred dogs and free-ranging pariah dogs or wild animals is like comparing apples and oranges.
If a pet dog contracts a nasty infection or disease and their immune system isn’t strong enough to fight off the infection on its own, we take them to the vet for antibiotics and supportive medical treatment. If a street dog contracts such an infection or disease, they will probably die and their genes will be lost.
If a dangerous disease becomes a threat to our pet dogs, we develop a vaccine to protect them against it so they ideally won’t contract it at all. If a street dog is exposed to such a disease, if they do become infected, they will probably die and their genes will be lost.
If a pet dog breaks a leg, suffers from crippling joint disease, or becomes paralyzed by IVDD, we bandage them up, give them NSAIDs and pain medication, insert pins surgically or give them artificial joints, and buy them slings and wheelchairs so we can help them move around. If a street dog suffers a similar injury, they will be unable to forage efficiently and evade predators. They will probably die and their genes will be lost.
If a pet dog easily succumbs to heat stroke and/or can’t breathe efficiently, we give them air-conditioned living spaces, oxygen cages, or even surgical correction of their deformities, all while feeding them nutritious meals every single day. If a street dog suffers from these conditions, they may die of heat stroke and will be unable to forage efficiently or evade predators. They will probably die and their genes will be lost.
If a pet dog is unable to mate, conceive, or whelp naturally, we use artificial insemination, hormone treatments, and perform C-sections so they can produce puppies. If a street dog has no practical reproductive ability, they die without passing their genes onto the next generation.
We dote on our canine companions to no end and spend thousands of dollars on their medical care while veterinary and genetic research accelerates further to give us even more treatments and cures for their ailments. We give them every opportunity to thrive and pass their genes onto the next generation, if we so desire.
The street dog has nothing. If he’s lucky, he’ll get handouts and possibly be vaccinated through a feral dog vaccination program. If he’s not, he will die young from an easily preventable or treatable disease, or through starvation, predation, extreme weather conditions, or natural disaster. His genes are at risk for being lost from the gene pool at every turn, and he only reproduces when he is able to overcome the many challenges of survival and has just enough luck to avoid any random events which would kill him regardless of his genetic make-up.
Like I said: apples and oranges.
“So, it if natural selection has very little bearing on how our pet dogs survive and reproduce, does that mean we can breed them to be handicapped just because it strikes our fancy?”
NO! That is beside the point of this article.
Just because natural selection has little impact on the form and function of our selectively bred dogs does not mean it’s ethically acceptable to breed them to be crippled and diseased, whether we may do so intentionally or not. We should make every effort to minimize the suffering we inflict upon living beings, whether they exist in a wild or domestic context.
“It sounds like you really don’t like selectively bred dogs and think that dogs that are born via natural selection are superior. What gives?”
There is a strong cultural movement in the Western world to make “natural” synonymous with “good, healthy, efficacious, superior” and “artificial” synonymous with “bad, unhealthy, ineffective, inferior.” But the truth is that “natural” and “artificial” (or “manmade”) are factual descriptors with no inherent value judgments attached one way or the other.
Quite recently, one of my friends attempted to sell me one of those “all natural” health products which is supposed to promote weight loss, improve immune function, boost GI health, cure cancer etc. When I expressed my concern that this product had not been empirically proven via research study to be safe or effective and had not been approved by the FDA for the prevention, treatment, or cure of any disease, her response was that it was illogical for me to be willing to put “manmade chemicals” (FDA-approved drugs) into my body while being skeptical of “plant-based remedies” like her product. I reminded her that heroin, cocaine, and ricin were allalso plant-based products which were obviously much more harmful than helpful. Though “natural” might seem to be synonymous with “safe and healthy,” basic analysis of this line of thought proves that it’s not the case. “Natural” means just that: “natural.” Nothing else.
The same is true when comparing natural and artificial selection. They are simply descriptors of two different modes of genetic selection with distinct attributes.
Natural selection creates animals with a broad stroke of the brush. The only selection criteria is whether or not an animal is able to survive and reproduce. It is indiscriminate about specific traits beyond their effect on reproductive fitness.
Artificial selection creates animals with a greater degree of precision regarding specific traits. It enables us to breed dogs of all shapes and sizes fitting a wide variety of niches and uniquely equipped to perform special tasks, regardless of whether those specific traits would be useful in a survival context apart from human intervention. That is its beauty and its power.
I for one am much more comfortable with my docile, diminutive GMO dogs snuggling with me in bed than I would be a full-blooded wolf. I have no doubt that my dogs would die quite quickly in the wild. They’re tiny, lightly coated, thin-boned, and not too bright. But they aren’t living in the wild. They’re living with me in my home, and my requirements for the ideal dog which fits my personality and lifestyle are quite different from those traits which would be most useful for survival in the wild. Their value and usefulness is completely relative, which is why it only matters if their traits are well-suited to my life since the judge of that is ME!
At the same time, artificial selection is a double-edged sword. Just as it can produce dogs with very specific desirable traits, it’s equally capable of producing dogs which suffer from very specific diseases and deformities, all passed down via genetics.
This is why we need to be acutely aware of the features and distinctions between natural and artificial selection, rather than conflating them as being more or less the same. Ideally, dog breeders should strike a balance between maintaining enough homozygosity to produce specialized dogs of a consistent type and maintaining enough heterozygosity to reduce the risk for producing dogs with genetic diseases. It’s not a matter of all-or-nothing, purely natural versus wholly artificial, or good versus bad. It’s a balancing act, one that we need to be thoroughly familiar with in order to breed the best dogs, whatever that may look like to an individual person.
So stop talking about natural selection as being the ideal we need to mimic as dog breeders or as if it had a great degree of relevance to dog breeding at all. Start talking about making smart breeding choices to artificially select our way to healthy, happy dogs.