A Breed Apart II

The culmination of the previous post is this statement:

Our phylogenetic, clustering, and principal components analyses all suggest a genetic split within the breed between working and show Border Collies that is probably as large as the genetic distances between some breeds.

Is this true? Is this significant? Is this a problem?

(1) Is there a “genetic split” in Border Collies?

No. But there is most certainly clustering. The use of the word split is misleading (if not inappropriate) and certainly not supported by the data. Split implies a disconnect, a divide, a severance. Irreconcilable differences. Speciation, or perhaps in this case breed genesis. If you want to call the clustering a split, you must concede that there is nothing preventing that split from being filled.

Whereas speciation has a specific genetic mechanism (two different species can not mate with each other to produce fertile offspring), there is no such genetic condition to define breed. A breed is what WE make of it.

All Border Collies (working, trialing, show and sport) come from the same stock just one hundred years ago. Unlike English Shepherds or Australian Shepherds or McNab dogs, the clustering and refinement of the Border Collie factions has not included outcross breeding with other breeds. Barbie Collies and Coyote Collies have no genes that weren’t there in the original stock 100 years ago. They are both every bit as Border Collie as the foundational stock was and the odds that significant mutations have occurred within remote clusters in 100 years is minute at best.

The driving force behind differences between the breed today and 100 years ago is not injection of DNA from other breeds.

[2013 update: In the years since I wrote this article I have amassed a rather extensive pedigree database of Border Collies from all three continents.  I have found that there are actually Keplie ancestors listed in the earliest stud books in Australia as well as undocumented farm dogs listed within US pedigrees that follow the original import of dogs from the UK but predate the solidification of “Border Collies” as a breed and the keeping of stud books.  While I can’t calculate the genetic impact of this admixture, it does open the door to genetic differentiation via unique genetic stock that I discounted in the original article.]

The notion of a genetic split is thus laughable, since there’s nothing genetic preventing anyone from bridging those gaps. Unlike a real tree that is permanently severed when a branch splits off, no physical barrier prevents the un-clustering of the breed or the preservation of genetic diversity. The branches will only split when true speciation happens, and that’s unlikely. You can still breed a Chihuahua with a Mastiff if you want to, let alone a Barbie Collie and a Coyote Collie.

Working and Show Border Collies certainly aren’t a different species and they are hardly a different breed. I myself have bridged this “gap” by breeding a dam with Australian show lines and American sport lines with a stud who has strong US and UK working lines.

(2) If it’s not a split, what is it?

What we have here is evidence of clustering due to linebreeding, inbreeding, selection bias, a small sample size, and geographic isolation. The scientists have purposely sought out a very specific line of dogs (they attend Sheepdog trials in the US) to the exclusion of others. And they are comparing this group to a mere 5 Oz show dogs. If your sampling practices aren’t random and don’t include enough individuals, your data isn’t likely representative of the breed as a whole.

Trying to measure this “split” is asinine given the reality of the data. Selecting the most linebred samples from the most distant continents is more likely to reveal the greatest genetic distance between lines in a diverse breed, not reveal an unfilled gap worthy of the label “genetic split.”

(3) Why does this clustering exist?

There are 3 main geographic areas for Border Collies and 4 main activities: {UK, USA, OZ} and {Working, Trialing, Show, and Sport}. I’ve made the case before that we shouldn’t confuse trailing with working, they are not the same. If we have 4 activities from 3 locations, we have 12 possible pairings and among those twelve, the two which anyone can tell you are the most distinct are US working lines and OZ show lines.

The researchers are cherry picking those two groups and are somehow surprised that there’s a difference. The US has the weakest and smallest pool of show lines and Australia has the weakest offering of working lines. Specifically the import/export of those two groups are low. Thus, the geographical separation means that those two groups aren’t likely to meet and breed. Popular dogs in each area after the initial exodus of dogs from the UK to OZ are less likely to drive the breed in the remote location.

As breeding fads pare down the genetic diversity in both of those groups, it’s not surprising to see that they appear to be split. But they did not grow apart, the branches between them have simply been trimmed by obsessive inbreeding. Those two groups are not any farther apart now than 100 years ago, the diversity of dogs between them has simply been pruned.

A century ago, you wouldn’t say we have a breed spit, you’d say we have a genetically diverse breed. Now, the trimming of the Border Collie family tree has left us with a less diverse breed in both the number of alleles and in the enforcement of homozygosity. The diversity of genes is less on both a breed wide basis and in individual dogs.

The best way to combat the problems of inbreeding and isolation is to not inbreed and overcome isolation by breeding from diverse stock.

(4) What does the chart tell you about Border Collies?

When I look at this tree, I notice a limited diversity of lines branching off to the right, a horribly clustered group of clonelike dogs fighting for the same spot on the center top, and a hand full of show dogs down the trunk that are no further from the major cluster than many dogs along the branch to the right.

The heavy clustering at the top is much more detrimental to the health and diversity of the gene pool than any distance between clusters. Great distance between clusters is a wonderful thing. It allows for outcrossing to dogs that are still under the Border Collie umbrella, keeping healthy hybrid qualities alive within a breed that is still a single breed. Free from constraints of political organizations that say you can’t outcross with other breeds and your gene pool can only shrink and not grow.

There will come a day when we will miss what genetic diversity remains in our dogs and wish for the day when they all weren’t so homozygous and inbred. One means of slowing our approach to that day is to allow greater and greater outcrossing and by maintaining the diversity you already have.

(5) Is clustering/splitting a problem?

Australian Shepherds are a younger and still more diverse breed than Border Collies. You can see it in their tree and you’ll see it evidenced in other genetic studies of the dog where the only errors in identifying a dog’s breed by its DNA came from the Aussies. They’re harder to pin down because they are not as homogenous as older and more inbred breeds.

Despite having smaller numbers sampled, we can clearly appreciate that the Aussie tree is has more diversity and branches in the Noise Phobia study than the other breeds in the chart:

We can see several clusters located at the extremes of this tree and the distance between the cluster at the bottom left and the top right is pretty large. Do we see the researchers aghast at a rift in the Aussies? No.

All these dogs are Aussies, and we’d look at this tree and say that they are a diverse breed. But what if we simply trim some branches. What would we say about this new tree?

This scenario could happen if we didn’t sample dogs from the missing branches or perhaps if we exported only dogs from this smaller tree to another continent and thus limited the gene pool. Would we now say that the cluster at the bottom right is any less of an Aussie than the branches at the top left? Is there now a breed split simply because we have a smaller cherry picked sample? Should we revoke our naming of some of these dogs as Aussies? Should we prevent the cluster at the bottom right from breeding with the top branches?

Clustering becomes a problem when you lose diversity and increase homozygosity. Inbreeding creates clustering.

Splitting only becomes a problem when humans decide that they want to accentuate the differences and throw away diversity by trimming distant branches and preventing or discouraging those diverse animals from being bred. Exacerbating these slits creates bureaucratic rifts, not genetic rifts in breeds.

(4) Why should we emphasize splits if they’re not that important?
Sadly, the sheeple want to exacerbate and create splits in the breed and to use rules and institutions to enforce a physical divide that nature can not. They want to use the ABCA registry as a pruning device to keep the show dogs out and kick out any sport dogs that are dual registered with the AKC. They want to use the Registry as their tool to prevent others who do not abide by their philosophy, it is not enough for them to live their philosophy, they must force it on others as well.

They do this in the name of the breed, but it’s essentially fascism. There’s nothing preventing sheeple from breeding and buying the dogs they want within an open registry and inclusive gene pool. They are hacks and fools if they are unable to research their breeding and buying choices any more deeply than looking at the three letters in front of a dog’s registration number. In their own bigotry, they want their rivals out of the gene pool and they want what should be an information keeping service to be a defacto segregation service.

One such sheeple is Dr. Melanie Lee Chang. She’s essentially a dog sport person who sides with the sheeple camp. Her oldest Border Collie is a red (gasp!) rescue dog whom she has trained to a CGC (gasp! title chaser!) and a Novice Agility Tunnelers title (gasp! dog sport!), but her newest Border Collie is a trained imported trialing dog from the UK.

This is an excellent example of why trialing and the trialer jock sniffers are a sport and not work, as I doubt Melanie owns a farm or sheep or owns Border Collies for a higher purpose than playing at herding. It is clearly a hobby and she’s clearly bought into the elitist notion that putting up the big bucks for a European model is the way to go.

I’ve run into Dr. Chang on the BC Boards and besides being hostile to logic and debate, she also displayed many of the misandrist qualities so common to other women on those boards. She was particularly off put by my stature as a large built man.

Now it shouldn’t surprise you to find out that the hyping of the “genetic split” notions in the Noise Phobia study was done by none other than Dr. Chang!

Dr. Melanie Lee Chang: Dr. Chang is a postdoctoral scholar in Dr. Hamilton’s laboratory at UCSF. She holds PhDs in physical anthropology and ecology/evolutionary biology from the University of Pennsylvania, with primary research interests in systematics and phylogenetically-informed reconstructions of character evolution and population history. At this time, Dr. Chang’s primary canine project is investigating the genetic background of noise phobia in Border Collies. She has three Border Collies of her own, Solo, Fly, and Jett, with whom she has trained and competed in agility, flyball, and sheepdog trialing.

You don’t suppose that Dr. Chang’s own Border Collie politics have influenced her interpretation of the data and the selection bias present in the method for her experiment, do you?

And while we’re on the topic of Dr. Chang, I have to wonder what’s up with the name of her blog, CAVEDOG Canis Soloensis. While I know that intellectuals just love using Latin to make themselves sound erudite and thus smarter than you, it helps if you get the Latin right. Canis does mean dog, but Soloensis does not mean cave. Not even close.

Dr. Chang obviously got the name from Homo Erectus Soloensis, commonly known as Javaman, trying to equate Cavemen — most often referring to european Neanderthals or Cro-Magnon man — with Javaman. But Javaman wasn’t really a cave man, and Soloensis isn’t descriptive of a cave or of cavemen. You might mistake the Latin for “solitary,” but just as Neanderthalensis means “from the Neander Valley,” Soloensis means “from Solo.” And Solo is a city on the island of Java in Indonesia.

So for someone who has a PhD in physical anthropology and evolutionary biology and is all up tight about not calling barbie collies border collies, you’d think she’d get the name right. I had to dig deeper.

It turns out that the good doctor rescued a rebound dog named Franklin and renamed him Solo.

“I named the dog a name I didn’t even particularly like. It was just the first one that came to me and it didn’t matter”

Like many incongruous retcons, the name came first and the logic came later. In true showple fashion, which is strange for a sheeple, Franklin became Ashfall Solo River.

Solo is named for Ashfall Fossil Beds, Nebraska, a remarkable lagerstätten preserving loads of three-toed horsies, and Indonesia’s Solo River, where fossils of Homo erectus were first found.

Well, let’s giver her credit for a clever retcon that links her passion for anthropology and Border Collies, but question why someone who knows so much about evolutionary bottlenecks would want to create one in her own breed of choice and force her politics on others.

* * *
Comments and disagreements are welcome, but be sure to read the Comment Policy. If this post made you think and you'd like to read more like it, consider a donation to my 4 Border Collies' Treat and Toy Fund. They'll be glad you did. You can subscribe to the feed or enter your e-mail in the field on the left to receive notice of new content. You can also like BorderWars on Facebook for more frequent musings and curiosities.
* * *