Sunday, March 6, 2011

Just the basics: How to read a pedigree

I was trying to refer a puppy-seeker to some basic online information that explains how to read a pedigree chart, and was shocked to find that, while there are many sites out there that have complicated and pretty complete information on how to interpret the titles and abbreviations in a dog pedigree, there are none that do a clear job of the simplest principles. They all assume a knowledge base that most people who are not involved with animal breeding or genealogical research do not have.

One essential thing that many puppy-buyers don't understand is that a pedigree is not the same as registration papers. Every higher animal on the planet has a pedigree, whether or not it was ever written down or recorded -- it is just a chart that tells us who an animals' ancestors are. Registration papers may or may not come with a pedigree, and just show that an individual animal is recorded with a specific registry and (usually) that it is considered "purebred." Registration papers are often used as proof of ownership.

Let's start with a very simple pedigree chart:


Click to embiggen. Control-click (most browsers) to pop up a larger version in a new tab.

This is a copy of the three-generation pedigree for one of my own dogs, Moe. It is formatted by a free online dog breeders' website, but this does not matter. The pedigree is just information about the dog's ancestry, and a handwritten one on lined notebook paper would be laid out just the same and could have the same information on it. I entered the names into the form myself. The basic information that every pedigree must have (or it's not a pedigree) is the names of the animal's ancestors in a form that shows the relationships. (That counts for human animals, too.) Livestock animals might have a number rather than a name.

The pedigree reads from left to right. The name of the animal whose pedigree it is (let's call him Bubba) is either to the far left, or, as in this case, above the chart of ancestors. That saves horizontal space on the printed page. If this is a registered animal, this will usually be his or her registered name.

There will be two names in the far left column or the next column to the right if Bubba's name and info takes up the far left . The sire's (father's) name will be above, and the dam's (mother's) name will be below. The convention is always the male parent directly above the female parent. Moe's father is Shooter and his mother is Pipistrelle.

Then in the next column, you will see four names. The two above center line are the paternal grandparents -- Moe's father's father and mother. The two below center line are the maternal grandparents -- Moe's mother's father and mother.

Moe's paternal grandparents are "It'za Demost Happy Fella" and Brighton. His maternal grandparents are Dust-Dee and Cocoa.

This brings up a point that often confuses people looking at pedigrees. In many breeds the dog's registered name bears no resemblance to the name that people who know the dog will refer to him by. Often there is a kind of embedded "code" that is decipherable to insiders, including the breeder's kennel name and sly references to his ancestry. Strange spelling and punctuation is common. "It'za Demost Happy Fella" was called Chaz. I only know this because I've been told -- you can't derive it or guess it from the registered name. His owner followed the show-dog convention of using a very convoluted string of words to "name" her dog. Her kennel name was "It'za." As you can see, the owners of Moe's other ancestors followed the English shepherd owners' convention of fairly simple* registered names that are the same as, contain, or are similar to the dog's "call" name. Red Bank Shooter is called Shooter, and his owner's kennel and farm name is Red Bank. Houlahan's Pipistrelle is called Pip, and her owner is me, Heather Houlahan. My kennel and farm name is Brandywine. I didn't breed Pip, so she doesn't carry my kennel name.

In the third column, you will see eight names -- Moe's great-grandparents, all laid out so that you can see which grandparent each pair produced.

I've used a three-generation pedigree here so that it is easy to read and uncluttered.

The standard number of generations for a dog pedigree that is provided when you buy a puppy is generally five -- up to the great-great-great grandparents. That also happens to be the most extensive pedigree that can generally fit on a single sheet of paper (usually legal-sized) in readable type if you put Bubba's name above the "crane's foot"** of ancestors. Breeders who are researching their dogs' genetics will use spreadsheets or online databases to go back many more generations than can be fit on a sheet of paper. Some registries or services will sell up to an eight-generation pedigree printed on really big paper.

Now, let's get onto intermediate pedigree-reading. You can bail here if you found out what you need to know about reading a basic pedigree.

Here's a screen shot of the "top half" of Moe's five-generation pedigree as it is displayed by the English Shepherd Club Registry's database. When we talk about the "top half" of a pedigree, we mean the animal's sire's pedigree. Right now we are looking at Moe's father and his ancestors -- his mother and her ancestors are the "bottom half" of the pedigree. So we are really looking at Shooter's four-generation pedigree.

This is the "bottom half" -- Pip's four-generation pedigree.


Together they make Moe's five-generation pedigree. There are a lot of bells and whistles on this pedigree chart. You can see registration numbers, birthdates, colors, health information, owners' and breeders' names. I could configure it differently and get photos, where available, of three generations of ancestors. You won't get that level of detail on most pedigree charts from other registries, or most handwritten or home-produced pedigrees from breeders. You'll see that as you go further back in the generations, there is less information provided for each dog. This is necessary in order to fit all the names on a page.

Having a lot of information right on the pedigree chart is very useful when you don't personally know the animals. What a person or registry chooses to put on a pedigree tells you something about their priorities. This ESCR pedigree includes hip-health information, because hip dysplasia is a genetic health problem in our breed. If an owner didn't check his dog's hips or won't publish the results, that raises suspicions. If the hip information was good or bad and the owner published it, that gives important information about that dog that can help someone make buying and breeding decisions down the road. The rest of the information is mostly to help people identify the dog precisely (registration numbers) or find out more (owner and breeder information).

An AKC pedigree will include any show championships that the dogs have won. Show-ring wins are important to the AKC. The English Shepherd Club does not think that show-ring results tell us anything about the quality of the dog, so those will not ever appear on an ESCR pedigree. If a breeder is generating her own pedigree charts by hand or computer, she may include more information than a list of names; what she chooses to include may tell you about her priorities.

In order to read you will have to click to embiggen. Control-click to open in a new tab (on most browsers). But you might want to print them out on two pages and follow along.

A five-generation pedigree includes slots for sixty-two ancestors -- two in generation one, four in generation two, eight in generation three, sixteen in generation four, thirty-two in generation five.

However, it would be highly unusual for a purebred or purpose-bred dog's five generation pedigree to have sixty-two unique names in it. There will be repeats.

If a dog's name is repeated on both top and bottom of the pedigree, then Bubba is inbred on that dog.

If you look only at Moe's three-generation pedigree, at top, you would conclude that Moe is not inbred at all. All of his grandparents and great-grandparents are unique individuals, with no names repeated. His parents and grandparents were not cousins.

But if you look closely at Moe's five-generation pedigree, you will see that Kaschak's Brandy is both his paternal great-grandfather and his maternal great-great grandfather. Moe is inbred on Brandy in generations three and four. This is not very inbred by purebred dog standards.

He is also inbred on Butcher's Sam Odie, who occurs twice (top and bottom) in generation five. Again, this is pretty far back. It statistically makes Sam Odie the equivalent of a great-great grandfather, or 1/16th of Moe's genome -- 6.25%. By contrast, Moe is (statistically) 18.75% Kaschak's Brandy. Knowing something about Brandy -- his health, appearance, and behavior -- is much more likely to tell us about Moe than will knowing the same things about Sam Odie.

If a dog occurs multiple times only on the "top" (father's pedigree) or only on the "bottom" (mother's pedigree) then Bubba is not inbred on that dog -- the parent whose pedigree includes that repeat may be, but Bubba is not.

Moe's paternal great-grandmother Naomi is very inbred on Mohn's Boodie -- he is both of her grandfathers -- in other words, her parents were half-siblings. But Naomi's son Chaz (the guy with the long unpronounceable name) is not inbred on Boodie (as far as we can see from this pedigree) because his father, Kaschak's Brandy, does not have Boodie as an ancestor. Moe is not inbred on Boodie (as far as we can see from this pedigree) because Boodie does not appear in his mother's pedigree.

I hope that has been helpful information about how to read a basic pedigree chart. I've used a dog's pedigree here because that is the species I know best. An animal's detailed pedigree can be like a Russian novel for someone who is an expert in the specific breed. It takes many years of living with and learning widely about a breed of dog, horse, cow, etc. for a printed pedigree to easily give up its information that way. But a beginner can at least look at a five generation pedigree and get a sense for whether the animal was very inbred (as was Naomi) and question why. You can try to find the owners of the animal's parents and grandparents and find out more about them, such as their health and what kind of temperaments they have or had.

A pedigree can be a full of mistakes or frank lies. This is true of an official-looking printed pedigree certificate with gold-filigree borders as much as it is of a hand-scrawled pedigree on a brown paper bag. A breeder may not know that his bitch was bred by a different dog than the one he listed in the pedigree -- maybe the sneaky neighbor dog got in, or the stud dog's owner was careless in the kennel or dishonest. A breeder may lie about paternity or even maternity. Puppymillers use registration papers fraudulently, so a registry pedigree on a puppymill purchase is likely to be complete fiction. Most dog registry organizations can't or don't check the DNA for most litters, and there are even ways to get around those for the devotedly dishonest. An old mistake or lie is impossible to check. And both individuals and registries can introduce clerical errors; I have one pedigree that was produced and sold by the United Kennel Club in which a dog is listed as his own father (same registration number and all). Most pedigrees are probably mostly accurate, but it is important to note that they are just information that is reported and recorded by fallible human beings. If you can't trust the honesty or the competence of the people who provided the information, then you can't trust the pedigree to give you information about your animal.
_______________________

*Sometimes too simple. When the dog's full registered name is simply "Max," there is a high probability that there are going to be more of them. The common practice of using the owner's surname and the dog's simple name is better, but there turns out to be a bunch of Smiths, Wilsons, Thomases, Johnsons, etc. Sometimes this creates a great deal of confusion.

** The English word "pedigree" derives from Old French "pied de gru" or "crane's foot." This refers to the way the chart branches, literally looking like the foot or footprint of a bird. Pretty cool!

17 comments:

  1. This form should also be familiar to anyone who has done any serious genealogy.

    ReplyDelete
  2. It's somewhat different from the way most human family trees are charted, as it doesn't include anything but lineal ancestors, and relies on position rather than shapes/symbols to indicate gender and relationship.

    I deliberately did not get into the issue of the more comprehensive "genticist's pedigree" formats that one rarely sees for dogs unless one is studying specific genetic disease.

    ReplyDelete
  3. Beautiful explanation, Heather. My mother traced her genealogy back several generations. She made a copy and presented it to me as "half my pedigree".

    ReplyDelete
  4. Why didn't you include the formula to calculate the inverse ratio of championship ancestors to working ability?

    And no discussion of all those annoying acronyms like OFA, MDR, CERF and the like?

    Yeah, yeah I know, you don't usually find those on an AKC pedigree, but you will often find them on a good pedigree.

    ReplyDelete
  5. C'mon, J. There is teh Googlez.

    ReplyDelete
  6. Screw patience. Let's KILL something!

    (Like a pack of sleaze bag greeders...)

    ReplyDelete
  7. You have no idea how much of a help it was to read this. I know about pedigree reading (learned it in the Stallion section of the horsie magazines I read as a child), but didn't quite understand about how the whole "inbred to" thing worked. It'll definitely take some time for all this to sink in!

    ReplyDelete
  8. For racing greyhounds, the Greyhound-Data site has pedigrees that show the calculated coeffiency of inbreeding. There are a number of studs that are used a lot - they have 1000s of offspring, but there are a lot of dogs in the gene pool so the deleterious effects aren't as bad as they might be with a smaller breeding population. There's no testing for various conditions - their biggest health risks are osteosarcoma and lousy teeth - so you wouldn't have any idea about health concerns (although the cancer risk is common enough in all large breed dogs and the bad teeth are very common in all racing greys). Each pedigree also has 32 generation sire and dam lines, and that's an interesting trip to see how dogs from around the world can be direct ancestors to the dog in question.

    ReplyDelete
  9. Cool Kate.

    Both the health issues you cite in greyhounds are later-in-life problems -- so a purely winning-motivated kennel owner does not care, and in most cases, the dogs are out of their hands by the time it comes up. There's also a strong profit motive to conceal any genetic issues, just as there is a profit + pride motive to do so for show dogs *and* working dogs.

    I deliberately did not broach the COI issue because, for someone new to pedigrees, (head explodes).

    I think if I tackled it in the future it would be in the form of a guest post from someone who could lay it out more clearly than I think I can.

    ReplyDelete
  10. Laura has been pestering me to add COI calculations to the ESC Registry. I've wrote some code that computes it, so it's just a matter of figuring it out how and when to compute, store, and display it. Unfortunately I've got several more important things on the TO DO list, so COIs won't happen for a while.

    ReplyDelete
  11. I would love to have COI calculations in the dB.

    Realistically, it's a minority of ES for whom there are complete-enough pedigrees for the number to be meaningful.

    COI is a much more important issue in a long-closed gene pool.

    ReplyDelete
  12. If someone could specify just "how complete" a pedigree needs to be for the COI to be of any value, I could run some queries and figure out how many dogs actually would have an interesting COI. A ten generation pedigree has 1023 slots. What fraction of them need to be known for the COI to be interesting? Is there a reference to back up that choice?

    ReplyDelete
  13. I just really want to thank you for the word "embiggen".

    ReplyDelete
  14. It's a perfectly cromulent word.

    ReplyDelete
  15. Ah. I was sure too little television viewing would stunt my growth and now I've got it confirmed. Good thing I've got the internet! I shall do better, I swear.

    ReplyDelete
  16. Wow, I was just going to comment on "embiggen" also. "A noble spirit embiggens the smallest man." Love it! Kate

    ReplyDelete

I've enabled the comments for all users; if you are posting as "anonymous" you MUST sign your comment. Anonymous unsigned comments will be deleted. Trolls, spammers, and litigants will be shot.