Afghan Hound Times
(Afghan Hound Database and Breed Information Exchange)
User Login Home Register Community Forum Members Pages
Forgot User Name | Forgot Password | Update Your Registration | Create A Members Page |


*****NOTE. Since posting this article I have received expert and informed information from Jim Hickie (Gengala Afghans, Australia), who is considerably more qualified and knowledgeable than I on this subject. Jim Hickie has kindly provided an expert and very interesting article on AI. (please click here for Jim Hickies article) and is a recommended read prior to continuing with this page.

I considered re-writing my own text in the light of Jim Hickies kind and helpful response. However, I have decided to let my own text stand on the basis that it Is probably reasonably indicative of a lay perspective and is therefore a counterbalance to the more expert view. Obviously I would want to publish the best informed and accurate statements on this subject and for that reason I post Jim Hickies response to my own text first (ie the expert view and comment), and then follow this with my own original lay notes. I think this will work, let me know if you have any views. Also, again, our sincere thanks to Jim Hickie for his feedback and valuable writings. Thanks, Steve.

  • PART 1 Jim Hicke's comments on Steve Tillotson's AI Article

I have downloaded your A.I. article and pondered over its content and am intrigued by your use of words such as "conventional" and "normal" in relation to Afghan pedigrees. In the breeding of Laboratory animals and insects there is nothing abnormal in the types of pedigree that you have chosen to analyse (even without storing male genetic material over long periods of time).

The problem with Wright's Coefficient of Inbreeding is that it is a mathematical formula that makes no allowance (or weighting) for the degree of dominance that a particular sire or dam may have. Similarly with Wright's Coefficient of Relationship - it is only useful if you have a knowledge of the degree of dominance that a particular individual may have. Let me quote an example.Ch. Jorogz Heartbreaker was (and through A.I. still is) a most significant sire in the Afghan scene. He was inbred to Ch.Mecca's Falstaff who was a most significant sire in the 1970's. Heartbreaker is the result of inbreeding to a dominant sire. Now Heartbreaker is a great sire because he was closely bred to Falstaff - he was in effect a great dog with a great pedigree. No doubt if you research your database long enough you will find many dogs that are even more closely bred than Heartbreaker ( i.e. with higher coefficients of inbreeding) but who have made virtually no contribution to the breed.It is the dominance (for particular breed characteristics) that is the key to the success of a dog or bitch as a parent.

I regard these formulae - Wright's along with the Bruce-Lowe figure system (as applied to dogs)- as being useful tools for telling you how you have arrived at a particular situation rather than indicating how to breed the next generation.

Naturally the principles mentioned above have been the key to our success. If you research our most successful pedigrees you will find that our stock accumulates a bitch Ch. Kabik's Mindy in the pedigrees. Mindy was a Falstaff daughter and the basic reason for the successes of the Kabiks in the 70's and 80's. Similarly if you call up the (extended ) pedigree of our current top bitch - Ch Gengala Right This Way (A.I.) you will see that in spite of the fact that she is the result of an A.I. mating she is in effect fully linebred on the lines that have been good producers for us and that Mindy continues to figure prominently in the background. Again with the winning puppies from our most recent A.I. litter - also by Rhayders the Right Stuff (U.S.A.) - a similar degree of concentration on Mindy will be revealed. (Gengala Encore & G.Calculated Risk.)

On the other hand I do not want to give the impression that this is a formula for success - obviously one's selection criteria play a most important part in producing the next generation. I seem to have wandered somewhat. The real point at issue is that formulae such as Wright's are useful tools in a breeding program but they cannot indicate the direction that a breeder should go.

Also your concerns about a parent appearing in a pedigree many generations later after its initial appearance does not in any way diminish the validity of the mathematical results of Wright's formulae. Remember that the initial observations which led to Wright's development of these algorithms were based on laboratory experiments where it was (is) not uncommon to have ancestors reappearing in pedigrees as parents many generations later.

Regarding your comments about A.I. breeders keeping accurate records of their results, I should indicate the procedures that take place here.(Australia)
(1) Progeny of A.I. litters have (A.I.) incorporated in their registered names as an affix.
(2) On the Registered pedigree the sire has his country of registration listed in parentheses after the name.These two iniatives highlight to the pedigree analyst the exact state of how a dog was produced.

  • PART 2 Steve Tillotson's AI Article

In The Beginning - (Note - If you read Jim Hickies article you will see that the first recorded AI in dogs was undertaken in Italy around 1780, so the science is not as new as my writings indicate). The first successful AI breeding in Afghans (or perhaps any canine AI breeding) that I am aware of occured in 1976, undertaken by the leading pioneers Wendye and Stuart Slatyer of Calahorra Afghan Hounds in Australia. The "sire" of this very first AI litter was the English Ch Saringa's Abra Cadabra. It so happens that Abra Cadabra's pedigree itself is very interesting. The pedigree contains the three American imports to the UK - Ajman Branwen Kandahar, Wazir of Desertaire and Zardeeka of Crown Crest. Ajman lines are strong Grandeur, Wazir is strong Dutch Crown Crest, and Zardeeka contains Dutch lines as well as the silver Crown Crest lines. These three American imports are intermixed with traditional English lines of that era. Many of these lines, especially the Carloway's would fit very well with the Walliwog of Carloway and Mazari of Carloway imports to Australia and which exist in other Calahorra lines. So even at this pioneering stage, the prospects must have been tremendously exciting. In a later article I will look at other AI breedings from other parts of the world, but seeing as Wendye has kindly sent me details on her AI program it is convenient to start with Calahorra.

Turning The Clock Back - I dont know about you, but I do not find this an easy subject to get my head around. For example, the Slatyers have successfully bred several other AI litters since 1976, some of which involve semen collected nearly 20 years ago. If we could agree (for the sake of discussion) that the "normal" interval between generations in a pedigree is between 3 - 7 years, (so lets average it at 5 years), then in a four generation pedigree you probably expect to find ancestors dating back some 20 years earlier. Now, if you then use AI collected from a dog 20 years earlier, this pushes elements of the pedigree back to 40 years. As if that is not enough to cope with, if the most recent AI breeding involves a pedigree which already contains earlier AI bred ancestors then it becomes even more difficult to comprehend.

An Example AI Pedigree - For convenience of the following discussion we'll major on Dennis Smith's (Thailand) young puppy dog - Calahorra Social Contract, or Masoud as he is affectionately known.

Parents Grandparents Great-Grandparents Great-Great-Grandparents
Calahorra Below Zero The Flying Dutchman of Isfahan (us) Am Ch Charikar Wazir Am Ch Charikar Dominator Joh Cyn
Crown Crest Lululuv
Bardi Vdom Dutch Ch Nabob Vdom
Dutch Ch Ineke Vdom
Aust Ch Calahorra Fledgling Aust Ch American Eagle at Calahorra Am Can Ch Wildenaus Bonivant
Cadburyhill Beliatrix
Austral Ch Calahorra Rowena Calahorra Anglo Saxon
Aust Ch Calahorra Requiem
Calahorra High Society Calahorra Kasanova Calahorra Wherlwind Aust Ch Calahorra Harlequin
Calahorra Vapour Trail
Kaskarak Best of British Ch Montravia Kaskarak Hitari
Kaskarak Coppelia
Aust Ch Calahorra Fledgling Aust Ch American Eagle at Calahorra Am Can Ch Wildenaus Bonivant
Cadburyhill Beliatrix
Austral Ch Calahorra Rowena Calahorra Anglo Saxon
Aust Ch Calahorra Requiem

Web Pedigree Produced By "The Kennel Club Breed System"

The Conventional Pedigree View - If we viewed the above as a "conventional" breeding it would be interesting enough, but, maybe not particularly startling. The level of inbreeding (coeficient) is 14.6 percent which is typical for a half brother/half sister (linebreeding) mating. The Swedish/Dutch lines behind The Flying Dutchman Of Isfahan and the English lines behind Calahorra Kasanova (dam side) being an outcross and introducing "new" genes

Conventional Inbreeding Analysis - Well there's a word, in fact two, "new" and "genes". Lets explore this further.Firstly inbreeding influences. The list of "common ancestors" and their relative contribution is as follows;

Calahorra Fledgling             12.50
Flying Dutchman Isfahan          1.17
American Eagle Calahorra         0.78
Seven "other common ancestors    0.15
If you are eagle eyed you will have spotted (for example) that Saringa's Abra Cadabra (mentioned earlier) is a common ancestor, both sides of the pedigree in the fifth generation, so why is he not included in the list. Its simply the rules of calculating the inbreeding, he is behind Calahorra Fledgling, and this is the only pedigree route to Abra Cadabra, so he's not shown. By contrast , and by way of an example, the other seven common ancestors include Horningsea Tigers Eye (0.04 percent contribution). Tigers Eye registers because he contributes directly on his own (for example, Tigers Eye is in the pedigree of Abra Cadabra and in the pedigree of Montravia Kaskarak Hitari). This is complicated already and we havn't even given consideration to AI in this context yet.

Flying Dutchman Inbreeding Contributer - Well lets do that next. Flying Dutchman is the Grandsire of Calahorra Social Contract, BUT he also appears at the 6th generation behind Calahorra Renior, and in the 7th generation behind Calahorra Paradys Lost. Whilst you now might say "but Flying Dutchman is behind the main common ancestor Calahorra Fledgling so why count him in the inbreeding calculation as you excluded Abra Cadbra on this basis" - there is a difference. There are TWO different pedigree routes to Flying Dutchman, one direct via Calahorra Below Zero , and the ones via Fledgling - this is an example of the subtleties of the inbreeding calculation.

What Does All This Mean? - What does all this scientic jibberish mean? A good question. In my "lay" understanding (I do not purport to be a a Geneticist, or qualified in any such respect), it means, common ancestors provide the "inbreeding" in a pedigree. The higher the "inbreeding" from any particular common ancestors, the higher the "possibility" that the common ancestors gene's will be passed onto the next generation. Normally (excepting and allowing for "prepotency" of the dam, stud or othr ancestor), one would anticipate getting a majority of Fledgling's genes transferred on to Social Contract. Here's the rub - at first glance, the "single" appearance of Flying Dutchman in the pedigree, a "seemingly non common ancestor at 5 generations" would imply a contribution, but not as important as Fledglings. But there are several dimensions which alter this.

Dimensions To AI - Firstly, Flying Dutchman is in fact a common ancestor, we can calculate his inbreeding at 1.17 percent, which "is significant" in terms of the level of inbreeding (he is second highest contributor) in the entire known pedigree. Secondly, and this is the bit I cant get my head around, Flying Dutchman "should be about 7-14 years old in pedigree terms", in fact he was born in 1968, which makes him "nearly 27 years old", in effect, 5-7 generations ago is where he should be. In fact that is exactly where he "also appears" behind Calahorra Renior and Calahorra Paradys lost. Putting it another way, Calahorra Below Zero's sire, is also his great great great grandfather (a superhuman feat methinks). Yet another dimension is the interval between Flying Dutchman's appearences in the pedigree. He appears in the second generation, his next appearance is in the sixth generation - a gap of four generations. Again, this would not occur in a "conventional" pedigree. A "scientific question". Sewell Wright, a Chicago Geneticist developed the "coeficient of inbreeding" algorithm in 1922 - long before we had thought about using AI. Geneticists today state that Wrights algorithm is as valid as when first developed. It is a mathmatical algorithm, precise - the science of genetics is precise but "random" in its behaviour (dominent, recessive gene behaviour etc). My guess is that AI does not invalidate Wright's algorithm, and that "conventional" inbreeding calculations can be used when dealing with AI breeding. However, that is the "mathmatical" view, the genetic view may not be so clear.

A Precedent Exists? - The nearest thing I have seen to this situation occurs in the early American Afghan pedigrees, particularly with Badshah Of Ainsdart who has been recorded as appearing in a single pedigree across 5/6 generations. This would have occured when he was mated, late in life, to a second or third generation bitch of his own line. But even here there's a difference. In the case of Badshah, whilst he straddled several generations, he was consistent by his presence in contigious generations. With Social Contract (Masoud) there's a gap of three generations in Flying Dutchman's appearences in the pedigree.

Significant But How? - I think this "generation leap backward", especially when combined with "generation gaps in the pedigree" is significant, but I dont know how it would manifest in terms of the product of the AI mating. Is this "delayed reaction linebreeding" or is it "an injection of ""new/outside genes"". As if there were not already enough complications in this discussion, it is exacerbated when we look at Fledglings pedigree, he is of course the most important common ancestor..

AI Within AI Within The Same Pedigree - Fledgling's (dam side) grand-dam is Calahorra Anglo Saxon - an AI breeding out of Saringa's Abra Cadabra, who was born in 1970 (similar era to Flying Dutchman with some common lines to Vdom). Thus we have a "doubling up" in the pedigree of AI bred lines. As stated earlier, Flying Dutchman is behind Anglo Saxon at 2nd and 3rd generations.

Lets Talk Real Dogs - I have seen a couple of AI bred Afghans, but only briefly and from a distance, so I have no basis on which to make a meaningful statement. Perhaps, after this article I will pay more attention and give it more thought! I hope AI breeders, their owners, and breeders of stock produced from AI lines will keep good records. The technique of AI is becoming established more and more, but its still new and small scale. If we record and study this closely we might learn an awful lot more about breeding of Afghans generally. In particular it might help us understand how particular "traits and characteristics" are handed down (Mendelian Inheritence). Its certainly a fascinating area of our world of Afghans. I hope somebody out there with actual experience and knowledge of Afghan AI breeding will tell us more. Thanks. Steve.

Steve Tillotson May 1996


Library Of Articles/Main Menu Toolbar