What if the Turkish shepherd's dog is "white"?
I'll assume you know a little bit about classical (or Mendelian) genetics. Classical genetics is a blend of knowledge and theoretical assumptions supported by some experimentation with a few more assumptions thrown in. We know that each of the major genes has dominant and one or more recessive forms. The dominant forms may not be dominant over genes that are at another location (locus) on the chromosome, and may be affected by these other genes - so genes may be suppressed (hypostatic) by genes that override them (epistatic). Decisions are made between the the interaction of genes and sometimes with the environment and process of aging, toward the resulting outcome. update 30May2006: This previously linked article has gone to pay mode but related topics can be found here.
Molecular genetics is what much of current genetic study is based on. It is the study of DNA and exploring how sequencing and regulatory actions affect expression of genetic traits. Many but not all of findings have clarified some of the theories of classical genetics in the coat colors of dogs. Sometimes discovering a genotype as it exists on the chromosome doesn't really help explain why the phenotype (the end result of how it is expressed on the animal) is different, because the interaction of various genes with each other and how they regulate each other, are not yet understood.
Shepherd's Breeds of Turkey
The genes that control "white" colored dogs are still under exploration in molecular biology. Different breeds may tend to have different prevailing genetic codes for white hair color when it occurs in their breeds due to restrictions in their genepools and breed standards. Unfortunately, breed enthusiasts sometimes create breed standards without an understanding of the genetics for color in a breed. Color genetics can sometimes be in opposition to someone's "say so". Breed standards are heavily influenced by personal likes and dislikes, and sometimes misinformation about breed history or the health issues associated with certain pigment traits.
The landrace shepherd's breeds of Turkey have long been bred without a pedigree basis. Ephemeral pockets of isolation, intermixing and changing types across Turkey have existed for hundreds of years. In general, traditionally in Turkish agriculture, the sire is considered the important donator of genes - female lines are considered less important. Since multiple sire breedings are common with the female, because litters can be so mixed up, it does appear that her contribution to the puppies is less important than what the male gives! Keeping this concept of agricultural genetics in mind, it may be a little easier to see why things can be so inside out. The puppies in a traditional litter may be culled to the ones that resemble the favored sire. The various sizes, colors, types of puppies in the litter may be selected based on the mindset of the shepherd. Cross breeding to wolves was not that uncommon and these elements of breeding should be kept in mind when the "controversies" about Turkish dogs are in consideration. The problem is, everyone has their own ruler, but do not always base their ruler on the facts.
The Akbash is a recessive color variation of several morphologically varied regional strains of shepherd's dogs which carry a certain set of recessive genes that stop production of pigment in the hair. Since they are recessives that override the pigment producing genetics of the dogs, you end up with pups that generally resemble the parents in color.
The Kangal Dog likewise, is a nation-wide generalized blend of regional types and subtypes from all over Turkey which have the basic theme of being fawn, although there are brindle, pintos, "whites" and other pigment genetics in the phenotype. Essentially all of these dogs are a sable color (fawn) with various alleles that act on the base color. (think of the Anatolian Shepherd's Dog except with a color restriction when you think of the Kangal Dog, they are also a generic cobbling of many regional types, restricted to fawn due to Western influenced preferences)
The wide variety of styles in these dogs in Sivas alone is not consistent. From across Turkey, Kangal Dogs have been selected from the Ankara Zoo, from Izmir, Istanbul, Konya, Erzurum, Black Sea region among others -- although quite a bit of lip service has been given to the notion that they are a REGIONAL BREED. Rose tails (tightly curled), looser wheel tails, saber type tails, brush tails, plume tails, various foot shapes, textures of coat, degree of feathering on tail and hindquarters, and other variables all come to show that the dominant fawn(sable) color is really the only theme that matters. Here you can see that enthusiasts don't give much weight to geographical regions. A prized dog in this story (Arak) mentioned with a famous Turkish author, comes from Denizli. (more on other dogs and Arak can be seen an article here)
If you need a map to spot all those cities and regions, you can find one here.
A reaction to the increased mixing of various fawn genepools so that everyone "could be a Pepper too"... evolved a patent which claimed authenticity for the dogs from one region. An associate in Turkey who prefers the more mastiffy and loose skinned dogs, believes that this type represent the original Sivas dogs that he always knew. He is not alone. However, many of the other Sivas dogs are small and lighter boned compared to these other strains (some of these measurements are listed in the "patent").
Increased scrutiny by people with little knowledge of genetics, but whose passion for their opinions are strong, may have an effect in the establishment of more Turkish breeds. Possibly a new Turkish Mastiff breed. Turkish enthusiasts hope to have FCI affiliation of their own kennel club within the next few years.
The "white" conundrum...
I'll try to explain the basic actions of why Turkish dogs can be white, without going into too much genetic lingo (a challenge, since it will frustrate those who know the lingo), but a basic understanding of Mendellian recessives and dominants is needed.
There are only two types of pigment. Eumelanin is black or blackish(deep brown). Eumelanin is responsible for black and dark brown colors in the coat, on the nose, and eyerims, whiskers, etc. Phaeomelanin is the second pigment, which is yellowish to red.
Pure white hair represents absence of all pigment.
An illustration of a hair with visible pigment granules in the center part of the hair strand (medulla)
First, the white Turkish dogs are NOT genetically white with respect to the above two pigments. They are actually colored dogs which have inherited at least two (there may be more than two!) recessive functional detours in the development of their pigment.
I'll call these functional detours for now. The genes with which these are associated are the recessive extension gene (known symbolically as "e") and another pigment derailer that has functional similarities as the albino (or "c") and the "p" gene. The latter two genes or whatever other gene that functions to derail color development are still not completely understood.
The recessive forms of these functional detours to pigment development are epistatic (overriding, even though they are recessives) to other genes that want to make color in the coat. As long as those genes are present, the color cannot develop fully.
Not all the white dogs of the original foundation stock bred true (uniformly white litters of white puppies), which simply means that the pair of phenotypically white dogs which were bred, did not actually have the same functional detours (the same recessives) that changed their phenotype in pigmentation.
It is partly because of the fact that some imports did not breed "true", that some of the major early rifts (and 2) in the Akbash clubs started. The breed founders did not understand white genetics in their dogs but needed to 'stick to their story' so accusations were made that some dogs were not "pure" Akbash.
Historically, when small pockets of these morphologically variable shepherd's dogs lost some of their pigment diversity, the small clusters of white or cream dogs were believed by tourists to be a distinct 'breed' apart from the surrounding dogs, simply because they were "white". Simple as that. The fact that certain recessives will override the genetics of other colors wasn't as important as the mere fact that the dogs were "white".
Theories developed to somehow connect these isolated pockets of white dogs to other white LGD breeds simply because of the similar color. There was no logic or history behind the theory, but it seemed like a good idea since most people did not ask questions. Those that did, didn't get a reasoned response, but verbally attacked!
How scattered pockets of white dogs could be historically and genetically distinct from the rest of the multi-colored dogs in the same regions is never explained aside from more theories piled on other theories. When those multi-colored dogs were bred to "white", they produced some "white" puppies, which when bred to each other produced more "white". When the dogs are increasingly homozygous for the factors that derail pigment development in hair, that's only what is genetically expected.
Most people are not dog historians nor do they understand genetics. It's like a child's Sesame Street style game: which of these looks like the other. Voila, it's a breed!
These dogs however were not morphologically distinct from the colored shepherd's dogs in the same areas. They had the same wide ranges in structure as all the rest of the working shepherd dogs in the area. Rough coat, shorter coats, slender builds and some had more substance. Some had slender feet and others had large paws, with or without dewclaws. Some had distinct tail wheels, and others did not. Large muscular skulls and more refined skulls. Spotted and fully pigmented fawn working dogs were part of the region and modern enthusiasts treated them as thought they were unrelated.
To this date, there is still no definitive work that can show that these dogs have an association with other white LGD breeds but it's a popular notion. Researchers who quote studies regarding predation issues may often merely cite from the founder's comments, not bothering to do any research behind the dogs, thereby further promoting the confusion and heresy.
The use of genetic markers to provide an argument that white haired dogs are distinct from the colored ones in Turkey has been done. Small statistical samples with related dogs are a noteable a feature of such studies. As previously mentioned, no one actually knows which of the microsatellites (from DNA) tested are associated with the genes that code for pigment detours that lead to "white"-- yet these are the sort of tests that are being used to "prove" that the Akbash is distinct from the colored dogs by some enthusiasts.
I've communicated with a few scientists that are familiar with the use of microsatellites to associate groups of dogs together - for example, it has been shown that Belgian Tervs and Sheepdogs are actually the same breed by genome although American breeders have separated them as breeds anyway. When a puppy of the "wrong" color is born in a litter, it is taken to another country to get the proper Belgian paperwork so that the dog can be registered the way the Americans want the colors sorted.
In Turkey, when white puppies are born in the fawn litters, depending on the breeding center, they are sorted by color and each put into the appropriate breeding program selected for that color. TIGEM is the acronym for the Turkish Ministry of Agriculture (General Directorate for Agriculture Enterprises). A lot of pressure exists now to ensure that there is a wide genepool for the development of Turkey's new breeds.