Ever since sighthounds, greyhounds or Sloughis/Salukis were imported from their countries of origin in Asia and Africa into the western world, numerous myths about their origin have abounded. Clearly, except for the Greyhound and some smooth-coated dogs around the Mediterranean Sea, no other dog of the 420 or so known breeds (of which the AKC recognizes less than half) resembles these hounds. Their swiftness and stamina, their elegant build and their noble attitude all contribute to the fascination these breeds have generated among the dog fanciers in the western world. 
History shows sighthound-like dogs depicted in ancient images in the countries of origin that resemble in many ways sighthounds still living there today. Consequently, some fanciers assume that indeed Salukis are ancient and may have arisen some 2,000 or more years ago and may be the oldest breed of domesticated dogs. The images found in Egypt and Tepe Gawra suggest that such smooth and coated sighthounds might be even older and date back to the earliest civilizations known to modern man. 
        Be this as it may, the checkered history of many countries in North Africa and Southeast Asia raises certain doubts about the undisturbed continuity of the breeding of these hounds across such long expanses of time. In addition, the vast geographic area over which these sighthounds are spread, combined with limited accessibility in certain remote areas, raise issues of geographic isolation and limited genetic exchange which eventually could alter the genetic composition of various local populations. Nevertheless, based on certain features, in many cases not shared with more distantly related breeds, it seems reasonable to group all these hounds of the Orient and North Africa into a single group with other comparatively built dogs: the lop-eared sighthounds of many national and international dog registries of today (Afghan Hound, Azawakh, Saluki, Sloughi)
        Among the various sighthounds found throughout Africa and the Middle East, none has stirred more passion than the Sloughi/Saluki/Saluqi group of hounds.  In part, the debate revolves around the name, which describes in Arabic the local sighthound or windhound. Since in a given area only one of those hound types typically exists, there was no need to have more than the generic name that was equally used to describe the specific local variety. Of course, this linguistic confusion of generic and specific, combined with various transliterations of Arabic into European languages and the fact that both the colloquial “Sloughi” and the classical “Saluki” may be used to some extent interchangeably, have confused this matter even more. This confusion is well summarized in Waters’ book on the Saluki ("The Saluki in History, Art and Sport," by Hope and David Waters, David & Charles, 1969), and the reader is referred to this book for more detail.
In the modern era, the Sloughi of North Africa (henceforth simply referred to here as Sloughi) was first imported from the French North African colonies to France in the 1800s. These dogs were well known in France since 1850 through the books of General Daumas and also featured at international dog shows. Holland also was a major home of the Sloughi, brought back at the end of the 19th century from Africa by Auguste Le Gras.  Feathered sighthounds were portrayed on the European continent earlier than 1700, and were sporadically seen in the UK in the 18th century. It was around the turn of the 20th century that serious breeding started, following the first imports of “Slughi Shami”to England by Lady Florence Amherst (as she then named them) , and later on the feathered sighthounds of Brigadier Lance. These dogs provided the ancestors of the modern Saluki in the West.
        After the First World War ended and with the demise of the Ottoman Empire, both the French and the British had access to the Middle East and brought back many Salukis (named thus in England since 1923). Surprisingly, it appeared that the Saluki came in two varieties, a feathered variety and a smooth variety.  Logically, having a sighthound from the Middle East that had features in common with the smooth Sloughi of North Africa and the smooth sighthound-like dogs painted on the murals of ancient Egypt raised a number of questions about the relatedness of these dogs.
Different opinions developed about this subject over the past century and are summarized as follows:

1) Sloughis of North Africa and Salukis both belong to the group of oriental lop-eared sighthounds but are not more closely related to each other than the Saluki is with his eastern sighthound relatives, the Afghan Hound or the Tazi and Taigan of the steppe. The arguments in favor of this scenario assume the geographic isolation of the Sloughi in the vast expanse of Northern Africa and rely on the morphological differences, now also reflected to some extent in written standards of the breeds. It assumes that the Sloughi and Saluki evolved in their respective countries of origin and should be kept distinct. We will refer to this idea as the “Sloughi hypothesis”

2) In contrast to this idea is the assumption that the different geographic varieties are tightly interconnected through interbreeding and all Southwest Asian and African sighthounds form a single large “breed”. The underlying assumption here is that western breeds known as Afghan, Saluki and Sloughi are artifacts of the small extracted gene pool used to found those varieties now registered as distinct breeds in the West, and do not reflect the situation in the countries of origin which show continuity in both phenotype and the underlying genotype. We will refer to this hypothesis as the “Saluqi hypothesis.”

3) A variation of this latter idea is the assumption that among Oriental and African hounds three varieties are particularly related to each other: the smooth and feathered Salukis of the Middle East and the smooth-coated Sloughis and Azawakhs of Africa. This idea assumes that the African Sloughi and Azawakh were derived from the Saluki following the Arab invasion of North Africa in the 7th century.  We will refer to this hypothesis as the “Saluki hypothesis.” 

        Numerous arguments have been raised over the years by proponents and opponents of each of these three hypotheses. With the recent completion of a large analysis of the mitochondrial DNA of 654 dogs of some 124 known breeds, as well as other indigenous dog and wolf populations, the implications of these opinions can for the first time be tested. The samples processed included 8 Sloughis, 15 Salukis, 6 Taigans, 1 Tazi, 10 Basenjis, and 5 Afghan Hounds.

        The test consisted of sequencing an approximately 582-base-pair stretch (* SEE FOOTNOTE) of mitochondrial DNA. This genetic material is only inherited through the female lines, as male sperms lose their mitochondria upon entering the female egg. In other words, such a study provides insights exclusively into the female background of a given population, no matter what the male looked like and what his genetic material was. Analyzing the male genetic background of a population requires sequencing of the male-specific y-chromosome that is only inherited from male to male without contribution from the female. Such an analysis is on its way, but data are not yet available (P. Savolainen, personal communication).

We used the mitochondrial DNA data provided by Dr. Savolainen and looked at how they fit with the three ideas outlined above. We wanted to know:

1) Are Sloughis and Salukis genetically distinct or not?  Genetic distinctness would support the Sloughi hypothesis (Fig. 1), genetic similarity the Saluqi (Fig. 2) or Saluki hypothesis (Fig. 3).

2) Do Sloughi and Saluki share genetic material with other breeds that co-exist in the same geographical area and potentially crossed with them in the distant past?  Are these two breeds embedded genetically into their geographical origin while maintaining their unique looks.  Logically, the more genetic material is shared between these breeds and nearby totally different looking breeds, the less likely it is that consistent selective breeding took place over centuries.

3) Are there distinct genes in Salukis and Sloughis not shared by other breeds?  If the majority or all of the genetic sequences in a given breed are unique, it would indicate a long history of breeding from distinct females and would support the hypothesis that Sloughis and Salukis are distinct breeds. 

The Sloughi: 
        The six sequences found in the eight Sloughis are 50% unique to the Sloughi and not shared by any other breeds analyzed thus far (A4, A10, A33; 2 animals, 2 animals and 1 animal respectively).  Of the three sequences shared with other breeds, one is shared with another African breed, the Basenji (A5, 1 animal) but also with three breeds from Japan, Tibet and Siberia. Another sequence is shared with eight other breeds (A22, 1 animal):  Boxer, English Setter, Irish Wolfhound, Nguni and Sica (Africa), Scottish Deerhound, Saint Bernhard, and Tibetan Terrier.  The last sequence (B1, 1 animal) belongs to a different sequence group and is the only one shared with the Saluki, but also with 23 other known breeds, as well as a large group of dogs from China and Indonesia.  Among the breeds are the Dachshund, Doberman Pinscher, Finnish Spitz, Pekingese, Golden Retriever, Samoyed and Shiba (Fig. 4).  These sequences are provided by Sloughis that maternally originated, recently or in the past, in North Africa, and of which some represent several well established European and American lines.

        Dr. Savolainen and colleagues calculated that a single base substitution *(foot note) reflects a mutation rate of about 26,000+-8,000 years.  The three unique sequences found in Sloughis are one or more substitutions removed from other breeds. This suggests that the descendants of these three lines may have stayed in North Africa for several thousand years.  This suggestion is supported by the sequences shared with the Basenji, the Nguni and the Sica, African breeds that are now found only south of the Saharan desert.  At the same time the geographical range of the other breeds who share this sequence (A5) suggests that it might be very old and thus widespread across Eurasia and Africa. The two other shared sequences suggest different things.  The A 22 sequence could be related to the invasion of North Africa by Germanic tribes in 5th-6th century or could simply reflect the fairly widespread distribution across Europe, Africa and Asia of this sequence and might not be related to any specific historic event.  The B1 sequence shared with the Saluki does not necessarily reflect the invasion of North Africa by the Arabs and their hounds in the 7th century, but could be related to events long lost in the past, as it is shared with other widely distributed breeds of dogs. Unfortunately this study cannot tell us the following: was the maternal line a Sloughi or were the mitochondria contributed by the other breeds, perhaps now extinct?  Nor can we know the timing of these events.
         These results strongly support the idea of geographic isolation with some incorporation of three widely dispersed female sequences for the Sloughi.  The unique sequences strongly support the notion that the Sloughi is a genetically unique population of sighthounds and the sequences shared with the Basenji, Sica and Nguni indicate that this breed is, on the maternal side, embedded in Africa, possibly for thousands of years.

The Saluki:
        The eight sequences found in the 16 Salukis are 25% unique to the Saluki and not shared by other breeds analyzed thus far (A43, B15, 1 animal each).   Six sequences are shared with other breeds (Fig. 5). For example, A9 (1 animal) is shared only with the Basenji, A26 (2 animals) only with the Wirehaired Fox Terrier of Britain, and A15 (1 animal) is shared with the Korean Jindo and Akbash from Turkey.  The three sequences shared with numerous other breeds constitute about 37.5 %.  B2 (7 animals), is shared with the Afghan, the Akbash, the Canaan Dog, the Taigan and the Tazi from Kazakstan. B1 (2 animals) is the only one shared with the Sloughi, but it is also shared with 23 other known breeds as well as a large number of dogs from China and Indonesia. A11 (1 animal) is shared with over 35 breeds (Akbash, Akita, Basset Griffon, Borzoi, Canaan Dog, Chow-Chow, Fox Terrier, Siberian Husky, Jindo, Kangai, Pekingese, Rottweiler, Taigan, Thai Ridgeback, Whippet, and several dogs from China). All sequences were derived from Salukis out of the Middle East or bred in Europe and America.
         These results show that the genetic background of the Saluki is one of the most complex of all breeds in this study.  Clearly, the Saluki is not a distinct population derived from a single “Saluki-Eve”. Local Middle Eastern breeds such as Akbash and Canaan Dog appear several times among the Saluki’s related breeds.  Whether this is because of incorporation of Saluki-like dogs into these breeds or vice versa remains unclear. The Saluki shares sequences also with the Afghan, the Taigan, the Tazi and the Borzoi.  Since these sequences are also shared by other, non-sighthound breeds, it is difficult to understand the direction of breeding:  a sighthound female crossed with those other breeds or a female of those other breeds crossed with a sighthound male.  If such crosses happened, they may have been several thousand years ago, as the timeline of these events can not be resolved with this study. The Afghan and Taigan share a sequence unique to both, suggesting some, possibly ancient, relationship, in addition to sequences shared with several other breeds. 
         This study indicates that the Saluki is a local Middle Eastern breed which has been expanded genetically by incorporation of three minor and three major maternal lines shared with few or with many other breeds.  A single sequence is shared with the Sloughi, but is also shared with 23 other breeds and a large group of indigenous dogs from China and Indonesia.  .  Therefore, this single shared sequence does not prove an affinity of maternal lines specific to the Sloughi. This contrasts with the more profoundly shared maternal lines of the modern-day Saluki with local Middle Eastern dogs such as Akbash and Canaan Dogs, but also other Eastern sighthounds.

---

Saluki    ~    Sloughi

The Saluki and Sloughi compared.
        These results reveal that the Sloughi breed is not integrated in the Saluki’s gene pool, therefore cannot be derived from the Middle Eastern Hounds brought by the Arabs when they invaded North Africa some 1300 years ago. If this happened, it was of such a minor scale that it had virtually no impact on the maternal lines of the Sloughis of this study.  In fact, the three distinct maternal lines of the Sloughi suggest that those maternal lines might have been geographically isolated in Africa for several thousand years. If combined with the shared sequence with the Basenji and other African breeds, up to 67% of the female population of the Sloughi might have been in North Africa for a very long time. 
        The results for the Saluki indicate that the breed derives up to 62.5% from possibly local females, which, however, may be common to other local dogs such as Akbash and Canaan Dog.  The number of sequences shared by Saluki and Akbash is larger than those shared with any of the geographically nearby sighthounds, including the Sloughi.  These results do not support the idea that the Saluki of today has been “purebred” over the last several thousand years and thus might be the oldest purebred dog breed, except if we assume that such integration happened long ago.  The fact that non-sighthound breeds such as Akbash and Canaan dog share more maternal lines with Salukis than Salukis share with other sighthounds does not provide support for the Saluqi hypothesis. 
To conclude, these results support the idea that the Sloughi and Saluki of today are distinct breeds with genetic ties to each other that are no more profound than those of either to many other breeds.
         We hope that this essay will provide for a more rational basis of future discussions.  We deliberately refrained from quoting sources for the various ideas in order to allow everyone the opportunity to reconsider those ideas in the light of these genetic results.  It needs to be stressed, however, that these conclusions are based on a limited sample size and might undergo some revision once more samples are included.  Nevertheless, the genetic data presented here for the first time support the notion previously espoused by F. Amherst and others that these sighthound breeds are distinct. We propose that these breeds should remain as separate breeds until full resolution of their genetic background is completed. A more complete resolution of all the questions raised here will have to wait until the paternal side is known through analysis of the male Y chromosome.  Finally, correlating the genetic differences described here to differences in phenotype will only be possible after the dog genome * (footnote) is completed and the relationship of various genes to phenotype have been clarified, research that will require many more years. 

Acknowledgments
We thank Dr. P. Savolainen for improving an earlier version of this manuscript.

References:
Amherst F. 1907. Oriental Greyhounds. In Cassell's New Book of the Dog, ed. R Leighton, pp. Ch LVI. London: Cassell & Co
Savolainen P, Zhang YP, Luo J, Lundeberg J, Leitner T. 2002. Genetic evidence for an East Asian origin of domestic dogs. Science 298: 1610-3
Waters H, Waters D. 1969. The Saluki in History, Art and Sport. Newton Abbot: David & Charles. 112 pp.

Base substitution.
        Altering the information encoded in the orderly arrangement of bases on the DNA requires altering the sequence of bases.  Such an alteration is called a mutation.  Such mutations can do a number of things.  In many cases, mutations change a single base.  Thus, instead of an A it will substitute that for a G, for example.  Such base alteration rates typically range in the order of several hundreds or thousands of years, depending on the sequence in question. 

Dog genome.
       The genome is all the inherited information or DNA carried from one generation to the next in a given being.  The genome of four mammals has thus far been sequenced, humans, mice, rats and dogs.  A comparison revealed that dogs and humans share approximately 90% of their genes.