Saturday, 22 June 2013

Tarpan-like horse breeds

In this post, we looked at the phenotypic and behavioural traits of the Tarpan, Equus ferus ferus, the western wild horse subspecies. Today I’m going to introduce a number of horse breeds which apparently share some characteristics with the Tarpan. But before we go into this, it’s important to know where the wild horse was domesticated and how it left its traces in the modern equine stock.

Equus ferus ferus seemingly is the wild type of the modern domestic horse and there are no hints yet that the Przewalski horse played a significant role in horse domestication. The Tarpan was domesticated in the western Eurasian steppe approximately 6000 years ago, as it is indicated by genetic [1] and archaeological [2] evidence. Interestingly, the modern domestic horse has a considerably diverse mitochondrial DNA while the Y chromosome diversity is very low, what suggests that only few wild stallions but a larger number of wild mares was used in domestication [3, 4, 5]. Studies like Jansen et al. and others suggest that there was local wild mare introgression within Europe. Ancient breeders probably included wild mares into their stock to continue breeding on a larger scale, but avoided working with wild stallions because of their intractable behaviour [3, 4, 5].
What does this genetic data tell us? Apparently most if not all horse breeds originated in the Eurasian steppe, but local female influence was seemingly common as well. The very low Y diversity of modern horses reveals another important implication: if any horse breed is a direct descendant of a European wild horse population, we would expect a significantly higher Y chromosome diversity in these breeds, with haplogroups completely unknown in other horses. However, most studies on horse genetics are concerned with mDNA, and Lindgren 2004 used a set of 15 horse breeds that might not be representative enough. 6 of the included European breeds belong to the northern pony type which possibly all are related and not “divergent” as the study says (Exmoor Pony among them), and robust horse breeds from eastern Europe such as the Konik or Hucule are missing. So I think the hypothesis that no horse breeds are the result of comparably recent and significant wild horse introgression requires more testing (although we can tentatively rule it out for the breeds tested in Lindgren 2004).

Considering that the Tarpan is the ancestor of all domestic horses, when we see traits in horses that resemble what we know about the Tarpan (be it behaviourial or phenotypic) we can carefully assume that these features are inherited by wild Tarpans. I want to cover wild behaviour in horses in a following, separate post, so we’ll focus on the phenotype, ecologic adaptions and breeding history of the primitive horse breeds here.
Since the Tarpan displayed a number of colour morphs, the horses that we can consider primitive/less-derived/Tarpan-like (however you want to call it) don’t have to look very much alike, there is actually a diverse set of good “candidates”. Let’s start from west to east.

On Great Britain you can find a lot of robust pony breeds with primitive colour and body features, of which the most unchanged one is the Exmoor Pony. They seemingly all descend from a feral pony type that once ranged on the entire island [6]. The primitive British Pony/Exmoor Pony is a comparably small-bodied (130 cm at the shoulders), stocky horse with a robust skull and is of a bay/seal brown coat colour with a light mealy mouth and a light underbelly. These horses have a dark eel stripe, but it is (like in most non-dun horses) not easy to recognize. The Exmoor Pony and its relatives are known to be very robust and hardy. They perfectly cope with the harsh british winters, while more derived breeds from the mainland do not nearly as good [6]. Some primitive British ponies have a black colour, like the Fell Pony (descending from the extinct Galloway Pony). 

Exmoor Ponies
Fell Pony
Quite possibly, British ponies also influenced other small northern horses, like the Gotland or the Fjord horse (the latter is bay dun-coloured, by the way) [5].

Fjord horses
Primitive Asturcons, there are black ones as well

Some Iberian ponies, like the Garrano, Pottoka, Asturcon and others, resemble the Exmoor pony-type horses closely. In some of these, like the Pottoka, this resemblance might be caused by British introgression during the 20th century [7], but others like the Asturcon have a longer history. Those primitive Iberian Ponies either are bay or black, and some of them (f.e. Garrano), have feral populations.
Except for the lack of the dun colour, this western pony type resembles the descriptions of wild horses quite good (stocky and small body, large robust head, small eyes, short mane, wild colour present etc.). However, as I discussed in the previous post, the presence of wild horses lacking the dun factor cannot be ruled out yet.

The area of the former Austrian monarchy also has a set of primitive and hardy breeds with a diverse ancestry. The Noriker horse is a large (but there also was a smaller version of that breed called the Abtenauer) and robust working horse from Austria that was influenced by Italian and Spanish horses in the past. It shows (besides a number of domestic colours) bay, black and leopard spotted individuals. They have a robust head with small eyes and are tolerant against cold temperatures. 

Noriker horses with wild horse colours and other primitive features
The robust Hucul horse, from the eastern Carpathians, is a breed that likely descends from horses coming from the steppe. It’s size range is within that of the wild horse, and it displays a number of colours including bay dun, black dun, bay and black. The dun individuals have often have a prominent eel stripe and leg and shoulder striping. It is more gracile than the primitive western ponies.

Tarpan-like Hucul horses
The Konik is similar to the Hucule, but certainly has the most romantic story regarding its origin. Allegedly it descends from wild horses that were donated to polish farmers who tamed and crossbred them, and after that, a breeding-back project allegedly purged out the domestic influence, resulting in the modern Konik. The reality behind this story is a lot less romantic, but we will look into this in a future post. The Konik is a robust polish pony which’s colour is predominantly black dun. It’s phenotype varies from a more gracile riding horse type with a gracile head and a long mane that does not truly resemble the wild horse in proportions, and a robust type with wild horse-like proportions and skulls. It also has different expressions of black dun, ranging from a very light gray (not likely for wild horses), to a grayish brown resembling a mouse’s coat (very likely for wild horses).
In Germany, there are two popular derivations of the Konik. One is the Heck horse which is the result a breeding-back attempt by the Heck brothers and has influence of other ponies like the Gotland and Icelandic horse, but also the Przewalski horse. It is not necessarily more Tarpan-like than usual Koniks. The Dülmen Pony has a long history but the modern population is a mix of Exmoor, Shetland and Konik (with the Konik having the largest influence in this breed) and displays a diverse phenotype.

Koniks with very wild horse-like coat colours (Photo: Volker Kirchberg)
Moving more eastwards, there is the Yakutian horse of the Sakha Republic, Russia. It is comparably large (150 cm), but has a primitive stocky body and is well-adapted to extreme coldness. Its coat colours (in pure examples of the breed) can be bay or various dun colours, with or without pangare. Selective breeding could breed out a Yakutian lineage showing wild horse colour traits exclusively.

Yakutian horses
Another primitive Asiatic horse is the Mongolian horse. It is a robust and stocky breed, reaching 130 to 145 cm in height. Among a number of domestic colours, it contains many of wild horse coat colours, such as bay, bay dun, black and leopard spotted (I’m not sure if there are black dun Mongolian horses). Of the Asiatic horse breeds, it has the greatest genetic diversity [8], indicating that it might be an ancient lineage. And their location isn’t far away from the domestication centre of all horses.

Monoglian horses
Surprisingly, there are primitive horse breeds even outside the old world. Mustangs have a diverse and controversial ancestry, some of them display a rather wild horse-like phenotype (despite their comparably gracile build) and they have experience in surviving in nature over generations. One of the most interesting Mustang lineages is the Pryor Montain Mustang, a population that is known living feral for at least 200 years and descended mainly from the Spanish Barb and other lightly built horses. They range from 130 to 150 cm in height and some of them show a bay, bay dun, black or black dun colouring, including the wild markings. Some Sulphur Mustangs also show wild horse colour traits and a primitive body conformation. Surely even ancient-looking Mustangs have more gracile proportions than the primitive western pony breeds, but this may be due to adaption living in a more open habitat. However, the hypothesis that Tarpans with that proportions existed should be proven by osteometric data, since even the Przewalski horse (native to a steppe habitat) has a stockier build than mustangs do.

Pryor Mountain Mustangs with Tarpan-like coat colours

The Sorraia is a breed from the south of Iberia, and fanciers claim it is a surviving descendant of the Tarpan because of its black dun coat and the fact that the breed as such was not known before the 20th century. However, it is likely that the breed has an Arabic ancestry because of its tall and gracile build and the strongly convex head[7]. Some Mustang lineages, like Sulphur and Kiger, apparently share a common ancestor with the Sorraia, as there is a large resemblance between individuals of these breeds. Older genetic studies noted similarities of the Sorraia with the Konik or Przewalski horse, while more recent studies refuted this [5, 10]. Furthermore, the Sorraia is related to the Lusitano [9]. All in all, they do not differ from other Southern Iberian horse breeds [7]. Apart from that, the breed went through a severe bottleneck during the 1930s, the modern stock descended from only 12 individuals. Sorraia usually display a kind of weird body conformation with a straight long neck and thin legs, unlike primitive horses and what is described for the European wild horse. While fanciers hail the Sorraia as a surviving wild horse, some resemblance in colour is simply not enough for such a hypothesis (no matter how prominent the striping might be), especially since wild horse colour traits are apparently widespread among living horses.

As you see, there is a number of horses with a Tarpan-like phenotype, while the genetic distance between them and the wild horse is probably roughly the same. However, this aspect needs more testing. If some of the primitive horse breeds turn out having a bigger Y chromosome diversity with haplogroups completely unknown in other breeds, we might assume that there was more introgression than in other domestic lineages.
In future posts, we’ll have a look at which of these horse breeds might be best for being released into nature as an authentic proxy for the Tarpan. 


  • [1] Warmuth et al.: Reconstructing the origin and spread of horse domestication in the Eurasian steppe. 2012
  • [2] Outram, A.K., Stear, N.A., Bendrey, R., Olsen, S., Kasparov, A., Zaibert, V., Thorpe, N. and Evershed, R.P. The Earliest Horse Harnessing and Milking. 2009
  • [3] Lindgren et al.: Limited number of patrilines in horse domestication. 2004
  • [4] Jansen et al.: Mitochondrial DNA and the origins of the domestic horse. 2002
  • [5] Cieslak et al. 2010: Origin and History of Mitochondrial DNA lineages in domestic horses
  • [6] Baker, Sue, 2008: Exmoor Ponies: Survival of the Fittest – A natural history.
  • [7] Royo et al.: The Origins of Iberian horses Assessed via Mitochondrial DNA. 2005.
  • [8] Udina et al.: Computer Analysis of D-loop of Mitochondrial DNA variation in Asian horse breeds. 2002.
  • [9] Luis et al. A lost Sorraia maternal lineage found in the Lusitano horse breed. 2006
  • [10] Aberle, Diestl: Domestication of the horse: results based on microsatellite and mitochondrial DNA markers. 2004


  1. Nice post!!!!! What do you think about Retuerta horse that is claimed to be the most ancient breed in Europe?? Keep it up! Your blog is becoming a wealth for informations!!

    1. Thanks! Well, "ancient" is an ambiguous term. It either means that a breed is old (what does not necessarily imply it resembles its wild ancestor) or that it is "primitive". But for the Retuerta, I only know of a study from 2006 (can't recall the citation, sorry) that it has little influence from other Iberian breeds. That doesn't necessarily mean it is more primitive than other horses, and their phenotype doesn't resemble the wild horse that closely as well. That's why I didn't include it in this selection here.

  2. Great photos. I really want to learn more about horse breeding.


    1. The Sorraia is not a particularly interesting breed to me. Only their coat resembles the wild horse (and only one colour morph), and everything else is lacking evidence, also concerning their breed history.

  4. + Alter Real (?).



    1. Thank you for your comment. But: I don't know if all the "Anonymous" here are the same person, but I don't really appreciate comments that merely contain a link and no text at all, especially when it leads to websites repeating the same anecdotical stories that are repeated over and over. I am sorry.

  7. Two breeds of horses with unusual colors:
    British Spotted Pony
    Tiger Horse

  8. The Tibetan Pony:
    These two Tibetan breeds of horses also appear to be Tarpan-like:
    Nangchen Horse
    Riwoche Horse

  9. Pachyornis have you read the paper of Michael Cieslak, 2011 on coat colour research in which he stresses the importance of finding the gene that causes "Dun" ?

    In regard of the sorraia (I can call myself a sorraia-fancier (I'm not the mr. anonymous here above though) but I like to bias between liking the sorraia as a "horseman" as a pretty elegant horse, and the fact that it is, or is not, going back straight to predomestic horses, and that more research could shed a different light on this strain of horse), so in regard of the sorraia I'd like to mention that biologists/ecologist/other-concerned-scientists for horses seem to accept only one conformation type (the stocky poney type) in contrast to what they accept for other species. Allthough there is the theory of Ebhardt on 4 ancient types, a more heavy built draft like type, a compact poneytype, a small refined arabian type and a larger refined steppetype. The first two are more located in colder, wetter, shaded ecotopes, the latter two are more located to hotter, drier opener ecotopes (based on volume/surface energyconsumption). So in that light you could see a horse like the sorraia as produce of different surrounding than were you situate the english and other ponies.
    Cieslak points out two predomestic different populations, one on the eurasian steppe, the other on the iberian peninsula as glacial refuge (and probably a smaller third one in east-asia).

    You compare the sorraia to the arabian based on it's slender type and on its convex head, while the typical head of an arabian is quite the opposite, really concave (allthough in the last decades it is really overemphasized and is becoming quite caricatural). I think the sorraia has more the type of head you find in iberians (lusitanos and pre, but mind the mix with arabians here), and of course they are related, but also the type of head you find in the Akhal Teke, a real desert/steppe horse in turkmenistan. I know arabians, lusitanos, pres, akhalteke are strictly domestication nominations, but should we reduce all physionomic differences to human selection ?

    apart from that i have some thoughts on the colors pangare-bay of the exmoor and the grulla of konik/sorria/tarpan too. Pruvost, Cieslak et Al. searched for the allels E and e on the extension locus coding for black and non black, and for the allels A and a coding for bay and non bay. But there are 2 other allels at play on the agouti locus and that is A+(wildbay) and At (Sootybay), for the latter there is a genetic test now, but not yet for A+
    If exmoors would be tested, they would rather rest for At Ii think, since that gene is causing a shading effect on bay horses from the top down, leaving ampits, belly, loins and mouth rather "mealy".
    I think the gene for dun is rather to be found related to agouti than to other genes that code for dilutions of black in other species, while it was on those genes (fi. lavender in chickens) that Cieslak did his research but found nothing for dun there.

    I would like to show some images but i think I better send those privately, is that possible ?

  10. Hi, I plan to cover the Sorraia in a future post soon, I somehow never got to it. Ebhardt's hypothesis has no fossil support and therefore is rather baseless. I do not doubt at all that different environments do have an influence on the body conformation of horses, perhaps we can use living wild equine as an example. Nevertheless, there is no fossil evidence from the Holocene for a tall, slender predomestic horse that would resemble the Sorraia in conformation. I know Cieslak's paper and I am aware of the genetic separation of the Iberian wild horse population and the rest of the Equus ferus ferus subspecies, but that alone is no indication of a gross phenotypic change (many other large mammal species were isolated on Iberia too, including aurochs and wolves, for example). After all, Bergmann's rule (which is not as universal as it is often stated) would indicate that southern wild horses would have been actually smaller than northern ones.
    You can send me those images to daniel-foidl( )itmed-at ("-" is for ".", bracket for "@").

  11. Thanks for your comment.
    I am very impressed for this blog.

    Arabian horses

  12. The mtDNA patterns found in Sorraias can also be found in some other horse breeds. This is only natural, as our domestic horse breeds are usually a mix of populations, types, and breeds, a mixture of genotypes is therefore usually found in horse breeds, and the genotype of the Sorraia, or Tarpan, in horse breeds is thus easily explained.
    The mtDNA results are consistent with the presumed role of the Sorraia as one of the ancestors of those breeds. All 18 Sorraias sampled for this study had one of two mtDNA patterns (A1 and JSO41), both on the same branch in the phylogenetic network, and closely related to one of the two major mtDNA types found in Lusitanos and Andalusians (A3). The genotype found in Sorraias is completely unrelated to the other genotype found in Iberian horses (D1). This is remarkable regarding the status of the Sorraia.

  13. That’s so great, amazing stuff of high quality; I appreciate you on your working.

  14. This article gives the light in which we can observe the reality. This is very nice one and gives indepth information. Thanks for this nice article. Horse Barn

  15. What do you think about Mülen pony? It seems striking that nobody talks about him, despite the interesting story that he has (...and its morphology, its natural behabior...). of course I'm not trying to say that it may be the last European wild horse (e know everything is mixed with domestic horses).