Thursday, 19 October 2017

Wild horse series Pt.II: The Konik, Exmoor and Sorraia myths

In my 2017 wild horse summary I had a look at what the evidence suggests regarding the life appearance, extinction and population genetics of European wild horses. There are some open questions, but also some things that we can say with certainty. One of those things is that European wild horses are as extinct as the aurochs. Nevertheless, there are three horse breeds that are often referred to either as surviving wild horses or near-wild horses, having a special status among European horses in being particularly close to the original wild form. Those three breeds are the Polish Konik, the English Exmoor pony and the Portuguese Sorraia, and each one of those has its own proposed background story that is supposed to link them closely to European wild horses. Those stories, however, are only purported and believed by advocates of those breeds, people involved in various projects, and are spread exclusively in non-scientific text books, public relations articles, or on signs in zoos while nothing of them is replicated in scientific literature, and none of those breeds are considered zoologically special among domestic horses. Advocates of the breeds blame a “conservative scientific mainstream” that is biased against the idea that the European wild horse might still be extant in some form, while the truth is that the background stories for the purported special status of the Konik, Exmoor or Sorraia are largely based on wishful interpretations, Chinese whispers and fabrications, actually making those stories myths that do not withstand objective examination.
Here I want to present and subsequently dismantle those myths. When not relying exclusively on sources that repeat the same stories over and over, but actually provide a deeper background knowledge, backed up by more evidence than usually delivered, it becomes rather obvious that the usually purported background stories for those breeds are not tenable. Yet those stories are pretty persistent and widespread, and I hope this post might become a little contribution to clear things up. 

The Konik myth: The Polish game park at Zamosc was the place were Europe’s wild horses survived longest. In the year 1806, these wild horses were donated to local farmers of the Bilgoraj region and incorporated to their farm horse stock. In the 1920s, the Polish agriculturist Tadeusz Vetulani started a breeding-back project using wild horse-like individuals from the Bilgoraj region and bred them selectively for a wild horse-like nature, eliminating the domestic influence their genome experienced in the hundred years before. He called the result “Konik”.
The truth behind it: Whether or not the supposed wild horses at Zamosc were truly wild and not feral, it is unlikely that they left a noticeable trace in the local farm horses. This claim is based on a notion by Julius Brincken in a book from 1826, but this book is full of errors, misinterpretations and fabrications, so it is not a reliable source. Furthermore, the Zamoski family was at war with the local farmers due to social unrests in the 1780s and 1790s, so it is unlikely that they would have provided them with generous gifts in the form of livestock. And even if they did, it is unlikely that the farmers tolerated a strong wild influence in their horses as it would have been a throwback in their productivity and suitableness for agricultural work, as wild horse hybrids used to show intractable behaviour [2]. And even if they incorporated wild horses, and perhaps only mares, into their stock, not much would have been left after 100 years. So it is very unlikely that the horses of those particular rural regions of Poland had a notable wild horse influence in the 20th century [1,2].
In Poland, there was/is a landrace called Panje horses. Those horses were very robust, of a small and stocky body and varying in colour between black dun, black, sorrel and bay. In an 1921 article, Panje horses were considered as possible wild horse descendants (which is to be refuted for the reasons stated above) by the researchers Gabrowski and Schuch. This drew the attention of Polish agriculuturist Tadeusz Vetulani to the Panje horse as “wild horse relict”, he tried to back up this suspicion with cranial measurements (that have been questioned [1]), and coined the name “Konik”, which successively replaced the name Panje horse. Private and public studs were created in order to preserve and spread the landrace. In 1927, Vetulani started an experiment that is often considered a “breeding-back” experiment, but was actually more of a dedomestication attempt, using Konik/Panje horses that he considered wild horse-like. But Vetulani’s herd was only one of many Konik lineages back then, and the Second World War created a lot of confusion, as many Koniks were moved all over Poland and Germany during and after the war. After the war, the stud at Popielno was the most important Konik breeding site. They performed two separate ways of breeding: one to continue Vetulani’s way of few human interference, and one of traditional indoor breeding and commercial sales. Popielno was one of many breeding sites, and when other countries started to become interested in this breed, they purchased from any stud available [2].
Thus, neither was Vetulani’s experiement an experiment of selective breeding, nor was it the ancestral stock of the modern Konik population but merely one of many lineages [2]. He did not create the breed, but merely coined the modern name of the breed. The fact that 10-5% of the modern Konik still show a black or sorrel colour or white streaks on the forehead (respectively) reveal the mixed origin of this landrace [2], and there is no compelling, not even plausible, evidence that their ancestral stock was strongly influenced by wild horses [1,2].
The Heck horse, which is often presented as a “recreated Tarpan” in Germany and also other countries, is in a sense a washy Konik – it was created by the Heck brothers in the 1930s and 1940s by crossing Icelandic horses, Gotland ponies and Dülmen ponies with a Przewalski stallion. Later on, Koniks were used massively as breeding stallions on Heck horses, so that they are heavily influenced by that breed, hence the large optic resemblance [8]. The same was the case in the Dülmen pony, which is why the three breeds sometimes are regarded as the Konik group [8]. The Liebenthaler horse can also be regarded as a member of this group.

The Exmoor Pony myth: The Exmoor Pony is a remnant population of a wild horse type or at least a feral type of very original western European horses that once ranged on the entire British island. Several other populations of this primitive British horse type have been intermixed with derived horse breeds, creating the modern British pony breeds such as the Welsh Pony, New Forest Pony, Dartmoor Pony and others. The Exmoor Pony however is the only population that retains a stable, wild horse-like appearance with a brown colour, countershading + white muzzle and a sturdy body. The similarity to other northern ponies such as the Gotland or Faroe Pony, and especially some primitive Iberian breeds such as the Garrano and Pottoka, endorse the hypothesis of a north-western European wild horse/primitive horse type that once was found in this region and is most authentically represented in the form of the Exmoor pony.
The truth behind it: At first, this scenario sounded convincing to me. But what is most important to note first is that equines disappeared from the archaezoological record of the British isle at the end of the Mesolithic until domestic horses were introduced by the Celts [3,4,5]. Therefore, the Exmoor pony cannot be a remnant wild horse population or wild horses with domestic introgression, they have to be of domestic origin. But this alone does not rule out that they are the last representative of a homogeneous, feral and primitive horse population that once ranged across the entire British isle. One argument is their homogeneously small body, comparably short mane, the brown colour with countershading plus white muzzle. However, careful examination suggests that the population at the Exmoor never was homogeneous after all[4]. There are no helpful references prior to the 18th century on the colour variants found in the Exmoor population. Between 1805 and 1809, 81 Exmoor ponies were sold from the moor. Their colour was documented and included black, grey, bay, dun, chestnut and one piebald individual. There is no evidence that purposefully “non-pure” Exmoor ponies were caught in this case. Two illustrations in the Illustrated London News from 1835 clearly show horses with a long mane, their colours are probably implemented to be brown with a white blaze and sock, grey and a black one [4]. A notion by Worthley Axe in the year 1906 is even more revealing: “… the majority of the so-called Exmoors are simply mongrels” [4]. The Acland herd, which made a considerable contribution to the modern Exmoor population, also included a number of greys and blacks in 1900. There is a record that suggests that black Exmoor ponies were selected out because they lacked the expression of the white muzzle, indicating that artificial selection started back then. Furthermore, the Exmoor pony population went through several bottlenecks, the most severe in course of the Second World War. A stud book for the breed was set up in 1921, at first black and grey individuals were tolerated but selected out later. It is thus far more in line with the evidence that artificial selection and genetic bottlenecks created the homogeneous external appearance we see in the modern Exmoor pony, and there is no evidence that it was homogeneous prior to 1906. Most likely the Exmoor pony that is always brown in colour with countershading, a white muzzle and no white markings is an invention of the 20th century [2]. It would therefore also be more parsimonious to assume that other British pony landraces like the Welsh pony, Dartmoor pony or New Forest pony never were homogeneous either. Furthermore, the use of British ponies including the Exmoor pony on Iberian breeds like the Pottoka in the 20th century is well-documented[2], which is at least partially responsible for the optical resemblance. Thus there is no empiric basis for a once feral, primitive free-ranging horse type that looked like the Exmoor and ranged across Great Britain or even whole western Europe.
Also, which is important to note, Exmoor ponies most often display a colour variant that is called seal brown, At. This is an allele that has not been identified in wild horses yet, in contrast to the two wildtype alleles bay A+ and black a [6,7], and therefore most likely is a domestic colour. The black individuals, which have been actively purged from the population, however, would have displayed a wildtype colour variant [6,7].

The Sorraia myth: The Portuguese agriculturist Ruy d’Andrade spotted a herd of strongly striped, free-ranging horses in a remote region in Portugal. He considered them to be a remnant population of the zebro, a strongly striped Iberian wild horse type. He was unable to find them again, but collected a number of farm horses from that region that he considered to be closest to those horses, and started to breed them. This is the modern Sorraia horse, by some even communicated to be identical to the zebro and thus a wild horse.
The truth behind it: The claim that the Sorraia is a surviving Iberian wild horse is flawed by the story itself – d’Andrade did not catch any wild horses and started breeding them, but merely collected farm horses he considered to be reminiscent of the horses he spotted. He collected four local stallions and seven mares, and a Criollo stallion was added later on. Therefore, the Sorraia is a descendant of domestic farm horses. But is it possible that is particularly strongly influenced by an optically identical Iberian type of strongly striped wild horses, the zebro? First of all, there are not any references suggesting a survival of wild horses on Iberia into the beginning of the 20th century. C.H. Smith reports free-ranging horses of black dun and bay dun colour with wild markings that ranged from the Camargue, parts of Spain to the Ardennes, Great Britain and Scandinavia (see here), but also describes them as “sturdy mountain-forest ponies”, and it is not clear if those are wild or feral horses anyway. Regarding the actual zebro, it seems that this population of equines vanished in the 16th century [8] and many authors tend to consider them feral donkeys instead of horses. Indeed a genetic test of the skeleton of the supposedly last zebro turned out to be a donkey [9]. The strongly striped ashy grey colour of wildtype coloured donkeys is very reminiscent of the black dun colour scheme in horses.
The allegedly strong stripe pattern of the Sorraia is often purported as its trademark and an indicator of particular primitiveness that links it to the zebro. However, stripes are part of the bay dun and black dun colour scheme and found in many horse breeds. Therefore, it does not give the Sorraia any special status (see a very strongly striped Hucule here, for example).
Ferus-type wild horses are usually assumed to have been comparably small and sturdy in build. The Sorraia, however, is comparably large (140-150cm at the withers) and lanky. As there is no comprehensive archaezoological record of predomestic wild horse skeletons that has been osteometrically examined (only some single fragmentary specimen) it can neither be proven nor ruled out that Iberian wild horses were larger and more gracile in build than other Eurasian wild horse types, but it would be necessary to have evidence at hand for making such a claim.
More importantly, genetic information has revealed the Sorraia as a domestic horse [10].

I am going to come back to genetic studies more extensively in the next part of the wild horse series 2017. A closer look at the respective history of those three breeds that are often supposed to be wild horse relicts, near-wild horses or recreated wild horses alone is sufficient to show that there is actually nothing of substance that suggests that these three breeds deserve a special status among robust landraces of European domestic horses. But this of course provokes the question which horse breeds are best suited to be used as a substitute for the European wild horse in ecologic restoration. Thus, the upcoming post is on this question as much as on the domestication of the horse in general.

Similar but older posts: 


[1] Cis van Vuure: On the origin of the Polish Konik and its relation to Dutch nature management. 2014.
[2] Tadeusz Jezierski, Zbigniew Jaworski: Das Polnische Konik. 2008.
[3] Baker, Sue, 2008: Exmoor Ponies: Survival of the Fittest – A natural history.
[4] Peter Green, 2013: The free-living ponies within the Exmoor National Park: their status, welfare and future. A report to the Exmoor moorland landscape partnership. 
[5] Bunzel-Drüke, Finck, Kämmer, Luick, Reisinger, Riecken, Riedl, Scharf & Zimball: „Wilde Weiden: Praxisleitfaden für Ganzjahresbeweidung in Naturschutz und Landschaftsentwicklung“. 2010
[6] Pruvost et al.: Genotypes of predomestic horses match phenotypes painted in paleolithic works of cave art. 2011
[8] Bunzel-Drüke, Finck, Kämmer, Luick, Reisinger, Riecken, Riedl, Scharf & Zimball: „Wilde Weiden: Praxisleitfaden für Ganzjahresbeweidung in Naturschutz und Landschaftsentwicklung“. 2010
[9] L. Orlando et al.: Revising the recent evolutionary history of Equids using ancient DNA. 2009.
[10]  Lira et al.: Ancient DNA reveals traces of Iberian Neolithic and Bronze Age lineages in modern Iberian horses. 2009

Friday, 13 October 2017

Some more photos of Watussi x Maremmana

Claus Kropp from the Auerrind project recently published a post on the use of Watussi in the project on the Auerrind blog, including a recent photo of Thando and his half-Maremmana offspring. 
Thando (©
Apollo (©
Ambra (©
I think that Apollo might approach the age of reaching its final colour, but it is obvious that he is darker in colour than its halfblood sister. This is likely mostly due to its Maremmana ancestry, but sometimes also Watussi cattle show a tendency of sexual dichromatism. The zebuine hump seems to be either lacking or very weakly developed yet, but this is irrelevant as those individuals are F1. Surely the hump might or might not reappear in future cross individuals, depending on coincidence. One always has to pick the right individuals once the genes become split up and distributed among the offspring. Regarding the colour, I think this combination bears the potential for very dark or even black bulls (probably with a colour saddle) in F2 and subsequent combinations. Watussi might pass on alleles for very dark bulls, but is unable to express it since their coat colour seems to lack the black pigment eumelanin. Maremmana enables that, but carries dilution alleles that remove red pigment, pheomelanin. Watussi, however, contributes the wildtype alleles that enable the expression of pheomelanin. Thanks to the 2. Mendelian rule, it is possible that true F2 individuals might end up showing the right wildtype aurochs colour setting. 

I am confident that it is possible to boost the horn size of the Auerrind crossbreeds using Watussi without adopting the negative traits of the breed using careful selection and also a bit of luck. It worked successful with Heck cattle of Neandertal descent, and in Hortobagy there are a couple of qualitatively aurochs-like Taurus cattle with Watussi influence. 
Thus, I am looking forward to watching those individuals grow up and to see their respective offspring with great interest. 

Monday, 9 October 2017

Drawing: Central Europe's Holocene megaherbivores

I recently did another drawing showing Holocene Europe’s megaherbivores of the nemoral mixed-forest zonobiome, drawn to same scale. It shows, from left to right: fallow deer, Dama dama; wild boar, Sus scrofa; red deer, Cervus elaphus; elk, Alces alces; roe deer, Capreolus capreolus; aurochs, Bos primigenius primigenius; wisent, Bison bonasus; European wild horse, Equus ferus ferus.

I know of course that wild boar are actually omnivores. One might ask why I included fallow deer, as it was introduced in historical times just like the mufflon. In contrast to the mufflon it has a Pleistocene interglacial record in Central Europe, which is why I included it here. The question is whether Dama dama would have recolonized Europe without human influence. In order to answer the inevitable question “but where are elephants and rhinoceroses?”, my drawing intends to illustrate only those herbivores which had a solid presence in Holocene Europe. It is not known for certain that the European continental elephant species Palaeoloxodon antiquus as much as the two rhino species of the genus Stephanorhinus have been driven to extinction by men and would still be extant without anthropogenic influence today. It is plausible, I even consider it likely, but not proven. One might also ask where water buffalo and Equus hydruntinus, the European wild ass, are. There is only one Holocene remain for the genus Bubalus in Europe (Austria, to be precise) which has been tentatively assigned to this genus by Erich Pucher in 1993, nowadays he would reclassify it as Bos (personal communication). Bubalus murrensis might have been hunted to extinction (although its record is really rare and I do not know of literature mentioning its association with kill sites), but here goes the same as for elephants and rhinos. Europe being recolonized by extant members of Bubalus without anthropogenic influence is speculation. Equus hydruntinus has a solid Holocene evidence in Neusiedl, Austria (Pucher, pers. com.), but as part of a different biome – the Puszta steppe.
When I refer to a mixed-forest zonobiome I do not intend to imply it was naturally all forested. I consider the classical hypothesis of this biome being one closed canopy forest falsified by a number of facts. However, I also consider it an exaggeration to speak of a European savannah, as there is also evidence pointing against such a hypothesis. To me it is most likely that this biome was a mosaic of open, semi-open and closed landscape, depending on a variety of factors also including such as latitude, precipitation, flooding and humidity, natural disasters and so on. This a complex, big and (needlessly) heated debate that is way to extensive for this post and my drawing. I could have drawn my megaherbivores on a grassy background, which would be only one of several habitats we might have encountered those animals, but sporadic oak trees and hazel bushes are certainly a realistic background as well considering they have been pushed in those habitats as civilization progressed and the space for wild herbivores became increasingly limited.

I decided to give my wild horse a bay dun colouration, as bay dun is suggested to have been most frequent in Holocene wild horses by genetic evidence. Historic evidence, which is dubious, would have favoured black dun, but both colour variants plus the non-dun variants are proven for ferus-type wild horses. For details, see my wild horse summary 2017.

And here a close-up view of the aurochs:

I am very happy with this one, I managed to get it exactly the way I imagine Holocene European aurochs bulls to have looked like. It is based on my reconstructions of several skeletons that I published in July 2017 and I think it bears a striking similarity to many Lidia bulls. I am going to post a refined version of that close-up drawing soon.

Friday, 6 October 2017

Wild horse series pt.I: European wild horses - a summary

It has been long ago that I posted something extensive on wild horses here on my blog, which is why I decided to do a summary of all my past posts on the biology of western Eurasian wild horses. I have been digging through a lot of literature since 2013 and am also going to add some new information. I hope you find useful.

Definition & population genetics

At first I want to define what kind of wild horses I have been focusing on. As long as we regard Przewalski’s horses as members of the same species as domestic horses, Equus ferus, which is what I do, this species is composed of two major clades: one leading to the Przewalski’s horse, Equus ferus przewalskii, and one leading to domestic horses including their wild progenitors, Equus ferus ferus (=Equus ferus caballus). Genetics suggest a separation of both lineages between 160.000 [1] and 38-72.000 years [2]. On my blog, I have been focusing on these extinct wild horses that spread from Iberia to the Eurasian steppe and were domesticated about 6000 years ago.

This is where it gets complicated now, as there are several more or less open questions: How long did true predomestic ferus-type wild horses survive in Europe? Are the historic accounts referring to true predomestic wild horses, hybrids with feral horses or feral populations and wild horses were already extinct by that time after all? Where was the line between the geographical range of the ferus-subspecies and przewalskii-subspecies? (the later question is hardly addressed in the literature).

Not only are these questions hard to answer, there are also some uncertainties on how to refer to these populations. A very popular term is “Tarpan” (in the German-speaking non-scientific literature, it is also very common to distinguish between a “forest tarpan” and a “steppe tarpan” – a distinction that mostly came about by “copy+paste” and is baseless [3]). This term is problematic, as there is no certainty on what kind of animals it originally described. It has become a trend to either condemn the term, or to strictly define it as the term for predomestic, ferus-type wild horses, no matter what it originally described. You have to choose it based on your taste. But if you choose for the latter option, you have to be aware of the fact that “tarpan” has no tradition of being colloquially used for the native European wild horses. It never was a colloquial term outside of the steppes in the way “deer”, “wolf” or “aurochs” was. It has been introduced in scientific literature in 1763 [3], so it is just convention-based. In my 2013 post What do we mean by “Tarpan”? I chose this option and continued to used this term for another few blogposts. However, nowadays I avoid using the word “tarpan” because of all the confusions and also the connection to baseless concepts like “forest vs steppe tarpan” or the Konik-story and others (more on that in an upcoming post). Then I increasingly referred to the ferus-type wild horses as the European wild horse(s), or the Western wild horse(s) (in contrast to the Eastern wild horse, the Przewalski’s horse).

Today I still like to use the latter two terms, usually in plural, as we do not know how many local types there were. Ancient DNA samples indicate two separated predomestic populations on western Eurasia for the Holocene, one on the Iberian peninsular and one extending to the Eurasian steppe [4]. That does not mean that the Pyrenees were the mark were we can find phenotypic differences between the horse populations. It means that at least for some time it was a reproductive barrier. It is possible that there were indeed phenotypic differences between the wild horses dwelling the European mixed forest regions and those in the steppes, as both are two very different biomes. For that, we have to look at the evidence. Furthermore, any differences between wild horses of the European mixed forest biome* and the western steppe biome must have been more or less continuous, because so were the populations.
*I do not claim this biome was densely forested all over. This is not likely by current evidence, but neither is a “European savannah”. Rather, I assume a mosaic of open, semi-open and closed habitat for this biome.

Coat colour genetics

Historic references are a problematic source as it is not clear what kind of free-roaming horses they actually referred to. Ancient DNA, however, can at least provide a clue on the colour phenotypes present in predomestic European horse populations. A number of studies have been conducted in the past, identifying colour alleles in prehistoric European wild horses.
We mainly have to deal with four colour phenotypes that have been proven for predomestic horses, that I briefly introduce now:
Bay: This phenotype is caused by the expression of the dominant allele A and recessive non-dun1. Wild markings like the dorsal stripe are still visible (non-dun2, however, is a domestic mutation that does not show the wildtype markings).
Bay dun: This colour is produced by the bay base colour plus the dominant Dun allele. Wildtype markings are more or less clearly visible (dorsal stripe, shoulder stripes, leg stripes)
Black: Black is recessive under bay and caused by the a allele.
Black dun: Black dun is caused by a black base colour plus the dun allele. It is also referred to as “grullo”, “blue dun”, “mouse dun” and other terms and present in various shades. Wild markings are more or less clearly visible.
(There is also evidence for the so-called “leopard spotted” wild horses [5], but I leave this colour variant aside for now)
Bay dun is also found in other wild equines such as the onager, kiang and Przewalski’s horse, and therefore was ancestral for the ferus-lineage. It is also dominant over all the other colour phenotypes. Bay was the only Agouti allele in the Pleistocene, as the black variant is a mutation that occurred on the Iberian peninsular during the Holocene, where it started to outnumber bay and spread eastwards [6], indicating some selective advantage. What is interesting is that the non-dun1 mutation existed 42.700 years ago at least, and side by side with Dun[7]. Therefore in the Pleistocene, wild horses of the same population either showed a bay dun or bay phenotype (the former might have been more common as it is dominant), while for the Holocene of Europe we have four possible phenotypes – bay dun, bay, black dun and black – as there was no regional or chronological gradient for the dun factor found. In sample of Pruvost et al. for Holocene wild horses of Europe and the steppe, the bay allele was found more often, while in Ludwig et al., the black allele was found to be prevalent on the Iberian peninsular (75%).

This of course provokes two very interesting questions: what was the selective advantage of the black allele (Ludwig et al. speculate that this colour variant was of advantage in an increasingly forested habitat in the Holocene, I consider that plausible), and why where ferus-type wild horses one of the very few large vertebrate species that apparently was heterogeneous in colour? One explanation for the latter question is that maybe we are looking at a mutation-and-selection process that was under its way there and perhaps, if the populations had not been driven to extinction by man, would have become homogeneous in colour after a couple of further millennia of evolution.
Back in 2013, I illustrated the four respectively five colour phenotypes that occurred in the wild ferus subspecies of the Holocene:

The drawing also shows countershading in combination with the white muzzle (“mealy mouth”). It is, based on parsimony, ancestral for the ferus-lineage as well, but as far as I know the genetic background for it has not been resolved yet and also not been tested in aDNA samples, which is why its presence on my scheme is speculative. Countershading can be visible on a black dun phenotype, but never is the white muzzle. This drawing that I recently did shows a bay dun wild horse without countershading and white muzzle (it is actually part of a larger drawing that I am going to present in the next few days):

As an interesting side fact, the sorrel mutation of domestic horses apparently found its way into wild horse populations via introgression (Pruvost et al.).

Coat colour is of course only one aspect of the animals’ life appearance. For the morphology of ferus-type wild horses, we have to look at other sources. If you wonder why I illustrated my wild horses with a falling mane, I am going to come back to that later on.

Osteologic material

Due to the large osteological similarity between wild and domestic horses, it is very difficult to distinguish between both types in the subfossil record. Usually, when the bones are associated with human material, they get assigned to the domestic type unless they show signs of hunting, but the only way to distinguish them safely is on genetic level (I do not know how the studies for wild horse coat colour genes detected wild horse material, but I hope it was one genetic level). In the literature you barely find any remarks on the morphology of Holocene ferus-type wild horses based on osteologic material, and you also never see mounted skeletons of such horses in museums. While it is probably true that the mixed-forest biome is not the primary habitat of horses (I am not claiming that this biome naturally consisted only of closed forests, not at all), Holocene European wild horses could not have been that dramatically rare that one does not even have enough material for a couple of mounted skeletons. I suspect the true reason is that most Holocene subfossil equine material not directly associated with humans have been assigned “Equus sp.” and have been getting dusty in collections since then. I suggest testing some of this material and collect a number of genuine subfossil wild horse specimen, and I am confident that there is enough material for at least a couple of mounted specimen that could give us a clue on the actual morphology of Holocene European wild horses.
The only remarks describing the morphology of alleged European wild horse material were referring to comparisons with Exmoor and Konik bones – not surprisingly, the authors reported large similarity [8,9]. I say “not surprisingly” because the morphology of Holocene ferus-type wild horses was probably comparable to a robust pony type as this body type is also typical of Pleistocene remains and the Przewalski’s horse. However, one has to be careful of self-fulfilling prophecies – that is, if there were indeed ferus-type wild horse remains with a warmblood-like morphology, they might automatically become assigned to warmblood-type domestic horses because our conception of wild horses expects us only to find pony-like remains. What I am saying is that it cannot be bad to strengthen this conception on empirical grounds by examining the subfossil horse record more thoroughly, also to get an idea of regional variation (f.e. it could be possible that wild horses got larger and more long-legged towards the Eurasian steppe as the habitat becomes more open).

Historic accounts of free-roaming horses in Europe and the steppes

As genetic information so far only told us about the coat colour of ferus-type wild horses in Europe, and osteologic material has not been examined thoroughly enough yet to draw solid conclusions on morphology, size and regional variation, we also have to rely on human sources of information: that is, contemporaneous art and written sources. I know of no historic artworks showing supposed European wild horses (except for a “Tarpan” by C. H. Smith from the 19th century, that is way too generic). I do not use Pleistocene cave paintings as source material as the climate and therefore the biome and consequently the wild horse type of the Pleistocene in Europe were not comparable to those of the Holocene.

Written historic sources are problematic for the reasons I outlined above in the beginning of the post. It is not known when original, un-hybridized ferus-type wild horses died out. It could be that they died out in prehistoric times already, and that all historic accounts refer to feral domestic horses that exhibited more or less primitive traits. It is also possible that these populations were hybrids of wild and abandoned/escaped domestic horses. For the texts referring to free-roaming horses of the Russian steppe it becomes increasingly complicated because it is not known how far westwards the original range of the Przewalski’s horse extended, and if those wild horses described in the texts were actually either Przewalski’s horses or hybrids between those and feral horses, and not ferus-type wild horses.

One aspect why some authors discard all historic accounts of free-roaming horses a priori is that they seem to mention falling manes. A falling mane has long been viewed as a domestic trait, as all extant wild equines have a short erect mane. However, this conception is out-dated. Frozen Pleistocene mummies assigned to Equus lambei, a taxon suggested to be conspecific with Equus ferus [10], have a falling mane, showing that this trait arose well before domestication. It has been suggested that a falling mane and a bushy tail evolved as a protection from precipitation [9], and indeed all extant equines with an erect mane live in arid environments with low precipitation. So it is not unlikely that the Holocene wild horses of Europe had a falling mane. It is also most likely that the ancestral population of the domestic horse and the last common ancestor with E. lambei had a falling mane, otherwise this trait evolved twice.
Therefore, a reported falling mane is not an indication of feral horses. There are also authors that read “short mane” (a term used in some of these accounts) as an evidence for erect manes in Holocene wild horses of Europe, but “short” does not automatically mean “erect”, it can also describe a short but falling mane. And most commonly, the sources write of “short and frizzy” manes, indicating the manes were actually falling or at best semi-erect (what also occurs in some domestic horses).

It is not impossible that genuine, un-hybridized wild horses survived in Europe and the steppes until the 19th century. As the subfossil record of predomestic equines is understudied, there is no hint that these populations disappeared before historic times. It is possible, but the contrary is just as likely. However, it is questionable how pure these populations were. Wild horses were reported to steal domestic mares from farms, which is one of the reasons why they were persecuted by man, and must have led to occasional domestic introgression into wild populations. But the influence of escaped or abandoned cavalry horses that ran wild must have been more dramatic, something that happened for centuries. Wild and domestic forms usually tend to hybridize wherever they share the habitat, which is also apparent in canines and pigs. Furthermore, the notion by Pruvost et al. that the sorrel mutation found its way into predomestic populations shows that hybridization did took place.
I believe the truth is that we cannot know what the true identity of the historic free-roaming horse populations of Europe and the western Eurasian steppe was. A number of the historic accounts describe horses with morphological and behavioural traits expected or suspected for wild horses: a small, sturdy body, short mane, large head, wildtype colour (as explained above, there are four to five wildtype colour variants for ferus-type wild horses), shy behaviour that is either untameable or tameable to a certain degree after some time (other wild equines also can be tamed more or less). Due to the morphology and behaviour that is described, I think it is quite plausible that at least some of these populations represented actual wild horses, with a varying degree of influence from domestic horses. Nevertheless, suspected wildtype traits described in those populations would also be found if they descend from rather primitive landraces. And once again, we have to beware of the danger of self-fulfilling prophecies or circular reasoning.

In my 2014 post Western wild horses: What does the evidence actually say? I presented all sources of written accounts on putative ferus-type wild horses that I was able to find. I noted whether I got it from a primary source (therefore the actual text itself), a secondary source (a text that describes what is written in the actual source) or a tertiary source, in order to prevent the danger of Chinese whispers. Another inevitable problem is the language issue, especially with sources from the antiquity. Furthermore, there can be confusion over equine colour terminology because the terms changed over time. For example, “dun” used to refer to bay dun exclusively, and it is still used this way in the UK today, while in the strict modern terminology backed up by a genetic basis “dun” refers to a modification of pigmentation that can act upon variable base colours. Thus notions of an “Isabella” or “tan” colour from the 18th century must not be the same genetic colour variant as in modern terminology. The interpretation of “mouse-colour”, which is mentioned rather commonly in those historic texts, is subjective as well. It is usually interpreted as referring to black dun horses, which is plausible, but it is also possible that individual authors had other colours in mind.
I copy the summary of historical sources from 2014 into this post, also giving a personal interpretation. A P indicates a primary source, a S a secondary source, and a T a tertiary source.

Herodot, 5th century before Christ: lightly coloured (leukos) wild horses in the Ukraine. (S) [8] I am not a classical philologist, but I read that “leukos” can either be translated with “white” or “light-coloured”. A white colour would indicate a domestic origin of the horses described, light-coloured can indicate dun (and therefore wildtype colour) among other colours.

St. Isidore, 600: Iberian wild horses: colour like a donkey, ash-coloured. (P) [11] He is probably referring to black dun horses, as this colour scheme is rather similar to that of a donkey.

Albertus Magnus, 13th century: greyish-coloured wild horses with a dorsal stripe in Germany (S). [8] Very likely black dun as well.

F. Chr. Dahlmann, 1840: Large numbers of wild horses that were hunted lived in Denmark of the 12th century (T). [8] No information on the looks of the horses is given.

H. Röslin, 1593: Wild horses were still living at Vogesen, Elsass-Lothringen. They were faster and wilder than deer and difficult to catch. Once caught, they got tame after some time (S). [8] No information on the looks are given, could be either feral or wild horses.

Balthasar Haquet, 18th century: Wild horses at Zamosc: small, blackish brown, large and thick heads. Mane and tail comparably short, stallions had a beard. Were used in fight shows with predators and showed great bravery (S). [8] The described morphology suits the prevalent conception of wild horses, and “blackish brown” colour could either refer to black dun or other (in this case domestic) colour variants.

Eugeniusz Rozdzynski, 1721: Wild horses at Zamosc: tan and isabelline in colour (T). [8] “Tan” can indicate a lot of colour variants (bay, seal brown, dark expressions of black dun, chestnut), and isabelline could either indicate isabelline in the modern sense or any light colour, in which case dun variants would be among them. Thus, the description could refer to the wildtype colour variants, but also others.

Rytchkof, 1762: colour dun or bluish, other shades exceptional (S). [11] “Dun or bluish” most likely refers to both bay dun or black dun, or various expressions of black dun.

Samuel Gottlieb Gmelin, 1768: Wild horses at Woronesh, Russia. The largest of those wild horses barely reached the size of the smallest Russian domestic horses. The ears were very pinned and the size of domestic horses or sometimes longer and hanging down. The eyes were fierily. The mane was short and frizzy, the tail shorter than in domestic horses. The were typically mouse-coloured, but white or grey horses were also reported. The belly was ashy at the base, the legs were black from the knees downwards. The hair was long and dense during the winter and felt more like fur than horse hair. (P, S). [8] It seems plausible that the author was indeed referring to black dun, white and grey. The morphology fits what is expected for wild horses.

Berenger, 1771: Ural. middling size, roundish and short, big heads, ewe-necked and of a bluish grey colour (S). [11] Once again plausible wild horse morphology. Bluish grey colour could either refer to expressions of black dun or grey.

Peter Pallas, 1771: Free-roaming horses at eastern Prussia and western Siberia. Considered them feral domestic horses. Resembled Russian farm horses, but they had thicker heads, pinned ears, short and frizzy manes and shorter tails. They were of a greyish brown colour and had lighter coloured legs, brown and greys would appear. The colour of the head was white/light on the snout and black towards mouth. Black horses were rare, and there were no piebald ones. They lived in herds of 20 individuals. (S) [8,11] “Greyish brown colour” probably indicates black dun, but “lighter coloured legs” would not fit this colour variant, as it includes dark coloured legs. However, as he also mentions other colour variants, it is not clear which of those had lightly coloured legs. Both “brown” (might refer to bay) and black colours are possible for ferus-type wild horses.

Kajetan Kozmian, 1783: Wild horses at Zamosc: small, strong limbs, enormous strength and uniform dark mouse colour (S). [8] The described morphology is plausible for wild horses, dark mouse colour probably refers to black dun and the notion that the colour was uniform might indicate that the herds he saw were not admixed.

C. H. Smith, 1841: (Probable) Wild horses in the Russian steppe: “coupled with different proportions and position of the ears, an arched or plane forhead, a straight or curved nose, a difference of colour in the eyes, of the skin, of the hoofs, the constancy of their liveries, of their marks, in a streak along the back and bars on the limbs, of dappled croups and shoulders, or of dark uniform colours, dense or thin manes and tails, although traits now mixed,” […] “All seem to refer to a sturdy form of mountain-forest ponies, still found in the province of Cordova, in the Pyrenees, the Vogesian range, the Camargue, the Ardennes, Great Britain, and in the Scandinavian highlands: all remarkable for an intelligent but malicious character, broad forheads, strong lower jaws, heavy manes, great forelocks, long bushy tails, robust bodies, and strong limbs; with a livery in general pale dun, yellowish brown and a streak along the spine and cross bars on the limbs, or the limbs entirely black, as well as all the long hair and mostly having a tendency to ashy and gray, often dappled on the quarter and shoulders”. […]“These horses are evidently again referred to by Andr. Schneebergius, who states, that “there were wild horses in the preserves of the prince of Prussia, resembling the domestic, but mouse-coloured, with a dark streak on the spine, and the mane and tail dark;” […] “Real Tarpans are not larger than ordinary mules, their colour variably tan, Isabella or mouse, being all shades of the same livery, and only varying in depth by the growth or decrease of a whitish surcoat, longer than the hair, increasing from midsummer and shedding in May: during the cold season it is long, heavy, and soft, lying so close as to feel like a bear’s fur, and then is entirely grizzled; in a summer much falls away, leaving only a certain quantity on the back and loins: the head is small, the forehead greatly arched, the ears far back, either long or short, the eyes small and malignant, the chin and muzzle beset with bristles, the neck rather thin, crested with a thick rugged mane, which, like the tail, is black, also the pasterns, which are long: the hoofs are narrow, high and rather pointed; the tail, descending only to the hocks, is furnished with coarse and rather curly or wavy hairs close up to the crupper; the croup as high as the withers: the voice of the Tarpan is loud, and shriller than that of a domestic horse; and their action, standing, and general appearance, resembles somewhat that of vicious mules.” (P) [12] For Smith’s full text, go here.

C. R. Darwin, 1868: “It seems that not very long ago a wild breed of dun coloured horses with a spinal stripe was preserved in the royal parks in Prussia. I hear from Hungary that the inhabitants of that country look at the duns with a spinal stripe as the original stock, and so it is in Norway”. (P) [11] He is probably referring to the population at Zamosc.

Heptner, 1989: Last living Tarpan (Dubrowka Tarpan), died 1918. It was 140-145 cm tall, had a large head, small ears, short neck, mouse-coloured coat, broad dorsal stripe, faint shoulder stripes, black mane, tail and lower legs, semi-erect mane, broad and arched front head and a straight head profile. (S) [13]

I have not commented the remarks of Smith and Heptner yet because I treat the problematic of the steppe horses as a separate issue. At first, I want to come back to the horses of Europe western to the steppes. The colour that is mentioned most frequently is probably black dun. While in genetic samples bay was the prevalent base colour, there was a tendency of the black allele becoming increasingly common, as mentioned above. If this trend continued to historical times, the black allele might have outnumbered the bay allele. So the prevalence of black dun horses in written accounts is plausible for wild horses. For Iberia, St. Isodore mentioned probably black dun horses exclusively (“donkey coloured”), while Smith describes a southwestern European horse type that might have displayed both bay dun and black dun (“pale dun and yellowish brown” vs. “ashy to grey”). Therefore, the described coat colours often match with what is expected from genetic information. But it cannot be denied that many of the sources also mention colour variants not proven for wild horses, i.e. most likely domestic colours, such as white/grey and perhaps the notion of “isabelline” and “tan” horses (however, both words can also indicate bay and dun horses, which would be wildtype colour variants). It is of course possible that all those free-roaming horse populations were feral, primitive rural horse type and hence the wild horse-like behavioural and optical traits. But it is, in my opinion, just as likely that they do represent wild horse populations that mixed with abandoned/escaped domestic horses to a varying extent. It is a tricky situation: it would be as if it was not known when genuine aurochs eventually disappeared, but there would be historic accounts of very aurochs-like bovines ranging freely, looking a lot like aurochs but some individuals having diverging colour or horn variants. True aurochs, feral primitive cattle, or hybrids?
Only a genetic test of a representative sample of individuals from those populations could provide clarity.

For the western steppes, it becomes even more complicated, as it is not clear how far westwards the Przewalski’s horse originally ranged. Artworks from the 4th and 5th century before Christ resemble Przewalski’s horses in having an erect mane and a short-haired tail base [14], which makes it likely to me that this subspecies extended at least until the south-western edge of the steppes. The records of putative wild horses from the western steppes could easily also refer to Przewalski’s horses or hybrids of Przewalski’s horses with feral horses. Especially the notion of the “Tarpan” resembling “vicious mules” is reminiscent of the Przewalski’s horse with its standing mane and large, donkey-like head. It was probably not coincidental that this equine was initially described as Asinus przewalskii.

The records of three particular individuals from these free-roaming steppe horses are problematic too. One of them, the “Krim Tarpan”, was suspected to be a feral horse already back in the 19th century. The purportedly last individual of these populations, the “Dubrowka Tarpan”, fits the conception of a wild horse except for its comparably large size (see above), at least according to Heptner 1989. It is possible that it either a) was indeed a wild horse (wild horses are usually thought to be smaller, but as I wrote above there is no comprehensive subfossil record of articulated skeletons ferus-type wild horses from which we could get an idea of their size, especially not from the steppe, so we cannot rule out that there were 145cm tall wild horses) b) a hybrid between a ferus-type wild horse and a feral horse, c) a hybrid between a Przewalski’s horse and a feral horse, d) a feral horse with some primitive traits. The most famous of those “tarpan” individuals is the “Cherson tarpan”, which was caught as a foal in the Ukraine and exhibited in the Moscow Zoo. It is famous as it is the only “tarpan” that was photographed. It displayed aggressive behaviour and was castrated, but otherwise showed none of the suspected wildtype traits clearly. The mane was opulent and long, the legs were long too, and the colour could have been either bay, seal brown or a very dark shade of black dun to my eyes. With a size of 133cm at the withers, it fitted the expected height for wild horses, but so do many rural horses. It could simply be a feral horse, based on its looks I see nothing compelling that would qualify it as a genuine, un-hybridized wild horse.

As you see, there are a lot of open questions, room for interpretation and subjectivity on this subject. Some of these can only be solved by conducting research (such as identifying complete osteologic material of genetically proven predomestic horses, to get an idea of their actual morphology), others probably have to remain unanswered (such as the status of the historic free-roaming horse populations, as no specimen are preserved).
The true nature of Holocene ferus-type wild horses is of course relevant for ecologic restoration in order to chose horse breeds that serve as the most authentic proxy. For a number of breeds there are background stories purported by their advocates as the most authentic and most “wild” European horses today, which are mostly based on Chinese whispers, arbitrary interpretations and fabrications. The ferus-type wild horse is as extinct as the aurochs. I already wrote several posts on this subject. When I can find the time, I am going to do a similar summary on this issue.

For the posts that I wrote in the past and used as a basis for this summary, go here:


[1] Ryder et al.: A massively parallel sequencing approach uncovers ancient origins and high genetic variability of endangered Przewalski’s horses. 2011.
[2] Orlando et al.: Recalibrating Equus evolution using the genome sequence of an early Middle Pleistocene horse. 2013.
[3] Tadeusz Jezierski, Zbigniew Jaworski: Das Polnische Konik. 2008.
[4] Cieslak et al.: Origin and history of mitochondrial DNA linages in domestic horses. 2011.
[5]  Pruvost et al.: Genotypes of predomestic horses match phenotypes painted in paleolithic works of cave art. 2011
[6] Ludwig et al. 2009: Coat color variation at the beginning of horse domestication
[7] Imslandet al.: Regulatory mutations in TBX3disrupt asymmetric hair pigmentation that underlies Dun camouflage colour inhorses. 2015.
[8] Tadeusz Jezierski, Zbigniew Jaworski: Das Polnische Konik. 2008. Polish Academy of Sciences
[9] Baker, Sue: Exmoor Ponies: Survival of the Fittest – A natural history. 2008.
[10] Kirkpatrik, Jay F.; Fazio, Patricia M.: Wild horses as Native North American Wildlife.2005
[12] Charles Hamilton Smith, 1841: The Natural History of horses, with Memoir of Gesner
[13] Hardy Oelke: Wild horses then and now. Kierdorf-Verlag.
[14] Cis van Vuure: On the origin of the Polish konik and its relation to Dutch nature management. 2014.