The Sorraia is one of those horse breeds of which some of its advocates claim it is a remnant of a special wild horse population. According to them, the Sorraia descends from supposedly wild horses spotted in the 1920s in a rural region in Portugal, is phenotypically identical with a local wild horse type and genetically similar to the Polish Konik and the Przewalski, endorsing its status as a wild horse. With this post I try to give an overview over which arguments supporting a special status for the Sorraia among other Iberian horse breeds are verifiable, and what seems to be a colourful or wishful interpretation based on my research for this post.
I want to note that all this is my personal interpretation of data that was available for me, if anyone has additional material or disagrees with my reasoning for good reasons, I would like to read about it.
First of all, the Sorraia does not descend from any members of that one free-ranging horse herd spotted by Ruy d’Andrade as he was unable to find them again. The Sorraia actually descends from farm horses he thought to resemble those horses which he collected for a breeding program in 1937. All Sorraia descend from four stallions and seven mares of local farm horses, to which a Criollo stallion was added later on.
There are no reports of a feral horse population resembling those spotted by d’Andrade, not even in C. H. Smith’s extremely comprehensive work The Natural History of Horses with Memoir of Gesner. D’Andrade is the only person who saw them, and he saw them only once. I think it is much more likely that he saw just a herd of escaped local farm horses, rather than a population of free-ranging horses that remained undetected for decades or centuries, were spotted by one man and disappeared into the void again. The only references claiming there was an existence of these free-roaming horses prior to d’Andrade’s encounter that I know of is either d’Andrade himself or Hardy Oelke (who is one of the main contemporary Sorraia advocates), so I don’t know of a neutral reference. Sorraia fanciers also claim that the so-called “zebro” (“encebro” in Spain) was a wild Sorraia-type horse that disappeared in historic times. The Zebro was in fact more likely an Iberian wild ass, a late-surviving strain of E. hydruntinus  or maybe a population of feral donkeys (their wild nature is no argument against that, as feral animals do in fact re-develop such wild traits). The European wild ass survived the longest on the Iberian Peninsular, the last of these Zebros survived into the 16th century in a game reserve in Murcia .
As you see, the modern Sorraia is not older than 80 years, and neither does descend from a wild or feral horse population but from local farm horses (just like the Konik, by the way). Nevertheless, there is still the possibility that these farm horses may have been genetically influenced by surrounding wild horses, so let’s have a look at what genetic studies say about this horse breed so far.
There is some debate on whether the Sorraia is the ancestor of other, similar Iberian-based horse breeds like the Lusitano or various Mustang lineages. This is in fact not really relevant for the question stated in the headline of this post; domestic breeds do give rise to other breeds, that does not place them closer to their wild type. But since the population that is called Sorraia originated when d’Andrade started his breeding programme in 1937, it is chronologically not possible that the modern Sorraia gave rise to similar Mustangs and the Lusitano, though I think its possible that they derive from a common stock, as genetic similarities between the Sorraia and the Lusitano[2,3,4] and phenotypic match with some Mustangs (f.e. Kiger) suggest. It is not relevant for the wild horse debate anyway.
If the Sorraia is a true wild horse, or at least strongly wild horse-influenced as advocates claim, there should be a clear genetic indication for it. A connection with the Konik, Mongolian and Przewalski horse was suggested in Jansen et al. 2002, but later studies found the Sorraia to be more or less closely related to other Iberian horses [3,4,5,6]. Luis et al. 2007 write: “Because of its low variability, the Sorraia usually separated from the other [Iberian] breeds when only one type of marker was used, seen also in […]. However, when we used all three types of markers the Sorraia clustered with the other Iberian and Iberian influenced-breeds […] However, new results from mtDNA analysis revealed a lost haplotype in the Sorraia breed, which is included in the mtDNA haplogroup recognized for heaving high frequency in Iberian horses, and the result obtained here indicate closeness of the Sorraia with the other Iberian breeds, which fits historical documentation.” Lira et al. 2009  compared mitochondrial haplotypes of Iberian horses with those of ancient domestic and wild horses from the Neolithic to the Bronze age and Middle age. They find no relationship between the Sorraia and the ancient samples and conclude: “Lastly, our data do not support the Sorraia horses as a primitive predomestic lineage, in contrast to a previous claim (d’Andrade 1945).” The only similarity of the Sorraia with an ancient sample is with one that is from the middle ages, what does not implicate the Sorraia is an ancient, isolated breed. They consider the introgression of local wild mares for all Iberian horses is possible, without giving the Sorraia a special status.
Sorraia advocates argue that the wild horse samples are too small and the genetic diversity in Sorraia too limited to give an accurate clue about its relation to predomestic horses. However, if the Sorraia is truly a remnant of the wildtype of Equus ferus ferus, then it there should be at least some genetic markers totally unique to the Sorraia – unless domestic introgression and genetic bottlenecks purged them out (if that’s the case, it is the question if it still can be considered a representative of the wildtype then). Advocates of the breed argue that absence of evidence is not evidence of absence, and that is true. But it is no good basis for assuming presence either.
Pleistocene horses portrayed in southern European cave paintings are used as a comparison for the Sorraia, and the fact that these painted horses show leg and shoulder stripes is used as an argument to support the primitive nature of the Sorraia. But first of all, keep in mind that the end of the glacial brought a dramatic shift in the range of the Eurasian mammal fauna and terrestrial biomes, and the Pleistocene horse types probably represented a type different from that inhabiting Holocene Iberia. In my personal opinion, it is likely that some of these paintings might represent members of the przewalskii subspecies with standing manes rather than ferus with falling manes. But if we are willing to use these paintings as a reference, what do they tell us? That wild horses had striped legs and shoulders to a certain degree, what is typical of many dun horses and therefore no argument in favour of any breed.
The Sorraia is considered a primitive horse particularly because of its coat colour. It is predominantly grulla-coloured, and grulla is often considered the standard colour of western wild horses (although genetic data suggests that a large portion was coloured bay/bay dun [8,9] and grulla is only one of the colours mentioned for historic wild horses). Black dun has many different shadings, varying from light grayish, dark gray to yellowish to almost optically brown. Some Sorraia apparently also show a bay dun colour (see below). Bay dun and black dun are wildtype colours, and any domestic mutations present in the breed could be explained by domestic influence if we assume the Sorraia is a wild or ancient horse, so the coat colour does not contract this hypothesis. I have to note two things though: 1. the striping of the Sorraia isn’t as exceptional as often stated, stripes are part of wildtype colour + dun expressions and shown by other grulla horses, like the Konik or Hucule (considering that the stripes were regarded as a trademark of the breed right from the beginning, it is also possible that the breeders selected for prominent striping). The there are no photos supporting the claim that the stripes in Sorraia were more prominent in the past. 2. prominent stripes are not the unmistakeable indication of European wild horses as one might think because of the seductively strong striping of the African ass and, obviously, zebras. In fact, historic references for wild or free-ranging horses don’t mention them that often, while the dorsal stripe is nearly always remarked. Also in the Przewalski’s horse the stripes are not that prominent (anymore?).
|Striped black dun Hucul, there are even more strongly striped Hucul|
|This Sorraia seems to be bay dun ©Lynne Gerard|
The optic similarity with the Konik, a primitive horse breed from eastern Europe and thus far removed from the Sorraia, provokes the assumption that their phenotypic traits are inherited from a common ancestor since genetic exchange seems very unlikely. But in fact this similarity is grounded almost exclusively on the coat colour features (as primitive Koniks are smaller and more sturdy in build), which are regulated by only two loci (Agouti and Dun) since both breeds are non-sorrel horses. Bicoloured manes and tails, dorsal and leg stripes are part of the expression of black dun (=grulla). Black dun colours are not rare in rural horses and ponies, and it is one of the wildtype colours of Equus ferus; thus its presence in two compared horse breeds is not a surprise and also no indication of a common, predomestic ancestry.
While the coat colour is no argument against a status of the Sorraia as an ancient horse, their body conformation seems problematic. The Sorraia is a quite large (140-150 cm), lightly built horse with a proportionally small and slender head. Historic references for central and eastern European, Russian and even western European wild horses describe a horse type with a body conformation that is quite the opposite – small, stocky, with strong legs and a proportionally large and robust head. However, Sorraia fanciers argue that the Sorraia is a remnant of a specialized Iberian subtype of the wild horse, and there are no historic references describing wild horses from the western Iberian peninsular as far as I know, so that this is not necessarily contradicted. Genetic data suggests that the extinct wild horse population was separated into two main populations: one huge, panmictic population eastern to the Pyrenees inhabiting the Eurasian steppe, and one smaller population western to the Pyrenees inhabiting the Iberian peninsular . So the hypothesis there was an Iberian wild horse that slenderer and taller than other members of Equus ferus seems at least possible. But is there evidence for it? This is hard to tell because it is difficult to tell wild horses apart from domestic horses in the subfossil record of the late Holocene without genetic testing. I know that d’Andrade did measurements with skeletal elements of the Sorraia and supposed wild horses, and I do not dispute his professional competence, but I am not sure if it was possible to safely distinguish between domestic and wild horses back this time (except for some unambiguous cranial features, such as enlarged eyes and a reduced brain case). Furthermore, the Sorraia remains he worked with were prior to the bottleneck event, so modern Sorraia might display different features due to genetic drift. I know of Ebhardt’s division of 4 predomestic subtypes of Equus ferus that includes one type resembling the Sorraia, but it seems that there is low scientific support for it in modern scientific literature (although it is popular among hippologists). I haven’t seen any recent studies using or referring to it. Of course it is likely that different environments had a different impact on wild Equus ferus over its large geographic range, but the proposal of regionally differentiated body conformations should be grounded on a osteologic basis or on reports of people having seen these animals in flesh. Of course it is possible that the western wild horse was regionally differentiated, but other large herbivores with a large geographic range such as Red deer or moose don’t display any radical differences in body conformation when comparing their various subspecies (island dwarfism in case of Red deer on Sicily is another story), so it does not imply it.
Sorraia.org claims that the Danish wild horse skeleton from 11.100 years BP is identical in proportions and skull shape to Sorraias. I do not really see that, it would be more conclusive for me if it would be a member of the przewalskii clade and its head is clearly more robust (sorraia.org provides a superimposition with a Sorraia skull photo from d’Andrades work, but this historic skull hardly fits modern Sorraia’s heads because the lower jaw is more robust; perhaps an effect of the genetic bottleneck?). I will do a life reconstruction of that skeleton soon to get a better idea of what it might have looked like in life. Nevertheless, not all Sorraia have a gracile head. Looking at this dental, claimed by sorraia.org to be an example of a gracile-jawed wild horse remain, and comparing it with an Exmoor pony skull from Sue Baker’s book, it appears to me that this wild horse had a skull only insignificantly longer (=more gracile) than the Exmoor. To affirm the question if there are definite wild horse remains that endorse the hypothesis that there were predomestic horses that had a body conformation similar to the Sorraia, a new osteometric study comparing a representative sample of confirmed wild horse and Sorraia skeletons would be needed. Sorraia advocates sometimes also refer to Equus stenonis, but this is a huge and most likely incorrect stretch, as E. stenonis was only distantly related to E. ferus and disappeared in the late Pliocene. Furthermore, their hypergracile skulls have a concave profile, the opposite of what we see in the Sorraia.
|Skull of a Sorraia Mustang mare ©Lynne Gerard|
|Skull of a Sorraia Mustang stallion ©Lynne Gerard|
While looking for good Sorraia skull photos to compare them with wild horse skulls, I stumbled across something interesting. Lynne Gerard published photos of skulls of two of her Sorraia Mustang individuals, a stallion and a mare, on her blog. While neither of the photos shows a lower jaw in a useful lateral perspective, we are able to see that the size of the orbital openings is well larger than in wild horse skulls (I am using the Danish specimen, and in the lack of other useful European species, also American wild horses as a reference, which where probably conspecific). The increase of the eye size, as well as a reduction of brain size, is a typical artefact of domestication, visible in virtually any domesticated larger mammal, and shared by other domestic horses – even primitive ones like the Exmoor pony.
I want to make clear that I am not biased against the Sorraia. The idea that Iberia had a unique subtype of the European wild horse differing in having a taller, much slenderer body than other wild horses is exciting, but I think the evidence for that is rather thin. And if there was, I do not think the Sorraia would be a predomestic remnant population of this type, but more likely a domestic breed with introgression of these horses. Nevertheless, I think that the western Iberian wild horse was not drastically different from the wild horses formerly inhabiting the Pyrenees and Northern Spain as described by C. H. Smith, unless there is there is evidence for the contrary – I think this is the most parsimonious and most likely assumption. An osteometric evaluation of confirmed wild horse remains from various places of Europe would be very helpful to gain information about the appearance of Holocene wild horses and regional diversification.
The hypothesis that the Zebro was a Sorraia-type horse rather than a wild for feral ass seems far-fetched as well. If this was not the case, then I assume that there was no feral population with a long history in the area d’Andrade spotted those nebulous horse herd, as there are no other references to it. Most likely he saw an abandoned farm horse herd. The Sorraia itself is not older than 80 years, since it was created in the year 1937 by collecting local farm horses resembling those spotted by d’Andrade. Genetically, there is no solid evidence for the Sorraia having a special status on the domestic equine tree. Absence of evidence is no evidence of absence, but proposals like the Sorraia being a predomestic horse should have a more solid basis than this.
It seems that the Sorraia breed is “just” a domestic horse breed, albeit a hardy and robust landrace. In my personal opinion, the Sorraia and Lustano as much as similar mustangs like Kiger, Sulphur and Pryor might all derive from a historic domestic breed/type of horses with an appearance very similar to their modern counterparts, what is inferred by their very similar phenotype and also genetic similarities.
 Bunzel-Drüke, Finck, Kämmer, Luick, Reisinger, Riecken, Riedl, Scharf & Zimball: Wilde Weiden: Praxisleitfaden für Ganzjahresbeweidung in Naturschutz und Landschaftsentwicklung. Arbeitsgemeinschaft Biologischer Umweltschutz
 Jansen et al.: Mitochondrial DNA and the origins of the domestic horse. 2002
 Luis et al.: Genetic diversity and relationships of Portuguese and other horse breeds based on protein and microsatellite loci variation. 2007
 Luis et al.: A lost Sorraia maternal lineage found in the Lusitano horse breed. 2006
 Lira et al.: Ancient DNA reveals traces of Iberian Neolithic and Bronze Age lineages in modern Iberian horses. 2009
 Royo et al.: The origins of Iberian horses assessed via mitochondrial DNA. 2005
 Cieslak et al.: Origin and History of Mitochondrial DNA lineages in Domestic horses. 2010
 Ludwig et al.: Coat colour variation at the beginning of horse domestication. 2009
 Pruvost et al.: Genotypes of predomstic horses match phenotypes painted in paleolithic works of cave art. 2011
Cieslak et al.: Coat colours and mitochondrial lineages of ancient horses to document domestication. 2011