Often in science, there is the problem that a wide-ranging proposal that has a lot of echo in and outside the scientific community (often more in the latter) is not based on a very solid ground. But because of its wide-ranging nature and – I dare not to say sensational – implications, it gets a lot of attention and becomes adopted in many further sources even if the wide-ranging conclusion was premature. But unfortunately, when one or several new studies come out that refute the original proposal, it does not get nearly the same resonance. This way, exciting but not factually correct “new discoveries” become mainstream in and outside the scientific community although they have been refuted already in sometimes several works. One example might be the hypothesis that the wisent is an aurochs hybrid, which is not correct according to recent research. Another example is in the focus of this post, namely the hypothesis put forward by Gaunitz et al. (2018) that the Przewalski’s horse is not a genuinely wild animal but the feralized descendant of early domesticated horses from Botai site, Kazakhstan. For brevity, I call it the Gaunitz hypothesis in this post.
The hypothesis
The 2018 Gaunitz et al. study looked at 42 ancient horse genomes, including 20 from the eneolithic Botai site. This site was of particular interest for the authors because Botai was or is considered the earliest archaeological evidence for horse husbandy (note: not necessarily domestication in a genetic sense). The big surprise of the study was that all Przewalski’s horses examined in the analysis cluster within the Botai-Borly complex, therefore the authors jumped to the conclusion that the Przewalski’s horse is not a genuine wild animal but a feral descendant of the earliest horse domestication event [1]. Further studies seem to confirm the deep nesting of the PH within the Botai-Borly complex, most notably Librado et al. [2].
Why should a domestic history for the PH surprise us that much? Obviously, because the PH does not show any obvious signs typically associated with domestication in mammals: it has a larger brain volume than domestic horses, a skull that is not paedomorphic at all with small eyes and a large viscerocranium, it has a much higher aggression potential than domestic horses even in captivity and does not have the same mutation accumulation as domestic horses have developed during domestication (known as the costs of domestication [3]). Ever since the 20th century, the PH was used as a model for a wild horse opposed to the domestic horse. Apart from that: if the PH is not genuinely wild, it would mean that there are no genuinely wild, never-domesticated horses left on earth. So it is well worth investigating whether they truly have a domestic past or not. And it did not take long until the conclusion by Gaunitz et al. was called into question.
Were Botai horses domesticated?
In 2021, a study by Taylor & Barron-Ortiz was published that called the domestication of the Botai horses into question. They argued that the tooth recovered from the site that seems to show bit wear does not do so, as teeth of Pleistocene wild horses show the same wear pattern [4]. Furthermore, they argue that the equine lipids found on pottery from Botai do not necessarily have to be milk lipids (which would indicate that the mares there would have been tame enough to be milked) but could also be from meat or fat [4]. Also, the human teeth from the site do not show traces of milk proteins, which would be expected if they consumed milk on a regular basis [4]. Lastly, the authors argue that the age structure of the horse remains at the site is more consistent with mass harvesting of wild horses rather than raising horses under human custody, and they mention a source describing horse remains of signs of being hunted (see below). Thus, the authors conclude that Botai is not the earliest site with evidence of horse domestication but rather shows a culture that was heavily dependent on mass harvesting of wild horses present in the region [4]. Therefore, the hypothesis that the PH is a feral descendant of once domesticated horses loses its basis.
However, the paper by Taylor and Barron-Ortiz did not remain uncontested. In the same year, Outram et al. published a preprint that took their argumentation apart [5]. The Botai tooth does show a unique abrasion pattern consistent with bit wear, namely parallel abrasion at the front of the second premolar of the lower jaw, while similar wear is absent from other teeth in the mouth. This pattern was not found in the wild horses examined by Taylor and Barron-Ortiz, what actually endorses and not refutes the assumption that Botai horses were at least tamed [5].
The lipids from the pottery, so the rebuttal, are more likely to be from milk than from meat or fat because horse milk is produced during summer and therefore has a different deuterium signature compared to meat and fat, which take longer to build up in the body [5]. Furthermore, Outram and others argue that the sample size of two human teeth is not enough to refute milk consumption for the Botai people [5].
What about the signs of hunting on horse remains? The study referred to by Taylor and Barron-Ortiz was by Olsen (2005) and shows rather unambiguously that horses were hunted at the Botai site [6]. Harpoon projectiles were found and many horse remains at Botai show wounds consistent with these on rib bones and skulls, which indicates that these horses were hunted [6]. This, according to Outram et al., does not refute that there were domesticated horses at Botai, only that wild horses were hunted there too [5]. While that may be true, it has massive implications for the Gaunitz hypothesis: if not all horse remains found at Botai were from early domesticated ones but also from genuine wild horses that were hunted, the conclusion that the PH necessarily descendants from these early domestic horses loses its ground, assuming that the horses at Botai were all from the same region and thus genetically similar.
New data: it was a sample size problem
The last puzzle piece that refutes the Gaunitz hypothesis was a study by Lira Garrado et al. (2025). It actually focused mainly on the phylogenomics of the Iberian wild horse, including 87 genomes of this variant, but also included 261 other ancient horse genomes, which is a rather large sample size [7]. As it turns out, the PH is not nested in a monophyletic Botai-Borly clade, but rather both PH as much as Botai-Borly horses cluster with other old, predomestic lineages from the Ural and Central Asia; with these other old lineages, they form a much broader Central Asian/Uralic horse cluster [7]. This means that Botai/Borly horses and the PH do not form an exclusive monophyletic clade together, but rather descend from a common, large Central Asian wild horse population. There is therefore no evidence that the PH was ever domesticated; rather, it is the last remnant population of this large Central Asian wild horse population.
Literature
[1] Gaunitz, C., Fages, A., Hanghøj, K. E., Albrechtsen, A., Khan, N., Schubert, M., et al. (2018). Ancient genomes revisit the ancestry of domestic and Przewalski’s horses. Science, 360(6384), 111–114.
[2] Librado, P., Khan, N., Fages, A., Kusliy, M. A., Suchan, T., Tonasso-Calvière, L., et al. (2021). The origins and spread of domestic horses from the Western Eurasian steppes. Nature, 598, 634–640.
[3] Moyers, B. T., Morrell, P. L., & McKay, J. K. (2018). Genetic costs of domestication and improvement. Journal of Heredity, 109(2), 103–116.
[4] Taylor, W. T. T., & Barrón-Ortiz, C. I. (2021). Rethinking the evidence for early horse domestication at Botai. Scientific Reports, 11, 7440.
[5] Outram, A. K., Bendrey, R., Evershed, R. P., Orlando, L., & Zaibert, V. F. (2021). Rebuttal of Taylor and Barrón-Ortiz 2021 Rethinking the evidence for early horse domestication at Botai. Zenodo.
[6] Olsen, S. L. (2006). Early horse domestication: weighing the evidence. In S. L. Olsen, S. Grant, A. M. Choyke, & L. Bartosiewicz (Eds.), Horses and Humans: The Evolution of Human–Equine Relationships. BAR International Series 1560 (pp. 81–113). Oxford: BAR Publishing.
[7] Lira Garrido, J., Tressières, G., Chauvey, L., Schiavinato, S., Calvière-Tonasso, L., Seguin-Orlando, A., et al. (2025). The genomic history of Iberian horses since the last Ice Age. Nature Communications, 16, 7098.
