Monday 25 December 2017

What if the aurochs was genetically reconstructed?

This post plays with the possibility of a genetically resurrected, genuine aurochs. This possibility emerged since a full genome has been resolved in 2015. This animal would be a true aurochs in its original form, not domestic cattle with a number of morphological/optic wildtype traits that might be considered dedomesticated at some remote point of time of natural selection. What to do with such a reconstructed aurochs, and would it make existing “breeding-back” projects unnecessary?

Each time it comes to genetically reconstructing an extinct animal, there are people questioning the purpose of doing so. Well, reconstructing a genuine aurochs would of course be for the same purpose of breeding aurochs-like cattle and using it for nature restoration. It serves an ecologic good and is about authenticity.
A number of recent posts covered the probable genetic differences between aurochs and domestic cattle, especially this one. I outlined why I believe that we definitely need more than mere matches on SNPs and haplotypes as the genetic differences between aurochs and cattle include a lot of aspects that have a dramatic influence on the organism as a whole (neurology, endocrinology, ontogeny, sexual dimorphism, immunology) and are responsible for the big differences in morphology, development and behaviour that we see between the wild animal called aurochs and its domestic counterparts. It is very likely that  these differences are caused by a high number of loci whose individual role in the whole have not been resolved yet – let alone the identification of wild vs. domestic alleles on these loci and their distribution among domestic cattle today. This means that true breeding-back of a genuine, pre-domestic aurochs is not possible by the current state of knowledge and that it would take a very long time, dozens of decades, because of the very high number of loci involved. Not to speak of the fact that many of the original defining wildtype alleles probably have been lost in the dramatic and intense process of domestication anyway.

Possibilities to genetically reconstruct an aurochs

So that true “breeding-back” is most likely not feasible as the necessary material is a) not identified and traced down in modern cattle and b) very likely not fully preserved anymore, how to get to a genuine, pre-domestic aurochs then? Luckily, and finally, the full nuclear genome of a British aurochs bull from 8.000 years ago was fully resolved in 2015 (see here), after the full mitochondrial genome of the aurochs had already been known since 2011.
This makes it feasible to truly revive a genuine, pre-domestic aurochs individual. The fact that it is a male is especially advantageous, as male-specific aurochs genes on the Y chromosome are known this way that otherwise would be unknown, which would have an impact on sexual dimorphic traits. But how to get from the genetic information to a life aurochs? Two methods are usually mentioned for that are:  
a) Cloning. Both bantengs and gaurs have already been cloned successfully using a domestic cow as a surrogate. For an aurochs, a domestic cow would be even more closely related. For cloning, having the genetic information only is not enough. For the process of somatic nucleus transfer, an actual aurochs tissue cell with an intact nucleus that contains not only the DNA molecule but also the associated proteins that are necessary for the regulation of the cellular processes are needed in order to produce a viable embryo. As far as I know, what we have is not enough to produce a clone, let alone the usual obstacles of cloning (somatic nucleus transfer has a rather low success rate, and cloned individuals often suffer from developmental problems).
b) Genome editing with CRISPR-Cas9. This new method provides a lot of new possibilities regarding genetic engineering and also the recreation of recently extinct species. It enables to do very precise modifications of the DNA molecule and to change or exchange genes point by point. By this technique it would be possible to take the DNA of a surrogate cattle and exchange the domestic alleles with aurochs alleles, producing a viable cell with aurochs DNA, and eventually leading to an aurochs calf being born by a suitable surrogate mother. The mitochondria however will still be of domestic origin, but they should have only a minor influence on the phenotype of the organism. As for epigenetics, perhaps the difference would be small as well as aurochs and domestic cattle are closely related, but we shall see. However, as the DNA is from a somatic cell of a grown aurochs and not from an embryo or calf, the individual might experience similar developmental issues as clones do.

If a genuine aurochs was born

But when finally a genuine aurochs individual is successfully recreated and born and growing healthy, what do with it? It will be a male, and of course more than just one individual with this genome can be created. At first, it will be most interesting to see it growing and become a mature individual. Now that we have something directly to compare with, we would see how far “breeding-back” results or any primitive breed match the goal. Perhaps also some minor surprises about its life appearance would revealed. Those reconstructed aurochs bulls could go to zoos for display just like any wild animal. They could even be used as new breeding bulls for any “breeding-back” herd, be it in grazing projects, zoos or private farms in order to successively increase the aurochs-likeness of those herds. One would have to be aware of the fact that it is a wild animal though, with the behaviour and dangerousness of a wild bovine.

Building up a population

Theoretically, even a female can be constructed by removing the Y chromosome and replacing it with a duplication of the X chromosome of the aurochs bull. With that, you can theoretically build up a herd, but the genetic diversity would of course be extremely narrow. Probably a lot of the first-generation crosses if not all would end up stillborn or as miscarriages because lethal alleles have a high chance of homozygosity in the offspring. Furthermore, these aurochs, which are basically from 8.000 years ago, might also be vulnerable to many modern cattle diseases. The only way to create a viable population of aurochs would be to acquire more genetic diversity. If one is optimistic, perhaps the full nuclear genome one or even more further aurochs individuals, in the ideal case one of another time and region, can be resolved. In this case it might be enough – if we get, say, the full genome of five individuals from different regions of Europe and different geological times, we would probably have a way higher genetic diversity than in the Wisent already. But if one is pessimistic, that one British aurochs bull might remain the only individual for which the full genome can be resolved. In this case, there would be only one way to achieve more genetic diversity in order to create a viable aurochs population: controlled, systematic hybridization with domestic cattle.

I would use a technique called absorptive breeding, where the hybrid offspring is constantly back-crossed with pure individuals of the parental generation until the level of “purity” desired is achieved. This way, a fifth-generation hybrid would be genealogically almost 97% pure. Most alleles would be aurochs alleles, but it would also include domestic cattle alleles (however, genealogy is not everything – theoretically it is also possible that such an animal still has 50% domestic alleles, although statistically unlikely). It would be best to not to use just any domestic cattle, but to use robust and healthy breeds to introduce necessary immunological alleles for the resistance against modern cattle diseases, f.e. Chillingham cattle and other robust northern breeds, and primitive landraces.
I would build up a herd composed of third, fourth and fifth generation hybrids created by absorptive breeding, keep them under semi-feral conditions (therefore, using them under the frame of a grazing project in a natural area would be the best option, although their behaviour would be challenge to conventional grazing projects) and then let them breed for themselves but would still select the offspring. This is necessary in order to achieve the goal: the goal is to add genetic diversity for resistance against diseases and overcome developmental problems, but at the same time not altering the biologic integrity of these wild cattle. That means artificial selection should focus on breeding out individuals with obvious signs of hybridization that would show in the phenotype (remember: phenotype not only refers to morphological/optical aspects, but also such as behaviour). The genuine, unhybridized aurochs individuals should always be the template for that. At the same time, individuals that cope well with the natural circumstances, are fully fertile, give birth to healthy offspring and apparently are resistant to diseases should remain in the population. Since people have to work with those cattle, tame, relaxed, less aggressive and less nervous individuals will of course be easiest to handle. However, one has to pay attention not to favour these individuals and therefore domesticate those aurochs hybrids once again (which would be rather ironic). One would have to regard them as wild bovines, keep and handle them like wild bovines. It works with wisents, bison and buffaloes, so it should work with aurochs as well.
After some generations, a viable population of aurochs would have been created. The likelihood of deleterious alleles becoming fixed homozygously has been drastically reduced by hybridization, and hybridization introduced a lot of new alleles for the resistance against diseases and cope with other obstacles, while artificial selection preserved the biological integrity of the wildtype. Furthermore, there would still be pure individuals to backcross with, even if you have to genetically produce them again. However, alleles for domestic mutations that would show in the morphology of the animals would remain in the population for some time unless they are effectively selected against. It would be comparable to the situation of many American bison, where you still occasionally see individuals with vestiges of hybridization with domestic cattle around 1900 (for example here), but invisible in the majority. This could include divergent horn shapes or colour variants, especially when the alleles are recessive.

The perception of resurrected aurochs

A lot of people, especially conservationists, will complain about those aurochs having domestic cattle introgression. They will argue that these aurochs are not “pure” and therefore not genuine. This exhausting mindset is the result of underestimating the role of hybridization in evolution and the importance of genetic diversity and might one day cost the wisent’s life as a species, a mindset that is part of the 19th and 20th century conception of nature as a static and “highly balanced” system to which we humans are the “destabilizing factor”, a conception that is not tenable for obvious reasons. Not only is hybridization a not as rare aspect of speciation as once thought (I plan to do a post on that), in this case we would have introgression into a wildtype from its domestic descendants. Hybridization between wild and domestic forms takes place everywhere they share the habitat, and is also evident for wolves, wild boars and also been proven for Holocene wild horses. Nobody would consider timber wolves or McKanzie wolves less “genuine” because their coat colour diversity is suggested to be influenced by ancient domestic dogs, or condemn wild boar populations because they share some domestic haplotypes – so why not accepting a population of genetically resurrected aurochs whose genetic diversity has been increased by controlled, systematic introgression of some of its domestic descendants?  

The idea of hybridizing those aurochs with domestic cattle in order to increase genetic diversity of course evokes the question which cattle would be best-suited to do this. As mentioned above, these cattle should a) be as healthy and genetically diverse as possible, b) as less-derived as possible in order to introduce as few derived or deleterious alleles as possible – two points that often go hand in hand. Probably it would be best to choose a handful of less-derived, robust and healthy landraces to hybridize with. Once again a balancing act has to be done. If one would hybridize only with a herd of a breed like Maronesa, the impact on the visible traits would be very low in the outcome, but the gain of genetic diversity would be narrow. So more than one breed has to be chosen, and from more than just one region. You can find suited cattle even in Asia in the form of the turano-mongolian breeds (see here). But importing sufficient numbers of cattle is very costly and effortful, which is also one of the reasons why many grazing projects and “breeding-back” projects use breeds that are already local.

And here is where already existing “breeding-back” cattle are important and which is why I think they would not at all become meaningless if genuine aurochs would one day be reconstructed. If such an aurochs project is to be set up on a large, semi-natural area somewhere in Europe outside the Mediterranean range, “breeding-back” cattle would be the most aurochs-like cattle available and since they are composed of a number of landraces, would introduce genetic material of a number of breeds. If such a project came about in Germany or the Netherlands, and one choose to use Taurus and Tauros individuals as a base, you would have the genetic material of at least eight breeds already (Heck cattle, Sayaguesa, Chianina, Maremmana, Limia, Maronesa, Pajuna, Highland). This also means that the range of undesired domestic traits is large, though, but it will be minimized by absorptive breeding and selection, just like in the case of bison with domestic cattle introgression.  

What to do with resurrected aurochs?

So if one finally has the desired viable population of genuine aurochs, what to do with them? At first, it might be wise to create a herd book for them, perhaps right from the beginning, like it has been done for a number of endangered species. This can be advantageous to maintain an overview over the genetic composition of the animals after they are spread to other breeding sites. Surely some of them can be sold to zoos for preservation and education. They can also be used for grazing projects and private breeding by aurochs fanciers, but I doubt that this will be a common use for genuine aurochs, especially the latter one. These animals would display wild behaviour and would be difficult and dangerous to handle just like wisents, bison and wild buffaloes. There is a reason why domestic cattle are used that commonly in grazing projects, but there are only a handful that work with wisents, although the wisent is a native wild animal here whose reintroduction should be preferred over that of a domestic proxy.  

The ultimate goal would be to not only create a population that can be considered “pure” or more or less pure in the eye of enthusiasts but also in those of the IUCN. A recognition of those reconstructed aurochs as aurochs and wild animals from an “official” organization would be a great help for the acceptance of these cattle as wild animals and legitimate their reintroduction into nature areas. Zoos would also help to spread and increase the population number, where they could also be used for education.




Tuesday 28 November 2017

Thoughts for a Fighting cattle project

In a previous post, I presented an aurochs reconstruction I did recently. I wrote that it matches 100% what I imagine a Holocene European aurochs to have looked like, and that there are obvious similarities to many Lidia bulls (Iberian fighting cattle). Of course there is the danger of a Pygmalion effect: I might draw my aurochs Lidia-like, because I consider Lidia aurochs-like. This cannot be ruled out completely. But I always try to follow the evidence and not to be guided by preconceptions when doing illustrations. In the post linked above, I explained why I drew the bull the way I did. Here is the illustration I am talking about:

Actually, the fact that Lidia is one of the least derived cattle breeds left today did not just come to my mind because of that drawing, but was apparent to me right from the beginning when I started to become interested in the aurochs. I use to explain the looks of the aurochs to people as “much like a big, long-legged and large-horned Iberian fighting bull”. Also, aurochs expert Cis van Vuure comes to the conclusion that this Iberian breed is the most aurochs-like breed in his 2005 book “Retracing the aurochs: History, morphology and ecology of an extinct wild ox”. And also Lutz Heck considered the breed very reminiscent of the extinct wild form and relied heavily on it in his breeding project (his lineage diminished, modern Heck cattle most likelyhas no influence from the Iberian fighting bull).

I am going to sum up the aurochs-like pro’s of Lidia cattle:
- Muscular, athletic body shape very reminiscent of wild bovines
- High processus spinosi in the shoulder region (“hump”), probably as large as in the aurochs
- Often elongated skull shape, straight to convex snout and often prominent forelocks, sometimes also on neck and face
- Horn shape reminiscent of that of the aurochs with an inwards curve, although usually not entirely as prominent as in the aurochs  
- Short dewlap and scrota and small udders in less derived individuals
- Wildtype colour plus sexual dichromatism (albeit reduced) present in the breed
- Robust landrace, used to live outside with little husbandry all year round

Especially the first four points are very advantageous for “breeding-back”, as those traits are rarely found in any other primitive breeds to the same extent. All of them give primitive Lidia individuals a very aurochs-like appearance, and I am going to link some of those less-derived aurochs-like Lidia individuals here:

Bulls:

Cows:

For a video of aurochs-like Lidia bulls, go here for example. Another video of bulls with an exquisite anatomy and also respectable horns is here.

It is not only their less-derived anatomy and looks that make Lidia special, but also their hardiness and independency. As I reported in 2013, the three Lidia cows in the German Lippeaue reserve were the only cattle that were never seen to make use of supplementary feeding.

Despite of their many advantages, this breed also has considerable disadvantages. First of all, Lidia are quite small cattle. According to the Domestic Animal Diversity Information System, bulls reach a withers height of 130cm, and cows 110, which is 10-15cm smaller than Heck cattle and almost half a meter smaller than an average Holocene European aurochs. But even more important, Lidia cattle are being bred for aggression and “fighting spirit” and consequently are very aggressive and nervous, difficult to handle and often attack easily. Additionally to these two major disadvantages, Lidia bulls also tend to be short-legged compared to aurochs.

But of course it is the task of crossbreeding and selection to eradicate such negative traits, just as too small horns or undesired colour schemes. This has been tried in the Lippeaue, where three pure Lidia cows have been used for breeding. But they were not really satisfied with the results. For once, the crossbreeds remained comparably small, even if they had portions of Chianina or Sayaguesa, two large breeds, in their genome. The crossbreeds also displayed overly nervous or aggressive behaviour. This goes for the half-Lidia bulls as much as those that were only quarter Lidia such as Latino. One half-Lidia bull was sold to Hortobagy, Larus (which looked quite good), and Istvan Sandor told me it was the most aggressive bull he ever experienced. Another half-Lidia bull was sold to Denmark and slaughtered because he was too difficult to handle. I had a look at many Lidia cross individuals in Matthias Scharf’s photo archive, and many of them did look more muscular than crossbreeds without Lidia, but they also tended to develop a rather elongated trunk and did not possess the slim athletic waist of pure Lidia. As a side note, even pure Lidia bulls often grow rather hefty at higher age, and the crossbreeds seemingly tended to go that direction right from the beginning. So crossing-in Lidia was not that successful as hoped, which is why the ABU diluted the influence of this breed in their herds, and Matthias Scharf even said that he considered most Lidia-crosses to be “small, ugly and mean”. Currently, they have one good half-Lidia cow in their herd that also has a comparably relaxed behaviour.
Larus, a cross between a Lidia cow and a Dutch Heck bull (photo by Istvan Sandor)
So does this mean that Lidia is, despite its obvious qualities, not a useful breed for “breeding-back”? The Tauros Programme, for example, will not use this breed, especially because of its behaviour. Considering the difficulties in handling Lidia crosses, and the disappointing looks of many of the cross results concerning size and body shape, I would say that Lidia is indeed perhaps not worth the effort in conventional crossbreeding and keeping them in grazing projects.
Lidia x (Heck x Chianina) cow
Nevertheless, I also think that it would be a shame not to make use of the potential that is in the breed. Maybe the crossbreeding results are too double-edged to speak of a “successful use” of this breed in crossbreeding, but perhaps a project working exclusively with good, useful individuals of that breed and trying to eradicate the negative traits of Fighting cattle would be fruitful. I am thinking of a project that would at first try to find a sufficient number of aurochs-like Lidia cattle such as those on the photos I linked above that have the right colour setting, an athletic body, elongate skulls and good horns, and trying to acquire individuals that are as large as possible, as long-legged as possible and perhaps not overly aggressive or nervous. For this, I would try to search breeding sites all over Spain and Portugal and also include the Casta-Navarra breed. Selecting them for couple of generations for the good traits and against the three undesirable Lidia traits (small size, short-leggedness, aggression/nervousness) might produce some good results. When selecting against their aggressive and nervous behaviour, I would try to pay attention to not domesticate them any further by selecting explicitly on tame and docile behaviour, as pleiotropic effects can affect other traits as well as we see in all domestic animals (see the Dedomestication series). The question is, then, how to handle the cattle. It would be very difficult to next to impossible to keep them like in grazing projects or private farms. It might be wisest to handle them just like other Lidia breeders do, or perhaps like American farmers keep their bison, also with the necessary equipment. After all, a true wild aurochs probably could not be handled much differently, and it is no coincidence either that most natural grazing projects work with cattle instead of wisent. Of course it would be more advantageous if their behaviour was suitable for grazing projects, as grazing projects provide the most abundant opportunities to spread the breed and keep the animals under semi-natural conditions in herds of sufficient size, but this post focuses only on a scheme to breed a very aurochs-like strain of Lidia and not how to spread it. If it is indeed possible to breed a more relaxed and less dangerous behaviour but maintaining the wild cattle-like body conformation, it would be pleasant of course and perhaps enable the use of these cattle in grazing projects. However, I think it is questionable if this is possible within a few generations as fighting cattle have been bred for aggression for dozens of generations.

The idea of establishing an aurochs-like cattle population using Iberian Fighting cattle only is not new. Lutz Heck considered this breed almost indistinguishable from the aurochs, so that it is surprising that he still bred with other breeds as well. Cis van Vuure also proposed a project working with the most primitive Lidia cattle only in order to achieve maximum resemblance to the aurochs. I also remember that in email conversations it was once considered to put such a project into practise, but as several years passed now and I discovered no mention of it on the web I assumed this idea has died a silent death.

Also, I have my doubt that it is possible to achieve all aurochs-like traits using Lidia only. While some Lidia grow larger than the norm, I would be very surprised if selective breeding is able to raise the size range from 130cm for bulls to 160cm at least within a couple of generations. The same goes for horn size – while Lidia are not small-horned, the largest horns of this breed are still only at the lower size range of aurochs horns, perhaps still a bit too small when considering the size of the horn sheath. Also, the breed might need a boost in leg length, even after selective breeding. Furthermore, the problem with breeding for quantitative traits is that it takes comparably long and other traits might get neglected, including also genetic diversity (Lidia is already inbred).
Thus, crossing-in breeds that might help to overcome these deficiencies and also to increase genetic diversity is suggestive. The first option that comes to my mind are Chianina or Maltese cattle that would add the large size, slender posture and long legs. Maltese cattle have the advantage of introducing colour dilution alleles to a lesser extent and having longer snouts, but are far less available, with not even hundred individuals and only on Malta. The big disadvantage both breeds have are the very small horns. The horn size of Lidia already has to be improved (although some individuals, like these young bulls, have comparably large horns), and crossing-in one of these breeds would introduce alleles for very small horns, and since horn size is probably controlled by several gene loci it would be even more difficult to breed for the desired horn size. So it would be advantageous to add another breed to increase horn volume. Which breed would be best to do this job? One idea would be very large-horned Texas Longhorn with the right colour setting. However, the disadvantage would be that they are small-sized as well and have very outwards-facing horns. Another possibility are large-horned Heck cattle from lineages with a comparably stable inheritance; their disadvantages are their very domestic, heavy body and paedomorphic skulls. Using Watussi would greatly increase horn size, but also introduce many undesired traits concerning body shape, dewlap, colour, fleshy hump, skull shape et cetera, which is why this breed would have to be used wisely and with patience. If one cannot decide among existing stable breeds which one to chose in order to increase horn size, one could wait for other projects that are already in progress now to produce results that can be used for this job. Such as the Auerrind project, which has just started and is using Watussi. Perhaps in 15 years, when also a Lidia project would see first results, there might already be good Watussi-influenced Auerrind cattle that could be used in order to increase horn size in an improved Lidia strain without introducing too many undesired traits. This is just brainstorming – in practice it depends on many factors of course.

I would not suggest to start crossbreeding right from the beginning and to a larger extent, otherwise I would not be talking about a Lidia project here. It is about preserving, concentrating and improving the aurochs-like traits present within the Lidia population, and using other breeds only after a phase of selection when selective breeding has been proven to not being able to reach body size, leg length and horn size to a desirable degree. Another positive side effect of crossing-in other breeds would be that at least some individuals might display a more relaxed behaviour. I would use other breeds only carefully and perhaps use only F1 crossbred individuals in order to not diminish the advantageous Lidia traits. The point at which I would start to make use of such crossbreeds is when selective breeding has been shown to not reach all traits to a desirable extent, which would be after a couple of generations and therefore about 15 years or so (which would equal about five generations).

How could such an aurochs-like Lidia strain be named? For Maremmana, some people like to speak of the optically less derived individuals or herds of “Maremmana primitiva”, although those Maremmana primitiva do not represent one gene pool or breeding line. In the case of this Lidia lineage, on the other side, it would be one population/lineage. So perhaps calling them “Lidia primitiva” would be an idea.
I think that such a Lidia project could achieve a high level of authenticity regarding looks and anatomy and perhaps also behaviour (yes, exaggerated aggression is not “natural” behaviour; but tameness and docility are not either; you certainly could not have used an aurochs for draft work or have milked them). There is also the chance that with selection against exaggerated aggression and crossing-in of other breeds Lidia primitiva might once have a behaviour that is suitable to be used in grazing projects and other aurochs projects.

This post is just a thought-experiment, but I hope that it inspires and perhaps other people have similar ideas in order to make use of the potential we find in Spanish fighting cattle.



Saturday 18 November 2017

Wild horses pt. IV: The Przewalski's horse - is it still a wild horse?

This title is surely a bit provocative – it is of course zoological consensus that the Asiatic Przewalski’s horse, Equus ferus przewalskii, is the last living genuine wild horse that is extant today after the last western wild horses disappeared. However, advocates of a number of horse breeds that they purport as living remnants of the European wild horse or at least being strongly influenced by original wild horses, sometimes put the status of the Przewalski’s horse as a genuine wild animal into question. They argue that decades of breeding in captivity and introgression from domestic horses has altered the nature of the Przewalski’s horse, and claim that the situation is comparable to what has happened to their favoured “wild” breed: an original wild population has been influenced by domestic horses and artificial selection. As a consequence, they argue that if the Przewalski’s horse still deserves status as a genuine wild horse, which is zoological consensus, then so does their breed of choice. However, the often purported background stories for those breeds being “near-wild horses” are not tenable after objective examination for a number of reasons (see here), but what about the arguments against the original, wild status of the Przewalski’s horse?

The lineage of the Przewalski’s horse separated from that of domestic horses several millennia ago. The exact point of separation varies from study to study, depending on the molecular chronometer and its calibration. The maximum I found was 160.000 [1] and the minimum 38.000 [2] years ago. This comparably long reproductive separation resulted in a different karyotype, the Przewalski’s horse having 23 pairs of chromosomes and the domestic horse 24 due to a fission or fusion (depending on what is the plesiomorphic state), but they intermix readily and without fertility problems. In the millennia of living side by side in the Eurasian steppe, the Przewalski’s horse contributed genetically to the domestic stock (which is not only genetically [3] but also optically apparent, f.e. see some Mongolian horses), and vice versa. The Przewalski’s horse gene pool was introgressed by domestic horses, especially in the 20th century. Photographs of wild herds from 1954 showed individuals of divergent colours (Wikipedia), indicating admixture. The whole modern population descends from 13 founding individuals, one of them was a domestic Mongolian stallion [4]. Does this mean that the original, genuine Przewalski’s horse is lost and the modern population is an altered result of intermixture?
Orlando et al. 2015 made a genomic study compromised of a large sample of Przewalski’s horses, post and prior to the bottleneck (including the holotype specimen), domestic horses plus a late Pleistocene wild horse as outgroup. The result is that although there are genetic traces of intermixture, also including such having an effect on the phenotype such as an allele associated with increased withers height, there are still lineages in the population that are virtually free of admixture[3]. Also, height is a highly multifactorial trait, therefore it cannot be claimed that Przewalski’s horses are taller now due to admixture because of one allele – the average withers height is still between 122-142cm according to English Wikipedia, between 120-146 according to German Wikipedia (note that there is also sexual dimorphism in size). Przewalski’s horses are still uniform in their typical colour, sturdy build, robust head shape, erect and short mane, short-haired tail basis, a very lightly coloured almost white winter coat, and other typical morphologic differences to domestic horses such as thicker hooves (Wikipedia). I would even say that domestic cattle left a bigger trace on modern American bison than domestic horses did on the Przewalski’s horse, yet nobody is questioning the bison’s status as a wild animal. Also I found no source stating that Przewalski’s horses with a domestic karyotype have been observed.

Yes, the Przewalski’s horse seemingly intermixed with domestic horses continuously after their point of separation, but I see no compelling evidence that this fact altered the genetic integrity of this wild subspecies. Furthermore, domestic animals introgressed the gene pool of their wild counterparts everywhere they shared the habitat – this evident in European wild boar that show deviant, domestic colours and there is also the hypothesis that American wolves inherited black and other colour variants from domestic dog introgression several millennia ago (this might also explain blue-eyed wolves). Yet nobody is calling their wild animal status into question.

If Przewalski’s horses indeed lost part of their wild animal nature due to domestic introgression and being bred in captivity for a number of decades, it might be helpful to look at a checklist of aspects typical signs of domestication:

- Morphological paedomorphy
- Behavioural paedomorphy
- Reduced brain volume
- novel morphological/optical traits (very typical: colour variants, particularly white spots)
- Earlier maturity and increased litter size (the latter aspect is not true of domestic horses either, so let us ignore it for now)

Przewalski’s horses do not show any signs of morphological paedomorphy, not even if you compare photos of the early 1900s to modern individuals. Przewalski’s horses still always have the robust, donkey-like skull with small eyes and their proportions do not seem to be altered as well. I have not found any remarks in the literature stating that Przewalski’s horses as a whole lost brain volume; domestic horses have about 14% less brain volume than Przewalski’s horses [5]. Captive Przewalski’s horses also have 14% less brain capacity than wild counterparts [5]. Since there are no separate genetic lineage between wild and captive Przewalski’s horses, this should be applicable to phenotypic plasticity. According to Wikipedia, earlier maturity in captive Przewalski’s horses has been reported, but explained with better nutritional conditions in zoos than in the wilderness and as far as I know the same phenomenon can be observed in other zoo animals. The behaviour of Przewalski’s horses and domestic horses is well comparable, but Przewalski’s horses have a way higher aggression potential than domestic horses, especially the stallions. This is universal for this subspecies and evident in zoos as well as grazing projects. I once was told that zookeepers are more afraid of Przewalski’s horses than lions. Przewalski’s horses can be tamed and ridden to a certain degree, but this is also true for zebras, including the quagga.
There are not any novel traits found in any Przewalski’s horse, such as a new colour variant, or fur modifications. There are occasionally individuals showing a white streak along the face or white socks, which is applicable to introgression from domestic horses.
Looking at some deer populations which have been kept in game parks for many generations, we see incipient signs of domestications, such as new colour variants or typical domestic spotted patterns, or beginning paedomorphic skull shapes – you can find this in some roe deer, red deer and fallow deer in European game parks and this is what I would call an early state of slow domestication. But we do not see that at all in Przewalski’s horses.
All in all I think there is not one compelling reason to claim that the original Przewalski’s horse is gone, that it has been altered by man and hybridization, or that it is on the edge of domestication. I see nothing that calls their status as a genuine wild animal seriously into question, especially when we look at other wild animals. And even if the critics were right, it would not make any of the domestic horse breeds praised as near-wild horses “wilder” than they are (or not are, actually).

Literature

[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] Orlando et al.: Evolutionary genomics and preservation of the endangered Przewalski’s horse. 2015.
[4] Bunzel-Drüke, Finck, Kämmer, Luick, Reisinger, Riecken, Riedl, Scharf & Zimball: „Wilde Weiden: Praxisleitfaden für Ganzjahresbeweidung in Naturschutz und Landschaftsentwicklung“. 2010
[5] Röhrs, Ebinger: Are zoo Przewalski’s horses domesticated horses? 1998.