When is a Horse a Horse? The Species Definition Problem

In my introduction to the origin of horses I suggested that a horse is a horse of course, unless of course it isn’t a horse. But that begs the question, how do we know what the boundaries are of what constitutes a horse?  Although I suggested that young-earth creationists (YECs) have diverse views of what constitutes a horse “kind” I did not intend to suggest that defining any taxonomic category whether species, genus or family is easy.  In fact biologists and philosophers have debated the definition and meaning of species for the past century.

In a general biology class in high school or college you probably learned that species are “groups of actually or potentially interbreeding natural populations, which are reproductively isolated from other such groups.”  This is called the biological species concept (the BSC) and was proposed in its modern form by Dr. Ernst Mayr more than 50 years ago.  This definition is not without its detractors in part because it only best applies to higher animals and doesn’t work so well for most other life-forms on earth.  I teach an upper-level biological systematics course in which my students and I spend upwards of three-weeks discussing more than a dozen definitions of species and their philosophical roots.  Students in that class hopefully leave with an appreciation for the fact that the exact definition of a species—if one is even possible—for all organisms is far from clear.  Part of the problem is that identifying where one species starts and another ends can be very difficult and this difficulty extends to all taxonomic categories.

Regarding horses, or equines more broadly in this case, the biological species concept (BSC) can be practically applied to define modern species of equines.  Horses and donkeys can interbreed to produce mules but mules are sterile providing evidence that donkeys and horses are “reproductively isolated” from one another.  Horses and zebras can interbreed and produce offspring despite having different chromosome numbers. However there is typically an added caveat of the BSC that is usually not stated in the simple definition and that is that interbreeding “occurs in natural conditions.”  Yes, a horse and zebra can interbreed, but in nature even if they are in close proximity to one another interbreeding is an exceedingly rare circumstance and therefore the horse and zebra genetic pools remain genetically distinct over time maintaining them as separate species.

Another example would be coyotes and wolves, they can interbreed and there have been cases of natural hybrids, however, these two “species” have lived in the same geographical areas for thousands of years and they generally won’t associate with each other in nature due to multiple behavioral differences.  As a result they are effectively isolated from one another and thus maintain persistent genetic differences.

So a horse is a horse because it recognizes itself as a horse and won’t, except under very unusual and unnatural conditions, interbreed with other animals that are not of similar morphological and genetic stock.  Most of the 20 or more species concepts that have been developed to define the limits of species would recognize Equus ferus (wild horse + domesticated horses) as a species.  But clearly, these horses are more similar to zebras, quaggas and donkeys than they are to kangaroos and bears.  Those similarities suggest that though zebras, donkeys and horses are each distinct species today they likely share a common ancestor in the past.  Even many YECs now appear to agree these species are all equines and share a common ancestor.

These equines illustrate why the biological species have been so difficult to define.  If there multiple species of equines alive today but they were all derived from a single common ancestor in the past, during the time that this single ancestor was “evolving” into seven or eight recognized living species there would have been points along the way in which the distinction between one and then two and then three species would have been difficult to pin down exactly.  This is because they would go through periods where they may have become isolated but hadn’t yet developed any significant genetic or morphological differences.

We can see this process of speciation playing out before in the present.  Today there is named subspecies of horse (Equus ferus) called Przewalski’s horse. These are the only truly wild horses as opposed to feral horses like the mustangs of the western US which are domesticated horses that have escaped back into the wild.  Przewalski’s horses are not easily tamable and so truly wild. They are genetically (see below) most closely related to domesticated horses and can produce fertile offspring despite having one more chromosome than domesticated horses.   And yet, interbreeding in nature is very rare and thus genetic exchange, especially given the different chromosome numbers, doesn’t really happen. Over time this genetic difference between domesticated horses and Przewalski’s horses will increase but on what future day will we call these two separate species?  Some might call them species today but most taxonomists agree that they don’t yet deserve species rank.  What is clear though is that segregated gene pools are in the process of becoming more and more genetically isolated and at some point will have

obtained the types of mutation that will either prevent fertile offspring or will alter them behaviorally  (think of coyotes and wolves) so that the opportunities for future genetic exchange are reduced even further. Each newly gained bit of difference becomes a step toward the formation or “creation” of a new species.  If we look back in time, at some point zebras were probably just an interesting variant within a larger population of animals most of which did not become zebras.  They are like the Przewalski’s horse but at a time slice in the past we may have called all the species of zebras just one species with subvarieties.  Going back further in time there may have been a time when there was only one clear species will only scattered variation that had not yet coalesced into clumps of genetically and morphologically distinct subspecies.  So even if we define a species today, looking back through time that species will not always be as distinct as it is today.

How different are equine species today?   

I can illustrate how genetically distinct horses are from other “horses” in the figure below. I generated this genetic similarity tree by comparing complete mitochondrial genomes (just over 16,000 total base pairs for each sample) between hundreds of breeds of horses and other species of living equines.  The figure below is a comparison of total genetic similarity of this genome with the branches/lines representing the total genetic differences between a particular sample and all other samples.  It is hard to read the names but I have labeled the general groups. What you see is that all of the domestic horse breeds are remarkably similar genetically and thus very closely related to one another. The other “species” of horses are genetically distinct from horses and each species is nearly as different from one another as domestic horses are from those other species.   This is similar to what one would observe with domestic dog breeds vs. wolves and coyotes and other canines.

You can see that “wild horses” which are Przewalski’s horses are not much different from domesticated horses but as I said before, they are truly wild and so behaviorally differentiated from other horses.   There are three accepted species of zebra each of which is genetically distinct from one another.  While domestic horses have what appears to be considerable visible variability that variability is only skin deep. The zebras are far more genetically distinct from one another than any of the domesticated horses even though visually you may think they look very similar.  Yes, looks can be deceiving.

Preview of Part III: In Search of the Equine Common Ancestor

In the next part of this series on the horse we will look at what happens when we extend the idea of equine ancestry to the fossil record of horses. I have mentioned that YECs don’t seem to have a problem of all living species of horses having a common ancestor likening them all to be in the same created kind.  But when the fossil record enters the story things start to get a little more confusing for YECs.  How far can we push species concepts (or in the case of creationists the kind/barmin concepts) to accommodate greater and greater divergence of characters as being associated with a common ancestor? As we will see, YECs disagree about how far these boundaries can be pushed resulting in a diversity of opinions about the rate and mechanisms of evolution that are needed to explain present day species.