This post first appeared on November 9, 2011
It has been exciting to see that the recent paper on horse color and cave art has gotten a lot of attention in the mainstream press. It is great to see scientists like Rebecca Bellone, the lead researcher from The Appaloosa Project, recognized for their work. I also love the idea that an area of study traditionally connected with agricultural and veterinary science could be used to better understand seemingly unrelated fields of archeology and art history. It makes sense that in understanding the horse, whose history is so intertwined with our own, we gain insight into ourselves.

That is the larger picture. From the smaller picture that is the focus of this blog, the study offers some big surprises.

To adequately explain, I’ll need to expand on the comments that were made in the earlier post on gene locations. In 2009, the paper “Coat color variation at the beginning of horse domestication” was published. In that study, ancient remains were tested for the presence of color mutations. The range of tests available at the time included:
Extension (black/red)
Agouti (bay/black)
Cream
Silver
Frame
Sabino1
Tobiano

Those tests determined that all but the frame gene were present among the early domesticated horses. That is certainly in keeping with the theory that frame is a New World mutation. It also showed that in the wild populations – horses living somewhere between 15,000 and 3,100 BC and predating domestication – the only mutation was black. Black horses were found among the wild populations in Romania, Ukraine and the Iberian Peninsula. The other populations, which included remains from Siberia and Germany, were entirely bay.

The two patterning genes, Tobiano and Sabino1, were found in the remains of domesticated horses. That is in keeping with the idea that spotting mutations are linked to selection for tameness. The Russian Farm Fox study is often cited as a good example of this, but most people familiar with newly introduced “pocket pet” species have seen this in action. It usually doesn’t take long after a species becomes popularized before spotting patterns begin to appear.

What makes the cave painting study so fascinating is that the appaloosa patterning gene was found in a wild population. And it wasn’t just one horse. Of the thirty-one samples, six were carrying the mutation for leopard complex (Lp). Were someone to assemble a random sampling of modern domestic horses, it would be unusual to say the least to find that kind of ratio of appaloosas to non-appaloosas.

Also interesting is the fact that while there were six leopard complex horses, there were no chestnuts. Chestnut is found in the Przewalski Horses, where it has been documented as far back as the early twentieth century in skins taken in Mongolia. In the cave art study, there was a single Romanian sample that tested as carrying chestnut, so the mutation did exist at least in that population.  It would seem to be rare compared to leopard complex, and not nearly as old. The German samples with leopard complex date between 15,000 and 11,000 BC, whereas the Romanian with the chestnut allele is 4,300 BC. This is interesting when one considers how in many primitive European pony breeds, chestnut is non-existent, or when found is considered proof of foreign influence. It also gives a certain level of credibility to claims made by both Gypsy Cob and British Spotted Pony breeders that appaloosa coloring was once part of the native population.

It is also interesting that this mutation occurred in a wild population, and was obviously perpetuated, despite the fact that homozygous leopard complex horses have a defect. They are blind in low-light situations, which should act as a negative selection factor. None of the horses tested as homozygous for leopard complex.

Limitations

As exciting as the results of the study are, some limitations have to be remembered. Probably the most important is that this was a really small sample set. Getting usable genetic material from ancient remains is difficult, which is why there are only 31 samples. Broken down by time frame and geographic location, you end up with even smaller groups. These tests can certainly confirm that a mutation was present, but it is hard to draw any firm conclusions about the whole of the ancient horse population based on so few animals.

We also only have a partial picture, because we only have a partial set of color tests. The previous study, done in 2009, used some of those same samples. Without the leopard complex test, which was not yet developed then, we only knew that the wild horses were bay or black. With the new test, we now know that, yes, they were bay and black – and some were varnish roans (leopard complex). We don’t yet know if they had the patterning genes that turn leopard complex into true leopard patterns, though certainly the cave paintings would suggest that this was so. Likewise, we assume that the original horses were dun since that coloration is associated with wild equines, including the last remaining wild horse. A completely reliable test for dun is not yet available, so that part of the picture is incomplete as well. Those bays and blacks, now known to in some cases be bay or black varnish roans, may later prove to be dun and grulla varnish roans – or not!

We know they were not silver or cream, since those can be and were tested. But as new color tests are developed, we may later learn that some of those horses were also roan or grey or splash. It may be that varnish roan will eventually lose its place, but for the moment it is the oldest tested pattern.