Tag Archives | true roan

A family of diverse colors


It was perhaps a bit rotten of me to bring up the tobianos and dark-headed roans when talking about the confusion about what to call horses with sabino patterns. Instead of saying, “No, this really is pretty simple,” I opted to point out that it is even more complicated. Now that I am feeling a little less mischievous, I probably should attempt to clarify things a bit.

I began this (meandering) train of thought with a post about the change from Mendelian genetics to molecular genetics. That is, a shift from analyzing colors using visual identification and statistical analysis to understanding colors based on the changes to the genetic code. Using the first method, colors were grouped a certain way that has become familiar to many horsemen. When the colors are grouped according to the gene where the mutation occurs, however, they sort a little differently than expected. Colors that look quite different – colors that aren’t even thought of as belonging to the same basic category of modification, like tobiano and dark-headed roan – can be mutations to the same gene.

The technical term for this situation is allelic heterogeneity. In plain English, what that means is that there are a number of different options for one gene. In the case of these particular colors and patterns, the gene where they are found is called KIT.  Tobiano, roan, sabino and dominant white are all alleles at KIT. As different as they look from one another, they can be thought of as belonging to the same family. This may seem a bit esoteric, but it has a couple of implications for breeders.

The Spotted Saddle Horses pictured at the top of this post display a type of tovero pattern that is very common in their breed. The ragged, torn outline of their spots is typical of what happens when tobiano is paired with Sabino1. It is a compound heterozygous pattern. That is, both copies of the KIT gene have a mutation, but they are different alleles. If that same horse had two copies of tobiano (two of the same allele), we would call him a homozygous tobiano. Instead these horses have one tobiano and one Sabino1 (two different alleles for the same gene). A horse has two copies of a given gene, but they only get to give one of them to each of their offspring. So like the homozygous tobiano, if they were bred to a solid horse all their offspring will be pintos, but only half will be tobiano. The other half will be sabino.

Bred to solid mates, half the offspring of the toveros above should have this kind of pattern – Sabino1.

Allelic relationships like this are important to breeders because it means that under most circumstances, the patterns that result from combinations of alleles are not going to breed true. That might not be important if all that matters is that the resulting foal have a pinto pattern, because a compound heterozygote is going to produce a patterned foal 100% of the time. But if a breeder wants to duplicate the original combination, that might matter quite a lot. And if the other “pattern” is something that would not qualify as a pinto, like dark-headed roan or one of the more minimal versions of sabino, then the 50/50 nature of the inheritance might be a problem.

Breeders have noticed that some combinations, like tobiano roan, are difficult to get consistently. That is because this same splitting of the two alleles occurs; the horse can only give one but not both, so the only way to repeat the combination is for the other parent to contribute the second allele. The fact that some of these alleles look so different from one another makes the relationship between the colors less obvious. Knowing why Sabino1 toveros do not produce their own color when bred to a solid mate allows breeders to pick crosses that stack the deck in their favor. (A cross to the same Sabino1-tobiano combination, for instance, would give the desired pattern 50% of the time.)

The connection between these seemingly different colors might also make it easier to understand some of the quirks within some of these patterns. One of the most common questions I get from breeders of tobianos is about roan patches, or roaning in the colored areas of the coat. It is a relatively common occurrence in tobianos, and it often causes breeders to inquire if their horse might carry some kind of sabino pattern. In many cases, it appears that the roaning is just part of the tobiano pattern itself.

Dexter has diffused roaning throughout the dark areas of his coat, with somewhat greater concentrations of white hairs around the borders of his spots

Here the roaning is mostly limited to one patch, though colored specks remain inside the roaned area

When tobiano is understood to be a mutation to the same site as both roan and sabino, irregularities like these seem less surprising than when tobiano is thought of as something wholly separate. Likewise, the idea that tobianos might be more prone to white on the face than solid horses seems less outrageous. Tobiano is related to a whole group of patterns that can quite rightly be described as doing just that, after all! (For newer readers, more on my scandalous views on tobiano face white can be read here and here.)

In fact, knowing that these colors are alleles of the same gene is useful because it encourages us to think about them in a different way. If we know that KIT mutates frequently, giving a surprising number of white and sabino variations, what about roan? Roan has proven problematic when it comes to testing, which suggests there is more than one version of the color. It is also true that there are quite a few instances of spontaneous “roans” in a variety of breeds. These have been dismissed in the past as not “true roan” because they came from non-roan parents. But what if they are just one of many roan mutations? And what about the various forms of white ticking, like rabicano and salpicada? Are they roan variants on KIT, too? Given what is known about the white mutations, that seems like a reasonable theory.

Taken as a group, many of these colors and patterns blend together with a lot of overlapping traits. Which brings me back to the original question, which is what to call them all. I’ve skipped over the more pressing problem of sabinos and dominant whites in this post, but I wanted to highlight the connection between these different colors and introduce the idea of compound heterozygosity. It is an idea that is pretty important to the situation with the sabinos. I had hoped to wrap this subject up with just one more post, and start posting some less in-depth topics, but it is probably obvious why I have avoided posting about this before. It is not a subject that lends itself well to brevity! So next up, the other group of KIT mutations and some ideas about what to call them. I promise, eventually I will get back to some easier topics!


Continue Reading

So what happened to all the sabinos?


As I mentioned in the previous post, sabino was once a catch-all term for a group of patterns that had a pretty broad range of expression. The first use of the term in an English-language book on horse color that I have found was Reiner Geurts’ Hair Color of the Horse, published in 1977. (The original Dutch book was published four years earlier.) Although it appears that Geurts took the term from an earlier Dutch paper written by J. K. Wiersema, he does state that it is an American term. I suspect that  by this he may have meant South American, since it is a Spanish word that does not appear to have been widely used in the United States until the late 20th century. This is how Geurts characterized sabinos:

[The term refers to] a kind of piebald with often indistinct and irregularly bordered white radiating from under the breast, the belly and the extremities, and thence spreading laterally up the ribs and sometimes to other parts of the body. The rest of the coat is often ticked with white, or roan.

In 2005, scientists identified the mutation responsible for one of the sabino patterns. In their paper, the authors’ criteria for sabino was consistent with Geurts’ definition:

Horses characterized as having a Sabino white spotting pattern had three of four of the following characteristics: (1) two or more white feet or legs, (2) blaze (white patch extending the length of the face), (3) jagged margins around white areas, and (4) spots or roaning in the midsection.

The identified pattern was named Sabino-1, because the researchers expected to add more patterns to the series. From the same paper:

Consequently, we anticipate future reports for other genes encoding different Sabino phenotypes (possibly SB2, SB3, and so on)… The variant of KIT described in this study can explain some but not all the phenotypes described as “Sabino.”

That last sentence proved to be true. The resulting test is useful for only a small percentage of horses with a sabino pattern. So what happened? Where are the additional sabino patterns that were predicted?


The simple answer is that many sabinos, like the two Paint Horses pictured above, have patterns that have not yet been formally identified. That is, the mutation that causes them has not yet been found, which is necessary for the development of a test. But it is also true that along the way, the approach to naming the different KIT mutations seems to have changed.

Around the same time that Sabino-1 was identified, researchers began looking at a suspected Dominant White family in the Franches-Montagnes breed. Just like other researchers in the past, they indicated that the mutation did not always produce a completely white horse. The paper on Sabino-1 noted that these horses fit the phenotype for sabino, but that the pattern of inheritance matched the one described in the original studies of Dominant White. In 2007, a paper was published identifying the Franches-Montagnes mutation along with three others. These were named White-1 through White-4. Because these particular mutations consistently produced white or near-white mutations, and because the patterns appeared to be homozygous lethal, it made sense to categorize them as Dominant White and not Sabino.

It should be noted, however, that the original paper outlining those first four Dominant White families did anticipate the possibility that future dominant white (W) mutations might not necessarily be lethal. In fact, it was not clear if all four of the initial mutations were lethal. Nonetheless, some saw the traditional distinction between the non-lethal Sabino patterns and the homozygous lethal Dominant White as significant for determining the appropriate category for a given pattern.

This became important when some of the patterns identified in subsequent studies more closely resembled Sabino than White. This was particularly true of the fifth White family, which originated with the Thoroughbred stallion Puchilingui. Puchilingui had what would have been called a “sabino roan” pattern, but his offspring were often loud, patchy sabinos.


Sato, pictured above, has what many would consider a typical pattern for a horse with the W5 mutation. It seemed that most horses that inherited this particular Dominant White mutation had a sabino phenotype. There were, however, some members of the family that could be described as white. This is the white Puchilingui daughter Shew.


With the identification of the W5 mutation, some began to refer to all highly-contrasted, patchy patterns as Dominant White while others (including many breeders) continued to prefer the term Sabino, even for those families with known mutations assigned a number in the W series. The confusion was increased with some of the later additions, particularly the two most recent Arabian mutations, W15 and W19, which did not seem to produce white phenotypes – or did so only very rarely. Would there be any new sabino patterns? Was the single incompletely dominant pattern likely to remain the first and only named Sabino pattern?

This question came up again when it was discovered that the reason some of the Puchilingui horses were white was that they carried two different KIT mutations. In addition to the already-named W5, white horses like Shew had another mutation that was initially overlooked because it was not thought to significantly alter pigmentation. It would appear that on its own, the W5 mutation produced what would have been called Sabino. Loudly marked, but still Sabino. It took adding a “quieter” pattern to produce the white or nearly-white phenotype.

The additional mutation was said to be “common and widely distributed”, and to have a “subtle white-increasing effect”. Of the 145 horses in the study, 52 were found to carry this particular mutation. Although only a handful of tested horses have published photographs, those that have been identified appear to be minimally-marked sabinos of the “flashy white markings” variety. In assigning the name W20, the authors gave the following reasoning:

We termed the new KIT variants W18 – W20 to provide a simple and unambiguous nomenclature for future genetic testing applications.

Some have interpreted this decision to mean that the name Sabino has been abandoned. It is an understandable conclusion, if horses like these two are not going to pick up the numbering in the Sabino sequence.


Yet perhaps overlooked is that in that same paper, the term Dominant White was also dropped. The previous seventeen mutations in the W series are referred to as Dominant White, but that phrase appears in scare quotes. Instead, the patterns are referred to as “White-spotted”. Sabino appears as a descriptive term (“sabino-like”) for one of the mutations described in the paper. The approach appears to be to number the new mutations to KIT consecutively in the W-series, regardless of the phenotype. That means that the “official” notation for the mutation may not correspond with the visual appearance of the horse. Given that diverse colors like tobiano and dark-headed roan have been found to share this same molecular location, it might be helpful to develop additional terminology that communicates visual differences that breeders find significant. From a molecular standpoint, dark-headed roan is one type of depigmentation, and is therefor a form of “white spotting” (and has been referred to in this way by scientists). A breeder interested in a predictable amount of white (either enough or not too much!), however, might have quite a different idea of what is and what is not white spotting.

It may be that phenotypes that fall clearly into existing groups, like roan (Rn), will retain their unique names. Unfortunately the appropriate category for many of the resulting colors – including some of the roan phenotypes as well as those factors that produce “ordinary” white markings – is likely to be less clear. While adding to the list of W mutations might provide “clear and unambiguous nomenclature” for tests, breeders and owners may find that they need additional names and categories to effectively communicate with one another.

Continue Reading

Color and pattern variations


As I mentioned in the previous post, one of the things that attracted me to the original Sponenberg book was that it proposed a system for categorizing colors. In later books, the author revised his system from one that grouped things by point color to the more familiar base colors and modifiers that most of us use today. Not only did that system make it easier to explain coat color inheritance to others, but it promised a common language so that horsemen involved with different breeds might find it easier to communicate with one another about the subject. Needless to say, a lot of us who talked about the subject spent a lot of time explaining that no, there was no “Arabian chestnut” that was different from, say, “Morgan chestnut”. Chestnut was chestnut. Color from breed to breed was more alike than it was different.

In some sense, that is still true. We know that the vast majority of chestnut horses, like the American Curly Horse above, carry the same mutation to the MC1R gene. Somewhere in their distant past, chestnut horses all trace back to a common ancestor – the founder – that carried that original change. Although this “common ancestor” story has made news lately, especially when it comes to hair and eye color in people, as I mentioned during the discussion about founders, that is actually what would be expected. It would be expected that grey horses would trace back to the one horse that carried the initial mutation for grey, and that tobianos would trace back to the original tobiano. The question is, would this happen more than once? Could their be multiple grey mutations? Is there more than one tobiano?

In the case of chestnut, it turned out there were at least two. In 2000, another form of chestnut (ea) was identified in the Black Forest Horse. Technically the ‘new’ mutation does not cause the horse to produce red hair; horses with this version have the original mutation for red. They just also have another mutation that interferes with the test used to detect the original mutation. What is important, though, is that the two forms are functionally equivalent. That is, horses that carry the common mutation (e) and those that carry the more rare form (ea) do not look different from one another. They both can be light or dark in shade, or have pale or self-colored manes and tails. Whatever variation there may be from one chestnut to the next, it would seem the controls for those traits are to be found elsewhere in the genetic code.

A good example of functionally equivalent mutations can be found in dogs. There are three known variations that produce chocolate in dogs – bS, bd, and bc. Given the isolated nature of the different dog breeds, it is thought that there may be other less common mutations as well. Like the two chestnut varieties in horses, though, these mutations are not thought to produce a visually different outcome. From a testing standpoint, the versions of chocolate are slightly different, but the results are the same. For that reason, test results in dogs are usually reported simply as b, without indicating the specific version. (I should note that as far as I know, the different forms of chocolate in dogs are independently arising mutations that are similar, rather than the mutation-then-another-mutation situation with chestnut in horses.)

This Boykin Spaniel falls on the darker end of the spectrum for chocolate, but not quite so dark as the German Shorthair pictured in the earlier post on diluted dogs.

This Australian Shepherd illustrates the paler end of the shades of chocolate. It does not appear that shade varies according to the type of brown mutation present.

The idea that there are equivalent mutations that produce the same visual result makes testing for colors more complicated because it can produce confusing test results. Prior to the discovery of the alternate form of chestnut in horses, it was possible to get test results back on a chestnut that indicated the horse was bay or black. So far the spread of that particular mutation has been fairly limited to some of the European draft breeds and some of the rustic Spanish ponies. (The latter are thought to have the mutation from more recent introductions of draft horse blood.) It is not hard to imagine a situation where an equivalent mutation might be more widespread, however, leading to confusing test results.

Even greater confusion might result if you had mutations that already produced a pretty broad range of outcomes. The visual range of a chestnut horse, or a chocolate dog, is pretty subtle compared to the range of what we have called sabino in horses. Both of the horses below have patterns that have, in the very recent past, been called “sabino”. Not many horsemen would think that two mutations producing patterns of white this different from each other could be thought of as “functionally equivalent.”



To confuse things further, there are mutations to the same gene that have not, by tradition, been put in the same general category. It is likely that this horse, a true dark-headed roan, has a mutation to the same gene as the sabino horses above. Again, this is not something most people could think of as equivalent. Tobiano, the pattern many horsemen think of as the  “opposite” of sabino and the other patterns in the overo category, also alters that same gene (KIT). Meanwhile the splashed white patterns, which are thought of a single category, include mutations that occur in two different locations (MITF and PAX3). So things that look completely different can be mutations to the same gene, while things that look alike can be mutations to different genes. 


And that is the challenging part about where things stand in color genetics at the moment. Advances in molecular research are expanding our understanding at a phenomenal rate. In just a short time, more than thirty white patterns have been formally identified, and we know there are many many more. We know more about how these different patterns are related than ever before. The downside is that what has been discovered does not exactly match up with the structure that many of us have come to depend upon when explaining white patterns. Just naming the different patterns is a challenge, if you believe that the primary benefit of a uniform naming system – one that is used across breeds and countries – is clarity. This challenge of names and structure will be the topic of the next post.

Continue Reading