If I was drawing up a color classlist, I would not group palomino with chestnut (although admittedly some pale flaxen chestnuts are hard to tell apart from palos). My own preference is to group the palos, buckskins and cremellos together and then the duns - and I like to have the grullas in with the duns (they are black duns) and not in the "Other Color" class. But most classlists are not that extensive, and I have often just had a palo/buckskin/dun.
For myself, I lump in three genes into a general heading of "dilution factors"; the cream gene (Cr), the dun gene (D), and the chocolate gene (b). I don't really think of the silver dapple gene (Z) as a dilute, because I find using this terminology leads to confusion (silver dapple does more that just dilute - or lighten - color), but technically it *is* a dilution gene.
Cash For Gold
Palomino Quarter Horse stallion customized from the Stablemate Quarter Horse Stallion by Elaine Lindelef and owned by Kate Cabot.
The first gene, Cr, is responsible for palominos and buckskins when the horse inherits one Cr gene. Two Cr genes gives cremellos (if the horse would have been a chestnut) and perlinos (would have been bay). Cr (in "one dose") ONLY dilutes red pigment, which means chestnuts are totally diluted (gold body, white mane), and bays only have the red hair diluted (gold body, black mane). A black horse, which has no red pigmented hair, will not *show* the Cr gene, although he can have it - it just isn't visible. Blacks with this hidden Cr gene (remember it's not recessive - it's just not visible) will breed just like a palomino or buckskin would; half their non-black foals will be dilutes. If the horse inherits two Cr genes (CrCr), it will be what is commonly termed a "double dilute", and will be near-white with blue eyes (ALWAYS blue). When there are two Cr genes in the animal, even the black pigment is diluted, so it is hard to tell cremellos from perlinos without test breeding; both colors look nearly white. Double dilutes produce 100% dilutes.
Grulla Quarter Horse stallion; Rio Rondo QH#1 custom-painted by sculpting artist Carol Williams and owned by Kathy Blair.
The second gene, D, is responsible for all duns. Unlike Cr, dun is not pigment specific, but rather LOCATION specific. It will dilute the body color, and leave the point color dark - whatever color that is. So D will make chestnuts red duns (tan with red points), bays yellow duns (tan with black points), and blacks grullas (smokey with black points). Unlike Cr, doubling the dun does not have any effect on the visible color, although a horse carrying two D genes will produce 100% duns. The most distinctive thing about duns is the presence of primitive marks (linebacks, barring, cobwebbing, etc.). Although these are a tell-tale sign of the gene, they are thought to be the result of a separate gene (primitive marks such as linebacks DO occur occasionally in non-duns) that is strongly linked to the color. This primitive gene (in it's most extreme form) is thought to be the source for what I term "zebra" brindles (as opposed to "rabicano" brindles).
Now it is really hard to tell some buckskins (Cr) from some duns (D). Usually the presence of primitive markings are a giveaway, but this isn't 100% certain because some buckskins can have them (just as some bays do). But it makes it far more likely that they are duns. Another way is by test breeding or studying production records. If a horse has ever produced a cremello, he HAS to be carrying the Cr gene. If he has bred 100% buckskins (and has a fairly large number of foals), he is likely a dun (and a homozygous one at that). Also, checking his breeding records to see if he's produced palominos (he's Cr) or red duns/grullas (he's D) will help. If you are good at telling fine nuances of color (most artists are very sensitive to this), a Cr horse will tend to have a more golden cast, while a D horse will be a more 'dusty' muted shade.
Also remember that nothing prevents a horse from inheriting BOTH genes, D and Cr. Although it is not common because few breeds have both genes, it does happen. My observations have lead me to believe that when this does happen, the dun will excercise a kind of dominance and "override" the Cr from completely manifesting. A chestnut Cr horse that would be a uniform gold with a white mane and tail, will, when D is added, turn a more dusty shade, and a faint lineback will usually appear (not as visible if Cr wasn't present, but still there), and the legs will usually have some darkening, but again not as much if the horse were just dun (they look slightly smutty tan), and often the mane will have a few darker hairs or even be dusty tan. These horses are what most people term claybank duns, and there are several good examples of these in the Sponenberg book.
Those are the most common dilution factors. The third is the chocolate gene (b), and it is so rare as to not be an issue with most horses. Chocolate (b) is much like the same color in dogs like Labradors - it turns the skin that would be blackish grey a chocolatey pink color, the fur chocolate brown and the eyes a golden or hazel color. If you have ever seen a pinkish-skinned (kind of pumpkin) palomino with amber eyes, you have seen this gene --- and I suspect from the breed standard that calls for pumpkin skin and amber eyes, the Cream Draft actually carries this rare gene. I've also seen it in Tennessee Walkers (where it is called Champagne) and Saddlebreds, but it is hard to say much about it since it's rare and largely undocumented. I have meant to ask Dr. Sponenberg if more had been learned about it, but keep forgetting!
The (much discussed here!) silver dapple (Z) is indeed a diluting factor, but as I said before, classifying it as such does lead to confusion with a color already REALLY prone to generating confusion! In a way, silver dapple could be looked upon as the opposite of the Cr gene. It only "dilutes" BLACK (rather than red) pigment. So a black horse is completely diluted, and a bay horse has only the black mane, tail and legs diluted. And just as a black horse can carry an "invisible" palomino gene, a chestnut horse (with no black pigment) will carry the invisible silver dapple.
But the confusing part about silver dapple is that it doesn't just "dilute" the black; it changes it. Silver dapple (Z) makes the black a chocolatey, mocha color (ranging from coffee-with-cream to nearly black), often with creamy mocha dapples (although sometimes with none). The points end up nearly white cream. The face usually remains almost black (I refer to this as 'masking' - silver dapples all have masking to some extent). If the horse is bay (black only on the points), it will turn the legs some variety of chocolate or dapple chocolate, and the mane/tail cream (usually with black roots). Doubling the Z gene doesn't affect the color as far as I can tell, but only creates a homozygous silver dapple. The non-chestnut offspring of a homozygous silver dapple will all be silver dapple.
Whew. More than anyone ever wanted to know about dilutes. :^) Sorry for being so long-winded, but I couldn't think of a more concise explanation!
PS - The basis for my understanding of color is the writings of Dr. Sponenberg (and a few of his predecessors), and I would HIGHLY recommend his book Horse Color, which I believe Cascade carries. I try to avoid it, but some of what I say on Haynet regarding color are my own theories or conclusions, and I have no scientific proof beyond my own study, correspondence and observation. I can say that I've done a lot of study (as one truly possessed by this particular subject!), but in the end they are still just theories and may be disproved over time and with more numbers/data to work with.