Acceptable colours in the shepherd coat are black and tan. Black, created by the pigment eumelanin, varies in its degree of extension over the body, while tan, created by the pigment phaeomelanin, varies in its richness. A well pigmented dog will show black extending over at least 50% of its body, with a strong, uniform tan ranging from a golden, "autumn grass" colour to a rich, reddish mahogany. These colours are arranged in four patterns. They are, in order from most to least dominant: sable, black and tan (B&T), bicolour and black.

1. Rich saddle sable, black sable, faded sable.
Sable is the original wolf colour, with a paler ground colour overlaid with a mantle of black tipped guard hairs. In the shepherd this ground will be tan, with an overlay which is heavy and may appear from a distance to be black. Willis discusses, but is not convinced of, the existence of two alleles for sable: grey and yellow. However, the same variation in colour can occur in B&Ts, and can be perfectly well explained by other, better documented genetic influences. I do suspect however, that there may be two sable alleles affecting the degree of extension of the black mantle: one for the saddle or blanket marked sable, the other for the so-called black sable, which shows distinct black markings on its pasterns, toes, stifles, and hocks, and often black overlay on the throat and chest. Some black sables are so dark that they appear black, until they move and the lighter under fur shows through. Two such alleles, if they exist, would coincide with the genetic distribution of pigment in the B&T and the darker bicolour.

I don't believe a sable actually improves pigmentation in breeding just because it's a sable. A sable can be as poorly pigmented as any other pattern. I do think that a sable must be more melanistic (dark) than an average B&T just to appear equally dark, and therefore brings greater pigmentation to a breeding because it is genetically darker. But the same results could be achieved by breeding to an equally melanistic B&T.

2. Rich B&T, melanistic B&T, faded B&T.
B&Ts also show a rich tan ground, but have black saddles or blankets made up of guard hairs which are black to, or close to, the skin. The black mantle should cover the nape of the neck, the shoulders and distinctive harness markings, back and sides down to nearly the belly, croup and tail. Often tan or grey shows through at the neck and down the tail, but if this is noticeable over the croup or along the sides, or the tail lacks a black tip, colour paling is present. "Salt and peppering" of the withers and back is quite common in bitches, even very dark ones, but is more commonly seen in faded males than dark ones. A very pale B&T can easily be mistaken for a light sable, but if they're that hard to tell apart, they're both very faded in pigmentation and should be faulted accordingly. Very dark B&Ts are often advertised as bicolours, and can have heads, necks and bodies that are nearly solid black. However, they usually show some tan shading about the base of the ears, and have few or no black hairs down the pasterns, or on the toes or hocks.

3. Rich Bicolour, Melanistic Bicolour, faded bicolour.
Typically, the bicolour is a black dog with tan points much the same as a doberman. It is in fact the same gene. A very dark one with a strong black mask and much extension of the black may appear to be solid black, with a black undercoat and only some tan shading on the feet. Theoretically, bicolours always show tan around the vent, but I have seen a virtually black dog with just enough tan around the vent (and between the toes) to suspect that it was a bicolour, but not enough to be sure. However, a pale bicolour can look strikingly like a black sable, showing much black over a light grey undercoat, with tan or grey over the nape of the neck and along the harness lines, and black markings on the legs and feet. These dogs have the genetics for fading of the black mantle, and while they can still appear dark to the eye, will produce colour fading.

4. Black
A black is a black is a black. Some blacks may show tan hairs between the toes or on the pasterns, and some may even show a bit of grey undercoat over the neck and shoulders. Maur Ray kennels, an American show kennel specializing in blacks in the 1940's, suggested that blacks bred of blacks for several generations showed fewer and fewer light hairs, until they were absolutely pure black. Breeding to a black to improve pigmentation however, is not recommended unless its pedigree is made up of dark dogs of other patterns. It is not uncommon for pale B&Ts to throw blacks, who will in their turn throw pale B&Ts. A black will not reliably show the genetics for absence of mask, paling tan or fading mantle, and can easily hide and pass on the genes for fading colour.

5. Muzzle pigmentation
The black mask, so necessary for strong shepherd expression, appears to be a separate dominant gene. It is quite possible for a very strongly pigmented dog to lack mask, or for a fading dog to show a strong mask. The mask involves the muzzle up to and around the eyes. The dark brow band is genetically related to the mantle, so that a dog with a faded mantle will probably have a tan or nearly tan forehead, while the dog with the strong, heavy black mantle will show a black brow. A bicolour with no mask will be marked like a Doberman; black through the face with a tan muzzle and throat, and the black with no mask will be impossible to distinguish. In sables, the brow takes on the badger-like sable effect, while the muzzle does not.

6. Black & Silver, Red and faded black.
Paling pigmentation is not desirable, but its various manifestations can be rather complicated. A fading of the black mantle to little more than a smudge is decidedly unattractive, but on a tomato red dog seems less apparent and is often forgiven. The dog with the strong black mantle and mask, but with a creamy or even silvery ground colour may possess a gene for partial albinism, and is equally undesirable. Under certain circumstances, this dog may produce whites. White markings are not desirable of course, although a small white spot on the chest is common even in dark phase wolves, and may as well be ignored so long as it doesn't spread further. Pink toenails may signal the appearance of a more invasive white spotting gene. Deep black nails are an indicator of overall mantle pigmentation, as dogs with poor saddles usually have fleshy or dusky nails. White can occur as a result of several different genes, some associated with colour fading (appearing creamy, with dusky lips and nose), some not (appearing milky, with black lips and nose), but under the standard, white is white and is not acceptable. However, according to Willis, deafness is the result of a double recessive merle white which does not occur in shepherds, so white shepherd puppies are not at risk of this defect.

7. Four patterns of Blue. 8. Four patterns of Liver.
The blue and liver genes, once rare, now seem to have captured the imaginations of pet breeders, if my search of the internet is any indication. If otherwise well bred, they can be quite striking in appearance. They are not expressly forbidden, but the standard demands the nose leather be black, and these dilutions show blue and brown leather. The blue gene causes the black pigmentation in both hair and skin to turn steel grey, often very dark, while the tan ground remains much the same, or takes on a slightly silvery cast. The liver gene does much the same, turning black pigmentation brown. In both, the eyes take on a peculiar, silvery yellow colour. Both dilutions are simple recessives and can occur in all patterns, and when these recessives link, the result is the silvery fawn of the Weimaraner and Isabella Doberman Blues and fawns, according to Doberman breeders, tend to have poor hair coats, a good enough reason for culling them, but the so-called lethal gene once thought to be associated with liver appears not to exist. Blue, liver and white puppies should be placed in loving homes where they will be neutered and live as valued pets. Puppies should never be euthanized because of their colour.

9. B&T with bright points, Sable with bright points, Willis pg 132*
One oddity of colour seen occasionally is the tendency for the tan ground colour to divide into two different tones of tan. The "points" of lighter markings most obviously seen in the bicolour seem to exist genetically in all patterns. I've seen B&Ts and sables with strong mantles and deep, rich tan, but showing well defined points of cream on the feet, chest, throat and cheeks; what the Germans seem to call "bright". In one family I've observed, this trait has been remarkably tenacious, and great care should be taken to select partners who have tan which is uniformly rich.

10. Brindle B&T
I can't resist mentioning the brindle, which was one of the founding patterns of the breed and seems to have become extinct. The brindling gene affected the tan ground, so that B&T's and bicolours showed a dark striping or marbling (the same effect seen in brindle boxers) over the tan which was not at all unattractive. I wouldn't hazard to guess what a brindle marked sable would look like. Hopefully, if the pattern somehow reappeared, it would not be discarded as atypical.

Eye colour is unrelated to coat pigmentation. A very dark dog can have very light eyes, while a very pale dog can have very dark eyes. The standard says eyes should be as dark as possible, but a dead black eye is rather expressionless. Others suggest the eye should harmonize with the overall colouring, but a lighter face is even more attractive with a rich, medium to dark brown eye, so there is really no reason to settle for a lighter eye. In very melanistic dogs even the gums can be black, and black spotting of the tongue is not uncommon. The skin under areas of black, whatever the pattern, will be an ice blue, but under areas of tan will tend to be pinker. All areas of exposed skin, except for the ear flap and vent, should be black, although in dark dogs, even these can take on a blackish cast. Nails should always be black.

* Willis, Malcolm B., B.Sc. Ph.D., 1977, The German Shepherd Dog: Its History, Development and Genetics. New York; Arco Publishing. p. 132.