Contact Information

Coats, Colors and Markings

by David E. Smith, MD

Coat Texture

Bella and Gabby After Grooming

The photo on the left shows John and Diane Parks two female PWDs immediately after grooming. Gabby is on the left and has a tight curly coat while Bella on the right has classic wavy coat. The best assessment of coat type is immediately after a bath and grooming as a "dirty" wavy coat tends to curl around the dust and dirt particles. When the coat is clean after bathing the sheen of the wavy coat is apparent while a curly coat has a dull appearance.

The PWDCA Standard approved by the American Kennel Club describes the PWD coat as "A profuse, thickly planted coat of strong, healthy hair, covering the whole body, evenly, except where the forearm meets the brisket and in the groin area, where it is thinner. No undercoat , mane or ruff. There are two varieties of coat:

Curly - compact, cylindrical curls, somewhat lusterless. The hair on the ears is sometimes wavy.

Wavy - Falling gently in waves, not curls, and with a slight sheen.


When we started in PWDs it was fairly easy to tell coat types. However in the 1980s the health problem of "adult hair loss" arose. Dermatologic studies by Dr. Miller showed that the hair loss was due to "follicular dyspasia". There was no histologic characteristics of inflammation in the hair follicles such as we see with "sebaceous adenitis" in Standard Poodles. DNA science at this time was very unsophisticated and crude so pedigree analysis was undertaken to try to determine the inheritance of this disease. It became apparent that "follicular dysplasia" usually occurred with tight curly to tight curly breedings. Due to these findings most conscientious breeders did wavy to curly breedings to try to prevent this disease. These wavy/curly breedings tended to produce a spectrum of transistional coats. When you couple these transistional coats with puppy coats it can frequently be very difficult to characterize an eight-week old puppy's coat.

Jane Anderson an Australian breeder (Bluegrace PWDs) has made the observation: "The interesting thing with curly puppies, is they will often look like straggly wavy puppies for a long time until their adult coat comes through. I like to shave off curly boys completely at 12 weeks of age to allow a stronger coat to come through earlier. Seeing a wavy and a curlier PWD pup side by side at a show, it is quite likely you won't be able to tell the difference in coat type. But these differences will become more apparent as the pups age. Sometimes it can take 3-4 years for the curls to come through properly, confusing both owner and breeder as to whether the dog is a wavy or a curly! And just to confuse people more, the dog can have a curly head, yet a wavy front. The key to deciding which coat type the dog is, is to look at the luster of the coat. Curly coats will not shine, while wavy coats do."

Carla Molinari's description of coat types in her 1993 book provides insight into the variations in coat seen in Portugal. "...The curly coat is found mainly in two types: very thick curly–as in a Poodle–which is composed of thick hairs of strong texture. This type of coat is rather opaque, easy to sculpture but difficult to keep in condition if left to grow long. This not a desirable type of coat, especially when it is too thick. The other variety is made of a tighter curl, with a better luster but not as thick in texture. When kept long it tends to fall in clusters, it mats easily and is difficult to condition. This is a much more typical coat texture of the curly variety. In the wavy coat we find again two varieties (or degrees) of waviness. In one the wave is slight and the hair is thick and almost "oily" looking. However when wet this coat curls instantly. The other variation shows a thick wave (almost a loose curl), where the hair forms wide S shapes which naturally stay curly even when the coat is brushed."

Recent revelations with modern genetic DNA testing, described below, has revealed that curlies have a allele variation in the KRT-71 gene on Canine Chromosome 27 (I hate to use the word "mutation" as it is interpreted as "defective".) This gene is one that contols keratin production. A similar genetic KRT-71 variation in the mouse produces an abnormally short keratin in the center of their hair. These mice are born with hair but lose their hair as they age. This is very similar to the course of PWD follicular dyplasia. Clarifying the gene variation responsible for "follicular dysplasia" will focus on the KRT-71 gene.

The rapid advances in DNA technology coupled with the mapping of the complete canine genome in the past decade have provided an understanding of the genetics of coat texture and length. The breakthrough article published by Edouard Cadieu and nineteen other genetic scientists in Science in 2009 revealed that the seven different phenotypes that distinguish more than 80 AKC breeds are controlled by only three genes: RSPO2 (R-spondin-2 gene), FGF5(Fibroblast growth factor 5 gene), and KRT71(keratin-71 gene). More than 1000 purebred dogs' DNA was involved in this study including 75 PWD sample provided by the Georgie Project at the University of Utah (Drs. Gordon Lark and Kevin Chase). The ancestral short-haired dog has the "wild-type" of these three genes. Variants of these three genes account for 95% of the variety in coat length and textures that we see in dog breeds today.

A Short (Ancestral Dog) - - - Basset Hound, Beagle
B Wire - + - Australian Terrier
C Wire and Curly - + + Welsh and Other Wire Coated Terriers
D Long + - - Golden and Flat Coated Retrievers
E Long with Furnishings + + - Bearded Collie, Wavy PWD (CC KRT-71)
F Curly + - + Irish Water Spaniel, Curly Coated Retriever
G Curly with Furnishings + + + Curly PWD (TT or TC KRT-71), Bichon, Poodle
Interestingly none of the eighty breeds studied had a KRT-71 mutation alone.

A better understanding of how these three genes work to control coat length and texture is to use classic Mendelian genetic locus descriptors.

The L (length) locus:
The alleles at the L locus (the fibroblast growth factor-5 gene) determine the length of the animal's coat. There are two known alleles that occur at the L locus:
L= Short coat (ancestral/"wild-type")
l = Long coat L is dominant to l.

The R (curl) Locus:
The alleles at the R locus (the keratin-71 gene) determine whether an animal's coat is straight or curly. There are two known alleles that occur at the R locus-curly may be TC or TT for KRT-71 while wavies are homozygous for the wild-type allele CC:
R= Curly
r = Straight
R is dominant to r.

The W (wired) locus:
The alleles at the W locus (the R-spondin-2 gene) determine the coarseness and the presence of "facial furnishings" (e.g. beard, moustache, eyebrows). There are two known alleles that occur at the W locus:
W= Wire (hair is coarse and facial furnishings present)
w= Non-wire (hair is not coarse and facial furnishings are not present)
W is dominant to w. Animals that are homozygous for l (i.e. l/l) and possess at least one copy of W will have long, soft coats with furnishings, rather than wirey coats. PWDs that are homozygous for w (i.e. w/w) have improper coats.

Let us now put this new knowledge into the two types of classic PWD coats as well as improper coats. All PWDs have long hair that continues to grow so they must have two copies of the little "l" gene written l/l. Wavy coated PWDs are obviously r/r while curlies may be homozygous R/R or heterozygous R/r. The heterozygous curly (ll/WW/Rr) may demonstrate incomplete dominance accounting for the transistional coats we see, although this is just my hypothesis. Improperly coated PWDs have to possess two copies of the little "w" gene written w/w. We can DNA test for the Carrier state W/w using the Optigen IC-13 test. Non-carriers are obviously W/W. Improper coated wavies are ll/rr/ww while improper coated curlies are ll/RR/ww or ll/Rr/ww.

My belief is the two ends of the PWD coat spectrum are wavy and curly with multiple transitional coats in between. The in-between or transistional coat has been called by breeders "loose curly, wavy with mousse or wavy with texture". For breeding purposes I tend to look at body coat as the representation of the two genetic coat types but recent bathing can even change the appearance of the body coat. The headshots below illustrate the striking difference in coat on the head of three of our dogs. On the left in the photo, GCh Dacher's Go For The Gold "Bud" is a classic wavy with a very shining coat and a natural part in his hair on the top of his head. In the center of the photo, Dacher's Quailchaser "Crystal" is a tight curly. Note the dome of her head has tremendous body almost like a topiary and the coat is dull (lusterless). On the right in the photo, GCh Dacher's Force Five "Fresco" is a typical loose curly. His coat is more shiny than a tight curly and has body so that it does not naturally part.

Coat Types

Color and Markings

The genetics of coat color in dogs has been extensively studied for decades. Clarence C. Little, a mouse geneticist at the Jackson Laboratory at Bal Harbor Maine, provided the basis of our understanding of color genetics in his classic text "The Inheritance of Coat Color In Dogs" that was published in 1957. He described eleven different genetic loci that control color and the pattern of markings in domestic dogs. Two additional genes have been postulated since Little's work to explain other variations. Remember what we see is the phenotype (appearance) but what is important is the genotype. Each parent provides a puppy with on copy of the thirteen genes responsible for color and markings. Eight of these genes have been identified and five are still hypothetical to describe marking patterns seen in dogs. The table below summarizes these genes and their effects.

Nine Important Color Gene Locus Alleles
Gene Alleles Origin of Name Type Effect of Gene
A   Agouti Modifier Tan/sable markings (pheomelanin production)*
B   Black/Liver Base Color  
  B Black   Creates black coat BB or Bb
  b Brown (Liver)   Homozygous bb creates brown
C   Concentration Modifier Variation cause decreased concentrations of pigment–C normal
  cch Chincilla   Warm cream to white: Poodle Varieties (Probable White PWD)
  ce Extreme   Extreme reduction in color: West Highland White Terrier*
  cb Blue Eyed Albino    
  ca Pink Eyed Albino    
D   Dilution Modifier dd causes dilution of eumelanin; amber eye color
E   Extension Melanin Production Extension of dark color mask on muzzle and ears *
G   Graying (Early) Modifier  
  G Early Graying   Born black or brown but grey as they age (Kerry Blue Terrier)
  g No graying   Recessive no graying
K   blacK Pigment Switcher ß-defensin interacts with Mc1r membrane receptor
  KB Black   Mutation that blocks Mc1r (solid black-eumelanin produced)
  kbr Brindle   Recessive for brindle*
  ky Yellow   Recessive for yellow(wild-type allows pheomelanin production)*
S   Spotting Distributor S, si, sp, sw–see discussion below
T   Ticking Distributor Presumed gene for coat ticking has not been isolated
  T Ticking   Causing ticking (small areas of pigment) in white areas
  t Non-ticking   Non-ticking effect

* Not present in the PWD–M (Merle) and H (Harlequin) are also not present

It seems incredible but these genes affect only two types of pigment–eumelanin (black/brown pigment) and pheomelanin (sometimes spelled phaeomelanin) (red pigment). These pigment are present in skin, hair follicles and hair. There are three color coats in PWDs–black, brown and white. Black is created by two genes the KB and the B or liver gene. A black base coat in a PWD requires one copy of the dominant black gene KB and one copy of the dominant B gene. Believe it or not a brown PWD is really a dominant black KB with two copies of the recessive liver gene bb. Most brown PWDs have a graying or silvering gene as well. The puppies are born a deep rich chestnut brown but will tend to develop a pewter or lighter brown as adults. The photos below show the brown color variations as a puppy matures.

Brown Coat Color

A completely white coat in a PWD is rare in the United States, but does occur in Portugal. Having bred standard poodles for many years white and cream coats have pigmented bluish skin. The hair color is basically an extreme reduction in pigment in the hair (eumelanin in white Poodles and pheomelanin in creams). It is also controlled by the C locus that drastically reduces the amount of pigment in the poodles hair. The genetics pf white coat coloring is complex and has not been studied in the PWD but actual DNA genetic tests have been developed in poodles to predict color inheritance. A link to the VetGen DNA test is provided at the end of this article.

White markings are very different than a solid white coat. White spotting as described below have no pigment in their hair and the skin around these hair follicles is pink(lacking pigment).There are four distinct patterns of markings–solid, Irish-marked, parti-colored and extreme white. Solid color dogs do not have absolutely 100% of a single color hair. Almost all "solid" colored dogs will have small amounts of white on their feet, tail tip, chin and even a small patch of white on the chest. The reason for small amounts of white on solid colored dogs has been ascribed to the fact that melanin producing cells originated from the embryonic neural crest cells. The neural crest cells begin in the embryos head and spine area and have to migrate down to the feet, abdomen and chest during the development of the fetus. These solid dogs are basically 95-98% one solid color. These small amounts of white are developmental and not genetic. Irish-marked dogs are said to have "tuxedo-markings". A full or partial white collar defines Irish-markings. This is usually accompanied by long "opera gloves" on the front legs and white socks on the rear legs. A white tail tip, large white chest marking, white face blaze also are seen with this pattern of markings. When the white markings break into the body hair the dog should be considered a parti-colored. Extreme white is uncommon in the PWD.

Darby Irish Markings

The striking bitch pictured to the left Ch Dacher's Nobreza Dare Devil "Darby" is an example of an Irish marked Dog.

The gene responsible for white patterning was defined by C.C. Little as the S locus. This locus has been termed "the white spotting gene". Each PWD has two copies of this gene. Four alleles of this gene determine the pattern of markings seen in many dog breeds including PWDS:

S —Solid Color (no white)

si —Irish Marked (<40% White)

sp —Piebald/Parti-colored (40-80% White)

sw—Extreme White (>80% White)

The white spotting alleles are all thought to be examples of incomplete dominance. This means that a heterozygous dog will express its most dominant gene, but will also be affected by the recessive one. A bitch like Darby with classic full Irish markings most likely possesses two copies (homozygous) of the si allelle but a dog who is heterozygous S/si will not be a solid color but have partial white Irish markings such as a large white chest and white front socks. This explains the large variation in white markings in PWDs.

Color in the United States and International Standards

AKC Portuguese Water Dog Standard Approved 01/15/1991

Black, white, and various tones of brown; also combinations of black or brown with white. A white coat does not imply albinism provided nose, mouth, and eyelids are black. In animals with black, white, or black and white coats, the skin is decidedly bluish.

FCI Portuguese Water Dog Standard Approved 09/12/1966

Colour : Either self-colour or combinations. Colours are black, white and various shades of brown; also combinations of black or brown with white. A white coat does not imply albinism provided nose, mouth and eyelids are black.
In animals with black, white, or black and white coats the skin is decidedly bluish.

This quote from the standard issued by the CPC regarding combinations with white is not very clear, but in "The Portuguese Water Dog" Carla Molinari clearly states that:

"An official statement by the Clube Português de Canicultura(CPC) sets the quantity of white on a black or brown coat to a maximum of 30%, while areas where white markings are permissible are defined as follows: muzzle, head, neck, chest, front legs, underbelly, rear legs, tip of tail."

This clarification from the CPC has been incorporated in the FCI(International) breed standard as the illustration below from the Australian breed standard shows. NOTE in this illustration a full white collar Irish marked PWD would exceed the 30% rule as shown as incorrect at the top left.

Correct Markings Australian Standard

Why is there a difference in acceptance of excess white markings in the PWD standards?

The only explanation that I have received over my years in the breed is that in Portugal the PWD had to work in the water and on open boats with little shelter. Since white markings on the back and head are associated with pink skin the dog was more likely to get sunburned in these areas. A sunburned dog would not be able to carry out his work. White on the paws, abdomen, and chest would not be susceptible to sunburn on these areas as they are not exposed normally to the sun. The Portuguese fisherman clearly only cared about color if it effected their canine companions ability to work on their fishing boats at sea. However this is only a proposed reason.

Thoughts on Breeding for Color and Markings

White markings as demonstrated above are carried by a recessive gene. This is similar to brown that is recessive to black. If you want to breed a brown dog with Irish markings you lose tremendous genetic diversity. I feel that in breeding healthy PWDs with sound temperaments means that color and markings are of secondary consideration.


Coat Variation in the Domestic Dog Is Governed by Variants in Three Genes. Science 2009

A β-Defensin Mutation Causes Black Coat Color in Domestic Dogs. Science 2007

DNA Testing For K-Locus
VetGen DNA CHROMAGENE Color Testing For Poodles
The Genetics of Cream Coat Color in Dogs

Created January 3, 2012