Genetic Color  Inheritance


My own Pedigree
Grafical version 1
Grafical version 2

Grafical version 3



About Us

Pre history  

Tibet Dog

Our Dogs
Foto"s Offspring

In loving memory


Heritage Type 1
Heritage Type 2



Breed Standard


Hoek van Holland 1
Hoek van Holland 2

Our other Passion
Magical Realism



Webcam 1






Black with  si-gene


White-Brown- with sp-gene  

Brown with d-gene is grey-isabella

Black with si-gene
White-black  sp-gene
White-black  sp-gene
Black en Tan with at-gene
 Foto = G v/d Vegt/K Butenhoff
Liver en Tan with  at-gene

Foto =G v/d Vegt/K Butenhoff
Black-Grey with d-gene
Foto = G v/d Vegt/K Butenhoff
White-Black wich Tikking 
Black-Grey is looking as Bleu-Grey
White-Black with d-gen
is White-Grey.
Albino = ca gen, 
born in the USA 2007.

Dutch Shepherd, wich kbr-gene
Out of Brown x Brown













There exists rather what misunderstanding about the colour code of a Newfoundlander.
With my statement mentioned below I try to bring clarity in
wrong colour and colour codes. 

What is now a wrong colour? 

A wrong colour is a not desired colour, which is not acknowledged by the Council of Management and the FCI ( Federation Cynologic International ) and is not described in the race standard of the race association of the NNFC (Dutch Newfoundland Club).  

Through new sciences study's of the last 5-year: are through the "University of Saskatcheawan Department of Animal and Poultry Science by Professor Dr. Mrs. Sheila Schmutz" are based on DNA and phenotypic comparisons other conclusions regarding the genes of the A locus B , C , E , S and a K locus have been added.

In the old situation it was :

Thus therefore the genes lie which stipulate the colour of the coat, as far as I know, on 9 loci : A,B,C,D,E,G,M,S,t
For the Newfoundlander applying loci are
therefore A , B ,C , D , E , S , t.

Well the new situation is now :

For instance, the genes that determine coat color, to my knowledge, at 10 loci a, KB, B, C, D, E, G, M, S, t.
The force for the Newfoundland locussen are
a, KB, B
, C, D, E, S, t.
The As-gene is disappeared on the A locus, KB stays for dominance (blac K=solid) and kbr for brindle is placed in the eumelanine dark  pigmentation and the ky-gene is responsible for Fawn.
On the
light pigmentation phaeomelanine, brindle is on the E locus Eebr is and the suggested Cch-gene for dilution was not found. 
ee-gene is in homozygote form responsible for a white/creme dog as the Canadian Shepherd
and yellow for the Labrador by exemple.
si-gene is divided by de Undersides or Pseudo Irish-Spotting what came out in practice that it autosomal recessive inheriting is.  
sw-gene they could not find it.

On a locus "a place on the DNA string" is only room for Maximun two genes, "of every parent one for each gene" .

The genes below stand in
ORDER of DOMINANCE, such I have seen sometimes in homozygous and 
what also possible is, in
heterozygous (impure
) form.

The colour code is in homozygous form (pure) for the black = KBKBBBCCDDEESStt.   
And for brown
KBKBbbCCDDEESStt and for the black and white is this KBKBBBCCDDEEspsptt.

Dark pigment ( eumelanin) is black and brown, yellow and red are light pigments (phaeomelanin).

§       ay gene : oppressed the genes which carry the dark colours ( KBB and b, E ) and  give yellow or red colour at the hairtip
y comes from yellow ) and give the dog the expression of a brown one without the nose and mask

§          at gene : is standing for black/liver and tan.

§         KB gene : is standing for dominance (blac K = Solid) dark pigment over the whole body.
In some
breeds, such as the Newfoundland dog, all dogs homozygous for K
In such races, only the Ee locus determines whether the dog than
dark pigment or light pigment gets.
KB KBgenotype is fixed and the B locus determines whether they are black or brown.

§         kbr gene : br stands for the brindle pattern as exemple the Dutch Shepherd (wrong color by the Newf "see photograph").

  • ky-gene : stays for Fawn (yellow-brown) on the Tan points. Dogs which have two recessive alleles (ky/ky) can express a variety of phenotypes.
    All Black-and-Tan dogs or dogs with Tan points are
    kykyBBCCDDEESStt  : = Black and Tan (see photograph).            atatkykybbCCDDEESStt : = Liver and Tan (see photograph).

  • B-gene : is standing for black.

§         b gene : is standing for brown. Three different mutations in this gene ( bs,bd,bc ) can produce brown.
Based on the brown Newfs they have tested the bc allele, is the typical brown allele in this breed (read The Mythe of Brown down under).

§         C gene : complete pigmentation, gives all deep red and brown races.

  • ca-gene :  gives no  pigmentation, complete albinism, gives white with red eyes.
    Most researchers have the same opinion that albinos by mutations in the tyrosinase (TYR) on the C locus lies.
    Tyrosinase is a gene that causes albinism in mice, humans and

§         D gene : is standing for intensive pigmentation .

§         d gene : is standing for dilution black will be blue-grey and brown will be silver-brown/isabel.

§         E gene : is standing for dark pigment ( eumelanin ) concerning the coat.

  • e- gene : prevents the formation of dark pigment so the bright red or yellow as in the Labrador or white/creme by the Canadian Shepherd.
    Does not occur at the Newfoundlander, but I use this code for the theory explanation.

§         S gene : is standing for divided stopping the white colour, except sometimes a white mark on the chest and a couple of white toes .

§         si gene : stands for Irish spotting  "pattern or UnderSides or Pseudo-Irish Spotting"  is a white spot on his chest, white feet and tailtip, sometimes with some white on the muzzle in a more or less symmetrical pattern.

§         sp gene : piebald spotting, a large white mark or white without pattern.


  • NOTE: Some of these alleles at the S locus may have incomplete or co-dominance
    A dog that is (si-sp) may show both patterns simultaneously, for example (see photograph).

§         T gene : is standing for Ticking .

§         t gene : is standing for reducing Ticking.


The key to understanding dog genetics is simply this: there are two types of pigment which create coat colour in dogs (and most other mammals).  Pigment is just the thing that gives each strand of hair its colour, just like pigment in paint or dye, or pigment in your own hair. 
All coat colours and patterns in dogs are created by these two pigments.  
of the pigments has a "default" colour, and it can then be changed by various genes. 
gene: suppresses the genes that contribute dark color (KB and b) and gives yellow to reddish color to the hair, leaving a black dog looks brown. so y = yellow light comes from pigment.



The Newfoundland Dog falls into this category dark pigment.

Eumelanin is black pigment. All black areas on a dog are caused by cells producing eumelanin. However, there are genes which turn eumelanin into other colours - liver (brown), blue (grey), or isabella (a dusty pale brown) If a dog has any of the genes to turn its black eumelanin into liver, blue or isabella then all of the black in its coat will be changed. This is because these genes restrict and/or alter the production of eumelanin, so the cells aren't able to produce full-strength pigment. We call blue and isabella dogs "dilutes" for this reason.  
They either lack the genes which tell their cells how to produce proper eumelanin, or their cells know how to produce it but can't.

As well as being found in the coat, eumelanin is present in the other parts of the dog that need colour - most notably the eyes (irises) and nose. 
The nose will be black, liver, blue or isabella depending on the type of eumelanin the dog can produce.  The colour in irises is produced by layers of pigment, and brown eyes like most dogs have (and like many people have too) are caused by black eumelanin in those layers.  When a dog has altered/restricted production of eumelanin, the irises are also unable to produce full-strength eumelanin.  This means that the dark pigment in the eye becomes lighter, and the eyes turn into a light brown colour, known as amber or gold . This isn't really a colour we find in human irises, so it's hard to illustrate it with an example. The closest we have are hazel eyes, which are a mixture of green and brown, and so appear lighter than normal brown eyes.  Dogs, however, cannot produce that green colour, only brown and blue.

This shows the variation in colour of eumelanin. 

Black is the default, then it can be turned into liver by one set of genes, and black and liver can be turned into blue or isabella by another set of genes.  Blue is diluted black and isabella is diluted liver.

When we talk of dogs that are "black pigmented ", "liver pigmented", etc, we mean that is the colour of eumelanin that the dog can produce.  Sometimes these dogs have no eumelanin at all in their coats ( their skin cells produce only the other type of pigment, phaeomelanin), 
but we can tell what their "pigment colour" is by looking at their nose. 
A black nose means the dog produces black eumelanin, and so on. 
It's confusing to talk of a dog's "pigment colour" like this, because as we know, eumelanin isn't the only type of pigment. 
 But it's common practice, and I'll be using such terms a bit on this site.



The second type of pigment, less important than eumelanin, is phaeomelanin. This is red pigment. 
The term "red" covers everything from deep red ( like Irish Setters) to light cream, encompassing gold, yellow and orange. 
Whenever we talk of red, unless we're talking particularly about Setters, we mean the whole range of tan colours.

Phaeomelanin is produced only in the coat
It does not occur in the eyes or the nose, so any genes which affect the colour/intensity of phaeomelanin will not affect the eyes or nose.  
Only eumelanin occurs in those areas, and so only genes which affect eumelanin can affect the eye or nose colour.

This shows the variation in colour of phaeomelanin.
Unlike eumelanin, it doesn't occur in two distinct colours (black and liver, with dilutes counting as shades of those), but rather just one colour , which varies in intensity. The most intense phaeomelanin colour is Irish Setter red.  The default colour is probably golden, with different genes causing it to be more or less intense (ie telling the cells to produce a higher amount of pigment particles, so making the colour stronger, or a lower amount, making the colour weaker, so lighter).


Isabel color

Black and Gray/Bleu color

Creme and Brown color

Gray color

So far so good, but this doesn't seem to explain all the coat colours in dogs - how about white


White is not really a color, so white hairs on animals is not caused by pigment, but a lack of pigment. It is a lack of both eumelanin and phaeomelanin. White areas on animals are simply caused by the pigment cells, there is no creation. Sometimes the entire animal affected by e/e-gene (read Canadian Shepherd), or as in albinos ( ca gene) where no pigment is present and sometimes just parts of the dog without pigment is influenced, like dogs with white markings where si and sp gene responsible. It can affect the production of eumelanin in the eyes and forenose which nose pink and the eyes to blue change (or red in the right albinos).  This type of white has no effect on eumelanin, so any black / brown / blue / isabella areas on the coat will remain pure, and the eyes and forenose, paw pads will be too. Research has shown that a recessive e allele at the Extension (E) gene is at least partially responsible for cream and white coat color by the Canadian Shepherd. The (E) gene, now identified as the Melanocortin 1 receptor (MC1R) gene, is one of the two genes known to code for alleles that are absolutely fundamental to the formation of all German Shepherd Dog colored coat variations. When the recessive allele is inherited from each breeding pair parent, the e/e genotype offspring of certain breeds, including white coat dogs from German Shepherd breed lines, 
always have cream or white colored coats. 



The colour genes in dogs do two things - they determine the eumelanin and phaeomelanin colours/shades, and they control the distribution of these two pigments. They tell certain cells to produce eumelanin, others to produce phaeomelanin, and sometimes they tell them to not produce pigment at all.  Exactly which cells are told to produce what is determined by the exact set of genes, although it can be random to a certain degree (eg puppies may have slightly different white markings to their parents, or patches in different places). Sometimes genes can even tell cells to switch which type of pigment they are producing every once in a while.  This means that as a hair grows, it becomes banded with black and red, because the cell produces black (eumelanin) for a while, then changes to red (phaeomelanin), then back to black, etc. It's a bit like when you highlight your hair and after a while the roots start to show through. The overall colour of an animal with this sort of black and red banding will generally be a muddy brown from a distance, and close up you will be able to see the black parts of the hairs. It's called agouti (aw-gene) , and it's the colour of wild rabbits and mice, as well as a large amount of other mammals.  It's popular amongst wild animals because it provides very good camouflage. It also occurs in dogs, but it looks a bit different (like the colour of a wild wolf rather than a rabbit) and isn't very common.

All of this probably seems a bit confusing at this point.  Many genetics sites gloss over the types of pigment for precisely this reason. However, I've decided to include this information because it really is crucial to your understanding of dog genetics (and the genetics of other animals too).  Once you've grasped the idea of the two types of pigment, and understand how they work, the rest should be plain sailing.

The collective code of the Newfoundlander: : ayatKBkbrkyBbCcaDdEESsispTt is heterozygous (impure).

This code gives the genotypic yellow (
ay) pigment over the whole body, black/liver and Tan (at), dominance black is (KB), brindle (kbr), Fawn (ky) black (B), brown (b)
Albinism (ca), intensief  intensieve pigment ( D ), grey/diluted ( d ) , eumelanine pigment ( E ), and white ( si, sp ) - in several forms inherit - and (T) for Ticking.

The yellow does not have been confused with small e-gene, like the Labrador, in homozygous form (ee) gives this a junction of a black and a brown 
Ee x Ee) a quarter (ee) yellow Labrador and creme/white by the Canadian Shepherd (e-gene = light

The K-locus : KB kbr ky

Code KBKBBBCCDDEESStt is a dominance black dog.
KBkbrBBCCDDEESStt is a black dog and carrier of brindle.  
kbrkbrBBCCDDEESStt is black and leader of brindle. ( see foto )
kbrkyBBCCDDEESStt is black brindle and carrier of Fawn
atatkykyBBCCDDEESStt is black and Tan and leader of Fawn only on the Tan points.
KBBbCCDDEESStt is a black dog and recessieve of brown.
KBKBbbCCDDEESStt is a brown dog .  
kbrkbrbbCCDDEESStt is brown and leader of brindle. (orange)
kbrkybbCCDDEESStt is brown brindle and carrier of Fawn
kybbCCDDEESStt is brown and Tan and leader of Fawn only the Tan points.

The genes on
B/b-locus and Ee-locus are complementary, such as keys and lockers or as yin and yang.

In some breeds, such as the Newfoundland dog, all dogs are homozygous for KBKB. In such races, only the Ee locus determines whether the dog gets than black pigment or light pigment. The KBKB genotype is fixed and the B locus determines whether they are black or brown.
Although C.C. Little has initially thought that the A
s-gene was also the cause of "Solid" black, but by DNA testing showed that here the
KB allele is responsible for, so we recommend actually  KBKB allele for this kind of black.  But it doesn't mean that no white marking may be present.
Some people call this
"black self".
The K-
locus is located on an entirely different chromosome than the agouti locus.

A brown Newfoundlander (si x si-gen) with white "Irish spotting" (code KBKBbbCCDDEEsisitt) (see photograph). 
Don’t confused this with a white-brown Newfoundlander (

§         Code KBKBbbCCDDEEspsptt (wrong colour) (see photograph).

§         Code KBKBBBCCddEESStt = grey with black nose capsule, and foot soles and lips (wrong colour) (see photograph).

§         Code KBKBbbCCddEESStt = grey with brown nose capsule and foot soles and lips (wrong colour) (see photograph).

Is it now a white-black or black-white Newfoundlander (see photograph)?

The proposition that the piebald "sp-gene" is not responsible for a white mark, but for a white dog with coloured marks.
I can not explain in the genetics.

First, even for all clarity, the code: KBKBBBCCDDEEspsptt , therefore in my opinion a black-white Newfoundlander and I will try to explain why I think so. 
The genes
BB ( dark black pigment) and EE (proportional partitioning) are responsible for black, CC for complete pigment and DD for intensive pigment. Therefore a black dog which is stands under influence of spsp-gene for fur ,to hold back the white on some part of the coat
, and tt-gene for hold the Ticking.

S-locus: S, si, sp

S = stop of white, si, sp, are genes which brings white in order of dominance, therefore everybody talks about a white factor and white is NO COLOUR but the cause that these genes the pigment widen-honoured which is in a dog present, as above already mentioned.

Everybody is talking about a black dog with no pigment mark "so white". 

si-gene  do that mostly in symmetrical pattern form "white feet, tail point, chest mark".

sp-gene do that without pattern, in fact random white  marks are falling in the coat.

The genes on the S-locus are not complementary to other genes in the code such as the KBB/b/D locus and the Ee locus, such as locks and keys or Yin / Yang and can move independently of the other genes between the code and manifest themselves "as exemble" as a dominant dog :                 "  KBKBBBCCDDEESsitt " but carriers of the white - si or sp factors.

NOTE: Some of these alleles at the S locus may have incomplete or co-dominance
A dog that is (si-sp) may show both patterns simultaneously, for example (see photograph).

There may be genes independently modify (modifications change a part of an assignment, so the interpretation and implementation differ) from the original manner specified place for example in the SS-genes (white patch) or a black and white litter (spsp-genes) where there is not one in the same pigmentation and the Irish Spotting marks (sisi-genes) is one white paw can be more white than the other instance.

And Ticking -TT-gene - gives therefore mark colours in the white and tt-gene hold back this.

If you cross therefore two black dogs with the white inherit factor, the recessive sp-gene = KBKBBBCCDDEESsptt  x  KBKBBBCCDDEESsptt  you get :

§         25% black = SS pure breed ,

§         50% black = Ssp impure breed

§         25% fur = spsp pure breed for fur, without fixed pattern (see photograph)

§         A cross of two furs from the white factor spsp x spsp = 100% fur without fixed pattern

§         Two black with white mark and white tail point, feet and white chest mark: sis sisi = 100% have this property in more or less in the symmetrical pattern .

§         Two black with the Ssi x Ssi = 25% blacks SS pure, 50% are black recessive Ssi impure, 25% sisi  pure black with white marks.

  • In theory, genotypical dominance black fur with spsp  x  SS =  Ssp = 100%  black but 100%  recessive for sp - gene.

    Conclusion: this is a black Newf where the full pigment is
    back in the coat  =

    Preferably in the fur 2/3 white marks" (see photograph). 

  • Code for Albinos is: KBKBbbcacaDDEESStt  of brown parents or KBKBBBcacaDDEESStt  from black parents.
    Albinism absence of dye (pigment, see above about the gene).
    In the typical albino lacks pigment in the retina and the iris of the eye and soles
    / paw pads, nose , pink lips and eyelids.
    Albinism is often paired with deaf and / or blindness.
    (see photograph).

Take a brown / grey dog with the dd gene (wrong colorr) a clear genotypical dominance brown dog = SS
KBKBbbCCddEESStt   x    KBKBbbCCDDEESStt = 100% brown =  KBKBbbCCDdEESStt  = 100%  recessive = d- gene for dilution  so gray.
D is dominant over d and you see the pigment is back and so a brown dog.  
See example
below combined with dark brown puppies
in the F1 generation. In the F2-generation your got  +/_  25%  gray ones. 

The theory above is as good on the colour grey X black.

      I want to explain the behaviour of the travelling of the genes of the white si-gene "Irish Spotting" what is travelling for generations and is coming out if both parents are carrier (= 25% who has it). Yes it is also travelling even after using dominant black dogs ( KBKBBBCCDDEESsitt ) except with dogs who are ( pure black by breed ) from this si-gene.

KBKBBBCCDDEESStt  =  pure black by breed. You will know that after mating the dog with a black and white one ( KBKBBBCCDDEEspsptt ). 
If all the puppies are black
( KBKBBBCCDDEESsptt ) but carrier of the sp-gene, then he/she is real dominant = totally black = KBKBBBCCDDEESStt. 

If you mate a real dominant black dog.
( KBKBBBCCDDEESStt  = genotypical black ) with also a dominant black dog but carrier of the si-gene 
( KBKBBBCCDDEESsitt ) then 50% of the puppies are carrier and so is this si-gen travelling from generation to generation.

§         Did you know that all colour code combinations of the Newfoundlander in pure breed and impure breed (homozygoot:  KBKBBBCCDDEESStt and heterozygote "collective code" :  ayatKBkbrkyBbCcaDdEESsispT amounts, so far as I can see, more than a 500.

§         This way there is still a couple wrong colour codes.

§         Code ayayKBKBBBCCDDEESStt, is a black-yellow/brown (agouti Fawn) pigment dog "looks like a brown one"with a black forenose and mirror/mask.

§         Code ayatKBKBCCDDEESStt, is a black yellow/brown = agouiti Fawn pigment dog "looks like a brown one "with a black forenose and mirror, but by heterozygote form because the ay-gene are dominance over the at-gene.

§         This one with the small b for brown on the locus ayatKBKBbbCCDDEESStt, therefore a liver colour with more yellow "looks like a yellow dog" with a liver coloured nose capsule and mirror. 
Do not confuse this dog with a brown dog which coat is
influenced by salt water of the sea, or the sun that gives a temporary
pigment colouring.

§         Code atatkykyBBCCDDEESStt, is for Black and Tan "brown-yellow colouring and looks like crème colour to halfway the legs and/or the mask " 
(see photograph).

§         Code atatkykybbCCDDEESStt, is for liver and Tan "brown-yellow colouring at brown, looks like "crème-grey" also called “sandy” on the legs and/or the mask" (see photograph).

§         Code KBKBBBCCDDEEspspTt, is white-black with Ticking (heterozygote “impure” form).

§         Code KBKBBBCCDDEEspspTT, is white-black with Ticking (homozygote “pure” form
 (see photograph).

  • Code kbrkbrBBCCDDEESStt , this are like on a fenotypical Dutch Shepherd brindle ( zie fotograph ).

  • Code  kbrkbrbbCCDDEESStt = leader of ( orange ) brindle.

  • Code , KBKBBBCCddEEspsptt is white-grey without Ticking.

  • Code for Albinism is : KBKBbbcacaDDEESStt from brown an from black parents is: KBKBBBcacaDDEESStt.  (see photograph)

However, the indication is that black/liver only shows themselves in homozygote “pure” form, therefore both parents are carrier of the at-gene, ( is also present by the Tibet Dog,  see also , “ Pre history / Primal Source /Tibet Dog"                            

 By strong INBREEDING there is a great risk in the future to get back this wrong colour, especially with combinations of the 
Old American / and Dutch line.
For example :

The hair:

K = short hair  = dominance

k = long hair

F = phenotypic calculation of the Augustine monk Gregor Johan Mendel (Czechia 1822 --.1884).

By means of F1-generation becomes 100% short hair, therefore K x k = 100% short hair = Kk.

In the F2 generation becomes the KkxKk = 75% short hair ( of which, 25% KK is pure breed, 50% impure breed Kk, Kk ) and  25% long hair  kk  homozygoot  pure breed.  ( 3: 1 ) .

Conclusion: Long hair x long hair = is therefore ALWAYS long hair.  

The Mythe of brown

There lives a myth within the cynology  color inheritance that the combination of brown on brown pigment dilution would occur, but brown, like black too dark pigment, and now owns one dog once more pigment than others. This is in large breeds recognized by the lips
`Flesh reddish discoloration on the tongue 'and' brown or black discoloration or` eyes.

Tyrosinase Related Protein 1 (TYRP1) is the gene responsible for brown coat colors in dogs (and mice and cattle and cats). 
Three different mutations in this gene
( bs,bd,bc ) can produce brown. We have studied a litter of Cheasapeake Bay Retrievers who had different shades of brown, but all had the same genotype of b alleles. Based on the brown Newfs they have tested
the bc allele, is the typical brown allele in this breed such as Sheherazade of The End of the Line
her genotype is bc/bc, E/E.

Her genotype is bc/bc, E/E.  
The Newfoundland is
homozygous for the rarer brown mutation. 
Dogs which have any of
these mutations on both chromosomes would also be brown, i.e. bs/bd.

Dilution arises at the hands of small d the recessive gene that prevents the dog from a phenotypically black black gr
ey / blue and will see a brown dog brown grey / isabella will see.

If you have two dark brown dark brown crosses are you going to record but the chance of a light brown, natural to dilute 25% (Mendel's Law) is present as both parents are carriers "D
d" of the small d-Gene
See the link to the most beautiful brown bitch (from brown x brown) of Switzerland.

Experience teaches us that much (more) lighter tan resulting from the combination black x brown. And that has everything to do with the fact that "it especially American dogs" with gr
ey bred and thus more carriers.
So the
risk of dilution is black as they grow bigger if not as brown.

HealthGene Dog Coat Color DNA Testing
With the new science, the dog can be tested on KBBb on the Dd-gene and sp-gene.

Eye colour

The standard color of the eyes is brown for dogs.

However, a number of genes affect the eye color.
Each of the pigments is a "standard" color, and it can then be modified by different genes
as described above is particularly the
KBB/b-gene and gene-Dd. The iris color is produced by layers of pigment, and brown eyes like most dogs (and like many people have) caused by black eumelanin these layers. When a dog has changed the production / limited eumelanin, the irises are not able to full strength to produce eumelanin. This means that the dark pigment in the eye is converted into a lighter brown color. Besides the fact that the coat that eumelanin is present in other parts of the dog that need color, especially the eyes (irises), including amber or gold. This is actually not a color you see in people and therefore difficult to preview it. The best example is that there are eyes of hazel, which is a mix of green and brown and look therefore lighter than brown eyes. Dogs can not produce the green color only brown and blue. The eye color depends on the amount of production levels of the pigment (melanin) in the irrissen the dog regardless of coat color. Only the pigment melanin is found in those areas, and therefore only genes that affect melanin can affect the eyes and / or the nose color. A light eye can be obtained by the recessive trait (for) parents ("carriers of less dark pigment"). A brown dog, the color is often lighter and is only 3 years with pigment strength. This is an indication that light is inherited eye dominant, even a black Newfoundland dogs with lighter eyes and is very disruptive to the expression. Amber eyes usually occur when the eumelanin in the coat is diluted or changed by recessive genes in the KBB/b or Dd series. In other words, all the brown dogs (bb) have amber eyes, and so are gray / blue dogs and Isabella (dd). Occasionally have dogs with black pigment amber eyes, but in general they are only found in brown and dilute specimens. Amber eyes range from light brown (overlaps with lighter eyes, sometimes in black-pigmented dogs to yellow, yellow-green or gray. Amber is an orange-yellow color that's named for the color of amber. 
Amber is usually warm yellow to dark red in color, but there are more green, blue or black colored species

Occasionally have dogs with black pigment amber eyes,
but in general they are only found in brown and dilute specimens.

Conclusion: that light eye obtained by a deficiency of melanin (dark pigment) in the area of the iris so the dog may have lighter eyes. This also applies to dogs of the light pigment pheomelanin which is solely responsible for the color of the coat, 
the sight here is the melanin pigment in those areas responsible.

Copyright ©
The End of the Line
Auteur: Koos van Rijn

Hoek van Holland

Note :

The first gene that causes at least some spotting patterns in dogs was identified and published in 2007
Furthermore potential mutations causing some forms of spotting have been identified in this gene.
This gene is MITF, microphthalmia associated transcription factor. 
MITF is a critical gene in the pigmentation pathway.
(Prof. L. Andersson/mr Karlsson)
Department of Medical Biochemistry and Microbiology, Uppsala University, Uppsala, Sweden

MITF has been shown to cause spotting of the piebald or random type in crossbreds by Rothschild and colleagues. It also caused the Landseer pattern in Newfoundlands. However the causative mutation was not identified in their study.

Another study presented by Karlsson on behalf of the Broad Institute and the Univerisity in Uppsala, Sweden focussed primarily on Boxers. They have shown that MITF is the gene causing the solid, flashy, and virtually white forms in Boxers and Bull Terriers. They suggest that two mutations in the promoter region of the MITF form that works in the pigmenation pathway (MITF-M) are necessary for white in these breeds. One of these is called a SINE (short interspersed nucleotide element) and the other is a string of repeated alleles that varies in length in different spotted dogs, which they call a "Length Polymorphism or LP". Their data suggest that the LP is found to be longer in breeds with white markings than in dogs with no white.

In cattle and pigs, some spotting patterns have been shown to be caused by the KIT gene.
Although the spotting pattern of a Landseer Newfoundland looks somewhat like the spotting pattern of Holstein dairy cows
this pattern was NOT caused by the KIT gene, based on a recent study we completed

Therefore Landseer Newfoundland dogs are s/s based on the nomenclature of Mr. Winge or sp/sp using the nomenclature of Mr. C.C. Little.  
genotype is SINE / SINE at MITF.

This form of spotting appears to be inherited as an autosomal recessive allele of a number of MITF.
Testing for carriers of piebald spotting in several breeds is now available at HealthGene.

White Undersides or Irish Pattern

Clarence C. Little (1957) discusses the "pseudo-Irish Spotting" phenotype in the flashy Boxer and then raises the perplexing pattern he refers to as "Irish spotting", in dogs such as Basenji. He also suggests that plus and minus modifiers are likely. He says the si causes this pattern and that is recessive to the S allele for solid, but dominant to the allele causing piebald spotting (sp).

MITF and White Spotting in Dogs: A Population Study

This study was designed to determine if one of the variants found in our laboratory, or previously reported in microphthalmia-associated transcription factor (MITF), was associated with one or more spotting patterns in dogs. None of the rare variants found in the coding sequence consistently occurred in dogs of any particular spotting pattern. However, an insertion of a short interspersed nucleotide element (SINE) over 3000 bp 5′ of the MITF-M start codon (Karlsson et al. 2007) did fit with random spotting in many dog breeds. Most (319/324) dogs of 45 breeds fit 1 of 2 inheritance patterns. All dogs that were homozygous for the SINE had white markings that either covered at least the ventral surface (mantle pattern) or most of the body (piebald or extreme white spotting). In most breeds, dogs heterozygous for the SINE insertion were solid colored or had minimal white, such as on the toes, but in some others, heterozygotes had white undersides, often with a white collar in the pattern called pseudo-Irish by mr Clarence Charles Little (1957). However, none of the 15 dogs of 5 breeds in which all individuals have markings known as Irish spotting had the SINE insertion. Finally, we studied RNA expression in skin. The 2 MITF-M forms, M+ that contains an extra 18 bp that adds 6 amino acids between exons 5 and 6 and the M− form, were present. MITF-M is considered to be specific to melanocytes but was found in skin from a white Samoyed.  A putative pseudogene containing exon 1M was also identified.

As a resource for building a dog genetic map and as a tool to study the genes responsible for behavioral and morphological differences in the dog, an intercross was created between a male Border Collie (b) and a female Newfoundland (a). The Newfoundland parent had a small patch of white on the chest and was otherwise completely black. The Border Collie used in this cross had markings characteristic for the breed - black with white markings on the face, chest, neck, tail tip, ventral abdomen, all four digits, and extending up the front legs to the carpals. These markings have many similarities to the white spotting patterns of other mammals. The Border Collie's sire and dam had the same markings, consistent with homozygosity for the causative loci. This cross provides the opportunity for analyzing the inheritance of the white spotting pattern exhibited by the Border Collie. Six F1 animals were produced which had medium-sized white patches on their chests. These six dogs were intercrossed to produce 25 of F2 progeny. In the F2 generation, 7 out of 25 had markings like the Border Collie parent, consistent with the phenotype being caused by a recessive allele of a single locus.
This form of
spotting appears to be inherited as an autosomal recessive allele of a number of MITF

So with the Mendelian calculation you may say therefore in the white factor S-Locus : SS x sisi = 100% Ssi in the F1 generation.
And in the F2 generation you gat Ssi x Ssi = 25% = SS = black , +/_  25% = have the sisi pattern "Undersides or pseudo Irish Spotting" of the grand-father. The other 50% = Ssi = black and carrying "recessieve" for Undersides or pseudo Irish-Spotting.

The SINE insertion 5′ of MITF-M first described by Karlsson et al. (2007) was associated with white markings in many and diverse breeds in this study, suggesting that it is an “old” mutation. There is considerable debate about the age of particular breeds.

  However, the Chinese Shar-Pei, Akita, and other Asian dogs are typically considered to be among the oldest breeds ( H.G Parker et al. 2004
" Provisional hypothesis The Tibetan mastiff  have a lot compare with the Newfoundland" .

The SINE insertion has been found in individuals with white markings in these breeds.

Science Links about this issue are :

Undersides or Pseudo Irish Spotting (2000)

MITF and White Spotting in Dogs: A Population Study (2009)

Prof. Dr. Sheila M. Schmutz University of Saskatchewan Saskatoon, Canada S7N 5A8 updated in january 9, 2012