The Merle Reference Centre

Merle is caused by a SINE insertion in the PMEL gene. What matters is not whether the insert is present but how long it is. Labs that offer length-based testing report that length in base pairs, and the number places the dog in one of seven classes.

The right-hand column is the phenotype when the other parent contributes a plain non-merle allele. Two merle alleles together behave differently, which is what the calculator models.

The single most useful thing a breeder can know about merle is where the visible threshold sits. It is between Ma and Ma+, not at the bottom of the scale.

• Mc, Mc+, and even a single Ma produce no visible pattern at all. The dog looks solid.
• Two Ma together begin to express, weakly, with no body white.
• Ma+ is a genuine threshold. Near the top of its range it shows a muted, often diluted pattern. Lower down it may show nothing.
• M and Mh express clearly.

This is why a solid-looking poodle can produce merle puppies, and why no coat inspection can rule merle out. Only a test can.

Most labs still report merle as a simple presence or absence, typically as M/mor “merle detected”. That answer is not wrong, but it collapses seven classes into two categories and discards the number that determines risk.

A binary result cannot distinguish a cryptic Mc from a harlequin Mh, even though those two dogs carry entirely different consequences for a pairing. If you are breeding merle deliberately, ask your lab whether they report the base-pair length, not just the presence.

Because reported values vary between labs by roughly two base pairs, a result sitting near a class boundary should be read as the more cautious, longer class. The calculator applies that tolerance automatically when you enter an exact value.

“You can’t see it” is not one situation. Three different, non-overlapping things can keep a merle invisible, and they carry different implications for a pairing.

• Cryptic (Mc, Mc+) — the insert itself is too short to ever produce a pattern, no matter what it’s paired with. A property of the allele.
• Non-expressing (a lone Ma or Ma+) — the allele is pattern-capable, but a single copy against a weak partner isn’t enough to show. Pair it differently and it can express. A property of the pairing, not the allele.
• Hidden— the allele is fully expressing, but there is no eumelanin in the coat left for it to pattern: an ee dog, or a sable with high phaeomelanin expression. See the next section.

All three look the same at a glance, a solid-looking coat, but only a DNA test tells you which one you’re actually looking at, and that distinction matters for what the dog can produce.

A dog that is ee at the E locus produces no eumelanin in its coat. Merle patterns eumelanin, so on an ee dog there is nothing for it to pattern. The coat shows no merle whatsoever.

A sable dog with heavy phaeomelanin expression can hide merle the same way, for the same reason — not because the dog is ee, but because there is too little visible eumelanin left in the coat for merle to pattern. The result looks the same either way: a coat that reads entirely clear over a fully expressing merle allele.

The allele is completely unaffected. The dog is a merle, passes it on at the normal rate, and can produce a double merle if bred to another merle. This is one of the most common routes to an accidental double merle litter, because both parents looked entirely clear.

The poly-A tail of the insert is unstable. It shortens readily during cell division and has never been confirmed to lengthen. When that happens in body cells you get a somatic mosaic, often visible as tweed. When it happens in the reproductive line you get a germline mosaic.

A germline mosaic carries a second, shorter merle allele in its sperm or eggs that may never appear on its own DNA report. It can therefore produce a puppy with an allele neither parent was known to carry.

Roughly ~18% (about 1 in 5–6) of merles are mosaic, and ~50% of mosaic dogs' minor allele falls in the Mc range. Those are population figures, not per-litter odds, and that distinction matters: there is no honest per-cross probability for mosaicism. Any tool that offers you one is inventing it. The calculator treats it as a caution layer rather than a number.

When two expressing merle alleles meet, pigment is deleted across the body rather than patterned. Because the same pigment cells are involved in the inner ear and in eye development, extensive pigment loss carries a genuine risk of deafness and of ocular defects.

Double merle is not the same thing as homozygous.Double merle means both alleles fall in the pattern-capable classes (Ma, Ma+, M, Mh) — it says nothing about whether the two alleles match. M/M is both homozygous and double merle. M/Mais double merle but not homozygous — two different alleles, both pattern-capable. A homozygous Mc/Mc, by contrast, is not double merle at all, since neither allele is pattern-capable. What drives the risk is which two classes are paired, not whether they happen to be identical.

BAER testing (brainstem auditory evoked response) is the only objective way to establish whether a dog hears, and it detects unilateral deafness that behavioural observation reliably misses. A dog deaf in one ear usually behaves entirely normally. If a pairing could produce extensive white, BAER testing the resulting puppies is the responsible step, and an ophthalmologic examination is worth discussing with your veterinarian.

Not every merle pairing carries this risk. The calculator separates the genuinely critical combinations from the ones that merely need care, so the warning stays meaningful.

Cryptic merle (Mc, Mc+)

The two shortest classes. Their insert is too short to ever produce a visible pattern, no matter what allele they are paired with — a property of the allele itself. A cryptic dog is still genetically a merle and still passes the allele on; it just never shows on its own.

Non-expressing merle

A single Ma or Ma+ paired with a non-merle or cryptic allele. Unlike Mc/Mc+, these are pattern-capable lengths — a property of the pairing, not the allele. Two Ma together begin to express weakly, and Ma+ shows once paired with another expressing allele.

Hidden merle

A merle whose pattern is masked by the dog’s base coat, not by the allele itself: an ee (red, apricot, or cream) dog, or a sable with high phaeomelanin expression, has too little visible eumelanin left in the coat for merle to pattern. The allele is fully expressing genetically — there is simply nothing to show it on.

Ghost merle

An informal synonym for cryptic merle. Name the allele class (cryptic Mc or Mc+) instead of the vague label.

Atypical merle

The Ma and Ma+ classes. Historically described as odd or washed-out merles, now understood as a length effect rather than a separate gene.

Harlequin (Mh)

The longest class, 269 bp and above. Expresses as mostly solid pigment with random merle patches and can carry extended white. The most unstable class, and every Mh pairing warrants expert review.

Double merle

A dog carrying two pattern-capable alleles (Ma, Ma+, M, or Mh) — a genetic definition, not a description of matching alleles. Not the same as homozygous: M/M is both homozygous and double merle, but so is the heterozygous M/Ma, while a homozygous Mc/Mc is neither pattern-capable nor double merle at all. Risk scales with which two classes are paired, not with whether the two alleles match.

Tweed

A patchwork of differently toned areas within a merle coat, produced when the insert shortens in some cell lineages during development.

Somatic mosaic

A dog in whom the merle insert shortened in some body cells after conception. The coat may show more than one merle expression, but only the germline determines what is passed on.

Germline mosaic

A dog carrying a second, shorter merle allele in its reproductive cells. It can produce puppies with an allele that never appeared on its own DNA report, which is why no calculator can give an exact per-cross probability.

Minor allele

The shorter, secondary merle allele found in a mosaic dog, alongside the primary reported one.

SINE insertion

The short interspersed nuclear element inserted into the PMEL gene that causes merle. Its length, specifically the length of its poly-A tail, determines the class.

Poly-A tail

The unstable run of adenines at the end of the insert. It shortens readily and is why merle length changes between generations, and why it has never been confirmed to lengthen.

• Whether a specific dog is a merle. Only a DNA test can establish that.
• How much white a given puppy will carry. The amount of white is influenced by modifiers that are not currently testable.
• Whether a mosaic parent will pass its hidden minor allele in any particular mating.
• Whether a dog hears. That requires BAER testing, not observation and not a pedigree.

The length-allele framework follows the work of M. Langevin and colleagues on PMEL SINE insertion length. The pairing risk bands used in the calculator are Poodle Genetics Lab’s own precautionary framework: no published study assigns risk bands to M-locus pairings, and Langevin’s 181-dog, 14-breed cohort did not include Poodles. Treat a band as a structured starting point for a conversation with a geneticist or reproductive veterinarian, not as a validated clinical score. This page is educational and does not replace laboratory testing or veterinary advice.