Koi color genetics: white, red, black, and yellow inheritance

An ornamental-fish breeding researcher walks through the genetics of color in koi — the four base colors, the modifiers, the inheritance patterns, how breeders plan matings to produce specific varieties, the role of the white ground, the red and black pattern genetics, and the four mistakes that turn a controlled koi breeding project into a project that produces mostly kawarimono.

Koi breeding is a colour business. The fish is functionally identical to a common carp, but the price difference between a 50-cent fingerling and a $50,000 show fish is entirely a function of the colour pattern, the colour quality, the skin quality, and the body conformation. The hobbyist who wants to breed koi for a particular variety — say, a Showa with a clean white ground and bold red and black pattern, or a Kohaku with a crisp step pattern of red on pure white — needs to understand the genetic architecture of the four base colours and the modifiers that act on them. This article is the working breeder's view of that architecture.

We will go through the four base colours (white, red, black, and yellow), the major colour-pattern genes, the inheritance patterns that breeders have documented over the last century, the practical strategy for producing predictable varieties, and the four mistakes I see in beginner koi projects that produce mostly kawarimono (the catch-all "miscellaneous variety" category).

The four base colours and their genetic basis

White (shiro). The white in koi is not actually a colour — it is the absence of colour. The skin cells (chromatophores) are present but contain no pigment, so the underlying skin tone shows through. The white ground is critical to the appearance of Kohaku, Sanke, Showa, and most other varieties. The genetics of the white ground is governed by a series of alleles that breeders conventionally label as:

  • W: dominant white, which produces a uniformly white fish
  • w: recessive "non-white," which allows other colours to express
  • The classic Kohaku and Sanke fish are heterozygous (W/w) at this locus, which produces the white ground plus the ability to show red or black patterns

A truly homozygous W/W koi would be a completely white fish (Platinum Ogon). A w/w koi would have no white ground and would express the underlying colours all over the body, which is not what most breeders want. The Kohaku and Sanke and Showa are all W/w.

Red (aka, beni). The red colour comes from carotenoid pigments deposited in the chromatophores. The genetics of red is governed by a series of alleles, conventionally:

  • R: dominant red pattern
  • r: recessive non-red (no red, or very limited red)

The red colour is also affected by modifiers that determine the intensity, the depth, the edge quality (crimson vs orange vs scarlet), and the size of the red patches. The most prized red is a deep, lustrous crimson with crisp edges and uniform colour across the entire patch. The genetics of red quality is complex and involves many modifier genes, but the basic presence/absence of red follows a relatively simple dominant pattern.

A Kohaku is a koi that is W/w (white ground) and has at least one R allele (red pattern). A Sanke is a Kohaku with the addition of small black sumi spots. A Showa is a Sanke with more extensive black patterning.

Black (sumi). The black colour comes from melanin-producing cells called melanophores. The genetics of black is more complex than white or red because:

  • There are at least three different "types" of black in koi: sumi (deep, glossy black), nisai sumi (a softer black that often changes as the fish matures), and bonyan sumi (a black that appears and disappears with age, particularly in Showa).
  • The pattern of black is highly variable and is affected by many modifier genes.
  • Some black patterns are stable from year one; others undergo dramatic change as the fish matures (the famous "Showa no nisai" problem, where the two-year-old Showa looks very different from the four-year-old Showa).

The conventional genetics:

  • B: dominant black (sumi)
  • b: recessive non-black
  • M: a modifier gene that affects the depth and quality of the sumi
  • The B allele in the homozygous state (B/B) tends to produce very heavy, often undesirable black coverage. The heterozygous (B/b) is what most breeders want, producing a controlled amount of sumi in the right places.

A Sanke is W/w, R/r (or R/R), and B/b with limited sumi. A Showa is the same with much more extensive sumi and the sumi traditionally wrapping around the body in a "motoguro" pattern at the base of the pectoral fins.

Yellow (ki). The yellow colour comes from carotenoid pigments as well, but a different metabolic pathway from red. The genetics of yellow:

  • Y: dominant yellow
  • y: recessive non-yellow

A pure yellow koi is a Ki Utsuri or a Yamabuki Ogon. The most famous use of yellow in classic varieties is the Ki Bekko (white with yellow and black spots, less common) and the yellow ground varieties.

The yellow colour is also subject to dietary influence — koi fed carotenoid-rich diets develop more intense yellow. A pale yellow koi can sometimes be improved with diet, but the genetic potential sets the ceiling.

The major varieties and their genetic formulae

Let me work through the standard varieties, the genetics that produce them, and what the breeder needs to know to plan matings.

Kohaku (white with red pattern). W/w + R (at least one). The white must be clean, the red must be deep and crisp, the pattern must be balanced and free of "secondary" red spots (shimis) that appear as the fish grows. The Kohaku is the foundation variety in koi breeding, and the most difficult to do well. A poor Kohaku has muddy white and soft red. A great Kohaku is one of the most beautiful fish in the world.

Taisho Sanke (white with red and black). W/w + R + B (limited). The Sanke adds small black sumi spots over the back, ideally balanced with the red pattern. The sumi must be stable — not the nisai sumi that disappears as the fish matures.

Showa Sanshoku (white with red and extensive black). W/w + R + B (extensive). The Showa is similar to the Sanke but with much more black, including the traditional motoguro (black at the base of the pectoral fins) and sumi that wraps around the body. The Showa is the variety most subject to nisai sumi changes.

Tancho (white with a single red spot on the head). A Kohaku variant where the red is restricted to a single round spot on the head. The genetics of the tancho pattern is not fully understood but appears to involve a separate modifier gene that restricts red expression to the head. The Tancho is highly prized in Japan and the genetics are tricky — breeding two Tanchos does not reliably produce all-Tancho offspring.

Ogon (solid metallic colour). A koi with a single uniform metallic colour — typically Platinum (white Ogon), Yamabuki (yellow Ogon), or Orange (orange Ogon). The Ogon phenotype is governed by a dominant metallic gene (M) that produces the reflective scales. The ogon is homozygous in most lines.

Utsuri (black with red, white, or yellow pattern). A koi with a black ground and a contrasting colour (red, white, or yellow) pattern. The Utsuri genetics involve extensive black (B/B or B/b with modifiers) plus a colour pattern gene.

Bekko (white, red, or yellow with small black spots). A simpler version of the Sanke or Utsuri, with small sumi spots and a clean white, red, or yellow ground.

These are the major varieties. There are dozens of others, but most are variants or combinations of these.

How inheritance actually works in practice

Koi breeders have been documenting crosses for over a century. The patterns are well established.

Kohaku × Kohaku: 100% Kohaku (assuming both parents are W/w, R/r). The offspring vary in pattern quality, intensity of red, and sharpness of edges, but the variety is consistent. This is the standard pure-line Kohaku breeding.

Kohaku × Sanke (or vice versa): Offspring are a mix of Kohaku, Sanke, and occasionally fish with very heavy sumi. The Sanke parent contributes the B allele, which is expressed in about half the offspring. The breeder who wants to introduce sumi into a Kohaku line uses this cross, then selects for the best Sanke-patterned offspring for the next generation.

Kohaku × Showa: Offspring are roughly 50% Showa and 50% Kohaku, with some intermediate patterns. The Showa parent contributes the heavy sumi modifier. The breeder who wants to "improve" a Kohaku line by adding the depth of colour from a Showa uses this cross, but then has to manage the resulting line carefully to avoid losing the Kohaku's clean white.

Sanke × Sanke: Mostly Sanke, with occasional heavy-sumi offspring and occasional fish with very little sumi. The breeder selects for the best Sanke-patterned offspring.

Showa × Showa: Mostly Showa, but the nisai sumi problem becomes significant. The two-year-old Showa can look completely different from the four-year-old Showa. The breeder has to either grow the fish to four years before culling (expensive in space and time) or learn to read the nisai sumi patterns and cull at one year (a skill that takes years to develop).

Ogon × Ogon: All Ogon. The metallic gene is fully dominant.

Kohaku × Ogon: Offspring are roughly 50% Ogon-metallic and 50% Kohaku-non-metallic. The metallic gene (M) is dominant. The breeder who wants to introduce the metallic quality into a Kohaku line uses this cross, but the resulting "Kinsui" (metallic Kohaku) is technically a different variety.

The pattern that emerges is that koi colour genetics is more like cattle breeding than like Mendelian pea experiments. The major colour genes follow relatively simple dominant/recessive patterns, but the modifiers that determine pattern quality, colour depth, edge crispness, and skin quality are many, mostly uncharacterised, and heavily influenced by selection over generations.

Planning a breeding program

A serious koi breeding program has three elements: the brood stock, the mating plan, and the selection criteria.

The brood stock. A small number of high-quality fish of known variety, with known parentage if possible, and ideally with known performance as parents. The breeder who buys unknown fish from a mixed auction is gambling. The breeder who buys from a known line with a documented history is investing. Most serious breeders maintain their own brood stock rather than buying in.

The mating plan. The breeder decides which varieties to produce, in what proportions, and which parent fish to use. The plan is based on the genetics above. A typical program might include:

  • 60% pure-line Kohaku (Kohaku × Kohaku) to maintain and improve the Kohaku line
  • 20% Sanke (Kohaku × Sanke or Sanke × Sanke) to maintain the Sanke line
  • 10% Showa (Showa × Showa or Kohaku × Showa) to maintain the Showa line
  • 10% experimental crosses (e.g. introducing a new line or trying a new variety)

The selection criteria. The breeder must cull ruthlessly at each stage. A single Kohaku spawning can produce 50,000 to 100,000 eggs. Of those, perhaps 30,000 hatch. Of those, perhaps 20,000 reach the culling stage at one year. Of those, perhaps 200 to 500 are worth growing to two years. Of those, perhaps 20 to 50 are worth growing to three years. Of those, perhaps 2 to 5 are worth selling as show-quality fish. The culling rate is 99.99%. The breeder who does not cull at each stage ends up with a muddy, mixed population that is mostly kawarimono.

The four mistakes

Mistake 1: Breeding fish of unknown variety or parentage. This is the most common beginner mistake. The breeder buys fish at auction, breeds them, and is surprised when the offspring are not the variety expected. The lesson: know the parentage. If you don't know the parentage, you cannot predict the offspring.

Mistake 2: Selecting only on colour and not on body conformation. The colour is what makes the koi valuable, but the body conformation is what makes the koi a koi. A koi with perfect colour and a crooked spine is worth almost nothing. A koi with average colour and a perfect body is worth more. The breeder who selects only on colour will produce a population of fish with increasingly poor body conformation.

Mistake 3: Failing to cull at the nisai stage. The two-year-old fish looks different from the four-year-old fish. The nisai sumi in a Showa may be heavy and beautiful, then disappear by year four, leaving a fish with insufficient sumi. The breeder who culls at one year based on the nisai appearance is culling the wrong fish. The breeder who grows to three or four years and culls at the adult appearance is culling correctly but spending a lot on space and food. The economic balance depends on the operation.

Mistake 4: Crossing too many varieties and producing kawarimono. Kawarimono is the catch-all "miscellaneous" category. Fish that are not clearly Kohaku, Sanke, Showa, or any other standard variety are kawarimono, and they are typically worth much less than the standard varieties. The breeder who crosses Kohaku with Showa, then crosses the offspring with Sanke, then crosses the next generation with Ogon, is producing a mixed population of fish that are mostly kawarimono. The lesson: stick to the standard varieties, maintain pure lines, and use experimental crosses sparingly.

Closing the loop

Koi colour genetics is a sophisticated field with a long history. The major colour genes (W, R, B, Y) follow relatively simple inheritance patterns, but the modifiers that determine quality are many and complex. The breeder who understands the major genes, who plans matings based on the desired variety, who maintains pure lines, and who culls ruthlessly at each stage, will produce a population of fish that approaches the standard varieties. The breeder who ignores the genetics, who crosses randomly, and who does not cull, will produce mostly kawarimono. The colour is the same in both cases. The difference is the planning.

Liu Wei

Liu Wei

🐠 Ornamental fish genetics & breeding professor

Liu Wei is a doctoral supervisor in ornamental fish genetics and breeding at the Ocean University of China, with research spanning koi color inheritance, fancy goldfish line history, and marker-assisted selection in ornamental species.

→ View all 27 articles by Liu Wei