Genetics

Betta Genetics: Color, Fin, and Pattern Inheritance

Four pigment layers, a chromosome-level genome assembly, and the transposable element that produces the marble gene. The genetics behind every betta strain.

Published Reading time 4 min
A male double-tail betta with a fully split caudal fin, showing the DT genotype.
A double-tail male. The split caudal is a single-gene trait (DT) with characteristic spinal-stress side effects in homozygotes. One of the cleanest Mendelian phenotypes in the hobby. Photo: Pharaoh Hound via Wikimedia Commons, CC BY-SA 3.0 / GFDL 1.2+.

Betta coloration comes from four pigment layers stacked in the skin: black (melanophores), red (erythrophores), yellow (xanthophores), and iridescent blue-green (iridophores, which produce structural color from crystalline platelets, not a pigment). Visible color is the net result of which layers are expressed, at what density, and in what pattern. The 2020 chromosome-level genome assembly (PubMed 32385046) and the 2022 phenotypic-diversity paper (PubMed 36129976) mapped the principal genes. Marble pattern is now known to derive from a transposable element in the Kit Ligand A gene. This page is the map.

The four layers

Melanophores (black/brown). Express the black pigment eumelanin. Controlled at the b (blond) locus and the bl (melano) locus. Melano homozygous (bl/bl) produces maximal black but also infertile females.

Erythrophores (red). Express red pigment. Controlled by the non-red locus (nr): NR/NR or NR/nr gives red, nr/nr gives cambodian (red-less). Extended red alleles produce deeper spread.

Xanthophores (yellow). Usually underlies reddish colors; in the absence of erythrophores produces yellow fish.

Iridophores (iridescent blue, green, steel). Structural color from layered purine crystals. Blue (ST/ST), steel (ST/st), green (st/st) at the steel locus. Opaque gene (Op) produces the metallic intensified look (dragon scale, etc.).

Major gene loci, summary

LocusAllelesEffect
SteelST, stIridophore color (blue/steel/green)
Non-redNR, nrRed presence (red/cambodian)
MelanoBL, blFull black when bl/bl (infertile females)
BlondB, bReduced melanophore density
OpaqueOpIntensified iridophore (dragon)
MarbleMbUnstable pattern (transposable element)
CrowntailCTReduced fin webbing
Double tailDTSplit caudal (single allele: broader dorsal)

Spokes

  • black melano, the infertile-female problem, why it happens, how to breed around it.
  • iridescence, blue, steel, green, dragon, opaque; the structural color system.
  • dragon scale, intensified iridophore, selective breeding approach.
  • marble gene, the transposable element in Kit Ligand A and why marble pattern shifts during the fish’s lifetime.
  • color morphs, catalog of common color morphs with genotype notes.
A male crowntail betta with characteristic spike-like fin ray extensions and receding webbing.
A crowntail male. The reduced-webbing phenotype is recessive at the CT locus. Crowntail × crowntail breeds true; crowntail × halfmoon produces no crowntail offspring. Photo: Anandarajkumar via Wikimedia Commons, GFDL 1.2+.

The 2020 genome

The 2020 chromosome-level genome assembly resolved 21 linkage groups and catalogued 27,000+ protein-coding genes. Critical findings:

  • Sex chromosomes are not well-differentiated; sex determination is polygenic or environmental influence matters.
  • Color-related loci cluster in a few linkage groups.
  • The Kit Ligand A gene on one chromosome carries a transposable element insertion polymorphism that generates the marble phenotype.

Before 2020, betta genetics was largely hobbyist linkage mapping from breeding crosses. The genome assembly made formal mapping possible.

The 2022 phenotypic architecture paper

Genome-wide association mapped many visible traits to specific loci. Notable:

  • Body color variation maps to a smaller number of major loci than hobbyists had assumed. A handful of genes explain most of the visible variation.
  • Tail shape variation is highly heritable but polygenic; no single “halfmoon gene.”
  • Aggression variation is weakly heritable (much more environmental than genetic).
  • Fighting-line bettas show signatures of strong artificial selection over centuries.

Practical implications for breeders

Predict what you can. Single-locus traits (crowntail, double tail, melano) have predictable Mendelian ratios. Plan crosses accordingly.

Track what you can’t. Multi-locus traits (halfmoon tail geometry, dragon intensity, specific color hues) require long-term selection across generations. One spawn doesn’t tell you genotype; 3-5 spawns from the same pair do.

Keep records. Note the parents, the fry outcomes, the best fish retained. Over generations the pattern emerges.

Stabilize lines. A “halfmoon yellow dragon” line takes 5 to 10 generations of selection to fix. Most hobbyist breeders never reach the stable point; they buy stable lines from established breeders and work within them.

The genetic weight of selective breeding

The genetic-architecture paper documented inbreeding signatures in modern show strains. Narrow founding populations plus decades of selection produce:

  • Reduced heterozygosity (genetic diversity loss).
  • Accumulated recessive load (higher deformity rates).
  • Shorter lifespans in heavily selected strains (compared to wild-type).
  • Compromised immune function in some lines.

This is a real cost. Hobbyists who value animal welfare should prefer simpler strains (veiltail, plakat) over extreme elaborations (rosetail, king crowntail).

What color is my fish?

Basic color identification:

  • Solid red: body and fins red, full coverage. NR present, extended red alleles.
  • Royal blue: deep blue body, iridescent. ST/ST or ST/st, melanophores present.
  • Steel blue: slate gray-blue. ST/st heterozygous.
  • Green: green iridescent body. st/st.
  • Cambodian: pale body, red or blue fins. nr/nr (no body red), or reduced melanin.
  • Black melano: solid black, female infertile. bl/bl.
  • Dragon: metallic opaque iridescence over color. Op present, multi-generation selection.
  • Marble: unstable pattern of color and white. Mb present (transposable element active).
  • Cellophane: transparent fins, flesh-colored body. Absence of most pigments.

For deeper identification, see color morphs.

Genetics is a rabbit hole. The four-layer frame is enough for most hobbyists; the per-spoke detail is for breeders planning specific lines. Start with black melano if you want the most-discussed genetic case first, or marble gene for the most recent peer-reviewed finding in the field.

Frequently asked

Why is my red betta producing blue fry?
Your red fish carries a non-red allele, probably at the red-spread locus or iridophore locus. Color expression in one parent doesn't prove genotype. Track fry outcomes to map what alleles the parent actually carries.
What's the difference between melano and black lace?
Melano (bl/bl) is true black with infertile females. Black lace is a similar-looking pattern from different alleles without the infertility. Don't confuse them for breeding projects.
Is dragon scale a single gene?
No. Dragon scale is intensified iridophore expression, likely polygenic. Stabilizing dragon requires multiple generations of selection.
Can I predict fry color from parents?
Partially. Basic layers are predictable. Specific patterns require pedigree knowledge. A named-line pair produces more predictable fry than a random pair.