Japanese Ornamental Koi Carp: Origin, Variation and Genetics

Full text

1
Japanese Ornamental Koi Carp: Origin, Variation and Genetics
Servaas de Kock
KoiNet
PO Box 1643
Gans Bay, 7220, Cape, RSA
Phone: +27 82 440 6770
Fax: +27 86 518 8568
Email: sdk@koinet.net
Boris Gomelsky
Aquaculture Research Center
Kentucky State University
103 Athletic Drive
Frankfort, KY 40601, USA
Phone: 1 502 597 8114
Fax: 1 502 597 8118
Email: boris.gomelsky@kysu.edu
Introduction
The ornamental form of the carp, Cyprinus carpio L., provides a rich source of investigation for science
and commercial possibilities alike. For researchers, however, there is also the challenge of cultural and
language barriers that add a dimension of the mystic. On the other hand, many koi hobbyists and

2
professionals are not familiar with published scientific information on koi genetics. This chapter intends
to better equip all kinds of readers with an understanding of the origin, variation and genetics of koi.
This chapter was written collectively by a koi professional and writer, who has run a large koi
farm and authored several books on koi keeping (De Kock and Watt 2006), and a fish geneticist, who has
studied inheritance of different traits in koi for 20 years.
Emergence of koi
At least three strains of common carp were traditionally reared by farmers in the mountainous Yamakoshi
region known as Yamakoshi Nijuusongou (Twenty Villages) between Nagaoka and Ojiya Cities in
Niigata Prefecture north-west of Tokyo (Japan) during the first year of Tenmei era (1781-1788) (Koshida
1931). The fry were usually bought from merchants at Uono river and other rivers and brought inland to
be reared in wooden house ponds or small terraced paddy fields and reservoir ponds, to supplement their
diet when isolated during high snowfalls every year. The farmers eventually started to culture carp
themselves, sharing the paucity of resources amongst them. There are three native feral variants namely
asagi (multicolor), yamato (reddish) and magoi (blackish) of which the first two found favor in the
culture ponds of the Japanese because of good growth and size (Okada 1960, Okubo and Sato 1975,
Suzuki and Yamaguchi, 1980). The word magoi (true carp) being a general term referring to the wild carp
in the Japanese rivers. In the Nishikigoi fraternity, however, koi ancestry is traced to the Asagi, Tetsu
(iron) and Doro (rust) variants of the magoi. A 1913 catalog of fishes of Japan (Jordan et al. 1913) lists
under Cyprinus carpio the Sarasa as a form found in Shinano Prefecture, the river running through the
Yamakoshi region.
Theories abound why this area in particular is recognized as the birthplace of Nishikigoi bordering
on the mystic. The unusual environmental conditions are often cited, but an extended period of drought
and famine during 1781-1788 impacted on stock that reduced the gene pool, occurrence of genetic
mutations, idle time and the curiosity of man are probably the main reasons. According to local memory,
it was shortly after this that colored aberrations appeared. These were called kawarigoi (changed carp)

3
and included all types of single and patterned forms including higoi (red carp), shirogoi (white carp),
magoi (true carp or black carp), asagi (blue carp), ironashigoi (colorless carp) and moyogoi (patterned
carp) (Koshida 1931, Kuroki 1981, Kuroki 1987, Kataoka 1989, Hoshino and Fujita 2006). During the
pre-1915 period the development concentrated on the red and white pattern – first the simpler Sarasa and
later the more florid Sakura. As the color drawings used for identification during the1914 Taisho Expo in
Tokyo clearly show, these were not modern koi (Figure 1). Most of the effort went into stabilizing the
red-white pattern on the dorsal parts and these fish were known by their different pattern name and
collectively as irogoi (color carp), hanagoi (flower carp) and moyogoi (patterned carp). The Shusui
originated about the same time from crossing domestic cultured asagi with imported leather carp from
Europe (see below).
Figure 1. Catalog drawings of some fish at the 1914 Taisho Expo in Tokyo showing the state of
development. A – Kohaku; B – Gonzo Sanke; C – Kurokihan (Classic Ki Utsuri) (Reproduced with
permission from owner Yokio Isa).
Irogoi improved rapidly in the period after 1915 to about 1942 with the creation of the early
Kohaku out of the Sakura and tri-color variety Taisho Sanshoku (a white koi with red pattern and black

4
patches) and other varieties through crossing and selection. Largely still isolated and under harsh
environmental and political conditions, limited natural resources and no technological assistance, peasants
steadfastly kept up development with a new vigor created by the public exposure of the 1914 Expo. By
1929 the first Yamakoshi Farming Koi Show was held in Higashiyama. The Showa Sanshoku (a black koi
with white and red superimposed) tricolor was established by 1935. Thus the popular gosanke group that
includes the three major colour types Kohaku, Sanke and Showa, was formed (Figure 2).
Figure 2. Gosanke. The three koi varieties most in demand today. A – Kohaku, one year-old (tosai),
18cm TL; B – Sanke, 65 cm TL; D – Showa, 57 cm TL. The Sanke and Showa are two year-old
(nisai) immature females. Note the different red hues of these “unfinished” fish. Color not yet fully
developed and body conformation incomplete.
Development came to a virtual halt during the Second World War, but then resumed though still
mainly isolated and practiced as a hobby. Only with improved air travel (around 1950) and packing
material and technology (around 1963) did exports become viable which further stimulated production.
Improved marketing by organized koi shows (around 1960) further increased demand. As the
international marketing machine gained momentum from 1970 to 1990, koi became a global commodity.

5
Masayuki Amano claimed in 1971 that up to 87% out of 908 families in the Yamakoshi region were
producers of ‘fancy carp’. Development was stepped up and resulted in as many as 120 different popular
varieties known today. Production increased in several other centers including Hiroshima, Okayama,
Karume, Shizuoka and Yamanashi where the land was more flat and the conditions more favorable and
warmer. Breeders like Manabu Ogata, Kentaro Sakai and Michio Maeda added a different approach to
production and marketing by producing both quantity and quality to an ever-discerning world market. A
new name, Nishikigoi (錦鯉, brocaded carp) emerged when efforts were made to promote the trade in the
1970’s. This name was originally expressed in 1918 by Mr. Kei Abe, but never caught on. He was head of
the Niigata Prefecture’s Fisheries Department and according to Koshida (1931), was earlier responsible
for introducing Mendel’s law and the principles of inheritance of genetic traits to the farmers of
Yamakoshi during lecture tours. Dr. Takeo Kuroki was instrumental in the formation of the Zen Nippon
Airinkai (ZNA) in May 1968 to promote the hobby. In 1970 the breeders and trade followed with the Zen
Nihon Nishikigoi Shinkokai (All Japan Nishikigoi Promotion Association) that in 2003 became the
International Nishikigoi Promotion Center (INPC).
As the hobby grew internationally, so farms sprung up all over the world driven by the need to
meet demand. The respected Peter Waddington (Waddington 2009) estimated that the period of the
‘serious koi hobbyist’ in Japan was from 1965 to 1990 and peaked in 1978. He wrote: “After 1990, the
domestic market for koi continued to decline annually to become as it is today in 2008 when only the very
low grade koi produced by the breeders are sold to pet shop suppliers in volume.” and “the Japanese
Nishikigoi market in and around 1969 formed almost 100% of production in the world whilst in 2007 it
only amounted to under 20% and that includes koi of ALL prices and qualities." That may be true with
respect to the traditional markets in the West. He need not have worried. New markets in south-east Asia
are creating a seemingly insatiable market for the higher quality product and fuel further development.
The idiosyncrasies of different cultures also drive a continued range of varieties. The Japanese will

6
remain ahead in the developmental and technological departments of this industry for a long time to
come.
Genealogy of Nishikigoi
Various attempts were made to piece together the genealogy of koi based on breeders’ notes and other
accounts (Koshida 1931, Kuroki 1987, Watt and De Kock 1996, Hoshino and Fujita 2006). There are also
numerous accounts from anonymous writers in the Japanese Nichirin and Rinko magazines, but they have
no supporting evidence. The resultant tree, which is shown in Figure 3, remains shrouded in myth, and
questions need to be asked if the endeavor is worth it, but it remains of interest to the hobby and keen
hobbyists. For breeders it may be a tool to plan their breeding, whether of value or not. For inexperienced
breeders and hobbyists, it may help to put the different varieties in perspective.
Figure 3. Gene tree of Nishikigoi. (Adapted and redrawn after Koshida 1931, Kuroki 1987, Watt
and De Kock 1996, Hoshino and Fujita 2006).

7
The main events leading to the development of Nishikigoi today are summarized in Table 1 from
data in an anonymous document in the files of the Japan Nishikigoi Promotion Association, together with
extensive studies by Hoshino and Fujita (2006) and some other authors.
Table 1 – Summary of the key historical events* in the emergence of koi in Japan.
Variety
Date
Place
Originator
Remarks
1781+
severe drought and famine
Irogoi
1804+
Nijuu Yamakoshi
emergence from probably Asagi
Asagi Magoi
1868
Ueda river area
Jirohachi Seki
improving of Asagi line as
ornamental
Gosuke Sarasa
1890
Ojiya
Kunizou Hiroi
forerunner of Kohaku (Sakura)
Gonzo Sanke
1898
emergence from Asagi magoi
Ki Utsuri
1898
Takezawa
Komasaburo
Hoshino
emerged from Tetsu magoi
Doitsugoi
1904
Fukagawa
Shinnosuke
Matsubara
introduction of German leather
carp
Shusui
1908
Akiyama Fish
Farm
Kichigoro
Akiyama
German leather carp x Asagi
Akame Kigoi
1914
Akiyama Fish
Farm
red eye Kigoi emerged
Sanshoku
precursor
1914
Uragara
Heitaro Sato
Shirogoi x Kohaku = tejima-
Kohaku
Sanshoku
1917
Yamakoshi
Eisaburo Hoshino
tejima-Kohaku x Bekko
Ki Utsuri
1920
Yamakoshi
Sato Yohei
Ki Bekko x Magoi
Ki Utsuri
1920
Yamakoshi
Eisaburo Hoshino
Ki Utsuri x Ki Utsuri (fixed)
Kigoi
1922
Yamakoshi
Yoshiichi Matsui
improving of Akame Kigoi (red
eye)
Kin Kabuto
1924
Yamakoshi
Isematsu Takano
Kinbo x Fijigin
Shiro Utsuri
1925
Yamakoshi
Kazou Minemura
Ki Utsuri Sanke x Shiro Bekko
Jintoro Showa
1927
Yamakoshi
Jukichi Hoshino
Ki Utsuri x Matsukawabake x
Kohaku
Ginrin Kohaku
1929
Yamakoshi
first exhibition of ginrin
Ossa Showa
1935
Ojiya
Kishichirou
Hoshino
Showa x Shiro Utsuri (fixed)
Doitsu Sanke
1941
Yamakoshi
Tahichi Kawakami
Ogon
1947
Ojiya
Sawata Aoki
Kinkabuto x Ginfushi
Doistu Yotsushiro
1950
Kunio Hiroi
Karasu x Hajiro appear
spontaneous
Kumonryu
1953
Yamakoshi
Kunio Hiroi
Doistu Yotsushiro x Karasu
lineage
Oranji Ogon
1956
Yamakoshi
Masamoto
Kataoka
Asagi x Ogon (1963 fixed)
Yamabuki Ogon
1957
Yamakosh
Masamoto
Kataoka
Doitsu Ogon
1958
Yamakoshi
Tomisaku Sakai
Black Doitsu x Ogon
Kin Utsuri
1958
Yamakoshi
Kumazo Takahashi
Ogon x Ki Utsuri
Doistu Showa
1959
Yamakoshi
Kanekichi Fujii

8
Hariwake Ogon
1960
Yamakoshi
Tomisaku Sakai
Ogon (4x individuals)
Midorigoi
1960
Kiichiro Suzuki
Matsuba Ogon
1960
Ojiya
Eizaburo Mano
Kujaku
1961
Ojiya
Toshio Hirasawa
Matsuba Hariwake x Shusui
Parachina Ogon
1963
Uodu, Toyama.
Tadao Yoshioka
Akame Kigoi x Nezu Ogon
Yamatonishiki
1965
Seikichi Hoshino
Koshi no Hisoku
1965
Uodu, Toyama
Tadao Yoshioha
Shusui x Yamabuki Ogon
Beni Hikari
1990
Ojiya
Doitsu Goshiki x Doitsu Kujaku
Kinkiryu
1990
Yamakoshi
Ichiji Watanabe
Doitsu kinryu x Doitsu kinshowa
Kinryu = Kumonryu x Gin
Matsuba
Kikokuryu
1992
Ojiya
Haruo Aoki
Kumonryu x Kikusui
Beni Kikokuryu
1992
Ojiya
Haruo Aoki
Kumonryu x Kikusui
Ginka
1998
Ojiya
Kiyoshi Kase
Ginrin Hajiro x Goshiki x
Motoguro Kujaku
*Adapted from data from various anonymous sources, the Japan Nishikigoi Promotion Association, Rinko
and Nichirin magazines, and Nishikigoi Mondo by Hoshino and Fujita (2006).
Evaluation criteria
Because koi is an ornamental fish, its value is affected by perceived qualities and breeders aim to improve
those qualities through parentage and selection criteria. While color and color pattern play a part in
determining the value and marketability of mass produced, commercial fish, those are not prime factors in
determining the value of individual, high quality fish from well-known producers. These breeders created
lines that have shown consistent, predictable and sought after qualities. Great value is attached to such a
lineage (De Kock and Watt 2006). The essential elements determining the value of fish that breeders try
to improve genetically are listed below.
Body conformation. From a human perspective, the koi should have a symmetric, tub-like body
with dynamic, broad shoulders. The body gives the producer insight into the potential for growth over the
adult life of the koi and, therefore, is one of the most important value determinants. Late maturation and
lower fecundity ensure a longer ‘show life’. Males with a plumper body form can compete with the
females in the show arena for a longer period, but reproduce better in an artificial, one-on-one
environment without the competition of slimmer suitors.
Swimming style. A quality of ‘graceful’ swimming is sought. For older fish words like ‘dignity’
or ‘character’ may also be added. Swimming style is the aesthetic value of the biokinetic expression of

9
movement, directly related to skeletal and muscular qualities of the fish. ‘Gracefullness’ in the swimming
style refers not only to powerful but fluid movement but also to the calming effect it has on the viewer.
Feeding behavior is the primary cause of this movement, and possibly tameness brought on through
domestication and repeated handling may also help in creating this illusion. To swim ‘lively’ fish must be
healthy and in an environment conducive for feeding i.e. low in ammonia concentration and high in
dissolved oxygen. Failing these conditions the koi will exhibit a ‘lethargic’ or ‘listless’ swimming style.
On the other end of the scale ‘frantic’ swimming is exhibited when the fish are frightened or tries to
escape near-toxic, unfavorable water conditions.
Color quality. The brightness is determined by the number of chromatophores in the skin and
their propensity for collecting pigments, normally during their development but also later, from the
environment. The purity of color to separate into aesthetically pleasing colors and patterns as viewed
through different skin layers. Males tend to be brighter at an early age increasing their initial show value.
Color durability. The ability to maintain or increase the pigment ‘loading’ of the chromatophores
through synthesis and the ability of pigment cells to migrate through different skin layers to ‘develop’ a
pleasing pattern can add to the value of an individual. Intracellular reaction of chromatophores - in
particular melanophores - and likely cyanophores to environmental stressors can reduce the value. This
could be under both genetic and environmental control.
Color distribution. An aesthetically pleasing pattern to match the particular variety is normally
determined by selection. Patterns are not repeatable, but pattern types, pattern edging and shapes do run in
families and can be recognized and thus predicted to an extent. Therefore specific bloodlines are more
sought-after than others.
These attributes encompass the guidelines set out to nationally accredited judges when judging at
koi shows worldwide. As described in Hoshino and Fujita (2006) and De Kock and Watt (2006) judges
using a 100-point system award up to 50% to ‘body shape’ evaluation and 20% to colour evaluation.
Pattern, ’gracefulness’ and ‘dignity’ carry a 10% contribution each to the overall evaluation. The
importance of an aesthetic pleasing body and movement is left in no doubt.

10
Intentional hybridization and line breeding lead to improvement and proliferation of varieties
while stringent selection and culling regime led to increase in quality and marketability of the variety.
This labor-intensive task calls for skilled people of which few are outside of Japan. In Taiwan and
perhaps a handful of dedicated producers worldwide, follow the tradition of improving their production
and skills. For the rest the resultant breeding is without purpose other than for a multitude of small,
colorful fish and thus the world perception of “quality” is compromised (De Kock and Watt 2006).
Nishikigoi do not breed true and always require severe culling during the first culling to remain
true due to the variety’s constrains. The single colored types like Ogon, Purachina and Chagoi may be
exceptions in terms of color, but they do also require rigorous selection for sought-after qualities. For
example, at least 60-65% of a Kohaku paring is removed at about 40 days of age. Likewise, 75-80% of a
Sanke cross is removed at first culling at 25-30 days of age. With Showa only the black pigmented fry
being 5-50% survive the first culling by the 3-5th day after hatching. At least two more culling operations
are carried out before the age of six months illustrating how much value is placed on the individual gene
make-up.
The making of color
While the morphological variation was important in the creation of koi as they are today, the expression
of color is certainly central to defining koi as ornamental. Skin pigment cells produce the various colors.
Erythrophores and xanthophores contain carotenoid pigments and produce various shades of red or
yellow. Melanophores are black and contain melanin. In addition, further hues like green and light blue
are formed due to the superposition of pigment cells in the skin.
Leucophores contain granules of guanine resulting in a white appearance. Iridophores contain
iridescent plate-like crystals of guanine resulting in a shining, metallic appearance of the skin also known
as hikari. In combination with yellow, the effect can be golden (e.g. Ogon) and with white it could be
silver like Purachina. Platelets of guanine form in the scale structure a radiant, reflecting surface known

11
as ginrin. On a red or yellow background there is a golden reflection and when overlaying black, blue or
white, silver is reflected.
Cyanophores contain a blue pigment of unknown chemical composition. The melanophores and
cyanophores pigment is stored in fibrous organelles that can open or contract under control of the central
nervous system. Chromatosomes are derived from the neural crest and migrate to their positions during
ontology of the embryo. The melanophores and cyanophores follow the internal route while the other
future pigment cells migrate alone to the outer layers of the dermis. Probably this is the reason that red,
yellow and white appeared to be overlaying black and blue initially.
Aspects of chromatophores that are exploited during breeding and selection are: (1) pigment cell
formation and migration; (2) pigment synthesis; (3) pigment cell translocation; (4) pigment cell
interaction between the two main groups of chromatophores (xanthophore-like and melanophore-like
cells) and (5) time dependency of all these over the life of the individual.
Inheritance of color traits
This part of the chapter contains a review of known scientific information on the inheritance of color
traits in koi.
The first studies on color inheritance in koi, which became known to English-speaking scientific
community, have been performed in Russia (former Soviet Union). In the middle 1960s the Japanese
government gave a group of koi to Soviet government as a gift. These fish were delivered to the Research
Institute of Freshwater Fish Culture, which is located not far from Moscow, where Dr. V. Katasonov has
performed series of studies on color inheritance and described several color-determining genes in koi.
Katasonov (1978) has shown that melanin formation in wild-type color carp is controlled by dominant
alleles of two duplicate (i.e. having similar action) genes designated as B1/b1 and B2/b2. The presence of
one dominant allele at any gene in fish genotype results in appearance of wild-type color. When wild-type
color common carp, homozygous for dominant alleles at both genes (genotype B1B1B2B2), was crossed
with koi (genotype b1b1b2b2) the F1 progeny consisted of wild-type color fish only (genotype B1b1B2b2).

12
When F1 fish were crossed, the phenotypic ratio 15:1 was observed in F2 progeny; this ratio is typical for
duplicated gene action. The scheme and results of these crosses are given in Fig. 4.
Figure 4. Inheritance of wild-type color in common carp (duplicate genes action).
The presence of melanin in wild-type color fish is visible already at late embryonic and larval
stages while koi embryos and larvae are yellowish and transparent because of lack of black pigment. The

13
control of wild-type color in common carp by two duplicated genes has been confirmed later in several
further studies. For example, Cherfas et al. (1992) crossing wild-type color female, heterozygous for two
genes (genotype B1b1B2b2), with koi male (genotype b1b1b2b2) obtained a 3:1 phenotypic ratio in the
resulting progeny. This ratio is typical for a test cross in case of duplicated genes.
Later Katasonov et al. (2001) have suggested the existence of third alleles at these two duplicated
genes (designated as B'1 and B'2, respectively), which control the presence of melanin in lower
(underlying) skin layer; the dominance relationships between the three alleles was proposed as B > B' > b.
Katasonov (1974, 1978) also identified that blue body color in koi is controlled by the recessive allele of
gene R/r. Blue koi with genotype rr do not have red and yellow pigments. Combination of recessive
alleles of the three genes - r, b1 and b2 (genotype rrb1b1b2b2) gives white color of fish. White fish are
characterized by the lack of both black and red (yellow) pigments. Katasonov (1973, 1974) also revealed
that the inheritance of the trait “design” which is typical for metallic koi. This trait manifests as a yellow
stripe along the dorsal fin combined with a specific ornament on the head and is controlled by the
dominant allele of gene D/d. Metallic koi have genotypes DD or Dd while non-metallic koi (genotype dd)
are homozygous for recessive allele. When metallic koi are crossed with wild-type color common carp so
called “ghost koi” (or “ghost carp”) appear. Ghost koi (Fig. 5) have traits which are controlled by
dominant alleles of different genes - clearly visible “design” from metallic koi and black pigment melanin
from wild-type color carp. Although ghost koi is not accepted by koi show standards, this color type is
pretty popular in some countries.

14
Figure 5. Ghost koi – the hybrid between metallic koi and wild-type color carp.
Katasonov (1973, 1974, 1976) has also identified trait “light color” in koi. The light color of fish,
which possess this trait, is caused by permanent contraction of melanophores. Fish having this trait are
heterozygous for gene L/l (genotype Ll). Homozygotes for dominant alleles (genotype LL) are inviable
and perish at larval or fry stage.
Among multicolor traits in koi the inheritance of the white-red (Kohaku) color complex and
Bekko type of black patches, which is observed, for example, in Taisho-Sanke or Shiro-Bekko, have been
investigated most frequently. Some information was obtained on the inheritance of the Utsuri type of
black patches, which is observed, for example, in Showa-Sanshoku or Shiro-Utsuri. Iwahashi and Tomita
(1980) in a study performed at Niigata Prefectural Inland Water Fisheries Experimental Station (Japan)
have investigated inheritance of the Kohaku color complex. White-red (Kohaku) fish, as well as solid

15
white (Shiromuji) and solid red (Akamuji) fish originating from Kohaku, were reproduced in three
generations and the ratios of color phenotypes in progenies have been investigated. In progenies obtained
by crossing of white-red (Kohaku) fish all three color types (solid red, white-red and solid white) were
observed. The most numerous class was white-red fish (68-76%); percentages of solid red and solid white
fish varied from 18 to 21% and from 6 to 11%, respectively. Percentages of Kohaku have not been
increased in generations as a result of selection. In progenies obtained by crosses of solid-white fish,
proportions of solid red fish were low (0-6%); while the percentage of solid white fish increased from
30% in first generation to 62 and 80% in second and third generations, respectively; correspondingly
percentage of white-red fish decreased in the last two generations. Crossing of solid red fish gave only
solid red fish in progenies in all three generations. Based on their obtained results, Iwahashi and Tomita
(1980) concluded that solid red fish originated from Kohaku are pure breed but solid white fish originated
of Kohaku are hybrids; Kohaku are also hybrids of solid red and solid white fish. Also, it was suggested
that white-red color complex is controlled by numerous genes of white and red colors.
Gomelsky et al. (1996) have investigated color variability in normal (amphimictic) and meiotic
gynogenetic progenies produced from seven females of several multicolor traits (white-red - Kohaku,
white-black - Shiro-Bekko, and white-red-black - Taisho-Sanke and Tancho-Sanke) in a study which was
performed in Israel. Normal progenies obtained from Kohaku parents consisted of three color types –
solid white, white-red (Kohaku) and solid red; the white-red class was most numerous with proportions
varying from 45.7% to 57.2% (see below for data on color variability in gynogenetic progenies). Six color
types were observed in progenies obtained from white-red-black (Taisho-Sanke and Tancho-Sanke) and
white-black (Shiro-Bekko) parents: solid white, white-red, solid red, white-black, white-red-black and
red-black. The white-red-black tricolor type was most numerous in all progenies, including those from
white-black bicolor parents. In all normal and gynogenetic progenies obtained from white-red-black and
white-black females, the ratios of white : white-red : red and white-black : white-red-black : red-black did
not differ significantly. On this basis it was concluded that the occurrence or absence of black patches on

16
the body and the white-red color complex are inherited independently and these traits were discussed and
further studied separately (Gomelsky et al. 1996).
Based on large variability in ratios of white : white-red : red fish among progenies and in rate of
development of red coverage in white-red fish (from a single tiny spot to nearly complete coverage) it
was suggested that the white-red color complex is controlled by many genes (Gomelsky et al. 1996). It
was also suggested that low development of red color in white-red parents resulted in low development of
red color in the progenies (and vice versa). This suggestion was confirmed in a further study on
inheritance of white-red (Kohaku) complex in koi performed in Israel (Gomelsky et al. 2003). Color
variability was investigated in three progenies obtained by crossing of Kohaku parents with different rates
of red-area body coverage. It was shown that the lowest percentage of red fish and highest proportion of
red fish were recorded when white-red parents with weak development of red color (1-3% of body
coverage) were used. In this study the percentage of coverage by red patches was also measured in a
sample of white-red fish from each progeny (Gomelsky et al. 2003). For measurement of red-area
coverage, outlines of the entire body and red patches for both sides of the fish were traced on the plastic
film with a permanent marker. The relative red-area coverage was determined as ratio of red patches
outlines area to the area of total body outlines. The distribution of white-red fish in sample with regard to
red-area coverage, together with the white : white-red : red ratio in progeny was used to estimate the color
class distribution in the entire progeny (including the solid white and solid red individuals). In two from
three investigated progenies obtained by Kohaku x Kohaku crosses, the frequency of solid red fish was
much higher than the frequency of adjacent classes of white-red fish. Thus, in these progenies fish were
clearly divided into two groups: fish with white background color (solid white and white-red) and solid
red fish. As an example, distribution of fish in one progeny from this study, which was obtained by
crossing of Kohaku female and Kohaku male with 42.0 and 30.4% of body coverage with red patches,
respectively, is presented in Figure 6. From 215 fish analyzed in this progeny, 11 fish (5.1%) were solid
white, 143 fish (66.5%) were white-red and 61 fish (28.4%) were solid red; the coverage of red patches
was measured in 60 white-red fish from the progeny. The division of fish into two groups - with white

17
background color (solid white and white-red) and solid red - is clearly visible in the class distribution
presented in Figure 3. It was suggested that the appearance of these groups might be explained by the
existence of some major color-determining gene(s), which determine(s) the background color (either
white or red) of the individual fish (Gomelsky et al. 2003). The development of red patches in fish with a
white body background is polygenic and controlled by many genes with alleles that either maintain the
white color or induce the appearance of red patches.
Figure 6. Distribution of fish in progeny obtained by Kohaku x Kohaku cross with regard to
development of red color (after Gomelsky et al. 2003, with modifications).
Later Novelo and Gomelsky (2009) used image analysis software for determination of red-area
coverage in white-red koi) in a study performed in the United States. The color variability distribution

18
was described for one progeny obtained by crossing Kohaku x Kohaku; which was similar to the one in
the previous study (Gomelsky et al. 2003), where fish were clearly divided into two groups: fish with
white body color background and solid red fish. These studies (Gomelsky et al. 2003, Novelo and
Gomelsky 2009) have shown that measuring the relative red body coverage provides an opportunity to
investigate the development of red color in progenies obtained from Kohaku parents as a trait with
continuous variability and provides further information for the better understanding of inheritance of this
trait.
As mentioned above, the black pigmentation in wild-type color common carp appears during
embryonic development and hatched larvae already have well-developed melanophores. Gomelsky et al.
(1996) have described a different mode of development for Bekko type black pigmentation in koi. All
hatched larvae obtained in crosses of parents with black pigmentation of Bekko type (Taisho-Sanke,
Shiro-Bekko, Tancho-Sanke) expressed no black pigmentation, and developed it later, about one week
after transition of larvae to active feeding. Only fish having black pigment at the fry stage develop black
patches later. In two progenies obtained by crosses of white-red-black (Taisho-Sanke x Taisho-Sanke) or
white-black (Shiro-Bekko x Shiro-Bekko) parents the ratios of pigmented and non-pigmented fry did not
differ significantly from Mendelian 3:1 ratio. On this basis it was suggested that presence of Bekko black
patches is controlled by dominant allele of one gene (Bl/bl). Also it was shown (Gomelsky et al. 1996)
that the ratio of pigmented : unpigmented fry more correctly reflected original segregation since during
further rearing in ponds the proportion of fish with black patches sometimes increased apparently due to
their better survival compared with fish without black patches. Gomelsky et al. (1998) have reported the
results of a further study (performed in Israel) on inheritance of the Bekko type of black pigmentation in
koi. Parents with Bekko pigmentation (Taisho-Sanke and Shiro-Bekko) and without black pigmentation
(Kohaku) were used in crosses; a total of 22 progenies were obtained and analyzed. In this study the
segregation of fish with and without black pigmentation was usually recorded at the fry stage, at the end
of nursing period. In five of six progenies obtained by Bekko x Bekko crosses typical Mendelian 3:1
segregation was observed. In 15 of 16 progenies obtained by Bekko x Kohaku crosses Mendelian

19
segregation 1:1 was observed; in one progeny obtained by Bekko x Kohaku cross only black pigmented
fry were observed. The data obtained in this study confirmed earlier made suggestion (Gomelsky et al.
1996) that the Bekko type of black pigmentation in koi is controlled by the dominant allele (Bl) of one
gene Bl/bl. Fish with Bekko pigmentation have genotype Blbl (most of fish used in crosses) or BlBl (one
identified fish). It is known that the rate of Bekko pigmentation in koi is highly variable – from a few
dispersed black spots to many black patches covering large areas of the body. Gomelsky et al. (1998)
noted that the major gene Bl/bl determines only the presence or absence of black patches while the rate of
their development is apparently under control of multiple genes.
David et al. (2004) reported results of studies (performed in Israel) on inheritance of red color
development (Kohaku color complex) as well as Bekko and Utsuri types of black pigmentation in koi.
About 40 crosses of white-red (Kohaku), solid white (the authors called this group “transparent”) and
solid red fish parents in different combinations (including 20 crosses Kohaku x Kohaku) were obtained
and analyzed. The thirteen crosses of randomly chosen Kohaku parents gave three phenotypes in
offspring: solid white, white-red and solid red; proportions of these three phenotypes were highly variable
among crosses; for example, the proportion of white-red offspring ranged from 10.6% to 67.6%. The
same three phenotypes were also found in crosses solid white x solid white, solid white x Kohaku, solid
red x Kohaku, and solid white x solid red while crosses solid red x solid red resulted in red progeny only.
The proportions of phenotypes were studied also in progeny of Kohaku parents with various levels of red
coverage and an effect of individual parents was detected. Based on obtained results David et al. (2004)
concluded that at least three genes, having intra- and inter-locus interaction, control development of red
color in koi. The authors proposed that Tp (for transparent) is a locus that dominantly controls the absence
of red color (David et al. 2004). When transparent parents are heterozygous at the Tp locus, they can have
colored offspring. It was also suggested that two other genes (Ra and Rb for red) control the extent of red
color, ranging from none to complete red coverage. The completely red parents are probably homozygous
to one or the two R genes, and thus have only red progeny. It was also suggested that some transparent

20
(solid white) fish that are determined by the Ra and Rb loci rather than by the Tp dominant allele should
exist.
David et al. (2004) showed that Utsuri and Bekko patterns of black pigmentation appear at
different development stages – around hatching and 14-day post-hatching, respectively; and this implies
that these patterns are controlled by different genes. Nine Bekko x Bekko crosses and 14 Bekko x no
black pigmentation crosses were performed and segregations in progenies (Bekko : no black
pigmentation) were recorded and analyzed. An average about 75% of the offspring had the Bekko pattern
in nine crosses between Bekko parents, fitting a Mendelian 3:1 ratio. When segregations in progenies
were separately analyzed, two of the nine crosses deviated from a 3:1 but fitted to a 2:1 ratio, four
deviated from a 2:1 but fitted to a 3:1 ratio and three fitted both ratios. From 14 crosses of Bekko x no
black pigmentation, segregations in 13 crosses fitted a Mendelian 1:1 ratio while one cross deviated from
this ratio. David et al. (2004) noted that the obtained results are in agreement with those of Gomelsky et
al. (1998) who suggested the Bl gene having a dominant black-pattern allele. The authors (David et al.
2004) also suggested that the inheritance of the Bekko pattern may be somewhat more complicated;
namely some selection against fish with BlBl genotype might be operating (probably during maturation).
This suggestion was based on several observations. First, none of 17 Bekko parents used in crosses had
BlBl genotype (all parents were heterozygotes Blbl). The author also referred to the study of Gomelsky et
al. (1998) where out of 17 fish parents only one was shown to be a BlBl homozygote. Second, the
variation in the proportion of Bekko offspring was higher in crosses between two Bekko parents than in
crosses in which only one parent was a Bekko. Third, some Bekko x Bekko crosses gave segregations
close to a 2:1 ratio (suggesting absence of homozygotes BlBl).
David et al. (2004) analyzed segregations in nine progenies obtained by Utsuri x Utsuri crosses.
Progenies were separated into dark and light pigmented larvae, which were grown separately to the
juvenile stage. The light pigmented larvae developed into transparent (solid white), Kohaku and solid red
while dark-pigmented larvae developed into several color phenotypes including Utsuri but also wild-type
carp color, grey, brown and combination of these colors. Among offspring of nine crosses between Utsuri

21
parents, the proportion of Utsuri juveniles ranged from 0% to 16%, with an average of 4%. Also David et
al. (2004) presented data on pigmentation of the hatched-out larvae in crosses where Utsuri parents were
used. The average proportion of dark-pigmented larvae in 30 crosses of Utsuri x Utsuri was 24.5%,
ranging from 1% to 54.7%. In 14 crosses where one parent was Utsuri and the other had no black
pigmentation the average proportion of pigmented larvae was 15.3%, ranging from 0% to 40.9%. Based
on the obtained data, David et al. (2004) suggested a complex genetic control of the Utsuri type of black
pigmentation with possible environmental effects.
Recently Gomelsky and Schneider obtained data on timing of black pigmentation appearance and
segregation of pigmented and non-pigmented larvae in progenies obtained by crossing of Utsuri parents in
a study performed in the United States (unpublished data). One Hi Utsuri (black-red) male was crossed
individually with two Hi Utsuri females; also one Hi Utsuri female was crossed with Kohaku male (no
black pigmentation). Larvae at hatching had already segregated to pigmented and non-pigmented groups
as was described by David et al. (2004). Appearance of larvae at 3d day after hatching from Hi Utsuri x
Hi Utsuri progeny is shown in Fig. 7. Observed segregations of pigmented and non-pigmented larvae
have also been in the range indicated by David et al. (2004) for corresponding crosses. Proportion of
pigmented larvae in two crosses Hi Utsuri x Hi Utsuri were 19.2% and 29.5%, while only 5.6% of
pigmented larvae were observed in Hi Utsuri x Kohaku progeny (Gomelsky and Schneider, unpublished
data).

22
Figure 7. Pigmented and unpigmented larvae from Utsuri x Utsuri progeny at 3d day after hatching.
The information presented above shows that the number of studies, which were devoted to
investigation of color inheritance in koi, is not numerous. The scientific information on genetics of color
traits in koi is scarce if to compare, for example, with goldfish (Smartt 2001) or some aquarium fish.
Apparently, the main reasons of paucity of genetic studies in koi result from two peculiarities of koi
culture and evaluation. First, koi culture in large extent is based on strict and continuous culling. Most of
koi color types are not true breeds, i.e. if fish of the same type are crossed; the progeny will be variable
and will contain different color types. Second, multicolor koi are characterized by individual variability;
the value of multicolor individual fish depends on color pattern, which is determined by size and location
of color patches on the body. It is obvious that variability of these traits cannot have strict genetic
determination.
The obtained data on color inheritance show that during development of koi the accumulation of
mutations of different genes, which impacted synthesis of the same pigments, has occurred. These

23
mutations have been of both lost-of-function (stopping production of corresponding pigment) and gain-
of-function (initiating production of pigment) types. It is very demonstrative in case of genetic control of
melanin production in koi. Synthesis of melanin typical for wild-type color common carp is suppressed
by mutations of corresponding genes. However, accumulated mutations of other genes resulted in
appearance of different types of black pigmentation (Bekko and Utsuri), which are typical for koi. The
same regulating mechanisms can be suggested, for example, for red pigment. Loss-of-function mutation
of gene controlling synthesis of red pigment could result in appearance of fish with white background
color. Further mutations of other genes causing appearance of red patches on white background color
could result in Kohaku color type. The combination of mutations of genes controlling synthesis of
different pigments could give known color variability of koi.
Variation and inheritance of scale cover types
There are four main types of scale cover in koi (and common carp). In scaled fish (wagoi) even rows of
scales cover the whole body. Mirror (or scattered, kagamigoi) fish have large ("mirror") scales, which
are scattered on the body and do not fine favor with the collector. The mirror fish demonstrate high
variability with regard to the reduction of scale cover. Some mirror fish have the body almost completely
covered with big scales. In linear fish, (also called kagamigoi) large scales form an even row along the
lateral line. In general, there are no scales on the bodies of leather fish (also called nude, kawagoi);
several scales may be found near the base of fins.
The type of scale cover in koi (and common carp) is determined by an interaction of two genes
having two alleles each (S/s and N/n) (Kirpichnikov 1981, 1999). Fish of different scale cover types have
the following genotypes: scaled - SSnn or Ssnn, mirror - ssnn, linear - SSNn or SsNn , and leather - ssNn.
Fish having the dominant allele N in the homozygote state (genotype NN) are unviable and perish at the
time of hatching. The expected phenotypic ratios in all possible crosses are presented in Table 2.

24
Table 2. Expected phenotypic ratios in crosses of koi with different scale cover types.
Parents and type of crossing
Phenotypes (%)
Scaled
SSnn and/or Ssnn
Mirror
ssnn
Linear
SSNn and/or SsNn
Leather
ssNn
1. Scaled x Scaled:
SSnn x SSnn, SSnn x Ssnn
Ssnn x Ssnn
100
75
0
25
0
0
0
0
2. Scaled x Mirror:
SSnn x ssnn
Ssnn x ssnn
100
50
0
50
0
0
0
0
3. Scaled x Linear:
SSnn x SSNn, SSnn x SsNn
Ssnn x SSNn
Ssnn x SsNn
50
37.5
0
12.5
50
37.5
0
12.5
4. Scaled x Leather:
SSnn x ssNn
Ssnn x ssNn
50
25
0
25
50
25
0
25
5. Mirror x Mirror:
ssnn x ssnn
0
100
0
0
6. Mirror x Linear:
ssnn x SSNn
ssnn x SsNn
50
25
0
25
50
25
0
25
7. Mirror x Leather:
ssnn x ssNn
0
50
0
50
8. Linear x Linear:*
SSNn x SSNn, SSNn x SsNn
SsNn x SsNn
33.3
25
0
8.3
66.7
50
0
16.7
9. Linear x leather:*
SSNn x ssNn
SsNn x ssNn
33.3
16.7
0
16.7
66.7
33.3
0
33.3
10. Leather x Leather:*
ssNn x ssNn
0
33.3
0
66.7
* In these crosses 25% of offspring (NN) die; the ratios among viable fish are shown.

25
In koi, variants with reduced scale cover are known since 1908 when Kichigoro Akiyama crossed
German leather (nude) carp of the Aishgrunder strain with the endemic Asagi strain. As mentioned above,
leather carps have ssNn genotype, that is they are mutants for both (S/s and N/n) genes. Therefore, it is not
surprising that soon the full range of scale cover types appeared and Japanese names for the various
phenotype subclasses like Yoroi-goi (armour plated scales), Aragoi (jumbled scales), Ishigaki-rin (stone
wall scales) etc. came in use. Variants with reduced scale cover can be found in all color koi types. This
shows that scale cover types and color traits are inherited independently. Variants with reduced scale
cover in koi are called doitsu (which is the literal translation of “Germany”) that reflect the history of their
appearance in Japan (Figure 8).
Figure 8. Some doitsu koi indicating scale phenotypic feature mostly used to identify the genotype.
A. Kikusui, leather (nude, kawagoi); B. Doitsu Sanke, linear (kagamigoi); C. Shusui, mirror
scattered, kagamigoi).

26
It is known (Kirpichnikov 1981, 1999) that the genes for scale cover have wide pleiotropic effect,
i.e. they influence many traits. The effect of allele N is especially strong. Linear and leather carps
(genotype Nn) have retarded growth rate and decreased survival as compared with scaled and mirror carps
(genotype nn); also, allele N influences many physiological, morphological and meristic traits in fish. The
pleiotropic effect of allele s is much weaker although some differences between scaled (SS or Ss) and
mirror (ss) fish are observed. More detailed information on pleiotropic effect of genes for scale cover can
be found in Chapter 1. The Doitsu varieties are less appreciated by the Japanese because of body shape
considerations, but in most other countries the blight color qualities are sought after. Growth depression
and poor health is common among doitsu and a higher probability of individuals suffering from a range of
morphological defects can be found. These traits include reduction of fins, barbels and lateral line, and
spinal deformities.
Most doitsu varieties like Kikusui, Kumoryu, doitsu Purachina, etc. are of the leather class. Shusui
is from the mirror class, and it may account for its superior morphologic appearance. There is no
particular rule since breeders mostly are oblivious of the genetic makeup of their brood stock.
Development continues to improve these qualities in doitsu koi mainly in terms of growth and body
conformation. The Shusui is selected from one of the sub types of the resultant mirror types. In Table 2,
cross 5, mirror x mirror, will yield the highest proportion Shusui to select from. For the best possible
improvement in body conformity and growth, scaled or linear females can be used that are tested to be
heterozygous for the S/s gene. The linear phenotype is not generally accepted and culled out. While
stringent culling is needed with every mating, many less attractive phenotypes slip through, especially
when breeding is intended for export to countries with different preferences.
Cross leather x leather (cross 10 in Table 2) produces the highest proportion (66.7%) of leather
offspring. However, for non-Shusui type doitsu the common breeders practice is crossing females of the
mirror type with males of the leather type (cross 7 in Table 2); this cross gives 50% mirror fish and 50%
leather fish. The 50% mirror component is culled out except a few future mirror females selected for best
body conformation and color aspects, but ignoring scale cover. The 50% leather component should

27
produce the marketable product as well as future male parents, while selecting for all the qualities
demanded by the variety. Because of possible resemblance between some mirror and linear fish, the test
crosses of presumably mirror females are sometimes performed for confirmation of their genotypes.
Crossing of scaled females, tested to be heterozygous for the S/s gene, with leather males (cross 4 in
Table 2) is also used in breeders practice for production of leather fish.
Variation in scale reflection
The ginrin is phenotype with iridophores in the scales containing reflecting platelets composed primarily
of guanine (De Kock and Watt, 2006, Gur et al. 2013). Ginrin varieties are known in all color types and
independent of scale cover type. For show purposes, ginrin varieties are collectively known as Kinginrin
since it emulates silver when the reflective scale is against a black or white background and gold when
viewed against a yellow or red background. The term literally describes the golden and silver visual effect
on the appearance of the fish.
Based on appearance, ginrin is classified in two main groups (De Kock and Watt 2006). In the first
group the guanine forms a lumpy pearl-like deposit in the centre of the scale, giving it an additional
dimension referred to as pearl ginrin. In the second group the guanine platelets are flat, forming mirror-
like structures which are judged like a diamond on its reflectivity from multiple angles. Dia-gin and Beta-
gin are the most well-known of the later group. Beta-gin first appeared in Niigata after the turn of the
century, and it is speculated that it was introduced via the German Aischgrunder imports of 1904. Dia-gin
appeared much later in the Hiroshima region and has more popular appeal due to multiple reflective
surfaces imitating diamonds. The ginrin also has quantitative properties ranging from just having one
reflective scale, to a few rows, to down to the lateral line, and more rarely completely wrapping the body.
It also seems that the development of ginrin is later and the culling regime needs to be adjusted
accordingly. Ginrin, like the phenotypes with reduced scale cover, also finds expression in other
morphological changes noticeably in body conformation, growth and health.

28
Figure 9. Ginrin. A. Beta-gin, also called Niigata ginrin is more lumpy and dull; B. Dia-gin or
Hiroshima ginrin has larger crystal platelets reflective in many angles and is very popular.
Variation and inheritance of fin length
The long-fin variety of koi, sometimes called butterfly koi, is a relatively new morph that has been
developed in the last several decades. In the United States, long-fin koi have been developed at Blue
Ridge Fish Hatchery (North Carolina, USA) from the middle 1980s by crossing of normal short-fin koi
with the long-fin common carp of Asian origin (LeFever 1991, 2010). This long-fin common carp
apparently originated from Indonesia, where a local long-fin strain of common carp (also called
“kumpay”) has been described. There is information that the long-fin morph of koi was obtained in Japan
in the early 1980s by crossing koi with Indonesian long-fin carp. In Japan, long-fin koi are not allowed to
be demonstrated at koi competitive shows, yet some breeders do produce a range of popular varieties for
the export market. These include all the scale types including wagoi, doitsu and ginrin. Long-fin koi are
popular in the United States and most other countries.
Inheritance of long fins in koi was investigated by Gomelsky et al. (2011) in a study performed in
the United States. Fish segregations with regard to the presence or absence of long fins in two progenies

29
were recorded and analyzed. In the first progeny, produced by crossing a long-fin koi female with a
regular short-fin koi male, the observed segregation of long-fin fish : short-fin fish did not differ
significantly from the 1:1 Mendelian ratio. In the second progeny, produced by crossing the same long-fin
female with a long-fin male, the observed segregation of long-fin fish : short-fin fish did not differ
significantly from the 3:1 Mendelian ratio. Based on these data, it was concluded that the appearance of
long fins in koi is under the control of a dominant mutation of one gene (Lf/lf ). Fish with genotypes LfLf
and Lflf have long fins, while fish with genotype lflf do not have this trait. Analysis of fish variability in
the second progeny has shown that fish with genotypes LfLf and Lflf did not differ with regard to relative
length of the tail; this indicated that the Lf allele is characterized by complete dominance over the allele lf.
The authors noted that since appearance of long fins is controlled by a dominant mutation, it would be
relatively difficult to develop a true-breed, long-fin line since the crossing of two heterozygotes (Lflf )
gives the mixed progeny (Gomelsky et al. 2011). Development of a true-breed, long-fin line could be
achieved only by identifying heterozygotes Lflf by means of test crosses and removing them from the
stock.
Color variability in gynogenetic and triploid progenies and production of clones in koi
Similar to other fish, it is possible to produce in koi gynogenetic and triploid progenies. Gynogenesis is
embryo development under control of only maternal heredity. In order to induce gynogenetic
development in koi (as well as in any other fish reproducing by normal sexual mode) eggs should be
inseminated by genetically inactivated (usually by irradiation) sperm. For production of viable
gynogenetic diploids (2n), haploid chromosome set (n) in eggs should be doubled. This can be achieved
by suppression of either second meiotic division (meiotic gynogenesis) or first mitotic (cleavage) division
in haploid embryos (mitotic gynogenesis) by strong physical treatments (usually by cold or heat shocks).
It needs to be noted that in spite of the fact that fish obtained by induced gynogenesis are originated from
female parent only, their genotypes do not copy maternal genotype. During egg development, the
recombination of genetic material by chromosome crossing over occurs; therefore gynogenetic fish

30
resulted from suppression of the second meiotic division in eggs (meiotic gynogenesis) do not have
maternal genotype. Fish obtained by mitotic gynogenesis are homozygous for all genes because in this
case homologous chromosomes are completely identical as products of mitotic reduplication in haploid
embryos. Normally, many genes in fish genotypes are at heterozygous state, therefore completely
homozygous mitotic gynogenetic fish cannot have maternal genotype. By suppression of 2nd meiotic
division in eggs after insemination of them with intact (non-irradiated) sperm triploid (3n) fish can be
obtained. More detailed information on production of gynogenetic and triploid progenies in common carp
and koi can be found in Chapter 5 of this book. The data on color variability in gynogenetic and triploid
koi are presented below.
Taniguchi et al. (1986) have investigated color variability in three progenies obtained from
Kohaku female: normal (amphimictic) progeny obtained by crossing Kohaku female with Taisho Sanke
male, triploid progeny obtained with the same cross and suppression of 2nd meiotic division in eggs by
cold shock, and gynogenetic progeny produced by genetic inactivation of sperm by irradiation and
suppression of 2nd meiotic division in eggs in a study performed at Kochi University (Japan). Area of red
and black spots of on the body was determined by measuring free-hand drawing with an image analyzer.
There was no difference between fish from gynogenetic and normal progeny in the frequency distribution
of red color area; the authors noted that this suggests that genes for red (R) and white (W) have no allelic
relationship. Black spots were not observed in fish of gynogenetic origin; it proved genetic inactivation of
male’s chromosomes. The area of black spots was lower in triploids compared with the diploids. The
authors noted that this difference may be due to male parent’s allele (B) being suppressed as a result of
the additive effect of the female parent’s allele (b) which negatively affects black color development
(Taniguchi et al. 1986).
Gomelsky et al. (1996, 2003) have observed that proportions of white-red fish in meiotic
gynogenetic progenies produced from Kohaku females varied from 47.8 to 88.9% (with mean value for
six progenies 77.8%) and were usually higher than proportion of white-red fish observed in crosses of
Kohaku parents. Gomelsky et al. (1996, 1998) have recorded that proportion of pigmented fish in meiotic

31
gynogenetic progenies obtained from females with Bekko type of pigmentation (Shiro-Bekko and Taisho-
Sanke) have varied from 63% to 78% with a mean value for six progenies of 71.7%. It is known that the
frequency of heterozygotes in meiotic gynogenetic progenies depends on the crossing-over frequency
between the gene and the centromere (see Chapter 5 of the book). Based on the observed proportion of
fish with black pigmentation in meiotic gynogenetic progenies (71.7%) obtained from heterozygous
(Blbl) females, the recombination frequency between gene Bl/bl and centromere was calculated to be
about 0.4 (Gomelsky et al. 1996, 1998). In contrast to meiotic gynogenesis, mitotic gynogenesis results to
homozygosity for all genes. Therefore mitotic gynogenetic progeny produced from heterozygous females
Blbl should, theoretically, consist of two types of homozygotes (BlBl and blbl). In one from the two
mitotic gynogenetic progenies obtained from heterozygous females Blbl the ratio of pigmented and un-
pigmented fish was very close to the expected ratio 1:1; however, in the second mitotic gynogenetic
progeny the strong prevalence of pigmented fish was observed. It was suggested that low viability of un-
pigmented gynogens in this progeny was due to the action of some lethal genes, which were expressed
only in the absence of dominant allele Bl (Gomelsky et al. 1998).
Meiotic and mitotic gynogenetic progenies in koi have been obtained and investigated at Niigata
Prefectural Inland Water Fisheries Experimental Station (Sato and Amita 2001, Sato 2013). In meiotic
gynogenetic progenies obtained from Kohaku females, proportions of white-red fish varied significantly
and sometimes were higher than typical proportion of white-red fish in Kohaku x Kohaku crosses (60-
70%). On the contrary, proportions of white-red fish in mitotic gynogenetic progenies obtained from
Kohaku females were very low (0-1%) while proportions of other two color types, solid red and solid
white fish, were variable; proportion of solid red fish in these progenies varied from approximately 50%
to 95%.
Clones (or groups of genetically identical individuals) in fish can be produced by obtaining the
second consecutive gynogenetic generation from females of mitotic gynogenetic origin (for details see
Chapter 5). A clone in koi has been produced for the first time at Niigata Prefectural Inland Water
Fisheries Experimental Station in 1998 by inducing meiotic gynogenesis from a solid red female obtained

32
earlier by mitotic gynogenesis from Kohaku female. All fish of the obtained clone have been solid red,
the same as female of mitotic gynogenetic origin, from which this clone has been originated (Sato 2013,
S. Sato, personal communication).
As mentioned above, offspring obtained by induced gynogenesis do not have maternal genotype.
Therefore the method for production of clones in fish by consecutive gynogenetic generations does not
allow copying of valuable show-quality koi. The following question arises: is there any other
biotechnological method which does give opportunity to clone grand champions? Theoretically it is
possible using so-called somatic cell nuclear transfer (or SCNT) technology. In 1997 the first cloned
mammalian animal (sheep Dolly) was obtained by this method (Wilmut et al. 1997). In this case the
nucleus from cultured specialized (mammary) somatic tissue derived from adult sheep was transferred
into an egg with preliminary removed original nucleus. The production of Dolly became a widely known
scientific achievement and drew attention by public media. Ironically, similar pioneer study in fish was
reported more than decade before first cloning of mammal from adult somatic cell. Chinese scientists
(Chen et al. 1986) have succeeded to produce adult crucian carp by transferring the nucleus from kidney
tissue (derived from adult fish) to an egg after removal of the original nucleus. In the last 10-15 years the
technology of somatic cell nuclear transfer in fish has been developed further (mostly in model species,
zebrafish and medaka) by applying some advanced biotechnological methods (Lee et al. 2002,
Wakamatsu 2008, Siripattarapravat et al. 2009). Recently Tanaka et al. (2012) reported the results of a
study performed in Japan on attempt of cloning high-quality goldfish breed ‘Ranchu’ by fin-cultured cell
nuclear transplantation. As was reported, several embryos with transferred nuclei have reached hatching
stage. In future, somatic cell nuclear transfer technology could be applied for obtaining genetic copies of
high-quality koi. However, research and development in Nishikigoi is a slow process because females
require several years to mature. Also, application of different experimental treatments to eggs and early
embryos could seriously affect other evaluation criteria (besides color pattern) such as body
conformation. Therefore, it is doubtful that koi cloning can be effectively done commercially in the near
future.

33
Acknowledgements
The authors thank Katshusi Takeda, Ronnie Watt, Pierre Jordaan, Kiyoko Fujita, Gertan Agenbach and
Toru Inouye for valuable and essential assistance, Shoh Sato for providing unpublished data and some
articles, Yuka Kobayashi for interpreting articles in Japanese, Alexander Recoubratsky and Viktor
Dementyev for providing some articles and Charles Weibel for help in formatting of figures.
References
Chen, H., Y. Yi, M. Chen and X. Yang. 1986. Studies on the development potentiality of cultured cell
nuclei of fish. Acta Hydr. Sin. 10:1-7 (in Chinese); Republished in English: 2010. Int. J. Biol. Sci.
6: 192-198.
Cherfas, N.B., Y. Peretz and N. Ben-Dom. 1992. Inheritance of the orange type pigmentation in Japanese
carp (Koi) in the Israeli stock. Isr. J. Aquac.-Bamidgeh 44: 32-34.
David, L., Blum, S., Feldman, M. W., Lavi, U. and Hillel, J. (2003). Recent duplication of the common
carp (Cyprinus carpio L.) genome as revealed by analyses of microsatellite loci. Molecular
Biology and Evolution.
David, L., S. Rothbard, I. Rubinstein, H. Katzman, G. Hulata, J. Hillel and U. Lavi. 2004. Aspects of red
and black color inheritance in the Japanese ornamental (Koi) carp (Cyprinus carpio L.).
Aquaculture 233: 129-147.
De Kock, S. and R. Watt. 2006. Koi: A Handbook on Keeping Nishikigoi. Firefly Books Inc., New York.
Gomelsky, B., N.B. Cherfas, N. Ben-Dom and G. Hulata. 1996. Color inheritance in ornamental (koi)
carp (Cyprinus carpio L.) inferred from color variability in normal and gynogenetic progenies.
Isr. J. Aquac. - Bamidgeh 48: 219-230.
Gomelsky, B., N.B. Cherfas and G. Hulata. 1998. Studies on the inheritance of black patches in
ornamental (Koi) carp. Isr. J. Aquac.-Bamidgeh 50: 134-139.
Gomelsky, B., N. Cherfas, G. Hulata and S. Dasgupta. 2003. Inheritance of the white-red (Kohaku) color
complex in ornamental (Koi) carp (Cyprinus carpio L.). Isr. J. Aquac.- Bamidgeh 55: 147-153.

34
Gur, D., Y. Politi, B. Sivan, P. Fratzl, S. Weiner and L. Addadi. 2013. Guanine-based photonic crystals in
fish scales form from an amorphous precursor. Angew. Chem. Int. Ed. 52: 388-391.
Hoshino, S. and S Fujita. 2009. Nishikigoi Mondo (English version). NABA Corporation, Japan.
Iwahashi, M. and M. Tomita. 1980. Genetical studies of nishikigoi (Cyprinus carpio). II. On the heredity
of Kohaku. Rep. Niigata Inland Waters Fish. Exp. Stn. 8: 74-79 (in Japanese with English
summary).
Jordan, D. S., S. Tanaka and J. O. Snyder. 1913. A catalogue of the fishes of Japan. J. Coll. Sci. Imp.
Univ. Tokyo 33 (article 1): 1-497.
Kataoka, M. 1989. Nishikigoi dangi (Nishikigoi discussion). Takayoshi Kataoka Kumagaya-shi, Saitama,
Japan (in Japanese).
Katasonov, V.Y., 1973. Investigations of color in hybrids of common and ornamental (Japanese) carp: I.
Transmission of dominant color types. Genetika 9: 59– 69 (in Russian with English summary).
Katasonov, V.Y. 1974. Investigations of color in hybrids of common and ornamental (Japanese) carp: II.
Pleiotropic effect of dominant color genes. Genetika 10: 56–66 (in Russian with English
summary).
Katasonov, V.Y. 1976. The lethal action of the gene of light pigmentation in the common carp Cyprinus
carpio L. Genetika 12: 152-161 (in Russian with English summary).
Katasonov, V.Y. 1978. A study of pigmentation in hybrids between the common carp and decorative
Japanese carp: III. The inheritance of blue and orange patterns of pigmentation. Genetika 14:
2184-2192 (in Russian with English summary).
Katasonov, V.Y., V.N. Dementyev and A.V. Klimov. 2001. Genetics of color in the ornamental carp: the
inheritance of underlying dark coloration. Russian Journal of Genetics 37:1210-1211.
Kirpichnikov, V.S. 1981. Genetic bases of Fish Selection. Springer-Verlag, Berlin.
Kirpichnikov, V.S. 1999. Genetics and Breeding of Common Carp. INRA, Paris.
Koshida, S. 1931. Fish Farming as Side Business for Farmers. Niigata Farmers Association, Niigata,
Japan (in Japanese).

35
Kuroki, T. 1981. Manual to Nishikigoi. Shin Nippon Kyoiku Tosho Co., Shimonoseki-city, Japan.
Kuroki, T. 1987. Modern Nishikigoi, 2nd Ed., Shin Nippon Kyoiku Tosho Co., Shimonoseki-city, Japan.
Lee, K.Y., H. Huang, B. Ju, Z. Yang and S. Lin. 2002. Cloned zebrafish by nuclear transfer from long-
term-cultured cells. Nat. Biotechnol. 20: 795-9.
LeFever, W. 1991. The new butterfly koi. Tropical Fish Hobbyist (November):78-83.
LeFever, R. 2010. The origin of butterfly koi. Pond Trade Magazine (March/April):12-14.
Novelo, N.D. and B. Gomelsky. 2009. Comparison of two methods for measurement of red-area coverage
in white-red fish for analysis of color variability and inheritance in ornamental (koi) carp
Cyprinus carpio. Aquatic Living Resources 22:113-116.
Okada, Y. 1960. Studies on the freshwater fishes of Japan II. Special Part. J. Faculty of Fish. Prefec.
Univ. of Mie 4: 1-860.
Okubo, E. and R. Sato. 1975. Racial study on “Asagi” carp from the Matsushiro. Bull. Freshwater Fish.
Res. Lab. 25: 73-81 (in Japanese with English summary).
Sato, S. 2013. Breeding in Nishikigoi, Cyprinus carpio of Niigata prefecture. Fish Genetics and Breeding
Science 42: 81-84 (in Japanese with English summary).
Sato, S. and K. Amita. 2001. Induction of mitotic-gynogenetic diploid Nishikigoi, Cyprinus carpio. Rep.
Niigata Inland Waters Fish. Exp. Stn. 25: 1-5 (in Japanese with English summary).
Smartt, J. 2001. Goldfish Varieties and Genetics. A Handbook for Breeders. Fishing News Books -
Blackwell Science, Oxford, UK.
Siripattarapravat, K., B. Pinmee, P.J. Venta, C. Chang and J.B. Cibelli. 2009. Somatic cell transfer in
zebrafish. Nature Methods 10: 733-735.
Suzuki, R and M. Yamaguchi. 1980. Meristic and morphometric characteristics of five races of Cyprinus
carpio. Jap. J. Ichth. 27: 199-207.
Tanaka, D., A. Takahashi, A. Takai, H. Ohta and K. Ueno. 2012. Attempt at cloning high-quality goldfish
breed ‘Ranchu’ by fin-cultured cell nuclear transplantation. Zygote 20: 79-85.

36
Taniguchi, N., A. Kijima, T. Tamura, K. Takegami and I. Yamasaki. 1986. Color growth and maturation
in ploidy-manipulated fancy carp. Aquaculture 57: 321-328.
Waddington P. 2009. Article in Koi-Bito. http://www.koi-bito.com/forum/main-forum/9409-peter-
waddington-article.html
Wakamatsu, Y. 2008. Novel methods for the nucleus transfer of adult somatic cells in medaka fish
(Oryzias latipes): use of diploidized eggs as recipients. Dev. Growth Diff. 50: 427-436.
Watt R, and S. De Kock. 1996. Living Jewels: Koi Keeping in South Africa. Delta Books, Johannesburg,
RSA.
Wilmut, I., A. E. Schnieke, J. McWhir, A. J. Kind and K. H. S. Campbell. 1997. Viable offspring derived
from fetal and adult mammalian cells. Nature 385: 810-813.
Keywords
Koi, ornamental carp, Nishikigoi, color types, koi varieties, Japan, history, genetics, inheritance, fish
color, scale cover types, ginrin, gynogenesis, clones