GENETIC COLOR POLYMORPHISMS ON COMMON CARP STOCKS IN INDONESIA

Common carp is the most important species of fish culture in Indonesia, especially in the island ofJava, The objective ofthis study was to clarify the cellular basis ofcolor inheritance among body morphs of this fish. Mendelian test by single pair intraor inter-color matings of green, yellow and red ehowed that the Fr Beneration gave evidence that these phenotypes had t,he genotypes proposed by previous workers. However, this study proposed other gene symbols for scale type, i.e., T (non-transparent) for wild type and I for transparent. Two loci (R, and r,, R, and rr) with additive gene effect were probably involved in the production of yeliow/red pigment, which were present in the yellow and red morphs. The dominant alleles controlled the amount, of pigment produced. XEY\ilORDS: oolor inheritauce, genetics, colnnoon carp.


INTRODUCTION
Domesticated populations of carp in Indonesia have striking color polymorphisms.Matricia (1990) reported that higher proportion of green common carp were cultured by most farmers in West Java, West and North Sumatera.However, some farmers in West and East Java specialized in producing orange or red colored carps.
According to Gustiano & Phang (1993), specific color selection in each area is influenced by the consumer preference.It is therefore important to elucidate the Mendelian inheritance of color in Indonesian common carp.Katosonov (1978) working on crosses between Japanese Koi and wild type common carp from the Amur River in Russia showed that the latter was dominant over light colors (orange/yellow and white) of Koi carp.Orange color which was due to a lack of black pigment, is the joint result of 2 recessive alleles O, and b) at 2 independent loci.Genes determining orange color may produce another coloration when combined with other color genes.He suggested that b, and b, genes might, together with the recessive gene (rfof blue color, gave the white eolor.The genotype of the white Koi might be b,b,brbrrr.He also reported that appearance ofblack pattern did not appear to be connected with the inheritance of orange and white.The autosomal D and d gene pair was responsible for the presence and absence of the specific pattern on the body of the fish (Katosonov, 1973).Another autosomal gene pair (L and I) determined dark (common) or light color of the entire bodv.
WohHarth & Rothbard (1991) and Oherfas et al. (1992) studied the inheritance of the orange colored Koi carp in the Israeli stock and their results fitted the genetic model proposed by Katosonov (1978).A number of eolor mutants, including gold and gray, had appeared among common carp in Israel (Moav & Wohlfarth, 1968) and Poland (Wlodek, 1968).All were receseive to the wild type coloration, and were inherited independently.
Based on the above information a study was conducted to confirm the inheritance of the green, yellow and red morphs of common carp cultured in lndonesia.

MATERIALS AND METHODS
Three uniform color morphs of common carp, green, yellow and red, were used in this study.
The green stock, Racljaclanu, was obtained from an isolated farm in Kuningan, West Java.The red stock, Canglvingan, was broughtfrom Yogya.
karta and the yellow one, Sinyonyo, was from Sukabumi, West Java.They were all stocked at the Oijeruk research station, about l0 km from Bogor, West Java, for spawning.

Single-pair Matings to Produce F, Generation
Five intra.colorsingle-pair matings were set up for each (green, yellow and red) morphs to produce the F, progeny, during June lggl.September lgg2.Green ( G ) R€d (R) were scored when their colors developed (10 to 21 top with used ftoftobons' Newly hatched frv days after hatching) using a color standard (on remained in the hapaup to i3 days and were then their phenotypes) for textile.transferred into 40 mz earthen pond, filled with The other data recorded were mating number, water' containing chicken manure (15O0 g/m)'^of parent number (fin clip code), female ;;il;, 60 cm dgqth Fry were left to grow in the pond for number of progeny, and color phenotypes.Th#; t0 to Jl davs after which they were individuallv broocl segregation data for color phenotyp;; f"t: ::g"t-d for color phenotype' oommercial powrlered each maring were tested by Ohi.squar" #;it;;; {t:n,M las given as supplementarv ftr.tl afl'er to determine if the observed ratio fit ttte e*p*cied. 10days of stocking' Spawning and Fry Rearing Natural spawning was carried out in a hapa (2xtx1 mB size), suspended in a 40 m2 earthen pond.Each spawning used 6 kakaban; (1x0.il m'?) which are palm fiber mats, for the attachment of eggs.
These hahabans were kept floating on the water surface in the hopo during spawning.They were inspected regularly and were turned over when the lower side was full of eggs.
Each spawning used equal ratio of body weight of one male to one female.Hatching of the eggs in the hapa occurred af,ter 2 days.After hatching of the eggs, the kahabans were removed and the hqpo was shelteredfrom direct sun light by covering the

Green X Green Mating
From 5 matings, all F, were green colored,like their parents (fable 2).Therefore, all parenus were homozygous for green coloration.However, they were segregated into two scale phenotypes' The first was the non'tranparent (wild) green scale and the other was the transparent green ccllor scale.
Microscopic observations showed that, the scales ofthe transparent green phenotype did not have the pigment cell type, iridophores, as light reflecting cells, which were present in the non- tranparent green phenotype.Ohi'square tests showed that aII the observed F, ratios of non' *=not significantly different (P20.05),**=highly significant different (P=0.0 l) transparent green : fransparent green, except mating 4, fit the expected ratio of l3 : 1.The F, from mating 4 had excess of non.transparen[ green.These results showed that the green parents were probably heterozygous with the non.
transparent wild gene being dominant to the transparent gene.

Yellow X Yellow Mating
The results of yellow X yellow single.pairmatings are shown in Table 3.The F, progeny from mating I and 2 were all yellow.
The F, progeny from matings ll, 4 and 5 showed yellow or white background (Tables l3 and 4).
There were more progenies with yellow color than white.Chi-square tests showed that they deviated significantly from the expected 3 : I ratio.Among the progenies of matings i) and 4, there was an excess ofyellow and deficiency of white phenotype.
This deviation from the il : I ratio may be due to low viability of white progeny.The F, offspring of mating 5 consisted of about I times more yellow than white phenotype.These results showed that the parents were heterozygous for yellow color genes.
IFR Journal Vol, V No.l, 1999 F, Frogenies of 4 matings also showedpresence or absence of black color pattern on the backgroundcoloration (Iable 5).With pattern and without pattern among the F, from matings 2 and 5 did not fit the expected ratio from Ohi-square test.The deviation from the expected ratio in mating 2 was due to an excess of without pattern and deficiency of with pattern phenotype.In contrast, in mating 5 there was more off spring with pattern compared to with expected numbers.
In mating ll and 4, the ratio between without pattern and with pattern did not fit a I : I ratio.f)eviation from expected numbers was due to an exeess of without pattern phenotypes in mating 3.
The reverse was found in mating 4. Thus the results showed that there was no specffic relation.
ship between the with pattern and withrutpattern phenotypes to survival rate.The gene for pattern was dominant to without pattern.Variation in number of with pattern and without pattern probably was due to the effect of background color (yellow and white) since in this analysis the data was pooled according to presenee or absence of pattern..ds sown in the previous paragraph, in background analysis, yellow showed better   survival than white background.Another reason could be that the pattern is a multigenic trait.
Thus the observed segregation data did not fit the expected ratios based on single gene inheritance.

Red X Red Mating
Four matings of red X red gave all F, progenY with red background (table 6) showing that the parents were homozygous for red color.
However, the F, progeny from mating 2 segregated into 2 phenotypes, 2,087 red 660 red with greenish band along the dorsal region.The latter is referred to as the greenish red color pheno' type.The Ohi-square test showed that the ratio of red to greenish red offspring from this mating fit the expected il : 1 ratio.This gave evidence that both parents of mating 2 were heterozygous with the gene causing the greenish band being recessive to the no banding allele.Green X Yellow Cross All three single-pair matings of this cross gave only green F, progeny showing that green color Table 6.F, offspring of red X red matings.
was epistatic to yellow (Table 7).The yellow phenotype was the result of combination of two recessive genes, b, andb, (Katosonov, 1978).Thus, the green female parent contributed the dominant B, or B, gene that controlled black pigmentation.
However, the reciprocal cross of yellow female X green male gave one quarter yellow and three quarters green progenies (Table 8).There was no significant difference from the expected ratio of ll green : 1 yellow, except in mating iJ where there was an excess of green and deficiency of yellow' These results showed that green male parents were most probably heterozygous for green color and yellow females homozygous for yellow color.This study proposed that the green genotype was BlbrB2bzR-and yellow was b,brbrbrRR following the mbdel proposed by Katosonov (1978).

Green X Red Cross
This crossing gave similar result to those of the green with yellow cross, All F, Progeny from red male X green female crosses were green showing that green color was epistatic to red (Table 9).The mechanism of epistasiri was Mating Breeder number number    probably similar to the crosses between yellow male and green female.
In the reciprocal cross using red female as spawner, the F, progeny consisted of one quarter red and three quarters green.All Chisquare tests showed no significant dilferences at P=0.05 (Iable l0).The results were similar to those of the cross between yellow female and green male.Thus, these results supported the possibility of heterozygosity of all green male parents used in this cross.

Red XYellow Cross
All F, progeny from the cross of red male with yellow f6male had red color backround (Iable 11) with exception of mating 2 which also had 29 yellow progeny.The red color was most probably controlled by more than one gene locus with DISCUSSION Mating of green X green showed that all the parents were homozygous for the green color.However, they were heterozygous for the scale type sinee the F, progeny showed approximately 3 : I ratio between non'transparent (wild) and tranparent scale type (Iable 2).The green male and female parents used in these matings were probably heterozygous for scale type, with the non' transparent being dominant over the transparent scale type.Previous workers on common carp also reported that the wild phenotype is dominant over other mutant phenotypes (Katosonov, l9?8;Cherfas et al. 1990).The proposed gene symbols for the scale type is T (non'transparent) for the wild-type and t for transparent.Both parents of the green X green matings must be Tt hetero-zygOtes.additive gene effects which were involved in the production of yellow/red Pigment.
The reciprocal cross using green male with red female gave F1 progeny with red color background (fable l2).The results showed that most of red parents used in these crossings were homozygous dominant.
Genetic studies on goldfish showed that the character of transparent scale was controlled by a single autosomal gene (Kajishima, 19?7)'The dominant allele (g) controlled guanine synthesis in the iridophores while the recessive allele (g) lacked this potency thus causing scales to be transparent.Similarly, Moore (1974) working on The F, offspring from yellow X yellow matings could be classified into two groups based on their color backgrounds, yellow and white.All of the white progeny had black and yellowish pigmen- tation, in the iris of the eye.Study of albinistic common carp from Roosevelt lake, Arizona, by Johnson (1968) reported that none ofthem were true albinos, sinee all had dark, wine.redeyes.The present data showed that ratio between yellow and white was close to il: 1, except in matings 1, 2 and 5 (fable 4).However, Ohi.square tests showed that the observed ratio was significantly different from the expected.The proposed genotype for yellow parents for matings I and 2 was homozygous dominant, RR and heterozygous Rr for parents of the other three matings.
Deviation from the expected ratio in mating B and 4 wtrs due to an excess of yellow and deficiency of white phenotypes.Lower fitness of the white colored morph might be the cause of the observed deviation from the expected.Similar observations were ret)orted by Wohlfarth & Rothbard (lgg1).Katosonov (1978) and Oherfas et al. (lgg2) concluded that genes for white coloration in common carp were the most recessive among the color phenotypes.They reported that genes for white coloration, were b,b,brbrrr.A similar geno- type was also suggested by Yamamoto (lg?il) for albino goldfish. Katosonov (1976) reported that the light color genes in carp manifested lethal effects when homo- zygous.with mortality occurred during the post embryonic period, mainly at the fingerling stage of development.Bondari (1984) reported that albino fry channel catfish were also less viable than normal ones.According to Bridge and Limbach (1972), the rareness ofthe occurrence ofalbinos in natural rainbow trout populations was because that they were not well suited for natural stream or lake environments.
F, from yellow X yellow matings were classified and analyzed separately for background color and black color pattern.Wohlfarth & Rothbard (l ggl) also reported that the appearance of colored pattern did not appear to be connected with the IFR Journal Vol. V No.t, 1999 inheritance of color background.According to Katosonov (197;l), background color and color pattern were due to the presence of two unlinked, clearly autosomal gene pairs.This stuclv propnsed that without pattern genotype was controlled by the dominant gene (P) and with pattern by the recessive allele (p).This proposed genotype was probably similar to the gene f), which controlled pattern on common carp as proposed by Katosonov (1978).The F, results of presence or absence of pattern from yellow X yellow matings showed 2 kinds of ratios.The first was il : I showing that both the parents were heterozygous for gene pattern, pp.The seeond was I : I (table 5) which meant that in that particullar case one of the parents was homozygous recessive for pattern (Pp).However, that homozygous recessive parent for pattern was undetectable at spawning time.This could be possible because all the spawners were kept in a concrete tank for a long time before the crossing experiment.In some fishes.the color pattern, may disappear or becomes blurred during the long period of storage in concrete tanks where they were fed with commercial pellets.
In the red X red matings, all the parents were homozygous for red background eolor, since all the progeny were red colored.However.mating 2 produced some greenish red progen.yand the obseved ratio confirmed to il : 1 ratio between red and greenish red (lable 6).This shows that the red parents of this mating must be heterozygous, with the greenish band gene being recessive.
However, it was difficult to propose a gene model responsible for the band based on results from one mating.
Green X yellow and green X red crosses gave similar results.When green females were used in the erossing, all the progeny were green (Iable 7 and 9).The proposals of green backgound color (wild-type) being epistatic to light colors (yellow and red) are reported by previous workers on common carp (Katosonov, lg78;Taniguchi el ol., 1986; Oherfas et a1.,1990).
However when green colored male was used to cross with yellow or red females, the progeny produced were approximately one quarter yellow or red color, respectively (fables 8 and l0).These results showed that green parents were most pr-obably teterozygous (B,b,BrbrRr) while the yellow and red female were liomozygous dominant, ()rossing between yellow and red produced F, progeny with red background (fable I I ancl l2j indicating that red was dominant over yellow coloration.Ohemical analysis of the pigment cells (Gustiano, 1995) showed that red and yellow morphs had similar pigment contents.However, they differred in the density of carotenoids in their chromatophores.Red and yellow color were probably polygenic characters Each genetic combination produced a variation in the carotenoid content.Previous workers (Katosonov, 1978; Wohlfarth & Rothbard, l99l; Cherfas e.tal,199D, working on orange and yellow color of Koi carp, did not analyze them separately, but pooled them into one group.It was therefore difficult to confirm the results of the red X yellow cross.We proposed that two loci (R, and r,, R, and rr) were probably involved to form yellow and red phenotypes.In contrast to Iiatasonov (19?8) who proposed a single gene locus (R and r), our data gave indication that more than 1 locus were involved in the production of yellow and red pigments.These 2 dominant genes controlled the amount of pigments produced.
The other source of the intensity differences on red and yellow was probably due to environmental effect like diet, sunlight and water quality.
Proposed genotype (s) the various color pheno' type (s) of common carp are: Various color phenotypes of common carp cultured in Indonesia were mainly controlled by simple mendelian inheritance.Symbols for gene loci in common carp proposed are as follows : R, and 8,, for common green body background (wild type ailele), R, and R, for red or yellow.T for wild non-transparent scale type and P for unilbrm body color background.B, and B, are epistatic to the R, and R, genes.

Table 1 .
ln ter-color single.pairm atin gs were carried ou t, three pairs for each cross (Iable l).F, ofrspring ') Researcher ofthe Research Institute for Freshwater Fisheries.sukamandi Number of single-pair matings for each intra'color and inter-color matins.

Table 2 .
F, offspring of single.pairmatings of green x green common carp.

Table il .
F, offspring of single-pair matings of yellow X yellow.

Table 4 .
F, progeny with yellow background and white back.
ground from yellow X yellow matings.

Table 5 .
F, offspring with and without black pattern from yel' low X yellow matings.Mating Total number of Fr with F\'Fected X'

Table 7 .
F, offspring of single pair crossing of yellow male X green female.

Table 8 .
F1offspring of green male X yellow female.

Table 9 .
F, offspring of red male X green femele croosses.

Table 11 .
Fr offspring of red male X yellow female.

Table 12 ,
Fr offspring of yellow male X red female' Mating number Miller, reported that the transparent allele (tr) which caused absence of iridophores was inherited as a simple autosomal recessive allele.It was likely that the transparent gene acts to terminate proliferation of iridophores.The scales oftransparent green F, progeny found in this study were completely devoid of iridophores.