Hormonal Sex Manipulation and Evidence for

Female Homogamety in Channel Catfish



*Ecological Research Center, Department of Biology, Memphis State University, Memphis, Tennessee

38152, and United States Fish and Wildlife Service, Southeastern Fish Cultural Laboratory,

Marion, Alabama 36756

' Present address: United States Department of Agriculture, Agricultural Research Service, Catfish Genetics Research Unit, P.O. Box 38, Stoneville, MS 38776.

2 Present address: Texas Cooperative Fish and Wildlife Unit, Department of Range and Wildlife Science, P.O. Box 4169, Texas Tech University, Lubbock, TX 79409.

Copyright ~ 1990 by Academic Press, Inc.

All rights of reproduction in any form reserved.

Accepted June 29, 1989

Abstract.- The mechanism of sex determination in channel catfish Ictalurus punctatus was evaluated by hormonal and genetic methods. Aromatizable and nonaromatizable androgens, as well as an estrogen, caused feminization in fish fed steroids for 21 days after yolk-sac absorption. The effectiveness of 60 m g of ethynyltestosterone/g food decreased markedly when the experimental feeding period was shortened and was ineffective when the treatment lasted less than 12 days. Females from all-female populations produced by treatment with sex hormones were mated with normal males resulting in nine spawns with a sex ratio different from 1:1. The sex ratios were statistically similar to 3 male:l female in five spawns, both 2:1 and 3:1 in two spawns, and 2:1 in two spawns. These data are consistent with a model for female homogametic sex determination in channel catfish and suggest that the YY equivalent genotype is viable.

The phenotypic sex of many fish species can be experimentally altered during early development to produce monosex populations. Generally, in teleosts, treatment with androgens results in masculinization and treatment with estrogens results in feminization (Hunter and Donaldson, 1983). Paradoxical feminization (production of allfemale populations by treatment with androgens) has been reported in cichlids (Hackmann, 1974; Hackmann and Reinboth, 1974; Nakamura, 1975). Channel catfish, Ictalurus punctatus, were feminized by a wide range of concentrations of 17a -ethynyltestosterone as well as 17b -estradiol (Goudie et al., 1983).

Hormonal manipulation of phenotypic sex has been accomplished by injection into the egg, by dissolving the hormone in the water during the period of gonadal indifference, or by feeding hormone to young fish (Yamamoto, 1969). Treatment with hormones often produces fish that are sterile, intersex (ovotesticular), or have abnormalities of the genital ducts that prevent spawning (Yamamoto, 1969; Johnstone et al., 1979a, b). Morphological anomalies that would prevent spawning have not been observed in feminized catfish.

Genotypic sex cannot currently be determined in channel catfish by chromosomal analysis. However, the sex genotype and the genetic sex determination model can be determined by mating sex-reversed females with normal male fish and evaluating the sex ratio of the progeny.

This study evaluates the influence of a variety of androgens and time of treatment on sex determination in channel catfish. Breeding evidence is provided that supports a female homogametic sex determination model.


Hormone treatments. Channel catfish were spawned by traditional methods (Dupree and Huner, 1984). Hatched larvae were transferred in groups of 200 to 40-liter glass aquaria held at about 21. Experimental diets were prepared by dissolving hormones in 95% ethanol, spraying the food with the solution to obtain the desired concentration of hormone, and airdrying to evaporate the alcohol. Hormones were tested by feeding fish to satiation three times daily for 21 days starting after yolk-sac absorption. The first day that fish began to feed was designated as Day 1. The hormones tested were 17a -ethynyltestosterone (17-ET), 17a -methyltestosterone (17-MT), 1,4 17a -methyltestosterone (1,4 17-MT), 5a -dihydrotestosterone (DHT), 1a -methyldihydrotestosterone (1-MDHT), 17a -methyldihydrotestosterone (17-MDHT), 11-ketotestosterone ( 1l-KT), 11b -hydroxytestosterone (1l-HT), norethisterone (NE), and NE plus 17-ET.

In a second set of experiments a diet of 60 m g 17-ET/g food, prepared as described above, was administered at selected intervals within the first 21 days of feeding to delineate the minimum effective treatment period. Treatment intervals were from Days 1, 5, 9, 13, or 17 through Day 21, or from Days 1 through 8, 5 through 12, or 9 through 16.

Sac-fry were treated in a third experiment with 17-MT, 17-ET, or DHT on Days 1 and 2 after hatching and exposed under static conditions to 10 or 100 m g/liter of hormone for 2 hr. Water flow was stopped and sufficient hormone solution was added to the tank to produce the desired concentration. After 2 hr, water was turned on and the hormone washed out of the tank.

Progeny sex ratio evaluation. Four-year-old fish from all-female populations produced by 17-ET or 17b -estradiol (Goudie et al. 1983) were individually paired and spawned with normal males in wire cages placed in a pond. A spawning container was placed in each cage. Egg masses were removed and incubated in indoor hatching troughs. Spawning success from mid-May through mid-June under these conditions is normally about 60%.

Progeny from each treatment group were maintained and reared separately. Sex was verified in 10- to 15-cm fish by dissection. Females had paired tubular ovarian structures (Grizzle and Rogers, 1976); in males, gonadal structures were either not discernible or consisted only of opaque convoluted lines on the ventral surface of the kidney.

Statistical analysis. Sex ratios of control and hormone-treated groups were evaluated by x2 analysis based on an expected 1 male:l female sex ratio. The sex ratios of the progeny of hormone-treated females were tested for homogeneity by x2 analysis. When homogeneity was not indicated, groups were tested individually by x2 using Yates correction term with expected ratios of 1:1, 2:1, and 3:1 (Strickberger, 1985). A probability of 0.05 was considered to be significant.


Hormone Treatments

Every hormone fed to fry for 21 days (except 11-HT, which was ineffective) resulted in a significantly higher percentage of females than that in control populations (Table 1). Those most effective at producing females were NE, NE plus 17-ET, 17-ET, 17-MT, and 1,4 17-MT; 17-MDHT, DHT, 1-MDHT, and KT were markedly less effective. Treatment of yolk-sac fry with DHT, 17-MT, or 17-ET added to the water at 10 or 100 m g/liter did not alter the sex ratio (Table 2).

The feminizing effect of 17-ET (at a concentration of 60 m g food) decreased when fish were fed for less than 21 days and was absent when fed for less than 12 days (Table 3). Sex ratios were not significantly affected in fish fed from Days 13 to 21 or from Days 17 to 21, nor by 8-day treatments at the beginning, middle, or end of the 21-day period.

Progeny Sex Ratio Evaluation

Six of 15 spawns produced by crossing hormone-treated females with normal males grouped into a statistically homogeneous set and had a sex ratio of 1:1. The other 9 spawns were not statistically homogenous; however, when tested individually, 7 of the 9 spawns were not significantly different from a 3 male: 1 female sex ratio. Two of those 7 spawns and the remaining 2 spawns were not significantly different from a 2 male:1 female sex ratio (Fig. 1).






(m g/g food)

Fish sexed




Control 0 549 48
17a -Ethynyltestosterone 100 205 100*
Norethisterone 100 250 100*
Norethisterone plus 50  
17a -ethynyltestosterone + 166 100*
170a -Methyltestosterone 60 131 99*
1,4 17a -Methyltestosterone 60 153 97*
17a -Methyldihydrotestosterone 60 205 78*
Dihydrotestosterone 100 149 76*
1a -Methyldihydrotestosterone 60 130 71*
1 l-Ketotestosterone 60 97 63*
11b -Hydroxytestosterone 60 177 56

* Significantly different (P < 0.05) from 1:l sex ratio by x2 analysis.


Although the sexual phenotype of developing channel catfish could be altered by a variety of androgens and dosage rates, deviations from normal sex ratios were limited to feminization. Goudie et al. (1983) suggested that paradoxical feminization in channel catfish may be caused by endogenous aromatization of dietary androgens to compounds with estrogenic properties. Conversion of testosterone by aromatase is part of the normal biosynthesis of estrogen (Fishman, 1982).

Some natural and synthetic androgens are not susceptible to aromatization due to their molecular structure. Aromatizable androgens used in the present study were 17-ET and 17-MT; while DHT, 11-KT, and 11-HT were nonaromatizable (Crim et al., 1981; Callard, 1984). The two other forms of DHT—1-MDHT and 17-MDHT—are thought to be nonaromatizable due to their similarity to DHT and to the addition of methyl groups at positions that interfere with aromatase activity. Norethisterone, an estrogen component of contraceptive pills, has been found to inhibit the aromatase enzyme in vitro by some investigators (Osawa et al., 1982), but not by others (Brodie et al., 1983). Further, 1,4 17-MT has molecular similarities to both nonaromatizable androgens and to aromatase inhibitors (Numazawa et al., 1987).





Hormone Concentration

(m g/liter)

Fish sexed




Dihydrotestosterone 10 163 44.8
  100 150 48.7
17a -Methyltestosterone 10 101 49.5
  100 236 44.5
17a -Ethynyltestosterone 10 245 46.9
  100 281 45.5







Treatment period


Fish sexed




1-2 47 100*
5-2 62 97*
9-2 73 85*
13-21 156 53
17-21 74 55
1-8 69 58
5-12 62 58
9-16 59 61

Note. Groups significantly different from 50% females by x2 analysis (P < 0.05) are identified with an asterisk.

Predominantly female populations were obtained by treatment with nonaromatizable androgens, suggesting that androgen aromatization may not be necessary for paradoxical feminization in channel catfish. However, Nakamura (1981) reported that 11-KT in the diet produced all-male populations of Tilapia mossambica, a species that exhibited paradoxical feminization when sexually undifferentiated animals were treated with aromatizable testosterone propionate in the water (Hackmann, 1974). Furthermore, 11-KT treatment caused precocious appearance of male secondary sex characteristics, but no change in gonadal differentiation. Norethisterone was expected to have prevented the aromatization of 17-ET; however, NE when used alone had a feminizing effect. Treatment with nonaromatizable androgens resulted in the lowest percentage females of any of the feminizing treatments. Higher doses or earlier treatment with any of the feminizing nonaromatizable androgens might produce different sex ratios; however, these androgens apparently have a feminizing effect during the first 21 days.

The minimum effective treatment period for production of all-female populations with 60 m g 17-ET/g food must be close to the first 21 days of exogenous feeding. Reducing the duration of the treatment within any portion of the 21-day period reduced the percentage of females produced. This contrasts with phenocritical periods that are as short as 3 days in Hemihaplochromis multicolor (Hackmann and Reinboth,1974). The six spawns resulting from pairing normal males with females from an all female population had a sex ratio of 1:1 and must have been produced by genotypic females that were unaffected by the steroid treatment. Progeny from the nine spawns that had a sex ratio different from 1:1 must have resulted from breeding genotypic male-phenotypic female fish with normal males. These data are important for two reasons. First, they demonstrate that the hormone treatment does not differentially affect survival of genotypic males, leaving only female fish to develop. Second, this sex ratio is possible only if the genotypic sex determination model is equivalent to a homogametic female condition (XX). The resulting populations would have been composed of all-male offspring if males were the homogametic sex. Additional support for female homogamety is available: all channel catfish progeny produced by gynogenesis were females (Goudie, 1987), and the sex ratio of triploid channel catfish produced by second polar body retention was 1:1 (Wolfers et al., 1981). This female homogametic condition should be used only as a model because neither sex chromosomes nor sex-determining genes have been identified in channel catfish.

The 3 male: 1 female ratio of the progeny obtained from seven of nine sex-reversed females indicated that the male phenotype included both XY and YY genotypes. The normal male genotype (XY) should be present in two-thirds of the males and the YY genotype in one-third. Therefore, the YY equivalent male genotype must be viable in channel catfish, as has been reported in other fish species (Hunter et al., 1982; Nakamura et al., 1984; Chevassus et al., 1988). The YY genotype may not be as hardy as the normal XY genotype in all catfish families, which may explain why two of the populations were statistically similar to only a 2:1 sex ratio. In mammals, some nonsex-related genes essential for survival are located on the X chromosome; consequently, the YY genotype is lethal. Variability in YY viability has been found among families of medaka Oryzias latipes. The YY genotype is rare in families homozygous for xanthic coloration, but occurs with predictable frequency in families heterozygous for this trait (Yamamoto, 1975).

fig1.gif (11147 bytes)

FIG. 1. Percentage of male and female progeny of 15 spawns produced by mating females from hormone-treated groups with normal males. Six spawns were homogeneous and were similar to a 1:1 sex ratio. Horizontal bars indicate those families whose male:female sex ratios were not significantly different (P < 0.05) from the ratio indicated above the bar. The number of fish sampled is shown below each population.

Identification of the YY equivalent males in channel catfish can now be accomplished only by progeny testing. All-male progeny should be produced by the spawning of YY males with normal (XX) females or with sex-reversed (XY) females. In the first case, all offspring would have the XY genotype, and in the latter case, half of the progeny would be expected to have the XY genotype and half would be YY males. These techniques could be useful to the aquaculture industry because male channel catfish grow faster than females (Beaver et al., 1966; Brooks et al., 1982). All-male populations could be produced without sex steroids being administered to fish reared for food.


We thank M. Beck, C. Biggers, T. Tiersch, and T. Wong for their assistance. Supported in part by funds through the Commercial Fisheries Research and Development Act (PL-88-309), the Tennessee Wildlife Resources Agency, and USDA Grant 88-34123-3504.


Crim, L. W., Peter, R. E., and Billard, R. (1981). Onset of gonadotrophic hormone accumulation in immature trout pituitary gland in response to estrogen or aromatizable androgen steroid hormones. Gen. Comp. Endocrinol. 44, 374-381.

Dupree, H. K., and Huner, J. V. (Eds.) (1984). "Third Report to the Fish Farmers: The Status of Warmwater Fish Fanning and Progress in Fish Farming Research." U.S. Fish and Wildlife Service, Washington, DC.

Fishman, J. (1982). Biochemical mechanism of aromatization. Cancer Res. Suppl. 42, 3277s-3280s.

Goudie, C. A. (1987). "Gynogenesis and Sex Manipulation, with Evidence for Female Homogamety, in Channel Catfish." U.S. Fish and Wildlife Service Res. Info. Bull. 87-20.

Goudie, C. A., Redner, B. D., Simco, B. A., and Davis, K. B. (1983). Feminization of channel catfish by oral administration of steroid sex hormones. Trans. Amer. Fish. Soc. 112, 670-672.

Grizzle, J. M., and Rogers, W. A. (1976). "Anatomy and Histology of the Channel Catfish." Agricultural Experiment Station, Auburn University, Auburn, Alabama.

Hackmann, E. (1974). Einfluss von Androgenen auf die Geschlechtsdifferenzierung verschiedener Ci-chliden (Teleostei). Gen. Comp. Endocrinol. 24,44-52.

Hackmann, E., and Reinboth, R. (1974). Delimitation of the critical stage of hormone-influenced sex differentiation in Hemihaplochromis multicolor (Hilgendorf) (Cichlidae). Gen. Comp. Endocrinol. 22,42-53.

Hunter, G. A., and Donaldson, E. M. (1983). Hormonal sex control and its application to fish culture. In "Fish Physiology" (W. S. Hoar, D. J. Randall, and E. M. Donaldson, Eds.), Vol. 9B, pp. 223-303. Academic Press, New York.

Hunter, G. A., Donaldson, E. M., Goetz, F. W., and Edgell, P.R. (1982). Production of all-female and sterile coho salmon, and experimental evidence for male heterogamety. Trans. Amer. Fish. Soc. 111, 367-372.