Transactions of the American Fisheries Society 118:427-434, 1989


Influence of Sex on Growth of Channel Catfish


Ecological Research Center, Department of Biology
Memphis State University, Memphis, Tennessee 38152, USA


U.S. Fish and Wildlife Service, Southeastern Fish Cultural Laboratory
Marion, Alabama 36756, USA


Ecological Research Center, Department of Biology
Memphis State University


'Present address: U.S. Department of Agriculture, Agricultural Research Service, Catfish Genetics Research Unit, Post Office Box 38, Stoneville, Mississippi 38776, USA.

2 Present address: U.S. Fish and Wildlife Service Marquette Biological Station, Marquette, Michigan 49855, USA.

3 Present address: Texas Cooperative Fish and Wildlife Research Unit, Department of Range and Wildlife Sciences, Texas Tech University, Lubbock, Texas 79409, USA.

4 Reference to trade names or manufacturers does not imply endorsement of commercial products by the U.S. Government.


Abstract.—Families of channel catfish Ictalurus punctatus of the 1982 and 1983 year classes were maintained in separate ponds. Fish from 4 to 26 months of age were evaluated for weight, length, and sex during seven sampling periods. Sex did not influence the growth of fish weighing less than 50 g at 10 months of age. Subsequently, sex had an increasing influence with increasing age: by 26 months, males were 37% heavier and 10% longer than females. Fish whose sex was reversed from male to female with steroid hormones grew like normal females through the time of sexual maturity. Growth and reproductive characteristics were typical of the phenotypic sex rather than of the genotype. Monosex populations of male channel catfish could increase production in commercial aquaculture operations.

The production of channel catfish Ictalurus punctatus by commercial fish farmers continues to be high due to the application of intensive management techniques. These techniques seemingly have been exploited fully to achieve current production capacities. Further substantial increases in production may require changes in the fish used. Stock or strain selections for improved heritable characteristics in channel catfish may increase production (Bondari 1983, 1986; Dunham and Smitherman 1983), but genetic controls of components such as weight, length, and percentage of edible flesh are not understood completely. Increased growth by one sex compared to the other has been reported for several species, including channel catfish in which males have been reported to grow faster than females (Beaver et al. 1966; El-Ibiary et al. 1976; Brooks et al. 1982; Gjedrem 1983; Dunham et al. 1985). Determination of the age and size at which sex influences growth, and the extent of this influence, could have important management implications. In previous studies with channel catfish, relatively small numbers of fish were used for the size-groups evaluated, growth patterns were not followed over time, and family effects were not evaluated.

Monosex populations have been produced by placing sex steroids in the diet or in the water during the phenocritical period of development of sexually undifferentiated fry (Yamamoto 1969; Hunter and Donaldson 1983). The advantages of sex manipulation in economically important species have been discussed in relation to the control of unwanted reproduction, enhancement of reproductive potential, avoidance of precocious sexual maturation, and production of fish of only the faster-growing sex (Schreck 1974; Donaldson and Hunter 1982; Hunter and Donaldson 1983; Yamazaki 1983). Breeding studies of several species have indicated that some sex-reversed individuals are reproductively competent (Yamamoto 1969; Johnstone et al. 1979a; Hunter et al. 1982; Calhoun and Shelton 1983; Hunter et al. 1983; Johnstone and Youngson 1984): however, little information is available on the comparative growth and reproductive characteristics of hormone-treated and control groups. Thus, analyses of sex-reversed populations may yield information on the relative importance of the effect of genotypic or phenotypic sex on these characteristics.

The objectives of our study were to determine the influence of sex on growth in channel catfish, the age or size when differences related to sex are expressed, and the variability of the influence of sex on growth among families. Additionally, we examined all-female populations produced by sex steroid administration to determine the influence of male and female genotypes on the growth of fish with female phenotypes.



Culture and growth of families.—Spawning conditions, parental and fry characteristics, fry handling procedures, and culture conditions for fish used in this study were reported by Klar et al. (1988). Progeny from nine full-sib families of channel catfish were stocked in June at approximately 6 weeks of age into separate, unreplicated, 0.04-hectare ponds at rates of 25,000 (1982 year class) and 12,500 (1983 year class) fish per hectare. After one growing season, fish of the 1982 year class were restocked in June 1983 at 7,500 fish per hectare; three families were stocked in replicate ponds, and the other six families each were stocked into separate, unreplicated ponds. From September 1983 through July 1984, we sampled 4-, 10-, and 14-month-old fish of the 1983 year class and 13-, 16-, 22- and 26-month-old fish of the 1982 year class. Weight, standard length, condition factor (k = 102[weight, g]/[length, cm]3), and sex were determined for 50-100 fish from each pond. Sex was determined by dissection (Grizzle and Rogers 1976). Sample weights were used to estimate growth and to adjust feeding levels. Fish were fed twice a day with appropriate-sized pellets of a nutritionally complete floating catfish food (Ralston Purina Company4) at a rate of 2-3% of estimated biomass during May-October and ad libitum at other times.

Differences between mean weights of males and females within each family were evaluated by Student's t-test. A factorial analysis of variance was used to evaluate effects of sex and families on growth. Regression analysis was used to compare growth for all families.

Sex-reversed populations.—Two all-female populations with genetic females and sex-reversed genetic males were produced. Sex reversal in one group was by administration of the androgen ethisterone (17 a -ethynyltestosterone) to induce paradoxical feminization, as reported by Goudie et al. (1983). Genetic males in the second group were feminized with the estrogen estradiol. Control fish and the all-female groups were maintained in separate tanks at 21 + 2C from June 1980 to November 1982. Weight, standard length, gonad weight, and sex were determined for 10 fish from each group. Fish remaining from the control group (N = 93), and the androgen- and estrogen-treated groups (A:= 103 and 182, respectively) were stocked in November 1982 into separate 0.2-hectare ponds until the fish reached sexual maturity. At the beginning of the breeding season in May 1984, when the fish were 4 years old, 25 control females and 15 females from each of the hormone-treated groups were killed; body weight, length, and gonad weight were measured and sex was confirmed. Heparinized syringes were used to collect blood samples from vessels in the caudal peduncle of 50 additional fish in each group.

Males from the control group were mated with females from the estrogen- and the androgen-treated groups. Thirteen of the hormone-treated fish spawned; seven of these were identified by sex ratios of the progeny as sex-reversed (i.e., genetic males sex-reversed to phenotypic females). Body weight and spawn weight were measured, and blood samples were collected from these fish within 24 h after spawning for comparisons between normal females and sex-reversed females.

Sex-steroid analyses were conducted on unextracted plasma samples with commercial radioimmunoassay kits (validated for use on channel catfish plasma by percent recovery and accuracy determinations) for testosterone (Cambridge Medical Diagnostics) and estradiol (Ciba-Corning Diagnostics Corporation). The gonadosomatic (gonad weight/body weight) and spawn (spawn weight/body weight) indices were expressed as percentages and transformed into arcsines for analysis. Body weight, the two indices, and plasma concentrations of sex steroids of hormone-treated fish were compared to those of control fish by one-way analysis of variance, and differences among means were detected by Tukey's honestly significant difference procedure if treatment effects were significant. A probability level of 0.05 was considered significant for all tests.


TABLE 1.—Weights, standard lengths, and condition factors of channel catfish grown in ponds. Significant differences between sexes (P 0.05) are indicated by an asterisk (*). Number of fish of each sex examined as each age varied from 216 to 494.

Weight, g
1983 4 15.2 5.4 15.2 5.1 0.1
  10 42.9 16.2 41.3 16.2 3.8
  14 143.6 76.5 130.5 64.9 10.0*
1982 13 64.3 29.2 58.6 25.9 9.8*
  16 284.4 122.7 256.1 102.9 11.2*
  22 449.3 194.3 386.3 167.4 16.3*
  26 700.9 290.9 510.7 209.3 37.3*
Length, cm
1983 4 9.8 1.3 9.9 1.2 -0.9
  10 13.8 1.7 13.7 1.7 0.7
  14 20.5 3.1 20.0 2.8 2.7*
1982 13 16.0 2.3 15.5 2.3 3.0*
  16 26.0 3.8 25.3 3.6 2.8*
  22 30.3 4.4 28.9 4.2 4.8*
  26 35.4 5.2 32.2 4.7 9.7*
Condition factor, K = 102 (weight/length3)
1983 4 1.53 0.10 1.51 0.12 1.3
  10 1.57 0.14 1.54 0.11 1.9*
  14 1.54 0.11 1.53 0.11 0.7*
1982 13 1.47 0.10 1.47 0.13 0.0
  16 1.52 0.11 1.51 0.09 0.7
  22 1.52 0.18 1.50 0.10 1.3
  26 1.47 0.12 1.43 0.10 2.8*

a Difference = 100(male weight - female weight)/female weight.



Evaluation of Families

Length and weight of males and females did not differ significantly among fish less than about 50 g in weight or 14 cm long at 10 months of age (Table 1). However, males of both year classes were significantly larger than females at all ages greater than 1 year. Sex had an increasing influence on body weight and length with increasing age; males outweighed females by 37% and were 10% longer at 26 months of age. Differences due to sex were independent of pond effects, as indicated by the lack of significant interactions. However, growth rates of different families were highly variable, and the effects of sex were not always apparent. Regression analyses indicated similar growth rates for males and females of the 1983 year class (b = 0.400 and 0.364, respectively; b = regression coefficient) despite significant differences in weight at 14 months of age. However, growth rates were significantly different between males (b = 1.465) and females (b = 1.072) for the 1982 year class. Condition factors generally were similar for males and females; however, the values for males were significantly higher on three occasions.


fig1.gif (35309 bytes)

FIGURE 1.—Weights of channel catfish families of the 1983 year class maintained in nine separate ponds and sampled at three ages: 4, 10, and 14 months. Values represent means for 50-100 fish per family; vertical lines represent SD. Families in which significant differences (P 0.05) in weight occurred between the males and females are indicated by an asterisk.


Comparisons of males and females within individual families demonstrated that size differences between the sexes were variable, and not all families exhibited growth differences between sexes. Such differences were significant in only one of the nine families in the 10- and 14-month samples of the 1983 fish (Figure 1). The number of families with males larger than females in the 1982 year class increased with increasing age of fish (Figure 2). However, even at 26 months, males and females in two families did not differ significantly in size.

fig2.gif (65706 bytes)

FIGURE 2.—Weights of channel catfish families of the 1982 year class maintained in up to 12 separate ponds end sampled at four ages: 13, 16, 22, and 26 months. Values represent means for 50-100 fish per family; vertical lines represent SD. Families in which significant differences (P 0.05) in weight occurred between the males and females are indicated by an asterisk. After 13 months, fish from families 1, 4, and 6 were stocked into ponds 1 and 5, 3 and 4, and 6 and 9, respectively. No data are presented for pond 8 at 22 months and ponds 3 and 8 at 26 months due to fish mortality.


Evaluation of Sex-Reversed Populations

Body weights of estrogen- and androgen-treated females at 2.5 years were similar to those of female control fish, and all three female groups were significantly lighter in weight than male controls (Table 2). The mean gonadosomatic indexes of hormone-treated females were significantly higher than that of control females, and the index was lower in control males than in females of all groups. These patterns were still evident at 4 years of age.

Estrogen levels were similar in control and estrogen-treated females, but were significantly lower in androgen-treated females (Figure 3). Testosterone was significantly higher in only the estrogen-treated females. Estrogen, testosterone, and spawn weight did not differ between normal females and sex-reversed females (Table 3).


table2.gif (15459 bytes)

TABLE 2.—Body weight and gonadosomatic index (GSI; gonad weight/body weight) of hormone-treated (sex-reversed) and control channel catfish aged 2.5 and 4.0 years. For each age, means within a column with a letter in common are not significantly different (P 0.05).



fig3.gif (15299 bytes)

FIGURE 3.—Levels of sex steroids (ng/mL) in 4-year-old control and estrogen- and androgen-treated (sex-reversed) channel catfish sampled before spawning (May 15, 1984). Values represent means; vertical lines represent SD. Numbers of fish are in parentheses above treatments. Treatment groups with different letters for estradiol or testosterone are significantly different (P 0.05).





Female (XX) 6 23.5 11.3 1.24 0.82 0.28 0.24
Male (XY) 7 33.6 12.1 1.29 1.10 0.31 0.24

TABLE 3.—Spawn index (spawn weight/body weight) and plasma concentrations of sex steroids within 24 h after spawning of 4-year-old phenotypically female channel catfish with male or female genotypes. Monosex female populations were produced by early hormonal treatment (Goudie et al. 1983).




Sex had an increasing influence on growth with increasing size and age in channel catfish. After the first growing season, the size of fish was typical for that age, and sex had no apparent effect on growth, although there was a tendency for males to be heavier than females. Probabilities (from analysis of variance) that differences in weight related to sex were due to chance alone decreased from 78% at 4 months to 10% at 10 months. Brooks et al. (1982) reported similar findings in fish 6 months of age: 7 of the 10 largest channel catfish sampled in their study were males. Size grading of fish after the first growing season could result in unequal sex distributions and unequal mean growth among production groups. Differential growth during the second growing season resulted in males that were 37% heavier than females. Size-selective harvesting of food fish could remove a disproportionate number of males and could result in predominantly female populations with decreased subsequent growth to harvest.

The influence of sex was usually evident despite differences in growth by families in different ponds. However, the growth by fish within each sex was variable, and the sizes of males and females in a pond at a particular time overlapped considerably. The absence of any significant statistical interactions between sex and family indicated a consistent influence of sex on growth in the different families tested despite differences in growth that might have been due to pond effects.

The similarity of condition factors for males and females for the sizes examined indicated little difference in relative plumpness due to sex. The occasional significant differences we did observe represented a maximum difference of only 2.8%. The large sample sizes used enabled us to detect these small differences; however, these differences may not have practical application. The percent of edible portion in food-size channel catfish is similar for each sex because males typically have larger heads, but females have a larger percentage of viscera (El-Ibiary and Joyce 1978; Dunham et A. 1985).

Growth among families differed significantly in our study, as was reported by Klar et al. (1988). Selection of the largest fish (in the 90th percentile) for brood stock may not only select for males, and possibly result in a shortage of females (Brooks et al. 1982), but also may limit the number of families contributing to the genetic diversity of brood stock. Therefore, brood stock must be selected by sex, not just by size, and should include different genetic lines.

Successful hormone treatment to direct sexual development results in the production of phenotypically monosex male or female populations in which 50% of the fish are genetically opposite to their expressed phenotypic sex. Relatively few data are available on the influence of genotypic and phenotypic sex on the growth and reproductive performance of sex-reversed fish compared to untreated groups.

Although genetic control of growth and sex may be independent, in our study the expression of growth and reproductive characteristics was typical of the phenotypic sex, not the genotype. The weight, length, and gonadosomatic index of 2.5year-old females treated with hormones as fry were similar to those of control females, which was a pattern that persisted through the first reproductive season. Johnstone et al. (1979b) also found similar growth to maturation between sex-reversed and normal female brook trout Salvelinus fontinalis. Groups of sex-reversed chinook salmon Oncorhynchus tshawytscha had similar growth rates at 9 months and 2, 3, and 4 years of age, but the percent of mature fish at each age was affected by the early steroid treatments (Hunter et al. 1983). Growth rates of androgen-treated, sex-reversed, male (genetically female) fry of Tilapia aurea and T. nilotica, and manually selected male control groups were evaluated by Anderson and Smitherman (1978). Hormone-treated groups were similar, but grew significantly slower than the male control group, presumably due to the presence of the female genotype. On the other hand, Hansen et al. (1983) compared growth and food conversion of manually selected populations of male T. nilotica, male hybrids of T. nilotica x T. hornorum and T. hornorum x T. mossambica, sex-reversed male populations of T. nilotica, and untreated T. nilotica females. Sex-reversed male groups had higher weight gain and food conversion than the hand-selected male groups, and all the male groups grew faster than the female group. Because males of these species grow faster than females, it is difficult to explain the better performance of the sex-reversed groups, which contained 50% genetic females, over selected male groups. A residual anabolic effect of the early steroid treatment cannot be ruled out.

Sex-steroid profiles of channel catfish may reflect a residual effect of early hormone treatments that influenced metabolism of the hormones. The lower estradiol levels of Androgen-treated channel catfish when compared with those of estrogen-treated and control females might be expected, but the significantly elevated testosterone levels in estrogen-treated fish have no apparent explanation. The gonadal organization and metabolic capacity of steroid-producing cells need further study.

The reproductive performance of female channel catfish was not altered by hormone treatments to induce sex reversal. Sex-reversed females (genotypic males) spawned at the same rate, produced similar quantities of eggs and viable fry, and had physiological steroid levels that were statistically similar to those in known genetic females. Several other authors have demonstrated the successful reproduction of some sex-reversed fish (Yamamoto 1969;Johnstoneetal. 1979a;Hunter et al. 1982; Calhoun and Shelton 1983; Hunter et al. 1983; Johnstone and Youngson 1984), but they did not compare the reproductive characteristics of sex-reversed and normal fish.

Although the hormone-treated (sex-reversed) female populations in our study displayed growth and reproductive traits similar to those of the female genotype, it is not known whether sex-reversed and normal males also would be similar. If so, the significantly larger size of males in food-size channel catfish suggests that monosex culture of male populations could be advantageous. The use of monosex populations also could result in fish of more uniform size, which would be beneficial to producers and processors to meet specific production requirements.



We thank D. Campton, University of Florida, Gainesville, Florida; G. Carmichael, U.S. Department of Agriculture, Stoneville, Mississippi; and T. Tiersch, Memphis State University, Memphis, Tennessee, for reviewing the manuscript. This project was supported in part by U.S. Department of Agriculture grant 88-34123-3504.




Received October 17, 1988 -