The Progressive Fish-Culturist 50:173-178 1988
Comparison of Growth among Families of Channel Catfish
GERALD T. KLAR, NICK C. PARKER, AND
CHERYL A. GOUDIE
U.S. Fish and Wildlife Service
Southeastern Fish Cultural Laboratory
Route 3, Box 86
Marion, Alabama 36756, USA
Abstract.The progeny produced from 17 spawns of channel catfish (Ictalurus punctatus) in two year classes (1982 and 1983) were evaluated for growth in tanks (constant temperature, 20 ± 1°C) and ponds (ambient temperature). Significant differences in weight and length were observed among families of both year classes after 14-22 months of culture. No consistent pattern of growth was evident between siblings reared in ponds and those reared in tanks; siblings reared in tanks did not follow the same rank order for fish size as that followed by siblings reared in ponds. Performance was affected more by environmental than by genetic factors.
Differences have been reported among families of channel catfish (Ictalurus punctatus) in such characteristics as growth and survival (Bondari 1982, 1983a, 1984b; Durborow et al. 1985), effects of selection (Bondari 1983b,1983c; Dunham and Smitherman 1983), and inbreeding (Bondari 1981). These investigations indicated a potential for increasing yields of channel catfish by genetic manipulation and selection. Catfish farmers commonly attempt to improve their stock by selecting the largest fish from a year class for use as brood fish. However, farmers who raise 100,000 fish/ year may use as few as 8-10 female brood fish for their annual production even though it is known that this small number of females may result in decreased survival, growth rate, and food conversion efficiencyparticularly if they are siblings or from an inbred lot (Kincaid 1976, 1983). In addition to the effect on production, potential for selection is limited by reduced genetic variability.
The objective of this study was to compare the size of channel catfish progeny from the same brood in tanks and ponds, and to extend the comparison over several families, to demonstrate the relative influences of genetic and environmental factors on growth.
TABLE 1.Parent weight, spawn weight, and number of fry for different families of channel catfish in 1982 and 1983.
Adult male and female channel catfish (2 years old in 1982 and 3 years old in 1983) were paired in wire cages (2 ´ 2 ´ 2 m) in a 0.04-hectare pond and allowed to spawn naturally in metal containers (Table 1). Females from two families (83-3 and 83-7) that failed to spawn naturally were induced to spawn by the injection of 1,760 international units of human chorionic gonadotropin per kilogram of body weight. Spawns were collected daily, weighed, and placed in wire baskets suspended in an aluminum hatching trough. Rotating paddles provided continuous water circulation during incubation, and pond water (ambient temperature) was delivered to each incubator basket at a flow rate of 2 L/min. Spawns were treated daily with a prophylactic solution of 0.8% iodine to inhibit fungal development. Immediately after hatching, the number of fry in each spawn was estimated by water displacement and the fry were transferred to individual aluminum rearing troughs receiving aerated pond water. Fry reared in troughs for 6-10 weeks were fed fry starter diet (Ralston Purina Company, St. Louis, Missouri) ad libitum three to four times per day. They were then divided into two groups to be stocked into tanks or ponds. At the time of stocking, 20 fish were sampled from each family for determination of length, weight, and condition factor (100 [weight, g]/[length, cm]3). Each family was maintained and reared separately.
Tank culture.Fifty catfish fry from each family spawned in 1982 and 1983 were stocked into each of two 100-L circular tanks supplied with aerated well water (20 ± 1°C) at a flow rate of 3 L/min. Fish of the 1982 year class were fed a Purina diet at 3% of body weight per day for 308 d and 2% of body weight per day for an additional 224 d. Fish of the 1983 year class were fed at 4% of body weight per day for 406 d. Total fish weight in each tank was measured at 14-d intervals and feeding levels and feed pellet size were readjusted for each tank of fish.
On the final day of tank culture for both year classes, length, weight, and condition factors were determined for all remaining fish from each family. At that time, fish of the 1982 year class were 20 months old and those of the 1983 year class were 16 months old.
Pond culture.Siblings from each family stocked into tanks (two replicates) in 1982 and 1983 were also stocked into 0.04-hectare ponds (without replication). Fish of the 1982 and 1983 year classes were initially stocked at densities of 25,000 and 12,500 fish/hectare, respectively. After one growing season, the 1982 fish were randomly selected and restocked at 7,500 fish/hectare. Weights of 50-100 fish/pond, sampled about every 3 months, were used to adjust feeding levels. Feeding levels were also adjusted weekly based on an estimated food conversion of 1.5. The appropriate pellet size of a nutritionally complete catfish diet (Ralston Purina Company) was provided in two daily portions at 2-3% of the estimated body weight during warm weather. As water temperature decreased, fish were fed ad libitum. When channel catfish of the 1982 year class were 22 months old, and those of the 1983 year class were 14 months old, 50-100 fish from each family were measured and weighed. Ranking of families (by mean fish size) in ponds was compared with ranking of siblings reared in tanks within each year class.
Data analysis.Differences in the final size of fish in each family in tanks and ponds were evaluated by single-classification analysis of variance. Significant differences among family means were identified by Tukey's highly significant difference (HSD) test for multiple comparisons. Spearman's rank order was used to compare family ranking between tanks and ponds. Statistical comparisons were made only among families within each year class. Differences were considered significant at £ 0.05.
TABLE 2.Mean (SD) size and condition factor of channel catfish families at time of stocking into 100-L tanks and 0.04-hectare ponds (20 fish were sampled from each family).
At the time of initial stocking, mean weight of fry was positively correlated with the age of fry and weight of the parent female for both year classes (Tables 1, 2). Fry from different families and differed in age by as much as 12 d in 1982 and 26 d in 1983. Condition factor at the time of stocking was similar among families in 1982, but not in 1983. Fish from the families with relatively low average condition factors were among the youngest fry. Significant differences in mean weight were observed among families of the 1982 and 1983 year classes reared in tanks (Figures 1, 2). Significant differences were also observed among those same families reared in ponds. No correlation in ranking by mean weight was evident between families; reared in tanks and families reared in ponds with in either year class. Mean weight for 1983 year-class fish reared in tanks for 16 months and in ponds for 14 months was significantly correlated with age and female weight, but individual families performed differently in tanks and ponds with no correlation in ranking. The condition factors of 1982 year-class families reared in tanks were consistently lower than those of corresponding families reared in ponds (Table 3). Condition factors were also lower in six of nine families of the 1983 year crass reared in tanks compared to those reared in ponds.
TABLE 3.Mean (SD) condition factors of channel catfish families in the 1982 and 1983 year classes reared in 100-L tanks and 0.04-hectare ponds. Fish of the 1982 year class were sampled at 20 months of age in tanks and 22 months of age in ponds. Fish of the 1983 year class were sampled at 16 months of age in tanks and at 14 months of age in ponds.
FIGURE 1.Comparison of mean weight (vertical bars) + SD (vertical line above bars) of channel catfish families after the 1982 year class had been reared in 100-L tanks for 20 months or 0.04-hectare ponds for 22 months. Number of fish sampled was 50 from each family in ponds and 75-99 from each family in tanks. Values with the same letters within each culture condition were not significantly different; fish in tanks were not statistically compared to fish in ponds. Fish of family 82-3 did not survive for 22 months in the pond.
The channel catfish families used in this study were taken from a typical lot of fish maintained at the Southeastern Fish Cultural Laboratory, Marion, Alabama. The fish may have been inbred to an unknown extent because a relatively small number of fish were retained for brood stock from year to year. Kincaid (1976, 1983), who reported the effects of inbreeding on populations of rainbow trout (Salmo gairdneri) and reviewed the effects of inbreeding on fish populations in general,
wrote that, as the level of heterozygosity decreases due to inbreeding, fish lose the ability to compensate for environmental differences. Thus, slightly different conditions that might be encountered from pond to pond or from tank to tank could elicit different responses, even within the same family. Bondari (1984a), who compared the growth of one inbred and one outbred channel catfish family, found that the inbred family performed better than the outbred family. Results from the present experiment suggest that environmental influences affected performance more than genetic factors. To fully describe the range of variation in response among families, one must compare several families of similar age over several months. Differences that are apparent during short-term experiments may be eliminated over the long term by environmental variables.
FIGURE 2.Comparison of mean weight (vertical bars) + SD (vertical line above bars) of channel catfish families after the 1983 year class had been reared in 100-L tanks for 16 months or 0.04-hectare ponds for 14 months. Number of fish sampled was 50 from each family in ponds and 61-97 for families in tanks. Values with the same letters within each rearing unit were not significantly different; fish in tanks were not statistically compared with fish in ponds.
Within the 1982 year class, fish of one family (82-1) were largest (in both length and weight) in tanks as well as in ponds. No other consistent or pronounced family effects on performance were apparent. In addition, no overall correlation was noted between family and performance for either year class. Within the 1983 year class, replicates of the same family in tanks performed differently. When McKay et al. (1984) compared the growth of families of rainbow trout at different temperatures, they found that families differed in their performance in two different growth periods and that they did not rank identically at two temperatures within the same growth period.
Fry weight for both year classes was correlated with age at the beginning of the experiment; selection of the fastest-growing family 6-10 weeks after hatching, if selection were based solely on size, would be precluded if the ages differed by more than 12 d. A difference in age of 26 d among 1983 year-class families led to weight differences at 14 and 16 months of age. This observation suggests that care should be taken in choosing families of different ages for use in comparison or selection programs. Additionally, a mixture of small numbers of families of different ages may cause large size variation among fish in commercial production.
A combination of selection between and within families for genetic improvement of body weight and length was suggested by Bondari (1983c) for channel catfish. Bondari (1983b) also suggested segregating fingerlings into small, medium, and large sizes and using only the medium- and largesized fish in selective breeding. Kincaid (1983) reported that inbreeding depression can negate the effects of selection for desirable traits, such as growth. When only a few fish are used in a selection program, undesirable traits are combined and may offset any advantage of selection for desirable traits. Consequently, there would be no genetic advantage to selecting the largest fish of the fastest-growing family in an attempt to maximize growth of channel catfish. In addition, if fish were segregated according to size as fingerlings, a small number of fish from each of a large number of families should be used to produce the next generation. A selective breeding program involving mass selection and line crossing, outlined by Schom and Bailey (1986), may be appropriate for catfish farmers wishing to minimize inbreeding while maximizing genetic gain in production.
Based on results of this study and previously reported data in the literature, we recommend the following for channel catfish culture. (1) Select fingerlings for culture in tanks from a brood line that has previously performed well in tanks. (2) Select fingerlings for culture in ponds from a brood line that has previously performed well in ponds. (3) Fry differing in age by more than a few days (12 d in this study) should not be mixed and reared together, so that size variation within a population is minimized. (4) A small number of fish from each of a large number of families should be used as brood stock to maintain genetic diversity, or (5) inbred lines should be maintained and then crossed to maximize genetic gain of selective breeding.
We thank Rex Dunham, R. Oneal Smitherman, and Douglas Tave of Auburn University, Auburn, Alabama, and Bill Simco, Memphis State University, Memphis, Tennessee, for their critical reviews of an early draft of this manuscript.
Bondari, K. 1982. Cage performance and quality comparisons of tilapia and divergently selected channel catfish. Proceedings of the Annual Conference Southeastern Association of Fish and Wildlife Agencies 34:88-98.
Bondari, K. 1983a. Caudal fin abnormality and growth and survival of channel catfish. Growth 47:361-370.
Bondari, K. 1983b. Genetic and environmental control of fingerling size in channel catfish. Aquaculture 34:171-176.
Bondari, K. 1983c. Response to bidirectional selection for body weight in channel catfish. Aquaculture 33: 73-81.
Bondari, K 1984a. Growth comparison of inbred and random bred catfish at different temperatures. Proceedings of the Annual Conference Southeastern Association of Fish and Wildlife Agencies 35:547-553.
Bondari, K. 1984b. Performance of albino and normal channel catfish (Ictalurus punctatus) in different water temperatures. Fisheries Management 15: 131-140.
Dunham, R. A., and R. O. Smitherman. 1983. Response to selection and realized heritability for body weight in three strains of channel catfish, Ictalurus punctatus, grown in earthen ponds. Aquaculture 33: 89-96.
Durborow, R. M., J. W. Avault, Jr., W. A. Johnson, and K. L. Koonce. 1985. Differences in mortality among full-sib channel catfish families at low dissolved oxygen. Progressive Fish-Culturist 47:14-20.
Kincaid, H.L. 1976. Effects of inbreeding on rainbow trout populations. Transactions of the American Fisheries Society 105:273-280.
Kincaid, H. L. 1983. Inbreeding in fish populations used for aquaculture. Aquaculture 33:215-227.
McKay, L. R., G. W. Friars, and P. E. Ihssen. 1984. Genotype ´ temperature interactions for growth of rainbow trout. Aquaculture 41:131-140.
Schom, C. B. and J. K. Bailey. 1986. Selective breeding and line crosses to reduce inbreeding. Progressive Fish-Culturist 48:57-60.