Experiments are being conducted to determine the genetic risk of transgenic channel catfish, Ictalurus punctatus, which possess the rainbow trout growth hormone gene to wild populations of channel catfish. If transgenic channel catfish containing the rainbow trout growth hormone gene were to be accidentally released into the natural environment from either a research or commercial aquaculture facility, several possible scenarios exist. These transgenic fish may be eliminated by predation or starvation before they reproduce. They may not be able to reproduce or spawn effectively. They may reproduce but catfish possessing the rainbow trout growth hormone gene (or other growth hormone genes) be eliminated by random genetic drift because of their low population numbers. These transgenic fish could spawn and the foreign gene be established at low frequency based upon initial gene frequencies resulting from the surviving and reproducing escaped fish. The transgenic fish could segregate, spawn with themselves and form a reproductively isolated, sympatric population. Lastly, the transgenic catfish could increase in numbers or proportion compared to non-transgenic catfish.

The last example has the greatest potential for genetic impact, be that positive or negative, on natural populations. The last scenario, increase in transgenic fish numbers in the wild, will only occur if the transgenic fish possess traits that allow them greater predator avoidance (therefore, increased ability to survive to sexual maturity), greater ability to forage on natural food, or a greater ability to spawn and produce young. These three traits are the key to fitness, the ability to transmit ones genes to the next generation.

We have determined previously that transgenic channel catfish or common carp, Cyprinus carpio, produce salmonid growth hormone have increased growth rate, increased feed conversion efficiency, altered survival, altered body composition, and altered body shape under aquaculture conditions. However, fitness traits of transgenic channel catfish have not been measured.

Key words: Transgenic, channel catfish, growth hormone, fitness, predation

Salmonid growth hormone genes, RSVLTR-rtGH₁ cDNA, RSVLTR-rtGH₂ cDNA and RSVLTR-csGH cDNA, have been transferred to channel catfish, Ictalurus punctatus. One-nine copies of the foreign DNA were inserted in either head-to-tail tandem array at single insertion sites or single copies at multiple insertion sites. All P₁ transgenic catfish evaluated produced salmonid growth hormone regardless of the construct. Five P₁ X P₁ matings have been accomplished. The spawning rate and fertility of these P₁ transgenics in artificial spawning conditions was comparable to that of normal channel catfish. In two of three years, 100% spawning and 100% hatch were obtained. Percent transgenic progeny observed in the five matings were 20, 52, 7, 47 and 0% which was lower than the 75% inheritance expected assuming the P₁ brood stock had at least one copy of the foreign gene integrated and were not mosaics in the germ-line. At least seven of ten P₁ were mosaics and a minimum of 2 of 10 P₁ did not possess the salmonid growth hormone gene in their germ-line.

P₁ transgenics grew at the same rate as their non-transgenic full-siblings which is not surprising since the P₁ were mosaics. F₁ transgenic progeny in two families possessing RSVLTR-csGH cDNA grew 26% faster to 40-50 g than their non-transgenic full-siblings when evaluated communally. One F₁ progeny group produced by a RSVLTR-rtGH₁ cDNA X RSVLTR-csGH cDNA mating and one F₁ progeny group (parents either RSVLTR-rtGH₁ cDNA or RSVLTR-csGH cDNA) grew at the same rate as normal full-siblings when grown communally to 25 g. In families where F₁ progeny grew faster than controls, the range in body weight and coefficient of variation for the transgenic full-siblings was less than that for controls. In families where F₁ progeny grew at the same rate as controls, range in body weight and coefficient of variation were similar for transgenic and normal individuals. The percent deformities observed in P₁ transgenics, 13.6%, was higher than in microinjected P₁ non-transgenics, 5.1%. Percent deformities in transgenic and control F₁ channel catfish was not different, 0.5 and 2.8%, respectively.

The RSV-LTR (Rous sarcoma virus-long terminal repeat) promoter, a constitutive promoter used in these experiments, appears to be effective in fish. Channel catfish and common carp express salmonid growth hormone genes driven by RSV-LTR. Northern pike, Esox lucius, and goldfish, Carassius auratus, also express recombinant genes driven by the RSV-LTR promoter. The mouse metallothionean promoter does not appear to consistently drive expression in fish (McEvoy et al., 1988; Rokkones et al., 1989). The SV-40 promoter usually does not allow expression (Guyomard et al., 1988; Stuart et al., 1988) with the lone exception being the successful expression of catalase in transgenic tilapia (Indiq and Moav, 1988).

Results we have obtained for F₁ transgenic common carp (Zhang et al., 1990; Chen et al., 1993) and channel catfish (Hayat et al., 1991; Dunham et al., 1992) containing salmonid trout growth hormone gene were similar. The presence or absence of increased growth can vary among families, and may be related to family effects, genetic background, epistasis or dosage effects of the foreign growth hormone gene expression. Apparently, a combination of both family selection as well as gene transfer is needed to optimize growth increase from the insertion of salmonid growth hormone genes.

Fish culture in the United States and the world is increasing rapidly. Purposeful or accidental (such as flooding or escape during harvest) introduction of the domestic or transgenic fish may then allow mixing of the domestic (or transgenic) and natural populations.

The population genetics of such mixing of domestic and wild populations of fish and channel catfish is not well understood. If the domesticated or transgenic fish do not survive or reproduce, no effect on gene frequencies of the wild population will be observed.

Several factors might affect the establishment of domestic populations by accidental or intentional release and their opportunity or ability to interact with and influence wild populations. These include size of fish, number of fish stocked, number of repeat stockings or releases, timing of stocking or release, selective value of the new genotype and other environmental variables. These variables have not been completely evaluated, but a growing base of data illustrates their importance and function (Pycha and King, 1967; Plosila, 1977; Kulzer et al., 1985; Philipp et al., 1985; Norgren et al., 1986; Isley, 1987; Maceina et al., 1988; Smitherman et al., 1989a;b; Dunham et al., 1992). These studies indicate that it is difficult to establish a new genotype, even wild rather than domesticated one, in an established natural population (Fraser, 1972; Norgren et al., 1986; Smitherman et al., 1989a;b).

One example illustrating this point is that of the massive stockings of Florida largemouth bass, Micropterus salmoides floridanus, into established northern or native largemouth bass, M. salmoides salmoides, populations (Kulzer et al., 1985; Norgren et al., 1986; Dunham et al., 1992). Genes from Florida largemouth bass were established in these populations at varying levels, and in some cases not all. Key factors in the establishment of Florida alleles were total numbers of fish stocked, number of years since initial stocking, number of repeat stockings, elevation, age of the lake and water clarity.

Size of fingerling or sub-adult fish stocked has an effect that is not well defined. Kulzer et al. (1985) did not find a correlation between size of fingerling or sub-adult largemouth bass stocked and the success of the introduction in largemouth bass. However, the results were confounded by lack of replication and a multitude of additional variables. Studies on the success of stocking large and small sub-adult trout has had contradictory results (Anderson, 1962; Buettner, 1962; Pycha and King, 1967; Plosila, 1977; MacLean et al., 1981). The large sub-adults had greater survival when introduced in one study and lower survival in another. The main conclusion of these studies is that it is difficult to genetically impact established natural populations of fish.

Another possible interaction between domestic and wild populations of fish is the establishment of sympatric, but reproductively isolated populations. Although strains of fish usually do not have reproductive isolating mechanisms preventing them from interbreeding, occasionally behavioral mating blocks prevent or decrease the rate of inter-strain matings. Marion channel catfish females preferentially mated with their own strain rather than Kansas males (Smitherman et al., 1984), and Ghana strain of Oreochromis niloticus was more likely to mate with its own strain than other strains (Smitherman et al., 1988). The existence of reproductively isolated, sympatric populations of trout (Brown et al., 1981; Ryman and Stahl, 1981; Lerder et al., 1984), especially brown trout, Salmo trutta, is well-documented. Some strains of domestic and wild rainbow trout are sympatric, but reproductively or near to reproductive isolation. This occurs because of behavioral differences including temporal or spatial differences in spawning (Smitherman et al., 1988).

In addition to other factors, size often plays an important role in reaching sexual maturity and reproductive success. Often the fastest growing individuals within a population will reach sexual maturity the earliest (Gall, 1986; Dunham, 1990), although early sexual maturity does not necessarily correspond to an increased growth rate alone (Childers, 1967; Dunham and Smitherman, 1984; Gall, 1986; Dunham and Smitherman, 1987; Dunham, 1990). Hulata et al. (1985) reported a negative association between growth rate and onset of sexual maturity in Cyprinus carpio, and this same relationship between growth rate and sexual maturity has been documented in channel catfish (Dunham and Smitherman, 1987; Dunham, 1990).

Egg production also is related to size in channel catfish, and it is likely that transgenic channel catfish may produce more eggs than a normal channel catfish of the same age. However, we have observed that the maximum relative size difference between the transgenic and non-transgenic carp occurs by the time they reach 200 g. When these fish are three years old and have reached a size of 2 kg, the relative size difference is the same or slightly less, and preliminary evidence in catfish indicates that relative size differences of transgenic and control fish occur early, and then remain unchanged. This growth age relationship would help negate any size-related reproductive advantage of the transgenic catfish.

To gain a reproductive advantage in the natural environment, a specific genotype would need to be vastly superior to that of other genotypes. The ability to produce large numbers of offspring is not necessarily advantageous because of the tremendous mortality of young fish in the natural environment. In a typical Alabama reservoir, only 0.04% of the channel catfish larvae produced survive to be adults. Among the factors causing natural mortalities and affecting the survival of catfish during their early life history stages are predation and availability of food. Unless a large reproductive advantage existed, the ability to forage and avoid predators are likely of more importance than reproduction for fitness in the natural environment.

Predator avoidance probably has both behavioral and growth components. Aggressiveness or lack of wariness can lead to mortality due to predation by insects, fish, animals, angling or netting. Additionally, increased growth can be disadvantageous making a fish susceptible to selective harvest (Ricker, 1975; Hulata et al., 1985) by either animal predators or man. Conversely, rapid growth at certain life stages is probably important to reach a size that allows avoidance of small predators.

The ability to forage and find feed is, of course, critical to survival in natural settings. The ability of a genotype well adapted or selected for feeding in the culture environment is not necessarily well adapted for feeding in the natural environment since the genetic value or rank of a particular genotype may change as the environment changes. The best genotype for one environment may not be the best for another.

Genotype-environment interactions are prevalent in aquaculture experiments when different genotypes are compared utilizing either natural food organisms or artificial supplemental feeds (Wohlfarth et al., 1975; Wohlfarth et al., 1983; Khater, 1985; Dunham, 1992). High performance triploid oysters have no advantage compared to normal oysters when food is limiting (Dunham, 1992).

Environmental units. Environmental risk data for transgenic fish has not previously been attempted. The proposed experiments will be conducted in 0.04-ha earthen ponds that have been designed specifically for confinement of and approved by USDA for transgenic fish research. It is important to conduct such risk data in confined ponds to mimic as closely as possible the natural environment. This makes ponds a more appropriate test environment than aquaria or tanks since ponds can more closely represent the ecological complexity and diversity of the natural environment. This consideration is especially important because of the potential of genotype-environment interactions. The relative performance of genotypes can change with change in environment. Genotype-environment interactions are common in fish experiments. Experimental ponds will be fertilized, habitat, prey and predators introduced to simulate as closely as possible natural conditions.

Spawning ability and mating preference. Transgenic and non-transgenic channel catfish were grown to a mean weight of 1.5 kg for reproduction experiments. Ten pairs of transgenic and ten pairs of controls were stocked into three replicate ponds each in March prior to spawning season. Spawning habitat and twenty spawning containers were placed in each pond at this time. When the water temperature reached 20C the spawning containers were checked every 5 days for eggs. The coldest temperatures that channel catfish eggs have been detected previously was 21C. The spawning containers were checked until water temperatures reach 33C, the maximum spawning temperature for catfish.

When eggs were detected, a small sample was removed for analysis for rtGH genes and for staging the developmental age of the eggs. The transgenic and control brood fish were also artificially spawned in the laboratory to ensure that transgenic and control fry were available for the remaining experiments.

Foraging ability. Four ponds were prepared and fertilized to provide a natural food base and plankton levels similar to that of the natural environment. Two-thousand transgenic fry and 2,000 controls were stocked in each pond at swim-up, the stage of first feeding. At four months a sample of 100 of the surviving fish from each pond were individually weighed and fin clipped for DNA analysis. Mean body weight and survival of transgenic and control catfish were determined. Since food will be limiting, as is the case in natural systems, we expect the majority of the fish will not survive. The fish in the ponds will be resampled at 6 months and at 12 months, which correspond to periods after shifts in prey items. When the fish reach 12 months of age, a fish forage base will be established, and the fish sampled again at 18 months of age.

Transgenic and control fish were stocked similarly in four additional 0.04-ha ponds. The fish in these ponds were fed and cultured under normal aquaculture conditions with supplemental feed for comparison to the fish grown under the natural foraging system. Three-hundred fish/pond were weighed, fin clipped for DNA analysis, and heat branded for future identification at 4 months of age. These fish will be sampled again at 6, 12 and 18 months of age in parallel with fish grown on forage.

When these experiments are completed the foraging ability, survival and growth of transgenic channel catfish possessing the rainbow trout growth hormone gene will have been established. Comparisons between fish grown under natural and commercial conditions will have been made to determine any genotype-environment interactions for survival and growth in these two environments.

DNA analysis for rainbow trout cDNA. Genomic DNA was extracted from pectoral fin clips of presumptive transgenic fish following the methods of Maniatis with some modifications. Pectoral fin tissue was lysed in 4 ml of 10 mM Tris buffer (pH 8.0) containing 0.1 M EDTA (pH 7.0), 0.5% SDS and 200 g/ml proteinase K, and digested at 50C for 15 hrs. Samples were extracted with phenol, phenol/chloroform and chloroform/isoamyl alcohol. Then, DNase free Rnase A was added at a final concentration of 100 g/ml and samples will be incubated for 3 hrs at 37C. DNA was extracted again with phenol/chloroform and chlolroform, and dialysed against several changes of TE buffer (10 mM Tris-1 mM EDTA, pH 8.0). After ethanol precipitation, the DNA pellet was dissolved in TE buffer and concentration determined.

To detect the presence of rainbow trout cDNA in presumptive transgenic fish, the polymerase chain reaction (PCR) amplification/Southern blot hybridization method was used. In this method, two synthetic oligonucleotide fragments (20 mers) derived from the sequences flanking the RSVLTR and rtGH cDNA were used as amplification primers and the genomic DNA of the presumptive transgenic fish as a template for PCR amplification. The amplification conditions were: denaturation at 90C for 2 min. annealing at 50C for 2 min and synthesis at 72C for 2 min, and amplifying for 30 cycles. Following amplification, the products were resolved by electrophoresis on 1.0% agarose gels, transferred to nylon membranes and hybridized to radio-labelled rtGH cDNA.

Genomic DNA (18 g) of some individuals was denatured in 0.5 M NaOH at 37C for 10 min and spotted onto a nylon membrane using a commercial dot blot apparatus (BRL, Bethesda, MD). The membrane was soaked in 1.5 M NaCl, 0.5 M NaOH for 2 min, neutralized in 1.5 M NaCl, 0.5 M Tris-HCl (pH 7.2), 1 mM EDTA for 5 min. After air drying, the membrane was irradiated on a short wavelength UV transilluminator for 5 min.

The NdeI-HindIII and KpnI-HindIII restriction fragments isolated from the pRSVrtGHcDNA construct were labelled with [-³²P] dCTP either by nick translation or by random priming, then purified through a sephadex G-50 spin dialysis column and used as probes. Prehybridization was carried out in 20 mM Tris-HCl, pH 7.5, 0.1% SDS, 5X SSC, 10X Denhardt and 100 g/ml of denatured calf thymus DNA at 42C with constant shaking for 3 hrs. Hybridization was carried out in the same buffer containing denatured probe at 42C with constant shaking overnight. Membranes were washed twice in a solution containing 2X SSC and 0.1% SDS for 15 min at room temperature, and twice in a solution of 0.2X SSC, 0.1% SDS for 15 min at 65C.

For Southern blot analyses, DNA (18 g) was digested with HindIII or BamHI under conditions indicated by the supplier (Boehringer Manheim) and electrophoresed on a 0.8% agarose gel. DNA samples in the gel were transferred to nylon membrane by diffusion, and hybridized to ³²P labelled DNA probes prepared as above.

Copy number estimates were based on a haploid DNA content of 1.2 picograms for ictalurid catfish, using as standard the DNA content of one sea urchin sperm (0.98 X 10⁹ bp). The molecular weights of catfish genomic DNA and pRSV rt GH1cDNA were determined and used to estimate the amount of exogenous DNA (equivalent to one DNA molecule) present in 18 g of catfish genomic DNA. The total number of DNA copies were estimated from the optical density profiles of dot blot autoradiographs scanned under visible light.

Data analysis. Growth rate of transgenic and control fish will be compared with student's t test within each experiment. Genotype-environment interactions will be evaluated with analysis of variance. Spawning percentage, percent successful hatches, and survival in all experiments for transgenic and control fish will be compared with chi-square analysis.

Data is currently being analyzed. Definitive results and conclusions have not been reached.

Related research. Preliminary results indicate that transgenic channel catfish containing rainbow trout growth hormone gene are vulnerable to bacterial pathogens such as Edwardsiella ictaluri. No increases in reproductive ability have been observed.

Common carp model. Related parallel experiments on common carp containing rainbow trout growth hormone gene are being conducted. These fish function as an experimental model for the channel catfish research. Fecundity of F₁ transgenic and control common carp was not different. F₂ transgenic common carp were fewer than expected from F₁ matings. This indicates loss of the transgene during meiosis, or death of gametes, embryos or young transgenic fish. Once the transgenic individuals survived and reached 30 g, their survival and tolerance of low oxygen was superior to non-transgenic full-siblings. Feed conversion efficiency was lower for transgenic common carp compared to controls. Transgenic individuals had larger heads, thicker and deeper bodies than non-transgenic controls, however, no differences were found in dressout percentage and condition factor. Transgenic common carp had higher protein and less fat in their flesh compared to non-transgenic full-siblings.

This research was supported by USDA/CSRS 92-39210-8267.

Anderson RB (1962) A comparison of returns from fall and spring-stocked hatchery-reared lake trout in Maine. Trans. Am. Fish. Soc. 91:425-427.

Brown EH Jr., Eck GW, Foster NR, Horrall RM, and Coberly CE (1981) Historical evidence for discrete stocks of lake trout (Salvelinus namaycush) in Lake Michigan. J. Fish. Aquat. Sci. 38:1747-1758.

Buettner HJ (1962) Recoveries of tagged, hatchery-related lake trout from Lake Superior. Trans. Am. Fish. Soc. 90:404-412.

Chen TT, Powers DA, Lin CM, Hayat M, Chatakondi N, Ramboux A, Duncan PL, and Dunham RA (1993) Expression and inheritance of pRSVrtGH₁cDNA in common carp, Cyprinus carpio. Mol Mar Biol and Biotech. 2:88:95.

Childers WF (1967) Hybridization of four species of sunfish (Centrarchidae). Ill. Nat. Hist. Surv. Bull. 29(3).

Dunham RA (1990) Production and use of monosex or sterile populations of fish in aquaculture. Reviews in Aquat. Sci. 2:1-17.

Dunham RA (1992) Outlook for genetics research and application in aquaculture. Proceeding Agric. Outlook '92 New Opportunities for Agriculture. USDA, 68:137-148.

Dunham RA, Eash J, Askins J and Townes TM (1987) Transfer of metallothionein-human growth hormone gene into channel catfish. Trans. Amer. Fish. Soc. 116:87-91.

Dunham RA, Ramboux AC, Duncan PL, Hayat M, Chen TT, Lin CM, Kight K, Gonzalez-Villasenor I and Powers DA (1992) Transfer, expression and inheritance of salmonid growth hormone genes in channel catfish, Ictalurus punctatus, and effects on performance traits. Mol. Mar. Biol. and Biotech. 1:380-389.

Dunham RA and Smitherman RO (1984) Ancestry and breeding of catfish in the United States. Ala. Agric. Exp. Sta. Auburn Univ. Circ. 273.

Dunham RA and Smitherman RO (1987) Genetics and breeding of catfish. Southern Cooperative Series Bulletin 325. Auburn University, Auburn, AL.

Dunham RA, Turner J and Reeves WC (1992) Introgression of the Florida largemouth bass genome into native populations in Alabama public lakes. N. Amer. J. of Fish. Man. 12:494-498.

Fraser JM (1972) Recovery of planted brook trout, splake and rainbow trout from selected Ontario lakes. J. Fish. Res. Board Can. 29:129-142.

Gall GAE (1986) Sexual maturation and growth rate. In G. E. Dickerson and R. K. Johnson (eds.). Third world congress on genetics applied livestock production. 10:401-410.

Guyomard R, Chourrout D and Houdebine LM (1988) Production of stable transgenic fish by cytoplasmatic injection of purified genes. In: Proceedings of the UCLA Symposium on Gene Transfer and Gene Therapy. Los Angeles, California, USA; UCLA Press.

Hayat M, Joyce CP, Townes TM, Chen TT, Powers DA and Dunham RA (1991) Survival and integration rate of channel catfish and common carp embryos microinjected DNA at various developmental stages. Aquaculture 99:249-255.

Hulata G, Wohlfarth G, and Moav R (1985) Genetic differences between the Chinese and European races of the common carp, Cyprinus carpio L. 14. Effects of sexual maturation patterns. J. Fish Biol. 26:95-103.

Indiq FE, and Moav B (1988) A prokaryotic gene is expressed in fish cells and persists in tilapia embryos and adults following microinjection. In: Reproduction in Fish: Basic and Applied Aspects of Endocrinology and Genetics. Paris, France; INRA Press. pp. 221-225.

Isley JJ, Noble RL, Koppelman JB and Philipp DP (1987) Spawning period and first-year growth of northern, Florida and intergrade stocks of largemouth bass. Trans. Amer. Fish. Soc. 116:757-762.

Khater AA (1985) Identification and comparison of three Tilapia nilotica strains for selected aquacultural traits. Ph.D. Dissertation. Auburn University, AL, USA.

Kulzer KE, Noble RL, and Forshage AA (1985) Genetic effects of Florida largemouth bass introductions into selected Texas reservoirs. Proc. Annu. Conf. SE Assoc. Fish and Wildl. Agenc. 39:56-64.

Lerder SA, Chilicote MW, and Loch JJ (1984) Spawning characteristics of sympatric populations of steelhead trout, Salmo gairdneri, evidence for partial reproductive isolation. Can. J. Fish. Aquat. Sci. 41:1454-1462.

Maceina MJ, Murphy BR and Isley JJ (1988) Factors regulating Florida largemouth bass stocking success and hybridization with northern largemouth bass in Aquilla Lake, Texas. Trans. Amer. Fish. Soc. 117:221-231.

MacLean JA, Evans DO, Martin NV, and Desjardine RL (1981) Survival, growth, spawning, distribution, and movements of introduced and native lake trout (Salvelinus namaycush) in two inland Ontario lakes. Can. J. Fish. Aquat. Sci. 38:1685-1700.

Maniatis T, Fritsch EF, and Sambrook J (1982) Molecular cloning: a laboratory manual. Cold Spring Harbor, NY. Cold Spring Harbor Laboratory.

McEvoy T, Stack M, Keane B, Barry T, Sreenan J, and Gannon F (1988) The expression of a foreign gene in salmon embryos. Aquaculture 68, 27-37.

Norgren KG., Dunham RA, Smitherman RO and Reeves WC (1986) Biochemical genetics of largemouth bass in Alabama. Proc. Annu. Conf. SE Assoc. Fish and Wildl. Agenc. 40:194-205.

Philipp DP, Childers WF and Whitt GS (1985) Correlations of allele frequencies with physical and environmental variables of largemouth bass, Micropterus salmoides (Lacepede) J. Fish. Biol. 27:347-365.

Plosila DS (1977) Relationship of strain and size at stocking to survival of lake trout in Adirondack lakes. N. Y. Fish Game J. 24:1-24.

Pycha, RL and King GR (1967) Returns of hatchery-reared lake trout in southern Lake Superior. J. Fish. Res. Board Can. 24:281-298.

Ricker WE (1975) Computation and interpretation of biological statistics of fish population. Bulletin 191. Department of Environment, Fisheries and Marine Services. Ottawa, Canada.

Ricker WE (1981) Changes in the average size and average age of Pacific salmon. Can. J. Fish. Aquat. Sci., Vol. 38, pp. 1636- 1656.

Rokkones E, Alestrom P, Skjervold H and Gautvik KM (1989) Micro-injection and expression of a mouse metallothionein human growth hormone fusion gene in fertilized salmonid eggs. J. Comp. Phys. B 158, 751-758.

Ryman N and Stahl G (1981) Genetic perspectives of the identification and conservation of Scandinavian stocks of fish. Can. J. Fish. Aquat. Sci. 38:1562-1575.

Schneider JF, Hallerman EM, Yoon SJ, He L, Gross ML, Liu Z, Faras AJ, Hackett PB, Kapuscinski AR, and Guise KS (1989) Transfer of the bovine growth hormone gene into northern pike, Esox lucius. J. of Cell Bioch. 13 B (abstract).

Smitherman RO, Dunham RA, Bice TO, and Horn JL (1984) Reproductive efficiency in the reciprocal pairings between two strains of channel catfish. Prog. Fish-Cult. 46:106-110.

Smitherman RO, Dunham RA and Norgren K (1989a) Annual Performance Report - Florida. USFWS. Atlanta, GA.

Smitherman RO, Dunham RA and Norgren K (1989b) Annual Perforamnce Report - North Carolina. USFWS. Atlanta, GA.

Smitherman RO, Khater AA, Cassell NI and Dunham RA (1988) Reproductive performance of three strains of Oreochromis niloticus. Aquaculture. 70:29-37.

Stuart GW, McMurray JV, and Westerfield M (1988) Replication, integration and stable germ-line transmission of foreign sequences injected into early zebrafish embryos. Development. 103:403-412.

Wohlfarth G, Moav R and Hulata G (1975) Genetic differences between the Chinese and European races of common carp. II. Multi-character variation--A response to the diverse methods of fish cultivation in Europe and China. Hered. 34:341-350.

Wohlfarth GW, Moav R and Hulata G (1983) A genotype-environment interaction for growth rate in the common carp, growing in intensively manured ponds. Aquaculture. 33:187-195.

Yoon SJ, Hallerman EM, Gross M, Liu Z, Schneider JF, Faras AJ, Kapuscinski AR and Guise KS (1990) Transfer of the gene for neomycin resistance into goldfish, Carassius auratus. Aquaculture. 85:21-23.

Zhang P, Hayat M, Joyce C, Gonzalez-Villasenor LI, Lin CM, Dunham RA, Chen TT, and Powers DA (1990) Gene transfer, expression and inheritance of pRSV-rainbow trout-GHcDNA in the carp, Cyprinus carpio (Linnaeus). Mol Repro Develop. 25:3-13.