Previous Page Table of Contents Next Page


D.O.F. Skibinski
Molecular Biology Research Group, School of Biological Sciences, University of Wales
Swansea, Singleton Park, Swansea SA2 8PP, U.K.


Genetical techniques and theory play an important role in fisheries enhancement programmes. They allow the monitoring of genetic variation within hatchery stocks and thus help in the detection of the negative effects of inbreeding. Molecular genetic approaches allow the measurement of differences between populations which might reflect adaptive differences and thus be important when fish are chosen for broodstock purposes. Genetic markers play an important role in monitoring the dispersal of released fish and in assessing the extent of interbreeding between released and wild fish and the damage that this might cause. Finally, genetical techniques such as selective breeding are important in developing new types of fish, for example sterile strains, that might be important in enhancement programmes. This paper reviews these applications of genetics in enhancement programmes and makes recommendations based on considerations of current genetical theory and experimental results.


Much information is now available on the genetic effects of introductions of fish and shellfish for the purpose of enhancing fisheries. Many reviews of the subject, some emphasising genetic effects have appeared in the last decade. Of particular value are the reviews of Allendorf (1991) and Hindar et al. (1991). Many of the recommendations made at the end of this paper can be traced to these reviews. Enhancements are usually made with the aim of providing more food fish for capture. In this respect they appear to have had some success. Unfortunately the available evidence suggests that the impact of releases on indigenous fish populations tends to be negative.

Salmonids are particularly well studied. Recent reviews of supplementation of salmonids from hatcheries to increase the abundance of natural populations indicate that most efforts have either been unsuccessful or are controversial (Miller et al., 1990, and Steward and Bjornn, 1990, cited in Brannon, 1993). The Miller et al. (1990) study reviewed over 300 hatchery supplementation programmes in arriving at this conclusion.

The main genetical concerns of fishery enhancement programmes are the evolution of genetic differences between wild and hatchery stocks, the erosion of genetic variation within hatchery stocks through inbreeding and genetic drift, and the interbreeding between wild and hatchery stocks causing deterioration of the wild stocks (Stickney, 1994). These concerns are developed and extended in the present paper in four main areas. The first is the range of molecular techniques used for assaying genetic variation in populations. The second is the empirical evidence of a genetical nature concerning the success or otherwise of enhancements. The third is the variety of approaches by which cultured fish used for enhancement might be profitably modified through selective breeding or other genetical techniques. The fourth is the evidence, derived in large part from population genetics theory, on the optimal size for broodstock populations used in enhancement programmes.


Possibly the two most important uses for molecular genetic techniques in fisheries are to mark released fish and to measure differences among wild populations and hatchery stocks. Hindar et al. (1991) stress the importance of marking released fish genetically so that future consequences of releases can be monitored. Reduced performance can result when using fish sampled from one region to stock another region. This has been reported in salmonids (e.g. Fleming and Gross, 1993). Hindar et al. (1991) cite studies suggesting an inverse relationship between geographic or genetic distance between released and local populations and the success of the released fish. This implicates local adaptation as an important factor governing success of the enhancement programmes, and points to the desirability of establishing hatchery populations locally for each river or strain.

There are several recent reviews of the applications of molecular genetic techniques in fisheries and related areas (Skibinski, 1994; Ward and Grewe, 1995; Ferguson, 1995). Allozyme studies are reviewed by Ward et al. (1992, 1994). An example of use of an allozyme marker is provided by the study of Chilcote et al. (1986) in which a rare allozyme variant was used to mark hatchery bred steelheads. Unfortunately such studies are relatively rare because allozymes do not often show large frequency differences between populations in many fish species. Allozymes have however, proved particularly useful for comparing variation between populations and strains. Durand et al. (1993) analysed variation at 18 allozyme loci in wild and three generations of hatchery stocks of the black pearl oyster (Pinctada margaritifera) from Japan initiated from 40–50 individuals. The observed heterozygosity values in the wild and hatchery samples were 0.237 and 0.236 respectively, although the hatchery stocks had decreased numbers of alleles. The authors were able to conclude that, within this time scale, the small number of hatchery parents had not eroded overall genetic diversity. However, some rare variants were lost in the hatchery, an expected consequence of reduction in population size. In a study of white seabass (Atractoscion nobilis), Bartley and Kent (1990) assayed variation at 19 polymorphic allozyme loci in 13 wild and in six hatchery samples derived from 20 broodstock fish maintained for several years. The average heterozygosity in the wild samples ranged from 0.033 to 0.064 and in the hatchery samples from 0.024 to 0.060. Again, the hatchery samples lacked some rare alleles present in the wild samples, but the data indicate that at least in the short term genetic diversity was maintained quite well. Ferguson et al. (1991) compared levels of allozyme variation in cultured stocks of brown trout (Salmo trutta) and rainbow trout (Oncorhynchus mykiss) with the wild populations from which they were derived. Allele frequencies in hatchery and source populations were similar in rainbow trout but some significant differences were detected at two out of eight polymorphic loci in brown trout. Again, some rare alleles had been lost in the hatchery fish. The significance of losing rare alleles is that these could be of adaptive value for the species if environmental conditions change. This adaptive potential is lost by the hatchery fish.

The development of DNA technology over the last two decades has led to the possibility of developing unlimited numbers of genetic markers for plant and animal genomes. The main sources of markers are mitochondrial DNA (mtDNA), minisatellites, microsatellites, introns, and anonymous nuclear sequences assayed using highly specific PCR primers or by using the RAPD technique.

Mitochondrial DNA has an effective population size one quarter that of nuclear genes and thus might be expected to show greater population divergence than nuclear genes. Billington and Hebert (1991) reviewed patterns of mtDNA variation in 40 fish species. Considerable divergence of local populations was observed in line with this expectation. Hatchery stocks tended to be fixed for haplotypes common in the wild populations from which they were derived. The adaptive consequence of this is unclear, however, a positive feature is that fixed mtDNA variants might be used as genetic markers of the hatchery stocks. This level of discrimination is not often achieved by allozyme markers which tend not to show large frequency differences between hatchery stocks (Ferguson et al., 1985).

Minisatellite markers are based on length variation of tandem DNA repeats up to one hundred or so base pairs and present some technical difficulties in isolation and development for a specific organism. Many minisatellite loci are highly variable and because of this are useful in parentage analysis or for marking individual families. Unfortunately the highly variable loci are less useful for discriminating populations unless large sample sizes are used. Large numbers of alleles can also lead to difficulties in scoring and interpretation of experimental data. Minisatellites have been developed successfully for salmonids (e.g. Prodohl et al., 1995) and have been used to study population differences (e.g. Taylor, 1995).

Microsatellite markers are based on length variation of tandem repeats of usually 2–5 base pairs. They have a number of advantages in aquaculture and fisheries research over other molecular markers (O'Reilly and Wright, 1995). They are abundant in the genome, thus the number of markers is potentially unlimited. Microsatellite loci display varying levels of polymorphism. The highly polymorphic loci are of use in parentage studies, the less variable loci are more useful in discriminating populations. The assay of microsatellite variation is based on the PCR technique, thus only small amounts of tissue, for example from fish scales, are needed as a source of DNA.

A number of studies have assessed the utility of microsatellite markers in fish population genetics. For example, although allozymes and microsatellites show similar patterns of differentiation of Irish and Spanish populations of Atlantic salmon, microsatellite loci show higher levels of variation (Sanchez et al., 1996). Atlantic salmon show low levels of genetic differentiation relative to other salmonid species using allozymes and mtDNA. However, with microsatellites, McConnell et al. (1995) were able to discriminate clearly between Canadian and European fish. In cod, microsatellite loci have provided evidence of population structure at a finer geographical scale than that shown by other techniques (Ruzzante et al., 1996). Tessier et al. (1995) found significant differences between wild and first-generation hatchery fish in Atlantic salmon using microsatellites.

The RAPD technique (Williams et al., 1990) uses short PCR primers of arbitrary sequence to amplify regions in the nuclear genome in which the primers are, by chance, closely spaced. A multibanded gel pattern is produced that can be used for discriminating genetic differences between individuals and populations. For example, Bardakci and Skibinski (1994) used the technique for discriminating between species and subspecies of tilapia. Because the technique can rapidly generate large numbers of markers it has been used in gene mapping or in attempts to locate genes affecting fitness or performance characters. For example, Allegrucci et al. (1995) used RAPD to study the effects of acclimation to freshwater in European sea bass (Dicentrarchus labrax). Starting samples (S89 and S90) of about 60,000 fingerlings were reared in seawater up to a size of 70 g. Samples of 7000 (from S89) and 2600 (from S90) fish were then acclimated to freshwater (A89 and A90) over a period of two years with 94% and 75% mortality in the A89 and A90 samples respectively. Of 126 polymorphic RAPD markers scored in at least 39 individuals per sample, two showed significant changes in marker frequencies between S89 and A89 and 13 showed significant changes between S90 and A90. However, only one marker showed concordant changes in both acclimated samples, suggesting that the changes during the acclimation period were due either to drift or differences in the starting samples. The disadvantage of RAPD analysis is that great care is required to obtain reproducible results, and genetic interpretation of RAPD patterns is problematical without breeding experiments. Although the RAPD technique and refinements of it are likely to be used widely in future, strategic research on reliable codominant markers such as microsatellites might pay greater dividends in the long term.

The failure to detect variation in molecular markers between populations should not necessarily be taken as evidence that genetic differences for other characters are also absent. Gharrett and Smoker (1993) have emphasised the importance of taking account of geographic differences in polygenic characters in fisheries management practices. For example salmon populations frequently show local differences in characters such as timing of migration. Enhancement programmes must take account of such differences if adaptive genetic variation in the species as a whole is not to be eroded.


There are a number of substantial reviews of the effects of fish introductions (see Allendorf, 1991). For example, a supplement of the Canadian Journal of Fisheries and Aquatic Sciences (Vol. 48S1, 1991) published the proceedings of an International Symposium on “The Ecological and Genetic Implications of Fish Introductions”. The genetic effects of cultured salmonids on natural populations have been well studied and are reviewed by Hindar et al. (1991).

Cultured salmonids are being released into the wild either deliberately or as escapees from closed culture conditions. The released fish may be genetically different from local fish because they are derived from genetically distinct source populations, or because they have been modified during a selective breeding programme, or because they have been modified by drift or unintentional selection in culture. Often, the released fish outnumber local fish. Although the intention is that released or escaped fish will be harvested, they may reproduce successfully and come to represent a large proportion of the naturally spawning population. For example, in rivers in southern Norway up to 80% of spawning Atlantic salmon were escapees (see Gausen and Moen, 1991).

Hindar et al. (1991) use 33 published studies to review the impact of releases in relation to interbreeding, competition and contamination effects on genetic structure and performance. They conclude that when an effect is detectable, the indigenous fish most often perform better, and these effects have a genetic basis in part. They conclude that enhancement programmes have, in general, a negative effect on indigenous local populations. The same general conclusion is reached in the review by Allendorf (1991). Hindar et al. (1991) admit that there exist opposing views, and that some authors argue that large-scale releases might be neutral or have beneficial effects by introducing new variation into local populations. Many of these enhancement programmes were not designed with the aim of assessing genetic effects, thus clear-cut evidence is not easy to obtain. Following the approach of Hindar et al. (1991), some of the more important effects on genetic structure and performance are outlined below.

3.1 Effects on genetic structure

Displacement: Enhancement programmes can lead to decline or displacement of indigenous populations, effectively changing genetic structure. Mass transfer of chum salmon to rivers in the USSR was followed by dominance of donor gene frequencies in some areas. Subsequently the donor fish disappeared but total returns dropped substantially (Altukhov, 1981). Even if there is no introgression, indigenous populations may be reduced by increased competition from the larger total number of fish and by increased rates of harvest. For example, Evans and Willox (1991) reported that hatchery reared lake trout (Salvelinus namaycush) can cause the decline of indigenous stocks because of increased exploitation even if the introduced fish do not reproduce.

Introgression: There are many examples of introgression between donor and indigenous populations. In salmonids, the effects are rather unpredictable and include the continued coexistence of both types of fish. However, if introgression occurs, poorer performance of cultured fish can be transferred to the indigenous local populations. If there are differences between the introduced and indigenous fish at only a few loci, introgression leads to a situation where the resulting population will have essentially no individuals with genotypes that occurred in the original populations. If the indigenous population had well adapted genotypes the introgression will clearly have eroded this variation. Management programmes have been responsible for the introduction of the Florida largemouth bass (Micropterus salmoides floridanus) into populations of the northern subspecies M. s. salmoides. Philipp (1991) reported results of growth trials of experimental populations including both species, their hybrids, and introgressed individuals. The greatest growth was always achieved by salmoides, suggesting the potential negative effects of hybridisation and introgression in the wild. In principle, although introgression could bring new and increased genetic variation into the recipient population to provide raw material for adaptive evolution the consequences cannot be predicted a priori, and there are unlikely to be short-term advantages. Benefit might more easily arise where the introduced population differs from the indigenous population at a single locus carrying an allele that confers a major advantage, for example to cold tolerance. However, again the effects are unpredictable and might be unwelcome.

Reduction of geographic diversity: Movement of escaped fish between hatcheries is expected to lead to substantial gene flow and homogenisation of gene frequencies and genetic structure between geographically distant populations and hatcheries. This would frustrate plans for hatcheries to produce releases of locally adapted fish.

Hybridisation between species: There are a number of reports of hybridisation between different salmonid species when fish are released beyond their normal range. In one instance (Garcia de Leaniz and Verspoor, 1989), the released fish interbred in an area where the two species normally coexist. This suggested that the natural isolating mechanisms in the released fish might have been altered genetically during culture.

3.2 Effects on performance traits

Population size: Enhancement programmes, though causing population expansion initially, might lead subsequently to lower population sizes and hence lower yield. In coho salmon, the density of fish in stocked streams exceeded that in unstocked streams in the summer following release. Subsequently the density of juveniles in stocked streams declined and was attributed to earlier spawning of hatchery fish which contributed little to recruitment (Nickelson et al., 1986).

Disease resistance: Cultured fish derived from one region and transferred to another can bring parasites and diseases with them. Atlantic salmon from the Baltic have resistance to the parasite Gyrodactylus salaris. Transfers from the Baltic to Norwegian rivers have dramatically reduced local populations which lack this resistance (Bakke et al., 1990). More than 30 local populations have been wiped out by this parasite (Heggberget et al., 1993).

Inferior adaptation: A number of studies provide evidence of fish showing poor adaptation to the environment where they were released. There is evidence that this lack of adaptation can have a genetic basis in part and can evolve in hatcheries. Green (1964) demonstrated reduced swimming stamina of cultured brook trout compared with wild fish. In steelheads, hatchery fish tend to have a narrower spawning window. If this characteristic were transferred to wild fish, adverse environmental conditions at the time of the window, e.g. floods, may destroy a large proportion of eggs or juveniles. In a recent Norwegian fishery enhancement scheme with brown trout (Salmo trutta), 70,000 fingerlings were released from 10 localities (Fjellheim et al., 1995). Compared with wild fish, the stocked fish had lower growth and food consumption, higher mortality (99% compared with 79% for wild) and moved less away from the point of release.

Negative effects of releases are not confined to circumstances where the released fish have the possibility of interbreeding with local fish of the same species. The pressure group Greenpeace is opposed to fisheries enhancement except in circumstances when a population of the species is no longer extant in a region (Greenpeace Principles for Ecologically Responsible Fisheries, preliminary document - February 1996). However, negative effects of introductions can occur even in areas where the introduced species is not endemic. For example, the introduction of the Nile perch in Lake Victoria is believed to have played a part in the extinction of 260 endemic species (Leveque, 1995). Clear Lake in California originally had twelve native species. Sixteen species have been introduced causing the extinction of five of the twelve indigenous species (Moyle, 1986). Even if there are no direct effects, such as through predation, the reduction of population sizes of indigenous fish caused by introductions might considerably reduce genetic diversity in the endemic species thus affecting their ability to adapt to environmental change.


Fish used as broodstock for enhancement purposes could be subject to artificial selection for improvement of performance characters. An example of this in practice is Iceland's stocking programme for Atlantic salmon which includes a selective breeding programme for growth rate and rate of return. However, such selection programmes are not common at present in fish (Welcomme and Bartley, 1997). One potentially undesirable consequence of maintaining stocks over long periods of time for enhancement purposes is that they might adapt by natural selection or unintentional artificial selection to the culture conditions (Allendorf, 1993). Genetic changes, whether due to artificial selection, natural selection, or genetic drift are likely to lead to loss of genetic variation and poorer adaptation to natural conditions. On release these cultured stocks will pose a threat through introgression to natural populations. Undesirable effects might be minimised by supplementing breeding stocks with wild fish, however, outbreeding a selection line will reduce the response to selection. Genetic structure of cultured stocks can be conserved to some extent by good hatchery management, reducing stock transfers, and reducing transfers of eggs and fingerlings between hatcheries (Philipp et al., 1993).

Doyle et al. (1991) propose that selective breeding programmes on aquacultural species should aim to develop improved strains adapted specifically to local conditions. Farmers would select the best strains for their conditions by trial and error from those strains available. This system emphasises genotype environment interaction, and would conserve considerable genetic diversity within the gene pool of the species. In the context of enhancement programmes, releases of different locally adapted strains would preserve genetic variation and adaptability globally.

Molecular genetic markers facilitate the identification of chromosomal segments containing loci controlling quantitative characters (QTLs) and eventually the identification of the loci themselves (Haley, 1995; Falconer and Mackay, 1996). QTL identification has perhaps been most successful in plants. However, the association of performance characters with market loci has also been reported in pigs (e.g. Clamp et al., 1992) and cattle (e.g. Rocha et al., 1992). Specific mtDNA haplotypes have also been associated with spawning time, body weight and condition factor in two commercial strains of rainbow trout (Danzmann et al., 1994; Danzmann and Ferguson, 1995). The association of DNA markers with QTLs has raised the possibility that such markers might be incorporated in selection programmes (Lande and Thompson, 1990). This would effectively increase the heritability as marker phenotypes would not be affected by environmental variation. An approach towards this marker-assisted selection (MAS) in fish has been made in rainbow trout by Herbinger et al. (1995). Five microsatellite loci were used to mark parents and progeny in a factorial cross. The progeny were reared together, and the markers then used to trace the parents of the largest and smallest fish. Significant differences for growth and survival of progeny were observed between sires and between dams. A breed improvement programme has been initiated on the basis of these results.


A number of genetic manipulation techniques are being actively investigated in aquacultural research. These include ploidy and chromosome manipulation and transgenesis. In relation to enhancement programmes, the use of these techniques to develop conditionally sterile strains of fish might be attractive because threats to indigenous populations through introgression would be removed. Since the released fish could not breed, any undesirable effects could be eliminated by halting release of modified fish. An example of such an approach is provided by the use of grass carp, made sterile by triploidization, for weed control (Wynn, 1992). Sterile triploid rainbow trout form a large proportion of production in the UK (Bye and Lincoln, 1986). Chromosome manipulation through gynogenesis and sex reversal to produce monosex strains with improved performance, as in tilapia (e.g. Mair et al., 1995), might be incorporated in enhancement programmes. Sterile fish could still cause considerable problems to local populations, for example through competition for food or if sterile males compete in courtship with fertile wild males.

Much research has now been carried out on transgenic fish (see Maclean and Rahman, 1995). For example, salmonids with impressive growth enhancement have been produced (Devlin et al., 1994). However, there exists considerable concern over the advisability of aquacultural production of transgenic fish because of the potentially damaging ecological effects of escapes (Kapuscinski and Hallerman, 1991). Transgenic strains are likely to be impoverished genetically and thus will have the same deleterious features as inbred cultured stocks. In addition, there will be the unpredictable consequences of transfer of the transgene to native populations. In these circumstances it appears difficult to justify the use of transgenics in enhancement programmes unless sterile strains can be produced. In any event, careful step-by-step risk assessments would be needed. Ironically, molecular genetic techniques would be of great value in risk assessment by providing markers for monitoring escapes and introgression of transgenes.


Hybridisation between related species might be considered for enhancement purposes if the hybrid shows evidence of heterosis. For example, the saugey, a cross between walleye and sauger, is superior to both parent species and is being stocked in rivers for recreational purposes (White and Schell, 1995). A danger of increased risk of introgression might occur in localities where the two species occur naturally. In Bangladesh, the Asian catfish (Clarias batrachus) is more favoured by consumers than the introduced African catfish (C. gariepinus), but grows more slowly. Rahman et al. (1995) show that offspring of the cross Clarias batrachus × C. gariepinus show heterosis with respect to performance characters. Exploitation of this result might be advantageous for aquaculture production but might put the native Asian catfish at risk through introgressive hybridisation. The dangers are likely to be less when the species hybridise at a low level naturally, for example the brown trout and Atlantic salmon, and where natural selection has already resulted in the development of isolating mechanisms.


One of the key problems in enhancement programmes concerns the choice of an appropriate size for the broodstock population. Large populations are costly for maintenance. Thus the aim is to have a population as small as possible yet avoid problems caused by loss of genetic variation through drift and inbreeding. Similar considerations arise in selection programmes and when establishing captive breeding populations of threatened animals. Of particular importance is the effective population size (Ne), rather than the census number (N), as this is central to determination of the level of genetic variation within the population. Unequal sex ratio, temporal fluctuation in N, bottlenecks in N, and variance in reproductive success between individuals reduce Ne more than they reduce N. This is because Ne depends on the geometric mean of N not the arithmetic mean. The relationship between the number of adults in a population and the effective size is given by Ne/N = 4/(2+V) (Wright, 1938), where V is the variance in reproductive success. If, in a broodstock maintenance programme, all matings are controlled and the numbers of progeny per mating contributing to the next generation equalised, then V = 0 and Ne = 2N. Drift in such a population would actually be less than in an idealised random mating population of equal numbers of males and females.

How big does Ne have to be to maintain a level of genetic variation approaching that in wild populations? Franklin (1980) and Soule (1980) estimated that an Ne of 500 would probably maintain adequate genetic variation for quantitative characters. This number is based on a consideration of the balance between the loss of genetic variation by drift and its gain by spontaneous mutation. Empirical observations from a variety of plants and animals suggest that of the total genetic variation (Vg), the proportion arising by spontaneous mutation each generation (Vm) is about 10-3 times the environmental variation (Ve). To maintain a typical heritability value (Vg/(Ve+Vg)) of 0.5 by a balance between genetic drift and mutation, Ne needs to be about 500. However, many mutations are detrimental and unlikely to reach high frequencies within populations. Thus those that contribute to the observed quantitative variation are likely to be quasineutral, having small positive or negative effects on fitness. Lande (1995) estimated, on the basis of empirical results with the fruit fly Drosophila, that quasineutral mutations only represent 10% of Vm. Thus the estimate of Ne needs to be revised upwards by a factor 10 to 5000. These considerations of the balance between mutation and drift apply to long-term maintenance of an isolated population and assume that the dominance of drift in a small population is not acceptable from a management viewpoint. A much smaller population size could be acceptable if the broodstock is maintained for only a few generations and then replenished or replaced by wild fish. However, if the replacement fish are themselves largely derived from past releases, decline in the level of genetic variation by drift might not be avoided completely.

Most populations are probably under some form of stabilising selection and Lande (1995) showed also that the estimate of Ne = 5000 for maintenance of potentially adaptive genetic variation holds also in this circumstance. This figure is considerably larger than most fish broodstock population sizes. Heterozygous advantage and other forms of balancing selection are very efficient at maintaining high levels of genetic variation even in small populations. If heterozygous advantage were widespread, smaller broodstock population sizes would suffice. A number of studies have provided evidence for allozymes that individuals that are heterozygous grow faster than those that are homozygous (e.g. Danzmann et al., 1988). This could provide the empirical basis for heterozygous advantage for some allozyme loci. However, for quantitative characters there is little evidence to suggest that new mutations exhibit heterozygous advantage.

In large populations, deleterious mutations can be removed efficiently by natural selection. In small populations these can be fixed more easily by drift and thus lead to a reduction in reproductive rate compared with larger populations. It is shown that populations with Ne less than 100 are at high risk of extinction over a time scale of 100 generations, but that this risk would be increased substantially by even a doubling of the mutation rate, caused for example by human pollution. This estimate does not take into account the effects of environmental fluctuation and thus may underestimate the risk. This result also points to the undesirability of establishing broodstock populations with 10s of individuals.

Direct measurement of Ne is difficult as it requires estimates of fertility for all individuals in the population. However, indirect estimation is possible. For example, Hedgecock et al. (1992) estimated Ne from data on temporal variation in gene frequencies. The method makes a number of assumptions including selective neutrality of the variation and absence of migration. Results for 16 shellfish stocks indicate Ne values of less than 100, with 13 having values less than 50. In most circumstances, in natural populations, the ratio Ne/N would be in the range 0.25–0.75 (Nunney, 1992). However, in the study of Hedgecock et al. (1992) an extreme value of 10-6 was reported for oysters. This was attributed to high variance in fecundity which should in theory depress the value of the ratio. The high variance was in turn attributed to high female fecundity in this species. However, Nunney (1996) has shown recently that there is no clear connection in theory between high fecundity and low values for the ratio.

Hatchery mating strategies have a marked influence on Ne and levels of genetic variation. Many broodstock populations are founded from small numbers of individuals. For example in the Philippines, the widely cultured Israel strain is derived from a single introduction of 100–200 fry, possibly from a single family (Pullin, 1988). Eknath and Doyle (1990) estimated that Ne of hatchery stocks of carp in Karnataka state in India varied between 3 and 30. The problem of population bottlenecks might be particularly acute in highly fecund species such as carp where there is a danger of some families contributing disproportionally to future generations. Combining gametes of different individuals prior to fertilisation might also inflate the variance of familial contribution, and there is evidence that this procedure has the effect of reducing levels of genetic variation and Ne (e.g. Withler, 1988). Doyle and Talbot (1986) argue that broodstock management practices in Asia frequently exert strong negative selection on early growth rate and this also leads to smaller than anticipated Ne values. A number of studies have now been carried out in which genetic markers have been used for identification of the parentage of progeny. In an aquacultural application, Herbinger et al. (1995) made a 10×10 diallele cross on a rainbow trout farm followed by communal rearing. Using microsatellites as markers it was subsequently possible to trace 91% of the progeny fish to only one or two couples out of the 100 possible couples. This suggests that the potential for genetic drift or inbreeding with communal spawning due to small Ne is very substantial. It thus appears that in practice broodstock sizes are smaller than are required for long-term maintenance of adaptive genetic variation.

Broodstocks will usually be much smaller than the wild populations they are to enhance. Thus if they contribute in a major way to naturally spawning populations they represent a form of bottleneck through which the wild populations are funnelled. Waples and Do (1994) considered this phenomenon explicitly in simulations of a model in which captive broodstocks were used to supplement natural populations of Pacific salmon. The aim was to determine the relative importance of factors affecting the level of inbreeding and consequent loss of genetic diversity in the natural populations. The most important factor was whether the wild population remained large after supplementation. If the population increases substantially in size as a result of enhancement but then drops in size to its original level, a large proportion of the population will be derived only from the broodstock and Ne will be reduced. Maintenance of a large population size post-supplementation to minimise this effect might depend on improving or extending the habitat to facilitate a larger carrying capacity. Of relevance here is the study of Fjellheim et al. (1995) which reported inferior performance of released fish in a brown trout enhancement scheme. The authors concluded that the natural recruitment of brown trout was close to the carrying capacity and thus enhancement was not an effective strategy. They suggested that effort would be better devoted to improving the habitat.

Waples and Do (1994) also suggested that marking of fish released from the hatchery is important in maintaining Ne as it enables these fish to be avoided as broodstock in later years, thus helping to avoid an increasing proportion of the population being descended from a small number of individuals. The effect of mating strategies was also investigated. Sib-avoidance was found not to affect the rate of inbreeding compared with random mating for the simple reason that it does not change allele frequencies within the population as a whole.


Welcomme and Bartley (1997) review the problem of measuring the cost effectiveness of fisheries enhancements. The value of fish are dependent on circumstances, for example being greater for recreational fisheries than for commercial food fisheries. However, they argue that many cost benefit analyses have demonstrated that stocking can be highly profitable. For example, stocking of oxbow lakes in Bangladesh gave income over expenditure ratios of 1.78. However, it might sometimes be difficult in such analyses to gauge the environmental cost or the cost of detrimental genetic changes to natural populations.

Given that enhancement programmes are to be actively pursued, genetical work should play a part in the areas outlined in this paper. Genetical work is usually expensive. Institutionalised selective breeding programmes in fish can be expensive in terms of manpower and infrastructure. Molecular genetic techniques are expensive in terms of equipment, consumables and training of highly specialised personnel. To weigh these costs against benefits is quite difficult. For example, some doubts have been raised about the value of MAS given the high costs of DNA analysis (Strauss et al., 1992). It seems that the technique will be most effective when trait heritabilities are low, QTLs account for much of the genetic variance, selection intensity is high within families, many progeny are examined, and the genetic base is such as to minimise linkage equilibrium between QTLs and the genetic background. There are many unknowns here in relation to the probability of realising benefits even though financial costs can be estimated. Another example is transgenic work. The financial costs are high and can be estimated, whereas the benefit, though potentially very great, is not easy to predict.


Enhancement programmes have a number of possible genetic consequences. What happens in the end depends on the values of the theoretical parameters of population genetics - selection, mutation, migration and population size as well as on historical and local environmental factors and the uncertainties of human behaviour. For example, consider the problem of predicting future changes in gene frequency in introduced fish caused by selection. This requires some quantitative knowledge of selection coefficients acting at individual loci. These coefficients which might affect viability or fertility, though measurable in principle given sufficient scientific resources, could not be measured in practice in nature to sufficient precision to allow predictions. Hence there will always be an element of uncertainty about what will happen, genetically or to performance, when fish are released. A further problem for prediction concerns the difficulty in assessing the significance of genotype environment interactions when the fish used for enhancement are genetically variable. For example, any variable that might affect the success of stocking, e.g. stocking rate, might have an interaction with a genetic variable, that is some genotypes might have different norms of reaction in terms of return when plotted against stocking rate. Under this scenario of high uncertainty, possibly the best that can be done is to use the principles of risk assessment to determine policy, taking account of all available information, and to be cautious.

A further important consideration in relation to recommendations is whether an enhancement programme has the primary objective of helping the recovery of threatened wild populations of a species, or is simply a means of increasing the amount of food (Ferguson, 1995). In relation to the second objective, it is clear that putting large numbers of juvenile cultured fish into a body of water might have a negative impact on indigenous populations of the same species. However, if the fish are released into an area free of indigenous populations this will not be a problem, though other species might be affected.

The following recommendations, pertaining to genetic aspects of enhancement, can be made.


Allegrucci, G., A. Caccone, S. Cataudella and J.R. Powell. 1995. Acclimation of the European sea bass to freshwater: monitoring genetic changes by RAPD polymerase chain reaction to detect DNA polymorphisms. Marine Biology 121:591–599.

Allendorf, F.W. 1991. Ecological and genetic effects of fish introductions - synthesis and recommendations. Can. J. Fish. Aquat. Sci. 48(S1): 178–181.

Allendorf, F.W. 1993. Delay of adaptation to captive breeding by equalizing family size. Conservation Biology 7:416–19.

Altukhov, Y.P. 1981. The stock concept from the viewpoint of population genetics. Can. J. Fish. Aquat. Sci. 38: 1523–1538.

Bakke T.A., P.A. Jansen and L.P. Hansen. 1990. Differences in the host resistance of Atlantic salmon, Salmo salar L., stocks to the monogenean Gyrodactylus salaris Malmberg, 1957. J. Fish Biol. 37:577–587.

Bardakci, F. and D.O.F. Skibinski. 1994. Application of the RAPD technique in tilapia fish: species and subspecies identification. Heredity 73: 117–123.

Bartley, D.M. and D.B. Kent. 1990. Genetic structure of white seabass populations from the southern California Bight region: applications to hatchery enhancement. COFI Report No. 31:97–105.

Billington, N. and P.D.N. Hebert. 1991. Mitochondrial DNA diversity in fishes and its implications for introductions. Can.J.Fish.Aquat.Sci. 48(S1):80–94.

Brannon, E.L. 1993. The perpetual oversight of hatchery programs. Fisheries Research 18: 19–27.

Bye, V.J. and R.F. Lincoln. 1986. Commercial methods for the control of sexual maturation in rainbow trout (Salmo gairdneri R.). Aquaculture 57:299–309.

Chilcote, M.W., S.A. Leider and J.J. Loch. 1986. Differential reproductive success of hatchery and wild summer-run steelhead under natural conditions. Trans. Am. Fish. Soc. 115: 726–735.

Clamp, P.A., J.E. Beever, R.L. Fernando, D.G. McLaren and L.B. Schook. 1992. Detection of linkage between genetic markers and genes that affect growth and carcass traits in pigs. J. Anim. Sci. 70:2695–2706.

Danzmann, R.G. and M.M. Ferguson. 1995. Heterogeneity in the body size of Ontario cultured rainbow trout with different mtDNA haplotypes. Aquaculture 137:231–244.

Danzmann, R.G., M.M. Ferguson and F.W. Allendorf. 1988. Heterozygosity and components of fitness in a strain of rainbow trout. Biol. J. Linn. Soc. 33:285–304.

Danzmann, R.G., M.M. Ferguson and D.M. Heculuck. 1994. Heterogeneity in the distribution of mitochondrial DNA haplotypes in female rainbow trout spawning in different seasons. Can.J.Fish. Aquat. Sci. 51:284–289.

Devlin, R.H., T.Y. Yesaki, C.A. Biagi, E.M. Donaldson, P. Swanson and W.K. Chan. 1994. Extraordinary salmon growth. Nature 371:209–210.

Doyle, R.W. and A.J. Talbot. 1986. Effective population size and selection in variable aquaculture stocks. Aquaculture 57:27–35.

Doyle, R.W., N.L. Shackel, Z. Basiao, S. Uraiwan, T. Matricia and A.J. Talbot. 1991. Selective diversification of aquaculture stocks: a proposal for economically sustainable genetic conservation. Can. J. Fish. Aquat. Sci. 48(S1):148–154.

Durand, P., K.T. Wada and F. Blanc. 1993. Genetic variation in wild and hatchery stocks of the black pearl oyster, Pinctada margaritifera, from Japan. Aquaculture 110:27–40.

Eknath, A. and R.W. Doyle. 1990. Effective population size and rate of inbreeding in aquaculture of Indian major carps. Aquaculture 85:293–305.

Evans, D.O. and C.C. Willox. 1991. Loss of exploited, indigenous populations of lake trout, Salvelinus namaycush, by stocking of non-native stocks. Can. J. Fish. Aquat. Sci. 48(S1):134–147.

Falconer, D.S. and T.F.C. Mackay. 1996. An Introduction to Quantitative Genetics. Longman Group Limited, Harlow, England.

Ferguson, M. 1995. The role of molecular genetic markers in the management of cultured fishes. In: Molecular Genetics in Fisheries (Carvalho, G.R. and T.J. Pitcher, eds.): 29–54. Chapman and Hall, London.

Ferguson, M.M., P.E. Ihssen and J.D. Hynes. 1991. Are cultured stocks of brown trout (Salmo trutta) and rainbow trout (Oncorhynchus mykiss) genetically similar to their source populations. Can. J. Fish. Aquat. Sci. 48(S1):118–123.

Fjelheim, A., G.G. Raddum and B.T. Barlaup. 1995. Dispersal, growth and mortality of brown trout (Salmo trutta L.) stocked in a regulated west Norwegian river. Regulated Rivers: Research and Management 10: 137–145.

Fleming, I.A. and M.R. Gross. 1993. Breeding success of hatchery and wild coho salmon (Oncorhynchus kisutch) in competition. Ecological Applications 29: 305–327.

Franklin, I.R. 1980. Evolutionary changes in small populations. In: Conservation Biology: An Evolutionary Perspective (Soule, M.E. and B.A. Wilcox, eds.): 135–149. Sinauer Associates, Sunderland, Massachusetts.

Garcia de Leaniz, C. and E. Verspoor. 1989. Natural hybridization between Atlantic salmon, Salmo salar, and brown trout, Salmo trutta, in northern Spain. J. Fish. Biol. 34:41–46.

Gausen, D. and V. Moen. 1991. Large-scale escapes of farmed Atlantic salmon (Salmo salar) into Norwegian rivers threaten natural populations. Can. J. Fish. Aquat. Sci. 48: 426–428.

Gharrett, A.J. and W.W. Smoker. 1993. A perspective on the adaptive importance of genetic infrastructure in salmon populations to ocean ranching in Alaska. Fisheries Research 18: 45–58.

Green, D.M. 1964. A comparison of stamina of brook trout from wild and domestic parents. Trans. Am. Fish. Soc. 93:96–100.

Haley, C.S. 1995. Livestock QTLs - bringing home the bacon. Trends in Genetics 11: 482–487.

Hedgecock, D.V., V. Chow and R.S. Waples. 1992. Effective population numbers of shellfish broodstocks estimated from temporal variance in allele frequencies. Aquaculture 108: 215–232.

Heggberget, T.G., B.O. Johnsen, K. Hindar, B. Jonsson, L.P. Hansen, N.A. Hvidsten and A.J. Jensen. 1993. Interactions between wild and cultured Atlantic salmon: a review of the Norwegian experience. Fisheries Research 18: 123–146.

Herbinger, C.M., R.W. Doyle, E.R. Pitman, D. Paquet, K.A. Mesa, D.B. Morris, J.M. Wright and D. Cook. 1995. DNA fingerprint based analysis of paternal and maternal effects on offspring growth and survival in communally reared rainbow trout. Aquaculture 137: 245–256.

Hindar, K., N. Ryman and F. Utter. 1991. Genetic effects of cultured fish on natural fish populations. Can. J. Fish. Aquat. Sci. 48: 945–957.

Kapuscinski, A.R. and E.M. Hallerman. 1991. Implications of introduction of transgenic fish into natural ecosystems. Can. J. Fish. Aquat. Sci. 48(SI): 99–107.

Lande, R. 1995. Mutation and conservation. Conservation Biology 9: 782–791.

Lande, R. and R. Thompson. 1990. Efficiency of marker-assisted selection in the improvement of quantitative traits. Genetics 124: 743–756.

Leveque, C. 1995. Role and consequences of fish diversity in the functions of African freshwater systems - a review. Aquatic Living Resources 8: 59–78.

Maclean, N. and A. Rahman. 1995. Transgenic Fish. In: Animals with Novel Genes (N. Maclean, ed.): 63–105. C.U.P.

Mair, G.C., J.S. Abucay, J.A. Beardmore. and D.O.F. Skibinski. 1995. Growth performance trials of genetically male tilapia (GMT) derived from YY-males in Oreochromis niloticus L.: On station comparisons with mixed sex and sex reversed male populations. Aquaculture 137: 313–323.

McConnell, S.K., P. O'Reilly, L. Hamilton, J.N. Wright and P. Bentzen. 1995. Polymorphic microsatellite loci from atlantic salmon (Salmo salar) - genetic differentiation of North American and European populations. Can. J. Fish. Aquat. Sci. 52: 1863–1872.

Miller, W.H., T.C. Coley, H L. Burge and T.T. Kisanuki. 1990. Part I: Analysis of salmon and steelhead supplementation: Emphasis on unpublished reports and present programmes. In: Analysis of Salmon and Steelhead Supplementation (Miller, W.H., ed.): 1–46. Technical Rep., US Department of Energy, Bonneville Power Administration, Division of Fish and Wildlife, Project 88–100.

Moyle, P.B. 1986. Fish introductions into North America: patterns and ecological impact. In: Ecology of Biological Invasions in North America and Hawaii: 27–43. Springer-Verlag. New York.

Nickelson, T.E., M.F. Solazzi and S.L. Johnson.. 1986. Use of hatchery coho salmon (Oncorhynchus kisutch) to rebuild wild populations in Oregon coastal streams. Can. J. Fish. Aquat. Sci. 43: 2443–2449.

Nunney, L. 1992. Estimating the effective population size and its importance in conservation strategies. Transactions of the Western Section of the Wildlife Society 28: 67–72.

Nunney, L. 1996. The influence of variation in female fecundity on effective population size. Biol. J. Linn. Soc. 59: 411–425.

O'Reilly, P. and J.M. Wright. 1995. The evolving technology of DNA fingerprinting and its application to fisheries and aquaculture. J. Fish Biology 47: 29–55.

Philipp, D.P. 1991. Genetic implications of introducing Florida largemouth bass, Micropterus salmoides floridanus. Can. J. Fish. Aquat. Sci. 48(SI):58–65.

Philipp, D.P., J.M. Epifano and M.J. Jennings. 1993. Point/counterpoint: Conservation genetics and current stocking practices - are they compatible? Fisheries 18: 14–16.

Prodohl, P.A., J.B. Taggart and A. Ferguson.. 1995. A panel of minisatellite (VNTR) DNA locus-specific probes for potential application to problems in salmonid aquaculture. Aquaculture 137: 87–97.

Pullin, R.S.V. 1988. Tilapia Genetic Resources for Aquaculture. ICLARM Conference Proceedings 16. ICLARM, Manila, Philippines. 108p.

Rahman, M.A., A. Bhadra, N. Begum, M.S. Islam and M.F. Hussain. 1995. Production of hybrid vigor through cross breeding between Clarias batrachus Lin. and Clarias gariepinus Bur. Aquaculture 138: 125–130.

Rocha, J.L. and 4 others. 1992. Statistical associations between restriction fragment length polymorphisms and quantitative trait loci. J. Anim. Sci. 70: 3360–3370.

Ruzzante, D.E., C.T. Taggart, C. Cook and S. Goddard. 1996. Genetic differentiation between inshore and offshore atlantic cod (Gadus morhua) off Newfoundland - microsatellite DNA variation and antifreeze level. Can. J. Fish Aquat. Sci. 53: 634–645

Sanchez, J.A., C. Clabby, D. Ramos, G. Blanco, F.Flavin, E. Vazquez and R. Powell. 1996. Heredity 77: 423–432.

Soule, M.E. 1980. Thresholds for survival: maintaining fitness and evolutionary potential. In: Conservation Biology: An Evolutionary Perspective (Soule, M.E. and B.A. Wilcox, eds.): 151–170. Sinauer Associates, Sunderland, Massachusetts.

Skibinski, D.O.F. 1994. The potential of DNA techniques in the population and evolutionary genetics of aquatic invertebrates. In: Proceedings of the Conference on Genetics and Evolution of Aquatic Organisms, 10–16 Sept 1992, Bangor, UK. (Beaumont, A., ed.): 177–199. Chapman and Hall, London.

Steward, C.R. and T.C. Bjornn. 1990. Part II: Supplementation of salmon and steelhead stocks with hatchery fish: A synthesis of published literature. In: Analysis of Salmon and Steelhead Supplementation (Miller, W.H., ed.): 1–26. Technical Report., US Department of Energy, Bonneville Power Administration, Division of Fish and Wildlife, Project 88–100.

Stickney, R.R. 1994. Use of hatchery fish in enhancement programs. Fisheries 19: 6–13.

Strauss, S.H., R. Lande and G. Namkoong. 1992. Limitations of molecular marker aided selection in forest tree breeding. Can. J. For. Res. 22: 1050–1061.

Taylor, E.B. 1995. Genetic variation at minisatellite DNA loci among North Pacific populations of steelhead and rainbow trout (Oncorhynchus mykiss). J. Hered. 86: 354–363.

Tessier, N., L. Bernatchez., P. Presa and B. Angers. 1995. Gene diversity analysis of mitochondrial DNA, microsatellites and allozymes in landlocked Atlantic salmon. J. Fish Biology 47: 156–163.

Ward, R.D. and P.M. Grewe. 1995. Appraisal of molecular genetic techniques in fisheries. In: Molecular Genetics in Fisheries (Carvalho, G.R. and T.J. Pitcher, eds.): 29–54.

Waples, R.S. and C. Do. 1994. Genetic risk associated with supplementation of Pacific salmonids: captive broodstock programs. Can. J. Fish Aquat. Sci. 51(S1): 310–329.

Ward, R.D., D.O.F. Skibinski and M. Woodwark. 1992. Protein heterozygosity, protein structure, and taxonomic differentiation. Evolutionary Biology 26: 73–159.

Ward, R.D., M. Woodwark. and D.O.F. Skibinski. 1994. A comparison of genetic diversity levels in marine, freshwater, and anadromous fishes. Journal of Fish Biology 44: 213–232.

Welcomme, R.L. and D.M. Bartley. 1997. An evaluation of present techniques for the enhancement of fisheries. (MS).

White, M.M. and S. Schell. 1995. An evaluation of genetic integrity of Ohio River walleye and sauger stocks. American Fisheries Symposium 15: 52–60.

Williams, J.G.K., A.R. Kubelik, K.J. Livak, J.A. Rafalski and S.V. Tingey. 1990. DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acids. Res. 18: 6531–6535.

Withler, R.E. 1988. Genetic consequences of fertilizing chinook salmon (Oncorhynchus tshawytscha) eggs with pooled milt. Aquaculture 68: 15–25.

Wright, S. 1938. Size of population and breeding structure in relation to evolution. Science 87: 430–431.

Wynn, F. 1992. Controlling aquatic vegetation with triploid grass carp. World Aquaculture 23: 36–37.

Previous Page Top of Page Next Page