Human activities can cause the erosion of genetic variation and the extinction of fish via several mechanisms. These include:
Pollution and other environmental changes that stress a population and cause differential mortality, extinction, or both;
Fishing pressure which can favour some genotypes over others;
Artificial selection and domestication which can result in conscious or unconscious inbreeding and genetic impoverishment;
The introduction of exotic species and diseases.
In this section the above factors are reviewed in the light of recent experience in fisheries and aquaculture.
Pollution of different kinds (e.g., eutrophication, toxins such as mercury, DDT, PCB) have had disastrous effects on many fish communities. In most western countries toxins in single fish species have made them unfit for human consumption. In some cases legislation has been successful in reversing this situation. For example, a large part of Lake Vänern, Sweden (the third largest lake in Europe) has been blacklisted, but the situation has improved by changing the water purification procedures of the pulp mills.
The paramount problem in Scandinavia and elsewhere has been airborne acid rains (SO2) from industries using fossil fuels as a source of energy. For example, large parts of southern Norway and southwestern Sweden have become almost devoid of fish due to acid rains originating from Great Britain and the Federal Republic of Germany. Species of fish differ in their response to acidification. Roach (Leuciscus rutilus) and most salmonids are extremely sensitive to decreasing pH. The addition of lime has been tried as an “artificial breathing” agent (1 million Swedish Kroner a year in Sweden), but it is obvious that this problem cannot be solved without some kind of international negotiation. Several charr populations have recently become extinct due to their sensitivity. The primary effect of acidification is apparently on the early reproductive stages.
Pollution by nutrients (especially PO4) from human communities has been most disastrous to densely populated areas in temperate, sub-tropic and tropic areas. Despite this, in arctic or alpine areas with extremely oligotrophic environments, eutrophication by adding nutrients has been promoted as a method of increasing productivity. This is now being studied experimentally in Scandinavia as well as in Canada (Milbrink and Holmgren, 1981; Schindler and Fee, 1974).
In some cases, “nutritional pollution” has been successfully treated. In all these cases, the construction of purification plants has resulted in significant improvement in water quality. In Lake Washington (U.S.A.), the spawning run of Pacific salmon species is now restored and natural spawning is occurring. Likewise, there is now a natural recolonization of native species in the lower Thames in England. The construction of purification plants in all communities around Lake Malaren in Sweden has made it possible to reintroduce Baltic salmon and sea-run brown trout into a stream that discharges through Stockholm.
Hydroelectric power development has also affected the fish populations in impounded lakes and the river stretches between the power plants. An important observation in Scandinavian waters is that there are sub-populations of salmonids that are affected in different ways. For instance, brown trout is often represented by different populations in different sites within a stream, apparently with little interbreeding. During the construction of dams the populations spawning in the outlet area become extinct. Arctic charr and whitefish are often represented by two or more discrete populations which are affected in different ways. On the whole, pelagic populations are less affected than benthic ones.
There are only a few rivers remaining in Sweden which support natural spawning of salmon. In the Baltic the majority of salmon originate from Swedish hatcheries. The smolts released have been carefully selected, so as to represent the original native stocks. There is evidence for genetic change, however, mainly toward smaller size and earlier homing (earlier maturity), this is probably an effect of overfishing.
The landlocked salmon of Lake Vänern are represented by two sub-populations which do not seem to interbreed. These are: the “Klarälven” salmon, which spend three years in the River Klaralven and return for spawning after three to four years, and the “Gullspångsälven” salmon which spend two years in the River Gullspångsälven and return from the lake after four to five years. The growth rates of the two populations are strikingly different. Tagging experiments with reared smolt have shown that after 40 months in the lake the salmon from Gullspångsälven attain a weight of about 5 kg, whereas the salmon from Klaralven reach only 2.5 kg. They are also segregated spatially during their life in the lake, suggesting an inherent capacity for both growth and habitat selection.
Both populations are endangered (Figure 1), mainly by hydroelectric development. The fastgrowing population from the River Gullspångsälven is very close to extinction, but efforts are being made to save it, both by hatchery operations and introductions into other lakes.
Overfishing for Baltic salmon has resulted in slower growth, smaller size at maturity and early homing. Very similar results were obtained by Gwahaba (1973) in the case of Tilapia nilotica in Lake George, Uganda (Figure 2). He claimed that the restricted size of nursery zones, fish killing storms and predation have been factors in the decreased propagation, but an increased commercial fishery appeared to be the main reason for the changes.
Turner (1976) provides another example of ecosystem change by intensive fishing on Lake Malawi cichlids. Whereas at the beginning the catch was dominated by large species, small species now predominate. This example dramatically shows the effects of “overfishing” on an entire fish community. It is predictable that in the case of continuous fishing with nets of unsuitable meshes, the larger species may be threatened with extinction, in part, because the more abundant smaller species can outcompete the small stages of the former large ones.
Similar observations were made comparing discrete populations of Arctic charr which are subjected to different exploitation rates. Figure 3 illustrates that the esteemed fast-growing population matures earlier than the slow-growing, unexploited one (Nilsson and Filipsson, 1971). In this case the difference is no doubt genetically determined (Nyman, 1972).
Since the early 1970's much information has accumulated on the serious consequences of introductions of exotic (foreign) fish on native ichthyofauna (e.g., Walford and Wicklund, 1973). For example, many formerly fishless, deglaciated lakes were stocked by the Lappish people and European settlers in Scandinavia and in North America by the Indians and European settlers. As hatchery technology has improved, the practice has spread.
Figure 1. Catches of land-locked salmon in Lake Vänern, Sweden, 1881–1977. From Nilsson, 1979.
Figure 2. The percentage of females of Tilapia nilotica with mature ovaries at different lengths. The maturity curve for the 1971–72 survey (O) is compared with the one drawn earlier by Fry and Kimsey (1960), (•).
Figure 3. Differences in age and length of two population of Arctic charr at maturity in Lake Ovre Björkvattnet, Sweden. The ordinary charr are faster growing and heavily exploited. From Nilsson and Filipsson, 1971.
The rate of such introductions has greatly increased in recent years. According to Everett (1973) rainbow trout (Salmo gairdneri) was introduced to the Lake Titicaca Basin in Peru and Bolivia in 1942. In the 1950's the following trout species were also introduced: S. trutta, Salvelinus fontinalis, S. namaycush and the atherinid Basilichthys bonariensis (Villwock, 1972). By 1972 two of the most valuable (largest) species of the endemic genus Orestias in Lake Titicaca became practically extinct by predation or food competition from the trout species. Sporozoon parasites which were introduced passively together with the exotic species account for the dramatic population declines in the majority of endemic species of Orestias (Frey, 1975).
Another well known example is that of Lake Lanao on the island of Mindanao in the Philippines. According to Villaluz (1966: cited by Frey, 1969) personnel associated with the State University of Marawi City on Mindanao introduced Clarias batrachus (Siluridae), Ophiocephalus striatus (Ophiocephalidae) and Tilapia mossambica (Cichlidae). Even more serious was the accidental introduction of Glossogobius giurus (Gobiidae) at the same time. According to Frey at least some benthic and pelagic species of the endemic genus Barbodes (Puntius) have become rare as evidenced by their virtual disappearance from the local fish markets.
Both these species groups, the endemic Orestias in Lake Titicaca, as well as the Barbodes in Lake Lanao, were the principal economic and nutrition resources of the local residents. These cannot be replaced by the introduced species (Villwock, 1972) because the introduced species are not readily accepted by the local population. The effects on the genetic resources of the endemic species of Orestias or Barbodes are irreversible. Some species are evidently extinct and even those which may have survived have probably been diminished to one or a few small, disjunct populations.
Game fishermen and aquarists have been responsible for numerous cases of introductions which have led to detrimental changes in the native fauna. Studies in Scotland and Scandinavia have shown that native species may become extinct by ecological competition with exotic ones. For example, Svardson (1979) described the effect of introductions of species of whitefish (Coregonus) and consequent extinction of the Salvelinus alpinus complex in Scandinavia. As a result of competition between the two forms shown in Figure 4 the size at maturity, age, weight and gill-net catch of charr decreased continuously resulting in its eventual extinction around 1965 (Nilsson, 1967).
Hybridization between native and related, introduced species (Svärdson 1979; Moyle, 1976) also has serious consequences on genetic resources. In the second half of the 19th century different trout species were introduced into California, confusing the already complex situation in the native trout fauna (Hoopaugh, 1974: cited by Moyle, 1976). Similar events have been reported by Moyle (1976) for species of chub (Gila) and sticklebacks (Gasterosteus). The most frequent result is swamping of the genome of the native species. The large number of hybridization events reported from Florida were caused by aquarists or the aquarium fish industry (Courtenay, et al., 1974). Twenty exotic species and five hybrid populations have established themselves as breeding populations.
Moyle (1974) reported on fish introductions in California for weed and insect control. In addition to several killifish and two species of Tilapia (T. mossambica, T. zilli), the mississippi silverside (Menidia audous, Atherinidae) was introduced illegally in 1967. Since then it has become the most abundant and widespread species. The Tilapia as well as Menidia have affected the relative abundance of the different fish and the total number of zooplankton organisms to which native fish were highly adapted, although it is difficult to predict the ultimate consequences for the native species.
Both food competition and predation on the larvae by exotics of native species have often led to extinction. Within less than a decade, mosquito fishes (Gambusia, Mollienisia) introduced into southern European countries and North Africa were responsible for the decline of the native Aphanius sp. population (Cyprinodontidae) (Villwock, 1977 and unpublished data). Similarly, Cyprinodon in the Colorado River system is endangered by direct competition of introduced live-bearing forms.
Figure 4. Decrease in the gillnet catch of charr (Salvelinus alpinus) as the catch of the introduced whitefish (Coregonus sp.) increased, Lake Västansjö, North Sweden. From Svärdson, 1976.
Accidental introductions have occurred in several ways. Among these are uncleaned gillnets (“egg pollution”) and the escape of bait (Johannes and Larkin, 1961). Another danger is the escape of stocked and exotic species from aquaculture ponds, for instance by flooding during rainy seasons. Introductions have also occurred via canals or tunnels used for shipping and power plants. The most famous case is that of the landlocked sea lamprey (Petromyzon marinus) which entered the Upper Great Lakes through the Welland Canal. Similar movements have occurred into the Mediterranean Sea since the opening of the Suez Canal.
Many introductions have been economically or aesthetically successful. For example, most species of North American salmon, as well as the black bass, have been introduced into European fresh waters, and several Oncorhynchus species have been introduced into the Northern Russian marine waters and to the Baltic. Kokanee (Oncorhynchus nerka) has been introduced into Scandinavian freshwaters. Although most of these introductions have failed, black bass in Germany and rainbow trout in Great Britain have been successfully established. Brown trout and carp from Europe have been introduced into North America, New Zealand and most alpine areas where Europeans have settled. The introductions of lake trout (Salvelinus namaycush), which began in Fennoscandia in the 1960's, seem to have been successful in some Swedish and Finnish waters.
Little is known about the impact of these introductions on the genomes of the species, but where intense culturing and domestication is practised, for instance in rainbow trout, genetic change is inevitable. The same should be true for brown trout in Europe and elsewhere.
A very intense study on the effects of transplanting Coregonus and Salvelinus species has been carried out in Sweden and Italy (Svärdson, 1979; Nilsson, 1978; Berg and Grimaldi, 1966; Nyman, 1972). Briefly it has been shown that introductions of “exotics” have four possible results:
extinction of “ecological homologues”. The most drastic example has been the extinction of Arctic charr (Salvelinus alpinus complex) by introductions of certain whitefish stocks;
hybridization with concomitantly profound effects of the genetics of the original fish population;
failure of the introduction, in part, because of competition from established resident species;
coexistance, which means that the introduced species has found a “vacant niche” in the community, with an interactive niche segregation as an obvious result.
Figure 5 illustrates a common result of the introduction of species in northern Scandinavia and their impact on the zooplankton community. Apparently the grazing of zooplankton plays a great part, the most efficient zooplankton feeders becoming dominant (Nilsson and Pejler, 1973; Svärdson, 1976).
Introductions of exotic prey species is now a common practice, especially in North America and Scandinavia. In Canada and Sweden, for instance, introductions of glacial relicts (especially the crustacean Mysis relicta) have been practised for several decades (Fürst et al., 1978). Such introductions have on the whole seemed successful. In many cases, however, a strong impact on the native zooplankton community has been observed.
Artificial selection is usually associated with desirable genetic changes. However, in many cases the outcomes of artificial selection may be neither desirable nor predicted. This is especially true when brood stock selection and management lead to loss of genetic variation. By their very nature, specific examples of genetic impoverishment due to such practices are not generally available. Nevertheless, there is sufficient background knowledge to warrant caution. For example, genetically based performance under one set of conditions (i.e., hatchery) may not be correlated with performance under a different set of conditions (stream, lake, natural area). If the goal is to release stock in a different environment from that where they are bred, then the brood stock selection practices must be designed to avoid unconscious selection and inbreeding.
 |  |
| Heterocope saliens Daphnia longispina Polyartemia | Heterocope saliens Daphnia longispina |
 |  |
| Heterocope saliens Daphnia galeata | Heterocope appendiculata Daphnia cristata |
Figure 5. Model of the “dimensions” of the niches of brown trout (Salmo trutta), Arctic charr (Salvelinus alpinus) and whitefish (Coregonus sp.) in allopatry and sympatry, and the dominant species of zooplankton. (After Nilsson and Pejler, 1973)
In this regard, the accessibility of brood stock to human managers does not ensure its genetic superiority. For example, it has been shown that large carp that are chosen from a population of identically aged individuals do not necessarily represent superior genotypes (Wohlfarth and Moav, 1969). Such individuals may represent an exposure to a favourable set of environmental conditions which magnify a small, initial, non-genetic, size differential into a large one. The use of a brood stock composed of such individuals would lead to little or no genetic progress, and could, if the number of selected individuals is small, lead to inbreeding and genetic drift (Section 3). Moreover, large individuals could represent behaviourally aggressive phenotypes under genetic control. The choice of such individuals could lead to an undesirable general level of agressiveness in the population.
In some cases brood stock selected from culture units containing mixed year classes (as is probable in undrained tropical ponds) probably represent a mixture of both superior and inferior genotypes. Consequently little genetic progress, and even erosion, can be expected by the haphazard selection of brood stock. For example, in the case where only gravid females from undrained, net harvested ponds are used for brood stock, they may represent both large young animals (superior genotype) and large older ones (inferior genotype). In addition, it may be that genetically inferior (i.e., slow growing but “net avoiding” males) are reproducing more often with these females.
It is in part for reasons such as these that Tilapia pond culture has failed in some tropical areas. In harvesting Tilapia ponds it is not uncommon to select the large fish and to return the small ones. This procedure favours the reproduction of small and slow-growing fish, and it has been shown experimentally that such selection can produce a genetic shift in the population toward genetically smaller (and less desirable) fish. The problem in Tilapia is compounded by mouth breeding and defence of young until they reach a size safe from cannibalism. The result is overpopulated ponds with stunted individuals. Cannibalism can be genetically and economically advantageous in some aquaculture systems.
Domestication is defined as a genetic selection process mediated by geographic and reproductive isolation, inbreeding, and small population size to produce profound and desirable evolutionary changes in a genetic stock. Domestication presumably involves at least the ability to live under artificial conditions through most parts of the life cycle. It further would usually be expected to include genetic adaptation such as to crowding, handling, and artificial diets. Genetic improvement, in the sense of enhanced growth, modification of body form and some loss of fear-flight behaviour are also usually assumed. The anthropological aspects of this are well documented (Zeuner, 1963), but not much scientific scrutiny has been directed at the domestication process itself because of the limited opportunity to study, document, and control the phenotypic and genotypic changes that occur in the domestication of livestock, poultry, and companion animals. The opposite is true in aquaculture.
There are probably a number of reasons why man has not domesticated as many fishes as he has mammals, birds and plants. Because of the difficulties of transporting live fish from place to place over land, domestication could only have proceeded in sedentary cultures. There is indeed evidence of fish culture in Sumeria contemporaneous with very early stages of agricultural development, pre-dating the reported beginnings of aquaculture in China in 600 B.C.
It is also likely that attempts to apply the criteria for selection learned in animal and plant breeding to fish or to cold-blooded vertebrates generally were counter-productive. Whatever the reason, and with the notable exception of the goldfish, ornamental carp, and perhaps the common carp, few fish could be considered domesticated even though some strains of trout, for example, are clearly much more adapted to hatchery conditions than their wild counterparts (Moyle, 1969; Hines, 1976). Other species like the Chinese carps, Indian carps, Tilapia sp. and American catfishes are becoming domesticated.
As pointed out in Section 5.4 culture conditions themselves, along with the conscious selection imposed by man, will foster domestication. By and large the outcomes of this process are not only desired but predictable. Workers might be cautioned, however, that, history rarely reports failures in such attempts. In this regard researchers, culturists and breeders should develop the perspective that these efforts represent, in many cases, the initial steps of incipient domestication of a new animal group. Their scientific scrutiny should be designed to monitor and document this process.
It is recognized that in view of the inevitable operation of natural selection in culture environments (Section 5.4) that propagation without domestication is, perhaps, impossible. However, in some circumstances breeding without genetic change is the goal, as in the case of stock enhancement programmes of natural fisheries, in which it is desirable to preserve the inherent, undomesticated genotypes. Thus care must be taken to avoid the development of traits that have historically charracterized domestication (e.g., loss of awareness of predators, agressivity, irritability). Such losses may prove useful in brood stock management because these traits result in increased accessibility and tractability to human management. However, development of such traits will be counterproductive to the goals of stock enhancement of natural fisheries.
In some cases of stock enhancement, behavioural genetic change under domestication is useful and desirable. For example, impounding of areas in coastal reclamation projects or inland reservoirs can create vast areas of semi-natural aquatic habitat useful for aquaculture. In such situations, the loss of migratory and other inappropriate social behaviours may be desirable.
It might be expected that the scientific knowledge now available will result in relatively rapid domestication of aquatic species as suggested by Moav et al. (1978). Techniques such as gynogenesis are likely to facilitate the domestication process dramatically. The normally high fecundity of fish species has been cited as contributing to rapid domestication. Whether or not this happens, it is surely essential, as was pointed out in the recommendations of the 1976 FAO Technical Conference on Aquaculture held in Kyoto, that fish breeders provide very careful documentation of the breeding history of strains undergoing domestication. This documentation, as now recognized by plant and mammal breeders, will be of great importance when it later becomes desirable to regain genes, for example, for disease resistance, which have been lost in the domestication process. It has also been proposed (Malecha et al., 1980) that there is an opportunity in the domestication of aquatic species to study the early stages of the process of domestication itself, in a way no longer possible with existing livestock.
