¿ìè¶ÌÊÓƵ

NORWAY is breeding a super salmon. The aim is to raise fish that are
perfectly adapted to life in cages, be they in the frozen reaches of Finnmark,
in the Arctic Circle, or in the warmer waters of Rogaland 1000 kilometres
south. The ideal salmon would tolerate sharing its cage with hundreds of
similar fish yet show no signs of stress. It would resist disease and grow
fast, efficiently converting its food into the desired balance of flesh
and fat. It would not mature sexually until it is large enough to be slaughtered
and sold. Norway has a prototype of this super salmon, which is already
transforming the country’s fish-farming industry. But as scientists press
on in their search for a still better salmon, conservationists fear that
such a fish could spell disaster for the country’s wild Atlantic salmon.

Constant breeding of like with like produces a line in which desirable
genes accumulate. Breeders find it difficult to sort out which of the traits
they wish to build into a fish are inherited and which are a product of
the fish’s environment (see ‘How to build a better fish’, ¿ìè¶ÌÊÓƵ,
22 April), but they are slowly working towards fish with a uniform collection
of the genes that they want.

Norway is unique in having a national breeding programme for salmon.
So far, it has succeeded only in producing fish that grow faster, but that
is a great success for fish farmers. Each year, the new generation of farmed
salmon grows faster than its predecessor by 10 to 12 per cent. Trygve Gjedrem,
who heads the breeding programme at the Institute of Aquaculture in As,
near Oslo, says: ‘It’s fantastic. If growth continues at its present rate,
it will double in 18 years. It is much more rapid than for farm animals.’

The success of the breeding programme has made fish farming a boom industry.
Just five years ago, Norway produced 17 000 tonnes of farmed salmon. This
year, the Norwegian Fish Farmers’ Association expects the farms to produce
80 000 tonnes. There are now 747 fish farms along the Norwegian coast and,
as more people see the possibility of making a quick krone, applications
for more farm licences flood in.

Salmon farming is good news for the treasury, too. More than 90 per
cent of farmed salmon and rainbow trout is destined for export. In 1987,
salmon and trout exports earned Norway $320 million. ‘Aquaculture is great
for coastal development,’ Gjedrem says. ‘For the first time, people are
moving out instead of inland.’ The rapid rise of the industry, however,
worries conservationists concerned for the future of Norwegian wild Atlantic
salmon.

Those fears centre on the growing number of farmed salmon that escape
from sea cages and find their way to Norwegian rivers. About 400 of the
country’s rivers contain wild salmon. The populations, or stocks, of salmon
in each river are genetically distinct. Indeed, one river may contain two
or three different stocks, each spawning in a particular stretch of the
river.

Once in a river, the escaped fish compete with wild salmon for scarce
food and spawning grounds. Diseases transmitted to farmed fish, for example
through the import of infected eggs or fry, may be new to wild salmon.

According to Lars Hansen, a senior researcher at the Directorate for
Nature Management, in Trondheim, the most recent catch statistics from the
Norwegian commercial fishery show that 20 per cent of the fish caught this
year had escaped from fish farms. Dagfinn Gausen, also at the directorate,
has been recording the number of farmed fish that invade Norwegian rivers.
The register for 1987 shows that 23 of 54 rivers across six counties contained
reared salmon. Of 615 salmon examined, 83 had escaped from farms.

Rivers vary greatly in the numbers of escape fish that they harbour.
The closer a river is to a fish farm the more likely it is to be invaded.
According to Gausen, samples taken from the River Oselva, for example, suggest
that the proportion of wild fish has plummeted to a mere 20 per cent; 42
per cent of fish sampled from the River Etnelva are of reared origin. In
1987, 15 to 20 per cent of the fish examined from 54 Norwegian rivers were
escapees.

Life on a fish farm is very different from life in the wild. Atlantic
salmon spawn in fresh water and stay there for at least a year and perhaps
as long as eight years. These juvenile fish, growing in fresh water, are
known as parr. Before embarking on the next stage of their life cycle, in
which they migrate to sea to feed and mature, the parr undergo certain physiological
and behavioural changes that equip them for a marine environment. They develop
specialised ‘chloride cells’ on their gills that actively excrete sodium
salts. These cells proliferate before migration. Their kidneys virtually
stop producing urine and the fish drink more to compensate for the dehydrating
effects of salt water. Pigments in the retina change to enable the fish
to see in the bluer water. The smoky-blue markings on the body fade and
the fish turn to silver-grey. They are then known as smolts.

Smolts may travel thousands of kilometres at sea to reach their feeding
grounds. Most remain at sea for two or three winters, although some have
been known to stay for as long as five years. Then, with extraordinary precision,
the fish make the long journey back to their natal rivers to spawn. Once
in fresh water, they virtually stop feeding yet may swim more than 30 kilometres
a day, leaping through waterfalls and other obstacles to reach their particular
spawning grounds in the head waters.

A regulated life

Farmed Norwegian salmon, in contrast, begin life in fresh-water tanks
at one of two national breeding centres. All fish at the centres are descended
from 40 stocks of wild salmon. In the mid-1960s, anglers and river managers
helped biologists to collect eggs over a four-year period from the 40 stocks,
from rivers all over Norway. From the fish collected in each of the four
years, known as a year-class, breeders reared full-sib and half-sib families.
By 1975, the fish had matured, and breeders made the first selection for
parents of their would-be super salmon. The criteria for selection today
are fast growth and good survival, or resistance to disease. Breeders select
fish that mature late so that the animals devote their energy to producing
flesh rather than eggs or sperm. The quality and colour of meat are also
important in choosing potential parents.

Each year, the two breeding centres send smolts from 150 families to
five test stations in different parts of Norway, chosen to provide different
environmental conditions, such as water temperature and day length. The
test stations nurture the fish for two years, during which researchers record
how fast they grow and mature, and their mortality at various stages. The
results are fed into a computer and the biologists at the breeding centre
then select the top 10 families within that year-class, and the largest
fish in each family, to provide parents for the next generation of salmon.
Because detailed records exist for every fish since the beginning of the
breeding programme, breeders can trace the pedigrees of the fish they select
and so avoid inbreeding.

In this way, the farmed fish are consistently selected for traits of
commercial importance; 1987 marked the fourth generation to be artificially
selected. The centres grow the fish to smolt stage and supply more than
200 million to fish farms throughout Norway. Demand for the selected smolts
outstrips the centres’ capacity to produce them. They make up the shortfall
by supplying ‘multiplying stations’ and smolt producers with some 40 cubic
metres of eggs (about 200 million eggs) each year.

Once in a sea cage, the fish are fed concentrated foods, tailored to
their patterns of growth and weight. They are given antibiotics and fungicides
to cure bacterial diseases and fungal infections. Territorial behaviour
– essential for survival in the wild but a nuisance for fish farmers – is
fast disappearing. After only three generations, farmed salmon tolerate
more than 10 times the stocking rate that their wild ancestors could. Gjedrem
is proud to say: ‘It’s a very different fish we have today than when we
started. They are not so afraid when humans approach them. They are not
aggressive, but calm.’ But breeders are still ignorant about whether these
are genetic or behavioural changes.

At first glance, the possibility that escaped reared fish might succeed
in the wild seems remote. Domestication makes some wild instincts redundant.
For example, one facet of territorial behaviour in the wild is the incessant
competition for food. Fish farmers overcome this by supplying their salmon
with more than enough food. Another facet of territoriality, competing for
spawning areas, is of little use to farm fish because they are slaughtered
prior to mating.

Hansen and colleagues from the Directorate for Nature Management and
the University of Oslo showed that fish escaping late in life from farms
become ‘homeless’; after feeding at sea, they return to the vicinity of
the sea cage from which they escaped, but they lose the precision that is
characteristic of their wild counterparts. The result is that they enter
rivers at random to spawn. Salmon biologists still do not know exactly how
salmon home to their natal rivers. However, Hansen and others have experimental
evidence that homing relies on a sequential imprint of sensory cues, which
the smolts pick up as they leave their natal rivers. Odours, sights, water
temperature and other sensory markers probably combine to provide the imprint.
Hansen believes that the smolts record the cues, as if on a video camera,
and then replay the recording in reverse when they are ready to return to
their natal rivers. Farmed fish receive no opportunity for such sequential
imprinting and so late escapees do not ‘home’ to any particular river on
their way back from the sea.

Although biologists have no means of measuring ‘wildness’, some suspect
that traits which contribute to wildness are weakened with each new generation
of farmed fish. Could such pampered, domesticated fish have any chance of
successfully competing in the wild? For most salmon rivers in Norway, Gjedrem
thinks not. ‘In rivers with a large enough population to spawn successfully,
I see very little problem,’ he said. Unfortunately, most of the rivers have
small populations. According to Gausen, half the salmon rivers of Norway
produce a catch of just 50 fish per year.

Experiments with other fish related to salmon bear out the danger to
such small wild stocks. In a report for Britain’s Nature Conservancy Council
on the genetic impact of farmed Atlantic salmon on wild populations, Peter
Maitland, a fisheries consultant based at Stirling in Scotland, mentions
recent American research on the relative reproductive success of farmed
and wild steelhead trout. Although hatchery trout succeeded in producing
only about 30 per cent of the number of offspring that wild fish produced,
the high density of hatchery fish resulted in hatchery spawners outnumbering
the wild ones by four-and-a-half to one. Another experimenter in Montana,
in the US, found that populations of wild rainbow trout increased by almost
900 per cent four years after a river stopped being stocked with rainbow
trout from a hatchery.

Just moving wild stocks from one river to another can have disastrous
effects. Researchers in the Soviet Union, believing the chum salmon in the
River Kalininka to be superior to that in the River Naiba, transferred more
than 350 million eggs of the Kalininka stock to the Naiba between 1964 and
1971. According to Maitland’s report, only 10 to 20 per cent of the expected
number of Kalininka fish returned initially and the number of Naiba fish
returning to the river also fell. The population dropped from 650 000 spawners
in 1968 to about 35 000 in 1980. By 1985, Naiba salmon were virtually extinct.
‘In some places, and in some years,’ Maitland says, ‘the input of farmed
fish may be so enormous that the sheer number could easily deplete food
resources, particularly if in the same year wild fish had to overcome naturally
occurring environmental problems.’

The arrival of the prototype super fish is not the only threat to wild
salmon. Wild Norwegian salmon have already suffered a shattering blow from
Gyrodactylus salaris, a parasite that first appeared on fish farms and has
since all but eradicated some stocks of wild salmon. The parasite is believed
to have been imported from an infested hatchery in Sweden. It first appeared
in the wild in 1975 when parr from the River Lakselva, in northern Norway,
were found to be heavily infested. Two years later, there were hardly any
parr left in the Lakselva whose name, ironically, means ‘salmon river’.

By 1982, the parasite had spread to seven more rivers. The latest count
is 33 infested rivers. In the five best-studied rivers, the number of salmon
parr is fast approaching zero. Experts estimate that the commercial fisheries
lose between 260 and 520 tonnes of salmon to the parasite each year and
they expect the annual catch to drop still further.

Between 1980 and 1982, researchers from the Directorate for Nature Management
examined 200 salmon rivers. They made smaller studies of infested rivers
over the next four years. They found that all but three of these rivers
had been stocked with fish from hatcheries infested with the parasite. One
of the three may have received fish from Sweden, while the other two may
have been infected by anglers with contaminated equipment or by fish farmers
transporting infested salmon smolts from hatcheries to fish farms and changing
their water in the rivers. The parasite is not a great problem for fish
farmers as they can kill the parasite by adding small amounts of formalin
to the tank water. The only solution in the wild is to remove and treat
affected fish, treat the entire river with vast quantities of a potent toxin
of plant origin, called rotenone, and then restock with healthy wild fish.
The procedure is expensive and not suitable for all rivers.

Acid rain and snow, too, have taken their toll. According to Gausen,
about 25 stocks ‘are nearly or entirely extinct in southern and western
sections of the country, which are heavily affected by acid precipitation’.
On top of this is the politically sensitive issue of overfishing both by
Norway’s commercial fishery and by fisheries in the open sea off Greenland
and the Faeroes. Salmon from rivers in Europe and North America feed around
Greenland where fishermen take enormous catches with long drift nets. In
1971, the total catch from waters off Greenland was 2689 tonnes. As in the
commercial fishery in Norwegian home waters, the catch includes both stocks
with a large number of spawners and small, already more vulnerable stocks.

Meanwhile, the fish farming industry moves from strength to strength.
A surplus of about 6 million smolts was produced in Norway last year. Farmers
want to release excess smolts at sea and then catch them when they return
to the area, suitably fattened. They estimate that even if only 5 per cent
of the fish returned, the catch would be large enough to make the exercise
worthwhile. Such a practice would increase the pressure on the smaller wild
stocks that already suffer from the undiscriminating nets of the commercial
fishery. Research suggests that the fish not caught at sea would end up
in rivers and, again, might compete with wild stocks.

Of greater importance for the future is the risk that if farmed salmon,
bred for a set of genes that fits them for farm life, mate with local stocks
of fish, then the wild fish might lose their adaptations to local conditions.
The wild salmon have accumulated genes that suit them to local conditions,
as a result of natural selection, in the same way that farmed fish have
accumulated genes as a result of unnatural selection by the breeder. Interbreeding
with the wild populations might create a genetic uniformity that spreads
far beyond the confines of the fish farm.

Kjetil Hindar, of the Directorate for Nature Management, like many other
biologists in countries that have Atlantic salmon, has been studying genetic
diversity in wild and farmed fish by looking at variations in the chemical
composition of certain proteins. Using the technique of gel electrophoresis,
it is possible to identify different forms of a particular protein, usually
an enzyme, in blood and muscle tissue from fish. The different forms reflect
differences within the genes that govern the synthesis of these proteins.
The technique relies on the fact that certain amino acids are electrically
charged. The quantity and distribution of charged amino acids affect the
rate at which proteins migrate across a gel that is exposed to an electric
field.

Because different stocks may have different forms of the same protein,
biologists can, if they choose a wide enough range of protein markers, create
a statistical picture of variation between stocks. Hindar has confirmed
that stocks even within the same river but which spawn at different sites
are genetically distinct.

Part of such genetic differences will come from natural selection since
the end of the most recent Ice Age. Each stock has adapted to the peculiarities
of water temperature, rate of flow, and the availability of food and spawning
grounds. A particular combination of those environmental factors will favour
the survival of fish with a genetic make-up best suited for survival under
those specific conditions.

Norway’s breeding programme selects for traits of commercial value.
At present, the programme avoids matings between fish in different year-classes,
which would make them still more uniform, but, Gjedrem says, this may change
in the future. So far, Hindar has shown that genetic variability within
reared fish is as great as that in the wild. However, because populations
of fish in single rivers are highly adapted, they are still threatened by
the greater genetic uniformity of farmed fish. In his booklet on the genetics
of salmon stocks, Noel Wilkins, of the National University of Ireland, in
Galway, writes: ‘From our knowledge of quantitative traits in domestic livestock
we know that constant mating of like with like may build up a selected line
in which the desired genes will accumulate at the many loci governing the
selected trait. Dilution of the local stock with strays or introductions
may disrupt the process of local adaptation by disrupting the accumulation
of suitable genotypes’ (Salmon Stocks: A Genetic Perspective, Atlantic Salmon
Trust). The problem may worsen when breeders mate fish from the different
year-classes of farmed fish. Because so many of Norway’s rivers have small
stocks, Gausen says, the ‘wild salmon are very vulnerable to genetic change’.
If enough interbreeding took place, wild salmon could become indistinguishable
from farmed salmon. He believes that: ‘The loss of locally adapted traits
and genetic combinations – co-adapted gene complexes – will destroy wild
salmon as we know them today.’

Moreover, one large population would be more vulnerable to diseases
and harmful changes to the environment than would many, distinct populations.
Should breeders ever need it, they would have nowhere to turn for fresh
genetic material for new desired characteristics. Fortunately, the threat
to wild salmon prompted the Directorate for Nature Management to create
a sperm bank for wild stocks in 1986. Today, the bank has sperm samples
from 857 salmon from 69 stocks.

Despite the potentially devastating effects of genetic erosion and the
increasing number of escaped farmed fish found in Norwegian rivers, little
research has been carried out on the interaction between wild and farmed
fish. Hansen and his colleagues want to know the proportion of escaped farmed
fish that spawn in Norwegian rivers. How many breed successfully? What happens
to a wild population when it is outnumbered by invaders from fish farms?
What are the genetic effects of interbreeding between reared and wild fish?
Salmon biologists in all the salmon-producing countries are asking the same
questions. No one has the answers. Yet biologists sense danger. John Thorpe,
of the Department of Agriculture and Fisheries for Scotland, said: ‘Introgression
of wild by farmed salmon is a risk that many people are worried about at
the moment – it’s very much a gut reaction, though, because we have no data.’

One measure that could be taken in the future is to ensure that farmed
fish are infertile. Researchers at the Department of Agriculture and Fisheries
for Scotland and elsewhere have shown that by treating newly fertilised
eggs with heat, they can produce triploid fish. These fish have an extra
set of chromosomes and are infertile. Although this would not solve the
problem of competition with small wild stocks it would protect against genetic
erosion of wild stocks through interbreeding.

The lack of research has provided the fish-farming industry with a standard
response when biologists and conservationists voice their fears. It simply
says there is no evidence to suggest that farmed fish affect wild stocks.
‘They are correct,’ says Hindar, ‘but if we wait for data it may be too
±ô²¹³Ù±ð.’

* * *

Legal aid for the world’s wild salmon

LAWS are already in the making in Norway to cope with some of the threats
to wild salmon. A Salmon Act, proposed in 1985 but not yet passed by parliament,
will make it illegal to move wild stocks from river to river. Populations
must be enhanced with fish of the same stock. The aquaculture industry,
however, can still obtain permits to import salmon eggs, fry and smolts.

The council of the North Atlantic Salmon Conservation Organisation (NASCO),
at its annual meeting held in Reykjavik in June 1988, considered developing
an international code of practice or a series of recommendations to minimise
the impact of aquaculture on wild stocks. It is also to review existing
codes of practice in its nine member parties (Canada, Denmark, the EEC,
Finland, Iceland, Norway, Sweden, the US and the USSR).

A Federal Code of Regulation in the US prohibits the introduction of
salmonids with the virus that causes whirling disease. But each state varies
in its ability to detect fish diseases and in its regulations on fish introductions.

Canada’s Fisheries Act of 1977 covers importation and transfers between
provinces of live and dead products of cultured salmonids and eggs of wild
salmonids. Permits for transfers are granted only if the products originate
from sources that meet specified health standards. Canada’s Department of
Fisheries and Oceans has established Introduction and Transfer Committees
to advise on the implications of importations and transfers.

The Norwegian government banned the use of drift nets in its commercial
fishery in 1986. NASCO’s West Greenland Commission agreed at the meeting
in June 1988 to limit its total catch of salmon to 2520 tonnes for the three
years 1988 to 1990. This works out at an average annual catch of 840 tonnes
and must not exceed that figure by more than 10 per cent in any given year.

Sea ranching of Atlantic salmon, in which hatchery smolts are released
directly to the sea and harvested on return at maturity, has so far been
carried out commercially in Iceland and on an experimental basis in Norway
and Scotland. Lars Hansen, overseeing the Norwegian experiment for the Directorate
for Nature Management, said that ranching was profitable at the national
level but not at an individual level. That is because private enterprises
would have to compete with the commercial fishery for fish returning from
sea. Hansen says it could be economic if smolts are released, say, at the
bottom of a fjord and caught in a trap on their return.