¿ìè¶ÌÊÓƵ

On all sides, the world’s fish are beleaguered. Freshwater fish in particular
are losing their habitats, as rivers are dammed, diverted, and generally
dried up. Many fish are over-fished – a problem that has provoked alarming
conflicts, even between different kinds of biologist: fishery managers want
to maximise yield, while conservationists are interested in biological diversity.
Even the idea of ‘sustainable’ yield is less helpful than it may seem, for
although fishermen may assiduously try to control their catch of commercial
species, they may also fish many a ‘minor’ species to extinction in the
‘bycatch’. Huge numbers of fish, in rivers and lakes worldwide, are threatened
by species introduced from elsewhere – Nile perch into Lake Victoria; brown
trout into the Southern Hemisphere; big predatory ‘sports’ fish such as
bass and catfish, into warm waters throughout the world.

In theory, there are two ways to approach conservation: saving species
and saving habitats. In either case, we need to know what exists and where
each kind is living. With fish, as speaker after speaker lamented at the
recent meeting of the Fisheries Society of British Isles at the University
of Lancaster*, fundamental knowledge is sadly lacking. There are 24,000
species listed, but another 100 species of teleost alone, (the principal
fish of bony fish), are described each year. The true total is probably
around 35,000.

Remarkably, some of the new found species are under our very noses.
Peter Miller of the University of Bristol, and P S Economidis from the Aristotle
University of Thessaloniki, recently described a whole new genus of goby
from the west of Greece, the birthplace of biological science. The fish
are gobies; and the genus is Economidichthys, characterised by a peculiar
glandular organ around the anus, which is unique among teleosts and of unknown
function. The genus so far includes two species: E pygmaeus and E trichonis,
which they described only this year, and which is probably Europe’s smallest
species of vertebrate. Female E trichonis are only 27 millimetres long;
the males a little bigger.

Greece has five species of goby in three genera, of which Economidichtys
is one. Presumably, originally, they all arose from one ancestral population
which then was divided as interlinking waterways dried up or were blocked
by shifts of landscape. The first major split, so Miller suggests, may have
occurred in the mid-Miocene epoch, 12 to 14 million years ago, and then
another in the late Miocene, 6.5 million years ago. That is plenty of time
for separated populations to have evolved into separate species.

Indeed, freshwater fish may have a greater tendency than any other animal
to become divided into separate populations, precisely because connecting
waterways can so easily become impassable. But if the separation occurred
only recently, the different populations may not have had time to form new
populations. Thus does fish classification become immensely complicated
– and so correspondingly does conservation. If we set out to save a species,
we should in some cases ideally save many different populations of that
species, each of which may be ecologically and physically distinct. If we
set out to save a habitat, we should ascertain (given that we cannot save
all habitats and are obliged to select) whether the population within is
simply the same as one elsewhere, or is very different and particularly
worthy of protection.

The brown trout of northern Europe make the point. The first invaded
from the south only 13,000 years ago, after the last Ice Age. The original
invaders have also divided into many different populations, which are now
distinct in colour, form (morphology) and way of life. But although anglers
and naturalists have in the past divided Europe’s brown trout into many
different species (about 50 have been recognised since 1758, when modern
systems of classification began) the different types are still sufficiently
similar, genetically, to be able, in theory, to interbreed. Modern biologists
therfore classify all brown trout as one species, Salmo trutta. But ecologists
recognise that the differences between the various brown trout populations
are very significant.

Just how diverse these trout populations are is shown by the studies
of Andrew Ferguson and his colleagues from Queen’s University, Belfast.
They studied the brown trout in Lough Melvin, on the border between Northern
Ireland and the Republic of Ireland. The lough contains three populations,
known to anglers as the gillaroo, the sonaghen and the ferox. They are quite
distinct from each other, both in appearance and way of life. The gillaroo
has big red spots, and big, numerous teeth, for grasping snails and other
prey from the lake bottom. The sonaghen has black spots, fins, and long
gill rakers, with which to filter zooplankton from the mid-waters. Ferox
is a predator, with a wide head to hold the muscles for its powerful jaws,
for gripping the char and perch on which they feed.

Although the three may theoretically interbreed, and as adults they
occupy the same space, the breeding habits of the three in practice keep
them separate. Salmonids (members of the Salmonidae, including salmon and
trout) return to the river of their birth to spawn, and only a few stray
each year. So once a population is established in a particular river, it
remains genetically isolated, as surely as if it lived on an island, even
though the adults may intermingle in a lake or in the sea. Ferguson and
his colleagues have shown that the sonaghen and ferox breed in different
in-flowing rivers while the gillaroo breeds in the out-flowing river, the
Drowes. When the gillaroo hatch, they have to swim upstream to reach Lough
Melvin, where they spend their adult life; which is most unusual among European
salmonids, though many North American rainbow trout populations do the same.

How, though, did Lough Melvin first acquire its three separate Salmo
trutta populations? In theory, the original post-glacial ancestral brown
trout may have entered the lough (from the sea) and then diverged into three
groups. Alternatively, successive waves of trout may have entered at different
times, already differentiated.

In practice, Ferguson and his colleagues’ studies of the genetics of
hundreds of brown trout populations throughout Europe suggest that what
really happened was a combination of the two. Soon after the Ice Age, they
suggest, the ancestral brown trout entered Lough Melvin; and this evolved
into the modern ferox. Later, a second population of brown trout evolved
elsewhere in Europe, which matured earlier than the original type, and hence
had a competitive advantage. A few thousand yeras after the first brown
trout invaded Ireland, this second wave arrived; and differentiated into
the gillaroo and the sonaghen. In the millennia after the Ice Age, the level
of the land in Northern Europe slowly rose, as the weight of ice was removed
from it. As the land rose, lakes such as Melvin became increasingly inaccessible.
After the second population of brown trout arrived, no further natural invasions
were possible.

Clearly, the diversity of trout in Melvin depends upon the rich geological
and biological history of the lake. Their continued survival, however, depends
on good luck. In many Irish waterways other freshwater species have been
introduced by humans. One common introduction is the pike, which is considered
an excellent ‘sport’ fish. Had pike been introduced in Lough Melvin, they
would surely have consumed, and out-competed, the ferox. Another common
introduction is farm-bred brown trout, for ‘re-stocking’. These genetically
generalised fish (largely derived from Denmark) would surely have swamped
the three natives, had they been introduced.

Ferguson’s work at Lough Melvin, and Miller’s and Economidis’s work
in Greece, show that, wherever possible, the fishes’ habitats should be
preserved intact. Each of the two species of Economidichthys has an extremely
limited distribution. E pygmaeus occurs only in two river systems, plus
Lake Trichonis, in southwestern Greece; and the tiny E trichonis lives only
in Lake Trichonis. The waterway where the first known E pygaeus came from
in 1929 (though then called Gobius) has now been drained. We will never
know how many other gobies have disappeared in similar circumstances before
being described. Lough Melvin’s delicate interactions of trout could theoretically
be wrecked by a single ill-considered introduction. Yet it is not easy simply
to create ‘reserves’ elsewhere for fish, as is commonly done for mammals.
If fish are taken from their native river and put into a new one, either
they may not adapt, or if they do, they may in their turn compromise the
ecology of their new home. Many a species nowadays is endangered in its
proper home, but is a menace somewhere else. The one hopeful prospect (suggested
by Peter Maitland, of the Fish Conservation Centre at Stirling, Scotland)
might be to use newly created waterways, such as reservoirs, as reserves
for threatened populations.

The second key lesson, from the work of Ferguson and his colleagues,
is that standard classifications may be less than helpful for serious conservation.
If we set out simply to preserve Salmo trutta, then we may simply produce
some generalised gene-pool, with fish that do not exactly resemble any animal
in the wild, and are not adapted to any particular place. Alternatively,
we might preserve just one or two populations, and lose the enormous diversity
of the rest. It is essential, argues Ferguson, not simply to break away
from standard classification; but to quantify the genetic differences between
populations, so that we know which are most similar, which are most different,
and which (since we probably cannot save everything) are most worthy of
preservation. Direct analysis of DNA, or of different forms of the same
protein from different populations, is enabling such quantification.

Indeed, Gary Meffe of the Savannah River Ecology Laboratory, University
of Georgia, who studies the extremely vulnerable desert fish of the southwest
US, sees conservation as an exercise in conserving genetic variation; both
because genetic variation underpins biodiversity, and because all creatures
need as broad a genetic base as possible, if they are to continue evolving.
If the individuals we conserve from any one species between them contain
only a small fraction of the total species’ gene-pool, then their ability
to adapt to changing circumstances in the future will be severely limited.
In 1988, he and his colleague R C Vrijenhoek proposed mathematical models
to express the ways in which genes might, in principle, be distributed through
a species.

In their ‘Death Valley’ model, Meffe and Vrijenhoek envisaged a species
divided into totally separate populations – such as might well occur in
pupfish, for example, living in streams and pools in the southewestern deserts.
Each population would be likely to be small, and small populations are particularly
likely to lose rare variants (alleles) of genes because the rarities will
be contained within only a few individuals, and these may die before reproducing.
Such loss is called ‘genetic drift’. Natural selection would also act powerfully
upon such populations, as unfavourable genes would also be rapidly selected
out. Such small populations as this, with a limited range of genes, would
also be particularly prone to in-breeding depression. If two individuals
who each carry a deleterious allele mate, their offspring are liable to
carry a double-dose of that allele, and which means they are likely to suffer
some disability. Small populations are liable to persist for a long time
only if they have been purged of deleterious alleles.

At the other end of the spectrum, is the ‘Stream Hierarchy’ model. This
envisages that different fish populations are separated for much of the
time, but that they come into contact now and again; for example in times
of flood. So there is a continual, though interrupted, flow of genes between
them.

In practice, the total genetic variation within a species is the combined
total variation within any one population and the variation between populations.
By studying the variations of particular enzymes (allozymes) of different
populations, it is possible to work out whether, in the past, those populations
have been totally separate, Death Valley style; or whether there has been
some gene flow between them. One such study (by Vrijenhoek and his colleagues)
was of the endangered Sonoran topminnow, Poeciliopsis occidentalis, in southern
Arizona and northwestern Mexico. This showed that the general principle
works; and in this particular case, it transpired that the diversity within
colonies accounted for 21.3 per cent of the total variation.

Such exercises as this are not just academic. They can, and should,
profoundly influence conservation strategy. If genetic studies reveal that
populations have in the past been isolated, then they should remain isolated.
Merging of populations would reduce the differences between them – and thereby
reduce an important source of total genetic variation. On the other hand,
populations that have been linked in the past may have come to depend upon
such linkage, as the constant inflow of genes reduces the dangers of inbreeding.
So if two previously linked populations become isolated (for example by
draining some linking waterway) then appropriate numbers of individuals
should be transferred from one population to the other.

Genetic studies based on analyses of proteins or DNA are also essential
to identify populations that contain unique alleles, and are therefore particularly
worthy of conservation. A study in 1989 nof 16 populations of the Pecos
gambusia (Gambisia nobilis) in Texas and New Mexico showed that the most
variable were those of the Toyah Creed drainage; so they were the best ones
to conserve.

It is in the United States that the greatest efforts are made to conserve
fish by breeding them in captivity, and then releasing them. The Dexter
National Fish Hatchery in southwest New Mexico is devoted entirely to rearing
endangered species. At any one time the DNFH maintains 10 to 20 types, of
which at least six have already been restored to their original ranges.

The aim of captive rearing is of course to conserve genetic diversity,
and studies of DNA and of protein variants can show which groups of individuals
are most genetically various, and which, therefore, should be used to ‘found’
a captive population. From 1976 to 1985, the DNFH raised a great many Sonoran
topminnows. But Vrijenhoek and his colleagues showed that the population
from which the captive creatures were bred – they all came from Monkey Spring,
Arizona – contained no detectable genetic variation. They therefore recommended
that the DNFH should start all over again, with some of the much more varied
topminnows from Sharp Spring. And this is what the DNFH did, starting in
1986.

Genetic studies can also be used to monitor captive populations, says
Meffe. After all, they may tend to lose genes by genetic drift, as wild
populations may do. Captivity may also accidently select for particular
traits – for example, for docility or a lowered resistance to disease as
animals are sheltered from selection pressures that would act upon them
in the wild. A study of captive Pecos gambusia showed that after six to
eight generations some rare alleles had indeed been lost.

Does it really matter, though, this genetic diversity? Again, Meffe
cited studies to show that the genetically diverse Sharp Spring topminnows
grew faster, survived better, reproduced faster, and were more stable developmentally
than the inbred Monkey Spring fish. Intriguing, too, is the study by Paul
Leberg, also of the Savannah River Ecology Laboratory. He established two
experimental populations of the eastern mosquitofish, Gambusia holbrooki
– one group founded by pairs of individuals who were siblings, and the other
founded by pairs of unrelated individuals. After three generations the populations
founded by unrelated animals were twice as large as those founded by siblings.
The more successful fish did not breed faster, but they survived better.
It has long been suspected that this might be the case; but this, says Leberg,
is the first formal experimental demonstration.

But in the real, harsh world, are the genetic studies that Meffe argues
are so necessary, actually possible? When populations have already been
devastated, and are on the brink of vanishing, conservationists just have
to do what they can. But, he says, ‘Populations in better condition should
be dealt with at the outset with a distant time-frame in mind, and with
the knowledge that management decisions made today will have genetic impacts
for millennia.’

In some instances, the case for captive breeding is problematical. Most
conservationists oppose plans for captive breeding of the coelacanth, the
extraordinary lobe-finned fish of the Comores Islands in the Mozambique
Channel. There are probably only 300 to 500 left in the wild. They may be
safe enough, even though there are so few, but attempts to pull some out
may well harm those that are left, both directly (as fish may be damaged
by unsuccessful capture) and indirectly (by reducing genetic diversity).
Neither is it known whether coelacanth will breed in captivity.

But the case for captive breeding of America’s endangered desert fish
is clear cut. Rearing in capitivity is in most cases known to be possible;
and several endangered habitats have already been made safe again by altering
drainage schemes or passing laws to prevent drainage, so that there are
places to reintroduce the fish to.

Lake Victoria’s haplochromines are equally worthy candidates for captive
breeding. If it is not done, then scores of them will simply go extinct.
On the other hand, Lake Victoria itself has been so compromised, that it
is hard to know what to do with the captively bred fish, except to keep
them in captivity.

Haplochromines are fish of the genus Haplochromis which with its allies
and the more famous Tilapia, belong to the ubiquitous tropical family, the
cichlids (cichlidae; pronounced sick-lid-ee). They are mostly small (five
centimetres or so) and not obviously prepossessing, but they were, until
recent decades, of enormous economic and social importance. Indeed the haplochromines
accounted for 80 per cent of the fish biomass of Lake Victoria, and the
locals caught them, dried them in the sun and wind, and relied upon them
as a prime source of protein and flavour.

As objects of biological interest, Lake Victoria’s haplochromines were
unsurpassed. They diverged to form about 300 different species, each kind
ecologically and behaviourally separate from the others; some in deep water,
some in the shallows, some feeding on plankton, some on each other’s young,
and so on. No one knows how they achieved this variety. One theory, proposed
by Humphry Greenwood, formerly of the Natural History Museum in London,
is that parts of Lake Victoria were at times divided into separate ponds,
in which different populations could diverge, as did the Greek gobies, Europe’s
brown trout, and the fish described by Meffe’s Death Valley Model. Some
biologists believe, however, that the variety arose in the lake as a whole.

But this ecological and evolutionary contemplation has been brought
to a halt, however. In the late 1950s, Nile perch, Lates were introduced
into Lake Victoria, from the Ugandan shores. Alois Achiend, now of the Lake
Basin Development Authority in Kenya, who at the time was working for the
Ugandans, says that nobody knows how the perch first got into the lake,
or who put them in. However, once it was clear that they were established,
the Ugandan government introduced them deliberately in May 1962; and the
Kenyans followed suit, with 300 fish, in September 1963.

Lake Victoria is on the site of a more ancient lake, which dried up.
In the Miocene, the ancient lake contained Lates, as do many other African
lakes now coexisting with many species of cichlid. But modern Lake Victoria
has never contained the perch, and its haplochromines evolved in its absence.
The Nile perch is one of the world’s supreme predators, which in a 20-year
life can grow to 200 centimetres and a weight of 100 kilograms. Thirty-five
kilograms is commonplace. The perch ate the naive and ill-adapted haplochromines,
and miltiplied. At first, the local fishermen hated them, because they broke
their nets. Now they have stronger nets, with bigger mesh. By the mid 1980s
haplochromines accounted for only 1 per cent of the catch, while the perch
made up 70 per cent.

Off their perch

Economically, the perch is a mixed blessing. It is too big and oily
simply to be sun and wind-dried. It needs smoking, which requires timber,
and local deforestation is a problem. Most local people seem to welcome
it, however. For the first time in their history, the Lake Victorian Kenyans
are now fish exporters – of canned Nile perch fillets.

Biologically, Lates in Lake Victoria have been a disaster. Extensive
surveys from the 1970s onwards by Dutch biologists in the Haplochromine
Ecology Survey Team (HEST), showed that by the mid 1980s almost 200 of the
300 haplochromines species were already extinct. A survey in 1986 by biologists
from the Natural History Museum in London suggested that inshore species
at least, living among the rocks where Nile perch do not care to penetrate,
have survived. But the mass – the offshore species – seem unequivocally
to have gone.

The implications of the Lake Victoria/Nile perch story are enormous.
First, though this is only one of many examples of introduced animals harming
a native ecosystem, it is probably the most spectacular ever. Very few vertebrate
groups are as varied as the haplochromines; never has a vertebrate taxon
been so comprehensively devastated in such a short time. Some biologists
(including Achiend) argue that overfishing would have wiped out at least
some haplochromines even without Nile perch. But none denies that the Lates
have had a tremendous influence.

Secondly, it is clear that the present, Nile perch-based ecology is
not stable. What do the Nile perch live on, now the haplochromines are scarce?
Their own young, is one possibility. It is clear, though, that the lake
is becoming eutrophic (too rich in nutrients) and anoxic, possibly because
the animals that fed upon the plankton – namely the haplochromines – have
been removed, so that the food chain no longer runs, as aquatic ecosystems
should, smoothly from plankton to top predator. Whatever the cause, ecologists
(and economists) can only stand and wait. Sooner or later, the Nile perch
population in its turn seems bound to crash. it is conceivable that the
entire lake, which is the size of Switzerland, could simply die.

This leaves the option of captive breeding. Haplochromines are not spectacular,
and have not engaged the aquarists’ fancy. But about 15 species are known
to be living in aquaria worldwide, notably at the New England Aquarium in
Boston, Massachusetts; at Bielefeld University in West Germany; Leiden University
in Holland; and at the Horniman Museum in London, where Gordon Reid is Keeper
of Natural History. Some of the 15 captive-bred species are probably now
extinct in the wild, and for Reid and his fellow aquarists, they pose several
dilemmas. Clearly, they do not want to stop breeding them; to do so would
be to write off entire species. But the prime reason for captive breeding
is eventually to return the animals to their native habitat, and it is hard
to see how Lake Victoria could ever be restored, or even remotely so.

Then again, there is the matter of logistics. If genetic diversity is
to be maintained then it is vital that the first generation of offspring
should be as large as possible. The theory is simple; each parent passes
on only half of its genes to each offspring, and the only way to retain
all (or the vast majority) of the genes in the founder generation, is to
ensure that they reproduce a lot. But aquaria are of finite size. Horniman
is a public museum, with a responsibility to show a variety of species.
It cannot fill all its tanks with haplochromines. Besides, breeding fish
is a tedious business, once the initial problems have been solved.

So what is to be done? One possibility, discussed by Reid and Chris
Andrews, curator of fish at the London Zoo, is to engage the help of amateur
aquarists, many of whom are at least as good as the professionals in fish
husbandry (See Box).

Finally, biologists, spurred not least by Rosemary Lowe-McConnell, formerly
of the Freshwater Biological Association, who did much of her work on Lake
Victoria, are now turning their attention to the other African lakes, such
as Malawi and Tanganyika, which are endangered for other reasons. Beneath
Lake Malawi, for example, there is oil, which various companies are seeking
to drill. Habitat protection is the ultimate goal of conservation, to which
all other approaches are subsidiary. The African lakes are among the most
diverse and vulnerable of ecosystems, but they are in the heart of an economically
deprived continent with a rapidly expanding population. For conservationists
worldwide, they present the greatest immediate challenge.

Taken all in all, it is clear that conservation is an uphill struggle.
Fish, all 35,000 species and their many variants, demonstrate all the problems
of conservation and their possible solutions, writ large: the lack of fundamental
knowledge; the need to conserve habitats; the need to identify which habitats
are worthiest of conservation, and to define the reasons for such decisions;
and the need, where appropriate, to develop methods of captive rearing,
for it is always in theory possible that wilderness may become available
once again. The FSBI is now beginning formally to address these issues.

Yet still there are many who ask, ‘why bother?’ Peter Miller has one
answer to this: that biodiversity is a source, and that it is incumbent
upon us to preserve resources for future generations. Philip Pister of the
California Department of Fish and Game, at Bishop, states the issue more
bluntly. Whenever he is asked, ‘What use is a pupfish?’, he replies simply,
‘What use are you?’

* * *

Can amateurs conserve rare fish?

If captive breeding is to make a significant contribution to the conservation
of fish, the captive populations must be large. Only then will they retain
a wide diversity of genes. Professionals aquarists, such as Gordon Reid
at Horniman Museum, London, and Chris Andrews at London Zoo, have only limited
space and resources, and wonder whether they should elicit the help of amateurs.
But there are pros and cons:

In favour of amateurs

Fish-breeding requires ‘green fingers’. Many amateur aquarists are extremely
skilled.

Animal breeding of all kinds requires time, commitment, space, and money.
Amateurs collectively lavish these in super-abundance, spending an estimated
7 billion Pounds on their hobby worldwide, and 100 million Pounds in Britain
alone. As Gordon Reid says: ‘What a resource!’

The success of all conservation depends largely on engaging public interest.
Most of that interest, however, is inevitably passive. By helping to breed
endangered animals, some at least of the public could become not simply
interested, but involved. Here is a way to build bridges between scientists
and the world at large.

Most fish kept by amateur (and professional) aquarists are still caught
from the wild. large captive-bred populations could reduce this drain. There
are dangers here, though (see ‘drawbacks’ below).

There are precedents for amateur involvement in other species. Many
hoofed animals, in particular, are now kept in significant numbers on private
ranches, especially in South Africa and the US. So far, these enterprises
have been the prerogative of people who are, effectively, landed gentry.
Fish could be kept for serious purpose by ‘ordinary’ people in ordinary
houses. So, for that matter, could many other kinds of animal, from red-kneed
tarantulas to parakeets.

Potential drawbacks

Not all endangered fish are particularly spectacular. Lake Victoria’s
haplochromines are not the most beautiful. Breeding, once the subtleties
are mastered, is a chore. Amateurs have a tendency to lose interest. On
the other hand, professionals have a tendency to run out of grants, so perhaps
it is tit for tat.

The whole point of conservation breeding, is to maintain a broad genetic
base. But amateur breeders, whether of fish, pigeons, or dogs, tend to try
to ‘improve’ their stock, which implies selection, and deliberate narrowing
of the genetic base. To be truly useful, conservation breeding has to be
carried out according to a strict strategy. How many hobbyists would be
prepared to accept such discipline?

Theoretically, the establishment of breeding populations of endangered
species might encourage a trade which at present does not exist. This in
theory could encourage capture from the wild. In the case of the most relevant
haplochromines, however, this danger does not exist, as they are already
extinct in the wild.

To enlist the help of amateurs might take up a great deal of professional
time. But most of the process of enlistment is organisational, and could
itself be carried out by amateurs. In principle, every hour of a professional’s
time could elicit several hundred hours of valuable help.

On the whole, the case for trying to enlist the help of amateurs seems
far stronger than the case against.

Colin Tudge is a freelance science writer and broadcaster.

‘The Biology and Conservation of Rare Fish: an international symposium’
was held by the Fisheries Society of the British Isles, at the University
of Lancaster, earlier this year.