91av

Shells, stuffed birds, dried fish and pickled reptiles all find their
place in natural history collections. It became fashionable 300 years ago
for European noblemen and rich merchants to accumulate natural curiosities.
Their collections sometimes continued to grow as state museums, incorporating
specimens from voyages of discovery in the 18th century, from colonial and
administrative ventures in the 19th, and from more formal scientific expeditions
in the 20th. They remain with us to this day, inconspicuous behind the public
galleries, but expanded out of all recognition. The Natural History Museum
in London, for instance, cares for some 67 million items housed in myriad
jars, boxes, bottles and drawers.

The pedigree of natural history collections is often revealed by their
homes:in cupboards and cases ofunfashionable design within elderly buildings.
Many specimens are stored in Victorian bottles, with labels laboriously
inscribed in long outdated styles of handwriting. A pervading smell of preservatives
reinforces the impression that biological collections are an unwieldy and
unnecessary relic of the past. It is almost as if physics were still being
pursued using the methods and facilities of Newton and Galileo. The spurious
air of leisured inefficiency is enhanced by the apparent randomness of some
collections – they often continue to grow in a haphazard fashion, new material
being acquired without any apparent relevance to current research.

Yet natural history collections are undoubtedly part of modern biology,
and are arguably one of the most cost-effective research tools. They have
a variety of uses that are important in their own right and underpin many
other activities. Among these are taxonomy – the recognition and description
of species and their variation – and identification of biological specimens,
such as those used in other areas of biological research or which are of
economical importance. Collections are also used for reconstructing phylogenies
– the historical relationships of species – and for studies of comparative
anatomy. Finally, they provide many insights into the natural history of
species. For instance, a preserved animal will often contain all its parasites
and its last meal and often information on its reproductive life as well.

Acquisition policies are opportunistic because collections increase
most economically by a mixture of planning and serendipity. Material needed
for particular research projects, and to fill gaps in representation, may
be collected and exchanged deliberately. But other specimens are accepted
as gifts or purchased when they become available, sometimes with no immediate
use in mind. Expeditions organised to obtain species needed for immediate
research often come back with many others besides, because collection is
cheap in these circumstances.

No business would be run like this. In terms of storage overheads, turnover
and all the other shibboleths of commercial practice, such procedures appear
eccentric. Nevertheless, museum collections produce results at remarkably
small cost. The alternative, to collect material when it is required, would
be enormously expensive.

One example makes the point. Some years ago, in northeast Nigeria, doctors
treated the bites of carpet vipers (Echis) with serum produced from local
snakes. For various reasons, this process could not be continued and antivenin
from Iranian carpet vipers was used instead. Many people died as a result.
Subsequent work at the Natural History Museum showed that this disaster
happened because the Iranian animals belong to a different species. The
matter was clarified in a couple of weeks by examining carpet vipers in
the collection obtained cheaply over the past century from a hundred or
so localities spread from Africa to Sri Lanka.

Material could have been collected afresh from these places, but it
would have taken a very long time and the labour and travel costs would
have been astounding. Fresh speci-mens of some taxonomic groups may not
even be obtain-able, especially at short notice. They may be endangered
or extinct, secretive and difficult to find, or confined to inaccessible
places.

Unlike most businesses, biological collections run on very long timescales.
Material is obtained when it is possible and convenient to do so, then stored
cheaply so that it is available for use as soon as the need arises, at costs
far below that of deliberate collection. Such long-term investments are
by no means uneconomic. Ask anyone whose family invested in Titian a few
centuries ago.

Clearly, collections cannot grow forever. But calling a halt now would
be a great mistake. Few collections are complete enough to enable many desirable
projects to be executed without obtaining additional material. If the world
is going into a period of mass extinction, our descendants will curse us
if we do not take the opportunity of collecting representatives of doomed
forms. Such collections are also necessary for monitoring and, if possible,
averting the impending crisis.

Collections in different countries undoubtedly overlap in their content
and often in their areas of research, and it is tempting to consider rationalisation.
But some repetition is necessary if collections are to be used to identify
biological specimens with confidence. The fact that collections hold similar
material does not mean that research done on them will be the same. Nor
does having material in one collection ensure that it is available to workers
in other countries. Loans are costly and may not even be permitted by the
specimens’ owners. Again, people will often collect more readily for their
local institution than for some distant supranational centre.

The presence of material of a group in several institutions ensures
the survival of some samples even if collections are destroyed – war, floods
and fire have all taken their toll in the past. Finally, taxonomy and phylogeny
– how species are historically related to each other – are complex disciplines
open to different interpretations. It is often helpful to have more than
one researcher examining a problem independently, just as in other areas
of science.

While they are cheap compared with many other scientific enterprises,
collections clearly cost significant amounts of money, and assessing costs
and benefits is not easy. The biggest problem is that the benefits are virtually
impossible to compute. They tend to accrue over very long periods, so it
is not appropriate for short-term customers to dictate overall strategies
for accession and research. Specimens may take over a century to give up
their information. The benefits to research also accrue slowly. For instance,
a conservation project may rely on recognising species first described many
years earlier.

Taxonomic papers often have a small initial readership but go on being
used and quoted for a very long time. The taxonomist George Boulenger stopped
working on lower vertebrates at the Natural History Museum in 1920, yet
60 years later, during the early 1980s, he was mentioned 350 times in five
years in the Science Citation Index. His predecessor, Albert Gunther, retired
in 1900 but still has 301 citations over the same period, and Edward Cope,
who worked on reptiles at the Philadelphia Academy and stopped at the same
time, has 330. As the index does not include citations for many of the journals
in which taxonomists publish, the real number of citations must be far higher.
Also where the original work is used at second and third hand, it is not
attributed at all. This continued use is not exclusively taxonomic. Boulenger,
for instance, is quoted in other contexts in 17 journals in the same period.

Collections continually surprise. Specimens originally valued for their
external appearance have later provided information about many sorts of
internal systems. Then light microscopes revealed fine structure, and successive
generations of biologists subjected samples to investigation by X-ray, scanning
electron micros-copy, carbon isotope studies and finally DNA investigations.
None of the resulting discoveries could have been predicted when the specimens
were collected.

It is difficult to overstate the amount of information that natural
history specimens contain. For instance, recent analysis of the bones collected
over 60 years ago has shown, through radiocarbon dating and some assessment
of diet from the ratios of stable carbon isotopes, that the two giant tortoises
now extinct survived on Madagascar until after the arrival of people on
the island, and that the species differed in the kind of plants they ate.

The high information content of natural history specimens means that
they can easily be put to uses beyond the particular piece of research for
which they were originally preserved. The Natural History Museum has long
held specimens of the tuatara or Sphenodon, a reptile found only in New
Zealand (see ‘Tuatara sheds its fossil image’, 91av, 20 October
1990). John Gray first described it from the museum’s specimens in 1842.
Later, in 1867, Albert Gunther used them to show that the tuatara was not
just ‘another lizard’ but the sole survivor of a reptilian order that is
otherwise known no later than the Jurassic period. These specimens have
been studied on numerous occasions since. As late as 1983, they were used
to explore muscle systems of primitive reptiles, and in 1989 they helped
to confirm there are in fact two very similar species of tuatara.

Taxonomy – recognising and describing new species and their variation
– constitutes one of the longest standing and most extensive uses of collections.
Estimates of the number of species that exist have recently been revised
(see ‘Counting species one by one’, 91av, 11 August 1990). Investigations
of insects in the canopies of tropical forests suggest that, instead of
the 3 million species of organisms previously estimated, there may be 30
million. Perhaps surprisingly, this has brought the value of conventional
taxonomy into question. Describing so many species seems an impossible task
for the foreseeable future. But the problem of huge numbers is substantially
confined to entomology and even here does not involve all insect groups.
The fact that there are a myriad of certain similar groups, such as beetles,
does not destroy the value of trying to describe and inventory other types
of organisms where most diversity actually lies. When it was realised there
are an infinity of stars, astronomers did not give up cataloguing the heavens,
they merely became more selective. Presumably, this is what will happen
in groups of organisms with many species.

The utility of taxonomy

Furthermore, the supposed crisis caused by the recognition that the
number of species may be much larger than we had realised is irrelevant
to many of the uses of collections. For taxonomy contributes to a huge variety
of activities. Conservation, for instance, is intimately linked to systematics.
At the simplest level, it is impossible to assess the conservation needs
of a species unless it is first recognised and differentiated from others.
For instance, recent concerns about the fate of the Italian agile frog (Rana
latastei) rest entirely on the fact that taxonomists had earlier distinguished
it from other similar species in Europe. The neglect of taxonomy can also
cause problems. The second species of tuatara Sphenodon gutheri which occurs
only as a small population on North Brother Island off New Zealand was recognised
in the 19th century, but a few years later other biologists erroneously
assigned it to the more widespread S. punctatus It was given no special
protection for more than 100 years and could easily have become extinct.

The study of natural history collections has sometimes been considered
rather archaic, lacking in real intellectual content, and peripheral to
the rest of science. The distinguished immunologist Peter Medawar once unkindly
described working out the historical relationships of species as ‘browsing
in the parish registers of evolution’. But the process of reconstructing
from biological specimens phylogenies that represent the routes by which
living things have evolved is not easy. It involves broad anatomical knowledge
and complex decisions. Today’s methodology has developed only recently and
continues to evolve in the wake of sophisticated computer programs and debates
on philosophical and logical issues. Probably the methods used in systematics
have changed more radically in recent times than those of most other biological
sub-disciplines. And the application of phylogenetic information to our
understanding of evolution is developing rapidly.

For instance, many phylogenetic studies contradict the simplistic notion
that evolution is always ‘progressive’, with steady increase in complexity
and sophistication. Jonathan Coddington, of the Smithsonian Institution
in Washington DC, for example, has shown that spiders that build simple
cobwebs arose from those that make complex orb webs. Similarly, the widespread
tail-shedding strategy of lizards, an effective antipredator device, has
been lost many times during evolution. And Ray Huey at the University of
Washington and Al Bennett have found that one group of Australian lizards
has evolved so that their optimal physiological performance no longer occurs
at their usual environmental temperature. Such revolutionary shifts in how
biologists in other disciplines see their results arose from systematic
studies and are a direct result of research based on collections.

It is paradoxical that the institutions apparently dedicated to killing
and preserving organisms should play such a significant role in their conservation,
but the damage they do to natural populations is minute compared with the
benefits. Museums should always take possible conservation problems into
account when obtaining specimens, and they nearly always do. The number
of organisms that end up in museums is infinitesimal compared with those
that die or are removed from the wild as a result of other human activities.

The absence of some of the superficial trappings of modernity can mask
the importance of natural history collections, and their sheer variety makes
them hard to encapsulate. They vary in overall quality, and among taxonomic
groups in such aspects as the information content of specimens, the proportion
of taxa still undescribed, the cost of collection and storage, the appropriateness
of taking in more material, and their relevance to human interests. In this
situation, it is easy to make inappropriate generalisations. In some ways,
collections are victims of their own success. Although the benefits they
produce are substantial, these are difficult to sum up because they occur
in so many areas, often at second or third hand, and may only become apparent
over time.

If the complexity of natural history collections is taken into account,
however, and they are judged on appropriate time scales, it is clear that
they are one of the key ways of understanding the diversity of the living
world and its origins. It is difficult to predict all their future uses
and benefits. But unless they are allowed to persist and flourish, we can
be sure that our knowledge of the natural environment and of our place within
it will suffer.

Nicholas Arnold researches reptiles at the Natural History Museum in
London.