91av

On the second floor of a postmodern building not far from Ascot racecourse,
it is forever June. A cloudy sky – for the records say that in England flaming
June is exceptional – looks down on plants typical of a weedy field. Snails
rasp at the leaves, earthworm casts litter the surface of the soil, aphids
suck the plant juices, and tiny wasps hunt for aphids to parasitise. The
day starts off at a cool 12 degree C and rises to a modest peak of 20 degree
C in the middle of the afternoon. And three times a day, at 7 o’clock in
the morning, 4.15 in the afternoon and 8.30 in the evening, it rains for
precisely one minute.

This is the Ecotron, a £1 million facility that is, for the first
time, giving ecologists the means to carry out experiments that are complex
enough to be realistic, yet simple enough to understand. It is housed in
the Natural Environment Research Council’s Centre for Population Biology,
within the Department of Biology of Imperial College, London, at Imperial’s
country retreat, Silwood Park.

GLOBAL GREENHOUSE

In essence, the Ecotron is no more than a big bottle – a closed community
in which everything that goes in can be controlled, and everything that
comes out measured. It has already answered one of ecology’s fundamental
questions: what does biodiversity do? The more species you have, the Ecotron
has revealed, the more efficiently organisms can harness energy from sunlight.
And scarcely had that paper been written when researchers at Silwood Park
set the Ecotron to work on another question: what would a global greenhouse
with twice the carbon dioxide and a 2 degree C rise in temperature do to
ecosystems?

Until now, ecologists have had three ways of learning about the living
world and how its components work together. Some did bottle experiments
in the lab, under tightly controlled conditions but usually involving just
two species; real life is a sight more complicated than that. Others looked
at ecosystems out in the field, with all their complex components but without
the ability to manipulate them. A third group modelled ecosystems within
computers, but were condemned by the bottle merchants for being far too
complex, by the field ecologists for being too simple, and by both for being
unrealistic.

HALFWAY HOUSE

The Ecotron has changed everything. It provides a halfway house between
the field and lab approaches, says Phil Heads, manager of the centre. ‘The
Ecotron can bring in more complexity under greater control, and point the
way towards what could be looked for in the field with much more confidence
than a bottle experiment that only has two species.’

Indeed, its very existence is altering perceptions about the way ecology
should be done. Ecologists have until now been handicapped by the small
scale of their investigations. Unlike physicists with their colliders or
oceanographers with their research vessels, ecologists have traditionally
gone off individually to work on their own little patch of the planet.

The Ecotron provides a much needed focus, says Heads. ‘It’s not astrophysics,
it’s medium-size science. But that is a scale that ecologists have not been
able to work at in the past.’ And it has brought ecologists running. ‘People
around the world are very keen to come here,’ says Hefin Jones, project
leader of the Ecotron. ‘Before, they would never have worked together on
the same project.’ The latest experiment has pulled in microbiologists,
chemists, experts on decomposition, botanists, entomologists and engineers.

Yet it began life as a piece of pub talk, at a winter meeting of the
British Ecological Society in the mid-1980s. ‘It was one of those good ideas
that still seemed like a good idea in the morning,’ recalls John Law-ton,
an ecologist then at the University of York. A proposal was drawn up and
submitted to the NERC, which duly funded the Eco-tron and the centre that
Lawton now heads.

The pub musings have become reality as a suite of 16 compartments rather
like supermarket cold stores, arranged in two banks of eight. Each compartment
is 2 metres square, with lights overhead, a water sprinkler, and various
air inlets and outlets. The identical compartments offer the chance to do
replicated experiments, something unprecedented in ecosystems research.
Alternatively, the walls between the compartments, though sealed, can be
moved or removed, allowing the researchers to create fewer, larger chambers
to study the effects of scale.

In the middle of the floor of each compartment is what the scientists
call the ecosystem container. Each one is rather like a big plant pot housing
a selection of organisms drawn from a pool of suitable species: ‘Ecotron-friendly’
is how Heads describes them. ‘Two or three years of desk research by our
botanist, along with preliminary screening and trials, went into choosing
a list of species that we could use.’

Plants are represented by 16 species. All are self-pollinating annuals,
so they can go through several generations in the Ecotron. Likewise, the
herbivores – three species of aphid, a whitefly, a leaf miner, a slug and
a snail – have lifestyles ideal for the Ecotron. Each of the insect herbivores
has a parasitoid wasp that preys on it, and there are earthworms and springtails
in the soil to help recycle plant debris.

WEEDY FIELDS

A key point about the Ecotron is that the communities it contains –
pedants insist they are not ecosystems – are not designed to mimic the
real world, but to be a simplified model. A visit to a typical weedy field
might not reveal the particular selection in an Ecotron pot, but everything
in the Ecotron is in the typical weedy field.

But what does the model do? The way diversity affects the performance
of an ecosystem has assumed huge importance, as human activities lead to
mass extinctions of species. Whether this matters is the question that the
Ecotron’s first experiment set out to address.

To do so, Shahid Naeem and his colleagues created model ecosystems of
different diversity, all populated from the list of Ecotron-friendly species.
The low-diversity group had 9 species, the medium-diversity group had 15,
and the high-diversity group 31. The larger groups included all the species
from the smaller ones or – to look at it the other way round – for the smaller
groups, the scientists removed species from an ecosystem with high biodiversity,
just as is happening in some natural ecosystems.

For 200 days the Silwood ecologists measured several aspects of the
structure and functioning of each model ecosystem. Almost all varied significantly
with diversity, but the most striking observations was that the more diverse
communities consumed more carbon dioxide and were more efficient at turning
it into plant tissue. Since the researchers had excluded all possible influences
apart from the number of species, the message seemed clear: more diverse
ecosystems are more productive.

A similar story is beginning to emerge from other studies of diversity
in ecosystems. A couple of months before Naeem’s paper appeared in Nature
earlier this year, David Tilman and his colleagues at the University of
Minnesota published their conclusion from a 12-year study of tall-grass
prairie. At the start of the study, in 1982, they marked out more than 200
plots and added a different amount of fertiliser to each one. Every species
in a diverse community such as a prairie represents a slightly different
way of making use of limited supplies of nutrients, especially nitrogen.
So adding nitrogen tends to favour those species that can best make use
of it, and these then crowd out the others.

In 1987, Tilman started to measure the biomass of plants on each plot.
Then, as fate would have it, the following year saw the severest drought
for 50 years. It knocked all the plots for six, but the more diverse systems
were not hit quite as hard as the others. During the drought, the most species-rich
plots produced about half of their pre-drought biomass, while the most impoverished
(those that had been dosed with nitrogen) produced just one-eighth of their
pre-drought biomass.

The occurrence of a 50-year drought smack in the middle of the study
period was obviously a stroke of luck. ‘We knew the baseline and the variability
before the drought,’ says Tilman. ‘So we were really lucky that it came
along when it did.’ Once the rain returned, Tilman was in a position to
ask whether recovery from the drought was in any way linked to diversity.
It was. By 1992, three years after the end of the drought, the species-rich
plots had returned to their former biomass while the other plots were still
below their former average.

The Ecotron shows that species-rich systems are more productive. The
prairie studies show that they are more stable, in that they are less affected
by an outside stress and recover from it more quickly. A third, older study
confirms this. Sam McNaughton looked at plant growth in the Serengeti National
Park in Tanzania, by studying plots in areas which naturally different diversities.
Part of each plot he fenced off to prevent grazing by migrant herds of zebra,
gazelle and wildebeest. When the herds had passed through, McNaughton compared
the plants in the grazed and ungrazed areas of different plots. The less
diverse had lost three-quarters of their biomass; the more diverse, only
one-quarter.

You don’t have to look far for clues to how diversity produces such
effects. In the Ecotron, plants in the richest communities also had the
greatest variety of shapes, which filled the space above the pots with leaves.
The explanation for the increased productivity seems obvious: a greater
diversity of plant shapes allows the plant community as a whole to make
more use of the available light.

What about the prairies? Here, the species-poor plots happened to be
dominated by grasses sensitive to drought. Among the species in the more
richly endowed plots there were more likely be some drought-resistant ones,
and the growth of these compensated for the loss of growth of more sensitive
species.

RIVETS AND PASSENGERS

These results have been seized upon because they address the fundamental
question of what biodiversity is for – in other words, why so many species
evolved and survive when many of them seem to carry out broadly similar
jobs within ecosystems. Ecologists have puzzled over this for decades,
without producing any overwhelmingly acceptable hypothesis.

Two broadly opposed schools of thought have emerged on the connection
between diversity and stability. In one of the models, each species in an
ecosystem can be likened to one of the rivets that holds an aircraft’s wings
to the fuselage: every rivet plays its part in keeping the plane intact,
and the loss of even one or two will weaken it. Lose more than a few, and
the system is in for a catastrophic decline. The alternative to the rivet
model is known as the passenger model. It sees the species as the people
on the plane – crew as well as passengers. You could remove most of them,
but as long as a few key individuals remain the system will still fly. On
the other hand, removing those key individuals will lead to disaster, no
matter how many other passengers remain.

AFFORDABLE LOSS

Which model is right is a question of some importance for our own future.
How many species can we afford to lose before the ecosystems needed for
human life start to collapse? If the rivet model is at work, we’re in trouble;
if there is redundancy, and the passenger model applies, why worry?

What experiments tell us is that there is a bit of redundancy but that
hanging on to all the species matters more than had previously been thought,
argues Stuart Pimm, an ecologist at the University of Tennessee in Knoxville
and an expert on ecosystem stability. In other words, species are more
like rivets than passengers. When Pimm and his colleague Julie Lockwood
reanalysed data from Tilman and from McNaughton, they found that the drought
resistance of an ecosystem increases with its diversity – but only so far.
Eventually it reaches a plateau. And from this Pimm concludes that ‘some
species do appear to be redundant when diversity is high, but below a certain
threshold, species become increasingly more important in maintaining resistance’.

Reassuring as this seems at first sight, there is little room for complacency
about the value of biodiversity, Pimm warns. ‘The threshold in each study
is close to the highest observed diversities. There is no wide plateau of
species richness where species do not matter. Species become important quickly.’

Like much good science, this new information relating diversity to stability
and productivity raises more questions than it answers. Take the wishy-washy
argument among armchair conservationists that can be summarised as ‘diversity
is our friend’. Bob May, a professor of ecology at the University of Oxford,
challenged that notion in the 1970s. He showed that complex ecosystems are
in theory more likely to fall apart than simple ones. ‘My view is largely
unchanged,’ he says. ‘In general, the more bits and pieces you put together
the harder it is to hold it all together. But on the other hand the course
of evolution is to keep injecting new things and trying to see how much
you can shuffle together.’

In other words, there are two opposing pressures at work in ecosystems.
One is to pump up species diversity to allow an ecosystem to make the most
of its resources. The other is to reduce species diversity to avoid generating
fragility. History, says May, may have selected a subset of complex ecosystems
that balance these two pressures.

Pimm agrees. ‘The communities we see in nature are the end product of
a fantastically rich and complex process of history,’ he says. Among these
processes are species change, invasions that succeed or fail, the relentless
addition and subtraction of components. Such complexity is hard to mimic:
artificial wetlands planted by land developers in the US as a substitute
for wetlands drained for housing are no real replacement, says Pimm.

If diversity increases resistance, then single-species ecosystems should
be the most vulnerable to catastrophic decline. This prediction is being
tested by the largest ecological experiment of all: agriculture. Planting
monocultures where there used to be diverse prairies that were productive
and stable opens them to the dire effects of drought. And because the new
plants are annuals, disastrous soil erosion may be one of them. Pimm is
putting such ecological insights to work to try and develop a productive,
persistent agriculture, which is bound to be diverse.

What next for the other researchers? Lawton would like to use the Ecotron
to study certain questions of scale. ‘I’d like to know whether it is possible
to keep more of the species at a larger scale,’ he says. There are complicating
factors at work here: theory suggests that the number of organisms that
can be packed into a given space might increase as the total area of an
ecosystem becomes bigger. Highly localised ecosystems might, therefore,
have disproportionately fewer individuals than larger ones. Habitat fragmentation,
in which a large area becomes carved up into small patches, could constitute
what Lawton calls a ‘double whammy for conservation’. A population hanging
on in the isolated remnants of an ecosystem could be, as Lawton explains,
‘more likely to wobble off into extinction quite by chance’. The Ecotron,
with its ability to change the absolute size of the ecosystems while holding
everything else constant could provide an answer.

Naeem has left the Ecotron to work with Tilman, who describes himself
as ‘an experimentalist at heart’, despite having spent 12 years studying
plots where the only experimental manipulation was adding nitrogen. They
plan to create plots of known diversity out on the prairie (see box: ‘Growing
an ecosystem’), and to impose droughts to order with movable rain covers.
Mesh will keep out insects and small mammals; Tilman has a hunch that they,
too, will be affected by the diversity of the habitat around them. ‘The
paper offered the plausibility of a link between diversity and stability,’
he says. ‘The experiment will cement it.’

WINDOWS OF CHANGE

Meanwhile, in the Ecotron it is still June, but June 2060. In half the
chambers the carbon dioxide has been doubled, to just over 550 parts per
million, and the pots enjoy an average temperature 2 degree C higher than
today’s. This is the NERC’s recommended ‘moderate’ scenario for research
on the greenhouse effect. Does Lawton have any hunch what that might do
to the model ecosystems? ‘I honestly don’t know,’ he replies. ‘At this stage,
when you look in the windows, they don’t look different. I hope to hell
something does happen.’ Then he corrects himself. ‘Actually a negative result,
at all trophic levels and in all parts of the system, would be thumpingly
interesting. So I really don’t mind what happens.

Jeremy Cherfas is a freelance writer and broadcaster

* * *

Growing an ecosystem: gardening plus science

Soil for the Ecotron comes from the fields around Silwood Park. But
before it goes into the pots, it is sterilised with methyl bromide to kill
everything in it, then washed with a brew of bacteria and fungi leached
from Silwood soil. Why kill everything, only to reinoculate? ‘It standardises,
makes it repeatable, a real experiment,’ says Hefin Jones, Ecotron project
leader.

Next come the plants. Six seeds of each species are placed on the surface
at positions determined with random numbers at whatever density the experiment
calls for. Of the six, only one is allowed to grow on; any others that germinate
are removed, again in the interests of replicability.

Then come the decomposers and the herbivores. The earthworms and springtails
need a supply of food, so they aren’t added to the pot until there is a
layer of leaf litter on the surface. And the insect herbivores must wait
until the plants are well established and growing strongly enough to withstand
their depredations. Finally, with the herbivores in place, the parasitoids
are introduced to prey on the plant eaters and keep them in check.

‘By three months,’ says Jones, ‘the whole community is set up. The exact
populations might vary, but they all establish.’ That reflects the trials
carried out to ascertain which species get on together and to draw up the
list of Ecotron-friendly components.

On the Minnesota prairie, David Tilman takes a similar approach in his
experiments aimed at confirming the link between diversity and stability.
‘We started with a 23-acre field,’ he says. The area was sprayed with glyphosate,
a total weedkiller, and the dead plants were burnt off. ‘Then we used a
road grader to peel back the upper 10 centimetres, which contained all the
seeds.’ The underlying soil was ploughed and thoroughly disced, to create
a fine tilth.

Meanwhile, hundreds of soil cores were taken back to the laboratory
over the winter, to check that the clean-up had indeed emptied most of the
seed bank. Finally, this spring, the researchers planted up the plots with
seeds of 1, 2, 4, 8, 16 or 32 species. There are 342 large plots, each 13
metres square. And there are 147 smaller plots 3 metres square: ‘those we
can tend, watering and weeding them’.

Tilman concedes that not all the plots, especially not the larger plots,
will be identical. ‘We’re getting replicability as best as we could,’ he
says. Jones, with his much smaller plots, can maintain tighter standards:
‘You have to make sure everything starts off exactly the same,’ he insists.

* * *

B-word or buzz word?

Measureable effects of the Earth Summit in 1992 are hard to find, but
there is at least one: Rio has made the word ‘biodiversity’ popular. A trawl
through a selection of British newspapers and magazines reveals that it
appeared in an average of 13 articles a quarter in the year before the conference.
While the conference was on, usage peaked at 201 articles – not surprisingly,
given that the signing of the Biodiversity Convention was one of the main
reasons for having the conference. But the B-word kept some of its new-found
popularity long after the diplomats had flown home, and usage now runs at
31 articles per quarter.

Biodiversity came to be considered a Good Thing. But what is it actually
good for? There have been many different attempts to pin down its value.
Ecological services have been high on the list: biodiversity regulates rainfall
and water supply; it enhances the fertility of the soil; it soaks up excess
carbon dioxide, mitigating the greenhouse effect; it provides the genes
to sustain our crops and the medicines to cure our ills; some of it is pretty
and people will pay to preserve that.

All of these virtues have a considerable intuitive appeal, even more
so to ecologists who have made a study of nature’s workings. But they have
not had much support from real data. And worse, from the point of view of
those who want to preserve biodiversity, these formulations encourage a
superficial trade-off based on the cost of maintaining these individual
benefits. Meanwhile we are left guessing at just how much biodiversity we
can afford to lose.