杏吧原创

THE airlock door slams shut. There are ants everywhere, crawling over your
shoes like a fine brown dust. Beyond the path everything is lush, green and
tropical, but through the fronds and the glass panels behind, you can still see
the Arizona desert that you left moments ago. It stays in view as you climb
through the tropical rainforest, pausing at a muddy lookout above the tree
canopy.

A second huge glass pyramid houses scrub desert, and between this and the
rainforest there is savanna, a mangrove estuary, an ocean complete with coral
reef and waves lapping gently against an idyllic white sand beach. There is an
orchard, fields of wheat and鈥攊ncongruously鈥攁 subterranean network of
concrete and steel housing the pumps, water pipes and heating systems that keep
the wind blowing and the rain falling.

Times have changed at this vast glorified greenhouse, dubbed Biosphere
2鈥攖he Earth is Biosphere 1. Just two years ago, no one鈥攏ot even the
resident crew of six biosphereans鈥攃ould enter or leave. It was sealed,
self-sustaining, a holistic prototype for a Martian space colony. It was also
vilified. The whole enterprise, run by a secretive eco-cult, was viewed by most
of the scientific community with scorn, tinged with regret for a good project
gone to waste.

Fresh start

Now the scientists have been given their chance. The owner of the site,
billionaire businessman Ed Bass, has turfed out the eco-cult along with the idea
of self-sustaining space travel. Bass was already worried about the image and
direction of the project in 1993 when geochemist Wally Broecker, from Lamont
Doherty Earth Observatory in New York, wrote to him, complaining that the
facility was being misused. On 1 April that year, when Bass decided to oust John
Allen, head of the cult and the project, Broecker was ready to step in. He
persuaded Columbia University to set up a consortium to run the Biosphere, and
assembled an impressive list of collaborators including many top names in
ecosystem research. Bass has given the new team five years to show what they can
do. And their plan? To flush Biosphere 2 with carbon dioxide, and use it to
predict the Earth鈥檚 future.

They will have their work cut out. Ever since atmospheric records first
showed beyond doubt that carbon dioxide levels were rising, ecologists and
agriculturalists have been trying to predict what effect this will have on the
world鈥檚 plants and trees. Because CO₂ is a fundamental requirement for
photosynthesis, scientists have long suspected that higher CO₂ levels
will fuel extra plant growth. Some have even hailed rising CO₂ as a
fine way to boost global harvests. But will this really happen? And even if it
does will all crops respond in the same way?

Others look to natural trees and shrubs around the world to help alleviate
the problems of global warming by soaking up some of the extra CO₂. But
if extra carbon is taken up by the natural biosphere how long will it stay
there? Will it be safely locked up in forests, or could it go quickly into the
soil and be remobilised back into the atmosphere? Could there be a CO₂
level that will kill off trees and shrubs, so that they would give up their
accumulated carbon in one catastrophic burst that would be the final straw for
the planet?

There have been many attempts to answer these questions. Ecologist Peter
Curtis from Ohio State University in Columbus estimates that by 1992 more than
1500 papers had been published on this topic, and a hundred more emerge every
year. Most researchers have based their studies on the CO₂ levels
projected for the Earth around the middle of the next century. Experiments have
taken place in everything from small glass containers kept indoors under
artificial light to outdoor suites of open-topped plastic chambers, continuously
flushed with air enriched with CO₂.

With so many different approaches it is almost impossible to draw firm
conclusions about how the global ecosystem will react to higher CO₂.
But something strange is already happening to the Earth鈥檚 carbon cycle鈥攖he
chain of processes that ceaselessly recycle carbon from air to land and back
again. Atmospheric CO₂ levels are not rising as fast as they should.
Every year over a billion tonnes of carbon that ought to end up in the
atmosphere go missing. Where, nobody knows. Boreal forest wood, plant roots,
soil are all candidate 鈥渟inks鈥� for harbouring this missing carbon.

Hopes are high that flushing Biosphere 2 with controlled amounts of CO
₂ will provide clues to help solve this and all the other conundrums about
how plants respond to higher CO₂. The beauty of Biosphere 2 is its
large size鈥攐ver three acres鈥攁nd the complexity of its ecosystems.
Also, virtually everything about the environmental conditions can be
controlled鈥攎aking it possible to pin down potentially important factors
that other approaches miss.

Previous studies have done their best. For nearly a decade, Bert Drake from
the Smithsonian Environmental Research Center in Maryland has been running an
experiment in a Chesapeake Bay salt marsh. Every spring he encircles marsh
plants with small, open-topped plastic chambers. Then he flushes half the
chambers with extra CO₂ and monitors them throughout the growing
season. Each year, plants in the enriched plots take up around 40 per cent more
carbon than those in the control plots, suggesting that wetland plants may well
help to scrub out some of the atmosphere鈥檚 CO₂ in the future. But
what鈥檚 intriguing is that the researchers are not sure what exactly happens to
the extra carbon taken up by the plants.

Though some of it disappears into the roots and leaves, Drake has not yet
discovered whether the rest of the extra carbon has stayed bound up in some
other form, such as organic matter in the soil, or has escaped back to the
atmosphere. In 1994 one of Drake鈥檚 collaborators, John Dacey from Woods Hole in
Massachusetts, showed that the enriched plots of salt marsh seemed to give out
more methane than the ambient ones, suggesting that at least some of the extra
carbon is being decomposed and reemitted to the atmosphere.

Natural systems

Bill Schlesinger from Duke University has also been trying to quantify how
much extra CO₂ will be taken up by plants. For the past two years, he
has been testing a prototype free-air CO₂ enrichment (FACE) experiment
on pine trees in Duke Forest, North Carolina. The idea is to make the enrichment
system as natural as possible by simply blowing CO₂ onto the trees from
vertical pipes arranged in a 30-metre circle.

In preliminary experiments, Schlesinger and his colleagues saw 33 per cent
more CO₂ in the soil pores of their enriched plot. This summer they
plan to run three sets of FACE rings, measuring all the possible outlet routes
for the carbon, to find out how much carbon will be locked up in the trees, how
much will pass into the soil, and how much will pass rapidly back into the
atmosphere.

Other ecosystems have also shown signs of taking up atmospheric CO₂.
In 1994, for example, Walt Oechel from UC San Diego reported results from a
three-year enrichment experiment using chambers constructed in Arctic tundra,
where large amounts of carbon are already stored below the surface. In the first
year, the enriched plants soaked up a sizeable amount of carbon, suggesting that
tundra too could contribute to mitigating global warming. But disappointingly
Oechel found that the growth spurt slackened off almost completely in the second
and third years.

Other studies have also seen hints of a similar acclimatisation, where plants
seem to lose interest in the extra CO₂ over time. But it is difficult
to know how seriously to take them. Earlier this year, Curtis published results
from a meta-analysis, in which he took 48 studies of trees enriched with CO
₂ and used statistical techniques to try to extract common factors. He found
that the type of chamber used can make a drastic difference to the results. All
the trees grew more effectively under higher CO₂, but after around 50
days the ones in small enclosed growth chambers under artificial light were
showing distinct signs of acclimation鈥攖heir response was only a third that
of the trees in the open-topped chambers. Curtis concludes that the further you
go from real conditions, the more likely you are to introduce glitches.

Complex problems

However, the more realistic conditions are, the less control you have over
them. Take the FACE experiments. Schlesinger admits that virtually all you can
control is the CO₂ level. 鈥淥ur forest is diverse the way nature is
diverse,鈥� he says. 鈥淚t also gets the diversity of drought and low soil nutrients
and pests and everything else that nature throws at it.鈥�

But many high-CO₂ studies suggest that factors such as water,
temperature and nutrient levels could play a vital role in determining how big a
response there is to high CO₂, and whether or not plants become inured
to the new CO₂ levels. 鈥淚f the planet begins to warm and we have a
two-factor experiment鈥攈igh CO₂ and high temperature鈥攚e
could be moving into a realm we don鈥檛 understand,鈥� says Schlesinger.

Attempts to work out how crops will respond to higher CO₂ levels
have the same problems. Bruce Kimball from the US Water Conservation Laboratory
in Phoenix, Arizona, has used FACE experiments to investigate the effects of
21st century CO₂ levels on both wheat and cotton.

The cotton results were fairly simple鈥攖he plants took up extra CO
₂, just as predicted, and by the end of the season yields in the enriched
plots were 40 per cent higher than those in the control plots. But wheat was
more complicated. Though in the middle of the growing season, the crop was
taking up 20 per cent more CO₂ than control patches, it also matured
faster, shortening its growth season. The ultimate yield was only 10 per cent
higher than the controls.

Kimball thinks changes in temperature and water levels could be the key to
these findings. He points out that higher CO₂ and lower water levels
both close stomatal pores on the surface of the plant leaves. When water escapes
through these pores it evaporates, cooling the leaves. But with the pores partly
closed, less water escapes and the leaf heats up. Kimball suspects that heat in
the wheat leaves speeds up all the physiological processes inside, including the
ones that cause the grains to mature and ripen. If this is right, the higher
temperatures expected through global warming could combine with higher CO
₂ to shorten the growing season still further.

Kimball and his team can test their hypothesis by measuring leaf
temperatures, but what they can鈥檛 do using FACE technology is deliberately raise
the temperature of the whole plot and see what effect it has. Enter Biosphere 2.
鈥淚t is a completely controlled facility,鈥� says Bruno Marino, who in 1994 gave up
his research post at Harvard to take over as director of science there. 鈥淵ou
can鈥檛 walk out to a natural prairie and control the temperature, the nutrients,
the species composition, the light conditions and the CO₂ above it for
long periods. But you can here.鈥� In principle, everything in Biosphere 2 can be
controlled, right down to the artificial hurricanes induced on the palm trees by
judicious application of hand saws. And it鈥檚 already set up to address
questions about increased CO₂ in both agricultural and natural
ecosystems.

That said there are many problems to be solved before Biosphere 2 is ready
for new, rigorous scientific research. One legacy of the facility鈥檚 original,
holistic purpose is that wilderness, agricultural areas and even the living
quarters of the erstwhile biosphereans are all connected, making individual
controlled studies virtually impossible. Isolating wilderness from agricultural
areas has proved relatively easy, as there were few connecting points between
them. But separating the individual wilderness environments is more difficult.
For now, researchers are using plastic sheeting to isolate the rainforest and
the desert in their respective pyramids. Plans are also afoot for more permanent
dividers to separate the agricultural area into three separate regions to be
treated with different levels of CO₂.

Another legacy of the original setup is that though the ecosystems in
Biosphere 2 were designed to be self-sustaining like the Earth, they weren鈥檛
designed to mimic Earth exactly. For instance, some ecosystems contain species
lumped together almost at random from widely different areas. Also the insect
population is now heavily skewed鈥�-ants and cockroaches have outperformed
everything else, and many original insect species are now extinct. 鈥淭he first
thing we have to do is establish whether what happens inside the dome is
applicable to outside,鈥� says Taro Takahashi, an oceanographer from Lamont, and
codirector of Biosphere 2. Marino believes a good way to do this is with
continuous 鈥渞eality checks鈥濃�攃omparing plant behaviour inside and outside
the Biosphere, and checking results with those from the FACE experiments.

But plant physiologist Guanghui Lin stresses that Biosphere 2 need not be
exactly like the real world. The point, he says, is to use it as a model to
learn more about the mechanisms that drive the way plants respond to increased
CO₂. The facility鈥檚 biggest advantage, according to Lin, is that it can
be closed allowing researchers to trace the effects of all the components in the
system. Marino relishes the prospect of using isotopes to trace individual
interactions. Rainfall is controlled in the Biosphere, so by tagging the water
it could be possible trace it through the system. It may also be possible to pin
down the crucial nutrient cycles by tagging nitrogen in N₂O, or carbon
in the added CO₂. This and other experiments should allow the
researchers to track down the mechanisms behind the changes they see, and use
models to apply the same mechanisms to the real world. Kimball sees 鈥減lenty of
promise鈥� in this idea. 鈥淢odels could provide the bridge between what鈥檚 going on
in there and outside,鈥� he says.

Rich soils

Other unique advantages of Biosphere 2 are its complexity and size. These
mean, says Lin, that it could be used to validate 鈥渟caling鈥�
models鈥攁ttempts to predict the effects of higher CO₂ by scaling
up from leaf measurements to whole ecosystems. 鈥淏iosphere 2 is the only place
where you can test at the leaf level, the individual plant level and the whole
ecosystem level, and even at the global scale with different ecosystems
interacting with one another,鈥� he says. Most other scientists agree that the
size of the Biosphere is a great asset. But Curtis, for one, is more sceptical
about the benefits of the complexity of the different ecosystems. Though he
admits there are many different plant species in the Biosphere, Curtis feels
that the key interactions take place in the soil, whose structure and
composition is much less variable.

In fact, the soil in the Biosphere has provided the team with a significant
headache, by artificially boosting 鈥渘atural鈥� CO₂ levels. Most CO
₂ enrichment experiments run at around 650 parts per million, but when
Biosphere 2 was first sealed back in 1991, CO₂ levels soared to 4000
ppm. It turned out that the soils used were extremely rich in carbon, and the
soil biota shifted this extra carbon into the atmosphere with gusto. Marino says
there is no prospect of replacing the soil鈥攊t would be far too expensive.
But he hopes that by separating off the different sections and employing
artificial CO₂ scrubbers, he should be able to keep the levels under
control. Drake is not so sanguine. He thinks the soil problem could undermine
attempts to apply Biosphere 2 results to the real world. 鈥淚 don鈥檛 want to throw
water on their parade when they are just starting, but it鈥檚 the kind of thing
that could ultimately sink such studies,鈥� he says.

In spite of these problems, the new team remain cautiously optimistic. Though
they have already done a few experiments in the Biosphere, they are taking their
time and consulting widely to decide on the main projects. And despite the
Biosphere鈥檚 somewhat shaky reputation, the response from the scientific
community has been enthusiastic. Takahashi believes this is because it offers
the chance to take the sort of scientific risks impossible in today鈥檚 funding
climate. 鈥淭here is increasingly a sense around the world that the scientific
community has to be more cautious and less bold and dreamy in its proposals.
Riskier projects don鈥檛 get funded when the money is tight,鈥� he says.

But the one thing that won鈥檛 happen is locking another team of biosphereans
in. 鈥淚t was a stunt,鈥� says Broecker. 鈥淚鈥檓 not interested in that.鈥� Matt Smith, a
marine biologist from the Smithsonian Institute, was brought in after the 1994
management change to replace one of the second team of biosphereans. The first
team contented themselves with trying to survive , but the second team was
supposed to do proper experiments. The problem, says Smith, was that with a crew
of just seven people鈥攁ll the building could support鈥攖here was
scarcely time to solve the running problems, let alone to do rigorous research.
In the end the mission was aborted, and no-one has been locked into the
Biosphere since.

Smith has high hopes that the new approach will produce better science. But
he has a word of warning for those who believe that such studies will make it
easy to deal with global warming. 鈥淧eople have tried really hard here. We have
something that鈥檚 cool, kind of fun, but it鈥檚 definitely not planet Earth. We
can鈥檛 turn three acres back into the Earth, so if you think you can go around
screwing up an entire planet and have it fixed, you can forget it.鈥�