AT FIRST glance, Cedar Creek Natural History Area looks much like hundreds of
other abandoned farm fields scattered across the state of Minnesota. Between the
oak copses that dot the landscape, prairie grasses nod in the summer breeze, and
many familiar wildflowers mark the land’s return to native prairie. But look
more closely and you’ll see that instead of growing randomly the plants form a
precise chessboard pattern of hundreds of squares just a few metres across. Over
there is nothing but little bluestem grass, here nothing but lupins, and in
another patch a mixture of two dozen species share their square of ground.
The burning question
Clearly, this is no ordinary field. Cedar Creek is the site of one of the
world’s most famous ecological experiments, a multimillion-dollar effort led by
David Tilman of the University of Minnesota in St Paul. Tilman created and
meticulously seeded these hundreds of plots in search of an answer to a question
burning in the minds of ecologists, conservationists and policymakers alike:
does an ecosystem with more species work better than one with fewer species?
That is, does it make more efficient use of sunlight and nutrients and produce
more biomass to support more vigorous populations of plants, animals and
microbes—and is it less fragile and vulnerable to damage when physical
conditions change? If so, extinctions caused by human activity risk more than
just the loss of species: they may threaten the biosphere’s ability to capture
energy through photosynthesis, cycle nutrients and resist or adapt to the
vagaries of climate.
An uncomfortable answer to this question is now beginning to emerge from
Tilman’s field at Cedar Creek and a handful of other experiments around the
world. In all but the simplest ecosystems, losing a few species probably matters
very little to how well the system functions, at least in the short term.
Advertisement
That result is sure to disappoint conservationists fighting to preserve rare
species. It suggests that the biggest benefit from biodiversity is likely to be
in farm fields and forest plantations. A few recent experiments, however,
suggest that those “extraneous” species in nature might turn out to be important
in the long term.
The picture looked a lot simpler four years ago, when two research groups
came out with evidence suggesting that ecosystems with a richer assortment of
species did indeed work better. One team, headed by John Lawton of Imperial
College, London, created miniature ecosystems of different diversities in a
sophisticated growth chamber called the Ecotron at the college’s field station
at Silwood Park, Berkshire. The more diverse systems were more productive,
Lawton’s colleague Shahid Naeem found, indicating they made more efficient use
of sunlight and nutrients. The second group was Tilman’s, which suggested that
more diverse patches of prairie—left over from an older experiment testing
the effect of nitrogen fertilisation—withstood drought better than those
with fewer species (see “How many species do we need?”, Â鶹´«Ă˝,
6 August 1994, p 36).
Critics quickly pointed out serious flaws in both studies. “It was giving the
conservationists exactly what they wanted to hear, but it was a bit alarming to
some of us ecologists who knew the results were very suspect,” says Philip
Grime, director of the Unit of Comparative Plant Ecology at the University of
Sheffield. The Ecotron experiment, for example, included some fast-growing plant
species in the high-diversity systems that were not found in any lower-diversity
systems. The presence of these species could account for the higher productivity
of the more diverse systems, says Grime. And in Tilman’s study, plots with fewer
species had usually also been fertilised more heavily, so he couldn’t be sure
that differences in diversity, not nitrogen, had caused the patterns of drought
resistance.
To clarify matters, Tilman started a new experiment in 1994 in which he
manipulated the number of species while controlling for all other environmental
factors. He stripped the prairie down to bare soil and marked out 342 plots 13
metres square and 147 plots 3 metres square. Then he reseeded the plots with
between 1 and 32 native prairie plant species. In one portion of the experiment
he selected species at random from a “pool” of 18 plant species. In the other
portion of the experiment, Tilman selected up to 32 species randomly from up to
four “functional groups”—warm-season grasses, cool-season grasses,
broad-leaved perennial herbs and nitrogen-fixing legumes. Many ecologists think
these functional groups, which define the general role a plant plays in the
ecosystem, are more important than a plant’s precise identity. Every summer
since, Tilman and a small army of colleagues have monitored the plots and kept
the smaller ones watered and weeded.
Tickets in the lottery
Now in its fifth summer, Tilman’s experiment shows clearly that plots with
more species have more total plant growth and are more efficient at taking up
nitrogen from the soil. Some critics worry that this may result from the Ecotron
problem in a slightly different guise: plots with more species have more tickets
in the “plant lottery” and so a better chance of including the largest, most
productive species. But if that were so, a few simpler plots growing the right
species ought to do just as well as more complex plots with the right species
plus a few extras. Instead, many two-species plots are more productive than even
the best monocultures, and many four-species plots do better than any
two-species plots. That’s hardly surprising, says Tilman, because the roots of
different species reach down to different depths in the soil, drawing on
different reserves of water and nutrients.
Species richness has other positive effects, too. At Cedar Creek far fewer
weeds invade the richer plots—a direct consequence of the smaller amount
of open space and scarcer nitrogen on such plots, says Johannes Knops, research
director for the site. Nor does fungal disease ravage rich plots to the same
extent as poorer ones, because susceptible plants are separated by resistant
species.
All of this sounds like a good reason to conserve as many species as possible
in natural ecosystems. But Tilman’s experiment really says nothing of the sort.
Most of the advantages of increasing diversity—greater productivity,
better nutrient use, resistance to disease and weeds—come with the first 5
or 10 species, just a small fraction of the total species richness of most
natural ecosystems. After that, adding more species brings diminishing returns,
and the benefit of diversity nears a plateau (See Diagram, p 32).
“The sad thing is that these results demonstrate that having a lot of species
really doesn’t contribute all that much to stability or productivity,” says
Michael Huston, an ecologist at Oak Ridge National Laboratory in Tennessee, and
one of Tilman’s most vocal critics. “You could sack the majority of species and
still maintain most production. The extinction crisis is a moral problem with
essentially no functional consequences.”
Even Tilman, who wrote just two years ago that widespread extinctions of
species could threaten the productivity of ecosystems, has backed off from that
claim. “The first few species you lose have relatively little effect, and the
last few species have a huge effect,” he says now.
Indeed, it could hardly be otherwise, argues Grime. “The minor components
simply account for a small part of the biomass. If they’re wiped out completely
or double in size, it’s a very minor change.”
And on the global scale, of course, temperature, rainfall and soil quality
dictate the productivity of a site far more than biodiversity does. At any given
site, however, plant diversity still makes some difference. A new European study
known as Biodepth, which Lawton leads, helps to clarify this distinction. In
1996, researchers established Tilman-style plant diversity experiments at eight
grassland sites across Europe, from Sweden to Britain to the Greek island of
Lesbos. At any site, plots with more species are more productive during the
first two years of the study. But when Lawton compared the most diverse plots
across all eight sites, he saw no pattern in the graph at all. “It just looks
like a currant bun,” he says. In other words, the physical differences between
sites were great enough to override any effect of diversity.
But species-rich plots may do better not because of the number of species in
itself, but merely because they contain representatives of just a few key
functional groups. To a first approximation, in fact, you might be able to put
together a productive ecosystem in the same way you might choose a Family Meal
at a not-too-authentic Chinese restaurant—by choosing one from Column
A, one from Column B, and so on. When Tilman and his colleagues put the Cedar
Creek data through a sophisticated statistical analysis, they discovered that
the most productive plots were those containing both a legume and a warm-season
grass. Adding a third functional group—cool-season grasses or perennial
broad-leaved herbs—boosted productivity further. Plots containing
additional species within each functional group did better still, but the gain
was minor compared with that from adding functional groups.
Key species
A second study also emphasises the importance of functional groups. Working
in a Californian grassland, David Hooper and Peter Vitousek of Stanford
University created plots containing from one to four functional groups of
plants—in this case, early and late-season annual herbs, clump-forming
perennial grasses and nitrogen-fixers. They found that productivity depended
strongly on which functional groups were present, and much less on the number of
groups. Although the researchers did not directly test differences in species
richness, their results suggest that the number of species probably matters less
than whether key species are present.
All the biodiversity experiments up till now have a gaping hole, however,
because they fail to explore the influence of animal diversity. Even the Ecotron
and other ecosystem-in-a-bottle experiments that modify the number of animal
species have always modified the number of plant species as well, so no one
really knows whether animal diversity has a stronger effect on how well the
ecosystem functions than plant diversity does. Certainly, as Grime points out, a
few rare species—beavers, for example, and some starfish in intertidal
zones—can sometimes transform an entire ecosystem by their activities.
But a field experiment that tried to manipulate the number of animal species
instead of plants would probably flop because of the huge technical problems.
How do you maintain 32 insect species in one plot and two in an adjacent one,
for instance, asks Nick Haddad, an ecologist at the University of Minnesota who
studies insects on Tilman’s plant plots. Lab experiments hold better promise
here—and the Ecotron team has just begun an experiment that varies the
number of earthworm species while keeping plant diversity constant. Results
should be available next year, says Lawton.
Lawton and others are quick to point out, however, that even if natural
ecosystems could indeed lose many of their species and still work reasonably
well, this does not mean we should stand aside and watch such species go
extinct. Biodiversity has many other benefits, including genetic resources for
plant breeding and new therapeutic drugs. It supports a tourist industry worth
many billions of dollars a year. And, at the top of most ecologists’ lists,
species are worth saving for moral and aesthetic reasons. “They’re exactly the
same reasons we conserve medieval cathedrals and Mozart concertos,” says Lawton.
“They enrich our lives.”
Barren fields
But the most powerful lesson from Tilman’s experiment, and others like it,
may apply not to the richness of pristine nature but to the poverty of the
ecosystems we have created—the farms and forest plantations that often
contain a single dominant plant species. Increasing the diversity of these
ecosystems could bring bigger harvests of food and timber, fewer pest problems
and less damage to the environment from overuse of fertilisers and
pesticides.
These and other ecosystems altered by humans are where environmentalists’
real work must be done, says David Wedin, one of Tilman’s long-time
collaborators who is now at the University of Nebraska in Lincoln. “Conservation
biologists and academic ecologists spend a lot of time on the conservation end,
which is critical. But what are we fighting about? Whether to set aside 6 per
cent of the land surface or 12 per cent,” says Wedin. “Where I see one of the
important missions of ecology is in that other 88 to 94 per cent of the land
surface. How do we use ecology to contribute to sustainable land
łľ˛ą˛Ô˛ą˛µ±đłľ±đ˛ÔłŮ?”
Here, Tilman’s results suggest that adding a half-dozen or so well-chosen
species to the mix could bring huge returns. Agronomists have known for a long
time that intercropping two species, such as maize and beans, can raise total
yields, but their experiments rarely go beyond two or three species. And most
farmers have largely ignored the benefits of intercropping because it’s too
difficult for heavy machinery to cultivate and harvest more than one crop at
once.
Forestry may be an easier place to start. “Forestry as a discipline has
basically copied what agriculture has done. It’s treating trees as long-lived
row crops. But if what’s happening in our grassland is also happening in forest,
we should be able to get a 60 to 70 per cent gain in yield by going to a
mixed-species forest,” says Tilman.
While this might bring immediate benefits, there is a problem with Tilman’s
approach. The assumption that any advantages of high diversity will show in an
ecosystem’s day-to-day functioning is a dangerous one because it distracts
people’s attention from the longer-term processes, warns Grime. He and many
others think that in the long term even rare species may play a valuable role in
keeping natural ecosystems going. “We’re forgetting that one of the great values
of diversity is that if you have a lot of species sitting around and one or two
get knocked out there will be some to take their place,” says Naeem.
Naeem, who is now at the University of Minnesota, recently found a hint that
“spare” species may indeed help ecosystems cope with changing conditions. In a
scaled-down version of the Cedar Creek experiments, Naeem and Shibin Li
constructed 318 ecosystems of microbes in Petri dishes. These have an advantage
over prairie plots because you can create many copies of particular ecosystems
and can include a wider range of variables—all on a single bench top. “I
can put in things like herbivores and predators and decomposers, whereas the
bulk of biodiversity experiments are focused on plants,” says Naeem.
Filling the gaps
The biomass of photosynthetic algae and bacterial decomposers varied less
from dish to dish in systems with more species per functional group, Naeem and
Li found. In a similar study of microbial microcosms by Jill McGrady-Steed and
her colleagues at Rutgers University in New Brunswick, the productivity of more
diverse microcosms showed less fluctuation from week to week and less variation
from dish to dish—presumably because as conditions change, different
species step into the leading role in each functional group. If these results
apply to nature on a larger scale, then a pool of “spare” species would help an
ecosystem weather catastrophic droughts or freezes, ensuring that even if pests
or diseases wipe out one dominant species, another will be around to take its
place. And some of today’s minor species may be the perfect candidates to become
tomorrow’s dominant ones as climates change and one ecosystem gives way to
another.
But there is likely to be a diminishing return from ecosystems containing
more and more “spare” species. Most people carry one spare tyre in their car,
for example, and perhaps a second in remote areas, but only the truly obsessive
carry a trailer load. Might the same argument apply for ecosystems? “The problem
is that we don’t have a very long-term vision,” says Naeem. “One spare tyre is
fine for you or me and my car. We know what tyres are, and we know how they
function. But when you’re trying to manage something as complicated as an
ecosystem for hundreds of years, it’s hard to know how many spares you
˛Ô±đ±đ»ĺ.”