Grazers Grass and Microbes

ellowstone National Park has been called "America's Serengeti" for a unique wildlife spectacle that rivals the annual migration of vast herds of zebra and wildebeest across the East African savanna. Each winter, elk and bison congregate by the thousands in the northern third of Yellowstone— the relatively low-lying, sagebrush-dotted grasslands that stretch along the Lamar, Gardner, and Yellowstone rivers and extend across the northern park boundary into Montana. There the animals paw away the snow to expose dried grasses that sustain them for up to 7 months. Around April, when the first flush of green grass emerges on the northern range, the animals graze it heavily and then begin migrating south, following the retreating snow and the advancing wave of greenery upslope.

Within decades after Yellowstone was declared the world's first national park in 1872, some managers and scientists became convinced that these wintering herds were too numerous for the health of the northern range. Perception that the winter range was overgrazed intensified with the drought of the 1930s and has persisted in some quarters ever since. For almost a century, managers aggressively controlled elk numbers, as well as fire and predators, eliminating gray wolves by 1926 and reducing elk numbers to a low of 4,000 by 1962. Soon thereafter, national television images of park rangers shooting elk sparked a public outcry. In 1968, yielding to public pressure as well as shifts in ecological thinking, Yellowstone adopted a largely hands-off policy of "natural regulation," relying on natural processes inside and hunters outside the park to limit elk numbers. Two decades later, the elk herds reached highs of 19,000 and the debate continues among range managers, soil scientists, wildlife biologists, and ecolo-gists about whether elk and bison are degrading the northern range.1

The controversy is part of a much larger conflict over evolving ideas of how to manage grazing lands sustainably, for livestock or for wildlife. The outcome is vital because grassland ecosystems cover about one-fourth of the earth's land surface. Barely two centuries ago, most of these grasslands still sustained large free-ranging herds of wild grazers, whether zebras or elk, saiga antelope on the Eurasian steppe or ecologically equivalent but hoofless kangaroos on the Australian savannas. According to the fossil record, wild grazers such as these have coexisted stably with grassland ecosystems for some 70 million years. Now, however, most of the earth's grasslands have been plowed up for crops or converted to range or pastureland.2 Cattle and sheep have degraded rangelands in many parts of the world in a matter of decades, stripping away plant cover, compacting the soil with their hooves, and leaving bare ground open to wind and water erosion. Today, Yellowstone and Serengeti are among the few surviving semi-natural grasslands where we can learn how grass and grazing animals coexist—lessons that link to life underground and that may help us better manage pastures and rangelands for the future.

The first such lessons began to emerge from the Serengeti in the 1970s, and they were startling. Serengeti National Park in Tanzania and the adjoining Masai Mara Game Reserve in Kenya form a tropical grassland three times the size of the temperate grasslands of Yellowstone and host 50 times as many ungulates (hoofed grazers). Yet the Serengeti's grasses are not passive victims of the grazers. In turn, the zebras and wildebeest that migrate by the millions across these plains do not rely passively on the rain and grasses to determine the amount of forage available to them. Instead, as ecologist Sam Mc-Naughton of Syracuse University showed, these animals actually stimulate grass production. On average, the grazed grasses produce twice as much greenery as ungrazed grasses, and that greenery is younger, denser, and more nutritious in grazed areas.3 Thus, the animals greatly increase the carrying capacity of their own habitat.

"It was some of the most exciting work I'd ever encountered in ecology," Doug Frank says, telling me how he had come across Mc-Naughton's findings in the 1980s. It is a June morning and we are chatting in front of the historic stone post office at Mammoth, the headquarters of Yellowstone.

"He was discovering all these really cool things about how plants and large herbivores in East Africa were highly adapted to one another, and I decided I wanted to see whether these same kinds of feedbacks are occurring in temperate systems like Yellowstone. It's really the only place left where we can get a glimpse of the kinds of ecological processes that occurred throughout the earth's temperate grasslands in prehistory."

At the time, Frank had a master's degree in plant ecology and was studying the effects of mountain goat grazing on alpine plants in Olympic National Park. McNaughton agreed to take him on as a Ph.D. candidate, and Frank spent the next 4 years working in the park each spring and summer as a graduate student. He has been coming back ever since. "This work continues to be very exciting," says Frank. Each season when he arrives in the park, "it's like coming home."

Today he is letting me tag along to one of the northern range field sites he has been using for the past 15 years. As we drive up the winding road past the crowds of tourists at Mammoth Hot Springs, Frank, now an associate professor at Syracuse alongside his mentor Mc-Naughton, recaps what he found in those first years.

The bottom line is that on Yellowstone's northern range, as on the Serengeti, grazing doubles the amount of plant-available nitrogen in the soil and stimulates more abundant and denser grass growth.4 Grasses that have been chomped by elk or bison produced

45 percent more greenery on average during the growing season than ungrazed grasses, Frank found. Further, his findings came in a drought year when the elk population had reached a high of 19,000 and bison, too, were at record numbers.5

"Once you see this very strong positive feedback effect that the animals are having, you have to begin to ask 'okay, but how in the world is this happening?' It really makes no intuitive sense. And that's when we started looking at the soil."

Historically, grazing research has focused on the animals themselves —everything from population densities to bite size to food preferences —and on the plants they consume—the makeup and diversity of the plant community, plant defenses, productivity. Now a growing number of researchers have come to realize that the soil community plays a crucial role in choreographing the complex duet of grasses and grazers. Plants are intimately linked to the soil community through feedbacks involving energy (carbon) transfers, decomposition and nutrient cycling, symbiotic relationships with mycorrhizal fungi and nitrogen-fixing bacteria, and inhibition by soil pathogens and root-feeders, as we've seen throughout this book. It's now increasingly clear that grazing animals, too, from elk to grasshoppers, can influence the feedbacks between plants and the belowground community.6 And humans alter this interaction by reducing or increasing populations of grazers or introducing nonnative grazers to the landscape.7 Thus, in the early 1990s, Frank and a number of other ecologists found themselves pulled, unexpectedly, toward the opaque frontier of the soil.

After only a few minutes, Frank pulls over to the left side of the road and parks the car in a small turnout. It's a quixotic summer day in the Rockies, windy and pleasantly cool, with sudden rainsqualls alternating with what is now brilliant sunshine. Before us, a grassy slope dotted with sagebrush drops down to a small plateau bordered by a dark stand of Douglas fir trees. No animals graze on the plateau.

"This site is grazed intensively in the springtime," Frank points out. "As the snow melts and this area begins to green up, we see a lot of evidence of grazing." But it is June, and the animals have moved up into the mountains and high meadows.

The perennial grasses on the plateau before us still look green and plentiful, and I wonder aloud why the animals migrate so far and return only when this grass has dried to hay and been buried by snow? It's a question Frank and his colleagues have spent years answering.

Essentially, the animals choose the richest, most productive sites in this dramatically variable landscape to graze, he explains. By measuring the total green matter—net aboveground primary productivity— a patch of land produces, and comparing that with how much the animals consume, he found that the more productive a patch is, the more of its grass the animals eat. And they eat the actively growing young shoots as soon as the grass begins to sprout, not waiting for large masses of greenery to accumulate.

"In Yellowstone, the growth of grasslands comes in a pulse of about a month and a half, and the migratory grazers track this pulse, this green wave of young tissue, from low to higher elevations, moving from winter range to higher summer range," he says. "We can see the migration, but we wanted to know what kind of benefit the animals are deriving from it."

The answer he's pieced together over the years is that the animals get more food and better nutrition per bite and thus a better diet for less effort by migrating. It turns out that the nitrogen and mineral content of grass blades on this range peaks a few days after green-up begins in the spring. And the young grass is denser and more highly concentrated than taller, older grasses, allowing the animals to get more of this highly nutritious food per bite.8 Efficient foraging is particularly important in spring when females are bearing and nursing calves.

On the Serengeti plain, migratory herds move seasonally across more than 100 miles of rolling savanna, following the rains and the "green wave" of plant growth that begins in Kenya's Masai Mara and sweeps southeast into Tanzania as the wet season progresses. Just as in Yellowstone, McNaughton has found that this pattern of movement allows young animals and pregnant and nursing females to graze forage with elevated levels of necessary minerals.9

At the richest sites in both Yellowstone and the Serengeti, the grazers may consume more than half of the grass produced each year. McNaughton, Frank, and their colleagues scoured the scientific literature to compare the amount of green matter produced with the amount consumed in a wide range of other ecosystems around the world. The exercise revealed that ecosystems fall into two very distinct categories. One group contains low-herbivory systems such as desert, tundra, temperate forest, tropical forest, and small remnant grasslands without large grazers. Here, on average, animals large and small consume only about 9 percent of the vegetation produced each year. The second group includes high-herbivory systems, mainly large grassland regions such as Serengeti and Yellowstone that still support abundant migratory herds, where consumption averages 55 percent.10

"How can an ecosystem that experiences this kind of intense, high, chronic herbivory be sustainable?" Frank asks. "Because this appears to be the case if you believe the fossil record." This is the overarching question that has brought him back to the park year after year.

As we begin walking downslope, I catch a glimpse of a wire-fenced enclosure below us, not readily visible to visitors from the road. Actually, scientists call it an "exclosure" because the fence is designed to keep elk, bison, and other large grazers out. It's a simple way of allowing researchers to examine grass production and other ecological processes in the same ecosystem with and without grazers. This 5-acre exclosure is one of eight scattered across the northern range from Mammoth east through the Lamar Valley. All were erected more than four decades ago, just a few years before the park stopped culling elk, and all serve as testimony to the long-running controversy over the impacts of grazing here.

What's most striking looking down from above is a contrast that rankles park critics. Dense thickets of 20-foot-tall willows cover much of the ground inside the fence. In the meadow beyond, where elk, bison, pronghorn antelope, and mule deer have been free to feed for 40-plus years, we see only what Frank describes as "little nubbins of willow" here and there.

At the foot of the slope we walk out across the meadow and approach the tall wire fence. The original wooden posts now sag outward in places, and newer metal fence posts have been placed at angles to buttress them. Nearby, a small stream trickles out of the ex-closure. The ground is spongy and damp, green with Kentucky blue-grass, rushes, clumps of wild iris, and a yellow-flowering cinquefoil. Inside the fence, wild rose bushes and irises crowd the openings in the willow thickets. The grass inside appears no higher than out, but dense gray thatch and standing dead grasses mingle with the greenery inside. Frank points out that far less litter accumulates out here in the grazed meadow. The result is dramatically warmer soil temperatures and increased soil microbial activity during the growing season, and colder winter soil temperatures in the grazed areas.

As we walk along the fenceline, Frank explains that soggy meadows or swales like this one readily support willows. If grazers are around to browse the growing tips, however, "a willow will survive, but it's going to remain small." Indeed, across the northern range, grazers have visibly altered the pattern of trees and shrubs since the early 20th century. Willows are short and sparse, and all tree-sized aspen and cottonwoods got their start before the 1920s. Elk browsing has prevented "recruitment"—growth of seedlings or suckers into mature trees and shrubs. And browsing by elk, pronghorn, bison, and mule deer has reduced the extent and height of sagebrush at lower elevations on the winter range.11 Those changes, along with questions about the state of the grasses and the soil, underpin the controversy about whether the herds are degrading the winter range.

Various experts still make dramatically different judgments about the landscape in which Frank and I are walking. To understand how some perceive ruin where others see vibrant health, it's necessary to delve briefly into the changing world of range science. In the 1940s, range scientists developed a system of rating the state of grazing lands that was based on then-current ideas in ecology about what a plant community on a given site should look like. Basically, the thinking was that a plant community on a site develops through a predictable succession of stages toward a single idealized climax or stable state unless it is disrupted, say by excessive grazing.12 The heavier the grazing, the more plant community succession is held back. Range specialists classified the condition of rangeland such as this plateau based on the percentage of the vegetation that matched what was believed to be the natural climax state. If too many climax elements appeared to be missing—say willow and aspen in the case of Yellowstone—just reduce or eliminate the grazing pressure, the idea went, and the damage would be reversed and succession toward a predictable endpoint would be resumed.

Ecologists have since moved away from the notion that all plant communities develop inexorably toward a single stable state, and for the past several decades, range scientists have been struggling over changing standards for interpreting what they see and classifying and managing range conditions.13 Nevertheless, with the elk population soaring in the 1980s, the debate over the state of Yellowstone's winter range grew loud enough that Congress in 1986 directed the National Park Service to "start a study on Yellowstone to see whether there is evidence of overgrazing [and] what should be done to avoid that."14 Work by Frank and a bevy of other researchers has mushroomed since that time. In 1998, again at the request of Congress, the park service asked the independent National Research Council to assess the findings to date.

The research council's 2002 report confirmed that the "condition of the northern range is different today than when Europeans first arrived in the area," and larger numbers of elk and bison share responsibility for that along with human development and possibly shifts in climate. But the report concluded that no major component of the ecosystem was in danger of being lost: "[A]lthough we recognize that the current balance between ungulates and vegetation does not satisfy everyone—there are fewer aspen and willows than in some similar ecosystems elsewhere—the committee concludes that the northern range is not on the verge of crossing some ecological threshold beyond which conditions might be irreversible." And the park's

"practice of intervening as little as possible is as likely to lead to the maintenance of the northern range ecosystem and its major components as any other practice."15

As for the grass: "The composition and productivity of grassland communities in the northern range show little change with increasing grazing intensity."16

Park critics were far from mollified, but in the winter of 1995-1996, a new twist was added to the ongoing saga. Gray wolves were reintroduced to the park after a 70-year absence and quickly flourished. Elk counts from the 1999-2000 to the 2002-2003 winters ranged from 11,700 to 14,500 animals. Now researchers are seeing signs that young willows, aspens, and cottonwoods are taking hold on the northern range, and not simply because wolves are helping to reduce elk numbers. A new wariness about wolves may also be driving elk to change their behavior since they can no longer linger fearlessly to browse along river bottoms and in open meadows where they make easy prey. Ironically, the too-many-elk voices are now being drowned out in the local media by hunters and outfitters who fear that wolves will eat so many elk that there will be none migrating out of the park in fall and winter to shoot.17

With wolves and elk in the public eye, new findings about how the soil community influences the interactions between elk and their food plants have gotten little attention.

Frank has come to the Mammoth exclosure today to arrange for some soil sampling, and he wants to avoid taking samples from sites near the fenceline that were fertilized with nitrogen during a previous project. He drops to his knees and begins crawling along the fenceline, poking his gloved hands into the grass under the wire to feel for metal pins that mark the boundaries of the old plots. I drop to my knees and grub in the grass, too, unsure what the pins might look like. It takes 10 minutes of crawling and searching before Frank spots the first crumpled and weathered copper identification tag attached to a pin. I get down inches from the ground to read the numbers etched on the tag.

"It looks like 'Study #GLLoi668'," I report. "Then it's got a number like 1-0034."

"Could that be 0037?" he says, studying a plot diagram. Yes, I allow.

Based on that number and the diagram, we search for the other corner pins.

This site and a handful of others throughout the northern range have been sampled and probed extensively over the past two decades, but except for the corner pins, few tangible signs of the studies remain. Since the mid-1990s, Frank and his colleagues have been delving beneath this ground to find out just how, as the National Research Council noted, these grasses can be so little affected by being literally half eaten each spring.

One plus for the grazed grasses is the dung and urine the animals leave behind. Grass wastes processed through the guts of elk and bison are much more easily broken down by decomposer microbes than uneaten grass litter. Thus, nutrients in these waste patches are rapidly made available to fertilize new plant growth. But Frank and others believed that this effect was too spotty to explain the luxuriant boost in grass growth they were documenting across the grazed landscape. The grazers had to be triggering some other positive effect be-lowground.

"It's kind of like peeling the layers of an onion," Frank explains. "You figure out the exterior layer, like enhanced productivity with grazing, but that's being controlled by something below. So you try to peel that next layer away just to get at the nexus of what might be happening. It's all very exciting."

And that layer led you into the soil? I ask.

"Kicking and screaming," he says, laughing easily. "The problem is, that's really hardcore biogeochemistry. If what turns you on is being out in the field, then when you decide that you want to do soil process work you have to spend an awful long time in the laboratory with pot experiments and extracting soils with dilutions, all that stuff."

Fortunately for Frank, "an extremely talented lab junkie" named Bill Hamilton was working as a postdoctoral researcher in his lab when he decided to find out just what happens belowground at sites like this each spring. Since herds graze a site early in the growing season before moving on, the ground is still moist and spring conditions are favorable for the grasses to rebound and compensate for the lost greenery. The question is, how do they go about it?

Hamilton took samples of Kentucky bluegrass (Poa pratensis) and fresh soil from a grazed area like the one where we're crawling around. In a greenhouse at Syracuse, he grew the grass in pots filled partly with field soil so that their microbial communities would be similar to those under grazed grasses on this range. Then he clipped the grasses with scissors to simulate grazing. To track what happened next, Hamilton spiked the carbon dioxide the potted grasses were breathing with a carbon-13 (13C) isotope.

You might imagine that the first act of a wounded plant would be to devote all the carbon it could capture through photosynthesis to building new greenery in order to capture more light, grow more stems and leaves, and so on until it had compensated for the lost growth. But as Hamilton tracked the fate of the 13C with a mass spectrometer, he found instead that all the action went underground.

Within 24 hours, the clipped grasses were pulsing five times as much sugar and other carbon-based substances from their roots into the soil as the unclipped grasses. This sudden bounty quickly spurred a dramatic population boom among soil microbes in the rhizosphere adjacent the roots, and these microbes set to work decomposing organic matter and releasing nitrogen and other nutrients. Later work showed that the increased root exudation and microbial activity lasted about 3 days before starting to taper off. But that frenzy of decomposition and increased nutrient release was enough to launch the plants toward regrowth and recovery. After a week, the clipped plants enjoyed higher nitrogen uptake, higher nitrogen content in their leaves, and a 24 percent higher photosynthetic rate than unclipped plants.18

Feeding extra sugars to soil microbes must be quite an expensive investment for an injured plant, I comment.

"Yes, so it's a very short-term pulse, and that makes sense," Frank replies. "But grasses exude a lot of carbon anyway, up to about 20 percent of the carbon that they assimilate, so it must be important to them." The strategy of ramping up carbon exudation in response to grazing makes sense if the grass finds nitrogen in shorter supply than carbon.

"That seems kind of counterintuitive," he says. "But I think that's what's happening. Before the plants are grazed, their growth is limited by the amount of carbon they can assimilate through photosynthesis. So they're concentrating on producing stem material and trying to grow tall and reach the light, in competition with their neighbors. Once the herd goes through and the plants are grazed, they're all grazed to about the same low level and there's less competition for light. So carbon may not be limiting any longer. What become limiting are the soil resources. And I think that's why you see grasses after they're grazed allocating this carbon resource belowground."

Nitrogen, as gardeners know, is the primary fertilizer of plant growth, and it's a limiting resource in all grassland ecosystems.19 Several studies have found the same postgrazing response in other members of the grass family, including corn plants and blue grama grasses attacked by munching grasshoppers,20 and even grassland plants attacked by root-feeding nematodes.21

"It appears that this is a robust response that grasses have to being defoliated," Frank explains. "So the question was, why are grasses doing this? They're losing their photosynthetic machinery by being grazed. It would seem that carbon is not what they want to lose, so why are they exuding it?"

Other studies had found that grazing can increase the number and activity of microbes in the rooting zone.22 By labeling the carbon and tracking the extra pulse into the soil, Hamilton was able to link it directly to both the microbial population boom and accelerated nutrient cycling.

The plants seem to get a second payoff on their investment when the microbial masses die or go dormant. "Adaptively, it makes sense for the plant to cultivate this large population through exudation,

Grasses grazed by bison or elk in Yellowstone National Park quickly ooze extra sugars from their roots, spurring activity throughout the soil food web, accelerating nutrient cycling, and enhancing release of nitrogen that fertilizes new grass growth. Dung and urine patches also speed decomposition and nitrogen release.

Grasses grazed by bison or elk in Yellowstone National Park quickly ooze extra sugars from their roots, spurring activity throughout the soil food web, accelerating nutrient cycling, and enhancing release of nitrogen that fertilizes new grass growth. Dung and urine patches also speed decomposition and nitrogen release.

then cut off the supply. The gravy train is over, they die, and you can get another big pulse of nitrogen out of the decaying microbes."

Of course, a microbial community in the soil doesn't boom in isolation. Tiny predators such as nematodes and protozoa arrive to graze on the microbes, and other soil animals such as mites or spiders consume these tiny predators, and so on, so that the nitrogen in mi-crobial carcasses may cycle through a complex underground food web before becoming available to fertilize plant growth. That happens when decomposer microbes finally break down the carcasses of these soil creatures and release the components into the soil in inorganic forms such as the nitrate that plants require. Although Frank and Hamilton did not investigate this cascade, an earlier study on the Yellowstone winter range found that the density of both bacterial-grazing and root-feeding nematodes was greater under grasses grazed by elk and bison than under ungrazed grasses.23 And others have reported increased activity by microbe-consuming soil animals in the rhizo-sphere of grazed plants.24 The consumption of microbes by nematodes, protozoa, or springtails actually increases the growth of both microbial populations and plants.25 Indeed, underground consumers such as nematodes can speed up nitrogen cycling by 20-50 percent over systems with only microbes.26 Thus, elk and nematodes may unwittingly interact to stimulate the growth of plants on whose productivity both ultimately depend.

"My big picture take is, plants are capable of manipulating the microbial activity in the rhizosphere to increase nutrient availability, and they do that in response to grazing," Frank sums up.

But the amount of carbon that grasses share with their microbes, routinely or in times of stress, is just one indication of how much effort grasses here expend on cultivating their ties with the soil. Frank reminds me that grasses always invest more effort in building roots than greenery.

"There's always more biomass belowground. Always. In grasslands, most of the allocation is belowground because water is a big limiting factor, so they have to allocate a lot of biomass to roots in order to get water," he explains. Nitrogen, too, as we've just seen. Yet until recently, most studies have shown that aboveground grazing reduces root production.

Frank and his colleagues recently discovered that, to the contrary, Yellowstone grazers stimulate productivity belowground as well as above. At this exclosure and others across the northern range, his research team sank clear tubes known as "minirhizotrons" into the soil both inside and outside the fences. Using a tiny video camera and a light source inside the tubes, they were able to zoom in on a field less than i/ioth of an inch square and photograph individual grass roots growing along the outside of the tubes. By photographing the same roots month after month and digitizing the photos back in the lab, they were able to calculate the growth or shrinkage of each root during the season.

"It's painfully time-consuming, but it's really the only accurate way to monitor belowground production," Frank says. "And to our great shock, we saw greater root production as well as aboveground production in the grazed plots. In fact, the rate of stimulation by grazing was seven times greater belowground than aboveground."27

When you put all these bits and pieces from 15-plus years of research together, it's possible to visualize a remarkable cascade of events taking place largely unseen on this range each spring as elk and bison crop the grass low and leave. The timing is just right: The soil is still moist from snowmelt, the sun is quickly warming the long-frozen ground, and the growing season is young. Structurally, the grass is primed for growth: Its growing tips hug the soil surface, safe from the teeth of grazers and ready to send out lateral shoots, or tillers. The taller, older greenery has been stripped away, allowing the new shoots with their higher photosynthetic rates greater access to light. The herds have tilled the earth with their hooves. Belowground, microbial populations are burgeoning as the creatures feast on highly digestible carbon: not only the extra bursts of sugars, carbohydrates, and proteins exuded by the roots of the injured grasses but also patches of dung and urine deposited across the surface by the animals. The microbes are eating and being eaten, fostering a complex underground food web that accelerates the nutrient cycle and frees up generous amounts of nitrogen. The grasses, in turn, soak up sunlight and nitrogen, build new roots, and recover their lost greenery.

After crawling around in some of this postgrazing greenery for a half hour, we've located only three copper-flagged pins and one un-flagged marker. But Frank feels that's enough to guide future sampling and avoid the fertilized plots. Besides, the sky is darkening and the wind is picking up.

As we head back upslope to the car, I ask whether he's excited about the soil organisms that have begun to enter his work and his life.

"Oh yeah, I am, very much so," he says, laughing easily. "I've just arrived there, realizing the soil is an important key to how everything else in the system is operating."

Each finding, it seems, is taking him farther underground. After discovering how grazers and grasses can manipulate activity in the soil community, Frank had begun to wonder whether thousands of years of such interactions might have altered the character and identity of the soil community in grazed grasslands. That's the layer of the onion he's peeling back now.

We reach the car and drive back downhill to the restaurant at Mammoth. Soon after we're seated, the sky darkens and a curtain of rain sweeps across the landscape, settling in for a long downpour that drives gaggles of wet tourists inside to join us.

I ask him about the next layer of the onion, which began with an idea for an experiment that came to him while attending a workshop on plant-herbivore interactions in Sweden.

"I remember describing this pot experiment to some folks over beers, and they said 'ah that's never going to work,'" he recounts, laughing. "They thought it was kind of a harebrained notion, but I did it anyhow, and it turned out really interesting. We asked a simple question: Do grazers affect the composition of the soil microbial community?"

As soon as he returned to Syracuse, Frank arranged to have soil samples and grasses taken from inside and outside the Mammoth ex-closure we have just left. The soil samples were chilled and shipped overnight along with the grasses to his lab in New York to ensure that the soil community remained intact. The experiment itself was straightforward. For 2 months, he grew grasses from inside and outside the exclosure separately in pots of soil from either inside or outside. The results were dramatic.

"I get a little tingling down my spine when I think about it," he laughs. The grasses from both sides of the fence did much better— meaning a 34 percent increase in aboveground growth—when they were grown with the soil community from the grazed areas outside the fence.28 "This is exciting because it suggests that somehow grazers are changing the composition of the microbial community, and this community is functionally different because it's facilitating plant growth."

But what was different about the community? Frank treated some of the pots with fungicide and the differences in plant growth disappeared, suggesting that the fungal community was responsible for the effect. Specifically, he believes the arbuscular mycorrhizal (AM) fungal community in the grazed areas may somehow provide more effective symbiotic partners for the grasses than the AM fungi on the roots of ungrazed grasses. Indeed, collaborator Catherine Gehring from Northern Arizona University extracted AM fungal spores from the soil and found that the abundance and diversity of spores were greater in the grazed soils and that the identity of the players in the spore community was also different. (Recall from the previous chapter that AM fungi are the most widespread of the mycorrhizal sym-bionts. They are much harder to study than ectomycorrhizae because they do not form mushrooms and their hyphae form branching structures called "arbuscules" within the root cells of the host plant rather than forming visible structures on the root surface.)

Frank believes that since grazing spurs root production and carbon dumping by Yellowstone grasses, this may also increase the "investment" the grasses make in the mycorrhizal partnership. And the AM fungi in turn may help the grasses make faster and better use of the nitrogen being liberated by the enhanced bacterial activity around their roots.29 The story is still unfolding, however. An earlier study found that grazing actually disrupts the mycorrhizal partnership.30

Despite the decisive effect of the fungicide on his pot experiments, Frank is not ready to write off the potential for important changes in the bacterial community in grazed grasslands, too. Wiping out the fungi might have eliminated closely associated bacteria that also have a hand in boosting plant productivity.31

More than a year later, I learn that one of Frank's graduate students, Stacey Massulik, and Syracuse soil microbial ecologist Andria Costello are using genetic probes to determine the richness and composition of the bacterial community inside and outside a number of exclosures. They are confirming that the composition of the bacterial community is indeed different in grazed and ungrazed areas, although the diversity of bacterial species is similar. More than half the bacterial species are found in either the grazed community or the ungrazed community, but not in both. Another of Frank's students, Tanya Murray, is finding evidence that grazers also have strong effects on the composition of the mycorrhizal community, although the diversity of AM fungi appears unaffected.

This work comes at a time when a growing number of researchers around the world are investigating the ecological consequences of changes in the identity or biodiversity of the living communities underground. Fungal disease agents in the soil, for instance, can play a powerful role in shaping the aboveground plant community by weakening their hosts and permitting their replacement by or coexistence with other plants.32 The identity and talents of my-corrhizal fungi in the soil community can also affect the diversity and productivity of the plant community, as well as the competition among plant species.33 Many other soil denizens, from microbes to nematodes, may also influence the makeup and workings of the above-ground community, through either negative or positive feedbacks.34

Clearly, large grazing animals in Yellowstone, with the interaction and cooperation of grasses and soil creatures, influence energy (carbon) and nutrient flows in a way that has made these grasslands sustainable over time. But that doesn't mean these ecosystems or others can't be disrupted.

"You can degrade any system, drive it down, if grazing intensities are too great," Frank tells me. "But at the current rates of her-bivory, that's not what we're seeing here. We're seeing stimulation."

I ask how managers could monitor or judge whether grazing had reached its limit.

"Reduced productivity would be a good indicator of overgrazing here, although measuring production in a grazed grassland is very time-consuming and expensive," he says. Certain types of human management practices are likely to decouple grazers from coevolved feedbacks with plants and soil. Herding and fencing of animals, for example, alter the natural migration patterns and the seasonal rhythm and timing of grazing on a range. Stocking cattle and sheep at high densities with the help of predator control, supplemental food and water, and veterinary care can also disrupt the natural feedbacks between grazers and the plant and soil communities. Ironically, without the natural intensity and timing of grazing to stimulate productivity, grazing-tolerant grasslands such as Yellowstone and the Serengeti may actually decline in carrying capacity and support fewer animals.35

The message from Yellowstone is not that grazing is "good" or "bad" for the range, but that we cannot understand how ecosystems will respond to grazing and other disturbances without considering what is happening below as well as above the ground. Plant-eating animals interact in complex ways with other forces, from climate and geology to plants and soil creatures, to shape ecological processes in grasslands, forests, and other systems. Exclosure experiments involving reindeer in northern Scandinavia, exotic deer in New Zealand forests, sheep and deer in the Scottish Highlands, moose in the boreal forests of Michigan, and many other habitats around the world show that grazers can have positive, negative, or even neutral effects on plant productivity and nutrient cycling.36

On the high arctic island of Spitsbergen, for instance, wild reindeer promote nutrient cycling and grass growth on the moss-dominated tundra largely via the dung they deposit. Richard Bardgett of Lancaster University in the United Kingdom and his colleagues added reindeer dung pats to some plots one summer and watched for three years before the impacts began to appear. By then, the patches where the dung had been added boasted a greater standing crop of grasses and enhanced microbial biomass in the soil. Also, the moss layer on these sites was significantly reduced, probably because of the enhanced activity of microbial decomposers.37

In the tallgrass prairies of the North American heartland grazing bison can spur positive feedbacks that increase plant species diversity, alter patterns of plant productivity, and speed up nitrogen cycling, apparently counterbalancing the effect of prairie fires and other stresses.38 Yet there are many instances around the world where large grazing animals reduce the productivity and diversity of plant communities, including a number of arid grasslands where heavy grazing has led to the spread of shrubs and trees at the expense of the

0 0

Post a comment