A Landscape of Fear
By Bruce Fellman
Picture yourself in a meadow in high summer. Tall downy grasses, amid a colorful patchwork of buttercups, knapweed, Indian paintbrush and butterfly weed, sway languidly under a sultry sun. Monarch butterflies, hummingbird moths and droning bumblebees skitter atop the flowers in jittery haste for nectar. Thoreau, you fancy, could have been inspired in that meadow to write “The Inward Morning”: How could the patient pine have known / The morning breeze would come, / Or humble flowers anticipate / The insect’s noonday hum ... It is the picture of peace.
Don’t believe it. Just below the brilliant spatter of green, yellow, purple and orange are scenes of wanton violence that often end in sudden death. A spider, its eyes glinting in the light, feels movement and gnashes its jaws in anticipation, waiting to ambush a meal. A nearby grasshopper inches along, calmly munching on grass and a wildflower, unaware of its fate.
It won’t be long.
Or maybe it will. The grasshopper feels something, too, and stops its forward progress. Like a human being about to confront a predator of the same species around the corner, the insect starts behaving in a way we’d find very familiar. Inwardly and outwardly, the grasshopper seems to be afraid.
Dror Hawlena, a Gaylord Donnelley Environmental Postdoctoral Associate at F&ES, is piecing together how grasshoppers react to spiders that are intent on making a meal of insect prey. The grasshopper’s physiological response to potential doom is, perhaps surprisingly, “quite similar to our own,” says Hawlena. “Grasshoppers may have different glands, but their stress reaction involves the same hormones and endocrine mechanisms found in humans. In fact, you see this in all vertebrates—and we’re now finding it in invertebrates, too. It’s one of the most fundamental responses, and the most critical. If you can’t respond quickly in an emergency, you die.”
That response, it turns out, is critical in another way. “How a grasshopper avoids a predator can lead to the transformation of an ecosystem,” says Oswald Schmitz, Oastler Professor of Population and Community Ecology. Schmitz, who oversees Hawlena’s research, along with that of numerous graduate students and undergraduates, is a pioneer in an effort to understand how fear and other so-called nonconsumptive effects (NCEs) help determine the species composition, functioning and structure of the natural world.
At its most basic, Schmitz’s work, now in its 17th year in the meadows of the Yale-Myers Forest, involves monitoring and recording what happens during the summer in a series of screened-in cages, each roughly one square meter in area and one meter high. Inside these microcosms, various combinations of spiders, grasshoppers and plants go about their business.
“This is a wonderful model system,” says Schmitz. But, he quickly adds, the research is about much more than the natural history of the organisms that he and his colleagues have had under consider-ation since 1993. “What we’re getting out of these small enclosures are broad conceptual insights that we can apply to much larger systems.”
Natural resource managers and policymakers alike can use fear factors and other NCEs to predict, with a much greater degree of accuracy than has previously been possible, what our impact on the planet might be. Taking these fundamental but, until recently, overlooked processes into account can, says Schmitz, lead to better ways of carrying out such seemingly unrelated projects as the management of timber harvests in the boreal forests of Canada, the reintroduction of wolves and other top predators to the national parks and the implementation of biological control programs in agriculture.
“You want to avoid unintended consequences,” says Schmitz. For example, when wolves were brought back to Yellowstone, one impetus, besides the restoration of a missing part of the ecosystem, was to put the brakes on an overabundance of elk. However, things didn’t work out quite as planned. To be sure, the wolves killed prey, but in short order the elk dealt with their newfound fear of predation by changing their feeding patterns and moving away from the more open range areas.
“Elk are important for maintaining the species diversity and productivity of grasslands,” says Schmitz. “There’s evidence that both declined after the wolves were brought back.”
That evidence was clearly present in the enclosures in the Yale-Myers meadows. But instead of wolves, which were in short supply in Connecticut, Schmitz worked with Pisaurina mira, the nursery web spider and a dominant predator in the meadows. There were other spiders haunting the grasses, but P. mira, which employs a sit-and-wait hunting strategy, was simply the one he encountered first when he began his work. “Choosing it turned out to be pure serendipity,” he recalls.
The most common grasshopper in the field, and the one Schmitz used in the research, was a red-legged species known as Melanoplus femurrubrum. Its favorite food is Kentucky bluegrass (Poa pratensis). The insect was less inclined to eat goldenrod, which was also very common, or a variety of wildflowers, such as Queen Anne’s lace, that flourished in the area and were well-represented in the enclosures.
When he started tallying up his initial results, he noticed something odd. “The spiders had no net effect on prey density,” Schmitz explains.
If the predators weren’t eating many of the herbivores, then there shouldn’t have been much difference between the enclosures that contained spiders and grasshoppers and those with grasshoppers alone. But he found measurable changes in prey behavior when spiders were present, and, by the end of the season, there were significant differences in the diversity and abundance of grassland plants between the two kinds of enclosures. The spiders, he said, were leaving a “predation signature.”
Schmitz had stumbled upon the unexpected ecological consequences of a landscape of fear.
“Many biologists grew up viewing prey simply as animals waiting to be eaten, and if you didn’t get eaten, you were fine,” says Evan Preisser ’98, an assistant professor in the Department of Biological Sciences at the University of Rhode Island and a co-editor of a 32-page special feature on NCEs in the September 2008 issue of the journal Ecology. “But that’s akin to saying that if you were in a mall when a fight broke out and someone got shot, you were okay as long as you weren’t hit. In reality, even though this had no direct impact on you, it probably would have terrified you and, as a result, caused you to make all sorts of changes in your life. And if everyone present that day reacted the same way, even small changes in behavior manifested over a large population could have profound and unexpected effects. NCEs are, in essence, the wages of fear. The genius of Os’ work is that it gives us a much more nuanced view of reality.”
In the enclosures that had spiders, the grasshoppers changed both what they ate and when and where they ate it. P. mira is typically most active early in the day, but when the spiders and the grasshoppers were together in the cages, the potential prey shifted its peak activity and feeding period from morning to midday. The grasshoppers also tended to congregate in the relative safety of the upper reaches of the goldenrod. These tactics helped the insects avoid their predators, but at a cost—the increased risk of stress, sometimes fatal, from the heat. In addition, the spider-confronted grasshoppers ate considerably less of the bluegrass they preferred and turned more to the goldenrod and the wildflowers.
By the end of the summer, the mini-meadows without spiders had different plant communities than those with resident arachnid predators. In the former, goldenrod, which can dominate its competition, held down the abundance and diversity of the grasses and wildflowers, such as clover, black-eyed Susan and wild strawberry. But where the nursery web spiders sat patiently and waited for a grasshopper meal, there was less goldenrod and a greater variety of the other plants.
This pattern held even in enclosures that contained spiders whose jaws had been glued shut by the researchers so that they could no longer hunt. The fear they continued to strike in their prey was enough to push the stressed grasshoppers into predator-avoidance mode—and the plant community into an altered state. (The operation, by the way, requires a gentle touch—the spider is held between two soft sponges while a dab of super glue is squirted between the jaws. This experimental strategy is possible only because nursery web spiders can go as long as two months between meals without starving to death.)
Schmitz looked at results he called “counterintuitive” and scratched his head. “If you’re not losing grasshoppers to predation, how can the spiders be affecting the plants?” he asked.
Then came an “Aha!” moment—the result, ironically enough, of a graduate school research project in behavioral ecology—and a bitter disappointment. A Canadian by birth, Schmitz started his research career in the late 1970s at the University of Guelph, where he studied the foraging behavior of deer and evaluated whether supplemental feeding programs were cost-effective. As a doctoral student at the University of Michigan, he examined deer foraging from an economics viewpoint, using a then-promising analysis tool called portfolio theory, to see whether deer were maximizing their returns by “investing” in a broad portfolio of food items that would minimize the risk of starvation.
“Unfortunately, by the time I finished my doctorate in 1989, this approach had been eclipsed by people interested in how avoiding the risk of predation shaped the behavior of animals,” says Schmitz. “I vowed to get out of behavioral ecology.”
But a landmark 1989 paper in Ecology by Thomas Schoener, a University of California at Davis community ecologist, changed his mind. Schoener, whom Schmitz calls his “academic grandfather,” and his colleague David Spiller looked at what happened to the animal and plant life on small islands in the Bahamas when the researchers manipulated the number of lizards present in experimental enclosures. The lizards preyed on web-building spiders, which, in turn, preyed on insects. Changing the number of lizards present not only had a direct impact on the spiders, but it also had an indirect effect on prey insects and the plants they consumed.
“The Schoener paper showed the value of looking at predators not just in terms of their interactions with prey, but also in terms of the myriad effects they could have in ecological communities of carnivores, herbivores and plants,” says Schmitz. “It offered me a way toward a more holistic perspective of entire systems.”
Aquatic biologists such as Bobbi Peckarsky, then at Cornell, had documented various aspects of these so-called trophic cascades in streams while investigating the interactions of trout, stoneflies and mayflies. But in the early 1990s, Peckarsky explained that her work and other aquatic studies like it weren’t commonly on the radar screens of terrestrial ecologists, who were still enmeshed in research about the direct effects of predation. “Os was the exception,” she says.
Schoener’s findings in the Bahamas and Schmitz’s conversations with Peckarsky and others provided “the impetus to get into food web work,” Schmitz says. Still, when he came to Yale in 1992, there was considerable doubt that indirect effects played a significant role in shaping terrestrial systems. “But no one had really looked,” he says.
The more Schmitz pondered his early results, the more he realized that the changes he observed in the grassland plants were not the direct result of prey being consumed and, thus, of there being fewer hungry herbivores in the meadow. Rather, he had documented an indirect effect: a predator causing a shift in prey behavior that, in turn, cascaded through the plants, as the grasshoppers switched to a less nutritious, but safer, diet.
This discovery and years of subsequent refinements of the key role played by what are now known as trait-mediated interactions—think, fear—almost didn’t happen. And here’s where serendipity came in. There’s another common spider in the meadow: Phidippus rimator. This jumping spider is an active predator, always on the hunt, and, unlike its sit-and-wait co-conspirator, never lingers in the same place for long.
When Schmitz examined this hunter’s impact on grasshoppers in the enclosures and in grasslands alike, he got different results. P. rimator knocked down prey numbers considerably, so there were fewer grazers to munch down the bluegrass. But because it was impossible for the grass-hoppers to determine where this predator was, the prey didn’t shift its behavior and diet toward the goldenrod and wildflowers.
Goldenrod is the dominant competitor in the meadow, and without herbivores to keep it somewhat in check, it eventually overshadowed its fellow plants. Paradoxi-cally, fewer grazers resulted in lower species diversity.
“Had I started working with the jumping spiders, I might never have encountered the behavior-shift story,” says Schmitz.
And he and Hawlena might never have become interested in looking at fear factors. The researchers already knew that grasshoppers reacted to the presence of a sit-and-wait spider by changing their behavior and moving elsewhere. The situation is very similar to what a human would do when walking down a dark street in a less-than-sterling neighborhood—first chance you get, you cross over to a seemingly safer and brighter location rather than stroll by that creepy-looking character hanging out on the corner.
“We don’t understand how grasshoppers detect the location of the spider,” says Hawlena, “but when we subject them to the risk of chronic predation, they react just like we would.”
Hawlena placed each overwrought insect inside a metabolic chamber—a transparent cylinder about 3.5 inches long and three-quarters of an inch in diameter—used to measure carbon dioxide output. “They breathe faster and their metabolism increases,” he explains.
If there’s no escape from a spider—Schmitz and Hawlena modeled this in field and laboratory setups, but it could easily happen in nature if spiders or any other predator, for that matter, were especially abundant—the grasshoppers do something else familiar. They crave carbs.
“Stressed grasshoppers need more energy,” says Hawlena, “and they get it from eating more goldenrod.”
Left to themselves, the grasshoppers would rather spend their time munching on nitrogen-rich grass that stimulates their own growth and reproduction, as well as results in high-quality fertilizer. But with spiders in the picture, their stress increases and their diet changes. “Their immediate needs become more important than the future,” says Hawlena.
They head for the goldenrod, whose chemistry—the carbon-to-nitrogen ratio is tilted toward carbon—provides the grasshoppers with the increased carbohydrates they require. So the insects cope with the metabolic demands of stress, but, according to research currently being reviewed for publication, there’s literally no free lunch. Grasshopper mothers stressed out by chronic predation risk having offspring that are 15 percent smaller than those of mothers living in spider-free zones. Also, mothers in the risk zone produce kids with poorer jumping ability. “This results in offspring that are more vulnerable to predation,” says Schmitz.
There’s another cost, as well—this one to the meadow. When it comes to fertilization, nitrogen trumps carbon. Compared to their grass-fed, living-is-easy counterparts, grasshoppers on the high-carb, anti-stress diet produce excrement that is harder to break down and less nutritious for plants. Grasshopper bodies even reflect this change in diet; in death, the corpses containing more carbon don’t offer the premium fertilizer benefits of those that had more nitrogen in them.
Schmitz estimates that grasshoppers, living and deceased, contribute between 5 and 10 percent to a grassland’s overall nutrient budget. So more spider-induced herbivore pressure on the goldenrod not only can alter the diversity and abundance of plant species in the meadow ecosystem, it can also result in poorer fertilizer. This, along with changes in the composition of the leaf litter and the soil bacteria, may eventually change the kinds and amounts of minerals available in the ground—and the varieties of plants and animals that can thrive.
And so, says Schmitz, “being scared to hell can significantly change an ecosystem’s nutrient budget and transform the way it looks and functions.”
Schmitz has synthesized nearly two decades of experimental and theoretical work in Resolving Ecosystem Complexity, a Princeton University Press book due out this summer. “The beauty of this work is that with the insights we’ve gathered, we can predict what an ecosystem will look like if we change it in some important way,” he says.
In the vast boreal forests of Canada, one company is already putting these insights into practice. Mistik Management oversees the planting and harvest of softwood trees on the company’s nearly 4.5 million acres of land in northwestern Saskatchewan. The typical logging operation, says Schmitz, who has worked with Mistik, involves clear-cutting and then replanting primarily with white spruce saplings. But there’s always been a problem with this strategy. “Aspen trees tend to come back first in a clear-cut and choke off spruce growth,” he says.
Based on his work in the enclosures, Schmitz had an idea. While moose and deer did a fine job of grazing aspen sprouts, the herbivores largely avoided young spruce, which are filled with noxious chemicals. But with wolves frequenting the clear-cuts, the fear-filled moose and deer stayed far away.
In an experiment, which involved acres instead of meters, Schmitz compared clear-cuts with areas in which Mistik cut most of the timber but left patches of standing trees. “These became refuges for the moose and deer,” he says.
The herbivores tended to avoid the open areas, and even when they entered them they were ineffective at keeping the aspen out. Not so when there were refuges. “By changing the way we harvested the landscape, we were able to let nature effectively manage for us,” says Schmitz. “And at no real cost.”
Mistik has since adopted the refuge technique. For this company, looking for a more sustainable way to manage its resources, the landscape of fear has become a landscape of opportunity.