horizons

Stressed Out by Earthquakes

Give Ross Stein a tabletop, some sandpaper, two bricks, a cable winch, and-most importantly-a rubber bungee cord, and he'll show you how the Earth makes earthquakes. Think of the bricks as masses of rock along a fault. The sandpaper mimics the friction generated by rubbing rocks, and the bungee cord stores and releases elastic energy just like rocks in the crust do.

To demonstrate, Stein, a geophysicist with the U.S. Geological Survey (USGS) in Menlo Park, aligns the bricks atop the sandpaper on the table and then connects them with the bungee cord to each other and to the winch mounted at one end of the table. He cranks the winch, tightening the bungee cord until the force pulling it equals the force holding the bricks in place. Then the bricks shift sideways with a sudden jerk.

Repeating the experiment a few times produces a staccato sequence of quakes. This somewhat crude simulation illustrates what Stein calls the "bungee cord connection": the idea that as an earthquake relieves stress from one part of a fault, it shifts stress to somewhere else. All things being equal, he contends, wherever stress increases even slightly becomes the next likely location on a fault to break.

By studying the pattern and distribution of earthquakes in California, Japan, Turkey, and other seismically active spots, Stein and his colleagues have developed a provocative approach to explain why quakes occur and even to forecast where-if not when-to expect them. Their hypothesized process, known as "Coulomb stress transfer," boils down to building up too much or too little stress. Any given earthquake, they say, can enhance or suppress the likelihood of future shocks, which seem to cluster in time and space.

"A fault that's close to failure is very sensitive to these small [stress] changes," Stein says, "and these small changes can trigger large earthquakes." Stress increases as slight as one bar-roughly equal to half the air pressure in one car tire-can be enough to push fault segments past the breaking point. One recent study even suggests that ocean tides can exert enough stress change on faults to cause a modest rise in the number of earthquakes.

Unable to measure stresses directly in the earth, Stein and collaborators at the USGS, Woods Hole Oceanographic Institution, and abroad take data from actual earthquakes, such as ground displacement and aftershock distribution, and use them to create computer models to help visualize where the ground has stretched and where strain has been trapped or released. From the models, they can calculate the probability of a given fault segment failing in the near future.

"It's interesting and exciting, but we're still trying to quantify just how well it works," says Caltech seismologist Tom Heaton about stress transfer. The idea makes sense and seems to work in some cases, says Heaton, but the method has yet to show a clear connection between stress and earthquake triggering. Nevertheless, he says, "it's one of the most hot and promising topics" in seismology.

California's 1992 Landers earthquake marked a watershed event in stress transfer research. Some 10,000 aftershocks occurred within a month of Landers, mostly in areas where calculations indicated the compounding of seismic stress. Countering the traditional view that the location of aftershocks follows the actual rupture path of a fault, Stein says, "We think aftershocks are simply telling you where stress has increased, and that can be well off the fault." Since Landers, stress transfer proponents have also proposed the existence of antishocks, quiet regions where earthquakes decline.

Some experts still doubt that a force as powerful as an earthquake could be turned on or off by seemingly insignificant stress changes. "Seismologists were really revolted by this idea," Stein recalls of the initial response. But the theoretical ground has been shifting, he believes, toward wider acceptance of stress transfer as a fundamental feature of earthquakes.

Consider, for example, last August 17 when the North Anatolian fault ruptured at Izmit, Turkey, unleashing a magnitude 7.4 quake that crumbled and toppled buildings and caused more than 15,000 deaths. Two years earlier, Stein and two colleagues had published a study about that fault's tendency to fail sequentially like a line of toppling dominoes, one temblor apparently setting up the next.

Between 1939 and 1992, the fault had generated ten earthquakes along more than 600 miles. "This classic case of progressive earthquake failure," the researchers wrote, "has tantalized earth scientists and terrorized insurance providers, for half a century." After presenting their calculated, cumulative stress changes-each quake, on average, transferred up to a fifth of the stress released at one location to another-the researchers forecast that the North Anatolian fault could fail near Izmit anytime within 30 years. The next domino appears poised to fall uncomfortably close to Istanbul, Turkey's densely populated capital, where Stein will return this spring to conduct further research.

Besides being an ideal testing ground for the stress transfer hypothesis, the North Anatolian fault may hold important lessons for California. The San Andreas fault shares several striking features with its restless Turkish counterpart: They have similar ages, lengths, and direction of motion; both have a slow creeping section in the middle and release temblors of similar strength.

But unlike the North Anatolian, the San Andreas resides in an intricate network of faults, which may interact in curious ways. In a review of seismic stress transfer research published last December in Nature, Stein describes how a stress shadow has settled over the San Andreas's Bay Area segment since the 1906 San Francisco earthquake. For 75 years prior to 1906, the San Andreas rocked the Bay at least 14 times, but the 75 years after 1906 saw no major quakes. This suggests that stress changes can have long-term effects on seismic activity.

Geophysicists like Stein must sometimes feel like ambulance chasers rushing to the scene of the latest disaster to learn what they can before the next big one. But in order for novel ideas like stress transfer to stand or fall, the Earth has to keep shaking. "Once we leave the world of bricks and bungees," says Stein, "we're really up a tree until another earthquake occurs."

When Coyote's Away, Cats Will Prey

RECENT RESEARCH IN SOUTHERN California's coastal sage scrub and chaparral suggests that small predators can have surprisingly big impacts on native birds, while large predators actually boost populations of birds and other potential prey. Biologist Kevin Crooks calls this seaside region "one of the world's largest epicenters of extinction," but relatively wild, steep-sided canyons break up the densely developed mesas surrounding San Diego. And in at least some of these fragments of natural habitat, the ever-adaptable coyote is still top dog. If coyotes get edged out and vacate a canyon, though, all hell breaks loose-ecologically speaking.

When the top predator disappears, says Crooks, smaller carnivores like foxes, skunks, raccoons, and house cats tend to take over, with consequences that cascade throughout the ecosystem. Even minor increases in the numbers and activity of these so-called mesopredators help push populations of such birds as the California gnatcatcher, California quail, and cactus wren closer to vanishing from a canyon. Roadrunners have already been forced from all but the largest canyons in San Diego County.

"You wouldn't expect that the presence of coyotes would benefit prey populations, but there's greater predation pressure from all these smaller predators," says Crooks, a postdoctoral researcher at the University of California at San Diego. The idea that larger carnivores might keep numbers of small carnivores in check-and that losing the top dog would free smaller predators to ride roughshod on prey-had been proposed more than a decade ago to explain the rapid decline in certain scrub birds noted by Crooks's former professor, Michael Soulé, and colleagues. Even as the "mesopredator release" hypothesis became widely embraced as part of carnivore conservation plans nationwide, however, its central tenets remained untested. Crooks set out to do just that. "Kevin's work put meat on the skeleton of the theory," says California Department of Fish and Game biologist Ron Jurek, while providing hard evidence as to how this particular ecosystem works.

Between 1995 and 1997, Crooks studied the dynamics of predators and their feathered prey in 28 habitat fragments that sampled the range of size, age, and isolation in San Diego's canyons. The presence of coyotes turned out to be a major predictor of bird diversity. Nearly half of the study plots contained at least one coyote full-time, while in other plots coyotes would come and go. Mesopredator activity increased wherever coyotes were absent, but even coyotes' part-time presence provided sufficient impetus for cats and skunks and such to lie low or stay away. After accounting for the likelihood of birds faring poorly in older and smaller fragments, Crooks found that mesopredators reduced bird diversity. Ten years after local bird numbers were first detected to be dropping, the population extinctions had largely continued.

Worst offender among mesopredators is the domestic cat. By surveying neighborhood residents, Crooks learned that a third of homeowners bordering the canyons kept cats, and most pet cats spent time outdoors. A canyon that might harbor a pair of coyotes or a few foxes thus became hunting ground for dozens of cats. The overwhelming abundance of cats, plus their habit of hunting for fun as much as for food, increased their toll on native prey. In a paper published last year in Nature, Crooks and Soulé estimate from survey results that cats took home about 840 rodents, 595 lizards, and 525 birds yearly from a 50-acre canyon (and other studies show that cats kill many more animals than they drop on their owners' doorsteps).

Coyotes hunt their fair share of native prey, too, but Crooks found that local coyotes also eat a wide variety of garden fruits and vegetables, as well as a considerable number of cats. Cat remains were in 21 percent of coyote scat examined during the course of Crooks's study, and coyotes killed a quarter of radio-collared pet cats that he monitored.

Much like mesopredators, cat owners seem to sense when coyotes might pass through their local canyon; this perceived threat causes people to restrict their pet's outdoor forays, which indirectly benefits biodiversity. "If you want to protect species," concludes Crooks, "you ought to foster coyotes." And keep the cats indoors.

Blake Edgar is Senior Editor of California Wild.