In October 1996, Hurricane Lili was a killer in Central America. In Cuba, it destroyed homes and cane crops. But at its final landfall, among the small islands of the central Bahamas, the great storm was a stroke of extraordinary fortune for three American scientists -- a catastrophe that created in hours the natural conditions they had speculated about for 20 years.
The resulting information may answer long-standing questions about catastrophic impacts on ecosystems, and could help conservationists plan better ways to preserve natural habitat.
"This was the experience, not of one lifetime, but of several lifetimes," said one of the three scientists, UC Davis ecologist Thomas Schoener, who is still amazed by the story two years later.
"I think it's unique in science," Schoener continued. "You have here a very destructive event that strikes any given site very infrequently, the experimental site was very well studied for a number of years, and the investigators not only were there during and after the hurricane, but also had completed their annual population censuses just hours before the storm arrived. It won't happen again in scientific research."
A description of the adventure from UC Davis ecologist David Spiller, Washington University evolutionary biologist Jonathan Losos, and Schoener appears in this Friday's issue of the journal Science.
On Oct. 18, 1996, Spiller and Losos had just returned to Great Exuma Island from surveying the spider and lizard populations on 19 smaller islands nearby. The work was part of a long-term study of the islands' food web.
On Great Exuma, the researchers found faxed weather bulletins from Schoener, who was in Davis, that warned of Hurricane Lili's approach. In darkness, as the winds rose to gale force, they helped their Bahamian friends board up buildings, then collected their irreplaceable notebooks and headed for a sturdy house on higher ground.
Soon, the researchers write in the Science article, "The highly improbable happened. ... Hurricane Lili, the first major hurricane to strike anywhere in the Exumas since 1932, passed directly over our study site with sustained winds of 90 knots [about 110 miles per hour] and a storm surge of nearly 5 meters [about 15 feet]."
They emerged after the storm, Spiller said, to find roofs missing, sailboats capsized and their 15-foot research boat resting atop a clump of trees.
The boat was recovered and, the Science article continues, "The next day, as soon as the storm subsided, and for 3 days thereafter, we recensused populations on all the islands."
Then, and in repeated observations during the next year, the researchers found evidence for a hypothesis advanced by Schoener in 1983: When Anolis lizards are missing from small Caribbean islands with lizard-friendly habitat, hurricanes are to blame.
Their unprecedented before-and-after data, the authors write, also helped them evaluate other, more general hypotheses about the ecological impact of a natural catastrophe. They concluded:
- Larger organisms were more resistant than small ones to the immediate impact
of a moderate disturbance, but the more prolific organisms recovered faster. On
islands that were partly protected, and hence only moderately disturbed, 34
percent of lizards were washed away by the storm, compared with 79 percent of
spiders. Yet after a year the numbers of lizards had not changed, while the
number of spiders had returned to pre-hurricane levels.
- Local extinction risk was related to population size when the disturbance was
moderate, but not when it was catastrophic. On partly protected islands no
lizard population was wiped out, even though some islands previously had fewer
than five lizards. Nine of 22 spider populations were eliminated. On exposed,
hence catastrophically disturbed, islands all lizards and spiders became
extinct, even though some populations previously had 140-190 individuals.
- After a catastrophic disturbance, the organisms that could disperse the most readily recovered the fastest. In one year, spider numbers on the exposed islands reached about one-third of their pre-hurricane levels, but there were no lizards. Spiller said in an interview that the spiders were probably blown onto the depopulated islands and thrived in the absence of the predatory lizards.
Understanding such effects is more than academically interesting, Spiller said. It could help shape programs for preserving natural communities.
"For instance, if you have an endangered species in a floodplain, these results should be of great concern. Areas that are physically protected may be worth a lot more in the conservation scheme than exposed areas," Spiller said. "That factor usually isn't taken into consideration."
Spiller and Losos will return to Great Exuma this October, in the last weeks of the 1998 hurricane season, to continue following the recovery of the biota on the devastated islands.
Note: David Spiller will be in Fort Lauderdale, Fla., Wednesday afternoon, July 29, through Friday afternoon, July 31, at 954-527-2700. He can be reached at UC Davis, starting Monday, Aug. 3, 530-752-1925, daspiller@ucdavis.edu.
Contacts:
- Thomas Schoener, evolution and ecology, (530) 752-7465, twschoener@ucdavis.edu
- Jonathan Losos, Washington University, St. Louis, Mo., 314-935-6706
- Sylvia Wright, News Service, 530-752-7704, swright@ucdavis.edu
Journal
Science