Abstract
Lizard and spider populations
were censused immediately before and after Hurricane Lili on islands differentially
affected by the storm surge. The results support three general propositions.
First, the larger organisms, lizards, are more resistant to the immediate
impact of moderate disturbance, whereas the more prolific spiders recover
faster. Second, extinction risk is related to population size when
disturbance is moderate but not when it is catastrophic. Third, after
catastrophic disturbance, the recovery rate among different types of organisms
is related to dispersal ability. The absence of poorer dispersers,
lizards, from many suitable islands is probably the result of long-lasting
effects of catastrophes.
Non-Technical Summary
We had been conducting long term experiments on ecological interactions between two species of Anolis lizards, and the effects of these interactions on lower food-web levels, for three years in the vicinity of Georgetown, Great Exuma, Bahamas. Great Exuma is a long and narrow island; our experimental islands, which were quite small (officially termed "rocks") were distributed on the southwest and northeast sides of Great Exuma.
In October of 1996, we returned to Great Exuma and completed our annual census of the islands, estimating lizard and spider densities and collecting data on plant herbivory on each of 19 islands. Just as we finished our work, Hurricane Lili moved across Cuba and headed toward Great Exuma. As we sat and waited, Lili came closer and closer, failing to change directions, as so many other hurricanes had done in the very busy 1995 and 1996 hurricane seasons. Rather, Lili scored a direct hit, with the eye of storm passing over Georgetown in the early hours of October 19th. Thus, by sheer good fortune, we were able to collect a nearly unparalleled set of data, with censuses both immediately before and after the hurricane, replicated on 19 islands.
Lili was a Class 3 hurricane, with sustained winds of 90 knots and a storm surge of five meters. As soon as possible (a day after the storm), we retrieved our boat from the small tree in which it had landed and returned to the islands. We first visited the islands on the northeast side of Great Exuma. Because the storm had come from the southwest, Lili had passed over Great Exuma before striking these islands. What we found on the islands was clear evidence of a major disturbance: vegetation was denuded, branches were broken, and lizard and spider densities were down. Nonetheless, the effects were clearly quantitative in nature. Moreover, the effects were greater on spiders than on lizards. Among spider populations, those that survived had higher pre-hurricane population sizes than those species that perished.
We then returned to the islands on the southwest. There, the story was entirely different. Islands were barren, looking as if a bomb had exploded. Almost all vegetation was completely destroyed; only lowlying plants remained. None of the spiders or lizards previously on the island remained (although, paradoxically, one species of spider never previously recorded on any islands was found in low numbers on six islands; most likely, these spiders were blown from Great Exuma by the back end of the hurricane).
The much greater devastation on the southwest islands can be explained by the storm surge that accompanies a hurricane. Our experimental islands were extremely low-lying. Hence, the storm surge, which raised sealevels five meters, completely inundated the islands and kept them underwater for several hours.
We returned to the islands several times over the course of the next year. One year after the storm, densities and diversity had increased to a much greater extent for spiders than for lizards. None of the islands on which lizards had been extirpated were recolonized. Consequently, we can understand the absence of lizards from many islands on which they used to occur 10,000 years ago (when the entire Great Bahama Bank was united into a single large landmass) and on which experimental introductions indicate they can thrive. Because overwater dispersal occurs at a low rate, hurricanes coming through the islands every 50-100 islands are probably sufficient to ensure that most small islands do not harbor lizard populations.