W T tORLDWIDE DECLINES IN AMPHIBIAN POPULATIONS ATtracted wide attention following a workshop sponsored by the National Research Council (NRC) (1, 2). The declines seem to be general, but some regions and many taxa are apparently unaffected (1). Are the declines due to normal climatic fluctuations, which produce droughts and frosts that can severely reduce population size locally, or are there other, more insidious causes? The report of census data for four amphibians (three salamanders and a frog) gathered during a 12-year period of continuous (daily) monitoring at the Savannah River Site in South Carolina contributes importantly to this question (3). Substantial variation was found in breeding population sizes and juvenile recruitment. Population declines were found in some years, but these were coupled with increases; there is a complex dynamic associated with many factors, especially local moisture regimen. The new data, the most extensive available, raise the possibility that declines are the coincidental effect of population fluctuations. If so, the consequences will vary greatly from place to place. The southeastern United States is a veritable carpet of amphibians, from the abundant salamanders of the highest Appalachian peaks to dense lowland populations of both frogs and salamanders. Should populations decline to local extinction in this area, chances of recovery are high. However, local extinctions have more profound implications in other parts of the world, where species are specialists for habitats that are localized or badly fragmented; under such circumstances, opportunities for recolonization are low to nonexistent (4). It is significant that participants at the NRC workshop, who heard a brief summary of the new data, considered it unlikely that global declines were coincidental, because of the number and widespread distribution of the reports, but population fluctuations clearly must be taken into account. Concern about declines arises because amphibians are abundant, integral components ofmany diverse ecosystems. They are local top carnivores that are major consumers of invertebrates, especially insects. General declines would have widespread consequences and might indicate more general environmental problems (5). Some declines have been dramatic. Ranid frogs have all but disappeared from southern California (6). By the late 1980s a montane frog (Rana muscosa) had disappeared from 98% of the ponds in which it had been studied in the mid-1970s in SequoiaKings Canyon National Park (1, 4). In Oregon, populations ofRana cascadae, monitored since the mid-1970s, have suffered about an 80% disappearance (1, 7). The golden toad (Bufo periglenes), endemic to the Monteverde Cloud Forest Preserve in Costa Rica, has not bred in its traditional breeding sites since 1987; many other frog taxa in the region have experienced declines and local disappearance during this period (1, 8). The gastric-brooding frog (Rheobatrachus silus), which lived in rivers in relatively undisturbed regions of Queensland, Australia, has not been seen since 1979, and several other sympatric species offrogs also are thought to be extinct (9). Sometime after 1981, 8 of 13 species of frogs that had been present in Reserva Atlantica, Brazil, disappeared, including unusual diurnal hylodine frogs, whose birdlike calls were distinctive features