Alexander, M. A., and J. K. Eischeid, 2001: Climate variability in regions of amphibian declines. Conservation Biology, 15, 930-942.
We explored the relationship between amphibian declines and climate variations in Colorado (U.S.A.), Puerto Rico, Costa Rica-Panama, and Queensland (Australia) through two sources of data: output from the National Center for Environmental Prediction "reanalysis system" and area-averaged station data. The reanalysis system merges observations from airplanes, land stations, satellites, ships, and weather balloons with output from a weather-forecast model to create global fields of atmospheric variables. Station data consisted of temperature and precipitation measured with thermometers and rain gauges at fixed locations. Temperatures were near normal in Colorado when the amphibian declines occurred in the 1970s, whereas in Central America temperatures were warmer than normal, especially during the dry season. The station data from Puerto Rico and Australia indicated that temperatures were above normal during the period of amphibian declines, but reanalysis did not show such a clear temperature signal. Although declines occurred while the temperature and precipitation anomalies in some of the regions were large and of extended duration, the anomalies were not beyond the range of normal variability. Thus, unusual climate, as measured by regional estimates of temperature and precipitation is unlikely to be the direct cause of amphibian declines, but it may have indirectly contributed to them. Previous researchers have noted that the declines appear to have propagated from northwest to southeast from Costa Rica to Panama and from southeast to northwest in Queensland, Australia. Wind has the potential to transport pathogens that cause amphibian mortality. The mean direction of the near-surface winds tended to parallel the path of amphibian declines from July-October in Central America and from May-July in Australia. The wind direction was highly variable, however, and the propagation rate of amphibian declines was much slower than the mean wind speed. In addition, the most likely pathogen is a chytrid fungus that does not produce desiccation-resistant spores. Thus, if wind is involved in the propagation of amphibian declines, it is through a complex set of processes.