Amphibian Declines and Climate Disturbance: The Case of the Golden Toad and the Harlequin Frog

Abstract

The endemic golden toad ( Bufo periglenes ) was abundant in Costa Rica’s Monteverde Cloud Forest Preserve in April–May 1987 but afterwards disappeared, along with local populations of the harlequin frog ( Atelopus varius ). We examine the possible relationship between these sudden declines and unusually warm, dry conditions in 1987. For our analyses of local weather patterns, we define a 12‐month (July–June) amphibian moisture‐temperature cycle consisting of four periods: (1) late wet season; (2) transition into dry season; (3) dry season; and (4) post‐dry‐season (early‐wet‐season) recovery. The 1986–1987 cycle was the only one on record (of 20 analyzed) with abnormally low rainfall in all four periods, and temperature anomalies in 1987 reached record highs. Flow in local aquifer‐fed streams during the dry season and post‐dry‐season recovery period reached a record low. This climate disturbance, associated with the 1986–1987 El Niño /Southern Oscillation, was more severe than a similar event associated with the 1982–1983 El Niño , though this earlier oscillation was the strongest of the past century. Demographic data for one harlequin frog population, gathered during these two climatic events, support the hypothesis that in 1987, shortly before the population collapsed, the frogs underwent an unprecedented shift in distribution within the habitat in response to desiccating conditions. The juxtaposition of these rare demographic events suggests they were causally linked yet sheds little light on mechanisms underlying the sudden decline. While desiccation or direct temperature effects may have been factors leading to high adult mortality, moisture‐temperature conditions may have interacted with some other, unidentifted agent. We discuss two hypotheses concerning possible synergistic effects: In the climate‐linked epidemic hypothesis , microparasites are the additional agent. In the climate‐linked contaminant pulse hypothesis , atmospheric contaminants scavenged by mist and cloud water in montane areas reach critical concentrations when conditions are abnormally warm and dry.