Numerous species of amphibians declined in Central America during the 1980s and 1990s. These enigmatic declines mostly affected highland-stream amphibians and have been primarily linked to chytridiomycosis, a deadly disease caused by the chytrid fungus Batrachochytrium dendrobatidis (Bd). Since then, the majority of field studies on Bd in the Tropics have been conducted in midland and highland environments (>800m) mainly because the environmental conditions of mountain ranges match the range of ideal abiotic conditions for Bd in the laboratory. This unbalanced sampling has led researchers to largely overlook host-pathogen dynamics in lowlands, where other amphibian species declined during the same period. We conducted a survey testing for Bd in 47 species (n=348) in four lowland sites in Costa Rica to identify local host-pathogen dynamics and to describe the environmental space for Bd. We hypothesized differential host-pathogen dynamics occur across study sites because they exhibit latitudinal and altitudinal variation. Additionally, we hypothesized that that all study sites would exhibit low levels of Bd prevalence and intensity of infection suggesting that post-decline amphibian communities in lowlands persist with stable enzootic infections of Bd. In addition, we also expected a higher prevalence of Bd in amphibian assemblages occurring in permanent streams than in ephemeral ponds and terrestrial assemblages, as it has been found in previous studies.
Results/Conclusions
We detected Bd in three sampling sites and 70% of the surveyed species. We found evidence that lowland study sites exhibit enzootic dynamics with low infection intensity and moderate to high prevalence (55 % overall prevalence). Additionally, we found evidence that every study site represents an independent climatic zone, where local climatic differences may explain variations in Bd disease dynamics. We also found high prevalence of Bd across habitats among habitats, with the highest prevalence of Bd in species found in in leaf-litter/canopy, however habitat was a non-significant predictor for our data.We recommend more detection surveys across lowlands and other sites that have been historically considered unsuitable for Bd occurrence. These data can be used to identify sites for potential disease outbreaks and amphibian rediscoveries.