Disease epidemics can act as extinction filters, changing community structure, species niche breadth, and species co-occurrence patterns. Rarely do data exist that are collected systematically before and after an epidemic documenting species losses and changes in community composition. We examined spatial and temporal patterns of a stream tadpole community before, during, and after a chytridiomycosis epidemic in Central Panama that caused widespread extirpations and declines in adult amphibian populations. We sampled four microhabitats in four streams monthly one year before the decline, one year after the decline, and an additional three times post-decline. This is the first fine-scale ecological assessment documenting niche breadth changes following a chytridiomycosis epidemic.
Results/Conclusions
Pre-epizootic, we sampled 898 individuals, representing 5 families and 14 species, with a habitat-weighted density ranging from 0.04 to 17.71 individuals m-2. Identical sampling effort one year post-epizootic produced 266 individuals, belonging to 4 families and 8 species, with a habitat-weighted density ranging 0.00 to 1.08 individuals m-2. Species richness (F1,27 = 21.24, p < 0.001) and habitat-weighted density (F1,27 = 39.01, p < 0.001) decreased post-decline. Microhabitat community composition changed significantly from pre- to post-decline (F1,36 = 7.45, p < 0.001). Extirpations were nonrandom across species and microhabitats, with spatial-specialist species and species-poor microhabitats experiencing the most extirpations. Species that survived the epidemic had smaller niche breadths (R2 = 0.696, p < 0.001), defined as spatial and temporal habitat use, and higher pairwise niche breadth overlap with other surviving species after the decline. Chytridiomycosis is a nonrandom extinction filter of adult amphibians, and we show that it is also nonrandom for larvae. Extirpations removed specialist species and narrowed the spatial and temporal ranges of generalist species, resulting in homogenization of communities and a loss of species richness. Our results inform community disassembly rules and identify vulnerable species and life stages that are most at risk to disease-related declines.