Discipline: Ecology Environmental and Earth Sciences
Subcategory: Ecology
Kristina Frogoso - University of Arkansas at Little Rock
Co-Author(s): Scott Connelly, Odum School of Ecology, University of Georgia, Athens, GA
Global losses in biodiversity are occurring at unprecedented rates. Declines in amphibian diversity play a large role in these losses. Many of the dramatic and on-going declines in amphibian populations are due to infectious diseases. One pathogen in particular that has been associated with widespread amphibian losses is the fungus Batrachochytrium dendrobatidis (Bd), which causes chytridiomycosis, an often fatal amphibian disease. A number of environmental factors may play a role in the prevalence and pathogenicity of Bd, and not all amphibian species are affected equally after exposure to the pathogen. Our study asks: Can we use environmental and amphibian natural history traits to help predict and prevent the decline of amphibians due to chytridiomycosis? A multiple linear regression model was created on RStudio to simulate the relationship between twelve environmental variables and the threat status of global amphibian species. Then, ten amphibian natural history traits were added in the explanatory variable to more fully characterize threat status. We made predictions of their threat status and compared the predicted to the actual threat status. Threat status was correctly predicted for 1799 species out of 2333, showing 77% accuracy. We applied the same methods to predict the threat status for species with unknown conservation status. The logistic regression analysis quantifying the relationship between natural history traits and predicted threat status showed that breeding site type (permanent or ephemeral) was the strongest predictor of threat status, with species using more permanent breeding sites tending to be more threatened (p < 0.05). This analysis showed that future research should focus on understanding of both environmental factors and an organism’s specific natural history characteristics to best predict the threat status for each species.
Funder Acknowledgement(s): National Science Foundation - Research Experiences for Undergraduates Program, Odum School of Ecology, University of Georgia, Dr. John Drake
Faculty Advisor: Jim Winter,