Modeling Infection Risk of Chytridiomycosis among Illinois Amphibians

Background/Question/Methods

Species level traits can predict susceptibility to Batrachochytrium dendrobatidis (Bd) or decline probability for amphibians. We conducted a 2 yr systematic survey of Bd distribution among Illinois amphibians. Our objective was to assess infection risk among species using disease data, population level features, and species life history characteristics. We identified 13 traits expected to increase infection based on two mechanisms: increased exposure to Bd (i.e., aquatic index, breeding period length, and overlap of aquatic-breeding timeframe with optimal Bd-growth air temperatures) and increased likelihood of decline based on ecological principals (i.e., species range area, longevity, adult body size, clutch size, and NatureServe conservation status, trends, and overall threat levels). We assigned aquatic index ranks (1 – 3) to species based on adult natural history, modifying the list created by Lips et al. (2003) to fit temperate species characteristics. We used linear regression to determine which species traits were associated with high zoospore infections among 1,436 amphibians sampled, representing 2 salamander and 16 anuran species. We used linear regression mixed models that nested observations by family to account for phylogenetic relatedness. Finally, we constructed multivariate mixed models to assess the complex interacting species traits. All models were assessed with Akaike’s Information Criteria (AIC).

Results/Conclusions

We identified longevity as the best predictor of average species infection intensity (ΔAIC = 597; F_1,1240 = 61.64, p < 0.001), with long-lived species typically having lower infection intensity. Multivariate analyses indicated that aquatic index and state conservation status also contributed to intensity levels. The high intensity levels of Acris crepitans (μ = 3037.17 zoospores) support our results since it has a very short lifespan (4 mos) and medium aquatic index (level 2). Further supporting our results, the long-lived Lithobates catesbeianus (10 yrs), with high aquatic index (level 3), had low intensity infections (μ = 663.40 zoospores). The directionality of our aquatic index result is surprising given the assumption that Bd is primarily an aquatic pathogen, and the numerous reports of higher extinction among more aquatic species. Species with the lowest state-level threat status had the lowest infection levels. This may indicate that species have suites of traits that put them at greater or lesser risk of endangerment from multiple mechanisms, and that sensitivity to one threat may mean sensitivity to others. Risk assessment can be a useful tool for managers in prioritizing monitoring efforts or other conservation activities.