Disease

Understanding the demographic consequences of interactions among pathogens, hosts, and weather conditions is critical in determining how amphibian populations respond to disease and in identifying site-specific conservation actions that can be developed to bolster persistence of amphibian populations. We investigated population dynamics in Boreal Toads relative to abiotic (fall temperatures and snowpack) and biotic (the abundance of another anuran host and disease) characteristics of the local environment in Wyoming, USA. We used capture-recapture data and a multi-state model where state is treated as a hidden Markov process to incorporate disease state uncertainty and assess our a priori hypotheses. Our results indicate that snowpack during the coldest week of the winter is more influential to toad survival, disease transition probabilities, and the population-level prevalence of the amphibian chytrid fungus (Batrachochytrium dendrobatidis) in the spring, than temperatures in the fall or the presence of another host. As hypothesized, apparent survival at low (i.e., <25 cm) snowpack (0.22 [CI: 0.15–0.31]) was lower than apparent survival at high snowpack (90.65 [CI: 0.50–0.78]). Our findings highlight the potential for local environmental factors, like snowpack, to influence disease and host persistence, and demonstrate the ecological complexity of disease effects on population demography in natural environments. This work further emphasizes the need for improved understanding of how climate change may influence the relationships among pathogens, hosts, and their environment for wild animal populations challenged by disease.

Tracking and understanding variation in pathogens such as Batrachochytrium dendrobatidis
(Bd), the agent of amphibian chytridiomycosis, which has caused population declines
globally, is a priority for many land managers. However, relatively little sampling of amphibian
communities has occurred at high latitudes. We used skin swabs collected during 2005?2017 from
boreal toads Anaxyrus boreas (n = 248), in southeast Alaska (USA; primarily in and near Klondike
Gold Rush National Historical Park [KLGO]) and northwest British Columbia (Canada) to determine
how Bd prevalence varied across life stages, habitat characteristics, local species richness,
and time. Across all years, Bd prevalence peaked in June and was >3 times greater for adult toads
(37.5%) vs. juveniles and metamorphs (11.2%). Bd prevalence for toads in the KLGO area, where
other amphibian species are rare or absent, was highest from river habitats (55.0%), followed by
human-modified upland wetlands (32.3%) and natural upland wetlands (12.7%)—the same rankorder
these habitats are used for toad breeding. None of the 12 Columbia spotted frogs Rana
luteiventris or 2 wood frogs R. sylvatica from the study area tested Bd-positive, although all were
from an area of low host density where Bd has not been detected. Prevalence of Bd on toads in the
KLGO area decreased during 2005?2015. This trend from a largely single-species system may be
encouraging or concerning, depending on how Bd is affecting vital rates, and emphasizes the
need to understand effects of pathogens before translating disease prevalence into management
actions.

Batrachochytrium dendrobatidis (Bd), a fungal pathogen that causes amphibian chytridiomycosis, has been implicated in population declines globally. To better understand how Bd affects survival and how threats vary spatially and temporally, we conducted long-term (range: 9–13 yrs) capture-recapture studies of boreal toads (Anaxyrus boreas) from three similar communities in western Montana. We also estimated temporal and spatial variation in population-level Bd prevalence among populations and the potential role of co-occurring Columbia spotted frogs (Rana luteiventris) in driving infection dynamics. Hierarchical models that accounted for detection uncertainty revealed Bd reduced apparent survival in one population that declined, was unassociated with survival in one stationary population, and was associated with increased survival in one population that is near extirpation. Despite different effects of Bd on hosts, pathogen prevalence was similar and synchronous across the populations separated by 111 – 176 km. Variation in Bd prevalence was driven partly by seasonal temperatures, but opposite the direction expected. Bd prevalence also decreased sharply over time across all populations, unrelated to trends in temperature, boreal toad survival, or infection dynamics of co-occurring Columbia spotted frogs. Toad Bd prevalence increased when frog abundance was high, consistent with an amplification effect. However, Bd prevalence of toads decreased as Bd prevalence of spotted frogs increased, consistent with a dilution effect. Our results reveal surprising variation in responses to Bd and show pathogen prevalence is not predictive of survival or population risk, and they illustrate the complexity in understanding disease dynamics across multiple populations.