Abstract. There are many examples where the use of chemicals have had profound unanticipated consequences, such as fertilizers reducing crop yields (paradox of enrichment) and insecticides increasing insect pests (by reducing natural biocontrol). Recently, the application of agrochemicals, such as agricultural disinfectants and fungicides, has been explored as an approach to curb the pathogenic fungus, Batrachochytrium dendrobatidis (Bd), which is associated with worldwide amphibian declines. However, the long-term, net effects of early-life exposure to these chemicals on amphibian disease risk have not been thoroughly investigated. Using a combination of laboratory experiments and analysis of data from the literature, we explored the effects of fungicide exposure on Bd infections in two frog species. Extremely low concentrations of the fungicides azoxystrobin, chlorothalonil, and mancozeb were directly toxic to Bd in culture. However, estimated environmental concentrations of the fungicides did not reduce Bd on Cuban tree frog (Osteopilus septentrionalis) tadpoles exposed simultaneously to any of these fungicides and Bd, and fungicide exposure actually increased Bd-induced mortality. Additionally, exposure to any of these fungicides as tadpoles resulted in higher Bd abundance and greater Bd-induced mortality when challenged with Bd post-metamorphosis, an average of 71 days after their last fungicide exposure. Analysis of data from the literature revealed that previous exposure to the fungicide itraconazole, which is commonly used to clear Bd infections, made the critically endangered booroolong frog (Litoria booroolongensis) more susceptible to Bd. Finally, a field survey revealed that Bd prevalence was positively associated with concentrations of fungicides in ponds. Although fungicides show promise for controlling Bd, these results suggest that, if fungicides do not completely eradicate Bd or if Bd re-colonizes, exposure to fungicides has the potential to do more harm than good. To ensure that fungicide applications have the intended consequence of curbing amphibian declines, researchers must identify which fungicides do not compromise the pathogen resistance mechanisms of amphibians.

Emerging infectious diseases (EIDs) are a salient threat to many animal taxa, causing local and global extinctions, altering communities and ecosystem function. The EID chytridiomycosis is a prominent driver of amphibian declines, which is caused by the fungal pathogen Batrachochytrium dendrobatidis (Bd). To guide conservation policy, we developed a predictive decision-analytic model that combines empirical knowledge of host-pathogen metapopulation dynamics with expert judgment regarding effects of management actions, to select from potential conservation strategies. We apply our approach to a boreal toad (Anaxyrus boreas boreas) and Bd system, identifying optimal strategies that balance tradeoffs in maximizing toad population persistence and landscape-level distribution, while considering costs. The most robust strategy is expected to reduce the decline of toad breeding sites from 53% to 21% over 50 years. Our findings are incorporated into management policy to guide conservation planning. Our online modeling application provides a template for managers of other systems challenged by EIDs.

Drought has fueled an increased frequency and severity of large wildfires in many ecosystems. Despite an increase in research on wildfire effects on vertebrates, the vast majority of it has focused on short-term (<5 yrs) effects and there is still little information on the time scale of population recovery for species that decline in abundance after fire. In 2003, a large wildfire in Montana (USA) burned the watersheds of four of eight streams that we sampled for larval Rocky Mountain tailed frogs (Ascaphus montanus) in 2001. Surveys during 2004?2005 revealed reduced abundance of larvae in burned streams relative to unburned streams, with greater declines associated with increased fire extent. Rocky Mountain tailed frogs have low vagility and have several unusual life-history traits that could slow population recovery, including an extended larval period (4 yrs), delayed sexual maturity (6?8 yrs), and low fecundity (<50 eggs/yr). To determine if abundance remained depressed since the 2003 wildfire, we repeated surveys during 2014?2015 and found relative abundance of larvae in burned and unburned streams had nearly converged to pre-fire conditions within two generations. The negative effects of burn extent on larval abundance weakened >58% within 12 yrs after the fire. We also found moderate synchrony among populations in unburned streams and negative spatial autocorrelation among populations in burned streams. We suspect negative spatial autocorrelation among spatially-clustered burned streams reflected increased post-fire patchiness in resources and different rates of local recovery. Our results add to a growing body of work that suggests populations in intact ecosystems tend to be resilient to habitat changes caused by wildfire. Our results also provide important insights into recovery times of populations that have been negatively affected by severe wildfire.