Summary1. The need to increase our understanding of factors that regulate animal population dynamics has been catalysed by recent, observed declines in wildlife populations worldwide. Reliable estimates of demographic parameters are critical for addressing basic and applied ecological questions and understanding the response of parameters to perturbations (e.g. disease, habitat loss, climate change). However, to fully assess the impact of perturbation on population dynamics, all parameters contributing to the response of the target population must be estimated. 2. We applied the reverse-time model of Pradel in Program mark to 6 years of capture-recapture data from two populations of Anaxyrus boreas (boreal toad) populations, one with disease and one without. We then assessed a priori hypotheses about differences in survival and recruitment relative to local environmental conditions and the presence of disease. 3. We further explored the relative contribution of survival probability and recruitment rate to population growth and investigated how shifts in these parameters can alter population dynamics when a population is perturbed. 4. High recruitment rates (0AE41) are probably compensating for low survival probability (range 0AE51-0AE54) in the population challenged by an emerging pathogen, resulting in a relatively slow rate of decline. In contrast, the population with no evidence of disease had high survival probability (range 0AE75-0AE78) but lower recruitment rates (0AE25). 5. Synthesis and applications. We suggest that the relationship between survival and recruitment may be compensatory, providing evidence that populations challenged with disease are not necessarily doomed to extinction. A better understanding of these interactions may help to explain, and be used to predict, population regulation and persistence for wildlife threatened with disease. Further, reliable estimates of population parameters such as recruitment and survival can guide the formulation and implementation of conservation actions such as repatriations or habitat management aimed to improve recruitment.

Chytridiomycosis is linked to the worldwide decline of amphibians, yet little is known about the demographic effects of the disease. We collected capture-recapture data on three populations of boreal toads (Bufo boreas [Bufo = Anaxyrus]) in the Rocky Mountains (U.S.A.). Two of the populations were infected with chytridiomycosis and one was not. We examined the effect of the presence of amphibian chytrid fungus (Batrachochytrium dendrobatidis [Bd]; the agent of chytridiomycosis) on survival probability and population growth rate. Toads that were infected with Bd had lower average annual survival probability than uninfected individuals at sites where Bd was detected, which suggests chytridiomycosis may reduce survival by 31-42% in wild boreal toads. Toads that were negative for Bd at infected sites had survival probabilities comparable to toads at the uninfected site. Evidence that environmental covariates (particularly cold temperatures during the breeding season) influenced toad survival was weak. The number of individuals in diseased populations declined by 5-7%/year over the 6 years of the study, whereas the uninfected population had comparatively stable population growth. Our data suggest that the presence of Bd in these toad populations is not causing rapid population declines. Rather, chytridiomycosis appears to be functioning as a low-level, chronic disease whereby some infected individuals survive but the overall population effects are still negative. Our results show that some amphibian populations may be coexisting with Bd and highlight the importance of quantitative assessments of survival in diseased animal populations.

Abstract. Two populations of boreal toads (Bt{/o boreas) experienced drastic declines in abundance in the late 1990s. Evidence supported the hypothesis of disease (the chytrid fungus. Batrachochytrium dendrobatidis) as the cause of these declines. but other hypotheses had not been evaluated. We used an II-year capture-recapture data set to evaluate weather and disease as causes of these declines. We developed sets of mathematical models that reflected hypothesized relationships between several weather variables and annual survival rates of adult males in these popUlations. In addition, models that reflected the possibility that the declines were caused by an introduced fungus were developed. All models were fit to the data and were evaluated using a model selection criterion (QAICJ. Our analysis provided strong support for the hypothesis of an introduced fungus and little support for the hypothesis that weather conditions caused the declines. Our results also suggest a strong, negative "marking effect" on survival rates of boreal toads. Modela veraged estimates of survi val rate are presented.