Novel infectious diseases, particularly those caused by fungal pathogens, pose considerable risks to global biodiversity. The amphibian chytrid fungus (Batrachochytrium dendrobatidis, Bd) has demonstrated the scale of the threat, having caused the greatest recorded loss of vertebrate biodiversity attributable to a pathogen. Despite catastrophic declines on several continents, many affected species have experienced population recoveries after epidemics. However, the potential ongoing threat of endemic Bd in these recovered or recovering populations is still poorly understood. We investigated the threat of endemic Bd to frog populations that recovered after initial precipitous declines, focusing on the endangered rainforest frog Mixophyes fleayi. We conducted extensive field surveys over four years at three independent sites in eastern Australia. First, we compared Bd infection prevalence and infection intensities within frog communities to reveal species-specific infection patterns. Then, we analyzed capture-recapture data of M. fleayi to estimate the impact of Bd infection intensity on apparent mortality rates and Bd infection dynamics. We found that M. fleayi had lower infection intensities than sympatric frogs across the three sites, and cleared infections at higher rates than they gained infections throughout the study period. By incorporating time-varying individual infection intensities, we show that healthy M. fleayi populations persist despite increased apparent mortality associated with infrequent high Bd loads. Infection dynamics were influenced by environmental conditions, with Bd prevalence, infection intensity, and rates of gaining infection associated with lower temperatures and increased rainfall. However, mortality remained constant year-round despite these fluctuations in Bd infections, suggesting major mortality events did not occur over the study period. Together, our results demonstrate that while Bd is still a potential threat to recovered populations of M. fleayi, high rates of clearing infections and generally low average infection loads likely minimize mortality caused by Bd. Our results are consistent with pathogen resistance contributing to the coexistence of M. fleayi with endemic Bd. We emphasize the importance of incorporating infection intensity into disease models rather than infection status alone. Similar population and infection dynamics likely exist within other recovered amphibian-Bd systems around the globe, promising longer-term persistence in the face of endemic chytridiomycosis.

We conducted robust design capture-recapture surveys for Mixophyes fleayi at three sites from 2017–2021. Swab samples were collected from individuals during each capture and analysed with qPCR, which informed infection status with Bd. From these results, we constructed a capture history for analysis. Some columns, corresponding to primary occasions, were padded with missing values to align primary occasions across sites. Details of the data handling are in the Quarto document m2-cmr.qmd. Additionally, up to 15 individuals of sypatric frog species were swabbed during each primary occasion at each site from 2019–2021. These data are analysed in the Quarto document m1-infections.qmd. For more details, see the manuscript.

All code for data manipulation, analysis, and figure creation is included in the two Quarto files. All analysis was conducted in R using RStudio. Data analysis was conducted using NIMBLE.