SYMP 16-1 - Developing advanced quantitative tools to improve ecological inference and forecasting for wildlife diseases

Surveillance programs are essential for monitoring species populations and detecting emerging pathogens; and to make inference about the presence of a target species, surveillance programs either rely on (1) observer expertise in identifying a species in the field or (2) molecular methods. However, observers and molecular methods rarely detect the target species perfectly. For example, molecular pathogen detection methods can result in misclassification (i.e., false positives and false negatives) or partial detection errors (i.e., detections with ‘ambiguous’, ‘uncertain’, or ‘equivocal’ results). Then, when data are to be analyzed, these partial observations are either discarded or censored. This method of discarding or censoring data, however, disregards information that could be used to make inference about the true state of the system. Therefore, during this presentation, I will highlight my work related to developing advanced hierarchical Bayesian models that improve parameter estimation, ecological inference, and system forecasting when observations are imperfect. First, I’ll be talking about how an apex predator (i.e., snakes) declined following the mass mortality of a prey item caused by a deadly fungal pathogen. We found that the loss of amphibians led to a >20% decline in the estimated richness of local snake species. Surprisingly, there was only anecdotal evidence that the snake species most severely impacted by amphibian losses had diets that heavily relied on amphibian prey. Some snake species that were thought to be generalists also decreased in occurrence, suggesting that there may be other, indirect effects of Bd epizootics that are not easily measured. The second project that I’ll dive into is my latest work related to accounting for ambiguous detections in qPCR, and examining the question, “how many samples do I need to collect to declare a site ‘pathogen-free?’. Here, we found that the presence of uncertain detections increased the variability of the resulting posterior probability distributions of pathogen occurrence, and that our estimates of required sample size were very sensitive to prior information about pathogen occupancy, pathogen prevalence, and diagnostic test specificity. Collectively, my work defines complex problems, I develop new approaches to answer those questions, and I am adapting theory, principles, and techniques into original ways.