Bondo, K.J., C.S. Rosenberry, D. Stainbrook, W.D. Walter. Comparing risk of chronic wasting disease occurrence using Bayesian hierarchical spatial models and different surveillance types. Ecological Modeling 493:110756.
Abstract
One of the most important measures in controlling wildlife diseases is understanding risk of disease transmission. Risk of wildlife disease transmission in relation to environmental variables is often modeled and predicted using Markov Chain Monte Carlo (MCMC) methods, which are unsuitable for large datasets and those covering large spatial extents. Integrated Nested Laplace Approximation (INLA) and stochastic partial differential equations (SPDE) have become popular alternatives to MCMC for Bayesian inference because of its fast computational time and ability to process large datasets. Studies investigating risk of disease in wildlife, to our knowledge, have not yet compared model performance of various Bayesian hierarchical spatial models over large spatial extents using real world data. Using chronic wasting disease (CWD) surveillance from white-tailed deer collected in Pennsylvania as a case study, we demonstrate how parameter estimates compare using MCMC, INLA, and INLA-SPDE spatial models, and use INLA-SPDE to model CWD over a much larger spatial extent than has been conducted previously for this disease to determine how surveillance type (e.g. hunter harvest, roadkill, or targeted) influences model parameters and predicted risk at locations not sampled. Covariates considered in the models included deer age and sex, elevation, slope, distance to streams, percent clay, and proportion of habitat classes in three categories known to influence deer movements (forest, open, and developed). We found INLA to produce comparable estimates to MCMC and permit modeling large datasets covering expansive spatial extents much faster and more efficiently than MCMC. We identified potential biases in surveillance types, indicating the value of including all surveillance types rather than only a single type in the models. Understanding biases between surveillance samples and tools available for modeling and mapping diseases of wildlife in relation to ecological variables at large spatial extents will help guide future surveillance efforts for CWD and other wildlife diseases.