Chronic wasting disease (CWD) continues to spread among free-ranging populations of cervids across North America and where endemic can result in long-term population declines of white-tailed deer (Odocoileus virginianus) and mule deer (Odocoileus hemionus hemionus). With no treatment, vaccine, nor environmental decontamination methods forthcoming, identifying, and implementing effective management strategies to limit the growth and spread of CWD is critical. I focus on three challenges associated with managing CWD in my dissertation. First, I expand our knowledge of CWD dynamics in deer, particularly on indirect disease transmission. Second, I provide insights into heterogeneity observed in disease dynamics in response to local conditions, such as local deer density and landscape characteristics. Last, I describe sources of uncertainty surrounding implementation of CWD management. These uncertainties affect our ability to: 1) implement management effectively, 2) measure the success of management efforts and, 3) learn from past experiences. A contribution of this work is an individual-based model (IBM) that depicts CWD dynamics in free-ranging deer. The model is spatio-temporal and can: 1) incorporate local conditions and individual deer variation into estimates of disease dynamics, 2) simulate and assess localized and fine-scale management scenarios, and 3) integrate and account for aspects of uncertainty, such as with management implementation. In Chapter 1, I present the IBM using the Overview, Design Concepts, and Details protocol. I use the model to explore how interactions among deer and interactions with their environment affect CWD dynamics and validate aspects of the model using literature and independently collected field data. My second chapter focuses on construction and evaluation of localized and fine-scale (e.g., targeting different areas) deer management strategies for CWD using this model. In the third chapter, I incorporate sources of implementation uncertainty into the management scenarios from Chapter 2 to assess impacts on management outcomes. I worked closely with collaborators at the Michigan Department of Natural Resources, United States Department of Agriculture’s Animal and Plant Health Inspection Service Wildlife Services, and United States Geological Survey’s National Wildlife Health Center to develop this model, construct realistic management scenarios, and identify measures of management uncertainty. The IBM developed in Chapter 1 reproduced individual-level deer processes and macro-level population and CWD dynamics observed in Midwestern white-tailed deer populations for 20 years, which was the temporal extent of available validation data. A sensitivity analysis of the model revealed that fall fawn (i.e., deer <1.0 year old) migration rate (immigration and emigration), disease mortality rate, and fawn mortality rate had the largest impact on CWD prevalence 20 years post initial introduction of CWD. Prion half-life, prion shedding rate, and deer group membership had minimal influence. The results generated in Chapter 2 suggest that CWD outbreaks are more likely to persist in exurban areas compared to suburban landscapes. Additionally, CWD prevalence rate increased faster in exurban compared to suburban areas. Initial density of the affected deer population did not influence long-term CWD prevalence nor the probability of CWD persisting in the population. Localized deer removal by means of ring culling was the most effective management scenario for reducing the probability of CWD persistence and long-term prevalence rate under all landscape and deer density conditions. Moreover, removing deer closer to the origin of the disease (i.e., smaller radii rings) further reduced the probability of disease persistence and prevalence rate. Current agency removal goals of 20–30% of the local deer population were not enough to reduce probability of disease persistence below 10% according to the model except when the outbreak occurred in a suburban landscape and the method of management was a 1.6-km2 ring cull. The level of deer removal required (i.e., > 60%) to reduce probability of persistence to <10% may not be feasible for wildlife agencies nor socially acceptable by the public. In Chapter 3, my model results suggest that partial controllability, i.e., discrepancy between management decisions and their realization, affects the success of deer removal efforts to control CWD dynamics. The IBM revealed thresholds for rates of deer removal and land access required to influence persistence and long-term prevalence of CWD. Land access and deer removal rates as low as 20% were capable of reducing long-term prevalence of CWD. Conversely, at least one of these rates had to be ≥70% and the other ≥30% to reduce probability of CWD persisting in the population. I conclude this dissertation with a discussion of my results, management implications, and future directions for the IBM. I focused on localized removal of deer as a method of CWD management, but future analyses can investigate the efficacy of other control strategies, such as regulations that alter harvest rates or supplemental feeding of cervids. By inputting region- specific data, data on localized cervid populations, and developing management scenarios relevant to management objectives, wildlife managers can use this IBM to make more informed decisions about managing cervid populations for CWD.
Read
