Optimizing Chronic Wasting Disease Surveillance: An agent-based modeling approach for captive deer facilities in Texas

Texas has the highest number of captive cervid facilities in the United States, and since 2012, chronic wasting disease (CWD) has been detected in several deer breeding facilities, at breeder release sites, as well as in free-ranging cervids. CWD often remains undetected in the early phase of the outbreak, mainly because of the logistical constraints on wildlife agencies' ability to match the sample size targets. This is further compounded by competing stakeholder objectives and perceptions. Consequently, there is considerable uncertainty about the status of CWD infection in many wild and captive deer populations in the United States. Uncertainty about the current status of CWD infection in captive deer facilities is a major challenge in mitigating the threats posed by the spread and persistence of CWD in Texas.

We have developed a pragmatic agent-based modeling tool to quantify the likelihood of undetected CWD in individual deer breeding facilities. The model simulates a realistic in silico age-structured population for individual breeding facilities. We used CWD testing data (ante-mortem and post-mortem) from 2012 to run model scenarios iteratively and estimated mean annual CWD detection probabilities for 10 breeding facilities in Texas. The mean annual CWD detection probabilities for the 10 deer breeding facilities were low (<15%). In other words, these facilities had a high likelihood of undetected CWD. One deer breeding facility had a CWD detection probability of 68% in 2022 when a large proportion of deer were tested antemortem. If only the adult deer were considered, 5 deer breeding facilities in 2016, and 4 each in 2021 and 2022 had a CWD detection probability over 50%. Overall, this model-based approach can be used to estimate the likelihood of undetected CWD in deer breeding facilities and prioritize CWD surveillance efforts