The control of wildlife disease outbreaks often involves culling, a method that is invasive and frequently clashes with other management objectives. Early intervention is crucial for diseases like anthrax and chronic wasting disease (CWD), which can establish environmental reservoirs. For diseases with latent stages or low prevalence, preemptive culling—removing herds with potential contact to infected populations—is a proactive measure. This approach is common in livestock management but raises significant political and conservation concerns when applied to wildlife. This study presents an alternative: proactive hunting surveillance, aiming for early disease detection without causing significant population decline. The focus is on CWD in wild reindeer in Norway, a disease spreading geographically and causing population declines in North America. The Norwegian government aims for eradication, making preemptive culling politically unfeasible and undesirable from a conservation standpoint. Current CWD surveillance relies on testing hunter-killed cervids; however, this approach is often not strategic, requiring massive sampling that could be unsustainable. This research explores targeted sampling of specific demographic groups (adults, particularly males) to enhance disease detection while minimizing negative population impacts. The innovative aspect is a detailed plan for harvest quotas to selectively hunt specific groups, optimizing disease detection while preserving the population. The study uses a simulation model to determine the optimal harvest size and demographic composition for rapidly confirming freedom from infection without significant population decline, merging selective harvesting principles with concepts of freedom from infection and risk-based surveillance.

The literature extensively covers culling as a disease mitigation tool in wildlife, highlighting its invasiveness and conflicts with conservation goals. Studies on anthrax and CWD emphasize the importance of early intervention to prevent the establishment of environmental reservoirs. Preemptive culling, widely used for livestock, has shown benefits but is not readily applicable to wildlife species of conservation concern. The epidemiology of CWD in cervids shows higher prevalence in adults, especially males, compared to younger animals. Research on targeted sampling for disease detection exists, but lacks strategic plans for selective harvesting to balance disease control with population preservation. The existing literature on selective harvesting in wildlife management focuses on maintaining sustainable yields, whereas this research incorporates the concept of freedom from infection, integrating veterinary epidemiology concepts into wildlife management.

The study used a three-component model: a population estimation model, a population simulation model, and a disease detection model. The population estimation model used data from four annual surveys (minimum counts, calving surveys, harvest statistics, and demographic structure counts) to estimate reindeer abundance, age and sex structures, and demographic rates (survival and reproduction). Bayesian inference within a stage-dependent two-sex matrix model (three age classes: calves, yearlings, adults) was employed. The population simulation model, parameterized by the population estimation model, simulated population dynamics under different harvest strategies, including stochastic variation in demographic rates. Harvest strategies varied by harvest rates, carrying capacity, minimum adult female population size, and adult sex ratio thresholds. Two main harvest strategies were compared: "ordinary" (historical harvest rates) and "proactive" (optimized for disease detection). The disease detection model, a stochastic scenario tree model, estimated the likelihood of detecting CWD infection in an individual based on age, sex, time since infection, and test sensitivity. The model accounted for the sex-specific pattern of CWD infection (higher risk in adult males) and the varying sensitivity of the ELISA test across different tissues and time since infection. The model output included surveillance sensitivity (probability of detecting at least one infected animal) and probability of freedom from infection (prevalence below a predefined design prevalence). Simulations were run for 1000 iterations for each scenario of harvest strategy and epidemiological parameters (relative risk, design prevalence, probability of infection introduction). The design prevalence was varied to simulate different levels of infection detection and risk assessment. The probability of infection introduction was adjusted for spatial proximity to a previously infected population. Key parameters included the operational sex ratio threshold (adult male to female ratio), minimum number of adult females, carrying capacity, and harvest rates for each sex and age class. The simulation model accounts for environmental variation using stochasticity in survival and recruitment. Adaptive management protocols are proposed for annual updates, accounting for variations in social response and quota filling. The model outputs were compared to assess the time needed to reach a given probability of freedom from infection for different harvest strategies.

The study revealed that proactive hunting surveillance significantly accelerates CWD detection compared to ordinary harvest surveillance. For Nordfjella zone 2 (small population), ordinary harvest would take 6 years for 90% confidence in freedom from infection and 11 years for 99% confidence. Proactive culling, targeting adult males to achieve a 1:5 male-to-female sex ratio, reduced this time to 3-5 years for 90% confidence and 5 years for 99% confidence. Similar improvements were observed in Hardangervidda (large population), where proactive hunting reduced the time to 90% confidence from 4 to 1-2 years, and to 99% from 10 to 3-5 years. The optimal strategy involved culling all available adult males in the first year, followed by female culling to maintain a specific sex ratio threshold. Maintaining ordinary harvest rates for calves and yearlings only slightly decreased the probability of freedom from infection. Proactive hunting increased sample size without reducing the female population. The relative risk of infection (higher in males) strongly influenced disease detection, particularly when adult males were heavily harvested. Higher design prevalence (the prevalence level to be detected) significantly increased the time to reach a given level of confidence in freedom from infection. The probability of CWD introduction, reflecting spatial proximity to a previously infected population, also affected the time to freedom from infection. Empirical data from Nordfjella zone 2 showed that supplementary culling by marksmen increased confidence in freedom from infection. In Hardangervidda, proactive hunting increased the certainty of freedom from infection considerably. The study demonstrated the effectiveness of this approach across various epidemiological parameters and population sizes. Simulations showed that including stochasticity in relative risk resulted in slightly wider confidence intervals for freedom from infection.

This study provides a valuable framework for managing wildlife populations when culling is used for disease control. The results demonstrate that proactive hunting surveillance, focusing on specific demographic groups, can substantially reduce the time required to reach a high probability of freedom from infection, compared to traditional surveillance methods. The approach integrates epidemiological principles (risk-based surveillance, freedom from infection) with wildlife management strategies (selective harvesting). The model's ability to incorporate stochasticity in demographic rates and relative risk enhances its realism and utility. The findings are relevant for managing other wildlife diseases with similar demographic patterns of infection, provided selective harvesting is feasible and a population model can be developed for prediction. The study also addressed practical considerations, including unintended consequences of skewed sex ratios and ethical implications of harvesting. The balance between disease control and population preservation is emphasized.

Proactive hunting surveillance offers a significantly more efficient approach to CWD detection in cervids compared to traditional methods. The methodology presented, incorporating a multi-component model, offers a valuable framework for managing wildlife populations while minimizing undesirable population declines. Future research should explore the approach's applicability to other wildlife diseases and species, considering species-specific reproductive biology and social dynamics. Further investigations into the optimal sex ratio thresholds and the management of potential unintended consequences of biased harvesting are also crucial.

The study relies on a model incorporating various assumptions, such as the demographic infection pattern of CWD and the test sensitivity. While efforts were made to account for stochasticity, uncertainties associated with model parameters and their estimation remain. The model's predictions depend on the accuracy of the population estimation model and the assumptions made regarding demographic rates and carrying capacity. Social and ethical factors related to implementation (e.g., hunter acceptance, orphaning of calves) are discussed but not fully incorporated into the quantitative model. The model might not fully capture the complexity of disease spread and the influence of environmental factors.