Modeling management strategies for chronic disease in wildlife: predictions for the control of respiratory disease in bighorn sheep

1. Controlling persistent infectious disease in wildlife populations is an on-going challenge for wildlife managers and conservationists worldwide.

2. Here, we develop a dynamic pathogen transmission model capturing key features of M. ovipneumoniae infection, a major cause of population declines in North American bighorn sheep (Ovis canadensis). We explore the effects of model assumptions and parameter values on disease dynamics, including density versus frequency dependent transmission, the inclusion of a carrier class versus a longer infectious period, host survival rates, disease-induced mortality and recovery rates, and the epidemic growth rate.

3. We compare the effectiveness of a suite of management actions following an epidemic, including test-and-remove, depopulation-and-reintroduction, range expansion, herd augmentation, and density reduction.

4. Our results suggest that test-and-remove, depopulation-and-reintroduction, and range expansion have the potential to facilitate recovery of persistently infected bighorn sheep herds post-epidemic. By contrast, augmentation could lead to worse outcomes than those expected in the absence of management. Management that improves host survival or reduces disease-induced mortality are also likely to improve population size and persistence of chronically infected herds.

5. Dynamic transmission models like the one employed here offer a structured, logical approach towards exploring hypotheses and can serve as a basis for planning field experiments and adaptive management. Models should be used iteratively with the field empirical approaches to triangulate on better approaches to wildlife management.

Methods
Data were gathered by Idaho Department of Fish and Game, Oregon Department of Fish and Wildlife, and Washington Department of Fish and Wildlife as part of the Hells Canyon Bighorn Sheep Initiative, under the coordination of E. Frances Cassirer. Field efforts consisted of at least biweekly (and often higher-resolution) aerial or ground survey of instrumented bighorn sheep ewes in the Hells Canyon system (specific herds indiciated in the Herd field). Dates of birth (DayInYear_birth) were taken to be the midpoint of the last survey of a ewe without a lamb and the first survey in which she had a lamb each year, unless lambs were clearly less than 72 hours old upon first observation. Lamb mortalities are recorded in the DayInYear_death field (the midpoint between the last observation the lamb was seen alive and the first observation in which it was not observed).

1. 管控野生动物种群中的持续性传染病，是全球野生动物管理者与保护工作者面临的长期挑战。
2. 本研究构建了可捕捉绵羊肺炎支原体（M. ovipneumoniae）感染关键特征的动态病原体传播模型，该病原体是导致北美大角羊（Ovis canadensis）种群数量下降的主要诱因之一。我们探究了模型假设与参数取值对疾病动态的影响，包括密度依赖传播与频率依赖传播、是否纳入携带类群与延长感染期、宿主存活率、疾病致死率与康复率，以及疫情增长速率。
3. 我们对比了疫情暴发后一系列管控措施的有效性，包括检测淘汰、种群清除后再引入、栖息地扩张、种群补充与种群密度降低。
4. 研究结果显示，检测淘汰、种群清除后再引入与栖息地扩张这三类措施，有望助力疫情后持续感染的大角羊种群恢复。与之相反，种群补充可能带来比未实施管控时更糟糕的结果。能够提升宿主存活率或降低疾病致死率的管控措施，同样有助于提升慢性感染种群的种群规模与存续能力。
5. 如本研究所用的这类动态传播模型，可为假说探究提供结构化且逻辑严谨的研究路径，同时可作为规划野外实验与适应性管理的基础。模型应与野外实证研究方法结合迭代使用，以多方验证并优化野生动物管理方案。

方法
本研究的数据由爱达荷州鱼类与狩猎局、俄勒冈州鱼类与野生动物局以及华盛顿州鱼类与野生动物局在E·弗朗西斯·卡西雷（E. Frances Cassirer）的协调下，作为地狱谷大角羊计划的一部分收集所得。野外调查工作至少每两周开展一次（通常频次更高），内容为对地狱谷区域内佩戴追踪设备的大角羊母羊开展航空或地面观测（具体种群信息详见Herd字段）。羔羊出生日期（DayInYear_birth）以母羊当年最后一次未携带羔羊的调查与首次携带羔羊的调查的中间时点为准，除非首次观测时羔羊明显不足72小时龄。羔羊死亡信息记录于DayInYear_death字段，该值为羔羊最后一次被观测到存活与首次未被观测到的中间时点。