Department of Plant Sciences, University of Cambridge, Cambridge, United Kingdom.
PLoS Comput Biol. 2012 Jan;8(1):e1002328. doi: 10.1371/journal.pcbi.1002328. Epub 2012 Jan 5.
Exotic pathogens and pests threaten ecosystem service, biodiversity, and crop security globally. If an invasive agent can disperse asymptomatically over long distances, multiple spatial and temporal scales interplay, making identification of effective strategies to regulate, monitor, and control disease extremely difficult. The management of outbreaks is also challenged by limited data on the actual area infested and the dynamics of spatial spread, due to financial, technological, or social constraints. We examine principles of landscape epidemiology important in designing policy to prevent or slow invasion by such organisms, and use Phytophthora ramorum, the cause of sudden oak death, to illustrate how shortfalls in their understanding can render management applications inappropriate. This pathogen has invaded forests in coastal California, USA, and an isolated but fast-growing epidemic focus in northern California (Humboldt County) has the potential for extensive spread. The risk of spread is enhanced by the pathogen's generalist nature and survival. Additionally, the extent of cryptic infection is unknown due to limited surveying resources and access to private land. Here, we use an epidemiological model for transmission in heterogeneous landscapes and Bayesian Markov-chain-Monte-Carlo inference to estimate dispersal and life-cycle parameters of P. ramorum and forecast the distribution of infection and speed of the epidemic front in Humboldt County. We assess the viability of management options for containing the pathogen's northern spread and local impacts. Implementing a stand-alone host-free "barrier" had limited efficacy due to long-distance dispersal, but combining curative with preventive treatments ahead of the front reduced local damage and contained spread. While the large size of this focus makes effective control expensive, early synchronous treatment in newly-identified disease foci should be more cost-effective. We show how the successful management of forest ecosystems depends on estimating the spatial scales of invasion and treatment of pathogens and pests with cryptic long-distance dispersal.
外来病原体和害虫威胁着全球的生态系统服务、生物多样性和作物安全。如果一种入侵生物能够无症状地长距离传播，多个时空尺度相互作用，那么识别出有效控制、监测和控制疾病的策略就变得极其困难。由于财务、技术或社会限制，暴发的管理也受到实际受感染区域和空间传播动态数据有限的挑战。我们研究了在设计预防或减缓此类生物入侵的政策时重要的景观流行病学原理，并以导致突然橡树死亡的疫霉菌为例，说明对其理解的不足会导致管理应用不当。这种病原体已经入侵了美国加利福尼亚州沿海的森林，而加利福尼亚州北部（洪堡县）一个孤立但快速增长的疫情焦点有可能广泛传播。由于有限的调查资源和进入私人土地的机会，这种病原体的一般性和生存能力增加了传播的风险。此外，由于调查资源有限且无法进入私人土地，因此未知隐匿性感染的程度。在这里，我们使用在异质景观中传播的流行病学模型和贝叶斯马尔可夫链 - 蒙特卡罗推断来估计疫霉菌的扩散和生命周期参数，并预测洪堡县感染的分布和疫情前沿的传播速度。我们评估了控制病原体在北部传播和局部影响的管理选项的可行性。实施单独的无宿主“屏障”由于远距离传播效果有限，但在疫情前沿实施治疗性和预防性治疗相结合可减少局部损害并控制传播。虽然这个焦点的规模很大，使得有效控制成本高昂，但在新发现的疾病焦点中及早同步治疗应该更具成本效益。我们展示了如何通过估计入侵和具有隐匿远距离传播的病原体和害虫的空间尺度来成功管理森林生态系统。
