This study presents a new method for detection of between-herd livestock movements to facilitate disease tracing and more accurately describe network behaviour of relevance for spread of infectious diseases, including within-livestock business risk-carrying contacts that are not necessarily recorded anywhere. The study introduces and substantiates the concept of grouping livestock herds into business-units based on ownership and location in the tracing analysis of animal movement-based contact networks. To test the utility of this approach, whole core genome sequencing of 196 Salmonella Dublin isolates stored from previous surveillance and project activities was combined with information on cattle movements recorded in the Danish Cattle Database between 1997 and 2017. The aim was to investigate alternative explanations for S. Dublin circulation in groups of herds connected by ownership, but without complete records of livestock movements. The EpiContactTrace R-package was used to trace the contact networks between businesses and compare the network characteristics of businesses sharing strains of S. Dublin with different levels of genetic relatedness. The ownership-only definition proved to be an unreliable grouping approach for large businesses, which could have internal distances larger than 250 km and therefore do not represent useful epidemiological units. Therefore, the grouping was refined using spatial analysis. More than 90% of final business units formed were composed of one single cattle property, whereas multi-property businesses could reach up to eight properties in a given year, with up to 15 cattle herds having been part of the same business through the study period. Results showed markedly higher probabilities of introduction of infectious animals between proposed businesses from which the same clone of S. Dublin had been isolated, when compared to businesses with non-related strains, thus substantiating the business-unit as an important epidemiological feature to consider in contact network analysis and tracing of infection routes. However, this approach may overestimate real-life contacts between cattle properties and putatively overestimate the degree of risk-contacts within each business, since it is based solely on information about property ownership and location. This does not consider administrative and individual farmers behaviours that essentially keep two properties separated. Despite this, we conclude that defining epidemiological units based on businesses is a promising approach for future disease tracing tasks.

中文翻译：

---

结合沙门氏菌都柏林基因组信息和接触追踪，以证实一种新方法，可以改善对牛群传染性传播途径的检测。

这项研究提出了一种新的方法来检测畜群之间的牲畜活动，以促进疾病追踪，并更准确地描述与传染病传播相关的网络行为，包括不一定记录在任何地方的畜牧业内部承担风险的联系人。该研究在基于动物运动的联系网络的跟踪分析中引入并证实了基于所有权和位置将畜群分组为业务单位的概念。为了测试该方法的实用性，将1997年至2017年间从以前的监视和项目活动中存储的196株都柏林沙门氏菌分离株的全核心基因组测序与记录在册的牛群活动信息相结合。 S. 都柏林的成群畜群通过所有权联系在一起，但没有牲畜运动的完整记录。EpiContactTrace R-package用于跟踪企业之间的联系网络，并比较共享遗传水平不同的S. Dublin菌株的企业的网络特征。对于大型企业而言，仅所有权的定义被证明是不可靠的分组方法，内部距离可能大于250 km，因此不能代表有用的流行病学单位。因此，使用空间分析对分组进行了细化。成立的最终业务部门中有90％以上是由一个单一的牛场组成的，而在一年内，多产权业务最多可以达到八处，在研究期间，多达15家牛群属于同一行业。结果显示，与不相关菌株的企业相比，在分离出相同的S. Dublin克隆的拟建企业之间引入传染性动物的可能性要高得多，从而证实了该业务部门是需要考虑的重要流行病学特征联系网络分析和跟踪感染途径。但是，由于这种方法仅基于财产所有权和位置的信息，因此可能会高估牲畜之间的实际接触，并可能高估了每个企业内的风险接触程度。这没有考虑行政和个人农民的行为，这些行为本质上使两个财产分开。尽管如此，