Spatial and network analysis of US livestock movements based on Interstate Certificates of Veterinary Inspection
Spatial and network analysis of US livestock movements based on Interstate Certificates of Veterinary Inspection
复制标题
DOI：
10.1016/j.prevetmed.2021.105391
复制
发表时间：
2021-06-03
影响因子：
2.6
通讯作者：
Hanthorn, C. J.
中科院分区：
文献类型：
--
作者：
Cabezas, A. H.;Sanderson, M. W.;Hanthorn, C. J.
牲畜迁徙是传染病在人群中传播的常见途径。要了解高传染性疾病在流行期间的国家传播风险，就需要了解牲畜的流动模式。社会网络分析(SNA)是一种帮助描述个体之间的关系以及这些关系的含义的方法。我们使用SNA描述了2015年4月1日至2016年3月31日期间美国各地牲畜流动的联系结构。我们描述了4种网络类型：肉牛、奶牛、猪和小型反刍动物。牲畜流动数据来自州际兽医检验证书(ICVI)，而县级农场人口数据来自国家农业统计局(NASS)。在所描述的网络中，节点由县表示，而弧线由节点之间的出货量表示；网络基于节点之间的出货量进行加权。在分析过程中，每年在县一级汇总流动数据。计算了中心性和凝聚力的衡量标准，并确定了所有网络中的贸易社区。在研究期间，总共记录了219,042次牛群迁徙，其中肉牛迁徙占所有迁徙的63%。美国至少有70%的县出现在每个网络中，但所有网络中弧线的密度都不到2%。在肉牛网络中，对外开放程度高的县与每个县的肉牛数量有很强的相关性(0.8)，而奶牛网与县级每平方公里的奶牛数量有很强的相关性(>；0.86)。所有网络都有4到6个大型社区(每个社区50个县或更多)，除了小型反刍动物网络中的社区外，其他所有网络在地理上都是聚集的。这些分析报告的产出有助于了解肉牛、奶牛、猪和小型反刍动物的接触网络结构。它们还可与模拟模型结合使用，以评估诸如口蹄疫等高传染性疾病在国家一级的传播情况，并评估干预战略的应用。
Livestock movements are a common pathway for the spread infectious diseases in a population. An understanding of livestock movement patterns is needed to understand national transmission risks of highly infectious diseases during epidemics. Social Network Analysis (SNA) is an approach that helps to describe the relationships among individuals and the implications of those relationships. We used SNA to describe the contact structure of livestock movements throughout the contiguous U.S. from April 1st, 2015 to March 31st, 2016. We describe 4 network types: beef cattle, dairy cattle, swine, and small ruminant. Livestock movement data were sourced from Interstate Certificates of Veterinary Inspection (ICVI) while county-level farm demographic data were from the National Agricultural Statistics Service (NASS). In the described networks, nodes are represented by counties and arcs by shipments between nodes; the networks were weighted based on the number of shipments between nodes. For the analyses, movement data were aggregated at the county level and on an annual basis. Measures of centrality and cohesiveness were computed and identification of trade-communities in all networks was conducted. During the study period, a total of 219,042 movements were recorded and beef cattle movements accounted for 63 % of all movements. At least 70 % of U.S. counties were present in each of the networks, but the density of arcs was less than 2% in all networks. In the beef cattle network, counties with high out-degree were strongly correlated (0.8) with the number of beef cows per county while for the dairy cattle network a strong correlation (>0.86) was found with the number of dairy cattle per km2 at the county level. All networks were found to have between 4 and 6 large communities (50 counties or more per community), and were geographically clustered except for the communities in the small ruminant network. Outputs reported in these analyses can help to understand the structure of the contact networks for beef cattle, dairy cattle, swine, and small ruminants. They may also be used in conjunction with simulation modeling to evaluate spread of highly infectious disease such as foot-and-mouth disease at the national level and to evaluate the application of intervention strategies.