BCMS Cattle Movements in the UK
Recent years have seen diseases of farm livestock in the United Kingdom having a huge economic impact, particularly foot and mouth disease in 2001, classical swine fever in 2000, and bovine spongiform encephalopathy since 1986.
The movement of farm livestock around the country is important to the economics of UK farming, but each movement clearly carries the risk of transmitting infection. This was most dramatically demonstrated during the 2001 FMD outbreak, when animal movements spread FMD to twelve distinct locations across the UK before the introduction of nation-wide movement restrictions on February 23, 2001.
It is a legal requirement in the UK that keepers of cattle report most movements of their cattle to the British Cattle Movement Service (BCMS). The RADAR Project makes this BCMS data available to researchers.
Research at Warwick based on this cattle movement data is aiming to understand what features of the way in which cattle are moved around the country are important for the dynamics of infectious diseases within the UK cattle herd, and how cattle movement patterns might be modified to reduce the risk of disease outbreaks. We are interested in a range of related topics, including the mathematical representation of the BCMS movement data, the simulation of diseases upon networks derived from it, the generation of model networks based on BCMS data for studying potential control strategies, and the policy implications of research upon cattle movement networks.
Cattle Movements in the US
Understanding animal movements can be crucial to predicting disease spread in livestock networks. Unfortunately, in the U.S. cattle industry, animal movements are poorly described except at coarse spatial (i.e., state) and temporal (i.e., yearly) resolutions. We attempt to fill this data gap by sampling paper Certificates of Veterinary Inspection (CVIs) to obtain premises level information on cattle shipments. Using CVI data, we create network models by aggregating premises at two spatial scales: the previously considered state scale and the higher resolution county scale. At both scales, the U.S. cattle network is well connected and exhibits a large giant strongly connected component (GSCC). However, in the state network, the GSCC contains virtually the entire network, while the county network exhibits numerous satellite nodes that either send to or receive from the GSCC. Centrality measures such as betweenness also suggest that the state level network may aggregate over potentially important heterogeneities at the county scale. Indeed, simulated disease spread across the county network indicates that epidemic behavior is highly dependent upon the site of introduction, whereas state level simulations show spread to the entire county, irrespective of introduction site. This work is part of a collaborative project withe Professor Colleen Webb of Colorado State University and Professor Uno Wennergren of Linkopings University funded by the Department of Homeland Security. The aim of the project is to determine the risk of spread of foot-and-mouth disease and bovine tuberculosis through the cattle movement network in the USA.
In Great Britain all pig holdings are registered and have their own unique county parish holding (CPH) number. All movements of pigs must be recorded (now entirely electronically) whether they involve the movement of one pet micropig to a vet or 1000 bacon pigs to slaughter. Like with cattle movements the data is available through the RADAR Project.
Unlike the cattle data, whereby every animal has it's own unique identifier a 'passport', this does not exist for pigs. Reasons for this may be the large numbers of pigs that are moved in anyone go in the pig industry, and the shorter life span of pigs in commercial farms. This makes the pig movement data less detailed.
Pigs in the industry can sometimes be moved between farms 2 or 3 times in a 6 month lifespan. A higher rate of movements leads to an increased risk of transmission. Having worked with the RADAR data, it has become apparent that the networks can be very large, primarily due to the pyramidal nature of the industry, whereby a small number of farms produce grandparent stock, feeding into another level of farms producing parent stock, and then further (often multiple) levels producing food stock.
The networks of pig farms can be complex, and do not apparently conform to “perceived wisdom”.
Before April 2012, movement data was recorded via paper forms sent to trading standards and then manually enetered into a database. The movement data had not been analysed before this time. Recent movement to electronic recording has brought to light some of the inconsistencies within GB pig movements.
Funded by: Wellcome Trust, BBSRC
WIDER people involved:
Ellen Brooks Pollock (Cambridge)
Mike Tildesley (Nottingham)