Descriptive and social network analysis of pig transport data recorded by quality assured pig farms in the UK.

The aim of this study was to examine farm-to-farm pig movement connections, and the usage of hauliers and abattoirs, for farms in the United Kingdom (UK), to determine the interconnectivity of the pig farm network and the implications this may have for the transmission and control of Salmonella, which was chosen as an example of an important endemic disease. Data were collected from three Quality Assurance Schemes on the use of abattoirs and livestock hauliers by commercial pig farms, and the supply of pigs to and from farms. The observed dataset had considerable variability in the number of direct (farm-to-farm) and indirect (via hauliers or abattoirs) connections between each farm over a 12-month period. The use of multiple hauliers and abattoirs by many farms resulted in considerable interconnectivity between farms due to these indirect routes. The network displayed a higher level of clustering, and short network distances between farms, than that shown by equivalent randomly generated datasets, indicating that infectious diseases might transmit quickly amongst farms but within certain clusters of farms, thus limiting the total number of farms affected. This structure and the occurrence of multiple indirect routes between many pairs of farms (via pig movements to other interconnected farms or from hauliers used by both farms), may indicate that targeting surveillance and controls on those farms with high network centrality characteristics (degree, betweenness) alone would not prove effective. Encouraging all farms within the network to reduce the number of connections might be a useful way of reducing Salmonella prevalence as it would reduce the effect of high prevalence on other farms within the network. The analysis also highlighted differences in the connections between geographical regions that were used to discuss the comparatively low Salmonella prevalence in Scotland detected by previous studies. Farms identified as belonging to large companies, as categorised by our analysis, were shown to mostly have movement connections within that company. However, small company farms had connections to farms belonging to large companies or other small companies, demonstrating that they may play an important role in creating links between large companies. The majority of farms in both Scotland and North-East and Eastern England belonged to large companies, and this finding is discussed in relation to how this may help explain the lower Salmonella seroprevalence in these areas. The study provides a first description of the characteristics of the UK pig movement network and the analysis has indicated a number of findings that might have implications for disease transmission and targeting surveillance and control. Crown