hide
Free keywords:
network analysis; pig trade network; animal movements; monthly networks; yearly networks; Germany
Abstract:
A major risk factor in the spread of diseases between holdings is the transport of
live animals. This study analysed the animal movements of the pork supply chain
of a producer group in Northern Germany. The parameters in-degree and
out-degree, ingoing and outgoing infection chain, betweenness and ingoing and
outgoing closeness were measured using dynamic network analysis to identify
holdings with central positions in the network and to characterize the overall network
topology. The potential maximum epidemic size was also estimated. All
parameters were calculated for three time periods: the 3-yearly network, the yearly
and the monthly networks. The yearly and the monthly networks were more fragmented
than the 3-yearly network. On average, one-third of the holdings were
isolated in the yearly networks and almost three quarters in the monthly networks.
This represented an immense reduction in the number of holdings participating
in the trade of the monthly networks. The overall network topology
showed right-skewed distributions for all calculated centrality parameters indicating
that network resilience was high concerning the random removal of holdings.
However, for a targeted removal of holdings according to their centrality, a rapid
fragmentation of the trade network could be expected. Furthermore, to capture
the real importance of holdings for disease transmission, indirect trade contacts
(infection chain) should be considered. In contrast to the parameters regarding
direct trade contacts (degree), the infection chain parameter did not underestimate
the potential risk of disease transmission. This became more obvious, the
longer the observed time period was. For all three time periods, the results for the
estimation of the potential maximum epidemic size illustrated that the outgoing
infection chain should be chosen. It considers the chronological order and the
directed nature of the contacts and has no restrictions such as the strongly
connected components of a cyclic network.
