Estimating time evolving cross-dependence of porcine reproduction and respiratory syndrome sero-prevalence in Danish swine herds

Porcine reproductive and respiratory syndrome (PRRS) has been a challenge for the Danish swine industry. It is important to identify forms of time-dependence in PRRS sero-prevalence among swine herds at country level, in order to allocate resources in an effective manner. This is of particular value if an increase of PRRS sero-prevalence occurs in a particular type of herds and control measures are taken to avoid disease spread downwards the swine production system.

The objective of this study was to describe the presence (or lack thereof) of the linear cross-dependence of PRRS sero-prevalence in Danish swine herds from 2012 to 2014. These temporal dependencies were calculated for Danish swine herds with different biosecurity status and directional trade contacts between them: from nucleus/breeding herds to multipliers and finisher herds and from nucleus/breeding herds to finishers via multiplier herds.

We used a method proposed for non-stationary time-series decomposition which allows an assessment of the existence of linear cross-dependencies between multiple of types of herds. The results demonstrate the existence of cyclic patterns of direct linear positive and negative dependence between PRRS sero-prevalence between multipliers and finishers herds.

This suggests that the associations between PRRS sero-prevalence time-series follow the swine production system downwards for the majority of the study period, i.e. an increase or decrease in PRRS sero-prevalence in multiplier is followed by a shift in PRRS sero-prevalence in finisher herds after a few weeks.

For specific time periods, however, these associations might have been influenced by other factors.