Bayesian kriging of seroprevalence to Mycobacterium avium subspecies paratuberculosis and Neospora caninum in Alberta beef and dairy cattle

Identifying spatial patterns of risk is important in the study of diseases with ecologic causes. Furthermore, relatively complex hierarchical modeling is required to determine how factors that are organized across levels interact, such as how an ecologic cause interacts with farm management and with animal characteristics. The objective of this study was to map the risk for Mycobacterium avium subspecies paratuberculosis (MAP - the causative agent of Johne's disease) and Neospora caninum (NC - the cause of neosporosis) infections in Alberta beef and dairy cattle. This objective utilized Bayesian generalized linear kriging to partition herd effects into a portion attributable to location and a portion that was independent of location. Seropositivity to NC in beef cattle showed strong support for spatial covariance, suggesting that ecologic causes were important for beef cattle but not dairy cattle. There was little evidence of spatial covariance for MAP seropositivity in either beef or dairy cattle.