Merging DNA typing and network analysis to assess the transmission of paratuberculosis between farms

Paratuberculosis, a chronic enteric infection caused by Mycobacterium subsp. paratuberculosis (MAP), is endemic in all farmed ruminant species in New Zealand. The use of genotyping in combination with network analysis of livestock movement events from one farm location to another has the potential to contribute to our understanding of between-farm transmission events. We studied a population of 122 farms from a corporate commercial livestock enterprise in New Zealand, trading with each other in near isolation from other commercial farms. The data consisted of longitudinal movements to and from these farms between 2006 and 2010, as well as the results of cross-sectional MAP screening and genotyping performed in 2010. We explored associations between past livestock movements and current strain type distribution in this population of farms using quadratic assignment procedure. Our results show that measures of farm clustering within the movement network were significantly associated with sharing of MAP strains. For example, farms closely related by trade were twice as likely to share the same strains of MAP (p = 0.033). Other covariates were also associated with the probability of sharing the same strains of MAP, such as being located on the same island (OR = 5.8 to 8.7, p < 0.01), farming the same livestock species and Euclidian distance between farms. The novel approach we used supports the hypothesis that livestock movement is indeed a significant contributor to farm-to-farm transmission of MAP.