Correlation among genetic, Euclidean, temporal, and herd ownership distances of porcine reproductive and respiratory syndrome virus strains in Quebec, Canada

Abstract

Background: Porcine reproductive and respiratory syndrome (PRRS) is a viral disease that has a major economic impact for the swine industry. Its control is mostly directed towards preventing its spread which requires a better understanding of the mechanisms of transmission of the virus between herds. The objectives of this study were to describe the genetic diversity and to assess the correlation among genetic, Euclidean and temporal distances and ownership to better understand pathways of transmission.

Results: A cross-sectional study was conducted on sites located in a high density area of swine production in Quebec. Geographical coordinates (longitude/latitude), date of submission and ownership were obtained for each site. ORF5 sequencing was attempted on PRRSV positive sites. Proportion of pairwise combinations of strains having ≥98% genetic homology were analysed according to Euclidean distances and ownership. Correlations between genetic, Euclidean and temporal distances and ownership were assessed using Mantel tests on continuous and binary matrices. Sensitivity of the correlations between genetic and Euclidean as well as temporal distances was evaluated for different Euclidean and temporal distance thresholds. An ORF5 sequence was identified for 132 of the 176 (75%) PRRSV positive sites; 122 were wild-type strains. The mean (min-max) genetic, Euclidean and temporal pairwise distances were 11.6% (0-18.7), 15.0 km (0.04-45.7) and 218 days (0-852), respectively. Significant positive correlations were observed between genetic and ownership, genetic and Euclidean and between genetic and temporal binary distances. The relationship between genetic and ownership suggests either common sources of animals or semen, employees, technical services or vehicles, whereas that between genetic and Euclidean binary distances is compatible with area spread of the virus. The latter correlation was observed only up to 5 km.

Conclusions: This study suggests that transmission of PRRSV is likely to occur between sites belonging to the same owner or through area spread within a 5 km distance. Both should be considered in the perspective of prevention.

Figure 1

Correlogram showing Mantel r statistic computed on binary matrices of genetic (≥98%, <98% homology) and Euclidean distances for different thresholds (km). Results from 7381 different pairwise combinations of 122 PRRSV ORF5 sequences. Dark dots indicate significant correlation (P ≤0.003) after 9999 permutations of matrices.

Figure 2

Correlogram showing Mantel r statistic computed on binary matrices of genetic (≥98%, <98% homology) and temporal distances for different thresholds (month). Results from 7381 different pairwise combinations of 122 PRRSV ORF5 sequences. Dark dots indicate significant correlation (P ≤ 0.004) after 9999 permutations of matrices.