Herd-level prevalence of bovine leukemia virus, Salmonella Dublin and Neospora caninum in Alberta, Canada, dairy herds using ELISA on bulk tank milk samples

Abstract

Endemic infectious diseases remain a major challenge for dairy producers worldwide. For effective disease control programs, up-to-date prevalence estimates are of utmost importance. The objective of this study was to estimate the herd-level prevalence of bovine leukemia virus (BLV), Salmonella Dublin, and Neospora caninum in dairy herds in Alberta, Canada using a serial cross-sectional study design. Bulk tank milk samples from all Alberta dairy farms were collected 4 times, in December 2021 (n = 489), April 2022 (n = 487), July 2022 (n = 487), and October 2022 (n = 480), and tested for antibodies against BLV, S. Dublin, and N. caninum using ELISAs. Herd-level apparent prevalence was calculated as positive samples divided by total tested samples at each time point. A mixed effect modified Poisson regression model was employed to assess the association of prevalence with region, herd size, herd type, and type of milking system. Apparent prevalence of BLV was 89.4, 88.7, 86.9 and 86.9% in December, April, July, and October, respectively, whereas for S. Dublin apparent prevalence was 11.2, 6.6, 8.6, and 8.5%, and for N. caninum apparent prevalence was 18.2, 7.4, 7.8, and 15.0%. For BLV, S. Dublin and N. caninum, a total of 91.7, 15.6, and 28.1% of herds, respectively, were positive at least once, whereas 82.5, 3.6, and 3.0% of herds were ELISA-positive at all 4 times. Compared with the north region, central Alberta had a high prevalence (prevalence ratio (PR) = 1.13) of BLV-antibody positive herds, whereas south Alberta had a high prevalence (PR = 2.56) of herds positive for S. Dublin antibodies. Furthermore, central (PR = 0.52) and south regions (PR = 0.46) had low prevalence of N. caninum-positive herds compared with the north. Hutterite colony herds were more frequently BLV-positive (PR = 1.13) but less frequently N. caninum-positive (PR = 0.47). Large herds (>7,200 L/day milk delivered ∼ > 250 cows) were 1.1 times more often BLV-positive, whereas small herds (≤3,600 L/day milk delivered ∼ ≤ 125 cows) were 3.2 times more often N. caninum-positive. For S. Dublin, Hutterite-colony herds were less frequently (PR = 0.07) positive than non-colony herds only in medium and large stratum but not in small stratum. Moreover, larger herds were more frequently (PR = 2.20) S. Dublin-positive than smaller herds only in non-colony stratum but not in colony stratum. Moreover, N. caninum prevalence was 1.6 times higher on farms with conventional milking systems compared with farms with an automated milking system. These results provide up-to-date information of the prevalence of these infections that will inform investigations of within-herd prevalence of these infections and help in devising evidence-based disease control strategies.