Seroprevalence, production impacts, economics and risk factors of Mycobacterium avium subspecies paratuberculosis in Canadian dairy cattle

The purpose of this thesis was to determine the seroprevalence, production impacts, economics and risk factors associated with Mycobacterium avium subspecies paratuberculosis (MAP) in Canadian dairy herds. Seroprevalence of bovine leukemia virus (BLV) and bovine viral diarrhoea virus (BVDV), and production impacts of BLV, BVDV and Neospora caninum (NC) were also estimated. Within 373 randomly selected Canadian herds (from New Brunswick, Nova Scotia, Prince Edward Island, Quebec, Ontario, Manitoba, Saskatchewan and Alberta), a serum sample was obtained from approximately 30 randomly selected lactating animals for ELISA testing for exposure to BLV, MAP and NC, while 5 unvaccinated cattle over 6 months of age were selected for virus neutralization testing for exposure to BVDV.

Seroprevalence at the cow level was 30.3% and 3.1% for BLV and MAP, respectively. At the herd level, 76.6% had at least one seropositive cow for BLV, 31.2% had at least one animal with a titer >1:64 for BVDV, and 18.9% had at least two seropositive cows for MAP. The Global Moran's I statistic for BLV and MAP indicated significant positive spatial autocorrelation of 0.16 and 0.15. The K-function for BVDV did not indicate any significant spatial autocorrelation. Using the spatial scan statistic, 4 and 1 significant clusters were detected for BLV and MAP-seroprevalence, respectively. No significant clusters were detected for BVD.

Regarding 305 day milk production (305 d milk), MAP-seropositivity was significantly associated with a lower 305 d milk of 212 kg in 4-plus lactation cows. NC-seropositivity was associated with a lower 305 d milk of 158 kg in primiparous cows compared to NC-seronegative primiparous cows. Cows in BVDV-seropositive herds (at least one unvaccinated animal with a titer ≥ 1:64) had significantly lower 305 d milk (by 368 kg) compared to cows in BVDV-seronegative herds. Among 1st lactation animals, NC-seropositivity was associated with a significant reduction of 5.5 kg and 3.3 kg of 305 d fat and 305 d protein yields, respectively. However, there were no statistically significant effects of MAP-seropositivity on 305 d fat or 305 d protein yields. BVDV-seropositivity (at the herd level) was associated with reductions in 305 d fat and 305 d protein yields of 10.20 kg/cow and 9.46 kg/cow, respectively. Regarding somatic cell counts, NC- (4+ lactation) and MAP-seropositivity were associated with a 0.094 decrease and 0.098 increase in the lactational mean of the log 10 somatic cell count. The cows in BVDV-seropositive herds had a significant increase in the mean log10 somatic cell count of 0.096. BLV-seropositivity was not associated with 305 d milk, 305 d fat, 305 d protein or mean log 10 somatic cell count. There was no significant effect modification of any of the outcomes examined from interaction between seropositivity for any of the pathogens.

For all reasons of culling, MAP-seropositive cows had a 1.38 (95% C.I., 1.05--1.81) times increased hazard of culling compared to MAP-seronegative cows. Seropositivity for the other pathogens was not associated with an increased risk of overall culling. Among cows that were culled because of either decreased reproductive efficiency or decreased milk production or mastitis, MAP-seropositive cows were associated with 1.55 (95% C.I., 1.12--2.15) times increased hazard compared to MAP-seronegative cows. Among cows that were culled because of reproductive inefficiency, NC-seropositive cows had a 1.43 (95% C.I., 1.15--1.79) times greater hazard than NC-seronegative cows. Among cows that were culled because of decreased milk production, cows in BVDV-seropositive herds had a 1.86 (95% C.I., 1.28--2.70) times increased hazard compared to cows in BVDV-seronegative herds. BLV-seropositive cows did not have an increased risk of reason-specific culling as compared to BLV-seronegative cows. No significant interaction on culling among seropositivity for the pathogens was detected, but only a limited number of cows tested seropositive for multiple pathogens. Results from our research will help in better understanding the economic impacts of these pathogens and justification for their control.

For the MAP-seropositive Canadian dairy herds, the mean loss per 61-cow herd was CD $2,992 (95% C. L, $143--$9741) annually, or CD $49 per cow per year. Herd additional culling losses were responsible for 46% (CD $1374) of the total losses from MAP. Decreased milk production, mortality and reproductive losses accounted for 9% (CD $254), 16% (CD $488) and 29% (CD $875) of the losses, respectively.

The following factors were significantly associated with the number of MAP-seropositive cows: "more than one cow in the maternity pen", "group housing in pre-weaned calves in winter", "purchasing open heifers", "direct contact with beef cattle", "BVDV-seropositive herds" and "not having proper BVD vaccination in calves (i.e. animals were not boostered 2-4 weeks after their first killed inoculation)" with count ratios of 1.69, 2.03, 2.34, 1.27, 1.41 and 1.77, respectively. The variable "herds having BVDV modified live vaccination in calves" was associated with 0.44 times fewer MAP-seropositive cows. There were significant regional differences in the number of positive cows per herd that were not explained by the management factors included in the final multivariable model. In the logistic portion of the final Zero Inflated Negative Binomial model, a one unit increase in the mean lactation number of cows tested in a herd was associated with an increase of 1.44 times the odds of being herds having at least one MAP-seropositive cow. Herds having "feeding total mixed ration" and "bedding that was not added frequently to calving areas (for each calving)" were associated with 3.1 and 2.7 times greater odds of being MAP-seropositive herds, respectively. The variables "NC-seropositivity" and "area of pasture (more than 200 acres)" were marginally significant (P 0.10) with MAP-seropositivity.

The results from this study will improve the knowledge of seroprevalence, production impacts, economical losses and risk factors of MAP in Canadian dairy industry and will assist management recommendations in national Johne's control programs.