The decision to cull a cow was made by the owner without knowledge of the N. caninum serological

status of the animals. The proportions of culled N. caninum-seropositive and seronegative cattle per 100 cow-years were, respectively, 22.22% and 23.60% at Farm I; 11.77% and BTK inhibitor 15.24% at Farm II; and 32.97% and 23.21% at Farm III. The mean ages at the time of culling the seropositive and seronegative cattle were, respectively, 4.69 ± 3.00 years (range, 3.29–9.14 years) and 4.83 ± 2.63 years (range, 0.57–9.87 years) at Farm I; 4.68 ± 3.76 years (range, 0.67–9.75 years) and 4.29 ± 3.34 years (range, 0.71–11.66 years) at Farm II; and 5.17 ± 2.82 years (range, 0.69–10.68 years) and 5.58 ± 3.68 years (range, 0.66–15.39 years) at Farm III. In all herds, there was no significance difference

(P < 0.05) in culling rate between the cattle that were seropositive and seronegative for N. caninum infection. There was a wide variation in N. caninum prevalence in the herds investigated and, within herds, variations were observed over the sampling times and stock classes. These values are within the BIBF 1120 manufacturer range of previous studies in Brazil, in which a wide range in prevalence values among cattle was also observed, from 0.0 to 91.2% ( Gondim et al., 1999, De Melo et al., 2001 and Guedes et al., 2008). The high vertical transmission rate of N. caninum at Farms II and III (83.33% at both farms) is very similar to the congenital infection values (81–95%) reported in other studies ( Paré et al., 1997, Wouda et al., 1998, Hietala and Thurmond, 1999 and Dijkstra et al., 2001). The lower value found at Farm I (50%) was also in agreement with other studies ( Bergeron et al., 2000, Chanlun et al., 2007 and Moré et al., 2010). Bergeron et al. (2000) and Chanlun et al. (2007) DNA ligase suggested that disparities between rates of vertical transmission may be explained by the variability in prevalence of seropositive animals. In fact, the high degree of correlation between the vertical transmission rate and the prevalence of seropositive animals at Farms II and III suggests that only in herds with high prevalence were high levels of transmission observed. Two explanations for this correlation

are worth examining: first, high herd prevalence of seropositive animals may reflect a high proportion of active versus latent infections; second, the positive predictive value of IFAT must be considered in interpreting herd results. In low-prevalence herds, like Farm I, the predictive value of a positive test is low, because of the high proportion of negative animals. Therefore, in low-prevalence herds, a higher proportion of false-positive results may be expected, in relation to high-prevalence herds. It appears that estimated vertical transmission rates are more reliable in high-prevalence herds than in low-prevalence herds. No association between age and congenital infection rate was found in the present study, as also reported by Paré et al. (1996).