Establishing a surveillance programme for Salmonella Dublin in Austrian dairy herds by comparing herd-level vs. individual animal detection methods

Due to its increasing occurrence in cattle farms in various countries, leading to significant economic losses in affected livestock, Salmonella enterica subspecies enterica serovar Dublin (S. Dublin) has become a highly investigated pathogen in cattle production. In Austria, there have been occasional human cases of S. Dublin as well as an increase in laboratory-confirmed cases in cattle, indicating the need for a screening programme to determine the current status in Austria. The aims of this study were, firstly, to determine the seroprevalence of S. Dublin in dairy herds through bulk milk screenings in two federal states (Salzburg, Tyrol) of Austria. Secondly, the study aimed to identify the infection status of the herds through individual animal and herd level detection, comparing microbiological, molecular and serological detection methods. The results of the study will allow the development of a sampling strategy for a surveillance programme in Austria. A total of 6973 dairy farms were tested through serological bulk milk screening. The seroprevalence for the federal state of Tyrol was 14.8 % and for Salzburg it was 18.2 %, resulting in an average seroprevalence of 16.5 %. At an individual animal level, 205 (11.3 %) animals tested positive for shedding of S. Dublin in the faeces through microbiological detection, and 268 (17.0 %) animals had positive values (ct value ≤ 38) by qPCR. The association between microbiological and molecular detection was statistically significant (p < 0.001), with a calculated kappa value of 0.65 ± 0.27 (p ≤ 0.001), assuming a substantial level of agreement. In 17 herds, where an individual animal tested positive for shedding of S. Dublin, environmental sampling and testing were carried out. At a herd level 16 (94.1 %) out of the 17 participating herds, tested positive for S. Dublin either microbiologically or by molecular assay in boot swab samples. Bulk milk samples from 14 out of the 17 participating herds were analysed for antibodies to S. Dublin and 12 samples (85.7 %) were positive. In total 111 (18.9 %) out of 587 blood samples tested positive for S. Dublin antibodies, demonstrating a statistically significant correlation (p < 0.001) both with microbiological (κ = 0.32 ± 0.49; p ≤ 0.001) and molecular (κ=0.23 ± 0.06; p ≤ 0.001) findings. It was possible to identify S. Dublin by culture from boot swabs in 14 (82.4 %) out of 17 herds and by molecular assay using qPCR in 15 (88.2 %) out of 17 herds, indicating a suitabl