Bulk tank milk quality data is unlikely to give useful information about dairy cow welfare at herd level.

This research communication explores the value of routinely collected bulk tank milk quality data for estimating dairy cattle welfare at herd level. Selected bulk tank milk quality parameters (somatic cell count, total bacterial count, urea, protein and fat contents) recorded during the years 2014-2016 in 287 Italian dairy farms were compared with the animal welfare data of each farm. The welfare assessment data were extracted from the database of the Italian Reference Centre for Animal Welfare (CReNBA), which includes the outputs of the application of the CReNBA welfare assessment protocol for dairy cows, used at national level for on-farm controls. The statistical analysis was carried out using the correlation coefficient for Kendall's Tau ranks, in order to investigate the presence of a categoric relationship between the selected bulk tank milk quality parameters and the overall animal welfare score or the scores of the single areas A (farm management and staff training), B (housing) and C (animal-based measures). Somatic cell count, total bacterial count, urea and proteins demonstrated only a few statistically significant and very weak correlations with farm animal welfare data, while no significant correlations were obtained for milk fat content. Given the weak correlations found, the selected bulk tank milk parameters seems to be able to provide only limited information about the welfare level of the herd, thus it could be difficult to use them for drawing up a pre-screening model for identifying herds at risk of poor welfare.

Diagnostic value of composite milk sample vs single quarter milk sample for the detection of Staphylococcus aureus intra-mammary infections in cattle.

Staphylococcus aureus (S. aureus) is one of the most important causes of mastitis in dairy cattle. Control and eradication programs of S. aureus intra-mammary infections (IMI) are based on different factors included the correct detection and management of the infected cows. The present study aimed at evaluating the efficacy of composite milk samples (CMS) analysis, compared to quarter milk samples (QMS) analysis, for the bacteriological detection of S. aureus intra-mammary infections. During 2016, 661 CMS (hygienically collected) and 2644 QMS (aseptically collected) were obtained from 661 cows in 5 herds. All the samples were submitted to S. aureus bacteriological culture and somatic cell count (SCC) analysis. QMS bacteriological analysis on blood agar plates was able to detect 236 cows excreting S. aureus, while the bacteriological analysis of CMS, using selective agar, identified 229 positive cows. The concordance was 95% with an excellent Cohen's κ (0.89). Relative sensitivity and specificity of CMS vs QMS, considered as the reference test, were 91.5% ± 2.1 and 96.9% ± 1.3 (CI 95%), respectively. In addition, the relative sensitivity of CMS improved as the number of infected quarters per cow and the number of colony forming units (cfu) per sample increased. The predictive value of CMS results was better when paired with SCC data, in particular CMS showed better negative predictive value when SCC was <200,000 cells/mL and better positive predictive value when SCC was>200,000 cells/mL. The probability for a cow to be S. aureus positive was 56.4% in case of SCC > 200,000 cells/mL, while it was 18.6% in case of SCC < 200,000 cells/mL. The average SCC in CMS was significantly higher in positive cows and the value rose as the number of infected quarters per cow increased. Given the intermittent excretion of S. aureus in milk from dairy cows, it could be more advantageous to carry out several serial CMS, rather than few QMS, being CMS an easier to collect and less expensive milk sampling method. Thus, bacteriological examination of CMS, combined with SCC data of the same sample, could be extremely useful for the success of S. aureus IMI control plans, because repeated CMS are easier to be performed and could be more easily proposed to the farmers.