Modelling the risk of bacterial infections, within the framework of the evaluation of the health and welfare of dairy cows

Among the main strategies proposed by FAO for sustainable livestock production are improving the efficiency of livestock farming and the use of local resources. This research work addressed both challenges with two case studies. The first deals with the evaluation of four health indicators (HI) limiting the productivity of dairy cattle farming, and it used data collected within the framework of the dairy herd improvement program (DHI) in the Piedmont region over a 5-year period. The incidence, prevalence and cure rate of mastitis, the incidence of hyperketonemia, the duration of the calving interval and the culling rate of fresh cows were estimated for more than 1200 Piedmontese herds. For each HI, risk factors acting at individual animal level were evaluated and the impact of farm management was estimated. The geographic distribution of each HI was explored to identify areas where the risk was higher and foci of aggregation of high or low rates of HIs. Finally, an attempt was made to use the indicators as an early warning system for the risk of poor welfare (PWR). The large dataset allowed a precise estimation of individual-cow risk factors and to determine that at least 20% of the risk, depending on the HI, derived from herd management. DHI did not include information about herd type, limiting the comprehension of its impact on risk factors. On the other hand, the used model was general enough to permit herd benchmarking on a regional scale. The results suggest that along with high-producing herds in lowlands, other worse-performing and less controlled herds remained in marginal territories. The lack of information about those last led to the poor performance of HI as screening tools for PWR, as more is needed to train the detection model. However, HI can be used to confirm the absence of PWR thanks to the high specificity of the analysis. The second case study regarded the comparison of dairy cattle farms in marginal areas of the northern Apennines. Sixteen farms from three neighboring areas were analyzed, collecting information on herd management, antimicrobial use and the prevalence of intramammary infections. From cows’ milk analysis, mastitis pathogens were isolated and subsequently evaluated for susceptibility to antimicrobials. Although it was not possible to distinguish geographic areas based on interview results, they evidenced some differences and a widespread lack of awareness about herd health problems. Antimicrobial use was significantly higher in the highest producing herds, and intramammary infections were the most 3 frequent reason for antimicrobial administration. Nonetheless, antimicrobial resistance was at a low level in all herds. The microbiological results highlighted that pathogens with different epidemiology characterized different geographical areas, not due to environmental factors, but to management characteristics of the farms. The least productive farms were characterized by the presence of contagious pathogens likely because they failed to control the spread of the infection. The factor most associated with this unequal distribution of pathogens was the employment of specialized veterinarians to counsel and help managing the herds. In conclusion, to improve the efficiency of farms located in marginal territories, it would be necessary to increase farmers' awareness regarding herd health problems. For this purpose, DHI and animal welfare assessment seemed a suited strategy.