Exploring the sources of variation of electrical conductivity and total and differential somatic cell count in Italian Mediterranean buffaloes.

In the buffalo dairy sector, a huge effort is still needed to improve mastitis prevention, detection, and management. Electrical conductivity (EC) and total somatic cell count (SCC) are well-known indirect indicators of mastitis. Differential somatic cell count (DSCC), which represents the proportion of neutrophils and lymphocytes on the total SCC, is instead a novel phenotype collected in the dairy cattle sector in the last lustrum. As little is known about this novel trait in dairy buffalo, in the present study we explored the nongenetic factors affecting DSCC, as well as EC and total somatic cell score (SCS), in the Italian Mediterranean buffalo. The data set used for the analysis included 14,571 test-day (TD) records of 1,501 animals from 6 herds, and climatic information of the sampling locations. The original data were filtered to exclude animals with less than 3 TD per lactation and, for the investigated traits, outliers beyond 4 standard deviations. In the statistical model we included the fixed effects of herd (6 classes), days in milk (DIM; 10 classes of 30 d, with the last being an open class until 360 d), parity (6 classes, from 1 to 6+), year-season of calving (11 classes, from summer 2019 to winter 2021/2022), year-season of sampling (9 classes, from spring 2020 to spring 2022), production level (4 classes based on quartiles of average milk yield by herd), and temperature-humidity index (THI; 4 classes based on quartiles, calculated using the average temperature and relative humidity of the 5 d before sampling). Average EC, SCS, and DSCC vary across herds. Considering DIM, greater EC values were observed at the beginning and the end of lactation; SCS was slightly lower, but DSCC was greater around the lactation peak. Increased EC, SCS, and DSCC levels with increasing parity were reported. Year-season calving and year-season sampling only slightly affected the variation of the investigated traits. Milk of high-producing buffaloes was characterized by lower EC and SCS mean values, nevertheless it had slightly greater DSCC percentages. Buffaloes grouped in the highest THI classes (classes 3 and 4) showed, on average, greater EC, SCS, and DSCC in comparison to the lower classes, especially to class 2. Results of the present study represent a preliminary as well as necessary step for the possible future inclusion of EC, SCS, or DSCC in breeding programs aimed to improve mastitis resistance in dairy buffaloes.

Exploiting machine learning methods with monthly routine milk recording data and climatic information to predict subclinical mastitis in Italian Mediterranean buffaloes.

Mastitis has detrimental effects on the world's dairy industry, reducing animal health, milk production and quality, as well as income for farmers. In addition, consumers' growing interest in food safety and rational usage of antibiotics highlights the need to develop novel strategies to improve mastitis detection, prevention, and management. In the present study we applied machine learning (ML) analyses to predict presence or absence of subclinical mastitis in Italian Mediterranean buffaloes, exploiting information collected the previous month during routine milk recording procedures, as well as climatic data. The data set included 3,891 records of 1,038 buffaloes from 6 herds located in Basilicata Region (South Italy). Prediction models were developed using 4 different ML algorithms (Generalized Linear Model, Support Vector Machines, Random Forest, and Neural Network) and 2 data set splitting approaches for the creation of the training and test sets (by record or by animal ID number, always with 80% of the data used for model training and the remaining 20% for model testing). Support Vector Machine was the best method to predict high or low somatic cell count at the subsequent test-day record in the validation set, and therefore it was used to estimate the contribution of each feature to the best model. Independently from the data set splitting approach, the most important features were somatic cell score, differential somatic cell count, electrical conductivity, and milk production. Among climatic data, the most informative were temperature and relative humidity. When the data were split by animal ID, an improvement in models' predictive performance on the test set was observed, suggesting this as the most appropriate data splitting approach in data sets with repeated measures to avoid data leakage. According to different metrics, Neural Network was the best method for making predictions on the test set. Our findings confirmed the promising role of ML methods to improve prevention and surveillance of subclinical mastitis, exploiting the large amount of data currently available to identify animals that would possibly have high somatic cell count the subsequent month.