Analysis of the influence of variation factors on electrical conductivity of milk in Murciano-Granadina goats.

The aim of this study was to obtain further knowledge on electrical conductivity (EC) of milk as a tool for detecting mastitis in goats. The effect of farm, parity, stage of lactation, and health status of the glands on EC, and the somatic cell count (SCC) of milk was analyzed. Additionally, relationships between EC and chemical composition and SCC were studied. Finally, characteristics of EC and SCC (sensitivity and specificity) as diagnostic tests used to detect mastitis were studied. One hundred and five Murciano-Granadina goats were enrolled in the study. Milk samples (by gland) were collected monthly for 7 mo on 3 farms in the southeastern Spain. To establish the health status, milk samples were aseptically collected before milking by gland. Foremilk (by gland) was collected to analyze EC, SCC, and chemical composition. Glands were classified according to the health status as free of mastitis, bacterial mastitis, or unspecific mastitis. The effects of farm, parity, and stage of lactation, as well as the interactions between health status and parity, parity and stage of lactation, and health status and stage of lactation were associated with EC. Changes in the milk's chemical composition (particularly of chloride ions) explained most of the variance in EC (R(2)=0.91). The strongest association between EC and SCC was found at SCC >2×10(6) cells/mL (r=0.42). The use of a single EC threshold for all animals and farms for detecting mastitis led to limited results for mastitis detection, which, in any case, favors negative predictive values over positive predictive values. This study revealed that factors, other than the health status, affecting EC hamper the use of an EC threshold for mastitis detection with sufficient specificity on all animals. Any detection system based on EC of milk should consider these factors, as well as specific variations for each of the animals.