Abstract

Milk is generally analyzed on a monthly basis in dairy animals under official recording schemes. Other than conventional composition, nowadays one sample of milk can provide a variety of interesting phenotypes. This has been mainly possible thanks to infrared spectroscopy, a rapid and low-cost technology implemented worldwide and accurately recognized as a gold standard for some traits (e.g., milk gross composition). Milk is secreted by the mammary gland and mirrors the physiological and metabolic status of the animal. Potentially, milk spectra can be used to predict biomarkers useful for health status assessment and fitness monitoring, such as ketone bodies, urea nitrogen, fatty acid profile, minerals, and protein fractions. In addition, it has been demonstrated that traits not directly measurable in milk can also be predicted with sufficient accuracy, e.g., blood parameters, reproductive traits, body weight, methane emissions, and feed intake. Finally, devices collecting milk spectra are commonly equipped with a flow cytometry machine that can quantify the milk somatic cell count (SCC), widely adopted as an indicator of udder health by farmers, veterinarians, researchers, and breeders. Therefore, milk sampling is a non-invasive procedure that allows for collection of phenotypes even at a population level. This medium can be exploited for diagnostic purposes in dairy species to track difficult-to-be-measured features at individual level, such as health, fertility, and efficiency.