Leveraging milk mid-infrared spectroscopy to authenticate animal welfare, farming practices, and dairy systems of Parmigiano Reggiano cheese

Abstract

Increasing consumer concerns underscore the importance of verifying the practices and origins of food, especially certified premium products. The aim of this study was to evaluate the ability of Fourier-transform mid-infrared (FT-MIR) spectroscopy to authenticate animal welfare parameters, farming practices, and dairy systems. Data on farm characteristics were obtained from the Parmigiano Reggiano Consortium in northern Italy. Animal welfare data were collected by trained veterinarians using the assessment protocol developed by the Italian National Reference Center for Animal Welfare (CReNBA), and bulk milk test-day data were obtained from the laboratory of the Breeders Association of the Emilia Romagna Region. A merged final dataset of 12,083 bulk FT-MIR spectra records from 949 farms was created. Using a nonhierarchical clustering approach, the farms were classified into 5 dairy systems: 2 traditional systems comprising farms located in either the Apennines or the Po Plain; 2 modern systems, one that used TMR and one did not; and one traditional dairy system comprising farms rearing local breeds. To evaluate the ability of bulk milk to capture differences in farming systems, we conducted an ANOVA on milk composition. The linear models included the following effects: season, dairy system, farm, and the interaction between dairy system and season. The effect of the dairy system was significant for all milk composition traits. A 10-iteration linear discriminant analysis was used to evaluate the discriminative ability of the spectra in classifying farming practices and dairy systems. The average results of the area under the receiver operating characteristic curve revealed good authentication performance for genetic type (0.98), housing system (0.91), and feeding system (0.89), and medium-low authentication performance for geographical area (0.70); poor results were obtained for the percentage of concentrate in the diet and animal welfare parameters (0.57-0.64). With regard to dairy systems, the best result was obtained when dairy systems were grouped into 2 simplified categories, traditional versus modern (0.89), instead of the 5 categories (0.87). The results of this study show that FT-MIR is a useful tool for authenticating farming practices and dairy systems, but not animal welfare as defined by CReNBA evaluation criteria. Our results show that infrared spectroscopy has the potential to authenticate dairy products and quality label certifications.

Keywords: Fourier-transform mid-infrared; animal welfare; dairy cow; discriminant analysis; quality labels.