Verifying origin claims on dairy products using stable isotope ratio analysis and random forest classification

Area of Science:

• Food science and analytical chemistry
• Application of stable isotope ratio analysis (SIRA)
• Machine learning in food authenticity

Background:

• Geographic origin claims on food labels require scientific validation to build consumer trust.
• Stable isotope ratio analysis (SIRA) leverages natural isotopic variations for origin verification.
• Dairy products like butter, cheese, and whole milk powder are key food items for origin authentication.

Purpose of the Study:

• To scientifically validate geographic origin claims for dairy products using SIRA.
• To assess the effectiveness of SIRA in differentiating dairy products from various global regions.
• To develop a predictive model for dairy product origin based on isotopic signatures.

Main Methods:

• Casein was isolated from butter, cheese, and whole milk powder samples (n=197).
• Stable isotope ratios for Carbon (δ13C), Hydrogen (δ2H), Nitrogen (δ15N), Oxygen (δ18O), and Sulfur (δ34S) were measured.
• A random forest machine learning model was employed to predict the region of origin.

Main Results:

• The SIRA-based random forest model achieved an overall accuracy of 88% in predicting product origin.
• High class accuracy was observed for Irish (95%) and US (94%) products.
• The model also demonstrated good accuracy for European (84%) and Australasian (71%) products.

Conclusions:

• SIRA is a robust analytical technique for verifying the geographic origin of dairy products.
• Machine learning models utilizing isotopic data can accurately assign dairy products to their region of origin.
• This approach enhances the reliability of food origin claims, supporting market transparency and consumer confidence.