Stable isotope ratio and elemental composition parameters in combination with discriminant analysis classification model to assign country of origin to commercial vegetables - A preliminary study

- Acceptable discrimination of vegetable products according to the country of origin was achieved.
- Combination of stable isotope and elemental composition was used.
- The classification efficiency of the proposed multivariate statistical models was 71-86%.
- Methodology could be used for establishment of databanks and creation of traceability models.
- Methodology could be used for rapid food screening purposes.

Recently, increased public attention has been paid to the geographical authentication of food, including vegetables, which are considered to be one of the major health-promoting components in a balanced diet. The purpose of the present study was to investigate the suitability of the use of isotopic compositions of light elements (δ13C, δ15N, δ18O, δ34S) in combination with multi-elemental fingerprinting (P, S, Cl, K, Ca, Mn, Fe, Zn, Br, Rb, Sr) to provide rapid, robust and inexpensive screening methods for distinguishing lettuce, sweet pepper, and tomato samples according to their given country of origin (i.e., Slovenia, Austria, Spain, Morocco, Italy, Greece), and thus ensuring their traceability in terms of their authenticity. The classification efficiency of the proposed multivariate statistical models using supervised pattern-recognition analysis, namely multivariate discriminant analysis, was sufficient for rapid and robust screening purpose. The predictions of the suggested discriminant analysis models per kind using cross-validation leave-one-out were 86.2%, 71.1% and 74.4% for lettuce, sweet pepper and tomato, respectively. The first use of the proposed methodology on vegetable samples on European and Mediterranean scales provides a valuable and necessary contribution to the development and implementation of a new national surveillance system that can be used to trace the geographical origins of vegetables.