Carrots cultivated with organic and conventional agricultural systems: Discrimination using spectroscopic data and chemometric analysis

POLITI, M.; DASKO, L.; ALONSO-SALCES, R. M.; SEGEBARTH, N.; MATTARUCCHI, E.; RENIERO, F.; GIORDANO, G.; MAQUET, A.; GUILLOU, C.

Geel (Belgium)

Workshop; International Workshop on Organic Food Authentication: Challenge or Utopia?; 2009

European Commission - DG Joint Research Centre, Institute for Reference Materials and Measurements (IRMM)

Following our previous investigations aiming at the identification of discriminant metabolites betweenorganically and conventionally grown crops (Maquet et a/., 2007), carrots (Dauczrs cnrota L.) weregrown and collected under controlled conditions during two years (2007 and 2008) in two differentselected environments (1 and 2), and by using both, organic and conventional agricultural systems.Carrot samples of both harvest years were dried, lyophilized, and powdered; additionally, part of 2008samples were directly frozen in the field and then lyophilized and powdered (fresh samples). Allpowdered samples were stored at -80 O C until analysis.Powdered carrot samples were extracted with deuterated methanol in an ultrasonic water bath and thencentrifuged. The methanol extracts were directly analyzed by proton Nuclear Magnetic Resonance(NMR) spectroscopy and Ultra-Performance Liquid Chromatography coupled with a QuadrupoleTime of Flight Mass Spectrometry (UPLC-QTOFIMS). Several data processing strategies were testedfor both, 'H-NMR and LC-MS data. Then, chemometric techniques such as principal componentanalysis (PCA) and partial least squares discriminant analysis (PLS-DA) were applied.Our results indicate that LC-MS data contains enough information to create a PLS-DA model todistinguish organically and conventionally grown carrots. However, PLS-DA does not provide anymodel using only NMR data for this purpose. LC-MS and NMR data together enable the constructionof a model leading to better predictions for both classes than that based only on LC-MS data. Theseresults are only preliminary given the limited number of samples available for this study. A muchlarger sampling associated with a carefully designed experimental plan is needed to assess the factorsthat impact the metabolic profile of carrots cultivated with organic and conventional agricultureregime. The current study has allowed the identification of the gaps and the improvement needed inthe applied methodology which, despite these limitations, appears to be promising.