Abstract: This study aimed to characterize the traceability of soybeans with similar regional characteristics and improve the accuracy and stability of mineral elements in tracing the geographical origin of soybeans. In this experiment, inductively coupled plasma mass spectrometry (ICP-MS) was used to analyze the contents of mineral elements in 42 samples collected from 9 farms in Bei’an city and 6 farms in Nenjiang county, Heihe city, Heilongjiang province. The obtained data were analyzed by analysis of variance (ANOVA), principal component analysis (PCA) and discriminant analysis (DA). The results showed that the first principal component for these soybean samples mainly included V, Fe, La, Ce, Pr, Nd, Dy, Er and Yb, with largest contribution rate of 28.390%. The second principal component mainly included As, Se, Lu, Ir and Au, with middle contribution rate of 14.435%. The third principal component mainly included Ca, Ni, Cd and Ba, with smallest contribution rate was 10.881%. Thus, analysis of mineral content data was useful for tracing the geographical origin of soybeans from different producing areas in the province. Six metal elements, Na, K, Mn, Rb, Ba and Au, were identified as indicators for traceability analysis. Our results demonstrated that the geographical origin of soybeans from the two main producing areas in Heilongjiang province was identified with 100% accuracy.

Key words: soybean, geographical origin traceability, mineral element, principal component analysis (PCA), discriminant analysis (DA)