Modeling of stable isotope and multi-element compositions of jujube (Ziziphus jujuba Mill.) for origin traceability of protected geographical indication (PGI) products in Xinjiang, China

•Stable isotope and multi-element compositions of jujube was used for its origin traceability.•Regional, varietal, inter-annual variations and the correlations with eco-environment were discussed.•OPLS-DA model significantly enhanced the traceability capability for Xinjiang jujube products.•The discrimination accuracies for two PGI jujube products in Xinjiang were higher than 85 %.•The strategy can be applied to combat mislabeling of PGI jujube products and ensure food safety. Jujube is widely grown in Xinjiang province, China, including two high-value PGI products, Charkhlik Hui jujube and Khotan Jun jujube. Origin mislabeling and substitution of PGI jujube by inferior products seriously harms their reputation and has potential food safety risks. In this study, stable isotope (δ13C, δ15N, δ2H, δ18O) and elemental (Na, Mg, Al, P, K, Ca, Ti, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Sr, Mo, Cd, Sb, Ba, Pb) compositions of jujube from five different regions across Xinjiang province were analyzed by orthogonal partial least squares-discriminant analysis (OPLS-DA) to verify the geographical origin of jujube and authenticate PGI products. A total of 167 Hui jujube and 156 Jun jujube samples, together with their associated soils were collected annually from 2013 to 2018. The regional, varietal and inter-annual differences of stable isotope and elemental compositions of jujube and their correlation with soil values were analyzed using one-way ANOVA and multivariate statistics. The discriminant accuracies of OPLS-DA modeling for both Hui and Jun jujube samples collected in 2016 were higher than 90 %. Over a longer five-year (from 2013 to 2018), the discriminant accuracies decreased slightly, but were still acceptable at 85 % and 75 % for the two varieties. The most important variables for discrimination models were Na, Al, Ba, and δ13C for Hui jujube, δ18O and δ15N for Jun jujube, respectively. The study show that this strategy holds good promise as a tool to combat mislabeling and fraudulent conduct and has the ability to protect PGI jujube produced in Xinjiang province.