【摘要】 为确保水产品的质量和消费者的权益，开发高效的原产地保护方法和真伪鉴别技术尤为重要。稳定同位素与元素分析技术凭借其快速、灵敏和准确的特性，在映射环境信息及生物食源结构特征方面展现出独特优势，已成为水产品溯源领域备受瞩目的方法。对稳定同位素、元素分析技术结合多元统计分析方法在水产品溯源中的应用原理及最新研究进展进行了综述与展望。综述表明，大多数研究中水产品溯源的准确率仅为60%及以上，因此现有溯源体系仍存在一定的局限性，建议深入探索元素/稳定同位素“指纹图谱”的形成机制，加速溯源技术的改进与创新，并引入现代机器学习算法与多模型融合技术。研究可为水产品的地理来源示踪与生产方式判别等提供新的视角，并为构建稳定、可靠的水产品溯源体系提供理论支撑。更多还原

【Abstract】 To ensure the quality of aquatic products and safeguard consumer’s rights, it is imperative to develop effective techniques to authenticate the geographical origin and identitification of aquatic products. Stable isotope ratio and element analysis techniques are considered as critical tools for tracing the origin of aquatic products. They can reveal environmental information and dietary structure characteristics related to organisms with the advantages of rapid, sensitive and accurate determination. These technologies have been extensively employed in tracing fish, crustaceans, echinoderms, and mollusks. It is noteworthy that the corresponding tracers, including elements and stable isotope ratios, have been widely applied in the traceability of aquatic products. Unique characteristics are observed in the composition and content of elements and stable isotope ratios within aquatic organisms from different habitats, facilitating the identification of origin and cultivation methods. This paper reviewed the recent advances in applications of stable isotope ratio and multi-element analysis techniques combined with multivariate statistical analysis for the traceability of aquatic products. In most studies, the accuracy of tracing aquatic products is only 60% or higher, and there are still some limitations within the current traceability system. First, the distributions of elemental and stable isotope compositions are influenced by various factors, including biological variations, aquaculture environments, and climate conditions. Thus, it is necessary to conduct an in-depth investigation to explore the formation mechanism of element and stable isotope “fingerprints”, understand the living environment and dietary structures of organisms, screen factors affecting the stability of methods, and select the most discriminative factor that is least influenced by other variations. Additionally, increasing the sampling size as well as extending the sampling regions will contribute to establishing a comprehensive and dynamic “fingerprint” database of elements and stable isotopes. Secondly, given the rapid development of the aquatic product industry and the growing diversity of aquatic species, current technologies are unable to meet the convenient and rapid detection requirements subjected to any time and spot. Hence, it is essential to continuously accelerate the improvement and innovative attempts in traceability approaches and explore the potential advantages of element and stable isotope analysis techniques, such as the extended use of compound-specific isotope analysis and the integration of multiple analytical methods(omics technology, molecular biology, etc.). Lastly, multivariate statistical analyses, such as principal component analysis and linear discriminant analysis, have been extensively applied to the traceability of aquatic products. However, modern machine learning algorithms with enormous data mining potential, such as support vector machines and artificial neural networks, remain underexplored. To achieve high discrimination precision, it is recommended to introduce more modern machine learning algorithms and merge multi-model techniques in the traceability of aquatic products. In general, this review provides comprehensive insight regarding geographical origin discrimination and cultivation method identification of aquatic products, serving as a theoretical basis for the establishment of a stable and reliable aquatic product traceability system.更多还原