AUTHOR=Xu Kaida , Wang Jiahao , Zhou Yongdong , Zhu Kai , Fang Guangjie , Wang Haoxue , Zeng Jiayin 

TITLE=A group identification method based on trace element levels of muscle tissue: a case study for Oplegnathus fasciatus in the Northern Zhejiang Waters

JOURNAL=Frontiers in Marine Science

VOLUME=Volume 12 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/marine-science/articles/10.3389/fmars.2025.1627012

DOI=10.3389/fmars.2025.1627012

ISSN=2296-7745

ABSTRACT=Group discrimination is a critical prerequisite for studying fish population dynamics, implementing effective fisheries management, and evaluating the outcomes of stock enhancement programs. However, research on group differentiation of Oplegnathus fasciatus (Kroyer, 1845), an economically important species in the East China Sea, remains limited across different geographical regions. We conducted a comparative analysis of trace element concentrations, including iron (Fe), manganese (Mn), copper (Cu), zinc (Zn), chromium (Cr), cadmium (Cd), and mercury (Hg), as well as metalloid arsenic (As), in the muscle tissues of three distinct groups of O. fasciatus: the farmed group, the Yangtze River Estuary group, and the Zhongjieshan group. Trace elements were detected using inductively coupled plasma mass spectrometry. Group classification methods such as cluster analysis, stepwise discriminant analysis, and random forest were used to differentiate the groups based on these eight trace elements. Consequently, significant differences were observed in the concentrations of Fe, Cu, Zn, Cr, As, and Hg among the groups (p< 0.05), while differences in Mn and Cr were not statistically significant (p > 0.05). Discriminant analysis showed that cluster analysis effectively differentiated the farmed group from the Zhongjieshan group. Stepwise discriminant analysis achieved an overall accuracy of 85.1%, while the random forest model demonstrated an accuracy of 92.86%. These findings suggest that trace element concentrations could be applied to effectively distinguish among groups of O. fasciatus, with the random forest model outperforming traditional statistical methods to some extent. Our results are relevant to the activities of stock release and fisheries management.