Artificial intelligence for identifying and conserving aquatic species

In a review published in the KeAi journal Water and Ecology, a team of researchers in China systematically summarized the current applications and challenges of machine learning algorithms in image analysis, acoustic identification, and ecological threat detection in the field of aquatic biology. The main findings are presented as follows.

1. Innovations and breakthroughs of artificial intelligence (AI) in aquatic species identification

Deep learning models such as as convolutional neural network (CNN) and recurrent neural network (RNN) have been extensively applied to the analysis of aquatic species imagery and acoustic data, enhancing identification accuracy and real-time capabilities. For instance, CNNs have achieved unprecedented precision in analyzing fish schooling behavior, automating microalgae identification, and assessing coral reef health. RNNs have also shown exceptional capability in classifying cetacean acoustic signals, enabling real-time and automated tracking of species movements.

2. AI-driven refinement of aquatic species conservation management

The deep integration of remote sensing technologies and AI algorithms has refined and enhanced the intelligence of aquatic species conservation strategies. By leveraging AI for the analysis of remote sensing data, it is now possible to proactively warn of harmful algal blooms, water pollution incidents, and invasive species distribution trends. Additionally, AI-driven automated monitoring systems can continuously track species distribution changes and ecological conditions, supporting the delineation of protected areas and optimal allocation of conservation resources.

3. Challenges and future prospects for AI in aquatic ecosystem conservation

Challenges such as insufficient data availability, varying data quality, and limited generalizability of models still constrain the effective application of AI in aquatic ecosystem conservation. To overcome these barriers, the authors advocate for the establishment of interdisciplinary, regionally inclusive ecological data-sharing platforms to provide high-quality training data. Further integration of AI techniques with environmental DNA (eDNA) and molecular ecology is recommended to enhance the interpretability and generalization capabilities of AI models.

###

Contact the author: Zhenbin Wu, State Key Laboratory of Breeding Biotechnology and Sustainable Aquaculture, Institute of Hydrobiology, Chinese Academy of Sciences, Wuhan 430072, China, wuzb@ihb.ac.cn

The publisher KeAi was established by Elsevier and China Science Publishing & Media Ltd to unfold quality research globally. In 2013, our focus shifted to open access publishing. We now proudly publish more than 200 world-class, open access, English language journals, spanning all scientific disciplines. Many of these are titles we publish in partnership with prestigious societies and academic institutions, such as the National Natural Science Foundation of China (NSFC).