1. Detection And Growth Estimation Of Indo-Pacific Eel (Anguilla marmorata Quoy & Gaimard, 1824) Using Machine Learning In Central Vietnam.
- Author
-
Kieu Thi Huyen, Ha Nam Thang, and Nguyen Quang Linh
- Subjects
ANGUILLA anguilla ,EELS ,ECOLOGICAL regions ,SPECIES distribution ,ECOLOGICAL models - Abstract
Context. The Indo-Pacific eel (Anguilla marmorata) is a widely distributed and commercially valuable species across ecological regions worldwide. Overfishing and habitat loss are leaving the Indo-Pacific eel in a risky situation and raising a high demand for conservation. Previous research has found relationships between the Indo-Pacific eel's migration patterns and environmental factors. However, there is still a need to advance the discovery of its spatial distribution by using diverse environmental and ecological datasets and modelling its growth in terms of different environmental characterizations. Aims & Methods. Here, we compared machine learning (ML) CatBoost (CB) and the multivariate linear model to investigate the relationship between spatial distribution, Indo-Pacific eel development stages, and environmental factors in central Vietnam. Key results. Our results show that CB detected the Indo-Pacific eel at high accuracy (Overall Accuracy (OA) = 0.9, F1 = 0.88, AUC = 0.97) and estimated the total length at different confidence levels (R2 ranging from 0.51 to 0.70), demonstrating superior performance to the multivariate linear model. Conclusions & implications. This study highlights the potential use of ML models in species distribution mapping and modelling growth patterns to support conservation efforts of Indo-Pacific eels in their natural habitats. [ABSTRACT FROM AUTHOR]
- Published
- 2024