A Low‐Impact ANN Model for Age and Maturity Estimation in Oblada melanura : Toward Sustainable Fish Assessment in the Mediterranean

This study presents a data‐driven, nonlethal approach for estimating age and maturity in Oblada melanura , a commercially important member of the family Sparidae inhabiting the Mediterranean Sea. A multilayer perceptron artificial neural network (MLP ANN) with architecture (1, 10, 2) was developed using 5112 records from published literature, institutional reports, and global databases. The model uses only total length as input, eliminating destructive sampling such as otolith extraction or gonad dissection. Trained and validated in MATLAB R2024a, it achieved very high predictive accuracy, with correlation coefficients exceeding 0.999 across training, validation, and test sets, and mean squared error below 0.009. This performance likely reflects the strong length–age relationship in O. melanura during its primary growth phase (2–6 years) in the Eastern Mediterranean. For maturity classification, accuracy reached 92.3% (precision = 0.91, recall = 0.93, F 1‐score = 0.92), based on histologically confirmed data from Syrian coastal populations. Results indicate that age and maturity can be predicted accurately under current sampling conditions, supporting ethical and sustainable monitoring. The method reduces time, cost, and technical complexity compared to traditional or deep learning–based approaches. When benchmarked against a linear regression model on the same test dataset ( n = 767), the ANN showed markedly superior performance ( R = 0.9995 vs. 0.9796; MSE = 0.0068 vs. 0.4041), confirming its ability to capture nonlinear biological patterns. Although the full dataset is not publicly available due to collaborative agreements, the methodology is fully described. The trained MLP model (1‐10‐2) is available from the corresponding author for noncommercial academic use under a standard data use agreement, with plans to release a standalone version via a public repository. Originally developed for threatened species, this framework demonstrates robust performance in O. melanura , highlighting its potential as a standardized, low‐impact tool for fish population assessment in the Eastern Mediterranean.