Are conventional methods sufficient to calculate growth parameters of Pontastacus leptodactylus (Eschscholtz, 1823)? A case study of artificial intelligence from Keban Dam Lake

Abstract

In this study, the length–weight relationships of Pontastacus leptodactylus, a freshwater crayfish species found in the Keban Dam Lake, were assessed using both conventional methods and artificial intelligence techniques. Throughout the research process, all biometric measurements of the crayfish were meticulously recorded, including TL, TW, and other biometric data. These measurements were analyzed using both the conventional length–weight relationship method and artificial neural networks. The results obtained using artificial neural networks and conventional methods were compared, and the analysis was based on MAPE and R2 performance criteria. The study showed that the ANNs method outperformed the conventional LWR method, showing more accurate results. The models employed to predict the length–weight relationships of the crayfish demonstrated high accuracy, and the Artificial Neural Networks method was identified as the most effective model. These results provide strong evidence that the ANNs method performs significantly better in predicting the LWRs of freshwater crayfish.

Keywords

• pontastacus leptodactylus
• freshwater lobster
• growth
• keban dam lake