Citronella oil as a natural anesthetic alternative to 2-Phenoxyethanol in Danube sturgeon (Acipenser gueldenstaedtii): Artificial neural network (ANN) optimization and multilevel physiological assessment

Abstract

This study aimed to optimize the induction and recovery performance of anesthesia in Danube sturgeon (Acipenser gueldenstaedtii) using citronella oil and 2-phenoxyethanol (2-PE), and to evaluate the hematological, biochemical, and histological responses that may occur after anesthesia. Concentrations of 300, 600, and 900 μL L−1 for citronella oil and 750, 1000, and 1500 μL L−1 for 2-PE were tested. An artificial neural network (ANN) based model was developed using the induction and recovery times. According to the model results, the most suitable candidate concentrations in terms of induction and recovery times were determined to be 550 μL L−1 for citronella oil and 1200 μL L−1 for 2-PE. In the hematological evaluation performed after anesthesia, all blood parameters were found to be significantly higher in the citronella group compared to the control group. In the 2-PE group, only platelet (PLT) levels showed a significant increase. In time-series biochemical analyses (0, 3, 6, and 9 h), it was determined that ALT levels were significantly affected by the anesthetic agent and time. In contrast, more limited changes were observed in AST, ALP, and total protein parameters. Histopathological examinations revealed no significant pathological findings in the gill tissue in either anesthetic application. However, the melanomacrophage centers observed in the liver tissue of the 2-PE group suggested a possible agent-related response at the hepatic level. In conclusion, the ANN approach enabled the determination of candidate optimum concentrations for both anesthetics. When hematological, biochemical, and histological findings are evaluated together, citronella oil appears to be a promising natural anesthetic for Danube sturgeon. It may also be a potential alternative to 2-phenoxyethanol. However, this finding needs to be supported by further validation studies.