Email this article (Login required)
Email the author (Login required) Using a modified growth model for the spotted rose snapper (Lutjanus guttatus) in a floating net cage system in northwestern Mexico
Submited: 2025-08-14 17:03:25 | Published: 2026-08-18 19:56:16
DOI: https://doi.org/10.3856/vol54-issue4-fulltext-3516
Abstract
Having the most appropriate model to describe farmed fish growth is a tool for improving responsible aquaculture production. Above 39,000 Lutjanus guttatus with an initial weight of 14 g were reared in three floating net cages at three harvest densities (15, 20, and 22 kg m-3) for 360 days of commercial culture. We compared Schnute's growth model with a new growth model for shrimps proposed by Ruiz-Velazco. The first model generates eight curves between sigmoidal, acceleration, and deceleration, resembling the von Bertalanffy, Gompertz, and logistic models. The second model is a modification used to estimate the weight of organisms under culture conditions. We selected the best model using information theory based on Akaike information criterion (AIC). The model with the lowest AIC at density D15 was Schnute case 1. At density D20, the Ruiz-Velazco model received the strongest support, with a probability of 98.2%. For the highest density (D22), support for the Schnute case 2 model declined to 63.3% (Table 2). The fitting results indicated that the best-fitting model varied with density, suggesting our results lacked consistency across the three datasets and prevented generalization to a single best model. These findings also suggest that density influences growth trajectories by altering underlying growth dynamics, implying that growth does not follow a universal pattern across densities. Therefore, researchers should test multiple models and use criteria such as AIC to select the most appropriate model when analyzing species growth under different conditions. Researchers should avoid assuming a default model for analyzing fish growth; generalization may be inappropriate, and more suitable models may be available in specific contexts. Adopting a multi-model inference approach is essential for effectively analyzing growth in aquaculture systems.
