A size-age model based on bootstrapping and Bayesian approaches to assess population dynamics of Anguilla anguilla L. in semi-closed lagoons

• Published in: Ecology of Freshwater Fish
• Main Authors: Aschonitis, Vassilis G., Merighi, Matteo, Gelli, Fernando, Rossi, Remigio, CASTALDELLI, Giuseppe, LANZONI, Mattia, GIARI, Luisa, FANO, Elisa Anna
• Other Authors: Castaldelli, Giuseppe, Lanzoni, Mattia, Giari, Luisa, Fano, Elisa Anna
• Format: Article in Journal/Newspaper
• Language: English
• Subjects: Anguilla anguilla; Bayesian probability; Bootstrap regression; Eel management; Stock assessment; Aquatic Science; Ecology; Evolution; Behavior and Systematic
• Online Access: http://hdl.handle.net/11392/2338314; https://doi.org/10.1111/eff.12269; http://www3.interscience.wiley.com/journal/117978938/toc

A size-age modelling technique is presented for assessing the vital rates, stock and recruitment of eel populations in semi-closed lagoons with fully monitored migration of silver eels. Data for yellow and silver European eels (Anguilla anguilla L.) were obtained in 2011 from the Comacchio lagoon (Italy). The analysis was performed in three steps: (i) correction of yellow eel data, which are affected by the fyke nets selectivity during samplings, (ii) estimation of survival curve, stock, recruitment and metamorphosis rates of the population (calibration using data from 2011) and (iii) validation of the model using the observed amount of silver eel migrating population of the next year. A bootstrap procedure was used to assess the level of uncertainty for each parameter using the 95% intervals of the highest posterior density distribution HPDD (Bayesian approach). The measured abundance of silver eels was 0.56 ind·ha-1, while the yellow eel abundance and recruitment were estimated by the model for 2011 at 8.77 ind·ha-1 and 5.99 ind. ha-1 respectively. The model performance during validation was satisfactory as the observed total mass of migrating population of 2012 (3777 kg) was inside the 95% HPDD intervals (3197-3839 kg) of model's predictions. The estimated stocks and recruitment were at least ten times lower from the respective estimations of previous studies of 1989 highlighting the crucial conditions of the population. The proposed modelling approach can provide significant information about eel population conditions, facilitating the evaluation of a range of management options in the context of eel conservation plans.