Individual-based modelling of small pelagic fish in the Adriatic Sea: Integrating stock assessments, ecophysiology of fish, and environmental forcings

Understanding population dynamics requires linking individual-level physiology to emergent population patterns. This study presents the first application of a dynamic energy budget individual-based model (DEB-IBM) to European pilchard (Sardina pilchardus) and European anchovy (Engraulis encransicolus) populations in the Adriatic Sea. The SPelAgent modelling framework, implemented in the Julia programming language, simulates fish energy acquisition and allocation throughout the life cycle as a function of environmental conditions, while prioritising simplicity in model structure and assumptions. To hindcast past population trajectories in the Adriatic Sea, SpelaAgent is forced with temperature and zooplankton time series from physical and biogeochemical reanalyses, and informed with age-specific fishing and natural mortality from stock assessments. SPelAgent captures energy allocation patterns, reproductive rates and size at puberty, although it overestimates the growth of sardine and older anchovy specimens; this could be improved through region-specific parameterisation. Model outputs qualitatively reproduce trends in catch data and stock-assessment biomass estimates, but their accuracy is limited by the need to control unrealistic population dynamics arising from simplified individual processes. The resulting differences in population structure between SPelAgent and the stock assessment lead to inconsistent outcomes when applying fishing mortality estimated from the stock assessment. This highlights the importance of explicitly representing prey dynamics and modelling fishing mortality within the model context. Overall, the model offers a mechanistic approach to studying small pelagic fish demography, accounting for individual variability and climate-driven effects on population dynamics, while prioritising simplicity in model structure and assumptions. Our results highlight key considerations for the development and application of individual-based bioenergetic models in fisheries ecology, which can serve as a valuable and complementary tool to current methods in fisheries management.