On the predictive potential of spatially-explicit IBMs: Can larvae fish IBMs be used to predict recruitment success, results from a multi-decadal simulation of Atlantic cod (Gadus morhua) in the North Sea

No abstracts are to be cited without prior reference to the author.

Recruitment of Atlantic cod (Gadus morhua) in the North Sea is determined by a chain of processes impacting survival of early life stages, which are difficult to derive from observations only. Here, we utilized a spatially-explicit Individual Based Model for North Sea Atlantic cod linked to a 3d ecosystem model (ECOSMO) to disentangle processes relevant for long-term changes in potential larval survival (PLS). The model was integrated over a 60-year time period (1949-2008) and the results were analysed using common statistical methods. The simulated long-term dynamics in annual PLS exhibited pronounced year-to-year variability and a significant decrease after 1989. Ten environmental parameters were analysed with respect to their relevance for PLS whereof only water temperature and the u-component of the velocity field showed the same significant shift after 1989. A subsequent correlation analysis with observed Atlantic cod recruitment in the North Sea revealed significant changes in correlation during the considered time period indicating general shifts in process interactions relevant for recruitment. Our analysis revealed that the model is able to produce a cod recruitment index with significant predictive capacity during specific time periods and hence.