Combining hydrographical particles-tracking models with spatial analyses to evaluate spatial dynamics of cod larvae and 0-group in the Barents Sea

Recruitment ecology of cod has been an important focus within the framework of GLOBEC. A large part of this work focused on understanding the spatiotemporal component of cod populations. For the early pelagic life stages, studies based on an individual-based platform have provided highly valuable insights combining oceanographic, behavioural and modelling approaches. The spatial modelling of observational data often fails to include densitydependent covariates, which for early pelagic life stages, originate from the combination of circulation patterns and eggs coming from the spawning aggregations. We performed this task combining a hydrographical particle-tracking model with spatial statistical analyses to investigate the relative contribution of hydrographical variables on the spatial distribution of cod larvae in the Barents Sea under two short-term climatic regimes in the period 1986-1991. The cod larvae distribution is modelled using eggs drifting from the spawning aggregations in the Lofoten Islands. We found that inter-annual variability in the spatial aggregations of the spawners influenced the distribution of larvae drifted. We have also shown how the spatial distribution of passive-drifting larvae can change over the two regimes (1986-1988 and 1989- 1991), being more upstream in cold periods. Though the currents pattern is the main hydrographical factor shaping the spatial distribution of larvae, the temperature modifies such distribution by affecting larvae survival. However, our study highlights the geographic extension of the temperature effect changed between warm and cold periods, with clear ecological implications in terms of growth and survival. This approach can be useful for other fish populations to further understand the underlying processes shaping the seascape of early life stages. Keywords: Barents Sea cod, larvae, hydrographical particles-tracking models, spatial analyses.