Modelling retention and dispersion mechanisms of bluefin tuna eggs and larvae in the northwest Mediterranean Sea

Abstract

Knowledge of early life history of most fish species in the Mediterranean Sea is sparse and processes affecting their recruitment are poorly understood. This is particularly true for bluefin tuna, Thunnus thynnus, even though this species is one of the world’s most valued fish species. Here we develop, apply and validate an individually based coupled biological–physical oceanographic model of fish early life history in the Mediterranean Sea. We first validate the general structure of the coupled model with a 12-day Lagrangian drift study of anchovy (Engraulis encrasicolus) larvae in the Catalan Sea. The model reproduced the drift and growth of anchovy larvae as they drifted along the Catalan coast and yielded similar patterns as those observed in the field. We then applied the model to investigate transport and retention processes affecting the spatial distribution of bluefin tuna eggs and larvae during 1999–2003, and we compared modelled distributions with available field data collected in 2001 and 2003. Modelled and field distributions generally coincided and were patchy at mesoscales (10s–100s km); larvae were most abundant in eddies and along frontal zones. We also identified probable locations of spawning bluefin tuna using hydrographic backtracking procedures; these locations were situated in a major salinity frontal zone and coincided with distributions of an electronically tagged bluefin tuna and commercial bluefin tuna fishing vessels. Moreover, we hypothesized that mesoscale processes are responsible for the aggregation and dispersion mechanisms in the area and showed that these processes were significantly correlated to atmospheric forcing processes over the NW Mediterranean Sea. Interannual variations in average summer air temperature can reduce the intensity of ocean mesoscale processes in the Balearic area and thus potentially affect bluefin tuna larvae. These modelling approaches can increase understanding of bluefin tuna recruitment processes and eventually contribute to management of bluefin tuna fisheries.

Introduction

The North Atlantic bluefin tuna (Thunnus thynnus thynnus, Linnaeus, 1758) is a large, highly migratory pelagic predator whose historical range encompasses the shelf and open sea areas of the North Atlantic, the Mediterranean Sea and the Black Sea (Mather et al., 1995, Fromentin and Powers, 2005). The species spawns in spring and summer in the Gulf of Mexico and Mediterranean Sea, after which adults migrate north for feeding (Mather et al., 1995, Cury et al., 1998). Major fisheries on bluefin tuna have existed for centuries but, despite the high economic importance of these fisheries, there are still major gaps in knowledge of bluefin tuna ecology (Fromentin and Powers, 2005).

Some of the largest gaps relate to recruitment and oceanographic processes affecting survival of the early life stages. For example, in the Mediterranean Sea some spawning areas were identified decades ago (e.g. near the Balearic Islands and Sicily; Mather et al., 1995, Fromentin and Powers, 2005), but others have only recently been identified (e.g. near Cyprus; Karakulak et al., 2004), while the presence of spawning in other areas may vary over time and is not well documented (e.g. the Adriatic Sea; Mather et al., 1995, Fromentin and Powers, 2005). In different locations, the survival probability of eggs and larvae is not known and could vary both within and among years possibly as a response to spatial and temporal variability in different environmental conditions.

To improve knowledge of the early life history of bluefin tuna and how it is influenced by environmental conditions, comprehensive field investigations have been conducted since 2001 around the Balearic Islands, northwest Mediterranean Sea (Alemany et al., 2006, Garcia et al., 2006b). The spawning habitat for bluefin tuna is characterized by the presence of numerous mesoscale oceanographic features that appear to aggregate and retain several larval species (Alemany et al., 2006, Garcia et al., 2006b). These frontal structures seem to concentrate bluefin tuna larvae at restricted spatial scales.

Given that temperatures during the larval production period are relatively high (Mather et al., 1995, Garcia et al., 2006b), bioenergetic demands of larvae for food (i.e. zooplankton) will also be high (Houde, 1989, MacKenzie et al., 1990). However, the Mediterranean Sea is characterized by relatively low levels of plankton production (Agostini and Bakun, 2002, Sabates et al., 2007a); hence, larvae have a bioenergetic requirement for high concentrations of prey in an environment that paradoxically, is relatively unproductive. As a consequence, localized areas with higher prey concentrations and physical processes which can aggregate and retain larvae together with their prey or advect larvae to relatively prey-rich areas, could increase survival probability (Sabates et al., 2007a). In the Mediterranean, mesoscale dynamics (Velez-Belchi and Tintore, 2001) may favor locally elevated plankton production rates (Agostini and Bakun, 2002), or produce “predator refuges” (Bakun, 2006). These phenomena could promote or suppress survival of eggs and larvae produced at different times throughout the year and in different areas.

In this study, we developed and applied an individual-based coupled biological–physical oceanographic model for bluefin tuna larvae around the Balearic Islands (Fig. 1A). We used this model to investigate the spatial, seasonal and interannual variability of egg and larval distributions in this area, and to determine the influence of regional climatic and hydrographic variations on their distributions. Numerical simulations were performed under simple biological constraints in order to identify bluefin tuna spawning grounds and the emerging patterns of aggregation, retention and dispersion of bluefin tuna larvae. We focused in particular on how mesoscale hydrographic activity affects the seasonal and year-to-year variability of spawning and subsequent larval distributions around the Balearic Islands.

However, before applying the model, we tested its numerical structure and performances with a detailed Lagrangian dataset on drift and growth of anchovy larvae collected in the summer of 2000 in the northwest Mediterranean Sea (Sabates et al., 2007b).