Modeling the spatial and temporal population dynamics of the copepod Centropages typicus in the northwestern Mediterranean Sea during the year 2001 using a 3D ecosystem model

Highlights

• Modeling of the population dynamics of Centropages typicus in the Mediterranean Sea

• Forcing of the population model by low trophic level plankton and physical models

• Distribution patterns of copepods influenced by physical, food and predator fields

• Seasonal dynamics of population are different considering coastal or offshore zones.

• Process of match–mismatch between developmental stages and their prey and predators

Abstract

The spatio-temporal dynamics of a simulated Centropages typicus (Kröyer) population during the year 2001 at the regional scale of the northwestern Mediterranean Sea are addressed using a 3D coupled physical–biogeochemical model. The setup of the coupled biological model comprises a pelagic plankton ecosystem model and a stage-structured population model forced by the 3D velocity and temperature fields provided by an eddy-resolving regional circulation model. The population model for C. typicus (C. t. below) represents demographic processes through five groups of developmental stages, which depend on underlying individual growth and development processes and are forced by both biotic (prey and predator fields) and abiotic (temperature, advection) factors from the coupled physical–biogeochemical model. The objective is to characterize C. t. ontogenic habitats driven by physical and trophic processes.

The annual dynamics are presented for two of the main oceanographic stations in the Gulf of Lions, which are representative of shelf and open sea conditions, while the spatial distributions over the whole area are presented for three dates during the year, in early and late spring and in winter. The simulated spatial patterns of C. t. developmental stages are closely related to mesoscale hydrodynamic features and circulation patterns. The seasonal and spatial distributions on the Gulf of Lions shelf depend on the seasonal interplay between the Rhône river plume, the mesoscale eddies on the shelf and the Northern Current acting as either as a dynamic barrier between the shelf and the open sea or allowing cross-shelf exchanges. In the central gyre of the northwestern Mediterranean Sea, the patchiness of plankton is tightly linked to mesoscale frontal systems, surface eddies and filaments and deep gradients. Due to its flexibility in terms of its diet, C. t. succeeds in maintaining its population in both coastal and offshore areas year round. The simulations suggest that the winter–spring food conditions are more favorable on the shelf for C. t., whereas in late summer and fall, the offshore depth-integrated food biomasses represent a larger resource for C. t., particularly when mesoscale structures and vertical discontinuities increase food patchiness. The development and reproduction of C. t. depend on the prey field within the mesoscale structures that induce a contrasting spatial distribution of successive developmental stages on a given observation date. In late fall and winter, the results of the model suggest the existence of three refuge areas where the population maintains winter generations near the coast and within the Rhone River plume, or offshore within canyons within the shelf break, or in the frontal system related to the Northern Current. The simulated spatial and temporal distributions as well as the life cycle and physiological features of C. t. are discussed in light of recent reviews on the dynamics of C. t. in the northwestern Mediterranean Sea.

Introduction

The distribution of marine zooplankton results from the combined effects of abiotic and biotic environmental factors (e.g., hydrodynamics, temperature, trophic conditions) and the biological processes occurring in each species throughout its life history (e.g., physiology and growth, buoyancy, behavior). Modeling the spatial distributions of zooplankton taking into account features of their ontogenetic development was initiated in the 1980s, such as in the models of Calanus marshallae in the Oregon upwelling area (Wroblewski, 1980, Wroblewski, 1982), and increased in the framework of Global Ocean Ecosystem Dynamics (GLOBEC) program (Carlotti and Poggiale, 2010, Carlotti et al., 2000, Gentleman, 2002).

Zooplankton population models take into account different types of structuring parameters, such as age (Wroblewski, 1980); age within stage, as for Pseudocalanus sp. (Davis, 1984) and Euterpina acutifrons (Carlotti and Sciandra, 1989); weight, as for Calanus finmarchicus (Bryant et al., 1997, Slagstad, 1981); and stage and weight, as for Pseudocalanus sp. (Eisenhauer et al., 2009, Fennel, 2001, Stegert et al., 2007). The level of complexity required depends on the objectives of the study. On one hand, a reduction in model complexity simultaneously allows computation of population models in a 3D numerical environment and avoidance of numerical diffusion within the demographic structure (Gentleman et al., 2008). On the other hand, a mechanistic representation of biological processes related to environmental factors including biological properties such as physiological (Caparroy and Carlotti, 1996, Carlotti and Sciandra, 1989, Carlotti and Wolf, 1998, Fennel and Neumann, 2003, Neumann and Fennel, 2006) and behavioral aspects (Slagstad and Tande, 2007) is also required to correctly represent population dynamics. Consequently, environmental factors differentially modify the biological processes occurring in each developmental stage in a species. For instance, temperature modifies stage-specific physiological rates and, consequently, the rates of growth and development. It also triggers a shift in biological phases from full activity to dormancy and vice versa (Halsband-Lenk et al., 2004). Temperature-dependent rates must be taken into account when copepod population dynamics are simulated at a seasonal scale and in areas characterized by contrasting temperature gradients. Trophic interactions and nutrition utilization represent complex processes that are stage dependent because they depend strongly on the predator–prey size relationship (Hansen et al., 1994). The best survival conditions for different developmental stages may occur in different habitats, and zooplanktonic population maintenance therefore has to be understood in the context of the regional dynamics of physical and ecosystems dynamics, including the role of various mesoscale features.

Zooplankton populations in the northwestern Mediterranean Sea (NWMS) are also quite dependent on hydro- and ecosystem dynamics in different ecoregions (Alcaraz et al., 2007, Boucher et al., 1987, Champalbert, 1996, Gaudy and Champalbert, 1998, Molinero et al., 2008, Sabatès et al., 1989). The local wind regimes and topography lead to cyclonic circulation of Modified Atlantic Waters from the Ligurian to the Balearic Sea (Millot, 1990). The Northern Current (NC) flows westward from the Ligurian Sea to the Balearic Sea, inducing permanent seasonally fluctuating Ligurian and North Balearic fronts (Birol et al., 2010), acting either as a temporary dynamic barrier between the shelf and the open sea or allowing cross-shelf exchanges between the Gulf of Lions and the offshore domain (Qiu et al., 2010). Moreover, the region features coastal upwelling (Albérola and Millot, 2003, Millot, 1979, Pinazo et al., 1996) and anticyclonic eddies in the western part of the Gulf of Lions, depending on the occurrence of Mistral and Tramontane winds (Campbell et al., 2013, Hu et al., 2009). In addition, the Rhone River represents an important source of fresh water, locally modifying the seawater density in the plume extension (Estournel et al., 1997), and it is also an important source of nutrients and organic matter, enhancing biological productivity within the shelf region (e.g., Blanc et al., 1969, Diaz et al., 2008, Tusseau et al., 1998).

To our knowledge, previous 3D ecosystem models applied to the NWMS have represented mesozooplankton based only on a single generic variable (for instance, Crise et al., 1998, Fontana et al., 2009, Lévy et al., 1998). Studying trophic interactions between the copepod population and its prey and predator species at relevant spatial and time scales is possible using a copepod population model coupled to a 3D-resolved ecosystem model. Fennel and Neumann (2003) and Neumann and Fennel (2006) have successfully replaced a bulk zooplankton variable in a biogeochemical model of the Baltic sea with a stage-resolved copepod model.

To model zooplankton distributions in the NWMS, the pelagic plankton ecosystem model (Auger et al., 2011) implemented in the Eco3M platform (Baklouti et al., 2006a, Baklouti et al., 2006b) has been used, coupled with the eddy-resolving regional circulation model Symphonie (Marsaleix et al., 2008), which reproduces the mesoscale features of an area. The copepod population module (Eisenhauer et al., 2009) has been applied to the calanoid Centropages typicus (noted C. t. below), a dominant neritic species in the Mediterranean Sea (Mazzocchi et al., 2007, Siokou-Frangou et al., 2010). Its biology and ecology are rather well documented (see special issue by Carlotti and Harris, 2007) compared to that of other copepods. The general seasonal distribution is briefly summarized here. It usually shows a wide peak of abundance in the spring–summer period at coastal stations monitored in Corsican and Balearic waters, the Ligurian Sea, Gulf of Lions and Catalan Sea (see Mazzocchi et al., 2007). A few spatial studies have been carried out at different times of the year. In the Ligurian Sea, reports are available for surveys conducted in December 1990 (Licandro and Icardi, 2009), without any occurrence of C. t., in April–May 1986 (Pinca and Dallot, 1995), when copepodite and adult stages were distributed over a large area between the French Côte d'Azur and Corsica, and in April 2011 (Molinero et al., 2008), during a short survey across the Ligurian front from Nice. Although it is difficult to compare data from different years and collected using different sampling strategies and methods, these studies show that adult and copepodite stages present spatially heterogeneous distributions, with incomplete overlap between the stages. Molinero et al. (2008) measured patches with dimensions of 4 km for copepodites and of 10 km for adults. Similarly, along a transect across the NC front from Barcelona in May 1989, Saiz et al. (1992) observed the highest densities of C. t. adults concentrated inshore from the density front. In five surveys performed by Kouwenberg (1994) between September 1986 and May 1988 and two surveys conducted in March–April 1998 and in January 1999 by Youssara (2002) and Gaudy et al. (2003), respectively, all in the Gulf of Lions, it was shown that (i) the winter population was maintained in coastal areas of the Gulf of Lions at low densities, except in the Rhone mouth area, and (ii) the spring populations were higher on the shelf than in the near open sea. A few other surveys covering the Gulf of Lions and the Algero-Provençal Basin (Furnestin, 1960, Plounevez and Champalbert, 2000) showed that C. t. has the highest densities in summer on the shelf and in the open sea and the lowest densities in the Northern Current. In late summer, the highest densities were observed in the open sea. On the shelf, C. t. is usually replaced by Temora stylifera in late summer and fall. Concerning its biological characteristics, C. t. exhibits a generation time of two weeks under optimal food and temperature conditions, which can extend for more than one month at 13 °C. At the end of fall and winter, when the population shows its lowest densities, it is maintained with a few individuals in the late copepodite and adult stages surviving with a flexible diet (Calbet et al., 2007, Carlotti et al., 2007). C. t. neither displays diapausing eggs, nor diapausing copepodite stages.

The objectives of the present work are to provide a 3D perspective of C. t. population dynamics in response to hydrographic events and the associated trophic supplies from spring to fall and to identify the potential habitats of overwintering late copepodite stages and adults. Our main goal is to understand the mechanisms guiding the seasonal distribution patterns of the developmental stages using realistic patterns of circulation and trophic conditions. The annual dynamics are presented for two main oceanographic stations that are representative of shelf and open sea conditions, while the spatial distributions over the whole area are presented for a few key dates during the year.