Control mechanisms in a complex marine ecosystem: less microplankton, less forage fish, more predators

Frédéric Diaz*1, Daniela Bănaru*1, Philippe Verley2, Yunne-Jai Shin3

1Aix Marseille Univ., Université de Toulon, CNRS, IRD, MIO UM110, 13288, Marseille, France.

2IRD, UMR 123 AMAP, TA40 PS2, Boulevard de la Lironde, 34398 Montpellier Cedex 5, France.

3IRD, UMR 248 MARBEC, Université de Montpellier, Bat. 24 – CC 093 Place Eugène Bataillon, 34095 Montpellier cedex 5, France.

The purpose of this study, conducted using the OSMOSE-GoL integrated ecosystem model, was to propose hypotheses in order to better understand why, over the past ten years, a more or less regular decline in catches of small pelagic fish (sardines, anchovies) by French artisanal fisheries has been observed in the north-western Mediterranean and more particularly in the Gulf of Lion

The model used in this study is an end-to-end model, which provides a numerical representation of the food web from plankton to certain predatory fish or cephalopod species in a given climate and fisheries context. The species represented are those with an economic and/or ecological interest, in this case about ten species of teleosteens and cephalopods representing about 70% of the catches in the Gulf of Lions. The particularity and originality of the model used lies in its ability to operate a totally bidirectional and dynamic coupling between the high and low trophic levels represented.

The results show that the diet of small pelagic planktonophagic fish (anchovies, sardines, sprat) consists mainly of microplankton (20-200 µm) and mesozooplankton (200 µm - 2 cm) with a maximum predation rate corresponding, respectively, to 20 and 30% of the natural mortality rates of these plankton categories. These predation rates may appear low in absolute terms, but they are sufficient to significantly impact annual and infra-seasonal variations in the abundance of plankton and their teleosteal predators. Commercial fish landings are also affected although, in the model, fishing effort was considered constant. The influence zone of the Rhone plume and the regions concerned by intense vortex hydrodynamic activity (i.e. in the northwestern part of the Gulf) are areas of privileged plankton predation because they are places where plankton is very abundant.

The model highlights both bottom-up (from the bottom of the food web) and top-down (from the top of the food web) controls within the ecosystem, and also allows further analysis of its complex functioning. Thus, the four simulated years show changing winter convections of varying intensity leading to significant variations in plankton size spectra and biomasses in early spring. Our results show that these variations in size structure could potentially lead to a change in the diet of small pelagic fish (bottom-up control). This would result in a decrease in the average sizes of these fish and their fishable biomass, which corroborates the landings curve observed in the field for several years.

If the biomass of plankton and plankton-bearing fish decreases, one would expect a decrease in the predators of these fish as well. However, over the duration of the simulation, there is a significant increase in predator biomass and therefore in top-down control. In fact, plankton species decrease in biomass, leading in parallel to a decrease in their predation on mesozooplankton. This category of plankton can therefore be consumed more by the juveniles of predators such as horse mackerel, mackerel, poutassous or cephalopods, which ultimately stimulates their recruitment. These predators see their biomass increase by 38% in our study.

In conclusion, small pelagic fish would be subject to a double pressure linked to both a change in the size and biomass of their prey, and an increase in the biomass of their main predators. What will become of these predatory fish later on and what will be their impact on ecosystem functioning in the short/medium term?

In the future, the end-to-end model developed in this study will have to be confronted with longer data series that integrate variations in fishing effort, the observed increase in biomass through predator migration (e.g. bluefin tuna), changes in nutrient inputs by rivers (e.g. the Rhone) as well as climate variations (temperature, wind regime, stratification, etc.).



Key words: end-to-end model, two-ways coupling, plankton, fisheries, food web functioning

Note: This manuscript is linked to the following research article: “Implementation of an end-to-end model of the Gulf of Lions ecosystem (NW Mediterranean Sea). I. Parameterization, calibration and evaluation” by Bănaru et al. (2019), Ecological Modelling 401, 1-19,