Microbes to Megafauna Modelling of Arctic Seas


We will develop two contrasting types of mathematical models of marine food webs to quantify the ‘end-to-end’ (microbes to megafauna) spatial and seasonal patterns of biomass, production and fishery yields in slope and shelf waters of the Atlantic-Arctic.

We will test the models by assessing their ability to explain contemporary patterns, and then use them to predict the impacts of multiple stressors associated with a warming climate, in particular changes in temperature, circulation and mixing, sea-ice cover, and freshwater inputs.

Finally, we will use the models to examine trade-off between the realisation of provisioning services (commercial and artisanal harvesting of fish, invertebrates and mammals), and cultural values (reputation, tourism) arising from the abundances of marine megafauna in the pristine Arctic environment.

Scientific aims

The project will first test, in a modelling context, the hypotheses that the environmental changes associated with a warming climate will result in a bottom-up trophic cascade from a) increased primary production, to b) increased zooplankton production, to c) increased fish and benthos production and harvesting potential, through to d) increased populations of charismatic marine megafauna. Secondly, the project will determine when and where the peak concentrations of algae, zooplankton and fish as feeding hot-spots for higher trophic levels are likely to occur in the future under warmer ice-free conditions. Finally we aim to maximise the reach of our findings by delivering them objectively and in plain language through a series of audio and video productions.

Scientific deliverables

We will deliver quantitative simulations of the end-to-end (microbes to megafauna) top-down and bottom-up cascading trophic effects of multiple stressors in the Barents Sea, Fram Strait, and the wider Atlantic-Arctic, to coincide with several components of the existing CAO field programme.

General approach

Two different types of E2E models will be used to represent the Arctic ecosystem – ECOSMO-E2E and StrathE2E –both of which are based on functional groupings of taxa. The former is a high resolution General Circulation Model – Nutrient Phytoplankton Zooplankton detritus – Higher Trophic Level (GCM-NPZD-HTL) type model, while the latter is a low-resolution exploratory type of model. By deploying the two approaches together we will be able to address model-dependent uncertainty, identify areas of consistency and divergence of results, and address questions at different scales.

Brief summary of the end-to-end models and how they will be used

The StrathE2E marine food web modelling concept is designed to simulate regional scale, macroscopic top-down and bottom-up cascading trophic effects with coarse spatial resolution (horizontally resolved into coastal and offshore compartments with 3 seabed sediment habitats in each compartment; vertically resolved into surface and deep layers in offshore waters). The mathematical formulation is based on a network of coupled ordinary differential equations representing the entire food web from nutrients and microbes though zooplankton and fish, to birds and mammals, including the effects of advection, mixing and active vertical and horizontal migrations. Living components are represented at low taxonomic resolution, focussing on fluxes of nitrogen between coarse functional groups, and simulating the general ‘shape’ of the food web rather than the detail. The scheme takes off-line output from GCMs in the form of volume fluxes and mixing rates between spatial compartments, but is not directly coupled to any GCM. The advantage is very fast run-times, enabling tens of thousands of model runs in a fraction of the time required for a single GCM-NPZ-HTL run. This has enabled the implementation of computational parameter optimisation methods to fit the models to observed data, sensitivity analysis, and computation of likelihoods for model outputs. The focus of existing uses of Strathe2E has been on UK shelf seas and the cascading implications of fisheries and fishing practices such as trawling and its impacts on the seabed, and discarding of unwanted catch. To this end, the StrathE2E model is coupled to a separate spatial model of fishing fleets, which generates harvesting, discarding and seabed disturbance rates for the ecology model based on data about the activities of a range of different fishing fleets.

ECOSMO-E2E is an integrated GCM-NPZD-HTL scheme, developed from an original GCM-NPZD model (ECOSMO) by the addition of functional groups representing benthos and fish. ECOSMO represents the three main nutrient cycles (nitrogen, phosphorus and silica), three functional groups of primary producers (diatoms, flagellates & cyanobacteria) and two zooplankton groups. Additionally oxygen, biogenic opal, detritus, dissolved organic matter and three groups of sediments are considered. As in StrathE2E, the biological reaction terms include consumption and loss, which incorporate natural mortality, predation and excretion, with grazing rates represented by type-II functional responses. Macrobenthos and fish functional groups are included following the basic principles of the ECOSMO formulation. While macrobenthos remain anchored and grow at the bottom of the water column, the formulation of fish requires additional assumptions to account for mobility and searching of the water column for food. ECOSMO-E2E provides spatially detailed information on HTL production potential and identifies areas expected to experience rapid change as a result of trends in physical drivers. This is exactly the issue of interest in the Arctic with respect to marine mammals and fisheries as sea-ice retreats. ECOSMO-E2E additionally gives a better representation of fish predation on zooplankton, which is one of the bottlenecks in most existing GCM-NPZD approaches. ECOSMO-E2E has been implemented for the North Sea and Baltic Sea. ECOSMO has been implemented for the Barents Sea and reproduces the main pattern of primary and secondary production, but lacks the formulations for HTL. Nevertheless, analysis of trophic amplification indicates a strong top-down response of ecosystem effects under climate change scenarios. Hence, there is a sound basis for a HTL coupled version in the Arctic.

Underpinning hydrodynamics for ECOSMO-E2E

The GCM component of ECOSMO E2E is the unstructured grid model SCHISM. SCHISM is a three-dimensional, baroclinic model, that has been tested and validated in a range of marine environments up to a horizontal resolution of less than 100m. Temperature, salinity and baroclinic flow dynamics are represented by a non-linear equation of state, a scale-independent turbulence model as well as tidal wetting and drying. The unstructured model grid enables an optimal resolution of the study area, which is independent of grid convergence due to spherical coordinate grids or refinements in curvilinear grids. In the vertical, SCHISM has a hybrid Localised Sigma Coordinate scheme in the vertical (LSC²), that ranges from sigma coordinates in shallow areas to geopotential coordinates with smooth transition in between for the numerical representation of density currents along the sea floor in stratified regimes. SCHISM is coupled to ECOSMO through the FABM framework, which allows for flexible extensions of the ecosystem model component, as required by our initiative. A present, ongoing development is the coupling of SCHISM to the FESIM sea-ice model, which completes the basic technical environment for unstructured, cross-scale ecosystem simulations of the Arctic seas.

StrathE2E can accept offline hydrodynamic driving data from any competent GCM, in the form of monthly averaged volume fluxes through external and internal model boundaries, and vertical mixing rates. For the existing UK shelf seas applications these have come from the NORWECOM and UK-POLCOMS models. In this project we will take data from SCHISM so that our two ecosystem models will be driven by exactly the same GCM visualizations of the environmental physics. However, StrathE2E will also accept hydrodynamic data from NEMO simulations of the Arctic and outputs from any of the other data-providing GCM models associated with CAO as these become available.

Adaptation of SCHISM-ECOSMO-E2E and StrathE2E to the Arctic ecosystem

Development of SCHISM-ECOSMO-E2E for CAO will be a compilation of the ECOSMO version for the Barents Sea, the ECOSMO-E2E model and the unstructured grid SCHISM GCM. The model domain will be constructed to cover the field-study regions of the CAO Programme. Additions to the food web will need to be Arctic fish and benthos groups, vertical and horizontal migration strategies for the fish following e.g. isolumes, temperature or food gradients, and the addition of sea ice biogeochemistry. Sea ice biology is key for understanding the Arctic ecosystem by accounting for parts of the primary production, and by defining nutrient availability in the pelagic ecosystem below ice and at the ice edge. Hence, we will evaluate existing sea ice biogeochemical models  to formulate sea ice algae as a functional group in ECOSMO-E2E and its coupling to the pelagic ecosystem.

Two new versions of StrathE2E will be configured, one covering the Barents Sea and another for Fram Strait. These are intended to map onto the field study areas of the existing CAO projects. Representing the dynamics of ice-algae as an explicit group of primary producers, nutrient consumers, and food source for zooplankton will also be a required development for StrathE2E. Other key structural developments of the StrathE2E model will be the representation of seasonal vertical migrations, and review of the groupings of species represented by the functional groups of birds and mammals. Representation of seabed disturbance processes and benthos mortality, which are explicitly represented in StrathE2E in connection with fishing activity, will also need to be re-considered in the context of sea-ice cover and grounding of ice.

Lead Investigators

  • Professor Mike Heath

    Co-lead investigator, University of Strathclyde

    My current research interests are the mathematical and statistical modelling of fish populations and fisheries, and the dynamics of ecosystems. I am co-lead investigator of the MiMeMo project, and a co-investigator on the DIAPOD project.

    View full profile
  • Dr Ute Daewel

    Co-lead investigator, Helmholtz-Zentrum Geesthacht Centre for Materials and Coastal Research

    My main research interests are focused on investigating biological-physical interactions effecting different trophic levels of the ecosystem by developing and utilizing relevant coupled model systems. One major purpose of my recent research is to overcome the limitations of lower-trophic-level models for simulating and understanding changes in the marine food web and higher trophic level production by developing the consistently formulated E2E model framework ECOSMO E2EI am a co-lead investigator in the MiMeMo project.

    View full profile

Related Articles

  • UK and Germany combine forces to fund crucial Arctic science

    For the first time, the UK and Germany have joined forces to investigate the impact of climate change on the Arctic Ocean. The UK’s Natural Environment Research Council (NERC) and Germany’s Federal Ministry of Education and Research (BMBF) have jointly invested almost £8 million in 12 new projects to carry… Read more

    03 July 2018