Programme Outline - Marine Ecosystems Research Programme

Document technical information

Format pdf
Size 1.3 MB
First found May 22, 2018

Document content analysis

Category Also themed
Language
English
Type
not defined
Concepts
no text concepts found

Persons

Grand Duke Konstantin Konstantinovich of Russia
Grand Duke Konstantin Konstantinovich of Russia

wikipedia, lookup

Organizations

Places

Transcript

Programme Outline
Revised from original Case for Support, September 2014
Marine Ecosystems Research Programme is jointly funded by the Natural Environment Research
Council (NERC) and the Department for Environment, Food and Rural Affairs (Defra)
Contents
MERP: PART A: Previous Track Record ........................................................................................................ 2
PART B. DESCRIPTION OF PROPOSED RESEARCH
1. Summary ............................................................................................................................................... 7
2. Introduction and Rationale .................................................................................................................... 7
3. Project overview and relationship to the Programme objectives ............................................................. 9
3.1 Project overview ........................................................................................................................................... 9
3.2 Relation to Programme Objectives ............................................................................................................. 11
3.3 Geographical focus ..................................................................................................................................... 12
4. Policy relevance and impact goals ......................................................................................................... 13
5. Collaborations with NERC, LWEC and other research programmes ......................................................... 13
6. Work module descriptions .................................................................................................................... 14
6.1 Module 1: Marine ecosystem data toolbox & application of macroecology ............................................. 14
6.1.1 Module Tasks ....................................................................................................................................... 14
6.1.2 Milestones ........................................................................................................................................... 16
6.1.3 Deliverables ......................................................................................................................................... 16
6.2 Module 2: Fieldwork to measure poorly known processes ........................................................................ 17
6.2.1 Tasks .................................................................................................................................................... 19
6.2.2 Milestones ........................................................................................................................................... 21
6.2.3 Deliverables ......................................................................................................................................... 22
6.3 Module 3: Ecological processes and their representation in models ......................................................... 22
6.3.1 Tasks..................................................................................................................................................... 23
6.3.2 Milestones ........................................................................................................................................... 24
6.3.3 Deliverables ......................................................................................................................................... 24
6.4 Module 4: Simulating and predicting ecosystem changes using a model ensemble ................................. 24
6.4.1 Tasks..................................................................................................................................................... 24
6.4.2 Milestones ........................................................................................................................................... 27
6.4.3 Deliverables ......................................................................................................................................... 27
6.5 Module 5: Linking macroecology and models to ecosystem services. ....................................................... 27
6.5.2 Milestones ........................................................................................................................................... 28
6.6. Module 6.................................................................................................................................................... 29
6.6.1 Tasks .................................................................................................................................................... 29
6.5.2 Milestones ........................................................................................................................................... 31
7. REFERENCES ......................................................................................................................................... 32
Appendix 1 Original Case for Support for WP2 Developing a model based understanding of ecosystem
service regulation .................................................................................................................................... 34
3
Marine Ecosystem Research Programme
MERP: PART A: Previous Track Record
Partners
Plymouth Marine Laboratory (PML) conducts interdisciplinary marine science to increase understanding of
the interactions between the marine environment and society in estuarine, coastal and shelf waters, as
well as the upper layers of the global ocean. PML is a NERC National Capability delivery partner, responsible
for long-term datasets and developing ecosystem models. Capabilities include data collection and analysis,
field and laboratory experiments, environmental impact assessments, satellite observation, ecosystem
modelling, method and tool development, socio-economic evaluation, programme co-ordination, policy
advice and capacity building.
Paul Somerfield is a Principal Investigator with >25 years’ experience in studying interactions between marine
organisms and their environment across scales of space, time and biological organisation using a range of
approaches.
Professor J. Icarus Allen (NERC band 4) is Head of Science for Modelling and Observing Systems at PML, leader
of PML’s ecosystem modelling NC and is an honorary visiting professor at the University of Exeter. He has
been involved in and acted as PI for over 35 national and EC projects, including coordinating the FP7 MEECE
and OpEc integrated projects. He is the PI for the modelling activities in the Shelf Seas Biogeochemistry
program, the Ocean Carbon theme of the National Centre for Ocean Forecasting (NCEO), Integrated Marine
Biogeochemical Modelling Network (i-MarNet). His primary expertise is in marine ecosystem modelling and
ecosystem response to climate change. He has 78 ISI publications.
Angus Atkinson has >25 years of experience in studying distribution and rate processes in zooplankton
and micronekton, particularly the diversity of feeding and trophic level interactions. Steve Widdicombe has
>20 years’ experience in using field observations and manipulative experiments to study the impact of benthic
organism behaviour and diversity on biogeochemical processes and ecosystem function focused on the
impacts warming and acidification.
Melanie Austen leads a broad spectrum of interdisciplinary research with >25 years’ experience originally
in marine ecology and for >10 years in ecosystem services interfacing between natural and social science
research.
Nicola Beaumont is a marine environmental economist with >14 years’ experience in the assessment
and valuation of ecosystem services in the marine environment. The PML Communications Group is
experienced in running large KE offices including that of NERC’s UK OA Programme. They have also had
significant input into other KE offices such as VECTORS and EPOCA and have produced a number of
successful science communication media, e.g the Other CO2 Problem animation, (20,000 hits on You Tube,
won the Bill Bryson Prize for Science Communication).
Jerry Blackford (NERC band 4) is a Merit scientist at PML. His principle interests are the impact of Climate
Change and Ocean Acidification on marine systems (OA) and environmental impact assessments for Carbon
Capture and Storage (CCS). He leads a NERC consortium on CCS impacts and a NERC-Defra funded project on
regional climate /OA modelling.
Jorn Bruggeman in trait based modelling and Dr Sevrine Sailley expertise on zooplankton and trophic
interactions, Dr Yuri Artioli provides expertise in modelling marine plankton diversity. Dr Luca Polimene, has
expertise in pelagic process modelling and Dr Nick Stephens in benthic modelling and nutrient cycling.
Momme Butenschön has >10 years of experience in ecosystem modelling leading the development of the
marine ecosystem model ERSEM. He with a particular focus on ecosystem trends, natural variability and
climate change impacts.
Ana Queirós is a benthic ecologist with expertise in studying complex animal- environment relationships
3
Marine Ecosystem Research Programme
focusing on benthic responses to acute and chronic anthropogenic disturbance and climate related impacts.
Pennie Lindeque is a molecular ecologist with >18 years research experience in a wide range of molecular
techniques; most relevant to this project is studying feeding interactions and the use of Next Generation
Sequencing.
The School of Biological Sciences at Queen’s University Belfast undertakes environmental research
spanning fundamental through to applied, empirical through to theoretical work. Strengths include a focus on
the drivers of biodiversity change, conservation biology and food web ecology, and ecosystem service
delivery at a hierarchy of scales.
Mark Emmerson’s work focuses on relationships between biodiversity and ecosystem functioning
using marine, terrestrial and freshwater systems to explore the consequences of biodiversity change using
food webs. He has published >40 papers on biodiversity and ecosystem functioning and food web
dynamics.
Nessa O’Connor is an early career researcher and lecturer. Her work focuses on the loss of primary
consumers and predators in coastal systems and, more recently, the effects of multi-trophic species
loss, and how they interact with perturbations on marine intertidal assemblages. She has published 17
papers and a book chapter.
The Department of Animal & Plant Sciences of the University of Sheffield is one of the UK’s leading
departments for whole-organism biology, . It has a thriving research programme which includes
macroecology, population and community modelling and ecology, and ecosystem services, with
expertise in terrestrial and marine ecosystems and in both theoretical and empirical approaches. The School
of Maths and Statistics has an international reputation for its methodological research in Bayesian statistics
and the handling of uncertainty in complex models, and applied research in ecology and environmental
science, among other areas.
Tom Webb, a Royal Society University Research Fellow since 2008, is a macroecologist with particular
expertise in cross-domain (marine—terrestrial) approaches to ecology and use of global biodiversity
databases.
Julia Blanchard is a Lecturer in Ecology and quantitative marine ecologist with expertise in macroecological
and food web models and indicators for understanding and predicting responses of marine ecosystems to
human and environmental pressures.
Paul Blackwell is a Bayesian statistician and modeller with expertise in stochastic processes, spatiotemporal modelling and computer-intensive methodology, with particular interest in applications in ecology
and environmental science.
Centre for Environment, Fisheries & Aquaculture Science (Cefas)’ marine ecosystem and biodiversityrelated science focuses on assessing, modelling and predicting the effects of fisheries or environmental
change on ecosystems and biodiversity. Expertise is in interpreting survey results, defining the relevance to
legislation and determining the implications for marine planning and management. Our applied science has
been influential nationally and internationally in providing scientific evidence to support decisions on
marine biodiversity conservation, marine spatial planning and implementation of the MSFD. We collaborate
internationally with leading universities and research institutes, and a full range of other stakeholders. We
direct and influence world-class research to ensure scientific advice is underpinned by the best available
evidence: eg, partners in many EU FP7 projects. Resources include RV Cefas Endeavour, two specialist
laboratories, comprehensive IT facilities, controlled environment aquaria and fish-rearing facilities. Cefas
bring to IMMERSE unique data archives, and specialist systems for interrogating and modelling.
3
Marine Ecosystem Research Programme
Michaela Schratzberger: 14 years’ post-doc; 35 peer-reviewed publications (PRP). Senior Benthic Ecologist,
specialising in changes in benthic community structure in response to environmental change; Science
Leader providing science leadership/direction within Cefas’ Environment & Ecosystem Division.
Axel Rossberg: 14 years’ post-doc; 53 PRP. Senior Ecosystem Scientist, expert in developing highly
simplified descriptions of complex systems; past10 years specialising in food webs and marine ecology.
Steven Mackinson: 13 years’ post-doc;52 PRP. Senior Fisheries Scientist, specialising in developing ecosystem
models and applying acoustic methods to study behavioural ecology.
Jeroen van der Kooij: 15 PRP. Pelagic Fisheries Ecologist with 11 years’ experience in modelling temporal and
spatial dynamics of populations; application of acoustic technologies to fisheries assessment.
John Pinnegar: 13 years’ post-doc;50 PRP. Programme Director for Marine Climate Change, providing
leadership and strategic planning of CC research, stable isotope analysis, multispecies and food web
modelling.
Dr Johan van der Molen (Cefas Pay Band 7) is a principal ecosystem modeller. He is Secretary for Coastal and
Shelf Seas in the Oceans Division of the European Geosciences Union, and a convener in its annual assemblies.
He has over 20 years of experience in numerical modelling of hydrodynamics, sediment transport,
morphodynamics, transport of fish eggs and larvae, and biogeochemistry.
Dr Sonja van Leeuwen (Cefas pay band 6) is a senior ecosystem modeller with experience in primary
production, eutrophication, coupling ERSEM to size-based higher trophic level models, and river runoff and
nutrient loads.
Dr John Aldridge (Cefas pay band 6) is a senior mathematical modeller with broad experience in physical and
biological marine systems, and specialising in benthic processes and phytoplankton and macroalgal
productivity. Other Cefas staff may contribute to the project where appropriate.
The School of Ocean Sciences of Bangor University conducts interdisciplinary research to provide the
science required to underpin the conservation and sustainability of natural resources. Bangor University has
internationally recognised expertise in interpretation and quantification of threats to ecosystems.
Jan Geert Hiddink is a benthic ecologist with >10 years’ experience in studying the effects of fishing gears on
the state and functioning of seabed ecosystems. He has published 46 peer-reviewed papers and grant capture
since 2006 exceeds £3.5M.
Peter G.H. Evans is a marine vertebrate ecologist with >35 years’ experience in studies of marine mammals
and birds; he has c. 200 peer-reviewed publications and is author or editor of 12 books. He is Scientific
Director of the marine environmental research organisation, Sea Watch Foundation and Honorary Senior
Lecturer.
The Centre for Ecology & Hydrology (CEH) is the UK’s Centre of Excellence for integrated research in
terrestrial and freshwater ecosystems and their interaction with the atmosphere. The organisation provides
National Capability based on innovative and interdisciplinary science and long-term monitoring. CEH has an
international reputation for long-term population and dietary studies of seabirds in UK shelf seas. The CEH
staff involved in IMMERSE all have extensive experience of multi-disciplinary research and will contribute indepth knowledge of avian predators, modelling and empirical skills.
Sarah Wanless has > 30 years research experience of climate and fishery effects on birds. She has published >
200 peer-reviewed papers (PRP) including several on bottom-up and top-down processes as causes of seabird
breeding failures.
Francis Daunt is a population ecologist with 15 years’ experience of research into drivers of seabird
3
Marine Ecosystem Research Programme
populations, in particular climate change, marine renewables and fisheries. He has published > 60 PRP and
coordinates CEH’s long term seabird study on the Isle of May.
Kate Searle is an ecological modeller specialising in the estimation and modelling of population processes
and dynamics, particularly the role played by extrinsic and intrinsic processes and spatial and temporal
heterogeneity.
The Institute of Biodiversity, Animal Health and Comparative Medicine of the University of Glasgow is a
multi-disciplinary research organisation carrying out a wide range of pure and applied work at levels
from viruses to natural ecosystems. One research focus of the institute comprises the effects of
environmental change operating at all levels of biological organisationlinked to studies of how individuals
cope with environmental fluctuations, and how in turn this influences population dynamics and species
interactions (including those between predator and prey). The contribution to the project will be investigating
seabird diets. The institute has a strong group (Nager, Monaghan, Furness) studying seabird ecology and the
analysis of their diets using a range of approaches.
Scottish Association for Marine Science (SAMS) is an independent marine research institute and state-ofthe art facility on the west coast of Scotland.
Mike Burrows is Head of Ecology & Principal Investigator in Ecological Processes.
Sheila Heymans has extensive experience in food web and ecosystem modelling using Ecopath with Ecosim
and ecological network analysis. She has co-authored 35 peer-reviewed publications on food web models
of both marine and fresh water ecosystems.
Sir Alister Hardy Foundation of Ocean Science (SAHFOS) operates the Continuous Plankton Recorder (CPR)
Survey. The CPR survey is recognised as the longest sustained and geographically most extensive marine
biological survey in the world. The dataset comprises a uniquely large record of marine biodiversity. The CPR
survey is of global importance for understanding natural variability and human-induced change and it is used
by scientists, policy makers and environmental managers across the world.
Martin Edwards is a macroecologist with >15 years’ experience studying environmental change on marine
ecosystems. He has written >60 peer reviewed papers and >50 policy related reports and assessments.
Pierre Helaouët is a numerical ecologist specialising in large-scale spatial and temporal statistical models.
Through two very successful rounds of SRIF funding, the Queen Mary University London School of Biological
and Chemical Sciences’ laboratories have been refurbished and re-equipped, with a total of £1.5 million
having been invested in the last 5 years. Facilities include a specialist marine plankton laboratory, sea water
aquaria, controlled temperature rooms, and phytoplankton and zooplankton culture facilities. All of these are
available to support this project.
Andrew Hirst has produced some of the first global accounts of growth, mortality and fecundity in
zooplankton. He has co-authored ~50 publications. Most recently this has included developing new models
and making new measurements to explain the Temperature-Size Rule. Hirst has had grants totaling ~£3.5
million funded and as PI on 4 grants e has a track record in managing NERC grants to their successful
completion .
The marine research group in the Department of Mathematics and Statistics from University of Strathclyde
was established in 2010 as part of the Scottish marine pooling (MASTS) initiative. The group focuses on
mathematical modelling of aquatic populations and ecosystems, hydrodynamic modelling, and statistical
3
Marine Ecosystem Research Programme
analysis of large data sets, with unique capability in fish stock assessment and fisheries modelling,
multi-species food web modelling and ecosystem modelling. The group has produced two modelling products
which are available as packages for the R statistical environment and form the basis for a range of UKRC and
EU funded projects. The team is led by Mike Heath (Prof) and Douglas C. Speirs (Lecturer), and comprises1
technical support staff, 5 post doctoral researchers including 1 senior researcher, and 8 post graduate
students supported by a range of internal, CASE and MASTS funding sources. Heath and Speirs have
jointly generated £1.255million of research income since 2008.
The National Oceanography Centre (NOC) is a NERC Research centre that maintains world class
oceanographic research, consisting of Liverpool (NOC-L formerly the Proudman Oceanographic Laboratory;
POL) and Southampton (NOC-S) sites. This work sits within the Marine Systems Modelling (MSM) group, which
has world-class expertise in high resolution ocean and Shelf Sea modelling of hydrodynamics, waves and
biogeochemistry, and uncertainty. At MSM at NOC-L has a substantial international reputation and
contributing to the development of operational shelf sea forecast services at the UKMO (based on NEMO) and
marine impact climate projections for UKCP09 and MCCIP.
NOC Principal investigator: Dr Jason Holt (NERC band 4) is associate head of the MSM. He specialises in the
synthesis of model and observations to develop our understanding of shelf-sea physical and coupled physicalbiological systems. He leads the modelling component in the NERC FASTNet (ocean-shelf exchange)
programme and the Next Generation Ocean Dynamical Roadmap Project. He has published 50+ peer reviewed
papers primarily on the modelling of hydrodynamics and ecosystems in shelf seas and ocean margins. Co-
Investigators: Dr. Hedong Liu (NERC Band 5) works on the development and applications of 3-D unstructured
and structured grid ocean models. He is currently a member of NEMO developer's committee and specialises
in state-of-the-art numerical methods.
Dr Sarah Wakelin (NERC Band 6) specialises in the modelling of physical and coupled physics-ecosystem
processes in shelf seas, including the impacts of climate change and the combine impacts of climate and
anthropogenic drivers.
3
Marine Ecosystem Research Programme
Marine Ecosystem Research Programme (MERP)
Formerly Integrating Macroecology and Modelling to Elucidate Regulation of Services from Ecosystems
(IMMERSE) and the WP2 Developing a model based understanding of ecosystem service regulation grants.
MERP was created when two grants were combined to make an overarching programme.
Full original WP2 Case for support is available as appendix 1.
PART B. DESCRIPTION OF PROPOSED RESEARCH
1. Summary
IMMERSE is a highly integrated project designed to develop new understanding of the processes governing the
dynamics of marine ecosystems, and how changes in them affect delivery of ecosystem services, from a whole
ecosystem perspective. It will make step-changes in:
•
marine macroecology, through applying the latest ecological theory coupled to novel integration of
existing data using ecoinformatic approaches
•
targeted field sampling and experimental studies of key features that are currently understudied
•
ecosystem modelling through state of the art application of an ensemble of ecosystem models
•
ecosystem services science through use of macroecology and models to hindcast and forecast
ecosystem states, indicators, and estimates of goods and services.
2. Introduction and Rationale
Marine ecosystems provide a range of services to humanity. These are highly dependent on biodiversity and
ecological functioning (Austen et al 2011). Most widely recognised are the provisioning services, particularly
wild-capture fisheries. Others range from those supporting human life (e.g. climate regulation) to those
enhancing the lives of individuals (e.g. leisure, recreation, tourism). Marine ecosystems are experiencing ongoing environmental change (e.g. Halpern et al 2008), ecosystem restructuring generated by fisheries (e.g.
Smith et al 2011), eutrophication, pollution and other environmental degradation (e.g. Diaz & Rosenberg
2008) climate-driven changes (e.g. Beaugrand et al 2009, Dulvy et al, 2008, Hoegh-Guldberg & Bruno 2010)
and growing human consumption and pressures (e.g. marine renewables). Understanding the consequences of
these changes, and designing, testing and refining potential management solutions to address them, is
important for the long-term delivery of services from marine ecosystems (Pinsky et al 2009). It is becoming
increasingly clear that human activities and environmental change that affect parts of the ecosystem (e.g.
fisheries management) can have much wider consequences for biodiversity and ecosystem services than
previously thought, due to interactions through food webs.
Marine food webs play a key role in regulating these ecosystem services but there are important gaps in our
knowledge and understanding of the way they might respond to environmental change. Although there is
evidence that marine food webs are affected by both ‘bottom-up’ (e.g. biophysical factors affecting primary
productivity) and ‘top down’ (e.g. top predators modifying the biomass of lower trophic levels) processes (e.g.
Ware & Thompson 2005, Baum & Worm 2011), existing knowledge is much greater for lower trophic levels
and associated biophysical factors. This is reflected in current ecosystem models such as the European
Regional Seas Ecosystem Model (ERSEM) that are limited in their ability to model ecosystem components
larger than small zooplankton. This means it is currently difficult to understand the relative roles of these
processes and hence the extent to which environmental change cascades through marine food webs and
affects ecosystem services. These processes are also scale-dependent. For example large-scale removal of top
predators (through fishing or other activities) can have a range of impacts across scales. However, scaledependence is poorly understood, making it difficult to quantify the large- scale impacts on ecosystem services
Marine Ecosystem Research Programme Part B
of changes at small spatial scales (e.g. marine conservation zones); and vice versa. Ecological patterns of most
interest (e.g. variation in biodiversity, fish production or nutrient cycling) are often generated by processes
operating at different spatial and temporal scales from those at which they are usually observed (Chave 2013).
In particular, policy decisions regarding the conservation of biodiversity and sustainable ecosystem service
provision typically require large-scale thinking (Kerr et al. 2007, Keith et al. 2012). To say anything meaningful
about the potential consequences of future ecosystem change for ecosystem services we need ecosystem
models. While ERSEM is currently limited to modelling lower trophic levels there are a number of models of
higher-trophic levels which can be coupled to ERSEM. Improving our understanding in the areas above will aid
in the development of more realistic marine ecosystem models and holistic modelling systems, which in turn
will provide important tools for exploring the impacts of potential future changes on marine ecosystems, and
testing their response to potential management solutions.
Although there is considerable activity and associated data in the North-East Atlantic region, from a whole
ecosystem perspective it tends to be fragmented, focused on limited descriptors of foodwebs (e.g. a single
tracer of trophic level and not others) or components of the system (e.g. a specific habitat or component of
the biota) or issues (e.g. fishery management or renewables), largely reflecting the interests and
responsibilities of individuals and organisations doing the work. There is widespread recognition among the
research and policy communities that a whole- ecosystem perspective is now required. The step-change
envisaged by the Marine Ecosystem Research Programme (MERP) is to integrate existing data, targeted new
data, and state-of-the-art ecosystem models, all within a common framework built around the latest and most
appropriate ecological theories, and to use these to improve our understanding of the whole marine
ecosystem rather than just parts of it, how it responds to changes in pressures, and the consequences of those
changes in terms of ecosystem services.
Modelling challenges: Macroecology is the study of the relationships between organisms and their
environment at large spatial scales to characterise and explain statistical patterns of abundance, distribution
and diversity (Brown and Maurer 1989). Scaling relationships are common at higher levels of biological
organization, (e.g. Brown, 1995). Body size is an important determinant of many ecological properties, ranging
from abundance and biomass to physiological and ecological rates and provides useful underpinning concept
for characterizing and modelling marine ecosystems. The widely observed macro-ecological patterns in log
abundance vs. log body mass of organisms can be explained by simple scaling theory based on food (energy)
availability across a spectrum of body sizes. For example quarter-power scaling with body mass applies to
virtually all organisms (West, Brown & Enquist, 1999) and for marine animals, metabolic rate and production
scales at three-quarters power (e.g. Brey, 1990; Warwick & Price, 1979). Allometric equations have been
derived for respiration, ingestion, excretion and photosynthesis for a variety of organisms of different sizes
(e.g. Verdy et al 2009, Moloney and Field 1989). Finally, the temperature dependence of gross primary
production and community respiration is consistent among the major types of ecosystems on the planet,
suggesting that the fundamental biochemical kinetics of respiratory metabolism are highly conserved and can
be scaled from organism to ecosystem levels (Yvon Durcher 2012).
Understanding the interactions of the various system controls, the consequences of change, and designing,
testing and refining potential management solutions, is important for the long-term delivery of services from
marine ecosystems. Dynamic models that link the physical, chemical and biological processes through food
web interactions provide a means of understanding how human impacts on different parts of the ecosystem
interact and of predicting the consequences of management actions.The traditional approach to modelling
marine plankton has generally been to build modelling frameworks by coupling bulk biomass functional type
(FT) models to 1D and 3D hydrodynamic models; The European Regional Seas Ecosystem Model (ERSEM) is
one such model. The ecosystem can be divided into four sections, the pelagic cellular ecosystem, the pelagic
Marine Ecosystem Research Programme Part B
mid trophic levels (zooplankton), the benthic ecosystem (bacteria, meiofauna, zoobenthos) and the higher
trophic levels (fish, birds, mammals); ERSEM describes the first three.
The cellular ecosystem is represented as a biogeochemical flux model where the cell is a black box whereby
the fluxes of carbon or nutrients are transferred across the cell walls via first order rate equations. Most often
defined in terms of biogeochemical functions (e.g. diatom, non-diatom, calcifiers, grazers etc.), the functional
groups are treated as passive tracers and are very sensitive to the biophysical environment they are placed in.
The pelagic mid trophic levels are particularly important yet are a weak component of all models. They are
particularly important because they provide both the top closure for the planktonic ecosystem and the link to
the higher trophic levels. This part of the foodweb is poorly represented in biogeochemical flux type models
because the Eularian continuum approach becomes inadequate as behaviour and individual history becomes
important. Grazers do not eat biomass; they eat individuals and the process is the sum of the interactions
between large numbers of organisms. These interactions are potentially dependent on combinations of prey
density (function of the number of potential prey), prey quality (nutritional status), prey type (species) and
behaviour (e.g. vertical migration, foraging strategies and defence mechanisms). The challenge is to represent
both biomass and population dynamics model in size / trait based foodwebs, coupled to biogeochemical
cycles. The benthic system is strongly influenced by the physical sediment structure, environmental conditions
and organic fluxes to the sediments. Benthic size spectra are more complex than their pelagic counterparts
(Schwinghamer 2011); three biomass peaks, are observed corresponding to bacteria, interstitial meiofauna,
and macrofauna, separated by low biomasses. Once again the requirement is to develop a size/trait based
foodweb model from bacteria to top predators. Only by viewing the water column and benthos as a whole
that will enable a more complete simulation and understanding of fish predating on benthic invertebrates,
macroalgae contributions and benthic ecosystem resilience through the modelling of larval and reproductive
abilities. Finally the Higher trophic levels are not a part of ERSEM, however the new model will be designed to
provide data products to facilitate 1 way coupling to a range of different models ranging from dynamic size
spectra and Ecopath with Ecosim to individual based models to further investigate the different model
ecosystem structures and the ecosystem dynamics represented.
3. Project overview and relationship to the Programme objectives
3.1 Project overview
MERP is structured around six work modules that will enable novel integrated marine ecosystem science to
propagate through to understanding of the dynamics of ecosystem services provided by marine ecosystems,
leading in turn to a mechanism for providing advice on the likelihood of changes in ecosystem services in
response to future changes in drivers.
A key characteristic of the project is integration. Findings and data from modules 1 and 2 will be used to
improve models in modules 3 and 4. Outputs from modules 3 and 4 will be fed back into modules 1 and 2 to
generate new hypotheses and to validate models, and outputs from modules 1, 2, 3 and 4 will feed into
module 5. To facilitate integration regular integration workshops will be held.
Marine Ecosystem Research Programme Part B
Module 1. Marine ecosystem data toolbox & application of macroecology
We shall re- organise existing data records on aspects of marine ecosystems from a wide range of sources
according to species traits (body size, habitat, feeding mode). We shall use this resource to conduct regional
macroecological analyses. Macroecology seeks to identify relationships between taxa and their environment
by statistical analysis of the emergent (observed) regional properties of the ecosystem.
Module 1 specifically addresses improving understanding of how the marine ecosystem as a whole responds
to specific ‘bottom up’ and ‘top-down’ perturbations through the novel combination of existing data with
recent theoretical advances from marine and terrestrial ecology.
Module 2. Fieldwork to measure poorly known processes
We have identified key components and properties of marine ecosystems which are currently under-sampled
and not adequately represented in existing ecosystem models. We shall conduct a programme of field surveys
and experiments to generate new understanding of these features and organise the information for inclusion
in models.
Module 2 specifically addresses improving understanding of marine ecosystem responses to specific ‘bottom
up’ and ‘top-down’ perturbations through the novel combination of existing long-term data, new field-based
and experimental observations with recent theoretical advances from marine and terrestrial ecology
Module 3. Ecological processes and their representation in models
In contrast to macroecology, simulation models seek to assemble ‘from the ground up’ representations of
relationships between taxa, so as to reproduce the observed properties of the ecosystem as a whole under
known driving conditions. We shall identify the implications differences in model structure have on
macroecological patterns and reduce structural model uncertainty.
Marine Ecosystem Research Programme Part B
Module 3 specifically addresses integrating improved understanding into marine ecosystem models using
concepts and expertise derived from non-marine models.
Module 4. Simulating and predicting ecosystem changes using a model ensemble
The predictive engine of the project will be an ensemble comprising well documented, contrasting, whole or
partial simulation models of marine ecosystems together with statistical models emerging from
macroecological analyses. Following the ensemble modelling principles defined by the IPCC for forecasting the
consequences of greenhouse gas emissions, we will apply parameter optimisation methods to fit each of the
models to historic time series of observations under known driving conditions, and then forecast ecosystem
states under scenarios of future conditions. The aim is to quantify the uncertainty in future predictions.
Module 4 specifically addresses integrating improved understanding of how the marine ecosystem as a whole
responds to specific ‘bottom up’ and ‘top-down’ perturbations into marine ecosystem models using concepts
and expertise derived from non-marine models, and the generation of predictions about the fate of marine
ecosystems and their services under different past and future scenarios, at local and regional spatial scales.
Module 5. Linking macroecology and models to ecosystem services
We shall map the outputs from the model ensemble onto an inventory of ecosystem services developed by
the National Ecosystem Assessment and the EU VECTORS programme (www.marine-vectors.eu). We shall
translate outputs from modules 1, 2, 3 and 4 into quantitative measures of goods and services and relevant
indicators of ecosystem status, in particular indicators that are defined in the MSFD. We will create an
integrated system capable of making forecasts of ecosystem status, goods and services for various scenarios of
future environmental conditions.
Module 5 specifically addresses translating improved understanding of the dynamics of marine ecological
communities into the currency of ecosystem services.
Module 6. Developing a model-based understanding of ecosystem service regulation
This work will our capacity to assess the structure of marine ecosystems by improving the way that
biodiversity and ecosystem function are represented in the European Regional Seas Ecosystem Model
(ERSEM). Cornerstone of this development is the transition of ERSEM from a one-size-fits-all model of fixed
complexity to a hierarchy of models that selectively inserts detail (e.g., species diversity) where demanded by
applications. The enhanced model will be used to explore the impacts of human-induced stresses and natural
variability on the structure of marine ecosystems (e.g., their species composition and size structure), and
develop links to project ecosystem services in order to meet future science and policy needs.
Module 6. Developing a model-based understanding of ecosystem service regulation
This work will our capacity to assess the structure of marine ecosystems by improving the way that
biodiversity and ecosystem function are represented in the European Regional Seas Ecosystem Model
(ERSEM). Cornerstone of this development is the transition of ERSEM from a one-size-fits-all model of fixed
complexity to a hierarchy of models that selectively inserts detail (e.g., species diversity) where demanded by
applications. The enhanced model will be used to explore the impacts of human-induced stresses and natural
variability on the structure of marine ecosystems (e.g., their species composition and size structure), and
develop links to project ecosystem services in order to meet future science and policy needs.
3.2 Relation to Programme Objectives
Modules 1, 2 and 3 directly support Objective 1 of the broader NERC Marine Ecosystems Research
Programme, which is to improve understanding of how the regulation of key ecosystem services such as food
production, macronutrient cycling and cultural values by marine food webs are affected by ‘top down’ and
‘bottom up’ driven cascades, scale- dependence in the underlying processes and functional diversity at
Marine Ecosystem Research Programme Part B
different trophic levels. This will be achieved by macroecological analysis of existing data, gathering of new
data on key missing elements, and analytical study of the consequences for cascades of alternative model
structures and mathematical formulations for representing key processes.
Modules 3, 4 and 5 support Objective 2, which is to integrate improved knowledge and understanding into
existing ecosystem models, and to explore the impact of environmental change on the structure, function and
services associated with marine food webs across scales. These modules will achieve this by analysis and
extension of a suite of 5 existing ecosystem models linked to, or driven by, outputs from ERSEM. The suite will
then be treated as an ensemble, driven by common drivers, optimised to hindcast past changes in west of UK
shelf ecosystems, and run forwards to quantify the uncertainty in future states under scenarios of climate
change and human activity .
Modules 4 and 5 provide a forward link to Objective 3, which is to apply models to test the impact of potential
management solutions, such as marine conservation zones, on the structure and function of marine food
webs, and explore the efficacy of specific indicators of Good Environmental Status (GES). Although the
Programme specification specifically excludes the conduct of Objective 3 in this project, the work proposed is
an essential precursor for it.
3.3 Geographical focus
This project will focus on the shelf seas off the western seaboard of the
British Isles (Fig. 1).
This region is characterised by highly-diverse foodwebs and supports important commercial fisheries. This
geographic focus enables us to capitalise on an exceptional archive of existing data. For instance, the UK
waters have more records in the Ocean Biogeographic Information System (OBIS, www.iobis.org) than any
other region (Tyler et al. 2012) reflecting the long history of marine biological exploration of the UK’s seas, and
the operation of a number of significant, long-term and/or spatially extensive monitoring efforts of all major
Marine Ecosystem Research Programme Part B
components of the ecosystem including fish (e.g. International Bottom Trawl Surveys, Cefas’ DAPSTOM
database of fish stomach content records and stable isotope analyses), plankton (e.g. Continuous Plankton
Recorder Survey), predators (e.g. Seabird Monitoring Programme, European Seabirds at Sea), Cefas benthos
surveys, and a network of comprehensive monitoring sites (e.g. Western Channel Observatory, L4 (Plymouth
Marine Laboratory), Cefas CSEMP (Clean Seas Environmental Monitoring Programme) stations in the Celtic
Sea, Irish Sea and Bristol Channel and Loch Ewe (Marine Scotland). Past (Tyler et al. 2012) and ongoing
(EMODnet biological attributes) data compilation on the biological traits of key components of these systems,
again mean that our project draws upon an exceptional base of existing data of international significance. The
region is accessible using available inshore and medium-range vessels with appropriate capabilities. Finally,
the focus on the western seas allows us to capitalise on the Cefas Poseidon project cruises in SW UK waters,
Cefas groundfish surveys and the field work planned in the NERC Shelf Seas Biogeochemistry Programme, to
provide an innovative characterisation of the whole system from biophysical ecosystem models through to top
predator dynamics. Such a complete picture would be impossible for less well- studied regions.
4. Policy relevance and impact goals
MERP will link services to variables within ecological models to assess how ecosystem services will change in
the future (incorporating IPCC climate scenarios), contributing to the objectives of the EU Biodiversity Strategy
and the Natural Environment White Paper. We anticipate that IMMERSE understanding, models and outputs
will support decision making (e.g. marine planning, licencing and conservation) of regulators under the Marine
and Coastal Access and the Marine (Scotland) Acts.
The UK and other European states are investing heavily in monitoring and consultations designed to quantify
descriptors of Good Environmental Status (GES) defined in the MSFD. Current specification of descriptors is
largely driven by the availability of data. We expect that our research will indicate new and different
measurements and assessments that could be implemented to provide more insightful and sensitive indicators
of GES. The MERP framework can be used strategically to project relative changes of descriptors of GES (e.g.
biodiversity, fish stocks, foodwebs, and seabed integrity) (Lassen et al 2013) under different management
scenarios (within programme Objectives 3). Outputs from this project will help inform future revisions of the
EU Data Collection Regulations (DCR) on the management and use of fisheries data, and support the revised
EU Common Fisheries Policy. Our impact goal is to have a decisive influence on the next generation of marine
monitoring programmes deployed in the UK and across Europe, and to provide a holistic framework to support
decision making in the marine environment.
5. Collaborations with NERC, LWEC and other research programmes
MERP builds on several recent marine research programmes, including Quest-fish (www.quest-fish.org.uk)
(ended 2010), (GLOBEC (www.globec.org) (ended 2010) and Oceans 2025 (www.oceans2025.org) (ended
2012). Several of the partners in MERP were leading participants in these programmes. The project also
complements ongoing work in international programmes where the UK has a significant (often leading) role:
VECTORS (www.marine-vectors.eu), DEVOTES (www.devotes-project.eu), Euro-Basin (www.euro-basin.eu),
GreenSeas (www.greenseas.eu), MEECE (www.meece.eu) IMBER (www.imber.info/index.html), ICED
(www.iced.ac.uk), EMECO (www.emecogroup.org/), EuroSITES (www.eurosites.info), ESONET (www.esonetemso.org), MarBEF+ (www.marbef.org), EuroMarine (www.euromarineconsortium.eu), EMODNet
(http://bio.emodnet.eu/) and EUR-Oceans (http://vds1719.sivit.org/eoc/). Again, IMMERSE scientists are
involved in these programmes.
The project has close working ties with the UK National Capability in ecosystem modelling, which is based on
the European Regional Seas Ecosystem Model (ERSEM). Outputs from ERSEM (50 yr hindcast) will inform
macroecological analyses and model development. Several of the models which will be used in the project
Marine Ecosystem Research Programme Part B
have been interfaced with ERSEM or actively use output ERSEM as input driving data. ERSEM is a
biogeochemical model that largely ignores the macroecology of larger bodied organisms which provide a
substantial part of ecosystem goods and provisions. IMMERSE will provide an understanding of the wider,
“end-to-end” food web to be elaborated and integrated into ERSEM in subsequent funding rounds. Data from
all modules will feed back into ERSEM validation and development. Momme Butenschön, a key member of
the ERSEM development team, is a member of the consortium to facilitate this integration. IMMERSE draws
heavily on NERC National Capability investment in observational systems around the UK, especially the
Continuous Plankton Recorder Survey and the Western Channel Observatory. The project will also capitalise
on the heavy UK investment in trawl surveys for fish and benthos (Cefas, MSS and AFBI). The project will also
interface closely with ongoing NERC Programmes (in particular Shelf Seas Biogeochemistry (SSB), and
Biodiversity and Ecosystem Services (BESS)). Members of the MERP consortium are involved in both these
programmes, and the former in particular is conducting field work in the same geographical region. The MERP
and SSB project will work closely together to share data and resources.
6. Work module descriptions
6.1 Module 1: Marine ecosystem data toolbox & application of macroecology
Co-Is: Webb (Sheffield), Burrows (SAMS), Daunt, Searle, Wanless (CEH), Edwards (SAHFOS), Emmerson (QUB),
Evans (Bangor), Hirst (QMUL), Nager (Glasgow), Pinnegar, Schratzberger, Van Der Kooij (Cefas), Somerfield
(PML)
Macroecology investigates patterns of ecological systems that emerge at large spatial or temporal scales, as
well as the interactions between these scales (Beck et al. 2012; Keith et al. 2012). By focusing on properties of
ecosystems that emerge at large scales from the complex interactions of many fine-scale processes,
macroecology complements the mechanistic ecosystem models described in Modules 3 and 4.
Macroecological scaling rules linking body mass with characteristics such as population abundance, individual
space use and foraging rate (Brown et al. 2004; Jetz et al. 2004; Pawar et al. 2012) can be used to
parameterise mechanistic ecosystem models, whereas model outputs can be tested against alternative,
independent macroecological patterns such as community size spectra and species-abundance distributions
(SADs) which have robust empirical support (Webb 2012).
Comprehensive macroecological analyses of the Western Seas are rare, and whole-system quantification of
key descriptors of food webs, size spectra, and SADs are lacking. Although this region is data rich (Fig 1), the
data are disparate, scattered, and contain gaps and biases. Creating Big Data by combining these outputs from
past projects provides significant opportunities (Hampton et al. 2013). We will synthesise existing data using
an ecoinformatics approach (Peterson et al. 2010; Michener & Jones 2012; Hardisty & Roberts 2013),
repurposing available data (see Outline Data Management Plan) to support our theoretical and empirical
modelling framework, developing tools to efficiently and transparently access and quality screen the data, and
identifying major gaps that will be addressed by compiling data from the literature and via the field sampling
described in Module 2. This data synthesis will feed directly into work to quantify major macroecological
patterns across functional groups, including how they vary in space and time in response to changing
environmental and anthropogenic pressures, and how patterns covary across trophic levels – in particular, the
top-down role of the spatial and temporal dynamics of vertebrate predators (seabirds, marine mammals, and
fish) on lower trophic levels. This provides both a means to test outputs of the models developed in Modules 3
and 4, and a route (task 1.3) to translating size-based model outputs into the species-based indicators of
Ecosystem Service provision (Module 5).
6.1.1 Module Tasks
T1.1 Ecoinformatics
Marine Ecosystem Research Programme Part B
The Western UK focus of this project enables us to capitalise on exceptional coverage of existing biodiversity,
ecological, environmental, and socio-economic data. The major large-scale, long-term surveys and atlases of
species distributions are listed in our Outline Research Data Management plan, together with the
organisations responsible for their management, and fixed-location monitoring time series, extensive
environmental records (e.g. temperature, ChlA, bathymetry, substrate) and information on human pressures
and ecosystem service provision (e.g. fisheries landings, nutrient cycling). Task 1.1 is designed to collate and
provide access to these data such that key macroecological patterns (T1.2, T1.3) can be directly compared with
the predictions of our model ensemble (Modules 3, 4). Our key innovation is to link these well-established
marine biological databases to facilitate such analyses. Together, the taxonomic name of an organism and its
location in space and time provide unique ‘access keys’ (Hardisty & Roberts 2013) as the basis for a suite of
data enrichments. Taxonomic names link directly to other resources including biological and physiological
traits, habitat affinities, and ecological function and a comprehensive UK list is already available in the World
Register of Marine Species. Where and when a biodiversity record was collected enables links to physical
environmental observations made in situ, obtained from remote sensing, or inferred from models (e.g.
ERSEM). These enrichments allow us to leverage existing biodiversity data into the development and testing of
the ecosystem models described in Module 3, including delimiting the spatial distributions of taxa of
contrasting mobilities (e.g. Isojunno et al. 2012; Wakefield et al. 2013), and quantifying macroecological
patterns, relationships and dynamics at various spatio-temporal scales (Task 1.2). They also identify priorities
for new data collection from the literature (e.g. predator life histories; Evans & Raga 2001) and from the
sampling programmes in Module 2 to fill key gaps. Important information relevant to the ecosystem models
and not included in existing data compilations are spatially- and temporally-resolved quantitative data on
resource utilisation for top predators and focused data collection from the literature will fill this gap across the
major foraging guilds, correcting for known biases of different methods of diet identification (Barrett et al.
2007). Throughout the focus is on developing open, lasting ecoinformatics tools, primarily in R, as this is both
more useful and more sustainable than developing new user interfaces which may not outlive the duration of
the funded project.
T1.2 Empirical Macroecology
We will develop, test and apply whole-system macroecological analyses (specifically based around the
fundamental parameters body size and temperature). Because organisms from different habitats and of
different sizes interact and are influenced by their environment in fundamentally different ways (Chave 2013)
we hypothesise that patterns at a single scale will differ between size classes, functional groups, and
habitats in gross and predictable ways. For instance, Pawar et al. (2012) derive the scaling implications of
differences between trophic interactions in 2D (benthic) versus 3D (pelagic) systems, while different
organisational levels – and predators versus prey – have different thermal dependence to their ecological
rates (Dell et al. 2011; Frederiksen et al. 2013). Our approach will be extended to consider differences
between habitats within systems, including comparing benthic systems with and without macrophytes,
predators with their prey, and the role of spatial and temporal heterogeneity in environmental conditions
applying the framework developed by Searle et al. (2010). We will examine the consequences of changes in
abundance of top predators on scaling relationships lower in the food chain, and the role of functional
diversity and spatio-temporal heterogeneity in lower trophic levels and climate on predator abundance,
richness, and population dynamics. There is a comprehensive understanding of scaling of feeding, growth and
metabolic rates for many terrestrial and aquatic organisms (Brose et al. 2008, Kiørboe & Hirst submitted), we
will test the generality of such scaling for benthic species (Emmerson & Raffaelli 2004, Twomey et al. 2013),
and organisms which move, or have trophic dependence, between these domains (e.g. benthic species with
planktonic larvae, seabirds that breed on land but forage at sea).
Comparisons will be based on meta-analyses and estimation of parameters from simple quantitative
Marine Ecosystem Research Programme Part B
macroecological summaries will include vital rates, life history, size spectra (slope, non-linearity estimated
using Generalized Additive Models (GAMs)), SADs (mean and variance of log-normal fits; Saether et al. 2013),
and species-area relationships (SAR, slope). Combined they will produce novel trivariate food web data
descriptions combining abundance, body mass and food web structure (Cohen et al. 2003). We will quantify
these relationships in different functional groups, e.g. benthos, zooplankton, demersal fish, seabirds and
cetaceans, and test weighting factors (e.g. typical distributional extent; Storch et al. 2012) to examine scale
independence across groups.
T1.3 Size, species and functional groups
To relate model outputs from Module 3 (e.g. biomass within functional groups (EwE), community size spectra
(Size spectrum models, fishSUMS)) to empirical descriptions of these same phenomena, we will undertake an
extensive comparison of different metrics for summarising community structure, focusing on comparing size
structure (slope and linearity of size spectrum, diversity and evenness across size classes) with characteristics
of the SAD (mean, variance) over space in contrasting environments, through time, and across major
functional groups, and building on recent efforts to reconcile size spectra with taxon-based trophic pyramids
(Trebilco et al. 2013) and species richness (Rossberg 2013). We will focus on datasets which include both
species identity and individual size (or where the latter can be accurately estimated) over extensive spatial
scales (e.g. IBTS, Cefas fish stomach content database) or at very fine temporal resolution (e.g. L4) to
empirically resolve patterns of covariation between different metrics, and to develop a simple conversion
between the currencies of modelling (mostly size-based), macroecology (size- or species-based), and
ecosystem services (mostly species- based). We will also examine how species-level variation in seabird and
marine mammal populations influences size-based properties of the system at lower trophic levels. We will
explore material flux through size-structured food webs, and test predictions of the Metabolic Theory of
Ecology (Brown et al. 2004) using biomass- and density-size distributions and our empirical descriptions of size
dependence of rate processes. Integrating with modules 2 & 3, we will quantify density-dependent aspects of
population size using the slope, shape and variation of density-size spectra.
Dependencies: T2.2
6.1.2 Milestones
No.
M1.1
M1.2
M1.3
M1.4
M1.5
M1.6
M1.7
Milestone
Identify, compile, format and initial quality control long-term datasets
for major functional groups, for use in model development and testing
Build R package to access data via web services
Produce a map of raw data availability across taxa and through time to
identify major gaps
Compile Species Abundance Distributions from diverse sampling
programmes across multiple taxa
Develop Shiny web applications to enable simple visualisations of
biodiversity data for non- academic users
Develop and test methods (including GAMs, GAMMs, Random Forests)
for the spatial interpolation of survey data
Produce synthetic data products in useable open formats satisfying EU
and UK drivers towards open science
Date
completion
(M12)
Lead
Partner
Sheffield
(M12)
(M24)
Sheffield
Sheffield
(M24)
Sheffield
(M36)
Sheffield
(M36)
Sheffield
(M48)
Sheffield
6.1.3 Deliverables
No.
D1.1
Deliverable
Inventory of large-scale, long-term datasets for
major functional groups in the Western Seas,
including known sampling issues and biases
Marine Ecosystem Research Programme Part B
Date
completion
(M12)
Dependencies
None
Lead partner
Sheffield
(Webb)
D1.2
D1.3
D1.4
D1.5
D1.6
D1.7
D1.8
D1.9
Submit R package for accessing marine taxonomy to
CRAN
Joint meeting with British Ecological Society
Macroecology Special Interest Group & National
Biodiversity Network on frontiers in marine
macroecology and data
Paper on macroecology of major functional groups in
the Western Seas: scale dependence and the role of
environmental covariates
(M12)
None
Sheffield
(Webb)
Sheffield
(Webb)
Sheffield
(Webb)
(M24)
D1.1, D1.2,
work
contributing to
D1.5
D1.1, D1.2
Submit additional R packages for accessing marine
macroecological data to CRAN
Report on methods for interpolating sampled
distribution and abundance data across taxa, and
advice on best practice
Report empirical relationships between size spectra
and species-level macroecological patterns, and
scale-dependence of species abundance distributions
across functional groups
Report on environmental and biological correlates of
regional differences and temporal variation in toppredator communities
Deliver synthetic data products and tools to BODC in
accordance with Research Data Management Plan
(M36)
D1.1
(M36)
D1.1, D1.2, D1.5
Sheffield
(Webb)
(M48)
D1.1, D1.5
CEH (Searle)
D1.1, D1.2,
D1.5, D1.6, D1.7
Sheffield
(Webb)
(M18)
(M24)
(M48-60)
Sheffield
(Webb)
Sheffield
(Webb)
Links to other Modules
Module 1 provides the empirical data underpinning the modelling in M3-4, and gaps identified will feed into
the survey designs in M2. The data management and ecoinformatics tools will provide immediate access to
the new data collected in M2. Task 1.3 is designed to enable transitions between the currencies of models
(M3-4) and Ecosystem Services (M5).
6.2 Module 2: Fieldwork to measure poorly known processes
Co-Is Atkinson, Queirós, Lindeque (PML), Burrows (SAMS) Emmerson, O’Connor (QUB), Hiddink (Bangor), Van
Der Kooij, Pinnegar (Cefas), Hirst (QMUL)
While the ecology of NW European shelf is generally well quantified (Module 1), for the purposes of modelling
and understanding services some key gaps remain. These gaps represent components and processes not
sufficiently understood to be resolved in ERSEM or the models studied in Modules 3 and 4, and form the basis
of the Module 2 Tasks.
Given what is known about area-specific production rates (Mann 1973) and relative habitat areas for
production (8000km2 kelp (Walker 1953), 133000km2 phytoplankton <20km UK coastline), intertidal and
subtidal macrophytes (mainly kelp) may contribute 45% of total primary production in UK coastal waters,
supporting the shellfish fishery worth £282M in 2011, greater than demersal and pelagic landings. Despite its
potential importance, macrophyte detritus is not included directly in any ecosystem model, including ERSEM.
The contribution of kelp forests in supporting culturally and commercially valuable inshore fisheries is not yet
known and will be a key outcome of this project.
Understanding how matter and energy flows between benthic and pelagic systems is vital to modelling marine
ecosystems, and thus how such processes (and the valuable ecosystem services they underpin) will be
impacted by future climate scenarios. While the ERSEM model and NERC’s Shelf Sea Biogeochemistry
programme emphasise physical/geochemical processes, many biological processes contribute to the
Marine Ecosystem Research Programme Part B
incorporation of primary production into benthic foodwebs (e.g. contact of live versus dead material with the
seafloor, trophic assimilation, bioturbation), and their relative importance is poorly known (Evrard et al. 2012).
The strength of benthic-pelagic coupling and asymmetry between detrital and primary producer energy
pathways is important for stability and resilience in simple consumer-resource and coupled size spectrum
models (Rooney & McCann, 2006, Blanchard et al. 2011). Within benthic-pelagic coupled systems, body size
dictates the pace of life (physiology, feeding and reproduction) and forms an underlying structuring element
across the >18 orders of magnitude of body mass (White et al. 2007). While this is well grounded in both the
terrestrial and aquatic literature (O’Gorman & Emmerson 2011), complete biomass spectra are not available
at whole system scales and very little is understood about the details of various types of coupling across all
trophic levels.
Similar size/trophic level organisms can have fundamentally different roles for ecosystem services. For
example, under the MSFD, the relative abundance of gelatinous zooplankton and fish larvae is proposed as an
indicator of Good Environmental Status. The suggestion that jellies are increasing due to degraded ecosystem
states, displacing fish larvae (Richardson et al. 2008), is debated. Our 4th objective is thus to quantify variation
in feeding and mortality as a function of species traits, e.g. as a function of body mass and feeding mode.
Central to IMMERSE is scale dependency and the sampling is orientated to encompass a continuum of time
and space scales. The Plymouth L4 shelf monitoring site has 25 years of weekly sampling across the whole
planktonic food web and seasonal benthic monitoring. This will form part of a UK-wide group of shore-to-30
km offshore transects quantifying the macrophyte derived subsidy to the food web (Fig. 1), embedded in
regional scale Cefas-Poseidon and NERC-Prince Madog cruises, and using data synthesised in Module 1. Our
approach will extend scales of space and time to shelf wide processes such as top predator foraging ambits.
Marine Ecosystem Research Programme Part B
6.2.1 Tasks
T2.1 Quantify the macrophyte versus phytoplankton contribution to the food web in relation to distance
offshore and at whole-shelf scales
This task will be approached at two scales. First we will estimate kelp production and compare with overall
primary production by phytoplankton at whole UK scales. Secondly we will measure the onshore-offshore
gradient in kelp subsidy to the benthic food web by following the assimilation of natural markers for kelp
versus phytoplankton up into the foodweb along five transects. Biomass and production of kelp will be
estimated on a UK scale by synthesising existing data on distributions (JNCC, NHM, Channel Coastal
Observatory), and production rates (Krumhansl and Scheibling, 2012). We will measure habitat extent for
kelp species using satellite data that show kelp beds (Casal et al. 2011, Byrnes et al. 2011) and use synoptic
data on primary drivers of distribution, ocean colour, wave fetch and temperature (Burrows 2012, Bekkby et
al. 2009), to drive measurements and collated abundance data into models of species distribution and
production. At local scales (1-20km) near Oban and Belfast, novel data from new technology (RPAs, ROVs,
and AUVs from developing SAMS capabilities and facilitated by AFBI) on the incidence of kelp along high
resolution environmental gradients will improve the performance of the distribution models.
We propose four primary sites across a latitudinal gradient with different kelp species and environmental
conditions (Fig. 1) where (1) Primary production will be estimated by: (a) in situ production and rates of loss
of fronds and whole plants, (b) in situ PAM fluorometry, and (c) floating mesocosms enclosing whole plants
in sheltered areas. (2) Direction and distance of detritus transport will be tracked with labeled fragments and
thalli using diver surveys and towed video. (3) Kelp bed monitoring will include estimations of blade damage
(to estimate detrital inputs from small particles to whole thalli) and rates of dislodgement and erosion
(particles). UK-wide estimates of total kelp production will be compared to those for phytoplankton primary
production based on current best spatially resolved ecosystem models (primarily ERSEM) and satellite
observations.
At the finer scale we will sample along gradients from the shore to ~30 km offshore using inshore boats of
AFBI (see letters of support), SAMS, PML and Prince Madog. This work, along a variety of habitats right
across the UK (Fig. 1) will quantify the relative roles of phytoplankton- and macrophyte-derived material in
supporting the food-web. We will conduct full depth CTD profiles to collect suspended POC with concurrent
dredge, sledge and core sampling sampling of kelp detritus, sediment POC and DOC, infauna, up to
commercial species such as Nephrops. Given the methodological challenges (e.g. Dethier et al. 2013) we will
use a combination of dietary approaches, namely mixing models based on contrasting kelp/phytoplankton
δ13C signatures, lipid/fatty acid markers and molecular approaches (18S RNA). The transport of coastal
detritus will be modelled with FVCOM, POLCOMS/NEMO. In combination these approaches will reveal the
spatial scale of kelp subsidy over 10s m to 100 kms in coastal areas, and provide an understanding of the role
of macrophyte-derived material in supporting the commercially valuable inshore food web.
T2.2. Quantify pathways by which sources of primary production enter benthic food webs
We will conduct a seasonal (4 times per year for 2 years) sampling programme extending the mechanistic
insights obtained in Task 2.1 and adding trophic information to the biomass spectrum sampling in Task 2.3,
by tracing trophic and non-trophic flows of macrophyto-detritus/phytoplankon through the benthic foodweb. We will use a combination of tracing techniques including stable isotope and phospholipid derived fatty
acid analysis (PLFA, Hardison et al. 2010, Evrard et al. 2012). Using an epibenthic sledge, the natural food
source in the bottom few cm of the water column will be sampled throughout the tidal cycle and compared
with the overlying water column derived from isotopic and FlowCAM analyses (Task 2.4). These same
techniques will be used to provide seasonal coverage of the quality and origin of detritus, e.g. phytoplankton
versus macrophyte, that supplies the benthic food web. Using sediment cores we will undertake seasonal
tracer experiments to 1) determine seasonal variation in the uptake of detritus material in to the benthic
foodweb (water, sediment, bacteria, meio- and macrofauna) by determination of δ13C- backgound
signatures; 2) Seasonally resolve changes in food-web structure with and without the experimental addition
of 13C enriched macrophyte material, illustrating the potential for bottom-up control of coastal food webs at
Marine Ecosystem Research Programme Part B
a fine spatial scale; and 3) measure bioturbation in parallel to clarify the relative importance of this nontrophic pathway in the overall uptake of macrophyte detritus by the benthic ecosystem.This work will inform
the trophic and non-trophic characterization of functional groups and their influence on ecosystem
processes in large-scale models like ERSEM (Module 3). All fauna will be measured to help generate biomass
trophic spectra (Tasks 1.3, 2.3). The range of oceanographic data available from the L4 site helps to provide a
mechanistic understanding, and placement of this work in the wider context within the onshore-offshore
network of transects (Task 2.1).
Dependencies: T1.3, T2.1, T2.3, T2.4
T2.3 Determine the complete biomass-size spectrum in the benthic and pelagic food web
We will capitalise on Poseidon, an annual pelagic sampling cruise undertaken aboard Cefas Endeavour (Fig. 1)
spanning the autumn bloom in 2015, using RV Prince Madog to take additional samples at a subset of the
stations. These stations will be selected to fit a hierarchical sampling design, where a range of distances
between sampling stations allow the analysis of the role of scale in the food web. Stations will also be
harmonised with those sampled the previous year on the NERC Shelf Sea Biogeochemistry Programme.
Together, the 2-ship cruise will quantify food webs and size-spectra over crossed gradients of primary
production: i.e. for Prince Madog (4 levels, as a bottom up driver) and fishing activities (4 levels, as an
anthropogenic top-down driver) resulting in 16 sampled benthic stations, c.f. ~70 pelagic stations for
Endeavour. Substantial gradients in primary production and fishing effort exist in the study area in the Celtic
Sea and past work (Hiddink et al. 2011) shows that this design allows estimation of large-scale effects of
fisheries.
Using these platforms we will quantify size-spectra for pelagic and benthic ecosystems using a wide variety
of sampling gears to capture the full size spectrum quantitatively. Water bottle profiles, an array of 4 net
sizes (80 μm up to pelagic trawls) and acoustic records will be processed respectively by flow cytometry,
flowCAM, zooscan and Echoview acoustics software. During the cruises, marine bird and mammal densities
will be determined using distance sampling line- transect methods (Thomas et al., 2010). At select food
patches where aggregations of top predators occur, sampling will take place at all trophic levels, and their
persistence determined over the short-term to establish spatio-temporal variation across the food chain, in
relation to measured or estimated (task 1.1) consumption rates (cf. Benoit-Bird & McManus, 2012).
Observations of higher predators including seabirds and seals and cetaceans are integral to these cruises,
and the complete biomass spectrum for larger, more mobile organisms will be supplemented by the array of
existing data compiled within Module 1. Benthic sampling will focus on important but hard-to-sample groups
(e.g. large bivalves and mobile hyper benthos) using targeted dredges and sledges. A second Prince Madog
cruise, during the spring bloom of 2016, will repeat this sampling coverage.
These regional cruises cover spring and autumn blooms only. We recognise that there are fundamental
seasonal shifts from putative export systems (large producers) to regenerating type systems (small
producers), and to capture the dynamics of the full seasonal size spectrum for the pelagic environment we
will sample the Plymouth L4 site monthly throughout 2015. This will use equivalent gears and analysis
methods as above to enable effective integration of the data. The advantage of the L4 site is that it has 25
years of weekly planktonic sampling with partial size spectra data collected under NERC National Capability,
forming a longer-term context for the 2015 intensive study season. Using data gathered in the NERC Oceans
2025 programme, supplemented by cores obtained in Task 2.2, we will construct a benthic biomass
spectrum for the L4 site that resolves seasonal variability. In conjunction with the tracer work in Task 2.2 this
will allow construction of feeding guilds.
Dependencies: T1.1, T2.2
T2.4 Quantify key, but poorly known, feeding and mortality parameters
Within MERP we have particular expertise in the measurement of diets, feeding and mortality using a wide
range of classic and state-of-the-art techniques (Atkinson 1996, Pinnegar et al. 2002 Hirst et al. 2004,
Emmerson & Raffaelli 2004, McLaughlin et al. 2010, O’Gorman & Emmerson 2010, Twomey et al. 2013).
Marine Ecosystem Research Programme Part B
Quantification of diets is non-trivial, and consequently we will target key species that require better
appraisal. Much diet/trophic level data are already available for fish (Module 1) but we know less about
feeding and diet at lower trophic levels. Smaller consumers can have multiple feeding modes (e.g. Kiørboe
2010), large ranges in predator-prey mass ratios (Wirtz 2011) and rapid isotopic turnover (Schmidt et al.
2003), hampering some methods applied to larger consumers. Likewise mortality is notoriously hard to
estimate, and the density-dependent terms often used to close ecosystem models (Module 3), including
ERSEM, are poorly grounded in evidence.
To address these issues we will:
1) employ trait-based approaches to experimentally estimate parameter values such as feeding- and
mortality rates as a functions of body mass. We will test these approaches for estimating parameter values
across benthic and pelagic food webs and validate size based feeding relationships that have been
documented extensively elsewhere (Brose et al. 2008, Rall et al. 2010).
2) Given constraints on size-based approaches at lower trophic levels we will use a model food web motif of
key species ubiquitous in UK waters to estimate the strength of predator-prey and competitive interactions.
This pelagic food web motif will comprise Oithona similis (tiny but numerically dominant ambush feeder),
Calanus helgolandicus CV/adults (biomass dominant, mainly suspension feeder and key dietary item), fish
larvae (larger raptorial, visually-feeding carnivores supporting recruitment to commercial fisheries), and
Ctenophores and Cnidarians (largest, often non-visual, passive ambush predators). Benthic species will
include amongst others Nephrops norvegicus, Necora puber, Buccinum undatum, and Asterias rubens,
reflecting dominant benthic species and those of commercial importance.
We will sample these across scales, namely across the diel cycle, across the seasonal cycle (during 12 visits of
L4 during 2015) and regionally (Poseidon cruises aboard CEFAS Endeavour in autumn 2015). For the MSFD
life-form pair of jellies and fish larvae we will combine approaches to their diet. We will apply stable isotopes
(bulk 13C and 15N) to estimate food resource base and trophic level, comparing this with a novel molecular
diet approach using Next Generation Sequencing (NGS) technology. At PML we have shown that high
throughput 454 pyrosequencing of nuclear small subunit 18S rRNA gene amplicons is a powerful tool for
estimating diversity and species richness of mixed zooplankton communities. NGS of 18S amplicons will be
used to assess gut content (see O’Rorke et al. 2012). The third diet method will be classic stomach content
analysis. For taxon-and size-specific information on the crustacean part of the diet, prey mandible size will be
converted to prey size (Atkinson et al. 2002). Laboratory feeding incubations at PML will compare functional
responses of O. similis, C. helgolandicus and Ctenophores based in experiments using the ambient seasonal
food assemblages with their naturally varying food types and sizes, providing estimates of consumer
satiation as input to Module 3. The PML flowCAM, which quickly and automatically counts, measures and
classifies thousands of food items will also enable us to better measure prey switching and the clearance
rates on items when in low abundance. Mortality of O similis and C. helgolandicus will be estimated by the
vertical life table approach, based on the weekly samples collected routinely at the L4 monitoring site (Hirst
et al. 2004). We will construct empirical models to tease out the effects of mortality, including predator
abundance, size, temperature as well as density-dependent effects. Likewise, functional responses of the
benthic species will be measured in the mesocosm facilities at QUB under the supervision of Co-I Emmerson.
6.2.2 Milestones
No.
M2.1
M2.2
M2.3
Milestone
Field programme to estimate scale and extent of macrophyte
subsidy to UK marine ecosystems
Experimental programme to determine rates and pathways of
incorporation of primary production at the base of marine foodweb
Field programme to quantify seasonal variation in body-size
spectra
Marine Ecosystem Research Programme Part B
Date completion
(M24)
Lead partner
PML
(M36)
PML
(M36)
PML
M2.4
Combined Poseidon/Prince Madog large-scale field programme
to characterise whole-ecosystem size spectra and spatial scaling
(M36)
PML
6.2.3 Deliverables
No.
D2.1
D2.2
D2.3
Deliverable
Report on the quantification of trophic and nontrophic benthic-pelagic coupling pathways for
macroalgae-derived carbon sources
Report on the dynamics of interactions between
gelatinous zooplankton and fish larvae
Report on “end to end” pelagic and benthic
biomass spectra across regional gradients and
seasons
Date
completion
(M36)
Dependencies
SAMS (Burrows)
(M36)
(M42)
D2.4
Report on the parameterisation of functional
responses of feeding and mortality related to traits
(M42)
D2.5
Parameterization of trophic and non-trophic
pathways of carbon assimilation in coastal benthicpelagic systems in ecosystem models
(M42)
Lead partner
PML (Lindeque)
M1 for provision
of data on larger
organisms to
construct full
“end to end”
biomass spectra.
Bangor
(Hiddink)
QMUL – benthic
(Emmerson)
PML - pelagic
(Atkinson)
PML (Quierós)
Links to other Modules M2 provides empirical data underpinning the modelling in M3-4, and data will feed
into the macroecological analyses in M1. The data management and ecoinformatics tools will provide
immediate access to the new data collected in M2. Data and information (in terms of appropriate measures
and indices) will inform spatio-temporal assessment of ecosystem services (M5).
6.3 Module 3: Ecological processes and their representation in models
Co-I Rossberg (Cefas), Blanchard, Webb (Sheffield), Heath, Speirs (Strathclyde)
Simulation models seek to capture the essence of relationships between taxa to reproduce properties of
complex ecosystems and thus to enable predictions of ecosystem responses to perturbations. Because
ecosystems are complex, this requires simplification, and the simplifying assumptions of a particular model
inevitably lead to imperfect predictions of true dynamics due to ‘structural uncertainty’ (Hosack et al. 2008).
Simplifications include the combination of taxa into ‘functional groups’ which occupy similar habitats, feed in
similar ways, or have similar body size; and assumptions regarding the functional responses governing intake
rates by predators in response to prey densities. Alternative formulations are used to reflect variation in
underlying biology and can have very different effects on trophic cascade propagation (Herendeen 1995,
2004; McCann et al. 1998). Most models include density dependent (self-limiting) terms applied to either
mortality or intake rates which confer numerical stability with highly significant consequences for both topdown and bottom-up propagation (Herendeen 1995; Heath et al. submitted; McCann et al. 1998), but
although meta-analyses have revealed some generalisations (Borer et al. 2005; Shurin et al. 2002) the basis
for different model formulations remains uncertain. Module 3 will characterise the relationship between
formulations of density-dependent population control mechanisms and the resulting properties of trophic
cascades to identify which response forms are appropriate in different circumstances. Our primary tool to
compare empirical macroecological data with simple analytic and complex numerical models will be
elasticity indices (Hill et al. 2008), defined as ratios of the form (% change in Y)/(% change in X) to quantify
the strength of the effect that a variable X (e.g. primary production) has on another variable Y (e.g. fish
production). The strengths of trophic cascades can be quantified by the elasticities of the effects that
changes at one trophic level have on other trophic levels. Observations and models (Ware & Thomson 2005;
Rossberg 2012) suggest that the elasticity indices for bottom-up effects are usually >1, whereas top-down
Marine Ecosystem Research Programme Part B
cascades in models are usually damped (elasticity < 1, Andersen & Pedersen 2010), but not always (Rossberg
2012).
6.3.1 Tasks
T3.1 Extraction of relevant elasticities from the empirical macroecological data
The statistical methodology to conduct empirical elasticity analyses is well developed (Qian and Giles, 2007).
The relevant macroecological data – spatially- and temporally resolved size spectra distributions across
major functional groups and top-predator density dependencies, including effects of resource heterogeneity
– will be obtained in Module 1 and Module 2 (T2.3, T2.4),. The focus will be on top- down and bottom-up
cascades. Approaches to take spatial and temporal correlations into account will be informed by analyses of
scale dependence in Modules 1 and 2.
Dependencies: T1.3, T2.3, T2.4
T3.2 Mathematical analyses of the effects differences in model structure have on elasticities in simple models
Mathematical elasticity analysis for simple models is facilitated by the fact that elasticity is determined by
the linear response of the system to pressures, enabling the use of Generalized Modelling (Gross et al. 2004)
to derive elasticity indices. This method measures the dependence of the elasticity coefficients for systemlevel responses on the elasticity coefficients for the direct density-dependent interactions among species or
functional groups. Large classes of conceivable functional forms for these interactions can therefore be
covered by a few analytic calculations. For example, the analysis of a simple, infinitely long food chain
(Rossberg 2013) reveals a strong dependence of elasticities for top-down cascades on consumer satiation,
i.e., on the elasticity of the dependence of consumer intake on resource abundance. Quantitative data on
functional response types, satiation, and density-dependent mortality determined in Task 2.4 is used to
constrain the parameter ranges relevant for the study area. Analyses will be extended to incorporate
phenomena such as benthic-pelagic coupling of food chains (Module 2).
Dependencies: T2.2, T2.4
T3.3 Numerical analyses of the effects of model structure on elasticities in complex models and comparison
with data
The numerical experiments required to probe the elasticities of macroecological responses of complex
ecosystem models to pressures are straightforward with fast model software and the knowledgeable model
operators. This task will therefore be conducted in one of the project’s integrative workshop that brings
together experts on a representative yet easily simulated sub-set of the candidate models to be included in
the model ensemble of Module 4, as well as empiricists with a good understanding of the macroecological
data used (Module 1). The workshop will evaluate the standard implementations of models as well as
variants with different representations of density dependencies, informed by Task 3.2. The resulting
enhanced understanding of the implications of model structure will be used to adapt models as required by
comparisons with data such as that obtained in Task 3.1, while leaving variations in model structure that are
not constrained by this comparison intact. General recommendations for representations of densitydependencies in models will inform the model ensemble in Module 4.
Dependencies: T1.2, T3.1, T3.2
T3.4 Comparison of macroecology in models and data beyond elasticity
The conceptual simplicity of elasticity analysis allows its broad application to various data and models, but it
cannot cover all macroecological patterns and processes of interest. Some models make much richer
predictions. Such model outputs will be confronted with the detailed macroecological pictures obtained in
Module 1, including relationships between food-web structure, size spectra, and SAD. This will lead to both
improvements in model structure and mechanistic interpretations of observations. Recent theory (e.g.
Rossberg 2013) will guide this comparison. Among the hypotheses we examine will be that modulations of
size spectra can generate modulations of richness along size such as wasp-waist food-web topologies (Cury
Marine Ecosystem Research Programme Part B
et al. 2000).
Dependencies: T1.2, T1.3,
6.3.2 Milestones
No.
Milestone
Date completion
M3.1
Numerical experiments to probe the elasticities of
macroecological responses of complex ecosystem models
Improvements in models and mechanistic interpretations of
observations
(M24)
Lead
partner
Cefas
(M36)
Cefas
M3.2
6.3.3 Deliverables
No.
D3.1
Deliverable
Report empirical values of relevant
elasticity indices confronted with
analytic results
D3.2
D3.3
Joint module 3/4 workshop
Manuscript on in-depth comparison
between models and macroecological
data
Date completion
(M18)
(M24)
(M36)
Related Milestone
Spatially or temporally
resolved size spectrum data
available from Modules 1
and 2
D3.1
Detailed macroecological
data from Module 1
Lead partner
Cefas/UoS
Cefas/SU
Cefas, SU
Links to other Modules Data provided by M1 will be used to calculate empirical elasticity indices.
Observations in M2 will support choices for variants of model structure. Constraints on model structure
derived in M3 inform construction of the model ensemble in M4.
6.4 Module 4: Simulating and predicting ecosystem changes using a model ensemble
Co-Is: Blanchard, Blackwell (Sheffield), Heath (Strathclyde), Heymans (SAMS), Rossberg, Mackinson (Cefas),
Wanless (CEH)
In this module we build the capacity in modelling to support predictions of ecosystem states and services.
There are many different ways in which complex ecosystems are conceptualised for empirical analyses and
development into models. By combining the ecoinformatics toolbox and new data collected (Module 1,2) at
different scales with these existing ecosystem models and their variants (Module 3) we will assemble a
whole ecosystem model ensemble. We will advance the predictive capacity of marine ecosystem models
beyond the state-of-the-art by developing a system for utilising the ensemble of models to deliver
assessments of the uncertainty in ecosystem forecasts (and hindcasts) analogous to the ensemble approach
employed by the IPCC to make prognoses of future climate change. We will objectively select models and
sub-models based on structural uncertainty analyses, their performance from elasticity analyses and
capacity to predict macroecological patterns, reducing key uncertainties of existing models and their
predictions, invigorating models with new process-based understanding and evaluating the appropriateness
and utility of specific models at different temporal and spatial scales.
Statistical inference methods will be used to establish an ensemble of ecosystem models and quantify the
uncertainty in predictions arising from structural differences within and among models (Gardmark et al.
2013, Purves et al. 2013). These methods are widely and routinely used in climate and environmental
modelling (e.g. Chandler 2013; Perkins et al. 2013, Hoeglind et al. 2013). Using the ensemble approach, this
Module will provide a systematic account of the uncertainty in predictions of ecosystem states under given
forcing, arising from the inherent structural uncertainty in models. Crucially, we will also provide tools to link
ecosystem properties and dynamical processes to ecosystem services.
6.4.1 Tasks
T4.1 Whole ecosystem model ensemble
Marine Ecosystem Research Programme Part B
Ecosystem ecology, food web ecology and macroecology have all influenced the modelling tools that are
currently being developed and applied to marine ecosystems around the UK. Our models (Box 1) have been
selected on the basis of representing state-of-the-art in ecosystem modelling, their computational efficiency,
their application to real ecosystems, our expertise, and their availability as open source code. EwE,
StrathE2E and size spectrum models already have already been interfaced with ERSEM and work is
Box 1: Model ensemble members
1) Ecopath with Ecosim (EwE) (Christensen &Walters 2004) incorporate diet and biomass data from fisheries and
oceanographic datasets and the literature to carry out a mass-balance network analysis (Ecopath), to simulate
dynamical changes in biomass of functional groups or species through time (Ecosim) and space (Ecospace) at a
regional scale. Models available include West of Scotland (Bailey et al. 2011), Celtic Sea (Lauria 2012) and Irish
Sea (Lees & Mackinson 2007).
2) StrathE2E (Heath, 2012) simulates the fluxes of nutrients (nitrogen) through ecosystems from dissolved
inorganic (nitrate and ammonia), through plankton, benthos and fish, to birds and mammals, regeneration
through excretion and mineralization of detritus in the water column and sediment and physical exchanges
across geographic boundaries. Models available include the North Sea, English Channel, Celtic Sea, Irish Sea and
west of Scotland (EU-FP7 project BASIN).
3) The Population-Dynamical Matching Model (PDMM, Rossberg et al. 2008) combines community assembly
and species biomass dynamics via body mass foraging/vulnerability traits and relative adult body masses. Recent
model applications include the Celtic Sea and the North Sea to study the Large Fish Indicator (Fung et al. 2013).
4) Size-spectrum models for pelagic (Law et al. 2009), coupled benthic-pelagic (Blanchard et al., 2009, 2011,
2012) and trait-based (Hartvig et al. 2011, Rossberg 2012)) communities predict changes in abundance across a
continuum of body masses through time and space (Castle et al. 2011). Models available include the North Sea,
Celtic Sea, deep waters off the West of Scotland (EU-FP7 Deepfishman) and global shelf seas (Blanchard et al.
2012, NERC Quest-Fish).
5) FishSUMS is a size-structured model that includes a suite of species and functional groups (Speirs et al. 2012).
Populations of 10-15 species are represented by a series of discrete length classes from egg to adult as well as
benthic and zooplankton resources. FishSUMS has been parameterised for the North Sea with an
implementation for the west of Scotland in progress.
6) Simple consumer resource models(e.g. Murdoch et al.2003) that follow energy pathways through focal
groups of species rather than whole webs will also be employed (particularly for studying top predators and
through development in Module 3).
ongoing.
Outputs of all models include time series of biomass, changes in mortality, food consumption and diets,
estimated for each functional group/species and catch for each group fished. Existing links to Marine
Strategy Framework Directive include the capacity of the models to output existing and proposed indicators
for GES. The models can all output macroecological abundance-body mass patterns, either as size spectra,
species abundance/biomass-mean body mass or trophic level directly or through conversion of relationships
between mean and variance of species body mass and/or trophic levels. Building on work in Modules 1 and 3
we will develop a standardised suite of model variants to investigate and quantitatively assess structural
uncertainty within and across these models (see also tasks 4.2, 4.3). We will combine model code with data
and develop data assimilation methods as part of the ecoinformatics toolbox (Module 1). We will also
engage with the international community and provide opportunities for inclusion of other models into the
ensemble (see letters of support).
Dependencies: T1.1, T1.2
Marine Ecosystem Research Programme Part B
T4.2 Bayesian data assimilation for ecosystem models
Although a subset of the models (StrathE2E, EwE and size spectrum models) have been formally subjected to
parameter optimisation to fit them to a suite of observational data, only maximum likelihood or least
squares estimation has been used and the nature of the model fitting algorithms has not been formally
evaluated. We will extend statistical parameter estimation to other models in our ensemble within a
Bayesian framework.
These methods require very high repetition of model simulations to explore multi-dimensional parameter
space, so computational efficiency is of paramount importance. The different models represent the system
differently, e.g. only as particular species or only as bulk biomass of groups or size classes. Hence the
observational data (Module 1) will need to be aggregated in different ways to compare with the different
models. Then, we will use a simulation-based algorithm such as Markov Chain Monte Carlo, particle filtering,
or a hybrid of both (Dowd 2007; Golightly & Wilkinson, 2011; Mortier, 2013; Weir et al., 2013), Approximate
Bayesian Computation (Sunnaker et al, 2013), or a combination of these. This will give us fully Bayesian
multi-model inference, while allowing us to take into account suitable prior information where available and
appropriate.
Dependencies: T1.2, T1.3,
T4.3 Simulations of past ecosystem responses at different spatial and temporal scales
Using information on parameter sets from the empirical work (Modules 1,2) we will ensure priors are
developed using biological realism (within metabolic and physiological constraints). Traits-based analyses
will benefit species/groups for which parameters are not available (Modules1,2). We will test the
performance of models at different spatial and temporal scales. Each model and their selected variants will
be fitted to data at one spatial/temporal scale (e.g. whole Celtic Sea and/or annual production rates) and
cross-validated using data from another spatial/temporal scale (e.g. L4 and/or seasonal production rates).
This will enable us to evaluate and rank the performance of the models and their variants across spatial
scales. This will allow further model testing and selection of the type of future scenario projections to be
made with each respective model. Model variants will incorporate alternative formulations based on new
information derived from MERP Modules 1, 2, and 3. We will specifically use data on macrophyte-derived
subsidies to the food web (EwE, Strathe2E, size spectrum models), detrital and predator bentho-pelagic
coupling (all models), complete regional and seasonal biomass structure (all, especially size spectrum
models), trophic level information (all models), role of functional diversity such as gelatinous zooplankton
versus fish larvae (all models), density-dependence and top predator foraging traits (all models), systemlevel responses to pressures (all models). The ecoinfomatics and macroecology (Module 1) will synthesise
these data with existing understanding to provide fundamental scaling considerations, for example on
functional feeding relationships, growth and mortality with temperature and body size, and provide data in
the form required here. Specific variants to include in the ensemble will be determined by linking Modules 14.
Dependencies: T1.1, T1.2,T1.3, T2.1, T2.2, T2.3, T3.1, T3.1, T3.3
4.4 Predictions of future ecosystem responses and services
We expect that models will respond differently to top-down and bottom-up perturbation scenarios due to
their different structures and processes. We will test this via a series of carefully constructed model
experiments utilising well-established, simple but well specified scenarios. These will capture a range of both
bottom-up and top-down pressures and we will do full factorial experiments to examine responses to single
and multiple drivers of change. They will include but not be limited to: (i) ocean warming, (ii) multispecies
fishing (synergies with the EU-FP7 MYFISH and BASIN), (iii) changes in nutrients from runoff (iv) changes in
the species, size and abundance of primary producers, (v) key species replacement at mid trophic levels, for
example increases in gelatinous zooplankton, (vi) temperature-related reductions in body sizes (vii) spatial
closures for example by Marine Protected Areas or fisheries management. Warming and nutrient scenarios
will use forecasts carried out under IPCC scenarios from POLCOMS-ERSEM (EU-FP7 BASIN). Predictive model
Marine Ecosystem Research Programme Part B
outputs will include those specified in Module 5 to quantify uncertainty in ecosystem services.
6.4.2 Milestones
No.
M4.1
M4.2
M4.3
Milestone
Specification of the alternative models and their driving and
fitting data that will form the ensemble and the baseline
runs
Comparative assessment of the baseline outputs from the
model ensemble as the basis for quantifying uncertainty
Predictions of the response of ecosystem states and services
in response to climate and anthropogenic drivers
Date completion
(M10)
Lead partner
Sheffield
(M20)
Sheffield
(M30)
Sheffield
6.4.3 Deliverables
No.
D4.1
D4.2
D4.3
D4.4
Deliverable
Report on model fitting methodology and results
Paper on model skill specifying the selected model
ensemble
Report on ensemble predictions of ecosystem
responses and services across scenarios.
Policy report on ecosystem service predictions and
uncertainty.
Date
completion
(M18)
(M24)
Dependencies
Lead partner
M4.1, M4.3
M4.2, M4.4
Sheffield
Sheffield
(M36)
M4.5
Sheffield/PML
(M42)
M4.5
PML/Sheffield
Links to other Modules This module will rely on inputs from M1, 2 and 3, and provide outputs to M5. Subprocesses and structures to include into models will be identified through empirical macroecological and
food web analyses (M1), dedicated experiments and observations (M2) and elasticity analyses (M3).
Methods for translation of model outputs into ecosystem services will be developed for M5 and integrated
into the models.
6.5 Module 5: Linking macroecology and models to ecosystem services.
Co-Is: Austen, Beaumont (PML) and all MERP Co-Is
The inclusion of ecosystem services within marine management and regulatory frameworks is limited by a
lack of understanding of the relationships between ecosystem structure and processes and how they
integrate to provide ecosystem services, and of the spatial and temporal scales that are key to these
relationships (Mace et al 2011, Austen et al 2011). As lead authors in the National Ecosystem Assessment
and lead researchers in projects such as EU VECTORS and UKERC we have already made significant progress
in identifying and understanding what services are delivered by marine ecosystems, the underlying functions
and processes and, in particular, ecologically-based indicators for provisioning and regulating services (e.g.
Austen et al 2011, Hattam et al submitted). Improved mechanistic understanding gained in MERP (Modules
1-4) will be used to improve understanding of the processes and structures that are relevant to ecosystem
services. This will enable us to translate outputs from macroecological analysis (Module 1) and predictions
derived from the model ensembles (Module 4) indicative of changes in these processes and structures into
the currency of ecosystem services. Bringing together natural scientists together with interdisciplinary
ecosystem service researchers working between and among natural and social science disciplines, we will
develop a framework to generate predictions about the fate of marine ecosystem services under different
past and future scenarios, at local and regional spatial scales, allowing exploration of the implications of
different management scenarios for ecosystem services and their socio-economic benefits (Programme
Objective 3.)
6.5.1 Tasks
T5.1 Translate understanding and predictions derived from macroecological and empirical approaches and
model ensembles into the currency of ecosystem services.
Marine Ecosystem Research Programme Part B
Ecosystem services that will be addressed are climate regulation (carbon flows and stocks); wild species
biodiversity; food provision (support for fish and shellfish stock and structure); and bioremediation and
detoxification of waste, as well as some elements of disturbance prevention or moderation, regulation of
water flows and coastal erosion prevention (marine aspects of hazard regulation). It is clear that both higher
trophic level organisms (e.g. seabirds and mammals) as well as large megafauna at lower trophic levels
(basking sharks) are important for the cultural services of leisure, recreation, tourism and aesthetic
experience which also provide health benefits (White et al 2010, 2012; Wheeler et al 2012) and MERP will
also address these aspects of cultural service provision.
Module 5 will utilise the MERP Stakeholder Panel to ensure that the research outputs are relevant. An
Integrating Workshop at the Kick-Off Meeting will be used to explain our approaches, including their
limitations and uncertainties, and refine them to ensure that they will have utility. A second workshop at the
end of the project will explain the outcomes and then explore their implications for current policy and
management.
Building on e.g. Austen et al (2011), Hattam et al (submitted) and Worm et al (2006) preliminary conceptual
models will be developed that link different aspects of biodiversity and ecosystem processes to each of the
ecosystem services. Such models are already being developed via UKERC and BESS support for carbon
elements of climate regulation based on coastal shelf carbon flows in ERSEM, and for bioremediation of
waste and greenhouse gas sequestration in estuaries. The understanding derived from macroecological,
empirical and model ensemble approaches (Modules 1-4) as well as their outputs (for parameterisation) will
be used to further develop these conceptual models. A variety of processes will be addressed for each
service. For food provision, for example, these include primary production, maintenance of food web
dynamics, nutrient cycling to maintain food web dynamics for target species, supply of larvae and gametes of
target species and rovision of suitable habitats. For climate regulation, processes include pelagic and
benthic fixation of carbon through photosynthesis, C storage over time in living biomass (e.g. kelp, fish,
mammals, benthic organisms, microbiota etc.), deposition and burial of carbon in seabed sediments through
bioturbation.
Through statistical analysis of model parameters we will define the key processes and biodiversity elements
that are essential to ecosystem service delivery and good indicators of changes in delivery of the different
ecosystem services, and at which spatial and temporal scales. We will also identify which current indicators
of GES can indicate change in ecosystem services. We recognize that neither the empirical results of Modules
1 and 2, nor the theoretical and modelling results of Modules 3 and 4, can directly represent all structures
and processes relevant for major ecosystem services. This will necessitate a triage of all marine ecosystem
services, identifying those directly represented by models (such as harvestable fish production), those for
which proxy indices can be defined that can be computed from the model data (e.g. for climate regulation
and for waste treatment and assimilation) or extracted through ecoinformatics approaches (e.g. wild species
diversity), and those that are inaccessible to our methodology (e.g. some cultural services). We will thus first
identify which are the key model and data outcomes of MERP that can be used to indicate changes in
ecosystem services. We will then use the model and data outputs (from Modules 1-4) to identify how
ecosystem services change in response to changes in food web structure, and generate predictions of
change in marine ecosystem services at local and regional spatial scales under different past and future
scenarios.
Dependencies: Module 5 will rely on the outputs for all modules to feed into their work.
6.5.2 Milestones
No.
M5.1
M5.2
M5.3
M5.4
Milestone
Integrating workshop with consortium and stakeholders
Develop conceptual models for different ecosystem services
Macroecological and ecosystem models linked to services
Final workshop
Marine Ecosystem Research Programme Part B
Date completion
(M6)
(M24)
(M36)
(M48)
Lead partner
PML
PML
PML
PML
6.5.3 Deliverables
No.
Deliverable
D5.1
Paper on conceptual models relating ecosystem structure
and processes to ecosystem services
Paper on analysis of changes in ecosystem services at
different spatial and temporal scales
D5.2
Date
completion
(M30)
(M42)
Related
Milestone
M5.2
Lead partner
PML
M5.3
PML
Links to other modules M5 will use data from M1 and M2 at different scales, mechanistic
understanding of processes from M1-3 and simulations of past and future ecosystem responses across
spatial and temporal scales from M4.
6.6. Module 6
6.6.1 Tasks
T6.1. Technical development of the software environment for the hierarchical model. We will establish a
version controlled development trunk for DivERSEM. Using a modular approach based on the standard
organism the code will be designed to provide a scalable model system, with a traceable hierarchy
whereby more complex foodweb structure can by systematically and coherently related to simple
foodweb structures. Funded though NC (PML).
T6.2. Evaluation of the ecosystem properties of the existing ERSEM configuration. To inform both the
model development and WP1 we will evaluate and existing hindcast of the NW European Shelf, in terms
of the model reproduction of observed size spectra, metabolic scaling,
observed ecosystem variability and shifts in trophic control. The focus will be on data rich regions
(PML, Cefas).
T6.3. Computational cost of advection. The horizontal advection of biological variables is a substantial
computational cost. This task involves improving the capacity of the model to scale and capability to
work with 10’s to 100’s state variables. Firstly we will optimise, for 10s-100s state variables, existing
schemes in NEMO (TVD and PPM) that are accurate, monotonic and non-diffusive. Then we will explore
the implications (on efficiency/accuracy) of combining these schemes with substantially cheaper but
lower fidelity methods (e.g. upwind and centred), using a master variable/subsidiary variable approach
and test these approaches in a frontal resolving NEMO model (~1.8km). Finally we will review and assess
other advanced transport methods, such as those used in multi-tracer atmospheric simulations (e.g.
Lauritzen et al 2010) (NOC).
T6.4. Revising the ERSEM standard organisms. The purpose of this task is to improve the trophic structure
in terms of organism size and function through revisiting the definitions of the standard organisms
combining pelagic and benthic approaches as appropriate, improving the individual process descriptions
based on, allometric scaling and metabolic theory where possible to define a baseline set of size scalable
core parameters. This will allow us to address whether size and function based relationships can describe
macroecological relationships and explain the flow of energy though the ecosystem. The performance of
this model will be evaluated against the standard SSB ERSEM model in 1D water column models.
• Autotroph’s: Analysis of nutrient-dependent growth and uptake in phytoplankton, reveals
physiological trade-offs in species abilities to acquire and utilize resources (Litchman et al
2007). Such trade-offs, arise from fundamental relations such as cellular scaling laws and enzyme
kinetics and define contrasting ecological strategies of nutrient acquisition and the subsequent
Marine Ecosystem Research Programme Part B
modulation of cellular stoichiometry with size (Finkel et al., 2010). Cellular stoichiometry provides a
link between bottom up control (availability of light and nutrient) and top down control in the
planktonic food web (zooplankton stoichiometric modulation of predation). Starting with the
standard phytoplankton organism from the SSB ERSEM model and drawing on the published analysis
of size based traits (e.g. Litchman et al 2007, Finkel et al 2010, Edwards et al 2012) the
parameterisation of the standard autotrophic organism will be refined to better represent these
tradeoffs. The applicability of the standard autotroph to
the benthic system (as developed in Blackford, 2002) will be further developed and evaluated. (PML).
• Heterotrophic Consumers: We will refine the pelagic and benthic foodweb, focusing on the
allometeric scaling of growth, ingestion and respiration to better resolve the observed size
structures. (Cefas, PML).
• Decomposers (Heterotrophic prokaryotes) Heterotrophic prokaryotes (bacteria and archaea)
play a crucial role in both the benthic (Lloyd et al., 2013) and pelagic systems (Carlson et al.
2007). ERSEM is being developed by including (and further refining) the formulation proposed by
Polimene et al. (2006). This describes the release of excess of carbon (overflow metabolism) which
decouples the uptake of carbon from cell growth (i.e. “net” production),
thus making the variability of Bacterial Growth Efficiency as a function of the environmental
conditions. We shall draw on this work to develop the basic standard decomposers for both the
benthic and pelagic systems. Further work will be undertaken to improve the parameterisations of
aerobic and anaerobic bacteria in the benthic system.
• We will explore the top closure of both the pelagic and benthic models starting with a simple
quadratic term, but also considering the external closure from and offline HTL model. In addition we
will make data available for 1 way coupling to drive external HTL models being applied in WP1.
Depending on the nature work proposed by the modelling activities in WP1 we will explore the
collaborative development of 2 way couplers with WP1 if required. (Cefas, PML).
T6.5 Biological traits. To further improve our capability to reproduce the observed spatial and temporal
heterogeneity in ecosystem structure this task focuses on improving the description of size / functional
class diversity, and within trait diversity. Although size is the major driver for many physiological and
behavioural parameters, other traits are not size-dependent and they can significantly contribute in
determining the fitness of the organisms and therefore potentially contribute to the general ecosystem
structure. We shall focus on expanding the biological trait diversity of autotrophs, zooplankton and
zoobenthos, and explore how these groups drive the energy supply into the ecosystem and its transfer to
higher trophic levels.
• Autotrophs: Diatoms play a crucial role in determining the ecosystem properties of both pelagic and
benthic ecosystems, the latter through the supply of fast sinking detritus in the spring. The current
version of ERSEM describes one category of diatoms (defined as silicate consumers). We will focus on
differentiating the size classes of diatoms. In addition we will parameterise traits in benthic
autotrophs (e.g. benthic diatoms and macroalgae). (PML, Cefas).
• Zooplankton For the heterotrophic grazers, the strength of the predator–prey interactions, are driven
by the choice of parameters and more specifically the food preferences (Sailley et al
2013). This illustrates the importance of equation and parameter choice as they define
interactions between FTs and overall food web dynamics (competition, bottom-up or top-down
effects). Related to this is the issue of the relationships between key processes such as prey
selectivity, ingestion and digestion to the amount of prey, its quality and its type (Mitra and Flynn,
2005). Drawing on work being undertaken in the SSB program, the description of zooplankton grazing
activity will be refined by including processes such as the stoichiometric modulation of predation
(Mitra, 2006; Mitra and Flynn, 2005) and prey palatability (i.e. production of toxins), which may heavily
contribute to promote algal blooms in coastal eutrophic areas (Mitra and Flynn, 2006). In addition they
Marine Ecosystem Research Programme Part B
exhibit a number of foraging behaviours ranging from ambush feeding (waiting for random prey
encounters), to cruise feeding (increasing prey encounters by swimming) and current feeding
(increasing prey encounters by generating a feeding current). Each strategy has a different set of
penalty costs, trading prey availability against metabolic costs of movement and increased risk of
predation. The current version of ERSEM describes zooplankton by size (e.g. micro, meso) but does not
take account of foraging behaviours. We will focus on differentiating the different feeding strategies by
initially differentiating between ambush feeders and filter feeders (PML).
• Zoobenthos: In an analogous manner to the zooplankton above, we shall focus on refining the benthic
foodweb, primarily by defining the macro fauna in terms of feeding and burrowing strategies. (PML,
Cefas).
• Adding species diversity: To fully explore the impact of changes in diversity on the resistance and
resilience of ecosystems requires representation of intra trait diversity. This trait variability will be
included in the model by randomly varying the value of some of the parameters of the modelled
organism (e.g. growth rate, carbon: nutrients ratio, food preference, clearance rate) based on the
variability of the allometric relationships observed in the literature. The variability of ecosystem
properties depending on these parameters will be compared to that one depending on size in order to
assess how much of the observed patterns can be explained by the size, and how much the other
traits contribute (PML).
T6.6. Simulations and synthesis: The DivERSEM model will be used to explore the importance of biological
traits in defining major ecosystem properties like diversity, primary and secondary production, energy
transfer efficiency and their temporal and spatial patterns.
• GOTM water columns with be set up for data rich sites (e.g. L4, Stonehaven, Cefas Oyster Grounds)
and validated. Sensitivity analysis will be performed to explore natural variability, trophic controls
and diversity/function relationships (PML, Cefas).
• A 40 year hindcast of the NW European Shelf will be made, exploiting the common model forcing data
base setup by the SSB project. This hindcast will be assessed in terms of its skill in reproducing macroecological properties (PML, NOC).
• We will also explore the effects of refining model resolution (from ~7km to ~1-2km), so as to better
represent the structure and dynamics of shelf sea fronts and investigate the consequences for
ecosystem structure and function (e.g. Holt et al 2004). For this we will exploit developments of fine
resolution NEMO models in FASTNet (NERC ocean –shelf exchange RP), but on a reduced area nested
domain covering the Celtic Sea, English Channel and Irish Sea (see Holt and James, 2006). The 3D
simulations will be evaluated using large spatial data sets e.g. Satellite Ocean Colour (NEODASS) and
the CPR. (NOC)
6.5.2 Milestones
No.
M6.1
M6.2
Milestone
Delivery of new model code with improved representation of
biodiversity
Finalized hindcast and re-analysis simulations
Date completion
(M30)
Lead partner
PML
(M36)
PML
6.5.3 Deliverables
No.
D6.1
D6.2
D6.3
Deliverable
Published version of the model code – submitted
manuscript
Community modelling tools available on the web
Paper(s) on new trait based models - submitted
manuscripts
Marine Ecosystem Research Programme Part B
Date
completion
(M24)
(M30)
(M30)
Related
Milestone
M6.1
M6.1
Lead partner
PML
PML
PML
D6.4
D6.5
D6.6
Paper on advanced advective methods - submitted
manuscripts
Hindcast simulations and re-analysis simulations, sub sets
of the data deposited with appropriate BODC data centre
Scientific paper(s) on the sensitivity and resilience of shelf
seas marine ecosystems, to climate change and other
anthropogenic drivers – submitted manuscripts
(M30)
NOC
(M36)
M6.2
PML
(M42)
M6.2
PML
7. REFERENCES
Andersen, K. H., & Pedersen, M. (2010). Proceeding of the Royal Society B, 277: 795-802.
Atkinson A (1996). Mar Ecol Prog Ser 130: 85-96.
Atkinson A, Meyer B, et al. (2002). Limnol Oceanogr 47:953-966.
Austen, MC, Malcolm SJ, et al. (2011). Marine. In: The UK National Ecosystem Assessment Technical Report.
UNEP-WCMC.
Bailey, N., D. M. Bailey, et al. (2011). Marine Scotland Science Report 0911. Aberdeen, Marine Scotland: 288.
Barrett, R.T. et al. (2007). ICES J Mar Sci, 64, 1675-1691.
Baum, J. & Worm, B. (2009). J Anim. Ecol.. 68, 299
Beaugrand, G., Luczak, C., and Edwards, M. (2009). Global Change Biology 15, 1790
Beck, J. et al. (2012). Ecography, 35, 1-11.
Bekkby, T., E. Rinde, L. et al. (2009). ICES J. Mar. Sci. 66:2106-2115.
Benoit-Bird, K.J. and McManus, M.A. (2012). Biology Letters, 8(5): 813-816.
Blanchard, J. L., Jennings, S., et al. (2009). Journal of Animal Ecology, 78(1), 270-80. doi:10.1111/j.13652656.2008.01466.x
Blanchard, J. L., Jennings, S., et al. (2012). Philosophical transactions of the Royal Society of London. Ser B,,
367(1605), 2979-89.
Blanchard, J. L., Law, R., et al. (2011). Theoretical Ecology, 4(3), 289-300.
Borer, E.T., Seabloom, E et al. (2005). Ecology, 86, 528-537.
Brose, U. et al. (2008). J Anim Ecol, 77, 1072-1078.
Brown, J. et al. (2004). Ecology, 85, 1771-1789.
Burrows, M. T. (2012). Marine Ecology Progress Series 445:193-207.
Byrnes, J. E., D. C. Reed,. (2011). Global Change Biology 17:2513-2524.
Casal, G., N. Sánchez-Carnero, et al. (2011). Estuarine, Coastal and Shelf Science 91:371-378.
Castle, M. D., Blanchard, J. L., & Jennings, S. (2011). Advances in Ecological Research, Volume 45, 41-66.
Chandler, R.E. (2013). Phil. Trans. R. Soc. A 371, 20120388.
Chave, J. (2013). Ecol Lett, 16, 4-16.
Christensen, V. & C. J. Walters (2004). Ecological Modelling172: 109-139.
Cohen, J.E., Jonsson, T. & Carpenter, S.R. (2003). PNAS, 100, 1781-1786.
Cury, P., Bakun, A., et al. (2000). ICES Journal of Marine Science: Journal du Conseil, 57, 603-618.
Dell, A. I., S. Pawar, and V. M. Savage (2011). PNAS 108,10591-10596.
Marine Ecosystem Research Programme Part B
Dethier, M. N., E. Sosik, et al. (2013). Marine Ecology Progress Series 478:1-14.
Diaz, R.J. & Rosenberg, R. (2008). Science 321, 926
Dowd, M. (2007).Journal of Marine Systems 68, 439‚-456.
Dulvy, N.K. et al. (2008). Journal of Applied Ecology 45, 1029
Emmerson MC, Rafelli D (2004) J Anim Ecol 73: 399-409.
Evans, P.G.H. and Raga, J.A. (Eds) (2001) Marine Mammals: Biology and Conservation. Kluwer
Academic/Plenum Press.
Evrard, V., M. Huettel, et al. (2012). Marine Ecology Progress Series 455: 13-31.
Frederiksen, M. et al. (2013). Global Change Biol, 19, 364-372.
Fung, T., Farnsworth, K. D., et al. (2013). Marine Ecology Progress Series.doi:10.3354/meps10305.
Gårdmark, A., Lindegren, M., et al. (2013).Ecological Applications, 23(4), 742-754.
Golightly A. & Wilkinson D. (2011) Interface Focus 1, 807-820.
Gross, T., Ebenhöh, W., & Feudel, U. (2004). Journal of Theoretical Biology, 227, 349-358.
Halpern, B. S. et al. (2008). Science 319, 948
Hampton, S.E. et al. (2013). Frontiers Ecol Env, 11, 156-162.
Hardison, A. K., E. A. Canuel, et al. (2010). Marine Ecology Progress Series 414: 41-55.
Hardisty, A. & Roberts, D. (2013) BMC ecology, 13, 16.
Hartvig, M., Andersen, K. H., & Beyer, J. E. (2011). Journal of Theoretical Biology, 272(1), 113-122.
Heath, M.R. (2012). Progress in Oceanography 102, 42-66.
Heath, M.R., Speirs, D.C. and Steele, J.H.(submitted). Ecology Letters.
Herendeen, R.A. (1995 J. Theor. Biol., 176, 13-26.
Herendeen, R.A. (2004) Ecol. Model., 171, 21-33.
Hiddink, J.G., et al. (2011). Journal of Applied Ecology,
Hill, R. C., Griths, W. E., Lim, G. C. (2008). Principles of Econometrics, 3rd Edition. Wiley, New York.
Hirst AG, Bonnet D et al. (2004) Mar Ecol Prog Ser 340: 189-205
Hoegh-Guldberg, O. & Bruno, J.F. (2010). Science, 328, 1523
Hoeglind, M., Thorsen, S.M., Semenov, M.A. (2013). Agricultural and Forest Meteorology 170, 103-113.
Hosack, G. R., K. R. Hayes, et al. (2008). Ecological Applications 18(4): 1070-1082
Isojunno, S., Matthiopoulos, J., and Evans, P.G.H. (2012). MEPS 448, 155-170.
Jetz, W. et al. (2004). Science, 306, 266-268.
Keith, et al. (2012). Biology Letters 6, 904-906.
Kerr, J. T., Kharouba, H. M., & Currie, D. J. (2007). Science, 316, 1581–1584
Kiørboe, T. & Hirst, A.G. (submitted). Am Nat
Kiørboe, T., H. Jiang, et al. (2010). Proceedings of the Royal Society B: Biological Sciences 277, 3229-3237
Krumhansl, K., and Scheibling, R. E. (2012). MARINE Marine Ecology Progress Series 467:281-302.
Lassen, H., S. A. Pedersen, et al. (2013). ICES Journal of Marine Science 70, 471-479
Lauria, V. (2012).Impacts of climate change and fisheries on the Celtic Sea ecosystem.School of Science and
Marine Ecosystem Research Programme Part B
Technology. Plymouth, Plymouth University. PhD: 248.
Law, R., Plank, M. J., et al. (2009). Ecology, 90(3), 802-811.
Lees, K. and S. Mackinson (2007). An Ecopath model of the Irish Sea: ecosystems properties and sensitivity
analysis. Lowestoft, Cefas: 49 || Mace, G. M., I. Bateman, et al. (2011). Conceptual framework and
methodology. The UK National Ecosystem Assessment Technical Report. UK National Ecosystem
Assessment, UNEP-WCMC, Cambridge. || Mann, K. H. (1973). Science 182, NO 4116, P 975-981. ||
McCann, K.S., Hastings, A. & Strong, D.R. (1998) Proc. R. Soc. Lond. B, 265, 205-209. || McLaughlin, Ó.,
Jonsson, T. & Emmerson, M.C. (2010). Advances in Ecological Research, 42: 171-264. || Michener, W.K. &
Jones, M.B. (2012) TREE, 27, 85-93. || Mortier, Frederic; et al. (2013). Methods in Ecology and Evolution 4,
316-326. || Murdoch W.W., Briggs C.J., Nisbet, R.M. (2003). Princeton Monographs. 456 pp. ||
O'Gorman, E.J. & Emmerson, M.C. (2010). Advances in Ecological Research, 42: 301-419. || O'Gorman, E.J.
and Emmerson, M.C. (2011). Oikos, 120: 520-528. || O'Rorke, R., Lavery, S., et al. (2012). PloSone, 7 (8)
e42757 || Pawar, S., Dell, A.I. & Van M Savage. (2012). Nature, 486, 485-489. || Perkins, S. E.; Pitman, A.
J.; Sisson, S. A. (2013). International Journal of Climatology 33, 1153-1167. || Peterson, A.T., Knapp, S., et
al. (2010). Systematics & Biodiv, 8, 159-168. || Pinnegar, J. K., S. Jennings, et al. (2002). Journal of Applied
Ecology 39(3): 377-390 || Pinsky, Ml. et al. (2011). PNAS 108, 8317 || Purves, D. (2013) Nature
(Comment).493: 295. || Qian, C., & Giles, D. E. (2007). Economics Letters, 94, 185-191. || Rall, B. C., O.
Vucic‐Pestic, et al. (2010). Global Change Biology 16, 2145-2157 || Richardson AJ, Bakun A et al. (2008)
Trends in Ecol Evol. 24: 312-322 || Rooney, N., McCann, K., Gellner, G., & Moore, J. C. (2006). Nature,
442(7100), 265-269. || Rossberg, A. G. (2012). Advances in Ecological Research, 46, 429-522. || Rossberg,
A. G. (2013). Food Webs and Biodiversity: Foundations, Models, Data. Wiley, ISBN 9-780470973-55-4 ||
Rossberg, A. G., Ishii, R., et al. (2008). Ecology, 89(2), 567-580. || Saether, B.-E., Engen, S. & Grotan, V.
(2013). J Anim Ecol, 82, 731-738. || Schmidt K, Atkinson A, et al. (2003). Limnol Oceanogr 48: 277-289. ||
Searle, K. R., Hobbs, N.T. and Jaronski, N.T. (2010). Oecologia 163, 815-824. || Smith et al. (2011). Science
333, 1147 || Shurin, J. B., Borer, E. T., et al. (2002). Ecol. Lett., 5, 785-791. || Speirs, D.C., Guirey, E.J., et al.
(2010). A length structured partial ecosystem model for cod in the North Sea. Fisheries Research 106, 474494. || Storch, D., Keil, P., & Jetz, W. (2012). Nature, 488, 78-81 || Sunnaker, M., Busetto, A.G., et al.
(2013) PLOS Computational Biology 9(1): e1002803. doi:10.1371/journal.pcbi.1002803. || Thomas, L.,
Buckland, S.T., et al. (2010). Journal of Applied Ecology, 47: 5-14. || Trebilco, R. et al. (2013). TREE, in press
|| Twomey, M., Jacob, U., Emmerson, M.C. (2013). Advances in Ecological Research, 47: 349-409. || Tyler,
E. H., P. J. Somerfield, et al. (2012). Global Ecology and Biogeography 21, 922-934 || Wakefield, E.D. et al.
(2013). Science DOI: 10.1126/science.1236077. || Walker, F. T. (1953). J Conseil:160-166. || Ware, D. M.,
and Thomson, R. E. (2005). Science, 308, 1280-1284. || Webb, T.J. (2012) . TREE, 27, 535-541. || Weir B.,
Miller R.N., Spitz Y.H. (2013). Bull Math Biol 75, 223‚-257. || Wheeler B, White MP, Stahl-Timmins W,
Depledge MH.(2012). Health & Place;18:1198-1201. || White MP, Smith A, et al. (2010). Journal of
Environmental Psychology 30:482-493. || White, E.P. et al. (2013). PeerJ PrePrints
doi:10.7287/peerj.preprints.7 || White, E.P., Morgan Ernest, S.K., et al. (2007) TREE. 22: 323-330. ||
White, M. P., Pahl, S., et al. (2012). Feelings of restoration from recent nature visits: Analysis of the English
Monitoring Engagement with the Natural Environment survey (2009-2011). || Wirtz KW (2011). Marine
Ecology Progress Series 445, 1-12.
Appendix 1 Original Case for Support for WP2 Developing a model based
understanding of ecosystem service regulation
Part 1: Previous Track Record: See individual CVs for further details and relevant publications
Plymouth Marine Laboratory (PML) is an International Centre of Excellence in Marine Science
& Technology and a Collaborative Centre of NERC that carries out innovative and timely fundamental,
strategic and applied research in the marine environment from the uppermost reaches of estuaries to the
Marine Ecosystem Research Programme Part B
open ocean. The research at PML contributes to the issues of global change, sustainability and pollution
delivering solutions for national and international marine and coastal programmes. PML is an
independent, impartial provider of scientific research in the marine environment with a focus on
understanding biodiversity and ecosystem function, biogeochemical cycling, pollution and health, and
forecasting the role of the oceans in the Earth System. PML has an outstanding reputation at a national
and international level for its capabilities in the modelling of regional marine ecosystems. PML Principal
Investigator: Professor J. Icarus Allen (NERC band 4) is Head of Science for Modelling and Observing
Systems at PML, leader of PML’s ecosystem modelling NC and is an honorary visiting professor at the
University of Exeter. He has been involved in and acted as PI for over 35 national and EC projects, including
coordinating the FP7 MEECE and OpEc integrated projects. He is the PI for the modelling activities in the
Shelf Seas Biogeochemistry program, the Ocean Carbon theme of the National Centre for Ocean
Forecasting (NCEO), Integrated Marine Biogeochemical Modelling Network (i-MarNet). His primary
expertise is in marine ecosystem modelling and ecosystem response to climate change. He has 78 ISI
publications. Co-Investigator: Jerry Blackford (NERC band 4) is a Merit scientist at PML. His principle
interests are the impact of Climate Change and Ocean Acidification on marine systems (OA) and
environmental impact assessments for Carbon Capture and Storage (CCS). He leads a NERC consortium on
CCS impacts and a NERC-Defra funded project on regional climate /OA modelling. Research CoInvestigators (PML scientists): Dr Jorn Bruggeman in trait based modelling and Dr Sevrine Sailley expertise
on zooplankton and trophic interactions, Dr Yuri Artioli provides expertise in modelling marine plankton
diversity. Dr Luca Polimene, has expertise in pelagic process modelling and Dr Nick Stephens in benthic
modelling and nutrient cycling.
Cefas (Centre for Environment, Fisheries and Aquaculture Science) is a multi-disciplinary research centre
providing services in fisheries management, environmental monitoring and assessment, and aquaculture
to Defra and a large number of clients worldwide. Cefas has an ocean-going research vessel and a
monitoring buoy network. Cefas has published 755 peer- reviewed scientific papers in the past five years
in the fields of marine ecology, fisheries science, climate change, impacts of marine policy and
ecotoxicology. Cefas scientists hold senior advisory positions in organisations such as ICES and the EU, and
honorary positions in universities in the UK. Cefas staff have significant experience in developing empirical
and numerical models to study environmental and fisheries effects of management scenarios. Cefas has a
long history in R&D and providing scientific advice to a range of UK Government Bodies ensuring the
sustainable use of natural resources. Cefas currently participates in joint NERC-Defra funded projects
(Shelf Seas Biogeochemistry Programme and EBAO). Cefas Principal Investigator: Dr Johan van der Molen
(Cefas Pay Band 7) is a principal ecosystem modeller. He is Secretary for Coastal and Shelf Seas in the
Oceans Division of the European Geosciences Union, and a convener in its annual assemblies. He has over
20 years of experience in numerical modelling of hydrodynamics, sediment transport, morphodynamics,
transport of fish eggs and larvae, and biogeochemistry. Co-I’s Dr Sonja van Leeuwen (Cefas pay band 6) is
a senior ecosystem modeller with experience in primary production, eutrophication, coupling ERSEM to
size-based higher trophic level models, and river runoff and nutrient loads. Dr John Aldridge (Cefas pay
band 6) is a senior mathematical modeller with broad experience in physical and biological
marine systems, and specialising in benthic processes and phytoplankton and macroalgal
productivity. Other Cefas staff may contribute to the project where appropriate.
The National Oceanography Centre (NOC) is a NERC Research centre that maintains world class
oceanographic research, consisting of Liverpool (NOC-L formerly the Proudman Oceanographic
Laboratory; POL) and Southampton (NOC-S) sites. This work sits within the Marine Systems Modelling
(MSM) group, which has world-class expertise in high resolution ocean and Shelf Sea modelling of
hydrodynamics, waves and biogeochemistry, and uncertainty. At MSM at NOC-L has a substantial
international reputation and contributing to the development of operational shelf sea forecast services at
the UKMO (based on NEMO) and marine impact climate projections for UKCP09 and MCCIP. NOC
Marine Ecosystem Research Programme Part B
Principal investigator: Dr Jason Holt (NERC band 4) is associate head of the MSM. He specialises in the
synthesis of model and observations to develop our understanding of shelf-sea physical and coupled
physical-biological systems. He leads the modelling component in the NERC FASTNet (ocean-shelf
exchange) programme and the Next Generation Ocean Dynamical Roadmap Project. He has published 50+
peer reviewed papers primarily on the modelling of hydrodynamics and ecosystems in shelf seas and
ocean margins. Co-Investigators: Dr. Hedong Liu (NERC Band 5) works on the development and
applications of 3-D unstructured and structured grid ocean models. He is currently a member of NEMO
developer's committee and specialises in state-of-the-art numerical methods. Dr Sarah Wakelin (NERC
Band 6) specialises in the modelling of physical and coupled physics-ecosystem processes in shelf seas,
including the impacts of climate change and the combine impacts of climate and anthropogenic drivers.
Recent Related Work
As a component of the NERC-Defra Shelf Seas Biogeochemistry program, PML, Cefas and NOC in
collaboration with the UKMO are developing the ERSEM model system to address the overarching
scientific goal of enhancing our capacity to assess the physical, chemical and biological controls on
biogeochemical cycling, with a focus on the NW European Shelf. This will allow quantification with
uncertainties the budgets of C, N, P, Si including their response to climate, natural variability and
anthropogenic stress. The work undertaken in the SSB project focuses on merging the existing versions of
ERSEM used by the UK community into a single code. The process developments will focus on extending
biogeochemical function in the plankton foodweb (e.g. calcifiers, Phaeocystis), including plasticity of
response in zooplankton grazing due to food quality, extending the description of benthic and pelagic
nitrogen chemistry to include denitrification and N2O production, and extending the description of the
benthic model to describe non-cohesive advective sediments as well as cohesive sediments. The program
will develop, two core modelling tools based on ERSEM as community model system. These will be made
freely available for academic use. The first is a relocatable biogeochemical water column model based on
GOTM-ERSEM. This system can by run on a PC and provides an entry level modelling tool for non-specialist
modellers to engage in process modelling. This system will be made available to all SSB scientists. We see
developing better synergy between modelling and experimental studies as key to progress over the next
decade. The second tool is the high resolution coupled 3D hydrodynamic biogeochemical model based
ERSEM-NEMO-shelf, providing the basis of a community modelling approach for looking at spatially
resolved system response. In addition we will provide a library of model skill assessment tools. Model data
sets provide an important resource across the spectrum of oceanographic science. Examples include use in
habitat identification, planning field work effort, the provision of boundary and environmental conditions
for local studies and lab based experiments.
Part 2 – Case for Support
Background: Biodiversity refers to the variability among living organisms at all scales of organisation and
includes the diversity within species, between species, and of ecosystems. Marine ecosystems are
extremely diverse and the functional role this diversity provides underpins major ecosystem services for
example: food production, climate regulation through the cycling of carbon and other macronutrients, and
a range of cultural values (e.g. recreation and tourism). In many regions marine ecosystems are in serious
decline, primarily as a result of over-harvesting, pollution, and the direct and indirect impacts of climate
change. Combinations of direct anthropogenic stresses and natural variability have in some locations been
associated with dramatic shifts in species composition, known as phase or regime shifts, which maybe long
lasting and difficult to reverse (e.g. Lindegran et al 2012; Mollman et al 2009, Mackas and Beaugrand
2009). Changes in species composition affect fundamental ecosystem properties including biodiversity and
size structure. For example changes in the physical and chemical environment (temperature, circulation,
light availability, nutrients) mainly affect primary production and thus impact the foodweb through bottom
Marine Ecosystem Research Programme Part B
up control, whilst impacts such as harvesting act on top predators thus altering top down grazing control
(e.g. Frank et al 2006). However the relative roles and interactions of these processes and hence the
extent to which environmental change cascades through marine food webs and affects ecosystem services
is still poorly understood. In addition secondary impacts on key system determinants such as reproductive
success, benthic- pelagic coupling and phenology must be considered in order to get a full picture of
ecosystem resilience or vulnerability.
These bottom up, top down and internal processes are inherently scale-dependent. For example, the
physical structure of the water column (e.g. stratification, fronts) and the distribution of benthic habitats
affect the spatial distribution and abundance of all trophic levels. The large-scale removal of top predators
through fishing or other activities can have a range of impacts across scales. In the benthic system the
sediments are also dynamic and heterogeneous with differences in particle composition, size and shape.
These initial physical variations are diverse leading to large spatial heterogeneity on relatively small submetre scales. Subsequent variations observed in the biology are even greater and result in a relatively
complex and diverse community structure.
Scale-dependence is poorly understood, making it difficult to quantify the large-scale impacts on
ecosystem services of changes at small spatial scales (e.g. marine conservation zones); and vice versa.
Finally there is growing evidence that the loss of biodiversity from marine ecosystems can adversely
impact ecosystem function and hence the way marine food webs regulate ecosystem services. Studies
suggested that species loss decreases the productivity of communities and how efficiently they capture
and consume limited resources (e.g. Cardinale et al 2006). The interpretation of these studies has and
continues to provoke considerable debate, and subsequent work produced several counter examples that
questioned the generality of these biodiversity effects. Consequences of biodiversity loss are likely to be
idiosyncratic, differing quantitatively and qualitatively between trophic groups and ecosystems (Cornell
and Lawton. 1992). Changes in diversity though the arrival of non-indigenous species (taxa ranging from
phytoplankton to fish) introduced by both natural and human vectors also impact on biodiversity,
communities, habitats and ecosystem functioning and hence economic impacts. Their increasing spread is
a major concern for many world and European seas (Costello et al 2010, Coll et al2010).
Modelling challenges: Macroecology is the study of the relationships between organisms and their
environment at large spatial scales to characterise and explain statistical patterns of abundance,
distribution and diversity (Brown and Maurer 1989). Scaling relationships are common at higher levels of
biological organization, (e.g. Brown, 1995). Body size is an important determinant of many ecological
properties, ranging from abundance and biomass to physiological and ecological rates and provides
useful underpinning concept for characterizing and modelling marine ecosystems. The widely observed
macro-ecological patterns in log abundance vs. log body mass of organisms can be explained by simple
scaling theory based on food (energy) availability across a spectrum of body sizes. For example quarterpower scaling with body mass applies to virtually all organisms (West, Brown & Enquist, 1999) and for
marine animals, metabolic rate and production scales at three-quarters power (e.g. Brey, 1990; Warwick
& Price, 1979). Allometric equations have been derived for respiration, ingestion, excretion and
photosynthesis for a variety of organisms of different sizes (e.g. Verdy et al 2009, Moloney and Field
1989). Finally, the temperature dependence of gross primary production and community respiration is
consistent among the major types of ecosystems on the planet, suggesting that the fundamental
biochemical kinetics of respiratory metabolism are highly conserved and can be scaled from organism to
ecosystem levels (Yvon Durcher 2012).
Understanding the interactions of the various system controls, the consequences of change, and
designing, testing and refining potential management solutions, is important for the long-term delivery of
services from marine ecosystems. Dynamic models that link the physical, chemical and biological
processes through food web interactions provide a means of understanding how human impacts on
different parts of the ecosystem interact and of predicting the consequences of management actions.The
Marine Ecosystem Research Programme Part B
traditional approach to modelling marine plankton has generally been to build modelling frameworks by
coupling bulk biomass functional type (FT) models to 1D and 3D hydrodynamic models; The European
Regional Seas Ecosystem Model (ERSEM) is one such model. The ecosystem can be divided into four
sections, the pelagic cellular ecosystem, the pelagic mid trophic levels (zooplankton), the benthic
ecosystem (bacteria, meiofauna, zoobenthos) and the higher trophic levels (fish, birds, mammals); ERSEM
describes the first three.
The cellular ecosystem is represented as a biogeochemical flux model where the cell is a black box
whereby the fluxes of carbon or nutrients are transferred across the cell walls via first order rate
equations. Most often defined in terms of biogeochemical functions (e.g. diatom, non-diatom, calcifiers,
grazers etc.), the functional groups are treated as passive tracers and are very sensitive to the biophysical
environment they are placed in. The pelagic mid trophic levels are particularly important yet are a weak
component of all models. They are particularly important because they provide both the top closure for
the planktonic ecosystem and the link to the higher trophic levels. This part of the foodweb is poorly
represented in biogeochemical flux type models because the Eularian continuum approach becomes
inadequate as behaviour and individual history becomes important. Grazers do not eat biomass; they eat
individuals and the process is the sum of the interactions between large numbers of organisms. These
interactions are potentially dependent on combinations of prey density (function of the number of
potential prey), prey quality (nutritional status), prey type (species) and behaviour (e.g. vertical migration,
foraging strategies and defence mechanisms). The challenge is to represent both biomass and population
dynamics model in size / trait based foodwebs, coupled to biogeochemical cycles. The benthic system is
strongly influenced by the physical sediment structure, environmental conditions and organic fluxes to the
sediments. Benthic size spectra are more complex than their pelagic counterparts (Schwinghamer 2011);
three biomass peaks, are observed corresponding to bacteria, interstitial meiofauna, and macrofauna,
separated by low biomasses. Once again the requirement is to develop a size/trait based foodweb model
from bacteria to top predators. Only by viewing the water column and benthos as a whole that will
enable a more complete simulation and understanding of fish predating on benthic invertebrates,
macroalgae contributions and benthic ecosystem resilience through the modelling of larval and
reproductive abilities. Finally the Higher trophic levels are not a part of ERSEM, however the new model
will be designed to provide data products to facilitate 1 way coupling to a range of different models
ranging from dynamic size spectra and Ecopath with Ecosim to individual based models to further
investigate the different model ecosystem structures and the ecosystem dynamics represented.
Aims and Objectives
The overarching scientific goal is to enhance our capacity to assess trophic and spatial controls on the
structure of marine ecosystems through improving the representation of biodiversity and ecosystem
function in ERSEM. This enhanced model will be used to explore the impact of anthropogenic stress and
natural variability on the structure of marine ecosystems and its consequences for ecosystem services, in
order to better meet future science and policy needs. To achieve this we will develop a traceable hierarchy
of models of different complexity based on ERSEM using a modular approach with consistent process
descriptions based on size, function (autotrophy, heterotrophy, decomposition), and biological traits (e.g.
feeding strategy, motility, physiology). Through species-neutral process descriptions, the approach will
allow scalable representation of diversity within and between functionally similar “guilds” of species.
Building on the ERSEM model developments being undertaken in the Shelf Seas Biogeochemistry Program
we will establish an ecosystem focused version of ERSEM called ‘DivERSEM’. This will form the basis of a
new community model system which will be made freely available to the UK and international research
communities.
The project will be delivered through 2 phases of model development, testing and application: Phase I
(M1-36), will develop and test v1 of DivERSEM while Phase II (M36-54) will iterate and develop a second
Marine Ecosystem Research Programme Part B
version v2 of DivERSEM exploiting the new data and knowledge from the program. Both phases will be
underpinned by activities to develop and maintain core model infrastructure, funded through National
Capability at PML and NOC. The proposed activities in Phase I are described in this document; Phase II is
outlined and will be detailed towards the end of phase1 in response to progress within WP1 and the
needs of stakeholders.
Scientific questions: Underpinning this is a suite of key scientific questions, which provide a
framework for model development, simulation experiments and analysis.
• Can the model reproduce the observed emergent macroecological relationships (e.g. body size
relationships with biomass and rates) and help to explain how these relationships emerge?
• To what extent and at which spatial and temporal scales can size and function based
relationships describe and explain the flow of energy though the ecosystem.
• Does the introduction of selected biological traits other than size (e.g. feeding, reproduction,
physiology, motility) improve our capability to reproduce the observed spatial and temporal
heterogeneity (e.g. seasonality, inter-annual and inter-decadal variability and biogeography) in
ecosystem structure?
• Can changes in control on the ecosystem (top down and bottom up) explain observed spatial and
temporal heterogeneity (e.g. seasonality, inter-annual and inter-decadal variability and
biogeography)?
• How do the spatial scales of pelagic and benthic habitats of lower trophic levels (phytoplankton)
relate to those of mid (zooplankton) and higher trophic levels (zoobenthos and fish) and vice
versa?
• How do changes in inter and intra guild diversity (including invasion and extinction) impact on the
resistance and resilience of ecosystems and hence biogeochemical function.
Technical objectives: Alongside the scientific goals there are technical objectives, which are required
to deliver the model system.
• To construct a version controlled software environment for a traceable hierarchy based on size
and function – this will allow scaling for a simple size based plankton / zooplankton model to one
with stochastic functional variability.
• To provide model outputs to facilitate the 1 way coupling of ERSEM to other ecosystem model,
e.g. those of higher trophic levels used in WP1.
• To explore differing approaches to the transport schemes to balance the computational cost of the
model with the accuracy of the simulation of various components.
Models will be developed within the Framework for Aquatic Biogeochemical Models (FABM 1), produced
within FP7 project MEECE, to enable the rapid, run-time combination of any number of stochastically
initialized models into a coupled ecosystem. This coupler will enable the software to scale from a classic
ERSEM configuration to stochastic multi-species setups using a single, unified code base.
Existing Modelling tools
ERSEM was originally conceived as a generic model, which when coupled to a qualitatively correct physical
model, is able to simulate the principle dynamics of the marine system across eutrophic to oligotrophic
gradients of the range of marine environments from the global ocean to coast shelf seas (e.g. Blackford et
al 2004, Blanchard et al 2012). ERSEM simultaneously describes pelagic and benthic ecosystems in terms of
phytoplankton, bacteria, zooplankton, zoobenthos, and the biogeochemical cycling of C, N, P, Si (e.g.
Baretta et al., 1995, Ruardij et al.,1995, Baretta-Bekker et al., 1997, Blackford 1997, Blackford et al., 2004).
The ecosystem as described by ERSEM is considered to be a series of interacting physical, chemical and
Marine Ecosystem Research Programme Part B
biological processes which together exhibit coherent system behaviour where state variables have been
chosen in order to keep the model relatively simple without omitting any component that exerts a
significant influence upon the energy balance of the system. ERSEM uses a functional group approach to
describe the ecosystem whereby biota are grouped together according to their trophic level (subdivided
according to first size, then trophic role and finally feeding method). All biological functional groups in
ERSEM are modelled according to the concept of the 'standard organism' (Baretta et al 1995). There are
three classes of standard organisms: autotrophs, heterotrophic consumers and decomposers. The
dynamics of biological functional groups are described by both physiological (ingestion, respiration,
excretion and egestion) and population processes (growth, migration and mortality). The differences
between the functional groups mainly lay in the rate constants, which are derived experimentally from the
literature or from allometric considerations (e.g. Moloney and Field 1989), and in the food components on
the uptake side. This concept works well for those groups where data are usually derived from population
or even community studies.
Physical models: Underpinning, any ecosystem, model is the ability to accurately represent the physical
environment at appropriate time/space scales and adequate process descriptions. Here we build on the
UK’s involvement in the NEMO 2 consortium. Originally a global ocean model
1 FABM (http://www.meece.eu/documents/deliverables/WP2/D2.14.pdf)
2 Nucleus for European Modelling of the Ocean (www.nemo-ocean.eu/)
NEMO has now been developed as a coastal-ocean model that includes (e.g.) tides and sophisticated
vertical coordinates (O’Dea et al 2012) and direct coupling to ERSEM (Edwards et al, 2012). These papers
describe the 7km resolution Atlantic Margin Model (AMM7) which is the chosen baseline 3D
configuration for this work. The sensitivity of the system to different processes/process descriptions and
parameterisations will be tested in 1D (GOTM 3-ERSEM) where thousands of different test runs are
tractable. The integrated impact of the processes on seasonal and inter-annual variability of the shelf will
be evaluated using the 3D AMM7 model (NEMO-ERSEM). We will explore the implications of moving to
DivERSEM on computational capability and ways of facilitating this, while retaining accuracy.
Darwinian ERSEM4: A version of ERSEM has been modified to include phytoplankton diversity; each of the
four phytoplankton groups have been substituted with 10 different sub type or ‘species’ whose
characteristics are selected to simulate diversity, using an approach analogous to that used by Follows et al
(2007). The community is then allowed to develop in response to its environment through intraspecific
competition for the same resources. The model was used in the FP7 MEECE project to explore the
consequences of the disruption of the local phytoplankton community by invasive species.
Model Development
In order to meet the challenges of the programme we require a scalable model system, with a traceable
hierarchy whereby more complex foodweb structures can be systematically and coherently related to
simple foodweb structures. Building on the concepts of the standard ERSEM organisms and the SSB
ERSEM size structured foodweb as a baseline we will establish and test a hierarchical framework which
describes three levels of organisation of the foodweb
1. Trophic structure in terms of organism size and function (here we refer to high level
ecosystem function, i.e. autotrophy, heterotrophy, decomposition).
2. Within size / functional class diversity, by subdividing by biological traits (e.g. feeding
strategy, motility, physiology).
3. Within trait diversity whereby intra- and inter-specific competition is described by defining a
set of species within each trait type, stochastically drawing parameters from a rule based
parameter space.
Marine Ecosystem Research Programme Part B
The capabilities of ERSEM will be expanded in two ways. First, distinctions within functional type (FT) size
classes (e.g., phytoplankton, zooplankton, bacteria, zoobenthos) that are currently present in the model
will be reduced by introducing unified models, suitable for broad classes of organism types. These will be
able to replicate the behaviour of the various original FTs if configured with appropriate parameter values.
This unification will enable div-ERSEM’s stochastic approach to cross FT boundaries, thus reducing the
emphasis on traditional, subjective, hard- coded subdivision of FTs. Second, the parameter space spanned
by the stochastic approach will be reduced by identifying a low number of principal traits (e.g., size,
feeding strategy) that determine other functional traits through trade-offs. Identification of principal traits
and trade-offs will be based upon trait value compilations published in existing literature (e.g., Bruggeman
2011, Edwards et al. 2013, Wirtz 2013). By breaking down model barriers between types and by focusing
on a low number of principal traits that characterize interspecific variability, div-ERSEM
3 General Ocean Turbulence Model www.gotm.com;
4 Darwinian ERSEM (http://www.meece.eu/documents/deliverables/WP4/D4.3_P3_InvasiveSpecies.pdf)
will be able to generate ecosystem models of varying degrees of complexity, merely by varying the
number of stochastically created model species. This approach will be used to assess the relationship
between diversity (number of species) and ecosystem functioning, and to identify the optimal level of
complexity required to describe ecosystems at the sites of study. It is anticipated that aspects of
DivERSEM (e.g. parameterisations of scale and diversity/function relationships) would be fed back to
improve the core biogeochemical model developed under the SSB programme.
Adaptive dynamics: In ecology, “adaptive dynamics” approaches have been used to summarize the
behaviour of multi-species models in terms of a few key statistics: total biomass, mean trait value, and
trait variance (a measure of functional diversity) (Abrams 1992, Norberg et al. 2001, Merico et al. 2009).
Critically, the dynamic behaviour of these statistics, rather than abundances of the species they describe,
is evolved in time. This drastically reduces the number of modelled tracers. A similar approach has been
used to describe the size distribution of the plankton community in terms of a dynamically evolving size
spectrum slope (Kriest & Oschlies 2007). These methods are approximations that maintain a traceable link
with multi-species models. It is therefore possible to evaluate their accuracy of approximation by directly
comparing their results with that of species-explicit simulations (Norberg et al. 2001, Merico et al. 2009).
Moreover, a key feature of adaptive dynamics approaches is that they explicitly represent functional
biodiversity in the form of a single tracer, i.e., trait variance. This makes it possible to directly (analytically)
assess the interaction between diversity and ecosystem functioning (e.g., bulk biogeochemical fluxes).
DivERSEM and similar approaches that offer refined representations of variability in size (Baird & Suthers
2007, Banas et al 2011, Ward et al 2012 ) and other functional traits (Follows et al. 2007; Bruggeman &
Kooijman 2007) do so by explicitly simulating large numbers of species. This is computationally expensive,
particularly in 3D models. Moreover, species-explicit approaches only allow empirical assessment of the
link between biodiversity and ecosystem functioning, by experimentally varying the number of simulated
species and observing the response in model behaviour. We will apply adaptive dynamics-type
approximations to aggregate multi-species configurations of DivERSEM in terms of total biomass, and the
mean and variance of a few principal traits (e.g. size). To embed these in spatially explicit hydrodynamic
models, we will use and refine existing variable transformation methods developed for water columns
(Bruggeman 2009). The performance of approximate methods will be assessed by comparing their results
with those of species-explicit DivERSEM simulations.
Physical scales and advance approaches to advection: Advective transport is a key element of biophysical
interaction that is often overlooked because our view of the seasonal cycle of plankton growth in shelf
Marine Ecosystem Research Programme Part B
seas has tended to focus on vertical mixing processes, for the good reason that these are largely the first
order controls on seasonal timescales. Moreover a ‘vertical only’ view fits nicely with timeseries
observations and computationally fast models such as GOTM-ERSEM. However, shelf sea production is
also characterised by strong horizontal gradients (e.g. at fronts and coastal regions of freshwater
influence), and in neglecting processes driven by these gradient we miss a crucial aspect of the system
and its response to change e.g. enhanced frontal production (Richardson et al 2000) and in doing so
possibly under-estimate the diversity of niches in the model. On longer (annual to decadal) timescales,
horizontal transport sets the overall elemental budgets the ecosystems have to work with (e.g. Holt et al
2012). In addition the numerical treatment of advection is the primary limit for 3D models to exploit
current increases in computer power, which now largely occur by increased parallelism rather than
computational speed. Hence, the challenge is to provide a computationally efficient horizontal transport
scheme that can accommodate a substantial increase in the number of state variables and maintain
accuracy in the representation of the key biophysical interactions. In tidally active shelf seas advective
transport generally dominates over horizontal diffusive transport and so is our focus here.
Data Requirements: data is the life blood of modelling and is required to both parameterise and verify
models. Observations may also be used to formulate derived functions or function-related parameters
that can be used to test or improve the functional representations in ERSEM. Assessing the skill of
DivERSEM will require new approaches to model skill assessment beyond those commonly used in
biogeochemical modelling. We will develop novel methods applying statistical techniques commonly
used for analysing biodiversity data (e.g. non parametric multi- dimensional scaling). More specifically, to
validate models we require information on the larger scale emergent patterns (time, space, and
emergent ecosystem properties such as biogeography, size spectra, trophic levels, connectivity, diversity,
and also key functions such as primary and secondary production, respiration and stability measures).
The interactions with WP1 are crucial to this aspect of the project.
Workplan
Task 1. Technical development of the software environment for the hierarchical model. We will establish
a version controlled development trunk for DivERSEM. Using a modular approach based on the standard
organism the code will be designed to provide a scalable model system, with a traceable hierarchy
whereby more complex foodweb structure can by systematically and coherently related to simple
foodweb structures. Funded though NC (PML).
Task 2. Evaluation of the ecosystem properties of the existing ERSEM configuration. To inform both the
model development and WP1 we will evaluate and existing hindcast of the NW European Shelf, in terms
of the model reproduction of observed size spectra, metabolic scaling, observed ecosystem variability
and shifts in trophic control. The focus will be on data rich regions (PML, Cefas).
Task 3. Computational cost of advection. The horizontal advection of biological variables is a substantial
computational cost. This task involves improving the capacity of the model to scale and capability to
work with 10’s to 100’s state variables. Firstly we will optimise, for 10s-100s state variables, existing
schemes in NEMO (TVD and PPM) that are accurate, monotonic and non-diffusive. Then we will explore
the implications (on efficiency/accuracy) of combining these schemes with substantially cheaper but
lower fidelity methods (e.g. upwind and centred), using a master variable/subsidiary variable approach
and test these approaches in a frontal resolving NEMO model (~1.8km). Finally we will review and assess
other advanced transport methods, such as those used in multi-tracer atmospheric simulations (e.g.
Lauritzen et al 2010) (NOC).
Task 4. Revising the ERSEM standard organisms. The purpose of this task is to improve the trophic
Marine Ecosystem Research Programme Part B
structure in terms of organism size and function through revisiting the definitions of the standard
organisms combining pelagic and benthic approaches as appropriate, improving the individual process
descriptions based on, allometric scaling and metabolic theory where possible to define a baseline set of
size scalable core parameters. This will allow us to address whether size and function based relationships
can describe macroecological relationships and explain the flow of energy though the ecosystem. The
performance of this model will be evaluated against the standard SSB ERSEM model in 1D water column
models.
• Autotroph’s: Analysis of nutrient-dependent growth and uptake in phytoplankton, reveals
physiological trade-offs in species abilities to acquire and utilize resources (Litchman et al
Marine Ecosystem Research Programme Part B
2007). Such trade-offs, arise from fundamental relations such as cellular scaling laws and enzyme
kinetics and define contrasting ecological strategies of nutrient acquisition and the subsequent
modulation of cellular stoichiometry with size (Finkel et al., 2010). Cellular stoichiometry provides a link
between bottom up control (availability of light and nutrient) and top down control in the planktonic
food web (zooplankton stoichiometric modulation of predation). Starting with the standard
phytoplankton organism from the SSB ERSEM model and drawing on the published analysis of size based
traits (e.g. Litchman et al 2007, Finkel et al 2010, Edwards et al 2012) the parameterisation of the
standard autotrophic organism will be refined to better represent these tradeoffs. The applicability of
the standard autotroph to
the benthic system (as developed in Blackford, 2002) will be further developed and evaluated. (PML).
• Heterotrophic Consumers: We will refine the pelagic and benthic foodweb, focusing on the
allometeric scaling of growth, ingestion and respiration to better resolve the observed size
structures. (Cefas, PML).
• Decomposers (Heterotrophic prokaryotes) Heterotrophic prokaryotes (bacteria and archaea) play a crucial
role in both the benthic (Lloyd et al., 2013) and pelagic systems (Carlson et al. 2007). ERSEM is being
developed by including (and further refining) the formulation proposed by Polimene et al. (2006). This
describes the release of excess of carbon (overflow metabolism) which decouples the uptake of carbon
from cell growth (i.e. “net” production), thus making the variability of Bacterial Growth Efficiency as a
function of the environmental conditions. We shall draw on this work to develop the basic standard
decomposers for both the benthic and pelagic systems. Further work will be undertaken to improve the
parameterisations of aerobic and anaerobic bacteria in the benthic system.
• We will explore the top closure of both the pelagic and benthic models starting with a simple quadratic
term, but also considering the external closure from and offline HTL model. In addition we will make data
available for 1 way coupling to drive external HTL models being applied in WP1. Depending on the
nature work proposed by the modelling activities in WP1 we will explore the collaborative development
of 2 way couplers with WP1 if required. (Cefas, PML).
Task 5 Biological traits. To further improve our capability to reproduce the observed spatial and temporal
heterogeneity in ecosystem structure this task focuses on improving the description of size / functional class
diversity, and within trait diversity. Although size is the major driver for many physiological and behavioural
parameters, other traits are not size-dependent and they can significantly contribute in determining the
fitness of the organisms and therefore potentially contribute to the general ecosystem structure. We shall
focus on expanding the biological trait diversity of autotrophs, zooplankton and zoobenthos, and explore
how these groups drive the energy supply into the ecosystem and its transfer to higher trophic levels.
• Autotrophs: Diatoms play a crucial role in determining the ecosystem properties of both pelagic and
benthic ecosystems, the latter through the supply of fast sinking detritus in the spring. The current
version of ERSEM describes one category of diatoms (defined as silicate consumers). We will focus on
differentiating the size classes of diatoms. In addition we will parameterise traits in benthic autotrophs
(e.g. benthic diatoms and macroalgae). (PML, Cefas).
• Zooplankton For the heterotrophic grazers, the strength of the predator–prey interactions, are driven by
the choice of parameters and more specifically the food preferences (Sailley et al 2013). This illustrates
the importance of equation and parameter choice as they define interactions between FTs and overall
food web dynamics (competition, bottom-up or top-down effects). Related to this is the issue of the
relationships between key processes such as prey selectivity, ingestion and digestion to the amount of
prey, its quality and its type (Mitra and Flynn, 2005). Drawing on work being undertaken in the SSB
program, the description of zooplankton grazing activity will be refined by including processes such as
the stoichiometric modulation of predation (Mitra, 2006; Mitra and Flynn, 2005) and prey palatability
(i.e. production of toxins), which may heavily contribute to promote algal blooms in coastal eutrophic
Marine Ecosystem Research Programme Part B
areas (Mitra and Flynn, 2006). In addition they exhibit a number of foraging behaviours ranging from
ambush feeding (waiting for random prey encounters), to cruise feeding (increasing prey encounters by
swimming) and current feeding (increasing prey encounters by generating a feeding current). Each
strategy has a different set of penalty costs, trading prey availability against metabolic costs of
movement and increased risk of predation. The current version of ERSEM describes zooplankton by size
(e.g. micro, meso) but does not take account of foraging behaviours. We will focus on differentiating the
different feeding strategies by initially differentiating between ambush feeders and filter feeders (PML).
• Zoobenthos: In an analogous manner to the zooplankton above, we shall focus on refining the benthic
foodweb, primarily by defining the macro fauna in terms of feeding and burrowing strategies. (PML,
Cefas).
• Adding species diversity: To fully explore the impact of changes in diversity on the resistance and
resilience of ecosystems requires representation of intra trait diversity. This trait variability will be
included in the model by randomly varying the value of some of the parameters of the modelled
organism (e.g. growth rate, carbon: nutrients ratio, food preference, clearance rate) based on the
variability of the allometric relationships observed in the literature. The variability of ecosystem
properties depending on these parameters will be compared to that one depending on size in order to
assess how much of the observed patterns can be explained by the size, and how much the other traits
contribute (PML).
Task 6. Simulations and synthesis: The DivERSEM model will be used to explore the importance of biological
traits in defining major ecosystem properties like diversity, primary and secondary production, energy
transfer efficiency and their temporal and spatial patterns.
• GOTM water columns with be set up for data rich sites (e.g. L4, Stonehaven, Cefas Oyster Grounds)
and validated. Sensitivity analysis will be performed to explore natural variability, trophic controls and
diversity/function relationships (PML, Cefas).
• A 40 year hindcast of the NW European Shelf will be made, exploiting the common model forcing data
base setup by the SSB project. This hindcast will be assessed in terms of its skill in reproducing macroecological properties (PML, NOC).
• We will also explore the effects of refining model resolution (from ~7km to ~1-2km), so as to better
represent the structure and dynamics of shelf sea fronts and investigate the consequences for
ecosystem structure and function (e.g. Holt et al 2004). For this we will exploit developments of fine
resolution NEMO models in FASTNet (NERC ocean –shelf exchange RP), but on a reduced area nested
domain covering the Celtic Sea, English Channel and Irish Sea (see Holt and James, 2006). The 3D
simulations will be evaluated using large spatial data sets e.g. Satellite Ocean Colour (NEODASS) and
the CPR. (NOC)
Outline workplan for Phase II (M37-M60) In the 2nd phase of the project we will introduce reproductive
processes to represent one of the largest trophic shifts in the benthic ecosystem, namely the release of
larval forms into the pelagic water column, a substantially large perturbation in organic material and a
transfer of energy from sediment to water column. Such dispersal events often affect the phenology of
various heterotrophic life cycles and in particular those of fish. We will also implement a dynamic top
closure scheme based on a dynamic size spectrum model (e.g. Blanchard 2008). Finally we will undertake a
set of hindcast and scenario experiments to further explore the sensitivity of the marine ecosystems of the
NW European Shelf to natural and anthropogenic drivers. The exact details of these scenarios will be
defined following consultation with WP1 and policy focused stakeholders.
National Capability: We assume the following additional support will be available to the project: PML 3
FTE NC, In addition model development will be underpinned by activities in the following NERC programs,
Marine Ecosystem Research Programme Part B
NCEO, UKOA, SSB, and FastNet.
Linkages within the Marine Ecosystems program: Model development is most effective when the nonmodelling scientists are stakeholders in the model development. To ensure full interaction with the wider
community we propose to hold a modeller / stakeholder workshop at project meetings. We propose
establishing explicit linkages by creating mini working groups whereby model developers and appropriate
PI’s or researchers, team up to focus on modelling key ecosystem properties.
Links with the wider international ERSEM community: We will use the inputs provided by the project to
cross fertilise ideas with the wider community and share code, to ensure that both the Div-ERSEM model
benefits from the international activity and vice versa (see letters of support).
Deliverables and legacy
• Published versions of the model code
• Community modelling tools available on the web; 1D Div-ERSEM-GOTM, 3D Atlantic
Margin Model -(Div-ERSEM NEMO-Shelf), model evaluation tools.
• Hindcast simulations and re-analysis simulations, sub sets of the data deposited with
appropriate NERC data centre.
• Publication of new trait based models
• Scientific publications on the sensitivity and resilience of shelf seas marine ecosystem, to climate
change and other anthropogenic drivers
Project and resource management
The PI will provide overall leadership of the project and assume responsibility for reporting OPMs to NERC
and the final report. The programme of research is described in the table 1. The project duration is five
years (Phase I 3years, Phase II 2 years) and will begin in January 2014. To coordinate the different
elements of the model development project meetings will be held in months 1, 6, 12, 18 and 24, 30 and
36. In addition the project teams will update each other on progress via monthly video conferences.
Table 1
Supporting Activities
T1 Setup and maintain hierarchical model tools
Phase I
T2 Analysis of existing ERSEM hindcast
T3 Advanced advective schemes
T4 Refining the standard organism
T5 Adding traits and trait diversity
T6 1D GOTM scenarios
T6 3D NEMO shelf scale hindcast
Phase II
Addition of reproductive processes
Dynamic top closure
Scenarios, reanalysis and synthesis
Marine Ecosystem Research Programme Part B
Year 1
Year 2
Year 3
Phase II
References:
Abrams P 1992, Adaptive Foraging by Predators as a Cause of Predator Prey Cycles, Evolutionary Ecology,
6(1), 56-72; Baird and Suthers 2007, A size-resolved pelagic ecosystem model, Ecol. Mod., 203(3-4), 185-203;
Banas N 2011, Adding complex trophic interactions to a size-spectral plankton model: Emergent diversity
patterns and limits on predictability, Ecol. Mod., 222(15), 2663-2675; Baretta et al 1995 The European
Regional Seas Ecosystem Model, A Complex Marine Ecosystem Model Neth. J. Sea Res. 33 (3/4): 233-246;
Baretta-Bekker J et al 1997 Microbial dynamics in the Marine Ecosystem model ERSEM II with decoupled
carbon assimilation and nutrient uptake. J. Sea Res 38, 195–212; Blackford et al 2004 Ecosystem dynamics at
six contrasting sites: a generic modelling study J. Mar Sys 52, 191-215, Blackford J 2002, The Influence of
Microphytobenthos on the Northern Adriatic Ecosystem: a Modelling Stud, Est. Coast, Shelf Sci. 55, 1, 109123. Blackford J 1997. An analysis of benthic biological dynamics in a North Sea ecosystem model. J. Sea Res.
38, 213– 230.; Barton A et al 2010, Patterns of Diversity in Marine Phytoplankton, Science, 327(5972), 15091511; Blanchard et al 2010 Potential consequences of climate change for primary production and fish
production in large marine ecosystems, Phil.Trans. R. Soc. B oi:10.1098/rstb.2012.0231; Blanchard et al 2008
How does abundance scale with body size in coupled size- structured food webs? J. Animal Ecol. 2009, 78,
270–280; Brey, T. 1990. Estimating productivity of macrobenthic invertebrates from biomass and mean
individual weight. Meeresforschung, 32, 329–34; .Brown J & Maurer B 1989 Macroecology: the division of
food and space among species on continents. Science, 243, 1145-1150; Brown, J. (1995) Macroecology.
University of Chicago Press, Chicago, Illinois, USA. Bruggeman J. & S. Kooijman(2007, A biodiversity-inspired
approach to aquatic ecosystem modeling, Limnol. and Oceanogr., 52(4), 1533-1544; Bruggeman J (2009),
Succession in plankton communities: a trait-based perspective, PhD thesis, 158 pp, Vrije Universiteit
Amsterdam, Amsterdam; Bruggeman J 2011, A phylogenetic approach to the estimation of phytoplankton
traits, J. Phycology, 47(1), 52-65; Cardinale B et al 2006. Effects of biodiversity on the functioning of trophic
groups and ecosystems, Nature doi:10.1038/nature05202; Carlson, et al 2007. Microbes and the Dissipation
of Energy and Respiration: From Cells to Ecosystems. Oceanography, Vol 20 n2; Coll M, et al. 2010, The
biodiversity of the Mediterranean Sea: Estimates, patterns, and threats. PLoS ONE 5(8): e11842.
doi:10.1371/journal.pone.0011842; Cornell, H & Lawton, J 1992. Species interactions, local and regional
processes, and limits to the richness of ecological communities – a theoretical perspective J. Animal Ecol., 61,
1–12; Costello MJ, 2010 A Census of Marine Biodiversity Knowledge, Resources, and Future Challenges.PLoS
ONE 5(8): e12110. doi:10.1371/journal.pone.0012110; Edwards K et al 2012, Allometric scaling and
taxonomic variation in nutrient utilization traits and maximum growth rate of phytoplankton, Limnol.
Oceanogr., 57, 554-566; Edwards, K et al 2012. Validation of the NEMO-ERSEM. Ocean Sci. 8, 983-1000;
Frank KT, et al 2006. Reconciling differences in trophic control in mid-latitude marine ecosystems Ecology
Letters 9, 10, 1096–1105; Follows M et al (2007), Emergent biogeography of microbial communities in a
model ocean, Science, 315(5820), 1843-1846; Holt, J.T., James, I.D., 2006. An assessment of the fine scaleeddies in a high resolution model of the shelf seas west of Great Britain. Ocean Modelling, 13, 271-291; Holt J
et al 2004. Advective controls on primary production in the stratified western Irish Sea: an eddy-resolving
model study. Journal of Geophysical Research, 109, C05024, doi: 05010.01029/02003JC00195; Holt J 2012.
Oceanic controls on the primary production of the northwest European continental shelf: model experiments
under recent past conditions and a potential future scenario. Biogeosciences, 9, 97-117; Irwin, et al JPR, 28
(5), 459-471; Finkel et al., 2010 J. Plank.Res 31 119-137: Kriest I & A. Oschlies 2007, Modelling the effect of
cell- size-dependent nutrient uptake and exudation on phytoplankton size spectra, Deep-Sea Res. Part I-, 54,
1593-1618; Lauritzen, P et al 2010 A conservative semi-Lagrangian multi-tracer transport scheme (CSLAM)
on the cubed-sphere grid. J. Comp. Physics, 229, 1401-1424; Lindegren M, 2012 Early Detection of Ecosystem
Regime Shifts: A Multiple Method Evaluation for Management Application PLoS 1, 7, 7 e38410; Litchman E,
et al 2007, The role of functional traits and trade-offs in structuring phytoplankton communities: scaling
from cellular to ecosystem level. Ecol Lett
10:1170–1181; Lloyd et al., (2013) Predominant archaea in marine sediments degrade detrital proteins Nature,
496, 215-218. Merico, A., et al (2009), A trait-based approach for downscaling complexity in plankton
ecosystem models, Ecological Modelling, 220, 3001-3010; Mitra A & Flynn K 2005, Predator-prey
interactions: is “ecological stoichiometry” sufficient when good food goes bad? J. Plankton Res., 27, 393–
399. Mitra, A. & Flynn K 2006, Promotion of harmful algal blooms by zooplankton predatory activity. Biol.
Marine Ecosystem Research Programme Part B
Lett., 2, 194–197. Mollmann C, et al 2009 Reorganization of a large marine ecosystem due to atmospheric
and anthropogenic pressure: a discontinuous regime shift in the Central Baltic Sea Global Change Biology
15, 1377–1393, doi: 10.1111/j.1365-2486.2008.01814.x; Moloney C & Field J 1989 General allometric
equations for rates of nutrient uptake, ingestion, and respiration in plankton organisms Limnol. Oceangr.
34, 1290-1299; Mackas L, Beaugrand G, 2010 Comparisons of zooplankton time series J. Mar. Sys. 79
(2010) 286–30; Norberg, J., et al 2001, Phenotypic diversity and ecosystem functioning in changing
environments: A theoretical framework, PNAS, 98(20), 11376-11381; O’Dea, E. et al 2012. An operational
ocean forecast system incorporating NEMO and SST data assimilation for the tidally driven European NorthWest shelf J. Op. Oceanogr. 5:3- 17; Ruardij, P Van Raaphorst, W 1995. Benthic nutrient regeneration in
the ERSEM ecosystem model of the North Sea. Neth. J. Sea Res. 33, 453– 483; Richardson, K., et al 2000.
Subsurface phytoplankton blooms fuel pelagic production in the North Sea. J. Plankton Res., 22, 1663-1671;
S Sailley 2013 Comparing food web structures and dynamics across a suite of global marine ecosystem
models Ecol. Mod. 261–262 43– 57. Schwinghamer P (2011) Characteristic Size Distributions of Integral
Benthic Communities. Canadian J Fisheries and Aquatic Sci. 1981, 38(10):
1255-1263, 10.1139/f81-167; Verdy et al 2009 Optimal phytoplankton cell size in an allometric model Mar
Ecol Prog. Ser. doi: 10.3354/meps07909; Ward, B. A., et al (2012), A size-structured food-web model for
the global ocean, Limnol. and Oceanogr., 57(6), 1877-1891; West, G. et al 1999 The fourth dimension of
life: fractal geometry and allometric scaling of organisms. Science, 284, 1677–1679. Warwick R & Price R
1979. Ecological and metabolic studies on free-living nematodes from an estuarine mud-flat. Est.
Coast.Shelf Sci., 9, 257–271; Wirtz K 2012, Who is eating whom? Morphology and feeding type determine
the size relation between planktonic predators and their ideal prey, Mar Ecol Prog Ser, 445, 1-12;YvonDurocher G et al 2012, Reconciling the temperature dependence of respiration across timescales and
ecosystem types Nature doi:10.1038/nature112.
Outline Data Management Plan
1. Data management procedures
All the data produced by this project will for numerical simulation outputs from the DiVERSEM – NEMO
Shelf Model system.
The model code will be maintained using a version controlled core trunk for DivERSEM, linked to the
NEMO Shelf trunk using the PUMA shared repository following JWCRP guidelines. All data outputs will be
in a standard platform independent NETCD format, with a header which describes the layout of the rest
of the file, in particular the data arrays, as well as arbitrary file metadata in the form of name/value
attributes. The contents of this header and metadata will be agreed with the relevant NERC data centers
at the beginning of the project.
The version controlled model codes, along with the parameter sets and external forcing will be archived at
PML, with mirrored copies at NOC and Cefas. Large shelf scale simulation such as those proposed here
produces TBytes of model output. It is not practical to store the complete runs on disk for long periods of
time. Therefore the full model runs will be archived to tape at PML as soon as they are completed. A set
of standard model validation metrics will be defined and applied to quantify the quality of the outputs.
Research Co-I Y Artioli, who is a member of the
PML data management committee, will act as data manager.
2. Existing datasets
We will make use of existing data sets held at BODC (e.g. Western Channel Observatory, shelf seas
biogeochemistry) to validate models and atmospheric forcing data held by BADC to force hindcast
simulations (e.g. ERA40). Satellite Ocean Colour data and SST will be requested from NEODAAS for
model validation.
3. Datasets likely to be created
Marine Ecosystem Research Programme Part B
Data Centre
Data set description
Release to data
centre
Reuse scenarios
BODC
40 year hindcast of
NW European Shelf
Ecosystem made with
DivERSEM-NEMO.
Outputs would
include, T, S, biomass of
all biological
variables, production
and grazing rates,
nutrients and other
biogeochemical
information. The
output frequency
would need to be
discuss to keep the data
set tractable and usable
M36 of the project
Model data would be
made available to
support shelf seas
marine ecosystem
research and the
development and
implementation of
marine environmental
policy
Marine Ecosystem Research Programme Part B

Similar documents

×

Report this document