Marine Policy 43 (2014) 382–386

Contents lists available at ScienceDirect

Marine Policy

http://d

1 w2 ht3 ht

journal homepage: www.elsevier.com/locate/marpol

Building ecological-economic models and scenarios of marineresource systems: Workshop report

a r t i c l e i n f o

Keywords:Ecological-Economic ModellingIntegrated AssessmentMarine ResourcesEcosystem Approach

x.doi.org/10.1016/j.marpol.2013.05.010

ww.ices.dk/community/groups/Pages/WGIPEMtp://www.ihp.fr/fr/ceb/mabies.tp://mpe2013.org.

a b s t r a c t

As part of the ecosystem approach to managing fisheries and other uses of marine ecosystems, there hasbeen a growing call for the development of integrated assessment tools to support the provision of bothtactical and strategic management advice. Of particular importance in this domain is the development ofmodels that capture the dynamic interactions between social and economic systems, and marineecosystems. In February 2013, a workshop jointly organised by the ICES working group on Integrative,Physical–biological and Ecosystem Modelling and researchers attending the “Mathematics of Bio-economics” initiative, a contribution to the international event “Mathematics of Planet Earth 2013”,brought together experts to discuss recent advances and key methodological challenges posed by thisfield of research. The manuscript provides a brief report of the key points discussed during the workshop,including identification of the research which may help progress both the development of thesemodelling approaches and their application to actual management decision problems.

1. Introduction

As part of the ecosystem approach to managing fisheries andother uses of marine ecosystems, there has been a growing call forthe development of integrated assessment tools to support theprovision of both tactical and strategic management advice [1]. Ofparticular importance in this domain is the development ofmodels that capture the dynamic interactions between socialand economic systems, and marine ecosystems, allowing identifi-cation of scenarios for the future, and evaluation of potentialresponses to alternative management strategies [2–4].

In February 2013, a workshop jointly organised by the ICESworking group on Integrative, Physical–biological and EcosystemModelling1 and researchers attending the “Mathematics of Bio-economics” initiative2 , a contribution to the international event“Mathematics of Planet Earth 2013”3 , brought together experts todiscuss recent advances and key methodological challenges posedby this field of research. The workshop combined presentations ofrecent work towards meeting a diversity of these challenges, andopen discussion of the key domains currently thought to be crucialto progress both the development of these modelling approachesand their application to actual management decision problems.

A first challenge addressed in the workshop relates to thegrowing demand for tools that, in evaluating the trade-offsassociated with managing marine resource systems, fully accountfor the multiple (economic, ecological and social) dimensions ofsuch trade-offs and the distributional impacts of scenarios acrossstakeholder groups. Presentations by Luc Doyen, on the viabilityapproach to ecological-economic scenarios [5–7], and by MartinQuaas, on the identification of winners and losers in the transition

.aspx.

towards sustainable fisheries [8–10], illustrated recent efforts atdeveloping such evaluations, while also taking into account thecomplex set of interactions and the multiple sources of uncertaintywhich characterise marine ecological-economic systems.

A second challenge relates to process understanding of marineecosystem uses (including but not limited to commercial fish-eries), how this can be modelled, and coupled to biophysicalmodels in order to gain better understanding of the potentialconsequences of alternative economic, environmental or manage-ment scenarios. Presentations by James Innes on modelling fishingbehaviour in the Australian Eastern Tuna and Billfish Fishery [11–13], and by Christian Mullon on modelling international tunacatching and trading as a global network [14] illustrated twoextremes of the spectrum over which such research has beendeveloping, in an effort to reduce a key source of uncertainty infisheries management [2,15].

A third challenge relates to the growing complexity of modelsthat couple representations of ecological, economic and socialprocesses, each of which may be affected by uncertainty, makingthe systematic exploration of sensitivity of model projections tothese different sources of uncertainty increasingly difficult. Apresentation by Stephanie Mahévas [16,17] on sensitivity analysisfor complex models illustrated the research efforts underway toaddress this issue, and develop formal methods which enable thesystematic evaluation of model projections to the assumptionsrelating to their parameters.4 The presentation of summary con-clusions from a recent meeting focused on “modelling from firstprinciples”, by Benjamin Planque, illustrated the need felt by somemembers of this research community to use simple, well-established models, to delimit domains within which projectionsfrom the more complex models should remain, and should provide

4 www.reseau-mexico.fr/welcome; www.isis-fish.org/.

O. Thébaud et. al. / Marine Policy 43 (2014) 382–386 383

additional understanding or predictive capacity. Models consid-ered here are based on few and simple (‘first’) principles withparticular emphasis on mathematical theory of viability [18,19],stochastic constrained models [20] and Liebig's law of limitingfactors [21].

Finally, a fourth important challenge relates to the key issuesthat need to be considered when models are expected to be usedin decision-support processes involving multiple stakeholders. Thepresentations of key steps in developing operational bio-economicmodels for fisheries management support5 by Claire Macher [22–25], and of the lessons learned from the SPICOSA6 Europeanresearch project in adopting a systems modelling approach toco-construct models of coastal zone management issues withstakeholders [29], both emphasised the role which models mayhave in assisting the management process, as well as the manypractical issues which need to be addressed for this role to beeffective. The presentation by Tony Smith also illustrated theimportance of “social licence to operate”, and the dangers of takingfor granted that evidence-based decision making can withstandpublic perceptions relayed through social networks [2].

Altogether, the approaches presented at the workshop pro-vided a good illustration of the diversity of modelling challengeswhich one must face when attempting to include “human dimen-sions” in models of marine resources management. The followingreport summarises the key methodological and practical pointsidentified by the group as requiring particular attention.

2. Model complexity: the value of formalization andparsimony

The group noted the increasing complexity which characterisesmodels that aim to couple dynamic representations of biologicaland economic processes, and the diversity of approaches tomanaging this complexity which characterises efforts in this space.There is clearly a trade-off associated with increasing the complex-ity of the processes simultaneously represented in models ofsocio-economic systems and ecosystems. Expanding this complex-ity in both dimensions comes at a cost in terms of (i) the ability tounderstand the causes of the behaviour of the model, (ii) theability to systematically assess the influence, on model projections,of the different sources of uncertainty in these processes and theirinteractions, and (iii) the ability for the models to be used intactical decision support.

As illustrated in Fig. 1, many models aimed at understandingthe ecological dynamics of marine resource systems have beendeveloped with only limited economic content. In a similarfashion, efforts to model the dynamics of socio-economic systemsoften only include limited ecological content. Most of the inte-grated modelling efforts to date have thus been centred onrelatively simplified representations of both ecology (i.e. a species,often a single species, focus) and economy (i.e. the fishingmortality “F” imposed by a fishing fleet, often a single fleet, focus).The models developed at this level have also been those used fortactical decision-support in fisheries management. The challengesidentified in the workshop imply moving towards the North-Eastpart of this figure, into a domain where the trade-offs in decidingon the level of economic and ecological complexity of models willbe guided by the questions which these models are asked toprovide answers to, the purposes of using a modelling approach toaddress these questions, and the context in which the models areto be developed and used.

5 www.umr-amure.fr/pg_partenarial_bioeco.php.6 www.spicosa.org.

The group discussed several principles which could be devel-oped into a set of guidelines to modelling in this domain, based onthe experience of participants. In particular, the group agreed thatreinforcing the mathematical focus of modelling efforts could bevery effective in ensuring clarity in the modelling assumptions,particularly in a multi-disciplinary context, as well as parsimony inmodel development. The importance of “starting simple” was alsostressed, as well as the need to justify the importance of adoptingmore complex representations, given the modelling objectives.The “first principles” approach, discussed during the workshop,would seem particularly relevant in this respect. In addition, thedevelopment of sensitivity analysis methods for complex modelsappears to be a key domain in which to invest research efforts, asthis may allow to better identify and rank the major sources ofuncertainty which warrant further empirical research.

3. Diversity of modelling approaches: the need for systematicdocumentation of model characteristics

The group also discussed the large diversity of approacheswhich are being followed, and which was well illustrated by thepresentations at the workshop. This diversity is apparent in termsof (i) the nature and scale of the economic and social processes (e.g. from models of micro-economic behaviour, to model of macro-economic systems, through the representation of strategicdecision-making using game theory) as well as ecological pro-cesses (from single population dynamics to multi-populationdynamics and trophic networks); (ii) the modelling approaches(optimal or viable control system dynamics, simulation, agent-based modelling, …); (iii) the diversity of tools/platforms/lan-guages; and (iv) the diversity of contexts in, and purposes forwhich the models are used.

The group considered that this diversity is a positive characteristicof the research area, as it seems unlikely that a single “one size fitsall” approach could suitably address the many issues and multiplescales which need to be considered. In addition, given the currentlimitations of knowledge, the group also acknowledged that a degreeof diversity, redundancy and modularity could be considered to be astrength. For instance, a multiple-models approach may allowidentifying the range of uncertainty relating to alternative scenarios.While such diversity may be difficult to justify in terms of researchfunding support, it could be expected to enable better adaptation toevolving needs, under the ecosystem approach to marine livingresources management, as societal demand and our understandingof marine ecological-economic systems progresses.

However, the group also considered the need to seek common-alities between the different approaches, which would assist in thelonger-term strengthening of integration between economic andecological models, particularly for decision support. To assist inthis, there may be a need to carry out more systematic reviews ofexisting models. Related to this, and to the importance of for-malizing model assumptions, the group agreed that efforts todocument existing models in a systematic fashion, including anevaluation of their strengths and weaknesses, would contribute toincreasing the knowledge base. An approach similar to the “Over-view, Design Concepts and Details (ODD)” protocol recommendedby [26] for documenting complex models may be particularlyuseful in this regard. For case studies for which several (docu-mented) models have been developed, a common operationalframework to run models could also be a promising tool tosupport decision-making. This approach would consist in (i) fixingcommon input-scenarios (including uncertainty) and outputsvariables; and (ii) running the available models to assess andcompare the consequences of a selection of management mea-sures, given the assumed uncertainty.

Fig. 1. Challenges for the ecological-economic modelling of marine living resource systems.

O. Thébaud et. al. / Marine Policy 43 (2014) 382–386384

4. Impact assessment and evaluation: the importance ofmulti-criteria approaches, and of considering thedistributional impacts of projected scenarios

The group also discussed the nature of the evaluation whichshould drive the development of integrated ecological-economicmodels of marine living resource systems. All agreed that thesemodels should allow for multi-criteria evaluations, as the ecosys-tem approach increases the likelihood that even simple resourcesystems will need to be managed considering multiple objectives,which cannot easily be combined into a single objective function.This is particularly important in a decision-support context. Insuch contexts, models should assist in developing operationalapproaches to tracking the performance of management strategiesacross a range of dimensions (i.e in defining operational manage-ment objectives). This should be true of models used in both ex-ante impact assessments, and in ex-post evaluations. In addition,models should also provide the ability to assess the distributionalimpacts of management strategies and scenarios, as these mayinfluence the extent to which these are deemed acceptable, andare supported in practice.

5. Stakeholder engagement: avoiding errors of the third kindand building trust in the models used for decision-support

As soon as models become part of a decision process, thequestion of stakeholder trust in models becomes crucial.The group highlighted the fact that such trust can be difficult toboth establish and maintain. In addition, once established, beliefsin model validity and accuracy may be difficult to change. Thegroup discussed the factors which may affect trust. An importantdimension of ecological-economic models, which several presen-tations at the workshop highlighted, is the fact that their devel-opment can often involve significant stakeholder engagement.Such engagement is expected to allow better inclusion of stake-holder knowledge in the representation of key processes, as wellas build a shared understanding and support for the questionsaddressed, representations adopted and methods used to assist indecision-making. This in turn can help reduce the risks that model

developers commit “errors of the third kind” [27], i.e. developmodels that provide the right answers to the wrong problems.Indeed, there has been an increasing tendency for strongerstakeholder engagement in the development of models relatingto natural resource management issues internationally, withresults showing the potential benefits of such an engagement.

The group discussed some of the participant's experience inthis domain, in particular the potential role for “developmentmodels” which can be used to interact with stakeholders at veryearly stages in the development of decision-support tools. InAustralia, single species assessments and more complex ecosys-tem models coexist. The latter can be used to explore questionsrelated to more complex sets of management arrangements. Insome cases, simpler qualitative models were initially developedwith stakeholders; these enabled a dialogue to be establishedwith stakeholders allowing the development of the more com-plex quantitative models. In effect, both simple and more com-plex models are used to inform management. In the Europeancontext, simple models linking biological and economic dimen-sions have been successfully developed and used to discusspractical fisheries scenario and management strategies, forexample in the French context. Stakeholder engagement at bothnational and international (Regional Advisory Committees) levelshas been crucial in the development and use of these models. Inturn, this has led to a number of practical issues that werehighlighted in presentations at the workshop. As in the Austra-lian context, the group observed that different models could beused to address different questions. For example, the assessmentof the new mixed fisheries management plans has entailed use ofmodels of mixed fisheries.

The group also considered the ways in which stakeholderengagement might best work. In the participant's experience,key principles which seemed to positively influence the outcomeof stakeholder engagement around model development and useincluded avoiding the imposition of a predefined model, andmaintaining an open approach to suggested changes and adapta-tions to existing modelling assumptions. In addition, participantshighlighted the importance of the particular skill set required fromindividual researchers who engaged in developing and runningmodels with stakeholders. Finally, the group also identified the

O. Thébaud et. al. / Marine Policy 43 (2014) 382–386 385

fact that this stakeholder engagement in model development andmodel use comes at a cost in research resources, which is still veryoften under-estimated in the planning of research projects.

6. Areas where strategic research is needed

In the course of the discussion, the group also identified anumber of key areas in which further efforts at developing formalmodels of processes at play would appear to be of strategicimportance, in the current context of ecological-economic model-ling of marine living resource systems. Given the composition ofthe group, most of these relate to key social and economicprocesses which influence these systems, and how they could beincluded in integrated ecological-economic models.

The areas identified include: (i) market dynamics at various scales,and how these drive the prices which influence resource harvestingand decisions on changes in fishing capacity in local fisheries; (ii) otherdrivers of individual resource harvesting patterns, including theinsights which may be gained from the application of game theoryand other conceptual frameworks addressing strategic behaviour; (iii)governance and management decision processes themselves, particu-larly as they relate to:

�

the allocation of catch shares in commercial fisheries (which isa multi-level process in the European context) and the influ-ence of access regulation systems such as quota tradingmarkets;

�

spatial planning and the determinants of spatially definedregulations and incentives;

�

the institutional and political determinants/constraints thatcharacterise resource management systems.

On the latter aspect, the group noted that the need to find formal,model-based approaches to represent such constraints has beenrecognised for several decades (e.g. [28]), and still appears to behighly relevant, although this has not yet translated widely intoformal approaches to ecological-economic models of marineresource systems.

Further discussion of these issues is expected to continue aspart of the activities of WGIPEM, as well as in the forthcoming2013 ICES Annual Science Conference in Reyjavik, Iceland, at whicha session on the theme “Modelling human behaviour as part ofintegrated models of marine ecosystems”, will take place.

Acknowledgements

Preparation of this report was carried out with the financialsupport of the CSIROWealth from Oceans Flagship. We are grateful tothe organisations that supported participation of the co-authors inthe 2013 meeting of the ICES working group on Integrative, Physical-biological and Ecosystem Modelling (WGIPEM), as well as to theMathematics of Bio-economics (MABIES) initiative and the InstitutHenri Poincaré for their support.

References

[1] Prellezo R, et al. A review of EU bio-economic models for fisheries: the valueof a diversity of models. Mar Policy 2012;36(2):423–31.

[2] Fulton EA, et al. Human behaviour: the key source of uncertainty in fisheriesmanagement. Fish Fish 2011;12(1):2–17.

[3] Wilen JE, et al. Avoiding surprises: incorporating fisherman behavior intomanagement models. Bull Mar Sci 2002;70(2):553–75.

[4] Reeves, S. A., Marchal, P., Mardle, S., Pascoe, S., Prellezo, R., Thébaud, O. andTravers, M. (2009) From Fish to Fisheries: The Changing Focus of ManagementAdvice, in Advances in Fisheries Science: 50 years on from Beverton and Holt

(eds A. Payne, J. Cotter and T. Potter), Blackwell Publishing Ltd., Oxford, UK.http://dx.doi.org/10.1002/9781444302653.ch7.

[5] Béné C, Doyen L, Gabay D. A viability analysis for a bio-economic model. EcolEcon 2001;36:385–96.

[6] Doyen L, et al. A stochastic viability approach to ecosystem-based fisheriesmanagement. Ecol Econ 2012;75:32–42.

[7] Gourguet S, et al. Managing mixed fisheries for bio-economic viability.Fisheries Research 2013;140(0):46–62.

[8] Voss, R., et al., Management targets and trade-offs in multispecies fisheries,unpublished manuscript.

[9] Stoeven, M. and M. Quaas, Privatizing renewable resources: who gains, wholoses?, in Economics working paper 2012, University of Kiel: Kiel.

[10] Quaas, M., et al., Windows of opportunity for sustainable fisheries manage-ment: The case of the Eastern Baltic Cod., unpublished manuscript.

[11] Dowling NA, et al. Assessing opportunity and relocation costs of marineprotected areas using a behavioural model of longline fleet dynamics. Fish andFisheries 2012;13(2):139–57.

[12] Innes, J., et al., How optimal is optimal: do behavioural adaptations by fishersaffect the cost of marine reserves?, unpublished manuscript: Brisbane.

[13] Pascoe S, et al. Economic and conservation implications of a variable effortpenalty system in effort-controlled fisheries. Appl. Econ. 2012;45(27):3880–90.

[14] Mullon C, et al. Modeling the global fishmeal and fish oil markets. NaturalResource Modeling 2009;22(4):564–609.

[15] van Putten IE, et al. Theories and behavioural drivers underlying fleetdynamics models. Fish and Fisheries 2012;13(2):216–35.

[16] Lehuta S, et al. Selection and validation of a complex fishery model using anuncertainty hierarchy. Fisheries Research 2013;143:57–66.

[17] Mahévas S, Pelletier D. ISIS-FISH, a generic and spatially explicit simulationtool for evaluating the impact of management measures on fisheriesdynamics. Ecol. Modell. 2004;171:65–84.

[18] Aubin J. Viability Theory. Boston: Birkhauser; 1991.[19] Aubin J. A Survey of Viability Theory. SIAM Journal on Control and Optimiza-

tion 1990;28(4):749–88.[20] Mullon C, et al. A minimal model of the variability of marine ecosystems. Fish

and Fisheries 2009;10:115–31.[21] van der Ploeg RR, Bohm W, Kirkham MB. On the origin of the theory of

mineral nutrition of plants and the law of the minimum. Soil Science Societyof America Journal 1999:1055–6263 1999:1055–62.

[22] Merzéréaud, M., et al., Description of the Impact Assessment bio-economicModel for fisheries management (IAM), in Amure Electronic Publications,Working Papers Series (available online)2011. p. 19.

[23] Simmonds, E., et al., Impact Assessment of Bay of Biscay sole (STECF EWG 11-01), 28 February - 4th March 2011., in JRC Scientific and Technical Report2011,:Copenhagen. p. 40.

[24] Raveau A, et al. A bio-economic analysis of experimental selective devices inthe Norway lobster (Nephrops norvegicus) fishery in the Bay of Biscay. AquaticLiving Resources 2012;25(03):215–29.

[25] Guillen J, et al. Estimating MSY and MEY in multi-species and multi-fleetfisheries, consequences and limits: an application to the Bay of Biscay mixedfishery. Marine Policy 2013;40(0):64–74.

[26] Grimm V, et al. The ODD protocol: A review and first update. Ecol. Modell.2010;221(23):2760–8.

[27] Kimball AW. Errors of the third kind in statistical consulting. J. Am. Stat. Assoc.1957;52(278):133–42.

[28] Pope JG. Fisheries resource management theory and practice. In: G.G. TJLB,editor. New Zealand fisheries: the resources and their management. Editor.Auckland: Trade Publications Limited; 1983. p. 56–62.

[29] Lample, M., et al., A sustainability-driven approach for participatory model-ling, 19th International Congress on Modelling and Simulation, Perth, Aus-tralia, 12–16 December 2012. http://mssanz.org.au/modsim2011.

Olivier Thébaud n, James InnesCSIRO Marine and Atmospheric Research, Brisbane, Australia

E-mail addresses: Olivier.Thebaud@csiro.au,thebaud.olivier@gmail.com (O. Thébaud)

Luc DoyenCNRS—MNHN, Paris, France

Michel LampleUniversity of Western Brittany, UMR AMURE,Brest, France Olivier Thébaud n, Claire Macher

Ifremer, UMR AMURE, Brest, FranceE-mail addresses: (O. Thébaud)

Stephanie MahévasIfremer, Unit of Fisheries Ecology and Modelling, Nantes, France

O. Thébaud et. al. / Marine Policy 43 (2014) 382–386386

Christian MullonIRD, UMR 212 EME, Sète, France

Benjamin PlanqueInstitute of Marine Research, Tromsø, Norway

Martin QuaasUniversity of Kiel, Department of Economics, Germany

Tony SmithCSIRO Marine and Atmospheric Research, Hobart, Tasmania, Australia

Youen VermardIfremer, Fisheries Research Laboratory, Boulogne-sur-Mer, France

Received 6 May 201325 May 201325 May 2013

Available online 2 July 2013

n Corresponding author at: CSIRO Marine and Atmospheric Research, Brisbane,Australia. Tel.: +61 457568144.