ChS

GSBa

b

c

d

e

f

g

h

i

j

k

l

m

n

o

p

q

r

Ss

t

u

a

AA

KISFES

0d

Ecological Modelling 222 (2011) 2456– 2468

Contents lists available at ScienceDirect

Ecological Modelling

journa l h o me pa g e: www.elsev ier .com/ locate /eco lmodel

ommon challenges for ecological modelling: Synthesis of facilitated discussionseld at the symposia organized for the 2009 conference of the Internationalociety for Ecological Modelling in Quebec City, Canada, (October 6–9, 2009)

.R. Larocquea,∗, D. Maillyb, T.-X. Yuec, M. Anandd, C. Penge, C. Kazanci f, M. Ettersong, P. Goethalsh,.E. Jørgenseni, J.R. Schramski j, E.J.B. McIntirek, D.J. Marceaul, B. Chenm, G.Q. Chenn, Z.F. Yangm,. Novotnao, N. Luckaip, J.S. Bhattiq, J. Liur, A. Munsons, A.M. Gordont, J.C. Ascough II u

Natural Resources Canada, Canadian Forest Service, Laurentian Forestry Centre, 1055 du P.E.P.S., P.O. Box 10380, Stn. Ste-Foy, Quebec, QC G1V 4C7, CanadaDirection de la recherche forestière, Ministère des Ressources naturelles et de la faune, 2700, rue Einstein, Quebec, QC G1P 3W8, CanadaInstitute of Geographical Sciences and Natural Resources Research, Chinese Academy of Sciences, 11A, Datun Road, Anwai, Beijing 100101, ChinaSchool of Environmental Sciences, Department of Environmental Biology, University of Guelph, Guelph, ON N1G 2W1, CanadaInstitute of environment sciences, Biology Science Department, University of Quebec at Montreal (UQAM), Case postale 8888, succursale Centre-Ville, Montreal, QC H3C 3P8, CanadaDepartment of Mathematics, Faculty of Engineering, University of Georgia, Athens, GA 30602, USAU.S. Environmental Protection Agency, Mid-Continent Ecology Division, 6201 Congdon Boulevard, Duluth, MN 55804, USADepartment Applied Ecology and Environmental Biology, Ghent University, J. Plateaustraat 22, B-9000 Gent, BelgiumUniversity of Copenhagen, Langkaer Vaenge 9, Vaerloese DK-3500, DenmarkDepartment of Biological & Agricultural Engineering Systems & Engineering, University of Georgia, 406 Driftmier Engineering Center, Athens, GA 30602-4435, USACentre d’étude de la forêt, Faculté de foresterie, de géographie et de géomatique, Université Laval, 2405 rue de la Terrasse, Quebec, QC G1V 0A6, CanadaDepartment of Geomatics Engineering, Schulich School of Engineering, University of Calgary, 2500 University Drive N.W., Calgary, AB T2N 1N4, CanadaState Key Laboratory of Water Environment Simulation, School of Environment, Beijing Normal University, Beijing 100875, ChinaNational Laboratory for Turbulence and Complex Systems, Peking University, Beijing, ChinaSoils and Crops Research and Development Centre, Agriculture and Agri-Food Canada, 2560 Hochelaga Blvd, Quebec, QC G1V 2J3, CanadaFaculty of Natural Resources Management, Lakehead University, 955 Oliver Road, Thunder Bay, ON P7B 5E1, CanadaNatural Resources Canada, Canadian Forest Service, Northern Forestry Centre, 5320-122nd Street, Edmonton, AB T6H 3S5, CanadaStinger Ghaffarian Technologies (SGT, Inc.), contractor to the U.S. Geological Survey (USGS) Earth Resources, Observation and Science (EROS) Center, 47914 252nd St., Sioux Falls,D 57198, USACentre d’étude de la forêt, Faculté de foresterie, de géographie et de géomatique, Université Laval, Quebec, QC G1K 7P4, CanadaSchool of Environmental Sciences, University of Guelph, Guelph, ON N1G 2W1, CanadaUSDA-ARS, Agricultural Systems Research Unit, Fort Collins, Colorado 80526, USA

r t i c l e i n f o

rticle history:vailable online 25 January 2011

eywords:SEM 2009ymposiaacilitationcological modelscientific discussions

a b s t r a c t

The eleven symposia organized for the 2009 conference of the International Society for Ecological Mod-elling (ISEM 2009) held in Quebec City, Canada, October 6–9, 2009, included facilitated discussionsessions following formal presentations. Each symposium focused on a specific subject, and all the sub-jects could be classified into three broad categories: theoretical development, population dynamics andecosystem processes. Following discussions with the symposia organizers, which indicated that theyall shared similar issues and concerns, the facilitated discussions were task-oriented around four basicquestions: (1) key challenges in the research area, (2) generating and sharing new ideas, (3) improvingcollaboration and networking, and (4) increasing visibility to decision-makers, partners and clients. Com-

mon challenges that emerged from the symposia included the need for improved communication andcollaboration among different academic disciplines, further progress in both theoretical and practicalmodelling approaches, and accentuation of technology transfer. Regarding the generation and shar-ing of new ideas, the main issue that emerged was the type of positive interactions that should beencouraged among potential collaborators. The usefulness of the Internet, particularly for the sharingof open-source software and conducting discussion forums, was highlighted for improving collaborationand networking. Several communication tools are available today, and it is important for modellers

∗ Corresponding author. Tel.: +1 418 648 5791; fax: +1 418 648 5849.E-mail address: Guy.Larocque@NRCan.gc.ca (G.R. Larocque).

304-3800/$ – see front matter Crown Copyright © 2011 Published by Elsevier B.V. All rights reserved.oi:10.1016/j.ecolmodel.2010.12.017

G.R. Larocque et al. / Ecological Modelling 222 (2011) 2456– 2468 2457

to use them more intensively. Visibility can be increased by publishing professional newsletters, main-taining informal contacts with the public, organizing educational sessions in primary and secondaryschools, and developing simplified analytical frameworks and pilot studies. Specific issues raised in eachsymposium are also discussed.

Table 1Symposia organized during the 2009 international conference of the InternationalSociety for Ecological Modelling (ISEM 2009).

Title Organizers

1- Earth surface modelling and globalecology

Tian-Xiang Yue (China)Sven E. Jørgensen (Denmark)Guy R. Larocque (Canada)

2- Network modelling and systems theory:advances in network environ analysis

Caner Kazanci (United States)John Schramski (United States)

3- Individual-based models forconservation

Eliot J.B. McIntire (Canada)Danielle Marceau (Canada)

4- Ecological accounting Bin Chen (China)Guogian Chen (China)

5- Forest simulation models for sustainableforest management under a changingenvironment

Changhui Peng (Canada)Guy R. Larocque (Canada)Daniel Mailly (Canada)

6- Ecological modelling for environmentalflows

Zhifeng Yang (China)

7- Modelling the consequences of climatechange on the agricultural landscape

Beata Novotna (Slovakia)

8- Modelling avian seasonal productivity Matthew Etterson (UnitedStates)

9- Research perspectives in carbon cyclemodelling to support sustainableterrestrial ecosystem management

Guy R. Larocque (Canada)Jagtar S. Bhatti (Canada)Jinxun Liu (United States)Alison Munson (Canada)Daniel Mailly (Canada)Andrew M. Gordon (Canada)Nancy Luckai (Canada)James C. Ascough II (UnitedStates)Louis Archambault (Canada)

10- Ecosystem modelling for decisionsupport in water management

Peter Goethals (Belgium)

11- Static and dynamic approaches to Madhur Anand (Canada)

1

2e(tpspsetccu9

gamasbl

and the applicability of models by non-modellers to meet soci-ety’s needs. There are still many challenges associated with thetheoretical and practical aspects of model development, including

modelling diversity and complexity

. Introduction

Eleven symposia on a variety of subjects were organized for the009 conference of the International Society for Ecological Mod-lling (ISEM 2009) held in Quebec City, Canada, October 6–9, 2009Table 1). In the year preceding the conference, the members ofhe organizing committee invited ecosystem modellers to submitroposals for the organization of symposia. The objectives of theymposia were to focus on a subject of interest to organizers andarticipants and provide a forum for constructive group discus-ions to develop new ideas or establish partnerships. Twenty-threecosystem modellers from five different countries responded tohe invitation. As indicated in Table 1, several subjects were dis-ussed at the symposia, which can be classified into three broadategories: (a) theoretical development (1, 2, 4, 10 and 11), (b) pop-lation dynamics (3 and 7), and (c) ecosystem processes (5, 6, 8 and).

The format of the symposia consisted of formal presentations byuest speakers invited by the organizers (Appendix A), followed by

discussion period. As indicated in Appendix A, more than 75 com-unications by 269 authors were presented. In order to encourage

ll the participants to propose solutions and share ideas, the discus-

ion sessions (which lasted between one and two hours) were ledy facilitators who were trained to lead group discussions and high-

ight the main ideas in a concise manner. This type of facilitation

Crown Copyright © 2011 Published by Elsevier B.V. All rights reserved.

in scientific conferences has proven its effectiveness and relevancein synthesizing the main ideas among large groups of participants(e.g., Larocque et al., 2009). The objective of this paper is to reportthe results of the facilitated discussions.

2. Methodology

The facilitators belonged to the Learning Organization Com-munity of Practice (LOCOP) of Natural Resources Canada1. Beforethe symposia, a member of LOCOP, Mrs. Deidre Moore2, inter-viewed the main organizer of each symposium to inquire aboutthe specific objectives and subjects presented in the formal pre-sentations. Following the interviews, the LOCOP came to theconclusion that the symposium organizers shared a common inter-est in the four following questions to be discussed among theparticipants:

1) What are the key challenges in this research area (ecosystemmodelling)?

2) How can we generate and share new ideas?3) How can we improve collaboration and networking?4) How can we become more visible to decision-makers, partners

and clients?

When the formal presentations ended in each symposium, thefacilitators divided the participants into four groups. The partici-pants in each group met together independently from the othergroups to hold discussions on each question and propose answers.Then, at the end of the discussion period, all the participants gath-ered together and voted on the most important ideas, suggestionsand solutions that came up during the discussions.

3. Results and discussion

3.1. Common suggestions and ideas

The initial data compilation phase indicated that many sug-gestions and ideas were similar in different symposia. To avoidredundancy in Section 3.2, this section describes the common out-comes from the symposia. During the subsequent data compilationphases, it was decided to classify the suggestions and ideas into dif-ferent broad categories to allow readers to better understand thediversity of opinions expressed by the participants.

Regarding the identification of key challenges in ecosystemmodelling (question 1), the suggestions and ideas from the differ-ent symposia were subdivided into five broad categories (Table 2).For information, society’s needs and technology transfer, the keypoints that summarize the ideas and suggestions gathered arethe importance of improved communication and collaborationamong modellers from different academic disciplines, the difficul-ties inherent to conducting the work required to test new ideas,

1 http://wiki.nrcan.gc.ca/index.php/AFC LoCOP.2 Library Manager, Natural Resources Canada, Laurentian Forestry Centre.

2458 G.R. Larocque et al. / Ecological Mod

Table 2Suggestions and ideas gathered during the facilitated discussion periods at the dif-ferent symposia organized at ISEM 2009 to identify key challenges in ecosystemmodelling (Question 1).

Information, society’s needs and technology transferAvailability of model librariesNew ideas are hard to work outFew colleagues in this fieldFragmented and isolated research effortsAdequating needs, projects and modelsFinding balance in economic, environmental and social aspects of our workAdaptative managementImprove scientific collaborationEnsure models are applicable to actual field scenariosMore, better, and user-friendly software (maybe econet does this)Sufficiently trained users of models we createLinkages between predictive models and decision support toolsTransfer information among scientists and from scientists to managers

(from theory to practice)Model development (theory and practice)

Elaboration of an accepted and widely used ecosystem theory, which canguide the modelling process

Integration of hierarchical spatial and temporal (dynamic) modelsMerging ontologiesCapture all ecosystem properties in our modelsEffectively work with large number of models and connecting modelling

with biological processes (interdisciplinary)Dealing with different scales and dimensionsHow to calibrate modelsGaps in knowledgeMaking the links between remote sensing and field dataPromote interdisciplinary research linking for enhanced sustainabilityEstimating covariances among model parametersGetting a better handle on variance in Start/End dates. Minimizing estimate

error/varianceCompeting risk issuesHierarchical modelsQuantifying spatio-temporal autocorrelation in vital ratesBayesian model averaging (weights)Non-probabilistic uncertainty analysisIs error so big that certain analyses are simply impossibleCorrectly integrating data from multiple sourcesAssessing commonalities compensating parameterBridge the gaps between field experiments and modellingPractical multiscale methods and data sharingBridge the gap between academic and practical approaches without

reducing the complexity of modelsImportance of data in modelling

Data limitations are still importantData qualityHow to get data for large-scale studiesHow to deal with large sample size requirementsData collection issues – scale VS scopeGetting good/appropriate data for the modelOpen datasets – no analysis, just descriptions

Dealing with uncertainty and validationError estimation in models and methodologyValidation proceduresEstimating and communicating uncertaintyStart collecting validation data. Find adequate datasets to validate modelsHow to reduce uncertainty

atcardctthpd

Link between lack of confidence and uncertainty about real systemsData uncertainty

commonly accepted ecosystem theory, the integration of spa-ial and temporal scales, and the application of methodologies andalibration approaches. Despite the fact that many datasets arevailable, data quality and suitability for model development stillemain a major issue in ecosystem modelling. The importance ofealing with uncertainty and validation was identified as a keyhallenge. In particular, it will be important for modellers to bet-er communicate what uncertainty is all about. Another suggestion

hat can be summarized in one expression, interdisciplinary science,ad the distinction of emerging as integral to the solution to allerceived challenges. When potential solutions were ranked, inter-isciplinary science also appeared at or near the top with regard to

elling 222 (2011) 2456– 2468

the level of importance in each area of discussion. These findingsagree with some of the latest efforts where research, conferences,and related publications in ecological sciences continue to addressinterdisciplinary science and management (Ewel, 2001; Leimu andKoricheva, 2005; Schramski and Gattie, 2009; Solidoro et al., 2009).While there is a need for new and innovative approaches to under-stand the complex structure of living systems, new mathematicalproblems arising from life sciences provide new opportunities toadvance quantitative sciences, which have mostly dealt with phys-ical problems over the last few centuries (Cohen, 2004). Clearly,these efforts and the findings from this symposium indicate thatinterdisciplinary collaboration is increasingly viewed as vital tothe advancement of network modelling and systems theory as wellas the mathematical and computational sciences used to supportthem.

The results of the discussion on the generation and sharing ofnew ideas (question 2) resulted in suggestions that were subdividedinto six broad categories (Table 3). Even though long-recognizedideas were discussed, such as interdisciplinary approaches andenhanced communications, participants highlighted the fact thatgenerating ideas is generally not a problem, but implementingthem is the true challenge. Specific attitudes, such as respect amongcollaborators, will be strongly desirable to improve how ecosystemmodellers work together to generate new ideas. The importanceof publications, including scientific and review papers, was onceagain recognized. However, it was felt that it would be importantto reduce the publication pressure by focusing more on quality, notquantity. This issue is in agreement with an editorial published inEcological Modelling, in which it was mentioned that papers pre-senting further advances in ecological modelling, not only papersthat use existing models with different datasets, were essential toincrease the scientific value of Ecological Modelling (Jørgensen et al.,2006). It is believed that Internet utilities should be used moreintensively for exchanges among ecosystem modellers, the shar-ing of open-source software and the organization of collaborativeefforts.

For question 3, “How can we improve collaboration and net-working”, the suggestions and ideas gathered were also subdividedinto six broad categories (Table 4). Several items had also beenidentified in question 2, but these were not necessarily overlap-ping because the ideas and suggestions originated from discussiongroups that met independently from each other. The role of ISEMin the organization of meetings and the development of Internetutilities was highlighted. For communications and Internet use,several ways to meet formally were suggested and they all relyon existing tools. If these Internet communications tools are usedmore often, it will become possible for collaborators to reduce theircarbon footprint. Also, the need for data and model sharing wasonce again identified, as well as the necessity to promote openaccess to scientific literature. Despite the fact that the use of thesetools may appear to advocate formal communications, the needfor informal discussions was also clearly identified. For collabora-tion and training, interdisciplinary approaches in student trainingand the establishment of exchange programs were advocated. Forcommunity actions, sharing data, methodologies and proceduresas well as networking, several ideas identified in question 2 wererepeated, which demonstrates their importance. However, someexisting initiatives were identified as examples that could be usedby the ISEM community, such as the National Center for EcologicalAnalysis and Synthesis (NCEAS), which provides funding, space andtools for proposals, or the USGS Northern Prairie Wildlife ResearchCenter (NPWRC), which maintains an important databank. For net-

working, some additional elements relative to question 2 wereidentified, including the establishment of networks of researchsites and measuring standards, common scientific questions andresearch framework. Several examples of existing networks of

G.R. Larocque et al. / Ecological Modelling 222 (2011) 2456– 2468 2459

Table 3Suggestions and ideas gathered during the facilitated discussion periods at the dif-ferent symposia organized at ISEM 2009 to answer the question on how to generateand share ideas in ecosystem modelling (Question 2).

InnovationApply theory to novel domains (like transportation)Establish on integrated theoretic framework for ecology (find out a general

rule for ecology)Renew the models (e.g. dynamic models), don’t always use same old

formulasPresent not only what works, but also what doesn’t

Interdisciplinary approachesCross-fertilization from other fields that also use models (e.g. economics)Bring together field ecologists, modellers and decision makersMultidisciplinary groupsCollaborative comparative model exercises in different ecosystemsSummer school cross disciplineEncourage co-ops and internshipsModify undergrad teaching methodsRequire student research for graduationMore student training (early, often, interdisciplinary)

Communications and publicationsWorkshopsUse open source and public access journal. Every paper should come with

data + model to produce reproducible science.Review paper; literature review (model types, model parameters, model

sensitivities)Organize symposia: small focused workshopsGenerating special issues in journalsSynthesis papers to identify “state of the art” in specific areasSpeculative blue-sky presentations at conferenceInnovative and cross-discipline workshops + conferencesFacilitate discussions at meetings/conferencesMore/smaller regional conference workshopsWorkshops at meetingsPublish papers which DO NOT fail “repeatability” criteriaIntegrated brainstorming exercises – community-based participation –

international project for collaboration and networking – more money (forconferences, visits between institutes, international exchanges) – increasecollaboration and networking – listservers (bank of email addresses) –more cross – fertilizing across disciplines

Balancing breadth and depth in academic trainingInvolve undergraduate and graduate students in idea generationEngage more undergraduates in research (e.g. USNSF REU’s)

Internet and Software resourcesOpen-source models & methods (transparency) and sharing of existing data

& models - “Open source software” for ecological models (e.g.http://ecobas.org).

Discussion forum and use of Twitter – using Internet tools to access otherideas (e.g. reading or creating blogs) – use technology (e.g.videoconferencing) more effectively (watch footprint) – chat line forecological modelling

Linking personal websitesOnline meta-data bank; data sharing

NetworkingFacilitate research networks and/or fund participation to conferencesMediated institutionalized modelling

Desired attitudesRespectful controversy generates ideasBetter know each otherImprove our communication skillsCollaborate on generic tools, models and methods of wider useMore joint research projects

iU(

b(ptip

Table 4Suggestions and ideas gathered during the facilitated discussion periods at the dif-ferent symposia organized at ISEM 2009 to answer the question on how to improvecollaboration and networking (Question 3).

Role of ISEMPromote networking in specialized topicsSociety website (LinkedIn, Facebook, Skype, membership list, personal

websites)Organization of meetings more oftenSpace on ISEM website and increasing Web presenceUse ISEM as a medium of communication

Communications and InternetInternet forums, websitesEstablish a platform for data sharingTeleconferencing and Internet tools (e.g., Skype webcam for Tokbox)Create common open-source software and modelsWeb conferencesDirectional website (to get information on the topic); Google waveWiki pages; Silvics WikiData and software sharingOpen access to all literatureVideo conferencing between field data collectors and model developersOpen source softwareGlobal databaseConferences where you can get feedback and have face to face communicationSmaller “regional” or “topic” conferencesGoogle/Facebook groupsCreate summary product following a workshopReduce institutional barriers to meeting

Collaboration and trainingExchange of researchers and graduate students (internship across levels)Student activities (meeting mixers, listservs)Interdisciplinary student trainingSummarizing informationEU networks of excellence and integrated projectsActively pursue/invite at all levels (take the initiative)Joint research programMeetings are important (especially for students)Work together for research project or student training

Community actionsActively seek feedback and survey communityEcological data sharing across fieldPromote international projectsFunding encourages collaborationLong-term funding commitmentsPossibility of feedback to published articlesPublications with failures as well as successesStructured dialogue on infrastructure and interoperability along with

feasibility of collaborationStronger top–down managementTechnical workshops on methodological issuesCreate incentives to collaborateGetting a cross-disciplinary perspective

Sharing data, methodologies and proceduresCreating standard data storage and metadata protocolsCommon conceptual framework with sharing of database and metadataUnited standards of measurementEnvironmental flows (common terms, definitions, applications, case studies)Model comparisonsIntegrate research with on the ground applicationsGreater use of shared model platforms, model & data repositories

NetworkingSetup networks of research sites and measuring standards

Collaborative modelling, apply everyone’s models to common datasets

nterest were given as examples: Waterfowl brood/pair surveys,S Fish and Wildlife Service (USFWS), Canadian Wildlife Service

CWS), and the Black Dutch Joint Venture (BDTV).Several issues were also debated for question 4, “How can we

ecome more visible to decision-makers, partners and clients”Table 5). The use of the Internet was discussed, but in a more

roactive way with decision-makers and the public compared withhe ideas and suggestions obtained for the other questions. Fornstance, the development of dynamic real-time websites was pro-osed. It was also suggested to use mass communication media to

Setup a common scientific question and research frameworkMulti-stakeholder workshops

get closer to the public, such as being present on television shows.The importance of transferring information to the public in sim-plified terms was highlighted. It would also be advantageous topublish newsletters and professional papers. Education in primaryand secondary schools was identified as an essential achievementto attract young minds. For decision-makers, it will be important

to develop analytical frameworks based on scientific data to assistthem during their reflection process for the analysis of model pre-dictions. In addition, special efforts will be required by showingpilot studies and making user-friendly applications available.

2460 G.R. Larocque et al. / Ecological Modelling 222 (2011) 2456– 2468

Table 5Suggestions and ideas gathered during the facilitated discussion periods at the dif-ferent symposia organized at ISEM 2009 to answer the question on how to becomemore visible to decision-makers, partners and clients (Question 4).

Communications and InternetSpecialized coursesPress conferences with the mediaGood institutional university websites with examples of models and

applicationsDocumentationPlan into research dissemination – outreachPublish in management journalsRegular meetings among scientists, policy makers, clients, NGOs

(non-governmental organizations). Invite them to conferences“Open houses” – scientific institutions have special days when public and

managers are invitedNGO to promote ecological modelling with end usersInfluence funding calls to get buy inCommunication at different levels in the communityBe proactive to policy making process via mass communication tools (TV

shows?)Publish non-scientific (professional) papers as well as more high-quality

papers (=sheer weight)Education from primary schoolsCreate interestProvide non science-heavy, non-tech. summaries; reduce scientific findings

to lay language in report form; talk to decision-makers in their language;publish in agency, NGO, etc. newsletters

Dynamic real-time websiteGenerate press releases with a follow up with mediaUse agencies as conduitsOrganize debates on TV between scientists and laypeople (politicians?)Build public relations (news releases, updates) into program structureUse agencies as conduitsUsing existing channels of publicity (e.g. other societies)Package output in relevant language (communicate in basic language)Communicate with decision makers to have their feedbacks or inputsMedia presence; Use models to help developing countries = media attentionMake sure concepts and ideas are communicated in terms that everybody

can understandEcological modelling web pagesSimple models to drive home key pointsSimple concepts (to facilitate communications)

Collaboration and trainingBasic courses in academic programsEmphasis on K-12 educationDeveloping framework for decision-makers based on scientific data

Community actionsFigure out and summarize who our decision-makers, partners & clients areAttend structured decision-making workshopsLearn how to “market” results to decision-makersAnticipate client needs 3–5 years aheadShow that we are members of an active scientific societyCommunity-based participationInvite local environmental or other citizen groups (stakeholders) to attend

ISEM conferencesPublic relations representative on executive councilInvite civil, architectural (etc.) firms to present their modelling projectsDetermine how visible do we really want to be and to whomIdentify stakeholders at beginning of project; get client participation and

buy-in at start of research projectLobbyingActive participation in community initiativesRepeated contact, follow up consultationGrassroots efforts of membersSalesmanship (learn to “sell” research ideas to broader community)Education to general publicMore partnerships = more credibilityLaypeople become the bridge between modellers and politiciansInvite decision-makers, politicians and business for ecological analysisUndergraduate training

Development and use of applicationsConvincing pilot studies and applicationsWell-explained case studies; showcases of applicationsAdded value of modelsEcological forensics/CSI (crime scene investigation)-styleInvolve stakeholders in research project, ask them their needsDynamic three-dimension tools based on scientific data

Table 5 (Continued.)

Build up global projects that aim at linking existing models to providedecision tools for stakeholders

Present examples for new scientific innovationsDevelop decision-making system based on scientific knowledge

3.2. Specific suggestions and ideas

In this subsection, specific issues that were raised in the sym-posia are discussed.

3.2.1. Earth surface modelling and global ecologyThe earth surface is the interface of the lithosphere, atmosphere,

hydrosphere and biosphere. Earth surface modelling is generallydefined as a spatially explicit digital description of an earth surfacecomponent or an ecosystem in terms of global ecology principles.The global-scale ecological unit of the biome has been extendedto include the human influence on ecosystems, which compriseshuman population density, land use and land cover to describeanthropogenic effects on earth surface (Alessa and Chapin, 2008).Climate and geology have shaped ecosystems and evolution in thepast, but human forces may now outweigh them across most ofthe Earth’s land surface (Ellis and Ramankutty, 2008). Earth sur-face modelling cannot be conducted by dealing with global controlsalone while ignoring local complications, or by treating local casestudies separately from global factors. The key to a sustainableenvironment is to think globally and act locally; the key to under-standing global ecology is to think globally and locally all at once(Phillips, 2002).

However, various problems remain in earth surface modelling.For example, no global models have yet achieved a satisfactorylevel of dynamic integration between the biophysical earth sys-tem and the human socioeconomic system. A global digital terrainmodel with high accuracy has yet to be completed or combinedinto related global models, and statistical transfer functions stillneed to be developed. An alternative means of solving these prob-lems is to develop high-accuracy and high-speed methods for earthsurface modelling, which could deal with huge amounts of dataand multiscale issues in three dimensions, under consideration ofa ground- and satellite-based global observation system with anoptimal data-sharing mechanism (Yue, 2010).

3.2.2. Network modelling and systems theory: Advances innetwork environ analysis

Data collection, availability and uncertainty constitute auniquely common and key challenge identified at the sympo-sium. The search for novel ideas is a never ending venture. Forexample, the development of high throughflow experimental tech-niques (Fernandes, 1998), such as microarrays, has played animportant role in recent advances in genetics, microbiology andmedicine. Similarly, accurate and abundant data are required tomodel complex environmental phenomena for which feasible solu-tions usually involve both quantitative and experimental research.As such, advances in new experimental techniques and new quan-titative approaches capable of dealing with incomplete and noisydatasets (e.g., stochastic methods, artificial intelligence, Bayesianapproaches, etc.) are needed to advance network modelling andsystems theory in environmental sciences (e.g., Bonavito et al.,1994; Aalders and Aitkenhead, 2006).

The need for interdisciplinary science is clear (Schramski and

Gattie, 2009; Cohen, 2004). Building successful collaborationsamong scientists with different backgrounds, skills and interestsis a key challenge. Derry et al. (2005) used cognitive science,“a field that attempts to promote cross-disciplinary integration

l Mod

otofimfcwpcr

3

ssvftbilan

iottotbitmtPmccsspgNrRwmustg“dfb

tta

3

r

G.R. Larocque et al. / Ecologica

f concepts, methods, epistemologies, language, data, and infras-ructures for research and education on cognition, to shed lightn the nature and complexity of interdisciplinary work.” Theacilitated discussion elicited several somewhat novel solutions,ncluding interdisciplinary student training and specific engage-

ent of undergraduate students in interdisciplinary research toacilitate the development of future scientists capable of communi-ating and working effectively across subject areas. Furthermore, itas noted that a commitment to ongoing and persistent interdisci-linary workshops, brainstorming sessions, and catalytic meetingsan help researchers establish successful interdisciplinary collabo-ations.

.2.3. Individual-based models for conservationIt has been recognized recently that conservation problems,

uch as restoration or preservation of critical habitats for wildlifepecies at risk, may be better addressed by modelling the indi-iduals, since it is important to include mechanisms, i.e., theactors that drive the response of the animals to changes inheir environment. Therefore, a critical component of individual-ased models is often capturing animal movement. The challenge

n simulating animal movement is to understand the interre-ated factors (internal and external) that explain how and whyn animal moves, which generates the observed patterns inature.

Several algorithms of varying complexity and behavioural real-sm have been proposed to achieve this goal. A discussion occurredn the merits and limitations of different movement algorithmso adequately mimic animal movement. A consensus emerged onhe necessity to compare algorithms for specific applications andbtain quality data for their validation. Another pervasive issuehat was discussed was the qualitatively dissimilar models madey “ecologists” and “non-ecologists.” For the former group, min-

mizing the number of parameters used is paramount, whereashe latter group focuses more on pattern-oriented modelling. This

ight cause parameter inflation, which is known to allow a bet-er fit of models, but in general it reduces the power of prediction.articipants further discussed whether, at least conceptually, theodeller and the programmer should be separated. The model is

onceptual and the programmer is the person who translates thisoncept into code. Some suggested the modeller and programmerhould be different people as they have different skills; however,ome particular experiences demonstrated that it may be moreroductive to combine these two types of expertise using pro-ramming tools, such as domain-specific languages. For example,etLogo (ccl.northwestern.edu/netlogo/) is used by some, and was

eported to be easy to learn to build relatively complex models.epast (repast.sourceforge.net) is used by the “non-ecologists” andas reported to be powerful, but with a steep learning curve thatay not be convenient for “ecologists.” SELES (www.seles.info) is

sed by several of the ecologists, and is reported to have sometrengths but also a steep learning curve. Several participants men-ioned that research teams often lack people who could bridge theap between the disciplines, resulting in models that are wildlyoff” because the different groups are not able to understand theomain-specific language of the others. An important niche existsor people who can communicate across traditional disciplinaryoundaries.

A last issue concerns granting agencies that often seem reluc-ant to fund research that involves using simulation models to testheory. It was concluded that perhaps it was a matter of maturitynd demonstrations of success that could help change this attitude.

.2.4. Ecological accountingVarious attempts have been made to describe the processes and

esponses of the growth and development of ecosystems (Fath et al.,

elling 222 (2011) 2456– 2468 2461

2001; Jørgensen, 1997), among which ecological accounting servesas a necessary step to incorporate and quantify the contributions ofsocial, economical and environmental issues. Different from con-ventional accounting and environmental accounting, the focus ofecological accounting is to develop practical models and methodsto count the energy, materials and information flow and storage aswell as structure of the ecosystem based on systems and thermody-namic perspectives to establish better strategies for environmentalmanagement (Chen et al., 2009). Its central defining characteristic isto be ecologically relevant and cost-effective in terms of biophysicalmetrics.

To quantitatively assess different socio-economic or compoundecosystems, holistic evaluation based on systems analysis and ther-modynamics can be added to the body of knowledge on the poorcoherence between economic profitability and ecological sustain-ability (Chen and Chen, 2006). Since many methods have beenapplied to ecological accounting, we need to integrate them intoa self-consistent framework to evaluate the genuine wealth ofnations and regions, which is not an easy task due to variousprinciples, scales, orientations and, sometimes, even beliefs (Chenand Chen, 2009). Also, advances in modelling methodology, inte-grated socio-ecological models and eco-sustainability models mustbe included in ecological accounting to connect different sub-ecosystems and present more dynamic and holistic details thattake into account the complexity of ecosystem evolution. Partic-ularly, we need to provide a forum (e.g. websites, FacebookTM orTwitterTM) for both researchers and practitioners who would liketo exchange knowledge, perspectives and ideas for the applicationof ecological models in ecological accounting using a common plat-form and discuss the most recent advances in simulation modelsand assessment methods from both a theoretical and a practicalperspective via suitable media.

Finally, more professionals from universities, governments andthe private sector should be responsible for, involved in or inter-ested in ecological accounting as well as conventional accountingand environmental accounting to make effective comparisons,present and share new ideas, innovations, trends, experiences andconcerns in environmental management so as to transform thegreen GDP into a ‘greener GDP’ or ‘ecological GDP’, based on whichwe could finally establish an important interactive (multi-agent)platform to make societal, economic and environmental develop-ments in an ‘ecologically profitable’ way.

3.2.5. Forest simulation models for sustainable forestmanagement under a changing environment

Forest simulation models are used to predict tree and standproductivity. This information is essential to ensure sustainable for-est management. However, the changing environment creates newconditions that increase the degree of difficulty in model develop-ment. There are several key challenges in this research area. Overall,three ideas were prevalent, with the first being that the participantswould like to have more opportunities to hold informal discus-sions about modelling, have more time to discuss and exchangenew ideas and to hold more coordinated discussions. Validationprocedures were quoted as important. To find adequate datasets tovalidate model processes is essential. However, there is confusionon the meaning of validation (Rykiel, 1996). Finally, estimating andcommunicating uncertainty of model simulations still remains amajor challenge (Larocque, 2008; Larocque et al., 2008, 2011).

Among the other key challenges identified was the questionof sustainable research funding from external grant organiza-tions. Participants felt, however, that there is not much that can

be done about it. This will remain an issue for the forthcomingyears. Partnerships among different research teams may help toalleviate the problem of sustainable funding. Another issue iden-tified was related to the connection of models with biological

2 l Mod

pimiasot

aoapnaml

3

iisaedeDeaam

tscswcflfldNviapfara

cndfdfdes

3a

c

462 G.R. Larocque et al. / Ecologica

rocesses. Examining more biological causality may suggest tak-ng new directions, i.e. towards process-based models and hybrid

odels (Kimmins et al., 1999; Peng et al., 2002). Finally, anotherssue was related to bridging the gap between field experimentsnd modelling, which implies that modellers and experimentershould communicate and work more closely together. In total, 60%f the challenges identified were methodological and communica-ion initiatives.

The main idea proposed during discussions on the generationnd sharing of new ideas was that it is important to present notnly what works, but also what does not work for the models. It islso important to mention that this was recognized by a large pro-ortion of the participants. A model or a research project that doesot deliver the expected results may provide as good information as

model or research project that does. Science has evolved becauseany hypotheses, which were entirely logical, were rejected fol-

owing testing and evaluation.

.2.6. Ecological modelling for environmental flowsDetermining environmental flows has become a fundamental

ssue around the world in dealing with water conflicts amongndustry, agriculture, energy production and ecosystem protectiontakeholders. In recent years, more and more objectives have beendded to environmental flow assessments. The complexity inher-nt to ecological responses to hydrological alteration has made itifficult to quantify the environmental flows required to maintaincosystem health based on the demands of an individual species.ifferent types of ecological modelling exercises have been used innvironmental flow assessments around the world because of theirbility to represent the complexity and processes of ecosystems,nd their relationship with human activities. Research on ecologicalodelling for environmental flows has only begun.The main challenges faced in the modelling of environmen-

al flows are the integration of different models with differenttructures and scales, and data scarcity for the establishment andalibration of integrated models. First of all, ecological modellinghould not only focus on describing eco-hydrological processeshen determining environmental flows; it should also be con-

erned with identifying the relationship between environmentalows and other sectors of water utilization. Second, environmentalows should be studied as part of an integrated system. Moreover,ata scarcity seems to be an eternal topic in ecological modelling.ecessary information should be provided for model construction,alidation, and so on. There should be a trade-off analysis in ecolog-cal modelling application between making a model more complexnd detailed, which requires more information, and keeping it sim-le so that it becomes easier to be used in practice and requiresewer data, which often makes it less accurate. Data collection is

money and time consuming job, so developed regions should beesponsible for helping developing regions, both technologicallynd financially.

Considerable work remains in the development of ecologi-al modelling for environmental flows. Things we can do in theear future include the integration of new ideas from differentisciplines, objectives from different stakeholders, models with dif-erent structures and scales, and complex and simple models forata requirements. Integrated ecological modelling will be help-ul through its application in environmental flow assessment andecision-making. Connections that can be made between the differ-nt types of ecological models for describing complex ecosystemshould be discussed at the next conference.

.2.7. Modelling the consequences of climate change on thegricultural landscapes

Agricultural areas are among the landscapes most vulnerable tolimate change. The long-term changes caused by climate change

elling 222 (2011) 2456– 2468

will affect natural environmental parameters which, in turn, caninduce changes on the total water balance of river basins, soil waterregimes, water movement through soil profiles as well as in plantproduction potential. Mathematical model applications have thepotential to produce a prognosis of the future management of anagriculture landscape. During the symposium, hydrological modelswere discussed as suitable tools to study the impacts of climatechange on hydrologic and water balance. The discussions can besummarized as follows:

• Spatial interpretation of the metrological input factors of riverbasins together with the application of geostatic methodsimprove the quality of output parameters.

• Regional interpretation of climate change impacts with respect tomonthly components of individual elements of water balance andtheir variability provides the basis for effective implementationand adaptation measures.

• The analysis and quantification of the extreme phases of dis-charges within the soil, run-off, and snow submodels for thepresent and for future time horizons specify the probable courseof extreme events.

• The evaluation of the different methodologies of nitrous oxideemission from arable soils under climate change conditions canbe used to clarify uncertainty in the results in order to project andformulate adaptive measures more precisely.

The participants to the symposium agreed on the need to:

1) protect and rationally use water resources, with emphasis onthe different aspects of soil exploitation and protection;

2) develop and adopt efficient mitigation measures in advance;3) conduct feasibility studies and develop wide information tools

to keep agricultural producers and stakeholders informed on theissue and its evolution.

3.2.8. Modelling avian seasonal productivitySustainable management of animal populations requires knowl-

edge of how the natural environment and anthropogenic stressorstogether influence population persistence. The integration of suchknowledge typically occurs through the application of demo-graphic models that combine estimates of vital rates (e.g. survival,fecundity and growth) to estimate the growth rate of a popula-tion. However, in many cases, the vital rates themselves are notdirectly estimable and must, in turn, be estimated via a model. Thisis typically the case with avian fecundity.

What are the key challenges in this research area? Overwhelm-ingly, participants identified unifying methodologies as the mostimportant challenge to further development and adoption of avianfecundity models, followed by collection of high-resolution fielddata. These topics were given priority in the resulting synthesis(Etterson et al., this volume, Sections 3 and 4). Five topics receivedtwo votes each, including distinguishing between pattern and pro-cess, estimating covariances among model parameters, estimatingjuvenile survival after fledging, and competing risks. A further eighttopics received a single vote each, including getting models adoptedby researchers, developing models for tropical birds, better under-standing of start and end dates of the breeding season, renesting,hierarchical models, quantifying spatio-temporal autocorrelationin vital rates, estimating temporary emigration in long-lived birds,and discordance between seasonal and lifetime fecundity in long-lived birds.

3.2.9. Research perspectives in carbon cycle modelling to supportsustainable terrestrial ecosystem management

The development of carbon cycle models for terrestrial ecosys-tems has been an active field of research for the last few decades.

l Mod

Tetfita

1

2

3

3m

i(cc

G.R. Larocque et al. / Ecologica

hese models share a common basic structure, which is the mod-lling of the carbon flow among the different compartments inerrestrial ecosystems. Several models of the latest generation areairly complex and are increasingly used to address environmentalssues, such as global climate change and effects of human interven-ion. Despite the fact that significant accomplishments have beenchieved, several important challenges still remain.

. How to handle and reduce uncertainty. This issue still remainsto be addressed in more depth. Very few carbon cycle process-based models provide uncertainty estimates due to the lackof understanding of several processes, the presence of manysources of uncertainty (model structure, measurement errors,temporal and spatial scales, and imprecise evaluation of natu-ral variability), the limited availability of mechanistic equationsto represent processes, and unknown errors associated withparameters (Larocque et al., 2008). The use of carbon cycle mod-els is increasing in importance to evaluate the impacts of climatechange on the dynamics of terrestrial ecosystems. For policymaking and adaptation, it is essential that carbon cycle modelsprovide answers on the potential long-term effects on net pri-mary productivity, carbon partitioning and potential for carbonsequestration. However, carbon cycle models are still in theirinfancy. Significant progress must be achieved to improve theirpredictive capacity. While research efforts continue, it is impor-tant for model users to understand the limitations of availablecarbon cycle models, and providing uncertainty estimates alongwith predictions would be a significant step in the right direction.

. Bridge the gap between academic and practical approaches with-out reducing the complexity of models. The majority of carboncycle models are characterized by complex structures. Theirdevelopment is conducted mainly in academic circles where themain objective is to better understand the mechanisms that gov-ern the dynamics of terrestrial ecosystems. In general, runningthese models requires a lot of data to provide initial estimatesof the different pools and parameters, which are not necessar-ily easy to obtain for application to actual field scenarios. Foreffective use in policy making, it will be important to find waysto simplify model initialization without compromising modelcomplexity.

. Better linkage between field experiments and modelling. Fieldexperimentation and model development are often performedindependently. For many ecologists and decision-makers, fieldexperimentation is merely considered as a simple step to collectdata for parameter estimation. In reality, field experimentationcan provide important information on the biological consistencyof the structure and representation of processes and contributeto better dealing with uncertainty. For instance, soil carbon poolsand below-ground processes are still poorly characterized. Esti-mates of soil carbon contents in field experiments are usuallyreported as a single number. In reality, soil carbon consistsof many complex organic compounds with different chemicalproperties and turnover rates. The majority of soil carbon cyclemodels partition soil carbon using only three different classes.Given the complexity of soil organic matter, it is not evident thatusing three pools is sufficient to represent the complexity of soilcarbon dynamics.

.2.10. Ecosystem modelling for decision support in wateranagement

The management of water resources must deal with several

ssues, such as pollution, habitat deterioration or invasive speciesBoets et al., 2010). Ecosystem modelling may contribute to betterapturing the quantitative insights in the interactions between theomponents that are involved.

elling 222 (2011) 2456– 2468 2463

Specific ideas and suggestions on the key challenges includedthe integration of habitat suitability (HSM), food web and spatial-explicit dynamic models, a better capacity to work effectively witha large number of models (development of a large set of modelscombined with user-oriented evaluation and selection) and theevaluation of models based on a more diverse set of criteria: special-ized versus parsimonious, development and calculation time, dataneeds for training and simulation, model band width (predictiverange and potential of extrapolation, e.g., Goethals et al. (2007)),ecological relevance, complexity and transparency (Mouton et al.,2009), user-convenience for training, validation and simulation.Multidisciplinary research was particularly discussed to generateand share new ideas, but also to integrate ecological models withhydrological, environmental, and other models as illustrated forinstance in van Griensven et al. (2006). The participants felt thatcross-fertilization from other fields was essential to make progress,both from a theoretical as well as an applied perspective.

3.2.11. Static and dynamic approaches to modelling diversity andcomplexity

Various static and dynamic approaches are available to modeldiversity and complexity in ecological systems. Static approachesare characterized as addressing one point in time, one spatial loca-tion and/or one organizational scale, while dynamic approachesgenerally involve temporal fluctuations, several spatial loca-tions or spatial scales (e.g. local, regional, global), and/or severalorganizational scales (e.g. individual, populations, communities).Definitions of diversity and complexity have seen recent exten-sions and thus should include not only measures such as richnessand Shannon diversity but also measures that incorporate spa-tial structures/configurations, fractals and connectedness in linewith suggestions recently made by ecologists attempting to applycomplex systems science tools and methodologies (Anand et al.,2010). However, improvements from information-theoretical aswell as graph-theoretical approaches can still be made in helpingto improve our definitions of ecological diversity and complexity.

One theme that emerged with respect to the efforts to link staticand dynamic methods was the sampling methodological issuesrequired to move across scales. Several papers presented novelstatistical methodologies which incorporate scale, space and/ortime, moving away from traditional static approaches. The othertheme pertained to modelling issues that try to link spatial pat-terns with dynamics (statistical, computational and mathematical).It was concluded that static measures of diversity and complex-ity could depend greatly on temporal structures and in particularthat spatiotemporal interactions needed to be better understood asthese interactions could result in critical behaviour and/or regimeshifts in ecosystems. Similar calls have been made recently in theliterature (White et al., 2010). Areas of application in which linkingstatic and dynamic approaches were shown to be particularly use-ful were restoration ecology and conservation biology. However,key challenges in this research area include: (1) linking statisticalapproaches and modelling approaches and (2) linking theoreticaland applied approaches. The participants felt strongly that muchinsight can be gained by increasing communication between thestatistical ecology and ecological modelling communities. We sug-gest having further session devoted to this topic.

4. Conclusion

Several ideas and suggestions were discussed at the different

symposia that took place during the 2009 conference of the Inter-national Society for Ecological Modelling (ISEM 2009). As indicatedby the synthesis of the discussions in this paper, common as well asdifferent trends emerged, and this is normal for a conference such

2 l Mod

ametc2focto

464 G.R. Larocque et al. / Ecologica

s ISEM 2009. The large number of common trends suggests thatodellers from different disciplines related to terrestrial and water

cosystems face similar challenges and problems, despite the facthat they have to deal with specific issues related to their own dis-iplines. A major advantage of a scientific conference such as ISEM009 is to provide an invaluable opportunity for delegates comingrom different parts of the world to explore and appreciate vari-us points of view, discuss and exchange different ideas, theories,

oncepts or methodologies, and establish long-lasting relationshipshat may generate good collaboration. We believe that this was onef the great achievements of this conference.

Earth surface modelling and globalecology

Network modelling and systemstheory: advances in network environanalysis

Inco

YUE-HASM based ecosystem scenarioson global level.

Network mutualism in ecological andtrophic networks.

Prep

Yue, T. X. and Z. M. Fan Fath, B.D. and U.M. Scharler MSelf-organization of tropical seasonal

rain forest in southern China.Comparative storage network environanalysis of a seven-compartmentmodel of nitrogen flow in the NeuseRiver estuary: time series analysis.

AthhBC

Lin, H., M. Cao and Y. Zhang Whipple, S.J., S.R. Borrett and B.C.Patten

MMM

Modelling and mapping topographicvariations in forest soils at highresolution: a case study.

Dynamic network analysis:throughflow, storage, cycling andindirect effects.

Mthb

Bhatti, J.S., P.N.C. Murphy, M.Castonguay, J. Ogilvie and P. Arp

Kazanci, C., Q. Ma, B. Tollner, L.Matamba, J. Schramski and B. Patten

MC

Modelling sediment-depositionpatterns of a flash flood in aresidential area using Erosion 3D.

Evidence for the dominance of indirecteffects in trophically-based ecosystemnetworks.

Incopth

Arévalo, S.A., J. Schmidt, W. Schmidtand A. Michael

Salas, A.K. and S.R. Borrett BADCB

Modelling spatial structure inecological data usingeigenfunction-based spatial filteringmethods.

Reconnecting environs to theirenvironment.

Hcoin

Blanchet, F.G. Borrett, S.R., M.A. Freeze and A. K. Salas LaValidation and uncertainty analysis of

an ecosystem model forwheat-maize double croppingsystem over the North China Plain.

Sufficient conditions for thresholdinsensitivity of network environanalysis dominance indicators toboundary input.

Blod

Mo, X. Freeze, M.A., S.R. Borrett and A.K. Salas RUsing spatial statistics and landscape

index to quantitatively analyze thespatial characteristic on simulationimage.

Do implicate forces in nature keepsystems at or near steady states?

SuuwLa

Wang, J.H., Y. He and W. Ren Patten, B.C. LaCLaMD

Modelling and simulating ecosystemservices? a brief summary of recentstate of the art and exampleapplications for ecosystemmanagement.

Structural and functionaldecomposition of ecological networksand their system-wide properties.

Aseanag

Seppelt, R. Gaff, R. and C. Kzzanci CPCMD

elling 222 (2011) 2456– 2468

Acknowledgments

Sincere thanks are extended to the members of the LearningOrganization Community of Practice (LOCOP) of Natural ResourcesCanada at the Laurentian Forestry Centre: Marie Anick Libo-iron, Deirdre Moore, Isabelle Lamarre, Sébastien Clément, CarolineGosselin, Josyanne Lamarche, Lucie Sokolyk and Lise Caron. Inparticular, Deirdre Moore held meetings with the symposium orga-nizers to develop a discussion plan, Marie Anick Liboiron organizeda special training session for the facilitators, and Sébastien Clémentspent considerable time writing a synthesis of the discussions held

in all the symposia.

Appendix A. List of symposium presentations at ISEM 2009.

dividual-based models fornservation

Ecological accounting

redicting spatial and temporalcolonization of large carnivore

opulations: wolves in the Italian Alps.

Ecological accounting based on exergy

arucco, F. and E.J.B. McIntire Chen, B.n agent-based model to investigatee impact of increased levels of

uman presence on wolf behaviour inanff and Kootenay National Parks,anada.

Ecological economic evaluation ofBeijing urban ecosystem based onemergy as embodied cosmic exergy:historical prospective stuty1978–2004.

orshed, A. Sk., M. Musiani, G.cDermid, M. Hebblewhite and D. J.arceau

Jiang, M. -M., G. -Q. Chen, B. Chen andZ. -F. Yang

oving through complex landscapes:e bane and promise of individual

ased models.

Effect of anaerobic fermentationresidues on a chromium-contaminatedsoil-vegetable system.

cIntire, E.J.B., C. Schultz and E.E.rone

Duan, N., X. Liu and C. Lin

dividual-based models, functionalnnectivity conservation and

erformances of ecological networks:e TenLamas project.

Network calculation of embodiedcosmic exergy and their consumptionfor Chinese economy 2005: aninput–output analysis.

aguette, M., A. Coulon, V.M. Stevens,. Chaput-Bardy, C. Turlure, J. Clobert,. Schmeller, C. Bleay, J.M.J. Travis, S..F. Palmer, T. Hovestadt and K.A.arton

Chen, Z.-M., B. Chen and G.-Q. Chen

ow to achieve generalization innservation models: Applications todividual-based modelling.

Ecological design and evaluation forthe new industrial district ofCaofeidian.

tombe, G., L. Parrott and D. Fortin Li, C. and Y. Yangiased correlated random walk vs forayop: which movement hypothesis

rives a butterfly metapopulation?

Research on the risk assessment ofGreen Manufacturing decision-makingunit and its application.

ompré, G. and E.J.B. McIntire Tan, X., L. Cheng and Y.Y. Wangpporting ecosystem management

sing agent-based modelling ofhalewatching activities in the St.wrence estuary.

Emergy and exergy synthesis of fourforest restoration modes in lowsubtropical China.

ndry, J.-A., L. Parrott, C. Chion, C.avalcante de Albuquerque Martins, P.montagne, S. Turgeon, R. Michaud, D.arceau, N. Ménard, G. Cantin and S.ionne

Lu, H.-F., Z.-H. Wang and B. Chen

nalysis of emerging primary andcondary patterns for the selectiond validation of a spatially explicitent-based model.

An analysis of the urban metabolism ofBejing based on emergy synthesis.

hion, C., P. Lamontagne, S. Turgeon, L.arrott, J.-A. Landry, D. Marceau, C.avalcante de Albuquerque Martins, R.ichaud, N. Ménard, G. Cantin and S.ionne

Zhang, Y., Z. Yang and B. Chen

G.R. Larocque et al. / Ecological Modelling 222 (2011) 2456– 2468 2465

Earth surface modelling and globalecology

Network modelling and systemstheory: advances in network environanalysis

Individual-based models forconservation

Ecological accounting

Finding important sub networks inecological network analysis: findingcoefficient vectors using fluxdecompositions.Luper, D., J.R. Schramski, C. Kazanci andH.R. ArabniaConservation laws: interdisciplinarymodelling lessons in engineering andecological compartment analysis usingfirst principles and a case study ofecosystem energetics.Schramski, J.R., E.W. Tollner, D.K.Gattie, B.C. Patten, and J. Jambeck

Forest simulation models forsustainable forest managementunder a changing environment

Ecological modelling forenvironmental flows

Modelling the consequences of climatechange on the agricultural landscape

Modelling avian seasonal productivity

Seasonal fecundity models: Criticalassessment of questions and direction.

Modelling forest growth, productivityand carbon dynamics under achanging environment using hybridmodelling approach.

Environmental flow requirements forthe habitat of Chinese shrimp (Penaeuschinensis) in the Yellow River Estuary,China.

Assessment of climate change impactson water balance in an agriculturalriver basin.

Grzybowski, J.A. and C.M. Pease

Peng, C. Sun, T.and Z. Yang Novotná, B., E. van Bochove and B.Viliam

Which life history componentsdetermine breeding productivity forindividual songbirds? A case study ofthe Louisiana Waterthrush.

Effects of climate change and CO2

fertilization on water use efficiencyof terrestrial ecosystem of China.

The effects of environmental flowsregulation on food webs in theBaiyangdian wetland.

Model evaluation of nitrous oxide(N2O) emissions from arable soils ofSlovak Republic in condition of climatechange.

Mattsson, B.J.

Jiang, H., Q. Zhu, C. Peng and J. Liu Chen, H. and Z. Yang Horák, J. and B. SiskaMarkov chain estimation of avianseasonal fecundity

Challenges in the development andapplication of forest simulationmodels under a changingenvironment: Using different modeltypes to answer resourcemanagement questions.

Seasonal environmental-flow demandcalculation of reed community(phragmites australis var.baiyangdiasis) under differentmeteorological conditions in thebaiyangdian lake, China.

Etterson, M.A., R.S. Bennett, E.L.Kershner and J.W. Walk

Larocque, G.R., L. Archambault, C.Delisle, N. Luckai and S. Adhikari

Wang, Q., Z. Yang and J. Liu

Retrospective estimation of breedingphenology of American Goldfinch(Carduelis tristis) using patternoriented modelling.

Simulating the effects ofpre-commercial thinning on forestgrowth of jack pine stands under twoscenarios of climate change using aprocess-based model.

Estimation of dynamic environmentalflows of Baiyangdian Lake, a shallowfreshwater lake.

Potvien, A., M. Etterson and R. Regal

Wang, W., X. Zhou, C. Peng, G.R.Larocque, D. Kneeshaw, X. Lei, T.Zhang and J. Sun

Yang, W. and Z. Yang

Food limits annual fecundity of amigratory songbird: an experimentalstudy.

The evolution of a growth model:PrognosisBC.

Study on sustainable water use basedon ecological network analysis withthe consideration of environmentalflow using ecological network analysis.

Nagy, L.R. and R.T. Holmes

Marshall, P.L., A.-A. Zumrawi and V.M.LeMay

Li, Y., Z. Yang and B. Chen

Calculations on the back of an envelopemodel: Applying seasonal fecunditymodels to species’ range limits.

IVY: An individual-tree growth modelfor complex-structured stands.

Development of an inexactoptimization model for waterresources management with theconsideration of uncertainenvironmental flows.

Olsen, B.J., M.A. Etterson and R.Greenberg

2466 G.R. Larocque et al. / Ecological Modelling 222 (2011) 2456– 2468

Earth surface modelling and globalecology

Network modelling and systemstheory: advances in network environanalysis

Individual-based models forconservation

Ecological accounting

Groot, A. Cai, Y. and G. HuangAnnual productivity in Greater SnowGeese: which fecundity parameter isthe best predictor and why?

Adapting a spatial forest simulationmodel for new managementpractices: approach and challenges.

Gauthier, G., C. Juillet, J. Bêty and M.Morrissette

Mailly, D.Seasonal fecundity of the commonLoon.

Sensitivity of forest landscape modelsto individual growth rate functions:do small scale processes matter forlarge scale question?

Evers, D.

Elkin, C., C. Temperli, B. Reineking andH. Bugmann

Shifts in predation risk atcamera-monitored Northern Bobwhitenests as a result of meso-mammalianpredator removal in the southeasternU.S.A.

The importance of using the correcttree architecture instructural-functional process-basedmodels.

Ellis-F., S.N., M.J. Conroy, W.E. Palmer,and J.P. Carroll

Schneider, R.Non-linear patterns of parasitism andnest survival: implications forestimating avian productivity.

Modelling diameter growth inresponse to varying silviculturaltreatments within mixed-speciesstands located in coastal BritishColumbia.

Post van der Burg, M., L.A. Powell andA.J. Tyre

Rathbun, L.C., V. LeMay and N. SmithModelling the effects of chemicalexposure on avian seasonalproductivity: Importance of differencesin breeding strategies.Bennett, R.S. and M.A. Etterson

Research perspectives in carbon cyclemodelling to support sustainableterrestrial ecosystem management

Ecosystem modelling for decisionsupport in water management

Static and dynamic approaches tomodelling diversity and complexity

Modelling carbon budget in fire-proneecosystems: Translating acrossscales from global to landscape.

Applications of habitat suitabilitymodels in marine management andresearch

What structures beetle communities inboreal forest?

Mouillot, F. Willems, W.M.M., S. Degraer, V.Vanlancker, E. Verfaillie, M. Vincx andP.L.M. Goethals

Blanchet, F.G.

Data-model fusion approach in globalcarbon cycle research: recentdevelopment and future challenges.

Inverse analysis of the structure of theVenice Lagoon ecosystem.

Regional pattern design for ecosystemrestoration: a case study.

Peng, C. Brigolin D., C. Savenkoff, M. Zucchetta,F. Pranovi, P. Franzoi, P. Torricelli andR. Pastres

Ma, K., M. Anand, G. Li and B. Fu

Quantifying the uncertainty of thecarbon dynamics of Canada’smanaged forest.

Plugging the holes in Lake Victoria:Summarising what is known about itand what has yet to be studied.

Evolutionary biogeography improvesbiodiversity prediction in the BrazilianAtlantic rainforest.

Metsaranta, J., W.A. Kurz, G. Stinson,E.T. Neilson, G.J. Rampley, C. Smythand C. Shaw

Downing, A.S., W.M. Mooij, E.H. vanNes, J.H. Janse and M. Scheffer

Carnaval, A.C. and C.C. Moritz

Modelling carbon stocks in unmanagedforests in northeastern Ontario,Canada.

Application of decision trees to analyzethe ecological impact of invasivespecies in Polder lakes in Belgium

Multifractal analysis of forest spatialdistribution pattern over time.

Ter-Mikaelian, M., S.J. Colombo and J.Chen

Everaert, G., P. Boets, K. Lock and P.Goethals

Zhang, Y., K. Ma and B. Fu

Using the Carbon Budget Model of theCanadian Forest Sector (CBM-CFS3)to examine the impact of harvest andfire on carbon dynamics in selectedforest units of the Canadian Borealshield.

Species abundance patterns in benthicmacro-invertebrate communities insteams in response to disturbance.

Assessing mortality across habitatamount and configuration usingindividual based models.

Luckai, N., G.R. Larocque, L.Archambault, D. Paré, R. Boutin, A.Groot and A. Innerd

Cho, W.-S., T. Van Nguyen and T.-S.Chon

Ribeiro, M.C., A.C. Martensen, M.-J.Fortin and J.P. Metzger

G.R. Larocque et al. / Ecological Modelling 222 (2011) 2456– 2468 2467

Earth surface modelling and globalecology

Network modelling and systemstheory: advances in network environanalysis

Individual-based models forconservation

Ecological accounting

Simulations show deteriorating carbonbalance of Rocky Mountain forests inthe next century.

Validation of ecological models tosupport water management:integration of mathematical, ecologicaland management-oriented criteria.

Modelling impacts of disturbance ondiversity using a markov model ofsuccession in the boreal forest.

Boisvenue, C. and S.W. Running Goethals, P. and A. Mouton Richardson, M.J., M. Anand and H.J.Eberl

Simulating the dynamics of forestvegetation diversity under clearcutting and natural development atdifferent climate change scenarios incentral European Russia.

Fire feedbacks with vegetation and thespatial structure of savannas

Khanina, L., M. Bobrovsky, A. Komarov,A. Mikhailov, V. Shanin and S.Bykhovet

Beckage, B. and C. Ellingwood

Modelling carbon and nitrogensequestration in boreal forestecosystems of central Russia underdifferent climate change scenariosand forest management regimes.

Facilitation and the emergence ofcommunity structure inmetacommunities.

Shanin, V.N., A.S. Komarov and S.S.Bykhovets

Filotas, E., M. Grant, L. Parrott and P.A.Rikvold

Effective use of an experimental designin balsam fir (Abies balsamea (L.)Mill.) and black spruce (Piceamariana (Mill.) B.S.P.) forestecosystems located along a climaticgradient for the development of acarbon cycle process-based model.

A model study of storm surge effectson a mangrove - hardwood hammockecotone

Larocque, G.R., D. Paré, R. Boutin, L.Sarr, V. Lacerte and C. Ansseau

DeAngelis, D.L., Teh, S.Y. Su and .L daSilveira Lobo Sternberg

Modelling soil carbon dynamics underdifferent traditional systems ofagriculture and forestry in EuropeanRussia.

Bobrovsky, M., A. Komarov and V.Shanin

Dissolved organic carbon and Ndynamics in forest streams, inrelation to geospatial andhydrothermal controls.

Castonguay, M., T.A. Clair and P.A. ArpModelling of elements dynamics in

mineral and organic forest soils: theRomul model expansion.

Komarov, A., O. Chertov, Y. Khoraskina,S. Bykhovets, A. Larionova and M.Bezrukova

Modelling the decomposition and Nmineralization in ramin wooddowels (Gonystlylus bancannus)using a common climate function forthe LIDET sites across the USA.

Smith, A.C., J.S. Bhatti, M.E. Harmonand P.A. Arp

Estimating California forest biomasschange using process model and landcover disturbance data.

Liu, J., J.E. Vogelmann, Z. Zhu, C.H. Key,B.M. Sleeter, D.T. Price, J.M. Chen,M.A. Cochrane, J.C. Eidenshink, S.M.Howard, N.B. Bliss and H. Jiang

Effects of climate, age and stand originon the above ground NPP (ANPP) ofjack pine stands along the borealforest transect case study (BFTCS).

Bhatti, J.S., T. Varem-Sanders and M.

R

A

A

Wang

eferences

alders, I.H., Aitkenhead, M.J., 2006. Agricultural census data and land use modelling.Computers, Environment and Urban Systems 30 (6), 799–814.

lessa, L., Chapin III, F.S., 2008. Anthropogenic biomes: a key contribution to earth-system science. Trends in Ecology and Evolution 23, 529–531.

Anand, M., Gonzalez, A., Guichard, F., Kolasa, J., Parrott, L., 2010. Ecological sys-

tems as complex systems: challenges for an emerging science. Diversity 2, 395–410.

Boets, P., Lock, K., Messiaen, M., Goethals, P.L.M., 2010. Combining datadriven meth-ods and lab studies to analyse the ecology of Dikerogammarus villosus. EcologicalInformatics 5, 133–139.

2 l Mod

B

C

C

C

C

D

E

E

F

F

G

J

J

K

L

468 G.R. Larocque et al. / Ecologica

onavito, N.L., Gordon, C.L., Inguva, R., Serafino, G.N., Barnes, R.A., 1994. Maximumentropy-driven Bayesian reasoning in data classification. Telematics and Info-matics 11 (4), 295–308.

hen, B., Chen, G.Q., 2006. Ecological footprint accounting and analysis of the ChineseSociety 1981–2001 based on embodied exergy. Ecological Economics 61 (2–3),355–376.

hen, B., Yang, Z.F., Hao, F.H., Chen, G.Q., 2009. Emergy-based energy and materialmetabolism of the Yellow River basin. Communications in Nonlinear Scienceand Numerical Simulation 14 (3), 923–934.

hen, G.Q., Chen, B., 2009. Extended-exergy analysis of the Chinese society. Energy34, 1127–1144.

ohen, J.E., 2004. Mathematics is biology’s next microscope, only better; biology ismathematics’ next physics, only better. PLoS Biology 2, 2017–2023.

erry, S.J., Schunn, C.D., Gernsbacher, M.A., 2005. Interdisciplinary Collaboration:An Emerging Cognitive Science. Mahwah. Lawrence Erlbaum Associates, NewJersey.

llis, E.C., Ramankutty, N., 2008. Putting people in the map: anthropogenic biomesof the world. Frontiers in Ecology and the Environment 6, 439–447.

wel, K.C., 2001. Natural resource management: the need for interdisciplinary col-laboration. Ecosystems 4, 716–722.

ath, B.D., Patten, B.C., Choi, J.S., 2001. Complementarity of ecological goal functions.Journal of Theoretical Biology 208, 493–506.

ernandes, P.B., 1998. Technological advances in high-throughput screening. Cur-rent Opinion in Chemical Biology 2, 597–603.

oethals, P.L.M., Dedecker, A.P., Gabriels, W., Lek, S., De Pauw, N., 2007. Applica-tions of artificial neural networks predicting macroinvertebrates in freshwaters.Aquatic Ecology 41, 491–508.

ørgensen, S.E., 1997. Ecological modeling by ‘Ecological modelling’. Ecological Mod-elling 100, 5–10.

ørgensen, S.E., Fath, B.D., Grant, W., Nielsen, S.N., 2006. The editorial policy of Eco-logical Modelling. Ecological Modelling 199, 1–3.

immins, J.P., Mailly, D., Seely, B., 1999. Modelling forest ecosystem net primaryproduction: the hybrid simulation approach used in FORECAST. Ecological Mod-elling 122, 195–224.

arocque, G.R., 2008. Forest models. In: Jøgensen, S.E., Fath, B.D. (Eds.), Ecologicalmodels of 17 Encyclopedia of Ecology, vol. 2. Elsevier, Oxford, pp. 1663–1673.

elling 222 (2011) 2456– 2468

Larocque, G.R., Bhatti, J.S., Ascough II, J.C., Liu, J., Luckai, N., Mailly, D., Archambault,L., Gordon, A.M., 2011. An analytical framework to assist decision makers in theuse of forest ecosystem model predictions. Environmental Modelling & Software26, 280–288.

Larocque, G.R., Bhatti, J.S., Boutin, R., Chertov, O., 2008. Uncertainty analysis in car-bon cycle models of forest ecosystems: research needs and development of atheoretical framework to estimate error propagation. Ecological Modelling 219,400–412.

Larocque, G.R., Mailly, D., Gaudreault, M., 2009. Forest modelling in Quebec: context,challenges and perspectives. Forestry Chronicle 85, 702–705.

Leimu, R., Koricheva, J., 2005. Does scientific collaboration increase the impact ofecological articles? Bioscience 55 (5), 438–443.

Mouton, A.M., De Baets, B., Van Broeckhoven, E., Goethals, P.L.M., 2009. Prevalence-adjusted optimisation of fuzzy models for species distribution. EcologicalModelling 220, 1776–1786.

Peng, C.H., Liu, J.X., Dang, Q.L., Apps, M.J., Jiang, 2002. TRIPLEX: 1 A generic hybridmodel for predicting forest growth and carbon and nitrogen dynamics. Ecolog-ical Modellling 153, 109–130.

Phillips, J.D., 2002. Global and local factors in earth surface systems. Ecological Mod-elling 149, 257–272.

Rykiel jr, E.J., 1996. Testing ecological models: the meaning of validation. EcologicalModelling 90, 229–244.

Schramski, J.R., Gattie, D.K., 2009. Cross-disciplinary informed ecological networktheory. Ecological Modelling 220, 3111–3112.

Solidoro, C., Bandelj, V., Cossarini, G., Libralato, S., Canu, D.M., 2009. Challenges forecological modelling in a changing world: Global changes, sustainability andecosystem based management. Ecological Modelling 220, 2825–2827.

van Griensven, A., Breuer, L., Di Luzio, M., Vandenberghe, V., Goethals, P., Meixner, T.,Arnold, J., Srinivasan, R., 2006. Environmental and ecological hydroinformaticsto support the implementation of the European Water Framework Directive forriver basin management. Journal of Hydroinformatics 8 (4), 239–252.

White, E.P., Ernest, S.K.M., Adler, P.B., Hurlbert, A.H., Lyons, S.K., 2010. Integratingspatial and temporal approaches to understand biological diversity. Philosoph-ical Transactions of the Royal Society B 365, 3633–3643.

Yue, T.X., 2010. Surface Modeling: High Accuracy and High Speed Methods. CRCPress, Boca Raton, FL, 2010.