November, 2010 Dear workshop participants, Re: Understanding the State of Tauranga Harbour, and Identifying the Most Important Issues that Impact on its Health, using Mediated Modelling Thank you for committing your time to the project: “Manaaki Taha Moana,” using the Mediated Modelling (MM) component to help us better understand what things have the greatest impact on the health of the Tauranga Harbour, and how those things interact with each other. Attached is the list of confirmed participants. I believe this is a rich group able to take a broad perspective on issues related to Tauranga harbour historically, now and in the future. I’m asking that in thinking about your contributions to this process, you think beyond your specific personal and institutional interests and apply your thinking and experience toward the question of what is good for Tauranga Harbour as a whole. To keep this process as open and inclusive as possible, interested people beyond the participant list may observe the workshops. All preparations and outcomes of each workshop will be published on the MTM website (www.mtm.ac.nz). This is an iterative process and there will be plenty of opportunities during breaks and between workshops for observers to provide feedback on critical ideas. The Foundation for Science and Technology (FRST) is funding this research programme (Manaaki Taha Moana (MTM) at Massey University). There are other components of the MTM research that are happening alongside the MM workshops, including collection of both Matauranga Maori and published western science literature about the state of the harbour. The mediated modelling workshops are designed to bring representatives from all important groups that impact on the harbour together, to help come to an understanding of the state of the ecosystems in the harbour, and how other cultural, economic and political issues impact and rely on the state of the harbour. Within the mediated modelling format, it is my task, as the facilitator of the workshops, to provide the participant group with the ‘building blocks’ of how the different components can be put together, and the steps that need to be taken to jointly develop a simulation model as a group. Exactly how this participant group chooses to take those steps is unknown at this stage, and will evolve through discussion at the workshops themselves. I will facilitate that discussion and continue to interpret and summarise the essence of your discussion in modelling terms and draw your attention to the evolving model, with help from the modelling team (Aaron, Sarah, Mark and myself), and data gatherers (Lydia, Eric, Sarah and Derrylea). The model will be visible at all times during the workshops. Meetings with individuals or sub-groups between workshops will be organised as needed to follow up on workshop issues and/or prepare for subsequent workshops. The construction of a group model (and associated learning) is time consuming and there is no quick fix. It is about synchronising the assumptions, clarifying definitions and communicating inter-linkages between social, economic, ecological and cultural aspects 1
in simple yet meaningful ways. If successful, the model building process will help stakeholders in Tauranga moana better understand the dynamics of the harbour, and help to identify the most critical factors that warrant detailed investigation in the next stage of our Manaaki Taha Moana research programme. The model itself is of secondary importance; its prime importance is as a means to maintain an in-depth and fact-based discussion. In the longer term, the model is expected to assist with adaptive management. It is possible that stakeholders may wish to meet with project team members between workshops to resolve specific model-related issues. This intention is usually brought forward during the workshops, to make it clear to the entire group what progress can be expected at the next meeting. Requests for meetings will be announced to the group in order to maintain an open dialogue and avoid a sense that covert developments are skewing the direction of the overall model. At the beginning of each workshop, the progress made in the development of the model will be presented, and participants will have a chance to point out concerns, mis-interpretations or mis-representations. A group model-building process is an on-going discussion. However, with the assistance of the Manaaki Taha Moana research team, Iâ€™m confident the model clean-up between workshops will happen within the comfort zone of the group. For the first workshop, you do not need to bring data or documents. Please, come with an open mind and the willingness to listen for possibilities and to brainstorm constructively. The appendices provide additional detail about the content of the first workshop, mainly in the form of excerpts from my book, â€˜Mediated Modeling: a systems dynamics approach to environmental consensus buildingâ€™, published by Island Press, Washington, DC. Appendix 1: Confirmed participant list and project team Appendix 2: Schedule, including topics for the first workshop Appendix 3: Workshop guidelines Appendix 4: Introduction to systems dynamics thinking and software Appendix 5: Preliminary model Appendix 6: Defining success of a mediated modeling process and Action Plan Appendix 7: Adaptive Capacity Appendix 8: Survey Reading the attachments of this communication is not required to participate effectively at the workshops. The attachments are intended to provide an opportunity to prepare for those who wish to and will be posted on the website. I am looking forward to working with you on this important issue. Sincerely,
Marjan van den Belt, PhD Director, Associate Professor Ecological Economics Research New Zealand Massey University 2
Mediated Modelling Participants at 1st Workshop, 17 November 2010 Organisation
Tauranga City Council Tauranga City Council Western BOP DC EBOP Regional Council Royal Forest and Bird Forest and Bird DOC DOC Zespri Ballance Fertilizers Federated Farmers Toi Te Ora Public Health Customary Fish Committee Polytechnic Institute Tangata Whenua Tangata Whenua
Graeme Dohnt Jane Groves
Drainage Services Manager Pollution Prevention Officer
Glenn Ayo Rob Donald
Environmental Officer Policy/Science Leader
Al Fleming Eila Lawton Chris Clark Dan Rapson Alistair Mowatt Represented Barry Roberts Trieste Ngawhika Paul Borell
Bay Of Plenty Representative Kaimai Mamaku Campaign Committee Biodiversity Manager Marine Ranger Innovation Leader â€“ Sustainability
Dean Tully Sarah Wairepo Tracey Ngatoko Carlton Bidois Paula Werohia Kevin Haua
Marine Department Also research team, WakaTaiao Also research team, WakaTaiao
Tangata whenua Tangata whenua Manaaki Taiao, Tangata whenua Tauranga Environment Ctr 2 x other tangata whenua reps
BOP Fed Farmers Tauranga/Katikati branch Health Protection Officer Board Member
Also research team, WakaTaiao Also research team, WakaTaiao Ngati Pukenga Customary Fisheries Ctee; Manaaki Taiao Trustee
Other groups who will be represented in subsequent workshops: Port of Tauranga, Landcare Trust Representatives from other sector groups who use or impact on the harbour have also been invited, and we are awaiting responses. Research Team Members at Mediated Modelling Workshop 1, as observers/researchers only: Massey: Marjan van den Belt (facilitator); Murray Patterson; Derrylea Hardy Cawthron Institute: Eric Goodwin Waka Taiao: Lydia Hale (+ others listed in participant list) WakaDigital: Aaron McCallion and Mark Berry
Appendix 2 Schedule and specifics of the first workshop Workshop 1: 9am-5pm, Wed 17 Nov 2010; Level 1, 12 Elizabeth Street, Tauranga Topics: o Welcome o Karakia o Mihi, introduction of participants o Workshop guidelines: (a) the rights and responsibilities (b) behavioral guidelines, (c) consensus enhancing procedures (d) modeling oriented guidelines: Appendix 3 o Introduction to the software (STELLA): Appendix 4 o Presentation of preliminary-model: Appendix 5 o Problem definition or “questions the model should answer” o Model scope, time horizon, time step and model sectors o Using stocks, flows, auxiliary variables, information connectors to start developing a qualitative model structure. o Choose preliminary topics for subsequent workshops, such as a focus on each of the model sectors presented in the preliminary model. Workshop 2: 9am-5pm Wednesday 15 December 2010; location to be determined. Topic: Use the modelling building blocks to continue with the qualitative model structure and define/evaluate quantitative data and information. Based on workshop one, there may be a specific topic for this workshop, such as “land use”, or “solutions”, for example. Define data gathering needs. Workshop 3: 9am-5pm Wednesday 19 January 2011; location to be determined. Topic: An update on the model, and how data was incorporated, will be presented. Continue with quantitative model building. There may be a specific topic for this workshop, such as “water quality and harbour indicators”, for example. Define data gathering needs. Workshop 4: 9am-5pm Wednesday 16 February 2011; location to be determined. Topic: An update on the model, and how data was incorporated, will be presented. Continue with quantitative model building, and quantify relationships. Start developing scenarios. There may be a specific topic for this workshop such as “socio-economic impacts”, for example. Data gathering needs defined again. Workshop 5: 9am-5pm Wednesday 16 March 2011; location to be determined. Topic: Simulation and adjusting the model, identifying gaps and findings. There may be a specific topic for this workshop such as “cultural values impacts”, for example. Data gathering needs defined again. Workshop 6: 9am-5pm Wednesday 21 April 2011; location to be determined. Topic: Run final simulation of model; present findings/recommendations of key issues for further detailed research in case studies; and possibilities for group action beyond the workshops.
Rights and responsibilities •
It is each participant’s right and responsibility to be unique. Each participant is respected for the perspective he/she brings to the table. Creativity stems from divergent ideas. Participants also have the responsibility to communicate their perspectives as concisely and clearly as possible. We would like to encourage participants not to use jargon, so that everyone is able to understand the discussions. Nobody knows everything, but together a group knows more than any one person alone. Ideas generated in the group belong to the entire group, not to any specific individual. Assume that all those present are the right persons for the task. Opinions about the balance of the group can be stated through this pre-workshop survey, and will go on record. However, once the workshops begin, we encourage all participants to embrace the group that is there, so that the energy of the workshops is focused on positives and new possibilities.
Behavioral guidelines •
We encourage participants to actively listen to the views of other people at the workshop. When you do contribute yourself, please try to keep to a maximum two minutes per discussion point. Everybody will then get the chance to contribute his/her share, without any one person or viewpoint dominating the workshop Creativity can only flow when destructive criticism is withheld. Withhold judgment on what another participant is saying, until that participant has made him/herself understood. Allow ideas to exist and grow, take them in, actively listen, listen for possibilities, allow for the possibility of being inspired, even when you would prefer to immediately shut out the ideas based on your own rationale. Sorting through the options presented during the workshops will be done in due course. Please ask questions for understanding, rather than for the purpose of invalidating a contribution. When you do voice disagreement, please do so without being disagreeable. Emphasise the situation rather than the people involved. Don’t become personal in disagreements. Maintain a focus on a logical train of thought, rather than “You are wrong” statements. Such a statement will invoke a request (by the workshop facilitator) to substantiate with “evidence” to the group. Instead, encourage coparticipants to feel competent; value disagreements as a source of creative ideas. In the workshops, we will focus on that which is equally good for all. This goes back to “increasing the pie” before “dividing the pie”. Often there is an intricate ongoing tension between “what is best for the overall objective” versus “what is best for me or the group I’m representing”. A focus on the latter is naturally present during the MM workshops. However, the challenge is to allow a focus on “what is best for the overall objective” to co-exist. In addition, “what is best for me” is often relative to the extent to which other stakeholder’s needs are satisfied. One way to open the pathway toward “what is best for the overall objective” is to concentrate on underlying concerns and interests rather than on stalemate positions. Another way is to emphasise future improvements rather than dwelling on the past. 5
There is no need to impress others. The ability to explain complex information in lay terms is more valuable than to confuse people with expert language.
Consensus enhancing procedures and conflict handling strategies • •
Free discussion geared toward creativity is the primary goal. Consensus is never a requirement, but it is a secondary goal as the crucial aspects of an issue are systematically lifted out of the free discussion and narrowed down to consensus. A request for a show of hands (up in the air) is used in order to identify the perception in the group about a specific issue. Even though not every participant may be equally happy with a specific solution, it may be necessary for everyone to “live with” a proposed step. This procedure may generate a temporary verdict, which usually has some participants going away after the workshop to study that subject in order to expand on their beliefs and provide the group with more facts at the next workshop. Interestingly enough, the “voting” procedure is usually not required when addressing qualitative matters, but mainly reserved for quantitative matters. Consensus is achieved in the absence of “reasoned and paramount objections”. In mediated modelling, a group will seek unanimity; however, an absolute consensus on every issue during the entire modelling process is impossible. In the interest of time, groups continue if there is “no reasoned and paramount objection”. However, when turning toward the absence of reasoned and paramount objections too early, the group runs the risk of ending up with solutions that all can live with, but the solutions may lack teeth and may therefore not be useful. Ad hoc meetings with individual participants, groups of stakeholders or experts may occur between the workshops as the need arises. In the data gathering and quantification stage, ad hoc meetings with small groups are often required to figure out specific problems or to determine the most elegant reflection of a specific discussion. At this point, the whole group has identified specific challenges and is looking for common answers. It is important that small steps are taken, both during these ad hoc meetings and between the meetings of the larger group, in order to retain the cohesive understanding of the participants.
Modelling guidelines •
The final model is a joint product of the team learning experience. The team learning is as important as the model, as an end product. The final model itself serves as a vehicle to recreate the insights gained during the workshops and communicate them to other people. A model is always an abstraction of reality. A model can only be evaluated for the purpose for which it was designed. Synthesis is the art of leaving things out. A minority of the variables that could be chosen should explain the majority of the system’s behaviour. A relatively small part of all possible variables that could be incorporated in a model are chosen to explain the behaviour of the whole system. This means that those constructing a scoping model should aim for simplicity and elegance, not for a high degree of detail.
Introducing system dynamics thinking
The modelling approach that will be used is called ‘system dynamics’. The introduction to system dynamics thinking on the High Performance Systems website (www.hpsinc.com/index.aspx) states: “To make sense of reality, we all simplify it. These simplifications are called mental models. We simulate our mental models in order to determine which course of action to implement, which alternative to choose, which strategies will best achieve our objectives. History shows that our choices and decisions often leave us with holes in our feet because: 1. The assumptions constituting the mental models we build are not sufficiently congruent with the reality they are seeking to represent 2. Our simulations of these models do not correctly trace out the dynamic consequences implied by the assumptions in the models. ‘Systems Thinking’ is an approach that can help us to construct mental models that are more likely to be congruent with reality, and to then simulate these models more accurately. Systems Thinking thus increases the likelihood that we will produce the consequences that we intend.”
From a system dynamics perspective, we are interested in non-linear behaviour within a system often explained by feedback loops and time lags.
Introduction to STELLA
The system dynamics software used is called STELLA. The software can be found at High Performance Systems, Inc: http://www.hps-inc.com/index.aspx. A run-time only version is downloadable free of charge and allows you to run models, but not save the changes to a model. You are welcome to download this software and learn how to use it. However, this is not a requirement of your participation. The research team running the Mediated Modelling will be responsible for building the computer model, and for explaining at each workshop what has been done in each step as the model is developed, to reflect the decisions made by participants in each workshop. In STELLA, there are three communicating layers that contain progressively more detailed information on the structure and functioning of the model (see Figure 1). The lowest layer contains the difference equation, generated by the model structure in the middle level. This level shows the model structure by icons. The graphic representation of these units are connected and manipulated on the screen to build the basic structure of the model. This process is made transparent to a group when the computer screen is projected. The middle layer is displayed during the construction phase. Icons represent the basic structure of the model and provide an input pathway for subsequent data. Once the basic structure of the model is laid out, initial conditions, parameter values and functional relationships can be specified. Input data can be entered in graphical or tabular formats.
The highest layer is the "user interface". In the final stage users can easily access and operate the model from this level. With the use of slide-bars, a user can also immediately respond to the model output by choosing alternative parameter values as the model runs. The model output can be generated in tabular or graphical form.
High Layer Map Containing Dialog Boxes, Graphs, Tables and Input-Output Devices
Model Construction Layer Containing Icons for Stocks, Flows, and Information Arrows
Model Equations Including Algebraic, Graphical and Logical Functions
Figure 1. STELLA Modeling Environment (Source: Costanza & Ruth, 1998)
Middle Layer: The four model icons The first task after defining the questions the model should answer will be to choose some model sectors. These are broad conceptual areas or domains that should be included in the model, guided by the questions the model should answer.
In principle, there are four model building blocks or icons: stock, flow, auxiliary variable, information connector (See Figure 2). A stock is a variable important enough to be explained within the model. A stock represents a state variable that embodies an aspect of the system under study. A flow represents the rate of change of a stock. The auxiliary variable defines the rate of change. Information connectors link the auxiliary variables, flows and stocks. Figure 2 Model icons
A preliminary model was presented in Workshop 1 to: (1) establish a point of reference; (2) demonstrate the type of output that can be obtained from a model structure; (3) function as a starting point from which the discussion takes off during the first meeting, if a group wishes to do so. Groups usually reject the preliminary model as a starting point (as did our group), as there is value in starting from scratch and focusing a group on the bigger picture, rather than zooming in on specifics of a preliminary model structure too early. Figure 3
Preliminary model outline: water quality component
Donâ€™t take this preliminary model outline too seriously, as it changed during the first workshop, based on the surveys and the participant group deciding not to lock itself into this preliminary model, and instead starting with a blank page. This picture, and the following ones, are included to give you an example of how these models may look.
Example of a graph and slide bars on the â€œuser-interfaceâ€? part of a model: 1: Total market value of Longfin eel harvest in NZ$
2: Market value of Longfin eel harvest in NZ$ per year
2 2 2 2
2030 4:40 p.m. Fri, 8 Oct 2010
Years Economic Value Improved Farm Management ie percent reduction in N loading rate
Annual Ag area growth rate in percent 0 U
Ri ver Sector N load in kg per year
N loading rate kg per hectare per year
Annual longfin eel harvest rate in percent
Total N load in kg
River flow in million cubicmeters per year
Annual growth of troutin numbers in millions Trout
Nitrogen N Soluti ons Periphyton Bi omass in N concentration ~ grams per square meter gram per cubicmeters Effect of Trout on Di ssolved Oxygen in ~ grams per cubicmeters Longfin eel
Sediment Annual growth of Ag area i n hectares per year
Annual longfin eel rate in percent Ag production rategrowth kg Ag area in hectares per hectare per year
Annual growth of longfin eel in millions
Improved Farm Management ie Improved Domestic percent reduction in N l oadingWastewater rate Discharges
Longfin eel populati on in millions Annual longfin eel harvest rate in percent moratorium on harvest of eel
Annual Ag area growth rate in percent Ag production kg per year
Annual harvest of longfin eel in millions Ecosystem Services
Values Total market val ue of Total market value of Ag production in Longfin NZ$ eel harvest in NZ$ Market value of Ag production in NZ$ per year
Price of Ag production NZ$ per kg
Food Longfin eel in kg per year Market value of Longfin eel harvest i n NZ$ per year Price of longfin eel NZ$ per kg
Cultural value index of eel population
Example of a Model structure, using the four icons in STELLA.
Because mediated modelling is a process-oriented tool, its success or failure has to be measured in process terms as well. In the short term, a mediated modelling project is successful when a group of about 20+ diverse stakeholders has a better shared understanding of a complex problem and reaches consensus on a course of action to address this problem. Individuals in the group become aware of the interdependency among group members and between economical, ecological, social and/or other relevant aspects. They develop a greater appreciation for the roles played by the other members of the group, and an improved sense of how common goals can be achieved. Sustained interaction among the group members to collaboratively implement, monitor and assess the recommendations is a medium term sign of success. The group has formed new working relationships among its members that are likely to change the way decisions are made in the future. Decisions are evaluated on their cohesiveness concerning short, medium and long-term impacts from a dynamic perspective and are forged from a shared vision. At this level, a result is evaluated in gradations according to its position on the spectrum between “where we are” and “where we want to be”. If the gap is closed, the problem is “solved”, if the gap became smaller, the situation has “improved”, and if the gap is unaffected or widens “insights are gained”. A long term success is characterised by the ability of the model building process to incorporate newly gained information as it becomes available over time and to function as the backbone of the stakeholder-involved, collaborative learning process. However, a model should be abandoned when the reason for building it is accomplished. Hanging on too long inhibits innovative thinking and learning.
Characteristics of an Action Plan Each action point should include Who, What, Where, When and Why. The “why” is specified with a link to the model, where possible. This may involve extrapolation from a specific local solution and envision what is possible at a wider scale. Actions will be recorded and may be brought into the process at any point; beginning and end.
Appendix 7: Adaptive Capacity We hope the model developed by workshop participants will be a tool that can be used by tangata whenua and other decision makers and users of Tauranga harbour to better understand the interactions within the harbour, and how different activities in Tauranga and its catchment impact on the health of the harbour. We hope that the model, and the process of discussion and collective decision making during the workshops themselves, will be a powerful force for improved understanding about Tauranga harbour, its ecosystems, and the services that the community gains from the harbour. Armed with a better understanding of how those ecosystems and services of Tauranga harbour are being enhanced or degraded over time, people in Tauranga can be better equipped and informed to make decisions about the use of the harbour, the types of activities that the Tauranga community want to encourage and support to ensure the health of the harbour is improved to the level that the community desires. Throughout the course of the workshops (November 2010 to April 2011) there will be opportunity for people to learn new skills in both ‘systems thinking’, ‘systems dynamics modelling’, ecosystems and ecosystem services of coastal and marine areas and associated catchment, as well as the important relationships between various aspects of the harbour and how changing one thing will likely impact on other aspects of the harbour, for better or for worse. Participants, or other interested people, are encourage to download the trial version of “stella” on our website – this is the systems dynamics software that we will use in this research, and to make changes to the various iterations of the model that are developed and updated at each workshop.