Mental models have played important roles in the history of Chilika: a 1000 km2 coastal lagoon in the Mahanadi delta, India. For instance, it was hoped that the legalisation of shrimp aquaculture in 1991 would bring economic prosperity by diversifying local livelihoods and boosting annual fishery production. Instead, benefits were reaped by non-native aquaculture entrepreneurs, triggering cultural and socio-economic instability. The institutional settings soon adjusted, largely due to local pressures and scientific contributions of the newly formed Chilika Development Authority (CDA), leading to the banning of shrimp aquaculture in 2001.
Mental models also prompted studies of Chilika’s sediment dynamics in the 1990s, resulting in the new tidal outlet which has since increased fishery productivity 10-fold. Going forward, a balance exists between the institutional-led discouragement of juvenile catch and the desires of some fishers to maximise hauls.
I (very excitedly) travelled to Chilika in early 2016, hoping that both my mental and system dynamics model (SDM) would benefit from exposure to the system and its people. Until February 2016, my SDM was projecting future fishery production from empirical data and published work only. Interviews could tap into decades of experience working, living and ‘dancing’ with the system, as the pioneering system dynamicists Donella Meadows would say. I concentrated on how Chilika’s fishers, scientists and governors perceive the causes of the 1990s collapse, the subsequent recovery and the lagoon’s future. I also hoped the insights would help model evaluation and provide governance scenarios for simulation.
In practice, various barriers exist to integrating qualitative data into SDMs. For example, SDMs assume lumped populations making the same decisions, different to agent-based modelling which can simulate individual decisions. Yet workarounds exist, like disaggregating populations and/or estimating proportions making a decision for a given condition. For example, the former principle splits Chilika’s fishers into traditional and motorised fleets, associated with different fishing schedules and catch capacities; the latter workaround estimates the proportion of traditional fishers purchasing motorised boats for a given average income.
Furthermore, interviews may provide a quantity of opinions which cannot all be incorporated into the model’s finite structure. Therefore, it is useful to consider the rationale bounds of each stakeholder to understand how each mental model is shaped. Regional scientific experts may possess holistic system understanding, whilst fishers live and breathe the conditions important to their activities. Prior to the interviews, I was debating spatially disaggregating the fisher population into northern, central, southern and outer channel fleets. But from the fisher interviews I learnt northern sector fishers commute south to exploit the relatively abundant fish stock, dispelling my preconceived idea that fishers rigidly stick to their locality.
Overall, the field visit exposed me to different qualitative insights not acquirable from my desk. Understanding that traditional fishing communities may collectively begin using motorised boats when socio-economically favourable has highlighted how fishers adapt to intensify practices. Paradoxically, fishers exhibited environmental stewardship during the 1990s collapse by limiting their days fished, doing their bit to calm extraction stresses.
And finally, discussions with state and district level policymakers helped design feasible management approaches to test within the model (e.g. continued ecological restoration, bans, alternative livelihoods). The issue of policy implementation and adherence was continuously stressed, meaning any policies simulated in the SDM must be framed as ‘if all fishers complied with regulations, the resulting dynamics may be as follows…’, which is important for model design and scope. A big thank you to all who shared their mental models with me!