Applying a Systems Approach to Marine Conservation
Evan O'Donnell

Evan O'Donnell Former Strategy Associate

Tags for this post
June 10, 2015

Applying a Systems Approach to Marine Conservation

School of fish

At The Rockefeller Foundation, we take a systems approach to our work.

This means we try to understand how different actors operate together and the different incentives that drive them, so that we can identify what types of interventions can most powerfully “re-wire” the system to operate differently and toward better outcomes. We believe this approach ultimately helps us achieve impact most effectively, but it comes with some challenges. Because systems are inherently complex, distilling the most critical cause-and-effect relationships can be a difficult and time-consuming endeavor.

Modeling & Development Strategy

We use modeling as a tool in our strategy development process to help overcome that complexity. The term “modeling” refers to the representation of something (in our case often a defined system) in order to understand its various characteristics. It helps us get a directionally accurate, if not perfect, assessment of how a system works. We can then play out “what if” scenarios to test the impact of various interventions. If done correctly, modeling can help us gain more confidence in our strategies and identify new opportunities and potential risks.

The Mindoro Fishery

For our Oceans & Fisheries initiative, we have used modeling to help us understanding the particularly complex linkages between marine ecosystems and human economic and social systems. We partnered with Icosystem, a company that designs software simulations that reflect complex network connections, to model the supply chain of an illustrative small-scale fishery. We wanted to understand how various interventions would impact the marine ecosystem and livelihoods of small-scale fishers. The model was based off the Mindoro yellow fin tuna fishery in the Philippines, and built iteratively over several months using a variety of quantitative and qualitative data sources.


Icosystem's software program models the economic environment of a fisheries supply chain. Image credit: Icosystems
Icosystem’s software program models the economic environment of a fisheries supply chain. Image credit: Icosystem


The model delivered some critical insights that relate to marine conservation more broadly, a couple of which are highlighted below:

Virtuous cycles are possible: With the right incentives and fishing controls in place, small-scale fisheries like Mindoro can naturally maintain a virtuous cycle where positive ecological, economic, and social outcomes reinforce each other. In other words, it just takes an initial intervention to get the fishery back on track, assuming fishing effort remains constant—long-term subsidy is not necessarily required to achieving sustained impact.

Input controls in isolation are insufficient to achieving impact: In many cases, effort controls (i.e., restrictions on number of boats and people who are able to fish) on their own will be insufficient. When the population depending on fishing for livelihoods or access to food is extreme, effort controls will need to be complemented with alternative livelihood options for fishers to restore a fishery. Unless alternatives are put in place, fishing effort controls, such as quotas that limit the number of fishers allowed to fish, will negatively impact incomes or incentivize fishers to circumvent these controls and restrictions.

Naturally, our model relied heavily on the data sources we had access to and assumptions made. Model assumptions and outputs should always be tested with experts and compared to existing knowledge. We view modeling as a complement, not a replacement, for more traditional forms of analysis and understanding of context-specific variables. Nevertheless, these insights contributed to our understanding of the underlying dynamics of this particular fishery, and the probable outcomes of various combinations of interventions. It also revealed potential negative unintended consequences that demand our attention.

All in all, modeling provided us a useful sandbox to test ideas and diagnose important sensitivities in the system, and proved to be a useful complement to existing analyses and learning activities we utilize in developing strategy. For more information about the model we developed, how it was developed, critical assumptions, and some of the key findings, please visit:


Share on Google+Share on FacebookTweet about this on TwitterShare on LinkedInShare on StumbleUpon
Tags for this post