Using Fuzzy Cognitive Maps to Engage Stakeholders in the Search for Solutions to Emerging Environmental Challenges

Abstract

Inland freshwater salinization is an emerging environmental problem that threatens freshwater ecosystem health, agricultural productivity, and the sustainability, reliability, and safety of drinking water supplies around the world. This body of work focuses on the Occoquan Reservoir in Northern Virginia and uses fuzzy cognitive maps (FCMs) to characterize stakeholder understanding of the social-ecological system (SES) for freshwater salinization in the Occoquan. A total of thirty-five stakeholders from in and around the Occoquan system participated in this work by constructing personalized mental models for salinization in the region. These FCMs were evaluated with respect to (1) their ability to reveal barriers and opportunities for stakeholder-driven, bottom-up management of the FSS, (2) their ability to capture physically meaningful information about the SES when aggregated to represent collective perspectives, and (3) their dynamics - e.g., are feedback loops between environmental outcomes and management actions present in stakeholder FCMs and, if so, what might they imply for managing the SES? My results suggest that FCMs can reveal barriers and opportunities for collective salt management, with preferred modes of governance and uncertainty about the sewershed constituting barriers, and the watershed (specifically winter anti-icing and deicing activities) representing an area of agreement where collective action is more likely. FCMs were found to capture meaningful biophysical information about the system (e.g., reproducing the balance between watershed and sewershed salt sources). Because different FCMs captured different, but equally valid, salt source ratios, their comparison provides an opportunity for perspective taking and learning. My results indicate that dynamic motifs like feedback loops are relatively rare in individual FCMs, but emerge when multiple perspectives are aggregated. This suggests that stakeholders collectively (if not individually) perceive a dynamic and interconnected SES for freshwater salinization in the Occoquan. However, gaps in systems understanding remain that could constitute barriers to effective management - for instance, the absence of decision rules for managing salts in stormwater, one of the three principal water streams that contribute salts to the Occoquan system. In addition, not all stakeholders felt that the feedback loops articulated by others (or collectively) made sense for the system. While this may in part be due to the introduction of noise during aggregation, cognitive bias and the presence of relationships that were simply viewed differently also played a role. Recognizing these diverse "truths" is likely to be important for collective decision making and collaborative governance of freshwater salinization in the region. In performing the analyses required for this work, it became obvious to me that the accessibility of advanced simulation tools for working with FCMs is somewhat limited. To address this, I developed the R package fcmconfr, which provides tools for FCM aggregation and dynamic simulation, along with features for incorporating uncertainty in individual FCMs and propagating that uncertainty through to the aggregate. It is my hope that this contribution will make FCM analysis more straightforward and enable researchers and practitioners from a variety of backgrounds to use FCMs in support of stakeholder and community-engaged research.