Testing stakeholder engagement in ecological risk analysis: a case study of genetically modified maize and South African biodiversity.

University of Minnesota Ph.D. dissertation. May 2010. Major: Conservation Biology. Advisor: Professor Anne R. Kapuscinski. 1 computer file (PDF); xi, 160 pages, appendices I-VII.Ecological risk analyses (ERA) are traditionally conducted by a narrow range of biological scientists with limited stakeholder involvement. Different knowledge types and epistemologies necessary for understanding how stressors move through complex socio-ecological systems are generally excluded, as well as broader societal concerns of interested or affected parties. Calls for stakeholder engagement in ERA aim to address such deficiencies, but little real-world evidence exists on: 1) how to design participatory ERAs, 2) how ERA results change with inclusion of diverse participants, and 3)how to facilitate social learning fundamental to such collaborative endeavors. Social learning occurs when people engage with each other and share diverse perspectives and experiences to develop a common framework of understanding and foundation for collective action. Such learning is critical for efficacious participatory ERAs because participants must engage with different disciplines, epistemologies and worldviews to understand socio-ecological systems in which risks manifest, how the risk situation occurred, and then develop joint support for specific risk governance actions. I tested a participatory ERA process specifically designed to engender social learning through open communication, constructive conflict and extended engagement, in two workshops analyzing potential impacts of genetically modified (GM) maize on South African biodiversity. Workshop 1 involved four biologists, who were then joined by 18 diverse participants in Workshop 2, and I compared the ERA process and results between the two. Workshop 2 participants generated a larger and more comprehensive set of hazards and a more in-depth understanding of the agro-ecological system, creating a robust information base for the final risk assessment. Social learning occurred, as participants engaged with new information and diverse perspectives, began thinking systemically and modified their risk perceptions of GM maize. Participants did not, however, develop a shared understanding of the ERA process or highest priority risks to biodiversity. These results suggest that it is possible to implement participatory ERAs that generate useful risk-relevant information, and carefully designed participatory processes can produce social learning about other stakeholders, complex socio-ecological systems, and the topic of risk in short time periods. However, longer engagement is needed to build a shared understanding of the risk situation and possible solutions. Such learning-focused participatory ERAs should help transform risk governance from an unreflective approach (e.g., using risk assessments conducted by a small set of experts focusing on the stressor's most obvious attributes) to a reflexive approach, which not only aims to widen the scope of impacts evaluated but also considers how societal values and norms influence the conception and handling of risk