Addressing human-induced uncertainty in fisheries management : social scientific and interdisciplinary solutions using Bayesian belief networks
Päivi Haapasaari · Työväentutkimus Vuosikirja · 2012
The complexity, ambiguity and various sources of uncertainty related to fisheries systems are increasingly acknowledged. This has led to questioning the conventional practices of producing the knowledge base for fisheries policy. The prevailing practice relies on biological stock assessments, whereas uncertainties stemming from the behavior of humans are usually ignored. Focusing on biological management advice has further led to defining management objectives and related reference points in biological terms only. Currently, both scientists and managers call for expanding the practical science-policy cycles to incorporate social sciences and economics and to analyze the different types of knowledge in integrated frameworks. In this thesis, the potential of Bayesian belief networks (BBNs) to broaden the knowledge base of fisheries management is discussed. Applications to social sciences and to interdisciplinary settings are demonstrated in relation to Baltic salmon and Central Baltic herring fisheries. BBNs are based on the idea of structuring problems into acyclic cause-effect relationships and quantifying the relationships with values expressing subjective degree of belief. With their subjective perspective to knowledge, BBNs have features in common with the constructivist and hermeneutic theories of social sciences. This facilitates applications of BBNs to social sciences, and further enables combining social knowledge with biological and economic knowledge. An interdisciplinary model provides a framework to examine interactions between various uncertainties, objectives, and stakeholder interests, and thereby to anticipate consequences of decisions prior to their implementation. Addressing implementation uncertainty by quantifying fishers potential reactions to management measures can question decisions calculated optimal by biological or bio-economic models and turn attention to options that fishers support. BBNs provide a decision tool and a device for participatory problem framings, and an illustrative focus of discussion for adaptive co-management processes. The probabilistic basis of the approach implies that it does not involve a claim for truth but provides a framework to address variables and interrelationships that are considered relevant by scientists and other stakeholders, and further to update a model in order to learn about the system that it represents. The thesis acknowledges the difficulty related to interdisciplinary collaboration caused by the differences in disciplinary practices and paradigms and the scarcity of integrative tools. Through a focus on our research process related to Baltic salmon management, the thesis analyzes what kind of interdisciplinarity between natural scientists, environmental economists and social scientists grew from the need to better understand the complexity and uncertainty inherent to the Baltic salmon fisheries and how divergent knowledge was integrated to support science-based decision making. It is concluded, that interdisciplinarity is an extensive learning process that takes place on three levels: between individuals, between disciplines, and between types of knowledge. Such a learning process is facilitated by formulating a global question and by agreeing a common approach at the outset of a process.