Fishery managers face two problems that are endemic to all renewable resource management: how much of the resource should be extracted, and how should resource users be managed to ensure efficiency and fairness. The predominant fishery management approach addresses these problems through fish stock assessment and resource economics. However, my review of the literature and analysis of the situation in the Danube Delta Biosphere Reserve show that both methodologies face serious difficulties: they deal inadequately with uncertainties about the causes of observed behaviour and the likely effects of different policies; they are too focused on readily measurable objectives; and they do not address the effects of the institutional context on management. In Chapter 3, I examine previous applications of Multiple Criteria Decision Analysis (MCDA) with a view to see if they can be applied to fishery management. My analysis shows that until now MCDA has been used to address only the first two sets of fishery management problems: systematically incorporating uncertainty and multiple objectives into policy development. I also argue that existing proposals for the use of Decision Analysis can be classified as variations of one version of MCDA, namely Multiple Stakeholder Decision Analysis (MSDA). The main problems that remain to be resolved relate to the interaction between experts, stakeholders, and managers when there are conflicting interpretations of evidence, and situations of high institutional inertia. In Chapter 4, I examine these problems within the context of ecological management experience and New Institutional Economics. I argue that for complex problems, such as those in the Danube Delta, management that aims to attain narrowly defined optimal fishing yields through command and control measures is unfeasable and undesirable. A more promising approach would seek to strengthen resilience, promote organisational variety, and increase the leverage of stakeholders over those who provide services for them. When one seeks to achieve such a transformation of management, I argue that the intervention needs to take into account the specific institutional circumstances of the client. In Chapter 5, I show how management procedures, problem perception, and strategy development are influenced by organisational structure and the hierarchical position of managers. That is why decision analysis interventions must address both technical as well as institutional needs of clients. In Chapter 6, I discuss Decision Conferencing, an alternative MCDA approach, and argue that it is more suitable for dealing with management problems such as those of the Danube Delta Biosphere Reserve. Decision Conferences can provide a structure for expert, manager, and stakeholder interaction and can lead to the transformation of social realities. In Chapters 7 and 8, I review the context and concrete environmental and institutional problems that led to the first Decision Conference on an environmental management problem. I report the processes of the Decision Conference, the agreements reached, and anlyse both the short and medium term effects of the intervention. On the basis of that evidence I make claims about the general utility of the approach. The thesis concludes with proposals to improve Decision Conferencing through a framework that provides guidance for context specific process management and helps to ensure that a requisite variety of viewpoints are incorporated into management strategy development.