Aquaculture and coastal fisheries are increasingly expected to support food production, livelihoods, and ecosystem recovery while responding to climate change and other interacting pressures. Yet management strategies often rely on implicit assumptions about how these systems change, recover, or resist intervention. For example, persistent low production may be interpreted as a poverty trap, declining fish stocks as an alternative stable state, and slow recovery as management failure. Each interpretation implies a different intervention strategy, but the underlying dynamic assumption is rarely made explicit or tested.

This presentation introduces a regime-strategy-action cycle that uses dynamical systems modelling to connect resilience narratives with intervention choices and adaptive management. The approach distinguishes among qualitatively different dynamics, including single and alternative stable states, endogenous cycles, bounded irregular variability, long transients, and moving targets created by changing external drivers. These regimes imply different management priorities: maintaining or shifting an attractor, crossing a threshold, dampening harmful oscillations, buffering unpredictable variability, managing a slow trajectory, or tracking changing environmental conditions.

Using examples from aquaculture and coastal fisheries, we show how persistent low performance, variability, or slow recovery can arise from different underlying dynamics. Each dynamic regime implies a different intervention logic, making diagnosis central to strategy selection. The approach links regime assumptions to management strategies and actions, as well as to minimal monitoring cues and exit ramps that indicate when the assumed regime and corresponding strategy should be revised. It also links each regime assumption to minimal monitoring cues and exit ramps, i.e., observations that indicate when the current explanation or strategy should be revised. Rather than prescribing a single pathway to resilience, it supports adaptive management by making assumptions explicit, matching actions to system behaviour, and enabling strategies to change as evidence accumulates.