The Open Ocean assessment aimed to clarify the relationship between human and natural systems, to help identify why particular indicators are proposed and their relevance, where assumptions have been made, and where gaps exist in knowledge and data. These interactions between humans and natural ecosystems were captured in a "conceptual framework", based on the idea of "causal chains" (Figure on top). The framework was centred on the vulnerability of both natural systems to external pressures and consequences for the sustainable production of ecosystem services, and of humans to ecological changes. In brief, human activities have associated stressors that in turn impact natural systems and this in turn affects the delivery (and value) of services to people (starting in Box 1 and going clockwise). Ultimately, the assessment wanted to know how people are affected (Box 5: Human Wellbeing). However, indicators for all elements of the human and natural systems could not be developed - as the systems and their interrelationships on different time and spatial scales are complex and require time to develop. So the Open Ocean Assessment developed rapid "early indicator" metrics that are earlier in the causal chain.

Understanding and modelling this causal chain allows an assessment of the relationship between indicators earlier in the causal chain while keeping in mind the ultimate goal. This allowed an identification of data sources and gaps, of assumptions made, of some factors peripheral to the central framework that came into play, and of natural points of intervention for management. In many cases for the open ocean, data on the state of the natural ecosystem was localized or non-existent and we know more about the stressor (for example fishing) than the state itself. So, the various indicators and parts of the assessment address specific elements, and the overall risk analysis attempts to tie these together focused on the issues that require global-scale governance frameworks to address.

For the human system, all interactions between boxes are strongly mediated by socio-economic factors. Governance is defined broadly as including government, markets, and civil society, operating at global, regional, national, and local scales. It has strong connections to human-wellbeing. As well, effective governance is fundamental to achieving healthy ecosystems (inclusive of people), and in this context, links to sustaining ecosystem services (Box 4) in addition to other politically-negotiated goals.

For the natural system, concentration is on stresses associated with human activities (Box 2, for ocean side may be from ocean-based activities like fishing and land-based activities like carbon emissions), how they affect the state of the ecosystem under consideration (Box 3, modulated by the ecosystem vulnerability), which may lead to changes in the ecosystem services (Box 4, eg. fish catch). Finally, crossing the natural-human system boundary, the changes can lead to consequences for people, buffered or exacerbated by their vulnerability (surrounding Box 5). Natural variability, whether a regular seasonal change or more complex interaction within the natural system, will need to be evaluated separately from the interaction with the human system, so that the impact of a change in the human system - through a change in governance or a particular GEF intervention, can be separately identified. It is also important to characterize natural variability in order to understand which ecosystem state changes require or can be subjected to management.

The conceptual framework identified the protection of ecosystem services as the main pathway to mitigate consequences for people, under some other internationally-recognized value systems for management (protection of biodiversity, endangered species, natural heritage sites). However, the goal of management is not focused on sustaining ecosystem services – instead on directly conserving ecosystem state. In systems where thresholds might exist but uncertainty is high, and where future benefits are unknown, such a conservative approach has been politically negotiated.

In the context of a future GEF intervention, the full framework could be useful in determining the main points of intervention in the human system to help manage a positive outcome via the environment (the natural system). These assumptions and scenarios will have to be scientifically tested and validated.