To truly manage marine ecosystems and all the multiple ocean-uses associated with them, we need to adopt ecosystem-based management (EBM). We also need to ensure that we cover the full range of features in socio-ecological systems (SES) associated with the ocean. To do so, a key means of implementing EBM operationally is to use models, and these need to be coupled models across multiple disciplines. For it is only when we present economics and related human dimensions do people truly care about and pay attention to our work on these SES. 

My assumptions are:

• Everyone in the practice and discipline of managing marine ecosystems and their uses is past the need to justify, define, rationalize, and explain EBM, and generally agrees it is a good idea to do EBM;
• Everyone holds the same as true for the concept of SES;
• Everyone understands the value, rationale, benefits and reasoning of using models for EBM applications; and
• Everyone recognizes that there are many, good, extant modeling tools that are available to use to make EBM operational. 

Yet, despite the growing prevalence of my assumptions, the use of the appropriate models to manage marine ecosystems has not been fully widespread, and certainly are not commonly used when making operational decisions. Furthermore, the applications that simultaneously consider economics and human dimensions are also relatively limited. 

Reflecting on the literature, case studies, and personal experiences, there remain barriers to the operational use of ecosystem models for marine ecosystem-based management. Beyond the usual 5-10 key criteria to make these models more widely acceptable, I submit that that a clearer understanding of what operational use means, both from a scientific/technical perspective and from a regulatory/legal perspective in a decision-support setting would be beneficial. I note this as often the statistical determinants for model use can be too constraining for decision-support contexts. 

The other thing I submit is that an ongoing, dynamic model taxonomy would probably be helpful, namely where the classes of models need to be mapped to particular and salient applications given the many dimensions of modeling. The reason is that all too often the classes and families of model use are applied incorrectly to different problems, and from that we end up in the modeling black hole of “too little data, too little precision, and too much uncertainty” that stimies the use of models for SES when we likely don’t need to fall into such a black hole trap, especially when the regulatory expectations are not requiring the same level of constraint to support decision-making.

I consider a few select case studies that have overcome these impediments and that have also jointly considered economics, human dimensions and other features of SES beyond the more common biogeochemical modeling.