Managing ecosystems under a changing climate requires flexibility in order to facilitate resilient ecosystems that satisfy desired ecological and societal goals. For example, the combination of high fishing rates, poor climatic conditions and high predation pressures are likely to produce less favorable management outcomes than the same fishing rates and predation pressures under good climatic conditions. This type of observation motivates the need for an approach to management that includes dynamic reference points that reflect the variable marine environment and a coupled social-ecological system. Identifying such ecosystem reference points in relation to climatic variables or key ecological species is a primary goal, but a critical gap, at this time in many PICES member countries. To move forward on this front, we need 1) methodologies for determining how ecological (e.g. trophic) interactions can be directly included in establishing reference points, 2) an examination of how climate variability and change might (should?) be incorporated into the determination of biological reference points, and 3) a methodological framework for identifying non-linearities, that might lead to surprises, in common ecosystem indicators.