In December 2014, the Puget Sound Leadership Council adopted the 2014-2016 Biennial Science Work Plan, a document identifying decision-critical science for Puget Sound recovery. PSI Research Scientist Nick Georgiadis was lead author on the report in collaboration with the Puget Sound Partnership and its Science Panel. In the report, Georgiadis addresses the challenge of managing large scale ecosystems in the face of scientific uncertainty. Read an excerpt from a summary of the Biennial Science Work Plan below.
By Nick Georgiadis
According to the latest State of the Sound report (2013) from the Puget Sound Partnership, most indicators of Puget Sound’s vitality (so-called ‘Vital Signs’) have not advanced much, if at all, towards their recovery targets. Some, such as orcas, have lost ground. Progress may be slow because capacity and resources are spread too thinly over too many targets and across too large an area for recovery gains to be detectable. Even if this is true, in complex systems like Puget Sound, narrowing the scope of the recovery strategy to a feasible yet functionally effective number of targets turns out to be one of the greatest challenges of all. This is because the number of potential targets is large, many of them interact with each other, and much of the information is lacking by which targets could be ranked by potential to yield greatest recovery gains.
This assessment cited several additional checks on recovery progress in Puget Sound: research goals are insufficiently specified, the effectiveness of actions inadequately measured, and too often it is left implicit how results from research should influence or be applied to recovery. These are all symptoms of a more fundamental reason for slow progress towards recovery: that adaptive management (AM) is difficult to apply at the ecosystem level. No approach to recovery guarantees success, but, more than any other, AM increases the likelihood that progress can be made under uncertainty. If everything was known about how to achieve a given recovery target, the process would resemble, say, construction of a bridge. For ecosystem targets, however, it is rarely the case that enough is known, and no new knowledge is needed, to attain a recovery goal. For most, there is uncertainty about how outcomes are affected by natural processes, how these are impacted by humans, or how restoration and protection actions bring about recovery. AM tackles uncertainty head-on by ensuring that we learn by experience. If progress is elusive, AM helps us understand why, and advances in understanding help to justify continued funding.
In Puget Sound, AM was nominally adopted as the default approach, but has been applied in a patchy and incomplete manner, partly because of insufficient resources. The Science Panel endorses application of AM, and specified how it should be supported in two ways.
Initial emphasis should be on completion of ‘implementation strategies’ for each recovery target. These would distill existing concepts about how recovery should be achieved: the mechanistic theories, causal pathways, and actions by which recovery targets are expected to be met. In this context, therefore, the term ‘implementation strategy’ includes and entails the integration of science to address critical uncertainties.
It was further recommended that implementation strategies for each recovery target be designed by a separate ‘recovery group’. The Puget Sound recovery community of scientists, managers, policy makers and practitioners is large and well versed in contributing to advances by consensus of expert opinion. Most targets are already attended to by groups operating in varying degrees of formality. The emphasis on creating implementation strategies by separate recovery groups is intended to meet several needs. One is to integrate science and policy more directly into the practice of recovery. A second is to specify strategies (that can be supported by policy) for attaining each recovery target, while implementing a style of adaptive management that is more informative about effectiveness and progress. The third need is to find additional ways to advance the research agenda to define decision critical science for recovery, given the wide array of what are essentially equal priorities, and in the face of shrinking budgets. The only way to achieve this exacting agenda is to spread the task of planning recovery strategies and guiding their implementation among groups of practitioners with the requisite experience and resources.
Used in this way, implementation strategies should address and resolve many of the difficulties relating to identification and selection of decision-critical research that were highlighted in Parts I-III of this report. Once an implementation strategy is designed and documented it becomes possible to separate layers showing the junctures at which (1) research is needed to resolve a critical uncertainty (including models and social science), (2) monitoring is needed to assess effectiveness of actions, (3) policy changes are required; (4) costs can be estimated to assess cost-effectiveness, and (5) time will be needed for social and ecological processes to deliver expected outcomes. In this way, not only are the essential elements of recovery (science, policy, monitoring, etc.) featured, but their integration and interactions can also be represented. For any given time and purpose (such as the preparation of a BSWP), the list of research priorities can be drawn up simply by combining the ‘research layers’ from each implementation strategy. The list of monitoring and policy priorities would be similarly derived. It becomes at least conceivable to estimate and compare the cost and cost-effectiveness of alternate paths and actions, order actions into a logical time frame, and expose common and conflicting goals within and among strategies. To date, this level of integration among recovery strategies has yet to be achieved.