Adaptive management of large aquatic ecosystem recovery programs in the United States.

Adaptive management (AM) is being employed in a number of programs in the United States to guide actions to restore aquatic ecosystems because these programs are both expensive and are faced with significant uncertainties. Many of these uncertainties are associated with prioritizing when, where, and what kind of actions are needed to meet the objectives of enhancing ecosystem services and recovering threatened and endangered species. We interviewed nine large-scale aquatic ecosystem restoration programs across the United States to document the lessons learned from implementing AM. In addition, we recorded information on ecological drivers (e.g., endangered fish species) for the program, and inferred how these drivers reflected more generic ecosystem services. Ecosystem services (e.g., genetic diversity, cultural heritage), albeit not explicit drivers, were either important to the recovery or enhancement of the drivers, or were additional benefits associated with actions to recover or enhance the program drivers. Implementing programs using AM lessons learned has apparently helped achieve better results regarding enhancing ecosystem services and restoring target species populations. The interviews yielded several recommendations. The science and AM program must be integrated into how the overall restoration program operates in order to gain understanding and support, and effectively inform management decision-making. Governance and decision-making varied based on its particular circumstances. Open communication within and among agency and stakeholder groups and extensive vetting lead up to decisions. It was important to have an internal agency staff member to implement the AM plan, and a clear designation of roles and responsibilities, and long-term commitment of other involved parties. The most important management questions and information needs must be identified up front. It was imperative to clearly identify, link and continually reinforce the essential components of an AM plan, including objectives, constraints, uncertainties, hypotheses, management actions, decision criteria and triggers, monitoring, and research. Some employed predictive models and the results of research on uncertainties to vet options for actions. Many relied on best available science and professional judgment to decide if adjustments to actions were needed. All programs emphasized the need to be nimble enough to be responsive to new information and make necessary adjustments to management action implementation. We recommend that ecosystem services be explicit drivers of restoration programs to facilitate needed funding and communicate to the general public and with the global efforts on restoring and conserving ecosystems.