SYMP 8-5
What's more important for managing ecosystems, knowing where you are or where you are going?

Tuesday, August 12, 2014: 3:40 PM
Gardenia, Sheraton Hotel
Iadine Chades, EcoSciences Precinct - Dutton Park, CSIRO, Dutton Park, Australia
Jacob LaRiviere, Ecology and Evolutionary Biology, University of Tennessee, Knoxville, TN
Alan Hastings, Department of Environmental Science and Policy, University of California, Davis, Davis, CA
Carl Boettiger, Environmental Science, Policy, & Management, UC Berkeley, Berkeley, CA
Background/Question/Methods

Natural resource managers need guidance on how to best invest their scarce resources to maximise the chance of saving species. Decision theory is helping decision-makers prioritise biodiversity threat management across time and space but a major drawback of current decision approaches is their need for “data-hungry” models that simulate how a system will behave under different management decisions. This currently prevents managers from using decision tools when data are scarce or resource dynamics are uncertain, such as under global environmental change. However, managers must still make conservation decisions with imperfect data and information or else risk losing species to extinction. They must adapt their conservation decisions as they simultaneously manage and reduce uncertainty over time. We first provide a methodological background on the current methods available to managers to deal with different types of uncertainty and highlight their strengths and weaknesses. We then study whether it is more important to reduce the uncertainty about the current state of an ecological system or to reduce the uncertainty about the future states of the systems (parametric or model uncertainty).  We use a Kulback-Liebler divergence metric to compare these otherwise incommensurate manifestations of imperfect information. We use partially observable Markov decision processes and adaptive management methods to compare the performance of management strategies under varying magnitude of uncertainty.

 Results/Conclusions

Our results suggest that both kind of uncertainty play an important role in the management of ecosystems but the magnitude of the uncertainty was the key driver of performance that managers should consider. While state uncertainty (not knowing where we are) can be reduced by investing in improved surveillance in critical states, parametric and model uncertainty (not knowing where we are going) require accurate monitoring of the system which might not be an available option. Our approach is general and can be applied to a range of management problems where managers are faced with the dilemma of choosing which kind of uncertainty they should account for to best conserve species.