USING BAYESIAN DECISION NETWORKS TO GUIDE RESTORATION OF FRESHWATER MUSSELS IN ILLINOIS

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Tools aiding in decision-making regarding restoration of freshwater mussels are needed. Bayesian approaches use both empirical data and prior knowledge to estimate likelihood of management outcomes when structural uncertainty is high. We used Bayesian decision networks to determine optimal management for two Rivers in Illinois. Management options were no action, propagation of juveniles, relocation of adults, release of inoculated host fish, or dam removal. We considered target species Ellipse (Venustaconcha ellipsiformis) and Spike (Eurynia dilatata), and tested for sensitivity to 1) dataset (long term presence, current presence, and current abundance), 2) streams (two target streams, six non-target streams, or both), and 3) expert opinion (median, minimum, or maximum). Maximum models tended to choose No Action less often, and predicted higher likelihood of mussel establishment after restoration. Models were more sensitive when using only target streams. Propagation of juveniles was most often recommended for Ellipse. For Spike, propagation of juveniles or no action were chosen. Use of all stream data and median expert opinion values is recommended. Bayesian decision networks offer a useful tool for restoration of freshwater mussel species, but care should be taken when choosing data sources.
Original languageEnglish (US)
Title of host publicationSociety for Freshwater Science 2019 Annual Meeting, Salt Lake City, Utah
StatePublished - 2019

Keywords

  • INHS

Fingerprint

Dive into the research topics of 'USING BAYESIAN DECISION NETWORKS TO GUIDE RESTORATION OF FRESHWATER MUSSELS IN ILLINOIS'. Together they form a unique fingerprint.

Cite this