Abstract
<em>Abstract</em>.—Harvest of Muskellunge <em>Esox masquinongy </em>within the Ceded Territory of northern Wisconsin is managed using a quota system where safe harvest levels are established for individual populations based on estimates of adult abundance. When a reliable population estimate is not available for an individual lake, linear regression is used to predict adult abundance based on loge lake surface area (LRSA model). However, the amount of variation explained by the LRSA model is relatively low (<em>r</em><sup>2</sup> < 0.49). Our objective was to determine if an alternative random forest (RF) analysis incorporating 24 predictor variables related to lake characteristics increased the accuracy of predicted estimates of adult abundance compared to the LRSA model. Random forest analysis increased predictive accuracy (measured as mean absolute error) by 45% and selected lake surface area, percent sand, muck, and gravel substrates, year of estimate, stocking intensity, shoreline length, and percent of shrub and grassland habitat in the watershed as important predictor variables. On average, use of RF analysis resulted in an 18% increase (range = –60% to 200% change) or an additional one fish per lake-year (range = –17–47 fish) in safe harvest compared to the LRSA model.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
