Evaluating the Ability of Specialization Indicators to Explain Fishing Preferences

Abstract

Understanding the predictive ability of recreation specialization to explain behavior is important for wildlife and fisheries management given the widespread use of specialization to capture diversity among outdoor recreationists. Using allocation of days among fishing opportunities in a discrete choice experiment, we studied the extent that specialization predicted preferences for attributes describing the opportunities. Latent class modeling revealed that three groups of anglers optimally captured preference diversity in our sample. To this base model, we sequentially added 11 metrics of angler specialization and used information theory to select the metric that best predicted group membership, namely centrality to lifestyle. Weaker evidence existed for the specialization dimensions “importance of catch,” “specialized gear use,” and a multidimensional self-classification approach, whereas indices of skill, media use, trophy fish, and harvest orientation were not supported. General specialization constructs such as centrality to lifestyle, therefore, might be best suited for predicting general fishing preferences and subsequent behaviors of anglers.

Keywords:

• discrete choice
• information theory
• preference heterogeneity
• recreational fishing
• specialization

Acknowledgments

Ben Beardmore is now at the Center for Limnology, University of Wisconsin-Madison. Funding for the diary portion of this study was provided by the European Financial Instrument for Fisheries Guidance and the State of M-V, with additional funding provided to Robert Arlinghaus through a grant within the Pact for Innovation and Research by the Leibniz-Community (www.adaptfish.igb-berlin.de) and another by the Federal German Ministry of Education and Research (www.besatz-fisch.de, grant 01UU0907). The Social Sciences and Humanities Research Council of Canada funded Ben Beardmore. We thank Malte Dorow for his assistance with the diary data and in developing the survey, Don Anderson for the experimental design, and the reviewers and guest editors for their insightful comments to improve the manuscript. We also thank the survey company, USUMA GmbH, for executing the data collection.

Notes

Atlantic cod (Gadus morhua), Atlantic herring (Clupea harengus), the group of marine flatfish species (e.g., founder – Platichthys flesus), garfish (Belone belone), common carp (Cyprinus carpio), a group of coarse fish species (i.e., small bodied cyprinid species such as roach, Rutilus rutilus), European eel (Anguilla anguilla), Eurasian perch (Perca fluviatilis), northern pike (Esox lucius), and zander (Sander lucioperca).

TABLE 2 Standardized Attribute Levels Used in the Discrete Choice Experiment of Fishing Site Selection for German Anglers

A table presenting the descriptive statistics by which realistic attribute levels were defined for each species may be obtained from the corresponding author.

For brevity, Table 3 presents models specifying only one to five classes, as these were sufficient to establish the best-fit model; however, the authors assessed models including up to 20 classes, consuming all available degrees of freedom. These additional results may be obtained from the corresponding author.