Analysis of logbook accuracy for blue marlin (Makaira nigricans) in the Hawaii-based longline fishery with a generalized additive model and commercial sales data

  • Walsh W
  • Ito R
  • Kawamoto K
 et al. 
  • 47

    Readers

    Mendeley users who have this article in their library.
  • 20

    Citations

    Citations of this article.

Abstract

Blue marlin, Makaira nigricans, catch (number per set) in the Hawaii-based longline fishery from March 1994 through June 2002 was analyzed by integrated use of observer reports, commercial logbooks, and auction sales records. The objective was to provide a corrected catch history for this species in this fishery during this period. The study was conducted because official statistics compiled from the logbooks are known to be biased by billfish (Istiophoridae) misidentifications. The initial step entailed fitting a generalized additive model (GAM) of blue marlin catch to environmental and operational data gathered by fishery observers during 8397 longline sets deployed by commercial vessels. The GAM included nine significant predictors and explained 41.1% of the deviance of observed blue marlin catches. The GAM coefficients were then applied to the corresponding predictors in the logbook reports from unobserved sets to estimate catches in evaluations of the accuracy of data from unobserved sets (N = 87 277 longline sets on 8437 trips; 95.4% of unobserved effort). This was done by regressing the logbook catch data on the predictions, using the residuals to identify trips with systematic misidentifications, and then checking their logbooks against sales records from the public fish auction in Honolulu. The large majority of the misidentifications consisted of striped marlin, Tetrapturus audax, reported as blue marlin, with lesser numbers of shortbill spearfish, T. angustirostris, logged as blue marlin, and blue marlin logged as either striped marlin or black marlin, M. indica. An estimate obtained by use of the GAM and observer data indicated that the nominal catch of blue marlin was inflated by 29.4%. The 95% prediction limits about the GAM-generated estimate (34 201-41 507 blue marlin) did not include the catch total from the logbooks (48 911 blue marlin). The corrections also refined understanding of the distribution of blue marlin by reducing the impression that large numbers of blue marlin are sometimes caught north of Hawaii in the autumn and early winter months. There was no evidence of widespread underreporting of marlins. We conclude that this study significantly improved the accuracy of logbook data for blue marlin and should also contribute to improved understanding of the ecology and distribution of blue marlin. We infer that self-reporting could yield accurate marlins catch data if species identifications were improved because there was no apparent underreporting problem. Finally, we recommend that logbook data accuracy receive serious attention in the context of stock assessments. © 2005 Elsevier B.V. All rights reserved.

Author-supplied keywords

  • Blue marlin
  • Corrected catch history
  • Fish auction data
  • Fishery observer data
  • Generalized additive model
  • Logbook accuracy

Get free article suggestions today

Mendeley saves you time finding and organizing research

Sign up here
Already have an account ?Sign in

Find this document

Authors

  • William A. Walsh

  • Russell Y. Ito

  • Kurt E. Kawamoto

  • Marti McCracken

Cite this document

Choose a citation style from the tabs below

Save time finding and organizing research with Mendeley

Sign up for free