Aller au contenu principal

    Fishing for structure; can we describe normal patterns in toothfish fishing operations using catch and effort data?

    Demander un document de réunion
    Numéro du document:
    WG-FSA-16/36
    Auteur(s):
    J.M. Fenaughty and K. Large
    Soumis par:
    Alistair Dunn (Nouvelle-Zélande)
    Approuvé par:
    Alistair Dunn (Nouvelle-Zélande)
    Point(s) de l'ordre du jour
    Résumé

    We describe the typical steps involved in demersal longline fishing operations for toothfish in CCAMLR fisheries and link those steps to the variables recorded as part of the CCAMLR catch and effort data reporting system. We then describe statistical properties of the recorded variables and how they may vary among gear type and other factors that make them useful to understand fishing activities and for error trapping or data validation.

    During this process we identified some unusual values for some variables; some are explainable while others are obvious errors. We suggest that identifying values outside normal maximum ranges can be used to target additional error-checking or seek additional information using other associated vessel records. Other unusual values, although potentially in error, are more difficult to evaluate given the wide range of influences that may affect recorded values for a given parameter.

    There are strong functional relationships identified between some of the variables. For example, a catch with high numbers of large fish increases the time taken to haul a line as each fish is required to be gaffed aboard and removed from the line. This effect is compounded by fish size and the requirement to tag fish, which further slows the hauling process. We investigate some of these relationships and their use in identifying erroneous data. The analyses show that many of the useful metrics that could be derived, are likely to be influenced by multiple variables. Therefore, multivariate analyses are needed to control for the influence of multiple variables before we interpret patterns in the data. We also recommend additional information be included in the data recording forms to improve interpretation for some variables.