Skip to main content

    Performance metrics to index the spatial coverage of mark-recapture data

    Request Meeting Document
    Document Number:
    WG-SAM-16/13
    Author(s):
    C. Marsh, A. Dunn and S. Mormede
    Submitted By:
    Dr Rohan Currey (New Zealand)
    Approved By:
    Dr Rohan Currey (New Zealand)
    Abstract

    Understanding the spatial distribution of the release of tagged fish and the subsequent spatial coverage of fishing effort to recapture tagged fish is an important consideration when interpreting the biomass estimated from mark recapture data. In 2015, the Scientific Committee identified that measures of spatial overlap and potential bias in the development of tag-based biomass estimates are an important focus topic for WG-SAM (SC-CAMLR-XXXIV, 2015, para 3.83). This paper outlines developments towards new biomass-weighted spatial overlap summary statistics for tagging data. The method calculates two statistics: (i) a single measure of the degree of spatial overlap between the release of tagged fish and subsequent fishing effort for tag recovery, labelled the tag spatial overlap (TSO) statistic and (ii) a measure of the potential bias in the biomass estimate calculated from non-homogenous spatial mark-recapture data, labelled the tag spatial bias (TSB) statistic.

    We apply the method to a single case study area, SSRUs 88.2H to illustrate its use. We found that the median tag spatial overlap statistic in 2012 was 70% (95% CIs 53–80%), and the median tag overlap bias was 88% (95% CIs 64–114%). In 2014 the median tag spatial overlap statistic 60% (95% CIs 50-69%), and the median tag overlap bias was 192% (95% CIs 141-268%).

    The TSO statistic and TSB statistic provide a simple; and intuitive approach to indexing the degree of spatial overlap and potential bias from mark recapture data, with few assumptions required on the distribution of fish, movement, or catch history. Potential uses of these statistics could include a measure of the total spatial distribution of effort through time in research or developing fisheries, and a measure of the change in bias of mark-recapture estimates through time.