The IBTS database as maintained at ICES Headquarters has been submitted to a quality check with reference to reported size distributions and species identifications. The check is based largely on a comparison of reported size distributions with available information on maximum length in the literature and on consistency in species compositions reported by individual countries (and sometimes years and areas). The results are alarming. For a large number of species the database contains unrealistic information. In many cases errors may be tracked to simple input errors, but clearly there have also been, and still are, severe species identification problems among all countries participating in the survey. Misidentifications involve errors in abundance of at least two species and, based on the limited analysis presented, a conservative guess is that at least 30% of all species reported may be affected. This not only means that trends in abundance of individual species may be biased, but, more importantly, it reduces the value of the IBTS database for diversity studies of the North Sea fish community to virtually zero. I present a strong plea for a comprehensive check of all data in the existing database, with an appropriate correction of uncertain identifications to higher taxa to reflect inherent uncertainties, and to embark on a serious quality control programme to ensure that future surveys yield only reliable information at the species level. The scientific credibility of ICES is at stake, because it is a waste to maintain a flawed database at ICES Headquarters. |
