Optimisation of survey design and calculation procedure for the International Herring Larvae Survey in the North Sea

摘要

The ICES International Herring Larvae Survey Program (IHLS), regularly carried out since 1972 in the North Sea and adjacent areas, provides an abundance index to be utilised by the ICES Herring Assessment Working Group. This is one of three indices for assessing the state of the herring stock components in the North Sea by means of VPA (Virtual Population Analysis) related numerical procedures. Due to a substantial decline in ship time and sampling effort since the end of the 1980s, it was necessary to revise the IHLS survey design and the related procedure for calculating larval abundance indices, as major constraints of the formerly used indices (LAI, LPE) had been strongly violated. Using the historical herring larvae data base, which was transferred from Aberdeen (Scotland) to Kiel (Germany) in 1997, the current article analyses statistically the effects of this decline in effort, derives the optimum effort and presents its optimal allocation to specific sampling units defined by space and time. All investigations have been performed on the basis of an abundance index (MLAI(year)) introduced in 1994 (Patterson and Beveridge1994), which was derived from a linearized multiplicative model approach (log-transformed multi-factor ANOVA). This MLAI approach was modified by introducing a weighting factor per sampling unit, taking into account differences in the spatial coverage of sampling and in the variation among stations within these sampling units.