Standardization of bigeye and yellowfin tuna CPUE by Japanese longline in the Indian Ocean which includes cluster analysis

Standardizations of Japanese longline CPUE for bigeye and yellowfin tuna in multiple Indian Ocean regions were conducted using generalized linear models (GLM) with log normal errors. The models incorporated fishing power based on vessel ID where available, and used cluster analysis to account for targeting. The variables year-quarter, vessel ID, latlong5 (five degree latitude-longitude block), cluster and number of hooks were used in the standardization. The numbers of clusters selected varied among regions and species, but in all cases were either 4 or 5. Dominant species differed depending on clusters. The effects of each covariate differed depending on species and region. The CPUE trends were similar to those estimated last year, though with some differences due to the inclusion of vessel effects and cluster variables.