Abstract
Reports by commercial fishers in New Zealand's ling ( Genypterus blacodes ) bottom longline fishery were analysed to identify the influence of official fishery observers on reported weights of bycatch. Observer data show that on every day ling were taken by bottom longline, both individual transferable quota (ITQ) and nonquota (non‐ITQ) bycatch were reported. However, commercial fishers reported ITQ bycatch on 82% of days fished, and non‐ITQ bycatch on only 50% of days fished. Frequencies of reported bycatch were compared for fishers’ data with and without an observer aboard for the Chatham Rise, Campbell Plateau, and Bounty Plateau areas. Significant underreporting of bycatch was indicated for some bycatch fish species in several areas in the absence of an observer. A sample selection model of reported spiny dogfish ( Squalus acanthias ) bycatch was used to model the effects of observers on whether spiny dogfish bycatch was reported at all, and the effects of observer presence and other independent variables on the quantity of spiny dogfish bycatch reported. The evidence reported here illustrates that misreporting of bycatch is common and can introduce significant errors into total catch estimation that can have serious implications for fishery sustainability.