Estimating catches of marine and freshwater recreational fisheries in the Netherlands using an online panel survey

In this study we describe a two-phase survey design and implications of approaches to non-response adjustments on estimates of the total catch taken by Dutch recreational fishers, including marine catches for Atlantic cod and European seabass and European eel in freshwater. The survey comprised three main elements which were executed online: a screening survey to estimate the characteristics of the population of recreational fishers (number of fishers, their demographic profile and stated fishing avidity); a 12 month logbook survey to estimate effort and catch rates; and non-response follow up surveys to adjust for non-response. A response rate of 80% was achieved for the screening survey and, following non-response adjustment and limited data imputation, 89% for the logbook survey. Some logbook participants reported no fishing activity (drop-outs) and were removed from the analysis. In addition, logbook data were weighted in accordance with the stated avidity distribution in the population to address potential response bias based on avidity. Imputation and weighting for avidity influenced the catch estimates a little, whereas the removal of the fisher drop-outs was influential, linked to the rates of fisher drop-outs (18% for freshwater and 55% for marine fishers). Freshwater recreational fishing was more popular than marine fishing; 9.7% of the Dutch population participating in the former and 4.1% fishing in marine waters. In total an estimated 53.6 million freshwater fish were caught (2.6 million retained) and 13.6 million marine fish were caught (9.6 million retained). Respective catch estimates for Atlantic cod, European seabass and European eel were 0.70, 0.35 and 1.23 million fish (0.53, 0.23 and 0.34 million retained). We conclude that the survey design using an online panel may serve as an example for future surveys because of its efficacy to collect a rich set of data at relatively low cost compared to traditional survey methods.