There is growing recognition of the need to monitor recreational fishing activity as an essential component of fisheries management and stock assessment. However, in the absence of mandatory reporting systems it is necessary to apply survey approaches to gather statistics for the sector. In practice, all survey methods are subject to sampling and non-sampling error. Non-sampling error includes coverage and response errors which, if not controlled or minimised, can introduce significant biases. Furthermore, as people are becoming increasingly reluctant to participate in surveys, declining response rates represent a growing challenge to the on-going provision of reliable statistics.
Telephone surveys have been used widely in the past, often with limited regard to the impact of non-sampling errors. In this presentation we examine how telephone survey response profiles have changed over two decades and compare key performance metrics based on sampling from licence registries and general population sampling from phone directory lists. Our findings demonstrate that it has been possible to maintain high response rates over time, although coverage issues are an increasing problem with general population surveys. The value of a comprehensive registry of fishers that includes phone contact details cannot be underestimated as a sampling frame for data collection.