Abstract
Mail surveys have been used over the last 17 years to estimate the recreational catch of western rock lobster Panulirus cygnus in an Australian fishery. Because nonresponse bias has been recognized as having a likely effect on these survey results, different inducements have been used over the years to increase returns. If these inducements have succeeded in reducing the effects of nonresponse bias, a catch model built from the resulting catch estimates would need to include inducement as a factor. At the end of the 2002–2003 season, 5,000 recreational license holders were sent identical surveys, along with one of the six different inducements that had been tried in previous seasons. The results showed that none of the inducements was likely to have affected total catch estimates in previous seasons. The age of the licensee and the license type held were identified as possible strata for future mail surveys to help reduce nonresponse bias.