38 



Fishery Bulletin 100(1) 



opinions about the reporting procedure. All participants 

 were asked the same questions from a standardized sur- 

 vey script. During the interview no information was pro- 

 vided to the anglers about the study design. 



For statistical analysis each release site was treated as 

 a replicate. By nesting site within estuary, within state, 

 differences associated with each site, estuary, and state 

 could be treated in the analysis to assess influence of the 

 reward messages. The study design was a 2x2 factorial de- 

 sign (state and reward) with three levels of nesting (state, 

 estuary, and site) (Table 1). Owing to differences in growth 

 rates, insufficient numbers of legal-size fish were available 

 to stock all estuaries during the same month. Thus one 

 estuary in each state was stocked in the fall of 1996 and 

 the remaining estuaries were stocked the following spring 

 and summer However, each stocking group was available 

 for capture during the fall season when fishing pressure 

 is heaviest (Wenner'). Percent return data were arcsine 

 square-root transformed prior to analysis. Return data 

 were analyzed by using a two-way analysis of variance 

 ( ANOVA) with significance determined at P<0.05. The ini- 

 tial analysis examined all reported or "cumulative" data. 

 The data were then partitioned in two additional ways: by 

 single returns and survey data. 



'Wenner, C. 1997. Personal commun. South Carolina Depart- 

 ment of Natural Resources, 217 Ft. Johnson Rd. Charleston, SC 

 29422-2559. 



Single returns 



This data set was the most restrictive. The assumption 

 was that the partitioned data would be free of any poten- 

 tial bias associated with captures of multiple fish, or with 

 monetary rewards or interactions with project staff 



Survey data 



The data were partitioned according to the angler's 

 answers during the interview to determine whether the 

 inducement of a $100 dollar reward changed his or her 

 reporting behavior This data set included all tags reported 

 individually, all tags of the same message reported as mul- 

 tiples, and all $100 tags. However, it excluded "reward" 

 tags in instances where answers during the interview sug- 

 gested that the angler's behavior had been changed by 

 capturing a fish with a "$100 reward" tag. 



Mean data for each of these analyses were reported with 

 standard errors. 



Results 



Nearly 95% of tags that were returned were reported 

 within 160 days after release of fish. More fish with 

 "reward" tags were reported than those with "$100 reward" 

 tags in one of the 12 release sites. Overall in SC, 151 

 anglers reported capture of 203 fish with tags. Anglers 

 reported capture of 1-9 red drum per trip. One hundred 



