Davis and Berkson Effects of a simulated fishing moratorium on the stock assessment of Pogrus pagrus 



587 



Red porgy is a relatively well-studied species; more 

 than 30 years of information are available. However, the 

 majority of the data for this species is a direct result 

 of fishing operations. Fishery-dependent data for red 

 porgy include landings (number offish) from commer- 

 cial, recreational, and headboat fisheries, as well as 

 large sample sizes of lengths from these fisheries (Table 

 2; Huntsman et al., 1978; Low et al., 1985; NMFS^). 

 The headboat fishery also provides an abundance index 

 based on catch-per-unit-of-effort data. Although fishery- 

 dependent data provide large sample sizes and a rela- 

 tively long time series, there are a number of problems 

 associated with this information source. Fishery-depen- 

 dent statistics have been affected by frequent changes 

 in regulations, economics, technology, and gear (Collins, 

 1990) that may cause difficulties for comparisons of 

 landings, abundance indices, or lengths among years. 

 Moreover, harvest data have inherent biases due to 

 the nature of the fishery, which attempts to maximize 

 catch, selects fishing areas nonrandomly, and targets 

 larger size classes of fish. 



In an attempt to compensate for some of the biases 

 potentially present in harvest data, the National Ma- 

 rine Fisheries Service (NMFS) instituted the Marine 

 Resources Monitoring, Assessment, and Prediction 

 (MARMAP) program in 1979. During this annual sur- 

 vey, fishery-independent data are collected on a variety 

 of reef fishes in waters from North Carolina to Florida 

 (Collins, 1990; Harris and McGovern, 1997). In the 

 MARMAP survey standardized gear types are used 

 and known amounts of effort are applied to sample ran- 

 domly selected sites between May and August (Collins, 

 1990), during which information on length, weight, age, 

 sex, fecundity, and level of maturation are collected. 

 An abundance index is also calculated for each gear 

 type. 



'' NMFS (National Marine Fisheries Service). Fisheries Sta- 

 tistics Division. 2005. Website: http:www.st.nmfs.gov/stl 

 [accessed on 31 January 2005]. 



The MARMAP survey and other well-designed fish- 

 ery-independent surveys are extremely useful for stock 

 assessments because standardized gears are used with 

 known efforts in documented locations, which constitute 

 a statistically valid experimental design that facili- 

 tates comparisons among years, areas, and gear types 

 (Collins and Sedberry, 1991). However, the MARMAP 

 survey does not provide information on harvesting, and 

 the survey's sampling methods are inherently different 

 from fishery-dependent sources. Therefore, MARMAP 

 data cannot serve as an absolute substitute for fish- 

 ery-dependent information in existing stock assess- 

 ment strategies. In addition, MARMAP sample sizes 

 are much smaller than those from fishery-dependent 

 sources, particularly for length-frequency data; 92% of 

 samples for length-frequency estimates were obtained 

 from fishery-dependent sources (Table 2). 



In a highly restricted or closed fishery where fishery- 

 dependent data are limited, information from fishery- 

 independent sources may be the only data available 

 to managers. This decreased amount of data could 

 affect stock assessment results, including the estima- 

 tion of stock status indicators and benchmarks used in 

 management. It is therefore important to understand 

 how to best use both fishery-independent and fishery- 

 dependent data in assessments of red porgy and other 

 species, both in unrestricted and heavily restricted 

 fisheries. 



During a moratorium or other period of strict regu- 

 lation, the reduction in fishery-dependent data may 

 increase variability surrounding status indicators and 

 management benchmarks. We quantified the increase 

 in variability for red porgy by using stock assessment 

 model simulations and projections, and we identified 

 how the uncertainty surrounding estimates could af- 

 fect management decisions. By evaluating the effects 

 of a lack of data before harvest restrictions are put in 

 place, managers will be able to identify future data 

 needs, improving their ability to assess and manage 

 species of concern. 



