FISHERY BULLETIN: VOL. 79, NO. 3 



12 July 1979 in the lower portion of Chincoteague 

 Bay, Va. (Figure 1). Located within the summer 

 distribution of this species, the bay supports a 

 relatively large number of young sandbar sharks 

 from early June through September (pers. obs.). 

 Average water depth of the bay is 2 m, but many 

 areas with strong current flow have depths as 

 great as 12 m. A tidal inlet connects the bay with 

 the Atlantic Ocean and tidal range varies from 

 0.75 to 2.00 m. Salt marshes with numerous tidal 

 creeks, brackish to seawater salinities, and other 

 conditions which seem typical of the nursery 

 grounds of this shark along the middle Atlantic 

 coast also characterize the area. 



A 4.9 m outboard motorboat was used as a fish- 

 ing platform and sharks were caught using sport 

 fishing rods with 3/0 Penn^ reels. Terminal 

 tackle consisted of a 0.3 m wire leader with a 

 straight-shank, ball-eye fishing hook. To increase 

 fishing effort, two leaders were attached (0.5 m 

 apart) on each fishing line. To facilitate captures 

 over the entire size range of sharks in the area, 

 each line was rigged with a 3/0 and a 8/0 hook. A 

 lead sinker or cork float was attached to adjust the 

 lines to the desired fishing depth. Cut pieces of 

 freshly frozen Atlantic menhaden, Brevoortia 

 tyrannus, were used as bait and each hook was 



''Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



Figure L — Chart of the Chincoteague Bay study area (modified 

 from National Ocean Survey Charts 12210 and 12211). x's give 

 the locations of the 10 24-h fishing stations. 



rebaited hourly. Nothing was thrown into the 

 water to attract sharks except the bait. Upon cap- 

 ture each shark was brought into the boat where it 

 was identified, sexed, and measured. The shark 

 was then either tagged and released or sacrificed 

 for stomach content analysis. 



The type of fishing conducted fell into two cat- 

 egories. On 10 occasions, continuous fishing sta- 

 tions of approximately 24-h duration were com- 

 pleted (three in 1977, six in 1978, and one in 

 1979). These stations occurred at 10 different lo- 

 cations (Figure 1), each with a water depth >3 m. 

 On all stations, the boat was anchored and three 

 rods (rigged and baited as described above) were 

 used with a rod fishing at the surface, middepth, 

 and bottom. Fishing technique was standardized 

 from station to station. After catching a shark, 

 the rod was immediately replaced by another so 

 that fishing effort was not interrupted. On 

 52 other occasions (20 in 1977 and 32 in 1978), 

 shorter fishing periods were completed at ran- 

 domly selected locations. The duration of these 

 fishing periods and the time of day that they were 

 conducted varied, but in a random fashion. The 

 same fishing gear was employed at these times, 

 but shallow as well as deep areas were fished. 

 These shorter fishing periods are hereafter re- 

 ferred to as miscellaneous fishing stations. 

 During both types of fishing, data concerning 

 tidal current flow and water temperature were 

 collected. 



Strikes by fishes not captured were not in- 

 cluded in the analysis of data. For each fishing 

 station, catch per unit of effort (CPUE) was cal- 

 culated by dividing the total number of sharks 

 caught by the total number of hours spent fish- 

 ing. CPUE is expressed as the number of sandbar 

 sharks caught per hour of fishing using the three 

 fishing rods previously described. CPUE was 

 also calculated for various comparison categories 

 within fishing stations such as day and night 

 periods. For comparison categories involving 

 fishing depth, CPUE at each depth was based on 

 one fishing rod. Where mean CPUE is referred to 

 in the results, the value is the arithmetic mean of 

 CPUE values calculated for each fishing station. 

 Because values of CPUE were not normally dis- 

 tributed, nonparametric methods of data analysis 

 were employed. The tests used are described by 

 Hollander and Wolfe (1973). Due to the large 

 number of statistical comparisons made, many of 

 the test names and probability values are given 

 in tables rather than in the text of the results. 



442 



