Buckel et al.: Foraging habits of Pomatomus sa/tatrix 



759 



Compared with the nearshore phase of the early 

 life history of bluefish, our knowledge of the forag- 

 ing ecology and predatory impact of bluefish on the 

 continental shelf is poor. Here we quantify the diet 

 of YOY and adult bluefish and determine YOY blue- 

 fish prey type and size selectivity patterns, foraging 

 chronology, daily ration, and biomass of prey con- 

 sumed during the autumn migration on the shelf. 



Methods 



Study area and collections 



YOY and adult bluefish and their potential prey were 

 collected on the U.S. east coast continental shelf dur- 

 ing the autumn of 1994 and 1995 aboard the research 

 \essel Albatross IV. Collections were made during 

 the National Marine Fisheries Service, Northeast 

 Fisheries Science Center's (NMFS-NEFSC) autumn 

 bottom trawl survey cruises at predetermined sta- 

 tions from Cape Hatteras, NC, to Nova Scotia. 

 Cruises began in early September and ended in late 

 October in both years (6 September to 28 October 

 1994 and 5 September to 27 October 1995). 



Descriptions of the survey design and the trawl 

 characteristics can be found in Azarovitz (1981). 

 Briefiy, tows were made with a no. 36 Yankee trawl 

 equipped with rollers with an opening 3.2 m high, 

 10.4 m wide, with 12.7-cm stretched mesh in the 

 opening, with 11.4-cm stretched mesh in the codend, 

 and with a 1.25-cm stretched mesh lining in the 

 codend and upper belly to retain YOY fishes. Tows 

 were 30 minutes in duration at 3.5 knots in relation 

 to bottom and were conducted on a 24-h basis. 



Dietary analyses 



Adult and YOY bluefish were distinguished by us- 

 ing length-age relationships. Munch and Conover ( in 

 press) examined the annual size distributions of blue- 

 fish from autumn bottom trawl surveys conducted 

 since the 1970s in four different regions of the shelf: 

 SOC=South of Chesapeake Bay; C-D=Chesapeake 

 Bay to Delaware Bay; SNE=Southern New England 

 (Delaware Bay to Narragansett Bay); and Georges 

 Bank. On the basis of these size distributions and 

 backcalculated sizes at age 1 (Chiarella and Conover, 

 1990, and references therein), they classified YOY 

 bluefish in autumn as those fish <300 mm FL. We 

 classified adults as those bluefish >300 mm FL and 

 we distinguished spring- and summer-spawned YOY 

 bluefish for each region in 1994 and 1995 on the ba- 

 sis of bimodality in length-frequency distributions 

 (Munch and Conover, in press). We also adopted 



Munch and Conover 's (in press) geographical bound- 

 aries for this analysis. In 1994, summer-spawned blue- 

 fish sizes by region were as follows: SOC, FL<120 mm; 

 C-D, FL<160 mm; and SNE, FL<160 mm. In 1995, 

 summer-spawned bluefish sizes were as follows: 

 SOC, FL<100 mm; C-D, FL<150 mm; and SNE, 

 FL<140 mm. Spring-spawned bluefish were those 

 fish larger than the summer-spawned cohort by re- 

 gion but <300 mm FL. 



Diets of spring- and summer-spawned juveniles 

 and adult bluefish were quantified. Bluefish taken for 

 stomach content analysis were wet weighed (±1.0 g) 

 and measured for fork length, FL (to 1.0 mm). Stom- 

 achs were removed at sea and preserved in 10% for- 

 malin buffered with sea water. On some occasions, 

 whole fish were either frozen or presei-ved in lO'^ for- 

 malin and then processed in the laboratory. Stomach 

 contents of bluefish were identified to the lowest pos- 

 sible taxon, enumerated, blotted dry, weighed (±0.01 g), 

 and measured ( TL for fish prey and mantle length for 

 squid, ±1.0 mm). Eye diameter (±0.1 mm) and caudal 

 peduncle height (±0. 1 mm ) were measured for partially 

 eaten bay anchovy and butterfish prey and converted 

 into TL from linear regression equations (Scharfet al., 

 1997;Scharfetal., 1998a). 



Each trawl containing bluefish provided us with a 

 group or "cluster" of bluefish for a given station. 

 Therefore, for our analysis, the mean and variance 

 of diet indices were calculated with cluster sampling 

 estimators developed at the NEFSC-NMFS, Woods 

 Hole, MA (Cochran, 1977, Fogarty, unpubl. data). 

 These calculations are described in a previous study 

 that examined bluefish diet in the Hudson River es- 

 tuary (Buckel et al., 1999). 



Net feeding 



During a 30-minute tow, bluefish may feed within 

 the trawl (an activity known as net feeding). Net feed- 

 ing could bias the diet index estimates or affect gut 

 fullness level estimates (or do both). We tested for 

 such bias by classifying fish or squid prey as either 

 1) "fresh" or 2) "digested" during our stomach con- 

 tent examination. "Fresh" prey had no sign of diges- 

 tion and "digested" prey were either partially or well 

 digested (e.g. they were anywhere from starting to 

 lose skin to being only identifiable by skeleton or 

 shape). The percent occurrence of four dominant prey 

 (bay anchovy, striped anchovy, butterfish, and squid) 

 were compared between fresh and digested catego- 

 ries in YOY bluefish. If the percent occurrence of each 

 prey category within a cohort and region was simi- 

 lar for fresh and digested prey, it would suggest that 

 either there was no net feeding or that it did not af- 

 fect diet index estimates. 



