772 



Fishery Bulletin 97(4), 1999 



catchability or availability of the different prey 

 groups to the trawl are the same. We do not have 

 estimates of the catchability for each of the prey spe- 

 cies used in this analysis. Our results would be bi- 

 ased if these differ. For example, catchabilities for 

 squid and butterfish may be higher than that of bay 

 anchovy and may explain the strong selection for bay 

 anchovy. 



Prey-size analysis 



Although there was a weak positive linear relation- 

 ship between ingested bluefish prey sizes and blue- 

 fish length, it was mainly due to prey other than bay 

 anchovy. The slope of the regression of bay anchovy 

 length on bluefish length alone was 50% and 75% 

 less than the slope of the regression of all prey on 

 bluefish length in 1994 and 1995, respectively. The 

 sizes of bay anchovy taken by the smallest YOY blue- 

 fish were similar to those taken by the largest YOY 

 bluefish, especially in 1995. Juanes and Conover ( 1995 ) 

 and Scharf et al. ( 1997 ) saw a similar relation for blue- 

 fish feeding on piscine prey in New York estuaries. 



Most ratios of bay anchovy length to bluefish length 

 were similar to those seen in past estuarine work 

 (Juanes and Conover, 1995; Scharf et al., 1997). 

 Juanes and Conover (1995) hypothesized that the 

 shift in spring-spawned bluefish diet from Atlantic 

 silversides to bay anchovy in Great South Bay, NY, 

 was a result of the higher capture success on the rela- 

 tively smaller bay anchovy. The capture success of 

 bluefish feeding on bay anchovy is likely higher than 

 that of bluefish feeding on the relatively larger but- 

 terfish and squid. Our size selectivity analysis shows 

 that even within bay anchovy there is preference for 

 smaller sizes (Table 7). 



The relatively small size of the 1995 SNE sum- 

 mer-spawned bluefish (1994 mean FL=135 vs. 1995 

 mean FL=70) influenced the extent of piscivorous 

 feeding in this cohort. The capture success was plot- 

 ted for bluefish feeding at different ratios of prey 

 length to bluefish length from laboratory work on 

 Atlantic silversides (Scharf et al., 1998b). Summer- 

 spawned, small-size bluefish in 1995 would have 

 experienced greatly reduced capture success when 

 feeding on bay anchovy. 



Feeding chronology, daily ration estimates, and 

 impacts on bay anchovy 



Bluefish appear to feed mainly during the day or at 

 dawn or dusk. This was also noted in past work on 

 juvenile bluefish in estuarine environments (Juanes 

 and Conover, 1994; Buckel and Conover, 1997). In 

 the Hudson River estuary, the diel feeding patterns 



of bluefish appeared to be linked to their diel move- 

 ment patterns ( Buckel and Conover, 1997 ). This may 

 also be the case for shelf bluefish; Munch ( 1997 ) found 

 that the catch per unit of effort (CPUE) ofYOY blue- 

 fish in NEFSC-NMFS bottom trawl collections was 

 significantly higher in diurnal than in nocturnal col- 

 lections. Munch (1997) hypothesized that this diel 

 movement to near bottom habitat (which makes them 

 available to bottom trawling) during the day may be 

 in response to the vertical migration of bay anchovy. 

 Bay anchovy exhibit diel vertical migration and are 

 significantly more available to bottom trawl sampling 

 during the day compared with at night ( Vouglitois et 

 al., 1987). 



Our daily ration estimates for spring-spawned 

 bluefish (3.7-9.0 g Ag • d) • 100) are similar to past 

 estimates of daily ration for September. Measure- 

 ments of bluefish maximum daily ration from a labo- 

 ratory mesocosm experiment (18.5-23°C) for mid- 

 September were -10 g/(g  d) • 100 (Buckel et al., 

 1995). Field estimates of bluefish daily ration in the 

 Hudson River estuary were 10.1-12.6 in mid to late 

 August and 7.3 in mid-September (Buckel and 

 Conover, 1997). Mean spring-spawned bluefish size 

 was larger and mean temperatures were slightly 

 lower on the shelf compared with bluefish sizes and 

 temperatures in the above laboratory and field ex- 

 periments. This finding may explain our slightly 

 lower daily ration estimates for spring-spawned blue- 

 fish on the shelf in the autumn. Hartman and 

 Brandt's (1995b) laboratory estimates of maximum 

 daily ration (C,,^^^) are also similar to our field esti- 

 mates of spring-spawned bluefish daily ration esti- 

 mates on the shelf From their experiments, esti- 

 mates of C„|g, for bluefish ranging from 100 to 350 g 

 at 20°C (temperature at which bluefish were found 

 on the shelD are 7.5 to 12.5 g/(g • d)  100. 



Owing to low sample sizes and an inadequate diel 

 record, the daily ration of summer-spawned bluefish 

 was determined only in the SNE region in 1994 and 

 1995. The daily ration estimate in 1994 was lower at 

 8.5 g/(g • d) • 100 than the estimate of 12.5 g/(g  d ) • 

 100 in 1995. The mean size offish in this cohort varied 

 substantially between 1994 (31.2 g) and 1995 (4.8 g). 

 The size and the temperature difference (1994= 

 18.9°C vs. 1995=20. 7°C ) may explain the differences 

 in daily ration. The daily ration estimates of these 

 fish are slightly lower than would be predicted from 

 laboratory estimates of maximum daily ration of 

 bluefish at these sizes and temperatures (Buckel et 

 al., 1995; Hartman and Brandt, 1995b); however, 

 both shelf estimates fall into the 95% confidence in- 

 terval of the summer-spawned bluefish daily ration 

 estimate (19-20 September 1992) in the Hudson 

 River (Buckel and Conover, 1997). 



