102 



Fishery Bulletin 91(1), 1993 



continued to feed mostly on copepods. The 

 largest size-class of summer-spawned blue- 

 fish increased their consumption of fish prey 

 and continued to take copepods and miscel- 

 laneous crustaceans. 



The diet of spring-spawned bluefish (19.0- 

 39.9 mm) was similar to the diet of same- 

 sized fish from the summer-spawned cohort. 

 There was a marked difference, however, in 

 the amount of fish prey consumed by the 

 40.0-49.9 mm size-class. This difference was 

 also evident in the largest size-class, despite 

 small sample sizes in both cohorts. The diet 

 of the two largest size-classes of spring- 

 spawned bluefish also contained a relatively 

 high percentage of crab larvae. 



Mean prey length increased significantly 

 with predator size in both spring- and sum- 

 mer-spawned bluefish (Fig. 4). Variance about 

 the regression, however, was much greater 

 for spring-spawned fish. Since transforma- 

 tion did not correct for inequality in the vari- 

 ances, we did not attempt further statistical 

 analyses of these data. 



Bluefish belonging to the spring-spawned 

 cohort were caught either close to shore 



(<8km) or at distances > 100 km from shore (Fig. 1). Piscivorous 

 individuals from this cohort were larger in size (x TL=51.6 mm) 

 and generally captured closer to shore than non-piscivorous in- 

 dividuals (x TL=37.1mm). The difference between the mean TL 

 for piscivorous fish versus non-piscivorous fish was significant 

 (P<0.001, df=50, *„=6.22). 



In comparison, piscivorous bluefish from the summer-spawned 

 cohort were more evenly dispersed across the shelf (Fig. 2). Mean 

 TL of piscivorous individuals (41.0 mm) was significantly greater 

 than that of non-piscivorous fish (x TL=33.7, P<0.001, df=92, 

 t s =5.04). 



Pomatomus saltatrix tended to feed on larger teleost prey with 

 increasing size. A linear regression of prey fish length on preda- 

 tor length was highly significant (r 2 =0.46, n=25, P<0.001; 

 Fig. 5). With increasing body size, P. saltatrix also tended to feed 

 on multiple teleost prey. More than one fish prey was found only 

 in P. saltatrix >49 mm. The maximum number of intact teleost 

 prey found in an individual predator (49 mm) was three. A t-test 

 for the difference in means between the size of fish that con- 

 sumed multiple fish prey (57.1 mm) and the size of those that fed 

 on single prey (44.8 mm) was significant (P<0.05, df=30, ^=3.265). 



Diel cycle 



Gut fullness pooled across all fish examined in the study versus 

 time of day is depicted in Fig. 6. Gut fullness was evaluated as 

 the ratio of total prey dry weight to individual P. saltatrix dry 

 weight (X100). Feeding peaked during 1600-2000h (the time of 

 sunset during the study period ranged from 1945 to 2020 h). A 

 nonparametric Spearman Rank Correlation test (Sokal & Rohlf 

 1981) resulted in a significant, positive relationship (r s =0.5818, 

 P<0.001, /? = 189) between time of day and gut fullness. The high 

 proportion of food in the gut at approximately 2400 h may be 

 attributed to food that was consumed during the evening crep- 

 uscular period, since prey from this time-period were generally 

 in an advanced stage of digestion. Experimental laboratory data 



25 



E 20 



£j 15 



i, 



Y=J.5107+(.1902)X 

 r 2 -.462 n-25 (P<0.001) 



20 



30 



40 



50 



60 



70 



BLUEFISH TOTAL LENGTH (mm) 



Figure 5 



Regression of prey fish total length (mm) on bluefish 

 Pomatomus saltatrix total length (mm I. 



