Marks and Conover: Ontogenetic shift in diet of young-of-year Pomatomus saltatrix 



103 



400 600 1200 1600 2000 2400 



TIME 124 hours) 



Figure 6 



Gut fullness expressed as a ratio of total prey dry 

 weight igi divided by individual dry weight (g) 

 ( x 100) for bluefish Pomatomus saltatrix, as a func- 

 tion of time of day. Squares represent the mean; 

 vertical lines represent ±1SE. Sample sizes are pro- 

 vided above each vertical line. 



showed that time to 90% digestion at 21°C is 

 5-7 h post-feeding in 57-199 mm bluefish 

 (Marks 1991). 



A two-way test for independence (adjusted 

 G-statistic, G od =5.98, P<0.05; Sokal & Rohlf 

 1981) showed that the occurrence of fish in 

 the stomachs depended on time of day. 

 Piscivory was restricted to daylight hours, with 

 some incidence recorded prior to 2400 h but 

 none during 2400-0400 h. A slight increase 

 (12%) was observed at 0800 h; the maximum 

 evidence of piscivory (42%) occurred at 1600 h. 



Mouth morphology 



Teleost prey had the greatest maximum thick- 

 ness of all prey types. The relationship be- 

 tween thickness offish prey at the widest point 

 and P. saltatrix mouth width was positive, but 

 the regression was marginally non-significant 

 (P=3.91, ra=23, 0.05<P<0.1). 



The log-log regression of mouth width on 

 total body length was highly significant 

 (r^.857, rc=203, P<0.001) and showed no evi- 

 dence of an abrupt change in mouth width 

 coincident with the onset of piscivory (Fig. 7). 

 The regression slope (1.06) suggests isometric 

 growth of mouth width in relation to body size. 

 A SYSTAT piecewise linear-regression proce- 

 dure (Wilkinson 1987) was employed to test 



for the existence of more than one slope in the line. The test 

 was run with an undetermined intersection as well as with an 

 a priori intersection at 30 mm (the size of initiation of piscivory). 

 No significant differences in slope were observed, and the con- 

 clusion was that a single regression line adequately described 

 the relationship between mouth width and total body length. 



Discussion 



Our results indicate that during the oceanic phase of the early 

 life history, P. saltatrix gradually shifted its diet from small 

 prey (crustaceans, mainly copepods) to fish and crab larvae. 

 The diet appeared to shift over a size range that corresponds 

 to the period of inshore migration. 



The smallest juvenile P. saltatrix foraged predominately on 

 copepods. A diet high in crustacean prey is characteristic of the 

 larval and juvenile phases of many piscivorous fishes, includ- 

 ing the Atlantic cod Gadus morhua (Last 1980) and the large- 

 mouth bass Micropterus salmoides (Keast 1985). 



Overall, fish prey became a substantial component of the 

 diet at about 40-60 mm. The majority of teleosts consumed 

 were hakes (Urophycis spp.) or other unidentified gadids. Gadids 

 are among the most-abundant icthyoplankton group found in 

 the MAB during the spring (Kendall & Nalpin 1981, Morse 

 1989, Cowen et al. In press). 



Crab larvae initially appeared in the diet of bluefish <40 mm 

 and accounted for 9% of the diet by weight in bluefish >60 mm 

 (Fig. 4). The appearance of crab larvae as the third most-abun- 

 dant prey item is not surprising. The majority of crab larvae 

 were found in spring-spawned bluefish captured close to shore. 

 This may be a result of higher concentrations of crab larvae at 

 frontal zones in the nearshore environment. Friedland et al. 

 (1988) reported a high abundance of crustaceans in the diet of 

 inshore juveniles. 



