Fortier and Villeneuve Cannibalism and predation by Scomber scombrus larvae 



271 



affected by changes in the abundance of potential 

 prey and is thus suitable for describing ontogenetic 

 variations in prey selectivity and for comparing or 

 averaging selectivity among stations with different 

 prey densities (Pearre, 1982). A value of a f >\IN in- 

 dicates that prey j is selected for, whereas a <VN 

 indicates that the prey is rejected. 



All dissected mackerel larvae (feeding and non- 

 feeding) were included in the determination of diur- 

 nal patterns of feeding activity. Larvae with empty 

 guts occurred primarily during late-night and early- 

 morning hours (0000-0800 h). These larvae were 

 excluded from the calculations of feeding ratios (mean 

 number of prey organisms per gut of feeding mack- 

 erel) or of piscivory (proportion of feeding mackerels 

 with at least one fish larva in the gut). Note that the 

 calculation of a feeding ratio for Atlantic mackerel 

 larvae feeding on fish larvae was meaningless be- 

 cause the gut of mackerel larvae seldom (3% of all 

 cases) contained more than one fish larva. 



Results 



Zooplankton and fish larvae 



Copepods numerically dominated the zooplankton 

 (Table 2). The cyclopoid Oithona similis, harpacticoid 

 copepods, and the calanoid Pseudocalanus sp. were 

 particularly abundant. The low mean densities of 

 copepod eggs and nauplii ( <2/L ) may reflect the wide 

 depth interval sampled by the RMT ( 0-75 m ) because 

 these organisms are often concentrated at the 

 pycnocline or in the surface mixed layer. 



Representatives of at least 18 genera offish were 

 collected (Table 2). Silver hake, yellowtail flounder, 

 Atlantic mackerel, and redfish made up the bulk 

 (98%) of the larval fish assemblage. Atlantic mack- 

 erel, silver hake, and yellowtail flounder larvae were 

 found primarily over the shallow area of Sable Is- 

 land Bank (Fig. 2). Redfish larvae were more evenly 

 distributed in the sampling area than were the other 

 species. Vertically, Atlantic mackerel, yellowtail 

 flounder, and redfish larvae aggregated in the sur- 

 face layer (<20 m) at night and moved deeper in the 

 water column during the day (Fig. 3). In the morn- 

 ing and afternoon, Atlantic mackerel were concen- 

 trated in the thermocline at 10-20 m, whereas yel- 

 lowtail flounder and redfish were found in or below 

 the thermocline. At 2000 h, all four species were most 

 abundant in the thermocline. During the day, the 

 newly-hatched silver hake larvae were found prima- 

 rily in the surface layer (0-20 m). Too few larvae were 

 captured at night to confirm any diel migration pat- 

 tern in this species. 



Table 2 



Mean density ( ± standard deviation land percent composi- 

 tion of the zooplankton and larval fish assemblages at sta- 

 tions where Atlantic mackerel larvae, Scomber scombrus, 

 were captured (n = 14) from 18 to 23 July 1991, southwest 

 of Sable Island (Scotian Shelf). 



' Includes Sebastes mentella and Sebastes fasciatus. 



2 Includes unidentified fish larvae, pleuronectids, gonostromatids, 

 stichaeids, and Cyclopterus lumpus (lumpfish) in order of de- 

 creasing average density 



