268 



Abstract. —Predation and cannibal- 

 ism during larval life are potential de- 

 terminants of recruitment and year- 

 class strength in marine fish. We stud- 

 ied predation by larval Atlantic mack- 

 erel. Scomber scombrus, 3—14 mm in 

 length, on conspecifics and other fish 

 larvae over Sable Island Bank (Scotian 

 Shelf) in July 1991. With increasing 

 length, the diet of mackerel larvae 

 shifted from copepod nauplii to copep- 

 ods and fish larvae. Mackerel <5 mm 

 selected strongly for copepod nauplii 

 Mackerel >6 mm preyed indifferently 

 on copepods (i.e. in proportion to their 

 abundance in the plankton) but se- 

 lected strongly for fish larvae. The lar- 

 val fish prey of Atlantic mackerel lar- 

 vae comprised conspecifics (66%), yel- 

 lowtail flounder. Limanda ferrugmea 

 (18%), silver hake, Merluccius bili- 

 nearis ( I2 r i ), and redfish, Sebastes spp. 

 (4^) larvae. Predation was stage-spe- 

 cific: only the newly hatched larvae of 

 a given species were ingested. Based on 

 a conservative estimate of the transit 

 time of fish larvae in the gut (24 h), 

 estimates of prey mortality by Atlantic 

 mackerel larvae ranged from 7% to 

 57%/d (average 31%) of the standing 

 stock of newly hatched larvae of suit- 

 able prey size. Piscivory was limited at 

 densities offish larvae <0.1/m ! . Preda- 

 tion on fish larvae by Atlantic mack- 

 erel larvae declined with increasing 

 density of nauplii in the environment 

 and with increasing number of alter- 

 native copepod prey ingested. Because 

 the capacity of Atlantic mackerel lar- 

 vae to digest fish larvae was limited, 

 cannibalism was density independent 

 and predation on other species was 

 depensatory. Hence, despite its poten- 

 tial importance during the first days 

 after hatching, cannibalism and preda- 

 tion by Atlantic mackerel larvae 3-14 

 mm long do not appear to contribute to 

 the density-dependent regulation of At- 

 lantic mackerel and other prey species. 



Cannibalism and predation on fish 

 larvae by larvae of Atlantic mackerel, 

 Scomber scombrus: trophodynamics 

 and potential impact on recruitment* 



Louis Fortier 

 Ann Villeneuve 



GIROQ. Departement de biologic University Laval 

 Ste-Foy, Quebec GIK 7P4. Canada 



Manuscript accepted 26 October 1995. 

 Fishery Bulletin 94:268-281 1 1996). 



Several lines of indirect evidence 

 suggest that predation is a major 

 determinant of early survival in 

 marine fish (Bailey and Houde, 

 1989): 1) the high mortality rate of 

 eggs and yolk-sac larvae in the field 

 (Fossum, 1988; Incze and Campbell, 

 1989); 2) the high estimates of pre- 

 dation mortality obtained by sub- 

 tracting starvation mortality from 

 total mortality (Hewitt et al., 1985; 

 Leak and Houde, 1987); 3) the vul- 

 nerability of fast-growing, healthy 

 larvae to predators in mesocosm 

 experiments (0iestad, 1985); 4) the 

 wide range of potential predators of 

 prerecruit fish in the sea (Hunter, 

 1984; Bailey and Houde, 1989); and 

 5) the variety of spawning strate- 

 gies that can be interpreted as re- 

 ducing the vulnerability of eggs and 

 larvae to predators (Bailey and 

 Houde, 1989). A recent field study 

 comparing concurrent measure- 

 ments of predation mortality and 

 total mortality confirmed that 

 predators such as medusae can 

 have a significant impact on the 

 early survival of bay anchovy, 

 Anchoa mitchilli, in Chesapeake 

 Bay (Purcell et al., 1994). Yet, in 

 most ecosystems, the actual preda- 

 tors of fish larvae, the trophody- 

 namics between predators and fish 

 larvae, and the role of predation in 

 regulating year-class strength re- 

 main unclear (Hunter, 1984; Ander- 



son, 1988; Bailey and Houde, 1989; 

 Heath, 1992). 



Scombrid larvae often prey on the 

 larval stages of their own and other 

 fish species (Hunter and Kimbrell, 

 1980; Jenkins et al., 1984; Finucane 

 et al., 1990; Young and Davis, 1990). 

 In particular, Atlantic mackerel, 

 Scomber scombrus, larvae capture 

 fish larvae sometimes more than half 

 their own length (Grave, 1981). The 

 diet of Atlantic mackerel larvae 

 changes markedly during ontogeny. 

 First-feeding larvae may be phy- 

 tophagous, young post-yolk-sac lar- 

 vae feed primarily on copepod nau- 

 plii and copepodites, and Atlantic 

 mackerel >6.5 mm in length add con- 

 specifics and other fish larvae to their 

 diet (Grave, 1981; Peterson and Ausu- 

 bel, 1984; Ware and Lambert, 1985). 



Because the probability of en- 

 counter between planktonic larvae 

 increases with increasing density, 

 Grave (1981) proposed that intra- 

 cohort cannibalism could be a 

 simple mechanism for the density- 

 dependent regulation of year-class 

 strength in Atlantic mackerel. In 

 the present study, we examined the 

 diet of Atlantic mackerel larvae and 

 their selectivity for prey on Sable 

 Island Bank (Scotian Shelf) in July 



Contribution to the research program of 

 GIROQ (Groupe interuniversitaire de 

 recherches oceanographiques du Quebec). 



