FISHERY BULLETIN: VOL. 85. NO. 2 



DISCUSSION 



Larval northern anchovy feeding ecology was 

 similar to other larval fishes (Laurence 1977; Haeg- 

 ler and Outran 1978; Werner and Blaxter 1980; 

 Theilacker and Dorsey 1980; Blaxter and Hunter 

 1982; Eldridge et al. 1982; Houde and Schekter 

 1981, 1983). As the larval northern anchovy grew 

 they selected increasing larger prey, and on the 

 rotifer diets, their feeding and growth rates in- 

 creased with increasing concentrations of prey. In 

 addition to the growth response to different prey 

 concentrations, I observed differences in growth due 

 to prey type. Fish grew the fastest on the copepod 

 diet where one-third fewer calories were available 

 than were available in the high-density rotifer diet 

 (4 X 10-2 vs. 13 X 10-2 cal/mL for prey <150 ^m). 



Length at age obtained for the first 2 weeks by 

 larvae raised on copepods agreed with previous 

 studies where northern anchovy were fed copepods 

 and Gymnodinium was the first food (Kramer and 

 Zweifel 1970; Hunter 1976). However, the larvae 

 raised on copepods did not put on as much weight 

 per unit length as their counterparts fed rotifers and 

 Gymnodinium. Between days 5 and 9, the faster 

 growing, copepod-fed larvae had significantly lower 

 size-specific weights, and the weight exponents esti- 

 mated for the first 2 weeks were lower, 2.9, than 

 for larvae feeding on the high-density rotifer diet, 

 3.2. Lasker et al. (1970) found an exponent of 3.3 

 for northern anchovy fed on a high-density veliger 

 and Gymnodinium diet. 



These weight exponents estimated for northern 

 anchovy are lower than the exponents of about 4 

 reported for deep-bodied fish larvae (haddock, 

 flounder, cod, and scup; Laurence 1979). Likewise 

 the exponents obtained for Atlantic herring ranged 

 between 3.8 and 4.7 (reviewed by Checkley 1984). 

 Weights of fishes used to estimate the exponents 

 in the other laboratory studies ranged between 20 

 and 10,000 /.<g, whereas the northern anchovy 

 weights ranged between 20 and 100 ^g. Differences 

 in growth rates occur as larval fish grow (Zweifel 

 and Lasker 1976), and length- weight relations may 

 change over the size range. Additionally, larval mor- 

 phology is an obvious important component in the 

 length-weight relation, and experimental variables 

 may further complicate the relation. Moksness 

 (1982) reported a low weight exponent of 2.6 for 

 capelin, Mallotus villosus, from the field and from 

 a large rearing basin. Both Atlantic herring and 

 capelin larvae are similar in morphology to north- 

 ern anchovy. 



In previous laboratory feeding studies of larval 



northern anchovy, percent of body weight eaten per 

 day was usually higher than consumption rates, 

 31-86%, I estimated. Reported values were 126- 

 144% for northern anchovy (Hunter 1972), 197- 

 440% for bigeye anchovy (Chitty 1981), and 20-295% 

 for bay anchovy (Houde and Schekter 1981). Con- 

 sumption varies with prey concentration and tem- 

 perature, and the differences may be due to the 

 experimental conditions. In the other studies, tem- 

 peratures were higher and food concentrations were 

 both lower and higher, with some including Chlorella 

 blooms. However, it is likely that the differences are 

 due to use of average food weights in the other 

 studies. First-feeding northern anchovy larvae select 

 small prey (Table 7; Fig. 1), and if average prey 

 weight is larger than those actually being eaten by 

 the larvae, their consumption may have been over- 

 estimated. For example, in Table 2, the average full 

 stomach (Cj^ax) of 4.25 mm fish (day 4-5) fed 25 roti- 

 fers/mL contained 15.55 rotifers while the average 

 4.75 mm fish (day 5-6) contained 7.79 rotifers. Using 

 the width-specific weights (Table 1), I converted the 

 15 small rotifers to 1.8 /ig and the 8 larger ones to 

 2.1 Mg. 



The exponent for the regression relating oxygen 

 consumption to northern anchovy weight was 0.834 

 for larvae weighing <0.14 mg and 0.697 for larger 

 larvae. Exponents have been reported for larval 

 winter flounder, Pseudopleuronectes americanus 

 (0.74; Laurence 1975); larval bay anchovy, Anchoa 

 mitchilli (0.8; Houde and Schekter 1983); larval sea 

 bream, Archosargus rhomboidalis, and larval lined 

 sole, Achirus lineatus (0.838 and 0.942; Houde and 

 Schekter 1983). Brett and Groves (1979) suggested 

 a weight exponent of 0.86 for adult fish. There is 

 considerable variation in respiration rates in the 

 literature, probably depending on experimental con- 

 ditions. The respiration rates given here for north- 

 ern anchovy range between 3.1 and 6.9 ^L/mg per 

 hour for 0.02-2.7 mg larvae, and they are compar- 

 able to rates given for other fishes of similar age 

 and size (reviewed by Theilacker and Dorsey 1980). 



The evacuation rates determined here, 1.15-2.73 

 hours, for actively feeding northern anchovy larvae 

 are comparable to rates of 1-3 hours estimated by 

 Arthur (1976) for field samples. Particle residence 

 time in the gut depended on prey density, prey type, 

 and experimental design, e.g., feeding vs. nonfeed- 

 ing larvae. Werner and Blaxter (1980) also showed 

 that evacuation rates for herring fed Artemia were 

 more rapid at higher prey densities, and because live 

 and undigested prey were defecated, assimilation 

 must have been low at these high prey densities. 



Northern anchovy larvae, like herring, are con- 



224 



