FISHERY BULLETIN: VOL. 84, NO. 4 



the smallest larval size group (6.7 mm) were the 

 prey. 



Predator Behavior 



Sighting distances, persistence of the attack, at- 

 tack speed, and other characteristics of predator 

 behavior were not well documented in our experi- 

 ments because our focus was on the larvae. Such 

 information could be quite useful if one were to 

 develop a predation model for northern anchovy lar- 

 vae, using northern anchovy or chub mackerel 

 predators. We provide some general observations 

 on the behavior of the predators. 



Chub mackerel attacked 6.7 mm larvae from a 

 shorter distance than larger larvae (£-test, P = 0.05), 

 but no statistically significant trend was evident 

 when northern anchovy were predators. Mean at- 

 tack distances were a poor measure of sighting 

 range as they included repeated, short-range attacks 

 on the larger larvae. We observed both predator 

 species swimming within 2-3 dm of the smallest lar- 

 vae without attacking them, whereas larger larvae 

 were always attacked from this distance indicating 

 that sighting distances may be shorter for small 

 larvae. 



Adult northern anchovy usually attacked a larva 

 only once during a feeding sequence, and if the larva 

 escaped, it was rarely attacked again or pursued. 

 On the other hand, if the chub mackerel did not cap- 

 ture the larva on the first attack, it usually turned 

 and attacked again. Chub mackerel usually chased 

 an escaping larva until it was captured. The attack 

 speeds of adult northern anchovy, although not 

 measured, seemed to be similar over a wide range 

 of larval prey sizes, whereas the attack speeds of 

 chub mackerel clearly were faster when attacking 

 larvae greater than about 10 mm SL than when at- 

 tacking smaller larvae. 



DISCUSSION 



Factors Affecting Larval Vulnerability 



A low level of responsiveness seems to be the 

 dominant feature of the vulnerability of northern 

 anchovy larvae to fish predators over the smallest 

 larval size classes we tested (6-10 mm SL). Presum- 

 ably northern anchovy larvae <6 mm would respond 

 even less frequently, as Webb (1981) found that only 

 9% of 2.9 mm northern anchovy larvae responded 

 to the aquarium fish Amphiprion percula, whereas 

 about 30% of 6 mm larvae did so. During this period 

 vulnerability of northern anchovy larvae to fish 



predators seems to be primarily a function of visual 

 detection by the predator, because when the larvae 

 are detected they have a low probability of escap- 

 ing. Our data on predation rates and maximum at- 

 tack distances indicate that predation in the sea on 

 the small, young larval stages might be lower than 

 expected because of the short range at which such 

 larvae may be detected. Thus, factors that affect the 

 distance at which larvae are detected by predators, 

 such as larval size, visual contrast, and water clar- 

 ity (Vinyard and O'Brien 1976), may be the most im- 

 portant variables during the first 3 wk of life. As 

 larvae grow they more often respond to the attacks 

 of predators and escape them more frequently. 

 Maturation of visual and lateral line systems (O'Con- 

 nell 1981) may be the principal cause of this general 

 increase in responsiveness with larval length. Al- 

 though older larvae are more responsive, they are 

 also more readily detected by predators because 

 they are larger and have more pigmentation. Im- 

 proved avoidance behavior may not completely com- 

 pensate for the greater visibility of larvae in the 8-12 

 mm range, as our data on predation rates by north- 

 ern anchovy indicated that the rates of consump- 

 tion were highest for larvae in this range. 



Larvae longer than 20 mm responded more fre- 

 quently to northern anchovy than to chub mackerel 

 predators, possibly because chub mackerel attacked 

 such large larvae at much higher speeds. At higher 

 attack speeds, less time is available for the larvae 

 to respond; consequently, predators with the most 

 rapid attack speeds evoke the lowest proportion of 

 prey responses (Webb 1982). Thus one might expect 

 a larva to respond to small fish predators more fre- 

 quently than to larger ones, since attack speed 

 would be expected to increase with predator size. 

 This may explain why northern anchovy larvae 

 (2.9-12 mm SL) responded more frequently to the 

 small Amphiprion (44 mm) (Webb 1981) than they 

 did to either northern anchovy or chub mackerel 

 predators (Fig. 4). The pectoral swimming of 

 Amphiprion might also provide more cues of an im- 

 pending attack than did the swimming movements 

 of either northern anchovy or chub mackerel. 



In addition to size-specific avoidance capabilities 

 and visibility, many other larval characteristics 

 affect their vulnerability to predators. We briefly 

 consider here three of these: effects of starvation, 

 effects of the onset of schooling, and effects of varia- 

 tions in larval growth rates. Clupeoid larvae undergo 

 degradation of muscle and other tissues during star- 

 vation, and a reduced predator avoidance behavior 

 might be anticipated (Ehrlich 1974; O'Connell 1980). 

 In a preliminary experiment Folkvord (1985) 



866 



