RESPONSES OF NORTHERN ANCHOVY, ENGRAULIS MORDAX, LARVAE 

 TO PREDATION BY A BITING PLANKTIVORE, AMPHIPRION PERCULA 



P. W. Webb* 



ABSTRACT 



Responses of northern anchovy larvae, ranging from 0.29 to 1.2 cm total length, to attacks by a biting 

 planktivore, the clown fish, were recorded on video tape. Schlieren optics were used to simultaneously 

 view an opaque predator and transparent prey All fish were reared, and experiments performed, at 20° 

 C. The percentage of larvae responding to attacks increased from about 9% for 0.29 cm larvae to SO'/r for 

 1.2 cm larvae. Of these larvae responding to attack, 26 + 10% attempted to escape too late and were 

 caught. This proportion was not related to larval size. The direction of larval escape paths to the initial 

 orientation of the body was not related to larval size, but escape distances traveled and mean escape 

 speeds increased with size from 1.0 cm and 4.0 cm/s respectively for larvae 0.29 cm total length to 3.5 cm 

 and 8.2 cm/s in larvae 1.2 cm total length. Larval performance was not maximal except during rare 

 chases which occurred in 7% of attacks by the predator. 



Escape maneuvers in vertebrates are initiated by the magnifying retinal image of an approaching 

 object, called the looming effect, and calculated as the rate of change of the angle subtended by the 

 predator as seen by the prey In the present study apparent looming thresholds for larval avoidance 

 responses were calculated at the start of the response, and differ from the true response because there 

 must be a finite time difference between the stimulus exceeding the response threshold and the motor 

 response itself (response latency). Maximum likelihood mean apparent looming thresholds were 

 calculated for log-transformed data, assuming nonresponding prey had apparent looming thresholds 

 greater than the maximum actually observed. These mean apparent looming thresholds decreased with 

 larval length from approximately 32 rads/s at 0.29 cm to 1.7 rads/s at 1.2 cm. The most important 

 feature of the larval avoidance response was that an escape attempt should be made appropriately 

 timed to an attack. 



The method could be used to examine larval fish responses in other interactions where predation 

 events take place over a small distance. Examples are attacks by biting juvenile and adult fish, and 

 other planktonic invertebrate predators. Effects of larval density and alternate prey species could also 

 be evaluated. 



Predation and starvation are believed to be the 

 dominant factors contributing to the extensive 

 mortality offish eggs and larvae (Blaxter 1969; 

 Hunter 1977, in press; Hunter and Kimbrell 1981). 

 Following Hjort's critical period concept concerned 

 with the importance of first-feeding success to lar- 

 val survival (Hjort 1914), most attention has con- 

 centrated on feeding behavior and food consump- 

 tion (e.g., Hunter 1972, in press; Arthur 1976; 

 Lasker 1978). Fish eggs and larvae are also ex- 

 tremely vulnerable to a wide variety of fish and 

 invertebrate predators. A few experimental 

 studies have determined feeding rates for inver- 

 tebrate predators on eggs and larvae in the 

 laboratory (Lillelund and Lasker 1971; Theilacker 

 and Lasker 1974; von Westernhagen and Rosen- 

 thal 1976) and rates of egg cannibalism by fish in 



'Southwest Fisheries Center La Jolla Laboratory, National 

 Marine Fisheries Service, NOAA, La Jolla, Calif; permanent 

 address: The University of Michigan, School of Natural Re- 

 sources, Ann Arbor, MI 48109. 



Manuscript accepted: June 1981. 



FISHERY BULLETIN: VOL. 79, NO. 4, 1981. 



the field (Hunter and Kimbrell 1981). I know of no 

 studies that have examined responses of larvae to 

 a fish predator. 



Northern anchovy, Engraulis mordax, larvae 

 are exposed to a wide range of juveniles and adults 

 of fish and consequently to a range of feeding 

 habits on a continuum from filter-feeding to biting 

 (e.g., Leong and O'Connell 1969). The purpose of 

 the experiment described here was to develop a 

 method to study responses of northern anchovy 

 larvae to attacks by fish, and to determine how 

 avoidance responses and larval vulnerability to 

 capture change during early development. A bit- 

 ing fish planktivore was used as a predator in an 

 attempt to complement studies relating to filter- 

 feeding predators, often crudely simulated by 

 towed plankton nets (e.g., Webb and Corolla 1981). 



METHODS 



Northern anchovy larvae were reared from eggs 



727 



