FISHERY BULLETIN: VOL. 74, NO. 3 



amphipods) was used for each exposure period. 

 Mortality of larvae was measured every 2, 4, and 8 

 h by means of direct counts. All remaining larvae 

 were removed, and healthy, wounded, and dead 

 larvae were counted. The original number of 

 herring larvae then was restored before a new 

 experiment was started. Between experiments, 

 the hyperiids were provided with food in order to 

 reduce cannibalism. 



RESULTS 



Swimming and Feeding Behavior of 

 Hyperoche medusarum 



Two modes of swimming were observed: 1) 

 quick darting movements with the body kept in a 

 horizontal position; and 2) slow hovering, in which 

 the body was held in a vertical position, and the 

 pleopods beat continuously. The latter mode of 

 swimming was maintained for periods longer than 

 20 min, but the speed of swimming was slow 

 (about 10 cm/min). It was only during swimming 

 that Hyperoche would, by chance encounter, cap- 

 ture a herring larva. The amphipod usually 

 grasped the tail but attacks at the head and the 

 midportion of the larva also occurred. An attacked 

 larva did not survive long. The larva attempted to 

 shake the amphipod off for a few minutes, then 

 sank to the bottom where it was eaten by the 

 Hyperoche. Larvae were not always consumed. 

 Frequently, amphipods clung to a larva for only a 

 few seconds but the wound inflicted during this 

 process inevitably lead to the death of the larva. 

 Wounded larvae which were removed after ter- 

 mination of the experiment never survived for 

 more than 4-5 h when kept in separate beakers. 



Between swimming activities, the amphipods 

 either remained on the bottom (probably an ar- 

 tifact due to the small size of the beakers-in large 

 enough containers Hyperoche juveniles swam 

 continuously (Westernhagen 1976)), or attached 

 themselves with the posterior pereiopods to the 

 nylon gauze provided in the beakers for this 

 purpose and assumed a resting posture. This 

 posture has been described for Hyperia galba by 

 Bowman et al. (1963) and for Hyperoche ynedusar- 

 iim by Evans and Sheader (1972). The latter 

 authors defined the posture as an "inactive curled 

 position head and urus directed away from the 

 substrate it (the animal) sits on." Larvae that 

 bumped into resting amphipods were not pursued 

 or captured. 



Predatory Efficiency of 

 Hyperoche medusarum 



The results of all experiments were summarized 

 and presented as the number of larvae attacked 

 per hour at different predator and prey densities 

 (Figure 1). The number of wounded and killed 

 larvae was dependent on two factors, the density 

 of the herring larvae and the density of hyperiids. 

 With increasing numbers (predator or prey) larval 

 mortality per hour increased, reaching a value of 

 more than two larvae killed or wounded per hour at 

 the 16 Hyperoche and 50 herring larvae 

 combination. 



The number of larvae attacked per unit time (1 

 h) depended to a great extend on the duration of 

 the experiment (Figure 2). Experiments with 

 short exposure times (2 h) yielded for all larvae 

 and hyperiid combinations higher attack rates per 

 hour than experiments lasting 4 or 8 h. The mean 

 predatory efficiency of the hyperiids was affected 

 also by their density in each beaker. The number 

 of larvae attacked per unit time decreased as the 

 density of the predators increased (Figure 3). It is 

 for this reason that there are different values for 

 the number of herring larvae attacked per hour by 

 one hyperiid (Figure 4), (A) for the observation of 

 one single hyperiid, and (B) for the calculated 

 mean predation rate of a hyperiid from exper- 

 iments with 1, 2, 4, 8, and 16 Hyperoche. Yet both 

 curves show that an increase of a potential prey in 

 a constant environment beyond a certain density 



01 5 10 



Number of h#rr.ng larvae / 500 ml 



Figure 1. -Predatory efficiency of Hyperoche medusarum on 

 yolk-sac larvae of Clupea harengus pallasi at different predator 

 and prey densities. Water temperature: 9°C; total observation 

 time: 111 h; observation periods: 20. 



670 



