FISHERY BULLETIN: VOL. 74, NO. 3 



0.5 -1 



O.A 



a; 

 O 

 > 



-o 

 o 



D 



"o 

 "o 



XI 



E 



D 

 Z 



0.3 



0.2 



0.1 - o 



• _• 1 larva /500ml 



o — o 5 larvae/ 500ml 



»-* 10 larvae/ 500ml 



-i-a 50 larvae/ 500ml 



 50 larvae (mixed) 



1 — I \ 1 \ 



1 2 ^ 8 16 



Number of Hyperoche/500ml 



Figure 3. - Mean number of yolk-sac larvae of Clupea harengus 

 pallaxi attacked by Hyperoche medusa rum during 1-h exposure 

 time in an experimental volume of 500 ml at different lanal 

 concentrations. Water temperature 9°C. the "mixed" trial was 

 provided with 25 herring and 25 flatfish lanae (11 replicates, 64 h 

 total obser\'ation time). 



reduce the rate of predation (Holling 1966; Brandl 

 and Fernando 1974). 



It became evident through the experiments that 

 predation rate was also influenced by the number 

 of predators present in an experimental beaker 

 (Figure 3). Calculated mean individual predation 

 rates in experiments using 50, 10, and 5 larvae 

 decreased as the number of hyperiids in one 

 container increased. Lillelund (1967) observed the 

 same phenomenon in his experiment using cy- 

 clopids preying on larvae of Osvierus eperlanus, 

 and Salt (1967) noted the same trend in exper- 

 iments using the predatory protozoan Woodruffia 

 metabolica preying on Paramecium. We con- 

 sider this phenomenon an artifact caused by more 

 than one predator feeding on the same prey, an 

 event frequently observed at higher predator 

 densities. This is unlikely to occur in the natural 

 habitat, because a herring larva once killed by its 

 predator which is still attached to it would sink 

 down to the bottom out of the reach of the other 

 Hyperoche. 



0,5 -, 



0.4 - 



SO.3 



5 0.2 



E 



0.1 - 



01 



10 



50 



Number of herring larvae / 500 ml 



Figure 4. -Mean number of yolk-sac larvae of Ctupca hannquK 

 pallasi attacked per hour by one Hyjwrochi' at different larval 

 densities: 



A. data of actual experiments with single hyperiids; 



B. data obtained from mean values for experiments with 1, 2, 4, 8, 

 and 16 hyperiids/50() ml. 



The number of herring larvae attacked did not 

 increase proportionally with an increase of her- 

 ring larvae available for the predators (Figure 4). 

 This phenomenon has been termed "functional 

 response" (type 2 response) by Holling (1966), and 

 is believed to occur commonly in preying inverte- 

 brates. Similar responses are displayed by the 

 house cricket, Acheta dojuei^ticus (Pimentel and 

 Cranston 1960); Podiscus maculive7itri.s (Morris 

 1963); Acanthina sp. (Murdock 1969); Tortanus 

 discaudat^is (Ambler and Frost 1974); and En- 

 phausia pacifica (Theilacker and Lasker 1974). In 

 a typical functional response curve, the number of 

 prey eaten or attacked per predator increases to 

 reach or approach a maximum at an asymptote 

 (Murdoch 1971). Although the curves in Figure 4 do 

 not yet approach an asymptote due to insuflficient 

 prey density, the trend towards a maximum at- 

 tacking rate at a given prey density is noticeable. 



Hyperoche medumrum exposed to two species 

 of fish larvae clearly discriminated disproportion- 

 ately between these two. In Figure 3, the total 

 number of larvae attacked in trials providing 

 alternate prey at equal densities is given as 0.7 

 individuals/h. Of these, 0.055 were herring larvae 

 and 0.015 flatfish larvae. Discrimination between 

 two prey species, which is likely to occur only in 

 predators with searching and food selection 



672 



