FISHERY BULLETIN: VOL. 72. NO 1 



pelagic bait fish are capable of ranging beyond 

 sight of an object for periods of several minutes 

 or longer, they apparently require periodic 

 visual reconfirmation of the object's position in 

 order to maintain their orientation with it. 

 This assumption is supported by our observa- 

 tions that structure-attracted fish aggregations 

 leave the structures at night when low light 

 levels inhibit visual contact. Our observations 

 of coastal pelagic bait-fish behavior around 

 artificial structures also indicate that the struc- 

 tures can apparently be useful to these species 

 for predator avoidance. Schools of bait fish 

 associated with an artificial structure have been 

 observed to be immediately attacked by j^reda- 

 tors upon removal of the structure from the 

 water. Bait-fish schools threatened by the pres- 

 ence of feeding predators were observed to form 

 a tight ball or ring around the stinicture or swirl 

 in tightly packed formation making quick dart- 

 ing passes near the structure. On several occa- 

 sions, we have observed the attack behavior of 

 a predator to be interrupted at the moment the 

 bait fish darted past the structure. Mitchell and 

 Hunter (1970) describe laboratory experiments 

 in which splitnose rockfish (Sebastes diploproa) 

 and opaleye (Girella idgriccuin) were pursued 

 more often, for longer periods, and captured 

 more frequently by ocean whitefish (Caulolatilus 

 princeps) in an aquarium when kelp was absent 

 than when it was present. 



Our ))resent supposition as to the possible 

 mechanisms involved in the association of some 

 species of coastal pelagic schooling fish with ob- 

 jects in the sea are summarized in the following 

 working hypothesis: "Objects in the sea i)rovide 

 visual stimuli which attract certain species of 

 pelagic schooling fish and are used in conjunc- 

 tion with natural optomotor responses to provide 

 a spatial reference for orientation in the other- 

 wise relatively unstructured pelagic environ- 

 ment; however, in the presence of feeding preda- 

 tors stimulus priorities are restructured such 

 that the objects become useful for predator 

 avoidance." An increasing body of subjective evi- 

 dence is available to support much of this con- 

 jecture, but its verification lacks the requisite 

 quantitative experimental evidence. 



Purse Seine Operations 



The feasibility of harvesting structure-attract- 

 ed coastal pelagic bait fish with conventional 



tom-weight type purse seines was evaluated 

 during our development of the quantitative col- 

 lection procedures. Fish aggregations normally 

 showed little disturbance during purse seine sets 

 while the structure remained in the water. Fish 

 stayed with the structure even when it floated 

 at the surface after the counterweight was lifted 

 to .prevent its tangling the purse line. The fish 

 showed distress and attempted to escape the net 

 only when the structure was removed from the 

 water and the diving boat prepared to pass over 

 the corkline and reset the structure. The only 

 deviation from this pattern was observed when 

 bait fish were attacked by predators, i.e., little 

 tunny (E. alletteratus), which on several occa- 

 sions were following the seiner. On these occa- 

 sions, the predators scattered the bait fish during 

 the set and then escaped before the net was com- 

 pletely closed. 



During our experimental collections, we 

 utilized an additional small boat and several 

 men to handle the structures during the purse 

 seine operations. Sets have been made, however, 

 using only the seine skiff and its operator to 

 retrieve and reset the structures. These trials 

 indicate that in a commercial fishing operation 

 using artificial structures, fishing procedures 

 can be modified so that additional men and 

 equipment should not be required. The applica- 

 bility of structure-attraction techniques for aug- 

 menting purse seining during commercial fish- 

 ing operations, although technically feasible, 

 remains dependent upon the production poten- 

 tial of structures and their recruitment charac- 

 teristics in the geographical area under consid- 

 eration. 



SUMMARY AND CONCLUSIONS 



An evaluation of our diver estimates and purse 

 seine catch data indicates that a combination of 

 these techniques provides a more complete de- 

 scription of the artificial structure experimental 

 environment than either singularly. Our com- 

 parative results support the contention by Klima 

 and Wickham (1971) that quantitative diver 

 estimates tend to be conservative where large 

 fish schools are involved. Our divers were able 

 to qualitatively determine the major species 

 present at a structure, but were unable to reli- 

 ably establish the percent species composition 

 in mixed species schools. 



190 



