ECOLOGICAL INTERPRETATION 



Fishes were recruited rapidly to moored objects. 

 The number beneath objects moored o days and the 

 number beneath those moored 120 or more days did 

 not differ. Goose barnacles were attached to all four 

 objects moored 14 days or longer, but they were 

 found on only one drifting object. Thus, all but 

 one of the drifting objects probably had been at sea 

 not longer than 14 days. Rapid recruitment appears 

 to be characteristic of the formation of flotsam- 

 associated fish aggregations. Hardenberg (1950) 

 reported that Indonesian fishermen harvest the fishes 

 beneath their palm frond rafts at intervals of several 

 days. Gooding and Magnuson (see footnote 1) 

 stated that fishes were recruited to their raft minutes 

 after it was set adrift. 



Recruitment of fishes followed no particular 

 se(iuence. Small collections containing only a few 

 fish were not necessarily all of one species. The 

 species composition and order of recruitment of 

 fishes to two balsa logs moored 100 m. apart were 

 dissimilar. 



The same species dominated our collections in both 

 fall and spring. Similarities were marked between 

 tlie families and genera represented in our study area 

 and those reported from other areas (Uchida and 

 Shojima, 1958; Besednov, 19G0; Kojima, 1960a; 

 Mansucti, 19G3). Juvenile carangids were by far the 

 most important family in terms of abundance, num- 

 ber of species, and dominance. They were also by far 

 the most freciuently reported flotsam associate from 

 other areas. Other families of fishes commonly 

 encountered in this study and freciuently reported by 

 otlier authors included the Scombridae, Balistidae, 

 Kyphosidae, and Stromateidae. 



The presence of attached invertebrates on floating 

 objects appeared not to influence the occurrence of 

 fish species. The only drifting object that had goose 

 barnacles attached did not have a species composi- 

 tion different from that of objects without the 

 barnacles. Although barnacles were present on each 

 of four objects moored 14 or more days, the species 

 composition of the aggregations of fish did not differ 

 from those of objects moored much shorter periods. 

 Many of the flotsam-associated fishes were scliool- 

 ing species, ^\'c believe that the habit of schooling 

 and of association with drifting materials may be 

 related. Two scombrids, Katsinvonus pelamis and 

 Thunnits albacans, not only associate with inert 

 materials but also with large sharks and whales, and 



T. albacares is a common associate of porpoise schools 

 (Uda and Tsukushi, 1934; and unpublished logbook 

 records of the Inter-.\merican Tropical Tuna Com- 

 mission). The carangid, N'aucrates duclor, known for 

 its a.s.sociation with sharks (Dales, 1957), also was 

 found beneath flotsam in this and in other studies 

 (Murray and Hjort, 1912; Galea, 19C.n. Many 

 fishes school at times with fish other than their own 

 species. Katmnoonus ■pclamiH and Thunnus albacares 

 school together (Orange, Schaefer,and Larmie, 1957), 

 and juvenile Selar crumenophlhalmus and Dccaplcrus 

 sp. were observed schooling together in this study. 

 Trachurus sijmmetricus (Ayres), a carangid associate 

 of flotsam in southern California waters, schools with 

 Scomber japoniciis Houttuyn and Sardinops sagax 

 (Jenyns)'' and associates with jellyfish (Limbaugh, 

 1955). The tendencies of fishes to associate witii 

 living animals other than their own species and to 

 associate with inert drifting materials may be related. 

 Atz (1953) suggested, among other possibilities that 

 an aggregating companion for a schooling fish coifld 

 represent merely "a simple point of reference for 

 optical fixation". Flotsam could function in this 

 capacity for schooling fishes. 



A schooling mechanism cannot explain the pres- 

 ence of all the associated species. Some fishes did 

 not school, and for others alternative mechanisms 

 are etiually plausible. Mechanisms postulated by 

 other authors were: attraction to food, negative 

 phototactic response to the shadow of the object, 

 shelter from predators, presence of spawning sub- 

 strate, and parasite-cleaning symbiosis. 



Owing to the infreciuent occurrence of flotsam- 

 associated fishes in the stomachs of predators tlic 

 food hyjjothesis probably can be eliminated for 

 predacious adults. 



For juveniles and nonpiscivorous adult fishes the 

 food hypothesis would not apply, as the drifting 

 materials were usually devoid of attached inverte- 

 brates and algae that would provide food. That 

 fishes were attracted because plankton was more 

 visible in the sliadow cast by the object also seems 

 unlikely, because the fishes did not remain in the 

 shadow. Furthermore, plywood sheets that cast 

 large shadows were less effective in attracting fish 

 than were objects that produced smaller shadows. 



That all the juvenile fishes and adult Canlhidermis 

 7naculatuK and Scetnlor ocyurus swam toward and 

 beneath the object when predators appeared, sug- 



> Unpublished data. Bureau of Conunercial Fisheries. I.a JoUa. California. 



26 



U.S. FISH AND WILDLIFE SERVICK 



