to that described by ]\Iagniison and Prescott' for the 

 reproductive behavior of Pacific bonito, Sarda chil- 

 iensis (Cuvier). 



Nearly all the advilt male and female Canthidermis 

 macidatus captured in the fall were ripe. On one 

 occasion these fish showed what may have been 

 courtship behavior, but no spawning was observed. 



TRANSFER OF FISHES TO 

 OTHER OBJECTS AND HOMING 



Some species were attracted to the skiff when it 

 drifted alongside a floating object. Only the fishes 

 that occupied upper positions in the aggregation, 

 such as juvenile Caranx cabaUns, Pscnes pacificus, 

 Elagatis bipinnulahts, Kyphosus elegans, and Sectator 

 ocyurus, showed this behavior. The more deeply 

 positioned species, Selar crumenophthalmns. Decap- 

 terus sp., Pseudupeneus grandisquamis, and Chromis 

 atrilobata did not transfer to the skiff. Those species 

 most closely associated with the surface of the object, 

 Abudcfduf saxatilis and Blaiyriohis breripinnis. did 

 not transfer unless the original object was removed 

 from the water. Transfers to the skiff were only 

 temporary. The fishes swam beneath the skiff, 

 remained there a few minutes, and then returned to 

 the original object. Movements back and forth 

 from the object to the skiff lasted no longer than 

 30 minutes; thereafter, the fish remained beneath 

 the original object. 



Two attempts were made to transfer the fish 

 population of a log to a 4- by 8-foot (122 cm. by 

 243 cm.), 14-inch (T) mm.) thick plywood sheet. A 

 log with a fish population was attached to the ply- 

 wood sheet 24 hours; then the two objects were 

 separated. During daylight, underwater observa- 

 tions were made while the two objects were attached 

 and after they were parted. At no time did the 

 fishes congregate beneath the plywood sheet. They 

 remained beneath the original object during the 24 

 hours the objects were attached and after they were 

 separated. The experiment was repeated ; this time, 

 60-cm. sections of unraveled 3 2-inch (13 mm.) manila 

 line were attached at 10-cm. intervals in three rows 

 to the underside of the plywood. The rope pro- 

 duced a dense mass of filaments. After 2^ 2 hours 

 none of the fishes had transferred from the original 

 log to the plywood, but 1 hour after the plywood was 



s John J. Magnuson and John H. Prescott— Courtship, locomotion, feed- 

 ing, and miscellaneous beliavior of Pacific bonito (Sarda chiliensis). Man- 

 uscript. Bureau of Commercial Fisheries Biological Laboratory. Honolulu, 

 Hawaii. 



freed from the original object, over 100 C. caballus 

 had been recruited to the plywood. 



The failure of fishes to form a permanent associa- 

 tion with new objects, a skiff or plywood sheet, when 

 already associated with another object suggests that 

 a more familiar object may have a higher valence. 



Ten adult Canthidermis inaculaius were captured, 

 tagged, and released separately; four were released 

 7.5 m. from their original log, four at 15 m. and two 

 at 30.5 m. One hour and 30 minutes later, three ojt 

 those released at 7.5 m. and three released at 15 m. 

 had returned. Neither of the fish released at 30.5 m. 

 returned. Conceivably the fish planted at the great- 

 est distance could not see the log. The recapture of 

 fish planted at lesser distances suggested that they 

 may return to their log when it is within visual range. 



RUBBING BEHAVIOR 



Adults of Coryphaena Mppurus, Canthidermis mac- 

 idatus, and Sectator ocyurus frequently rubbed their 

 dorsal surface or sides against the logs and the skiff. 

 An entire school of adult Sectator showed this 

 behavior. 



SEA SNAKES 



^^'e frequently observed the sea snake, Pelamis 

 platurus, swimming near the surface. Often a small 

 school of fish of the genus Polydactylus was below a 

 snake. On three occasions the snake was feeding. 

 It began to swim backwards; the schooled fish 

 reversed direction and began swimming with their 

 heads oriented toward the tail of the snake. The 

 snake then captured fish from the school by a rapid 

 thrust of the head and anterior portion of the body, 

 directed either to the side and posteriorly or down- 

 ward and posteriorly. 



Klauber (1935) also observed P. platurus feed on 

 fish schooled beneath it. 



Klawe (1964) examined the stomachs of 56 P. 

 platurus from the eastern tropical Pacific. In the 

 22 which contained food, Polydactylus approximans 

 was the most abundant food, Pseudupeneus grandis- 

 quamis was second, and Mugilidae third. One indi- 

 vidual each of Selar crumenophthalmns, Caranx hip- 

 pos, and Fistularia corneta also were found. Except 

 for F. corneta we captured all of these species beneath 

 flotsam, and there was no difference in size between 

 the fishes we collected and those in tlie stomachs of 

 sea snakes. Apparently P. platurus takes advantage 

 of the habit of some species to congregate beneath 

 flotsam. 



FLOTSAM IN OFFSHORE WATERS 



25 



