Gooding and Magnuson reviewed the hypotheses 

 that have been advanced to explain this habit: (1) 

 attraction by food (smaller fish, algae, decaying 

 palm fronds, and plankton made more visible by the 

 shade of the object); (2) negative phototaxis in 

 response to the shadow cast by the object; (3) shel- 

 ter from predators; and (4) use of the object as a 

 spawning substrate. They also suggested an addi- 

 tional hypothesis that floating objects are cleaning 

 stations where pelagic fishes go to have their para- 

 sites removed by other fish. 



This paper provides information on the ecologv 

 and behavior of fishes associated with floating ob- 

 jects in the offshore waters of Central America. 

 Special attention is given to the frequency, abun- 

 dance, and size of the species which compose flotsam- 

 associated aggregations and how these characteristics 

 are related to the location and size of the object. 

 These studies are the framework upon which future 

 behavior investigations will be based. The aim of 

 our program is to determine whether a de\'ice can be 

 designed tliat will be maximally efficient in aggregat- 

 ing tuna and skipjack. The potential value to the 

 tuna fishery of establishing svich devices has been 

 discussed by .\lverson and W'ilimovsky (19G3). 



PROCEDURES 



Nearly all of our studies were in the offshore 

 waters of Costa Rica (fig. 1) because yellowfin tuna 

 and skipjack are often associated with the flotsam 

 in this region (logbook records obtained thiough the 

 courtesy of the Inter-American Tropical Tvuia Com- 

 mission). Several collections were near the coast 

 of southern Mexico and 1 near Cocos Islantl. Sam- 

 ples were collected by encircling flotsam and its 

 associated fauna with a small %-inch (11 mm.) 

 stretch-mesh purse seine, 12 feet deep (3.7 m.) and 

 1 10 feet (33.5 m.) long (.\asted, MS.)=. An average 

 of 60 percent of the fishes observed beneath an 

 object were captured in the seine. Fish larger than 

 100 mm. standard length may have escaped the net, 

 and fish smaller than 15 mm. occasionally swam 

 through the webbing. When the net was set, fish 

 tended to stay near the flotsam or even swim up- 

 ward. Thus, fish swimming at a depth greater than 

 the maximum depth of the seine also may have 

 been caught. Sampling errors due to fish escaping 

 from or entering the seined cylinder of water were 

 probably small. 



- Donald C". .\astp<l. A iitiniutilro piiran seine for captnrinK stuall pclaci<' 

 fishes. Maniiseript. Bureau of Commercial Fislierles Tuna Resources 

 Laboratory, I.a Jolla, Calif. 



Twenty-three purse seine collections of fishes were 

 made during April, ]\Iay, and June, 1903, and 47 

 during October. Of these samples, 02 were of fishes 

 associated with naturally occurring flotsam, and 8 

 were of fishes collected beneath moored logs, buoys, 

 and other objects. 



After a collection was made, the success of the set 

 was estimated, the object was described and meas- 

 ured, and motile and attached organisms were pre- 

 served. In the October studies, to determine the 

 rate and direction of movement of drifting materials, 

 all objects were tagged and marked with a small flag 

 prior to release. Underwater observations and 

 cinematic photographs were used to describe the 

 behavior and estimate the abundance of fishes. 



CHARACTERISTICS AND 

 DISTRIBUTION OF FLOTSAM 



Far more drifting materials were in the study area 

 in October than in the spring. The Gulf of Xicoya 

 was littered with floating logs and other plant debris. 

 The greater abundance of flotsam in October was not 

 surprising because rainfalls are usually heaviest 

 during this period (Peterson, 1900). 



Fish were not seen beneath the flotsam in the Gulf 

 and were only rarely as.sociated with inshore logs 

 between Cape Blanco and Piedra Blanca (fig. 1). 

 Northwest of this area, however, nearly every drift- 

 ing object encountered had its own associated fish 

 I)opulation. Most often these objects were aggre- 

 gated in areas of current convergence. 



During April, i\Iay, and June, currents in the area 

 usually set northwest at an estimated 2 knots; cur- 

 rents also set northwest during October but were not 

 as strong. Three logs tagged in October and later 

 recovered had drifted northwest at 0.28, 0. 15, and 

 0.33 knot. 



Only one of the drifting objects .sampled had at- 

 tached invertebrates — goose barnacles, Conchodcrma 

 virgnliim (Spengler). This species and other goose 

 barnacles of the genus Lepas were found in (juantity 

 on moored objects after 14 or more days. 



Adult and megalops grapsoid crabs of the genus 

 I'lagiisia were numerous on nearly all logs. In- 

 dividuals in the megalops stage freciuently were 

 swimming near drifting objects. 



SEASONAL VARIATION IN 

 OCCURRENCE OF FISH 



Over 12,000 fishes were captured beneath floating 

 objects in this study; 12 families and 32 species were 



14 



U.S. FISH AND WILDL1F1-: SEUVICE 



