836 



Fishery Bulletin 101(4) 



There have been numerous attempts to 

 equate flotsam structure (size, complexity, 

 orientation, etc.) with the number of associ- 

 ated fishes (e.g. Hunter and Mitchell, 1968; 

 Dooley, 1972; Wickham et al., 1973; Wickham 

 and Russell, 1974; Rountree, 1989; Druce and 

 Kingsford, 1995), but the results have been 

 equivocal, except when the analysis was re- 

 stricted to a single species (e.g. Histrio histrio, 

 Dooley, 1972;Decapterus punctatus, Rountree, 

 1989). Huge aggregations have been associ- 

 ated with very small objects — lATTC (Inter- 

 American Tropical Tuna Commission ) records 

 include a report of 55 metric tons of mostly 

 yellowfin tuna {Thiinnus albacares) fished 

 from beneath a 1-m length of floating poly- 

 propylene rope (Hall et al., 1999b); therefore, 

 despite the intuitive appeal, there is no clear 

 reason to expect that size of FAD per se is an 

 important factor in determining the size of 

 associated assemblages. Thus, object size re- 

 mains an unresolved problem in understand- 

 ing flotsam-associated communities. If there 

 are optimal FAD sizes, these may be species 

 specific, and economical FAD design depends 

 upon controlled experiments in the field. 



Fouling organisms (sessile invertebrates 

 and algae that colonize flotsam) are believed 

 to have a strong, positive effect on the sub- 

 sequent recruitment and retention of fishes 

 by commercial fishermen (Gaertmer and 

 Medina-Gaertner, 1999; Hall et al,, 1999a; 

 Hallier and Parajua, 1999; Suzuki, 1999). 

 However, prior to the results presented here, 

 there appear to have been no controlled tests 

 of the hypothesis that the presence of fouling 

 organisms enhances fish recruitment to a 

 floating object. I also compared the numbers and diversity 

 of fishes associated with FADs that are equipped with ar- 

 tificial (lead weight) fish versus FADs without these artifi- 

 cial fish. The latter experiment was intended to determine 

 the importance of prior recruits to subsequent patterns of 

 recruitment. To test a similar hypothesis over the short 

 term (hours versus days) and using living fish instead 

 of painted models, I also compared recruitment to FADs 

 enriched with real fish (juvenile Abiidefduf troschelii) to 

 unenriched FADs. 



I tested the hypothesis that each of these factors would 

 affect the number of fishes associated with FADs (combined 

 and individual species), as well as the species diversity of 

 FAD-associated fish assemblages. Both the size of these 

 FAD-associated fish assemblages and their species diver- 

 sity provide insight on recruitment processes and the use of 

 floating objects by fishes. Although the association of fishes 

 with floating objects has been well documented, very little 

 is known regarding the behavioral and ecological processes 

 behind these assemblages. The results reported in the pres- 

 ent study provide new information on the role of flotsam 

 and FAD characteristics in determining the number and 

 diversity of these assemblages, and some clues towards 



Figure 1 



Location of the study site at Achotines, Panama, Central America. 



understanding why and how fishes aggregate beneath 

 floating objects. 



Materials and methods 



Study site and FAD construction 



All research was conducted between July and October 1997 

 on the Pacific coast of Panama, Central America, from the 

 Inter-American Tropical Tuna Commission laboratory at 

 Achotines, near the tip of the Azuero Peninsula (Fig. 1). 

 Experimental FADs were constructed of three tuna purse- 

 seine buoys lashed together and anchored to the substrate 

 with a 25-kg cast concrete block unless otherwise noted 

 (Fig. 2). The length of the anchor lines allowed the FADs 

 to rest at the surface at all tidal heights. Each buoy was 

 roughly 25 cm in cross sectional diameter, and approxi- 

 mately 35 cm in length. The FADs were detachable from 

 their moorings by detaching a large (2 m diameter) loop on 

 the anchor line that held a 2-kg line weight (Fig. 2). This 

 design allowed me to change FADs for another treatment. 

 The FAD arrays were deployed nearshore (within 1.5 km; 



