346 



Fishery Bulletin 89(3), 1991 



Table 1 



Characteristics of yellowfin tuna sets examined for other-species associations according to log-fish, school-fish, and porpoise-fish 

 set types. 



'The Successful Set (SS) Ratio = fraction of total sets catching more than 1 ton of yellowfin and/or skipjack tuna. On porpoise-fish 



sets, the SS Ratio pertains to sets on Stenella dolphins only. 

 2 Geometric mean short tons of yellowfin tuna (YF) per set, based on successful sets only; n is the size of the sample. Not all successful 



sets caught yellowfin. 



3 The multiplication-division factor for mean YF tons/set giving its 95% confidence interval. 

 4 Observer T.M. Duffy (fleet data exclude these data). 



with tuna, as seen in this study, will be discussed in 

 terms of the ontogeny of yellowfin behavior. 



Results 



The fractions (p) of purse seine sets measuring the 

 association rates of different (grouped) species are 

 given in Table 2 (off Central America) and Table 3 (off 

 Mexico). These fractions (expressed as percentages) are 

 for the fleet sets, and they are summarized graphical- 

 ly in Figure 2. Figure 3 shows the association fractions 

 for the T.M. Duffy (TMD) sets. In the tables, those 

 species whose association rates differed significantly 

 (x 2 , P<0.05) among set types are indicated by aster- 

 isks. For these, one can determine which of the set 

 types was actually different by noting which of the 

 mean rates occurred outside the 95% confidence inter- 

 vals for the other set types. 



Sharks clearly stood out among the fishes found 

 associated with tuna. Overall, 40% of log-fish tuna sets 

 took sharks. There was a progressive decline in this 

 percentage to 10%, from log- to school- to porpoise- 

 fish sets. Off Mexico, however, porpoise-fish sets may 

 have been more likely to catch sharks (lower graphs 

 in Figures 2 and 3). How commonly sharks may asso- 

 ciate with tuna is emphasized by the TMD data (Fig. 

 3); individual fishing trips, of which this data from a 

 single observer are an example, can experience shark 

 occurrence rates of up to 90% on log-fish sets. The 



silky shark Carcharhinus falciformis was the most 

 commonly reported species (25.2%) (Table 4). These 

 sharks averaged 29.2 individuals in sets that caught 

 them. However the range was large, from 1 to 500 per 

 set. Other carcharhinids recorded were each under 10% 

 of the identified sharks. Of these, the whitetip shark, 

 frequently recorded with porpoise-fish sets (especial- 

 ly off Mexico), was most likely the oceanic whitetip 

 C. longimanus, (R. Rasmussen, SWFSC, pers. com- 

 mun.). Other less-common sharks included the whale 

 shark Rhincodon typus. This species can be locally com- 

 mon off Baja California, where it is often associated 

 with skipjack tuna. Sizes of the sharks caught with 

 yellowfin were seldom recorded; however, it is known 

 that medium- to large-sized specimens that are con- 

 sidered dangerous are often caught. The few sizes 

 recorded ranged from 1.7 to 2.1 m. Many were simply 

 described as "large." Others, described as "small" 

 or comprising "25 tons" of catch, were probably all 

 small-sized. 



Rays occurred mainly in school-fish and porpoise- 

 fish sets, the former more often. The records indicate 

 that most were medium- to large-sized manta rays 

 (Mobulidae). They occurred in groups of 1 to 2 on 

 average, though up to 12 were recorded in some sets 

 (Table 4). 



Billfish (Istiophoridae) co-occurred with tuna in about 

 9% of the sets overall and more often off Central 

 America than off Mexico. The association rates were 

 similar among all set types, except for the TMD data 



