504 



Fishery Bulletin 88(3), 1990 



of the bigeye tunas' large upward excursions con- 

 sistently terminated near the interface between the 

 thermocline and the mixed layer (Figs. 13, 14, and 15). 

 Examples of possible fly-glide behavior (Weihs 1973, 

 Carey and 01 sen 1982) were observed in tracks of 

 yellowfin and bigeye tuna, a very consistent example 

 being BE8401, which exhibited "sawtooth" oscOlations 

 with a period of approximately 4 minutes and an 

 amplitude between 35 and 45 m for the entire 11-hour 

 daylight segment of the track. Similar, persistent oscil- 

 lating daytime behavior was exhibited by fish YF8506 

 as it moved on a direct course away from S FAD (Fig. 

 18). For instance, between 0930 and 1815 this fish 

 swam on a straight southwesterly course at an ap- 

 parently constant horizontal speed of 1.02 m/second 

 (Fig. 5). Assuming this constant speed, the descending 

 angles during this period averaged 6.06° (±0.7 SD, 

 N = 11), and the climbing angles averaged 9.55° (+1.4 

 SD, N =11) with 100% of the chmbing angles being 

 greater than the descending angles. Also, of the mea- 

 surable oscillations occurring between 0915 and 1815, 

 39 out of 46 (85%) displayed longer falling than rising 

 phases, suggesting active upward swimming followed 

 by a comparatively passive downward glide. These 

 oscillations were characterized by constant rates of as- 

 cent and descent, with abrupt changes in direction link- 

 ing the falling and rising phases (Fig. 18). 



Discussion 



Using two tie wraps appears to be a satisfactory way 

 of attaching transmitters to tunas. Transmitters at- 

 tached in this way were carried successfully by four 

 captive fish for several weeks, and one of the tracked 

 fish was caught in good health (and still carrying the 

 transmitter) by a fisherman 2 weeks after we termi- 

 nated the track. The similarity of vertical and horizon- 

 tal movements across tracks also suggests minimal 

 alteration of normal behavior. 



The association of these tuna with a daytime range, 

 whether FAD or reef perimeter, was extremely strong. 

 In the case of FADs, several fish spent many hours 

 within a few meters of the mooring line. Similarly, none 

 of the reef-associated fish made significant offshore 

 movements during daylight hours. In fact, the along- 

 shore movements were remarkable for their fidelity to 

 the outer reef contour. Combining FAD-associated and 

 reef-perimeter fish, 10 of the 15 fish tracked in this 

 study moved within a well-defined home range during 

 daylight hours and two other fish were lost before any 

 diel patterns could be observed. Most of the fish made 

 nocturnal excursions away from their respective day- 

 time habitats. Similar, consistent diurnal behavior has 

 been previously observed in a 44-cm skipjack tuna. 



ELAPSED TIME (min) 



10 15 :0 25 30 35 40 



SURFACE 

 50 



E 



I 

 (- 

 a. 

 in 

 a 



100 

 150 

 200 

 250 

 300 

 350 



Figure 18 



Yellowfin tuna YF8.506 showed oscillations suggestive of "fly-glide" 

 behavior while moving on an essentially straight course away from 

 S FAD (cf. Fig. .5). All descending angles (a) were smaller than the 

 climbing angles (fi). 



which returned to the same reef after making offshore 

 excursions on each of six nights that it was tracked 

 (Yuen 1970). 



This skipjack tuna and the yellowfin tuna tracked in 

 the present study represent the smaller size classes of 

 these species. For fish of these sizes, the reef dropoff 

 probably represents a zone of enhanced prey density 

 where epipelagic species, such as bigeye scad Decap- 

 tuims crumenopthalmus, mackerel scadD. pinmdatus, 

 and flying fish (Exocidae), can be found in proximity 

 to reef species which are also prey for these timas (Hida 

 1973). The nighttime excursions away from the island 

 reefs and FADs may be foraging behavior targeting 

 on squid and shrimp, which come up from greater 

 daytime depths and which are important components 

 of yellowfin tuna diets (Reintjes and King 1953, King 

 and Ikehara 1956, Brock 1985). Even though the reef- 

 associated fish moved out into deeper water, the night- 

 time distribution of these yellowfin tuna was closer to 

 the surface than their daytime depths. 



The strong daytime association of yellowfin tuna with 

 the FADs, and their nighttime excursions away from 

 them, appears to be analogous to the behavior of island- 

 associated fish patroling the outer reef dropoff. That 

 is, yellowfin tuna appear to respond to FADs as outliers 

 of the natural island topography (e.g., as offshore pin- 

 nacles). However, the FAD-associated fish may be pay- 

 ing an energetics penalty because the forage resource 

 at FADs is probably considerably smaller than that 

 available at the reef perimeter. This is indicated by the 

 fact that yellowfin tuna caught at FAD locations have 

 significantly smaller volumes of food in their guts and 

 a significantly higher frequency of completely empty 

 stomachs than do fish caught away from FADs (Brock 



