NOTE Zeller: Ultrasonic telemetry: its application to coral reef fisheries research 



1063 



tion generally produced the largest error poly- 

 gon area and length estimates, regardless 

 of distance between sound source and track- 

 ing boat (Fig. 4). The 45°-90° transect com- 

 bination produced similar results, with the 

 exception of 50 m distance, where this com- 

 bination resulted in the smallest polygon 

 area and length (Fig. 4). Overall, however, 

 the 0°-90"" transect combination consis- 

 tently resulted in the smallest polygon pa- 

 rameter estimates, irrespective of distance 

 from which bearings were talien ( Fig. 4 ). The 

 best results for this angle combination were 

 obtained at distances of 50 to 75 m from the 

 sound source, with maximum-error polygon 

 diameters of 58.5 m and 34.5 m, respectively. 



Field tracking trials 



The two fish tagged with external transmit- 

 ters did not travel far. The larger specimen 

 traveled less than 25 m, whereas the smaller fish 

 moved a maximum of 60 m. Both fish spent con- 

 siderable time inside the shelter of the reef ma- 

 trix, with the result that a low acoustic signal 

 was received, accompanied by reduced reception 

 angles. This behavior was confirmed during vi- 

 sual observations. The tracking trial was termi- 

 nated after four days, and both fish were col- 

 lected. Both specimens exhibited aggravated ex- 

 ternal wounds. Repeated dislodgment attempts, 

 or snagging of the external transmitter on reef 

 substratum had resulted in enlarged external 

 wounds and loosened transmitters. 



The two fish with surgically implanted trans- 

 mitters took shelter inside the reef matrix im- 

 mediately after release (observed on snorkel) and 

 were not resighted. The signal weakened once 

 the specimens had taken shelter, and remained 

 weak for 48 hours. By the third day both signals 

 were received strong and clear in the same loca- 

 tion. Visual confirmation indicated that both 

 transmitters were inside a whitetip reef shark 

 (Triaenodon obesus). The immediate, postsurgery 

 release off! leopardus had resulted in mortality 

 due to predation within 48-72 hours after release. 



Conclusions and recommendations 



Transmitter placement and fish anesthetics 



The aquarium experiments illustrated that, al- 

 though surgical insertion is technically the most 

 difficult method, it provided the least side effects 



9 



-- 7 



UJ 



« 5 



o 



S 1 

 ai 



■i -1 

 m 



§ -5 



IS 



5 -7 

 -9 



4.62 



 Observer 1 

 n Observer 2 



-0.43 



0° 



-5.24 .562 

 45° 



w 



-2.71 



-0.29 



90° 



Angular transect 

 Figure 3 



Mean bearing errors i±SE, n=42) for each observer on each angular 

 transect, illustrating the difference in observed bias between the two 

 observers for the transect facing directly into the prevailing wind (i.e. 

 0"). Mean values are indicated. 



25,000 



0-90 transect 



0-45 transect 



45-90 transect 



600 



- 500 



50 



B 



75 



.Id 



100 



iIiiJIlJl 



125 



150 



175 



200 



400 



300 



200 



100 



0-90 transect 



0-45 transect 



45-90 transect 



jjojiJiM 



I 



50 



75 



100 



125 



150 



175 



200 



Distance (m) 

 Figure 4 



Maximum-error polygon parameters for each identical distance 

 pair for the three possible angular transect combinations. Er- 

 ror polygons were calculated by using the largest and smallest 

 angular bearings recorded from equidistant positions for each 

 of the three possible angle combinations during the evaluation 

 trials. Data pooled for observers. (Al Maximum error polygon 

 area (m-l. (B) Maximum diameter of error polygon (m). 



