5 



Channel 4 



Lll.llll|jJ.llilflJlLlll|lllllllll f lb.iiLv^ ^ 



Channel 5 



lll.li.l l. hill. ).  ..   



u ~i i r i i 1 1 1 1 



20 40 60 80 



NUMBER OF BOATWHISTLES / 6-MIN UTE INTERVAL 



FIGURE 2. — Histogram of frequency of occurrence (i.e., 

 "number" on Y-axis) of number of boatwhistles in 6-min inter- 

 vals (X-axis) for each of three toadfish. Silent periods of an hour 

 or longer were excluded from the analysis. 



many hours (Figure 1), the number of calls fluc- 

 tuated markedly. High rates of calling were often 

 strongly peaked, i.e., not maintained for long 

 periods. 



Discussion 



The only obvious feature of the data from this 

 study (Figure 1; Table 1) is its lack of patterning or 

 predictability. Clearly, the recordings indicate no 

 diel cycle. While they do not rule out the possibil- 

 ity of maximal or minimal periods of sound pro- 

 duction for a toadfish population (Breder 1968), it 

 appears unlikely that individuals would be syn- 

 chronized to any great degree. It is difficult to 

 reconcile these results with the periodicity of the 

 in-air respiration data of Schwartz and Robinson 

 (1963) and the impressions of Tavolga (1960) and 

 Schwartz and Robinson ( 1963) that the toadfish is 

 basically nocturnal. Squirrelfishes are active at 

 night, when they are least vocal (Winn et al. 1964; 

 Salmon 1967; Bright 1972; Bright and Sartori 

 1972), and likewise toadfish might not have a clear 

 vocalization rhythm, while maintaining rhythms 

 for respiration or other functions. 



The rate of calling by fish in this study was low. 



20 40 60 80 



NUMBER OF BOATWHISTLES /6-MINUTE INTERVAL 



FIGURE 3. — Histogram of frequency of occurrence (left axis) and 

 cumulative frequency of occurrence (right axis) of number of 

 boatwhistles in 6-min intervals combined for the three toadfish. 

 Silent periods of an hour or longer were excluded from the 

 analysis. 



and individuals lapsed into silence for long 

 periods. This result verifies our experience from 

 playback studies (Winn 1967, 1972; Fish 1972; 

 Fish and Offutt 1972); fish were often silent, forc- 

 ing us to sample many tiles to find a male calling 

 rapidly enough for use in an experiment. For this 

 reason preplayback calling rates, equivalent to 

 control calling rates, were biased upward. From 68 

 experiments, each with sample sizes ranging be- 

 tween 11 and 16, Winn's (1972) preplayback data 

 (recalculated) show a mean of 22.41 ± 4.3 (1 SD) 

 boatwhistles/3 min, or an average of 7.5 calls/min. 

 In his initial playback experiments, Winn (1967) 

 increased the calling rate to an average of 11.46, 

 11.70, and 11.48 boatwhistles/min by playbacks of 

 18, 26, and 36 boatwhistles/min. Playbacks of 10 

 calls/min did not increase calling. Fish (1972) 

 found that with optimally spaced playbacks, he 

 could increase their rate to 14 to 16 sounds/min ( 1 

 call every 3.7 to 4.3 s). He called this pace the 

 maximum sustained calling rate. Fish's data com- 

 bined with Winn's indicate that when competing 

 with other males, the toadfish does not grade his 

 output uniformly, but follows more of a step func- 

 tion, i.e., his calling is either facilitated or not. In 

 one chance encounter Fish ( 1972) observed a male 

 calling 25 times/min as a female approached his 

 shelter. 



Our fish called considerably below their 

 capabilities. However, calling rates of 11 and 

 12/min would suggest that the males were sexu- 

 ally receptive. It will take more work to establish 

 what is normal for the toadfish and what abiotic 



873 



