CHANNEL S 



I I I I I I I I I I I 1 I I I I I I | I I I I I I 



JUNE II JUNE It 



I I I I I I I I I I I I I I I 

 JUNE 13 



— Mt 



1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 " 



JUNE 13 JUNE 14 



I I I I I I I I I I I I I I I I I I I | ^^^^^^ 

 JUNEI3 JUNE M 



'"Uw 



1111 I I I I I 



JUNE H) 



J\H« 



J 



I I I I I I I 1 T I t I T I T I I I | n-^^^t^ 1 



LA 



«0 - 



to - 



, , , A^^MW unrn 



 ■■MM 



JUNE IS 



I I I I I I I I I I I I I TTTTTIITTlfT 



I I I I I I I I I I I 

 JUNE 14 



J«fcUL 



JUNE 19 



a to t4 4 



I i i TfvFAfA t r i . i i t r t i t ir l 



It 10 H 4 II 



FIGURE 1. — Temporal record of boatwhistle production for each of three toadfish. A missing baseline indicates gaps in the record, and 



the horizontal line below the baseline indicates the period of darkness. 



TABLE 1. — Number of boatwhistles produced during 24-h 

 periods by three toadfish. 

 [L is light, D is dark, and dash indicates no data.] 



minor feature of the record. Each of the fish pro- 

 duced different numbers of boatwhistles and exhi- 

 bited separate patterns of calling (Figures 1, 2; 

 Table 1) that were not obviously correlated with 

 each other. One fish (channel 4) boatwhistled 

 twice as much during the day as at night, while the 

 other two (channel 2 and 5, respectively) called 1.5 

 and 2.7 times more at night than during the day. 

 These ratios from Table 1 change to 2.24, 0.66, and 

 3.92, respectively when considered on a per-hour 

 basis. 



In order to see how fast individual fish called, we 

 constructed histograms of the frequency of occur- 



872 



rence of number of boatwhistles in the 6-min seg- 

 ments (Figure 2). Even though the distributions 

 for day and night were statistically different 

 (Kolmogorov-Smirnoff test), they were combined 

 in each of these histograms. Since these day-night 

 differences have already been mentioned and were 

 inconsistent between fish (Table 1), it seemed 

 reasonable to present differences between the fish 

 rather than differences between day and night. 



Data from the three channels were combined to 

 show the calling rate from all boatwhistles re- 

 corded in this study (Figure 3). It is obvious that 

 toadfish remain quiet for long periods (Figure 1). 

 For Figures 2 and 3, all quiet periods of 60 min or 

 longer were arbitrarily excluded. Still, zeros ac- 

 counted for close to 207c of all intervals measured 

 (Figure 3). From the cumulative percent curve 

 (Figure 3), it is striking how strongly the distribu- 

 tion is skewed toward the low end. Over 50% of the 

 intervals measured had «= 1 to 2 boatwhistles/min 

 (ca. 10 calls/6 min), and over 75% of the intervals 

 had =£4 to 5 boatwhistles/min. Only 10% of the 

 intervals contained calls emitted at a rate of 6 or 

 more per minute. Finally less than 17c of the in- 

 tervals contained calls emitted at a rate of 10 to 

 12/min. Although an animal may have called for 



