FISHERY BULLETIN: VOL, 77, NO, 4 



Figure 6. — Average speeds with which dusky dolphins traveled 

 at different depths. The least squares regression, fit to the means 

 shown, is statistically significant (P<0.01). Numbers over bars 

 represent number of observations in that category. 



10 12 



TIME (h) 



Figure 7. — Mean depth of water (a) and mean swimming speed 

 (h) of dusky dolphins as a function of time of day. Bars represent 

 95% confidence intervals for means and numbers above bars 

 represent the number of theodolite readings per hour interval. 



50 

 40 



O- 



W 20 



10- 



72 IJOpL 



OVERALL MEAN 

 DEPTH = 27 8 m 



ir 



35 525 1368 



n — I — I I ' I I ' I ' 1 ' 1 — r r I — 11' I ' I ' I ' I ' r 1 



J ASOND,JFMAM JJA SO NDJ FM, 



1974 



1975 



1976 



n — nr 



N D 



OVERALL MEAN 

 SPEED = 7 7km/h 



* 



36 



n 



, T — I — I 1 — I T 



JFMAMJ JASONDJFM 



1974 



875 



1976 



Figure 8. — Mean depth of water inhabited by dusky dolphins 

 for different months (a), and mean speed of travel for dolphins for 

 different months (b). Numbers represent number of theodolite 

 readings obtained per month; bars represent 95% confidence 

 intervals for means. 



tively shallow water compared with the following 

 months (Mann-Whitney [/-test, P<0.001). 



Dolphins usually moved more rapidly during 

 afternoon than morning (Figure 9b). The increase 

 in rapid movement per month appears related to 

 the amount of surface feeding bouts in that month. 

 Thus, in August 1974 and 1975 few surface feed- 

 ing bouts occurred, and there was no increase in 

 speed during the day. In September, some feeding 

 took place, and there was a small speed increase. 

 In October, November, December, and January 

 much surface feeding took place during one or both 

 years, and the afternoon speed increase was most 

 dramatic. In February, both surface feeding and 

 afternoon speeds were again down to pre-October 

 levels (August, September, and February after- 

 noon speeds are significantly different from Oc- 

 tober, November, December, and January after- 

 noon speeds, Mann-Whitney t/ -test, P<0. 001). 



From these data we concluded that dolphins 

 traveled faster at surface feeding times. This was 

 confirmed by comparing speed data of dolphins as 



876 



