the X = 32.4 sec mean dives during daylight hours (SD = 5.94, n = 10 dolphins, 

 291 30-min samples, P = 0.0006, df = 2, Kruskal-Wallis). The significance of the 

 test was mostly due to FB504, FB515, FB514, and FB522, which had night dives 

 longer than day dives by 5-10 sec (Fisher's LSD). Figure 9 shows the difference 

 between night and day dive durations for actual surfacing intervals rather 

 than the averaged samples. There was a lower relative frequency of 

 dives < 10 sec, and higher relative frequency of dives > 50 sec, at night. 



The dive durations of individuals were different (P < 0.0001, df = 9, 

 Kruskal-Wallis). Fisher's LSD revealed that dive durations of FB514 and 

 FB521 were different from those of most other radio-tagged dolphins. Dives of 

 FB514 were longer by 6-14 sec, and FB52rs dives were shorter by 8-20 sec. 



Mann- Whitney U tests indicated no difference in dive durations between 

 males and females, pregnant and not-pregnant females, or females with-calf 

 and without-calf. Dive durations were not linearly dependent on either age or 

 surface duration. 



Dolphins spent more time at the surface between dives at night than i n 

 the day (P = 0.0053, df = 3, Kruskal-Wallis). Mean nighttime surface duration 

 was 6.0 sec (SD = 4.8, n = 129 surfacings), mean daytime surface duration was 

 5.1 sec (SD = 2.92, n = 238). The surface durations of individuals were different 

 (P < 0.0001, df = 9, Kruskal-Wallis, followed by Fisher's LSD). Fisher's LSD 

 revealed that FB501, FB505, and FB5irs surface durations differed from those 

 of almost all other radio-tagged dolphins. FB501 spent 4.5-7.5 sec/surfacing 

 more at the surface, and FB505 and FB511 spent 3-4.5 sec/surfadng more at 

 the surface than most other radio-tagged dolphins. 



Females spent longer times at the surface than males (P = 0.0278, Mann- 

 Whitney U), but showed no difference in surface durations between pregnant 

 and not-pregnant females, or females with-calf and without-calf. Surface 

 durations were not linearly dependent on age. 



Longer night dive and surface durations imply less diving at night. This 

 is supported statistically by significantly lowner dive rates at night than during 

 day and crepuscular periods (P = 0.0003, df = 2, Kruskal-Wallis, followed by 

 Fisher's LSD). 



No differences were found in dive rate between sex, pregnant and not 

 pregnant females, or females with-calf and without-calf (Mann-Whitney U) 

 or among individuals (Kruskal-Wallis). E>ive rates were not related to age 

 (simple linear regression). As expected, there was a strong negative relation 

 between dive rate and dive duration (R'= 0.80, P = 0.0005, ANOVA, 

 n = 10 dolphins). 



15 



