226 
Fishery Bulletin 99(2) 
gici spp. group was made up of sightings of dwarf sperm 
whales (n=32), pygmy sperm whales (n= 7), and those small 
whales that could only be identified to the genus Kogia 
(n= 17). The spatial distributions of sightings suggest high 
use areas for each species (Fig. 4), but these are heavily in- 
fluenced by the distribution of the sighting effort. To better 
capture the true spatial distributions, the location of each 
sighting and 1-km transect section was projected onto the 
200-m isobath (by using the minimum distance to this iso- 
bath) and equal-effort sighting distributions were gener- 
ated with respect to the distance along the 200-m isobath 
(Fig. 4). Chi-squared analyses indicated that all of the spe- 
cies’ distributions, except the Kogia spp. group, were sig- 
nificantly different from a uniform distribution (PcO.Ol). 
The northwestern Gulf of Mexico (west of the Mississippi 
River Delta) had much lower group sighting rates of each 
cetacean species when compared with the northeastern 
Gulf (east of the Mississippi River Delta). To the south of 
New Orleans, the Mississippi Canyon and just seaward 
of the Mississippi River Delta were regions of high group 
encounter rates for bottlenose dolphins, Risso’s 
dolphins, and sperm whales. Just to the east of 
this region and south of Mobile Bay, pantropi- 
cal spotted dolphin sighting rates reached a lo- 
cal maximum. Along the steep upper continen- 
tal slope of the Florida Escarpment between 
Tampa and Key West, very high relative abun- 
dances of Risso’s dolphins, pantropical spotted 
dolphins, and sperm whales were observed. 
Examination of the mean, median, first and 
third quartiles, and standard deviation of the 
environmental variables for each species sug- 
gested significant interspecies differences (Fig. 
5). The null hypothesis of equal medians for 
each species was rejected for depth, depth gra- 
dient, surface temperature, and zooplankton 
biomass (Mood’s median test, .PcO.Ol). Similar- 
ly, the null hypothesis of equal “locations” was 
rejected with a Kruskal-Wallis test for depth, 
depth gradient, zooplankton biomass (PcO.Ol), 
and surface temperature (P<0.05). The bottle- 
nose dolphin had the lowest median depth, 
depth gradient, and surface temperature of all 
the species. The Risso’s dolphin had the high- 
est median depth gradient and surface tem- 
perature and the Kogia spp. had the highest 
median zooplankton biomass. The bottlenose 
dolphin’s median habitat was so different from 
the others that if this species was removed 
from each of the Mood’s median tests, the null 
hypothesis of equal medians between species 
would be rejected for only depth (PcO.Ol) and 
zooplankton biomass (Pc0.05). 
Despite clear heterogeneity of variances (Fig. 
5A), a one-way analysis of variance indicated 
that the cetacean distributions with respect 
to depth were significantly different. Further- 
more, a Duncan’s multiple range test suggest- 
ed species groupings by depth (Pc0.05) that 
were qualitatively accurate and in agreement 
with earlier results (Mullin et al., 1994). These 
species groupings were 1 ) bottlenose dolphins, 
2) Risso’s dolphins and Kogia spp., and 3) pan- 
tropical spotted dolphins and sperm whales. 
Bottlenose dolphins were encountered predom- 
inantly over the continental shelf and were 
never sighted seaward of the 750-m isobath. 
Risso’s dolphins and Kogia spp. were dis- 
tributed mostly over the upper continental 
slope, whereas pantropical spotted dolphins 
o 
CD £9 
CD Cl 
co E 
) o 
CO 
a> - 
Q_ CO 
11 
CD "O 
O J- 
co 
-1 
w 2 
CO 
27 
26 
25 
24 
23 
0.8 
0.6 
0.4 : 
0 . 2 - 
0.0 
8 
E 
c o 
6- 
o o 
c CD 
_cd 
Q- CO 
O CO 
O CD 
N £ 
4 
2 ■ 
31 
f 
o 
100 
f 200 4 
-C 
a 300 
400 
500 
V V V V 
Figure 3 
Surface temperature (raw temperatures in gray, 10 km radius average 
in black), surface temperature variability (standard deviation in 10 km 
radius), epipelagic zooplankton biomass and vertical temperature struc- 
ture along 27°N observed between 19 May and 1 June 1993. Surface 
temperature and surface temperature variability were obtained from 
the composite satellite image shown in Figure 2. The inverted triangles 
indicate CTD/XBT station locations from which the temperature section 
was derived. 
