234 
Fishery Bulletin 99(2) 
trinities for cetaceans through local trophic interactions. 
Likewise, the area west of the southwestern Florida shelf 
break may be another region of high productivity. The 
physical oceanography of this region is characterized by 
the formation of a cyclonic meander or eddy in the spring 
between the Loop Current to the west and the steep Flor- 
ida Escarpment to the east (Cochrane, 1972; Vukovich et 
ah, 1979; Vukovich and Maul, 1985). Maul et al. (1984) ob- 
served that bluefin tuna catch per unit of effort inside a 
cold-core meander in this region was three times higher 
than in the central Gulf the previous year. Between 83— 
86°W and 24- 27°N in oceanic waters, the sighting rates of 
Risso’s dolphins, pantropical spotted dolphins, and sperm 
whales were 3.8, 2.6, and 2.8 times higher than the aver- 
age sighting rate and 4.9, 3.0, and 3.3 times higher than 
the sighting rate outside of this region, respectively. 
Bottlenose dolphin 
The bottlenose dolphin’s distribution in the northern Gulf 
of Mexico is markedly different from the other species 
examined in our study. This species and the Atlantic spot- 
ted dolphin are the only cetaceans that are routinely 
encountered on the continental shelf (Fritts et ah, 1983; 
Mullin et ah, 1994; Jefferson and Schiro, 1997; Hansen et 
al. 3 ). Caution is warranted when interpreting the bimodal 
distribution of bottlenose dolphin sighting rates with 
respect to depth (Fig. 7A). Effort on the continental shelf 
was neither extensive nor distributed uniformly through- 
out the northern Gulf. During the CETAP study (Kenney, 
1990; CETAP 5 ), a distinct bimodal distribution of bottle- 
nose dolphins was observed north of Cape Hatteras. Bot- 
tlenose dolphins were concentrated during warm months 
in waters less than 25 m and year round near the 1000-m 
isobath and some groups were sighted in waters as deep 
as 4712 m (CETAP’). This bimodal distribution is sugges- 
tive of the inshore (coastal) and offshore forms of bottle- 
nose dolphins described by others (Norris and Prescott, 
1961; Walker, 1975; Leatherwood and Reeves, 1982; Shane 
et al., 1986; Kenney, 1990; Walker 1 '-) and supported by 
mitochondrial DNA (Dowling and Brown, 1993; Curry and 
Smith, 1997), hematological (Duffield et al., 1983; Hersh 
and Duffield, 1990), and morphological (Hersh and Duff- 
ield, 1990) evidence. The spatial distribution of bottlenose 
dolphin group sightings from aerial surveys on the conti- 
nental shelf of the Gulf of Mexico (Blaylock et al., 1995) 
and off the southeast U.S. coast south of Cape Hatteras 
(Blaylock and Huggard, 1994), however, was not character- 
ized by any large-scale discontinuities in bottlenose dol- 
phin distribution similar to those observed north of Cape 
Hatteras. 
The shelf bottlenose dolphins were found in regions with 
cooler than expected surface waters and high surface tem- 
12 Walker, W. A. 1981. Geographical variation in morphology 
and biology of bottlenose dolphins iTursiops) in the eastern 
North Pacific. Southwest Fisheries Center Administrative 
Report LJ-81-03C, 52 p. [Available from Southwest Fisheries 
Science Center, National Marine Fisheries Service, NOAA, P.O. 
Box 271, La Jolla, CA 92038.1 
perature variability. These oceanographic characteristics 
are consistent with the cool and fresh water side of fronts 
associated with river plumes and, indeed, sighting rates of 
the shelf bottlenose dolphins were particularly high near 
the Mississippi River Delta. Sighting rates of the shelf 
break bottlenose dolphins were more evenly distributed in 
the central and eastern Gulf and the high surface temper- 
ature variability observed near these dolphins suggests a 
potential association with shelf break fronts. 
Risso's dolphin 
Baumgartner (1997) examined the same 1992-94 spring 
cruise data used in our study with the intent of defining 
Risso’s dolphin habitat in terms of the physiography of 
the northern Gulf of Mexico. Using both univariate and 
bivariate analyses, he determined that the sighting rate 
of Risso’s dolphin groups between the 350- and 975-m iso- 
baths and in depth gradients exceeding 24 m per 1.1 km 
was nearly 5 times the average. Of the groups encountered 
outside this region, 40% were sighted within 5 km of it. 
Aerial survey data collected during all seasons between 
1992 and 1994 were used to independently assess this 
habitat model. Sighting rates from these surveys were 
nearly 6 times the average inside this core habitat, and of 
the groups encountered outside of this region, 73%< were 
sighted within 5 km of it. 
The distribution of Risso’s dolphin along the continen- 
tal slope has been noted in several studies (Wiirtz et al., 
1992; CETAP 5 ; Dohl et al. 6 ; Dohl et al. 7 ; Green et al. 8 ; Da- 
vis et al. 11 ) and some evidence exists to support this spe- 
cies’ association with the steeper sections of the upper con- 
tinental slope elsewhere. Off the Oregon and Washington 
coasts, Green et al. 8 observed that Risso’s dolphin encoun- 
ter rates over the continental slope (200-2000 m) were 
seven times greater than on the shelf and that the groups 
sighted on the shelf were very close to the shelf break. 
Compared with the northern Gulf of Mexico, almost the 
entire Oregon-Washington continental slope can be con- 
sidered steep with depth gradients in excess of 22 m per 
1.1 km (Fig. 11 in Green et al. 8 ). Dohl et al. 7 found a simi- 
lar distribution off central and northern California, where 
the majority of Risso’s dolphin sightings were between the 
183- and 1830-m (100-1000 fathom) isobaths. As is the 
case off Oregon and Washington, virtually all of the con- 
tinental slope off central and northern California can be 
considered very steep (Fig. 1 in Dohl et al. 7 ). The physiog- 
raphy of the northwestern Atlantic Ocean is much more 
like that found in the northern Gulf of Mexico and the 
CETAP study (Hain et al.. 1985; Kenney and Winn, 1986; 
CETAP 5 ) found Risso’s dolphins concentrated at the shelf 
break (mode of 478 sightings was 183 m depth) and dis- 
tributed over the entire continental slope (average of 478 
sightings was 1092 m). 
Baumgartner (1997) hypothesized that Risso’s dolphins 
aggregate along the upper continental slope because of the 
presence of a persistent ocean front separating the rela- 
tively cool and fresh waters of the continental shelf and 
the more warm and salty waters of the oceanic Gulf. This 
shelf break front may provide greater feeding opportuni- 
