MARINE MAMMAL SCIENCE, 15(4):1098-1114 (October 1999) 0 1999 by the Society for Marine Mammalogy

LONG DISTANCE OFFSHORE MOVEMENTS OF BOTTLENOSE DOLPHINS1

RANDALL S. WELLS Chicago Zoological Society, YO Mote Marine Laboratory,

1600 Ken Thompson Parkway, Sarasota, Florida 34236, U.S.A. E-mail: rwells@mote.org

HOWARD L. RHINEHART PETRA CUNNINGHAM

Mote Marine Laboratory, 1600 Ken Thompson Parkway, Sarasota, Florida 34236, U.S.A.

JOANNE WHALEY 134 Brushwood Lane, Palm Coast, Florida 32137, U.S.A.

MELODY BARAN CHRIS KOBERNA

Clearwater Marine Aquarium, Clearwater, Florida 34630, U.S.A.

DANIEL P. COSTA Biology Department,

University of California, Santa Cruz, California 95064, U.S.A.

ABSTRACT

Despite recent progress defining the morphological and genetic character- istics of forms of the bottlenose dolphin inhabiting offshore waters, little is known of their behavior or ranging patterns. Reports suggest that an “offshore form” exists between the 200- and 2,000-m isobaths in distinct Gulf of Mexico and western North Atlantic stocks, while one or more coastal forms inhabit the waters inshore. Two opportunities to track rehabilitated adult male bottlenose dolphins with satellite-linked transmitters occurred in 1997. “Rudy” stranded in N W Florida and was released in the Gulf of Mexico off central west Florida. He moved around Florida and northward to off Cape Hatteras, NC, covering 2,050 km in 43 d. “Gulliver” stranded near St. Au- gustine and was released off Cape Canaveral, FL. He moved 4,200 km in 47 d

This article is dedicated to the memory of Professor Kenneth S. Norris, the man responsible for opening the senior author’s eyes to the many differences between pelagic and inshore del- phinids in terms of how they “make their living.” Ken’s teachings about questioning paradigms and allowing the animal to tell the researcher the truths about itself have provided valuable guidance over the years and are evidenced by what we were able to learn from “Gulliver’s travels.”

1098

WELLS ETAL.: BOTTLENOSE DOLPHIN MOVEMENTS 1099

to a location northeast of the Virgin Islands. Gulliver swam through 5,000- m-deep waters 300 km offshore of the northern Caribbean islands, against the North Equatorial Current. These records expand the range and habitat previously reported for the offshore stock of bottlenose dolphins inhabiting the waters off the southeastern United States, underscore the difficulties of defining pelagic stocks, illustrate the success of rehabilitation efforts, indicate the value of follow-up monitoring of rehabilitated and released cetaceans, and expand our understanding of the long-range movement capabilities of a dol- phin species more commonly thought of as a resident in coastal waters. Key words: bottlenose dolphin, cetacean stranding, cetacean rehabilitation, offshore bottlenose dolphin stocks, satellite-linked radio tracking, dolphin release.

Opportunistic tracking studies of stranded and rehabilitated cetaceans when they are returned to the wild are increasingly being used to obtain information about little-known species or populations that live far from shore. Provided the animals have fully recovered from their ailments and they are returned to familiar waters, it can be argued that their behavior and activities soon after release are representative of “normal” members of their populations. This would open a window to the lives of dolphins that would otherwise be very difficult and expensive to study in the wild.

In the southeastern United States, bottlenose dolphins (Tarsiops tramatas) are distributed fairly continuously, though not necessarily with equal density, from bays, sounds, and estuaries offshore into the deep waters beyond the continental slope (Waring et al. 1997, Wells and Scott 1999). Two forms of bottlenose dolphins exist with different genetic profiles, parasite loads, stom- ach contents, and morphology and have been referred to as “coastal” or “off- shore” (Duffield et al. 1983; Hersh and Duffield 1990; Mead and Potter 1990, 1995). The “offshore form” has been reported to range primarily between the 200- and 2,000-m isobaths (Kenney 1990), in two distinct stocks ot popu- lations, Gulf of Mexico and western North Atlantic (Waring et al. 1997), while one or more “coastal” forms inhabit the inshore waters.

Offshore dolphins are generally longer, with proportionately shorter rostra, smaller flippers, wider skulls, larger nares, and wider rostra than coastal dol- phins (Hersh and Duffield 1990, Mead and Potter 1995). The modal length of offshore dolphins is about 15% greater than that of coastal dolphins (Mead and Potter 1995). Hersh and Duffield (1990) suggested that the dolphins inhabiting relatively warm, shallow waters have smaller body sizes and larger flippers because of increased demands for maneuverability and heat dissipation compared to offshore dolphins inhabiting colder, deeper waters, where less maneuvering is required and where heat retention may be more important.

Nearshore and offshore distinctions have also been indicated by genetic analyses. Duffield et al. (1983) and Hersh and Duffield (1990) identified dif- ferent electrophoretic profiles for hemoglobin from coastal dolphins, charac- terized by the speed of sample progression along the electrophoretic gel (0% “slow” vs. 100% “fast”) vs. offshore (70% “slow” vs. 30% “fast”). Subsequent mitochondria1 DNA sequence analyses and nuclear DNA analyses support the

1100 MARINE MAMMAL SCIENCE, VOL. 1 5 , NO. 4, 1999

distinctions between inshore and offshore forms in the western North Atlantic (Curry and Smith 1997, LeDuc and Curry 1997, Hoelzel et al. 1998).

The different forms of Tursiops exhibit differences in other physiological properties as well. Offshore dolphins show higher hemoglobin concentrations, higher hematocrits, and greater red blood cell counts than do the coastal dolphins (Duffield et al. 1983). These characteristics correlate well with the greater oxygen-binding capabilities that would be necessary if the offshore dolphins are diving deeper or longer for their prey (Hersh and Duffield 1990, Mead and Potter 1995). Offshore dolphins also differ from coastal dolphins in food habits and parasite burdens (Mead and Potter 1995).

Bottlenose dolphins with features found to be intermediate between those of the coastal and offshore forms have also been identified. Hersh and Duffield (1990) found an “intermediate” form of hemoglobin profile (35% “slow” vs. 65% “fast”). Mead and Potter (1995) noted that while their offshore series formed a coherent sample with relatively little variation, the coastal series appears to be a mixed sample from more than one population.

These suites of adaptations suggest different ecotypes tied to different phys- iographic and oceanographic conditions, but the specific features that allow the identification of separate populations within these ecotypes are more elu- sive. In coastal waters it has been possible to identify ranging, residency, and social association patterns that suggest possible management units having a genetic basis (Duffield and Wells 1986, 1991, in press; Wells 1986). For offshore dolphins, however, we lack basic information on distribution and ranging patterns that could define stock structure more directly and address questions such as whether environmental features such as water depth or sea- surface temperature can serve as predictors for distinguishing between stocks and the meaning of “intermediate” features relative to the offshore and coastal forms.

Two opportunities to learn from stranded bottlenose dolphins about “off- shore” patterns occurred during 1997. We were able to rehabilitate and release two dolphins: “Rudy,” an adult male with intermediate features, and “Gulli- ver,” an adult male with offshore characteristics. Both were tagged and tracked via satellite-linked transmitters for follow-up monitoring. The results of the tracks of Rudy and Gulliver not only indicated their successful recovery but also provided two of the first records of long-distance offshore movements by non-coastal bottlenose dolphins.

METHODS

Subjects and Case Histories

Rudy, a 270-cm, 227-kg male bottlenose dolphin (CMSC 9634) stranded on 23 December 1996 at Alligator Point near Panacea in the Florida Panhan- dle. He was rescued by a team from Gulf World and transported to the Clearwater Marine Aquarium (CMA) in Clearwater, Florida. He was held in a 640,000-liter pool, given antibiotics for pneumonia, and monitored for chang-

WELLS ETAL.: BOTTLENOSE DOLPHIN MOVEMENTS 1101

es in morbillivirus titer. During 85 d of treatment at CMA, his pneumonia resolved, he gained 63 kg, and subsequent tests indicated no change in mor- billivirus titer or symptoms of an active morbillivirus infection. Thus, he was cleared by the National Marine Fisheries Service for release.

Rudy’s release site was selected on the basis of information from genetic testing by D. A. Duffield of Portland State University and on blood parameters that indicated that he probably had inhabited a range that was offshore of the dolphins commonly found in coastal waters but inshore of the truly deep- water “offshore” form of the bottlenose dolphin (Duffield et al. 1983). Elec- trophoretic tests showed Rudy to have a hemoglobin profile intermediate be- tween “coastal” and “offshore” bottlenose dolphins. His hematocrit and red blood cell count were also considered to be intermediate (Table 1; Duffield et al. 1983, Hersh and Duffield 1990). Thus, it was decided that Rudy should be released in waters at least 33 m deep. Such depths occur fairly close to shore near the stranding site, but the long transport required to return Rudy to Panacea was not considered to be in his best medical interest. Alternatively, such depths could be found less than 50 km from shore in the Gulf of Mexico off Clearwater, involving a transport of less than two hours duration. The closer release site also provided water temperature comparable to that in which Rudy had been maintained at CMA. On 17 March 1997 Rudy was tagged, taken by a U.S. Coast Guard cutter and released approximately 46 km offshore. No other dolphins were nearby at the time of release. Because adult male bottle- nose dolphins are not uncommonly observed alone, this was not considered to be a serious concern.

The long distance from shore precluded close observation or tracking of Rudy after release, so he was fitted with a satellite-linked transmitter (PTT) to monitor his movements (Fig. la). A. Westgate of Duke University Marine Lab prepared the transmitter pack. A side-mount configuration was used, con- sisting of a flat-board ST-10 PTT (Telonics, Mesa, AZ) mounted in a low- profile rectangular lexan plastic box. Prior to tag attachment, the dorsal fin was cleaned with a topical antiseptic and the tagging site was injected with an analgesic. The PTT was attached to the side of the dorsal fin using three 8-mm-diameter delrin plastic pins, which passed through holes cut in the fin with a sterilized cork borer mounted on a cordless electric drill. A sterile hypodermic needle was used to probe the proposed site of each pin, to ensure that it would not impact any major blood vessels. The 3-mm-thick backing plate on the transmitter housing provided attachment points for the delrin pins. The pins passed through the backing plate and dorsal fin and were secured on the opposite side of the fin with a plastic washer and fastened with a bimetallic combination of washers and nuts. Both backing plate and plastic washers were lined with open-cell foam to reduce the possibility of abrasion. The PTT had a 17-cm whip antenna, measured 11 X 5 X 2 cm, and weighed approximately 150 g in air. To minimize the size of the PTT package, only one sensor was used, a surface-time counter, which provided a cumulative record of the time the tag was above the water’s surface. The value of the surface-time counter was transmitted twice during each signal, allowing de-

Tabl

e 1

. C

ompa

rison

s of

coa

stal

and

off

shor

e bo

ttlen

ose

dolp

hins

.

Mea

sure

men

ts

Coa

stal

(A

tlant

ic) O

ffsh

ore

(fro

m H

ersh

and

Duf

field

199

0)

Mea

n (S

D)

Mod

e R

udy

Gul

liver

M

ean

(SD

) M

ode

n =

9 a

dults

n

= 4

adu

lts

Tota

l len

gth

(cm

, mod

e fr

om M

ead

& P

otte

r 24

7.7

(7.0

) 25

0-26

0 27

0 28

9 28

9.3

(13.

5)

290

Ros

trum

tip

to

dors

al f

in t

ip

146.

2 (5

.5)

183

168.

5 (6

.9)

Tip

of

rost

rum

to

umbi

licus

11

3.1

(3.2

) 11

7 12

5.8

(6.3

)

1995

)

Coa

stal

(P

acifi

c) O

ffsh

ore

Blo

od p

aram

eter

s (f

rom

Duf

field

et a

l. 19

83)

Mea

n (9

5% C

L)

Ran

ge

Rud

y G

ulliv

er

Mea

n (9

5% C

L)

Ran

ge

n =

70

n=

6

100%

Fas

t 65

% F

ast

30%

Fas

t 30

% F

ast

Hem

oglo

bin

type

(el

ectro

phor

esis

) 0%

Slo

w/

35%

Slo

w/

70%

Slo

w/

70%

Slo

w/

Hem

oglo

bin

conc

entra

tion

(g/lO

O m

L)

14.5

(0.

4)

13-1

6 13

.6

20.4

18

.4 (0

.9)

18-2

0 H

emat

ocrit

(%

) 42

(1 .

O)

37-4

7 44

.5

58.8

52

(3.0

) 47

-56

Red

blo

od c

ell c

ount

(X

106/

mm

3)

3.61

(0.

10)

3.11

-4.1

4 3.

73

4.65

4.

49 (

0.30

) 4.

05-4

.83

WELLS ETAL.: BOTTLENOSE DOLPHIN MOVEMENTS 1103

Figwe I , a. Satellite-linked VHF transmitter mounted on Rudy. b. Satellite-linked VHF transmitter mounted on Gulliver.

tection of possible transmission errors. To conserve battery life, the tag had a salt-water switch, which prevented transmissions when the animal was sub- merged, and a duty cycle of six hours operation per day. The PTT was powered by two 2h A lithium cells which, under these operating conditions, should provide several months of battery life. The dolphin was freezebranded “406”

1104 MARINE MAMMAL SCIENCE. VOL. 15 . NO. 4. 1999

in 5-cm-high numbers on each side of the body (Irvine et al. 1982, Scott et al. 1990).

Gulliver, a 289-cm adult male bottlenose dolphin (MML 9703) stranded on Crescent Beach, near St. Augustine, Florida, on 30 January 1997. He was rescued by Marineland staff and treated for 39 d at their facility in St. Au- gustine. Gulliver was treated initially for respiratory and intestinal problems. He had a negative titer for morbillivirus. He began swimming almost im- mediately and accepted dead fish within the first four days, eating an average of 16 kg/d. Once his condition had stabilized, it was determined that Gulliver might benefit from a larger pool while completing his recovery. On 9 March 1997 he was transferred to Mote Marine Laboratory (MML) in Sarasota, Flor- ida, where he was treated in a 200,000-liter medical pool for the first month, then moved to a 800,000-liter rehabilitation lagoon. He weighed 272 kg upon arrival at MML. Antibiotics were administered for pulmonary and urinary tract infections that resolved by 11 April 1997. He was also treated for a possible gastric yeast infection. He was monitored for recurrent infection for one month following discontinuation of antibiotics, and no medical problems were de- tected. He consumed up to 10.0 kg of dead fish each day and gained 41 kg. Weight and blubber depth were determined to be comparable to records for other male bottlenose dolphins of similar length and in similar water tem- perature. Upon veterinary and National Marine Fisheries Service certification of his readiness for release, Gulliver was freezebranded “812” in 5-cm-high numbers on each side of the body.

Electrophoretic tests by D. A. Duffield, as well as blood parameter values and body measurements, all indicated that Gulliver was an “offshore” bottle- nose dolphin (Table 1; Duffield et al. 1983, Hersh and Duffield 1990), so it was determined that remote tracking via satellite-linked transmitter would be most practical for follow-up monitoring. The PTT and attachment, prepared by A. Westgate of Duke University Marine Lab, were nearly identical to those used on Rudy (Fig. lb). In addition, a small radio transmitter (standard Model 2 VHF radio transmitter, ATS, Ipsanti, MN, USA) was attached to Gulliver’s dorsal fin to facilitate tracking immediately following release and in the event that Gulliver moved near to shore. This tag transmitted in the 148 MHz range at 110 pulses per minute, without a salt-water switch or interrupted duty cycle. The VHF transmitter had a life expectancy of more than 50 d. The tag was attached to livestock ear tags (Jumbo Roto-tags, Dalton Supplies, Nettlebed, England) applied to the trailing edge of the dorsal fin. The VHF tag had a 33 cm-long whip antenna, measured 1.1 X 2.5 X 5.5 cm, weighed 15 g in air, and had an effective range of approximately 5 km at sea level. Similar roto-tag VHF transmitters have been used on bottlenose dolphins in Florida and North Carolina (Read et at. 1996).

O n 20 May 1997 Gulliver was placed in a water-filled custom-built con- tainer, transported by truck to Cape Canaveral, Florida, loaded onto a U.S. Coast Guard cutter, and taken about 70 km offshore into the Atlantic Ocean for release. During release by the crew of the cutter, one sling pole struck the

WELLS E T A L . : BOTTLENOSE DOLPHIN MOVEMENTS 1105

transmitter package and broke the nut off one of the three attachment pins, but this did not appear to interfere with the functioning of the transmitter.

Tracking and Analyses

The data from Service Argos consisted of time, latitude, and longitude, with indicators of the precision of the location based on signal quality (Class 3 = 1150 m SD, Class 2 = 1 3 5 0 m SD, Class 1 = 1 1 0 0 0 m SD, Class 0 = >1,000 m SD; precision of A, B, and 2 signals within Class 0 can be ranked as presented above, but not quantified). Location data were filtered to identify unreasonable movements; locations that resulted in average rates > 10 km/h were eliminated and average rates recalculated between previous and subsequent points.

Immediately following release, Gulliver was tracked to monitor his initial response to release. We used Telonics TR2 receivers and hand-held 3-element antennas operated from the forward deck of the Coast Guard cutter. Initial observations indicated a strong correlation between signals and surfacings for respiration.

Environmental data were obtained through databases available on the In- ternet. Bathymetry data were obtained from the NOAA National Geophysical Data Center ETOPOS database of 5-Minute Gridded Elevation Data (NOAA 1988). Depths were provided as averages over a grid of five minutes of latitude by five minutes of longitude. Sea-surface temperature data were obtained from processed Advanced Very High Resolution Radiometer (AVHRR) data from NOAA satellite images, as archived by the University of Rhode Island Grad- uate School of Oceanography.

RESULTS

Ten signals were received from Rudy through 28 April 1997, over the first 43 d following release. Only four signals resulted in location data, and none of these were of sufficient quality for precise estimates. Location data from days 10, 18, 25, and 41 indicated movements of at least 2,050 km, around the south tip of Florida to a point offshore of Cape Hatteras, NC (Fig. 2). These movements corresponded to the general paths of the Gulf of Mexico's Loop Current and the Atlantic Ocean's Gulf Stream. Too few data were avail- able to allow calculation of detailed rates of travel, but overall Rudy averaged at least 48 km/d, or 2.0 km/h. The northeastward currents of the Gulf Stream at its axis average 2.8-5.6 km/h, with slower currents away from the axis. All locations placed Rudy in waters with surface temperature of 23°C. His first three locations were in waters approximately 10-1 13 m deep, whereas his last location near the northeastern tip of a narrow plume of 23°C water was in waters approximately 3,400 m deep. Rudy had been maintained at CMA in waters of 18"-25°C (ave. = 21°C).

We tracked and observed Gulliver directly for 1.6 h immediately following release. His respiratory patterns immediately upon release were dramatically

1106 MARINE MAMMAL SCIENCE. VOL. 1 5 . NO. 4. 1999

Figure 2. Locations of dolphin Rudy as revealed from satellite-linked radio trans- missions.

different from those at the shallow rehabilitation facility. Lengthy dives were followed by two to six respirations in rapid succession. The overall average respiration interval was 44 sec (+60 sec SD, n = 67), nearly double that recorded while he was at MML (24 sec 'I 14 sec SD, n = 919), with a maximum dive duration of 208 sec, more than double his longest dive du- ration at MML (90 sec).

We received 105 satellite-linked signals from Gulliver through 5 July 1997, over the first 47 d following release. Location data were obtained from 76 signals, including ten locations with precision 5 1 km (days 5 , 7 , 26, 28, 29, 39, 40,41, 42,45). Filtering out location data resulting in unreasonable travel rates produced 69 usable locations, indicating a minimum distance covered of 4,200 km over 47 d. Gulliver's travels took him inshore towards Cape Ca- naveral, then northward over the continental shelf, passing within about 37 km of his stranding site on day 4, continuing to as far as Savannah, GA on day 7 , then offshore to the southeast (Fig. 3). When Gulliver reached the axis of the Gulf Stream, he moved with the current on days 9-12. He then reversed his course and on day 17 ( 5 June) he resumed his southeasterly heading through day 44. On this course, Gulliver was swimming against the North Equatorial Current. By day 20 (8 June) Gulliver entered waters more than 3,000 m deep and continued in waters more than 5,000 m deep through the last day of tracking (Fig. 4). Gulliver reached the easternmost extent of his

000'9-

OOO'S-

OOO'C-

P u OOO'E-5

3 - - 000'2-

000'1-

0

1108 MARINE MAMMAL SCIENCE. VOL. 1 5 . NO. 4. 1999

10.00 , 9.00 1 . ,

1.00 1 0.00 I

0 10 20 30 4 0 5 0

Days Since Release

Figure 5. Rates of travel by Gulliver.

track by day 45 and moved to the southwest and west through day 47 ( 5 July).

Once Gulliver turned away from the mainland U.S. shoreline on day 8, he never again approached any shore closely. During days 1-8, Gulliver was typically within 34 km of shore ( 5 2 0 km SD, range = 7-78 km). He then remained an average distance of 321 km ( k 7 7 km SD, range = 118-440 km) from the nearest mainland or island shoreline.

Gulliver traveled approximately 89 km/d, on average. He moved both with and against the current, and his rates of travel varied from day to day (Fig. 5). O n days 1-7 and 10-12 he moved with the Gulf Stream, where the northeastward currents average 2.8 to 5.6 km/h. Gulliver's speed was 1.8-3.4 km/h with the current. On days 12-14 he swam against the Gulfstream, and averaged 1.5 km/h. On days 23-44 he swam against the slower (0.9-1.9 km/ h) North Equatorial Current, and averaged 4.5 km/h.

Gulliver was released into 24°C water off Cape Canaveral, comparable to the temperatures in which he had been maintained at MML (22"-28"C, ave. = 2 5"C>, and he traveled through sea-surface temperatures of 19"-27°C (ave. = 24°C). Deviations in Gulliver's heading appeared to be related at least in part to changes in sea-surface temperature, especially those associated with approaching cold fronts. Sea-surface temperatures along Gulliver's route varied

WELLS ETAL.: BOTTLENOSE DOLPHIN MOVEMENTS 1109

"1 t 2MI

H

- 180

OJ 2; 1Bo

O ~ b " B 8 0 ~ ~ 6 9 9 p ~ ~ ~ 9 ~ ~ $ ~ ~ ~ ~ ~ D.ya Sin- Reha~e

P E A SURF. TEMP. 53 +HEADING DEFLECTION 1

Figwe 6. Gulliver's travels relative to sea-surface temperature. "Heading Deflec- tion" represents change from previous heading, as indicated from orientation of line connecting two consecutive satellite-fixed locations. Arrows indicate passage of major cold fronts (see text).

the most during the first half of the track (Fig. 6), when cold fronts from the northwest passed through his locations. The strongest fronts, as indicated by degree of change in surface temperatures and extent of cloud cover, were as- sociated with the strongest deflections in Gulliver's course (Fig. 6). Three of four major heading changes of more than 100 degrees were associated with the passage of strong fronts in late May and early June. Each change was to the south or southeast, toward warmer water. During the latter half of his track, when fronts were weaker, Gulliver maintained a relatively steady south- easterly heading through waters averaging 25"-26"C (Fig. 6).

DISCUSSION

In light of the increase in recent years in the number of releases of stranded cetaceans following rehabilitation (20 in the southeast U.S. during 1992- 1997), the National Marine Fisheries Service strongly encourages postrelease monitoring so they may evaluate the success, costs, and benefits of cetacean rehabilitation efforts. Satellite-linked transmitters first and foremost provided a means of maintaining contact with the rehabilitated dolphins for determi- nation of their postrelease survival. Track durations alone suggested but could not confirm that the dolphins had recovered successfully from the ailments that led to their strandings. An animal that was unable to feed adequately or

1110 MARINE MAMMAL SCIENCE, VOL. 15. NO. 4, 1999

was suffering from serious medical problems would likely not have survived to transmit signals for more than six weeks. Similarly, a compromised dolphin would likely not have been able to maintain the travel rates demonstrated by Gulliver, especially swimming against currents. Taken with the locational data, the regularity of Gulliver’s transmissions provided evidence that he was main- taining normal orientation in the water and engaging in normal locomotion. The poorer quality of Rudy’s transmissions and the fact that his average travel rate was comparable to the currents with which he was moving raise questions regarding his condition, but surface-time counter data were consistent with those from other cetaceans, indicating that he was spending no more or less time at the surface than might be expected of a normal, healthy dolphin.* The infrequency of Rudy’s signals was believed to be a result of the setting of the timing of the duty cycle.

Satellite-linked telemetry proved to be a cost-effective means of conducting follow-up monitoring of these two rehabilitated stranded animals, though in the case of Rudy the appropriate interpretation of the tracking data is unclear. Previous tagging efforts with healthy bottlenose dolphins indicated that the technique could be used to safely track this species for a month or more, the most critical period for a rehabilitated dolphin (Tanaka 1987, Mate et af . 1995). Satellite tracking had proved to be similarly effective in previous fol- low-up monitoring of rehabilitated deep-water cetaceans, including an Atlan- tic spotted dolphin, Stenefla frontalis (Davis et af. 1996), short-finned pilot whales, Gfobicephala macrorhynchzls (Mate 1989), and a harbor porpoise, Pho- coena phocoena (Westgate et af. 1998). This technique seems particularly well suited for dolphins moving through deeper offshore waters as compared to shallow inshore waters. Long distances from shore and vessel expenses may preclude conventional VHF tracking of dolphins in offshore waters. In shallow water dolphins that beat the packages on the sea floor or other objects may cause premature loss of the transmitter (Irvine et af. 1982, Scott et af. 1990, Read et al. 1997). To date, inshore bottlenose dolphins have been tracked via satellite for only 26 d (Mate et af. 1995) to 32 d (Read et al. 1997). In contrast, Rudy was tracked for 43 d, and Gulliver’s 47-d track occurred even though only two of the original three attachment pins were intact upon release.

The precision of the location data for both dolphins was disappointing. Field tests with manatees have quantified the error radii in marine habitats for satellite-linked transmissions through Service Argos (Deutsch et al. 1997). The proportions of Gulliver’s signals relative to these error radii were: 3% Class 2 (within 398 m), 7% Class 1 (908 m), 10% Class 0 (3,741 m), 22% Class A (3,073 m), 31% Class B (8,586 m), and 27% Class 2. Rudy provided 10% Class A, 30% Class B, and 60% Class Z transmissions. The higher mounting of the transmitter on Gulliver’s dorsal fin may have contributed to the higher frequency of received signals and the greater precision of his locations, as compared to Rudy and inshore Tzlrsiops tagged during June 1997 with the

Personal communication from Andrew Westgate, Duke University Marine Lab, Beaufort, NC 28516, Canada, December 1998.

WELLS ET AL.: BOTTLENOSE DOLPHIN MOVEMENTS 1111

same system (Read et al. 1997). In open-ocean situations where many envi- ronmental features occur on broader scales than in inshore waters, the impre- cision we experienced may be tolerable, however.

It is not possible to know with certainty that the observed movement pat- terns of Rudy and Gulliver were representative of those of other members of their populations. Bottlenose dolphins inhabit the waters through which Rudy and Gulliver traveled. They were released into habitats believed to be similar to those of their origin. Gulliver was released within 155 km of his stranding site, and he passed 37 km offshore of this site within the first four days postrelease. No detailed information is available on the behavior of Rudy for the first nine days following release, except that he moved southward around the tip of Florida at a rate of at least 75 kmid. Gulliver’s movements into and through shallow waters during his first eight days postrelease, before moving into deeper offshore waters, may have indicated some degree of initial disori- entatio’n due to release far from his stranding site. Movements into shallow waters by apparently disoriented dolphins released both into strange and fa- miliar environments have been reported (Irvine 197 1, Wells et al. 1998). How- ever, with the exception of Rudy’s final location, the subsequent movements of both dolphins placed the animals generally within the kinds of habitats predicted for them based on genetic and hematological characteristics (Hersh and Duffield 1990).

It seems reasonable to assume that the dolphins would seek out appropriate and familiar habitats and respond similarly to the same environmental features as before their strandings. Our observations suggest that these dolphins re- spond to a variety of environmental factors. Except for his last location, Rudy apparently remained in waters of depth comparable to that at his release site and predicted habitat, over the continental shelf. Gulliver eventually moved away from shore, into waters of great depth off the edge of the continental shelf and remained there, as predicted. The movements of both dolphins also appeared to be related to sea-surface temperature. Both were released into waters within a few degrees of the temperatures of their holding pools. Rudy remained in the same surface temperature (23°C) throughout his track, even when the availability of this temperature was reduced to a narrow plume in deep water off Cape Hatteras. The range of sea-surface temperatures in which Gulliver was. found progressively narrowed to 25-26°C over the course of the track.

If Rudy and Gulliver were exhibiting normal behavior representative of their populations/stocks of origin, then our findings would alter some of the established concepts of bottlenose dolphin stock structure in offshore waters. The dolphins exhibited much more mobility than has been previously docu- mented for this species. Rudy’s movements from the Gulf of Mexico into the Atlantic Ocean link two regions previously considered to be used by different continental shelf stocks (Waring et al. 1997). Gulliver’s movements to the southeast into waters more than 5,000 m deep differed significantly from the previously held concept of offshore bottlenose dolphins moving along the At- lantic seaboard in waters 200 m to 2,000 m deep (Kenney 1990). Gulliver’s

1112 MARINE MAMMAL SCIENCE, VOL. 1 5 . NO. 4. 1999

travels also took him outside of U.S. waters, suggesting the need for a different management approach than for dolphins remaining within U.S. waters.

ACKNOWLEDGMENTS

Without the dedicated efforts of the veterinary and animal care staff members and volunteers at Clearwater Marine Aquarium, Marineland of Florida, and Mote Marine Laboratory, there would have been no rehabilitated dolphins to track. In particular, Forrest Townsend, Charles Manire, Debi Colbert, David Smith, Tom Morris, Kumar Mahadevan, and Jay Gorzelany at MML made crucial contributions to the efforts to return Gulliver to the sea. The rehabilitation of Rudy was greatly facilitated by the efforts of Gulf World, R. T. Goldston, Dennis Kellenberger, Terri Hepburn, Melissa Taylor, Mike Walsh, and Forrest Townsend. Andrew Westgate prepared both trans- mitter packages and, along with Andrew Read and Tara Cox, processed the information resulting from Rudy and reviewed the manuscript. The U. S. Coast Guard graciously provided transportation to sea for both dolphins. We thank Dan Crocker and Danielle Waples of the University of California, Santa Crut, for their assistance in obtaining and processing the tracking data for Gulliver, and Faith Keller, Dirk Neumann, and James Ganong for their assistance with figure preparation based on these data. Michael Scott, as per usual, improved the quality of the manuscript through his reviews.

LITERATURE CITED

CURRY, B. E., AND J. SMITH. 1997. Phylogeographic structure of the bottlenose dolphin (Tursiops trancatus): Stock identification and implications for management. Pages 227-247 in A. E. Dizon, S . J. Chivers and W. F. Perrin, eds. Molecular genetics of marine mammals. Special Publication Number 3, Society for Marine Mam- malogy, Lawrence, KS.

DAVIS, R. W., G. A. J. WORTHY, B. WURSIG AND S. K. LYNN. 1996. Diving behavior and at-sea movements of an Atlantic spotted dolphin in the Gulf of Mexico. Marine Mammal Science 12:569-581.

DEUTSCH, C. J., D. E. EASTON, H. I. KKHMAN AND J. P. REID. 1997. Locational accuracy of the Argos satellite telemetry system in a marine environment: Impli- cations for spatial data analysis and wildlife management. Abstracts, Forum on Wildlife Telemetry: Innovations, Evaluations, and Research Needs. 2 1-23 Sep- tember 1997, Snowmass Village, CO. p. 24. Available on the Internet from http: //wWW.npwrc.usgs.gov/resource/tools/telemtry/argos. htm.

DUFFIELD, D. A., AND R. S. WELLS. 1986. Population structure of bottlenose dolphins: Genetic studies of bottlenose dolphins along the central west coast of Florida. Contract Report to National Marine Fisheries Service, Southeast Fisheries Center, 75 Virginia Beach Drive, Miami, FL 33149. Contract No. 45-WCNF-5-00366.

DUFFIELD, D. A., AND R. S. WELLS. 1991. The combined application of chromosome, protein and molecular data for the investigation of social unit structure and dy- namics in Tursiops truncatus. Report of the International Whaling Commission (Special Issue 13):155-169.

DUFFIELD, D. A., AND R. S. WELLS. In press. The molecular profile of a resident community of bottlenose dolphins, Tursiops trumatus. In C. J. Pfeiffer, ed. Cell and molecular biology of marine mammals. Krieger Publishing Company, Melbourne, IT.

DUFFIELD, D. A., S. H. RIDGWAY AND L. H. CORNELL. 1983. Hematology distinguishes coastal and offshore forms of dolphins (Tursiops). Canadian Journal of Zoology 61:

HERSH, S. L., AND D. A. DUFFIELD. 1990. Distinction between Northwest Atlantic

10 PP.

930-933.

WELLS E T A L . : BOTTLENOSE DOLPHIN MOVEMENTS 1113

offshore and coastal bottlenose dolphins based on hemoglobin profile and mor- phometry. Pages 129-139 in S. Leatherwood and R. R. Reeves, eds. The bottle- nose dolphin. Academic Press, San Diego, CA.

HOELZEL, A. R., C. W. POTTER AND P. B. BEST. 1998. Genetic differentiation between parapatric ’nearshore’ and ‘offshore’ populations of the bottlenose dolphin. Pro- ceedings of the Royal Society of London B 265:1177-1183.

IRVINE, B. 1971. Behavioral changes in dolphins in a strange environment. Quarterly Journal of the Florida Academy of Sciences 34206-212.

IRVINE, A. B., R. S. WELLS AND M. D. SCOTT. 1982. An evaluation of techniques for tagging small odontocete cetaceans. Fishery Bulletin, U.S. 80:135-143.

KENNEY, R. D. 1990. Bottlenose dolphins off the northeastern United States. Pages 369-386 in S. Leatherwood and R. R. Reeves, eds. The bottlenose dolphin. Ac- ademic Press, San Diego, CA.

LEDUC, R. G., AND B. E. CURRY. 1997. Mitochondria1 DNA sequence analysis indicates need for revision of the genus Tursiops. Unpublished meeting document SC/48/ SM27 submited to Scientific Committee of the International Whaling Commis- sion. 12 pp. Available from IWC, The Red House 135 Station Road, Histon, Cambridge CB4 4NP, UK.

WTE, B. R. 1989. Watching habits and habitats from Earth satellites. Oceanus 32:

MATE, B. R., K. A. ROSSBACH, S. L. NEUKIRK, R. S. WELLS, A. B. IRVINE, M. D. SCOTT AND A. J. READ. 1995. Satellite-monitored movements and dive behavior of a bottlenose dolphin (Tursiops truncatus) in Tampa Bay, Florida. Marine Mammal Science 11A52-463.

MEAD, J. G., AND C . W. POTTER. 1990. Natural history of bottlenose dolphins along the central Atlantic coast of the United States. Pages 165-195 in S. Leatherwood and R. R. Reeves, eds. The bottlenose dolphin. Academic Press, San Diego, CA.

MEAD, J. G., AND C. W. POTTER. 1995. Recognizing two populations of the bottlenose dolphin (Tursiops truncatus) off the Atlantic coast of North America-morphologic and ecologic considerations. IBI Reports 5:3 1-44. International Marine Biological Institute, Kamogawa, Japan.

NOAA. 1988. Data Announcement 88-MGG-02, Digital Relief of the Surface of the Earth. NOAA, National Geophysical Data Center, Boulder, CO.

READ, A. J., A. J. WESTGATE, K. W. URIAN, R. S. WELLS, B. M. ALLEN AND W. J. CARR. 1996. Monitoring movements and health status of bottlenose dolphins in Beaufort, NC using radio telemetry. Final contract report to the National Marine Fisheries Service, Southeast Fisheries Science Center, 2 17 Fort Johnson Road, Charleston, SC 29412. Contract No. 40-GENF-500160. 37 pp.

READ, A. J., A. J. WESTGATE, R. S. WELLS, D. M. WAPLES, B. M. ALLEN, M. D. SCOTT AND A. A. HOHN. 1997. Testing attachment techniques for satellite transmitters on bottlenose dolphins near Sarasota, Florida. Contract Report to National Marine Fisheries Service, Southeast Fisheries Science Center, Beaufort Laboratory, Beau- fort, NC 28516. Contract No. 40ETNF700036. 12 pp.

SCOTT, M. D., R. S. WELLS, A. B. IRVINE AND B. R. MATE. 1990. Tagging and marking studies on small cetaceans. Pages 489-514 in S. Leatherwood and R. R. Reeves, eds. The bottlenose dolphin. Academic Press, San Diego, CA.

TANAKA, S. 1987. Satellite radio tracking of bottlenose dolphins Tursiops truncatus. Nippon Suisan Gakkaishi 53:1327-1338. (In Japanese)

WARING, G. T., D. L. PALKA, K. D. MULLIN, J. H. W. HAIN, L. J. HANSEN AND K. D. BISACK. 1997. U. S. Atlantic and Gulf of Mexico marine mammal stock assess- ments-1996. NOAA Technical Memorandum NMFS-NE-114. NEFSC, Woods Hole, MA 02543. 250 pp.

WELLS, R. S. 1986. Population structure of bottlenose dolphins: Behavioral studies of bottlenose dolphins along the central west coast of Florida. Contract Report to

14-18.

1114 MARINE MAMMAL SCIENCE, VOL. 15. NO. 4, 1999

National Marine Fisheries Service, Southeast Fisheries Center, 75 Virginia Beach Drive, Miami, FL 33149. Contract No. 45-WCNF-5-00366. 70 pp.

WELLS, R. S., AND M. D. SCOTT. 1999. Bottlenose dolphin Tursiops truncatus (Montagu, 1821). Pages 137-182 in S. H. Ridgway and R. Harrison, eds. Handbook of marine mammals. Volume 6. The second book of dolphins and the porpoises. Academic Press, San Diego, CA.

WELLS, R. S., K. BASSOS-HULL AND K. S. NORRIS. 1998. Experimental return to the wild of two bottlenose dolphins. Marine Mammal Science 1451-71.

WESTGATE, A. J., A. J. READ, T. M. Cox, T. D. SCHOFIELD, B. R. WHITAKER AND K. E. ANDERSON. 1998. Monitoring a rehabilitated harbor porpoise using satellite telemetry. Marine Mammal Science 14:599-604.

Received: 13 December 1998 Accepted: 11 March 1999