Anita. Behav.

, 1976, 24, 849-857

ORIENTATION AND BEHAVIOUR OF HATCHLING GREEN TURTLES

(CHELONIA MYDAS) IN THE SEA
BY J. F R I C K
P.O. Box 293, Lincoln, Massachusetts 01773

Abstract. To investigate the initial stage of the 'lost-year puzzle' of sea turtle ecology, both hatchlings
from the natural nesting ground at Tortuguero, Costa Rica and hatchlings from Tortuguero eggs that
had hatched in a beach on Bermuda were tracked individually after their departure from these beaches.
Of the Bermuda beach most were tracked by swimmers equipped with face mask and flippers and
followed by a boat. Tracking off Tortuguero was done by an observer in a following boat. The data
showed that non-random departure courses were maintained even when the swimming hatchlings had
moved over the horizon from all fixed objects on the shore. Observations were made on swimming and
diving behaviour and on predator relationships of travelling turtles. The procedures described are
useful research techniques and will be used for more extensive tracking in future seasons.

One of the important remaining gaps in the travel steadily for at least 24 hr after leaving the
natural history of sea turtles is the so-called beach. It seems likely that if the initial course
'lost-year mystery', the lack of information on taken could be determined by tracking, this
the range and habitat of the young of any species might provide clues to the ultimate destination
from the time they enter the sea until they re- of the young. The present paper presents data
appear as yearlings. Insofar as there is any bearing on this problem.
theory to explain the puzzle, it is that the time of
early development is passed in drifting masses Methods
of sargassum weed (Carr 1966). Although a Bermuda
few hatchlings, all loggerheads (Caretta caretta), The work in Bermuda was carried out in
have actually been found off the Florida coast 1970 and 1971. The hatchlings used were allowed
in sargassum weed (Caldwell 1968), the theory is to move naturally to the sea after emerging
still a long way from substantiation, and the from the sand in which the eggs were incubated.
means of testing it are few. The present paper The hatchery was established on Nonsuch
presents information bearing on the above theory Island and Howard Bay as part of an effort to
and other aspects of the behaviour of hatchlings test the possibility of rehabilitating the green
during the first stage of their marine existence. turtle populations of the islands. Dr It. C.
The cues that guide newly emerged hatchlings Frick and the author with the collaboration of
from the nest to the sea have been thoroughly Professor Archie Carr and the Bermuda Depart-
characterized during the course of more than ment of Agriculture and Fisheries had developed
50 years of investigation. However, this tropo- this as a pilot reintroduction project. Each
tactic orientation in response to brightness summer several hundred eggs were transported
differences between the landward and seaward from Tortuguero to Bermuda where the young
horizons is of no use to a turtle in the open turtles hatched and were released in the sea.
ocean with no landward horizon in sight. Initial releases in 1968 suggested that the
Readers may gain access to the sea-finding travel of the hatchlings after passing through
literature via the following citations, but should the surf remains steady and oriented. Trial runs
keep in mind that this phase of the life cycles in 1970 showed that it was feasible for a good
of sea turtles is distinct from and of limited swimmer to travel indefinitely behind a hatchling
relevance to the open-sea migration described without either losing sight of it or permanently
in this paper (Parker 1922; Cart & Ogren 1960; influencing its orientation or behaviour. In all
Ehrenfeld 1968; Mrosovsky & Shettleworth tests, hatchlings were held from 2.7 to 25 hr
1968). in covered buckets lined with a piece of damp
The well known swim-frenzy, in which con- cloth or paper, until the trackers and boats
fined hatchiings move constantly against the were ready. Under natural conditions, hatchling
walls of a tank for days after they are taken from green turtles usually emerge from the nest
the nest, suggests that young sea turtles probably and go to sea at night, or just before dawn.

849
850 ANIMAL BEHAVIOUR, 24, 4

Visual tracking at this time would obviously against which the turtles are headed, and the
be impossible, because the hatchlings tend to presence of frigate-birds (Frigata magnificens)
orient toward an extraneous source of light, and which cause the turtles to dive.
using a lantern or a flashlight would therefore Once the young turtles had gone through the
be ruled out. Moreover, tracking at night would surf and had swum until they were clearly out
be dangerous for the swimmer since sharks of visual contact with each other, a hatchling
are common inshore after dark. However, was assigned to each of several trackers. The
natural emergences of young turtles do occur tracker, equipped with mask and flippers,
in the daytime, and daytime observations are followed his hatchling at a distance of 1.5 m to
therefore regarded as a valid record of natural 3 m. At this distance, the swimming turtle was
behaviour. easily visible against the sky from depths of
Hatchlings were released in groups on the approximately one and a half metres. To deter-
beach and allowed to crawl to water without mine whether the presence of a tracker influ-
interference. Distances travelled on the shore enced the continuity of direction of travel of the
were 6 m or more. The mean temperature of the subject, test runs were made in which the
Bermuda water for October is 22~ which is swimmer changed positions from right to left, or
about six degrees cooler than the waters off the moved up close behind or beside the hatchling.
ancestral nesting beach at Tortuguero during None of these manoeuvrings produced more
the hatching season there. This difference in than transitory changes of course. Although a
temperature should influence the swimming turtle's swimming rate could be accelerated by
speed of hatchlings. However, the average the swimmer if he approached the turtle
swimming rate of hatchlings off Bermuda un- rapidly from behind, even when a tracker
expectedly was higher than at Tortuguero. As is moved up closely and held a hand in front of
discussed later, this difference can be accounted the swimming hatehling, the interruption of
for by strong longshore current off Tortuguero, locomotion was only momentary, and the turtle

I
BERMUDA

A ~"'~"~ "~-"~ ~

, _ . , ~_~,~ . , , ,

', | ...... ;.. oo ',,

~ -
KILeUETERS PL W
64040'

Fig. I. Bermuda. The arrows show general directions taken by groups of

hatchlings released in 1971 at four sites, as follows: A. Nonsuch South
Beach; B. Nonsuch North Beach; C. Shelly Bay Beach; D. Long Bay Beach.
 FRICK: GREEN TURTLE HATCHLING ORIENTATION IN TIlE SEA 851

regularly resumed travel in the original direction. single turtle was tracked in 1970), Nonsuch
If the swimmer blocked the path of the turtle South Beach, Long Bay Beach and Shelly Bay
by swimming in front of it or to the side, it Beach face the open ocean. The fourth, Nonsuch
would increase its speed and make every North Beach, is in a small channel between
effort to regain its original heading. Castle Harbour and the ocean. Seventeen turtles
Boats followed close behind the trackers at all were tracked by swimmers followed by boats for
times. To learn whether the presence of a boat 1 to 4 hr. Nine were lost within the first hour.
influenced locomotion and heading, the tolerance
tests made by the trackers were repeated, with Tortuguero
similar results. Even when a boat was stopped At the Tortuguero nesting ground three turtles
directly across the path of a swimming turtle, were tracked just after being excavated from
it merely dived, came up on the other side, and nests that were ready to emerge naturally on the
continued on the original course at the same shore. All hatchlings were released on wave-
steady pace. washed sand 3 m or less from the sea because
The course of each turtle was plotted from the black volcanic sand of the beach was ex-
line-of-sight fixes made from the boats on tremely hot from the sun. Because of the shark
prominent objects on shore every 10 to 15 rain. menace there, tracking was done from a boat
The shore was visible from the deck of the boat in which a person operating a small outboard
long after it had disappeared at turtle-eye level. motor in the stern was guided in the path of the
During the seasons of 1970 and 1971, turtles hatchling by an observer in the bow. Successive
were released on five different beaches (Figs 1 to bearings of the positions of the boat were re-
6). Only two of these, Nonsuch South Beach corded at approximately 5 to 10-min intervals
and Howard Bay, are hatchery sites. Hatchlings by the observer with alidades separated by a
released on the other three beaches were trans- 750-m base line on shore. The approximate
ported there by boat. Howard Bay (where only a

J
o i
KILOMtTI~R

d:5
c,///
/,;',
,
!:,
i! e I i \\ %
\ \
,/, I ', \ \
ss
,// I t
,, %

-- t I II I I
I I # 5
t I I I
t
, I / i
A

t I
, I
', I

t I
2 I
4

Yg
X 164~
I b 4 ~ 40'

Fig. 2. Approximate courses of hatchlings released Fig. 3. Approximate courses of hatchlings tracked for
off Nonsuch South Beach and Howard Bay Beach in 6.00 km to 6.48 km from release point on Nonsuch
1970. South Beach, 1971.
852 ANIMAL BEHAVIOUR, 24, 4

speeds of a south-trending longshore current regardless of direction, with what seemed clearly
were measured by timing an empty corked non-random accuracy and consistency, None
bottle as it floated a known distance. veered more than 5 ~ to 10 ~ within the first 1.50
km. However, when tracked beyond this dis-
Results tance, six were deflected, probably due to
Bermuda currents, although these were not measured.
The travel paths. Of the eighteen turtles tracked When a course was obstructed by breaking
off the beaches facing the open ocean, six held reefs, masses of drift or the tracker-tender boat,
courses approximately at right angles with the the turtle in every case swam over, under or
shore of release, ten diverged at angles of from around the obstacle and continued on course.
5 ~ to 35 ~ east and west of this course, and two Once in the sea, the hatchlings of a given batch
circled aimlessly in Nonsuch Bay. Of the eight never remained together or appeared to influence
turtles released on Nonsuch North Beach, where each other in any way. Distances covered and
the outlook is cluttered with rocks and promon- directions taken are indicated in Tables I and II.
tories, three swam into the harbour against the Swimming rate. The average swimming rate
tide, and five swam toward the open sea. Although for the twenty-four turtles tracked for 0.60
six of the eight turtles tracked for more than 2 to 6.48 km was 1.57 k m per hour. This is much
km veered as much as 40 ~ east and west of their slower than the sudden spurts of speed that
initial headings, in the case of Nonsuch North captive hatchlings sometimes display, but faster
Beach where the turtles had a choice between than the deliberate paddling they do in foraging.
the open ocean and the harbour, the departure Breathing interval, swimming depth and diving.
heading was considered the course taken after At 5 to 10 s intervals the swimming hatchlings
the initial choice between these two opposing would rise to the surface, tread water momen-
directions had been made. Once having estab- tarily with all four limbs, breathe and then
lished a course, all the hatchlings held to it, swim on at an average depth of 20 cm, diving at
varying intervals to approximately three metres.

64o45 ' ]
64*50' I
01•2 o i 2
~ILO~ETERS
K*LOMETERg

fro
~" N~ -

"4,\ o'"

Fig. 4. Courses of hatchlings released at Shelly Bay Fig. 5. Courses of hatchlings released at Long Bay
Beach, 1971. Beach, 1971.
 FRICK: GREEN TURTLE HATCHLING ORIENTATION IN THE SEA 853

Resting and lodging. Most of the hatchlings

swam steadily throughout the observation
KILOU~T~ I periods. Turtle E stopped briefly at 14.00 hours,
after 5 km of steady swimming, and floated
. ; L \ <-'h on the surface with the flippers pressed fiat
along the carapace in the sleeping position.
Turtle S, (Plate XII, Fig. 7) after swimming for
.; ? ','2*t, L 5.86 km, crawled up into one of several patches
of sargassum weed and stayed there, swimming
----t II ! U I. . "* 't
from one small clump to another for approxi-
o;. iI ~ l.~.~'l~ mately one-half hour, after which contact with
,." .. o // 'l'l I t '. it was lost during a change of trackers. A
U .'! careful search of the weeds where this hatchling
//" ! ',\", was last seen, and of nearby patches, suggested
that it had suddenly left the vicinity; and though
L/ v
;,',; I t

.
.

.,,':,-"
I%
we searched out along the way it had previously
been swimming we were unable to re-establish
contact. This hatchling could have been taken
d ./ ,. by a predator. In 1970 another turtle tracked
! / I
I
off Nonsuch South Beach lodged in a small
patch of weeds after swimming for 2.30 km.
I/1 i t
Predation. Although it is generally assumed
I | o
that predation by fish is extremely heavy during
the period after hatchlings enter the sea, few
s ~ t definite observations have been recorded. During
tt ~ F
the present experiments, two of the turtles
e
tracked were lost to fish. Turtle B was seized by
a belonid (needle fish) Tylosurus sp., but was
ii lllll ~B
dropped when the tracker swam up and frigh-
tened the fish. Turtle K was eaten by a small
barracuda, Sphyraena barracuda. Two other
J 64*40'
turtles were approached and nudged by bonitos,
Seriola rivoliana, which quickly retreated as
Fig. 6. Courses taken by hatchlingsreleasedat Nonsueh the tracker approached. In other cases, fish
North Beach and Nonsuch South Beach, 1971. of various kinds swam beneath turtles but made
no move toward them.
Dives were sometimes caused by the approach Observations on turtles released directly into
of a boat. A similar reaction to over-flying the sea. Five hatchlings were released directly
aircraft suggested that hatchlings dive to escape into the sea, out of sight of land at turtle eye-level
predatory birds, which later was found to be (approx. 5 km) and without prior experience in
true at Tortuguero. In most dives the hatchling crawling on the beach or swimming. Immediately
resumed horizontal swimming at the maximum after release these hatchlings dived at a nearly
depth attained, until it again rose vertically vertical angle to depths of up to 3 m, sometimes
to the surface. The maximum duration of such
dives was 2.5 min. swimming aimlessly in circles or zigzags at this
Feeding. One aspect of the lost-year puzzle depth. After approximately one and a half minutes
is the uncertainty whether and how hatchlings they would struggle to the surface, paddle in
feed in the open sea. None of the turtles tracked circles breathing rapidly with their heads above
in the present tests had been fed prior to release; water, then repeat the same diving behaviour.
all, in fact, had fresh umbilical scars. The only Because their erratic behaviour made it difficult
ease of feeding observed during a tracking run to follow them, all but one were lost within the
was that of turtle E, which stopped swimming, first half hour of tracking. This one turtle was
dived down 1 m, and systematically ate a small observed to take an oriented heading for 10
comb-jelly. min after 0.5 hr of circling.
854 ANIMAL BEHAVIOUR, 24, 4

Table I. Preliminary Trials Off Nonsuch South Beach and Howard Bay Beach in 1970

Time Approximate
tracked distance
Approximate (Eastern tracked
Turtle Release site Date course Standard) (km)
1 Nonsueh S 18 Oct. Circled 12-00
(hatchery) aimlessly in Bay 12.40
2 Howard Bay 19 Oct. S.E. 14.30 1.18
(hatchery) 15.00
3 Nonsueh S 20 Oct. S.W. 11.30 1.82
12.40
4 Nonsuch S 20 Oct. S.W. 11'30 2.30
12.40
5 Nonsuch S 21 Oct. S.E. 10.15 1.48
11.10
6 Nonsuch S 21 O c t . Circled 10.15
aimlessly in Bay 11.00
7 Nonsuch S 21 Oct. S.W. 12.00 1.98
13.00

Tortuguero b y either a swimmer or observers in a boat.

Results of the tracking at Tortuguero are The data gathered so far, though not extensive,
shown in Fig. 8 and Table III. Although the suggest some generalizations that may bear on
travel paths recorded seem obviously non- the lost year puzzle in sea turtle ecology as
random, they were evidently influenced by a follows: (1) Green turtle hatchlings swim
southerly longshore current measured at speeds steadily on an oriented heading after crossing
from 0"37 to 0.65 km per hour which affected the surf and passing reefs or other shore zone
the speed of the turtle when it was swimming obstructions. (2) The departure courses are
against the current. The average speed for all approximate extensions of the paths taken by the
three turtles was 1.49 km per hour. Further work hatchlings in crawling to the sea from the nest.
is necessary in order to determine the extent of (3) The headings are not maintained by visual
this influence beyond the first 2 km at sea. reference to a fixed object on shore, although
Unlike the situation at Bermuda, the diving what determines a departure direction is still
behaviour and surfacing intervals o f the Tor- unknown. (4) The departing turtles swim
tuguero hatchlings travelling during the day- steadily, near the surface, mainly at depths of
light hours are affected by the predatory at- 20 cm or less. (5) Although the turtles may pause
tention of the frigate-birds that usually occur in their swimming to feed, only one case of
off a natural rookery and are present at Tor- feeding has been recorded, and the main drive
tuguero. Although no underwater observations at this stage appears to be to get beyond the
were possible because of the sharks that abound reef area to offshore regions o f lower predation.
in the area, it was possible to keep the hatchlings (6) The turtles are able to dive at depths o f 3 m
in sight from a position far enough away from or more, which may allow them to escape from
them that the frigate birds would attack freely. predation by birds. (7) Predation by fish occurs
The hatchlings showed obvious evasive activity, most frequently when the hatchling is crossing a
diving quickly with each near approach o f the reef area. (8) Travel may be temporarily inter-
bird. During these periods, the swimming rate rupted, even in daytime, for rest. During such
decreased. periods the turtle floats with flippers flattened
on the edges o f the carapace, in the typical sleep-
Discussion ing attitude. (9) Refuge is taken, at least occa-
The observations recorded here prove the sionally, in sargassum rafts encountered along
feasibility of the direct tracking o f hatchlings the way. (10) Unlike hatchlings released on a
 FRICK: G R E E N T U R T L E - H A T C H L I N G O R I E N T A T I O N I N THE SEA 855

Table II. Trials Off Four Sites of Release in 1971

Approximate
Hr between distance
Emerging Approximate Time tracked
Turtle Release site Date 1971 and release courses tracked (km)
A Nonsuch S 16 Oct. 15 S.E. 12.20 6.00
then S. 15.30
B Nonsuch S 16 Oct. 14,25 S.W. 12.05 2.38
(hatchery) then S,E. 14.00
C Nonsuch S 16 Oct. 15.5 S.W. 12.40 2.63
(hatchery) 14.30
D Nonsuch S 17 Oct. 25 S.W. 10.15 1.82
(hatchery) then S.E. 11,45
E Nonsuch S 17 OCt. 25 S. 10.35 6.48
(hatchery) then S.E. 14.30
F Nonsuch N 18 Oct. 14.25 S. 11.05 2.14
12.35
G Nonsueh N 18 Oct. 14.25 N. into harbour 11.05 1.36
then S.E. 12.30
H Nonsuch N 18 OCt. 14.25 S.W. 11.05 1.47
12.05
I Nonsach N 18 Oct. 14.25 N. into harbour 11.05 1 "42
then S.E. 12.30
J Nonsuch N 18 Oct. 14.25 S.E. 11.05 1 "49
12.05
K Nonsuch N 18 Oct. 14.25 S. 11.05 0.70
11.45
L Nonsuch N 20 Oct, 19.25 N.W. 14.15 0.60
then S.E. 15.00
M Nonmach N 22 OCt. 18 S.E. 15.15 1 "48
16.00
N Shelly Bay 23 OCt. 11 N.W. 11.00 1.50
12.20
O Shelly Bay 23 Oct. 12.5 W. 12.35 1.68
then N.W. 13.30
P Shelly Bay 23 Oct, 13.5 W. 13.40 5.16
then N.W. 15.30
Q Long Bay 24Oct. 2.7 N. 10.10 1.50
11.00
R Long Bay 24 Oct. 2.7 N.W. 10.10 1 "52
11.15
S Long Bay 24 Oct. 2.7 N.W. to 10.10 5.86
N.E. to N, 13.15

Table HI. Preliminary Tracking Trials off Tortuguero Beach Costa Riea, 1974

Current direction
Approx. and velocity
Approx. Time distance (nautical miles Surfacing
Turtle Date course tracked (kin) per hr) intervals
I 24 Sept. S.E. 9.17 1.65 0.65 Every 7.2 s decreasing to
10.22 4 s when frigate-birds
appeaxed.
II 25 Sept. N.E. 10.50 1.22 0.37 Not recorded.
then E. 11.50
III 26 Sept. S.E. 9.05 1.70 0.65 Every 8"3 s.
10.07
856 ANIMAL BEHAVIOUR, 24, 4

beach and allowed to go through the surf in a Carr, Peggy Hammond, Eric and Brooke
natural way, those released directly into the Herter, David Braga, Jim Burnett-Iterkes,
sea appeared frantic and completely disoriented. David Lonsdale, Simon Holland, Penny Salter,
This behaviour suggests that the experience of a Mr and Mrs Richard Slaughter, Rob Seaver,
natural ontogenetic sequence of orientation and Jesse Goodyear, David Wingate, Campbell
locomotor responses may be necessary develop- O'Connor, Deloy Robinson, Alex Shoumatoff,
mental grounding for the performance of later and my parents, Dr mad Mrs H. C. Frick. For
and quite different functions. assistance at Tortuguero, I would like to thank
In sum, the observations appear to indicate Chuck Carr and Judy and Josh Polan. I am
that hatchlings engage in active, energetic, grateful for the the information Dr Joseph
oriented travel for at least 4 hr after they enter Seronde provided on technical aspects of coastal
the sea. The fundamental adaptive reason for navigation. I would also like to thank Shirley
the juvenile travel-drive may be, as has been Robbins and Dawn Brown for help in typing
suggested, to reach longshore currents in which the manuscript. Professor and Mrs A. Carr
sargassum rafts serve as a refuge and feeding helped with the tracker-tender during the
place. Bermuda experiments. Professor Carr and Dr
David Ehrenfeld made helpful suggestions
Acknowledgments regarding organization of the manuscript.
I want to thank the many who assisted in In Costa Rica, logistic assistance and facilities
tracking hatchlings off Bermuda. In particular, were provided by the Caribbean Conservation
I would like to thank Steve, Tom and David Corporation, and by the Oceanography Section,

CERROI"ORTU6UERO
BOCA DE LAS LAGUNAS
^ DEL TORTUGUERO SCALE
!
I :==1
Ikm

/,,

11.50
Ll:40 ........ I
..am'"'"
m'""
-
10t02 lO:lT~,,.,.~l~
10:22

9:22
"'e-.. 9~47
"'..9"55

"'./0:O7

| TRIANGULATION STATIONS
LANDING STRIP
TRACK I
9 TRACK 2
9 TRACK 3 I~

83 ~ 50'

Fig. 8. Courses taken by hatchlingsreleased on Tortuguero Beach, Costa

Rica, 1974.
 FRICK: GREEN TURTLE HATCHLING ORIENTATION IN THE SEA 857

N a t i o n a l Science F o u n d a t i o n , N.S.F. G r a n t Carr, A. F. & Ogren, L. (1960). The ecology and migra-
G A 36638, u n d e r which closely related research tions of sea turtles. 4. The green turtle in the
Caribbean Sea. Bull. Am. Mus. nat. Hist., 121,
is being conducted. 1-48.
This study is catalogue No. 6 5 5 - - B e r m u d a Ehrenfeld, D. W. (1968). The role of vision in the sea-
Biological Station for Research. finding orientation of the green sea turtle (Ckelonta
mydas). 2. Orientation mechanism and range of
spectral sensitivity. Anita. Bekav., 16, 281-287.
Mrosovsky, N. & Shettleworth, S. (1968). The wave-
REFERENCES length preference and brightness cues in the
Caldwell, D. (1968). Baby loggerhead turtles associated water finding behavior of sea turtles. Bekaviour,
with Sargassurn weed. Q. Jl. Fla. Acad. Sci., 31, 32(4), 211-257.
271-272. Parker, G. H. (1922). The crawling of young loggerhead
Carr, A . F. (1966). Adaptive aspects of the scheduled turtles toward the sea. J. exp. ZooL, 6, 323-331.
travel of Chelonia. In: Animal Navigation and
Orientation (Ed. by R. M. Storm), pp. 35-55. (Received 27 June 1972; revised 14 May 1974;
Corvallis, Oregon: Oregon State University second revision 1 April 1975; third revision 17 September
Press. 1975; MS. number: M344)
 FRICK: G R E E N T U R T L E HATCt-ILING O R I E N T A T I O N IN THE SEA

/PLATE XII

Fig. 7. After swimming seaward for 5"86 km from the beach on the northwest coast of Bermuda
a green turtle hatchling approaches (upper), climbs into (lower) and rests in a large patch of
sargassum weed. Photograph credit to Richard Slaughter.

Frick, Anim. Behav., 24, 4