Crab island revisited: reassessment of the world's largest Flatback Turtle rookery after twelve years

13 Pages·1993·4 MB·

by  C L Limpus

CRAB ISLAND REVISITED REASSESSMENT OFTHE WORLD'S LARGKST
FLATBACK TURTL.E ROOKERY AFTER TWELVE YEARS
COLIN J LIMPUS, P.J. COUPERAND K L.D. COUPER
Liinpus,CJ.»Couper,P.J.&Couper,K.L.D. J9930630:CrabIslandrevisited:reassessment
ofiheworld'slargestFlattackTurtlerookeryaltertwelvevears.Memoirsofthe Queensland
Museum33(1)277-289. Bnsbnne. ISSN0079-8835.
Cr3b Island in northeastern GulfofCarpentaria supports the largest nesting population of
Natator depressus, a marine turtle endemic to the Australian continental shelf, and low
density nesting by Eretmuchelys imbricate. The reproductivestatusofthe Flatback Turtle.
Natarordepressus,at Crab Island, isreassessed alter 12 y, .., , ..Ion B surveyconducted
duringhighdensity'nestinginJuly 1991-N. depressushatchlingproductivityfromtheisland
continues to be high. The characteristic small size of nesting females and egg diametersof
N. depressus that breed in theCrab Island regionsuggests thai this population isadifferent
bieeding unit from that ofthe southern Great Harrier Keef The feeding areas stippl
turtles to the Crab Island region rookeries extend as far north as southern IrianJaya.
Namtordepressus, Erevnocheiys imhricata. CrabIsland, Que*\ \ustialm, nest-
ing* hatchlingpmductivtty. conservation Status,
ColinJ. Limpus, Department uj Environment andHeritage, POBus t55i Brisbane. AH
Street, Queensland4002,Ansouhu. H.J, andK.LJ). Couper, QueenslandMust urn. f'O8t?>
3300, Saudi Brisbane, Queensland 4J0I, Australia; 27October, 1992
Crab Island, in the north eastern GulfofCar- team countedall existing turtle tracks by spei
pentaria, supports the largest recorded nesting without attempting lo age the tracks. A sil
aggregation ofFlatback Turtles. Ncttatordepres- track included both the emergence and dfi
&
sus (Liinpus Pamienter, 1986). N, depressus is crawl ofa nesting turtle. All tracks were crossed
almost totally confined to the Australian ccHu> off in the sand above the high tide mark as they
Rental shelfand its breeding is restricted lo Aus- were counted so that previously recorded tracks
tralia(Limpuseta). 1988) Thebiologyofme could be recognised. Thereafter, a track census
turtle populations breeding on Crab fsland was was conducted daily: 'all turtle tracks and nests
reviewed by Limpus el al, (1983a) based on field from which hatchlings had emerged from the
Surveys of the nesting turtles in 1970-1979. Be- previous night were counted by species along the
cause the status ofibis significant turtle breeding western beach. Because two persons could not
population had not been reassessed since that monitorthe nesting behaviourofall turtlesas
time, a study to redescribe the reproductive biol- for a night along 6km of beach, a subset of the
ogy of the Crab Island turtles after a 12 year beach sectors (4 to 6) was selected for nightly
interval was undertaken in July 1991, during the measurement of nesting success of each tuitL
high density nestingperiodidentified in previous
Potential predators of the turtles, their eggs ot
studies (Limpus et ai. 1983a).
hatchliftgS were identified and quantified where
ible. Thisincludedcounting ofcrocodile and
Methods bird tracks. Turtles were recorded in the waters
A two person team camped on Crab island adjacent to the nesting beach during the daily
.luring 6-22July 1991 (P.J C and K.L.D.C). The census oftracks on those days when the weather
western ocean)beachoftheisland was measured was relative!} calm. As well, a male turtle was
along thie spring high tide level and subdivided captured bv 'beach jumping' lef Limpus &
into eleven numbered sectors (each 500m hmg) Reed. 1985)
fAlrlomdatsaourtehcolrodneodratlho(nFgigt.heIi)sluasnidngwearepesdcoomreetderb.y ocOcpuprorretduwnihsetnicvoLlaugngtienegromfenmebsetrinsgoffetmhaelQeuleuerlnlses
beach sector.The beachwidth withinsectors 1-10 land Turtle Research Ptojecl visited Crab Island
was measured from the spring high tide level to (28 December 1989and 15-17January, 1991 (and
the crest of the seaward dune adjacent al each the south beach al the mouth ofthe Jardine River
Sector mark. These were the seCtOR m whirl. -M on the mainland, 14km from Crab island (29
turtle nesting activity occurred. On arrival, the ruber 1987). Adult turtles weretagged using
r '
278 MEMOIRS OFTHEQUEENSLAND MUSI
serially numbered, selfpiercing, selflocking tags (sector6). There were nomangroves growingon
applied to ihe axillarv flipper tagging position ihe Western beach However, therewerestanding
(Limpus&Rced, 1985). InJuly l991,33gmoncl dead mangrove trunks on the northern part ofthe
il Band and Tag Company, Style western beach that were the result of pas! en-
1005, size681; SD=0.0ig ange - 3-2S-3.29, n croachment ol the island into the mangrove for-
= 50), inscribed with a return address on the est.Thedead mangrovesandbeachrockwerenot
reverseside(Ecology, Box 26Woden 2606Aus- an impediment to the nesting activities of the
tralia), were used. On the o4her visits. 4.1g tita- turtles.
nium turtle tags (Stockbrands Co P/L.. return Crab Island lies within a region with a distinct
address: Wildlife Box 155 NorthQuay 4002Qld summer wet season and winter dry season. The
Australia; Limpiis, 1992) were used. The turtles mean monthly rainfall and the mean daily mini-
weredouble tagged,one tag in each from Hipper mum and maximum temperatures for the years
No attempt was made to tag all turtles ashore in 1950-1991 at Thursday Island ranged front 4mm
any one night.Themeasurementstaken followed in Septemberto 419mm inJanuary (Fig. 2). Crab
thestandardmethodologyoftheQueenslandTur- Island has no surface freshwater for most ofthe
tleResearchProjectascksenbedby Limpus et aJ. year. There is an ephemeral freshwater swamp
(1983a): midlinecurvedcarapace length (CCL): system behind the dune in the northern end of
clutch count; egg diameter, egg weight (10 eggs sector 5 which supports at least one species trf
randomly selected per clutch); nest depth to the frog, Ufnnodynwtes ontatus (QMJ3 1 15 '
bottomoftheeggchamber: incubation andemer- This swamp was dry during the July visit and in
gence success of dutches: sand temperature at most years there would he insufficient rain to
50cm depth. San: _ and recently dead refill the swamp until January, which is the peak
hatcbljngs were collected and deposited in the of the wet season (Pig. 2). Monthly mean daily
Queensland Museum. maximum temperatures ranged from 31.2°C fo
Air temperature, rainfall and wind data have Novemberto27.7°CinJuly, while monthly mean
been obtained from the Bureau of Meteorology daily minimum temperatures ranged from 25 4"f
(Brisbane) for their weather station at Thursday in Decemberto22.5nCin July. Duringthe middle
Island (lO^S'S, I42°LVK, 60m elevation. 45km ol wetseason monthsofJanuary andFebruary the
n dab Island) This is the closest weulbei wind is predominantly from the west and north-
station to Crab Island for which there is a com- west and is mostly weaker than 20km/hr. Jn the
prehensive data set. Tidal data ior Booby Island middle ofdry season months ofMay to October
was used (Anon, 1990), Kuohy Island, 47km 10 Mien/ ait almost no westerly or northwesterly
the north west is Ihe closest standard port tidal winds; thedryseason windscomepredominantly
botch m.nk across open *at«i toCrib island
LU Mangroves L_ Woodland
RESULTS
ii£3 Openforest LJ Opensand
Study Site
Crab Island CI0DS9'S( I42°06'E. Fig. 1) is a
erescentic sand island that measured 6.2km in
length along thehigh lide level oftheouterwest-
ern beach on 6July. The inner eastern margin ol
the island was partly mangrove lined with wide
inter-tidal mud Hals. Tnc western beach was
c
posed of mixed siliceous sand and calcium eai-
bonate (broken mollusc) fragments and was
exposed to surf. Beach rock was exposed intern
dally in sectors 4 and 5. The vegetation of the
foredunenestinghabitatofthewestern beach was
variable: Melaleuca woodland with a grass and
500metres
herb undcrstorcy in sectors 5 and 6: open wood-
landofscattered ( (isuarinn equisetifolta (sectors
4 and 7); grassland dominated by Spinife.x hirsu-
t/t\, fpomoea pes-capr&e and Trihulus nsundes FIG, 1. Map ofCrab Island showing position ofnum-
(sectors 2-4, 5, 8-10), unvegetated sand dune bered beach sectorsand vegetation types
REASSESSMENTOFTHE LARGESTTURTLEROOKERY 279
500 35
M
30
M
400
e
25
n
300
20
m
15
200
10
100
I
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
^B
Rainfall (mm) Maximum temp. C)
(
Minimum temp. C)
(
FIG. 2. Monthly mean rainfall (mm)andmonthly mean dailyminimum and maximumtemperatures recorded ftl
Thursday Island Meteorological Station Bar the years 950-1991.
from the east and southeast and are mostly beachings of E. imbricata per night (SD = 0.93,
strongerthan 20km/hr. Mean sand temperatureat n = 14, range = 0-3) fora lotal of9 beachings.
50em depth within Ihe turtle nesting habitat was On the limited data from this visit, there is a
27.6°C (SD=0.49, range =26.4 - 28.5, n =44, ] suggestion that the maximum nightly nesting
- 5 measurements per night). density may follow the occurrence of afternoon
The mean beach width above the spring high high tides. Only N. depressus was recordedcom-
tide level was 31.3m (SD = 15.3, n = 10). The ing ashore for nesting during daylight hours (19
narrowest beach (8m) occurred at sector mark 3 1.0%] ofthe 1839 total beachings). Most (4 and
and thewidest(53m)atsectormark 10. Thereare 113) Of these daylight nesting emergences oc-
two high tides per 24hr at Crab Island (Fig. 3). curredon twodaysonly(8thand9lhJuly, respec-
During 6-22 July, 1991. the daily tidal range tively) with the remainderbeingsinglebeachings
varied from 3.46m (12th) to 2.36m (19th). The
oneach ofthe 10thand 20thJuly. All exceptone
variation in night time high lide height (0.27m)
ofthe daylight beachings occurred on days with
was less than that of the daytime high tides and
both sets oflow tides (Fig. 3). mid-afternoon high tides (Sth-lOlh) while ihe re
maindercoincidedwithamiddayhightide(20th).
Turtle nesting activity was restricted to the
NestingTurtles western (ocean) beach, except for one turtle
On the team's arrival at the island on 6 July, which came ashore on Ihe western side at die
d1e9p9r1e,ssthuesr,e1w3e5reno3t0i9detnutritflieatbrlaectkossvpiesciibelse).(1B7a4seN.d ntuarrnreodwtomtihdedlseeaoofnIthheeeiassltaenrdns(isdeec.toMro6s)tnaensdtirneg-
on nightly track counts from the 14 nights 6 19 activity occurred within the middle sectors 4-8
July(Fig.3),therewasanaverageof 132.7 teach- (Fig. 4) where the beach width ranged 26-43m
ings ofN. depressus per night (SD = 52.1 n = Thenumberofturtlebeachingsvariednightlyand
1,
14. range=68-235)from atotal of 1839 recorded among sectors. The densest activity occurred in
beachings. Similarly, there was an average of0.6 sector 5 (nightly mean number of tracks = 43.9,
1
280 MEMOIRS OFTHEQUEENSLANDMUSEUM
250
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Date (July 1991)
FIG. 3. Daily frequencyofbeachingsby nestingFlatbackTurtles,Natatordepressus, atCrabIslandduring6-19
July 1991. Thedaily variation in tideheight (in metres), measured at Booby Island is also shown. *denotes a
nighttimehigh tide.
SD = 25.49, n = 14, range = 21 - 102) while no and 6 £. imbricata). There were no recaptures of
tracks were recorded on any night in sector 11, any turtles that had been tagged during previous
Forthe sampleof 172 beachings by N. depres- visits to the island, nor were there migrant recap-
sus examined for nesting success, 132 females tures ofany turtles tagged atotherlocations. The
(76.7%) definitely laid eggs, 14(8.1%) could not size distribution ofthese nesting turtles are sum-
bedeterminedfornestingsuccessand26(15.2%) marised in Table 1 and Fig. 5. The mean CCL of
returned to the sea without laying eggs. Thus the nesting turtles was: N. depressus = 88.2 cm,
there was aprobability (±95% confidence limits) E. imbricata = 83.9 cm.
ofbetween76.7±3.2% and 84.8±2.8% that atur-
tle would successfully lay eggs when she came Turtles AdjacenttoBeach
ashore. When this nesting success probability is By day when theweatherwascalm during 6-21
applied to the track counts for the period 6-19 July 1991, there werenumerous sightings ofsoli-
July, it is estimated that between 1352 and 161 tary turtles surfacing to breathe, mostly within
N. depressus clutches were laid on the island 50mofthebeach. All wereadult sizedN. depres-
during the 14 nights ofobservation. sus. Upto 15 such turtles wereobservedadjacent
Forthe sample of6 beachings by E. imbricata to a single sector during a track census with 0-49
examined for nesting success during the same isolated turtles seen during a single track census
period, 4 laid and 2 did not lay. This represents a along the entire beach. The water was clear
nesting success of 67% for this species with an enough at high tide for adult sized N. depressus
estimated total of 7 clutches laid during the 9 resting on the bottom at 2m depth to be visible
beachings. from a dingy adjacent to the beach. In addition,
Overthe 15 nights, 6-20July, 489 nesting tur- mating turtles were observed in the surfadjacent
tles wereexaminedandtagged(483 N. depressus to the beach on four days (0-2 courting pairs per
REASSESSMENTOFTHELARGESTTURTLEROOKERY 281
TABLE 1. Curvedcarapace length (cm)ofnesting femaleturtlesonCrabIslandandadjacentbeaches sincethe
studies ofLimpusetal. (1983a).
Species Location Date Mean SD range n
15-16Jan 1991 90.3 2.49 85.0-95.5 18
CrabIsland
Natatordepressus 06-20Jul 1991 88.2 2.80 77.0-95.6 315
mouthofJardine River 29Dec 1987 88.8 0.47 88.5-89.5 3
— —
Eretmochelysimbricata CrabIsland 28 Dec 1989 86.0 1
06-20Jul 1991 83.9 2.41 79.5-86.5 6
daily trackcensus). All were mounted pairs ofN. venter to dorsum (Fig. 8), and was gripping the
depressusandno attendant males wereobserved. female's carapace margin viathe single recurved
One mounted male (B0314) was captured and clawoneach flipper. Themalewas mountedwell
measured : CCL= 83.1cm, tail length beyondthe forward on the female such that his head could
carapace = 22.0cm (Fig. 6). The adult male N. reston heranteriorcarapaceand dorsal neck. His
depressuswas similartotheadultfemale(Fig. 7) snoutcontactedherheadwhensheraisedherhead
in most external features: colour, low doming of fora breath.
the carapace along the midline and lateral up-
wards reflexing of the carapace (Limpus et al., AdultMortality, Injuries and Diseases
1988). The only observed external dimorphic No dead turtles were found along the beach on
characters were tail length (male very long vs arrivalon6July.Thehighestandnarrowestdunes
female short) and claws (male elongate and ontheisland wereinsector4. The 10mhighdune
strongly recurved vs female short and slightly frontinthisareawassteepenoughformostturtles
curved). The male was mounted on the female, to have difficulty in climbing it. One female N.
700
Pre 6th July
600
6-19 July
500
400
300
200
100
8 10 11
Beach sector
FIG. 4. Distributionofturtle tracksby beach sectoron Crab Island, 6-19July 1991
282 MEMOIRS OFTHEQUEENSLANDMUSEUM
50
Natator
Eretmochelys
40
30
20
10
J L .
75 80 85 90 95 100
Curved carapace length (cm
FIG. 5. Sizedistribution ofnestingfemaleNatatordepressusand Eretmochelys imbricata atCrab Island, 6-20
July 1991.
depressus was observed to climb this dune and carapacewasrecordedfor8 (1.7%)turtles, minor
attempt to nest inland of the dune crest. Later, carapace damage was recorded for 17 (3.5%)
afterwanderingtothe inlandbaseofthese dunes, turtles and loss ofone third orgreaterofaflipper
shewasabletofindherwaybacktothebeachand was recorded for 32 (6.6%) turtles. All of these
return to the sea. A search ofthe woodland and injurieswere fromwellbeforethecurrentnesting
forest areas inland of the dune crest throughout season as indicated by their completely healed
the island revealed the old fragmented skeletal state. None of the E. imbricata showed signs of
remains of two adult sized N. depressus lying significant injuries.
ventralsidedownbehind sector4. These remains Fibropapillomas were recorded on five of the
were consistent with the turtles having died after nesting N. depressus only: B505, 2 fibropapil-
being disorientedfollowing a nesting attempt in- lomas on neck; B535, 2 on right hind flipper;
land of the crest of the high dunes. No nesting B625, 1 onrightfrontflipper;B690, 1 ontheright
females died while ashore during the 6-21 July shoulder; Bl 165, 1 on neck. The largest fi-
1992 visit. bropapillomawas 2.4cmindiameter. Theventral
Onelargecrocodile,Crocodylusporosus,(total surface was not examined.
length 3+m) was observed on most days during
late mornings at the water's edge in sector 3.
While this crocodile was large enough to be a Clutches
predator ofadult turtles, no turtles showed signs For N. depressus, the mean clutch count was
of having been mauled by a crocodile. Several 55.9 eggs (Table 2, Fig. 9) and the mean egg
nestingturtlesdid,however,havehealedcrescen- diameter was 4.93cm (Table 2). Of the 32
tic damage to the carapace that was presumed to clutches counted at oviposition, one contained
haveresultedfrompastsharkorfishbites. ForN. oneyolkless eggandnonecontainedmultiyolked
depressus,lossoflargecrescenticpiecesfromthe eggs. During the laying of these 32 clutches the
REASSESSMENTOFTHE LARGESTTURTLEROOKERY
40emergedclutches had been laid by aturtlethat
had dug into an existing clutch (clutch distm
bancerateat laying=0.025), Because these latter
eggs were adjacent to another previously
emerged clutch they could not be counted accu-
rately. Counts were made from the remaining 3°
freshlyemerged clutches (Table 3). Therewas no
significant difference (t test, p,«0.05) between
mean clutch count measured at laying (Table 2)
and mean clutch count measured at emergence
(Table 3. Fig, 9), the latter having been laid ap-
proximately two months earlier. The mean depth
to the bottom of these nests was 58.3 cm (Table
FIBG0,361.4,AredtuulrtnimnaglteotNhaetsaelaoraftdeerpbreeisnsgusb,rotuagghtniuismflbie-<rrc: 3o)f.tThheesreeclwuetrceheyso.lkless eggs ( 1 per clutch) in two
lortaggingand measurement.
From the 2129 eggs in the 39clutches counted,
nesting females did not dig into any existing there was a mean hatching success of 81.84%,
dutches. while hatchling emergence from the nest to the
For E. imbricata> the mean clutch count was beachsurfacerepresented 7Js\56% oftheeggslaid
139.3, themeaneggdiameterwas 3.60cmandthe (Table 3). Egg mortaltly was distributed as fol-
mean egg weight was 26.1g (Table 2). None of lows; l().49r unhatched. 7.4% undeveloped and
the fourclutchescountedatovipositioncontained 0.3% predation by ghost crabs, Ocypodt so i3
yolkless or muitiyolked eggs. During none ofthe nests). Within the nests, there were dead hatch-
Five successful nestings observed did the turtle lings representing 2.4% ofthe eggslaid. AfurlIk i
dig into an existing clutch. 0.9% oftheeggs were represented by live hatch-
Representative eggs were collected 6 normal lings that may or may hot have escaped from the
:
N, depressus eggs, QMJ31745; 1 yolkless N. nests, had they not beenexcavated."This included
depressus egg QMJ31747; 3 normal £ imhri- 9 hatchlings that were tangled in a buried branch
cata eggs, QMJ?31746. over one nest.
Egg and Hatching Pkudation
EmergedClutches
There was m evidenceofknownegg predators
Eighty four freshly emerged clutches were such as varanid lizards or large mammals (pigs,,
identified to species by examination of eg dogs, cats) on the island.
hatchlings or hatchling tracks. All were N. de- Nankeen night herons (Nycticorax caledotu-
pressusclutches.Representative hatchlingN. de- cus) including adults and fledgedjuveniles wcie
pressus were collected : QMJ31748-31753. present in the nesting habitateach night and were
eStmreornggedwicnldutschpersevfeonrtaeldl bteheachaccsuercatotrescoonunmtooMf obibrsdesrvweedreeatmionsgthcaotncchelnitnrgatA'e,ddienprbeesascush.sTechteosres
nights. Within the relatively wind protected 4-5. At night, upto6 birdswere observed around
of scctoi 5. the nightly number of emerged a nest taking hatchlings as they emerged. This
clutches was counted on 6-15 July. The mean
number of N. depressus clutches emerging per
night in this sector was 4.1 (SD = 2.23, range -
0-8, n = 14 nights, 57 nests). The maximum
number of emerged clutches counted along the
entirebeachon a nightwith lightwindwas sc\en.
with fourofthoseinsectorfive. Hatchlingswere
oftenencounteredcrossing the beachat night but v.
none were seen by day.
Forty nesls from recently emerged clutches
were dug to assess hatchling incubation and
emergence success. Alter having been laid, two
of these clutches had been each dug into by an-
otherturtle, representing aclutch disturbance rale FIG. 7. A-dmli Fernnl preteuA n
after having been laid of 0.05, Another ol' these the sea after nestiii hiJ, I
284 MEMOIRS OFTHEQUEENSLANDMUSEUM
TABLE 2. Clutch counts and egg measurements re- freshly laid. Thecollectorexpressedtheviewthat
corded fromnesting turtles atCrab Island, 6-20July theseparticulareggs were ofmarginal quality for
1991.* denotes teneggs measured perclutch. eating and only halfofthe clutch was taken and
the remainderreburied.
Species Mean SD range n
clutchcount 55.9 9.57 34-74 32 Other IncidentalData
Natator
depressus egg 4.93 0.173 4.52-5.17 60* Duringavisitonthenightof29December1987
diameter!cmi to the south ofthe Jardine Rivermouth, three N.
clutchcount 139.3 10.30 123-151 4 depressus were tagged inthe first 100m and sev-
Eirmebtrmioccahtealys edigagmeter(cm) 3.60 0.071 3.49-3.77 30* eral others could be seen furtheralong the beach
eggweight 26.08 1.075 24.0-28.0 30* (E. Evans, in litt.). G. Kyriazis (pers. comm.)
walked the beach along the mainland coast from
(g)
1 to 3km south ofthe mouth oftheJardine River
species was observed also scavenging on turtle during early September 1991. His impression of
eggs that had been dug up by a nesting turtle. the track density at this time was that it was
Silver gulls {Lams novaehollandiae, 1 pair pre- comparable to the track density on Crab Island in
sent)flewaway withtwoN. depressushatchlings July 1991. There wasone freshly deadN. depres-
that had been released from an emerged clutch susthathaddiedonits returncrawltotheseaafter
excavated during the afternoon. nesting. Pigs were recorded to have destroyed
mostofthehundredsofturtlenestsseen on a6km
Other potential predators of turtle hatchlings section ofthis mainland beach in the early 1980s
within the nesting habitat and beach area were: (B. Gray, pers. comm.) and local residents con-
waterpython (Morelia mackloti, population esti- tinue to identify pig, dingoand varanidpredation
QmaMtJe31n7o5t4)a,tteesmtpuatreidn;e 1croscpoedciilmeen(Crcoocloledcytleuds, poefrse.ggcsomams.c)o.mTmhoensoiznetrhaisngbeeaocfhf(eGm.alKeyrtiuarztilse,s
porosus, minimum offive individuals, including
tagged while nesting on other occasions during
threewhose hindfootprintlengths were 27, 22.5 1987-1991 atCrabIsland(n= 19) andtheJardine
atunsd, s1e7v.e5rcaml)g,rboeuapcshofth2i-c3kbkinredes){,Bousrphrienyus(Pnaengdlieocn- TRihveermamjooruitthy(onf=tur3t)leasreenscuomumnatreirseeddonintThaebseleoc1-.
haliaetus, 1 pair), whistling kite (Haliastur casions were N. depressus.
sphenurus, 1 bird), white breasted sea eagle
(Haliaeetusleucogaster, 1 pair),frigatebird(Fre- Recaptures
gata sp., 1 bird). Potential predators of turtle One long distance migration has been recorded
hathlings intheadjacentsurfincluded: Australian from these taggings. An adult female N. depres-
pelican (Pelecanus conspicHiatus, flock of 10), sus(T40505), which had been tagged whilenest-
numeroussmallblack-tippedwhalersharks(Car-
ing at Crab Island on 15 January 1990, was
charhinus sp.). captured in a fishing net off Merauke (8°30*S,
On the team's arrival at the island, there were 140°22*E) in southern Irian Jaya on 10 January
two recently used fire places with broken N. de-
pressus egg shells and fish remains in sectors 5
and 7. During the 17 days the team was at the
island, three parties oflocal residents from adja-
cent mainland communities visitedtocollect tur-
tleeggs. Turtle nests were located by probing the
sandinoldbodypitswithaspear.All threeparties
directed their egg collecting activity in sectors
5-8, the area ofhighest density turtle nesting: 14
July three men collected an estimated equivalent
of four clutches ofN. depressus eggs; 17 July a
party of adults and children collected multiple
clutches of turtle eggs: 21 July three men col-
lectedanundeterminednumberofturtleeggs. On
the 22 July, a fourth party from two dinghies FIG. 8. Adult male Natator depressus, tag number
visited the island, but the purpose of their visit B0314, mounted on female prior to capture in ap-
was notassessed. One ofthe nestsdugon 14July proximately 0.7m deep water adjacent to the beach.
containedeggswithdriedeggshell, i.e. it wasnot The femalehasjust surfaced forabreath.
.
REASSESSMENTOFTHELARGESTTURTLEROOKERY 285
14
oviposition (na32)
12
emergence (n=39)
10
3
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95100
Clutch count
FIG. 9. Frequency distribution of Natator depressus clutch counts measured at oviposition and at clutch
emergence, Crab Island6-20July 991
1
1992. This turtle was eaten. The rceapturc was nesting was also recorded for the island. Speci-
made 310km from the nesting beach and two mens ofN. depressus were collected from Crab
years aftershe wastagged. This is the firstrecord IslandinJanuary 1981, butthenestingpopulation
ofN. depressus from the coastal waters of Irian was not described (Zangerl et al., 1988). The
Jaya. islandhasremainedunsurveyedformarineturtles
One remigration recapture has been recorded since August 1979. Limpus et al. (1989) identi-
from these laggings. N. depressus (BI 181) was fied that N. depressus nesting occurs widely in
recorded nesting back at Crab Island on 1 1 Sep- western Torres Strait, with another large nesting
tember 1992, approximately fourteen months af- concentralion at Deliverance and Kerr Islands
ter her initial tagging at Crab Island on 20 July The results of the present study allow for a re-
1991. evaluaiion of the status of this significant turtle
rookery after 12 years.
DISCUSSION Marine turtles typically lay multiple clutches
Marine turtle nesting at Crab Island was first within a single nesting season at approximately
quantified in November-December 1970 (Bus- two weekly intervals, except for Lepidochelys
lard, 1972). There were subsequent briefvisits to spp. (Hirth, 1980; Limpus et al., 1984). Not all
furtherquantifynestingdensityandtotagnesting individuals in a particular population begin or
turtles during 1976-1979 (Limpus ct af, 1983a). complete nesting al the same lime of Ihe year
As a result of these surveys, Crab Island was (Limpus, 1985). Atrookeries in eastern Australia
found to support all yearround breeding and was where Caretta caretia and Chelonia mydas have
identified as supporting the largest recorded N, a discrete nesting season, 60-70% of the total
depressus nesting population. Low density E. nestingpopulationforthe yearcan beexpected to
imbricata and sporadic Lep'tdochelys olivacea be recorded during a two week tagging censusat
1
286 MEMOIRS OFTHEQUEENSLANDMUSEUM
TABLE 3. Incubation and emergence success recorded from 39 freshly emerged A'atutor depressus clutches
excavated Forcountingon CrabIsland, 6-22July 1991-
Mean SD n range
emergedhalchlings 42.9 12.99 39 17-87
ir ._ 0.5 1.73 0-10
hatchliqgs in nest:
dead 13 2.51 39 0-1
unhitched 57 : 9S 'l.' 0-22
undeveloped 4.1 5.10 V) 0-30
predated 0.2 L) i9 39 0-3
clutch poi 5-1 b 39 34-£4
rtesidepth (cm) 583 7 :i "-! 43-71
1fetching success '£ 81 84 1729 39 26.5-100.0
Emergence succe 78.56 18.55 39 25.0-100.0
the peak ofthe nesting season (Limpus, 1985 and imhriuita, 1 L. tdivacetV- Bustard, 1972; Limpus
unpublished data). Therefore a two week census et al,. 1983a; present study). There have been
of the Crab Island nesting population at (he peak more than 50 recaptures ofN- depresses reported
ofthe nesting season also is presumed to sample from the east cuast prawn fishery from inside the
asimilarproportionofthetotalpopulation. How- Great Barrier Reef (GBR) and central Torres
ever. Lherc isall yearround nesting at Crab Island Strait, whichhadbeen laggedwhilenestingatthe
(Limpus et al., 1983a) and the present study may central Queensland rookeries (Limpus et al.,
C
not have sampled at exactly the peak ofthe nest- 1983b; J. Parmenter, pens. eomm.). This sug-
ing season. Each of these factors would reduce gests that the At depressus population nesting al
the proportion of the total population sampled Crab Island does not migrate from the feeding
during the two week census. Therefore, the esti- areas along the eastern Queensland coast. No
mated 1352-1611 clutches laid during the 14 Ed N. depressus from Crab Island has been
nights, 6-19 July 1991, indicate that in excess of ted from the northern prawn fishery of lhe
2200 female N. depresses were breeding Bl Crab Gulf oi Carpentaria and southern Arafura Sea
Island during the 1991 breeding season. While even though several Ihousand N. depressus are
there areinsufficient data to identify any changes captured annually as pad o^ the by-catch of the
in the size of Ibis nesting population since ii northern prawn fishery (Poiner & Hams, 1993).
last surveyed during 1976-1979 (Limpus et a!., This lack ol A/, depressus tag recoveries within
1983a), CrabIslandcontinuesJO support the larg- the northern prawn fishery contrasts with the
estrecordedbreedingaggregation for this marine occurrence of tag recoveries of C. careitu From
turtle However, it should be remembered that eastern Australian rookeries in the same Fishery
manyN. depressusnest on the adjacent mainland (Limpus et al.. 1992; Limpus & Reimer. 1993),
and that the species also nests on the islands This suggests that the Gulfof Carpentaria is not
extending tothe north throughout western Tones a significantfeedingarea lortheN. depressusthat
Strait (Limpus et al., 1989). The lotal annual nest at Crab Island. The one long distance tag
bleeding population for N. depressus in the Crab recovery demonstrates that the feeding distribu-
I ltkJ region (northeastern Gulf of Carpentaria tion oflheCrab Islandnestingpopulation extends
and western Torres Strait)remains imprecise, but northwards al least to the southern Inan Jaya
must number many thousands of females. There coast.
is still noevidence that Crab Island is ofanything Turtle nesting and hatehling emergence I.
but minorimportancefornestingby othermarine beenrecordedal Crab Islandon all previousvisits
turtle species. (April, May, July. August. November and De-
As of March 1993, there has been only one cember.) with a seasonal variability in density
reported feeding ground recapture, a female N. (Limpus et al, 1983a). The limited data from
depressesfrom the south eastern IrianJayacoast, 1991 also show seasonal variability in nesting
outofthemany turtlestaggedwhilenesting inthe density.Thehigh density nesting inJuly (average
Crab Island region (HMO N. depressus, 11 E. 133 beaehings per night on thewhole beach and