A comparison of sample and total census data for a population of the Eastern Longneck TurtleChelodina longicollisin a farm dam north-west of Sydney, New South Wales

5 Pages·1999·2 MB·English

by  Shelley Burgin

A comparison of sample and total census data for
a population of the Eastern Longneck lbrtle
Chelodina longicollis in a farm dam north-west of Sydney,
New South Wales
Shelley Burgin,l* Steve Emerton' and Malcolm Burgin'
'Centre for Integrated Catchment Management, University of Western Sydney-Hawkesbu~,
Richmond, New South Wales 2753
*Corresponding author
ABSTRACT
Interpretation of trap bias in sampling freshwater turtles has previously relied on indirect
methods. The data presented here provides the first comparison of a sample and total census
of a waterbody. Eastern Longneck Turtles Chelodina longicollis Shaw were sampled in a
0.1 ha farm dam for three days immediately prior to drainage and a population census was
subsequently carried out. Fyke nets proved to be an effective method of sampling the
population. During the sampling phase 78% of males (n = 11) and 81% of females (n = 21)
were captured, and there was no significant difference between the sample and the census
in terms of population size structure (x's = 1.12, excluding juveniles) or sex ratio (x2t = 0.008).
None of the three resident hatchlings were captured despite a high probability of capture
(p = 0.67). Hatchlings and juveniles combined made up 19.4% of the population and there
were no immature resident females and few immature males. Based on the population profile,
recruitment was intermittent and apparently infrequent with the greatest cohort being mature
females, although there were no immature females. Mature males were present in larger
numbers than immature males.
Key words: Chelodina longicollis, Turtle, Farm dams. Sampling.
INTRODUCTION when seine netting and muddling, compared
to fyke netting (pers. obs.), although Parmenter
Investigations of the population dynamics of (pers. comm.) was satisfied with the numbers
freshwater turtles need to address trap bias that he has obtained by seine netting. Beumer
since different capture methods yield different et al. (1981) incidentally caught C. longicolllc
class size distributions and sex ratios (Ream in fyke nets in all but two months of his
and Ream 1966; Thompson 1983). Several two year study (August 1975 and June 1976),
methods have been reported for catching and Houston (pers. comm.) regularly captured
freshwater turtles in Australia including seine, turtles in fyke nets during his study of the
fyke and gill netting, hand collection and Arafurian File Snake Acrochordus arafurae
muddling, i.e., foot sensing in the sediments McDowell in the Northern Territory. As a
(Chessman 1988a). Baited hoop traps, often result of these observations, we use fyke nets
modified after Legler (1960), are commonly to sample freshwater turtles and in this
used (e.g., Chessman 1984, 1988a,b; Georges study we investigated their effectiveness for
et al. 1986; Kennett and Georges 1990) and sampling a- population of C. longicollis in the
have yielded samples of Emydura krepii [Gray] Svdnev Reeion.
,
(Fraser Island) that are unbiased with respect
One way to address problems of sample bias,
to size distribution and sex ratios (Georges
due to netting, is to follow routine sampling
et al. 1993). The potential for biased popula-
with a total census. In this way the sample can
tion data for other species, such as the Eastern
be compared with the total population. A rare
Longneck Turtle Chelodina longicollis Shaw,
opportunity to do this occurred when we were
needs to be addressed.
notified that a dam in western Sydney was to
Since hoop nets need to be cleared and be filled in for a housing estate development.
reset at intervals of between 1 and 3 hours The dam was typical of thousands that are
(Chessman 1988h; Kennett and Georges 1990), found throughout peri-urban and rural
the mechanical movement required to clear properties across Australia, but even if it were
the nets could adversely affect turtle trapping typical only of peri-urban Sydney, it would still
success. Turtles also have the ability to move be a useful study because these dams are
out of hoop traps after capture, and thus we being in-filled in their hundreds, without
considered this approach to sampling turtles attention to the impact on turtles or other
to be sub-optimal. Few turtles were caught fauna dependent upon them.
Australian Zoologisf 31(1) 161
MATERIALS AND METHODS Comparison between sample and census
To compare size range between the sampled
Study site
population and the total population, a chiP
The dam was located on an abandoned analysis was undertaken on size (carapace
market garden at Quakers Hill (35 km north- length) and sex ratio with Yates' correction
west of Sydney). Evidence fmm a 1974 ortho- for continuity as appropriate (Zar 1984).
photographic map (1:25 000 Central Mapping Because of the small number of hatchlings,
Authority, Riverstone 9030-I-S, 1974) indicated they were excluded from the analysis. Post-
that it had existed for at least 20 years. Its hatchling individuals were arbitrarily divided
unvegetated banks had a slope of approx- into small and large males (small <135, large
imately 30". Its surface area was about > 140 mm) and females (small < 195, large
0.1 ha (37 m x 27 m). The dam was blanketed >225 mm).
with duckweed Lemna minor [L.] and had
an average depth of 1.8 m. The substratum RESULTS
consisted of fine silt over a clay base.
Thirty-seven C. longicollis were captured and
marked during the study period: 13 males,
Pre-drainage sampling
21 females and three small hatchlings. Twenty-
The study was undertaken during typical seven of these were captured in the three days
Sydney summer weather conditions: dry and of sampling before the dam was drained. A
sunny. For the three days prior to the total of 36 turtles were accounted for in the
commencement of drainage, eight fyke nets total census: 15 not released after the third
were set overnight in the shallows against day of sampling; 16 netted as the water
the dam's perimeter. Thereafter, nets were set receded and five sets of outward hound tracks
and adjusted to coincide with the receding in the silt (Table 1; Fig. 1).
water levels until too little water remained
Thirty-five per cent of the turtle population
to effectively set the nets. The nets were
was captured on the first day, 40% by the
emptied each morning and turtles were
end of the second day and 73% had been
weighed, measured (straight line carapace
sampled by day three when the draining of
length, breadth and height of animal in
the dam commenced.
mm), sexed (after Kennett and Georges 1990)
and uniquely marked (after Parmenter 1976). The density of turtles in the dam was
Samples from the first two mornings were calculated at 360 ha-'. During the sampling
released at the point of capture but were not phase of the study no hatchlings were
returned to the draining dam on the third caught but 78% of males (n = 11 and 81%
morning. (n = 21) of females were captured. The
sample did not differ significantly from
Total census sampling the population, either in terms of sex ratio
(x21 = 0.008) or size structure (x2s = 1.12,
Nets continued to be set at the water's
excluding juveniles). The probability of
edge and the captured animals removed,
capturing a hatchling during the three days
while sufficient water remained to set the
of sampling was calculated to he p = 0.67.
nets. In addition, the banks of the dam and
surrounding land were regularly searched
(approximately 80 person-hours) for animals. DISCUSSION
Approximately four hours after drainage
The concept of this study was to compare
began, a band of silt around the perimeter was
trapped turtles with a total census. Such data
exposed, allowing for an accurate account of
have previously not been collected for any
animals leaving the dam through observation
turtle species. In an ideal world this would
of distinctive turtle tracks in the silt.
have involved replication, both in other farm
A diesel powered pump with an estimated dams and natural waterbodies. However, it was
flow rate of 80 kL/h was used to drain the no small task to organize the drainage of
dam between 0730 and 1630 hours from the dam to ensure a total census and this
the 13 to 16 March, 1995. Two days after had not been achieved before. We therefore
drainage was completed, approximately 50 cm had the rare opportunity to study one dam's
depth of sludge was removed using an freshwater turtle population, before and after
excavator with a bucket capacity of 0.5 cubic draining. Sampling began within two hours of
metre. A tip truck distributed the silt in strips hearing that the dam was to be drained, after
on adjacent flat land where it settled to contact and support had been gained from
a height of approximately 45 cm. The entire both the dam's owners and the contractors. We
process was continuously monitored for signs have provided one data set. Replication will
of turtle activity until 23 March. have to await the opportunity arising again.
Carapace length (m m)
IS am ple -Census
Fig. I. Size class distribution of turtles sampled from a farm dam compared with a total census. Animals between
95-170 mm were male, all larger animals were adult females.
Table 1. Summary data of sampled population and total census of Chelodina longicollis
from a dam in Quakers Hill (Sydney), March 1995.
Population sample Total population
Parameter (n = 27) (n = 37)
body mass 125-1040 (S 650, sd 245) 5-1040 (R 567, sd 291)
carapace length 92.7-222.6 (X 177, sd 30) 31-222.6 (Tf 162.5, sd 49)
carapace breadth 74.7-155.5 (X 130, sd 20) 26.3-159.3 (fi 119.4, sd 34)
body thickness 32.6-64.6 (jr 55. sd 8) 13.3-64.6 (X 51.3, sd 14)
sex ratio 10 males : 17 females 13 males : 21 females
Burbridge (1967) reported that no data were cohorts being predominant. The data
available on any population of Australian presented here indicated that recruitment was
Chelidae. Since that time a number of authors intermittent.
have assumed, based largely on intuition, that
Male C. longzcollis mature at approximately
various techniques discriminate against the
seven years and females at 10-11 years of
capture of juveniles (e.g., Parmenter 1976;
age. Females usually lay one clutch of eggs
Chessman 1978). Others have compared
annually (Parmenter 1985) of between 6
captures with re-captured individuals to
and 23 eggs (Legler 1985). Without mortality,
assess trapping bias (e.g., Georges 1982;
recruitment to the population would be
Dalem 1998). Our study provided a unique
between 126 and 483 hatchlinglyear or a
opportunity to obtain an understanding of
standing stock in the dam of 630-2 730
the freshwater turtle population structure of a
juvenile and sub-adult females and 441-4 641
dam and compare it with a typical sampling
males. Although large numbers of nests
approach. The comparison showed that the
are destroyed annually (Thompson 1983), it
sample provided a good estimate of the sex
is widely accepted that predation is most
ratio and size of individuals resident in the
intense on the smaller cohorts. Previous
dam, and it was therefore concluded that fyke
researchers have assumed that their capture
nets provided a reasonable sample.
techniques have under estimated juvenile
Wong and Burgin (1997) suggested that numbers (e.g., Parmenter 1976; Chessman
population structure was dependent upon 1978; Thompson 1983). Evidence to support
the recruitment pattern. Georges (1985; this hypothesis is provided by a comparison
Georges et al. 1993) postulated that this of catch against number of animals caught
could occur as a "sustained trickle" of young, (e:g., Georges 1982; Dalem 1998), although
recruited annually, and this would result this approach does not address preferential
in equal numbers of juveniles and adults predation on smaller individuals. The data
in the population. Alternatively, recruitment presented here were the first to provide a true
may be intermittent. This would result in a picture of the structure of a resident popula-
relatively large number of particular age tion. Hatchlings and juveniles combined made
June 1999 Australian Zoologist 31 (1) 163
CIE31e~aenhP. mpertles). T h a ha lso m Mr Juhn
Wauteroons who, almg with the mawactam*
tderated pur e;scentrk&qty,
Barbrid-> A. A., 1867. The biology of s~u.thwe~6e1-n
A~trdiant orwises. PUl. She&, Univ~r~itoyf
uymm Awm& We'h.
B. C., lo&&F. oodl &%he &&e Wcbd Turk
Glh9Q&fo~%z&~~pl iZZ1s ~-gd!fnw:C hlid~ei)n
the &fu~~Vaayll eya %tramhd Mew 5~11tWh da.
Am. ,MSBBI Res, 11: @?34'%,
K%e~mtma, Q, 1Wk. ~i[* pd,erences of fkcsh-
vrata m&es in the 1l-y Walk~V lctolcia and Mew
South Widex. AS$R%@.. h.IS : 485-9 1.
Ghessman B. C., 19888 Seaswpal a d d 2d edvity of
fresbatm tutdes ain .T$e U ~ tVqi-cy , Victorkt d
WBWw A* . W@$l. rWj& 867.7.T@.
Dalm, A, la, G. R., I@%!@. Pemwpphy and mwempt
p&ews d w pqJa@ssi cif WWf mak&&wk8
up IQ@,4%9.f the pmpx&tim@ a d w ere @at, as *t- ~&ku@hdn@& e@'~ ~&g&.*i~t~atDn 4 @~>~&ILI7GaW~) hW M&~&L USV:
i-nbrizivdy e~peccedp, mat twiz&m~ Rtlcbrnm~A,
There was no sqpScant &i;ffereneh emeex
h e smpk ~ u dthe p~p~bUc)antm (t%u~e
i~ terns ~f sex ratla c i ~s he stm&rri$.
(eItclu6ing bat~hk@* Thre 4i6,~BqtTcm
was skexed t~;mfwxlsm ature fcrn+l.es iJ,a.,
r 1? 0 mm, Parrneig;te~ b885), No bnature
kmdca wae, present and xmi~e maim
(< GQm ~&, armenter;1 984) m ep reRnt fa
greater a~asbrst -ha immature males and
hatchlings but in s.m.der anmbgrs tbaa <Tear8 te4& okr. ,ihR rb mpSa.sb+ iRi?v, wHw. m*a nbd L~., 1986~. s ~ v ~
mature ferrrdes. The obsgmed turtle dasl'tfr (nseudia~.C ChdiA4 &am &&a C d Dta *
sf 860 haa1 Tab with33 &e r~dgew tim~ted Lakes of t&oJ ig 2h~itay.- A&t$ W&,, Rm. 132
For o%bm farm dams studied (Z36 had1, 9.01-08.
Che~smmlS T8 to ,000 Pmmmter 19?QT;). Geaxpesy 14a, WBI,-UB., C. J. qS1 Parme~??qC, - J.,
Wm. Wakmd%Cmrp d ths ?%&&a. 9.1P&Q8
in P~u%BL $&MWB&~ WeM : dw*hm b *R
d b p e . ~ . a2t ~G . ~ . B . W Q S W ~ E L .
These data pravtde evidence @act just. an B&[email protected]. ~tga3"stla ~ P@I~I5,h.'4 &~d ~ ew:
W m a .
estI-m&gj that &cr~ are siktmtid nt@iBe_r-s& f
eurtlm &sident ira farm dams. Mbe3t Emitted
data, the &ens4.ty gf tmtla was at the upper
end of the rmge of estkwtes of stwax6bg
stock of' tbrtIes iii farm dtm&o m abwbtre.
June 1999
. -
Legler, J. M., 1960. A simple and inexpensive devise Ream. C. and Ream.. R... 1966. The influence of ramolincr
for trapping aquatic tuder. h c . Utah Acnd. Sc. 37: methods on the estimation of population structure
63-66. in painted turtles. A m M idl. J. 325-28.
Legler, J. M.. 1985. Australian chelid turtles: repro- Thompson, M. B., 1983. Population of the Murray
ductive patterns in wide-ranging taxa. Pp. 117-23 River Tortoise, Em$ura (Chelodina): the effect of
in Biology of Aurtralarian Frogs and Rcptilcr ed by egg predation by the red fox. Hlpcs vulpcr. Aurt.
G. Grigg. R. Shine and H. Ehmann. Surrey Beatty Wild. Res. 10: 363-71.
& Sons: Chipping Norton.
Wong, P. and Burgin, S., 1997. Preliminary observations
Parmenter. C. 1.. 1976. The natural historv of the of the freshwater turtle. Chrlndina longicollir (Shaw),
~ustralian& shwater turtle, ChcIodinn l~ngicollir. in the Longneck Lagoon Catchment (Hawkeabury-
Ph.D. thesis, University of New England: Armadale. Nepean River, New South Wales). Hcrprto/oum
27(2): 13-16.
Parmenter, C. J., 1985. Reproduction and survivorship
of Chrlodina Irmgicollir vertudinata: Chelidae). 4. Zar, J. H., 1984. Biostatirtical analysis, second edition.
53-61 in Biology of Austrolnrinn Fmgr and Reptiles Prentice-Hall International: New Jersey.
ed by G. Grigg, R. Shine and H. Ehmann. Surrey
Beatty & Sons: Chipping Norton.
-
Book Reviews Book Reviews
- . ,
Nature Photoenohv bv, Ken Gritfiths. showing through their shiny black coatings. The
UNSW Press. results from these older cameras certainly speak
RRP $29.95. for themselves as the book is liberally padded
with many fine examples of his work and the
All of us have had that feeling when we are methods of how they were achieved.
standing out of doors, a camera in hand, and we
have a particular image that we want to capture The book is light and compact so you could
on film. With limited experience and often conceivably carry it around with you for a quick
inappropriate equipment we do the best we can. reference while actually out photographing a
Often we can be a little disappointed at the result. subject.
If yon have ever tried to creep-up on wildlife
As with any book there is a few minor problems.
and capture the perfect close-up, feeling like
The term "nowadays" appears a little too often
Dombrovskis or Parish, only to look through the
within the text for my comfort, very much
camera lens and be surprised at how small the
reinforcing his "old school" approach to the subject.
subject still seems, then this book is for you.
The use of colour highlighting in the text for
In the last 20 years wildlife photography has quick reference does not always work due to the
gone from being the hobby of a few dedicated solo pale nature of the colours chosen. The caption
eccentrics to being accessible to almost anybody format for his photos is not always standardized
who enjoys spending extended periods outdoors. with the result that sometimes the main text
appears to flow into a photo caption. With some
Ken Grifftths knows his craft. In studying this pictures the subject is without identification, a
book you get the feeling you are being guided detail that would not detract from the intended
by an old master. This is not somebody who has example. Finally a couple of his photographic
hauled large format equipment and exorbitantly examples are a little obscure as to the technical
expensive accessories into the bush. Nor is he a point they are trying to represent.
single camera make exponent, a "Nikon-man" or
"Canon-man". This is somebody who has developed It was a little disappointing that the section
an interest in photography and grown with that on ethics did not include the warning on the
interest over the years. possibility of predation of the subject, particularly
involving nestlings, when disturbing wildlife during
His book has patiently laid out the steps to take photography.
to get that bit extra out of your developing interest
Overall the book is an excellent example of
without having to apply for an overdraft or trade-
taking some of the mystery out of photography.
in your current equipment for the latest offerings
For any budding photographer, young or old alike,
from the camera shop. Griffiths points out that you
there is much to benefit by the reading of this very
can use the more modern "auto-everything"
accessible book.
equipment but the limitations become apparent if
you get serious in your approach to this subject. Nicholas Carlile
His equipment has the patina of many years of Biodiversity Survey and Research Division
faithful service with the brass-metal camera bodies NPWS
June 1999 Australian Zoologist 37(1) 165