Journal of Biogeography, 29, 633–640
Male-biased sex ratio in a small tuatara population
Nicola J. Nelson1*, Susan N. Keall1, Shirley Pledger2 and Charles H. Daugherty1 1School of
Biological Sciences and 2School of Mathematical and Computing Sciences, Victoria
University of Wellington, Wellington, New Zealand
Abstract
Aim We estimated population size, survival, longevity and sex ratio of tuatara (Sphenodon
guntheri Buller 1877) on North Brother Island and determined whether recruitment was
occurring, to aid management and identify potential problems for population viability.
Location This study was conducted on North Brother Island, Cook Strait, New Zealand.
This 4-ha island supports the sole remnant population of S. guntheri.
Methods Demographic variables were estimated using capture–recapture methods. Tuatara
were individually marked and recaptured during twelve trips, spanning a decade. Population
size was estimated for selected trips using the CAPTURE package and finite mixture models,
and survival was analysed using the Jolly–Seber (J–S) model. Longevity was estimated using
tuatara individually marked in 1957.
Results Approximately 350 adult tuatara profile likelihood interval (PLI) (294–427) inhabit
North Brother Island, and the sex ratio is strongly biased towards males (1.7M : 1F). Annual
adult survivorship is high (0.95) for both sexes and some tuatara live for at least 61 years.
Main conclusions The small size and biased sex ratio of this population may make it
susceptible to demographic stochasticity, Allee effects, and/or loss of genetic variation.
Harvesting for translocation could exacerbate such problems. In addition, tuatara have
temperature-dependent sex determination (TSD), with males produced at higher incubation
temperatures. Global warming may therefore skew the sex ratio further unless female nest site
choice or adaptive shifts in pivotal temperatures compensate for rising temperatures.
Keywords
Sphenodon guntheri, tuatara, reptile, mark-recapture, sex ratio, survival, population
size.
INTRODUCTION
The fate of small populations is uncertain, because even
normal variation in birth and death rates can lead to
extinction (Lacy et al., 1995). Demographic, genetic and
environmental factors, for example, unbalanced sex ratios,
random variation in reproductive success, survivorship at an
individual level and inbreeding can contribute to the
vulnerability of small populations. Temperature increases
as a result of global warming may be an additional problem
for reptiles with temperature-dependent sex determination
(TSD), if warmer temperatures result in skewed sex ratios.
Tuatara are threatened reptiles restricted to off-shore
islands of New Zealand. They are of high conservation
importance (Cree & Butler, 1993; IUCN, 1996), and
*Correspondence: School of Biological Sciences, Victoria University of
Wellington, PO Box 600, Wellington, New Zealand.
E-mail: [email protected]
2002 Blackwell Science Ltd
biologically significant as the sole living representatives of
the order Sphenodontia (Benton, 1990; Cree & Butler,
1993). There are two species of tuatara, Sphenodon punctatus and S. guntheri. Prior to the arrival of humans, tuatara
were widespread throughout mainland New Zealand (Cree
& Butler, 1993). While there are still more than thirty island
populations of S. punctatus, S. guntheri are primarily
represented by one small population on 4 ha North Brother
Island in Cook Strait (Cree & Butler, 1993). This restricted
distribution is the result of habitat destruction and predation
by introduced mammals (Cree & Butler, 1993). Invasions by
introduced mammals, particularly rats, remain as risks for
tuatara populations. Translocations of tuatara from North
Brother Island to two other islands have occurred recently,
but the status of these populations is unclear (Nelson et al.,
2002). The only population information available for
S. guntheri is an estimate of 300 tuatara on North Brother
Island obtained using a simple weighted mean Petersen
analysis of mark-recapture data (Thompson et al., 1992).
634 N. J. Nelson et al.
Tuatara are medium-sized nocturnal reptiles with TSD
(Cree et al., 1995). Maximum snout–vent length of female
S. guntheri is c. 213 mm (Thompson et al., 1992). Their
lengthy life cycle means that recovery from population
reductions is very slow. Tuatara reach sexual maturity at
about 15 years of age (Dawbin, 1982; Castanet et al., 1988),
and females breed about every 2–5 years (Cree, 1994).
Sphenodon punctatus on Stephens Island are generally considered to have determinant growth and attain full size in 25–
35 years (Castanet et al., 1988; Thompson et al., 1992; but
see Nelson et al., 2002), but potential longevity is not known.
Three quarters of all species known to have become globally
extinct since 1600 were island species (Jenkins, 1992).
Accurate biological information is important for good conservation management to allow evaluation of the viability of
small isolated populations and of past management practices
and to aid future management strategies (Hiby & Jeffery,
1987; Miller & Ford, 1988; Smith & McDougal, 1991). We
estimated population size, survival, longevity and sex ratio of
tuatara on North Brother Island and determined whether
recruitment was occurring, to aid conservation management
and identify potential problems for population viability.
METHODS
Study site
This study was conducted on North Brother Island, a 4-ha
Wildlife Sanctuary located in Cook Strait, New Zealand
(4107¢ S; 17427¢ E). The area searched for tuatara from
1996 to 1997 covered c. 2 ha of the area considered to be
tuatara habitat by Thompson et al. (1992; c. 2.2 ha). The
further 0.2 ha of tuatara habitat was not searched because of
inaccessibility. There are very few tuatara in the remaining
1.8 ha of the island, which is poor habitat for them
(Thompson et al., 1992).
Data collection
The data were collected on thirteen trips by many workers
(Table 1). The first capture data were recorded in 1957, and
twelve further trips to mark and recapture tuatara were
made between 1988 and 1997. Trips ranged from one to
seven nights in duration (Table 1).
Tuatara were located by searching, mostly at night, using
torch light. From March 1996, the search area was covered
several times each night. Each captured tuatara was held in a
cloth bag until measurements were taken. The capture
location for each tuatara was recorded on a grid system, the
capture site marked, and all individuals returned to the
capture location. Measurements recorded included snout–
vent length, vent–tail length, tail regeneration length and
weight. The sex of each individual was recorded, where
possible, based on development of secondary sexual characteristics, such as crest development and large head for males,
and pear shaped abdomen for females (Dawbin, 1982; Cree
et al., 1991). Tuatara were given permanent individual
marks by toe-clipping when first captured, and recaptures
were recorded. A unique number was written on the side of
each animal with a ‘Vivid’ marker pen; the number was
legible throughout the duration of the trip.
Analyses
Demographic variables for Brothers Island tuatara were
estimated using capture–recapture methods. The study used
Pollock’s robust design (Pollock, 1982; Pollock et al., 1990),
and data were analysed as suggested in Pollock (1982) to
allow for heterogeneity of capture probabilities among
individuals. Within-trip closed population models were used
to estimate the population size, while a between-trips open
population model was used to estimate survival.
Population size
The total number of tuatara on North Brother Island was
estimated using the closed population models in CAPTURE
(Otis et al., 1978; White et al., 1982) and finite mixture
models (Norris & Pollock, 1996; Pledger, 2000). The
software for CAPTURE is available as part of the MARK
program at http://www.cnr.colostate.edu/gwhite/mark/
mark.htm.
Table 1 Surveys of tuatara on North Brother Island and their use in demographic analyses
Trip no.
Date
Leader
1
2
3
4
5
6
7
8
9
10
11
12
13
August 1957
January 1988
December 1988
November 1989
October 1990
November 1990
November 1991
November 1993
November 1994
November 1995
March 1996
February 1997
December 1997
Barwick
Thompson
Carmichael
Daugherty
Gaston
Cash
Cash
Gibbs
Cash
Cash
Markwell
Nelson
Nelson
No. nights
Closed population analysis
Open population analysis
3
3
3
7
3
3
5
3
1
5
5
5
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
Sex ratio analysis
4
4
4
4
4
4
4
4
4
4
4
4
4
2002 Blackwell Science Ltd, Journal of Biogeography, 29, 633–640
Population size and sex ratio of tuatara 635
This combination of programs allowed for up to three
types of variability in capture probabilities. Capture probabilities could vary over time (t), because of heterogeneity
among individuals (h), and because of behavioural responses
to capture (b). If such effects are present but not allowed for
in the model, estimates of population size may be biased;
reptile behaviour is responsive to temperature, for example.
Maximum likelihood models were used throughout, with
M(0) (the null model, with no variation in capture probability), M(t) and M(b) from Otis et al. (1978) fitted by
CAPTURE, and M(h) fitted as in Pledger (2000). The
methods of Pledger (2000) were also used to fit models
combining two or more of the sources of variation, some of
these models having additive effects only [e.g. M(t + h)] and
others including interactive effects [e.g. M(t · h)].
Likelihood ratio tests were used to select the best model
describing the data (5% significance level; Pledger, 2000).
Only trips 11–13 were used for model selection, as at least
four samples are required (Table 1). The CAPTURE model
selection procedure was also tried for all trips with at least
three nights’ sampling, but this gave conflicting results.
Problems with this model selection procedure are detailed in
Stanley & Burnham (1998). We therefore assumed that the
best model(s) for trips 11–13 would also be appropriate for
estimating population size for the other trips.
Reduced data sets with less heterogeneity were also
analysed. Juvenile tuatara almost certainly have a lower
probability of capture than adults (Dawbin, 1982), so adults
were analysed alone to remove this source of heterogeneity.
There may be behavioural differences related to sex, so we
analysed adult males alone to remove this source of
heterogeneity. Another possible source of heterogeneity is
the presence of a few animals that are almost always
captured, a feature that may sometimes destabilize population estimates from heterogeneous models. Cormack (1992)
suggested splitting the population into two groups: a small
group with such high capture probabilities that they are
effectively censused, and a larger group with lower, more
homogeneous capture probabilities. The size of this larger
group is estimated using a simpler (homogeneous) model,
while the size of the small group is effectively known
without error. The total population size estimate is the sum
from the two groups, and its standard error is simply that of
the larger group estimate.
We estimated population sizes using the overall best
model [M(t); see model selection in the results]. It was
important to use the model that best fitted the data to obtain
the most accurate estimates. All population estimates are
presented with a 95% profile likelihood interval (PLI), the
set of possible values of the parameter that would be
acceptable under a likelihood ratio test (at 5% significance
level) for that value of the parameter (see Cormack, 1992).
Cormack showed the PLI to be more appropriate for
capture–recapture study estimates than the traditional symmetric confidence intervals (which may go outside natural
bounds on the parameters, giving a population estimate
lower than the number of animals seen, or a survival
probability estimate above 1).
2002 Blackwell Science Ltd, Journal of Biogeography, 29, 633–640
Assumptions of the closed population models are that
populations must be demographically closed, that no marks
are lost during the study, and that all marks are noted and
recorded correctly at each sampling occasion. Demographic
closure means that no birth, death, immigration or emigration
occur during the sampling period. The mark-recapture study
of tuatara on North Brother Island met the closure assumption because the short period of each sampling period (three to
five nights) meant that there was negligible probability of any
mortality or recruitment occurring. For the short-term nature
of the closed population analyses, the permanency of marks is
indisputable. New toe losses are easily distinguishable, and
‘Vivid’ marker labels provided a simple backup.
Survival
Capture history data from selected trips (Table 1) were
analysed using the Jolly–Seber (J–S) open population model
to estimate survival (Jolly, 1965; Seber, 1965). The J–S
model is appropriate if M(0), M(t), M(b) or M(t + b) is
chosen from the closed population analyses, as it allows for
time effects and short-term trap response (within-trip data
pooling means only the first capture per trip is relevant), but
not heterogeneity among individuals. Even with low levels of
heterogeneity, the J–S survival estimates are relatively robust
to bias (Pollock, 1982). The J–S model has the following
assumptions (White et al., 1982; Pollock et al., 1990):
within a trip, each animal must have the same probability
of (at least one) capture; between trips, each marked animal
must have the same probability of survival; marks must not
be lost or misread over time; and sampling periods must be
short with respect to the intervals between samples.
Sampling periods of up to 5 days, with at least 10 months
between samples, ensured a suitable sampling regime for
analysis by J–S. August 1957 data were not used because of
the long period to the next trip. Conversely, October 1990
data were not used as the interval until the next trip was
considered too short. November 1995 data were not used
because of the limited number of tuatara captured and the
removal of eighteen tuatara from the population. A tuatara
was recorded as captured on a given trip if it was captured at
least once in that trip.
Extra analyses were performed with two reduced data
sets: adult females only and adult males only. These two data
sets were compared in MARK using ‘Recaptures Only’ to
determine whether capture rates and/or survival rates varied
by time and/or sex.
As equal catchability is not met in most capture–recapture
studies (Carothers, 1973), a series of contingency chi-square
tests was performed to test the model’s goodness-of-fit
(Pollock et al., 1985). Data were pooled to generate summary tables of capture histories for each trip or capture
occasion (i) according to the following two formulations:
Formulation 1: New captures in each trip (i) were categorized by whether they were next caught in the following trip
(i + 1), or not until after that (i + 2 onwards), or never
recaptured. Recaptures in each trip were categorized in an
identical fashion.
636 N. J. Nelson et al.
Formulation 2: Tuatara captured in each trip (i) were
categorized by whether they were first captured in the
previous trip (i – 1), or earlier (before i – 1). Tuatara not
captured this trip, but caught next trip (i + 1), were
categorized by whether they were first captured last trip
(i – 1) or earlier (before i – 1).
The summary statistics were then compared with expected chi-square values to evaluate goodness-of-fit or
divergence from homogeneity of capture or survival
probabilities.
The closure assumption is relaxed for analysis by J–S
(Seber, 1973). Births, deaths and migration of tuatara
between capture occasions are allowed, but losses to the
population must be permanent (Pollock et al., 1990).
Sex ratio
Chi-square tests (Sokal & Rohlf, 1981) were performed on
the total number of individual adult males and females
captured on each trip (Table 1) to test for a 50 : 50 adult sex
ratio. Adult capture records from March 1996 and December 1997 were also compared using the population estimates
from model M(t) to test whether different capture probabilities for males and females explained any skew in the
apparent sex ratio. These two trips were chosen based on a
large number of captures.
RESULTS
A total of 2024 captures was made over the thirteen trips,
and 547 tuatara were marked (Table 2). The total number
of individuals captured per trip ranged from 24 to 243. The
greatest numbers of captures occurred on trips when five
nights were spent on the island. More adult males than
adult females were captured on each trip, except August
1957.
Trip
no.
1
2
3
4
5
6
7
8
9
10
11
12
13
Population size
Model selection
Based on likelihood ratio tests, the best models for estimating population size were M(t + h), M(t) and M(t + h) for
trips 11, 12 and 13, respectively. In all cases, time was by far
the most important single factor in improving the model fit.
Removal of animals with four or five captures over five
nights reduced the heterogeneity. Removing seven such
animals in trip 11 caused model M(t + b) to be selected. The
removal of one animal in trip 12 gave no change of model,
and removal of three animals in trip 13 caused M(t) to be
selected. Overall, M(t) looks like a suitable model after
removal and later addition of high-capture animals. The
appearance of a short-term behavioural trap response in the
trip 11 data (after allowing for a time effect in the model)
does not affect the choice of open population model, and the
J–S model allows for such an effect (see Methods).
Population estimates
We used model M(t) outputs, with removal and later
addition of highly catchable tuatara to estimate population
size for each trip (Fig. 1). These estimates ranged from 245
in January 1988 (trip 2) to 474 in November 1994 (trip 9).
There is some concern about the 1996 (trip 11) estimate. The
model selected for that trip, M(t + b), produced a population
estimate of 259 (PLI 242–308), which was well below
the estimate based on M(t). However, the J–S population
estimate for trip 11 was in line with estimates for trips 12–
13, so we think the extreme trap-shyness ‘detected’ in trip 11
after removal of heterogeneity was an artefact of an unlucky
sample and/or the removal and re-addition process. The
tuatara removed from the sample were those caught most
frequently, so the remaining sample has some bias towards
trap-shyness.
Date
Adult
females
Adult
males
No. of individuals
captured*
No. of
captures
August 1957
January 1988
December 1988
November 1989
October 1990
November 1990
November 1991
November 1993
November 1994
November 1995
March 1996
February 1997
December 1997
15
35
15
23
20
46
49
55
70
15
65
76
60
9
73
20
24
31
71
64
130
133
21
148
148
116
24
112
35
47
51
118
113
190
211
36
240
243
202
24
135
35
47
53
119
133
230
252
36
347
343
270
Table 2 Summary of tuatara capture data
1957–97. Sex was identified by development
of secondary sexual characteristics
*This column does not sum to the total number of individuals captured over all trips, because
of recaptures of some individuals on several trips.
Eighteen of these adults (eleven females and seven males) were removed from North Brother
Island as founders for a new population (Nelson et al., 2002).
2002 Blackwell Science Ltd, Journal of Biogeography, 29, 633–640
Population size and sex ratio of tuatara 637
Table 3 Comparison of survival models fitted to the data using
the ‘Recaptures only’ option in MARK. Models attempted had
survival rates depending on time and sex (t · s), time only (t), sex
only (s) or constant (Æ). The best model is that with the lowest AICc,
indicating the best compromise between simplicity and fit to the data
Survival (/)*
Recapture (p)*
k
DAICc
wi§
t·s
t·
t
s
Æ
t·
t
s
Æ
t·
t
s
Æ
t·
t
s
Æ
38
29
22
21
29
19
12
11
22
12
4
3
21
11
3
2
29.75
23.87
91.47
89.52
15.34
10.12
77.91
76.96
6.29
0.73
73.76
72.15
4.52
0.00
72.35
71.29
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.02
0.38
0.00
0.00
0.06
0.54
0.00
0.00
t
Figure 1 Population estimates based on M(t) for tuatara in the
study area on North Brother Island. Error bars show profile
likelihood intervals (PLI).
s
Æ
Population size and recruitment
While the estimated population size increased from January
1988 to November 1994, this is not a reliable indication of
an increase in population size (Fig. 1). Only the trips from
March 1996 to December 1997 (trips 11–13) were planned
for closed population analysis. Population estimates for
these trips are in close agreement, indicating reliable results
with small PLI. Hence, the most robust estimate for the
total number of tuatara in the search area of c. 2 ha is
about 400 (Fig. 1). About 317 (PLI 267–388) of these are
adults, based on the 1997 data where adults were analysed
separately (Fig. 1). Assuming that the density in the
unsearched 0.2 ha of tuatara habitat (which was unsafe to
search) is the same as in the search area, we estimate the
total population of adult tuatara on North Brother Island to
be about 349 (PLI 294–427).
Although juveniles are hard to detect on North Brother
Island, they were captured on all occasions and indicate that
recruitment is occurring in the population.
Survival and longevity
The Cormack–Jolly–Seber ‘Recaptures Only’ analysis in
MARK allows model selection based on Akaike’s Information Criterion (AIC, or its small-sample version AICc) or
based on likelihood ratio (chi-squared) tests. Using data for
adults, we tried models with capture probabilities and/or
annual survival dependent on time, sex, both or neither. The
best model based on AICc involves no sex differences in
capture or survival probability, and constant survival over
time (Table 3). The second best model, with AICc only 0.73
higher than the best, has capture probabilities varying
through time and annual survival constant through time
but varying with sex (with females having a slightly lower
survival rate). This is a less parsimonious model, and a
likelihood ratio test selects the simpler one (v2(1) ¼ 1.318,
P ¼ 0.2510). The same result was obtained using likelihood
ratio tests with a 5% significance level. Under the best
2002 Blackwell Science Ltd, Journal of Biogeography, 29, 633–640
s
s
s
s
*Factors affecting survival and recapture probabilities in the model.
Number of parameters.
Relative AICc.
§Akaike weights.
model, mean annual survivorship from 1988 to 1997 was
0.949, with a 95% PLI (0.935–0.962).
We know juvenile survival is lower than adults, because
inclusion of juveniles gave lower survival estimates, but the
low capture probabilities of juveniles (Dawbin, 1982) meant
there were too few juvenile captures for a separate analysis.
J–S requires homogeneity of capture and survival probabilities, but goodness-of-fit contingency tables show significant divergence of the data from homogeneity. Equal
catchability is an unattainable ideal in natural populations
(Carothers, 1973). Population estimates from J–S are likely
to be negatively biased in the presence of heterogeneity
unless there are high capture probabilities, but there is little
evidence of bias arising from unequal catchability or
permanent trap response in J–S estimates of survival (Carothers, 1973; Gilbert, 1973; Nichols et al., 1984; Pollock
et al., 1990).
Of twenty-four tuatara marked by Barwick in 1957, eight
(three males and five females) were recaptured between 1988
and 1997. Two males grew in length, indicating they were
not fully grown when first marked; the third male was not
measured in 1957. The two females with the longest capture
histories were fully grown when first marked in 1957 and
were last captured in 1993, suggesting that the minimum age
of these two females is 61–71 years.
Sex ratio
The estimation of sex ratio using only the animals captured is
supported by the MARK analysis showing no sex differences
in capture rate. The proportion of females observed was
638 N. J. Nelson et al.
found to vary over the thirteen trips (analysis of deviance,
v2(12) ¼ 22.23, P ¼ 0.035), but if the earliest trip (1957) is
excluded, trips 2–13 show no significant differences in the
proportion of females (v2(11) ¼ 14.91, P ¼ 0.1868). The
1957 trip is considered an outlier because of the unknown
basis for selection of animals during that trip and the
possibility of a change in sex ratio since that trip. We pooled
information from samples 2–13 to estimate the proportion of
females, counting each animal only once even if it was seen
on several trips, to remove dependence from the data. With
179 females and 303 males, this gave an estimate of 0.371
(SE 0.043) for the proportion of females, with a 95%
confidence interval (0.328, 0.414), which does not include a
50 : 50 sex ratio. This translates to an estimated male :
female sex ratio of 1.7 : 1, with 95% confidence intervals of
(1.42 : 1, 2.05 : 1).
DISCUSSION
The long-term prospects of S. guntheri on North Brother
Island appear good. They have high survivorship and are
long-lived. Recruitment is occurring, and few juveniles are
required to reach breeding age to maintain a population with
a high survival rate. They are legally protected, and there is no
reason to believe that survival and recruitment have declined.
The adult population estimate of c. 349 is slightly higher
than the estimate of 300 obtained by Thompson et al. (1992)
for the same habitat area. Re-analysis by CAPTURE of
Thompson et al.¢s 1988 data (Thompson et al., 1992) also
provided a larger estimate than their analysis, which used a
weighted mean Petersen method. A larger search area, more
capture occasions and analysis by a more powerful method
may have contributed to a more accurate estimate from our
study. The density of tuatara on North Brother Island is
approximately 159 ha)1, which is similar to the minimum
density of 100 tuatara ha)1 in good habitat on other rat-free
islands (Cree & Butler, 1993).
The total number of tuatara marked on North Brother
Island since 1988, 523, is higher than the upper profile
likelihood of our population estimate of 457. However, our
population estimate does not include juveniles, which
comprise about 10% of the marked individuals; also some
deaths will have occurred over the 10 years. It is also
possible that some individuals were mis-identified because of
natural losses of toes.
Mean annual survival of adult tuatara on North Brother
Island is c. 95% and appears to have been stable for the past
decade. Compared with other New Zealand reptiles, tuatara
are live longer. The longevity of North Brother Island
tuatara (at least 61–71 years) is almost twice that of the next
longest record of 36 years for a New Zealand reptile,
Hoplodactylus duvaucelii (Thompson et al., 1992). Tuatara
live as long as crocodilians in the wild (70 years) and
possibly as long as turtles, which can live beyond 100 years
(Pritchard, 1967; Graham & Hutchison, 1969; Webb &
Manolis, 1988). High survivorship and extended life span
are expected traits for the survival of a species that matures
late and has low reproductive output (Cree, 1994).
The sex ratio of adult tuatara on North Brother Island
appears to be skewed strongly towards males. This could be
true or could be an artefact reflecting the inability of the
methods to account for some animals in a population that
are effectively invisible (Otis et al., 1978). Some females, for
example, may leave their burrows less often than males and
be effectively uncatchable over a five-night trip.
As adult males and females are estimated to have similar
survival, an obvious explanation for the sex ratio is a bias in
recruitment. Sex of tuatara is determined by the temperature
of egg incubation (TSD; Cree et al., 1995), with more males
produced as the temperature increases. Consistently warm
soil temperatures could lead to a male-biased population. In
turtles with TSD, sex ratios are commonly biased towards
female hatchlings (Limpus et al., 1985; Bull & Charnov,
1989; Ewert & Nelson, 1991). In crocodilians, the sex ratio
may deviate from 50 : 50, but not necessarily with a female
bias (Ouboter & Nanhoe, 1989; Thorbjarnarson, 1990;
Cooper-Preston, 1991; Lance et al., 2000).
The key demographic threats to the North Brother Island
population of tuatara are the disproportionately small
number of adult females and small total number of tuatara.
An adult population of 349 tuatara is likely to represent a
much smaller effective reproductive population. How much
smaller is not known as key information required to calculate
this figure is unavailable for this species. The number of
reproductive females may be small enough to compromise
the viability of the population due to demographic stochasticity, loss of genetic variation, and/or Allee effects (Lacy,
1993; Stephens & Sutherland, 1999). In addition, global
warming could contribute to further skewing of the population sex ratio (because of TSD) if female tuatara are not able
to adjust their nesting behaviour according to environmental
cues. Predicted temperature increases (1.4–5.8 C; IPCC,
2001) may also result in reduced recruitment as a result of
lethal temperatures in nests.
The first tuatara were marked on North Brother Island
45 years ago, contributing to estimates of longevity, but it is
only work over the past decade that has allowed estimates of
population size, survival and sex ratio. Current estimates of
survival and longevity are consistent with the long-term
survival of the North Brother Island tuatara population.
However, a male-biased sex ratio, genetic effects of a small
population size, and environmental (e.g. global warming)
and chance events (e.g. rat invasion and fire) pose significant
risks. We recommend no further females be removed from
the island for translocation until more is known about
natural fluctuations in sex ratio and population size.
ACKNOWLEDGMENTS
We thank Victoria University of Wellington and The
Zoological Society of San Diego for financial support. The
New Zealand Department of Conservation provided permits
and logistical support. Te Atiawa Manawhenua ki te Tau
Ihu Trust allowed us to conduct research on North Brother
Island. Victoria University of Wellington Animal Ethics
Committee provided ethics approval for the project. We
2002 Blackwell Science Ltd, Journal of Biogeography, 29, 633–640
Population size and sex ratio of tuatara 639
especially thank all those who gave us access to their data, in
particular Dick Barwick, Mike Thompson and Bill Cash,
and all the enthusiastic tuatara searchers who contributed to
fieldwork. We thank an anonymous referee for comments on
the manuscript.
REFERENCES
Benton, M.J. (1990) Vertebrate palaeontology: biology and
evolution. Harper Collins Academic, London.
Bull, J.J. & Charnov, E.L. (1989) Enigmatic reptilian sex ratios.
Evolution, 43, 1561–1566.
Carothers, A.D. (1973) Capture-recapture methods applied to a
population with known parameters. Journal of Animal
Ecology, 42, 125–146.
Castanet, J., Newman, D.G. & Saint Girons, H. (1988) Skeletochronological data on growth, age and population structure
of the tuatara, Sphenodon punctatus, on Stephens and Lady
Alice Islands, New Zealand. Herpetologica, 44, 25–37.
Cooper-Preston, H. (1991) Geographic variation in the population dynamics of Crocodylus johnstoni (Krefft) in three
rivers in the Northern Territory, Australia. PhD Thesis,
University of New England, New South Wales.
Cormack, R.M. (1992) Interval estimation for mark-recapture
studies of closed populations. Biometrics, 48, 567–576.
Cree, A. (1994) Low reproductive output in female reptiles from
New Zealand. New Zealand Journal of Zoology, 21, 351–372.
Cree, A. & Butler, D. (1993) Tuatara recovery plan (Sphenodon
spp.). Threatened Species Recovery Plan Series no. 9.
Department of Conservation, New Zealand.
Cree, A., Daugherty, C.H., Schafer, S.F. & Brown, D. (1991)
Nesting and clutch size of tuatara (Sphenodon guntheri) on
North Brother Island, Cook Strait. Tuatara, 31, 9–16.
Cree, A., Thompson, M.B. & Daugherty, C.H. (1995) Tuatara
sex determination. Nature, 375, 543.
Dawbin, W.H. (1982) The tuatara Sphenodon punctatus:
aspects of life history, growth and longevity. New Zealand
herpetology (ed. D.G. Newman), pp. 237–250. New
Zealand Wildlife Service Occasional Publication no. 2. New
Zealand Wildlife Service, New Zealand.
Ewert, M.A. & Nelson, C.E. (1991) Sex determination in
turtles: patterns and some possible adaptive values. Copeia,
1991, 50–69.
Gilbert, R.O. (1973) Approximation of the bias in the Jolly–
Seber capture-recapture model. Biometrics, 29, 501–526.
Graham, T.E. & Hutchison, V.H. (1969) Centenarian box
turtles. Journal of the International Turtle and Tortoise
Society, 3, 25–29.
Hiby, A.R. & Jeffery, J.S. (1987) Census techniques for small
populations, with special reference to the Mediterranean monk
seal. Zoological Society of London Symposia, 58, 193–210.
IPCC (2001) Climate change 2001: the scientific basis. Intergovernmental panel on climate change. Cambridge University
Press, Cambridge.
IUCN (1996) Red list of threatened animals. IUCN, Gland,
Switzerland.
Jenkins, M. (1992) Species extinction. Global biodiversity:
status of the earth’s living resources (ed. B. Groomsbridge),
pp. 192–205. Chapman & Hall, London.
2002 Blackwell Science Ltd, Journal of Biogeography, 29, 633–640
Jolly, G.M. (1965) Explicit estimates from capture-recapture
data with both death and immigration-stochastic model.
Biometrika, 52, 225–247.
Lacy, R.C. (1993) Impacts of inbreeding in natural and captive
populations of vertebrates: implications for conservation.
Perspectives in Biology and Medicine, 36, 480–496.
Lacy, R.C., Hughes, K.A. & Miller, P.S. (1995) VORTEX: a
stochastic simulation of the extinction process, Version 7,
Users Manual. IUCN/SSC Conservation Breeding Specialist
Group, Apple Valley, USA.
Lance, V.A., Elsey, R.M. & Lang, J.W. (2000) Sex ratios of
American alligators (Crocodylidae): male or female biased.
Journal of Zoology, London, 252, 71–78.
Limpus, C.J., Reed, P.C. & Miller, J.D. (1985) Temperature
dependent sex determination in Queensland sea turtles: intraspecific variation in Caretta. Biology of Australasian frogs and
reptiles (eds G. Grigg, R. Shine, R. and H. Ehmann), pp. 343–
351. Royal Zoological Society, New South Wales.
Miller, C.I. & Ford, L.D. (1988) Managing for nature conservation. Bioscience, 38, 456–457.
Nelson, N.J., Keall, S.N., Brown, D. & Daugherty, C.H. (2002)
Establishing a new wild population of tuatara (Sphenodon
guntheri). Conservation Biology, 16, 1–9.
Nichols, J.D., Hines, J.E. & Pollock, K.H. (1984) Effects of
permanent trap response in capture probability on Jolly–
Seber capture-recapture model estimates. Journal of Wildlife
Management, 48, 289–294.
Norris, J.L. & Pollock, K.H. (1996) Nonparametric MLE under
two closed capture-recapture models with heterogeneity.
Biometrics, 52, 639–649.
Otis, D.L., Burnham, K.P., White, G.C. & Anderson, D.R.
(1978) Statistical inference from capture data on closed
animal populations. Wildlife Monographs, 62, 1–135.
Ouboter, P.E. & Nanhoe, L.M.R. (1989) Notes on the
dynamics of a population of Caiman crocodilus crocodilus
in Northern Surinam and its implications for management.
Biological Conservation, 48, 243–264.
Pledger, S. (2000) Unified maximum likelihood estimates for
closed capture-recapture models using mixtures. Biometrics,
56, 434–442.
Pollock, K.H. (1982) A capture-recapture design robust to
unequal probability of capture. Journal of Wildlife Management, 46, 752–757.
Pollock, K.H., Hines, J.E. & Nichols, J.D. (1985) Goodness of
fit tests for open capture–recapture models. Biometrics, 41,
399–410.
Pollock, K.H., Nichols, J.D., Hines, J.E. & Brownie, C. (1990)
Statistical inference for capture–recapture experiments. Wildlife Monographs, 107, 1–97.
Pritchard, P.C.H. (1967) Living turtles of the world. THF
Publications, Neptune City, NJ.
Seber, G.A.F. (1965) A note on the multiple-recapture census.
Biometrika, 52, 249–259.
Seber, G.A.F. (1973) Estimation of animal abundance and
related parameters. Griffin, London.
Smith, J.L. & McDougal, C. (1991) The contribution of
variance in lifetime reproduction to effective population size
in tigers. Conservation Biology, 5, 484–490.
Sokal, R.R. & Rohlf, F.J. (1981) Biometry. W.H. Freeman,
New York.
640 N. J. Nelson et al.
Stanley, T.R. & Burnham, K.P. (1998) Estimator selection for
closed-population capture–recapture. Journal of Agricultural,
Biological and Environmental Statistics, 3, 131–150.
Stephens, P.A. & Sutherland, W.J. (1999) Consequences of the
allee effect for behaviour, ecology and conservation. Trends
in Ecology and Evolution, 14, 401–405.
Thompson, M.B., Daugherty, C.H., Cree, A., French, D.C.,
Gillingham, J.C. & Barwick, R.E. (1992) Status and longevity
of the tuatara, Sphenodon guntheri, and Duvaucel’s gecko,
Hoplodactylus duvaucelii, on North Brother Island, New
Zealand. Journal of the Royal Society of New Zealand, 22,
123–130.
Thorbjarnarson, J.B. (1990) Ecology and behavior of the
spectacled Caiman (Caiman crocodilus) in the Central Venezuelan Llanos. PhD Thesis, University of Florida, Gainsville.
Webb, G.J.W. & Manolis, S.C. (1988) Australian saltwater
crocodiles (Crocodylus porosus). G. Webb Pty Ltd, Darwin.
White, G.C., Anderson, D.R., Burnham, K.P. & Otis, D.L.
(1982) Capture–recapture and removal methods for sampling
closed populations. Los Alamos, NM.
BIOSKETCHES
Nicola Nelson is a PhD student with interests in translocation, population viability, demography, sex determination and performance of tuatara.
Susan Keall is a technical officer with 15 years experience
in field and laboratory-based projects on the conservation
ecology of New Zealand amphibians and reptiles, in
particular, tuatara.
Shirley Pledger is a statistician with research interests in
capture–recapture models.
Charles Daugherty studies ecology and genetics of indigenous New Zealand species to support their conservation.
2002 Blackwell Science Ltd, Journal of Biogeography, 29, 633–640