650 words 3 hrs

Predictors of translocation success of captive-reared lizards: implications for their captive management

K. M. Hare1, N. Schumann2, A. J. Hoskins3, C. H. Daugherty1, D. R. Towns4,5 & D. G. Chapple2

1 Centre for Biodiversity and Restoration Ecology, Victoria University of Wellington, Wellington, New Zealand

2 School of Biological Sciences, Monash University, Clayton, VIC, Australia

3 CSIRO Townsville, Australian Tropical Science & Innovation Precinct, James Cook University, Townsville, QLD, Australia

4 New Zealand Department of Conservation, Auckland, New Zealand

5 Institute for Applied Ecology New Zealand, School of Sciences, Auckland University of Technology, Auckland, New Zealand

Keywords

locomotor performance; egg-laying skink;

Suter’s skink; life history; incubation

temperature; New Zealand; translocations;

captive rearing.

Correspondence

David G. Chapple, School of Biological

Sciences, Monash University, Wellington Rd,

Clayton 3800, Australia.

Email: [email protected]

Present address

School of Graduate Research, University of

Waikato, Hamilton, New Zealand

Editor: John Ewen

Associate Editor: Carolyn Hogg

Received 07 May 2019; accepted 24

September 2019

doi:10.1111/acv.12544

Abstract

Islands are biodiversity hotspots, but their native inhabitants are vulnerable to pre- dation from exotic predators. Conservation of island endemics has often involved translocating captive-reared populations to predator-free refugia. However, the long-term success of these translocations has rarely been assessed. We investigated the traits that maximize post-translocation survival in a cohort of captive-reared Suter’s skinks Oligosoma suteri and compared traits normally associated with sur- vival and persistence of lizards (body condition, speed and overall size) to that of wild-born skinks 6 years after release onto a predator-free island in the north-east of New Zealand. Our models showed that larger lizards, and lizards with lower body condition, had improved survival. While sprint speed of captive-reared lizards did not differ significantly to that of wild skinks, diving ability of captive-reared skinks was poor, with only female captive-reared lizards diving during trials. Our results indicate that the traits associated with higher survival after release are not necessarily obvious and may be influenced by adaptation to captive conditions. Long-term monitoring post-translocation is therefore vital to determine the success of the translocation.

Introduction

Globally, islands represent important biodiversity hotspots (Myers et al., 2000). They support a large number of endemic taxa, and often contain species or lineages that are predomi- nantly, or entirely, represented on islands (Cowie & Holland, 2003; Whittaker & Fern�andez-Palacios, 2007). Thus, when adjusted for area, islands collectively support a disproportionate number of the world’s species (Myers et al., 2000; Whittaker & Fern�andez-Palacios, 2007). Insularity has favoured selection for low reproductive output and high longevity (Cree, 2005; Whit- taker & Fern�andez-Palacios, 2007; Covas, 2012), while release from predation pressure on islands has resulted in a loss or modi- fication of anti-predator behaviours (Blumstein & Daniel, 2005; Whittaker & Fern�andez-Palacios, 2007). Combined with their restricted geographic ranges, these island-specific behavioural traits mean that island species are especially vulnerable to anthropogenic threats (IUCN, 2018).

A key threat to island fauna is the introduction of exotic species, which compete with island biota, alter habitat and prey upon native species (Towns et al., 2006; Doherty et al., 2016). Introduced vertebrate predators, especially rodents, pose a particular threat. These have established on numerous offshore islands where they prey upon na€ıve island fauna, reducing reproductive success and recruitment and displacing and suppressing island populations (reviewed in Towns et al., 2006). The consequences for island endemics are dev- astating (Towns et al., 2006; Stolzenburg, 2011), with intro- duced predators implicated in a considerable number of extinctions and declines. Indeed, a disproportionate number of island species have been lost to extinction (reviewed in Doherty et al., 2016). Accordingly, islands are also hotspots of extinction (Myers et al., 2000; Whittaker & Fern�andez- Palacios, 2007).

New Zealand is a world leader in conservation manage- ment of species threatened by introduced predators

Animal Conservation �� (2019) ��–�� ª 2019 The Zoological Society of London 1

Animal Conservation. Print ISSN 1367-9430

(Armstrong & McClean, 1995; Romijn & Hartley, 2016; Towns et al., 2016a). Management has focused on eradicat- ing mammalian predators from islands before translocating native species to these island refugia, either to reinforce existing populations or to found new populations (Towns et al., 2016a,b, Hitchmough et al., 2016a). Translocated indi- viduals may be bred in captivity (Seddon et al., 2007; Maran et al., 2009) and are often ‘head-started’ prior to their release (Connolly & Cree, 2008; Jarvie et al., 2015). While the suc- cess of wildlife translocations has traditionally been difficult to quantify, Miller et al. (2014) proposed a staged approach to define translocation success using four standardized crite- ria: survival and growth of individuals, evidence of repro- duction, population growth and population viability. Based on this approach, a translocation is considered a success when released populations meet or exceed one of the four stages during monitoring (Miller et al., 2014).

Miller et al.’s (2014) standardized criteria for translocation success were developed using data from the New Zealand herpetofauna, of which lizards are the major components. New Zealand has a diverse, endemic lizard fauna comprising ~104 species (61 Eugongylinae skinks, 43 Diplodactylidae geckos; van Winkel et al., 2018; Hitchmough, 2016). How- ever, 83% of species are regarded as threatened or at risk, with introduced mammalian predators representing the pri- mary threat (Tingley et al., 2013; Hitchmough et al., 2016a; Towns et al., 2016b). Consequently, many New Zealand lizard species have declined on the main islands (North Island, South Island) and primarily persist on smaller, off- shore islands (Chapple & Hitchmough, 2016; Towns et al., 2016b).

Conservation attempts have involved translocating lizards to offshore islands or mainland refugia, and in New Zeal- and, captive management is increasingly being used to found populations for translocation to new sites. However, the success of such translocations has generally been unde- termined or assessed over relatively short timescales; longer term monitoring (>4 years) is required to direct future translocation efforts (Romijn & Hartley, 2016; Towns et al., 2016a). We investigated the success of a transloca- tion of a population of captive-reared Suter’s skinks on an offshore island of New Zealand as well as the factors likely to predict success. To this end, we used as our case study a cohort of captive-reared Suter’s skinks (Oligosoma suteri) that were hatched from eggs of females collected from Green Island off the north-east coast of New Zealand and head-started to 18 months of age before being translo- cated to nearby Korupuki Island in 2001. This represents the only translocated population of captive-reared lizards in New Zealand for which long-term data are available. Loco- motor performance of recaptured individuals from this pop- ulation was compared to wild individuals captured from Korapuki Island in 2006 (Miller et al., 2010). The wild population is the result of a successful translocation of 10 male and 20 gravid female founders successfully translo- cated from Green Island to Korapuki Island in 1992 (see Towns & Ferreira, 2001 for details of this translocation). The translocation sites were similar in appearance, with

similar sized rocks. However, the Towns & Ferreira (2001) translocation of adults to the northern beach was north fac- ing, and likely received more sunshine hours, and had a gentle slope up to the forest; the second translocation of juveniles was to a southern facing beach with a cliff face up to the forest.

The nocturnal Suter’s skink is the only oviparous lizard species endemic to New Zealand (Cree & Hare, 2005). It is a medium-sized (maximum snout-vent length (SVL) of c. 108 mm, Hardy, 1977) skink that is restricted to rocky bea- ches of northern New Zealand (Hare et al., 2008a; Chapple & Hitchmough, 2017), presumably due to cooler tempera- tures at more southerly latitudes (Hare et al., 2004). It is vulnerable to introduced predators (Towns et al., 2003) and has experienced significant range reductions in recent years, with the species now largely limited to offshore islands (Towns & Ferreira, 2001; Towns et al., 2016b). Individuals are long-lived, reaching at least 12 y in the wild (Towns & Ferreira, 2001), and typically deposit their first clutch at ~33 months after hatching (Towns, 1975). The breeding sea- son is broadly mid-Spring to Summer, with ovulation occur- ring in October–November and oviposition in late December–early January (Towns, 1975).

Hare et al. (2012) reported the short-term (~12–18 months pre-release) outcomes of captive rearing prior to release of Suter’s skinks to Korupuki Island. Here we report on the longer-term outcomes (~5.5 years) of the translocated Suter’s skink population to assess the success of this translocation as per Miller et al. (2014), and explore the factors that influ- ence translocation success. Specifically, we examine:

1 the key factors that can be measured prior to release that impact survival of captive-reared lizards in the wild (sex, body condition, mass at hatching and at 12 months, and sprint speed at 4 months, selected because these are com- monly measured during translocation programmes and are therefore relevant to other programmes attempting to gauge translocation success), and

2 how locomotor performance of translocated captive-reared adults compares to wild individuals post-release.

We expected that over the timeframe of the study, evi- dence of survival, growth and reproduction (Miller et al., 2014) would be detected if this translocation has been suc- cessful, but longer term data would be necessary to evalu- ate whether this population has met Miller et al.’s (2014) remaining criteria for translocation success (population growth and viability). Our study will help determine whether captive rearing programmes represent a viable con- servation technique for Suter’s skinks, and will facilitate consideration of the potential impacts of captive rearing on the likelihood of translocation success in lizards. In addi- tion to furthering our understanding of what influences translocation success, much of the information revealed in this study will assist managers to predict the potential qual- ity of captive-reared founders (i.e. individuals that are likely to produce successful translocation outcomes). This will enable the most suitable candidates for release into the wild to be identified.

2 Animal Conservation �� (2019) ��–�� ª 2019 The Zoological Society of London

Translocation success in lizards K. M. Hare et al.

Materials and methods

Egg collection & incubation

Details of egg collection and incubation are described in full in Hare et al. (2002), and in Hare et al. (2004, 2008b). Briefly, eggs were collected from 58 female Suter’s skinks, sourced from Green Island, Mercury Island group, New Zealand (36°38’S, 175°51’E, 1–2 m a.s.l.) in Summer, 1999. The eggs were randomly assigned to one of three incubation temperatures (18°C, 22°C and 26°C) and one of two water potentials (�120 and �270 kPa). Each clutch (two to five eggs per clutch, Hare et al., 2002) was (where possible) rep- resented at all three temperatures, but at only one water potential per temperature. Water potential of the vermiculite incubation medium had no influence on any variables by 12 months of age (Hare et al., 2004; 2008b) and is not con- sidered further.

Juvenile maintenance

Immediately after hatching, we attempted to identify sex, and individuals were weighed (�1 mg), measured (SVL and tail length, �0.5 mm) and permanently marked by toe clip- ping. Skinks were initially housed in groups of three under the same conditions that gravid females were provided, in transparent boxes (215 9 330 9 110 mm, 7 L) with a lid containing 1 mm2 wire mesh (165 9 120 mm) for ventila- tion. At least 20 mm of moist, non-fertilized potting mix was used as the substrate and pieces of bark were provided for shelter. A heat strip at one end of each box provided a temperature range of 14–30°C. At 5–8 months of age, juve- niles were re-sorted by size into groups of 6–12 in metal enclosures (700 9 580 9 350 mm) with 1 mm2 mesh lids. The metal enclosures were lined with 50 mm of sand, with driftwood, dried seaweed and small stones for shelter. Juve- niles continued to be periodically re-sorted by size.

Juveniles were given free access to water and fed ad libi- tum with live food items and fish-based cat food. Live food was regularly supplemented with calcium powder and vita- min drops (Avi-CalTM; calcium borogluconate 200 g/L, cholecalciferol (Vitamin D3) 12 500 IU/L and magnesium sulphate 5 g/L). Day/night was simulated using UV bulbs (Duro-test� True-lite� power twist fluorescent tubes) on a 12L:12D light cycle.

Release, monitoring, survival and reproduction post-release

Of the 136 hatchlings, 94 (69.1%) survived the 18 months until release in early October 2001 (Hare et al., 2004; this study). These were released en masse on a south-facing beach (hereafter Release Beach) on Korapuki Island, Mer- cury Island group, New Zealand (36°39’S, 175°50’E 36, 1- 2 m a.s.l), ~1.75 km from the source location of Green Island. The release site is separated from other beaches on the island by steep cliffs; while Suter’s skinks have

previously been translocated to Korapuki Island (Towns & Ferreira, 2001), those skinks in previous translocations were released on a north-west facing beach ~200 m (coastal dis- tance) from Release Beach. At that time there was no evi- dence of existing Suter’s skink individuals occupying Release Beach (Towns unpubl. data). Both translocated pop- ulations of lizards were exposed to a suite of native preda- tors, and for New Zealand lizards it included birds, bigger reptiles and invertebrates (Hare et al., 2016).

We conducted a mark–recapture exercise in 2005 and 2007, approximately 4–6 years after release. Due to the cost and per- mits required to undertake research on protected islands in New Zealand, the timing and effort involved in the monitoring trips was opportunistic, and part of much larger scientific events. In 2006, we measured the locomotor performance of recaptured individuals, and were therefore unable to do a mark–recapture study as individuals were captured for the experiments. For mark–recapture studies, the site was monitored over 5 days on 18–22 November 2005 and 28 February–4 March 2007. Five pitfall trapping sites were established: three on Release Beach, and one each on beaches to the east and west of Release Beach (to determine if skinks were expanding around the island). On Release Beach, trapping sites were located: (A) at the initial release site, (B) 20 m west of the initial release site and (C) 40 m west of the initial release site. At Sites A and B, pitfall traps were set at 2-m intervals in three lines of five, positioned from the storm line (bottom of the cliff face) to the high-tide mark. At Site C and the western beach, we placed two sets of four traps midway between the storm line and the high-tide mark at 2-m intervals. At the eastern beach, we used a perma- nent lizard monitoring transect of 20 traps in five stations, each 20 m apart along the storm line with four traps set on the cor- ners of each 2 9 2 m station. All traps comprised 4l plastic paint pails that were provided with cover and baited daily with canned cat food. Traps close to the high-tide mark were cov- ered with galvanized netting to stop entry by large crabs, which are known to prey upon trapped lizards (Hare et al., 2016) and indeed even wild-ranging lizards (Bell & Bauer, 2008). We placed large flat rocks upon the traps, with gaps beneath, to pre- vent captured lizards from overheating in the sun. On the south-eastern beach, each of the four traps at each station was baited with either fish- or meat-based cat food, canned pear, or left unbaited. Except for traps placed on the south-eastern beach, all traps were set forc. 12 h, and checked in the early morning and late evening for 5 days. The south-eastern beach traps were checked at intervals of 48 h and set three times over 6 days.

Sex, mass (�0.5 g), SVL (�0.5 mm) and for females, gravidity status were recorded for recaptured individuals dur- ing each year of trapping. Sex was determined by eversion of the hemipenes in males. Gravidity status and an estimate of clutch size were obtained via abdominal palpation by a trained expert (KMH).

Locomotor performance

We used sprint speed and dive duration as measures of loco- motor performance post-release in 2006. The methods used

Animal Conservation �� (2019) ��–�� ª 2019 The Zoological Society of London 3

K. M. Hare et al. Translocation success in lizards

to measure these locomotor performance traits in other O. suteri individuals are described in detail in Miller et al. (2010). Briefly, both speed and diving trials were conducted under natural photoperiods (14:10 light:dark cycle, with sun- rise at ~0600 h) and at ambient air temperatures averaging 17.5 � 0.05°C. Three trials were conducted within 1 day for both performance measures, and all individuals were at least 2 days post-absorptive. Dive trials were conducted after com- pletion of sprint trials, with a minimum of 120 min rest between trials of the two performance measures.

Sprint speed was measured between 0930 h and 1230 h using a field portable timer (Huey et al., 1981). The order of individuals was randomized over the three trials, but in all cases, individuals were given at least 15 min rest between trials. A plastic racetrack (1.5 9 0.8 m), with five paired infrared lights (0.25 m apart and 4 mm high) over 1 m, was used to test sprint speed. The lights transmitted and received an infrared beam horizontally across the track, and the inter- ruption of each successive infrared beam stopped its paired timer. The fastest speed over 0.25 m was used in analyses since burst speed is likely to be a more ecologically relevant measure for species that live on rocky shores where long sprints are not necessary to reach cover. To encourage sprinting, we gently touched the tails of lizards with a paint- brush. We also recorded the number of pauses, scored when an individual ran forwards after its tail was touched before stopping and requiring another touch to continue, over 1 m. Performance in captive-reared individuals was compared with that of wild-born individuals, captured from the northern release beach; these data were published in Miller et al. (2010).

Voluntary diving trials were undertaken in an artificial rock pool between 1530 h and 1930 h. The order of individ- uals was randomized over the three trials and animals were given at least 60 min rest between each dive trial. The artifi- cial rock pool comprised a darkened round plastic bin (base 340 mm diameter; 20 L to fill line) filled with rocks (in a centre spire with one emergent rock), seaweed and intertidal seawater (mean temperature 16.7 � 0.07°C; modified from Hare & Miller, 2009). Skinks were placed on the emergent rock facing away from the researcher and diving was encour- aged by tapping on the tail. Some skinks did not dive, and instead swam about on the surface of the water; if these ani- mals did not dive after 10 sec, they were removed. If an ani- mal dived, we recorded dive duration (i.e. the time between submergence and emergence) using a manual stopwatch (ac- curate to 1 sec). Dive duration of captive-reared skinks was compared to that of wild-born individuals, using data from Miller et al. (2010).

Statistical analyses

In order to determine the factors that influenced long-term survival of captive-reared Suter’s skinks, we fitted a Baye- sian Cormack-Jolly-Seiber (CJS) survival model (hereafter, full model). To account for differences in trap days, detec- tion probability was included as a fixed effect with a random effect of year (eq. 1).

logit observationð Þ � 1 þ randomeffect year of surveyð Þ ((1))

Survival probability was modelled as a function of sex, mass at 12 months, mass at hatching and body condition (eq. 2). In order to account for the 5-year gap between release and the first trapping period, year since release was treated as a two level factor (i.e. Year 1 and the following years) within the survival component of the model (eq. 2). We calculated body condition according to whether individu- als were heavier or lighter than predicted for their length based on the residuals from the estimated mass using log/log allometric equations.

logit survivalð Þ � mass12months þ masshatching þ body cond12months þ factor yr since releaseð Þ (2)

The influence of sprint speed at 4 months on survival post-release (Hare et al, 2008b) was incorporated into a sep- arate model, following the same structure, using a reduced dataset to include only those individuals that were run in sprint speed trials upon recapture (hereafter, reduced model). Snout–vent length (SVL) was omitted from all models due to its high correlation with mass. Analyses were carried out using JAGS (Plummer, 2003) through R (R Core Team, 2017) using the r2hags package (Su & Yajima, 2015).

Sprint speeds of captive-reared individuals were compared to those of 63 wild-caught individuals, all males, from Miller et al. (2010) using a linear regression that modelled sprint speed against the interaction of snout–vent-length and type (captive-reared or wild-caught) (eq. 3). To enable a fair com- parison with adult wild-caught animals, sprint speeds of cap- tive reared individuals were those measured upon recapture as adults and not the juvenile sprint speed used in the sur- vival analysis.

sprint speed � snout � vent � length � factor typeð Þ (3)

Results

Survival, size and reproduction

In total, 21 captive-reared individuals were recaptured over the 3 years, with an additional three unmarked individuals captured in 2006 and 2007 (SVL range = 55–80 mm). Of the marked individuals, 17 captures of 10 individuals were made in 2005, 10 individuals were trapped in 2006 and 25 captures of 17 individuals were recorded in 2007. Recap- tured individuals were incubated at either 22°C (n = 12) or 26°C (n = 9); no individuals incubated at 18°C were recap- tured. All recaptures were from Release Beach, and most were from the original release site (transect A; 78% of cap- tures), with some from transect B (20 m south of the release site; 22% of captures), and none from transect C (40 m south of the release site). No Suter’s skinks were captured from beaches directly north or south of Release Beach.

Size of individuals increased during the course of the 7- year study, with growth rate averaging 7.1 mm per year. At

4 Animal Conservation �� (2019) ��–�� ª 2019 The Zoological Society of London

Translocation success in lizards K. M. Hare et al.

hatching, mean SVL and mass of the 21 individuals was 34.9 � 0.16 mm and 0.77 � 0.01 g, respectively, and by November 2005, all recaptured individuals had attained size- based maturity (Towns, 1975, Table 1). The average size and weight of individuals continued to increase during the course of the study (Table 1).

In November 2005, three of five females were gravid with 1–2 eggs, and in November 2006, all five females captured were gravid with 1–3 eggs. Two females were captured in both 2005 and 2006, and of these, one was gravid with one egg in both years, and the other was only gravid in 2006.

Factors influencing survival probability

The full model indicated strong support (at the 95% confi- dence level) for a positive influence of mass at 12 months on lizard survival, with larger lizards experiencing greater survival, but no influence of mass at hatching (Table 2, Fig. 1). In contrast, body condition negatively impacted sur- vival. This suggests that individuals who are heavier, but lighter than predicted for their length, are more likely to sur- vive (Fig. 2). These findings were replicated in the reduced model, which also did not show support (at the 95% confi- dence level) for a relationship between sprint speed and sur- vival (Table 3).

Locomotor performance comparisons

Gravidity in female skinks has been shown to affect sprint speed (Shine, 1980). Unfortunately, we only had sprint speed information for gravid captive-reared females (present study)

Table 1. Mean snout–vent length (SVL) and weight of Suter’s

skinks Oligosoma suteri throughout the course of the study. Values

in parentheses represent range

Date (age) SVL (mm) Mass (g)

Feb–Jun 2000

(hatching)

34.9 � 0.16 (28-38) 0.76 � 0.01 (0.33–0.92)

Feb–Jun 2001

(12 months)

49.9 � 0.3 (40–56) 2.07 � 0.04 (0.77–2.82

November 2005

(<6 years)

80.56 � 0.85 (75–84) 10.17 � 0.41 (8.5–13)

November 2006

(<7 years)

83.4 � 0.48 (82–87) 10.58 � 0.19 (9.75–11.8)

March 2007

(7 years)

84.81 � 0.59 (81–89) 10.81 � 0.38 (8.25–12.8)

Table 2. Results of Bayesian BJS survival model for the survival of

captive raised skinks released into the wild. Italicized and

underlined values indicate those where a strong level of support

(95% CIs do not cross 0) for their influence on survival were

detected

Mean SE 2.50% 97.50%

Mass (12 mo) 1.056 0.486 0.248 2.132

Mass (0 mo) 0.013 0.332 �0.590 0.707 Condition �0.945 0.489 �2.074 �0.157 Year(1) �1.313 0.320 �1.958 �0.711 Years(2:4) 82.229 58.697 7.444 222.100

Intercept (obs. probability) �1.402 0.562 �2.160 �0.635 Sigma 0.467 0.785 0.012 2.096

Figure 1 Estimated probability of survival from year one showing

individual influential responses for body condition (blue) and mass

(red) in Suter’s skinks (Oligosoma suteri) [Colour figure can be

viewed at wileyonlinelibrary.com]

Figure 2 Observed (points) and estimated (dashed line) mass of

Suter’s skinks (Oligosoma suteri) by snout–vent length. Estimated

line indicates allometric equation (log[mass] ~ log[length]) gener-

ated from the observed data. Point colour indicates the estimated

probability of survival based on fitted survival model. Note: X-axis

values have a precision s mm but have been ‘jittered’ to improve

visibility [Colour figure can be viewed at wileyonlinelibrary.com]

Animal Conservation �� (2019) ��–�� ª 2019 The Zoological Society of London 5

K. M. Hare et al. Translocation success in lizards

and non-gravid, wild females, caught from the northern Release beach (Miller et al., 2010). Therefore, to avoid the confounding effect of reproductive status, we did not com- pare speed in these. Hence, only males were included in sta- tistical comparisons of sprint speed. These comparisons indicated no significant differences in maximum sprint speeds between captive-reared and wild-born males (P > 0.05 for all covariates), with average sprint speeds of 1.05 � 0.06 m/s (range 0.89–1.21) and 1.12 � 0.03 (range 0.79–2.21) attained by captive-reared and wild-caught male, respectively. Captive-reared (gravid) and wild-caught (non- gravid) females achieved mean speeds of 0.74 � 0.13 (range 0.45–1.05) and 1.05 (0.87–1.27) respectively.

Captive-reared lizards rarely dived; of the 10 captive- reared individuals (five males and five gravid females) used in dive trials, only four (40%) dived in at least one trial. Two skinks (20%) dived in all three trials and six (60%) did not dive in any of the trials. Instead, for the majority of dive trials, lizards swam about the water surface. All individuals who dived were female, with dive durations averaging 124 � 35.5 s (range 22.2–185.8 s).

Discussion

Larger, leaner lizards have a higher probability of survival following translocation

Unlike previous studies on survival of this species, which investigated the factors influencing survival in captivity (e.g. Hare et al., 2004, 2008b), our study presents novel data on long-term post-release outcomes for captive-reared Suter’s skinks. These indicate that the best predictors of survival of Suter’s skinks post-translocation are size and body condition; both larger lizards and lizards with lower body condition at 12 months experienced greater survival. The magnitude of the size difference between skinks is large enough to be bio- logically significant, with a 16-mm and 2.05 g difference in SVL and weight, respectively, between the smallest and lar- gest individuals about 12 months old. These results have

implications for the increasing use of captive management for threatened species and suggest that larger, leaner lizards should be prioritized for translocations.

Higher survival might be expected in larger lizards for several reasons. Firstly, bigger individuals typically have lar- ger heads, with positive follow-on effects on bite force (Her- rel et al., 1999; Herrel et al., 2001; Verwaijen et al., 2002), prey handling efficiency and the ability to eat bigger and harder-bodied prey (Verwaijen et al., 2002). In addition, lar- ger lizards may have fewer predators than their smaller counterparts, with some predators preferentially selecting, or limited to, smaller prey individuals (e.g. Hasegawa, 1990; Webb & Shine, 1993). Finally, greater size confers a com- petitive advantage, with larger-bodied individuals more likely to win agonistic interactions (Tokarz, 1985; Cooper & Vitt, 1987; Sacchi et al., 2009), thereby maximizing their ability to secure key resources, such as territories (Sacchi et al., 2009), favourable perching sites (Tokarz, 1985), food (Stamps, 1977) and optimal basking and refuge sites (Car- others, 1981).

Large size also has reproductive benefits. Body size in skinks has a positive relationship with clutch and litter size (Cree & Hare, 2005; Chapple, 2006, 1994; Chapple et al., 2014), and Suter’s skinks are no exception, with females of a greater SVL producing significantly more eggs (Hare et al., 2002). Thus, not only did larger individuals experi- ence higher survivorship, but they likely also have greater reproductive output. This could potentially increase the over- all likelihood of translocation success (Towns et al., 2016a).

The reduced survival of skinks with higher body condition seems counter-intuitive. Lipid reserves of lizards provide an energy store during lean periods (Avery, 1970; Derickson, 1976), and support reproduction, growth and maintenance (Derickson, 1976), so it is logical that heavier translocated lizards, with their presumably greater fat reserves to draw on, should have improved survival. So why would Suter’s skinks experience the opposite following translocation? The answer may simply be that they are overly acclimated to captive conditions. These can influence phenotype, including social and anti-predator behaviour (Snyder et al., 1996; Con- nolly & Cree, 2008), and produce inactive, overweight indi- viduals (Connolly & Cree, 2008) that are less adept at capturing prey in the wild; when the fat reserves of these individuals are depleted, their survival is compromised. In contrast, Hare et al. (2012) found no influence of body con- dition index or size on survival post-release in captive-reared Otago skinks (O. otagense), potentially a function of differ- ences in sample size, species and/or the amount of time post-release over which monitoring was conducted. Nonethe- less, our results suggest that captive management can result in selection for unfavourable traits in lizards.

It is also possible that underlying factors that affect phe- notype, such as the rearing environment and maternal effects, are driving survival post-release. For example, maternal effects, which influence traits such as body size and condi- tion (e.g. Shine & Harlow, 1993; Wapstra, 2000), may be the causal drivers behind the relationships observed here rather than body size and condition per se. Regardless of the

Table 3. Results of Bayesian BJS survival model, on reduced

dataset including only those individuals where sprint speed

informations were detected, for the survival of captive raised

skinks released into the wild. Italicized and underlined values

indicate those where a strong level of support (95% CIs do not

cross 0) for their influence on survival were detected

Mean SE 2.50% 97.50%

Mass (12 mo) 1.805 0.702 0.621 3.402

Mass (0 mo) 0.005 0.533 �1.053 1.107 Condition �1.617 0.597 �2.883 �0.549 Sprint Speed �0.520 0.528 �1.656 0.441 Year (1) �1.867 0.438 �2.793 �1.085 Years (2:4) 83.890 60.963 7.400 229.463

Intercept (obs. probability) �1.369 0.450 �2.190 �0.551 Sigma 0.455 0.656 0.011 2.054

6 Animal Conservation �� (2019) ��–�� ª 2019 The Zoological Society of London

Translocation success in lizards K. M. Hare et al.

proximate mechanism, it is reasonable to use phenotypic traits, which can be easily and routinely measured by captive managers, as predictors of survival since phenotype is an expression of these underlying factors.

Lizards that were incubated at 18°C were not recaptured 4 or more years after the translocation, supporting Hare et al.’s (2004) suggestion that individuals incubated at this tempera- ture would not survive to maturity. Such low survival rates may be due to the smaller body sizes reached by these indi- viduals at 12 months (Hare et al., 2004) or a consequence of the increased rate of abnormalities from this incubation treat- ment (Hare et al., 2002). This result is congruent with previ- ous research, which has demonstrated a negative influence of incubation temperature extremes on survival and phenotype on reptiles (Noble et al., 2018), highlighting the importance of considering incubation regimes when designing transloca- tion programmes; individuals should not be incubated at tem- perature extremes, and if they are, should not be released due to reduced likelihood of translocation success.

As we predicted, the captive-reared Suter’s skinks translo- cated to Korapuki Island have met two of Miller et al.’s (2014) four criteria that indicate translocation success: sur- vival and growth, and evidence of reproduction. Individuals increased in body size throughout the course of our seven year study, reaching maturity by 2005 when the first recap- tures commenced. By this stage, most females were gravid with one or more eggs. Thus, this translocation can be tenta- tively deemed successful. Wild individuals translocated from Green Island to Korapuki Island have met all of Miller et al.’s (2014) criteria for success (Towns & Ferreira, 2001; Miller et al., 2011, 2014), with population growth detected within 3 years of release (Towns & Ferreira, 2001), and juveniles representing approximately 25% of the consistently high number of individuals captured since 2000 (e.g. 168 captured in 2006 from 30 founders released in 1992) during biennial monitoring (Miller et al., 2011, 2014). However, Towns & Ferreira (2001) cautioned that at least 20 years may be required before translocation success can be declared. Nonetheless, the success rate of translocations for captive-reared lizards has rarely been monitored and evalu- ated (e.g. Towns et al., 2016a). As such, our case study pro- vides valuable information that can inform future translocation efforts.

Captive-reared and wild individuals sprinted at equivalent speeds

Our study presents important data on the performance of cap- tive-reared skinks in the wild compared to that of wild-born individuals (Miller et al., 2010). Sprint speed of captive- reared males 5 years post-release was similar to that of wild males. Our result was unexpected in light of previous studies on Oligosoma skink species in which sprint speeds of cap- tive-reared lizards were considerably slower than those of wild individuals (Connolly & Cree, 2008; Hare et al., 2012). The slower speed has previously been attributed to heavier body mass of captive-reared skinks (Connolly & Cree, 2008;

Hare et al., 2012), reduced fear of humans or physiological changes owing to the restricted size of enclosures (Hare et al., 2012). These factors are unlikely to have influenced lizards in our study since they were not captive at the time of trials and had 5 years in the wild to develop their sprinting ability. The similarity between captive-reared and wild-born males in our study could represent a bias towards faster captive-reared adults if 1. sprint speed at 4months, which did not affect sur- vival probability, is not predictive of adult speed, and 2. adult survival is influenced by locomotor ability, resulting in only the fastest captive-reared males surviving to recapture. Alter- natively, maximum sprint speed in this species may simply not be a fitness-related trait. Instead, the proportion of maxi- mum speed used in nature, or ecological performance, could be more biologically relevant, and the trait upon which selec- tion acts (Irschick & Garland, 2001; Irschick, 2003; Husak, 2006). For example, Husak (2006) found that survival of yearling and adult collared lizards (Crotaphytus collaris) was dependent on the speed used by individuals to escape preda- tors, irrespective of their maximal sprinting capacity. Like- wise, Hoskins et al. (2017) proposed that the low repeatabilities of sprint speed in the closely related shore skinks (O. smithi) could indicate that skinks were not per- forming at their maximum capacity in field trials.

Diving was inhibited in captive-reared males

Only 40% (all female) of captive-reared skinks dived in one or more dive trials (cf 89% in Korapuki Island wild-born skinks: Miller et al., 2010), and while the proportion of females that dived was similar to the wild lizard average (89% of skinks dived in at least one trial), their dive dura- tions (22–186 s) were shorter than, or in the lower range of, dive durations by wild-caught lizards (71–1229 s: Miller et al., 2010). The willingness to dive may reflect predator escape behaviour in gravid females (Miller et al., 2010) since gravidity impedes sprint speed in lizards (Shine, 1980; Cooper et al., 1990). While the effects of small sample size cannot be discounted, it is possible that avoidance of diving by males and shorter dive durations by females were due to a lack of opportunity to acquire diving skills during crucial stages of development. Whatever the reasons for the diving behaviour observed here, our results imply a generally nega- tive impact of captivity on diving behaviour in Suter’s skinks, though additional research is needed to elucidate this further.

Conclusions

Our study has shown that the traits associated with post- translocation success may not always be intuitive, and could potentially be influenced by adaptation to captivity. We found that larger, leaner skinks had greater survival, suggest- ing that these should be preferentially selected for transloca- tions of this species, and that sprint speed of captive-reared individuals was equivalent to that of wild-born lizards. Our

Animal Conservation �� (2019) ��–�� ª 2019 The Zoological Society of London 7

K. M. Hare et al. Translocation success in lizards

study demonstrates the importance of long-term monitoring of translocated populations in order to ascertain success rates and identify key traits that positively influence translocation success. In turn, the most suitable candidates for future release can be determined. Similar quantitative assessments should be a fundamental component of future translocations globally.

Rearing animals in captivity until they attain a size where they are less likely to succumb to predation or starvation through the critical neonatal period (Ferguson et al., 1982) can improve survival post-release (Ferguson et al., 1982; Escobar et al., 2010). However, captive conditions can influ- ence phenotype in ecologically relevant ways (Burghardt (2013); Snyder et al., 1996; Connolly & Cree, 2008), and this must be addressed when establishing captive manage- ment programmes. Certainly, captive rearing should be con- sidered a last resort option only, and not all species are suitable for captive conditions. For example, western popula- tions of grand (O. grande) and Otago skinks (O. otagense) experienced poor survival and breeding in captivity (Hare et al., 2019). Therefore, where captive management is deemed appropriate, breeding facilities should provide ample opportunity for lizards to acquire the skills necessary to max- imize survival post-release. To that end, large enclosures comprising a variety of structurally complex microhabitats that closely mimic wild conditions, and allow individuals to display natural foraging, antipredator and social behaviours, should be established. For example, semiaquatic species would benefit from access to deep water to acquire diving skills, whereas arboreal species would benefit from climbing apparatus of different structures.

Acknowledgements

We thank all those who provided expertise, time and assis- tance, in particular Ian Atkinson, Iris Broekema, Chris Green, Sue Keall, Chris Longson, Kim Miller, Richard Moore, and Nicky Nelson. Funding was provided by the New Zealand Department of Conservation (DOC) and Victo- ria University of Wellington (VUW). The research was car- ried out with DOC approval, VUW Animal Ethics Committee approval, and after consultation with Ng�ati Hei, Ng�ati Maru and Ng�ati Whanaunga.

References

Armstrong, D.P. & McClean, I. (1995). New Zealand translocations: theory and practice. Pac. Conserv. Biol. 2, 39–54.

Avery, R.A. (1970). Utilisation of caudal fat by hibernating common lizards, Lacerta vivipara. Comp. Biochem. Physiol. 37, 119–121.

Bell, T. & Bauer, A.M. (2017). Predation on a free-ranging Raukawa gecko (Woodworthia maculata) by a purple rock crab (Leptograpsus variegatus). Biogecko Issue 4, 20–35.

Blumstein, D.T. & Daniel, J.C. (2005). The loss of anti- predator behaviour following isolation on islands. Proc.

Roy. Soc. B: Biol. Sci. 272, 1663–1668. https://doi.org/10. 1098/rspb.2005.3147

Burghardt, G.M. (2013). Environment enrichment and cognitive complexity in reptiles and amphibians: concepts, review, and implications for captive populations. Appl. Anim. Behav. Sci. 147, 286–298.

Carothers, J.H. (1981). Dominance and competition in a herbivorous lizard. Behav. Ecol. Sociobiol. 8, 261–266.

Chapple, D.G. (2003). Ecology, life-history, and behavior in the Australian Scincid genus Egernia, with comments on the evolution of complex sociality in lizards. Herpetol. Monogr. 17, 145–180.

Chapple, D.G. (2005). Life history and reproductive ecology of white's skink, Egernia whitii. Aust. J. Zool. 53, 353–360.

Chapple, D.G. & Hitchmough, R.A. (2016). Biogeography of New Zealand lizards. In New Zealand lizards: 109–131. Chapple, D.G. (Ed.). Switzerland: Springer.

Chapple, D.G., Miller, K.A., Chaplin, K., Barnett, L., Thompson, M.B. & Bray, R.D. (2014). Biology of the invasive delicate skink (Lampropholis delicata) on Lord Howe Island. Aust. J. Zool. 62, 498–506.

Connolly, J.D. & Cree, A. (2008). Risks of a late start to captive management for conservation: phenotypic differences between wild and captive individuals of a viviparous endangered skink (oligosoma otagense). Biol. Conserv. 141, 1283–1292.

Cooper, W.E. & Vitt, L.J. (1987). Deferred agonistic behavior in a long-lived scincid lizard Eumeces laticeps. Field and laboratory data on the roles of body size and residence in agonistic strategy. Oecologia 72, 321–326.

Cooper, W.E., Vitt, L.J., Hedges, R. & Huey, R.B. (1990). Locomotor impairment and defense in gravid lizards (Eumeces laticeps): Behavioral shift in activity may offset costs of reproduction in an active forager. Behav. Ecol. Sociobiol. 27, 153–157.

Covas, R. (2012). Evolution of reproductive life histories in island birds worldwide. Proc. Roy. Soc. B: Biol. Sci. 279, 1531–1537. https://doi.org/10.1098/rspb.2011.1785.

Cowie, R.H. & Holland, B.S. (2006). Dispersal is fundamental to biogeography and the evolution of biodiversity on oceanic islands. J. Biogeogr. 33, 193–198.

Cree, A. (1994). Low annual reproductive output in female reptiles from New Zealand. New Zeal. J. Zool. 21, 351–372.

Derickson, W.K. (1976). Lipid storage and utilization in reptiles. Am. Zool. 16, 711–723.

Doherty, T.S., Glen, A.S., Nimmo, D.G., Ritchie, E.G. & Dickman, C.R. (2016). Invasive predators and global biodiversity loss. PNAS 113, 11261–11265.

Escobar, R.A., Besier, E. & Hayes, W.K. (2010). Evaluating headstarting as a management tool: Postrelease success of green iguanas (Iguana iguana) in Costa Rica. J. Biodiver. Conserv. 2, 204–214.

Ferguson, G.W., Brown, K.L. & DeMarco, V.G. (1982). Selective basis for the evolution of variable egg and

8 Animal Conservation �� (2019) ��–�� ª 2019 The Zoological Society of London

Translocation success in lizards K. M. Hare et al.

hatchling size in some iguanid lizards. Herpetologica 38, 178–188.

Hardy, G.S. (1977). The New Zealand Scincidae (Reptilia: Lacertilia); a taxonomic and zoogeographic study. New Zeal. J. Zool. 4, 221–325.

Hare, K.M. & Miller, K.A. (2009). What dives beneath: diving as a measure of performance in lizards. Herpetologica 65, 227–236.

Hare, K.M., Daugherty, C.H. & Cree, A. (2002). Incubation regime affects juvenile morphology and hatching success, but not sex, of the oviparous lizard Oligosoma suteri (Lacertilia: Scincidae). New Zeal. J. Zool. 29, 221–229.

Hare, K.M., Longson, C.G., Pledger, S. & Daugherty, C.H. (2004). Size, growth, and survival are reduced at cool incubation temperatures in the temperate lizard Oligosoma suteri (lacertilia: Scincidae). Copeia 2, 382–390.

Hare, K.M., Daugherty, C.H. & Chapple, D.G. (2008a). Comparative phylogeography of three skink species (Oligosoma moco, O. smithi, O. suteri; Reptilia: Scincidae) in northeastern New Zealand. Mol. Phylogenet. Evol. 46, 303–315.

Hare, K.M., Pledger, S. & Daugherty, C.H. (2008b). Low incubation temperatures negatively influence locomotor performance and behavior of the nocturnal lizard Oligosoma suteri (Lacertidae: Scincidae). Copeia 2008, 16–22.

Hare, K.M., Norbury, G., Judd, L.M. & Cree, A. (2012). Survival of captive-bred skinks following reintroduction to the wild is not explained by variation in speed or body condition index. New Zeal. J. Zool. 39, 319–328.

Hare, K.M., Chapple, D.G., Towns, D.R. & van Winkel, D. (2016). The ecology of New Zealand's lizards. In New Zealand lizards: 133–168. Chapple, D.G. (Ed). Switzerland: Springer.

Hare, K.M., Borrelle, S.B., Buckley, H.L., Collier, K.J., Constantine, R., Perrott, J.K., Watts, C.H. & Towns, D.R. (2019). Intractable: Species in New Zealand that continue to decline despite conservation efforts. J. Roy. Soc. New Zeal. 49, 301–319. https://doi.org/10.1080/03036758.2019. 1599967.

Hasegawa, M. (1990). The thrush Trudus celaenops, as an avian predator of juvenile Eumeces okadae on Miyake-Jima, Izu islands. Jpn. J. Herpetol. 13, 65–69.

Herrel, A., Spithoven, L., Van Damme, R. & De Vree, F. (1999). Sexual dimorphism of head size in Gallotia galloti: Testing the niche divergence hypothesis by functional analyses. Funct. Ecol. 13, 289–297.

Herrel, A., De Grauw, E. & Lemos-Espinal, J.A. (2001). Head shape and bite performance in xenosaurid lizards. J. Exp. Zool. 290, 101–107.

Hitchmough, R.A., Adams, L.K., Reardon, J.T. & Monk, J.M. (2016a). Current challenges and the future directions in lizard conservation in New Zealand. J. R. Soc. N. Z. 46, 29–39.

Hitchmough, R.A., Patterson, G.B. & Chapple, D.G. (2016). Putting a name to diversity: taxonomy of the New Zealand

lizard fauna. In New Zealand Lizards: 87–108. Chapple, D.G. (Ed.). Switzerland: Springer.

Hoskins, A.J., Hare, K.M., Miller, K.A., Schumann, N. & Chapple, D.G. (2017). Repeatability, locomotor performance and trade-offs between performance traits in two lizard species, Oligosoma alani and O. Smithi. Biol. J. Lin. Soc. 122, 850–859.

Huey, R.B., Schneider, W., Erie, G.L. & Stevenson, R.D. (1981). A field-portable racetrack and timer for measuring acceleration and speed of small cursorial animals. Experientia 37, 1356–1357.

Husak, J.F. (2006). Does survival depend on how fast you can run or how fast you do run? Funct. Ecol. 20, 1080– 1086.

Irschick, D.J. (2003). Measuring performance in nature: Implications for studies of fitness within populations. Integr. Comp. Biol. 43, 396–407.

Irschick, D.J. & Garland, T. Jr (2001). Integrating function and ecology in studies of adaptation: investigations of locomotor capacity as a model system. Annu. Rev. Ecol. Syst. 32, 367–396.

IUCN (2018). The IUCN red list of threatened species. Version 2018–2.

Jarvie, S., Senior, A.M., Adolph, S.C., Seddon, S.C. & Cree, A. (2015). Captive rearing affects growth but not survival in translocated juvenile tuatara. J. Zool. 297, 184–197.

Maran, T., P~odra, M., P~olma, M.O. & Macdonald, D.W. (2009). The survival of captive-born animals in restoration programmes - case study of the endangered European mink Mustela lutreola. Biol. Conserv. 142, 1685–1692.

Miller, K.A., Hare, K.M. & Nelson, N.J. (2010). Do alternate escape tactics provide a means of compensation for impaired performance ability? Biol. J. Lin. Soc. 99, 241– 249.

Miller, K.A., Towns, D.R., Allendorf, F.W., Ritchie, P.A. & Nelson, N.J. (2011). Genetic structure and individual performance following a recent founding event in a small lizard. Conserv. Genet. 12, 461–473.

Miller, K.A., Bell, T.P. & Germano, J.M. (2014). Understanding publication bias in reintroduction biology by assessing translocations of New Zealand's herpetofauna. Conserv. Biol. 28, 1045–1056.

Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Fonseca, G.A.B. & Kent, J. (2000). Biodiversity hotspots for conservation priorities. Nature 403, 853–858.

Noble, D.W.A., Stenhouse, V. & Schwanz, L.E. (2018). Developmental temperatures and phenotypic plasticity in reptiles: a systematic review and meta-analysis. Biol. Rev. 93, 72–97.

Plummer, M. (2003). JAGS: a program for analysis of Bayesian graphical models using Gibbs sampling. In Proceedings of the 3rd International Workshop on Distributed Statistical Computing (DSC 2003)). Vienna, Austria.

Animal Conservation �� (2019) ��–�� ª 2019 The Zoological Society of London 9

K. M. Hare et al. Translocation success in lizards

R Core Team (2017). R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing.

Romijn, R.L. & Hartley, S. (2016). Trends in lizard translocations in New Zealand between 1988 and 2013. New Zeal. J. Zool. 43, 191–210.

Sacchi, R., Pupin, F., Gentilli, A., Rubolini, D., Scali, S., Fasola, M. & Galeotti, P. (2009). Male-male combats in a polymorphic lizard: residence and size, but not color, affect fighting rules and contest outcome. Aggress. Behav. 35, 274–283.

Seddon, P.J., Armstrong, D.P. & Maloney, R.F. (2007). Developing the science of reintroduction biology. Conserv. Biol. 21, 303–312.

Shine, R. (1980). "Costs" of reproduction in reptiles. Oecologia 46, 92–100.

Shine, R. & Harlow, P. (1993). Maternal thermoregulation influences offspring viability in a viviparous lizard. Oecologia 96, 122–127.

Snyder, N.F.R., Derrickson, S.R., Beissinger, S.R., Wiley, J.W., Smith, T.B., Toone, W.D. & Miller, B. (1996). Limitations of captive breeding in endangered species recovery. Conserv. Biol. 10, 338–348.

Stamps, J.A. (1977). The relationship between resource competition, risk, and aggression in a tropical territorial lizard. Ecology 58, 349–358.

Stolzenburg, W. (2011). Rat island: predators in paradise and the world’s greatest wildlife rescue. Sydney: Bloomsbury.

Su, Y.-S. & Yajima, M. (2015). R2jags: using r to run 'JAGS'. R package version 0.5-7.

Tingley, R., Hitchmough, R.A. & Chapple, D.G. (2013). Life history traits and extrinsic threats determine extinction risk in New Zealand Lizards. Biol. Conserv. 165, 62–68.

Tokarz, R.R. (1985). Body size as a factor determining dominance in staged agonistic encounters between male brown anoles (Anolis sagrei). Anim. Behav. 33, 746–753.

Towns, D.R. (1975). Reproduction and growth of the black shore skink, Leiolopisma suteri (lacertilia: Scincidae), in north-eastern New Zealand. New Zeal. J. Zool. 2, 409–423.

Towns, D.R. & Ferreira, S.M. (2001). Conservation of New Zealand lizards (Lacertilia: Scincidae) by translocation of small populations. Biol. Conserv. 98, 211–222.

Towns, D.R., Parrish, G.R. & Westbrooke, I. (2003). Inferring vulnerability to introduced predators without experimental demonstration: case study of Suter's skink in New Zealand. Conserv. Biol. 17, 1361–1371.

Towns, D.R., Atkinson, I.A.E. & Daugherty, C.H. (2006). Have the harmful effects of introduced rats on islands been exaggerated? Biol. Invasions 8, 863–891.

Towns, D.R., Miller, K.A., Nelson, N.J. & Chapple, D.G. (2016a). Can translocations to islands reduce extinction risk for reptiles? Case studies from New Zealand. Biol. Conserv. 204, 120–127.

Towns, D.R., Hitchmough, R.A. & Perrott, J. (2016b). Conservation of New Zealand lizards: a fauna not forgotten but undervalued? In New Zealand lizards: 293–320. Chapple, D.G. (Ed.). Switzerland: Springer.

Verwaijen, D., Van Damme, R. & Herrel, A. (2002). Relationships between head size, bite force, prey handling efficiency and diet in two sympatric lacertid lizards. Funct. Ecol. 16, 842–850.

Wapstra, E. (2000). Maternal basking opportunity affects juvenile phenotype in a viviparous lizard. Funct. Ecol. 14, 345–352.

Webb, J.K. & Shine, R. (1993). Prey-size selection, gape limitation and predator vulnerability in Australian blindsnakes (Typhlophidae). Anim. Behav. 45, 1117–1126.

Whittaker, R.J. & Fern�andez-Palacios, J.M. (2007). Island biogeography: ecology, evolution and conservation. 2nd edn. Oxford: Oxford University Press.

van Winkel, D., Baling, M. & Hitchmough, R. (2018). Reptiles and amphibians of New Zealand: a field guide. Auckland: Auckland University Press.

10 Animal Conservation �� (2019) ��–�� ª 2019 The Zoological Society of London

Translocation success in lizards K. M. Hare et al.