No place like home: an experimental comparison of reintroduction strategies using snakes

Transcription

1 Journal of Applied Ecology 2010, 47, doi: /j x No place like home: an experimental comparison of reintroduction strategies using snakes John H. Roe 1 *, Michael R. Frank 1, Scott E. Gibson 1, Omar Attum 2 and Bruce A. Kingsbury 1 1 Center for Amphibian and Reptile Conservation and Management, Indiana-Purdue University, Fort Wayne, IN, USA; and 2 Department of Biology, Indiana University Southeast, New Albany, IN, USA Summary 1. The practice of deliberately moving animals from one site to another for conservation is increasing as a tool to re-establish extirpated populations. Resource managers are faced with developing strategies for reintroduction attempts, but often lack experimentally derived evidence upon which to base decisions. 2. Using the northern water snake Nerodia sipedon sipedon in the USA, we compared the behaviour and performance of resident snakes with that of individuals translocated directly from the wild to a nearby nature reserve or reared in captivity prior to translocation. 3. Both translocated groups had low survivorship relative to resident snakes, but the proximal causes of their poor performance differed considerably. Captive-reared snakes exhibited restricted surface activity and movements and abnormal habitat use, and ultimately failed to maintain appropriate body temperature and body mass, with high mortality associated with the overwintering period. Wild snakes directly translocated to an unfamiliar site maintained body temperatures and growth comparable with residents, but their more extensive movements resulted in frequent excursions off reserve and high mortality. 4. Synthesis and applications. We contend that an individual s prior experience is an important factor in determining their behaviour and performance during the phase of early establishment at an unfamiliar site. This suggests the existence of common underlying mechanisms influencing the outcome of reintroduction attempts, and provides a potentially useful framework for improving reintroduction efforts. Resource managers would likely improve success of reintroductions by matching habitats (and associated resources and conditions) between source and release sites, by temporarily confining animals in enclosures that force new associations to be made while limiting exploratory wanderings, or by enrichment of environmental conditions in captivity. Key-words: captive breeding, head-starting, natal habitat preference induction, Nerodia sipedon, radio-telemetry, repatriation, reptile, restoration, survival probability, translocation Introduction The practice of moving animals for conservation or wildlife management purposes is increasing (Seddon, Armstrong & Maloney 2007). There are several reasons why resource managers may attempt animal translocation, defined as the deliberate movement of individuals from one part of their distribution to another where the species historically occurred or is currently present (IUCN 1998). Animals may be removed *Correspondence author: Department of Biology, University of North Carolina at Pembroke, PO Box 1510, Pembroke, NC 28372, USA. john.roe@uncp.edu from an area due to imminent threats to their survival (Tuberville et al. 2005; Griffiths & Pavajeau 2008), to mitigate conflicts (Sullivan, Kwiatkowski & Schuett 2004; Bradley et al. 2005), to assist colonization in response to climate change (McLachlan, Hellmann & Schwartz 2007), to introduce genetic variation (Madsen et al. 1999), or for population re-establishment (Pedrono & Sarovy 2000; Moorhouse, Gelling & Macdonald 2009). Several of these examples are typically part of larger conservation efforts to re-establish faunal populations or augment existing ones at sites that have been restored to address threats to the species or population s persistence. Attempts to reintroduce animals often fail to result in the establishment of viable populations for numerous reasons Ó 2010 The Authors. Journal compilation Ó 2010 British Ecological Society

2 1254 J. H. Roe et al. reviewed elsewhere (Dodd & Seigel 1991; Fischer & Lindenmayer 2000). This limited success has led to the development of many novel techniques for reintroduction programmes and calls for the application of more scientifically rigorous approaches to addressing problems (Fischer & Lindenmayer 2000; Seddon et al. 2007; Armstrong & Seddon 2008; Kingsbury & Attum 2009). In its simplest form, animals can be captured from one site and immediately released into another, but several manipulations may improve the likelihood of success. For instance, the identification of appropriate source populations and use of captive-breeding, along with modifications to the demography of release cohorts, the timing of their release, and use of temporary enclosures to acclimate and familiarize animals to site conditions are some examples of methods that have proven useful (Bright & Morris 1994; Sarrazin & Legendre 2000; Tuberville et al. 2005; Griffiths & Pavajeau 2008). An experimental approach to developing the most appropriate recipe for reintroduction has led to some recent successes in translocation programmes, providing encouraging evidence that translocation can be a useful conservation tool (Taylor, Jamieson & Armstrong 2005; Griffiths & Pavajeau 2008; Germano & Bishop 2009; Santos et al. 2009). Here, we experimentally test the feasibility of common reintroduction strategies by assessing the post-release behaviour, body temperature, growth and survival (collectively referred to as performance measures hereafter) of northern water snakes Nerodia sipedon sipedon (Linnaeus) translocated under different scenarios in the USA. One group was raised in captivity at accelerated growth rates throughout an early part of their life cycle prior to release in the wild a strategy known as headstarting (Pritchard 1979). The other consisted of snakes captured and immediately translocated to an unfamiliar site withoutcaptive-rearing astrategywerefertoasdirecttranslocation. Head-starting works on the assumption that by releasing animals into the wild at a larger size, when they are presumably less vulnerable to predation and closer to reproductive maturity, the reintroduced animals will have a higher probability of establishing a population. When combined with captive-breeding programmes, head-starting can also be used to arrange for multiple releases of numerous individuals with less stress on wild donor populations (Brown & Day 2002). The direct translocation of wild-caught animals represents a simpler (and perhaps cheaper) alternative to head-starting, assuming wild stock can still be attained without jeopardizing donor population viability (Dimond & Armstrong 2007). We add an additional layer of comparisons by studying snakes already residing in the release site, which serve as reference benchmarks against which to gauge the performance of the experimental release groups. The goal of this study is not to attempt a population re-establishment (see Seddon 1999). Instead, we take an alternative approach by intensively monitoring the post-release behavioural responses (e.g. movements, activity and habitat selection) and subsequent consequences (temperature regulation, vital rates) for several individuals during the early process of establishment at unfamiliar sites. This design allows us to test two central (yet unresolved) questions in reintroduction biology: (i) what impact does a period of captive-rearing have on the performance of snakes during early establishment at unfamiliar sites in the wild, and (ii) how do translocated individuals (captive-reared or not) perform relative to snakes already residing in the release site? This approach can lead to a better understanding of the mechanistic basis for a particular manipulation s outcome, providing valuable information on howtocautiouslyproceedwithandimprovetranslocation efforts. Materials and methods STUDY SITE AND MAPPING Snakes were studied from June 2008 to June 2009 in northeast Indiana, USA, at an c. 500 ha nature reserve managed by The Nature Conservancy. The core terrestrial habitat of the reserve consists primarily of hardwood forests interspersed with several old field or scrub-shrub patches, many of which have been recently replanted for reforestation. Scattered throughout the reserve are numerous water bodies, several of which have been restored or created in ongoing site restoration efforts. The study area is bordered by a creek, and the adjacent private lands are predominantly used for agriculture and low-density housing residences. Maps describing the distribution of macro-habitats in the study area were digitized from aerial photographs using ArcGIS 9Æ3(ESRI 2008). Habitat types were defined based on Cowardin et al. (1979): palustrine open (PO) wetlands dominated by open water and or herbaceous emergent vegetation; palustrine scrub-shrub forested (PSS PFO) wetlands dominated by trees and or shrubs; and river stream (RIV) an area of flowing water confined to the channel. Upland habitats (i.e. those outside of typical flood zones) were defined as upland forest (UFO) areas dominated by tree canopy cover; old field upland scrub-shrub (OF USS) primarily open habitats dominated by grasses, forbs, or low shrubby vegetation; and agricultural residential (AG RES) areas actively maintained for agricultural or residential purposes. ANIMAL CAPTURE, MAINTENANCE, AND RADIO- TRANSMITTER IMPLANTATION In July 2007, seven pregnant N. s. sipedon were captured from a site c. 5 km north of the study area and allowed to give birth in the laboratory at Indiana-Purdue University in Fort Wayne, Indiana, USA. Sixty neonates (30 male and 30 female) were retained and raised in captivity, whereas adult females and remaining neonates were immediately returned to their sites of capture. Captive snakes were housed individually in small plastic boxes ( cm) with access to a hide box and water bowl, with heat tape providing a thermal gradient on the floor of the box. The room temperature was set at 25 C, relative humidity typically remained between 30% and 60%, and lights were set to a 12L : 12D cycle. Snakes were fed several live fish 2 3 times per week. These husbandry practices were maintained throughout the entire 11-month-captive period in an effort to accelerate growth rates and maximize size and survivorship for release the following summer. Thus, no cooler overwintering period was employed. At the end of the captive period, 12 snakes, representing offspring of seven females, were selected for release. We did not select snakes randomly because we were limited to using only those large enough for transmitter implantation (see Table 1). We refer to this study group as head-starts hereafter.

3 Snake reintroduction biology 1255 Table 1. Details of the three treatment groups upon release Group n Sex ratio (F:M) Snout-to-vent length (cm) Transmitter weight (% of snake body mass) Head-start 12 9:3 47Æ3 ± 2Æ2 (43Æ1 49Æ8) 5Æ9 ± 0Æ6 (5Æ2 7Æ0) Resident 12 9:3 51Æ3 ± 2Æ9 (46Æ7 55Æ6) 5Æ4 ± 1Æ0 (3Æ9 7Æ2) Translocation 10 6:4 57Æ1 ± 9Æ9 (47Æ3 77Æ5) 3Æ7 ± 1Æ9 (1Æ2 7Æ0) Values are mean ± 1 SD and (range). In May 2008, we captured 12 snakes from the study area and 10 snakes from the same source population as the head-starts. Snakes captured from the study area are referred to as residents hereafter, and those captured from the same property as the head-starts, but not raised in captivity, are referred to as the translocation group hereafter. Snakes in all three groups (i.e. head-start, resident and translocation) were surgically implanted with radio-transmitters (model SB-2T, 5Æ1 g; Holohil Systems Ltd, Carp, ON, Canada) while under anaesthesia using a technique described in Roe, Kingsbury and Herbert (2003). We attempted to match the body size, sex composition and numbers of individuals in the three groups as closely as possible (Table 1). After a 7- to 11-day recovery from surgery, resident snakes were released in their wetland of capture, along-side an approximately equivalent number of snakes from the experimental release groups. FIELD DATA COLLECTION We located snakes at least once per week during the active season (May September), every 2 weeks around the times of hibernation ingress (October November) and egress (March April), and once per month for the remainder of the overwintering period (December February). At each radio-location, we determined the coordinate position using GPS held at the snake s location. GPS units typically had an error of less than 7 m. We then plotted location coordinates using ArcGIS 9Æ3. We also determined the macro-habitat type in which the snake was located and attempted to make a visual observation of the individual to confirm its status as alive or dead, and whether it was active above the surface, in the water, or buried under vegetation, debris, or in a burrow. If a snake was confirmed dead after an extended period below-ground or water in the same location, then we presumed it had been dead since the last confirmed visual observation or movement. Several additional variables were derived to describe the behaviour and performance of each snake. We calculated minimum convex polygons (MCP) to estimate the size of area used with the Hawth s Tools extension in ArcGIS, and measured linear range length as the straight-line distance between the two most widely spaced radio-locations. MCPs and range lengths were calculated using all active season locations and one overwintering location. Movement distances were measured as the straight-line distance between two sequential locations. A monthly index of surface activity was calculated as the proportion of radio-locations where the snake was visually confirmed on the surface as basking, foraging, mating, travelling or fleeing. Transmitter pulse rates were recorded and converted to temperature (to the nearest 0Æ1 C) using calibration curves provided by the manufacturer. We periodically captured snakes and measured their snoutto-vent length (SVL; to the nearest 0Æ1 cm) and mass (to the nearest 0Æ1 g) using fabric tape and an electronic balance. The date of hibernation ingress was the final day of confirmed surface activity in autumn, and hibernation egress was the first date of known spring surface activity. DATA ANALYSES We performed statistical analyses with spss 17.0 (SPSS 2007), the program mark 5Æ0 (White & Burnham 1999) and Compos Analysis 6Æ2 (Smith 2004). Where appropriate, we examined the assumptions of homogeneity of variances and normality; when data failed to meet assumptions, they were transformed to approximate normal distributions or equal variances. Statistical significance was accepted at the a 0Æ05 level except for multiple related comparisons, where the Dunn Sidak method was applied to constrain the experiment-wide Type I error to 0Æ05 (Quinn & Keough 2002). In all analyses, sexes were pooled within treatments to increase sample sizes and power. We used a multivariate analysis of covariance (mancova)to test for the effects of treatment group on movement rate (log 10 m day) and size of area used (log 10 MCP), with log 10 SVL and number of radiofixes as covariates. Mean monthly surface activity indices (arcsintransformed) were examined using analysis of covariance (ancova), with treatment group as the predictor variable and time of day as a covariate. The Dunn Sidak adjusted level of significance for this series of surface activity tests was 0Æ010. Variation among groups in dates of hibernation ingress and egress were examined with Mann Whitney tests, using Julian day counted from 1 January as the dependent variable. We used compositional analysis to investigate habitat selection (Aebischer, Robertson & Kenward 1993). Habitat selection was assessed by comparing the relative proportions of habitats used by the individual to the habitats relative availability on the site. Habitats that were not used were replaced with a value of 0Æ7%, which was identified as minimizing misclassification error rates (Bingham, Brennan & Ballard 2007). The study site was defined as the MCP drawn around the combined positions for snakes in all three treatment groups. Body temperature (T b ) and time of location were compiled into monthly mean values for each individual. We then used ancova to test for the effects of treatment on T b for each month separately, using time of day as a covariate. The Dunn Sidak adjusted level of significance for this series of T b tests was 0Æ010. Body condition indices (BCI) at time of release were compared among groups using the method described by Beaupre & Douglas (2009). BCI was estimated as the deviation in observed mass from that predicted by the regression relationship between body mass and SVL of all snakes. BCIs were then compared using anova. ancova was used to test for treatment differences in growth rates, using log 10 g day as the dependent variable and log 10 initial SVL as the covariate. Survival probabilities were estimated using known fate models in the program mark. Time periods when radio-signals could not be detected (i.e. transmitter failure or undetected long-distance movements) were censored from the analysis. Akaike information criterion (AIC) was used to rank candidate models; if competing models had AIC values of <2Æ0, they were considered as having some support. Time intervals were set as bi-weekly during the active season and condensed into a single period during hibernation. We defined the

4 1256 J. H. Roe et al. hibernation period as the time between the final date of autumn ingress and first spring emergence of resident snakes. We started with a fully-saturated probative model in which survival probability was dependent on treatment group and time, with SVL as a covariate. We then fit a series of reduced-parameter models and only retained factors having some support for use in later analyses. Variation among groups with respect to use of areas off the nature reserve was examined with a Fisher s exact test. A snake was recorded as using areas off reserve only if it was confirmed by visual observation to have moved beyond reserve boundaries. Surface activity index Hst Res Tra Results MOVEMENTS AND ACTIVITY Jun Jul Aug Sep Oct Month Movement rates and size of area used differed among treatment groups, with translocated snakes exhibiting the highest vagility and head-starts the lowest (Tables 2 and 3). Headstarts were also less active on the surface than residents and translocated snakes in both June and July (P 0Æ001 in both cases), but not from August to October (P >0Æ064 in all cases; Fig. 1). Information on hibernation ingress and egress could be obtained for only one translocated snake, so we went forward with analyses of head-started and resident snakes only. Dates of hibernation ingress were similar between treatment groups, with entrance dates of 8 November ± 3Æ3 days (mean ± 1 SE) for eight head-starts and 7 November ± 3Æ7 days for seven residents (Z = )0Æ495, P = 0Æ694). Mean dates of hibernation egress differed between treatments groups, with seven head-starts initiating spring activity on 17 March ± 0Æ4 days, and seven residents becoming active on 21 April ± 3Æ2 days (Z = )3Æ25, P =0Æ001). Table 2. Movement and spatial variables for resident snakes compared with head-started and directly translocated experimental release groups Group n Area used (ha) Range length (m) Movement distance (m per day) Head-start 12 2Æ3 ± 1Æ3 198 ± 63 6Æ9 ± 2Æ0 Resident 12 4Æ8 ± 1Æ7 390 ± 82 16Æ7 ± 3Æ4 Translocated 10 13Æ3 ± 3Æ9 707 ± Æ6 ± 3Æ0 Values are mean ± 1 SE. Table 3. Results of mancova for the effects of number of radio-fixes, body size [snout-to-vent length (SVL)] and treatment on movements and spatial ecology of water snakes experimentally introduced to an unfamiliar location Effect Wilks Lambda Num d.f. Den d.f. F P Radio-fixes 0Æ Æ21 <0Æ001 SVL 0Æ Æ122 0Æ886 Treatment 0Æ Æ67 0Æ001 Fig. 1. Monthly surface activity for resident (Res), head-started (Hst), and directly translocated (Tra) water snakes. The surface activity index is the mean (± 1 SE) proportion of radio-fixes where the presence of a snake on the surface was confirmed visually as either basking, foraging, mating or travelling. Number of snakes in each treatment group (head-start, resident, translocated) are as follows: June (4, 6, 4), July (11, 12, 10), August (11, 11, 9), September (11, 9, 6), October (11, 7, 3). HABITAT SELECTION We limited analyses to the four primary habitats used by the snakes, including palustrine open, palustrine scrub-shrub forested, upland forest and old field upland scrub-shrub. More than 99% of snake locations were within these four habitat types that together represented >91% of the 160-ha study area. Treatment groups exhibited different patterns of habitat selection (Table 4, Fig. 2). Resident snakes selected palustrine open and scrub-shrub forested wetlands equally, with little distinction between upland forest and old field upland scrubshrub habitats. Translocated snakes selected palustrine open wetlands over scrub-shrub forested wetlands, and old field upland scrub-shrub habitats over upland forests. Headstarts exhibited neither strong selection nor avoidance of any habitat type. Overall, head-starts spent c. 48% of the active season in terrestrial habitats compared with only 13 15% for resident and translocated snakes. Table 4. Selection rankings of the primary habitat types for resident water snakes compared with head-started and directly translocated experimental release groups Habitat rank (most fi least)* Head-start PO a UFO a PSS PFO a OF USS a Resident PO a PSS PFO a,b UFO b OF USS b Translocated PO a OF USS b PSS PFO b,c UFO c Common superscripts within a row indicate habitats that were ranked similarly. *Habitats include palustrine scrub-shrub forest mix (PSS PFO), palustrine open (PO), upland forest (UFO), and old field upland scrub-shrub mix (OF USS). See text for habitat descriptions.

5 Snake reintroduction biology 1257 BODY TEMPERATURE, BODY CONDITION INDEX AND GROWTH After accounting for variation in time of day, T b of the headstarts was lower than the resident and translocated groups from June to August (P 0Æ045 in all cases), but not from September to October (P 0Æ195 in both cases; Fig. 3). Initial BCIs were lowest in residents ()12Æ4 ± 4Æ6), intermediate in head-starts (4Æ1 ±1Æ8) and highest in translocated snakes (16Æ6 ±4Æ6; anova: F 2,31 =14Æ18, P <0Æ001).Wewereable to recapture ten head-starts, 11 residents, and eight translocated snakes for growth measurements. Growth rates measured as change in mass and length were correlated (r 2 =0Æ56, P < 0Æ001). After accounting for variation in SVL, growth rates of head-starts were lower than for resident and translocated snakes (SVL: F 1,25 =16Æ46, P <0Æ001; treatment: F 2,25 =3Æ56, P =0Æ044; Fig. 4). SURVIVORSHIP AND OFF-SITE EXCURSIONS In our probative model, we found little evidence that survival probability varied over time (DAICc > 2), so we went forward with simpler models examining variation among treatment groups and body sizes, holding time constant. The model with the most support identified residents as having higher survival probabilities than translocated and head-started snakes, though models with constant survival among groups and head-starts with the lowest survival also had some support (Table 5). Bi-weekly survival probabilities were 0Æ961 ± 0Æ017, 0Æ922 ± 0Æ033 and 0Æ912 ± 0Æ030 for resident, translocated and head-started snakes, respectively, corresponding to annual survival rates of 45Æ2%, 19Æ6% and 16Æ0%. Groups differed in their propensity to leave the nature reserve boundaries, with 40% of translocated snakes leaving the reserve compared with 8% for head-starts and 0% in residents (P =0Æ025). Selection metric Po Pss/Pfo Riv Ufo Of/Uss Ag/Res Habitat type Fig. 2. Comparison of habitat selection for resident (Res), headstarted (Hst) and directly translocated (Tra) water snakes. The habitat selection index was derived by subtracting the proportion of available habitat from proportion of habitats used by individuals. Habitats used more frequently than their availability (i.e. selected) are shown as positive, and those used less frequently than their availability (i.e. avoided) are shown as negative. See text for description of habitat types. Hst Res Tra Discussion The ultimate success of an animal reintroduction programme hinges first upon the short-term performance of individuals in an unfamiliar environment. Reintroduced animals must adopt behaviours suitable for the release site such that they can Growth (cm day 1 ) (a) 2 00 (b) Hst Res 1 50 Tb ( C) Tra Growth (g day 1 ) Jun Jul Aug Sep Oct Month 0 25 Hst Res Tra Treatment Fig. 3. Monthly day-time body temperatures (T b )ofresident(res), head-started (Hst) and directly translocated (Tra) water snakes. Values are mean ± 1 SE. Refer to Fig. 1 for sample sizes in each month. Fig. 4. Growth rates measured in (a, cm per day), and (b, g per day) for resident (Res), head-started (Hst) and directly translocated (Tra) water snakes. Values are mean ± 1 SE.

6 1258 J. H. Roe et al. Table 5. Models of survivorship probability (S) for resident snakes (RES) compared to head-start (HST) and translocated (TRA) experimental release groups. All models include initial body size as a covariate Model AICc DAICc Weight n Deviance S (RES > HST = TRA) 179Æ39 0Æ00 0Æ Æ33 S (Æ) 180Æ47 1Æ07 0Æ Æ43 S (RES = TRA > HST) 180Æ68 1Æ29 0Æ Æ62 S (RES > HST > TRA) 181Æ41 2Æ01 0Æ Æ30 S (RES = HST > TRA) 181Æ85 2Æ45 0Æ Æ78 maintain homeostasis, avoid predators, find shelter, forage, grow, survive the initial reintroduction period and establish some degree of site fidelity. Although not indicators of a successful reintroduction themselves, these are all prerequisite to the breeding and eventual persistence of the re-established population (Kleiman 1989; Stamps & Swaisgood 2007). We found marked differences in the movement, activity and habitat selection behaviours of our experimental release groups, as well as variation in important physiological consequences and vital rates including temperature regulation, growth and survivorship. The most important findings relevant to reintroduction programmes were that (i) head-started and translocated snakes performed relatively poorly compared with residents, (ii) the proximal causes underlying the performance of experimentally reintroduced snakes differed considerably, and (iii) the performance of reintroduced snakes appears to relate ultimately to the snake s prior experience, this last point suggesting a common underlying mechanism and framework for improving reintroduction efforts. REINTRODUCED COMPARED WITH RESIDENT SNAKES We addressed the concerns of whether the use of captive-bred or wild-caught snakes translocated to an unfamiliar site compromises an individual s ability to behave appropriately and survive their first year in the wild. Head-starts moved considerably less and traversed smaller areas than either of the other groups. Perhaps most strikingly, head-starts demonstrated little motivation (or ability) to select or avoid habitats in a manner consistent with other water snakes. As a result, head-starts spent nearly half of their active season in terrestrial habitats, which is uncharacteristic of N. sipedon (Roe, Kingsbury and Herbert 2003). Moreover, head-starts were rarely observed basking, foraging or travelling, all of which were surface activities commonly observed in resident snakes, and they exited hibernation refuges c. 1 month prior to resident snakes. As a consequence, head-starts were less likely to make exploratory movements, had active season T b considerably cooler than that preferred by N. sipedon (Brown & Weatherhead 2000), failed to gain weight, and had high rates of mortality. Wild-caught snakes directly translocated to unfamiliar environments also behaved abnormally with respect to residents. Translocated snakes moved more extensively and frequently used areas outside of reserve boundaries. Translocated snakes selected open canopied wetlands over scrubby or forested wetlands, whereas resident wild snakes selected both wetland types equally. However, translocated snakes largely avoided terrestrial habitats, remained active on the surface throughout the active season, maintained active season T b within the preferred range of N. sipedon (Brown & Weatherhead 2000), and grew appreciably, though mortality rates were higher than resident snakes. PROXIMAL CAUSES OF PERFORMANCE VARIATION Although survival was relatively low in both reintroduction groups, the proximal causes underlying performance variation differed considerably. Although our models did not detect significant variation in survival over time, there is evidence that head-starts did experience relatively high (and perhaps biologically significant) mortality associated with the overwintering period. From November to March, only 50% of head-starts survived compared with 100% in both other groups (Fig. 5). One head-start never emerged from its burrow, and the other three were found dead in the vicinity of their hibernation refuges shortly after they emerged, but before any of the resident snakes initiated spring activity. It is possible that the headstarts poor nutritional state led them to premature emergence and exposure to cold temperatures and predators in early Proportion known to be alive Jun-08 Jul-08 Aug-08 Sep-08 Oct-08 Nov-08 Dec-08 Month Hibernation Jan-09 Feb-09 Mar-09 Apr-09 May-09 Hst Res Tra Fig. 5. Proportion of resident (Res), head-started (Hst) and directly translocated (Tra) water snakes known to be alive over time. Snake status (alive or dead) was assessed at bi-weekly intervals except during the hibernation period (delineated by the vertical dashed lines), when status could not be consistently confirmed for all snakes below the surface. Note that survivorship values on this graph are not derived from our estimates generated by known-fate models in the program MARK. Jun-09

7 Snake reintroduction biology 1259 spring, as has been observed in other natricine snakes (Aleksuik & Stewart 1971; Brown & Weatherhead 1997; Shine et al. 2001). Mortality associated with overwintering could also result from a poor choice of microsites or an inability to time emergence with appropriate seasonal environmental cues. In contrast, the poor performance of translocated snakes most probably resulted from their extensive movements. Wide-ranging movement is a common response to translocation, presumably as individuals search for familiar sites or attempt to home (Reinert & Rupert 1999; Bright & Morris 2004; Sullivan et al. 2004; Butler, Malone & Clemann 2005; Tuberville et al. 2005). Such movements resulted in high movement rates and several snakes using areas outside of the nature reserve, both of which are likely to increase a translocated animal s exposure to predators and other threats (Meek et al. 2003; Moehrenschlager & Macdonald 2003). ROLE OF PRIOR EXPERIENCE A potential common mechanism underlying the post-release behaviours of head-started and translocated snakes is natal habitat preference induction (NHPI), a form of behavioural plasticity whereby an individual s earlier experience shapes habitat preferences later in life (Davis & Stamps 2004). NHPI theory predicts that when dispersing animals encounter unfamiliar environments, they will select habitats with cues resembling those they are familiar with, assuming their prior experience was positive (Davis & Stamps 2004; Stamps & Swaisgood 2007). The early habitat experience of head-starts was limited to simple, primarily terrestrial enclosures that offered few (if any) similarities to natural environments. As predicted by NHPI theory, head-starts failed to exhibit habitat selection or avoidance perhaps because there were no familiar habitat cues at the release site. It is also possible that captive conditions may have encouraged maladaptive behaviours in the wild. For example, during captivity snakes heated via conduction from the floor, whereas wild water snakes regulate T b via solar radiation during aerial basking or conductance in the water (Brown & Weatherhead 2000). Not surprisingly, retreat into subsurface burrows during the active season failed to translate into high T b for head-starts in the wild. In either case, the role of prior experience provides a likely explanation for some of the head-starts abnormal behaviours and poor performance in the wild, a contention supported by numerous studies demonstrating how conditions experienced in captivity influence habitat choice, locomotor performance, problem-solving ability, exploration, and foraging in snakes (Almli & Burghardt 2006; Aubret & Shine 2008) and other animals (Davis & Stamps 2004; Stamps & Swaisgood 2007). Although we do not know for certain the early habitat experiences of translocated snakes, all were captured from a complex of open canopy wetlands surrounded by old field and scrubby uplands. Consistent with NHPI, translocated snakes demonstrated strong preference for open wetlands, a pattern in contrast to the mixed use of wetland types (forested scrub, and open) by resident snakes. Translocated animals have been shown to use resources in a manner consistent with their original site (Rittenhouse et al. 2008), sometimes with negative consequences to their performance in the new environment (Warren et al. 1996; Kenward & Hoddler 1998). Associations with familiar hibernation sites may also account for the trend of translocated individuals maintaining surface activity later in theseason(fig.1).snakesincoldclimatesoftenmaintain strong fidelity to specific overwintering areas (Prior, Blouin- Demers & Weatherhead 2001; Harvey & Weatherhead 2006), and may travel long-distances in an attempt to return if displaced (Brown & Parker 1976). The long-distance movements, large areas traversed and off-site explorations were likely to be a result of translocated snakes searching for cues matching resource selection templates developed previously. It is also possible that introduced snakes were less healthy as a result of the translocation process or the conditions they experienced prior to release. For instance, parasites can impact growth rates and body condition of snakes (Madsen, Ujvari & Olsson 2005), and snakes translocated from another site or raised in captivity may have been differentially exposed to parasites and other pathogens. At the time of release, however, resident snakes actually had the lowest BCIs, and we observed no overt manifestations of disease or parasitism in any individuals. We later discovered the nematode Eustronglydies ignotus in some snakes in our captive colony, but this parasite was probably introduced via wild-caught prey from the release site which was fed to captives. This nematode is known to occur in the region of our study (Friend et al. 1999). MANAGEMENT IMPLICATIONS We view the causes responsible for the poor performance of head-started water snakes as presenting more problems for reintroduction programmes than for those translocated directly from the wild. Although translocated snakes had low survivorship, they demonstrated an ability to more effectively use resources at the release site. The number of translocated animals could be adjusted to meet desired survivorship and site-retention goals, but the release of numerous individuals can risk compromising the viability of source populations. In such cases, captive colonies could feasibly generate stock for reintroductions, but our captive-reared snakes lacked the behavioural competence for survival in the wild, and would thus be ill-suited as population founders in a reintroduction programme regardless of the number released. That an animal s prior experience appears to factor importantly in determining their behaviour and ultimate survival during the phase of early establishment at an unfamiliar site provides a potentially useful framework for improving reintroduction efforts (Davis & Stamps 2004; Stamps & Swaisgood 2007). Reintroductions are likely to be most successful when habitats (and associated resources and conditions) are matched between source and release sites (Rittenhouse et al. 2008). When such matches are not possible, the performance and site fidelity of translocated animals may be improved by temporarily confining them in enclosures that allow new associations to be made while avoiding the dangers of

8 1260 J. H. Roe et al. exploratory wanderings (Bright & Morris 1994; Tuberville et al. 2005). Such efforts could include identifying or creating suitable overwintering refuges for snakes (Kissner & Weatherhead 2005), as the consequences of not finding a suitable overwintering site can be particularly catastrophic (Macmillan 1995). Head-starting programmes using captive animals may also be improved by providing housing in environmentally enriched conditions that match release sites as closely as possible. Reintroduction programmes drawing from captive stock should be aware of the difference between outputs of animals and conservation outputs, as facilities and practices designed strictly for ease of animal husbandry will be different to those designed to produce animals best prepared for life in the wild the latter being the preferred strategy for reintroduction, but also the most costly (Miller et al. 1999; Brown & Day 2002; Alberts 2007). A clear limitation in this study was the small sample sizes, especially with regards to the details of how different demographic groups may respond to reintroduction. Given the typically strong influence of sex, size, age and reproductive status on behaviour and vital rates in snakes (Brown & Weatherhead 1997, 2000, Shine et al. 2001), it would be valuable to rigorously examine how these factors influence their performance at novel sites. Instead of closing the door on animal reintroductions, it is our hope that this research will open additional avenues for experimental approaches to reintroductions designed to be of utility to conservation professionals. Acknowledgements We would like to thank P. Bieberich, C. Smith, J. Gibson, A. Detwiller and A. Sadote for their assistance with field work and captive husbandry. We are also grateful to E. Mizell and J. Draper at The Nature Conservancy, and W. Cosper for access to field sites. Research was conducted under protocol issued by the Purdue Animal Care and Use Committee, and licenses and from the Indiana Department of Natural Resources. Funding was provided to BAK from the United States Fish and Wildlife Service, and by Indiana-Purdue University at Fort Wayne, The Nature Conservancy, and The Center for Reptile and Amphibian Conservation and Management. References Aebischer, N.J., Robertson, P.A. & Kenward, R.E. (1993) Compositional analysis of habitat use from radio-tracking data. Ecology, 74, Alberts, A.C. (2007) Behavioural considerations for headstarting as a conservation strategy for endangered Caribbean rock iguanas. Applied Animal Behaviour Science, 102, Aleksuik, M. & Stewart, K.W. (1971) Seasonal changes in the body composition of the garter snake (Thamnophis sirtalis parietalis) at northern latitudes. Ecology, 52, Almli, L.M. & Burghardt, G.M. (2006) Environmental enrichment alters the behavioural profile of ratsnakes (Elaphe). Journal of Applied Animal Welfare Science, 9, Armstrong, D.P. & Seddon, P.J. (2008) Directions in reintroduction biology. Trends in Ecology and Evolution, 23, Aubret, F. & Shine, R. (2008) Early experience influences both habitat choice and locomotor performance in tiger snakes. American Naturalist, 171, Bingham, R.L., Brennan, L.A. & Ballard, B.M. (2007) Misclassified resource selection: compositional analysis and unused habitat. Journal of Wildlife Management, 71, Beaupre, S.J. & Douglas, L.E. (2009) Snakes as indicators and monitors of ecosystem properties. Snakes: Ecology and Conservation (eds S.J. Mullin & R.A. Seigel), pp Cornell University Press, Ithaca, NY. Bradley, E.H., Pletscher, D.H., Bangs, E.E., Kunkel, K.E., Smith, D.W., Mack, C.M., Meier, T.J., Fontaine, J.A., Niemeyer, C.C. & Jimenez, M.D. (2005) Evaluating translocation as a nonlethal method to reduce livestock conflicts in the northwestern United States. Conservation Biology, 19, Bright, P.W. & Morris, P.A. (1994) Animal translocation for conservation: performance of dormice in relation to release methods, origin, and season. Journal of Applied Ecology, 31, Brown, C. & Day, R.L. (2002) The future of stock enhancements: lessons for hatchery practice from conservation biology. Fish and Fisheries, 3, Brown, W.S. & Parker, W.S. (1976) Movement ecology of Coluber constrictor near communal hibernacula. Copeia, 1976, Brown, G.P. & Weatherhead, P.J. (1997) Effects of reproduction on survival and growth of female northern water snakes, Nerodia sipedon. Canadian Journal of Zoology, 75, Brown, G.P. & Weatherhead, P.J. (2000) Thermal ecology and sexual size dimorphism in northern water snakes, Nerodia sipedon. Ecological Monographs, 70, Butler, H., Malone, B. & Clemann, N. (2005) The effects of translocation on the spatial ecology of tiger snakes (Notechis scutatus) in a suburban landscape. Wildlife Research, 32, Cowardin, L.M., Carter, V., Golet, F.C. & LaRoe, E.T. (1979) Classification of wetlands and deepwater habitats of the United States. US Fish and Wildlife Service, Washington, DC. Davis, J.M. & Stamps, J.A. (2004) The effect of natal experience on habitat preferences. Trends in Ecology and Evolution, 19, Dimond, W.J. & Armstrong, D.P. (2007) Adaptive harvesting of source populations for translocation: a case study using New Zealand robins. Conservation Biology, 21, Dodd, C.K. & Seigel, R.A. (1991) Relocation, repatriation, and translocation of amphibians and reptiles: are they conservation strategies that work? Herpetologica, 47, Environmental Systems Research Institute (ESRI) (2008) ArcGIS 9.3. ESRI, Inc., Redlands, CA, USA. Fischer, J. & Lindenmayer, D.B. (2000) An assessment of the published results of animal relocations. Biological Conservation, 96, Friend, M., Franson, J.C., Ciganovich, E.A., Redman, P.J. & Stenback, R.S. (1999) Field manual of wildlife diseases. General field procedures and diseases of birds. United States Geological Survey Biological Resources Division. Information and Technology Report Germano, J.M. & Bishop, P.J. (2009) Suitability of amphibians and reptiles for translocation. Conservation Biology, 23, Griffiths, R.A. & Pavajeau, L. (2008) Captive breeding, reintroduction, and the conservation of amphibians. Conservation Biology, 22, Harvey, D.S. & Weatherhead, P.J. (2006) Hibernation site selection by eastern massasauga rattlesnakes (Sistrurus catenatus catenatus) near their northern range limit. Journal of Herpetology, 40, IUCN (1998) Guidelines for Re-Introductions. Prepared by the IUCN SSC Re-Introduction Specialist Group. IUCN, Gland, Switzerland and Cambridge, UK. Kenward, R.E. & Hoddler, K.H. (1998) Red squirrels (Sciurus vulgaris) released in conifer woodland: the effects of source habitat, predation and interactions with grey squirrels (Sciurus carolinensis). Journal of Zoology, 244, Kingsbury, B.K. & Attum, O. (2009) Conservation strategies: captive rearing, translocation and repatriation. Snakes: Ecology and Conservation (eds S.J. Mullin & R.A. Seigel), pp Cornell University Press, Ithaca, NY. Kissner, K.J. & Weatherhead, P.J. (2005) Phenotypic effects on survival of neonatal northern watersnakes Nerodia sipedon. Journal of Animal Ecology, 74, Kleiman, D.G. (1989) Reintroduction of captive mammals for conservation. BioScience, 39, Macmillan, S. (1995) Restoration of an extirpated red-sided garter snake Thamnophis sirtalis parietalis population in the interlake region of Manitoba, Canada. Biological Conservation, 72, Madsen, T., Shine, R., Olsson, M. & Wittzell, H. (1999) Restoration of an inbred adder population. Nature, 402, Madsen, T., Ujvari, B. & Olsson, M. (2005) Old pythons stay fit; effects of hematozoan infections on life history traits of a large tropical predator. Oecologia, 142, Meek, W.R., Burman, P.J., Nowakowski, M., Sparks, T.H. & Burman, N.J. (2003) Barn owl release in lowland southern England a twenty-one year study. Biological Conservation, 109,

9 Snake reintroduction biology 1261 McLachlan, J.S., Hellmann, J.J. & Schwartz, M.W. (2007) A framework for debate of assisted migration in an era of climate change. Conservation Biology, 21, Miller, B., Ralls, K., Reading, R.P., Scott, M.J. & Estes, J. (1999) Biological and technical considerations of carnivore translocation: a review. Animal Conservation, 2, Moehrenschlager, A. & Macdonald, D.W. (2003) Movement and survival parameters of translocated and resident swift foxes Vulpes velox. Animal Conservation, 6, Moorhouse, T.P., Gelling, M. & Macdonald, D.W. (2009) Effects of habitat quality upon reintroduction success in water voles: evidence from a replicated experiment. Biological Conservation, 142, Pedrono, M. & Sarovy, A. (2000) Trial release of the world s rarest tortoise Geochelone yniphora in Madagascar. Biological Conservation, 95, Pritchard, P.C.H. (1979) Headstarting and other conservation techniques for marine turtles. International Zoo Yearbook, 19, Prior, K.A., Blouin-Demers, G. & Weatherhead, P.J. (2001) Sampling biases in demographic analyses in black rat snakes (Elaphe obsoleta). Herpetologica, 57, Quinn, G.P. & Keough, M.J. (2002) Experimental Design and Data Analysis for Biologists. Cambridge University Press, Cambridge, UK. Reinert, H.K. & Rupert, R.R. (1999) Impacts of translocation on behaviour and survival of timber rattlesnakes, Crotalus horridus. Journal of Herpetology, 33, Rittenhouse, C.W., Millspaugh, J.J., Hubbard, M.W., Sheriff, S.L. & Dijak, W.D. (2008) Resource selection by translocated three-toed box turtles in Missouri. Journal of Wildlife Management, 72, Roe, J.H., Kingsbury, B.A. & Herbert, N.R. (2003) Wetland and upland use patterns in semi-aquatic snakes: implications for wetland conservation. Wetlands, 23, Santos, T., Pérez-Tris, J., Carbonell, R., Tellería, J.L. & Díaz, J.A. (2009) Monitoring the performance of wild-born and introduced lizards in a fragmented landscape: implications for ex situ conservation programmes. Biological Conservation, 142, Sarrazin, F. & Legendre, S. (2000) Demographic approach to releasing adults versus young in reintroductions. Conservation Biology, 14, Seddon, P.J. (1999) Persistence without intervention: assessing success in wildlife reintroductions. Trends in Ecology and Evolution, 14,503. Seddon, P.J., Armstrong, D.P. & Maloney, R.F. (2007) Developing the science of reintroduction biology. Conservation Biology, 21, Shine, R., LeMaster, M.P., Moore, I.T., Olsson, M.M. & Mason, R.T. (2001) Bumpus in the snake den: effects of sex, size and body condition on mortality in red-sided garter snakes. Evolution, 55, Smith, P.G. (2004) Automated log-ratio analysis of compositional data: software suited to analysis of habitat preference from radio tracking data. Bat Research News, 45, 16. Statistical Package for the Social Sciences (2007) SPSS 17.0 for Windows. SPSS, Chicago, IL, USA. Stamps, J.A. & Swaisgood, R.R. (2007) Someplace like home: experience, habitat selection and conservation biology. Applied Animal Behaviour Science, 102, Sullivan, B.K., Kwiatkowski, M.A. & Schuett, G.W. (2004) Translocation of urban Gila monsters: a problematic conservation tool. Biological Conservation, 117, Taylor, S.S., Jamieson, I.G. & Armstrong, D.P. (2005) Successful island reintroductions of New Zealand robins and saddlebacks with small numbers of founders. Animal Conservation, 8, Tuberville, T.D., Clark, E.E., Buhlmann, K.A. & Gibbons, J.W. (2005) Translocation as a conservation tool: site fidelity and movement of repatriated gopher tortoises (Gopherus polyphemus). Animal Conservation, 8, Warren, C.D., Peek, J.M., Servheen, G.L. & Zager, P. (1996) Habitat use and movements of two ecotypes of translocated caribou in Idaho and British Columbia. Conservation Biology, 10, White, G.C. & Burnham, K.P. (1999) Program MARK: survival estimation from populations of marked animals. Bird Study, 46, Received 19 April 2010; accepted 25 August 2010 Handling Editor: Andrew Royle