Evaluation of headstarting and release techniques for population augmentation and reintroduction of the smooth green snake

Transcription

1 bs_bs_banner Evaluation of headstarting and release techniques for population augmentation and reintroduction of the smooth green snake A. B. Sacerdote-Velat 1, J. M. Earnhardt 1, D. Mulkerin 2, D. Boehm 2 & G. Glowacki 3 1 Department of Conservation and Science, Lincoln Park Zoo, Chicago, IL, USA 2 Department of Animal Care, Lincoln Park Zoo, Chicago, IL, USA 3 Department of Natural Resources, Lake County Forest Preserve District, Libertyville, IL, USA Animal Conservation. Print ISSN Keywords headstarting; reintroduction; reptile; snake; brumation; growth; soft release; release technique. Correspondence Allison B. Sacerdote-Velat, Department of Conservation and Science, Lincoln Park Zoo, 2001 N Clark Street, Chicago, IL 60614, USA. asacerdote@lpzoo.org Editor: John Ewen Received 29 October 2013; accepted 7 May 2014 doi: /acv Abstract Headstarting is a conservation technique for improving survival of species with high juvenile mortality by accelerating growth rate and increasing body size of captive-born young. With reptiles, headstarts are often kept active year-round to achieve body size goals and increase survival, omitting overwintering (brumation). As brumation is part of the life cycle of reptiles, there may be tradeoffs related to temperature response post-release when reptiles are kept active. Upon release into habitats, reptiles are either soft released, where acclimation is provided with in situ enclosures, or hard released without acclimation, directly into habitat. Soft releases have resulted in greater survival and site fidelity than hard releases, but evaluations with snakes are rare. We used a comparative approach to examine effects of brumation versus year-round activity on prerelease growth and survival of smooth green snake Opheodrys vernalis headstarts. We estimated short-term post-release daily survival rates of headstarts and compared movements of hard and soft released snakes. Despite decreases in body mass during brumation, prerelease body size, growth rate and survival did not differ among brumation treatments. Brumated headstarts exhibited rapid compensatory growth, attaining the size of active headstarts within 2 months of brumation. We observed qualitative evidence of reproductive potential in brumated snakes with the production of spermatozoa and unfertilized eggs, which was absent in active headstarts. The short-term survival rate of all headstarts during post-release tracking was 0.83 (±0.01), but we lacked power to examine differences in survival among release treatments. Daily movements did not differ among release treatments. Soft releases had slightly greater recaptures, facilitating monitoring. Although brumation comparisons produced equivalent prerelease growth and survival, as a precautionary measure for post-release survival, we recommend incorporating brumation into headstarting efforts. While further study with other reptiles is warranted, we recommend a comparative framework in planning headstarting efforts with additional species. Introduction Worldwide, snake populations have declined from habitat loss, exacerbated by climate change, invasive species and disease (Gibbons et al., 2000; Reading et al., 2010). In the Midwestern US, snakes suffer losses from conversion of habitats into intensive agriculture including the Eastern Massasauga rattlesnake Sistrurus catenatus (Szymanski, 1998; Sheperd et al., 2008), copperbelly water snake Nerodia erythrogaster neglecta (Roe, Kingsbury & Herbert, 2003) and plains garter snake (King & Stanford, 2006). Accompanying declines, there is interest in restoring habitats and populations with applied conservation techniques such as headstarting (King, Berg & Hay, 2004; King & Stanford, 2006; Germano & Bishop, 2009). Conservationists may use headstarting to augment or reintroduce populations to restored historical locales. Headstarting aims to improve juvenile survival by increasing growth rate and body size of the captive-housed young while they are most vulnerable, to attain greater survival and reproductive potential (Haskell et al., 1996). Common metrics of headstarting success are increased growth rate or size Animal Conservation 17 (Suppl. 1) (2014) The Zoological Society of London 65

2 Headstarting techniques in snake reintroduction A. B. Sacerdote-Velat et al. compared with wild young of similar age. This approach is supported by the silver spoon concept that abundant food resources early in life contribute to increased growth rate for juveniles, and greater body size later in life, despite future resource variation (Madsen & Shine, 2000). Recovery resources are limited and single approaches are frequently adopted, precluding comparisons of techniques. Headstarting efforts often keep reptiles active throughout winter to increase survival and accelerate growth (Haskell et al., 1996; King & Stanford, 2006; Roe et al., 2010). Because overwintering (or brumation) presents a mortality risk for juveniles, headstarting programs may skip brumation to reduce mortality, prolonging the feeding and growth window (Herlands et al., 2004; Mitrus, 2005; King & Stanford, 2006). However, reptiles must survive seasonality in resources and conditions (Bjorndal et al., 2003). Naivety to brumation may result in low overwintering survival or limited reproduction of headstarts post-release (Bona-Gallo & Licht, 1983; Whittier et al., 1987; Roe et al., 2010). While headstarting aims to increase body size, should reptiles be kept active, regardless of their natural history, when environmental variation may impact future survival and reproduction? Reintroductions or augmentations may use hard releases, where headstarts are placed directly into the environment (Bertolero, Oro & Besnard, 2007; Kingsbury & Attum, 2009), or soft releases, where headstarts acclimate in an enclosure within the site prior to full release. Soft releases provide acclimation to habitat while limiting predation and facilitating monitoring (Tuberville et al., 2005). Wandering movements may be common in hard releases (Plummer & Mills, 2000; Rittenhouse et al., 2007; Roe et al., 2010). Soft release programs have found smaller movements and reduced wandering from released animals than in hard releases, but few studies directly compare release approaches or focus on snakes (Tuberville et al., 2005, 2008; Roe et al., 2010). In 2010, Lincoln Park Zoo and Lake County Forest Preserve District in Illinois partnered to focus on regional recovery of smooth green snakes Opheodrys vernalis. Opheodrys vernalis are small, insectivorous, grasslanddependent colubrids with a large range in North America (Redder, Smith & Keinath, 2006). While few population studies exist, O. vernalis are thought to be declining rangewide, driven by habitat loss. The species is endangered, threatened or designated in Greatest Conservation Need in several states (Redder et al., 2006). To guide recovery efforts, we compared effects of brumating headstarts versus maintaining activity throughout winter on prerelease growth, survival and signs of reproductive development. We hypothesized that growth rates and survival of active snakes would exceed that of brumated snakes prerelease. We expected that brumated headstarts, having experienced environmental variation, would have greater post-release survival than active snakes, and might exhibit signs of reproductive development cued by the temperature changes. We used soft releases and hard releases to augment an extant population of O. vernalis where founders for our zoo breeding colony originated. We compared movements of soft and hard releases following full release to determine if hard releases wandered more than soft releases, which could increase risk of mortality, and decrease success of augmentation (Hester, Price & Dorcas, 2008). We expected hard releases daily movements to exceed soft release movements, and soft releases to remain closer to the enclosures because of familiarity with scents of conspecifics. We compared headstart movements with wild residents, expecting that residents would move greater distances than soft releases, but smaller distances than hard releases. Materials and methods In 2010, 41 neonates from a communal nest and 10 neonates from two founder dams were headstarted at the zoo (Sacerdote, Glowacki & Schmidtz, 2012). Snakes were individually marked with ventral scale cautery to monitor growth and survival. Juvenile diet alternated between three 4.76 mm crickets or three mm wax worms per snake, every other day. All prey were gut loaded and dusted with calcium 6 days per week and with a vitamin D/calcium supplement once per week. Twenty-five neonates were randomly assigned across litters to an induced brumation treatment (n = 12 females, n = 13 males). During brumation, snakes were group housed in six 37.8-L tanks of four to five individuals of mixed sex. Brumation occurred from 16 December 2010 to 4 March 2011, lasting 78 days with a mean (sd) temperature of 13.6 C (±2.9). While O. vernalis experience colder and longer brumation in nature, a mild temperature (below 15.5 C) was sufficient to maintain inactivity, and 2- to 3-month brumation is standard for inducing reproduction in captive snakes (Rossi, 1992). Sixteen neonates (n = 8 females, n = 8 males) were randomly assigned to an active treatment in which temperature (23 25 C), humidity (75 78%) and food were kept constant year-round. Active snakes were group housed in three tanks of four to five individuals of mixed sex. Upon warm-up, brumated snakes were provisioned with the same diet as active snakes and all snakes were maintained in mixed sex groups to permit breeding. In May 2011, all female snakes (12 brumated, eight active) were palpated for eggs, and radiographs confirmed if females were gravid. Gravid females were given nest boxes, and any resulting eggs were incubated at C with 75% relative humidity. Cloacal smears were collected on glass slides from all male snakes (13 brumated and eight active) to examine for spermatozoa. Body mass, snout-vent length (SVL) and survival were recorded monthly. We used multivariate analysis of variance (MANOVA) (Statistica V.6, StatSoft, Inc., Tulsa, OK, USA) to examine effects of brumation, sex and treatmentby-sex interaction on prerelease body mass, SVL, overall growth rates from hatching until release and growth rates during the 2 months post-brumation for all headstarts. MANOVA permitted examination of effects of independent brumation and sex variables on related dependent variables 66 Animal Conservation 17 (Suppl. 1) (2014) The Zoological Society of London

3 A. B. Sacerdote-Velat et al. Headstarting techniques in snake reintroduction Table 1 Group composition of headstart cohorts released for augmentation of a founder population in 2011 including brumation treatment, release type, sex and month of release Treatment June July August Total for season Brumated hard release female Brumated soft release female Active hard release female Active soft release female Brumated hard release male Brumated soft release male Active hard release male Active soft release male (body mass, SVL, growth rates) while controlling for correlation among dependent variables. Because of mechanical setbacks with the cooling system, the intended December 2011 brumation was delayed until 19 April June 2012, lasting 52 days with a mean (sd) temperature of 12.8 C (±2.4). Brumation comparisons and analyses were repeated for 2012 headstarts except SVL was not recorded to reduce disturbance to snakes. The cohort of headstarts included 15 active neonates (n = 8 females, n = 7 males) and 15 brumated neonates (n = 7 females, n = 8 males) randomly assigned to treatments across litters. We repeated the MANOVA for growth rate within 2 months of brumation with headstarts pooled across years to determine if the treatment effect would be upheld. In 2011, we augmented an extant founder population in a Lake County Forest Preserve. The site is a 543-acre preserve dominated by tallgrass prairie with pockets of sedge meadow, bur oak-shagbark hickory woodland and oak savanna. All releases occurred within an 8-acre patch of tallgrass prairie adjacent to a sedge meadow and 50-acre savanna. The preserve is bound by a railroad to the south and roads to the north, east and west. The release location was the area of greatest capture frequency of wild O. vernalis during previous monitoring in the preserve. Enclosures measuring 1.8 m 3 were constructed from m corrugated fiberglass panels, trenched into the soil at a depth of 0.3 m. Headstarts had access to soil, vegetation, water, cover and native insect prey which colonized enclosures following construction, prior to release. Enclosure tops of 1 4 wire mesh excluded predators that occur in the site including hawks, weasels, raccoons, coyotes, opossums, rodents and shrews. For augmentation, we released 18 headstarts (n = 9 males, n = 9 females) as either hard releases (n = 9) or soft releases (n = 9). Release groups were randomly selected from a subset of headstarts 9 g from both brumated and active treatments. Snakes were randomly assigned to release types in late June, mid-july, or mid-august, with six snakes released each month. Soft releases were placed in enclosures for 3 weeks of acclimation and hard releases were placed adjacent to ten m coverboards 1 m from the enclosures. For each release, groups consisted of one to two active and one to two brumated males and females in soft and hard releases (Table 1). We recognize that sample size is small but we were constrained by the number of young produced that were of sufficient size to radiotrack, and the need to retain the remaining headstarts for breeding, as another facet of the recovery. Radiotracking, enclosure and coverboard checks were conducted 5 days per week during the first week following each release and then at least three times per week through October. Visual encounter surveys were conducted following coverboard checks. The 0.31 g radio transmitters (Holohil Systems Ltd, Carp, ON, Canada, Model BD-2X) were externally attached upon full release of headstarts into the environment. We waited until full release to affix transmitters to soft releases because the battery life of the transmitters was limited to c. 18 days and we were interested in the movements of soft releases once they left the enclosures. Soft release survival during acclimation was assessed by recapturing snakes within each enclosure. The start date of tracking varied slightly within each release group of June, July or August because external transmitters were attached following ecdysis which was not synchronous across snakes. We used a taping technique to attach transmitters to 15 of 18 headstarts (Madrid-Sotelo & García-Aguayo, 2008; Wylie et al., 2011), with 10 g as a body mass minimum. Transmitters were attached by wrapping paper medical tape around the snake and transmitter to provide a cushioned, flexible attachment system. We wrapped green gaffer s tape around the outside of the paper tape to increase water resistance and reduce visibility of the paper tape. In trial attachments at the zoo, snakes either retained transmitters until ecdysis ( 3 weeks), or in two instances were able to remove transmitters after several days if the tape became saturated with water. While attachment was imperfect, taping was more successful than cyanoacrylate. Body size of O. vernalis precludes implanting transmitters. Battery life of transmitters coincided well with ecdysis frequency for headstarts at the zoo. In three cases, new transmitters were attached, prolonging tracking to days. Five wild residents (three males, two females) were opportunistically tracked to compare their movements with headstarts. Locations of all telemetered snakes were recorded with a Garmin etrex GPS unit (Garmin International, Inc., Olathe, KS, USA). We measured daily movements in ArcGIS V. 10 (ESRI, Redlands, CA, USA) and compared movements among soft releases, hard releases and residents with a one-way univariate analysis of variance (Statistica V. 6, StatSoft, Inc.), using Tukey s test for post hoc comparisons. We used t-tests to compare recapture frequency for five hard releases and six soft releases encountered after their transmitters had dropped off, and for three snakes (two hard releases and one soft release) that were too small for transmitters. Given the body size and transmitter constraints, we could not collect long-term telemetry data for overwintering survival or home range analysis. Instead, we examined shortterm survival following full release to examine fates of headstarts during their first weeks in a novel environment. Small sample size limited our power to examine group effects of release, sex and brumation on survival. However, Animal Conservation 17 (Suppl. 1) (2014) The Zoological Society of London 67

4 Headstarting techniques in snake reintroduction A. B. Sacerdote-Velat et al. Mean (SD) body mass (g) short-term known-fate information for captive-born headstarts following release was still of interest. We used nest survival analysis in Program MARK (White & Burnham, 1999) to estimate daily survival rates during tracking of headstarts, pooled across treatments. Nest survival analysis is well suited to short-term ragged telemetry datasets as it assumes that individuals enter the study on different dates and incorporates the known-fate approach, including the last date the animal was seen alive and the last date it was checked for, acknowledging that mortalities may occur between sampling intervals. Program MARK extrapolated the daily survival rate across the number of days of the study to estimate survival for that time frame (White & Burnham, 1999; Dinsmore, White & Knopf, 2002; Murray, 2006). We tracked snakes over a span of 106 days although each snake was tracked for a portion of that time. We continued post-release monitoring at the 2011 release site in subsequent years, using coverboard and visual encounter surveys for presence of headstarts. Results Aug-2010 Sep-2010 Oct-2010 Nov-2010 Dec-2010 Brumated females Brumated males Active females Active males Dec Jan-2011 Feb-2011 Mar-2011 Apr-2011 May-2011 Jun-2011 Jul-2011 Aug-2011 Sep-2011 Oct-2011 Nov-2011 Date Figure 1 Mean (SD) body mass (g) of brumated and active male and female smooth green snakes at Lincoln Park Zoo from hatching through November The arrows at 15 December 2010 and March 2011 indicate the beginning and completion of the brumation period at the zoo. In 2011, brumated headstarts decreased in body mass during brumation, losing % of their body mass over 78 days (Fig. 1). While there was no sex or treatment-bysex interaction effect on loss of body mass during brumation, males lost a mean (sd) of 7.0% (± 6.9) body mass while females lost a mean (sd) of 9.6% (±9.3) body mass (Table 2). Growth rates from hatching until release did not differ with brumation, but did differ with sex, with males growing g per day and females growing g per day (Tables 2 and 3). There were no effects of treatment, sex or treatment-by-sex interactions on body size among snakes prior to release in June 2011 (Tables 2 and 3; Figs 1 and 2). However, brumated headstarts exhibited remarkable compensatory growth following brumation, attaining the size of active headstarts within 2 months of warm-up (Figs 1 and 2). Active snakes continued growing at a fairly constant rate throughout the year until reaching adult size. For the 2012 headstarts, post-brumation body mass and growth rate from hatching to release varied with sex, but not with treatment or treatment-by-sex interactions (Table 4). Brumated snakes lost % of their body mass over 60 days, with males losing a mean (sd) of 10.7% (±3.1) and females losing a mean (sd) of 12.1% (±5.4) (Fig. 3). Brumated headstarts compensated for their loss of mass and attained the body size of the active snakes (Fig. 3). However, there were no effects of treatment, sex or treatment-by-sex interactions on growth rate during the 2 months following brumation (Table 4). When 2011 and 2012 snakes were pooled, the treatment effect on growth rate during the 2 months following brumation remained significant (F 1, 59 = 11.53, P < 0.01), with no sex effect (F 1,59 = 0.08, P = 0.77) or treatment-by-sex effect (F 1, 59 = 2.71, P = 0.10). Five of 12 brumated 1.5-year-old females produced clutches of unfertilized eggs at the zoo. Unfertilized eggs were smaller than fertilized eggs and were somewhat translucent. None of the eight active females produced eggs. Spermatozoa were observed in cloacal smears of 11 of 13 brumated males, but not from the eight active males. The juvenile ex situ survival rate from hatching until brumation was 0.89 in A few mortalities occurred during the first month post-hatching, prior to assignment to treatments, with three neonates that failed to thrive and two incidents of neonate cannibalism. All headstarts survived through brumation and until release. In 2012, all headstarts survived from hatching through brumation and then through release, regardless of treatment. Small sample size limited our ability to examine differences in short-term post-release survival across brumation and release treatments. The post-release daily survival estimate for headstarts pooled across treatments was (se ± 0.001), (95% confidence interval = ). Extrapolated across the 106 days of the study, the survival estimate from the first release through the first frost was (se ± 0.152), (95% confidence interval = ). We documented one predation event in each release treatment and each brumation treatment with a brumated soft release female and an active hard release male succumbing to small mammal bites in July and October respectively. Snakes from both release and brumation treatments were tracked to burrows following nights when temperatures dropped below 15.5 C, with no observed difference in temperature responses. Mean daily movements differed among headstarts and residents (F 2 = 3.87, P = 0.043) with Tukey s test indicating variation between soft releases and residents, but not between soft and hard releases, or hard releases and 68 Animal Conservation 17 (Suppl. 1) (2014) The Zoological Society of London

5 A. B. Sacerdote-Velat et al. Headstarting techniques in snake reintroduction Table 2 Comparison of mean metrics of growth for the brumation experiment Metrics mean (SE) Brumated Active Females Males Brumated females Active females Brumated males Post-brumation mass (g) (0.38) (0.48) (0.44) (0.43) (0.55) (0.68) (0.53) (0.68) Post-brumation snout-vent length (mm) (3.91) (4.88) (4.46) (4.39) (5.64) (6.91) (5.42) (6.90) Growth rate (g per day): 2 months post-brumation (0.001) (0.001) (0.001) (0.001) (0.001) (0.002) (0.001) (0.002) Growth rate (mm per day): 2 months post-brumation (0.0002) (0.0003) (0.0003) (0.0003) (0.0004) (0.005) (0.0004) (0.0005) Growth rate (g per day): hatching until release (0.001) (0.002) (0.001) (0.001) (0.002) (0.003) (0.002) (0.003) % Change (g) during brumation (1.68) (2.10) (1.92) (1.89) (2.43) (2.98) (2.33) (2.98) Dependent variables include brumation treatment, sex and sex by treatment interaction. Boldface values indicate significant differences among means. Active males Table 3 Univariate results for comparisons of mean growth metrics for the brumation experiment Metrics mean (SE) Brumation treatment Sex Sex-by-brumation treatment Post-brumation mass (g) F 1, 37 = 0.09 F 1, 37 = 0.14 F 1, 37 = 0.19 P = 0.75 P = 0.70 P = 0.66 Post-brumation snout-vent length (mm) F 1, 37 = 0.11 F 1, 37 = 0.0 F 1, 37 = 0.09 P = 0.74 P = 0.99 P = 0.75 Growth rate (g per day): 2 months post-brumation F 1, 37 = F 1, 37 = 0.39 F 1, 37 = 1.46 P < P = 0.53 P = 0.34 Growth rate (mm per day): 2 months F 1, 37 = 0.91 F 1, 37 = 0.39 F 1, 37 = 0.11 post-brumation P = 0.34 P = 0.53 P = 0.23 Growth rate (g per day): hatching until release F 1, 37 = 0.04 F 1, 37 = 5.28 F 1, 37 = 0.91 P = 0.84 P = 0.02 P = 0.34 % Change (mass) during brumation F 1, 37 = F 1, 37 = 0.45 F 1, 37 = 0.08 P < P = 0.50 P = 0.77 Dependent variables include brumation treatment, sex and sex by treatment interaction. Boldface values indicate significant differences among means. residents (Table 5). All telemetered snakes remained within 30 m of their release location. Recapture frequency for non-telemetered snakes did not differ significantly among hard and soft releases (t = 1.70, d.f. = 12, P = 0.114). However, there was a mean (sd) of13 (±10.4) recaptures of soft releases versus a mean (sd) of 6.2 (±3.2) recaptures for hard releases. The elevated frequency of recaptures for soft releases facilitated monitoring. Severe drought in 2012 limited snake activity during follow-up monitoring. Two soft-released brumated female headstarts from the 2011 cohort were recaptured in the site in 2012 before onset of the drought, and both appeared gravid with body masses > 20 g. Discussion Through comparative approaches, we found that brumated headstarts had equivalent prerelease size and survival rates to active headstarts through compensatory growth. While reproductive potential may not be readily assessed with juveniles, we observed that brumated headstarts displayed Mean (SD) snout vent length (mm) Sep-2010 Nov-2010 Jan-2011 Date Feb-2011 Apr-2011 Brumated females Brumated males Active females Active males Figure 2 Mean (SD) snout-vent length (mm) of brumated and active male and female smooth green snakes at Lincoln Park Zoo from October 2010 to May The arrows at 15 December 2010 and March 2011 indicate the beginning and completion of the brumation period at the zoo. Jun-2011 Animal Conservation 17 (Suppl. 1) (2014) The Zoological Society of London 69

6 Headstarting techniques in snake reintroduction A. B. Sacerdote-Velat et al. Table 4 Comparison of mean metrics of growth for the 2012 brumation experiment Metrics mean (SE) Brumation treatment Sex Sex-by-brumation Treatment Post-brumation body mass (g) F 1, 21 = 0.59 F 1, 21 = F 1, 21 = 0.80 P = P < P = 0.86 Growth rate (g per day): 2 months post-brumation F 1, 21 = 3.01 F 1, 21 = F 1, 21 = 4.34 P = P = P = Growth rate (g per day): hatching until release F 1, 21 = 0.15 F 1, 21 = F 1, 21 = 0.03 P = P < P = Dependent variables include brumation treatment, sex, and sex by treatment interaction. Boldface values indicate significant differences among means. Table 5 Summary of movements of radio transmitted hard release headstarts, soft release headstarts and wild resident Opheodrys vernalis in a founder population site Treatment Mean (SD) daily movement (m per day) Range moved (m per day) Tukey s post hoc comparisons Hard releases (n = 7) 2.24 (±0.91) Hard releases versus Soft releases d.f. = 16 P = 0.12 Soft releases (n = 7) 1.61 (±0.85) Wild residents versus Soft releases d.f. = 16 P = 0.04 Wild residents (n = 5) 4.90 (±0.90) Wild residents versus Hard releases d.f. = 16 P = 0.86 Mean (SD) body mass (g) Sep-2011 Oct-2011 Nov-2011 Dec-2011 Jan-2012 Feb-2012 Mar-2012 Apr-2012 May-2012 Jun-2012 Jul-2012 Aug-2012 Date Brumated females Brumated males Active females Active males Figure 3 Mean (SD) body mass of brumated and active male and female headstarts from September 2011 to August The arrows indicate the beginning and end of induced brumation in April and June signs of reproductive potential that were lacking in active snakes, including production of unfertilized eggs and spermatozoa. We were unable to track headstart overwintering fate, but only brumated headstarts have been encountered in subsequent years since release. However, detection probability is low for O. vernalis and non-detection is not equivalent to mortality. Because we observed equivalent ex situ survival and growth in both treatments, with no apparent advantage to maintaining year-round activity in snakes, we recommend incorporating brumation into headstarting efforts as a precaution for temperate zone snakes, where experience adapting to temperature extremes may be beneficial (Roe et al., 2010). Compensatory growth in brumated headstarts demonstrates the ability of snakes to adapt to environmental extremes (Bjorndal et al., 2003) and attain equivalent size to active headstarts. In 2011, brumated snakes exhibited rapid compensatory growth. While we observed compensatory growth in our 2012 experiment, the difference in growth rate among treatments following brumation was not as great as in This difference across years may relate to delayed brumation in The 2012 headstarts were active longer prior to cool-down than in 2011 and the overall body mass decrease during brumation was greater. If compensatory response is reduced or slowed with delayed brumation, active headstarts may take longer to recover from overwintering weight loss, post-release. However, when years are pooled, the greater growth rate of the brumated snakes in the 2 months following brumation remains significant. Similar compensatory growth was observed in hatchling jacky dragons Amphibolurus muricatus, maintained in high and low prey treatments in enclosures (Radder, Warner & Shine, 2007). Low prey hatchlings had lower growth rates than the high prey hatchlings. However, after 6 months with equivalent prey, the original low prey subjects demonstrated compensatory growth, attaining body size of high prey lizards. Similarly, chuckwallas Sauromalus obesus from high and low elevations with prolonged and abbreviated growing seasons respectively, showed greater nutritional uptake and greater small intestine mass from brumation until active season in the population with the abbreviated active season (Tracy & Diamond, 2005). As compensatory growth is a mechanism for surviving variable conditions, inducing brumation may provide beneficial experience adapting to resource variation post-release. 70 Animal Conservation 17 (Suppl. 1) (2014) The Zoological Society of London

7 A. B. Sacerdote-Velat et al. Headstarting techniques in snake reintroduction Headstarts did not reproduce during the first 2 years of the program. However, reproductive maturity for the species is estimated at age three (Seibert & Hagan, 1947; Hammerson, 1999). Several brumated headstarts demonstrated reproductive potential with production of unfertilized eggs and spermatozoa at 1.5 years of age, which was absent in active headstarts. While body size was similar across treatments, in 2011, brumated females were 11% heavier on average than active females within 2 months of brumation, which could confer a reproductive benefit, both in mate selection and clutch size. In 2012, active females were 8% heavier on average than brumated females within 2 months of brumation. With delayed brumation in 2012, females lost a greater percentage of mass as they did in 2011, so the compensatory response may have been slower. Despite the slightly larger size of active females in 2012, only brumated females produced eggs. Telemetry of headstarts and residents documented small movements of m per day across groups. Plummer (1990, 1997) found similar movements in a study of resident rough green snakes O. aestivus averaging 1.3 m per day, but nesting females increased movements to 30 m per day. Following our releases, all tracked headstarts remained within 30 m of the release site. The larger movements of resident snakes as compared with headstarts may relate to differences in ages of tracked snakes. While body size was comparable between residents and headstarts, residents were reproductive adults that may have been exhibiting movements relating to nesting or fall mating. No headstarts were gravid at release, so nesting-related movements were not observed. We did not observe differences between movements of soft and hard release snakes, which may relate to age, size, prey availability or presence of conspecifics in the augmentation site. Comparative releases in a site lacking scent cues of conspecifics may produce different results. However, apparent fidelity to the release site may relate to communal behavior in O. vernalis, such as nesting and denning (Cook, 1964; Fowler, 1966; Gregory, 1975; Sacerdote et al., 2012). Plummer (1997) observed non-random distribution in O. aestivus with individuals using a specific strip of shoreline vegetation and suggested that outside of communal behaviors, canopy-level vegetation structure on the microhabitat scale may influence clumped distributions, as O. aestivus are arboreal. In our release site, high density areas of O. vernalis occur in a homogenous mix of tallgrass vegetation, so clumped distributions may relate more to behavior or prey base. The apparent clumped distribution of snakes has management implications. A high degree of site fidelity was observed, with frequent recaptures of headstarts and wild snakes, which is desirable for augmentation and reintroduction. Snakes tended to aggregate which is imperative for breeding. However, restoration managers must be aware of this clumped distribution in planning restoration practices (e.g. burns) and maintain nearby refugia. Minimal dispersal from the release site may be explained by scent trails from conspecifics, impending fall mating or a high-quality resource base at the site. Small insectivorous snakes may be less subject to prey dispersal as compared with snakes that feed on more mobile prey such as rodents. Density of insect prey may be great enough in small habitat patches to minimize wandering movements following release. While the post-release estimate of reflects high daily survival, the large confidence interval for the duration of the season reflects great uncertainty regarding long-term fate of headstarts. Recaptures of 2011 headstarts in 2012 included two brumated females. However, with severe drought limiting activity through 2012, we cannot ascertain if lack of recaptures reflects low overwintering survival. Monitoring is ongoing to improve our knowledge of long-term survival of headstarts. Recovery focused on small cryptic snakes presents logistical challenges to monitoring success. Our comparative framework allowed evaluation and modification of headstarting and release practices for successive years of the recovery effort. Compensatory growth and equivalent survival of brumated headstarts was unexpected. However, we were constrained in addressing questions of survival with release type by the number of headstarts. While movements were similar for soft and hard releases during initial postrelease monitoring, this pattern may relate to communal behaviors and resource base of O. vernalis specifically, and results may vary by species, and with the presence of conspecifics in the release site. We found that the slightly greater likelihood of recapturing soft releases provided enough benefit to continue the soft release approach in our program. Additional monitoring is required to assess long-term survival associated with each headstarting and release treatment. Use of comparisons in rearing and releasing headstarted snakes provides insight into efficacy of applied conservation techniques. Generally, we recommend use of such comparisons for continued development of a model framework for reptile recovery and wildlife reintroductions. Acknowledgments We thank Lake County Forest Preserve District and Lincoln Park Zoo Conservation and Science and Animal Care staff for their support and input. References Bertolero, A., Oro, D. & Besnard, A. (2007). Assessing the efficacy of reintroduction programmes by modelling adult survival: the example of Hermann s tortoise. Anim. Conserv. 10, Bjorndal, K.A., Bolten, A.B., Dellinger, T., Delgado, C.A. & Martings, H.R. (2003). Compensatory growth in oceanic loggerhead sea turtles: response to a stochastic environment. Ecology 84, Bona-Gallo, A. & Licht, P. (1983). Effect of temperature on sexual receptivity and ovarian recrudescence in the garter Animal Conservation 17 (Suppl. 1) (2014) The Zoological Society of London 71

8 Headstarting techniques in snake reintroduction A. B. Sacerdote-Velat et al. snake (Thamnophis sirtalis parietalis). Herpetologica 39, Cook, F.R. (1964). Communal egg laying in the smooth green snake. Herpetologica 20, 206. Dinsmore, S.J., White, G.C. & Knopf, F.L. (2002). Advanced techniques for modeling avian nest survival. Ecology 83, Fowler, J.A. (1966). A communal nesting site for the smooth green snake in Michigan. Herpetologica 22, 231. Germano, J.M. & Bishop, P.J. (2009). Suitability of amphibians and reptiles for translocation. Conserv. Biol. 23, Gibbons, J.W., Scott, D.E., Ryan, T.J., Buhlmann, K.A., Tuberville, T.A., Metts, B.A., Greene, J.L., Mills, T., Leiden Y., Poppy S. & Winne, C.T. (2000). The global decline of reptiles, deja vu amphibians. Bioscience 50, Gregory, P.T. (1975). Aggregations of gravid snakes in Manitoba, Canada. Copeia 1975, Hammerson, G.A. (1999). Amphibians and reptiles in Colorado. Niwot, CO. University Press of Colorado and Colorado Division of Wildlife. Haskell, A., Graham, T.E., Griffin, C.R. & Hestbeck, J.B. (1996). Size related survival of headstarted redbelly turtles (Pseudemys rubriventris) in Massachusetts. J. Herpetol. 30, Herlands, R., Wood, R., Pritchard, J., Clapp, H. & Le Furge, N. (2004). Diamondback terrapin (Malaclemys terrapin) head-starting project in southern New Jersey, In Conservation and ecology of turtles of the Mid-Atlantic Region: a symposium: Swarth, C., Roosenburg, W.M. & Kiviat, E. (Eds). Salt Lake City, UT: Bibliomania. Hester, J.M., Price, S.J. & Dorcas, M.E. (2008). Effects of relocation on movements and home ranges of eastern box turtles. J. Wildl. Mgmt. 73, King, R., Berg, C. & Hay, B. (2004). A repatriation study of the eastern massasauga rattlesnake (Sistrurus c. catenatus) in Wisconsin. Herpetologica 60, King, R.B. & Stanford, K.M. (2006). Headstarting as a management tool: a case study of the plains gartersnake. Herpetologica 62, Kingsbury, B.A. & Attum, O. (2009). Conservation strategies: captive rearing, translocation, and repatriation, In Snakes ecology and conservation: Mullin, S.T. & Seigel, R.A. (Eds). Ithaca: Cornell University Press. Madrid-Sotelo, C.A. & García-Aguayo, A. (2008). A simple method for externally attaching radio-transmitters to snakes. North-West. J. Zool. 4, Madsen, T. & Shine, R. (2000). Silver spoons and snake body sizes: prey availability early in life influences longterm growth rates of free-ranging pythons. J. Anim Ecol. 69, Mitrus, S. (2005). Headstarting in European pond turtles (Emys orbicularis): does it work? Amphibia-Reptilia 26, Murray, D.L. (2006). On improving telemetry-based survival estimation. J. Wildl. Manage. 70, Plummer, M.V. (1990). Nesting movements, nesting patterns, and nesting behavior of green snakes (Opheodrys aestivus) revealed by radio telemetry. Herpetologica 46, Plummer, M.V. (1997). Population ecology of green snakes (Opheodrys aestivus) revisited. Herpetol. Monogr. 11, Plummer, M.V. & Mills, N.E. (2000). Spatial ecology and survivorship of resident and translocated hognose snakes (Heterodon platirhinos). J. Herpetol. 34, Radder, R.S., Warner, D.A. & Shine, R. (2007). Compensating for a bad start: catch-up growth in juvenile lizards (Amphibolarus muricatus, agamidae). J. Exp. Zool. 307A, Reading, C.J., Luiselli, L.M., Akani, G.C., Bonnet, X., Amori, G., Ballouard, J.M., Filippi, E., Naulleau, G., Pearson, D. & Rugiero, L. (2010). Are snake populations in widespread decline? Biology Letters 6, Redder, A.J., Smith, B.E. & Keinath, D.A. (2006). Smooth green snake (Opheodrys vernalis): a technical conservation assessment. (Online). USDA Forest Service, Rocky Mountain Region. Available: projects/scp/assessments/smoothgreensnake.pdf (accessed October 2011). Rittenhouse, C.D., Millspaugh, J.J., Hubbard, M.W. & Sheriff, S.L. (2007). Movements of translocated and resident three-toed box turtles. J. Herpetol. 41, 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, Roe, J.H., Frank, M.R., Gibson, S.E., Attum, O. & Kingsbury, B.A. (2010). No place like home: an experimental comparison of reintroduction strategies using snakes. J. Appl. Ecol. 47, Rossi, J. (1992). Snakes of the United States and Canada, Keeping them healthy in captivity. Vol. 1, Eastern Area. Malabar: Krieger. Sacerdote, A.B., Glowacki, G. & Schmidtz, T. (2012). Communal nesting in the smooth green snake (Opheodrys vernalis). Herpetol. Rev. 43, 661. Seibert, H.C. & Hagan, C.W. Jr. (1947). Studies on a population of snakes in Illinois. Copeia 1947, Sheperd, D.B., Dreslik, M.J., Jellen, B.C. & Phillips, C.A. (2008). Reptile road mortality around an oasis in the Illinois corn desert with emphasis on the endangered eastern massasauga rattlesnake. Copeia 2008, Szymanski, J. (1998). Status assessment for eastern massasauga rattlesnake (Sistrurus c. catenus). Fort Snelling, 72 Animal Conservation 17 (Suppl. 1) (2014) The Zoological Society of London

9 A. B. Sacerdote-Velat et al. Headstarting techniques in snake reintroduction MN: U.S. Fish and Wildlife Service, Endangered Species Division. Tracy, C.R. & Diamond, J. (2005). Regulation of gut function varies with life-history traits in chuckwallas (Sauromalus obesus: iguanidae). Physiol. Biochem. Zool. 78, 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). Anim. Conserv. 8, Tuberville, T.D., Norton, T.M., Todd, B.D. & Spratt, J.S. (2008). Long-term apparent survival of translocated gopher tortoises: a comparison of newly released and previously established animals. Biol. Conserv. 141, White, G.C. & Burnham, K.P. (1999). Program MARK: survival estimation from populations of marked animals. Bird Study (Suppl. ) 46, Whittier, J.M., Mason, R.T., Crews, D. & Licht, P. (1987). Role of light and temperature in the regulation of reproduction in the red-sided gartersnake (Thamnophis sirtalis parietalis). Can. J. Zool 65, Wylie, G.D., Smith, J.J., Amarello, M. & Casazza, M.L. (2011). A taping method for external transmitter attachment on aquatic snakes. Herpetol. Rev. 42, Animal Conservation 17 (Suppl. 1) (2014) The Zoological Society of London 73