reintroduction and head starting: tools for landing s t ...buhlmann et al.—2010 head-starting...

Herpetological Conservation and Biology 10(Symposium):436–454.Submitted: 7 May 2012; Accepted: 20 February 2015; Published: 28 June 2015.

Reintroduction and Head-starting: Tools for Blanding’sTurtle (Emydoidea blandingii) Conservation

Kurt A. Buhlmann1,5, Stephanie L. Koch2, Brian O. Butler3, Tracey D.Tuberville1, Veronica J. Palermo3, Brian A. Bastarache4 and Zachary A.

Cava2

1University of Georgia, Savannah River Ecology Lab, Aiken, South Carolina, USA2U.S. Fish and Wildlife Service, Sudbury, Massachusetts, USA

3Oxbow Associates, Inc., Boxborough, Massachusetts, USA4Bristol County Agricultural High School, Dighton, Massachusetts, USA

5Corresponding author, e-mail: [email protected]

Abstract.— We reintroduced Blanding’s Turtles (Emydoidea blandingii) to Assabet River NationalWildlife Refuge, Massachusetts, USA, evaluating the relative benefits and risks of using various lifestages of Blanding’s Turtles collected from a donor population within the same watershed, includingdirect-release hatchlings (released in autumn shortly after hatching), head-started hatchlings (raisedin captivity for 9 mo), juveniles, and adults. We developed a simple population model to evaluatewhich of several release strategies was most likely to result in a stable population at the recipient sitewhile minimizing negative impacts to the donor site. Model results suggested that annual releasesconsisting largely of head-started hatchlings were most likely to achieve our goal. We released 81direct-release and 161 head-started hatchlings at the refuge in 2007–2011. Head-started hatchlingswere larger (mean = 62.7 mm carapace length, 46.6 g) compared to direct-release hatchlings (mean =

36.3 mm carapace length, 8.8 g). Simultaneous radio-tracking of 12 translocated sub-adults has pro-vided useful information on habitat preferences that we used to select two sites within the refuge forfuture releases. We also released six head-started hatchlings with radio transmitters (one in 2009 andfive in 2010): one was found dead a year after release. We plan to continue monitoring efforts to assesssurvivorship, growth, and site fidelity of all released Blanding’s Turtles and to compare results amongthe head-started and direct-release hatchlings. We will update our models and reintroduction effortsbased on monitoring data.

Key Words.—Emydoidea blandingii; adaptive management; conservation; hatchlings; head-starting;reintroduction; translocation

Introduction

The ultimate goal of species conservation ef-forts should be the persistence of viable, self-sustaining populations in their native landscapes.Thus, identification and preservation of existingpopulations is the first conservation priority, fol-lowed by habitat restoration and management andthe establishment of conservation corridors con-necting habitat patches. However, in increasinglyhuman-altered landscapes, species of conserva-

tion concern often survive as non-viable popu-lations or as a few viable populations with in-sufficient connections between them. Populationmanipulations, such as reintroduction, may be ap-propriate under circumstances when the threatsand causes of decline have been removed but thepopulations are so small that they are likely todisappear without temporary intervention (Frazer1992; Heppell et al. 1996) or are unlikely or un-able to recolonize sites on their own.

Success of population manipulations such as

Copyright c© 2015. Kurt Buhlmann.All Rights Reserved. 436

mailto:[email protected]

Herpetological Conservation and Biology

reintroductions and head-starting requires ap-propriate planning to minimize negative geneticand disease consequences, as well as impactsto other native species. Long-term monitoring iscritical to evaluating success, as is applicationof an adaptive management framework to incor-porate new knowledge. For long-lived speciessuch as turtles, the time scale for judging suc-cess of population manipulations often exceedsthe tenure of the manager or researcher. Thus,these techniques are still largely viewed as ex-perimental (Seigel and Dodd 2000) and methodshave not been standardized. In general, lowersuccess rates have been reported with non-game(46%) than with game species (86%; Griffith etal. 1989) but few articles have been publishedregarding the success of freshwater turtle rein-troduction projects, including those with a head-starting component (Haskell et al. 1996; Mitrus2005; Vander Haegen et al. 2009). Here we sum-marize the development, implementation, andpreliminary monitoring results of a collabora-tive reintroduction project for Blanding’s Turtles(Emydoidea blandingii), emphasizing the head-starting component.

The Blanding’s Turtle is a medium-sized (to240 mm carapace length [CL]), semi-aquaticfreshwater turtle that inhabits wetlands through-out the upper Midwest and New England statesin the USA and southeastern Canada. Blanding’sTurtles are the most northern-restricted turtlespecies in North America with a latitudinal rangenot exceeding 900 km (Buhlmann et al. 2009).

Blanding’s Turtles are long-lived with a longgeneration time (minimum 37 yr; Congdon etal. 1993) and individuals have been known tosurvive in the wild > 70 yr (Breck and Mori-arty 1989; Congdon and van Loben Sels 1993).Blanding’s Turtles require 14–20 yr to reach sex-ual maturity and minimum size at maturity forfemales is 163 mm CL (Congdon and van LobenSels 1993). Congdon and Keinath (2006) sug-gested through modeling that one breeding fe-male is demographically equivalent to 100 eggs(i.e., 8–12 yr of her reproduction). Nest sur-

vivorship is highly variable and can range from0–100% of all nests in a population in a givenyear. In a long-term Michigan study, first yearsurvivorship of hatchlings was estimated at 26%(Congdon and van Loben Sels 1993), juvenilesurvivorship (> age 1 yr to maturity) averaged72% annually, and high annual adult survivor-ship (96%) characterized natural populations, asadults have few natural predators.

Blanding’s Turtles have been known from NewEngland since the early 1800s (Storer 1839) andthose populations are disjunct from the Midwest-ern portion of the range. The species is of con-servation concern in every New England state inwhich it occurs (Levell 2000). The main threatsto Blanding’s Turtles are road mortality and lossof wetlands and the upland habitats that con-nect them (Congdon and Keinath 2006; Compton2007; Beaudry et al. 2009). Although environ-mental regulations and local community actionhave sometimes been successful in protectingwetland habitats, loss of upland habitats and land-scape fragmentation continue through develop-ment and road construction. Known records forBlanding’s Turtles in New England total 180 ele-ment occurrences (EOs), although 169 of theseare represented by only one or a few animals(Compton 2007). Many of these EOs are obser-vations of turtles crossing roads. Only nine sitesin New England have documented 10–50 turtles(Compton 2007) and only two or three sites inNew England have more than 50 known animals.Thus, most of these sites do not represent long-term viable populations under current conditions.

We identified a new conservation site that webelieve contains the components necessary to sup-port a Blanding’s Turtle population. We are inter-ested in reintroduction because simply proposingto protect existing sites may not be enough tomaintain this species as a viable component ofthe New England landscape. Our objectives wereto: (1) present the decision-making process wecreated and implemented to evaluate the appro-priateness of reintroducing Blanding’s Turtles toa selected site; (2) identify the release scenarios

437

Buhlmann et al.—2010 Head-starting Symposium: Reintroduction and head-starting tools forBlanding’s Turtle.

most likely to achieve a self-sustaining popula-tion using a simple population model that con-siders both biological and logistical constraints;(3) highlight the use of head-starting as a criticalcomponent of our reintroduction strategy; and (4)summarize the preliminary reintroduction resultsof releases using different life stages of Bland-ing’s Turtles.

Materials AndMethods

Identification of a candidate recipient site.—Assabet River National Wildlife Refuge is an 880ha protected area in eastern Massachusetts, ap-proximately 40 km west of Boston, within theheart of the Blanding’s Turtle range in New Eng-land. The refuge was established in 2000 fromlands acquired from the U.S. Army-Fort Devenscomplex. Prior to the 1940s, agriculture and tim-ber activities were conducted and a commercialCranberry bog was in operation. During a periodof extensive use from the 1940s through 1970s,numerous bunkers, buildings, and hand-dug wellswere built. These human activities ceased priorto the transfer of the property to the U.S. Fishand Wildlife Service.

The refuge contains a diversity of wetlandhabitats, including numerous vernal pools,extensive shrub swamps, bogs, and a 24 haglacial lake (Aneptek Corporation, unpubl.report; Butler, unpubl. report). The terrestrialhabitat is primarily forested and is comprisedof White Pine (Pinus strobus) and mixedhardwoods. The size of the refuge, the diversityof wetland habitats embedded in a forestedmatrix, and minimal automobile traffic withinthe refuge suggest that the refuge could supporta sizeable Blanding’s Turtle population. Oneroad-killed female Blanding’s Turtle was foundimmediately east of the refuge in 2000 (pers.obs.). However, surveys have failed to documentextant Blanding’s Turtles within the refuge(Aneptek Corporation, unpubl. report; Butler,unpubl. report; Buhlmann and Gibbons, unpubl.report). Given the lack of sizable populations

nearby and the fragmented landscape betweenthe refuge and other known populations, naturalre-colonization is unlikely.

Evaluating the appropriateness ofreintroduction.—We constructed a “deci-sion tree” to evaluate whether reintroductionwas an appropriate conservation strategy forBlanding’s Turtles at the candidate recipient site.The decision tree consisted of an ordered listof questions (Table 1) to be addressed prior toinitiating any reintroduction activities. Majorfindings resulting from the decision tree processare briefly summarized in this paper. More de-tailed results were included in an EnvironmentalAssessment (U.S. Fish and Wildlife Service.2007. Establishing a population of Blanding’sTurtles (Emydoidea blandingii) on Assabet RiverNational Wildlife Refuge: final environmentalassessment. U.S. Fish and Wildlife Service.Sudbury, Massachusetts, USA.).

Population modeling to evaluate alternativerelease scenarios.—A simple population modelwas developed using VORTEX 9.7 (Lacy et al.2005; Miller and Lacy 2005) to: (1) determinethe numbers of hatchlings that the donor sitecould provide without negatively impacting thestability of that resident Blanding’s Turtle popu-lation; (2) estimate the numbers of hatchlings andthe duration of repetitive introductions that wouldbe necessary to establish a stable population ofBlanding’s Turtles on the recipient site; and (3)predict the relative efficiency of different releasestrategies, including direct release of hatchlingsvs. release of head-started hatchlings.

Life-history parameters were estimated fromthe literature and from unpublished field data,relying on data specific to Massachusetts popula-tions where available. When long-term data wereavailable (Table 2) to estimate specific life historyparameters, we based our models on those datarather than on shorter-term data (even if collectedfrom Massachusetts populations). We developedmodels for both the source population and the

438


Table 1. Identification of “decision–tree” criteria (U.S. Fish and Wildlife Service. 2007. Establishing apopulation of Blanding’s Turtles (Emydoidea blandingii) on Assabet River National Wildlife Refuge: finalenvironmental assessment. Sudbury, Massachusetts, USA.), which we evaluated prior to implementation ofreintroduction efforts of Blanding’s Turtles.

QuestionQ1: Is the species secure in the region?Q2: Is the proposed recipient site within natural range?Q3: Does the proposed recipient site currently have a population?Q4: Is appropriate habitat present on the recipient site?Q5: Is the recipient site secure and protected?Q6: Is the cause of the initial decline known?Q7: Has a donor site been identified?Q8: Have population impacts to the donor site been considered?Q9: Have genetic concerns been addressed?Q10: Have disease transmission concerns been satisfied?Q11: Has an appropriate life stage and protocol been selected?Q12: Has population modeling estimated the numbers of animals needed to achieve population viability?Q13: Is there a plan for habitat management at the recipient site?Q14: Is there a plan for habitat management at the donor site?Q15: Is there a monitoring commitment commensurate with modeling results/habitat management needs?

recipient population.The source population model was based on

long-term data collected from that site. We per-formed manipulations of the assumed populationgiven by the source model each year for 15 ybut we continued simulations for another 85 y.Model scenarios included no nest protection (i.e.,nest survivorship = 30%), nest protection of 30nests per year but no harvest (i.e., nest survivor-ship for first 30 nests = 85%, 30% for all othernests), and nest protection and harvest of 50, 100,or 150 hatchlings for release at the recipient site.We averaged population growth rates for eachscenario over the course of the simulation.

The recipient site model was based on astarting population of zero adults, 10 y of annualhatchling releases, and a simulation durationof 50 y. Model scenarios include release of 25,50 or 100 direct-release hatchlings, or 25 or 50head-starts. We limited the number of head-startsin our scenarios to 50 individuals, which weassumed to be the number we could feasibly rearin captivity (based on the resources availableat the beginning of the project). The model

assumed no mortality during the head-starting(captive) phase and that head-starts survivedsimilarly to wild-recruited turtles. Number ofadult turtles (averaged over 100 simulations)at 15, 20, 25, and 50 y were reported for eachscenario.

Collection and treatment of hatchlings fromdonor site.—Blanding’s Turtle nests at the donorsite have been periodically protected from preda-tors since 1987 using wire cage covers, which ar-tificially boost long-term nest success and hatch-ling recruitment. The number of nests protectedeach year varied (range 16–43), with hatch-ing success averaging 85% for protected nests(unpubl. data). During 1987–1990, we direct-released 235 hatchlings from their protected nestsinto donor site wetlands; in the six nesting sea-sons (2000–2005) immediately prior to start ofthis reintroduction project, we similarly released1083 hatchlings at the donor site. Since the startof our reintroduction project in 2006, donor sitenests were protected to produce hatchlings forboth the donor and recipient sites. To accomplish

439


Table 2. Baseline parameters used in development of repatriation models for Blanding’s Turtle (Emydoideablandingii).

Demographic parameter ValueLongevity 70 yFemale age at maturity 17 yMale age at maturity 14 y% females breeding 80%Average clutch size 10 eggsHatchling sex ratio 50:50Hatchling survivorship (age 0-1) 30%Juvenile annual survivorship (age 1-13) 78%Adult survivorship (age 14+) 96.5%

this, we visited known nesting areas in eveningslate May-June and searched for nesting females.When nesting females finished depositing theireggs, we installed predator excluders, collectedGPS locations at nests, and noted the identity ofthe female (or marked her by notching scutes,if previously unmarked). Beginning in early tomid-August, we returned to nests daily to collectemerging hatchlings.

After emergence, we measured all hatchlings(CL to nearest mm) and weighed them to nearest0.1 g. We gave each hatchling an individual codeby marking marginal scutes with cuticle scissorsor nail clippers. Generally, we immediatelydirect-released half of the hatchlings from eachnest (e.g., half of each female’s reproductiveinvestment) into the main wetland at the donorsite. We designated the other half for therecipient site and further divided this group; halfof these we direct-released into either TaylorBrook Wetland (2007, 2008) or Pump StationWetland (2009; Table 3). The other half ofthe recipient site hatchlings were retained forhead-starting through the winter in captivity.However, in autumn 2010, no hatchlingswere directly released at the recipient site; in-stead all were retained for head-starting (Table 3).

Selection of specific release sites.—Ourinitial release location at the refuge was TaylorBrook Wetland, an herbaceous emergent marsh

with clumped grasses, sedges, water lilies, andduckweed on the eastern side of the glaciallake. This marsh forms the headwaters of TaylorBrook. Water levels in Taylor Brook Wetlandvary seasonally but the wetland is often keptflooded by resident North American Beavers(Castor canadensis) that maintained a damdownstream; water levels often exceeded 2 m insome areas in the spring. We began releasinghatchlings and head-starts in Pump StationWetland in 2009, a scrub-shrub and marsh habitatwith unconsolidated bottom currently influencedby beaver activity. Pump Station Wetland wasdominated by shrubby, woody vegetation, in-cluding Leatherleaf (Chamaedaphne catyculata),Sweet Gale (Myrica gale), Alder (Alnus rugosa),Red Maple (Acer rubrum), Swamp Rose (Rosapalustris), Water Willow (Decodon verticillata),Sphagnum with patches of open water, andduckweed-covered pockets. At both release sitesin the refuge (Fig. 1), we released hatchlings intoareas of vegetative cover typical of Blanding’sTurtle habitat that consisted of emergent grasses,sedges and rushes, Leatherleaf, Sphagnum, andduckweed (Butler and Graham 1995).

Head-starting techniques.—During the win-ters of 2007–2008 and 2008–2009, we main-tained head-started turtles in 37.8 L aquariums ingroups of three to 10 animals. We fed hatchlingsReptoMin R© floating food sticks (TetraFauna

440


(a) Taylor Brook Wetland . (b) Pump Station Wetland.

Figure 1. Release sites of Blanding’s Turtles (Emydoidea blandingii) at the Assabet River National WildlifeRefuge, Massachusetts, USA. (Photographed by Kurt A. Buhlmann).

United Pet Group, Blacksburg, VA) and Gam-mare dried shrimp (Vitakraft Pet Products Co,Inc., Bound Brook, NJ) daily, and occasionallive insects and tadpoles (which we collected atthe recipient site). A calcium supplement wasoffered weekly. We included basking lights andstructures in each aquarium, and we used waterheaters when necessary to maintain water temper-atures at 24–29 ◦C. Air temperatures fluctuatedbetween 18–30 ◦C depending on the type of bask-ing light we used.

During 2009–2010, we maintained head-started hatchlings in a greenhouse at a local highschool. Initially, we maintained hatchlings in 41L plastic bins Sterilite Corporation, Townsend,Massachusetts, USA; 88.6 cm L x 42.2 cm Wx 15.6 cm H) of seven to eight animals (Fig. 2).Each bin was equipped with a basking site and aPowerSunT M lamp (Zoo Med Laboratories, Inc.,San Luis Obispo, California, USA), and we pro-vided plastic plants (Exo Terra, Mansfield, Mas-sachusetts, USA) for hiding and ease of cleaning.We cleaned bins each day. We set air temper-ature in the greenhouse to 26.7 ◦C. One weekprior to release, we set greenhouse air temper-atures to a cycle that mimicked day (23.9 ◦C)and night (18.3 ◦C) temperatures at the recipi-

ent site. We weighed and measured head-startedturtles weekly while in captivity. We separatedanimals by size (weight) as they grew to mini-mize injuries. We maintained head-started turtlesof the three smallest size categories (i.e., < 10 g,10–14.9 g, 15–24.9 g) in bins. As some turtlesstarted to outgrow the bins (after attaining ≥ 25 gbody weight), we moved them to 1,741 L aquacul-ture tubs, with turtles in each size category (i.e.,25–49.9 g, 50–74.9 g and ≥ 75 g) maintained indifferent tubs (Fig. 3), with appropriately shallowwater levels. We fed head-started turtles a turtle“jello” diet consisting of beef heart cut into cubes,dried krill or large shrimp with shells, commer-cial turtle pellets, shredded sweet potato, calcium,and multi-vitamin supplements (Exo Terra, Mans-field, Massachusettes, USA) incorporated into agelatin matrix. Items were mixed in a blenderand added to unflavored gelatin (Knox R©, KraftFoods Group, Northfield, Illinois, USA). We fedturtles twice daily on weekdays and once daily onweekends and school vacation. If an individuallost 5–7% body mass between weekly measure-ments, it was moved to isolation and offered theturtle jello diet and chopped earthworms until itsmass increased. We also occasionally fed turtlesReptoMin R© and live foods (insects, worms, small

441


Figure 2. Head-started Blanding’s Turtles (Emydoidea blandingii) being maintained in a high school green-house in winter 2009–2010. Newly-emerged hatchlings were initially reared in small, shallow plastic bins.(Photographed by Brian A. Bastarache).

Figure 3. Head-start rearing facilities for Blanding’sTurtles (Emydoidea blandingii) in high school green-house. Once head-started hatchlings start to grow,they were sorted by size and transferred to 1471 Ltubs, where they were maintained until release. (Pho-tographed by Brian A. Bastarache).

crayfish, and tadpoles). In the month before re-lease, meal worms (Tenebrio for small turtles andZophobas for the larger ones) replaced jello andReptoMin R© for the afternoon feeding.

We released head-started hatchlings in lateMay when freezing nights were no longer likelyto occur. The largest head-started hatchling waslarge enough (115.5 g) to hold a radio-transmitter(Model R1680, 3.6 g; Advanced Telemetry

Systems, Isanti, MN) at the time of release inlate May, and we maintained four additionalhead-start turtles in captivity for the summerand then released them with radio-transmitters.Radios and epoxy added 6–8 g to each turtleand were no more than 7% of the individualturtle’s body weight. We radio-tracked thehead-started turtle released in May all summerand we recaptured, re-measured, and fitted itwith a new radio prior to winter torpor.

Translocated sub-adults.—We captured sub-adult Blanding’s turtles at the donor site usingbaited hoop traps. We selected 12 of the captures(known ages 4–11 yr) that were large enough tocarry radio transmitters (Model R1680 weighing3.6 g or Model R1920 weighing 14 g, dependingon turtle weight; Advanced Telemetry Systems,Isanti, Minnesota, USA). We chose sub-adults toavoid removing reproductive individuals from thedonor population. We trapped May-June 2008,May 2009, and August 2010. We translocatedthese 12 sub-adults to the recipient site and re-leased them into the same wetlands as the hatch-lings. We radio-tracked these sub-adults at leastweekly to obtain information about habitat se-lection, site fidelity, and winter torpor that we

442


could not obtain from the direct-release hatch-lings and most head-start turtles, which were toosmall to carry radio-transmitters. Monitoring thehabitat selection by translocated sub-adults wasdesigned to help us determine the best future suit-able sites for release of head-started and direct-release hatchlings.

Results

Evaluating the appropriateness of reintro-duction.—We answered the 15 decision treequestions (Table 1) and determined that fewviable populations existed in the local landscapeand that the species was not secure in the region(Q1). The proposed recipient site (Refuge)was within the species’ natural range (Q2), didnot have an extant population (Q3), was largeenough with appropriate habitat (Q4), and wassecure (Q5). We were not certain of the initialcause of decline (Q6), but prior military useof the site and hydrologic manipulation forcranberry cultivation and other farming activities(e.g., ditching and draining of wetlands forarable land) may have had negative effects.There were no identifiable contemporary threatsto the site. We identified a suitable donor site(Q7) that had a sizeable Blanding’s Turtlepopulation and a history of nest protection,nesting habitat management, and populationmonitoring. In addition, we reserved half ofthe hatchlings from each clutch for release atthe donor site, thus minimizing impacts to thedonor population (Q8) as its viability remainsparamount to Blanding’s Turtle conservation inthe region. The donor site and recipient site arewithin the same greater river basin, and wouldlikely be part of the same metapopulation in anunfragmented pre-European landscape, thus bothgenetic (Q9) and disease transmission issues(Q10) are probably largely negated. Head-startedanimals were not exposed to other species duringcaptivity. Selection of appropriate life stages(Q11) and numbers needed (Q12) are addressedbelow. Habitat management projects were in

place at recipient (Q13) and donor (Q14) sitesand are discussed below, as well as adaptivemanagement plans for long-term monitoring(Q15).

Evaluation of potential release scenar-ios.—Our source model predicted a growthrate of 0.01 for the source population undera scenario of no manipulations (i.e., no nestprotection or harvest; Fig. 4). Scenarios basedon protection of 30 nests predicted populationgrowth rates of 0.009–0.011, depending on thenumber of hatchlings harvested for release at therecipient site (Fig. 4). Thus, protecting a smallportion (n = 30) of total nests should be adequateto compensate for animals harvested from thedonor site under any of the scenarios we modeled.In our recipient site model scenarios (Fig. 5),the predicted number of adult Blanding’s Turtlesbegins accelerating at Year 15 around the timewhen the first introduced turtles reach maturity(males mature at age 14 y, females at age17 y; Congdon et al 1993). By that time, allhatchling releases have been completed. Thenumber of adults then plateaus once all survivingintroduced turtles reach maturity, after whichtime population growth is dependent on theirnesting and reproductive success after populationmanipulations are discontinued. All releasescenarios predict a stable recipient populationby the end of the 50 y simulations (Fig. 5), buthead-starting appears to be twice as effective asdirect-release of hatchlings (assuming that head-started turtles do behave normally after release).For example, our model predicts that annualreleases of 50 head-starts per year results in apredicted population similar to directly-releasing100 hatchlings per year, and that it will reach astable population size sooner. However, becausehead-starting is more intensive in terms of timeand resources and because no data were availableto determine whether released head-starts werelikely to behave similarly to wild-recruitedBlanding’s Turtles, we chose a release strategyinitially comprised of both head-started and

443


direct-release hatchlings.

Direct-release hatchlings at recipientsite.—We released 81 direct-release hatchlingsfrom 36 nests in the refuge in autumns of2007, 2008, and 2009 (Table 3). Direct-releasehatchlings averaged 36.3 mm CL and 8.8 gover the three years, were similar in size amongyears, and were comparable in size to hatchlingsselected for head-starting (Fig. 6). A 2008 DRhatchling (#3294) was incidentally recaptured18 October 2010 in the Pump Station Wetland,although it had been originally released inthe Taylor Brook Wetland. This capture wasin the same area where translocated, radio-tracked, sub-adults hibernated in previous years.This head-started turtle had increased in sizefrom 37.9 mm CL and 10.1 g at time of releaseon 9 September 2008 to 88.4 mm CL and 105.4 g.

Head-started hatchlings.—Survivorship ofhead-started animals in captivity ranged from91–100% among cohorts (Table 3). Three of the2008 cohort died from accidental exposure to hotwater. One of the 2009 cohort died the first dayin captivity of unknown causes, and two drownedaccidentally after getting trapped under the bask-ing structures on 31 January and 30 April 2010.Head-started hatchlings averaged 36.8 mm CLand 9.3 g at hatching and 61.7 mm CL and 44.1g at the time of release (Table 3; Fig. 6). Weintroduced 161 head-started hatchlings from 59nests to the refuge during 2008–2011 after beingmaintained in captivity for 9 mo. In addition, wereleased seven head-started animals from 2006 inthe autumn of 2007; these were the only animalsmaintained in captivity for a full year (Table 3).

The largest head-started hatchling produced todate (#3253; 99.5 mm CL, 163.5 g at release)was from the 2008 cohort. It was fitted with atransmitter and released into the Pump StationWetland on 27 May 2009. This turtle was recap-tured on 13 October 2009 in the same wetlandand had decreased in weight from 163.5 g to143.0 g, although it was larger than a wild-caught

4–yr old turtle from the donor site (Fig. 7). Basedon telemetry data, we believe the animal survivedthe winter but then died of unknown causes; weretrieved it from the Pump Station Wetland on 25May 2010. In February 2011, all five head-startedhatchlings from the 2009 cohort were tracked tohibernation locations below the ice. Four werein the Pump Station Wetland and one was in theTaylor Brook Wetland. The latter was found deadin March, as was a Snapping Turtle (Chelydraserpentina), presumably as a result of winter kill.

Discussion

Appropriateness of reintroduction at thecandidate recipient site.—Our decision toreintroduce Blanding’s Turtles to the AssabetRiver National Wildlife Refuge appears to bean appropriate conservation action based on theresults of our decision tree. This large refugewith extensive and diverse wetland habitatsprovided, in our opinion, an opportunity forproactive Blanding’s Turtle conservation bypotentially increasing the number of viablepopulations. This becomes especially significantgiven that at most only three populations inNew England currently contain > 50 turtles(Compton 2007). Although we are not certainthat a previous population was extirpated fromthe site, the habitat currently appears appropriateand is no longer subject to heavy human use. Inaddition, a 2000 record for a road-killed gravidfemale Blanding’s Turtle (Massachusetts NaturalHeritage and Endangered Species ProgramDatabase, Westborough, MA) on the two-laneroad adjacent to the east boundary of the refugeargues strongly for the historical presence ofBlanding’s Turtles in this exact portion of theNew England landscape. The refuge is largeenough to encompass the large home ranges thathave been reported for Blanding’s Turtles (63.0ha, Piepgras and Lang 2000; 24.8 ha, Babbitt andJenkins, unpubl. report; 27.5 ha, Grgurovic andSievert 2005). The refuge can also accommodatesome, but not all, long distance movements

444


Figure 4. Effect of potential population manipulations on the growth rate of the source population ofBlanding’s Turtles (Emydoidea blandingii) released 2006–2011 at the Assabet River National Wildlife Refuge,Massachusetts, USA. The dark bar on the left predicts population growth rate under no manipulations (i.e., nonest protection, no harvest). The light bars to the right predict population growth rate under varying harvestscenarios when 30 nests are protected. Note that all predicted population growth rates are > 0.

Figure 5. Predicted number of adult Blanding’s Turtles (Emydoidea blandingii) at the Assabet River NationalWildlife Refuge, Massachusetts, based on introducing 25, 50, or 100 direct-release hatchlings or 25 or 50head-started turtles. Simulations were based on 10 y of releases with results after 15, 20, 25 and 50 y presented.The lines are color coded by the number of animals released in each scenario; dashed lines correspond todirect-release scenarios, solid lines to head-start scenarios.

445


Figure 6. Yearly (2007–2010) comparisons of initial hatchling sizes for cohorts of Blanding’s Turtles(Emydoidea blandingii) selected for direct-release in autumn (first column) and those retained for head-starting over one winter (second column). The third column in each yearly comparison shows the releasesize of the head-started turtles the following May. No direct releases were made in 2010. Means ± 1 SD arepresented.

446

Herpetological Conservation and BiologyTable

3.C

ompa

riso

nof

cara

pace

leng

th(C

L)a

ndm

ass

(g)o

f242

hatc

hlin

g(8

1di

rect

-rel

ease

d[D

R]a

nd16

1he

ad-s

tart

ed[H

S])B

land

ing’

sTu

rtle

s(E

myd

oide

abl

andi

ngii)

rele

ased

2006

–201

1att

heA

ssab

etR

iver

Nat

iona

lWild

life

Ref

uge,

Mas

sach

uset

ts,U

SA.D

ates

and

wet

land

sof

rele

ase

and

surv

ivor

ship

sfo

rHS

whi

lein

capt

ivity

are

also

give

n.R

elea

sesi

tes

wer

eTa

ylor

Bro

okW

etla

nd(T

B)a

ndPu

mp

Stat

ion

Wet

land

(PS)

.NN

=nu

mbe

rof

nest

s.T

hem

eans±

S.D

.and

rang

esar

epr

esen

ted

inea

chca

tego

ry.*

the

seve

nhe

ad-s

tart

edtu

rtle

sfr

om20

06co

hort

wer

ere

tain

edin

capt

ivity

foro

neye

arbe

fore

rele

ase.

All

subs

eque

nthe

ad-s

tart

sha

veon

lybe

enke

ptin

capt

ivity

over

one

win

ter(

appr

ox.

9m

o).

**th

ere

wer

eno

dire

ct-r

elea

ses

in20

10;a

llw

ere

kept

forh

ead-

star

ting.

Yea

rC

o-ho

rt

#D

RD

RX̄

CL

mm±

1SD

(ran

ge)

DR

X̄g

(ran

ge)

NN

DR

rele

ase

date

(Wet

land

)

#H

SH

S-re

leas

ex̄

CL

mm±

1SD

(ran

ge)

HS-

rele

ase

x̄g

(ran

ge)

#ne

sts

HS

win

ter

Sx(S

urvi

val

inca

ptiv

ity)

HS

rele

ase

date

(Wet

land

)

2006

*–

––

––

750

.4(4

6.0-

54.7

)

21.4

(17.

6-26

.8)

410

0%4

Oct

2007

(TB

)

2007

2536

.7±

0.9

(35.

4-38

.3)

9.1

(6.9

-10.

4)6

4O

ct20

07(T

B)

2250

.1±

6.6

(38.

0-65

.0)

22.8

(10.

9-41

.0)

510

0%22

May

2008

(TB

)

2008

3436

.4±

1.7

(32.

8-38

.9)

8.7

(6.6

-10.

7)20

9Se

p20

08(T

B)

3165

.2±

13.7

(41.

0-99

.5)

52.2

(10.

5-16

3.8)

2091

%3

died

19M

ay20

10(P

S)

2009

2235

.5±

0.9

(34.

2-37

.5)

8.7

(7.3

-10.

4)10

13O

ct20

09(P

S)47

68.9±

8.8

(50.

0-86

.0)

57.7

(19.

7-11

5.5)

1594

%3

died

19M

ay20

10(P

S)

2010

**–

––

––

5462

.5±

11.6

(40.

0-83

.0)

43.7

(11.

2-91

.1)

1595

%3

died

25M

ay20

11(P

S)

TOTA

L81

––

36–

161

––

59–

–

447


Figure 7. Head-started hatchling (#3253, left) Blanding’s Turtle (Emydoidea blandingii) from the 2008cohort, recaptured in October 2009, compared to a 4-yr old wild-caught turtle (right) from the donor site.(Photographed by Kurt Buhlmann).

reported (1400 m, Rowe and Moll 1991; 2900m, Piepgras and Lang 2000). It is inevitable thatsome of the reintroduced Blanding’s Turtles willmigrate off site and into unfavorable situations,but if an 880 ha refuge is not able to sustain aBlanding’s Turtle population, then the species isunlikely to persist anywhere in New England.

Evaluating alternate release strategies.—Weconstructed a simple population model based onavailable life-history and behavioral data to evalu-ate five logistically feasible release strategies. Webased implementation decisions on published lit-erature and our model results. We made an earlydecision not to use adult Blanding’s Turtles be-cause survivorship of adult turtles in the donorpopulation was of paramount conservation con-cern; removal of adults would be equivalent tolowering the annual survivorship there. In addi-tion, observations of homing and site fidelity intranslocated adults of other species (Tuberville etal. 2005) suggest that adults are more likely than

juveniles to disperse from reintroduction sites.Protecting nests and using half of each female’sclutch for the reintroduction and returning theother portion to the donor site was predicted tomaximize genetic diversity at the recipient siteyet also minimize loss of genetic diversity andenhance recruitment at the donor site. Similarly,the effective 50% nest survivorship resulting fromnest protection and the release of half clutchesdirectly to the donor wetland (eliminating nest-to-water journey) provided greater survivorshipfor the residents during their first year than typi-cally reported for un-manipulated clutches of thisspecies.

Thus, our reintroduction strategy included di-rect release of hatchlings in autumn, head-startingof hatchlings for 9 mo over-winter with releasein May, and translocation and radio-telemetry ofsub-adults of known age 4–11 y. We did not con-sider sub-adult translocation as a long-term man-agement strategy but rather as a means to obtaininformation about habitat preferences of juvenile

448


Blanding’s Turtles in the refuge. We used infor-mation from radio-tracking the translocated sub-adults and the larger head-starts to guide selec-tion of subsequent release sites for direct-releasehatchlings and head-started turtles.

Our population model assumed that head-started turtles would behave similarly and exhibitcomparable survivorship to wild-recruitedjuveniles of similar size. Based on this assump-tion and the predicted higher survivorship ofhead-started turtles relative to smaller direct-release hatchlings, our models suggested thathead-starting should be twice as effective inreaching the target population size as directreleasing hatchlings. Releasing either 100direct-release hatchlings per year for 10 y or50 head-start turtles per year for 10 y wereboth predicted to result in a population ofapproximately 30 adult turtles in 20 y. However,because head-starting is much more time andresource demanding, our initial approach hasbeen a combination of head-start and directrelease. However, all partners are now in placefor nest protection at the donor site and forstandardized head-starting of the hatchlings. Wewill continue to direct-release and head-start andto hopefully increase effort as personnel andfunding allow.

Head-starting techniques and survivorship incaptivity.—Overall size of head-started hatch-lings by time of release has increased with eachyear of the project (2007:1.39-fold; 2008: 1.78-fold; 2009:1.83-fold), with slight decrease in2010 (1.75-fold) presumably as a result of stan-dardization and improvement in husbandry tech-niques. Survivorship during captivity has rangedfrom 91–100% among cohorts. In comparison,Mitrus (2005) reported 85% survival during cap-tivity of his head-started European Pond Turtles(Emys orbicularis), a closely related species. Thefew mortalities to date in our project have re-sulted from accidents (i.e., drowning under bask-ing logs); our husbandry techniques have beenaltered to prevent such accidents. Our collabora-

tion with a dedicated local high school teacherand students to carefully maintain holding facil-ities, take weekly growth and weight measure-ments, and provide consistent monitoring andfeeding has been the key to a successful head-starting program. In addition, segregating head-started turtles by size helped ensure that smallerturtles were not harassed by larger turtles and didnot lose out in competition for food.

The turtle “jello” diet we used since 2009provides a diverse, nutritious diet and most hatch-lings fed on it readily. Use of live foods, such asmeal worms, was successful in stimulating thefew reluctant hatchlings to feed. Introduction oflive food items to all head-starts prior to releasemay help condition them to search for prey inthe wild. None of the head-started hatchlingsexhibited knobby shells or “pyramiding” ashas been described in a captive program whereturtles (i.e., desert tortoise, Gopherus agassizii)maintained their activity year-round and were fedfrequently (Jackson et al 1976). Our intentionwas to produce normal-looking turtles thatwere too big for American Bullfrogs (Ranacatesbeiana) and other presumed predators(Haskell et al. 1996). At the time of release, thehead-started hatchlings were in the size rangeof wild 2–4 y old turtles and presumably wouldexperience comparable survivorship.

Post-release survival of direct-release andhead-started hatchlings.—We have limited dataon survivorship of the direct-release hatchlings.The recapture of a 2008 direct-release hatchling(88 mm CL) in summer 2010 is worth note. First,the hatchling was slightly larger than the largesthead-started turtle from the 2009 cohort, butsmaller than the largest head-started turtle (99.5mm CL) from the 2008 cohort at time of release.Second, this hatchling had moved to the loca-tion where radio-tracked, translocated sub-adultshave also congregated, which requires overlandmovement across a dirt road. Third, we caughtthis hatchling where translocated sub-adults hadpreviously successfully spent the winter in torpor.

449


We have radio-tracked six head-started hatch-lings. One (#3253) died after 11 mo; it had lost12% body weight when recaptured in Septem-ber 2009 prior to winter hibernation and wasfound dead in Spring 2010. Its cause of deathis unknown. We released five head-started tur-tles from the 2009 cohort with radio-transmitters,which are being tracked. Mitrus (2005) reporteda first-year recapture rate of 0.24 and a second-year recapture rate of 0.43 for head-started Eu-ropean Pond Turtles (Emys orbicularis). VanderHaegen et al. (2009) reported a survival rate offirst-year head-started Western Pond Turtles thatranged 83–91%. These observed survival ratesare higher than the rate we used in our models,thus our models are conservative. Congdon andvan Loben Sels (1993) estimated 72% survivor-ship for juvenile (> 1 y age) Blanding’s Turtles.We have limited data to estimate survival of ourhead-starts after release, but we will continue tomonitor head-starts with a combination of radio-telemetry and mark-recapture to evaluate theirsurvivorship, habitat selection, site fidelity, andgrowth.

Our reintroduction project includes monitoringof direct-release hatchlings and head-starts at therecipient site, as well as monitoring and nest pro-tection at the donor site. Thus, we are working toenhance the population at the donor site, as wellas establish a new population at the recipientsite. This experimental design will allow us toevaluate the performance of head-started turtleswhen compared to direct-release hatchlings atthe same site (i.e., recipient site). Monitoring ofdirect-release hatchlings via mark-recapture atthe donor site will provide a control to comparesurvivorship of direct-release hatchlings at therecipient site.

Adaptive management and other manage-ment considerations.—Our ultimate goal is toestablish a viable, self-sustaining population ofBlanding’s Turtles at the recipient site. We willevaluate a suite of interim benchmark goals, in-cluding survivorship, site fidelity, and eventually

reproduction. To further increase the likelihoodof success, our reintroduction strategy was de-signed to accommodate adaptive management.For example, monitoring the head-started anddirect-released hatchlings will allow us to de-velop site-specific survivorship estimates and toevaluate our assumption that head-started tur-tles do not behave abnormally after extendedcaptivity. Such data can be used to update ourmodel and adapt our release strategies accord-ingly. Likewise, radio-tracking of translocatedsub-adult and head-started turtles in the first fewyears of the reintroduction project has alreadyled to the identification of future release sites forhatchlings.

Lastly, although monitoring of both the donorsite and recipient site populations are valuablecomponents of this reintroduction project, habi-tat management and communication with stake-holders should be continued and expanded. Thenew Blanding’s Turtle population will need nest-ing sites, and such habitat requirements have al-ready been addressed early in this reintroduc-tion project. Creation of new nesting sites hasbeen shown to successfully attract female turtles(Buhlmann and Osborn 2011). Our project repre-sents an important partnership among federal andstate agencies, private consultants, a university, alocal high school, Friends of the Assabet NationalWildlife Refuge, and numerous volunteers. Di-verse stakeholder involvement, presentations tothe public, and newspaper articles have garneredsupport for the reintroduction project, helping as-sure the project’s long-term viability. Sustainedproject effort and long-term monitoring will benecessary to determine whether this venture willultimately be successful. We suggest that our ap-proach can serve as a model for others who wishto consider reintroduction and head-starting asparts of their conservation strategy.

Acknowledgements.—We especially appreciatethe field work by Steve Ecrement, Beth Schlimm,Jason St. Sauver and the dedicated care of head-started hatchlings by J. St. Sauver, Eileen Mc-

450


Courty, and the students of Bristol County Agri-cultural High School. Andrew Grosse helpedwith the statistical analysis and production ofFigure 6, and Brett DeGregorio provided helpfulcomments on earlier drafts of this manuscript.Manuscript preparation by KAB and TDT waspartially supported by the Department of EnergyAward Number DE-FC09-07SR22506. Fundingand in-kind support was provided by NationalFish and Wildlife Foundation (Project 2009-0017-021), Oxbow Associates, Inc., The Friends ofAssabet River National Wildlife Refuge, theU.S. Fish and Wildlife Service, and SavannahRiver Ecology Laboratory. Research was con-ducted under permits 2011 – 064.11SCRA, 2010– 106.10SCRA, 2009 – 149.09SCRA, 2008 –163.08SCRA issued by Massachusetts Divisionof Fisheries and Wildlife. We greatly appreciatethe reviews of this manuscript by Russ Burke.

Literature Cited

Beaudry, F., P.G. deMaynadier, and M.L. Hunter,Jr. 2009. Seasonally dynamic habitat use bySpotted Turtles (Clemmys guttata) and Bland-ing’s Turtles (Emydoidea blandingii) in Maine.Journal of Herpetology 43:636–645.

Breck, B.J., and J.J. Moriarty. 1989. Emydoideablandingii (Blanding’s Turtle). Longevity. Her-petological Review 20:53.

Buhlmann, K.A., T.S.B. Akre, J.B. Iverson, D.Karapatakis, R.A. Mittermeier, A. Georges,A.G.J. Rhodin, P.P. vanDijk, and J.W. Gibbons.2009. A global analysis of tortoise and fresh-water turtle distributions with identification ofpriority conservation areas. Chelonian Conser-vation and Biology 8:116–149.

Buhlmann, K.A., and C.P. Osborn. 2011. Use ofan artificial nesting mound by Wood Turtles(Glyptemys insculpta): a tool for turtle conser-vation. Northeastern Naturalist 18:315–334.

Butler, B.O. and T.E. Graham. 1995. Early

post-emergent behavior and habitat selectionin hatchling Blanding’s Turtles Emydoideablandingi, in Massachusetts. Chelonian Con-servation and Biology 1:187–196.

Compton, B.W. 2007. Status Assessment forthe Blanding’s Turtle (Emydoidea blandingii)in the Northeast. Department of Natural Re-sources Conservation, University of Mas-sachusetts, Amherst, Massachusetts, USA.

Congdon, J.D., A.E. Dunham, and R.C. vanLoben Sels. 1993. Delayed sexual maturityand demographics of Blanding’s Turtles (Emy-doidea blandingii): implications for conserva-tion and management of long-lived organisms.Conservation Biology 7:826–833.

Congdon, J.D. and D.A. Keinath. 2006. Bland-ing’s Turtle (Emydoidea blandingii): A Tech-nical Conservation Assessment. USDA ForestService, Rocky Mountain Region, Species Con-servation Project. 53 p.

Congdon, J.D., and R.C. van Loben Sels. 1993.Relationships of reproductive traits and body-size with attainment of sexual maturity and agein Blanding’s Turtles (Emydoidea blandingii).Journal of Evolutionary Biology 6:547–557.

Frazer, N.B. 1992. Sea turtle conservation andhalfway technology. Conservation Biology6:17–184.

Griffith, B., J.M. Scott, J.W. Carpenter, andC. Reed. 1989. Translocation as a speciesconservation tool: status and strategy. Science245:477–480.

Grgurovic, M., and P.R. Sievert. 2005. Move-ment patterns of Blanding’s Turtles (Emy-doidea blandingii) in the suburban landscapeof eastern Massachusetts. Urban Ecosystems8:203–213.

Haskell, A., T.E. Graham, C.R. Griffin, and J.B.Hestbeck. 1996. Size related survival of head-

451


started Redbelly Turtles (Pseudemys rubriven-tris) in Massachusetts. Journal of Herpetology30:524–527.

Heppell, S.S., L.B. Crowder, and D.T. Crouse.1996. Models to evaluate headstarting as a man-agement tool for long-lived turtles. EcologicalApplications 6:556–565.

Jackson, C.G., Jr., J.A. Trotter, T.H. Trotter, andM.W. Trotter. 1976. Accelerated growth rateand early maturity in Gopherus agassizii. Her-petologica 32: 139-145.

Lacy, R. C., M. Borbat, and J.P. Pollack. 2005.VORTEX: A stochastic simulation of the ex-tinction process. Version 9.50. Chicago Zoo-logical Society, Brookfield, Illinois, USA.

Levell, J.P. 2000. Commercial exploitation ofBlanding’s Turtle, Emydoidea blandingii, andthe Wood Turtle, Clemmys insculpta, for thelive animal trade. Chelonian Conservation andBiology 3:665–674.

Miller, P.S., and R.C. Lacy. 2005. VORTEX:A stochastic simulation of the extinction pro-cess. Version 9.50 user’s manual. ConservationBreeding Specialist Group (SSC/IUCN), AppleValley, Minnesota, USA.

Mitrus, S. 2005. Headstarting in European PondTurtles (Emys orbicularis): does it work?Amphibia-Reptilia 26:333–341.

Piepgras, S.A., and J.W. Lang. 2000. Spatialecology of Blanding’s Turtle in central Min-nesota. Chelonian Conservation and Biology3:589–601.

Rowe, J.W., and E.O. Moll. 1991. A radio-telemetric study of activity and movementpatterns of the Blanding’s Turtle (Emydoideablandingii) in northeastern Illinois. Journal ofHerpetology 25:178–185.

Seigel, R.A., and C.K. Dodd, Jr. 2000. Ma-nipulation of turtle populations for conser-vation: half-way technologies or viable op-tions? Pp. 218–238 In Turtle Conservation.Klemens, M.W. (Ed.). Smithsonian InstitutionPress, Washington, D.C., USA.

Storer, D.H. 1839. Report on the Fishes, Reptiles,and Birds of Massachusetts. Dutton and Went-worth State Printers, Boston, Massachusetts,USA.

Tuberville, T.D., E.E. Clark, K.A. Buhlmann, andJ.W. Gibbons. 2005. Translocation as a conser-vation tool: site fidelity and movement of repa-triated Gopher Tortoises (Gopherus polyphe-mus). Animal Conservation 8:349–358.

Vander Haegen, W.M., S.L. Clark, K.M. Perillo,D.P. Anderson, and H.L. Allen. 2009. Survivaland causes of mortality of head-started West-ern Pond Turtles on Pierce National WildlifeRefuge, Washington. Journal of Wildlife Man-agement 73:1402–1406.

Kurt A. Buhlmann is a Conservation Ecologistwhose research interests include life history and evo-lutionary ecology with application for conservationand management of amphibians and reptiles. He hasworked with non-profit, state, and federal agencies onhabitat management projects, including prescribedfire and wetlands restoration. He is involved inreintroduction projects for tortoises and freshwaterturtles. He holds a B.S. in Environmental Studiesfrom Stockton State College (New Jersey), a M.S. inWildlife Sciences from Virginia Tech, and a Ph.D.

452


in Ecology from the University of Georgia. Heis a Senior Research Associate at the Universityof Georgia’s Savannah River Ecology Laboratory.(Photographed by George Cevera).

Stephanie L. Koch is a Wildlife Biologist withthe U.S. Fish and Wildlife Service at EasternMassachusetts National Wildlife Refuge Complex,where she oversees the biological programs on eightdifferent National Wildlife Refuges. She earned herPh.D. in Environmental Science at the University ofRhode Island (2010) and her B.S. in Wildlife Biologyat the University of Massachusetts (1996). Her inter-ests include research and management of freshwaterand coastal habitats with an emphasis on minimizingimpacts of habitat loss and disturbance to benefitnative wildlife species in decline. (Photographed byKurt A. Buhlmann).

Brian O. Butler is President of Oxbow Associates,Inc., an environmental consulting company locatedin Massachusetts that specializes in wetlands and

endangered species permitting and field studies. Hereceived a B.S. in Marine Biology from SouthamptonCollege of Long Island University and a M.S. fromWorcester State University. (Photographed by Kurt A.Buhlmann).

Tracey D. Tuberville is an Associate ResearchScientist at the University of Georgia’s SavannahRiver Ecology Laboratory, near Aiken, SouthCarolina. Tracey received her Ph.D. in Ecology(2008) and her M.S. in Conservation Ecology andSustainable Development (1998) from Universityof Georgia and her B.S. in Biology from FurmanUniversity (1993). Tracey’s research interestsare in applied conservation and managementresearch for reptiles and amphibians, includingtranslocation and reintroduction as conservationtools, use of microsatellite markers to understandpopulation ecology and individual behavior, andecotoxicology. (Photographed by Kurt A. Buhlmann).

Veronica J. Palermo worked as a Field Techni-cian at Oxbow Associates, Inc., a Massachusetts

453


based wetlands and wildlife consulting companyspecializing in rare amphibian and reptile study andmitigation. She earned her B.S. in biology from theUniversity of Massachusetts Lowell. (Photographedby Anonymous).

Brian A. Bastarache is the Natural ResourcesManagement Division Head at the Bristol CountyAgricultural School in Dighton, Massachusetts, USA.He earned an M.A. in Environmental Managementfrom Harvard University and a B.S. in Wildlife Biol-ogy from the University of Massachusetts at Amherst.In addition to teaching courses in wildlife biology,fisheries and outdoor skills, he oversees severalcooperative conservation projects in partnership withuniversities, corporations, government agencies andNGOs. (Photographed by Mike Gay.)

Zachary A. Cava is a Biological Science Techni-cian with the Henderson, NV branch of the U.S.Geological Survey (USGS), where he is assistingin a Desert Tortoise translocation project. Zach isespecially interested in herpetological conservationand in recent years has worked with DiamondbackTerrapins, western rattlesnake species, and Blanding’sTurtles, among other species. He received a B.A. inBiology from Ithaca College in 2009. (Photographedby Kristin Godfrey.)

Errata. Zachary A. Cava’s name was incor-rectly spelled with a middle initial of D instead of Ain both the author line and bio sketch of the originalversion of this manuscript. On 18 July 2015, thecorrection was made and corrected manuscript postedonline. The co-author’ name Veronica J. Palermo wasspelled wrong.

454

reintroduction and head starting: tools for landing s t ...buhlmann et al.—2010 head-starting...

Documents