Unexpected Shallow Genetic Divergence in Turks Island Boas (Epicrates C. Chrysogaster) Reveals Single Evolutionarily Significant Unit for Conservation Author(S) :R

Unexpected Shallow Genetic Divergence in Turks Island Boas (Epicrates c. chrysogaster) Reveals Single Evolutionarily Significant Unit for Conservation Author(s) :R. Graham Reynolds, Glenn P. Gerber, and Benjamin M. Fitzpatrick Source: Herpetologica, 67(4):477-486. 2011. Published By: The Herpetologists' League DOI: URL: http://www.bioone.org/doi/full/10.1655/HERPETOLOGICA- D-11-00014.1

BioOne (www.bioone.org) is a a nonprofit, online aggregation of core research in the biological, ecological, and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published by nonprofit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/ terms_of_use. Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research. Herpetologica, 67(4), 2011, 477–486 E 2011 by The Herpetologists’ League, Inc.

UNEXPECTED SHALLOW GENETIC DIVERGENCE IN TURKS ISLAND BOAS (EPICRATES C. CHRYSOGASTER) REVEALS SINGLE EVOLUTIONARILY SIGNIFICANT UNIT FOR CONSERVATION

1,3 2 1 R. GRAHAM REYNOLDS ,GLENN P. GERBER , AND BENJAMIN M. FITZPATRICK 1Department of Ecology and Evolutionary Biology, University of Tennessee, Knoxville, TN 37996, USA 2Institute for Conservation Research, San Diego Zoo Global, Escondido, CA 92027, USA

ABSTRACT: The Turks Island Boa (Epicrates c. chrysogaster) is endemic to the Turks and Caicos Islands and is currently known from only 11 islands. The subspecies has likely been extirpated from several islands in its historic range, and all remaining populations are threatened with extirpation owing to habitat loss, introduced feral predators, malicious killing, and vehicle strikes. To assist conservation efforts, we undertook a genetic analysis of 53 individual E. c. chrysogaster, representing five island populations, with the goal of identifying existing population structure and genetic diversity. For each snake sampled, we sequenced one mitochondrial and two nuclear genes, resulting in 1591 bp of sequence, and screened nine microsatellite loci. All individuals were found to be monomorphic at the four microsatellite loci that amplified, and only three individuals were found to vary (by a single nucleotide polymorphism) in either nuclear gene. Nine mitochondrial haplotypes were found, with a maximum sequence divergence of ,1%. Taken together, these data indicate shallow genetic divergence in this subspecies, possibly owing to a lack of historical population structure and small population size when the Turks and Caicos Banks were each single islands during the last glacial maximum. Epicrates c. chrysogaster appears to represent a single evolutionarily significant unit, a significant finding suggesting that conservation strategies focusing on ecologically intact populations might be more appropriate than alternate strategies involving near-impossible reversal of declining populations on heavily disturbed islands. In addition, reintroduction programs would likely not disrupt any significant historical population structure. Key words: Boinae; Caribbean; Epicrates chrysogaster; Population structure; Snake; Turks and Caicos Islands; West Indies

KNOWLEDGE of standing genetic variation andRedig,2001),GrayWolves(Canis lupus within and between populations of species is [Fritts et al., 1997]), Bald Eagles (Haliaeetus important in designing conservation strategies lecuocephalus [Nye, 1988]), Seychelles Kestrels (Allendorf and Luikart, 2007; Caballero et al., (Falco areaus [Watson, 1989]), and Turks and 2010; Frankham, 2006; Frankham et al., Caicos Rock Iguanas (Cyclura c. carinata 2002). When populations are separated from [Gerber, 2007]). However, many caution against each other, as in island archipelagoes, pro- translocations that might disrupt natural genetic cesses such as mutation, genetic drift, gene structure (Allendorf and Luikart, 2007; Avise, flow, local extinction, and recolonization 2004; Storfer, 1999). might dramatically influence the apportioning The West Indies is one of the world’s of genetic variation within geographically biodiversity hotspots (Myers et al., 2000; structured populations (Slatkin, 1987; Wade Smith et al., 2005). The region has a diverse and McCauley, 1988; Whitlock and McCauley, herpetofaunal assemblage and these species 1990; Wright, 1977). Thus, understanding the are of particular conservation concern owing to distribution of genetic variation in an archipe- their vulnerability to anthropogenic threats, lagic species can shed light on evolutionary especially habitat destruction and the intro- processes operating in that system and might be duction of invasive mammalian predators such used to develop conservation management as House Cats (Felis catus), Small Indian strategies that minimize disruption of those Mongoose (Herpestes javanicus auropuncta- processes. In particular, translocation has been tus), and Black Rats (Rattus rattus), among an important tool in the recovery of high-profile others (Corke, 1992; Hailey et al., 2011; species such as Peregrine Falcons (Falco Iverson, 1978; Smith et al., 2005; Tolson and peregrinus [Cade and Burnham, 2003]; Tordoff Henderson, 2006). The Turks and Caicos Islands (TCI), located at the southern terminus 3 CORRESPONDENCE: e-mail, [email protected] of the Bahama Archipelago, contain a unique

477 478 HERPETOLOGICA [Vol. 67, No. 4

FIG. 1.—Map of the Turks and Caicos Islands illustrating sampling locations (December 2007–March 2010) of Epicrates c. chrysogaster. Inset map shows the location of the Turks and Caicos Islands within the greater Caribbean region. Outlines indicate the approximate extent of the Caicos and Turks banks. Haplotype map is the TCS statistical parsimony network overlaid onto a map of the Turks and Caicos Islands. Circles represent individual haplotypes, the size of which corresponds to the number of sequences each haplotype contains. Circles are divided and shaded to show the proportion of each island population within each haplotype, and each island is assigned a unique shading pattern. Small black circles indicate unobserved haplotypes. Haplotypes are connected with a 95% CL. endemic herpetofauna (Reynolds, 2011; Rey- islands, including Grand Turk and South nolds and Niemiller, 2010a) that is threatened Caicos and populations on major islands are by increasing development and population likely in a sharp decline (Reynolds, 2011). growth. The TCI consist of more than 200 Turks Island Boas face numerous threats, islands, ranging in size from ,1hato including development and associated habitat .12,000 ha, distributed on two shallow banks loss, introduced predators, direct persecution, (Fig. 1). The Turks and Caicos Banks are 150- and vehicular traffic, and very little is published million-yr-old sediment and limestone plat- about the ecology, distribution, and natural forms that have remained near the ocean history of this species (Reynolds, 2011; Rey- surface due to reef building. Various portions nolds and Niemiller, 2010a). of the banks have been alternately emergent Other reptiles in the region, such as the and inundated throughout this time, though Turks and Caicos Rock Iguana, are facing both banks were fully emergent within the last similar threats and extirpations (Gerber and 8000–15,000 yr. The two banks are separated Iverson, 2000; Iverson, 1978; Reynolds, 2011). by the narrow and deep Turks Island Passage As a result, they are being intensively managed and have never been joined, and rising sea and translocated to some of the outer islands levels since the Wisconsin glacial maximum that are free from introduced predators and partitioned each bank into the present islands development (Gerber, 2007). To help develop (Fairbanks, 1989; Lighty et al., 1982; Welch conservation strategies for Epicrates, we sam- et al., 2004). The Turks Island Boa, Epicrates pled boas from islands on the Turks Bank and c. chrysogaster, is an endemic nocturnal– the Caicos Bank to assess standing genetic crepuscular boine and the largest native diversity among populations and to investigate terrestrial carnivore in the TCI. Boas have whether separate evolutionarily significant been reported from 11 islands in the TCI, units (ESUs; Crandall et al., 2000; Mortiz, while extirpations have occurred on several 1994; Ryder, 1986; Waples, 1991), or populations December 2011] HERPETOLOGICA 479

TABLE 1.—Sampling locations in the Turks and Caicos Islands (December 2007–March 2010) and sampling effort used to obtain tissue samples of Epicrates c. chrysogaster used in this analysis. Effort is represented as person-hours (p/h), the amount of time spent searching by all individuals that day. Two additional islands (Long Cay and Little Ambergris Cay) where Epicrates is thought to occur (Reynolds, 2011) were also searched but yielded no specimens. Note that four individuals sampled from Providenciales and five individuals sampled from North Caicos were wild-caught captives being held for educational purposes, with known locality information.

Island Date(s) Effort Captures Big Ambergris Cay December 2007 ,54 p/h 34 Gibbs Cay August 2008 18 p/h 3 North Caicos August 2008, October 2009 194 p/h 10 Middle Caicos August 2008, October 2009 78 p/h 1 Providenciales July 2008, October 2009, March 2010 28 p/h 5 Long Cay (Caicos) July 2008 5 p/h 0 Little Ambergris Cay December 2007, March 2009 9 p/h 0 Totals 27 sampling d 386 p/h 53 genetically differentiated enough to be managed and after clipping, and antiseptic dermal separately. In particular, we were interested in adhesive was applied to prevent infection. the genetic uniqueness of individual populations Whole genomic DNA was extracted using the and whether significant population structure Wizard SV DNA purification system (Promega, existed on the Caicos Bank or between the Turks Madison, Wisconsin) according to the manufac- and Caicos Banks. turer’s protocol and stored at 220uC. DNA Sequencing MATERIALS AND METHODS The polymerase chain reaction (PCR) was Sample Collection and DNA Extraction used to amplify a portion of the mitochondrial We visited 7 of the 11 islands from which genome (617 base-pairs [bp] of cytochrome B; Epicrates c. chrysogaster is currently known CytB), as well as two nuclear intron loci: 518 bp (Reynolds, 2011) December 2007–March 2010. of neutrophin-3 (NT3) and 456 bp of oocyte Sampling involved diurnal and nocturnal visual- maturation factor (c-mos). The CytB fragment encounter surveys conducted between 0700 and was amplified using primers CYTBECF (59- 0230h. Diurnal surveys included turning over GCCCCACCAACAAATCCTA-39)andCY- and replacing cover objects such as rocks, logs, TBECR (59-GGTCTGGTATGGGTGGAATG- and palm fronds; while nocturnal surveys were 39) developed from a congener sequence on conducted by walking in suitable habitat using Genbank (Campbell, 1997; U69803). The NT3 battery-powered headlamps and handheld fragment was amplified using primers NT3- flashlights. We found boas on five of the seven F3 (59-ATATTTCTGGCTTTTCTCTGTGGC- islands visited, though more focused sampling 39)andNT3-R4(59-GCGTTTCATAAAAA- occurred on some islands owing to logistics TATTGTTTGACCGG-39) from Noonan and (Table 1). Boas proved extremely difficult to Chippindale (2006a,b). The c-mos fragment was find on Providenciales and Middle Caicos, and amplified using primers CMOS-Fsnk (59- logistics prevented extensive sampling of East GCTGTAAAACAGGTGAAGAGATGCAG-39) Caicos, Gibbs Cay, and Little Ambergris Cay. and CMOS-Rsnk (59-AGCACGATGGGTG- Tissue samples were collected from wild E. TATGTTCCCCC-39) from Noonan and Chip- chrysogaster from five different island popula- pindale (2006a,b). Polymerase chain reaction tions (Table 1; Fig. 1): Big Ambergris Cay (n 5 reactions were done in an Eppendorf Master- 34), Gibbs Cay (n 5 3), North Caicos (n 5 10), cycler with the following conditions: denatur- Middle Caicos (n 5 1), and Providenciales (n 5 ation at 94uC for 5 min; followed by 35 cycles at 3, plus 2 offspring of one sampled adult). 94uC for 30 s, 52uC for 30 s (CytB and NT3) or Samples were preserved in 95% ethanol and 61uCfor30s(c-mos), 72uC for 1 min, and final stored at 220uC. Tissue was obtained by either extension at 72uC for 5 min. clipping the distal 3–4 mm of the tail or clipping Polymerase chain reaction products were 3–4 ventral scales. Tails were sanitized before visualized by gel electrophoresis and purified 480 HERPETOLOGICA [Vol. 67, No. 4 using exonuclease I/shrimp alkaline phospha- including Tajima’s D and Fu’s Fs,were tase (ExoSAP). Sequencing reactions were conducted for islands with multiple haplotypes resolved on an automated sequencer (Applied using ARLEQUIN, though it is acknowledged Biosystems Inc.; ABI 377) at the University of that these tests might be biased for some taxa Tennessee, Knoxville (CytB and NT3) or at the (Wares, 2009). High Throughput Genomics Unit at the University of Washington, Seattle (c-mos). Phylogeographic Analyses Sequences were aligned in Sequencher 4.6 To estimate the level of genetic partitioning (Gene Code Corporation), and ambiguous base among islands and between banks, we calcu- calls were manually verified by examining lated W-statistics in an analysis of molecular electropherograms. Sequence alignment files variance (AMOVA) framework (Excoffier et were verified and trimmed in MacClade 4.07 al., 1992). Significance of WST values were (Maddison and Maddison, 2005) by compari- determined via 50,175 permutations. It should son to published sequences for Epicrates. be noted that valid statistical hypothesis Representative sequences were deposited in testing cannot be performed for WCT when GenBank (JF812171–181; JN052211–16). the number of populations is ,7, because permutation tests will never result in a P-value Microsatellite Genotyping of ,0.05 (Fitzpatrick, 2009). However, the Polymorphic nuclear markers have not statistical significance of hierarchical structure been developed for E. chrysogaster; hence, is not necessarily the most relevant for PCR was used to screen nine microsatellite conservation questions. For conservation pur- markers developed for the congener E. poses, the important question is whether subflavus (Tzika et al., 2008b). Each primer the population on Gibbs Cay (or any other set included a 59 ‘‘tail’’ (GTTTCTT) added to population) is genetically distinct enough to the forward primer to force ‘‘plus-A’’ addition warrant independent management as an ESU. to improve genotyping consistency (Brown- Therefore, we discuss our results with respect stein et al., 1996). Polymerase chain reaction to proposed ESU criteria. For our data, these reactions were done in an Eppendorf Mas- include reciprocal monophyly of mtDNA gene tercycler, PCR products were visualized by trees and significant differentiation of nuclear gel electrophoresis, and those with distinct allele frequencies (Moritz, 1994). bands were genotyped on an EGeneH multi- capillary electrophoresis system at the Uni- RESULTS versity of Tennessee, Knoxville. Peaks were We resolved 617 bp of mitochondrial CytB, cleaned and called using Biocalculator (Qia- 518 bp of nuclear NT3, and 456 bp of nuclear gen, Crawley, UK). One PCR product for c-mos for all 53 individuals. For CytB, there each primer set that was successfully geno- were 13 variable nucleotide positions (2.1% typed was sequenced to ensure that the repeat of total nucleotide positions), while for NT3 region was included in the fragment (Gen- and c-mos there was 1 variable site per gene bank JN052213–16). (0.19% and 0.21% of total nucleotide positions, respectively). Nine haplotypes were Haplotype Analyses observed for CytB and two haplotypes for To eliminate redundant haplotypes, we used each nuclear gene (CytB nos. 1–9, nuc genes the online toolbox FaBox (Villesen, 2007) to nos. 1, 2). Big Ambergris Cay and North collapse sequences into haplotype groups. A Caicos were the only islands with multiple statistical parsimony network was created using CytB haplotypes (four haplotypes represented default parameters (95% probability criterion) on each island), and one sequence each from in the program TCS 1.21 (Clement et al., 2000). Middle Caicos and North Caicos collapsed Genetic variation of populations was estimated into Haplotype 1 along with sequences from as nucleotide (p) and haplotype (h) diversity Big Ambergris Cay (Fig. 1). Gibbs Cay and using ARLEQUIN (version 3.1.1; Excoffier Providenciales both had unique haplotypes for et al., 2005; Schneider et al., 2000). Tests CytB (nos. 8 and 9, respectively). For the for ‘‘neutrality’’ (mutation–drift equilibrium), nuclear gene NT3, two haplotypes were December 2011] HERPETOLOGICA 481

TABLE 2.—Results of analysis of molecular variance (AMOVA) for various groupings of Epicrates c. chrysogaster for samples collected December 2007–March 2010. Samples are from Big Ambergris Cay, North Caicos, Middle Caicos, and Providenciales on the Caicos Bank, and from Gibbs Cay on the Turks Bank.

Source of variation df % variation W statistic

Analysis across both banks Between banks 1 58.1 WCT 5 0.58 Within banks 3 20.7 WSC 5 0.49* Within populations 47 21.2 Analysis across all populations Among populations 4 61.1 WST 5 0.61* Within populations 47 38.8

* Significant at P # 0.05. Note that this is impossible for WCT for fewer than seven populations in permutation tests. observed, one in all populations and the DISCUSSION second exclusively on North Caicos. Two The use of population genetics in initial haplotypes were observed for c-mos as well, conservation assessment and planning can be a with Haplotype 2 exclusive to Big Ambergris powerful tool for devising conservation strate- Cay and Haplotype 1 found in all other gies that minimize disturbance of evolutionary populations. trajectories for island populations (Frankham, Of the nine microsatellite markers screened, 2006). When defining conservation strategies only four successfully amplified using PCR. All four markers proved monomorphic for all for archipelagic endemics, it is important to populations. Sequenced PCR products indi- recognize that separate ESUs might exist, cated that the repeat motif was included in particularly on islands with deep-water isola- each fragment; hence, all populations appear tion and long separation (Frankham et al., monomorphic at these microsatellite loci. 2002; Whittaker and Ferna´ndez–Palacios, Statistical parsimony showed a maximum of 2007). As part of a larger new initiative to seven mutational steps separating haplotypes, promote snake conservation in the TCI, we allowing a 95% parsimonious connection have conducted focused studies of the biology, (Fig. 1). Pairwise genetic divergence among ecology, and distribution of the Turks Island islands ranged from 0.002 (North Caicos– Boa (E. c. chrysogaster) in the region (Rey- Middle Caicos) to 0.009 (Gibbs Cay–Big nolds, 2011). A major component of this Ambergris Cay). Haplotype diversity was initiative was to assess the level of standing calculated for the only two populations with genetic variation among islands where this more than a single haplotype: North Caicos species occurs. (0.4167) and Big Ambergris Cay (0.6291). Our surprising result is that Turks Island Nucleotide diversity was estimated as 0.002 Boas exhibit ,1% sequence divergence at the for Big Ambergris Cay and 0.0017 for North mitochondrial gene CytB, with haplotype Caicos. Tests for deviations from neutrality sharing between some islands. Private haplo- were nonsignificant for Big Ambergris Cay types were found on Big Ambergris Cay, (Tajima’s D 5 1.07; Fu’s Fs 5 1.50) and North Providenciales, North Caicos and Gibbs Cay, Caicos (Tajima’s D 5 21.15; Fu’s Fs 5 0.45). though these were minimally divergent from Analysis of molecular variance (AMOVA); other haplotypes (,1%). For nuclear genes, Table 2) revealed that an estimated 58% of only three individuals were found to be mtDNA variation is explained by grouping variable, with one unique haplotype on North between the Turks and Caicos Banks, and this Caicos for the marker NT3 and one unique configuration maximizes WCT (highest of the haplotype on Big Ambergris Cay for the five possible permutations of islands between marker c-mos. For both markers, Haplotype banks; therefore, P # 0.2). Variation among 1 was shared across all populations, and populations within the Caicos Bank accounted Haplotype 2 differed by a single nucleotide for 20.7% of the variation (WSC 5 0.49, P , polymorphism. Low polymorphism in nuclear 0.001). When no partition between banks is introns is not entirely unexpected in an made, among-population variance (61.1%, intraspecific phylogeographic study (Hare, WST 5 0.61, P , 0.001) and within-population 2001; Wang and Shaffer, 2008), though our variance (38.8%) are similar. results indicate that what little variation is 482 HERPETOLOGICA [Vol. 67, No. 4 present is not idiosyncratic of a specific increased rates of evolution at c-mos for population. Additionally, several populations snakes relative to the orthologous gene in in our study are represented by only a few chickens (Gallus spp.). The limited availability individuals and we have limited power to of markers forced us to select slower evolving detect shared haplotypes; hence, the few nuclear introns (Hare, 2001; Wang and private haplotypes we found may well be Shaffer, 2008); however, CytB has been shared among populations. shown to evolve in snakes at similar rates to Microsatellite loci developed for the con- ND4, another commonly used mtDNA mark- gener E. subflavus are subject to ascertain- er for intraspecific studies (Kelly et al., 2003), ment bias (Ellegren et al., 1995) in both E. and has been used in other intraspecific chrysogaster and other boid species, because studies of Epicrates (Campbell, 1997; Tzika several of these loci will either not amplify or et al., 2008a). Hence, recent panmixia seems a are monomorphic, indicating that polymor- more plausible explanation for low observed phism at these loci is mostly idiosyncratic of divergence in Turks Island Boas. Both the E. subflavus. Booth et al. (2011) screened Turks and Caicos Banks were likely emergent these same loci in Boa constrictor and as recently as 8000 yr ago (Fairbanks, 1989; obtained amplification at only three of the Lighty et al., 1982; Welch et al., 2004). These nine loci, while Booth et al. (personal super-islands would have had low topographic communication) screened these loci in Epi- relief (Reynolds, 2011; Sealey, 2006), with the crates maurus and found that only one only barriers to gene flow in boas likely being monomorphic marker amplified. It is impos- salinas or brackish wetland areas. The low sible to determine whether the lack of level of differentiation between the Turks and variation in E. chrysogaster reflects low Caicos Banks is slightly more problematic, effective population size, low mutation rates, because frequent dispersal across the Turks or strong purifying selection. Nevertheless, Islands Passage seems unlikely even when these loci represent four independent nuclear both banks were fully emergent. Thus, a markers, and were found to be monomorphic plausible explanation is that one of the banks in E. chrysogaster and polymorphic in the was recently colonized by individuals from the related E. subflavus and B. constrictor. other. This could either have been accom- The only other intraspecific phylogeographic plished naturally or could have been human- study of insular Epicrates (Tzika et al., 2008a) mediated, as is hypothesized to have occurred found significant divergence (0.084 , FST , in Rock Iguanas (deliberate introduction from 0.76) between populations of the Jamaican Boa the TCI to Booby Cay, Mayaguana; Bryan et (E. subflavus). Though these populations occur al., 2007). Humans have occupied the Turks on the same island, uplift of mountains and sea- and Caicos Archipelago for at least 1300 yr level variation likely contributed to reduced (Keegan, 1992; Marvel, 2008) and have used gene flow between populations, and boas have the local reptiles, particularly iguanas (IUCN likely occupied Jamaica far longer than the TCI Iguana Specialist Group, 2003) and turtles (Tolson, 1987; Tzika et al., 2008a). (Lee and Ross, 2001; Seidel, 1996), as a food It is unclear why such low genetic diver- source. Hence, it is possible that people gence is found in Turks Island Boas, while moved boas as well; remains of at least two much higher genetic and morphological di- individuals have been found in middens of a vergence occurs in other endemic squamates Native American archeological site on Grand with similar distributions in the archipelago Turk (Carlson, 1999; Newsom and Wing, (Reynolds, 2011). We hypothesize that low 2004; Reynolds, 2011). diversity could be due to the following: (1) low In addition to the lack of geographic rates of evolution in the markers examined, or structure, the low level of variation within (2) recent panmixia when banks were emer- populations implies that population sizes have gent. Low rates of molecular evolution are been low for many generations. In particular, known for some taxa such as turtles (King and lack of variation at four microsatellites that Julian, 2004); however, recent studies (e.g., were variable in the related Jamaican Boa Hughes and Mouchiroud, 2001) have shown (Tzika et al., 2008a) and distantly related Boa December 2011] HERPETOLOGICA 483 constrictor (Booth et al., 2011) is exceptional. controversy still exists regarding appropriate- This result might be a consequence of long- ness for many taxa (Germano and Bishop, term decline and extirpation, or it might be 2008). The federally endangered Virgin Is- that these snakes generally occurred in low lands Boa (Epicrates monensis granti) has densities, even under the most natural of been successfully reintroduced to at least two circumstances. islands (Stevens Cay and Cayo Ratones) on the Puerto Rican bank after a rat and Turks Bank Boas mongoose removal program, and these popu- The recent discovery of Turks Island Boas lations appear to be thriving (USFWS, 2009). on the Turks Bank (Reynolds, 2011; Reynolds Our results raise the intriguing possibility of and Niemiller, 2010b), where they were relocation of boas to other islands in the TCI, thought to have been completely extirpated, including movement of boas from Caicos raised suspicion that the Turks Bank might Bank populations to islands without boas on harbor a unique lineage of boas due to the the Turks Bank. Rock Iguanas (C. carinata) physical separation of deep water and strong have benefited greatly from translocations currents between the banks. Gibbs Cay is an from the few remaining strong source popu- extremely small (,0.07 km2) island ,1.5 km lations to outlying islands where no intro- from Grand Turk. There are no other known duced predators occur and no human devel- extant populations on the Turks Bank, and we opment exists (Gerber, 2007). Long Cay, did not sample boas from the nearest known located on the Caicos Bank south of South populations of Long Cay and East Caicos; Caicos, previously had populations of both therefore, it is difficult to determine whether iguanas and boas, though an introduction of the Turks Bank Boas harbor a truly unique cats caused the extirpation of the iguanas and mtDNA lineage. However, our observation of perhaps boas, the latter of which have not very limited divergence from the Caicos been recorded there in some time (Reynolds, haplotypes still stands as likely evidence of 2011). Long Cay was the focus of an intensive recent divergence. Although this population is cat eradication campaign, which eliminated unique in its habitat and location, manage- the invasive predators and allowed reintro- ment as a separate entity is probably not duction of iguanas (Mitchell et al., 2002). If justified given the shallow divergence of boas were also extirpated, Long Cay would be mtDNA and lack of divergence of other an ideal island to relocate boas from other markers. That is, we infer that Epicrates c. threatened populations. chrysogaster is a single ESU including a small Though additional data might yield further number of insular populations that appear to resolution of the minimal genetic structure be genetically exchangeable. and divergence in this species, our surprising The ESU concept is not perfect (e.g., Waples, results provide information valuable to con- 2006), however, and we cannot evaluate wheth- servation planning efforts for this species. Our er any Epicrates population is locally adapted to data set is important not for indicating a lack its island to the point of not being ‘‘ecologically of population structure, as in a Turks Bank/ exchangeable’’ with other populations (Crandall Caicos Bank split, but in determining that any et al., 2000). Nevertheless, the best available divergence is minimal and likely recent. evidence is that all studied populations of the Reasons for low genetic diversity in Turks Turks Island Boa are genetically similar parts of Island Boas are not clear, but the implications a single ‘‘evolutionary legacy’’ (Waples, 1991). for conservation are important. This species should likely be managed as a single unit, and Conservation Strategies the feasibility of future reintroductions should Our results suggest several conservation be strongly considered to buffer against the strategies for Turks Island Boas, including loss of important remaining populations to possible translocation and in situ conservation development and predators. of ecologically intact populations. Transloca- Acknowledgments.—We thank the following individuals tion of reptiles has seen a marked increase in for assistance in the field (alphabetical order): K. Bradley, success in the last 20 yr, though some J. Burgess, C. Deal, M. Emery, Z. Harlow, T. J. Jett, 484 HERPETOLOGICA [Vol. 67, No. 4

C. Knapp, B. N. Manco, L. Mensen, N. Missick, M. CARLSON, L. A. 1999. Aftermath of a feast: Human Niemiller, T. Wagner, G. Waters, and M. Wilson. Thanks colonization of the southern Bahama Archipelago and as well to Ð. Hadzˇiabdic´ and the Trigiano Lab at The its effects on the indigenous fauna. Ph.D. Dissertation, University of Tennessee for assistance with microsatellite University of Florida, Gainesville, Florida, USA. analysis. We thank the Turks and Caicos National Trust CLEMENT, M., D. POSADA, AND K. CRANDALL. 2000. TCS: for logistical support and Blue Horizon Resort for A computer program to estimate gene genealogies. property access. We are especially grateful to the Turks Molecular Ecology 9:1657–1660. and Caicos Sporting Club at Ambergris Cay for logistical CORKE, D. 1992. The status and conservation needs of the support, property access, and air transportation. Funding terrestrial herpetofauna of the Windward Islands (West for this project was provided by the San Diego Zoo Indies). Biological Conservation 62:47–58. Institute for Conservation Research and the Offield CRANDALL, K. A., O. R. P. BININDA-EDMONDS,G.M.MACE, Family Foundation (GPG) and to the senior author AND R. K. WAYNE. 2000. Considering evolutionary (R. G. R.) by the following: The University of Tennessee processes in conservation biology: An alternative to Department of Ecology and Evolutionary Biology Sum- ‘‘evolutionarily significant units.’’ Trends in Ecology and mer Research Grant Program, the Sigma Xi Grant-in-Aid Evolution 15:290–295. of Research, The American Philosophical Society Lewis ELLEGREN, H., C. R. PRIMMER, AND B. C. SHELDON. 1995. and Clark Fund for Exploration and Research, The Microsatellite evolution—Directionality or bias. Nature American Museum of Natural History Theodore Roose- Genetics 11:360–362. velt Memorial Fund Grant in Support of Research, The EXCOFFIER, L., P. SMOUSE, AND J. QUATTRO. 1992. Analysis Mohamed bin Zayed Species Conservation Fund, and The of molecular variance inferred from metric distances University of Tennessee W. K. McClure Scholarship for among DNA haplotypes: Application to human mito- the Study of World Affairs. Methods involving animals chondrial DNA restriction data. Genetics 131:479–491. used in this study have been approved by the University of EXCOFFIER, L., G. LAVAL, AND S. SCHNEIDER. 2005. Tennessee Institutional Animal Care and Use Committee Arlequin ver 3.0: An integrated software package for (IACUC), Protocol no. 1701–0408. This work was population genetics data analysis. Evolutionary Biology authorized by scientific research permits (nos. 1–4 R. G. Online 1:47–50. Reynolds) from the Department of Environment and FAIRBANKS, R. G. 1989. A 17,000 year glacio-eustatic sea Coastal Resources, Turks and Caicos Islands Government. level record: Influence of glacial melting rates on the We are grateful to D. Simberloff, G. Cabana, J. Fordyce, Younger Dryas event and deep-ocean circulation. B. Moon, and two anonymous reviewers for comments on Nature 342:637–642. this manuscript. FITZPATRICK, B. M. 2009. Power and sample size for nested analysis of molecular variance. Molecular Ecology 18:3961–3966. LITERATURE CITED FRANKHAM, R. 2006. Evolutionary conservation genetics. Pp. 502–512. In C. W. Fox and J. B. Wolf (Eds.), ALLENDORF, F. W., AND G. LUIKART. 2007. Conservation Evolutionary Genetics. Oxford University Press, New and the Genetics of Populations. Blackwell, Malden, York, New York, USA. Massachusetts, USA. FRANKHAM, R., J. D. BALLOU, AND D. A. BRISCOE. 2002. AVISE, J. C. 2004. Molecular Markers, Natural History, Introduction to Conservation Genetics. Cambridge and Evolution, 2nd edition. Sinauer Associates, Sunder- University Press, Cambridge, UK. land, Massachusetts, USA. FRITTS, S. H., E. E. BANGS,J.A.FONTAINE,M.R.JOHNSON, BOOTH, W., D. H. JOHNSON,S.MOORE,C.SCHAL, AND E. L. M. K. PHILLIPS,E.D.KOCH, AND J. R. GUNSON. 1997. VARGO. 2011. Evidence for viable, non-clonal but Planning and implementing a reintroduction of wolves fatherless boa constrictors. Biology Letters. DOI: to Yellowstone National Park and central Idaho. 10.1098/rsbl.2010.0793. Restoration Ecology 5:7–27. ROWNSTEIN ARPTEN AND MITH B , M. J., J. D. C , J. R. S . 1996. GERBER, G. P. 2007. Turks and Caicos iguana transloca- Modulation of non-templated nucleotide addition by tion program, Bahama Archipelago. Re-introduction Taq DNA polymerase: Primer modifications that News 26:53–55. facilitate genotyping. Biotechniques 20:1004–1010. GERBER, G. P., AND J. B. IVERSON. 2000. Turks and Caicos BRYAN, J. J., G. P. GERBER,M.E.WELCH, AND C. L. Iguana, Cyclura carinata carinata. Pp. 15–18. In A. C. STEPHEN. 2007. Re-evaluating the taxonomic status of Alberts (Ed.), West Indian Iguanas: Status Survey and the Booby Cay Iguana, Cyclura carinata bartschi. Conservation Action Plan. IUCN the World Conserva- Copeia 3:734–739. tion Union, Gland, Switzerland. ´ ´ CABALLERO, A., S. T. RODRIGUEZ-RAMILO,V.AvILA, AND J. GERMANO, J. M., AND P. J. BISHOP. 2008. Suitability of FERNA´ NDEZ. 2010. Management of genetic diversity of amphibians and reptiles for translocation. Conservation subdivided populations in conservation programmes. Biology 23:7–15. Conservation Genetics 11:409–419. HAILEY, A., B. S. WILSON, AND J. A. HORROCKS (Eds.). CADE, T. J., AND W. BURNHAM (EDS). 2003 Return of the 2011. Conservation of Caribbean Island Herpetofaunas, Peregrine: A North American Saga of Tenacity and Volume I. Brill, Leiden, The Netherlands. Teamwork. Peregrine Fund, Boise, Idaho, USA. HARE, M. P. 2001. Prospects for nuclear gene phyloge- CAMPBELL, B. N. 1997. Hic sunt serpentes—Molecular ography. Trends in Ecology and Evolution 16:700–706. phylogenetics and the Boidae (Serpentes: Booidea). HUGHES, S., AND D. MOUCHIROUD. 2001. High evolutionary Ph.D. Dissertation, Queen’s University, Kingston, rates in nuclear genes of squamates. Journal of Ontario, Canada. Molecular Evolution 53:70–76. December 2011] HERPETOLOGICA 485

IUCN IGUANA SPECIALIST GROUP. 2003. Turks and Caicos NOONAN, B. P., AND P. T. CHIPPINDALE. 2006b. Vicariant Iguana, Cyclura carinata carinata conservation and origin of Malagasy reptiles supports late Cretaceous management plan, 2005–2009. Pp. 1–32. In Workshop Antarctic land bridge. American Naturalist 168: Proceedings from the IUCN/SSC Iguana Specialist 730–741. Group, Providenciales, Turks and Caicos Islands. NYE, P. 1988. Bald eagle restoration in northeast United Available at: http://www.iguanafoundation.org/down- States, 1972–1987. Pp. 264–269. In B. G. Pendelton loads/pdf/TCI-CAMP-4July2007_Sml.pdf. (Ed.), Proceedings of the Northeast Raptor Manage- IVERSON, J. B. 1978. The impact of feral cats and dogs on ment Symposium and Workshop. National Wildlife populations of the West Indian rock iguana Cyclura Federal Scientific and Technical Series No. 13. Na- carinata. Biological Conservation 14:63–73. tional Wildlife Federation, Washington, DC, USA. KEEGAN, W. F. 1992. The People Who Discovered REYNOLDS, R. G. 2011. Status, conservation, and intro- Columbus. University Press of Florida, Gainesville, duction of amphibians and reptiles in the Turks and Florida, USA. Caicos Islands, British West Indies. Pp. 377–406. In A. KELLY, C. M. R., N. P. BARKER, AND M. H. VILLET. 2003. Hailey, B. S. Wilson, and J. A. Horrocks (Eds.), Phylogenetics of advanced snakes (Caenophidia) based Conservation of Caribbean Island Herpetofaunas, on four mitochondrial genes. Systematic Biology 52: Volume II. Brill, Leiden, The Netherlands. 439–459. REYNOLDS, R. G., AND M. L. NIEMILLER. 2010a. Island KING, T. L., AND J. E. JULIAN. 2004. Conservation of invaders: Introduced amphibians and reptiles in the microsatellite DNA flanking sequence across 13 Emy- Turks and Caicos Islands. Reptiles and Amphibians did genera assayed with novel bog turtle (Glyptemys 17:117–121. muhlenbergii) loci. Conservation Genetics 5:719–725. REYNOLDS, R. G., AND M. L. NIEMILLER. 2010b. Epicrates LEE, D. S., AND J. P. ROSS. 2001. The Cat Island turtle, a chrysogaster (Southern Bahamas Boa), distribution. reptile of problematic origin, including a bibliographic Caribbean Herpetology 1:14. review of the genus Trachemys in the West Indian RYDER, O. A. 1986. Species conservation and systematics: region. Pp. 36–47. In C. Clark-Simpson and G. W. The dilemma of subspecies. Trends in Ecology and Smith (Eds.), Proceedings of the 8th Symposium on the Evolution 1:9–10. Natural History of the Bahamas. Gerace Research SCHNEIDER, S., D. ROESSLI, AND L. EXCOFFIER. 2000. Center, San Salvador, Bahamas. Arlequin: A software for population genetic data. LIGHTY, R. G., J. G. MACINTYRE, AND R. STUCKENRATH. Genetics and Biometry Laboratory, University of 1982. Acropora palmate reef framework: A reliable Geneva, Switzerland. indicator of sea level in the western Atlantic for the past SEALEY, N. E. 2006. Bahamian landscapes, 3rd edition. 10,000 years. Coral Reefs I, 25–30. Macmillan-Caribbean, Oxford, UK. MADDISON, D. R., AND W. P. MADDISON. 2005. MacClade 4 SEIDEL, M. E. 1996. Current status of biogeography of release version 4.07 for OSX. Sinauer, Sunderland, the West Indian turtles in the genus Trachemys Massachusetts, USA. (Emydidae). Pp. 169–174. In R. Powell and R. W. MARVEL, J. 2008. Our first colonists: The pre-Columbian Henderson (Eds.), Contributions to West Indian people of the Turks and Caicos. Pp. 79–86. In C. Mills Herpetology: A Tribute to Albert Schwartz. Society (Ed.), A History of the Turks and Caicos Islands. for the Study of Amphibians and Reptiles, Ithaca, New Macmillan, Oxford, UK. York, USA. MITCHELL, N., R. HAEFFNER,V.VEER,M.FULFORD- SLATKIN, M. 1987. Gene flow and the geographic structure GARDNER,W.CLERVEAUX,C.R.VEITCH, AND of natural populations. Science 236:787–792. G. MITCHELL. 2002. Cat eradication and the restoration SMITH, M. L., S. B. HEDGES,W.BUCK,A.HEMPHILL, of endangered iguanas (Cyclura carinata) on Long Cay, S. INCHAUSTEGUI,M.A.IVIE,D.MARTINA,M.MAUNDER, Caicos Bank, Turks and Caicos Islands, British West AND J. F. ORTEGA. 2005. Caribbean islands. Indies. Pp. 206–212. In C. R. Veitch and M. N. Clout Pp. 112–118. In R. A. Mittermeier, P. R. Gill, M. (Eds.), Turning the Tide: The Eradication of Invasive Hoffman, J. Pilgrim, T. Brooks, C. G. Mittermeier, J. Species. IUCN, Gland, Switzerland, and Cambridge, Lamoreaux, and G. A. B. da Fonseca (Eds.), Hotspots UK. Revisited: Earth’s Biologically Richest and Most MORITZ, C. 1994. Defining ‘‘evolutionarily significant Endangered Terrestrial Ecoregions. CEMEX, S. A. de units’’ for conservation. Trends in Ecology and C. V. Mexico City, Mexico. Evolution 9:373–375. STORFER, A. 1999. Gene flow and endangered species MYERS, N., R. A. MITTERMEIER,C.G.MITTERMEIER, translocations: A topic revisited. Biological Conserva- G. A. B. DA FONSECA, AND J. KE. 2000. Biodiversity tion 87:173–180. hotspots for conservation priorities. Nature 403: TOLSON, P. J. 1987. Phylogenetics of the boid snake genus 853–858. Epicrates and Caribbean vicariance theory. Occasional NEWSOM, L. A., AND E. S. WING. 2004. On Land and Sea: Papers of the Museum of Zoology, University of Native American Uses of Biological Resources in the Michigan 715:1–68. West Indies. The University of Alabama Press, TOLSON, P. J., AND R. W. HENDERSON. 2006. An overview Tuscaloosa, Alabama, USA. of snake conservation in the West Indies. Applied NOONAN, B. P., AND P. T. CHIPPINDALE. 2006a. Dispersal Herpetology 3:345–356. and vicariance: The complex evolutionary history of TORDOFF, H. B., AND P. T. REDIG. 2001. Role of genetic boid snakes. Molecular Phylogenetics and Evolution background in the success of reintroduced Peregrine 40:347–358. Falcons. Conservation Biology 15:528–532. 486 HERPETOLOGICA [Vol. 67, No. 4

TZIKA, A. C., S. KOENIG,R.MILLER,G.GARCIA,C.REMY, WARES, J. P. 2009. Natural distributions of mitochondrial AND M. C. MILINKOVITCH. 2008a. Population structure sequence diversity support new null hypotheses. of an endemic vulnerable species, the Jamaican boa Evolution 64:1136–1142. (Epicrates subflavus). Molecular Ecology 17:533–544. WATSON, J. 1989. Successful translocation of the endemic TZIKA, A. C., C. REMY,R.GIBSON, AND M. C. MILINKO- Seychelles kestrel Falco araea to Praslin. Pp. 363–367. VITCH. 2008b. Molecular genetic analysis of a captive In B.-U. Meyburg and R. D. Chancellor (Eds.), Raptors breeding program: The vulnerable endemic Jamaican in the Modern World. World Working Group on Birds yellow boa. Conservation Genetics 10:69–77. of Prey, Berlin, Germany. (USFWS) UNITED STATES FISH AND WILDLIFE SERVICE. WELCH, M. E., G. P. GERBER, AND S. K. DAVIS. 2004. 2009, Virgin Islands Tree Boa (Epicrates monensis Genetic structure of the Turks and Caicos rock iguana granti) 5 Year Review: Summary and Evaluation. US and its implications for species conservation. Pp. 58–70. Fish and Wildlife Service Southeast Region Ecological In A. C. Alberts, R. L. Carter, W. K. Hayes, and E. P. Services, Boquero´n, Puerto Rico. Martins (Eds.), Iguanas, Biology and Conservation. VILLESEN, P. 2007. FaBox: An online toolbox for fasta University of California Press, Berkeley, California, USA. sequences. Molecular Ecology Notes 7:965–968. WHITLOCK, M. C., AND D. E. MCCAULEY. 1990. Some WADE, M. J., AND D. E. MCCAULEY. 1988. Extinction and population genetic consequences of colony formation recolonization: Their effects on the genetic differenti- and extinction: Genetic correlations within founding ation of local populations. Evolution 42:995–1005. groups. Evolution 44:1717–1724. WANG, I. J., AND H. B. SHAFFER. 2008. Rapid color WHITTAKER, R. J., AND J. M. FERNA´ NDEZ-PALACIOS. 2007. evolution in an aposematic species: A phylogenetic Island Biogeography. Second edition. Oxford University analysis of color variation in the strikingly polymorphic Press, New York, New York, USA. strawberry poison-dart frog. Evolution 62:2742–2759. WRIGHT, S. 1977. Evolution and the Genetics of WAPLES, R. S. 1991. Pacific salmon, Oncorhynchus spp., Populations, Volume 3, Experimental Results and and the definition of ‘‘species’’ under the Endangered Evolutionary Deductions. University of Chicago Press, Species Act. Marine Fisheries Review 53:11–22. Chicago, Illinois, USA. WAPLES, R. S. 2006. Distinct population segments. Pp. 127–149. In J. M. Scott, D. D. Goble, and F. W. Davis (Eds.), The Endangered Species Act at Thirty: Conserving Biodiversity in Human Dominated Land- .Accepted: 19 June 2011 scapes. Island Press, Washington, DC, USA. .Associate Editor: Christopher Austin