Hybridization between two gartersnake species (Thamnophis) of conservation concern: a threat or an important natural interaction?

Hybridization between two gartersnake species (Thamnophis) of conservation concern: a threat or an important natural interaction?
of 15
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  RESEARCH ARTICLE Hybridization between two gartersnake species ( Thamnophis )of conservation concern: a threat or an important naturalinteraction? John S. Placyk Jr.  • Benjamin M. Fitzpatrick  • Gary S. Casper  • Randall L. Small  • R. Graham Reynolds  • Daniel W. A. Noble  • Ronald J. Brooks  • Gordon M. Burghardt Received: 23 June 2011/Accepted: 1 January 2012   Springer Science+Business Media B.V. 2012 Abstract  Distinguishing between hybrid zones formedby secondary contact versus parapatric divergence-with-gene-flow is an important challenge for understanding theinterplay of geographic isolation and local adaptation in thesrcin of species. Similarly, distinguishing between naturalhybrid zones and those that formed as a consequence of recent human activities has important conservation impli-cations. Recent work has demonstrated the existence of anarrow hybrid zone between the plains gartersnake( Thamnophis radix ) and Butler’s gartersnake ( T. butleri ) inthe Great Lakes region of North America, raising questionsabout the history and conservation value of geneticallyadmixed populations. Both taxa are of conservation con-cern, and it is not clear whether to regard hybridization as athreat or a natural interaction. Here we use phylogeo-graphic and population genetic methods to assess thetimescales of divergence and hybridization, and test forevidence that the hybrid zone is of recent origin. Weassayed AFLP markers and ND2 mitochondrial DNA(mtDNA) sequences from  T. radix ,  T. butleri , and theclosely related short-headed gartersnake ( T. brachystoma )throughout their North American ranges. We find shallowmtDNA divergence overall and high levels of variationwithin the contact zone. These patterns are inconsistentwith recent contact of long-diverged taxa. It is not possibleto distinguish true divergence-with-gene-flow from a long-term secondary contact zone, but we infer that the hybridzone is a long-standing, natural interaction. Keywords  AFLPs    Conservation genetics   Hybrid zone    mtDNA    Thamnophis Introduction Hybridization (interbreeding between genetically distinctpopulations; Harrison 1993) can severely affect the statusand recovery of threatened and endangered taxa (Rhymerand Simberloff  1996). Conservationists are often concernedabout genetic swamping, or loss of distinctiveness owing tohybridization and gene flow (Rhymer et al .  1994; Wolf et al . 2001; Allendorf and Luikart 2007). This is a particularly important consideration in cases of hybridization betweenintroduced and native species (Allendorf et al. 2001).However,evennaturalhybridizationcanindirectlyinfluencethreatened and endangered species by affecting the conser-vation status and legal protection of genetically mixedindividuals or populations (O’Brien and Mayr 1991; Allen- dorf et al. 2001; Schwartz et al. 2004; Haig and Allendorf  2006). Coming to grips with the ethical and legal questions John S. Placyk Jr., Benjamin M. Fitzpatrick, Gary S. Casper, andGordon M. Burghardt contributed equally to the content of thismanuscript.J. S. Placyk Jr. ( & )Department of Biology, University of Texas at Tyler,3900 University Blvd., Tyler, TX 75799, USAe-mail: jplacyk@uttyler.eduB. M. Fitzpatrick     R. L. Small    R. G. Reynolds   G. M. BurghardtDepartment of Ecology and Evolutionary Biology, Universityof Tennessee, 569 Dabney Hall, Knoxville, TN 37996, USAG. S. CasperUniversity of Wisconsin-Milwaukee Field Station,3095 Blue Goose Road, Saukville, WI 53080, USAD. W. A. Noble    R. J. BrooksDepartment of Integrative Biology, Science Complex,University of Guelph, Guelph, ON N1G 2W1, Canada  1 3 Conserv GenetDOI 10.1007/s10592-012-0315-4  raised by natural hybridization is an important challenge forconservation biology.Few government agencies have clear-cut policiesdefining the legal status of individuals whose ancestryincludes both a protected and unprotected species (Haigand Allendorf  2006). As a result, the consequences of hybridization for conservation are determined on a case-by-case basis. Ideally decisions are informed by scientificresearch regarding the impact of humans on the hybrid-ization process (e.g. is it a recent consequence of habitatmodification; Anderson 1948), the fitness effects of  hybridization (Fitzpatrick and Shaffer 2007; Muhlfeld et al. 2009), and impacts of hybrid genotypes on third-partyspecies in native communities (Ayres et al. 2004; Ryan et al. 2009). Here we use geographic analysis of mtDNA and AFLP variation to address questions about the srcin of a hybrid zone between two native snakes (both of conser-vation concern), and the potential value of mixed popula-tions as reservoirs of genetic variation. While mtDNAsequence data provides a historical perspective, its hap-loid and nonrecombinant mode of transmission make itimpossible to address questions about hybridizationwithout additional markers. Given this, we choose to useAFLPs to broadly sample the nuclear genome (e.g., Creeret al. 2004; Savolainen et al. 2006; Fitzpatrick et al. 2008; Nosil et al. 2009). We weigh the evidence for recent sec-ondary contact and admixture vs. a long-standing, naturalhybrid zone, and evaluate the genetic variability of thethreatened populations in the region of the contact zone.Specifically, we were interested in hybridization betweenButler’s gartersnake ( Thamnophis butleri ) and the plainsgartersnake ( T. radix ).Morphological and molecular evidence identify  T. radix as the sister group to  T. butleri  (Rossman et al. 1996;Alfaro and Arnold 2001; de Queiroz et al. 2002) and Rossman et al. (1996) have even suggested that  T. butleri  isa dwarfed (neotenic) derivative of   T. radix .  Thamnophisbutleri  is primarily found in the Midwest Region of theUnited States east of Lake Michigan with isolated popu-lations in southeast Wisconsin (Fig. 1; Rossman et al.1996; Harding 1997). Currently, it is listed as Threatened in Wisconsin (WI) (USA) by the WI Department of Natural Lake Superior L     a    k     e     H      u    r     o    n                             L                  a                         k                 e                         M                        i                  c                        h                        i                  g                         a                   n   L a  k e   E r  i e rNDrSDrMNrNErCOrIArOHrINrIL3rIL2rIL1brPA3brPA2brPA1bON16 OntarioIowaKansasMichiganOhioIllinoisMinnesotaMissouriNebraskaColoradoWisconsinMontanaManitobaSouth DakotaNorth DakotaWyomingIndianaKentuckySaskatchewanVirginiaPennsylvaniaNew YorkOklahomaWest VirginiaTennesseeNorth CarolinaNew Mexico 0 150 300 sretemoliK57 SamplesThamnophis radixThamnophis butleriThamnophis brachystoma    F   i  g   u  r  e   2   F   i  g   u  r  e   3 Fig. 1  Sites sampled for Butler’s gartersnake ( Thamnophis butleri ),plains gartersnake ( T. radix ), and short-headed gartersnake( T. brachystoma ) for this study along with the generalized historicalrange for each species (modified from Rossman et al. 1996). Samplelocations for Wisconsin and Ontario/Michigan are shown in Figs. 2and 3, respectively. Note that  T. butleri  and  T. radix  in Ohio areperipatric, yet there is no evidence that the ranges have everoverlapped.  Thamnophis radix  occupies a very specific habitat type incentral Ohio that is not used by  T. butleri . See text for details. Codesdenoting the location of each site are from ‘‘Appendix’’Conserv Genet  1 3  Resources (WDNR) ( and Endangered in Indi-ana (USA) by the Indiana Department of Natural Resources( ) and Ontario (Canada) by the OntarioMinistry of Natural Resources ( Thedecline of   T. butleri  is attributed mainly to habitatdestruction, with much of its preferred habitat (wetmeadows and prairies) being rapidly developed for com-mercial and residential purposes.  Thamnophis radix  iswidespread in the Great Plains of North America (Fig. 1),but is declining in many locations and is listed as a Speciesof Special Concern in Wisconsin by the WDNR andEndangered in Ohio (USA) by the Ohio Department of Natural Resources ( as a consequenceof limited habitat availability and habitat destruction(Dalrymple and Reichenbach 1981, 1984; Rossman et al. 1996). In Wisconsin,  T. radix  and  T. butleri  are sympatricand hybridize, but despite coarse-scale range overlap inOhio, sympatry between the two in that state has neverbeen recorded (Wynn and Moody 2006) most likely due tothe lack of suitable prairie habitat for  T. radix  to knownpopulations of   T. butleri .Morphological (Casper 2003), behavioral (Ford 1982; Kirby 2005), and molecular data (Fitzpatrick et al. 2008) support the existence of a narrow hybrid zone betweenWisconsin populations. Outside of this zone,  T. butleri  and T. radix  are ecologically and morphologically distinct(Rossman et al. 1996), and even some hybrid populationsshow evidence of bimodality, indicating that the distinc-tiveness of the two forms is maintained in the face of geneflow (Fitzpatrick et al. 2008). An important alternative is that secondary contact is very recent, and a process of ‘‘species collapse’’ is just beginning (Taylor et al. 2006; Seehausen et al. 2008). The hybrid zone is coincident withthe City of Milwaukee and a vast, growing network of suburbs extending west from Lake Michigan. Therefore,the taxonomically problematic hybrid populations arehighly threatened by habitat destruction. A scientifically justified decision regarding their conservation value is animmediate concern. Methods Study populations and sample collectionMolecular variation was examined in 549 individual snakesfrom 74 locations including 316  T. butleri  from 45 sites,105  T. radix  from 17 sites, 123 hybrids from 9 sites withinthe hybrid zone (‘‘Appendix’’; Figs. 1, 2, 3). We also included 5  T. brachystoma  from 3 sites as outgroups for themtDNA. In addition to the sequences generated during thecourse of this study, we also included 1  T. radix  sequence(GenBank Accession No. AF384853, Alfaro and Arnold2001), 1 Michigan  T. butleri  sequence (GenBank Acces-sion No. AF420094, de Queiroz et al. 2002) and 1  T. bra-chystoma  sequence (GenBank Accession No. AF420091,de Queiroz et al. 2002) from GenBank. While samplingwas concentrated in or around the hybrid zone in Wis-consin and Illinois to examine fine scale patterns of vari-ation in this area, additional sites were sampled from acrossthe ranges of both  T. butleri  (i.e. Michigan, Ohio, Ontario)and  T. radix  (i.e. Indiana, Ohio, Iowa, Colorado, Nebraska,Minnesota, North Dakota, South Dakota). Samples wereobtained from numerous sources (see Acknowledgmentsand ‘‘Appendix’’) as frozen muscle tissue or as tail tips orventral scale clips from live specimens subsequentlyreleased at the point of capture.Mitochondrial DNA amplificationGenomic DNA was obtained with the DNeasy  Tissue Kit(Qiagen). The 985 bases of ND2 that we examined werePCR-amplified using the forward primer L4437b (5 0 -CAGCTA AAA AAG CTA TCG GGC CCA TAC C-3 0 ; Ku-mazawa et al. 1996), which lies in the tRNA-Met upstreamof ND2 and the reverse primer Sn-ND2r (5 0 -GGC TTTGAA GGC TMC TAG TTT-3 0 ; R. Lawson, pers. comm.),which lies in the tRNA-Trp downstream of ND2. In eachcase, polymerase chain reactions (PCR) were conducted in25- l L volumes with 1.0  l L DNA, 1 9  ExTaq PCR buffer(PanVera/TaKaRa), 1.5 mM MgCl 2 , 0.2 mM dNTPs,0.2  l g/  l L bovine serum albumin, 0.1 mM each primer,and 1.25 units of ExTaq polymerase (Panvera/TaKaRa).Amplification conditions involved 30 cycles each consist-ing of 1 min of denaturing at 94  C, 1 min of primerannealing at 55  C, and 1.5 min of extension at 72  C. PCRproducts were cleaned prior to sequencing using ExoSAP-IT TM (USB Corporation).Sequencing reactions were carried out using the internalprimers H5382 (5 0 -GTG TGG GCR ATT CAT GA-3 0 ) andL5238 (5 0 -ACM TGA CAA AAA ATY GC-3 0 ) (de Queirozetal.2002)andBigDye  Terminatorv3.1CycleSequencingkits (Applied Biosystems), and read on an automatedsequencer (Applied Biosystems 3100, University of Ten-nessee Molecular Biology Resource Facility). Sequenceswere edited using the program Sequencher 3.1.1 (GeneCodes Corporation, Ann Arbor, MI). tRNAs were trimmedfrom our sequences and alignments were performed initiallyusing Clustal X (Thompson et al. 1997) and subsequently manually refined. Sequences were collapsed into uniquehaplotypes using Collapse v1.2 prior to analyses.Mitochondrial DNA analysesWe estimated a mitochondrial gene tree using the ND2 dataunder the criterion of maximum likelihood (ML) as Conserv Genet  1 3  implemented in PAUP* (Swofford 2002) and Bayesianinference of phylogeny (BI) implemented in MrBayes 3.1(Ronquist and Huelsenbeck  2003). The best fit model of  evolution (HKY ? G) and estimation of parameters for thedataset under the Akaike information criterion (with cor-rection for small sample size) were determined utilizingModeltest 3.7 (Posada and Crandall 1998). ML and BI analyses were rooted with published GenBank sequencesfrom  T. elegans , which represents a clade that is sister to a T. butleri ,  T. radix , and  T. brachystoma  clade (Alfaro andArnold 2001; de Queiroz et al. 2002). ML analyses were conducted with 1,000 randomsequence addition heuristic search replicates with tree-bisection-reconnection (TBR) branch swapping and col-lapsing all zero-length branches. Bootstrap analysis wasemployed to assess internal support for the inferred phy-logeny using 1,000 bootstrap replicates with simple taxonaddition heuristic searches, TBR branch swapping andcollapsing all zero-length branches.All Bayesian analyses were run in duplicate for 5 mil-lion generations and were inspected for stationarity(effective mixing and convergence to the posterior distri-bution) using Tracer v1.4 (Rambaut and Drummond 2007),with the initial 10% of these generations (parameters esti-mated prior to effective mixing and convergence of theMCMC chain) discarded as burn-in.Because we sampled many individuals with similar oridentical mtDNA haplotypes, we also estimated a haplo-type network using TCS 1.13 (Clement et al. 2000) to help visualize the distribution of mtDNA variation. Ambiguousconnections (loops or reticulations) in the gene tree wereresolved using approaches from coalescent theory(Crandall et al. 1994). In the case of DNA sequence datathis resolution generally involves a comparison of theprobabilities of whether a haplotype arose via mutationfrom either a high- or low-frequency haplotype, with fre-quency evaluated based on both numerical frequency andgeographic distribution. That is, abundant and widespread hb9hb8hb7hb6hb5hb4hb3hb1rWI6 rWI5rWI4rWI3rWI2rWI1hb11hb10bWI9bWI8bWI7bWI6bWI5bWI4bWI3bWI2bWI1bWI18bWI23bWI22bWI21bWI20bWI19bWI17 bWI16bWI15bWI14bWI13bWI12bWI11bWI10 DaneRockDodgeGreenSaukColumbiaJeffersonWalworthWaukeshaFond Du LacIowaRacineSheboyganAdamsWashingtonLakeMarquetteKenoshaOzaukeeLafayetteGreen LakeMilwaukeeMcHenryBooneWinnebagoStephensonJuneauJo Daviess 0 10 205 Kilometers SamplesThamnophis radixThamnophis butleri Fig. 2  Wisconsin sites sampled for Butler’s gartersnake ( Thamnophis butleri ), plains gartersnake ( T. radix ), and hybrids for this study.  Codes denoting the location of each site are from ‘‘Appendix’’Conserv Genet  1 3  haplotypes are more likely ancestors than rare and geo-graphically restricted haplotypes.mtDNA evidence for population structure was assessedby performing a Spatial Analysis of Molecular Variance(SAMOVA).  U -statistics were calculated using SAMOVA1.0 (Dupanloup et al. 2002). SAMOVA consists of identi- fyinggroupsofpopulationsthataremaximallydifferentiatedfrom each other based on genetic data and spatial relation-ships.SAMOVAallowsforanalyseswithoutrelyingonfieldidentification of individual snakes as either  T. butleri , T. radix , or hybrids based on their morphology. A priorigroups, as implemented when using Analysis of MolecularVariance (AMOVA), are not utilized in SAMOVA. How-ever, we also used AMOVAs to examine three specificrelationships associated with potential hybridization/recentdivergence in Wisconsin: (1) Wisconsin (WI)  T. butleri  vs.WI  T. radix , (2) WI  T. butleri  vs. hybrids (as identified bymorphology;Rossmanetal.1996;Casper 2003),and (3)WI T. radix  vs. hybrids.  P  values of the  U -statistics were esti-mated through 1,000 permutation replicates.Finally,althoughthereisnoexplicitphylogeographictestfor range expansion into a hybrid zone (see, for exampleBloomquist et al. 2010), we used generalized tests of muta-tion-drift equilibrium to evaluate whether mtDNA variationin the hybrid zone and neighboring sites in Wisconsin givesany indication of recent, non-equilibrium dynamics. WeusedDnaSPv5(RozasandRozas1999)tocalculateTajimas  D andFuandLi’s  D* and F* withassociated P values.Thesestatisticsaresignificantlylessthanzerowhentherearemanyrare variants, as expected following population growth or aselectivesweep,andsignificantlygreaterthanzerowhenrarevariants are uncommon, as expected from balancing selec-tion or following a persistent bottleneck (Tajima 1989; Fu and Li 1993). Range expansion and admixture would mostlikelycausenegative  D,D* and F* inpopulationsadjacenttothe contact zone, but positive test statistics in recentlyadmixed populations. Structured populations at mutation-migration-drift equilibrium tend to have more rare variantsthan unstructured populations, hence negative test statisticsfor all three analyses (Peter et al. 2010). Fig. 3  Ontario, Michigan, Ohio and Pennsylvania sites sampled for Butler’s gartersnake ( Thamnophis butleri ), and short-headed gartersnake( T. brachystoma ) for this study.  Codes  denoting the location of each site are from ‘‘Appendix’’Conserv Genet  1 3
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!