A molecular evaluation of conservation units, translocations, and habitat fragmentation for a threatened species the White Sands pupfish

1 A molecular evaluation of conservation units, translocations, and habitat fragmentation for a threatened species: the White Sands pupfish J.S Heilveil* and C.A Stockwell 6Department of Biological Sciences, Stevens Hall, North Dakota State University, Fargo, ND 58105; e7mail: Jeffrey.Heilveil@ndsu.edu; Craig.Stockwell@ndsu.edu 9* Current address: Department of Natural Sciences, 1200 Murchison Rd Fayetteville State University, 10Fayetteville, NC 28301, jheilveil@uncfsu.edu 11 12Keywords: conservation genetics, fragmentation, microsatellites, Cyprinodon tularosa, translocation, ESU 13 14Correspondence to: 15Craig A Stockwell 16Department of Biological Sciences, Stevens Hall, North Dakota State University, Fargo, ND 58105-5517 17Phone: (701) 231-8449, Fax: (701) 231-7149, e-mail: Craig.Stockwell@ndsu.edu 18 19 Running Title: Conservation genetics of White Sands pupfish 1 1Abstract 2Cyprinodon tularosa, a New Mexico state-Threatened fish, is restricted to four populations worldwide and 3has been subjected to both translocations and habitat fragmentation Two native populations at Malpais 4Spring and Salt Creek were previously recognized as ESUs and two non-native populations at Lost River 5and Mound Spring were recently derived by translocation from the Salt Creek ESU Further, the latter 6three habitats are fragmented Here we use a suite of fourteen recently developed microsatellite markers 7to evaluate overall genetic structure, ESU designations, introduction history, and effects of habitat 8fragmentation Overall FST was high (0.382); however, removing fish from Malpais Spring (a sensu lato 9ESU) reduced FST (0.058), confirming a previous ESU designation made based on a limited number of 10markers Genetic assignment tests (both Bayesian and Likelihood) suggest Salt Creek Lower (SCLW) as 11the source population for both the Lost River and Mound Spring populations Lost River showed the 12most divergence from SCLW, supporting anecdotal reports of a small number of founders (n=30) 13Populations from both Salt Creek and Lost River show significant differences between upper and lower 14populations, but no barrier effects were seen in either Mound Spring or Malpais Spring Based on these 15data, management recommendations for the species are discussed 1Introduction 2The patterns of temporal and spatial genetic structure are of critical importance to the conservation of rare 3and endangered species (e.g Waples 1991; Moritz 1994; Moran 2002; Moritz 2002; Bouzat and Johnson 42004) For example, the designation of conservation units is an important first-step in the genetic 5management of rare species (Waples 1991; Moritz 1994; Crandall et al 2000) This information can 6direct management decisions to preserve distinct evolutionary lineages, and help prioritize management 7actions such as the establishment of refuge populations (Stockwell et al 1998) 9Molecular data can also be useful for evaluating the effects of historic transplants and habitat 10fragmentation on genetic variation Historical translocation events can be important drivers in altering 11genetic structure, causing reductions in genetic diversity in recently introduced populations (e.g 12Stockwell et al 1996; Helenurm and Parsons 1997) and potentially compromising their evolutionary 13potential Landscape factors, such as anthropogenic or natural barriers, may further reduce genetic 14diversity (e.g Smith et al 1983; Templeton et al 1990; Keller and Largiadèr 2003; Yamamoto et al 152004) 16 17High levels of genetic diversity in a population can increase the ability to respond to novel threats 18(Frankham 1995; Amos and Balmford 2001) It is therefore important, when designing the conservation 19plan for a species, to maintain as much genetic diversity as possible within and among populations The 20maintenance of genetic diversity is influenced by gene flow; as only a few migrants between populations 21are necessary to ameliorate the negative effects of genetic drift (e.g Wright 1931, 1970; Mills and 22Allendorf 1996) When physical barriers are present, however, gene flow can be severely limited or 23eliminated In aquatic systems, the damming of rivers has been shown to result in reduced genetic 24diversity for above-dam populations (e.g Smith et al 1983; Yamamoto et al 2004) These barrier25induced reductions in genetic diversity can be especially important in species with limited geographic 26distribution 27 28Collectively, the issues of conservation units, translocations and habitat fragmentation are of particular 29importance to the conservation of western fishes (Minckley and Deacon 1991; Waples 1991; Minckley 301995; Allendorf and Waples 1996; Vrijenhoek 1996; Stockwell and Leberg 2002) Identification of 31conservation units has been particularly important for the conservation of pacific salmon (Waples 1991; 32Waples 1995) and desert fishes (Quattro et al 1996; Stockwell et al 1998; Parker et al 1999; Echelle et 33al 2000) Further, habitat fragmentation is of particular concern for many western fishes such as pacific 34salmon and various trout species (Nehlsen et al 1991; Moyle 1994) Finally, translocations have been 35extensively used, especially for the conservation of many protected fish species (Wiliams et al 1988; 36Hendrickson and Brooks 1991; Minckley 1995; Vrijenhoek 1996) 37 38Historically, many studies of fish species have been constrained by relatively low genetic variation at 39traditional markers (Echelle 1991); however, the development of hypervariable markers, e.g 40microsatellites, has allowed genetic structure to be studied in such species (Parker et al 1999; Martin and 41Wilcox 2004) 42 43One species of particular concern is the White Sands pupfish (Cyprinodon tularosa; Miller and Echelle 441975), a New Mexico state-listed Threatened species This species, which is restricted to two native and 1two introduced populations in southern New Mexico, has limited variation at 37 allozyme loci and a short 2segment of mtDNA d-loop (Echelle et al 1987; Stockwell et al 1998) Recently, a battery of 3microsatellites has been developed for C tularosa and its congeners (Jones et al 1998; Stockwell et al 41998; Burg et al 2002; Iyengar et al 2004), which allows us to evaluate the variation between native 5populations, as well as the effects of historic translocations and habitat fragmentation on genetic variation 6in this species 8Background 9Malpais Spring and Salt Creek, which harbor native populations of C tularosa (Pittenger and Springer 101999), are located on White Sands Missile Range (WSMR; Fig 1) In 1970, a population was established 11at Lost River on Holloman AFB, and a second population was established between 1967 and 1973 at 12Mound Spring on WSMR (Pittenger and Springer 1999) The source population(s) for the introductions 13at Lost River and Mound Spring was not documented, but 30 fish were reportedly used to establish the 14Lost River population (Pittenger and Springer 1999) 15 16Stockwell et al (1998) used genetic and ecological data to recognize two Evolutionarily Significant Units 17(ESUs; sensu lato) of White Sands pupfish, the Malpais Spring ESU and the Salt Creek ESU The Salt 18Creek ESU was shown to include the native populations at Salt Creek as well as the Lost River and 19Mound Spring populations both of which were descended from the Salt Creek population (Stockwell et 20al 1998) The designation of the two ESUs was based on fixed and nearly-fixed differences at a 21microsatellite marker and an allozyme marker, respectively Further, Malpais Spring and Salt Creek differ 22ecologically in terms of salinity, flow and parasite communities (Stockwell and Mulvey 1998; Stockwell 23et al 1998; Collyer and Stockwell 2004; Rogowski 2004; Collyer et al 2005) The earlier study did not 24evaluate the retention/loss of genetic variation in the introduced populations at Lost River and Mound 25Spring The genetic effects of habitat fragmentation on this species have not been well studied due to a 26lack of variable markers 27 28Within each population, there are varying degrees of habitat fragmentation (Figure 1) Malpais Spring 29experiences the least fragmentation, with continuous habitat between the springhead and a nearby playa 30Although Malpais Spring lacks physical barriers to migration, non-obvious barriers have been shown to 31significantly deter gene flow between sub-populations in other animals (e.g Hitchings and Beebee 1997) 32Indeed, Stockwell and Mulvey (1998) found a significant difference in PGDH frequencies between upper 33and lower Malpais Spring that was correlated with a sharp gradient in environmental salinity Earlier 34workers reported distinct ponds (Miller and Echelle 1975) in the southern end of the Malpais Spring 35complex, though water is currently diverted to the south where these isolated ponds would have existed; 36leaving the hydrological relationship between upper and lower Malpais Spring poorly understood 37 38Salt Creek is fragmented by a head-cut waterfall approximately 1.2 m in height (Stockwell and Mulvey 391998; Pittenger and Springer 1999) Salt Creek is further fragmented by a road culvert, but bi-direction 40migration is possible during and after high water events (J Pittenger, NMDFG, pers comm.) 41 42Lost River is intermittent, with permanent water in three different reaches The upper reach is 43characterized by a series of deep pools that are periodically connected following high water events The 44middle reach is bounded by a road culvert upstream and a dry playa downstream Below the playa, Lost 1River re-emerges and flows until it reaches the White Sands dune fields During high water events, the 2three segments are connected, but the road culvert prevents fish from moving upstream 4Mound Spring is composed of upper and lower pools, which differ in elevation by approximately 5meters A small amount of water trickles from the upper to lower pools, but it is unclear if fish migrate to 6the lower pools Upstream fish movement is not possible 8This study employed microsatellite DNA markers to assess the distribution of genetic variation in the 9species, with the goal of informing future conservation and management efforts Specifically, we 10attempted to evaluate the ESU designation of Stockwell et al (1998) and the effects of historic 11introductions and habitat fragmentation on genetic diversity in this species 12 13Materials and Methods 14Sampling and molecular techniques 15Forty fish were caught by minnow-trapping and seining during March, 2003, at each of the following 16eight locations: 1) Malpais Spring below the USGS flow gauge (MLUP), 2) Malpais Spring, lower pool 17(Jet Playa) (MLLW), 3) Salt Creek, above the waterfall and at terminus of upper road (SCUP), 4) Salt 18Creek, below Range Road 316 (SCLW), 5) Mound Spring, upper pool (MDUP), 6) Mound Spring, lower 19pool (MDLW), 7) Lost River, above the confluence of Ritas Draw and Malone Draw (LRUP), 8) Lost 20River, near its terminus at the dune fields (LRLW) Fish were subsequently sacrificed, frozen, and stored 21at -80 oC upon return to the laboratory 22 23Whole genomic DNA was extracted from fin tissue using DNeasy kits (Qiagen) and stored at oC 24Fourteen microsatellite loci previously shown to be polymorphic in C tularosa were used to assess 25genetic differentiation: WSP11 (Jones et al 1998); WSP2 (Stockwell et al 1998); WSP20, WSP23, 26WSP24, WSP25, WSP26, WSP30, WSP32, WSP33, WSP34 (Iyengar et al 2004); AC23, C509, and 27GATA02 (Burg et al 2002) (See Table 1) 28 29Amplification reactions were performed in 25 ul volumes using 2.5 ul 10x PCR buffer, 1ul mM dNTP 30mix, 0.175 ul AmpliTaq Gold polymerase (Applied Biosystems), ul template DNA, 0.25 uM unlabeled 31reverse primer, and dye-labeled forward primer (for concentration, see Table 1) The annealing 32temperatures and number of cycles varied for each primer set (Table 1) Automated fragment analysis 33was performed on a Beckman Coulter CEQ8000, using 600 size-standard (0.5 ul), for the following four 34groups of pool-plexed PCR products: 1) WSP2 (2.0 ul), WSP20 (0.25 ul), WSP26 (1.0 ul), C509 (1.0 ul); 352) AC23 (0.5 ul), WSP23 (0.5 ul), WSP24 (1.0 ul), WSP33 (0.25 ul); 3) WSP11 (1.0 ul), WSP30 (0.5 ul), 36WSP32 (0.25 ul); 4) GATA02 (1.0 ul), WSP25 (0.5 ul), WSP34 (0.5 ul) 37 38Data Analyses 39Measurements of Hardy-Weinburg Equilibrium (HWE), linkage disequilibrium, F-statistics, a locus-by40locus AMOVA, exact tests of sample differentiation, and genotype assignment tests were performed for 41each subpopulation using Arlequin (ver 3.0, Excoffier et al 2005) One of the loci, WSP11, was found to 42consist of a complex compound microsatellite and was therefore input into Arlequin as allele sizes, 43because the number of repeats could not be determined The ESU designation and source of the 44introduced populations were determined using the results of the genotype assignment tests For ESU 1designation, the average log-likelihood for each ESU was compared for all 320 fish Earlier work 2suggested that the Mound Spring and Lost River populations were established with fish from lower Salt 3Creek (as opposed to upper Salt Creek; Stockwell and Mulvey 1998) To further evaluate this hypothesis, 4we compared log-likelihood scores for SCUP and SCLW for each fish from Mound Spring and Lost River 5populations, pooling the upper and lower sub-populations together within each population 7Overall and locus-by-locus tests of population differentiation were performed using polymorphic loci to 8examine whether barrier-separated populations differed genetically Additionally, the number of alleles 9per locus and the specific alleles present in each population were used to examine the loss/retention of 10alleles for the introduced populations at Lost River and Mound Spring 11 12To corroborate the results from Arlequin, genotype data were analyzed using STRUCTURE (ver 2.0, 13Pritchard et al 2000; MCMC = 1,100,000 generations, burnin = 100,000 generations) This analysis uses 14Bayesian methods to assign individuals to population-groups disregarding geographic origin, allowing 15natural population divisions to be determined 16 17Results 18Overall genetic structure and ESU designation 19None of the loci showed consistent linkage disequilibrium, and therefore were assumed to be unlinked for 20all further analyses After bonferroni correction was performed, no population was significantly out of 21HWE For all populations, across all loci, FST = 0.382, indicating a high level of structure (Table 2a) 22When the Malpais Spring fish were removed from the analysis, FST dropped to 0.058 (Table 2b) Fixed 23differences between the two ESUs were observed for WSP-11 and WSP-20, and commonly occurring 24(frequency > 0.20) private alleles were observed for GATA02, WSP-24, WSP-25, WSP-26, and WSP33 25 26According to the genotype assignment tests, all fish were correctly assigned to their ESU of origin When 27fish from all eight subpopulations were analyzed in STRUCTURE, Malpais Spring Fish were clearly 28different from all other populations (Fig 2a), supporting the ESU designation (Stockwell et al 1998) 29Within the Salt Creek ESU, fish were assigned to their specific habitats most of the time; Salt Creek – 3077.5%; Lost River – 83.8%; and Mound Spring – 75% 31 32Introduction History 33According to the genotype assignment test, 85% (n = 80) and 87.5% (n = 80) of all fish were more likely 34to have originated in SCLW than SCUP for Lost River and Mound Spring, respectively We therefore 35considered SCLW to be the source population for the fish introduced to Mound Spring and Lost River To 36test for genetic divergence of introduced populations, we compared each subpopulation at Mound Spring 37and Lost River to their putative founding stock (SCLW) in terms of gene frequencies and allelic richness 38 39The exact tests of sample differentiation showed both subpopulations at Lost River to be significantly 40different from SCLW (Table 3), while neither subpopulation at Mound Spring (MDUP and MDLW) was 41significantly different from SCLW When the Salt Creek ESU was analyzed alone in STRUCTURE, fish 42from Lost River were assigned the highest likelihood of coming from a single population, while fish from 43Salt Creek and Mound Spring were relatively indistinguishable (Fig 3) 44 1Compared to the SCLW subpopulation, LRLW lost 0.9 alleles per locus, LRUP lost allele / locus, 2MDLW lost 0.6 alleles / locus, and MDUP lost 0.8 alleles / locus (Table 4) The lost alleles ranged in 3frequency at SCLW as follows: LRLW, 1.75% - 3.75%; LRUP, 1.75% - 15%; MDLW, 3.75% - 6.25%; 4and MDUP, 1.75% - 10% Not all of the SCLW alleles lost at LRLW were missing in LRUP, despite 5LRLW being the most likely source of introduction at Lost River 7Barrier effects 8The upper and lower subpopulations of Lost River and Salt Creek populations were found to be 9significantly different at 63% of the polymorphic loci, while significant differences were only seen in 8% 10of polymorphic loci at Malpais Spring, and loci at Mound Spring (Fig 4) Allelic richness was lower 11for LRUP at loci and SCUP at loci (though higher for locus; Table 5), as compared to their lower 12counterpart 13 14In both Mound and Malpais Springs, the loss of allelic richness above the barrier was balanced with an 15equal number of alleles present above the barrier that were absent in the lower sub-population (two and 16four alleles, respectively; Table 5) 17 18The probability of correctly assigning a fish to its own sub-population was generally high When 19genotypic assignment was restricted by drainage, correct assignment occurred for 83% of MLLW fish, 2078% of MLUP fish, 85% of SCLW fish, 95% of SCUP fish, 80% of LRLW fish, 78% of LRUP fish, 80% 21of MDLW fish, and 85% of MDUP fish (Fig 4) 22 23Discussion 24This multilocus microsatellite dataset revealed a number of interesting patterns, with important 25consequences for the conservation and management of Cyprinodon tularosa The high FST and correct 26assignment of fish supports the ESU designations of Stockwell et al (1998) The genetic signature of 27these differences was evident in over half of the loci with either fixed differences or commonly occurring 28alleles that were restricted to one ESU 29 30We echo earlier recommendations (Stockwell et al 1998) that additional efforts be made to secure both 31the Malpais Spring ESU and the Salt Creek ESU of C tularosa 32 33If refuge populations are to be established, then the Malpais ESU should be given a top priority for 34“replication”, as it is genetically distinct from all other C tularosa populations Although the Salt Creek 35ESU appears to be at lower risk, we suggest guidelines for the genetic management of this population to 36mediate the effects of historic translocations and barriers 37 38Within the Salt Creek ESU, we observed significant drift in the Lost River population as evidenced by 39exact tests of differentiation and by a 21.4 – 23.8% reduction in allelic richness These data suggest that a 40bottleneck occurred during the founding of this population, consistent with anecdotal reports of the Lost 41River population resulting from a translocation of 30 fish (Pittenger and Springer 1999) Additionally, the 42significant differences between LRLW and LRUP suggest a second bottleneck occurred in the LRUP 43subpopulation, a conclusion supported by an 11% reduction in allelic richness between LRLW and LRUP 1The reduced allelic richness, often cited as a signature of population bottlenecks (e.g Leberg 1992), is 2especially noteworthy because of the limited number of loci examined 4In contrast, the Mound Spring population did not significantly diverge from SCLW; however, allelic 5richness has been compromised for this population (14.3 – 19% reduction) The loss of alleles at Mound 6Spring and Lost River is consistent with theoretical expectations (Allendorf 1986) and earlier empirical 7studies (Stockwell et al 1996) 9The loss of allelic richness is of some concern, as it may compromise evolutionary potential One 10solution may be to use artificial migration to increase genetic diversity in both Lost River and Mound 11Spring This type of “genetic restoration” (sensu Hedrick 1995; Westemeier et al 1998) would increase 12the value of these populations as “genetic replicates” for the Salt Creek population; however, this 13approach should be balanced against the possibility that migrants may compromise local adaptation in the 14“refuge” populations (see Storfer 1999; Stockwell et al 2006) 15 16Within the Salt Creek ESU, rapid divergence at an allozyme locus and in body shape for the Mound 17Spring population appears to be due to altered selection from reduced salinity and rate of flow (Stockwell 18and Mulvey 1998; Collyer et al 2005) Because gene flow can compromise local adaptation (Storfer 191999; Stockwell et al 2006), we repeat the recommendation of Stockwell et al (1998) that Mound Spring 20population be treated as a separate Management Unit 21 22By contrast, the ecological similarity of Lost River to Salt Creek reduces the likelihood that Lost River 23has rapidly diverged from Salt Creek In fact, divergence at an allozyme locus has been modest, and there 24has been no divergence for life history traits (Rogowski 2004) or body shape (Collyer et al 2005) In 25terms of ecological replication, Lost River is very similar to Salt Creek in both salinity and flow 26(Stockwell and Mulvey 1998; Rogowski 2004) Although the possibility that Lost River has diverged for 27some unsurveyed trait can not be ruled out, we suggest that the advantages of genetic restoration outweigh 28the unlikely costs of gene flow 29 30While theory predicts that as few as one migrant per generation (Wright 1931; Mills and Allendorf 1996) 31should keep populations from diverging, this may not be sufficient to restore allelic diversity in a timely 32fashion As an alternative, translocating 100 fish from Salt Creek should theoretically allow alleles as rare 33as 0.5% in Salt Creek to be introduced into Lost River, greatly increasing the replication value of the 34population 35 36We therefore recommend one of two approaches be taken to restore genetic variation at Lost River One 37approach would be a one-time movement of 100 fish followed by movement of one migrant per 38generation (once a year, see Rogowski 2004) 39 40Although 1-10 migrants per generation is a more traditional rule of thumb, higher levels of gene flow may 41be necessary when a population has been isolated for an extended period, (Mills and Allendorf 1996), 42such as the case with Lost River A less aggressive approach would be to relocate 10 fish per generation 43(year) over at least 10 generations, followed by one migrant per generation (see Wright 1931; Mills and 44Allendorf 1996); however, the effects of genetic restoration would be delayed by at least one decade 2In both cases, we recommend that fish introductions be taken in a series of steps For instance, the first 3migration pulse could be targeted for lower Lost River with later additions progressing upstream This 4iterative approach would also allow managers to evaluate if the translocations are resulting in genetic 5restoration 7As previously mentioned, the genetic effects of fragmentation were only observed at Lost River and Salt 8Creek; in both cases, allelic richness was reduced for the upstream sub-populations Changes in genetic 9structure due to dams are common in fish (e.g Yamamoto et al 2004; Neraas and Spruell 2001) Gene 10flow directly from Salt Creek to the three segments of Lost River should be sufficient for managing the 11divergence between the habitat fragments The reduced variation of SCUP is likely due to the headcut 12waterfall It is not clear, however, if this waterfall is of recent anthropogenic origin, or whether the Salt 13Creek population has been historically fragmented In terms of population management, we believe the 14replication of fish in other habitats should offer sufficient insurance and thus not recommend any 15movement of fish within Salt Creek 16 17The lack of significant divergence between MDLW and MDUP and the moderate (8.1%) reduction in 18allelic diversity, leads us to conclude that one migrant per generation (Wright 1931, 1970; Mills and 19Allendorf 1996) could be applied in bidirectional fashion to keep these subpopulations from diverging 20from each other 21 22These data combined with information from previous studies (Stockwell and Mulvey 1998; Stockwell et 23al 1998; Collyer and Stockwell 2004; Rogowski 2004; Collyer et al 2005), have allowed us to better 24evaluate the genetic management of White Sands pupfish A similar iterative approach to management 25has previously been taken with other desert fishes such as the Gila topminnow (Poeciliopsis occidentailis 26occidentalis; Parker et al 1999) This approach is especially useful for species with complicated histories 27that are actively managed Maintaining the genetic diversity of threatened populations should enhance 28both the short and long-term prospects for such species 29 30Literature Cited 31Allendorf FW (1986) Genetic drift and the loss of alleles versus heterozygosity Zoo Biology, 5, 18132 190 33Allendorf FW, Waples RS (1996) Conservation and genetics of salmonid fishes In: Conservation 34 genetics; case studies from nature (eds Avise JC, Hamrick JL) pp.238-280 Chapman & Hall, 35 New York 36Amos W, Balmford A (2001) When does conservation genetics matter? Heredity, 87, 257-265 37Bouzat JL, Johnson K (2004) Genetic structure among closely spaced leks in a peripheral population of 38 lesser prairie-chickens Molecular Ecology, 13, 499-505 39Burg TM, Wilcox JL, Martin A (2002) Isolation and characterization of polymorphic microsatellite loci in 40 pupfish (genus Cyprinodon) Conservation Genetics, 3, 197-204 41Collyer ML, Novak J, and Stockwell CA (2005) Morphological divergence in recently established 42 populations of White Sands Pupfish (Cyprinodon tularosa) Copeia, 2005, 1-11 43Collyer ML, Stockwell CA (2004) Experimental evidence for costs of parasitism for a threatened species, 44 White Sands pupfish (Cyprinodon tularosa) Journal of Animal Ecology, 73, 821-830 1Crandall KA, Bininda-Emonds ORP, Mace GM, Wayne RK (2000) Considering evolutionary processes in evolutionary biology Trends in Ecology & Evolution, 15, 290-295 3Echelle AA (1991) Conservation genetics and genic diversity in freshwater fishes of western North America In: Battle against extinction (eds Minckley WL, Deacon JE) pp.141-153 University of Arizona Press, Tucson 6Echelle, AA, Echelle AF, and Edds DR (1987) Population Structure of Four Pupfish Species (Cyprinodontidae: Cyprinodon) from the Chihuahuan Desert Region of New Mexico and Texas: Allozymic Variation Copeia, 1987, 668-681 9Echelle AA, Van Den Bussche RA, Malloy TP Jr., Haynie ML and Minckley CO (2000) Mitochondrial 10 DNA variation in pupfishes assigned to the species Cyprinodon macularius (Atherinomorpha, 11 Cyprinodontidae): taxonomic implications and conservation genetics Copeia, 2000, 353-364 12Excoffier L, Laval G, Schneider S (2005) Arlequin ver 3.0: An integrated software package for 13 population genetics data analysis Evolutionary Bioinformatics Online, 1, 47-50 14Frankham R (1995) Conservation Genetics Annual Review of Genetics, 29, 305-327 15Hedrick PW (1995) Gene flow and genetic restoration: The Florida panther as a case study 16 Conservation Biology, 9, 996-1007 17Helenurm K, Parsons LS (1997) Genetic variation and the reintroduction of Cordylanthus maritimus ssp 18 maritimus to Sweetwater Marsh, California Restoration Ecology, 5, 236-244 19Hendrickson DA, Brooks JE (1991) Transplanting short-lived fishes in North American deserts: review, 20 assessment and recommendations In: Battle against extinction (eds Minckley WL, Deacon JE) 21 pp.283-293 University of Arizona Press, Tucson 22Hitchings SP, BeeBee TJC (1997) Genetic substructuring as a result of barriers to gene flow in urban 23 Rana temporaria (common frog) populations: implications for biodiversity conservation 24Iyengar A, Stockwell CA, Layfield D, Morin PA (2004) Characterization of microsatellite markers in a 25 threatened species, the White Sands pupfish (Cyprinodon tularosa) Molecular Ecology Notes, 4, 26 191-193 27Jones AG, Stockwell CA, Walker D, Avise JC (1998) The molecular basis of a microsatellite null allele 28 from White Sands pupfish The Journal of Heredity, 89, 339-342 29Keller I, Largiadèr CR (2003) Recent habitat fragmentation caused by major roads leads to reduction of 30 gene flow and loss of genetic variability in ground beetles Proceedings of the Royal Society of 31 London series B, 270, 417-423 32Leberg P (1992) Effects of population bottlenecks on genetic diversity as measured by allozyme 33 electrophoresis Evolution, 46, 477-494 34Martin AP, Wilcox JL (2004) Evolutionary history of Ash Meadows pupfish (genus Cyprinodon) 35 populations inferred using microsatellite markers Conservation Genetics, 5, 769-782 36Mills LS, Allendorf FW (1996) The one-migrant-per-generation rule in conservation and management 37 Conservation Biology, 10, 1509-1518 38Miller RR, Echelle AA (1975) Cyprinodon tularosa, a new cyprinodontid fish from the Tularosa Basin, 39 New Mexico Southwestern Naturalist, 19, 365-377 40Minckley WL (1995) Translocation as a tool for conserving imperiled fishes: experiences in the western 41 United States Biological Conservation, 72, 297-309 42Minckley WL and Deacon JE (1991) Battle against extinction University of Arizona Press, Tucson 43Moran P (2002) Current conservation genetics: building on an ecological approach to the synthesis of 44 molecular and quantitative genetic methods Ecology of Freshwater Fish, 11, 30-55 10 1Moritz C (1994) Defining ‘Evolutionarily Significant Units’ for conservation Trends in Ecology and Evolution, 9, 373-375 3Moritz C (2002) Strategies to protect biological diversity and the evolutionary processes that sustain it Systematic Biology, 51, 238-254 5Moyle, P (1994) The decline of anadromous fishes in California Conservation Biology, 8, 869-870 6Nehlsen W, Williams JE, Lichatowich JA (1991) Pacific salmon at a crossroads: stocks at risk from California, Oregon, Idaho and Washington Fisheries, 16, 4-21 8Neraas LP, Spruell P (2001) Fragmentation of riverine systems: the genetic effects of dams on bull trout (Salvelinus confluentus) in the Clark Fork River system Molecular Ecology, 10, 1153-1164 10Parker K, Sheffer RJ, Hedrick PW (1999) Molecular variation and evolutionary significant units in the 11 endangered Gila topminnow Conservation Biology, 13, 108-116 12Pittenger JS, Springer CL (1999) Native range and conservation of the White Sands pupfish (Cyprinodon 13 tularosa) Southwestern Naturalist, 44, 157-165 14Pritchard JK, Stephens M, Donnelly P (2000) Inference of population structure from multilocus genotype 15 data Genetics, 155, 945-959 16Quattro JM, Leberg PL, Douglas ME, Vrijenhoek RC (1996) Molecular evidence for a unique 17 evolutionary lineage of endangered Sonoran desert fish (genus Poeciliopsis) Conservation 18 Biology, 10, 128-135 19Rogowski, D L 2004 Effects of salinity on the White Sands pupfish Ph.D dissertation, North Dakota 20 State University, Fargo, 186pp 21Smith MW, Smith MH, Chesser RK (1983) Biochemical genetics of mosquitofish I Environmental 22 correlates, and temporal and spatial heterogeneity of allele frequencies within a river drainage 23 Copeia, 1983, 182-193 24Stockwell CA, Kinnison MT, Hendry AP (2006) Evolutionary restoration ecology In: Foundations of 25 Restoration Ecology (eds Falk DA, Palmer MA, Zedler JB) pp.113-137 Island Press 26Stockwell, CA and Leberg PL (2002) Ecological genetics and the translocation of native fishes: emerging 27 experimental approaches Western North American Naturalist, 62, 32-38 28Stockwell CA, Mulvey M (1998) Phosphogluconate dehydrogenase polymorphism and salinity in the 29 White Sands pupfish Evolution, 52, 1856-1860 30Stockwell CA, Mulvey M, Jones AG (1998) Genetic evidence for two evolutionarily significant units of 31 White Sands pupfish Animal Conservation, 1, 213-225 32Stockwell, C.A M Mulvey, and G.L Vinyard 1996 Translocations and the preservation of allelic 33 diversity Conservation Biology 10: 1133-1141 34Storfer A (1999) Gene flow and endangered species translocations: a topic revisited Biological 35 Conservation, 87, 173-180 36Templeton AR, Shaw K, Routman E, Davis SK (1990) The genetic consequences of habitat 37 fragmentation Annals of the Missouri Botanical Garden, 77, 13-27 38Vrijenhoek RC (1996) Conservation genetics of North American desert fishes In: Conservation 39 genetics; case studies from nature (eds Avise JC, Hamrick JL) pp.367-397 Chapman & Hall, 40 New York 41Waples RS (1991) Pacific salmon, Oncorhynchus spp., and the definition of "species" under the 42 Endangered Species Act Marine Fisheries Review, 53, 11-22 43Waples RS (1995) Evolutionary significant units and the conservation of biological diversity under the 44 Endangered Species Act In Evolution and the aquatic ecosystem: defining unique units in 11 population conservation (ed Nielson JL) pp 8-27 American Fisheries Society Symposium 17, Bethesda, MD 3Westemeier RL, Brawn JD, Simpson SA, Esker TL, Jansen RW, Walk JW, Kershner EL, Bouzat JL, Paige KN (1998) Tracking the Long-Term Decline and Recovery of an Isolated Population Science, 282, 1695-1698 6Wright S (1970) Random Drift and the Shifting Balance Theory of Evolution In: Mathematical Topics in Population Genetics (ed Kojima K) pp 1-31 Springer Verlag, Berlin 8Wright S (1931) Evolution in Mendelian populations Genetics, 16, 97-159 9Yamamoto S, Morita K, Koizumi I, Maekawa K (2004) Genetic differentiation of white-spotted charr 10 (Salvelinus leucomaenis) populations after habitat fragmentation: spatial-temporal changes in gene 11 frequencies 12 13 14 15Acknowledgements 16Special thanks are due to David Layfield for his assistance with collecting these data The authors would 17also like to thank Jeanne Dye and Hildegard Reiser(CES/CEV, Holloman AFB) and Robert Myers 18(Environmental Stewardship, Environmental Division in the Directorate of Public Works, WSMR), for 19arrangement of range visitation Janice Terfehr assisted with collecting fish This research was funded by 20DOD Legacy Resource Program Grant no DACA87-00-H-0014 administered by H Reiser and J Dye, 21(CES/CEV, Holloman AFB) and North Dakota EPA-STAR EPSCoR Grant to CAS Pupfish were 22collected on White Sands Missile Range under New Mexico State collecting permit 2887 R Myers, S 23Carman and Y Chen commented on an earlier version of this manuscript 12 1Figure Legends 2Figure Map of Cyprinodon tularosa habitats Solid lines represent permanent barriers to fish movement, while dashed lines represent barriers that not block fish movement during high precipitation events Modified from Stockwell et al 1998 5Figure Population assignment for C tularosa A STRUCTURE analysis for all sampled C tularosa, where individuals were assigned Bayesian likelihood scores for belonging to one of four populations LR = Lost River, ML = Malpais Springs, MD = Mound Spring, SC = Salt Creek 8Figure Population assignment within the Salt Creek ESU A STRUCTURE analysis for individuals sampled from the Salt Creek ESU, where individuals were assigned Bayesian likelihood scores for 10 belonging to one of three populations Drainage designations as per Figure 11Figure Percent of polymorphic alleles with significant barrier effects Barrier effects were considered 12 present when populations were significantly differentiated by exact tests of sample differentiation 13 in Arlequin 14Figure Genotype assignment assessing barrier effects Assignment likelihood scores were plotted to 15 determine how well individuals were assigned to their subpopulation of origin In all graphs, grey 16 triangles refer to individuals sampled from the lower subpopulation, while black circles indicate 17 individuals sampled from the upper subpopulation a) Malpais Spring, b) Salt Creek, c) Lost River, 18 d) Mound Spring 19 13 10 11 12 13 14 15 16 17 18 19 20 21 Table Table Primer information and running conditions for loci used to examine population structure in Cyprinodon tularosa Complete motifs for loci with compound microsatellites available from the authors upon request * = In addition to a compound microsatellite, fish from Malpais Spring have a deletion in the flanking region (see Jones et al 1998) Locus Motif [fwd primer](uM) Annealing T cycles AC23 (CA)n 0.2 50;53 5;30 C509 (CA)n 0.2 60 30 GATA2 WSP11 WSP2 (GATA)n Compound* Compound 0.4 0.4 0.2 50;53 61.3 55 5;30 30 30 WSP20 (GATA)n 0.2 61 32 WSP23 (TG)n-G-(GT)n 0.04 52 40 WSP24 WSP25 (CA)n Compound 0.2 0.02 55 55 32 32 WSP26 (CA)3(AC)n 0.02 50 32 WSP30 (CA)n 0.02 59 32 WSP32 (CA)n 0.02 59 40 WSP33 (GT)n 0.04 55 32 WSP34 (TG)n 0.04 61 32 14 Table 2a Table 2a Results of the locus-by-locus AMOVA performed on all populations of Cyprinodon tularosa 9Source of Variation 10 11Among populations 12Among sub-populations 13 /within populations 14Within sub-populations 15Total 16Fixation indices 17FSC 18FST 19FCT 20 Sum of Squares Variance Percentage Components of variation 709.514 1.44248 37.096 22.831 1519.800 2252.145 0.04129 2.40475 3.88852 1.062 61.84 0.01688 0.38158 0.37096 15 Table 2b Table 2b Results of the locus-by-locus AMOVA performed on of the Salt Creek ESU (Salt Creek, Lost River, and Mound Spring populations) of Cyprinodon tularosa 9Source of Variation 10 11Among populations 12Among sub-populations 13 /within populations 14Within sub-populations 15Total 16Fixation indices 17FSC 18FST 19FCT 20 Sum of Squares Variance Percentage Components of variation 42.854 0.09719 3.98 17.631 1089.562 1150.048 0.04473 2.29866 2.44057 1.83 94.19 0.01909 0.05815 0.03982 16 Table Table Results of exact tests of differentiation, reported as p-values, within the Salt Creek ESU Bolded p-values area significant at α = 0.05 LRUP MDLW MDUP SCLW LRLW 0.00000+-0.0000 0.32810+-0.0474 0.10075+-0.0214 0.02470+-0.0089 LRUP MDLW MDUP 0.10230+-0.0255 0.06430+-0.0177 0.01355+-0.0136 0.37860+-0.0383 0.54655+-0.0394 0.85890+-0.0562 17 Table 4 Table Loss of alleles in the Salt Creek ESU, as compared to the founding (SCLW) population a = one allele present that was absent in SCLW, b = Not the same alleles as missing in LRLW, c = The missing allele had a frequency of 1.5% at SCLW AC23 LRLW LRUP 1 b MDLW MDUP a SCUP GATA02 a a 2 a WSP11 0 0 WSP02 WSP23 2 a a 18 WSP24 c c c c 1 1 c WSP25 WSP30 WSP33 WSP34 a 0 0 0 0 a 0 1 MLUP MLLW SCUP SCLW LRUP LRLW MDUP MDLW Table Table Number of alleles per locus AC23 6 6 C509 2 1 1 1 GATA02 11 12 10 10 11 13 WSP11 2 2 2 WSP02 3 4 WSP20 1 1 1 WSP23 2 3 WSP24 2 2 19 WSP25 3 2 2 WSP26 2 1 1 1 WSP30 2 1 1 WSP32 2 1 1 1 WSP33 1 2 2 2 WSP34 3 3 AVERAGE 3.43 3.64 2.79 3.36 2.64 2.86 3.00 3.00 Figure 1 20 Figure 12 LR ML MD 21 SC Figure LR MD 22 SC Figure 23 Figure 24 ... 1998) 9Molecular data can also be useful for evaluating the effects of historic transplants and habitat 1 0fragmentation on genetic variation Historical translocation events can be important drivers... units, translocations and habitat fragmentation are of particular 29importance to the conservation of western fishes (Minckley and Deacon 1991; Waples 1991; Minckley 301995; Allendorf and Waples... translocations and habitat fragmentation on genetic variation 6in this species 8Background 9Malpais Spring and Salt Creek, which harbor native populations of C tularosa (Pittenger and Springer 101999), are