Mitochondrial differentiation in a polymorphic land snail: evidence for Pleistocene survival within the boundaries of permafrost M. HAASE,* B. MISOF, T. WIRTH,* H. BAMINGER* & B. BAUR* *Conservation Biology Group, Department of Integrative Biology, University of Basel, Basel, Switzerland Department of Entomology, Zoological Research Institute and Museum Alexander Koenig, Bonn, Germany
Keywords:
Abstract
AMOVA;
The genetic differentiation of populations having colonized formerly unsuitable habitats after the Pleistocene glaciations depends to a great extent on the speed of expansion. Slow dispersers maintain their refugial diversity whereas fast dispersal leads to a reduction of diversity in the newly colonized areas. During the Pleistocene, almost the entire current range of the land snail Arianta arbustorum has repeatedly been covered with ice or been subjected to permafrost. Owing to the low potential for dispersal of land snails, slow (re)colonization of the wide range from southern refugia can be excluded. Alternatively, fast, passive dispersal from southern refugia or survival in and expansion from multiple refugia within the area subjected to permafrost may account for the current distribution. To distinguish between these scenarios we reconstructed a phylogeography based on the sequences of a fragment of the cytochrome oxidase I from 133 individuals collected at 45 localities and analysed the molecular variance. Seventy-five haplotypes were found that diverged on average at 7.52% of positions. This high degree of diversity suggests that A. arbustorum is an old species in which the population structure, isolation and the hermaphroditic nature have reduced the probability of lineage extinction. The genetic structure was highly significant with the highest variance partition found among regions. Geographic distance and mitochondrial differentiation were not congruent. Lineages had overlapping ranges. The clear genetic differentiation and the patchy pattern of haplotype distribution suggest that colonization of formerly unsuitable habitats was mainly achieved from multiple populations from within the permafrost area.
Arianta arbustorum; dispersal; mitochondrial DNA; phylogeography; Pleistocene; polymorphism; refugia.
Introduction It is now widely recognized that most extant species in the northern hemisphere have undergone great alterations in their ranges due to habitat shifts induced by glacial periods (e.g. Riddle, 1996; Jaarola et al., 1999; Hewitt, 1999, 2000). Species went extinct over large parts of their range, some dispersed to new localities, some Correspondence and present address: Martin Haase, National Institute of Water and Atmospheric Research, PO Box 11-115, Hamilton, New Zealand. Tel.: +64 7 856 1737; fax: +64 7 856 0151; e-mail:
[email protected] Present address: Thierry Wirth, Max-Planck Institute for Infection Biology, Schumannstr. 21/22, D-10117 Berlin, Germany.
survived in refugia and then expanded again. Using molecular data, the phylogeographic approach has yielded great insights into the recent evolutionary history of many species (e.g. Avise, 1992, 1994, 2000; Taberlet et al., 1998; Ferris et al., 1999; Hewitt, 1999, 2000). Several molecular studies provide evidence that many species (re)colonized central and northern Europe from one or several southern refugia (for reviews see Hewitt, 1999, 2000). However, although equally important for our understanding of the composition of the present day faunas and floras and although there is increasing evidence for northern refugia (Stewart & Lister, 2001), the recent spread of animal and plant species having survived cold periods within the area affected by
J. EVOL. BIOL. 16 (2003) 415–428 ª 2003 BLACKWELL PUBLISHING LTD
415
416
M. HAASE ET AL.
Fig. 1 Three nominal subspecies of Arianta arbustorum: (A) A. a. arbustorum (Swiss Jura Mountains: Mont Raimeux); (B) A. a. picea (central Austria: Leobener Te¨rl); (C) A. a. styriaca (central Austria: RotenederSeitengraben). Scale bar ¼ 1 cm.
permafrost has received little attention (rare examples are provided by Gabrielsen et al., 1997; Fedorov & Stenseth, 2001; Stehlik et al., 2001). It has been suggested that the pulmonate land snail Arianta arbustorum (Fig. 1) has survived the last glaciation in refugia in central Europe, the Alps and on the northwest coast of Norway (Lozek, 1964; Ant, 1966, 1969; Walde´n, 1986; Do¨ppes & Rabeder, 1997). This hypothesis, however, is based on a small number of fossil findings in areas where the conditions for fossilization were appropriate. In addition, it is practically impossible to tell whether a particular population, of which there are fossil remains, survived or was even the source for other and new populations, or whether it became extinct. The current distribution of A. arbustorum ranges from the eastern Pyrenees throughout central Europe, the Alps and the Carpathians to the Ukraine, and populations occur on the British islands, Iceland and in Scandinavia. This area largely coincides with that part of Europe that was repeatedly covered with ice or subjected to permafrost during the Pleistocene ice ages (Hewitt, 1999). Nowadays, A. arbustorum occurs in moist habitats from the lowlands up to an elevation of 3000 m in the Alps (Ehrmann, 1933; Kerney et al., 1983; Fechter & Falkner, 1989). A. arbustorum is very variable with regard to shell shape and size and more than 20 nominal infraspecific taxa have been described (cf. Ehrmann, 1933; Baminger, 1997; see below). In general, the rate of expansion after glaciation appears to depend on the reproductive and dispersal capabilities of the organism apart from abiotic parameters such as the sharpness of the climatic change, the latitude (length of vegetation period) and the topography of the region (Hewitt, 1999). Individuals of A. arbustorum become sexually mature at 2–4 years with a shell width of 16–28 mm. Adults live another 3–4 years (Baur & Raboud, 1988). In the field, snails deposit one to three batches of 20–50 eggs per reproductive season (Baur &
Raboud, 1988; Baur, 1990). Paternity analysis in broods of wild-caught A. arbustorum showed a high frequency of multiple insemination (Baur, 1994). Selfing is rare in this simultaneous hermaphrodite and may occur after longterm isolation (Chen, 1993). The reproductive success of selfing individuals, however, is very low (Chen, 1994). The potential for active dispersal is low with average distances covered per year of 12 m in an alpine grassland (Baur, 1986). Wider stretches may be covered through downhill rolling and passive transportation (Baur et al., 1997). Streams, rivers and avalanches are the most important agents (Baur, 1986, 1993). Downhill displacement may be compensated for by negative geotactic orientation behaviour (Baur & Gosteli, 1986). Aerial dispersal through wind must be considered especially for juvenile snails (Ant, 1963; Vagvolgyi, 1975; Kirchner et al., 1997). Repeated restrictions of gene flow, bottlenecks and extended (re)colonization of habitats leave footprints in the genetic variation of populations. Simulations predict that slowly dispersing species maintain their original genetic variability in newly colonized habitats, whereas populations founded by leptokurtic dispersers are genetically depleted and homogeneous (Ibrahim et al., 1996). Several empirical investigations have confirmed these predictions for species having quickly recolonized Europe from southern refugia after the last ice age (e.g. Taberlet et al., 1998; Hewitt, 1999). Furthermore, species, which survived the last ice age in several southern European refugia and quickly expanded north after the retreat of the glaciers, formed well-defined hybrid zones in areas where the different lineages met (e.g. Taberlet et al., 1998; Hewitt, 1999; Jaarola et al., 1999). The current range of A. arbustorum is too wide to be actively (re)colonized from southern refugia within less than 18 000 years, the time since the peak of the last glaciation, because of the low potential for active dispersal. Furthermore, slow, passive dispersal from southern
J. EVOL. BIOL. 16 (2003) 415–428 ª 2003 BLACKWELL PUBLISHING LTD
Mitochondrial differentiation in a land snail
refugia cannot account for the current distribution. However, the following scenarios could explain the present-day distribution of this and other species with similar distributions and dispersal characteristics. If A. arbustorum has exclusively survived in areas south of the permafrost-affected regions with strong reduction of population sizes and subsequently rapidly expanded northwards by passive transportation, we would expect (i) little genetic differentiation in northern and/or alpine (re)colonized areas, and (ii) geographically separated lineages (‘southern refuge scenario’). On the other hand, if refugia within the permafrost area played an important role in the (re)colonization of its current range as paleontological findings suggest (see above), genetic differentiation should be more patchy and pronounced (provided that the refugial populations were differentiated). We would not expect strict congruence of phylogeny and geography as a consequence of recurrent range shifts of different phylogenetic lineages (cf. Hewitt, 1996) (‘persistence scenario’). In order to find evidence for either of these scenarios for A. arbustorum, we conducted a phylogenetic analysis based on a fragment of the cytochrome oxidase subunit I (COI) from individuals collected over a major part of the species’ range and established the geographic distribution of the phylogenetic lineages. In addition, we inferred the genetic structure of the populations by analysis of molecular variance (Excoffier et al., 1992). Based on these results, we interpret the evolutionary history of these phylogeographic units.
Material and methods Sampling and outgroup selection One hundred and thirty-three specimens of A. arbustorum were collected at 45 localities in Austria, Switzerland, Germany, Romania, Denmark and Sweden (Fig. 2) between 1997 and 1999. These samples include the nominate subspecies A. a. picea and A. a. styriaca. The most common and widespread nominate subspecies has a globular shell and a covered umbilicus (Fig. 1A; Bisenberger, 1993; Baminger, 1997). A. a. picea has also a globular shell, but is darker in coloration, more fragile and has relatively rounded whorls (Fig. 1B). This subspecies occurs mainly, but not exclusively, on crystalline mountain ranges in the eastern Alps of Austria (Klemm, 1974). A. a. styriaca is characterized by a very flat shell and an open umbilicus (Fig. 1C; for a more comprehensive diagnosis see Baminger, 1997). It inhabits mainly the rocky rift valleys of the Gesa¨use mountains in the calcareous Ennstaler Alpen of central Austria (Baminger, 1997). Ecotypes of A. a. arbustorum from central Austria identified by Baminger (1997), which may have flatter shells than the common form but still clearly differ from A. a. styriaca in shape and other traits, were not further distinguished.
417
A. arbustorum is highly variable. Based on shell characteristics, more than 20 nominal infraspecific taxa have been described (e.g. Clessin, 1882; Taylor, 1882; Ehrmann, 1933). Much of the variation can be explained by phenotypic plasticity (e.g. Baur, 1984; Burla, 1984) and thus many taxonomic distinctions are probably unjustified. Some taxa including A. a. picea and A. a. styriaca, however, have geographically confined ranges and may, therefore, be regarded as valid subspecies. We included these subspecies into our analysis, because they may be the result of Pleistocene isolation and thus informative for our question. As outgroup we used specimens from three populations of Helicigona lapicida, which belongs to the same subfamily, Ariantinae, as Arianta, and from two populations of Cepaea nemoralis, which belongs to the Helicinae. We did not include another species of Arianta because they all have a much smaller range than A. arbustorum and occur either in sympatry, peripatry or parapatry with the latter (cf. Grossu, 1983; Kerney et al., 1983; Gittenberger, 1991). Thus, it is very likely that A. arbustorum is ancestral to at least some of these species (but see Gittenberger, 1991). Details on sampling localities are listed in Table 1. Snails were brought to the laboratory alive where they were frozen at –80 C. DNA preparation, amplification and sequencing DNA was extracted from muscle tissue of the foot following van Moorsel et al. (2000). A fragment of COI of 663 base pairs length was amplified using the primers ACO (heavy strand) 5¢-CCTATTATAATTGGGGGTTTTG G-3¢ and BCO (light strand) 5¢- GTATCGGCTGTAAAATAAGC-3¢. Primers were designed from an alignment of the COI sequences of C. nemoralis and Albinaria coerulea, of which the whole mitochondrial genomes are known (Terrett et al., 1994; Hatzoglou et al., 1995). Each PCR reaction was carried out in a total volume of 16 ll containing 10 ng template DNA, 0.5 lm of each primer, 0.3 mM of each dNTP, 4 mM MgCl2, 1.5· Mg-free buffer and 0.75 units Taq polymerase (Promega). The PCR conditions were an initial denaturation for 5 min at 95 C followed by 40 cycles comprising denaturation for 1 min at 95 C, annealing for 1 min at 52 C and extension for 2 min at 72 C. The final step was an additional extension for 7 min at 72 C. The amplifications were conducted in a PTC-100 Thermocycler (MJ Research). Ethidium bromide visualized PCR products were excised from agarose gels, purified using the QIAquick purification kit (QIAGEN) following the protocol of the manufacturer with slight modifications (Locher, pers. comm., 1999) and directly sequenced from both directions on an ABI 310 sequencer (PerkinElmer). Cycle sequencing reactions were performed according to the manufacturer’s protocol (Perkin-Elmer) with some modifications. We used only half of the suggested amounts and half of the dRhodamine
J. EVOL. BIOL. 16 (2003) 415–428 ª 2003 BLACKWELL PUBLISHING LTD
418
M. HAASE ET AL.
Fig. 2 Map of central Europe showing localities of sampled Arianta arbustorum populations and distribution of phylogenetic lineages I–VI (grey subscripts or superscripts). The inset shows central Austria at a higher resolution. AC, central Austria; AE, eastern Austria; AV, Vienna (Austria); DA, Denmark; GE, Germany; RO, Romania; SC, central Switzerland; SE, eastern Switzerland; SJ, Swiss Jura Mountains; SN, northern Switzerland; SW, Sweden. Scales ¼ 500 km (Europe) and 50 km (inset), respectively.
Terminator Ready Reaction mix was replaced by half TERM (GENPAK). Sequence analysis Sequences were aligned by eye in the editor of PAUP* 4.0b4a (Swofford, 1998). Analysis of sequence polymorphism and tests for neutral evolution were carried out with DnaSP 3.14.3 (Rozas & Rozas, 1999) and Arlequin 2001 (Schneider et al., 2000). Saturation of substitutions was tested by plotting the absolute number of pairwise differences at third against first and second positions. Translation into amino acid sequences was done using Sequencer 4.0 (Kessing, 1999). The v2-test implemented in PAUP* 4.0b4a with exclusion of all parsimony uninformative characters (cf. Lockhart et al., 1994, 1995; Misof et al., 2001) revealed significant heterogeneity of base frequencies among the haplotypes of A. arbustorum (v2222 ¼ 279.25, P ¼ 0.0055) ignoring correlations due to phylogenetic structure. This result indicates that COI has evolved according to more than one substitution model within A. arbustorum. Thus, standard maximum likelihood reconstructions could not be applied. Relationships of all Arianta haplotypes were reconstructed applying maximum parsimony (MP) with
equally weighted characters and minimum evolution (ME) based on LogDet distances using PAUP* 4.0b4a. LogDet reconstructions are insensitive to inhomogeneous base frequencies across taxa in contrast to parsimony methods (Lockhart et al., 1994, 1995; Swofford et al., 1996). Parsimony analysis was performed after exclusion of uninformative characters with heuristic searches and TBR branch swapping. The starting trees were obtained via stepwise addition. LogDet reconstructions starting from neighbour-joining trees removed invariable sites in proportion to the base frequencies estimated from constant sites (Waddel & Steel, 1997). The proportion of invariant sites estimated after fitting a GTR + I + G model of substitution (Rodriguez et al., 1990) on the data using Modeltest 3.0 (Posada & Crandall, 1998) was 0.558. Again, searches for the shortest tree were heuristic applying TBR branch swapping. Robustness of the reconstructions was estimated by bootstrapping with 1000 replicates. Due to the large size of the data set, we applied NNI branch swapping in heuristic searches and set MaxTrees to 50. In the case of the geographically very close central Austrian Gesa¨use populations, it is plausible that ancestral populations still exist and, therefore, a haplotype network might represent relationships better than a tree
J. EVOL. BIOL. 16 (2003) 415–428 ª 2003 BLACKWELL PUBLISHING LTD
Mitochondrial differentiation in a land snail
419
Table 1 Localities and abbreviations of samples. In Arianta arbustorum, the abbreviations indicate region, locality and subspecies (a, arbustorum; p, picea; s, styriaca; as, hybrids between the nominate form and A. a. styriaca); in outgroup samples, genus species and locality. From each population of A. arbustorum three specimens were sequenced except for Rotenedergraben and Lake Bıˆlea where only two specimens were available. From Rotenedergraben hybrid zone we had four snails. Outgroup samples comprised one to three individuals (N). Locality
Geographical position
Elevation (m a.s.l.)
Abbreviation
Arianta arbustorum Central Austria (AC) Buckletschneidergraben Fehringerkreuzgraben Glanegg Grabnerhof Grabnerstein Gsenggraben Gstatterboden Haindlkar brook bed Haindlkar hut Above Haindlkar hut Heiliger Brunnen Hochtor Kainzenalblgraben Ko¨lbl-Wirt Langgriesgraben Leobener To¨rl Path to Leobener Oberhof Petergstammgraben Reichenstein Rotenedergraben Roteneder-Seitengraben Rotenedergraben hybrid zone Schloss Kaiserau Teufelsarsch Totes Gebirge
4732.9¢N, 4732.7¢N, 4732.4¢N, 4735.6¢N, 4737.9¢N, 4734.0¢N, 4735.5¢N, 4733.8¢N, 4733.8¢N, 4733.8¢N, 4729.8¢N, 4733.6¢N, 4733.8¢N, 4731.9¢N, 4733.4¢N, 4729.8¢N, 4729.8¢N, 4734.0¢N, 4733.2¢N, 4732.9¢N, 4732.9¢N, 4732.8¢N, 4732.7¢N, 4731.8¢N, 4733.4¢N, 4742.1¢N,
1435.0¢E 1435.0¢E 1440.1¢E 1429.6¢E 1430.1¢E 1436.0¢E 1438.2¢E 1437.0¢E 1436.9¢E 1436.8¢E 1440.0¢E 1438.3¢E 1435.5¢E 1437.0¢E 1433.9¢E 1439.8¢E 1439.6¢E 1427.8¢E 1435.4¢E 1432.7¢E 1435.3¢E 1435.3¢E 1435.2¢E 1429.0¢E 1439.8¢E 1403.3¢E
700 780 1800 630 1805 940 590 1120 1120 1140 1780 2260 840 870 980 1730 1730 680 800 1940 730 750 720 1130 1750 1550
AC-Bg-s AC-Fk-s AC-Gl-a AC-Gh-a AC-Gr-a AC-Gs-s AC-Gt-a AC-Hb-s AC-Hh-a/s/as AC-Hk-s AC-Hr-p AC-Ht-a AC-Ka-s AC-Ko-a AC-La-s AC-Le-p AC-Lp-p AC-Oh-a AC-Pg-s AC-Re-a AC-Rg-s AC-Rs-s AC-Ry-a/s/as AC-Sk-a AC-Ta-a AC-Tg-a
Eastern Austria (AE) Puchberg am Schneeberg
4747.8¢N, 1551.5¢E
800
AE-Ps-a
Austria: Vienna (AV) Friedhof der Namenlosen Klosterneuburger Au
4809.3¢N, 1630.5¢E 4818.8¢N, 1619.8¢E
155 165
AV-Fn-a AV-Kl-a
Denmark (DA) Lyngby
5547.5¢N, 1230.6¢E
50
DA-Ly-a
Germany (GE) Grafenmatt Zastler Tal Romania: South Carpathians (RO) Lake Bıˆlea Sweden (SW) Fiskeba¨ckskil
4751.1¢N, 801.9¢E 4753.1¢N, 800.5¢E
1270 1070
GE-Gm-a GE-Za-a
4536.0¢N, 2430.0¢E
2200
RO-Lb-a
5815.0¢N, 1127.8¢E
15
SW-Fi-a
Central Switzerland (SC) Gurnigel
4645.4¢N, 727.3¢E
1320
SC-Gu-a
Eastern Switzerland (SE) Bollenwees Hinter Selun Hoher Kasten Schiawang Strela Below Zwinglipass hut
4715.3¢N, 4709.7¢N, 4717.1¢N, 4648.9¢N, 4649.0¢N, 4714.0¢N,
1470 1680 1680 2350 2270 1950
SE-Bw-a SE-Hs-a SE-Ho-a SE-Si-a SE-St-a SE-Zw-a
925.6¢E 914.3¢E 929.1¢E 948.5¢E 947.7¢E 922.6¢E
J. EVOL. BIOL. 16 (2003) 415–428 ª 2003 BLACKWELL PUBLISHING LTD
420
M. HAASE ET AL.
Table 1 (Continued) Locality
Elevation (m a.s.l.)
Abbreviation
Switzerland: Jura Mountains (SJ) Mont Raimeux 4718.6¢N, 725.8¢E Weissenstein 4714.7¢N, 727.1¢E
1300 1330
SJ-Mr-a SJ-We-a
Northern Switzerland (SN) Baldegg Nuglar
4728.3¢N, 817.0¢E 4728.7¢N, 741.9¢E
560 430
SN-Ba-a SN-Nu-a
Species
Locality (region)
Position
Elevation (m a.s.l.)
Lyngby (DA) Schinznach Bad (SN)
5547.5¢N, 1230.6¢E 4727.5¢N, 810.1¢E
Fiskeba¨ckskil (SW) Flu¨h (SN) Mont Raimeux (SJ)
5815.1¢N, 1128.0¢E 4729.4¢N, 730.2¢E 4718.0¢N, 725.9¢E
Outgroup Cepaea nemoralis
Helicigona lapicida
Geographical position
(e.g. Crandall & Templeton, 1996). Thus, for a clade consisting of 27 haplotypes (see Results), whose base frequencies were homogeneous, a haplotype network based on TrN + I distances (Tamura & Nei, 1993) was generated using the program MINSPNET (Excoffier, 1993). This substitution model was selected by the program Modeltest 3.0 and the proportion of invariant sites was estimated to be 0.814. In order to compare rates of substitution among lineages relative-rate tests were conducted using the program RRTree 1.1 (Robinson et al., 1998). We defined six lineages according to the tree reconstructions and computed three (Ks number of synonymous substitutions per synonymous site; As number of synonymous transitions per synonymous site; B4, number of synonymous transversions per four-fold degenerate site) of the five substitution parameters implemented in the program without a guiding tree. There were too few synonymous substitutions per nonsynonymous site to compute also Ka (all substitutions) and Ba (transversions). As outgroup we used the three H. lapicida haplotypes. In a further analysis, the Arianta haplotypes from the Grabnerstein, which were part of the most basal ingroup-clade in our phylogenetic reconstructions, were the outgroup for a comparison of the remaining five lineages. Correlation of phylogenetic differentiation and geographic distances between populations was tested in two ways. We compared (i) the length of the MP trees with the lengths of trees in which the individuals from certain regions were constrained to be monophyletic using Templeton tests implemented in PAUP* 4.0b4a and (ii) matrices of pairwise FST values and geographic distances by Mantel tests (1000 permutations; negative FST values were replaced by 0) with Arlequin 2001. The analysis of molecular variance (A M O V A , 16 000 permutations) grouping the populations from 11 regions (Fig. 2, Table 1) was also conducted with Arlequin 2001.
Abbreviation
N
50 350
CN-Ly CN-Sb
2 3
10 440 1200
HL-Fi HL-Fl HL-Mr
2 3 1
Results Sequence polymorphism There were 75 haplotypes (TreeBASE accession number S842) among the 133 individuals of A. arbustorum sequenced. The subspecies of A. arbustorum are not treated separately in this section, because they did not form monophyletic groups (see below). Table 2 groups individuals from different localities with identical haplotypes. All but groups II and VII consist of individuals from central Austria. Group II contains one specimen from central Austria and one from Vienna; group VII contains eastern Swiss individuals. Within A. arbustorum, haplotypes differed at 1–88 positions (0.15–13.27%). The average number of pairwise nucleotide differences was 49.85 (7.52%), which is biased due to the overrepresentation of closely related samples from central Austria. The maximum genetic distances were found between an individual from Vienna (AV-Fn-a2) and populations in the central Austrian Gesa¨use mountains (AC-Bg-s1-3, AC-Re-a2, AC-Rg-s2,3). Eleven samples Table 2 Groups of specimens with identical DNA sequences (abbreviations as in Table 1). Group
Specimens with identical sequences
Regions
I II III IV
AC-Fk-s2; AC-Hh-as3; AC-Ry-a4 AC-Oh-a2; AV-Fn-a1 AC-Gt-a2; AC-Oh-a3 AC-Gl-a1-3; AC-Hb-s3; AC-Hk-s1,2; AC-Hr-p1,2; AC-Ht-a1-3; AC-La-s2; AC-Le-p2,3; AC-Lp-p1,3; AC-Rs-s1; AC-Ta-a1 AC-Gs-1-3; AC-Hk-s3 AC-Pg-s2; AC-Rs-s2; AC-Ry-s1; AC-Ry-as3 SE-Ho-a3; SE-Hs-a2 AC-Le-p1; AC-Lp-p2 AC-Gt-a3; AC-Oh-a1
AC AC, AV AC AC
V VI VII VIII IX
AC AC SE AC AC
J. EVOL. BIOL. 16 (2003) 415–428 ª 2003 BLACKWELL PUBLISHING LTD
Mitochondrial differentiation in a land snail
were monomorphic; in 18 samples haplotypes differed at 1–6 positions (0.15–0.91%) and at 16 localities haplotypes diverging at 16–79 positions (2.41–11.92%) were found. Within the outgroup, four haplotypes were found in C. nemoralis and three in H. lapicida (TreeBASE S842). The maximum number of differing positions was 15 (2.26%) and 11 (1.66%), respectively. The number of different sites between in- and outgroup ranged from 109 to 155 (16.44–23.38%). Four hundred and sixty-two sites (69.68%) were monomorphic within A. arbustorum. Of the 201 (30.32%) polymorphic sites, 168 (25.34%) were parsimony informative with only 16 positions occupied by all four bases. The total number of synonymous and replacement changes were 263 and 17, respectively. In nine cases, an amino acid change was due to a replacement on a first position, on a second position in seven cases and on a third position in a single case. The data set including the outgroup had 252 (38.00%) polymorphic sites of which 234 (35.29%) were parsimony informative. Base composition and nonstationarity will be discussed elsewhere. Plots of the absolute number of pairwise differences at third against first and second codon positions showed a moderate degree of saturation in third positions, especially so in comparisons between in- and outgroup and between outgroup genera. The coding region comprised positions 3–662. Thus, on the protein level the sequenced COI fragments were 220 amino acids long. Within A. arbustorum, there were only 14 haplotypes with 15 (6.82%) polymorphic sites. The maximum difference between haplotypes was 7 (3.18%) amino acids. Tree reconstruction The MP analysis yielded 8448 equally parsimonious trees (TL ¼ 765, CI ¼ 0.486, RI ¼ 0.878 and RC ¼ 0.427). The majority rule consensus tree is given in Fig. 3. The bootstrap support is considerable for most clades. Only the relationships between monophyletic groups from Switzerland, Germany, Romania and the periphery of central Austria, and Vienna were unresolved in the bootstrap consensus tree. In the majority rule consensus tree, neither the samples from Austrian nor Swiss localities turned out as monophyletic groups in accordance with expectations from geographical positions. Six lineages were identified. The geographical range of each lineage overlapped with that of at least one other lineage with one exception. The distribution of these six lineages across regions is illustrated in Fig. 2. The most basal clade comprised 27 haplotypes from the central Austrian Gesa¨use mountains. The samples from Germany were outgroup to the remaining haplotypes. A lineage formed by individuals from Vienna and three Swiss regions was sister clade to the three most derived clades. Among these, a monophyletic group consisting of the Romanian
421
specimens, one individual from Vienna and localities from the periphery of the region sampled in central Austria was sister group of two geographically peculiar clades. Two haplotypes of snails belonging to A. a. styriaca and one from the Teufelsarsch (nominate ssp.) grouped with individuals from Vienna and the Swiss North, East and Jura Mountains. The sister clade of this group comprised the samples from Sweden, Denmark, central Switzerland and eastern Austria. The ME analysis based on LogDet distances produced a single optimal tree with a score of 1.63316. The bootstrap support was in general lower than for MP. This analysis principally yielded identical groupings of haplotypes as MP (therefore not shown). These were different only in the relationships between the groups and within the clade comprising specimens from Vienna, central Austria and three Swiss regions defined through its sister group, which contains the samples from Denmark and Sweden. The 27 Gesa¨use haplotypes were sister group to a large clade divided into several subclades comprising the remaining samples. A group consisting of the Romanian and the peripheral Gesa¨use individuals plus one from Vienna was sister clade to the group comprising the German specimens, individuals from three Swiss regions and two Viennese haplotypes. This whole clade was in sister relationship to the most derived groups identified in the MP analysis. The high similarity of both tree reconstructions differing in their sensitivity to base composition indicates that the heterogeneity in base composition across populations was hardly confounding. The haplotype network depicting relationships among the 27 haplotypes from central Austria in the most basal clade of the above trees is given in Fig. 4. No ties and hence alternative connections were identified. The most common haplotype IV found in all three subspecies in 17 individuals from 10 localities formed the central node with 11 connections to other haplotypes. This suggests that haplotype IV is the most plesiomorphic one. The three subspecies were not monophyletic and most haplotypes of A. a. styriaca and A. a. picea did not differ from those of the nominate subspecies or differed only in a few positions. Some individuals of A. a. styriaca, however, were closer to specimens from Vienna and Switzerland than to other populations from central Austria. Neutrality and rates of evolution The null hypothesis of neutral evolution within A. arbustorum could not be rejected (Tajima’s D ¼ –0.417; Fu and Li’s test statistics: D * ¼ –0.024, F * ¼ –0.224; with outgroup: D ¼ 0.872, F ¼ 0.359; P > 0.1 in all cases). Using H. lapicida as outgroup, evolutionary rates between all six lineages could only be compared based on B4, the number of synonymous transversions per four-fold degenerate site, because of saturation of other types of substitution. No rate difference was found
J. EVOL. BIOL. 16 (2003) 415–428 ª 2003 BLACKWELL PUBLISHING LTD
422
M. HAASE ET AL.
Fig. 3 Fifty per cent majority rule consensus (MJR) tree. MJR consensus indices/bootstrap support (if >50; single figures represent consensus indices) above branches; roman numerals below branches identify phylogenetic lineages. Abbreviations of OTUs are as given in Tables 1 and 2.
J. EVOL. BIOL. 16 (2003) 415–428 ª 2003 BLACKWELL PUBLISHING LTD
Mitochondrial differentiation in a land snail
423
Fig. 4 Haplotype network of the most basal clade (lineage I) from central Austria identified in tree reconstructions. Bars crossing branches indicate one mutational step each. Abbreviations of OTUs are as given in Tables 1 and 2. Figures in parentheses denote number of localities/number of individuals/ subspecies.
(uncorrected P > 0.09 in all cases). Ks the number of synonymous substitutions per synonymous site, could be computed for all comparisons except those involving the clade containing the Scandinavian haplotypes. Again, no difference in rates was found (uncorrected P > 0.54 in all cases). Using AC-Gr-haplotypes as outgroup for rate comparisons among the remaining five clades and omitting all other sequences of the most basal clade, all three parameters, Ks, As the number of synonymous transitions per synonymous site, and B4, could be computed for all 10 comparisons. Evolutionary rates also did not differ among lineages in these cases (Bonferroni corrected P > 0.1 in all cases). Population structure and geographic differentiation The A M O V A showed that populations of A. arbustorum were significantly differentiated on all three levels of the hierarchical design (Table 3). The highest fraction of genetic variance (47.91%) was found among the 11
regions. The average number of migrants exchanged between two populations (M) inferred from the FST value of 0.733 using the formula M ¼ 0.5(1/FST)1) was 0.182, or one migrant every 5.5 generations. MP trees constraining the haplotypes from central Austria – all or only those from the core of the region – to
Table 3 Summary of analysis of molecular variance. All variance components and fixation indices were highly significant (P < 0.0001). Source of variation Among groups Among populations within groups Within populations Total
d.f.
SS
V
%
Fixation index
10 34
1429.96 982.59
13.55 7.19
47.91 25.43
FCT ¼ 0.479 FSC ¼ 0.488
88 132
663.33 3075.88
7.54 28.27
26.66
FST ¼ 0.733
d.f., degrees of freedom; SS, sum of squares; V, variance; %, percentage of total variation.
J. EVOL. BIOL. 16 (2003) 415–428 ª 2003 BLACKWELL PUBLISHING LTD
424
M. HAASE ET AL.
be monophyletic were significantly longer than the unconstrained tree (Templeton test, P < 0.0001). The same result was obtained when haplotypes from eastern Switzerland and the remaining areas of Switzerland were constrained to be monophyletic. A Mantel test comparing a matrix of pairwise FST values with a matrix of geographic distances over all 45 samples found a high correlation (r ¼ 0.452, P < 0.0001). However, this was due to the overrepresentation of samples from central Austria. Reducing the number of samples from this region randomly to five (the order of magnitude of the number of samples from other regions) no longer indicated isolation by distance (r ¼ 0.134, P ¼ 0.233). Thus, isolation by distance was observed only regionally. On a continental scale there was no correlation between genetic differentiation and geographic distance. This lack of correlation, the geographic overlap of evolutionary lineages, the clear genetic structure between the regions and the low exchange of migrants are in accordance with our expectations from the persistence scenario.
Discussion Sequence divergence COI is assumed to be a slowly evolving locus (e.g. Brown, 1985; Folmer et al., 1994; Simon et al., 1994). With a divergence of up to 13.27 and 3.18% on the nucleotide and amino acid level, respectively, A. arbustorum exhibits one of the highest intraspecific variabilities of COI known so far. Similar values have been reported from comparisons between congeneric species or above the species level (e.g. Marko, 1998; King et al., 1999; Lessios et al., 1999; Medina & Walsh, 2000). We designed our primers from published sequences of whole mitochondrial genomes of two land snail species and did not encounter problems such as unexpected deletions/insertions, frame shifts, stop codons or other sequence ambiguities indicating the presence of nuclear sequences of mitochondrial origin (Zhang & Hewitt, 1996). Therefore, it is unlikely that the variation recorded in A. arbustorum is affected by the presence of nuclear copies of mtDNA genes. The saturation of substitutions among the haplotypes analysed is probably a consequence of the relatively high sequence divergence. Saturation can mislead phylogenetic reconstructions (Felsenstein, 1978), but the degree of genetic divergence is far below the level of 30–40% above which saturation might become problematic (Yang, 1998). The only effect may be lower bootstrap support values in the phylogenetic reconstructions. Thomaz et al. (1996) found extensive variation in a fragment of 16sRNA in C. nemoralis. They discussed five potential, but not mutually exclusive reasons, for the extreme diversity: (i) presence of cryptic sibling species;
(ii) selection; (iii) rapid mitochondrial evolution; (iv) ancient isolation and divergence of populations; and (v) unusually structured or exceptionally large populations. We do not think that our samples represent already differentiated species, because the genetic differentiation is mainly a result of synonymous substitutions. Seventyfive DNA-haplotypes are translated into only 14 proteinhaplotypes, which in most cases differ in one or two amino acids. However, matings of individuals from geographically distant populations showed effects of outbreeding depression indicating incipient reproductive isolation (Baur & Baur, 1992|). We thus may witness incipient speciation within A. arbustorum. We have no indication for selection as driving force for the differentiation of COI within A. arbustorum. Our tests are consistent with a model of neutral evolution. Unfortunately, we cannot directly estimate the time of divergence of the Arianta haplotypes, because we are unable to calibrate the substitution rate. Assuming an average rate of 1–2.8% mitochondrial sequence divergence per million years (My) estimated for mitochondrial genes from other land snails (Murray et al., 1991; Douris et al., 1998), we arrive at a coalescence time of 8–2.8 My before present. The latter figure is well in accordance with the age of the oldest fossils from the late Pliocene (Wenz, 1923). Therefore, an accelerated substitution rate as was assumed for 16sRNA of other land snails (Chiba, 1999; Ross, 1999; Thacker & Hadfield, 2000) does not need to be invoked to account for the present mitochondrial divergence in A. arbustorum. We do not agree with Thomaz et al. (1996) that isolation of populations during glaciations was unimportant for shaping the genetic structure within land snails. Small populations may have suffered from loss of haplotypes through genetic drift or may have even gone extinct. On the other hand, lack of migration during isolation and postglacial expansion might both have contributed to the persistence of lineages (Avise et al., 1984; Hoelzer et al., 1998). Assuming that land snail populations are frequently arranged in stepping-stone patterns and migration among them is limited, Thomaz et al. (1996) argued that time to coalescence may be very long provided that the number of demes is high. Thus, haplotypes may persist for long periods of time (Slatkin, 1991; Hoelzer et al., 1998). This may also hold for A. arbustorum, although the above scenario is an oversimplification. A. arbustorum exhibits a variety of patterns of distribution and dispersal (Baur, 1986, 1993; Arter, 1990; Baur & Baur, 1993), which, at least locally, may result in higher migration rates and thus shorter periods of lineage sorting (Hoelzer et al., 1998). An important factor reducing the extinction risk of lineages might be the hermaphroditic nature of pulmonates. As a result of the predominantly maternal inheritance, mitochondrial evolution in gonochorists is ‘self-pruning’ (Avise et al., 1987). In contrast, in most
J. EVOL. BIOL. 16 (2003) 415–428 ª 2003 BLACKWELL PUBLISHING LTD
Mitochondrial differentiation in a land snail
outcrossing pulmonate land snails including A. arbustorum (cf. Chen, 1993, 1994), mating and sperm transfer are reciprocal. Each animal functions both as a male and a female, and consequently, the effective population size of mtDNA equals the effective population size of individuals (e.g. Hoelzer, 1997), assuming ideal populations. Thus, the population structure with numerous demes in combination with little migration, isolation during glaciations and the hermaphroditic nature of A. arbustorum are probably the main factors accounting for the persistence of the considerably diverse, divergent and diverging COI haplotypes in this rather old species with an origin in the late Tertiary. Samples containing haplotypes that vary at 16 and more positions can hardly be traced to an immediate common ancestor. Thirteen of these 16 samples are from easily accessible localities characterized by increased human activities. This suggests that in the recent past displacement by humans may have had some relevance for the dispersal of land snails. Donnelly & Tavare´ (1986) showed that frequency and age of an allele are positively correlated. The probability of an allele to produce mutational derivatives increases with time. Thus, haplotypes of high frequency and ancient origin usually occur in the interior of networks (Golding, 1987; Excoffier & Langaney, 1989; Crandall & Templeton, 1993). The position of the most common haplotype (IV) in the centre of the network connecting 27 central Austrian haplotypes forming the most basal clade in the trees suggests that it is the most plesiomorphic one. Genetic versus morphological differentiation COI differentiation is not congruent with shell differentiation in A. arbustorum. The three subspecies defined by shell characters do not host exclusive mitochondrial lineages. Furthermore, the most common haplotype, IV, is found in all three of them. A. a. picea and most A. a. styriaca individuals cluster among A. a. arbustorum within the basal clade from the central Austrian Gesa¨use. Obviously, shell differentiation is much younger and faster than that of COI. An alternative explanation would be introgression of COI haplotypes by hybridization combined with selection on shell shape. This probably holds also for A. a. styriaca individuals belonging to more derived clades. However, the present data do not allow the exclusion of either alternative explanation. Assuming that the subspecies are good evolutionary species in statu nascendi, we witness the fixation of an ancestral polymorphism or introgression of old haplotypes. Analyses of nuclear genes might help to clarify this issue. Genetic differentiation in space and time The present distribution of COI haplotypes hardly reveals the geographic pathways along which the gene has
425
changed and dispersed. Lineages were geographically overlapping, trees constraining samples from certain regions to be monophyletic were significantly longer than the most parsimonious trees, genetic and geographic distances were not correlated and the populations were genetically significantly differentiated with the highest variance partition found among regions. Only on a local scale may genetic differentiation be in accordance with geographical distance (see also Arter, 1990). These findings together with the high genetic diversity correspond to our expectations from the persistence scenario. Thus, the overall pattern of COI differentiation suggests that A. arbustorum has experienced a long preglacial history of cohesion. Many of the lineages surviving into present times must have had a wide distribution in preglacial periods. Considering the limited potential for active dispersal of A. arbustorum, the high degree of mitochondrial diversity in its present range was hardly generated by multiple waves of introgression from areas not affected by permafrost during interglacial periods. The present pattern of haplotype distribution favours the interpretation that (re)dispersion into formerly unsuitable habitats was largely – southern refugia may have played a role for populations occurring in the south of the species present range – achieved from multiple populations from within the permafrost area. As A. arbustorum is a rather cold tolerant species and nowadays occurs in the lowlands as well as in alpine regions (e.g. Kerney et al., 1983), it seems likely that this species has survived the Pleistocene glaciations in numerous refugia within the boundaries of permafrost as paleontologists have concluded from fossil findings (Lozek, 1964; Ant, 1966, 1969; Walde´n, 1986; Do¨ppes & Rabeder, 1997). The vegetation in these refugia was probably similar to today’s alpine vegetation (Frenzel, 1990). Our genetic investigation thus complements the paleontological approach, which is limited when it comes to the question of a species’ survival. As our analysis may suffer from the restricted sampling regime, an extended survey including populations from the entire range of the species and more individuals per site may find an unambiguous answer to the question how important refugia within the area affected by permafost really were for the (re)colonization of the present range. A more comprehensive and detailed investigation may even allow identifying certain refugia.
Acknowledgments We are grateful to the following people for assistance with collecting or providing material: K. Baminger, A. Baur, M. Baur, W. Meyer, H. Sattmann, also on behalf of all participants of a workshop in the Gesa¨use mountains organized by him in 1999, K. Schaefer, J. Vo¨lkl, K. Wunderle and R. Wunderle. C. van Moorsel generously provided her extraction protocol prior to publication. P. Mu¨ller and J. Spring are acknowledged for their technical support and maintenance of the
J. EVOL. BIOL. 16 (2003) 415–428 ª 2003 BLACKWELL PUBLISHING LTD
426
M. HAASE ET AL.
sequencer. Comments of D. Ebert, P. Jarne, T. Kawecki and M. Schilthuizen on an earlier draft improved the manuscript. The constructive criticism of two anonymous referees is gratefully acknowledged. Financial support was received from the Swiss National Science Foundation (grant no. 31-64855.01 to B.B.).
References Ant, H. 1963. Faunistische, o¨kologische und tiergeographische Untersuchungen zur Verbreitung der Landschnecken in Nordwestdeutschland. Abh. Landesmus. Naturk. Mu¨nster Westfahlen 25: 1–125. Ant, H. 1966. Die Bedeutung der Eiszeiten fu¨r die rezente Verbreitung der europa¨ischen Landgastropoden. Malacologia 5: 61–62. Ant, H. 1969. Zur Wu¨rm-glazialen u¨berdauerung europa¨ischer Landgastropoden in Eisrandna¨he. Malacologia 9: 249–250. Arter, H.E. 1990. Spatial relationships and gene flow paths between populations of the alpine snail Arianta arbustorum (Pulmonata: Helicidae). Evolution 44: 966–980. Avise, J.C. 1992. Molecular population structure and the biogeographic history of a regional fauna: a case history with lessons for conservation biology. Oikos 63: 62–76. Avise, J.C. 1994. Molecular Markers, Natural History and Evolution. Chapman & Hall, New York. Avise, J.C. 2000. Phylogeography. Harvard University Press, Cambridge. Avise, J.C., Neigel, J.E. & Arnold, J. 1984. Demographic influences of mitochondrial DNA lineage survivorship in animal populations. J. Mol. Evol. 20: 99–105. Avise, J.C., Arnold, J., Ball, R.M., Bermingham, E., Lamb, T., Neigel, J.E., Reeb, C.A. & Saunders, N.C. 1987. Intraspecific phylogeography: the mitochondrial DNA bridge between population genetics and systematics. Annu. Rev. Ecol. Syst. 18: 489–522. Baminger, H. 1997. Shell-morphometrical characterization of populations of Arianta arbustorum (L.) (Gastropoda, Helicidae) in the Ennstaler Alpen (Styria, Austria). Ann. Naturhist. Mus. Wien 99B: 497–519. Baur, A. & Baur, B. 1993. Daily movement patterns and dispersal in the land snail Arianta arbustorum. Malacologia 35: 89–98. Baur, B. 1984. Shell size and growth rate differences for alpine populations of Arianta arbustorum (L.) (Pulmonata: Helicidae). Rev. Suisse. Zool. 91: 37–46. Baur, B. 1986. Patterns of dispersion, density and dispersal in alpine populations of the land snail Arianta arbustorum (L.) (Helicidae). Holarctic. Ecol. 9: 117–125. Baur, B. 1990. Seasonal changes in clutch size, egg size and mode of oviposition in Arianta arbustorum (L.) (Gastropoda) from alpine populations. Zool. Anz. 225: 253–264. Baur, B. 1993. Population structure, density, dispersal and neighbourhood size in Arianta arbustorum (Linnaeus, 1758) (Pulmonata: Helicidae). Ann. Naturhist. Mus. Wien 94/95B: 307–321. Baur, B. 1994. Multiple paternity and individual variation in sperm precedence in the simultaneously hermaphroditic land snail Arianta arbustorum. Behav. Ecol. Sociobiol. 35: 413–421. Baur, B. & Baur, A. 1992. Reduced reproductive compatibility in Arianta arbustorum (Gastropoda) from distant populations. Heredity 69: 65–72.
Baur, B. and Gosteli, M. 1986. Between and within population differences in geotactic response in the land snail Arianta arbustorum (L.) (Helicidae). Behaviour 97: 147–160. Baur, B. & Raboud, C. 1988. Life history of the land snail Arianta arbustorum along an altitudinal gradient. J. Anim. Ecol. 57: 71– 87. Baur, B., Ledergerber, S. & Kothbauer, H. 1997. Passive dispersal on mountain slopes: shell shape-related differences in downhill rolling in the land snails Arianta arbustorum and Arianta chamaeleon (Helicidae). The Veliger 40: 84–85. Bisenberger, A. 1993. Zur pha¨notypischen Charakterisierung verschiedener Arianta-Populationen (A. arbustorum, A. chamaeleon, A. schmidti; Helicidae, Gastropoda). Ann. Naturhist. Mus. Wien 94/95B: 335–352. Brown, W.M. 1985. The mitochondrial genome of animals. In: Molecular Evolutionary Genetics (R.J. MacIntyre, ed.), pp. 95– 130. Plenum Press, New York. Burla, H. 1984. Induced environmental variation in Arianta arbustorum (L.). Genetica 64: 65–67. Chen, X. 1993. Comparison of inbreeding and outbreeding in hermaphroditic Arianta arbustorum (land snail). Heredity 71: 456–461. Chen, X. 1994. Self-fertilization and cross-fertilization in the land snail Arianta arbustorum (Mollusca: Pulmonata: Helicidae). J. Zool. Lond. 232: 465–471. Chiba, S. 1999. Accelerated evolution of land snails Mandarina in the oceanic Bonin islands: evidence from mitochondrial DNA sequences. Evolution 53: 460–471. Clessin, S. 1882. Helix arbustorum und ihre Varieta¨ten. Corr.-Bl. zool.-min. Ver. Regensburg 36: 35–46. Crandall, K.A. & Templeton, A.R. 1993. Empirical tests of some predictions from coalescent theory with applications to intraspecific phylogeny reconstruction. Genetics 134: 959–969. Crandall, K.A. & Templeton, A.R. 1996. Applications of intraspecific phylogenetics. In: New Uses for New Phylogenies (P.H. Harvey, A.J. Leigh Brown, J. Maynard Smith & S. Nee, eds), pp. 81–99. Oxford University Press, Oxford. Donnelly, P. & Tavare´, S. 1986. The ages of alleles and a coalescent. Adv. Appl. Probab. 18: 1–19. Do¨ppes, D. & Rabeder, G. 1997. Plioza¨ne und pleistoza¨ne Faunen terreichs. Akademie der Wissenschaften, Wien. Douris, V., Cameron, R.A.D., Rodakis, G.C. & Lecanidou, R. 1998. Mitochondrial phylogeography of the land snail Albinaria in Crete: long-term geological and short-term vicariance effects. Evolution 52: 116–125. Ehrmann, P. 1933. Mollusken (Weichtiere), II (1). In: Die Tierwelt Mitteleuropas (P. Brohmer, P. Ehrmann & G. Ulmer, eds), pp. 1–264. Quelle und Meyer, Leipzig. Excoffier, L. 1993. MINSPNET (http://anthropologie.unige.ch/ LGB/software/win/min-span-net/). Excoffier, L. & Langaney, A. 1989. Origin and differentiation of human mitochondrial DNA. Am. J. Hum. Genet. 44: 73–85. Excoffier, L., Smouse, P. & Quattro, J. 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics 31: 479–491. Fechter, R. & Falkner, G. 1989. Weichtiere. Mosaikverlag, Mu¨nchen. Fedorov, V.B. & Stenseth, N.C. 2001. Glacial survival of the Norwegian lemming (Lemmus lemmus) in Scandinavia: inference from mitochondrial DNA variation. Proc. R. Soc. Lond. B 268: 809–814.
J. EVOL. BIOL. 16 (2003) 415–428 ª 2003 BLACKWELL PUBLISHING LTD
Mitochondrial differentiation in a land snail
Felsenstein, J. 1978. Cases in which parsimony and compatibility methods will be positively misleading. Syst. Zool. 27: 401–410. Ferris, C., King, R.A. & Hewitt, G.M. 1999. Isolation within species and the history of glacial refugia. In: Molecular Systematics and Plant Evolution (P.M. Hollingsworth, R.M. Bateman & R.J. Gornall, eds), pp. 20–34. Taylor and Francis, London. Folmer, O., Black, M., Hoeh, W., Lutz, R. & Vrijenhoek, R. 1994. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol. Mar. Biol. Biotechnol. 3: 294–299. Frenzel, B. 1990. Die Vegetationsentwicklung im Eiszeitalter. In: Eiszeitforschung (H. Liedtke, ed.), pp. 69–90. Wissenschaftliche Buchgesellschaft, Darmstadt. Gabrielsen, T.M., Bachmann, K., Jakobsen, K.S. & Brochmann, C. 1997. Glacial survival does not matter: RAPD phylogeography of nordic Saxifraga oppositifolia. Mol. Ecol. 6: 831–842. Gittenberger, E. 1991. Altitudinal variation and adaptive zones in Arianta arbustorum: a new look at a widespread species. J. Moll. Stud. 57: 99–109. Golding, G.B. 1987. The detection of deleterious selection using ancestors inferred from a phylogenetic history. Genet. Res. 49: 71–82. Grossu, A.V. 1983. Gastropoda Romaniae 4. Ordo Stylommatophora. Superfam.: Arionacea, Zonitacea, Ariophantacea si Helicacea. Editura Litera, Bucuresti. Hatzoglou, E., Rodakis, G.C. & Lecanidou, R. 1995. Complete sequence and gene organization of the mitochondrial genome of the land snail Albinaria coerulea. Genetics 140: 1353–1366. Hewitt, G.M. 1996. Some genetic consequences of ice ages, and their role in divergence and speciation. Biol. J. Linn. Soc. 58: 247–276. Hewitt, G.M. 1999. Post-glacial re-colonization of European biota. Biol. J. Linn. Soc. 68: 87–112. Hewitt, G.M. 2000. The genetic legacy of the Quaternary ice ages. Nature 405: 907–913. Hoelzer, G.A. 1997. Inferring phylogenies from mtDNA variation: mitochondrial-gene trees versus nuclear-gene trees revisited. Evolution 51: 622–626. Hoelzer, G.A., Wallman, J. & Melnick, D.J. 1998. The effects of social structure, geographical structure, and population size on the evolution of mitochondrial DNA: II. Molecular clocks and the lineage sorting period. J. Mol. Evol. 47: 21–31. Ibrahim, K.M., Nichols, R.A. & Hewitt, G.M. 1996. Spatial patterns of genetic variation generated by different forms of dispersal during range expansion. Heredity 77: 282–291. Jaarola, M., Tegelstrom, H. & Fredga, K. 1999. Colonization history in Fennoscandian rodents. Biol. J. Linn. Soc. 68: 113–127. Kerney, M.P., Cameron, R.A.D. & Jungbluth, J.H. 1983. Die Landschnecken Nord- und Mitteleuropas. Paul Parey, Berlin. Kessing, B.D. 1999. Sequencer, Version 4.0 (http:// 160.111.138.66/Sequencer.html). King, T.L., Eackles, M.S., Gjetvaj, B. & Hoeh, W.R. 1999. Intraspecific phylogeography of Lasmigona subviridis (Bivalvia: Unionidae): conservation implication of range discontinuity. Mol. Ecol. 8: S65–S78. Kirchner, C., Kra¨tzner, R. & Welter-Schultes, F.W. 1997. Flying snails–how far can Truncatellina (Pulmonata: Vertiginidae) be blown over the sea? J. Moll. Stud. 63: 479–487.
427
Klemm, W. 1974. Die Verbreitung der rezenten Land-Gehe¨useSchnecken in sterreich. Denkschr. O¨sterr. Akad. Wiss. (math. nat. Kl.) 117: 1–503. Lessios, H.A., Kessing, B.D., Robertson, D.R. & Paulay, G. 1999. Phylogeography of the pantropical sea urchin Eucidaris in relation to land barriers and ocean currents. Evolution 53: 806– 817. Lockhart, P.J., Steel, M.A., Hendy, M.D. & Penny, D. 1994. Recovering evolutionary trees under a realistic model of sequence evolution. Mol. Biol. Evol. 11: 605–612. Lockhart, P.J., Larkum, A.W., Steel, M.A., Waddell, P.J. & Penny, D. 1995. Evolution of chlorophyll and bacteriochlorophyll: the problem of invariant sites in sequence analysis. Proc. Natl. Acad. Sci. USA 93: 1930–1934. Lozek, V. 1964. Quarta¨rmollusken der Tschechoslowakei. Geologische Zentralanstalt. Verlag der Tschechoslowakischen Akademie der Wissenschaften, Praha. Marko, P.B. 1998. Historical allopatry and the biogeography of speciation in the prosobranch snail genus. Nucella Evol. 52: 757–774. Medina, M. & Walsh, P.J. 2000. Molecular systematics of the order Anaspidea based on mitochondrial DNA sequence (12s, 16s, COI). Mol. Phyl. Evol. 15: 41–58. Misof, B., Rickert, A.M., Buckley, T.R., Fleck, G. & Sauer, K.P. 2001. Phylogenetic signal and its decay in mitochondrial SSU and LSU rRNA gene fragments of Anisoptera. Mol. Biol. Evol. 18: 27–37. Murray, J., Stine, O.C. & Johnson, M.S. 1991. The evolution of mitochondrial DNA in Partula. Heredity 68: 93–104. Posada, D. & Crandall, K.A. 1998. Modeltest: testing the model of DNA substitution. Bioinformatics 14: 817–818. Riddle, B.R. 1996. The molecular phylogeographic bridge between deep and shallow history in continental biotas. Trends Ecol. Evol. 11: 207–211. Robinson, M., Gouy, M., Gautier, C. & Mouchiroud, D. 1998. Sensitivity of the relative-rate test to taxonomic sampling. Mol. Biol. Evol. 15: 1091–1098. Rodriguez, F., Oliver, J.L., Marin, A. & Medina, J.R. 1990. The general stochastic model of nucleotide substitution. J. Theor. Biol. 142: 485–501. Ross, T.K. 1999. Phylogeography and conservation genetics of the Iowa Pleistocene snail. Mol. Ecol. 8: 1363–1373. Rozas, J. & Rozas, R. 1999. DnaSP version 3: an integrated program for molecular population genetics and molecular evolution analysis. Bioinformatics 15: 174–175. Schneider, S., Roessli, D. & Excoffier, L. 2000. Arlequin ver. 2001: a software for population genetics data analysis. Genetics and Biometry Laboratory, University of Geneva, Switzerland. Simon, C., Frati, F., Beckenbach, A., Crespi, B., Liu, H. & Flook, P. 1994. Evolution, weighting, and phylogenetic utility of mitochondrial gene sequences and a compilation of conserved polymerase chain reaction primers. Ann. Entomol. Soc. Am. 87: 651–701. Slatkin, M. 1991. Inbreeding coefficients and coalescence times. Genet. Res. 58: 167–175. Stehlik, I., Schneller, J.J. & Bachmann, K. 2001. Resistance or emigration: response of the high-alpine plant Eritrichium nanum (L.) Gaudin to the ice age within the Central Alps. Mol. Ecol. 10: 357–370. Stewart, J.R. & Lister, A.M. 2001. Cryptic northern refugia and the origins of the modern biota. Trends Ecol. Evol. 16: 608–613.
J. EVOL. BIOL. 16 (2003) 415–428 ª 2003 BLACKWELL PUBLISHING LTD
428
M. HAASE ET AL.
Swofford, D.L. 1998. PAUP. Phylogenetic analysis using parsimony, Version 4. Sinauer, Sunderland. Swofford, D.L., Olsen, G.J., Waddell, P.J. & Hillis, D.M. 1996. Phylogenetic inference. In: Molecular Systematics, 2nd edn (D.M. Hillis, C. Moritz & B.K. Mable, eds), pp. 407–514. Sinauer, Sunderland. Taberlet, P., Fumagalli, L., Wust-Saucy, A.-G. & Cosson, J.-F. 1998. Comparative phylogeography and postglacial colonization routes in Europe. Mol. Ecol. 7: 453–464. Tamura, K. & Nei, M. 1993. Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Mol. Biol. Evol. 10: 512–526. Taylor, J.W. 1882. Life histories of British Helices. J. Conch. 3: 241–259. Terrett, J., Miles, S. & Thomas, R.H. 1994. The mitochondrial genome of Cepaea nemoralis: gene order, base composition, and heteroplasmy. Nautilus 108(Suppl. 2):79–84. Thacker, R.W. & Hadfield, G.M. 2000. Mitochondrial phylogeny of extant Hawaiian tree snails (Achatinellinae). Mol. Phyl. Evol. 16: 263–270. Thomaz, D., Guiller, A. & Clarke, B. 1996. Extreme divergence of mitochondrial DNA within species of pulmonate land snails. Proc. R. Soc. Lond. B 263: 363–368.
Vagvolgyi, J. 1975. Body size, aerial dispersal, and origin of the Pacific land snail fauna. Syst. Zool. 24: 465–488. van Moorsel, C.H.M., van Nes, W.J. & Megens, H.-J. 2000. A quick, simple and inexpensive DNA extraction protocol for gastropods. Malacologia 42: 203–206. Waddel, P.J. & Steel, M.A. 1997. General time-reversible distances with unequal rates across sites: mixing G and inverse Gaussian distributions with invariant sites. Mol. Phyl. Evol. 8: 398–414. Walde´n, H. 1986. A comprehensive account of the late quaternary land Mollusca in Scandinavia. Malakol. Abh. Staatl. Mus. Tierkd. Dresden 11: 105–133. Wenz, W. 1923. Gastropoda extramarina tertiaria. Fossilium Catalogus I, 18: 335–736. Yang, Z. 1998. On the best evolutionary rate for phylogenetic analysis. Syst. Biol. 47: 125–133. Zhang, D.-X. & Hewitt, G.M. 1996. Nuclear integrations: challenges for mitochondrial DNA markers. Trends Ecol. Evol. 11: 247–251. Received 28 August 2002; revised 18 December 2002; accepted 27 February 2003
J. EVOL. BIOL. 16 (2003) 415–428 ª 2003 BLACKWELL PUBLISHING LTD
