Fragmented distribution ranges of species with little active dispersal capacity raise the question about their place of origin and the processes and timing of either range fragmentation or dispersal. The peculiar distribution of the land snail Tudorella sulcata s. str. in Southern France, Sardinia and Algeria is such a challenging case.
Statistical phylogeographic analyses with mitochondrial COI and nuclear hsp70 haplotypes were used to answer the questions of the species' origin, sequence and timing of dispersal. The origin of the species was on Sardinia. Starting from there, a first expansion to Algeria and then to France took place. Abiotic and zoochorous dispersal could be excluded by considering the species' life style, leaving only anthropogenic translocation as parsimonious explanation. The geographic expansion could be dated to approximately 8,000 years before present with a 95% confidence interval of 10,000 to 3,000 years before present.
Citation: Jesse R, Véla E, Pfenninger M (2011) Phylogeography of a Land Snail Suggests Trans-Mediterranean Neolithic Transport. PLoS ONE 6(6): e20734. https://doi.org/10.1371/journal.pone.0020734
Editor: M. Thomas P. Gilbert, Natural History Museum of Denmark, Denmark
Received: March 28, 2011; Accepted: May 8, 2011; Published: June 22, 2011
Copyright: © 2011 Jesse et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The study was supported by the research funding programme “LOEWE - Landes-Offensive zur Entwicklung Wissenschaftlich-ökonomischer Exzellenz” of Hessen's Ministry of Higher Education, Research, and the Arts. (http://www.hessen.de/irj/HMWK_Internet?uid=fa560c0b-ed11-9311-1010-43bf5aa60dfa). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Ever since people have had the ability to move around the planet, they have been introducing new species, either intentionally or accidentally, into new geographical areas . Likewise, the biodiversity in the Mediterranean Basin was shaped, at least since the Neolithic emergence (ca. 12.000–3.000 years before present) by human landscape management and species introductions which usually impacted the indigenous fauna and flora . Since the beginning of live-stock domestication about 10.000–9.500 years before present (BP) in the Eastern Mediterranean, approximately a millennium after the first domestication of crop plants, both were spread westwards with settlers disseminating the Neolithic culture. From then on, the Mediterranean region has served as sink and source for extensive exchange of biodiversity associated to human use .
But also species not directly involved in human exploitation were spread accidentally from the very beginning of human migrations, like weed propagules contaminating crop seeds or blind passengers in ship ballasts . In the Mediterranean, it is known that at least nine land snail species were dispersed as an unintentional by-product of bronze-age maritime copper and resin trade . Analysis of land snail shells found in a ship wreck indicated that they were carried off alive with scrub used to cushion heavy freight .
The possibility of (un)-intentional introductions in addition to natural range expansions of a species makes it often difficult to infer the direction and timing of dispersal events . Historic records of first sightings of a species in a new area are rare , in particular for organisms that are inconspicuous, not a pest or otherwise attracting human attention. Moreover such records are, by definition, non-existing for prehistoric times. In these cases, the (sub)fossil archaeological record can provide an estimate of the minimum time of residence in an area , . However, not all organisms are prone to fossilisation or their remains cannot be attributed to a particular species with the necessary certainty. To overcome these difficulties, various methodological approaches based on the distribution of genetic variation have been developed in the past 25 years. The rationale behind these phylogeographic approaches is the reconstruction of the demographic history of the extant populations from DNA sequence data .
The distribution of the land snail Tudorella sulcata s. str. in Southern France, Sardinia and Algeria - areas that are currently separated by an ocean and that were last in direct contact about 30 million years ago - raises the strong suspicion that this species could have been at some point in the past subject to passive dispersal. The rather large snail (15–20 mm height) has a low active dispersal capacity and a covert life style, burrowed mostly in the soil under rocks and shrubs, except for activity phases during wet weather periods . The first record of the species in literature is its description from Provence by Draparnaud in 1805, thus setting the minimum age for the introduction to Southern France. Due to the presence of cryptic species, it is unfortunately not possible to attribute fossilised shells to extant species in the genus Tudorella , . Consequently, we used statistical phylogeographic analyses to successfully answer two major questions:
- Where is the origin of the land snail species Tudorella sulcata and what was the colonisation sequence?
- When were the respective areas invaded?
The present study covered all known occurrences of the species Tudorella sulcata sensu stricto (Draparnaud 1805) as described in Pfenninger et al.  (Figure 1, Table S1). For each of the 28 sampled locations, at least 5 apparently living individuals were preserved in 90% alcohol. Further information about the sampling sites is given in the additional files.
The sampling sites comprise all known occurrences of T. sulcatus s.str.. Sampling sites in France are blue, Algeria red and Sardinia green, indicating pooling for gene-flow analysis.
For DNA extraction a part of the foot muscle was taken. Genomic DNA was extracted using the CTAB method. For 138 individuals, a 604 bp segment of the cytochrome oxidase subunit I gene (COI) was PCR-amplified. PCR conditions and primers were adopted from the study of Pfenninger et al. .
Additionally, a 219 bp fraction of the nuclear coded heat shock 70 kD protein (hsp70) was sequenced. The primers with the sequences HSP71Iifor TGCAGCATCTTCTGGATACG and HSP71Iirev GCTCTACGGCGTCTACGAAC were used for PCR amplification under the same conditions as for COI.
PCR products were purified using Pure Link™ PCR Purification Kit (invitrogen, Carlsbad, CA, USA). Ten nanograms per sample were subjected to forward and reverse cycle sequencing using the ABI Prism Big Dye terminator kit (Perkin-Elmer, Norwalk, CT, USA). Sequencing reactions were electrophoresed on an ABI 377 automated DNA sequencer.
As the nuclear hsp70 locus is diploid, we checked the trace files for dinucleotide ambiguities. We counted a site as heterozygous single nucleotide polymorphism when the same ambiguity occurred in both forward and reverse sequencing trace file. We inferred the haplotype phases of heterozygous individuals with the coalescent-based Bayesian method PHASE 2.1  as implemented in DNAsp 5.0 .
Sequences were aligned with Clustal W  and manually adjusted. Statistical parsimony (SP) cladograms were constructed for both genes separately using TCS vers. 1.21  with the connection limit was set at 95%.
Phylogeographic Model Selection (PMS)
To compare the relevance of nine explicit dispersal hypotheses, we applied a model selection approach , , introduced into phylogeography by Pfenninger & Posada  and further developed in Depraz et al. . The 28 and 17 sampling sites for mitochondrial and nuclear marker, respectively, were pooled into three geographic groups (France (F), Sardinia (S) and Algeria (A), Figure 1). Taking each of the three regions as potential origin in turn, we evaluated i) the hypothesis that each of the other two regions were independently colonised from there and ii) that either one of the two regions was colonised first and the remaining region from there. This resulted in nine hypotheses, which were translated into the corresponding gene-flow matrices. The analyses were carried out for the mitochondrial and nuclear fragment separately.
The maximum likelihood migration rate matrix of each model was then estimated using migrate-n version 2.3 . The first genealogy was started with a random tree. Initial theta and migrant values were generated from an FST calculation. A static heating scheme with four different temperatures was applied. We ran ten short chains with 4×104 generations each, from which 1000 trees were recorded in regular intervals after a burn-in phase of 20.000 generations. These were followed by three long chains of 105 generations, from which 103 trees were sampled after a burn-in period of 2×103 generations. Parameter estimates were gained from the last chain. Log likelihood estimates cannot be directly compared over different runs with migrate-n. We therefore ran a final analysis with an unconstrained migration model using the likelihood-ratio-test option to gain likelihood estimates that were comparable between the different hypotheses and their parameter sets. We used these estimates and the number of free parameters in each model to calculate the Akaike Information Criterion (AIC, ).
Dating demographic expansions associated with geographic expansions was performed with the extended version of the Bayesian Skyline Analysis  implemented in BEAST 1.5.2  for the COI fragment only, because no molecular clock estimate was available for the hsp70 locus. A skyline plot is a model of population size fitting a wide range of demographic histories. The Bayesian skyline model uses standard Markov chain Monte Carlo (MCMC) sampling procedures to estimate a posterior distribution of effective population size through time from gene sequences, given a model of sequence evolution. If a molecular clock rate is known for the sequence in question, the model can be used to put demographic events into a historical context . As we are dealing with intraspecific data, where systematic rate heterogeneity is not expected, we have chosen a strict molecular clock model with a fixed rate of 4.28×10−8 changes per site and year, as estimated specifically for this Tudorella lineage . Even though a strictly intraspecific calibration point would be more desirable, the use of a rate estimated for a terminal branch avoids the overestimation of divergence dates due to saturation associated with calibration points from deep interspecific nodes .
A series of initial runs performed to optimise priors and run-time parameter choice indicated that it was necessary to run 4×107 generations of the Monte-Carlo-Markov-Chain, sampling every 1000th generation, to obtain effective sampling sizes above 500 for all estimated parameters. We have chosen the General Time Reversible model with empirical base composition as site evolution model, a gamma distribution of rate heterogeneity with four rate categories and invariant sites, because initial runs indicated that no alternative models fitted the data significantly better. As tree model, the Extended Bayesian Skyline Model for mitochondrial data with linear growth between population size change events was applied with a UPGMA generated tree as starting point. The prior for the number of population size changes was a Poisson distribution with a mean of two, as we were expecting two expansion events (see results). A uniform distribution between 104 and 1012 with an initial size of 106 was set for the demographic population mean prior, as no information on the actual population size of T. sulcata was available. We tested the skyline model against a model of constant population size, using otherwise the same parameters and priors. The models were compared using log Bayes factors.
We obtained 138 COI sequences of 604 bp length (Table S1). All sequences aligned unambiguously and translated without stop codons. Statistical parsimony analysis resulted in a network of 13 haplotypes, defined by 12 singleton mutations and five parsimony informative (PI) sites (Figure 2a). Haplotype diversities for the three geographical regions are: Sardinia 0.160, Algeria 0.143 and France 0.067. Haplotype sequences are registered in GenBank, accession numbers are GU385953–GU385966.
Circles depict sampled haplotypes, their size is proportional to their frequency. Connecting lines correspond to single mutational steps. Haplotypes occurring in France are blue, Algeria red and Sardinia green.
For the hsp70 locus 38 individuals were typed (Table S1). The majority of sequences contained one or more ambiguous sites which indicated the presence of heterozygous individuals. We inferred ten different haplotypes (GenBank acc. nos. JF967635–JF967644), defined by two PI and 8 singleton sites (Figure 2b). The region with the highest haplotype diversity for this marker was Sardinia (0.005), followed by Algeria (0.004) and France (0.003).
Inference of dispersal routes
Model selection criteria preferred the S>A>F model for both loci (AIC = −33.2, LnL = 21.6 and AIC = −379,78, LnL = 194,89, respectively) over the remaining eight models (Table 1). The best supported model implies that the origin of the expansion was on Sardinia. Starting from there a first expansion to Algeria and a second expansion from Algeria to France took place.
The median estimate in the Bayesian skyline plot analysis for the beginning of population growth was approximately 8.000 years BP with a 95% confidence interval ranging from 10.000 to 3.000 years BP. Figure 2 shows the estimated increase of effective population size with 95% confidence intervals. The skyline model inferring a population expansion was decisively better than a constant population size model (log Bayes factor = 3.64).
Statistical phylogeography is a powerful tool to answer questions about a species' origin, dispersal routes and timing of expansions in a rigorous statistical framework and has been successfully established in the last years , , . However, such analyses demand a sampling of the whole range of the focal species. The distribution range of Tudorella sulcata comprising Southern France, Western Sardinia and Algeria, is well know due to a previous study  on the range wide analysis of the genus and could thus be comprehensively sampled.
The occurrence of Algerian mitochondrial haplotypes both on Sardinia and in France and the lack of shared haplotypes between France and Sardinia argue either for an Algerian origin or a connection of the areas through this country (Figure 2a). The hsp70 haplotypes showed a very similar distribution pattern. However, as expected for a nuclear locus with deeper coalescent times, most haplotypes are shared between regions (Figure 2b).
To distinguish among all nine possible expansion sequence models by measuring the fit of the data to the models, we relied on phylogeographic model selection. The advantage of this alternative to classical hypothesis testing is the possibility to evaluate several models simultaneously instead of testing repeatedly single hypotheses against a null model . Analyses of both independently inherited marker strongly supported an origin of Tudorella sulcata on Sardinia and an expansion from there first to Algeria and subsequently to France (Table 1). It is thus likely that our results reflect the true population history and are not due to the particular coalescence history of the respective genes.
Sardinia as origin of the species was also indicated by the highest haplotype diversity for both mitochondrial and nuclear marker in that region. The lack of support for France as origin of the species is also compatible with the results of a biogeographical analysis that excluded France as possible origin for this species .
The somehow counter-intuitive dispersal sequence suggested by our analysis, in addition to the species' biology makes a natural expansion very unlikely. Dispersal by log-rafting in the sequence Sardinia-Algeria-France is most implausible when considering the surface circulations in the Western Mediterranean Sea. Dense water from the Tyrrhenian Sea is transported in a current flowing from Sardinia northwards to Corsica and France and counter clockwise along the French and Spanish coasts until reaching Northern Africa near the Strait of Gibraltar . The Algerian Current is going 3–5 km per day eastwards in anticlockwise circulation along the northern continental slope, making drift from Algeria to France under regular conditions highly unlikely as it would require the snails to survive a several week long journey exposed to saltwater.
The maximum reported dispersal distance of light objects (<450 g) by extreme strong winds is 130 km . As the distances between Sardinia and Algeria (ca. 250 km) as well as Algeria and France (ca. 700 km) exceed this maximum by far, it is reasonable to dismiss the hypothesis of wind-borne dispersal.
The pattern observed in our analysis thus suggests transportation through biological vectors. Migrating birds are potential over-sea vectors for small invertebrates, but rather for freshwater species , . Usually eggs or larval stages are transported, attached to feathers of water fowl. Tudorella sulcata lives mostly buried in the soil and contact to migrating birds is hardly imaginable. Even though important bird migration routes run between central Africa and Europe, transport by birds is thus unlikely for this land snail and would not explain the colonisation sequence.
An ancient expansion (e.g. during the Messinian Salinity Crisis 5.96±0.20 Mya  was a priori unlikely in the light of the shallow divergence of the populations and the COI molecular clock rate estimated from phylogenetic analysis . After exclusion of the possible alternatives, anthropogenic passive dispersal remains as sole plausible mechanism.
Timing of the expansion
As a colonisation of new areas is logically linked to demographic expansion, the dating of the latter can give a time frame for these events. Bayesian skyline plot estimates suggested that a population expansion in Tudorella sulcata started between 10.000 and 3.000 years ago (Figure 3). This broad interval excludes historic times on one end and Pleistocene dispersal on the other. The highest probability density of the estimate lay around 8000 years BP. This period coincides on the one hand with the beginning of the Neolithic period in the Western Mediterranean and on the other with the Holocene climate optimum . The latter may have been causal for the onset of the observed population expansion, just like the postglacial warming was for the spatial and demographic expansion of several land snail species , , . However, if this climate change was indeed triggering a major population dynamic in T. sulcata, the subsequent cooling starting around 4000 ybp should have also left its mark as population decline. Additionally, a climate driven population size increase would not explain the transoceanic dispersal.
Historical development of population sizes through time based on Bayesian coalescence analyses of the COI locus. The analysis suggests a population increase between 3,000 and 10,000 years bp with the highest probability density around 8,000 years bp.
The Holocene optimum may have nevertheless played an indirect role for the colonisation of new habitats by triggering or facilitating the Neolithic period . Coming from the Near East (Levant), Neolithic colonists started around 10.500 years BP to migrate first into the Balkans and then continuously colonised the European continent in expansion waves westwards during the following 4.500 years, with a peak of migration around 8.000 years BP , , . First sea crossings in the Mediterranean are documented already for the Mesolithic Age by obsidian originating from the Cycladic island of Melos found in Mesolithic sediments of the Franchthi cave on the Peloponnesus . Numerous obsidian artefacts found in Italy, Southern France, Eastern Spain, Croatia, Greece, Tunisia and Algeria from far away sources document the efficiency of trans- oceanic transport during the Neolithic period . The synchrony of the demographic expansion with this phase of early human expansion and the exclusion of other factors suggests indeed that Neolithic settlers or traders were acting as vectors for the snails. Genetic analyses of Sardinian human populations connect Sardinia and Northern Africa through early human migrations and strengthen the above scenario , .
The secondary expansion from Algeria to France is more difficult to date because the installation of the rather small population there has not left discernable traces like a steepening slope in the demographic reconstruction curve even though it contributed to the general population size increase. In Southern France, the first Neolithic coastal settlements date to 7.700–7.600 years BP . Contacts between the French and Algerian coasts have been numerous since then. However, the private haplotypes found in France (Figure 1a+b) argue for a relatively early introduction.
Without further archaeological evidence, it is impossible to determine the purpose the snails were transported for. Unintentional shipping as in the case of the Uluburun wreck , transportation as trading good or simply because the light pink, regularly sculptured shells have a quite attractive appearance, can be imagined.
Passive dispersal of Tudorella species across marine barriers to the nearest opposite coasts appear to have happened several times . The presence of the same species in Morocco and Southern Spain (T. mauretanica) as well in Tunisia and Sicily (T. multisulcata) illustrate this . However, in these cases the Pleistocene sea level changes  may have facilitated short range human or animal passive transport or may have even allowed active dispersal between these regions.
This study indicates that the population histories of non-domesticated animals and plants may give hints on the direction and timing of early human expansion routes. It suggests that modern invasions, such as the well-known zebra-mussel proliferation in freshwaters worldwide  or the invasion of the Asian tiger mosquito in Europe , are only among the latest instances of a long series of anthropogenic introductions, shaping the distribution of extant biodiversity.
List of sampling sites used in the study.
Thanks are due to Sebastian Klaus and Frédéric Magnin for their help in the field. Rebecca Bloch and Tanja Zimmermann are acknowledged for their assistance in the laboratory.
Conceived and designed the experiments: MP EV. Performed the experiments: RJ. Analyzed the data: MP RJ. Contributed reagents/materials/analysis tools: RJ EV. Wrote the paper: RJ MP EV.
- 1. Wilson JRU, Dormontt EE, Prentis PJ, Lowe AJ, Richardson DM (2009) Something in the way you move: dispersal pathways affect invasion success. Trends in Ecology & Evolution 24: 136–144.JRU WilsonEE DormonttPJ PrentisAJ LoweDM Richardson2009Something in the way you move: dispersal pathways affect invasion success.Trends in Ecology & Evolution24136144
- 2. Zeder MA (2008) Domestication and early agriculture in the Mediterranean Basin: Origins, diffusion, and impact. Proceedings of the National Academy of Sciences of the United States of America 105: 11597–11604.MA Zeder2008Domestication and early agriculture in the Mediterranean Basin: Origins, diffusion, and impact.Proceedings of the National Academy of Sciences of the United States of America1051159711604
- 3. Welter-Schultes FW (2008) Bronze Age shipwreck snails from Turkey: First direct evidence for oversea carriage of land snails in antiquity. Journal of Molluscan Studies 74: 79–87.FW Welter-Schultes2008Bronze Age shipwreck snails from Turkey: First direct evidence for oversea carriage of land snails in antiquity.Journal of Molluscan Studies747987
- 4. Jesse R, Pfenninger M, Fratini S, Scalici M, Streit B, et al. (2009) Disjunct distribution of the Mediterranean freshwater crab Potamon fluviatile-natural expansion or human introduction? Biological Invasions 11: 2209–2221.R. JesseM. PfenningerS. FratiniM. ScaliciB. Streit2009Disjunct distribution of the Mediterranean freshwater crab Potamon fluviatile-natural expansion or human introduction?Biological Invasions1122092221
- 5. Pringle A, Vellinga EC (2006) Last chance to know? Using literature to explore the biogeography and invasion biology of the death cap mushroom Amanita phalloides (Vaill.). Biological Invasions 8: 1131–1144.A. PringleEC Vellinga2006Last chance to know? Using literature to explore the biogeography and invasion biology of the death cap mushroom Amanita phalloides (Vaill.).Biological Invasions811311144
- 6. Olson SL (2003) First fossil record of a finfoot (Aves : Heliornithidae) and its biogeographical significance. Proceedings of the Biological Society of Washington 116: 732–736.SL Olson2003First fossil record of a finfoot (Aves : Heliornithidae) and its biogeographical significance.Proceedings of the Biological Society of Washington116732736
- 7. Pregill GK, Steadman DW (2009) The prehistory and biogeography of terrestrial vertebrates on Guam, Mariana Islands. Diversity and Distributions 15: 983–996.GK PregillDW Steadman2009The prehistory and biogeography of terrestrial vertebrates on Guam, Mariana Islands.Diversity and Distributions15983996
- 8. Emerson BC, Paradis E, Thebaud C (2001) Revealing the demographic histories of species using DNA sequences. Trends in Ecology & Evolution 16: 707–716.BC EmersonE. ParadisC. Thebaud2001Revealing the demographic histories of species using DNA sequences.Trends in Ecology & Evolution16707716
- 9. Vela E, Magnin F, Pavon D, Pfenninger M (2008) Phylogénie moléculaire et données paléobiogéographiques sur le gastéropode terrestre Tudorella sulcata (Draparnaud, 1805) en France et en Algérie orientale. Geodiversitas 30: 233–246.E. VelaF. MagninD. PavonM. Pfenninger2008Phylogénie moléculaire et données paléobiogéographiques sur le gastéropode terrestre Tudorella sulcata (Draparnaud, 1805) en France et en Algérie orientale.Geodiversitas30233246
- 10. Pfenninger M, Vela E, Jesse R, Arantzazu Elejalde M, Liberto F, et al. (2009) Temporal speciation pattern in the western Mediterranean genus Tudorella P. Fischer, 1885 (Gastropoda, Pomatiidae) supports the Thyrrhenian vicariance hypothesis. Molecular Phylogenetics and Evolution 54: 427–436.M. PfenningerE. VelaR. JesseM. Arantzazu ElejaldeF. Liberto2009Temporal speciation pattern in the western Mediterranean genus Tudorella P. Fischer, 1885 (Gastropoda, Pomatiidae) supports the Thyrrhenian vicariance hypothesis.Molecular Phylogenetics and Evolution54427436
- 11. Stephens M, Donelly P (2003) A comparison of bayesian methods for haplotype reconstruction from population genotype data. American Journal of Human Genetics 73: 1162–1169.M. StephensP. Donelly2003A comparison of bayesian methods for haplotype reconstruction from population genotype data.American Journal of Human Genetics7311621169
- 12. Librado P, Rozas J (2009) DnaSP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25: 1451–1452.P. LibradoJ. Rozas2009DnaSP v5: a software for comprehensive analysis of DNA polymorphism data.Bioinformatics2514511452
- 13. Thompson JD, Higgins DG, Gibson TJ (1994) Clustal-W - Improving the Sensitivity of Progressive Multiple Sequence Alignment through Sequence Weighting, Position- Specific Gap Penalties and Weight Matrix Choice. Nucleic Acids Research 22: 4673–4680.JD ThompsonDG HigginsTJ Gibson1994Clustal-W - Improving the Sensitivity of Progressive Multiple Sequence Alignment through Sequence Weighting, Position- Specific Gap Penalties and Weight Matrix Choice.Nucleic Acids Research2246734680
- 14. Clement M, Posada D, Crandall KA (2000) TCS: a computer program to estimate gene genealogies. Molecular Ecology 9: 1657–1659.M. ClementD. PosadaKA Crandall2000TCS: a computer program to estimate gene genealogies.Molecular Ecology916571659
- 15. Johnson JB, Omland KE (2004) Model selection in ecology and evolution. Trends In Ecology & Evolution 19: 101–108.JB JohnsonKE Omland2004Model selection in ecology and evolution.Trends In Ecology & Evolution19101108
- 16. Stephens PA, Buskirk SW, Martinéz del Rio C (2007) Inference in ecology and evolution. Trends In Ecology & Evolution 22: 192–197.PA StephensSW BuskirkC. Martinéz del Rio2007Inference in ecology and evolution.Trends In Ecology & Evolution22192197
- 17. Pfenninger M, Posada D (2002) Phylogeographic history of the land snail Candidula unifasciata (Helicellinae, Stylommatophora): Fragmentation, corridor migration, and secondary contact. Evolution 56: 1776–1788.M. PfenningerD. Posada2002Phylogeographic history of the land snail Candidula unifasciata (Helicellinae, Stylommatophora): Fragmentation, corridor migration, and secondary contact.Evolution5617761788
- 18. Dépraz A, Cordellier M, Hausser J, Pfenninger M (2008) Postglacial recolonisation at a snail's pace (Trochulus villosus): confronting competing refugia hypotheses using model selection. Molecular Ecology 17: 2449–2462.A. DéprazM. CordellierJ. HausserM. Pfenninger2008Postglacial recolonisation at a snail's pace (Trochulus villosus): confronting competing refugia hypotheses using model selection.Molecular Ecology1724492462
- 19. Beerli P, Felsenstein J (2001) Maximum likelihood estimation of a migration matrix and effective population sizes in n subpopulations by using a coalescent approach. Proceedings of the National Academy of Sciences of the United States of America 98: 4563–4568.P. BeerliJ. Felsenstein2001Maximum likelihood estimation of a migration matrix and effective population sizes in n subpopulations by using a coalescent approach.Proceedings of the National Academy of Sciences of the United States of America9845634568
- 20. Akaike H (1974) A new look at the statistical model identification. IEEE Transactions on Automatic Control 19: 716–723.H. Akaike1974A new look at the statistical model identification.IEEE Transactions on Automatic Control19716723
- 21. Drummond AJ, Rambaut A, Shapiro B, Pybus OG (2005) Bayesian coalescent inference of past population dynamics from molecular sequences. Molecular Biology and Evolution 22: 1185–1192.AJ DrummondA. RambautB. ShapiroOG Pybus2005Bayesian coalescent inference of past population dynamics from molecular sequences.Molecular Biology and Evolution2211851192
- 22. Drummond AJ, Rambaut A (2007) BEAST: Bayesian evolutionary analysis by sampling trees. BMC Evolutionary Biology 7: 214.AJ DrummondA. Rambaut2007BEAST: Bayesian evolutionary analysis by sampling trees.BMC Evolutionary Biology7214
- 23. Ho SYW, Saarma U, Barnett R, Haile J, Shapiro B (2008) The effect of inappropriate calibration: Three case studies in molecular ecology. PLoS One 3: e1615.SYW HoU. SaarmaR. BarnettJ. HaileB. Shapiro2008The effect of inappropriate calibration: Three case studies in molecular ecology.PLoS One3e1615
- 24. Cordellier M, Pfenninger M (2009) Inferring the past to predict the future: climate modelling predictions and phylogeography for the freshwater gastropod Radix balthica (Pulmonata, Basommatophora). Molecular Ecology 18: 534–544.M. CordellierM. Pfenninger2009Inferring the past to predict the future: climate modelling predictions and phylogeography for the freshwater gastropod Radix balthica (Pulmonata, Basommatophora).Molecular Ecology18534544
- 25. Depraz A, Cordellier M, Hausser J, Pfenninger M (2008) Postglacial recolonization at a snail's pace (Trochulus villosus): confronting competing refugia hypotheses using model selection. Molecular Ecology 17: 2449–2462.A. DeprazM. CordellierJ. HausserM. Pfenninger2008Postglacial recolonization at a snail's pace (Trochulus villosus): confronting competing refugia hypotheses using model selection.Molecular Ecology1724492462
- 26. Millot C (1999) Circulation in the Western Mediterranean Sea. Journal of Marine Systems 20: 423–442.C. Millot1999Circulation in the Western Mediterranean Sea.Journal of Marine Systems20423442
- 27. Nathan R, Schurr FM, Spiegel O, Trakhtenbrot A, Tsoar A (2008) Mechanisms of long-distance seed dispersal. Trends in Ecology & Evolution 23: 638–647.R. NathanFM SchurrO. SpiegelA. TrakhtenbrotA. Tsoar2008Mechanisms of long-distance seed dispersal.Trends in Ecology & Evolution23638647
- 28. Bilton DT, Freeland JR, Okamura B (2001) Dispersal in freshwater invertebrates. Annual Review of Ecology and Systematics 32: 159–181.DT BiltonJR FreelandB. Okamura2001Dispersal in freshwater invertebrates.Annual Review of Ecology and Systematics32159181
- 29. Boag DA (1986) Dispersal in pond snails - potential role fo waterfowl. Canadian Journal of Zoology-Revue Canadienne De Zoologie 64: 904–909.DA Boag1986Dispersal in pond snails - potential role fo waterfowl.Canadian Journal of Zoology-Revue Canadienne De Zoologie64904909
- 30. Krijgsman W, Hilgen FJ, Raffi I, Sierro FJ, Wilson DS (1999) Chronology, causes and progression of the Messinian salinity crisis. Nature 400: 625–655.W. KrijgsmanFJ HilgenI. RaffiFJ SierroDS Wilson1999Chronology, causes and progression of the Messinian salinity crisis.Nature400625655
- 31. Rohling EJ, De Rijk S (1999) Holocene Climate Optimum and Last Glacial Maximum in the Mediterranean: the marine oxygen isotope record. Marine Geology 153: 57–75.EJ RohlingS. De Rijk1999Holocene Climate Optimum and Last Glacial Maximum in the Mediterranean: the marine oxygen isotope record.Marine Geology1535775
- 32. Pinceel J, Jordaens K, Pfenninger M, Backeljau T (2005) Rangewide phylogeography of a terrestrial slug in Europe: evidence for Alpine refugia and rapid colonization after the Pleistocene glaciations. Molecular Ecology 14: 1133–1150.J. PinceelK. JordaensM. PfenningerT. Backeljau2005Rangewide phylogeography of a terrestrial slug in Europe: evidence for Alpine refugia and rapid colonization after the Pleistocene glaciations.Molecular Ecology1411331150
- 33. Kalis AJ, Merkt J, Wunderlich J (2003) Environmental changes during the Holocene climatic optimum in central Europe - human impact and natural causes. Quaternary Science Reviews 22: 33–79.AJ KalisJ. MerktJ. Wunderlich2003Environmental changes during the Holocene climatic optimum in central Europe - human impact and natural causes.Quaternary Science Reviews223379
- 34. Davison K, Dolukhanov P, Sarson GR, Shukurov A (2006) The role of waterways in the spread of the Neolithic. Journal of Archaeological Science 33: 641–652.K. DavisonP. DolukhanovGR SarsonA. Shukurov2006The role of waterways in the spread of the Neolithic.Journal of Archaeological Science33641652
- 35. Turney CSM, Brown H (2007) Catastrophic early Holocene sea level rise, human migration and the Neolithic transition in Europe. Quarternary Science Reviews 26: 2036–2041.CSM TurneyH. Brown2007Catastrophic early Holocene sea level rise, human migration and the Neolithic transition in Europe.Quarternary Science Reviews2620362041
- 36. Broodbank C (2006) The origins and early development of Mediterranean maritime activity. Journal of Mediterranean Archaeology 19: 199–230.C. Broodbank2006The origins and early development of Mediterranean maritime activity.Journal of Mediterranean Archaeology19199230
- 37. Tykot RH (2002) Chemical fingerprinting and source tracing of obsidian: The Central Mediterranean trade in black gold. Accounts for Chemical Research 35: 618–627.RH Tykot2002Chemical fingerprinting and source tracing of obsidian: The Central Mediterranean trade in black gold.Accounts for Chemical Research35618627
- 38. Calo CM, Melis A, Vona G, Piras I (2008) Sardinian population (Italy): a genetic review. International Journal of Modern Anthropology 1: 39–64.CM CaloA. MelisG. VonaI. Piras2008Sardinian population (Italy): a genetic review.International Journal of Modern Anthropology13964
- 39. Tobias PV (2002) An Afro- European and Euro-African human pathway through Sardinia, with notes on humanity's world-wide water traversals and Proboscidean comparisons. Human Evolution 17: 157–173.PV Tobias2002An Afro- European and Euro-African human pathway through Sardinia, with notes on humanity's world-wide water traversals and Proboscidean comparisons.Human Evolution17157173
- 40. Fleming K, Johnston P, Zwartz D, Yokoyama Y, Lambeck K, et al. (1998) Refining the eustatic sea-level curve since the Last Glacial Maximum using far- and intermediate-field sites. Earth and Planetary Science Letters 163: 327–342.K. FlemingP. JohnstonD. ZwartzY. YokoyamaK. Lambeck1998Refining the eustatic sea-level curve since the Last Glacial Maximum using far- and intermediate-field sites.Earth and Planetary Science Letters163327342
- 41. Ricciardi A, Neves RJ, Rasmussen JB (1998) Impending extinctions of North American freshwater mussels (Unionoida) following the zebra mussel (Dreissena polymorpha) invasion. Journal of Animal Ecology 67: 613–619.A. RicciardiRJ NevesJB Rasmussen1998Impending extinctions of North American freshwater mussels (Unionoida) following the zebra mussel (Dreissena polymorpha) invasion.Journal of Animal Ecology67613619
- 42. Adhami J, Reiter P (1998) Introduction and establishment of Aedes (Stegomyia) albopictus Skuse (Diptera : Culicidae) in Albania. Journal of the American Mosquito Control Association 14: 340–343.J. AdhamiP. Reiter1998Introduction and establishment of Aedes (Stegomyia) albopictus Skuse (Diptera : Culicidae) in Albania.Journal of the American Mosquito Control Association14340343