Keywords: mitochondrial DNA, cytochromeb, control region, phylogeography, wild house mouse

Materials

Our new sequencing effort is based chiefly on samples of house mouse genomic DNA stored in the National Institute of Genetics, Mishima, Japan. These were collected in China, India, Russia and a variety of other countries, on expeditions organized by KM during 1983–2003 (MG series, stored in the National Institute of Genetics; and BRC Series, stored in the RIKEN Bio-Resource Center), and by HI and KT during 1989–1992 (HI series, stored in Hokkaido University). We also used DNA samples stored at Hokkaido University (HS series), including mice collected by PV (IZEA series, vouchered in the Institut de Zoologie et d'Ecologie Animale of Lausanne University) and KA (ANWC series, vouchered in the Australian National Wildlife Collection). Some of the same samples have been used in previous studies (for example,Yonekawaet al., 1988,1994,2003;Miyashitaet al., 1994;Nagamineet al., 1994;Tsuchiyaet al., 1994;Munclingeret al., 2002;Spiridonovaet al., 2004).

New sequences were generated for mtDNA control region (CR) from 212 house mouse individuals from 137 localities; the cytochromeb gene (Cytb) was also sequenced from a subset of 167 individuals from 106 localities (Supplementary Table S1). In our sampling, we strived to achieve maximum geographic coverage, at the cost of small sample sizes (frequently just one) for each locality. Although this reduced the scope for sophisticated analysis of population expansion scenarios, it increased the likelihood of detecting previously undiscovered components of mtDNA diversity. The geographic distribution of the new sequences is shown inFigure 1.

We downloaded a further 571 CR sequences and 41Cytb sequences ofM. musculus from public databases, drawn primarily from the work ofPrageret al. (1996,1998),Gündüzet al. (2005),Rajabi-Mahamet al. (2008) andBonhommeet al. (2011), along with representative sequences of closely related species for use as outgroups. Our sequence alignments for each mtDNA region are provided in Dryad repository: doi:10.5061/dryad.rf161.

Sequence analyses

The PCR and direct sequencing of the CR (around 800 bp;Yasudaet al., 2005) andCytb (1140 bp;Suzukiet al., 2004) were performed according to previously described methods. Two primers were used for sequence determination of CR inM. musculus; CR1: 5′-CATGCCTTGACGGCTATGTT-3′ and CR2: 5′-ATCGCCCATACGTTCCCCTT-3′. The double-stranded PCR product was sequenced utilizing the ABI PRISM Ready Reaction DyeDeoxy Terminator Cycle Sequencing Kit (Applied Biosystems, Foster City, CA, USA) and an ABI3130 automated sequencer. Sequences ofM. cypriacus,M. macedonicus,M. spicilegus andM. spretus were obtained from the databases and used as outgroups in the phylogenetic inference.

Phylogeny and divergence time estimation

Sequences were aligned by eye using MEGA5 (Tamuraet al., 2011). Before further analyses, we deleted tandem repeat sequences of 75–76 bp in CR of some MUS and some CAS haplotypes (identified byPrageret al., 1996,1998) and an 11-bp insertion in CR of some DOM sequences, while encoding the occurrence of these repeats into the taxon name to check for conformation with phyletic lineages.

To obtain a general impression of clustering topology, we constructed Neighbor-Net (NN) networks (Bryant and Moulton, 2004) for reduced data sets of 399 unique CR haplotypes and 98 uniqueCytb haplotypes, and using the default parameters of uncorrectedP distance and the EqualAngle algorithm, as implemented in SplitsTree 4.10 software (http://www.splitstree.org). The principal advantage of this hypothesis-poor method over others that generate dichotomous branching networks or trees is that NN networks illustrates all potentially supported splits among a group of sequences as a reticulation. The potential complexity of a data set is thus represented rather than reduced by this method, while any predominant network topology remains visible. Further insights into the structure of each of the CAS, MUS and DOM mtDNA lineages was obtained by constructing NN networks, together with Median-Joining (MJ) networks (Bandeltet al. 1999), as implemented in SplitsTree 4.10.

Maximum likelihood (ML) phylogenies were constructed for each of the CR andCytb data sets and for a concatenate data set for 30 individuals. We used the PhylML algorithm (Guindon and Gascuel, 2003) with the HKY substitution model, as implemented on the ATGC website (http://www.atgc-montpellier.fr/). A maximum parsimony (MP) method and the neighbor-joining (NJ;Saitou and Nei, 1987) method were taken for phylogenetic inference with concatenate sequences using PAUP 4.0b10 (Swofford, 2001). Bootstrap analysis was carried out with 1000 pseudoreplicates in the ML and NJ analyses and 100 pseudoreplicates in the MP analysis. Sub-groups are designated within each of the major mtDNA lineages only if there was moderate-to-good bootstrap support (=0.7–0.9) from the ML analysis, combined with concordant structure in the NN networks.

We estimated the age of the most recent common ancestors (TMRCAs) for mtDNA clades using theCytb sequences and a relaxed Bayesian molecular clock with uncorrelated rates (Bayesian evolutionary analysis by sampling trees (BEAST) v1.6.1,Drummond and Rambaut, 2007), as described previously (Nunomeet al., 2010b). For this analysis, we usedM. cypriacus,M. macedonicus,M. spicilegus andM. spretus, the remaining members of theM. musculus Species Group, as outgroup taxa. For the root node of theM. musculus Species Group, we assigned a previous value of 1.7 mya (95% highest posterior density: 1.45–1.95), which is based on molecular divergences of single-copy nuclear gene sequences (Irbp andRag1), calibrated against the known fossil record of the genusMus and other Murinae (Suzukiet al., 2004;Shimadaet al., 2010). The monophyletic setting was applied for clades of the lineages of the four subspecies groups (CAS, MUS, DOM and NEP) andM. m. subspecies. Then TMRCAs were estimated by the Bayesian Markov-chain Monte-Carlo (MCMC) method, using the HKY substitution model as selected under the Akaike Information Criterion in MrModeltest version 2.2 (Nylander, 2004). Analyses were run for 50 million generations from a UPGMA (Unweighted Pair Group Method with Arithmetic Mean) starting tree with sampling at every 5000 generations following 5 million burn-in generations. The convergence of MCMC chains and the effective sample size values >200 for all parameters were assessed using the software Tracer version 1.5 (Rambaut and Drummond, 2009;http://beast.bio.ed.ac.uk/Tracer). BEAST analysis was not performed with the CR sequences due to the greater inequality in branch lengths observed on the CR ML trees, compared with theCytb ML trees, suggestive of less regular substitution fixation rates over evolutionary time.

Assessment of historical demographical processes

The DnaSP programme, version 5.00.7 (Librado and Rozas, 2009), was used to estimate haplotype diversity (Hd), nucleotide diversity (π), mean number of pairwise differences among sequences (k) and Tajima'sD value. The same software was used for the analysis of mtDNA sequence mismatch distributions, measured as substitutional differences between pairs of haplotypes. Estimates of the expansion parameter tau (τ) were calculated using Arlequin version 3.5 (Excoffier and Lischer, 2010). Population expansion times were estimated under the assumption of a constant molecular clock and using mutation rates 2.5, 10 and 20% using the online tool developed bySchenekar and Weiss (2011); available athttp://www.uni-graz.at/zoowww/mismatchca lc/mmc1.php. The goodness-of-fit of the observed distribution to the expected distribution under the sudden-expansion model (Rogers, 1995) was tested by computing the sum of squares deviation (SSD).

Characteristics of major HGs

The NN network generated from theCytb data set features four well-differentiated HGs (Figure 2a), three of them correspond to the previously identified DOM, CAS and MUS lineages. The fourth HG includes two sequences derived from Nepalese mice, one reported previously byTerashimaet al. (2006; HS1467) and one new to this study (HS1523). For convenience, this HG is herein labeled NEP to indicate its geographic origin. NoCytb sequences are available for the GEN HG. The fourCytb HGs are equally divergent from a common central hub and outgroups either join this central hub or have an affinity with the MUS HG. The CAS HG appears to contain deeper lineage diversity than either the MUS or DOM HGs, both of which have distinctly brush-like terminal segments. Overall, the topology of the NN network for theCytb data set is suggestive of a more or less simultaneous diversification of an ancestralM. musculus stock into multiple evolutionary lineages and also indicative of much recent diversification in each of the DOM and MUS HGs.

The ML phylogeny generated from theCytb data set also features the same four clades with support values between 99% (DOM) and 100% (CAS) (Figure 3a). Monophyly ofM. musculus (sensu lato) is well-supported relative to the outgroups, but there is no support for any special relationships among the four HGs.

The NN network generated from the CR data set (Figure 2b) shows a well-differentiated cluster of DOM sequences but less marked segregation among the other groups, which now includes GEN. A network generated without DOM (Figure 2c) shows well-differentiated HGs for GEN and MUS, and a less cohesive cluster of 5–6 HGs that includes 4–5 that might be regarded as ‘CAS' (CAS-1, CAS-2, CAS-3, CAS-4 andAF074526) and one that includes the two NEP haplotypes (HS1467, Tukuche; HS1523, Kathmandu) as well as ‘CAS' types 13 (AF074524, Kathmandu) and 14 (AF074525, Nuwakot) fromPrageret al. (1998). The GEN and some CAS HGs are more divergent from the central hub than other HGs, but this may be due, in part, to missing data in some sequences obtained byPrageret al. (1998) from museum skins.

The ML phylogeny generated from the CR data set features five major clades with support values between 86% (CAS) and 100% (DOM) (Figure 3b). Monophyly ofM. musculus (sensu lato) is well-supported, but there is no strong support for any special relationships among the HGs, as well as in theCytb data set.

To further explore the phylogenetic relationships among the HGs, we constructed an ML phylogeny using concatenate sequences (CR+Cytb) for the 30 individuals represented in both the data sets. The resultant trees remain ambiguous for branching order among the four major lineages of CAS, DOM, NEP and MUS (Figure 4).

Genetic diversity in each of the main HGs is summarized according to a variety of standard parameters inTable 1. Excluding NEP wheren=2, forCytb the highest nucleotide diversity (Pi) is observed in DOM, followed by CAS and MUS; while for CR nucleotide diversity is highest in CAS, followed by DOM and GEN, with MUS once again the lowest. The average number of nucleotide differences (k) is the highest forCytb in DOM, followed by CAS and MUS; and the highest for CR in DOM, followed by CAS, GEN, and MUS. The number of distinct haplotypes (H) and number of polymorphic sites (S) both are clearly correlated with the total number of samples (N) in each of theCytb and CR data sets (Table 1).

Geographic distribution of major HGs

The newly determined haplotypes show geographic distributions largely consistent with expectation based on previous findings (Figure 1a). DOM haplotypes are concentrated around the Mediterranean region but show numerous widely dispersed outliers, including localities within MUS territory in western and northeastern Russia and in China and within CAS territory in the Philippines and Indonesia; MUS haplotypes are predominant in northern part of Eurasia excluding western Europe; and CAS haplotypes are predominant across South and Southeast Asia but with outliers in Japan, the Middle East and eastern Russia.

A more detailed mapping of new and previously published sequences from South Asia through to the Middle East illustrates the concentration of mtDNA diversity in southwestern Asia forM. musculus as a whole and additionally for sub-groups within CAS (Figure 1b; identity of sub-groups discussed below).

Genetic and phylogeographic structure of individual HGs

Divergence time among and within the HGs

Divergence estimates generated by BEAST for each of the four major HGs have central values that range between 0.37 and 0.46 mya (Figure 6); in each case, the 95% highest probability density values have spans of around±0.16 mya (Table 2). The TMRCA of sub-group diversification within each of the CAS, MUS, DOM and NEP HGs was estimated at 0.22±0.08, 0.15±0.07, 0.13±0.04 and 0.14±0.06 mya, respectively (Table 2).

The Tajima'sD values for all HGs and sub-groups were significantly negative, indicating various phases of rapid population growth involving mice with matrilines in CAS-1, CAS-1b, MUS-1, MUS-1b, MUS-1c and DOM (Table 3). We estimated the age of population growth in each of the phyletic groups under four different mutation rates ofCytb; 2.5, 10 and 20% (Table 3).

The mismatch distribution for the CAS-1Cytb data set (Supplementary Figure S1) shows a multi-peaked distribution, which is consistent with the notion of CAS-1 as a well-structured HG (Figure 5a). CAS-1b shows a good mismatch conformation to a model of recent population growth, with further support coming from a statistically significant negative value for Tajima'sD (Table 3). Tajima'sD was negative but not statistically significant for CAS-1a. In MUS, there is support for recent population expansion of MUS-1 as a whole and for each of MUS-1b and MUS-1c, each backed up by statistically significant negative values for Tajima'sD, though the SSD value for MUS-1 was significant (P<0.01), as evidence for departure from the estimated model of population expansion (Table 3). In contrast, there is no support for recent population expansion of either MUS-1a and MUS-2. The mismatch distribution for both DOM CR (data not shown) andCytb data sets (Supplementary Figure S1) shows near perfect conformation with the population growth and decline model provided by DNASP. Neutrality test statistics also point to a significant phase of population expansion in the recent history of DOM (Tajima'sD=−2.02,P<0.05), as concluded previously by others (Rajabi-Mahamet al., 2008;Bonhommeet al., 2011), though the SSD value was significant (P=0.024).

A homeland forM. musculus in southwestern Asia

The ancestral homeland ofM. musculus is most likely to coincide with a broad region of co-occurrence of the various phylogroups, and it should encompass or about the geographic range of the most restricted phylogroups, namely GEN and NEP of the Arabian Peninsula and Himalayan region, respectively. Under these criteria, the region of southwestern Asia, encompassing modern day Iraq, Iran, Afghanistan, Pakistan and northwestern India stands out as the most likely candidate area.Bonhommeet al. (1984) reached the same conclusion on different evidence, namely the higher levels of variation in nuclear genes among mice in this area compared with peripheral regions (see alsoSuzukiet al., 1986;Boursotet al., 1996;Boissinot and Boursot, 1997;Prageret al., 1998;Darvishet al., 2006;Duvauxet al., 2011). As discussed at length byPrageret al. (1998), mtDNA lineage boundaries in this area show general association with major geographic barriers (see alsoDuvauxet al., 2011). In particular, the Zagros Mountains divide DOM in the west from CAS in the east, while the Elburz Mountains divide MUS in the north from CAS in the south. Similarly, the mountain chains of the Hindu Kush separate populations of MUS and CAS in northern Afghanistan, though the present day distribution of the mtDNA haplotypes is not always associated with the mountainous range (for example, MUS in Kabul, Afghanistan,Figure 1a). A process of allopatric differentiation is indicated, as also suggested by the lack of overt ecological differentiation among the divergent populations.

Our divergence estimates of 0.37–0.47 mya (seeTable 2 for confidence interval) for the major mitochondrial phylogroups are in good accord with previous determinations (Rajabi-Mahamet al., 2008;Terashimaet al., 2006). Initial lineage diversification evidently predates the dispersal of modern humans out of Africa; hence it is likely that initial phases of range expansion were not mediated by human activity, unless of course the impact of early human populations on the environment was much greater than currently understood.

An interesting biogeographic observation is that the inferred place of origin ofM. musculus (i.e. southwestern Asia) lacks any other co-occurring mouse species belong to subgenusMus. In this regard, it differs from each of those found in peninsular India, whereM. booduga andM. terricolor of theM. booduga Species Groups are both found (Musser and Carleton, 2005); Indochina, which hosts a variety of species in theM. booduga andM. cervicolor Species Groups (Suzuki and Aplin, 2012); and eastern Europe, where other species of theM. musculus species group are present. It is tempting to speculate that the presence of these ecologically similar native species in surrounding areas formerly served to constrain the geographic distribution ofM. musculus.

The distribution and ecology of contemporarycastaneus populations in Asia provides further clues to its regional history. As summarized byMarshall (1977): 205–206 the only ‘outdoor commensal' (that is, agricultural field) populations of CAS mice are found in the semi-arid habitats of Pakistan (thebactrianus morphotype). Elsewhere on the Indian subcontinent and through into Southeast Asia, house mice are found only as ‘indoor commensals' furthermore, across Southeast Asia, they are generally confined to larger towns and absent in rural villages (see alsoAplinet al., 2006).Marshall (1977) attributed the absence of house mice from agricultural contexts in these areas to competitive exclusion by other species ofMus, notably members of theMus booduga Species Group on the Indian subcontinent and members of theMus cervicolor Species Group in South East Asia; and he attributed the absence of house mice in rural villages in Southeast Asia to the presence of commensal species ofRattus such asR. exulans.

Phylogeography of the CAS lineage

The CAS lineage has been subject to two different phylogeographic interpretations.Boursotet al. (1993,1996) proposed that the northern Indian subcontinent was both the place of origin ofM. musculus and the cradle of genetic diversity within this group. This model was based on the discovery in this area of numerous highly divergent mtDNA lineages (Boursotet al., 1996) and levels of nuclear diversity (as determined by allozyme electrophoresis) that exceeded those found in European populations ofdomesticus andmusculus (Dinet al., 1996). To explain these dual observations,Boursotet al. (1993,1996) proposed a ‘centrifugal' model of differentiation in which the ancestors of each of thedomesticus,musculus andcastaneus lineages dispersed to the west, east and north, each carrying a subset of the mtDNA and nuclear diversity, and subsequently undergoing local differentiation. They referred to populations in the ancestral area as ‘Mus musculus subsp.' and restricted use of the namecastaneus to populations in southern India, southern China and Indochina.Boursotet al. (1996) referred to the northern Indian and Pakistani populations as an ‘oriental group', whileYonekawaet al. (1994) subsequently applied the existing namebactrianus to these populations.

The version of CAS phylogeography byPrageret al. (1998) is based primarily on interpretation of mtDNA phylogeny. Although confirming a high diversity of mtDNA types in northern India and Pakistan, they regarded these to be part of a monophyleticcastaneus lineage distinct from each ofdomesticus,musculus and the newly recognizedgentilulus lineage of the Arabian Peninsula.Prageret al. (1998) developed a model of ‘sequential' derivation of the lineages to reflect their phylogenetic branching order—domesticus being the oldest branch, followed bygentilulus,castaneus andmusculus. They preferred to locate the ancestral pre-domesticus stock in the Near East, within the current range ofdomesticus, and regarded the progressive derivation of other lineages as a consequence of sequential dispersal events that took house mice south onto the Arabian Penisula, then east onto the Indian subcontinent and, finally, north through the mountains of northwest India and Pakistan to occupy the great Eurasian steppe. Within CAS,Prageret al. (1998) suggest a relatively long phase of regional diversification on the Indian subcontinent, followed by a ‘more recent' dispersal into the ‘humid lowlands of Southeast Asia'.

Our sampling for CAS is relatively extensive and the results go far towards illuminating the historical phylogeography of this HG. In keeping with the findings ofPrageret al. (1998) and contrary to the predictions ofBoursotet al. (1993,1996), we recovered reciprocal monophyly with good to excellent support among all of the major HGs, including CAS. AlthoughPrageret al. (1998) considered the branching order among the major HGs to be resolved, our larger CR andCytb data set fails to provide a robust phylogenetic structure at this level, although there is a suggestion of special affinity between CAS and MUS and between DOM and NEP. Like both groups of previous researchers, we found the highest mtDNA diversity and depth in CAS populations inhabiting the mountainous region of northwest India and Pakistan, with a loss of haplotype lineage diversity from north to south on the Indian subcontinent (Boursotet al., 1996) and from west to east into Southeast Asia (Prageret al., 1998). Despite this general agreement,Boursotet al. (1996) clearly regardcastaneus in their restricted application of the name to be a long-term resident of Southeast Asia, whilePrageret al. (1998) portray this as a relatively recent phase of dispersal ofcastaneus, though without specifying any time frame.

Low nucleotide diversity in the widely distributed CAS-1 sub-group is evident in this study. This is consistent with that observed by the recent work on thecastaneus subspecies group done byRajabi-Mahamet al. (2012) (see alsoBonhomme and Searle, 2012). Our results are suggestive of a relatively recent range expansion of CAS-1 to a large geographic areas covering the south and east Indian subcontinent, Southeast Asia, Indonesia, South China, Northeast China and the Russian Far East (Figure 5). On the other hand, the presence of the locally restricted phyletic group, CAS-1a is suggestive of stepwise historical range expansion of CAS-1. Haplotype diversity within CAS-1a, the sub-group found in mice from South China (Kunming and Guilin), northern Honshu, Hokkaido and South Sakhalin, was most likely produced by subsequent dispersal and is suggestive of several thousands of years ofin situ evolution. Furthermore, the location of the Japanese cluster at the far eastern periphery of the CAS distribution implies a significantly earlier onset for dispersal onto the Indian subcontinent and thence through to East Asia.

We suspect that the dispersal of CAS-1 mice occurred in response to ecological transformation of the landscape by early agriculturalists and the emergence of urban centers and trade networks. As has been postulated for the Middle East (Auffrayet al., 1990;Cucchi and Vigne, 2006), South Asian populations ofM. musculus are likely to have benefited from the creation of new agricultural landscapes, and the common practice of storing harvested grain inside villages and even inside houses provided the context for development of commensalism. Long-distance dispersal is part and parcel of commensalism, with mice being carried as stowaways during transport of grain, building materials, clothing and bedding (Pococket al., 2005).

Although the archaeological record of agriculture is less comprehensive for Asia than for the Middle East and Europe, there is good evidence for domestication of cereal crops, including rice and millet, by about 9000 years ago in several parts of South and East Asia (Khush 1997;Londoet al., 2006;Zhenget al., 2009;Molinaet al., 2011) and even earlier evidence for long-distance overland and maritime trade (Oka and Kusimba, 2008). Assuming that populations experienced a sudden or exponential growth, we calculatedτ values from theCytb sequences and estimated times since the onset of population expansions. Higher rates of mutation (for example, 10% or 20% per million years per lineage) rather than lower rates (for example, 2.5%) are considered to be realistic for assessing rather recent diversifying events (Hoet al., 2005). We obtained aτ value of 1.7 for CAS-1b (n=17), which under mutation rates (per million years per lineage) of 10 and 20% gives expansion times of 7600 and 3800 years, respectively (Table 3). Aτ value was not calculated for CAS-1a, but it too is likely to have commenced its dispersal and diversification in China, the Russian Far East and Japan in prehistoric times. In this regard, it is of interest to note archaeological evidence for rice cultivation along the upper Yangtze river (for example, Yunnan province, here represented by mice from Kunming) at 4500 years ago (Fulleret al., 2010) and recent genetic evidence from an intensive genome survey on wild and cultivated rice, suggesting the Pearl River (Guangxi province, here represented by mice from Guilin) in southern China is the place of the first development of cultivated rice (Bonhomme and Searle, 2012;Huanget al., 2012).

Comparatively recent long-distance dispersal of CAS mice most likely explains the detection of CAS-2 haplotypes at isolated localities in Taiwan and Vietnam, though we could not exclude out the possibility that these are relictual haplotypes, either rare survivors of an earlier dispersal of CAS-2 mice out of India that was swamped by a later CAS-1 dispersal or the last remnants of incomplete lineage sorting of an immigrant population with a mixture of CAS-1 and CAS-2 haplotypes. In the case of the individual from Taiwan, the fact that its nuclear genetic profile is fully consistent with other East Asian populations of CAS and differ from mice from India and Pakistan with the CAS-2 mtDNA haplotypes (Nunomeet al., 2010a; Kodamaet al., unpublished) suggests that we are not dealing with a novel invader but perhaps with a product of mtDNA introgression.

Phylogeography of the MUS lineage

Previous phylogeographic interpretations of the house mouse group do not vary much with regard to the geographic origin of the MUS HG.Boursotet al. (1993) speculated that ‘the cradle ofM. m. musculus could be in Transcaucasia or east of the Caspian Sea', whilePrageret al. (1993,1996) saw the origin of MUS as the product of northward dispersal from a proto-CAS population occupying the region east of the Caspian Sea, followed by range expansion. Both groups of researchers also agree that MUS populations subsequently dispersed west into central Europe and east into China and Japan, and this scenario has been adopted as paradigmatic by Japanese researchers interested in the origin of the indigenousmolossinus population (Yonekawaet al., 1988;Terashimaet al., 2006;Nunomeet al., 2010a).Yonekawaet al. (1988) postulated that MUS populations relatively recently expanded into China where CAS populations had already colonized, but few other workers have expressed an opinion on the earlier timing of the remarkable eastward expansion of MUS.Nunomeet al. (2010a) suggested a latitudinal division within MUS between the northern (MUS-I) and southern (MUS-II) groups, based on phylogeographic analyses of nuclear gene sequences, and posited that range expansion of the MUS HG from west to east across continental Eurasia followed separate northern and southern dispersal routes, with separate expansion again into eastern Europe.

Much of the interest in the geographic distribution of MUS has focused on its genetic interaction with mice of other HGs. In the European context, numerous studies have examined the evolutionary dynamics of a narrow hybrid zone with DOM that runs from Norway through Denmark, Germany and Austria to eastern Bulgaria (Hunt and Selander, 1973;Sageet al., 1993;Boursotet al., 1993;Joneset al., 2010). There are grounds to believe that initial contact may have occurred further west in Europe with the current position stabilizing after a period of eastward retreat ofmusculus (Gyllensten and Wilson, 1987). Whatever the case, the age of the contact zone is constrained by the timing of the DOM migrations along the shores of the Mediterranean, an event that is thought to date to within the past 2–3000 years (Cucchiet al., 2005).

In Transcaucasia, gene flow between complexly parapatric populations of MUS and DOM is thought to explain a 300-km wide zone of genetic admixture (Mezhzherinet al., 1998;Frismanet al., 1990;Milishnikovet al., 1990); however, an alternative interpretation attributes the genetic diversity to a high level of ancestral polymorphism in the regional MUS population (Milishnikovet al., 2004), equivalent to that observed among the ‘oriental group' of mice in northern India and Pakistan (Boursotet al., 1993,1996;Dinet al., 1996). This would be consistent with long residency of the MUS population in this area. MUS and CAS populations also come into secondary contact in China (Moriwakiet al., 1994); however, both the geography and the genetic outcome of these interactions remain poorly documented.

Our expanded sampling among eastern house mouse populations sheds significant new light on the evolutionary history of the MUS HG. We identify two major sub-groups within MUS—MUS-1 and MUS-2—and a total of three phylogeographic components within MUS-1: MUS-1a in Moldova, Ukraine, N Caspian Sea and Russian Siberia; MUS-1b in East Europe, Kazakhstan and China; and MUS-1c in Korea and Japan. The origin of the MUS-1 and MUS-2 sub-groups is ancient, with a divergence estimate from BEAST of 150 000±13 000 years (Figure 6,Table 2). Both sub-groups are represented in the area around the Caspian Sea, and it seems likely that both matrilines originated within this ancestral geographic area.

Rapid population expansion was inferred for each of MUS-1b and MUS-1c (Table 3). Estimates of expansion times for these lineages (Table 3) suggest an early expansion of MUS-1b in northern China (τ=4.9 CI: 2.9–6.5; for example, 21 000 and 10 800 years ago, with an assumption of the mutation rate of 10% and 20% per million years per lineage, respectively), followed by a later expansion of MUS-1c in northeastern Russia, Korea and Japan at (τ=1.5 CI: 0.7–2.5; for example, 6600 and 3300 years ago).

The notion of ancient population expansions in eastern Eurasia is clearly at odds with the conventional notion of a recent west-to-east dispersal of the MUS HG. However, other lines of genetic evidence similarly point to a long residency of the MUS HG in central Russia and the Far East. For example, the beta-hemoglobin gene (Hbb) shows contrasting predominant alleles in the lower Yellow River basin and in the remaining western portion of northern China (Hbb^p andHbb^w1, respectively;Miyashitaet al., 1994; see alsoMoriwaki, 1994); and mice from the eastern part of China are known to have relatively longer tails (tail ratio: ∼93%) than those from the rest of MUS territory in China (81%Tsuchiyaet al., 1994).

Finally, we note that the area in which MUS-1c is predominant—the Korean Peninsula and nearby continental area—harbors unique genetic components in both Y-specific gene sequences and nuclear gene sequences (for example,Nagamineet al., 1994;Terashimaet al., 2006;Nunomeet al., 2010a). Under the existing paradigm of west-to-east dispersal, these phylogeographic patterns might be attributed either to genetic drift following migration of ancestral populations with diverse genetic components or to multiple migration events by mice carrying different genetic components, perhaps by different routes. However, neither of these scenarios can readily account for the evidence of ancient population expansions within geographically restricted matrilines. Accordingly, we favor the alternative model of regional differentiation within a long-term resident population.

The fossil record should be able to arbitrate this issue, and it is of great interest to note that paleontologists have long recognizedM. musculus as a component of the Chinese mammal fauna since the middle part of the Middle Pleistocene (that is, ca. 500 000 years ago); for example,Zhenget al. (1997) and references cited therein. Although the taxonomic identification of the fossils might be challenged, the determination is at least plausible given the molecular evidence for early diversification among Chinesemusculus populations. However, there is a risk of circularity in such arguments and an urgent need for critical appraisal of the relevant fossils.

The European sub-group MUS-1a contains substantial haplotype diversity, including persistent ancestral haplotypes and two deeply divided haplotype series, each of which contains relatively shallow stellar clusters derived from populations near the western limit of the MUS geographic range. This pattern is suggestive of a broad westward expansion of a MUS population into eastern Europe, with limited filtering of haplotype diversity. As summarized byAuffrayet al. (1990), the long history ofM. musculus in Europe is dominated by large expansions and contractions of range driven by glacial cycles. At the height of the last glaciationM. musculus was rare or absent across most of eastern Europe, which supported a mosaic of periglacial forest-steppe, steppe and semi-desert habitats (Markovaet al., 2009). Refugial forest habitats were restricted to small patches in the Crimea, in the Transcarpathian region and in the Caucasus (Markovaet al., 2009), and it is of interest to note fossil occurrences ofM. musculus in the Carpathian–Balkan region during the warm interval (33–24 000 years ago) immediately before the last glacial maximum (Markovaet al., 2010). However, in view of the high level of genetic diversity within MUS-1a and the lack of a strong signal of recent population expansion, it seems likely that mice persisted in multiple localities, perhaps including both forest and semi-desert habitats. This issue warrants further consideration.

MUS-1a contains a discrete lineage characterized by a 75-bp duplication, first detected byPrageret al. (1998) in a mouse from Kishinev in Moldova. We found closely related haplotypes at low frequency in mice from eastern Europe (for example, Donetsk, Ukraine) and also from Khasan in Primorye, Russia (Figure 5b). Given the other evidence of regional differentiation of mtDNA within MUS, we are inclined to view MUS-1a as originally restricted to eastern Europe (Ukraine), with its more easterly occurrences being a product of long-range transport by modern means. The locality of Novosibirsk, for example, is situated on the Turkestan–Siberia Railway that was built in the early twentieth century (in 1930) and connects the Caspian Sea to localities in Central Asia. The link to the Primorye region of the Russia Far East is less readily accounted for by overland transportation but might be explained by the activities of the Russian government to introduce kazak and peasants to the Russian Far East in the late nineteenth century; upwards of 90 000 people (and perhaps a few mice) from Odessa in the Ukraine settled in the Ussuri Region of Primorye (http://www.fegi.ru/prim/geografy/etap.htm).

Phylogeography of the DOM lineage

Our small number of new DOM sequences contributes only a few insights into the history of this well-studied lineage (Gabrielet al., 2011;Bonhommeet al., 2011;Joneset al., 2010). The onset of the expansion of DOM is estimated to be 12 000 years ago as the youngest timing, assuming the mutation rate of 20% per million years per lineage (Table 3), which is harmonious with the recent arguments based on zooarcheological records (Cucchiet al., 2005;Rajabi-Mahamet al., 2008;Bonhomme and Searle, 2012).

We recovered the expected ‘Clade F' haplotypes from mice collected in North America, Australia and Africa (Senegal, Somalia) but also detected them in mice from several localities in Asia, namely Lanzhou and Xining in China, and Bogor on Java in Indonesia. At Bogor, CAS and DOM mtDNA haplotypes were found to co-occur in one population.

A high frequency of DOM haplotypes was also detected in the Russian Far East, thereby supporting previous claims of DOM–MUS–CAS interactions in this area based on studies of chromosomes, allozymes and random-amplified-polymorphic DNA (RAPD) markers (Frismanet al., 2011;Spiridonovaet al., 2011). Interestingly though, the DOM haplotypes recovered at Primorye (HS1466) and Sakhalin (HS3606, HS3607) are not ‘Clade F' but are related specifically to haplotypes from Cameroon (for example, AFWCMR41;Bonhommeet al., 2011). This connection is very likely explained by long-distance dispersals associated with human activities in modern times.

The detection of DOM haplotypes in numerous corners of the world is testimony to the ongoing dispersal ofM. musculus and encourages further study of the impact of occasional arrival of ‘exotic' mice on the genetic constitution of pre-established mouse populations (Rajabi-Mahamet al., 2008;Searleet al., 2009a,2009b;Gabrielet al., 2010;Bonhommeet al., 2011). To further illustrate this point, mice with both DOM and CAS mtDNA haplotypes have been captured in Japanese international ports (Tsudaet al., 2007) andNunomeet al. (2010a) provided robust evidence from their nuclear haplotype analysis of genetic introgression by DOM components of Japanese house mice. The extent to which genetic introgression may now be shaping the future evolution of the house mouse is an interesting topic—one that has bearing on other commensal mammals, including the black ratRattus rattus which also displays comparable signals of former geographic sub-division and recent intermingling as a consequence of commensalism and human-assisted dispersal (Chinenet al., 2005;Aplinet al., 2011;Bastoset al., 2011;Lacket al., 2012).

Concluding remarks

The expanded mtDNA data set raises a number of important new issues regarding the prehistory of the house mouse. Most significantly, it has identified one particular CAS sub-group (CAS-1) that has expanded into southern India, Southeast and East Asia and raised the possibility that this expansion is linked to the emergence of agricultural lifestyles and of Asian civilizations. Also of significance is our suggestion that MUS populations have a long history of residency in eastern Russia and China, contrary to the existing paradigm of recent expansion from west to east. Finally, our results emphasize the role of long-distance dispersal in shaping contemporary pattern of distribution and opportunities for interaction between each of the major lineages withinM. musculus.

Our study also demonstrates the value of continuing efforts to fill gaps in geographic coverage ofM. musculus mtDNA. Moreover, it highlights the need for ongoing field collection to increase local sampling and the need for more comprehensive assessments of population genetic history using nuclear markers. From our preliminary work with nuclear genes on this group, it is clear that much deeper divergence between subspecies groups is observed in some regions of the genome than in others (for example,Suzukiet al., 2004;Nunomeet al., 2010a) and is also evident that different markers can yield strongly contrasting phylogeographic structure, such as in southern China where CAS mtDNA dominates but both CAS and MUS components are detected in nuclear genes (for example,Nunomeet al., 2010a). Finally, it is worth mentioning the as yet unexplored potential for detailed study of Central and East Asian house mouse populations to reveal important new aspects of human history, including the emergence of agricultural lifestyles and of regional trade networks.

1. Abe K, Noguchi H, Tagawa K, Yuzuriha M, Toyoda A, Kojima T, et al. Contribution of Asian mouse subspecies Mus musculus molossinus to genomic constitution of strain C57BL/6J as defined by BAC-end sequence-SKIP analysis. Genom Res. 2004;14:2439–2447. doi: 10.1101/gr.2899304. [DOI] [PMC free article] [PubMed] [Google Scholar]
2. Aplin KP, Brown PR, Singleton GR, Douangboupha B, Khamphoukheo K.2006Rodents in the rice environments of LaosIn: Schiller JM, Chanphengxay MB, Linquist B, Appa Rao S (eds). Rice in Laos. International Rice Research Institute: Los Banos, Philippines; 291–308. [Google Scholar]
3. Aplin K, Suzuki H, Chinen AA, Chesser RT, ten Have J, Donnellan SC, et al. Multiple geographic origins of commensalism and complex dispersal history of black rats. PLoS ONE. 2011;6:e26357. doi: 10.1371/journal.pone.0026357. [DOI] [PMC free article] [PubMed] [Google Scholar]
4. Auffray J-C, Vanlerberghe F, Britton-Davidian J. The house mouse progression in Eurasia: a palaeontological and archaeozoological approach. Biol J Linne Soc. 1990;41:13–25. [Google Scholar]
5. Bandelt HJ, Forster P, Rohl A. Median-joining networks for inferring intraspecific phylogenies. Mol Biol Evol. 1999;16:37–48. doi: 10.1093/oxfordjournals.molbev.a026036. [DOI] [PubMed] [Google Scholar]
6. Bastos A, Nair D, Taylor P, Brettschneider H, Kirsten F, et al. Genetic monitoring detects an overlooked cryptic species and reveals the diversity and distribution of three invasive Rattus congeners in south Africa. BMC Genet. 2011;12:26. doi: 10.1186/1471-2156-12-26. [DOI] [PMC free article] [PubMed] [Google Scholar]
7. Boissinot S, Boursot P. Discordant phylogeographic patterns between the Y chromosome and mitochondrial DNA in the house mouse—selection on the Y chromosome. Genetics. 1997;146:1019–1034. doi: 10.1093/genetics/146.3.1019. [DOI] [PMC free article] [PubMed] [Google Scholar]
8. Bonhomme F, Catalan J, Britton-Davidson J, Chapman VM, Moriwaki K, Nevo E, et al. Biochemical diversity and evolution in the genus Mus. Biochem Genet. 1984;22:275–303. doi: 10.1007/BF00484229. [DOI] [PubMed] [Google Scholar]
9. Bonhomme F, Orth A, Cucchi T, Rajabi-Maham H, Catalan J, Boursot P, et al. Genetic differentiation of the house mouse around the Mediterranean basin: matrilineal footprints of early and late colonization. Proc R Soc B. 2011;278:1034–1043. doi: 10.1098/rspb.2010.1228. [DOI] [PMC free article] [PubMed] [Google Scholar]
10. Bonhomme F, Searle JB.2012House mouse phylogeographyIn: Macholán M, Baird SJE, Munclinger P, Piálek J (eds).Evolution of the house mouse (Cambridge series in morphology and molecules) Cambridge University Press: Cambridge, UK; 278–296. [Google Scholar]
11. Boursot P, Auffray JC, Britton-Davidian J, Bonhomme F. The evolution of house mice. Ann Rev Ecol Syst. 1993;24:119–152. [Google Scholar]
12. Boursot P, Din W, Anand R, Darviche D, Dod B, von Deimling F, et al. Origin and radiation of the house mouse: mitochondrial DNA phylogeny. J Evol Biol. 1996;9:391–415. [Google Scholar]
13. Bryant D, Moulton V. Neighbor-Net: an agglomerative method for the construction of phylogenetic networks. Mol Biol Evol. 2004;21:255–265. doi: 10.1093/molbev/msh018. [DOI] [PubMed] [Google Scholar]
14. Chinen AA, Suzuki H, Aplin KP, Tsuchiya K, Suzuki S. Preliminary genetic characterization of two lineages of black rats (Rattus rattus sensu lato) in Japan with evidence for introgression at several localities. Gene Genet Syst. 2005;80:367–375. doi: 10.1266/ggs.80.367. [DOI] [PubMed] [Google Scholar]
15. Cucchi T, Vigne JD, Auffray JC. First occurrence of the house mouse (Mus musculus domesticus, Schwarz and Schwarz 1943) in the Western Mediterranean: a zooarchaeological revision of subfossil occurrences. Biol J Linn Soc. 2005;84:429–445. [Google Scholar]
16. Cucchi T, Vigne JD. Origin and diffusion of the house mouse in the Mediterranean. Hum Evol. 2006;21:95–106. [Google Scholar]
17. Darvish J, Orth A, Bonhomme F. Genetic transition in the house mouse, Mus musculus of Eastern Iranian Plateau. Folia Zool. 2006;55:349–357. [Google Scholar]
18. Din W, Anand R, Boursot P, Darviche D, Dod B, Jouvin-Marche E, et al. Origin and radiation of the house mouse: clues from nuclear genes. J Evol Biol. 1996;9:519–539. [Google Scholar]
19. Drummond AJ, Rambaut A. Beast: Bayesian evolutionary analysis by sampling trees. BMC Evol Biol. 2007;7:214. doi: 10.1186/1471-2148-7-214. [DOI] [PMC free article] [PubMed] [Google Scholar]
20. Duplantier JM, Orth A, Catalan J, Bonhomme F. Evidence for a mitochondrial lineage originating from the Arabian peninsula in the Madagascar House Mouse. Heredity. 2002;89:154–158. doi: 10.1038/sj.hdy.6800122. [DOI] [PubMed] [Google Scholar]
21. Duvaux L, Belkhir K, Boulesteix M, Boursot P. Isolation and gene flow: inferring the speciation history of European house mice. Mol Ecol. 2011;20:5248–5264. doi: 10.1111/j.1365-294X.2011.05343.x. [DOI] [PubMed] [Google Scholar]
22. Excoffier L, Lischer HEL. Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Mol Ecol Resour. 2010;10:564–567. doi: 10.1111/j.1755-0998.2010.02847.x. [DOI] [PubMed] [Google Scholar]
23. Frazer KA, Eskin E, Kang HM, Bogue MA, Hinds DA, Beilharz EJ, et al. A sequence-based variation map of 8.27 million SNPs in inbred mouse strains. Nature. 2007;448:1050–1053. doi: 10.1038/nature06067. [DOI] [PubMed] [Google Scholar]
24. Frisman LV, Korobitsyna KV, Yakimenko LV, Vorontsov NN. Biochemical groups of house mice inhabiting the Soviet Union, in Evolyutsionnye geneticheskie issledovaniya mlekopitayushchikh: Tezisy dokladov (Evolutionary Genetic Studies in Mammals: Proc. Conf.), Vladivostok: Dal'nevost. Otd Akad Nauk SSSR. 1990;1:35–54. [Google Scholar]
25. Frisman LV, Korobitsyna KV, Yakimenko LV, Munteanu AI, Moriwaki K. Genetic variability and the origin of house mouse from the territory of Russia and neighboring countries. Russ J Genet. 2011;47:590–602. [PubMed] [Google Scholar]
26. Fuller DQ, Sato YI, Castillo C, Qin L, Weisskopf A, Kingwell-Banham E, et al. Consilience of genetics and archaeobotany in the entangled history of rice. Archaeol Anthropol Sci. 2010;2:115–131. [Google Scholar]
27. Gabriel SI, Jóhannesdóttir F, Jones EP, Searle JB. Colonization, mouse-style. BMC Biol. 2010;8:131. doi: 10.1186/1741-7007-8-131. [DOI] [PMC free article] [PubMed] [Google Scholar]
28. Gabriel SI, Stevens MI, Mathias ML, Searle JB. Of mice and 'convicts': origin of the Australian house mouse, Mus musculus. PLoS ONE. 2011;6:et622. doi: 10.1371/journal.pone.0028622. [DOI] [PMC free article] [PubMed] [Google Scholar]
29. Guindon S, Gascuel O. A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst Biol. 2003;52:696–704. doi: 10.1080/10635150390235520. [DOI] [PubMed] [Google Scholar]
30. Gündüz İ, Rambau RV, Tez C, Searle JB. Mitochondrial DNA variation in the western house mouse (Mus musculus domesticus) close to its site of origin: studies in Turkey. Biol J Linn Soc. 2005;84:473–485. [Google Scholar]
31. Gyllensten U, Wilson AC. Interspecific mitochondrial DNA transfer and the colonization of Scandinavia by mice. Genet Res. 1987;49:25–29. doi: 10.1017/s0016672300026690. [DOI] [PubMed] [Google Scholar]
32. Ho SYW, Phillips MJ, Cooper A, Drummond AJ. Time dependency of molecular rate estimates and systematic overestimation of recent divergence times. Mol Biol Evol. 2005;22:1561–1568. doi: 10.1093/molbev/msi145. [DOI] [PubMed] [Google Scholar]
33. Huang X, Kurata N, Wei X, Wang ZX, Wang A, Zhao Q, et al. A map of rice genome variation reveals the origin of cultivated rice. Nature. 2012;490:497–501. doi: 10.1038/nature11532. [DOI] [PMC free article] [PubMed] [Google Scholar]
34. Hunt WG, Selander RK. Biochemical genetics of hybridization in European house mouse. Heredity. 1973;31:11–33. doi: 10.1038/hdy.1973.56. [DOI] [PubMed] [Google Scholar]
35. Jones EP, van der Kooij J, Solheim R, Searle JB. Norwegian house mice (Mus musculus musculus/domesticus): distributions routes of colonization and patterns of hybridization. Mol Ecol. 2010;19:5252–5264. doi: 10.1111/j.1365-294X.2010.04874.x. [DOI] [PubMed] [Google Scholar]
36. Khush GS. Origin dispersal cultivation and variation of rice. Plant Mol Biol. 1997;35:25–34. [PubMed] [Google Scholar]
37. Lack JB, Greene DU, Conroy CJ, Hamilton MJ, Braun JK, Mares MA, et al. Invasion facilitates hybridization with introgression in the Rattus rattus species complex. Mol Ecol. 2012;21:3545–3561. doi: 10.1111/j.1365-294X.2012.05620.x. [DOI] [PubMed] [Google Scholar]
38. Librado P, Rozas J. DnaSP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics. 2009;25:1451–1452. doi: 10.1093/bioinformatics/btp187. [DOI] [PubMed] [Google Scholar]
39. Londo JP, Chiang YC, Hung KH, Chiang TY, Schaal BA. Phylogeography of Asian wild rice Oryza rufipogon reveals multiple independent domestications of cultivated rice Oryza sativa. Proc Natl Acad Sci USA. 2006;103:9578–9583. doi: 10.1073/pnas.0603152103. [DOI] [PMC free article] [PubMed] [Google Scholar]
40. Markova AK, Simakova AN, Puzachenko AY. Ecosystems of Eastern Europe at the time of maximum cooling of the Valdai glaciation (24–18 kyr BP) inferred from data on plant communities and mammal assemblages. Quat Internat. 2009;201:53–59. [Google Scholar]
41. Markova AK, Puzachenko AY, van Kolfschoten T. The North Eurasian mammal assemblages during the end of MIS 3 (Brianskian–Late Karginian–Denekamp Interstadial) Quat Internat. 2010;212:149–158. [Google Scholar]
42. Marshall JT. A synopsis of Asian species of Mus (Rodentia, Muridae) Bull Am Mus Nat Hist. 1977;158:173–220. [Google Scholar]
43. Mezhzherin SV, Kotenkova EV, Mikhailenko AG. The house mice, Mus musculus s.l., hybrid zone of trans-Caucasus. Zeitschrift für Söugetierkunde. 1998;63:154–168. [Google Scholar]
44. Milishnikov AN, Lavrenchenko LA, Lavrenchenko LA, Orlov VN. High-level introgression of Mus domesticus genes in Transcaucasian Mus musculus s. str. populations. Dokl Akad Nauk SSSR. 1990;311:764–768. [PubMed] [Google Scholar]
45. Milishnikov AN, Lavrenchenko LA, Lebedev VS. Origin of the house mice (superspecies complex Mus musculus sensu lato) from the Transcaucasian region: a new look at dispersal routes and evolution. Genetika. 2004;40:1234–1250. [PubMed] [Google Scholar]
46. Miyashita N, Kawashima T, Wang CH, Jin ML, Wang F, Gotoh H, et al. 1994Genetic polymorphisms of Hbb haplotypes in wild miceIn: Moriwaki K, Shiroishi T, Yonekawa H (eds).Genetics in Wild Mice Japan Scientific Society Press/S Karger: Tokyo, Japan/Basel, Switzerland; 85–93. [Google Scholar]
47. Molina J, Sikora M, Garud N, Flowers JM, Rubinsteina S, Reynoldsb A, et al. Molecular evidence for a single evolutionary origin of domesticated rice. Proc Natl Acad Sci USA. 2011;108:8351–8356. doi: 10.1073/pnas.1104686108. [DOI] [PMC free article] [PubMed] [Google Scholar]
48. Moriwaki K.1994Wild mouse from geneticist's viewpointIn: Moriwaki K, Shiroishi T, Yonekawa H (eds).Genetics in Wild Mice Japan Scientific Society Press/S Karger: Tokyo, Japan/Basel, Switzerland; xiii–xxiv. [Google Scholar]
49. Moriwaki K, Yonekawa H, Gotoh O, Minezawa M, Winking H, Gropp A. Implications of the genetic divergence between European wile mice with Robertsonian translocations from the viewpoint of mitochondrial DNA. Genet Res. 1984;43:277–287. doi: 10.1017/s0016672300026069. [DOI] [PubMed] [Google Scholar]
50. Moriwaki K, Shiroishi T, Yonekawa H. Genetics in Wild Mice. It's application to Biomedical Research. Japan Scientific Societies Press/S Karger: Tokyo, Japan/Basel, Switzerland; 1994. [Google Scholar]
51. Munclinger P, Bozıkova E, Sugerkova M, Pialek J, Macholan M. Genetic variation in house mice (Mus Muridae Rodentia) from the Czech and Slovak Republics. Folia Zool. 2002;51:81–92. [Google Scholar]
52. Musser GG, Carleton MD.2005Family MuridaeIn: Wilson DE, Reeder DM (eds).Mammal Species of the World3rd edn.The John Hopkins University Press: Baltimore, MD, USA; 894–1531. [Google Scholar]
53. Nagamine CM, Shiroishi T, Miyashita N, Tsuchiya K, Ikeda H, Namikawa T, et al. Distribution of the mollosinus allele of Sry the testis-determining gene in wild mouse. Mol Biol Evol. 1994;11:864–874. doi: 10.1093/oxfordjournals.molbev.a040169. [DOI] [PubMed] [Google Scholar]
54. Nunome M, Ishimori C, Aplin KP, Yonekawa H, Moriwaki K, Suzuki H. Detection of recombinant haplotypes in wild mice (Mus musculus) provides new insights into the origin of Japanese mice. Mol Ecol. 2010a;19:2474–2489. doi: 10.1111/j.1365-294X.2010.04651.x. [DOI] [PubMed] [Google Scholar]
55. Nunome M, Torii H, Matsuki R, Kinoshita G, Suzuki H. The influence of Pleistocene refugia on the evolutionary history of the Japanese hare Lepus brachyurus. Zool Sci. 2010b;27:7469–7754. doi: 10.2108/zsj.27.746. [DOI] [PubMed] [Google Scholar]
56. Nylander JAA. MrModeltest v2. Program distributed by author. Evolutionary Biology Centre. Uppsala University: Uppsala, Sweden; 2004. [Google Scholar]
57. Oka R, Kusimba C. The archaeology of trading systems Part 1: towards a new trade synthesis. J Archaeol Res. 2008;16:339–395. [Google Scholar]
58. Pocock MJO, Hauffe HC, Searle JB. Dispersal in house mice. Biol J Linn Soc. 2005;84:565–583. [Google Scholar]
59. Prager EM, Sage RD, Gyllensten ULF, Thomas WK, Huebner R, Jones CS, et al. Mitochondrial DNA sequence diversity and the colonization of Scandinavia by house mice from East Holstein. Biol J Linn Soc. 1993;50:85–122. [Google Scholar]
60. Prager EM, Tichy H, Sage RD. Mitochondrial DNA sequence variation in the eastern house mouse Mus musculus: comparison with other house mice and report of a 75-bp tandem repeat. Genetics. 1996;143:427–446. doi: 10.1093/genetics/143.1.427. [DOI] [PMC free article] [PubMed] [Google Scholar]
61. Prager EM, Orrego C, Sage RD. Genetic variation and phylogeography of Central Asian and other house mice including a major new mitochondrial lineage in Yemen. Genetics. 1998;150:835–861. doi: 10.1093/genetics/150.2.835. [DOI] [PMC free article] [PubMed] [Google Scholar]
62. Rajabi-Maham H, Orth A, Bonhomme F. Phylogeography and postglacial expansion of Mus musculus domesticus inferred from mitochondrial DNA coalescent from Iran to Europe. Mol Ecol. 2008;17:627–641. doi: 10.1111/j.1365-294X.2007.03601.x. [DOI] [PubMed] [Google Scholar]
63. Rajabi-Maham H, Orth A, Siahsarvie R, Boursot P, Darvish J, Bonhomme F. The south-eastern house mouse Mus musculus castaneus (Rodentia: Muridae) is a polytypic subspecies. Biol J Linn Soc. 2012;107:295–306. [Google Scholar]
64. Rambaut A, Drummond AJ.2009. Tracer v1.5. Available fromhttp://beast.bio.ed.ac.uk/Tracer.
65. Rogers AR. Genetic evidence for a Pleistocene population explosion. Evolution. 1995;49:608–615. doi: 10.1111/j.1558-5646.1995.tb02297.x. [DOI] [PubMed] [Google Scholar]
66. Sage RD, Atchley WR, Capanna E. House mice as models in systematic biology. Syst Biol. 1993;42:523–561. [Google Scholar]
67. Saitou N, Nei M. The neighbor-joining method: a mew method for reconstructing phylogenetic trees. Mol Biol Evol. 1987;4:406–425. doi: 10.1093/oxfordjournals.molbev.a040454. [DOI] [PubMed] [Google Scholar]
68. Sakai T, Kikkawa Y, Miura I, Inoue T, Moriwaki K, Shiroishi T, et al. Origins of mouse inbred strains deduced from whole-genome scanning by polymorphic microsatellite loci. Mammal Genome. 2005;16:11–19. doi: 10.1007/s00344-004-3013-9. [DOI] [PubMed] [Google Scholar]
69. Schenekar T, Weiss S. High rate of calculation errors in mismatch distribution analysis results in numerous false inferences of biological importance. Heredity. 2011;107:511–512. doi: 10.1038/hdy.2011.48. [DOI] [PMC free article] [PubMed] [Google Scholar]
70. Searle JB, Jamieson PM, Gündüz İ, Stevens MI, Jones EP, Gemmill CEC, et al. The diverse origins of New Zealand house mice. Proc R Soc B. 2009a;276:209–217. doi: 10.1098/rspb.2008.0959. [DOI] [PMC free article] [PubMed] [Google Scholar]
71. Searle JB, Jones CS, Gündüz İ, Scascitelli M, Jones EP, Herman JS, et al. Of mice and (Viking?) men: phylogeography of British and Irish house mice. Proc R Soc B. 2009b;276:201–207. doi: 10.1098/rspb.2008.0958. [DOI] [PMC free article] [PubMed] [Google Scholar]
72. Shimada T, Aplin KP, Suzuki H. Mus lepidoides (Muridae Rodentia) of Central Burma is a distinct species of potentially great evolutionary and biogeographic significance. Zool Sci. 2010;27:449–459. doi: 10.2108/zsj.27.449. [DOI] [PubMed] [Google Scholar]
73. Spiridonova LN, Chelomina GN, Moriwaki K, Yonekawa H, Bogdanov AS. Genetic and taxonomic diversity of the house mouse Mus musculus from the Asian part of the former Soviet Union. Russ J Genet. 2004;40:1134–1143. [PubMed] [Google Scholar]
74. Spiridonova LN, Kiselev KV, Korobitsyna KV. Discordance in the distribution of markers of different inheritance systems (nDNA mtDNA and chromosomes) in the superspecies complex Mus musculus as a result of extensive hybridization in Primorye. Russ J Genet. 2011;47:100–109. [PubMed] [Google Scholar]
75. Stoneking M, Delfin F. The human genetic history of East Asia: weaving a complex tapestry. Curr Biol. 2010;20:R188–R193. doi: 10.1016/j.cub.2009.11.052. [DOI] [PubMed] [Google Scholar]
76. Suzuki H, Miyashita N, Moriwaki K, Kominami R, Muramatsu M, Kanehisa T, et al. Evolutionary implication of heterogeneity of the nontranscribed spacer region of ribosomal DNA repeating units in various subspecies of Mus musculus. Mol Biol Evol. 1986;3:126–137. doi: 10.1093/oxfordjournals.molbev.a040384. [DOI] [PubMed] [Google Scholar]
77. Suzuki H, Shimada T, Terashima M, Tsuchiya K, Aplin K. Temporal spatial and ecological modes of evolution of Eurasian Mus based on mitochondrial and nuclear gene sequences. Mol Phylogenet Evol. 2004;33:626–646. doi: 10.1016/j.ympev.2004.08.003. [DOI] [PubMed] [Google Scholar]
78. Suzuki H, Aplin KP.2012Phylogeny and biogeography of the genus Mus in EurasiaIn: Macholán M, Baird SJE, Munclinger P, Piálek J (eds).Evolution of the House Mouse (Cambridge series in morphology and molecules) Cambridge University Press: Cambridge, UK; 35–64. [Google Scholar]
79. Swofford DL. PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods) Version 4. Sinauer Associates: Sunderland, MA, USA; 2001. [Google Scholar]
80. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S. MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol Biol Evol. 2011;28:2731–2739. doi: 10.1093/molbev/msr121. [DOI] [PMC free article] [PubMed] [Google Scholar]
81. Terashima M, Furusawa S, Hanzawa N, Tsuchiya K, Suyanto A, Moriwaki K, et al. Phylogeographic origin of Hokkaido house mice (Mus musculus) as indicated by genetic markers with maternal paternal and biparental inheritance. Heredity. 2006;96:128–138. doi: 10.1038/sj.hdy.6800761. [DOI] [PubMed] [Google Scholar]
82. Tsuchiya K, Miyashita N, Wang CH, Wu XL, He XQ, Jin ML, et al. 1994Taxonomic study of the genus Mus in China, Korea and Japan–morphologic identificationIn: Moriwaki K, Shiroishi T, Yonekawa H (eds).Genetics in Wild Mice Japan Scientific Society Press/S Karger: Tokyo, Japan/Basel, Switzerland; 3–12. [Google Scholar]
83. Tsuda K, Tsuchiya K, Aoki H, Iizuka S, Shimamura H, Suzuki S, et al. Risk of accidental invasion and expansion of allochthonous mice in Tokyo metropolitan coastal areas in Japan. Genes Genet Syst. 2007;82:421–428. doi: 10.1266/ggs.82.421. [DOI] [PubMed] [Google Scholar]
84. Yang H, Wang JR, Didion JP, Buus RJ, Bell TA, Welsh CE, et al. Subspecific origin and haplotype diversity in the laboratory mouse. Nat Genet. 2011;43:648–655. doi: 10.1038/ng.847. [DOI] [PMC free article] [PubMed] [Google Scholar]
85. Yasuda SP, Vogel P, Tsuchiya K, Han SH, Lin LK, Suzuki H. Phylogeographic patterning of mtDNA in the widely distribtuted harvest mouse (Micromys minutus) suggests dramatic cycles of range contraction and expansion. Can J Zool. 2005;83:1411–1420. [Google Scholar]
86. Yonekawa H, Gotoh O, Tagashira Y, Matsushima N, Shi L, Cho WS, et al. A hybrid origin of Japanese mice “Mus musculus molossinus”. Cur Topics Microbiol Immunol. 1986;127:62–67. [PubMed] [Google Scholar]
87. Yonekawa H, Moriwaki K, Gotoh O, Miyashita N, Matsushima N, Shi LM, et al. Hybrid origin of Japanese mice ‘Mus musculus molossinus': evidence from restriction analysis of mitochondrial DNA. Mol Biol Evol. 1988;5:63–78. doi: 10.1093/oxfordjournals.molbev.a040476. [DOI] [PubMed] [Google Scholar]
88. Yonekawa H, Moriwaki K, Gotoh O, Hayashi JI, Watanebe J, Miyashita N, et al. Evolutionary relationships among five subspecies Mus musculus based on restriction enzyme cleavage patterns of mitochondrial DNA. Genetics. 1981;98:801–816. doi: 10.1093/genetics/98.4.801. [DOI] [PMC free article] [PubMed] [Google Scholar]
89. Yonekawa H, Takahama S, Gotoh O, Miyashita N, Moriwaki K.1994Genetic diversity and geographic distribution ofMus musculus subspecies based on the polymorphism of mitochondrial DNAIn: Moriwaki K, Shiroishi T, Yonekawa H (eds).Genetics in Wild Mice Japan Scientific Societies Press/S Karger: Tokyo, Japan/Basel, Switzerland; 25–40. [Google Scholar]
90. Yonekawa H, Tsuda K, Tsuchiya K, Yakimenko L, Korobitsyna K, Chelomina GN, et al. 2003Genetic diversity geographic distribution and evolutionary relationships ofMus musculus subspecies based on polymorphisms of mitochondrial DNAIn: Kryukov A, Yakimenko L (eds).Problems of Evolution, vol 5Dalnauka; Vladivostok, Russia 90–108. [Google Scholar]
91. Zheng SH, Zhang ZQ, Liu LP.1997Pleistocene mammals from fissure-fillings of Sunjiashan hill, Shandong, China Vertebrata PalAsiatica 35201–216.(In Chinese with English summary). [Google Scholar]
92. Zheng YF, Sun GP, Qin L, Li C, Wu X, Chen X. Rice fields and modes of rice cultivation between 5000 and 2500 BC in east China. J Archaeol Sci. 2009;36:2609–2616. [Google Scholar]

Supplementary Materials