Diagnosis employing molar and cranial traits, and enumeration of external, cranial, postcranial, dental, reproductive, and arterial characteristics presented by Carleton and Musser (1984). Contents of subfamily generally as presented by them except that Acomys, Lophuromys, and Uranomys, formerly considered murines (e.g., Carleton and Musser, 1984; Ellerman, 1941; Misonne, 1969; Musser and Carleton, 1993), are excluded and treated under subfamily Deomyinae. Murinae is characterized by a cohesive cluster of external, cranial, postcranial, dental, reproductive, and arterial characteristics (Carleton and Musser, 1984), but derived molar conditions form the primary basis for defining the subfamily. Two neomorphic cusps, the anterostyle (t1) and enterostyle (t4), are present on the lingual border of M1and form two chevron-shaped, transverse lamina; both upper and lower molars lack longitudinal enamel crests (mures/ids) between lamina; and cusps on the lower molars are positioned opposite one another (Flynn et al., 1985; Jacobs et al., 1989; Freudenthal and Martin Suarez, 1999). Other derived cranial features include their modified carotid circulatory pattern (sphenofrontal foramen and squamosal-alisphenoid groove absent; stapedial foramen present) and a reduced tegmen tympani that does not contact the posterior squamosal (Bugge 1970; Carleton and Musser, 1984; our unpublished survey). Cladistic integrity of Murinae is also supported by results of mitochondrial and nuclear gene sequences, although generic sampling is still limited and some inquiries have used only Mus and Rattus to represent this huge subfamily (Adkins et al., 2001, 2003; Catzeflis et al., 1995; Conroy and Cook, 1999; DeBry and Sagel, 2001; Dubois et al., 1996, 1999; Fieldhouse et al., 1997; Furano et al., 1994; Graur, 1994; Hänni et al., 1995; Huchon et al., 1999; Jansa et al., 1999; Jansa and Weksler, 2004; Kass et al., 1997; Larizza et al., 2002; Martin et al., 2000; Michaux and Catzeflis, 2000; Michaux et al., 2001; Montgelard et al., 2002; Pascale et al., 1990; Robinson et al., 1997; Usdin et al., 1995; E. Verheyen et al., 1995; Verneau et al., 1997, 1998).
No general tribal arrangement of genera is available except for Australian and New Guinea groupings (Conilurini and Hydromyini for Australian species, Watts and Aslin, 1981; Hydromyini, Watts and Baverstock, 1994a, which includes Conilurini and Hydromyini of Watts and Aslin, 1981; and Anisomyini for some New Guinea genera, Lidicker and Brylski, 1987, Watts and Baverstock, 1994a), and African Arvicanthini (Ducroz et al., 2001). Attempts to reflect intergeneric relationships have been made by some muroid specialists and checklist compilers who separated genera into formal and informal categories. Alston used Hydromyinae, Phloeomyinae, and Murinae (although he included what are now non-murine forms) for genera we place in Murinae. Tullberg (1899) classified genera into Murini and Phloeomyini, and also included Otomyini within Muridae. Thomas (1896) listed Hydromyinae, Rhynchomyinae, and Phloeomyinae as coequal with Murinae in a larger Muridae. Miller and Gidley (1918) also recognized Hydromyinae, Phloeomyinae, and Murinae. Tate (1936:505) allocated murines to three groups: 1) those with "simple Rattus like molar teeth," generally the Murinae of Thomas (1896); 2) genera with "complexly folded molars," the Phloeomyinae; and 3) murines "with specialized multi-rooted molars having basin-like depressions with raised edges, a definite tendency for non-development or loss of the third molars, and a trend in the direction of an aquatic habitus," the Hydromyinae. Ellerman (1941) recognized four clusters: 1) Anisomys Group (Anisomyes), containing only the New Guinea Anisomys, a cluster Ellerman thought might deserve subfamily rank; 2) Mures Group, containing most murine genera; 3) Rhynchomyinae for the Philippine Rhynchomys; and 4) the Hydromyinae, which resembled Tate’s group. Simpson (1945) listed genera in his Muridae under Murinae, Phloeomyinae, Rhynchomyinae, and Hydromyinae (and also included Dendromurinae and Otomyinae). Misonne (1969) sorted all genera into four Divisions with subgroups: Lenothrix-Parapodemus Division containing the Lenothrix and Parapodemus groups; 2) Arvicanthis Division; 3) Rattus Division, which contains the Praomys, Maxomys, Rattus, Uromys, and Mus groups; and 4) Basin shaped molars Division, composed of Hydromyinae, Rhynchomyinae, and a third group composed of Echiothrix, Macruromys, and Crunomys. Chaline et al. (1977) allocated genera to either Murinae or Hydromyinae. The checklist by Pavlinov et al. (1995a) segregated genera into primary groups containing sections: 1) Micromys Group, with Pithecheir, Batomys, and Micromys Sections; 2) Apodemus Group, containing the Apodemus, Tokudaia, and Mus Sections; 3) Arvicanthis Group; 4) Rattus Group, composed of Rattus, Dacnomys, Praomys, and Uromys Sections; 5) Phloeomys Group; 6) Rhynchomys Group, which includes the Chrotomys and Rhynchomys Sections; and 7) Incertae sedis, composed of the Crunomys, Nesokia, Melasmothrix, and Hydromys Sections. We have borrowed Misonne’s (1960) primary category and arranged genera in 29 Divisions, each presumably monophyletic (Table 1). A few are defined by published diagnostic morphological and/or molecular traits and could be treated as formal tribes; contents of others derive from our unpublished research. Placement of some genera is provisional and integrity of those clusters require testing with a greater range of morphological and molecular data than is now available. Each of seven Divisions contains only a single genus because no reliable information is presently available that would support allocation to a larger group; some of these genera may have no close living phylogenetic relatives. Reasons supporting our dispositions are explained in the generic accounts. Future research results may indicate some genera be reallocated, and some of our smaller Divisions be combined with others to form larger monophyletic clades. Various phylogenetic analyses of mtDNA cytochrome b, 12S and 16S rRNA sequences, as well as nuclear IRBP sequences, for example, associate the African Aethomys, Micaelamys, Grammomys, Hybomys, Dasymys, Lemniscomys, Rhabdomys, Desmomys, Pelomys, Mylomys, and Arvicanthis either in a single clade or in smaller clusters within a large African clade (Castiglia et al., 2003; Ducroz et al., 2001; Lecompte, 2003), which would consolidate our Aethomys, Arvicanthis, Dasymys, and Hybomys Divisions. Results from albumin immunology suggested that those murine genera sampled form large monophyletic clades reflecting geographic origins–New Guinean, Australasian, Southeast Asian, and African (Watts and Baverstock, 1995b, 1996). DNA/DNA hybridization experiments (Chevret, 1994), analyses of mtDNA complete cytochrome b sequences (Lecompte, 2003; Lecompte et al., 2002b) and combination of cytochrome b and nuclear IRBP gene sequences (Lecompte, 2003), by contrast, indicate the African murines to be paraphyletic and reflect at least four separate colonization events.
Comparative chromosomal data provided in phylogenetic framework and other contexts for European species (Zima and Kral, 1984a), some Asian groups (Cao and Tran, 1984; Gadi and Sharma, 1983; Raman and Sharma, 1977; Rickart and Musser, 1993), some African species (Robbins and Baker, 1978), and murines in general (Viegas-Péquignot et al., 1983, 1985, 1986). Phylogenetic relationships among murines inferred from biochemical sources include the studies of allozymic variation (Bonhomme et al., 1985; Iskandar and Bonhomme, 1984), microcomplement fixation of albumin (Watts and Baverstock, 1995b, 1996), DNA-DNA hybridization experiments (Catzeflis, 1990; Catzeflis et al., 1987), amplification of ancient murine Lx family of long interspersed repeated DNA (L1, Line-1) and L1 retrotransposons (Furano et al., 1994; Pascale et al., 1990; Usdin et al., 1995; Verneau et al., 1997, 1998), and several mitochondrial and nuclear gene sequence studies (see above citations and those in subfamily accounts). Morphological bases for phylogenetic interpretation provided by relationship between body mass, testes mass, and sperm size in murines (Breed and Taylor, 2000); variation in sperm head morphology (Breed, 1991); comparative hair morphology (Keogh, 1985), soft palate topography (Eisentraut, 1969a), and digestive system anatomy (Perrin and Curtis, 1980). The discrepancy between the relatively rapid divergence of murine taxa as revealed by fossils and the much slower rates indicated by molecular data is reconciled by Jaeger et al. (1986) by postulating accelerated rates of evolution for certain proteins and a higher rate of nucleotide substitution in murines than ordinarily seen in other eutherians.
Taxonomy and geographic distributions of African murines are embedded in a vast literature. Phylogenetic affinities among African taxa have been based on DNA/DNA hybridization experiments (Chevret, 1994), microcomplement fixation of albumin (Watts and Baverstock, 1995a), spermatozoal morphology (Baskevich and Lavrenchenko, 1995; Breed, 1995a, 1995d), and all the chromosomal, molecular, morphological, and morphometric data referenced in the accounts of African species. Watts and Baverstock (1995a:431) viewed the evolutionary picture of African murines in the framework of "a period of rapid radiation from a single ancestor, beginning 8-10 million years ago and still continuing. As a result the systematics of these groups is confusing and will take the concerted efforts of workers in many different disciplines to clarify." Based on mtDNA cytochrome b and 12S and 16S rRNA sequences, Ducroz et al. (2001) identified an African murine clade (containing Aethomys, Grammomys, Hybomys, Dasymys, Lemniscomys, Desmomys, Rhabdomys, Pelomys, Mylomys, and Arvicanthis), which they isolated as Arvicanthini, and which partly accords with the evolutionary reconstruction of Watts and Baverstock. Chevret’s (1994) results from DNA/DNA hybridization experiments, however, indicate African murines to be paraphyletic, as does Lecompte’s (2003) analyses of combined mtDNA cytochrome b and nuclear IRBP sequences. Lecompte also recovered an arvicanthine group, and identified three other clades: Praomys group (Myomyscus, Stenocephalemys, Praomys, Mastomys, Heimyscus, Hylomyscus, Zelotomys, and Colomys), a Mus (Nannomys) cluster, and a clade containing only Malacomys.
Late Miocene is the earliest that murinaes are recorded from African strata (Geraads, 2001; Jaeger, 1977b; Mein et al., 1993; Winkler, 2001). Whether African murines were derived from a single or several ancestral groups or originated by one or more immigrations cannot be resolved by the meager number of Miocene taxa recovered from African sediments. Southern Asia is vaguely indicated as the source area, with arrivals in Africa occurring in middle to late Miocene-early Pliocene (Jacobs, 1985; Winkler, 1994, 2002). Late Miocene is the earliest documentation for Progonomys and Paraethomys in North Africa (Jaeger, 1977b; Mein et al., 1993) and Progonomys and Saidomys in Kenya (Winkler, 2001; Karnimata was used, but that is a synonym of Progonomys, according to Mein et al., 1993). Murines from earlier Miocene beds have been found only in N Pakistan, and among them is the middle Miocene Antemus, currently regarded as the earliest murine (Freudenthal and Martin Suarez, 1999; Jacobs and Downs, 1994). Fossils of living endemic African murines first appear in the Pliocene (Denys, 1999; Winkler, 2002; see generic accounts). Mastomys and Arvicanthis, now mostly occurring in Africa, are each represented by an extinct species that lived in Israel during the Pleistocene (Tchernov, 1968, 1996), and Pelomys by a Pliocene species on the Mediterranean island of Rhodes (de Bruijn et al., 1996). See Denys (1999) for a review of fossil African murine faunas and past habitats in relationship to the modern fauna and distributions.
All native New Guinea and Australian rodents are members of Murinae. Classical compendia based upon morphological and distributional studies for murines of New Guinea and adjacent archipelagos remain those of Tate (1951) and Laurie and Hill (1954), which have been largely supplanted by current faunal treatises (Flannery, 1995a, b) and taxonomic reports cited in the various generic and species accounts. Reports relevant to understanding phylogenetic relationships among New Guinea endemics documented comparative phallic morphology (Lidicker, 1968), chromosomal information (Donnellan, 1987), variation in immunological distances assessed by microcomplement fixation of albumin (Watts and Baverstock, 1994a), and sperm morphology (Breed and Aplin, 1994). Evolution of the spermatozoon in New Guinea and Australian murines combined is treated by Breed (1997) who compared the range in sperm head morphology with the phylogenetic frameworks constructed by Watts et al. (1992) for Australian murines and Watts and Baverstock (1994a) for New Guinea species based on analyses of albumin immunology. Swann et al. (2002) explored the cDNA nucleotide sequence encoding the ZPC protein of selected species of Pseudomys, Notomys, and Hydromys, and its significance. Albumin immunological evidence derived from New Guinea marsupials was used to determine time and mode of origin of the terrestrial fauna and is pertinent to inquiries into historical origin of the endemic murines (Aplin et al., 1993). Aplin (2004) has reviewed Australian-New Guinea murine diversity, phylogeny and biogeography.
Earliest Australian record of murines consists of more than a dozen species recovered from Pliocene strata, among them representatives of Leggadina, Pseudomys, and Zyzomys (Aplin, 2004; Godthelp, 1990; Rich et al., 1991). Such a high diversity of fossil taxa suggests arrival of murines in Australia either before the Pliocene, possibly 7 mya (Godthelp, 1990) or earlier, only about 5 mya (early Pliocene; Aplin, 2004; K. Aplin, in litt., 2004). That murines are immigrants to Australia is indicated by the lack of fossils before early Pliocene times (Aplin, 2004; Pledge, 1992:140; Rich, 1991; Rich et al., 1991). In the rich assemblage of local faunas from the Riversleigh region in NW Queensland, for example, in which fossils have been retrieved from late Oligocene to Pleistocene strata, earliest murines were recovered only from late Pliocene sediments, about 3 million years old (Aplin, 2004; Archer et al., 1991; Rich et al., 1991). Possible earliest record of murines on New Guinea still consists only of a sliver of incisor enamel from the Otibanda Formation, 3.0-3.3 myo (late Pliocene; Aplin, 2004; Flannery, 1995a; Pledge, 1992; Rich et al, 1991).
Phylogenetic relationships among Australian murines and between them and non-Australian species have been inferred from chromosomal data (Baverstock et al., 1977c-e, 1983a, b), microcomplement fixation of albumin to measure immunological distances (Baverstock et al., 1977a, b, 1980; Watts et al., 1992, Watts and Baverstock, 1994a, b; 1995b), and spermatozoal and male reproductive tract morphology (Breed, 1980, 1983, 1984, 1985, 1986, 1990, 1997, 2000; Breed and Sarafis, 1978, 1983). In addition to the large taxonomic literature (see species accounts), important regional publications summarize their history of discovery, distribution, conservation, and status in South Australia (Robinson et al., 2000, which includes summary of deposits from which subfossils were collected), Victoria (Menkhorst, 1995a; Seebeck, 1995a; Seebeck and Menkhorst, 2000), New South Wales (Dickman, 1994; Dickman et al., 1993, 2000a), Queensland (Dickman et al, 2000b), arid Northern Territory (Cole and Woinarski, 2000), monsoonal tropics of Northern Territory (Woinarski, 2000), Western Australia (How et al., 2001; Morris, 2000), Tasmania (Hocking and Driessen, 2000), Cape Range peninsula of NW Australia (Baynes and Jones, 1993), and Australia in general (Strahan, 1995). Other topics covered in general summaries focusing on native murines or including them in broader mammalian surveys are patterns and cases of extinction and decline in murines (Smith and Quin, 1996), extinctions on Australian islands with discussion of causes and conservation implications (Burbidge and Manly, 2002), factors influencing species richness on Australian islands (Burbidge et al., 1997), and how range changes in tropical NE Australia may be attributed to late Quaternary climatic changes (Winter, 1997).
Molecular data indicate that phylogenetic affinities of murines are with gerbils and deomyines. Analyses of the nuclear protein-coding LCAT sequences (Michaux and Catzeflis, 2000), and LCAT in combination with sequences of the von Willebrand Factor (Michaux et al., 2001) clusters gerbils with deomyines, which form a sister-group to murines in a monophyletic clade. This topography is also consistent with results of analyses of complete mtDNA cytochrome b (Martin et al., 2000) and nuclear IRBP sequences (Jansa and Weksler, 2004); the Murinae-Gerbillinae link was also revealed by analyses of DNA sequences from the nuclear genes GHR and BRCA1 (Adkins et al., 2003). From their molecular-clock estimate, Michaux et al. (2001) suggested the three subfamilies diverged from the muroid stock 20.8-17.9 million years ago (early Miocene), which is earlier than documented by the first definite murine fossils.
"What is the ancestor of the true Murinae?" (de Bruijn et al., 1996:255) is still unanswered and the subject of active inquiry. The earliest molars currently generally accepted as murine belong to Antemus chinjiensis from 13.75 million-year-old Siwalik strata in N Pakistan (Freudenthal and Martin Suárez, 1999; Jacobs and Downs, 1994). Originally described as the earliest and most primitive (dentally) murine or murid (Flynn et al., 1985; Jacobs, 1977, 1978; Jacobs et al., 1989, 1990), A. chinjiensis was regarded as a dendromurine by Brandy (1981). Earlier records of isolated molars identified as A. primitivus from Pakistan and A. thailandicus from Thailand have since been removed to Potwarmus, a dendromurine (Lindsay, 1988) or myocricetodontine (Tong and Jaeger, 1993). Jacobs and Downs (1994) recorded a morphological transition in Siwalik stratigraphy from "morphologically certain Potwarmus" (14.3 million years ago) to "certain Antemus" (A. chinjiensis,13.75 million years ago) and from that genus to Progonomys or Progonomys like forms between 12.5 and 11.8 million years ago; Siwalik Mus auctor, considered the earliest Mus by Jacobs (1978), was derived from Progonomys at 5.7 million years ago. "The series Potwarmus-Antemus-Progonomys-Mus appears to be one long sequence of anagenesis" (Jacobs and Downs, 1994:155; which is an oversimplification considering that several different late Miocene lineages, including one very similar to Mus, have been identified in Europe; Mein et al., 1993). However, Freudenthal and Martin Suárez (1999:403) cautioned that "This does not necessarily mean that Antemus is derived from Potwarmus, and that Muridae are derived from the Dendromurinae, but it may mean that we have come as close as possible to the origin of the Muridae, and that taxons [sic] more primitive than Antemus will be arranged in other families than the Muridae." For de Bruijn et al. (1996:253), morphology of M1 in Antemus, Potwarmus and similar forms (usually considered myocricetodontines) represent an evolutionary grade, and ". . . assignment to subfamily . . . remains more or less a matter of taste at this stage." The cladistic position of Potwarmus, or forms with similar dental morphology, may be significant to understanding origin of the Murinae. If that genus is a myocricetodontine and represents a morphological transitory stage between that large Asian and African group and murines, then the strong phylogenetic link between gerbillines and murines indicated by molecular data is reflected in the fossil record because the myocricetodontines are the postulated ancestral group from which gerbils evolved (see Gerbillinae introduction).
While the phylogenetic affinity of Antemus is equivocal to some, extinct Progonomys or Progonomys like taxa are universally accepted as the earliest forms exhibiting "an essentially modern grade of murine dental morphology" (Jacobs and Downs, 1994:151), and the first murines to have migrated out of Asia (presumably in the Late Miocene) "to arrive in Europe and Anatolia were at the Progonomys stage-of-evolution" (de Bruijn et al., 1996:255); Progonomys like morphology has also been identified in late Miocene deposits of North Africa (Jaeger, 1977b; Mein et al., 1993) and East Africa (Winkler, 2001, as Karnimata). Mein et al. (1993:62), however, pointed out that most of the early murine samples identified as Progonomys in the literature are not that genus but represent a cluster of separate lineages; by the early Vallesian of Europe (late Miocene, 11-10 million years ago), for example, several murine lineages were already present representing Apodemus, Progonomys, a Mus like form, a Parapodemus like morph, and an unidentified murine, and none "of these lineages has an ancestor-descendant relationship with any of the other ones, but they have a common origin." Fossil links between the diverse living Indo-Australian endemic murines and the late Miocene Siwalik forms have yet to be discovered; the oldest in that region are Pliocene and either too fragmentary to be identified, represent extant genera, or are closely related to them (Chaimanee, 1998; Godthelp, 1990, 1997; Pledge, 1992).
Murine rodents are indigenous to Africa (but not Madagascar), Europe (excluding Ireland), Middle East, Arabian Peninsula, N Eurasia, Indomalayan region, Ceylon, Taiwan, Hainan, Japan, Ryukyu Isls, Philippine Isls, and archipelagos stretching from the Sunda Shelf through Moluccas to the New Guinea area, Australia, and Tasmania. Occurrence of some species outside their natural ranges likely result from intentional or inadvertent anthropogenic processes: an Apodemus on Iceland; two Bandicota in the Indomalayan region; several Rattus and Mus in the Indoaustralian region and Pacific Isls; and Mus musculus, Rattus rattus, and R. norvegicus nearly worldwide outside of Asia.