Table of Contents

Origin of arvicolids:

Plain-Language &
Multilingual  Abstracts



Cheek Tooth Adaptations in Cricetids, Microtoid Cricetids and Arvicolids

Miocene and Pliocene Cricetids, Microtoid Cricetids and Arvicolids





Print article




Tooth group A: Brachyodont and mesodont molars with bunodont dental pattern

Megacricetodon Fahlbusch, 1964
Democricetodon Fahlbusch, 1964
Kowalskia Fahlbusch, 1969

Collimys Daxner-Höck, 1972
Copemys Wood, 1936
Figure 1, Figure 7

Among cricetids Megacricetodon and Democricetodon appeared together with Eumyarion, Anomalomys and Neocometes during the early Miocene of Europe (Late Orleanian, MN 4) as immigrants (e.g., Kälin 1999; Doukas 2003; Mein 2003), but only Megacricetodon and Democricetodon could belong to the stock that gave rise to microtoid cricetids, arvicolids and extant cricetids owing to their specific molar patterns. Democricetodon entered the fossil record slightly before Megacricetodon (Kälin 1999; Sesé 2006). Megacricetodon has relatively small molars characterised by a double-cusped antercone and anteroconid. Major evolutionary trends of Megacricetodon include (Kälin 1999): increase in size, differentiation of the anteroconid along with the reduction of the mesoloph(id)s as well as the reduction and thickening of the posterolophid. In addition, the mesosinus became directed more anteriorly in stratigraphically younger species.

In contrast to Megacricetodon, the molars of Democricetodon are relatively shorter and broader. The anteroconid is short and undivided. The paracone is double-cusped in primitive species or three-cusped in advanced forms. There are well-developed elongated mesolophs and mesolophids as well as external ridges ("Zwischensporne" in the sense of Fahlbusch 1964). Major evolutionary trends are restricted to size increase and the development of additional ridges.

Democricetodon and Megacricetodon are most common in the Middle Miocene of Europe. Important lineages have been reported for Megacricetodon in Central Europe (Kälin 1999; Abdul Aziz et al. 2007) M. aff. collongensis - M. cf. bavaricus - M. bavaricus - M. aff. bavaricus - M. lappi) and Western Europe (Kälin 1999: M. primitivusM. ibericus). Megacricetodon ranges in Europe from late Orleanian (late MN 4) to Early Vallesian (late MN 9), Democricetodon from late Orleanian (early MN 4) to Late Vallesian (MN 10) (Heissig 1990; Kälin 1999; Sesé 2006, Prieto et al. 2010). A similar age range is known from Asia. Among the last occurrences of Democriceton in Europe are the sites of Masia del Barbo (Spain) and Suchomasty (Czech Republic) from early MN 10 (Mein 2003).

Democricetodon, for which the center of development was likely central to eastern Europe (Freudenthal 2006), gave rise to Kowalskia, which first appeared as a morphotype in advanced populations of Democricetodon freisingensis and Democricetodon sp. during the late Early Vallesian (MN 9). However, derivation from another Democricetodon species seems also possible (Kälin 1999).

The double-cusped paracone and paraconid, the pronounced elongated mesoloph and mesolophid, and the relatively long and broad tooth crowns of M1 and m1 are typical traits that distinguish Kowalskia from other contemporary cricetid genera. In Europe, there is a gradual evolutionary sequence of taxa (lineage) within Kowalskia ranging from the late Early Vallesian (MN 9) up to the Early Villanyian (MN 16) (Daxner-Höck 1995). Kowalskia was one of the most common cricetids in the European Late Miocene (Daxner-Höck et al. 1996; Kälin 1999). Many species were described from the Late Miocene and Early Pliocene of western, southern and central Europe, among them (Daxner-Höck 1995; Kretzoi and Fejfar 2005): Kowalskia fahlbuschi (e.g., Rudabanya, Hungary, MN 9; Kohfidisch, Austria, MN 10), Kowalskia occidentalis (Crevillente 2, Spain, MN 11), Kowalskia schaubi (Csakvar, Hungary, MN 11), Kowalskia skofleki, Tardosbanya, Hungary, MN 12), Kowalskia nestori (Baccinello V-3, Italy, MN 13/14), Kowalskia browni (Maramena, Greeke, MN 13/14), Kowalskia polonica and Kowalskia magna (Podlesice, Poland, MN 14), and Kowalskia intermedia (Ivanovce, Slowakia, MN 15). In eastern Europe, Kowalskia progressa (MN 9?, Topachevskij and Skorik 1992) and Kowalskia moldavica (MN 10?, Lungu 1981) were reported. One tooth of Kowalskia polonica was found in the Polish locality Rebielice Krolewskie (MN 16b) (Pradel 1988), but according to Kowalski (2001) this record needs confirmation. According to Daxner-Höck (1995), at least two lineages are recognised in Central Europe: Kowalskia fahlbuschi that developed to Kowalskia polgardiensis, and Kowalskia schaubi evolving through Kowalskia skoflekiKowalskia magnaKowalskia polonica to Kowalskia intermedia.

Collimys has been reported from the Middle and Late Miocene of Central Europe. Dental traits are (Kälin 1999): the loss of the anterior protolophule in M2 and a flat chewing surface along with distinctly increased and thickened mesoloph(id)s. The stratigraphically oldest species is Collimys transversus from the Middle Astaracian (MN 7) at Steinheim am Aalbuch, Germany (Heissig 1995). Collimys transversus possessed brachyodont molars with a primitive crown pattern characterised, for instance, by the absence of the ectomesolophid in lower molars (Heissig 1995). The crowns of Collimys longidens from the Early Vallesian (MN 9) of Nebelbergweg, Switzerland, are slightly higher (mesodont) than those of Collimys transversus (Kälin and Engesser 2001). Much more derived is Collimys primus from the Early Turolian (MN 11) site Eichkogel, Austria, (Daxner-Höck 1972) that displays distinctly hypsodont cheek teeth. Among Cricetini, the Middle Miocene Collimys developed first tendencies toward hypsodonty followed by Cricetulodon and Rotundomys that occurred in the Late Miocene (Kälin 1999). Collimys is likely to have evolved from Democricetodon.

Democricetodon, Megacricetodon, and Kowalskia have a more cosmopolitan fossil record than does Collimys for which localities are confined to Europe. Democricetodon and Megacricetodon were widely distributed across Asia during the Middle Miocene. Important records of Democricetodon and Megacricetodon are known, for instance, from many early to late Middle Miocene sites of China (e.g., Fahlbusch 1969; Qui et al. 1981; Qui 1990; Lindsay 1994; Qiu and Qiu 1995) and Pakistan (e.g., de Bruijn and Hussein 1984; Jacobs and Lindsay 1984; Lindsay 1987, 1988, 1994), those of Kowalskia from Late Miocene and Early Pliocene localities in China (e.g., Lindsay 1994; Qiu and Qiu 1995; Qiu and Storch 2000).

Both Democricetodon and Megacricetodon appeared in China for the first time during the late Early Miocene (late Shanwangian, MNU 4), equivalent with the late Orleanian (MN 4, and perhaps parts of MN 3) in Europe (Qiu and Li 2003). Important sites of this temporal interval are, for instance, Sihong (Democricetodon sp., Megacricetodon sp.) and Gashunyinadege (Democricetodon cf. lindsayi, Megacricetodon cf. sinensis) as well as from Wuerte in northern China (Megacricetodon sp.) (Qiu et al. 1999; Qiu and Wang 1999; Qiu and Li 2003). Megacricetodon yei is known from the first sand bed of the early Middle Miocene Halamagai Formation exposed at Site XJ 98018 of the Tieersihabahe locality in the northern Junggar Basin, China (Bi et al. 2008). According to Qiu and Qiu (1995), the evolutionary level of Democricetodon sp., Megacricetodon sp. from Sihong corresponds approximately with that of Megacricetodon collongensis and Democricetodon brevis of the European mammal zone MN 4. As revealed by the Tunggur local fauna (Qiu et al. 1999; Qiu and Li 2003), Megacricetodon sinensis, Megacricetodon pusillus, Democricetodon lindsayi and Democricetodon tongi were distributed in the northern parts of China during the late Middle Miocene (late Tunggurian, NMU 7), roughly correlatative with the European late Astaracian (MN 7, MN 8). Democricetodon sp. was reported from the early Late Miocene locality Amuwusu, Nei Mongol (early Baodean, NMU 8), the age of which correlates probably to the early Vallesian (MN 9) of the European mammalian Neogene chronology (Qiu and Wang 1999; Qiu and Li 2003; Qiu et al. 2003). According to data given by Lindsay (1994), Qiu and Qiu (1995), Qiu et al. (1999), Qiu and Li (2003), and Qiu et al. (2003), Megacricetodon apparently disappeared in China during the late Middle Miocene (late Tunggurian, late Astaracian, MN 8), followed by Democricetodon in the early Late Miocene (early Baodean, early Vallesian, MN 9) (Mein 2003).

The early dispersal history of Democricetodon and Megacricetodon in Europe and in the northern and eastern parts of Asia appears to be roughly concordant with the fossil record of Asia Minor (Anatolia). Democricetodon seems to appear there for the first time during the early Agenian (MN 1), followed by Megacricetodon during the Early Orleanian (MN 3) (Ünay et al. 2003). However, these early occurrences remain to be confirmed by further material and study. Both genera apparently survived in this region until the Late Astaracian (MN 8) (Sümengen et al. 1990; de Bruijn et al. 2003; Ünay et al. 2003). Based on the cricetid record, the fauna from Pasalar, Anatolia, Turkey (Peláez-Campomanes and Daams 2002), containing Megacricetodon andrewsi and Democricetodon brevis can be correlated with the late European mammal zone MN 6 (Early Astaracian). Democricetodon dispersed during the Early Miocene through Saudi Arabia into North Africa prior to about 14 Ma (Jebel Zelten, Libya: Savage 1990; Winkler 1994. However, this record is considered "Cricetidae gen. et sp. indet." by Wessels et al. (2003). Democricetines are also known from the Early Miocene of East Africa and Namibia (Mein 2003) and the Middle Miocene of East Africa (Winkler 1994; Mein 2003).

As in Europe, Kowalskia appears in the fossil record of eastern Asia distinctly later than Megacricetodon and Democricetodon. One of the earliest species, Kowalskia hanae, is from the middle Late Miocene Shihuiba site in southern China (middle Baodean, NMU 10), a possible equivalent of the Middle Turolian (MN 12) in Europe (Qiu and Li 2003), as is Kowalskia gansunica from Songshan (Lindsay 1994; Qiu and Li 2003). Possibly slightly older records of Kowalskia are from Shala (Qiu and Wang 1999) that is middle Baodean (early NMU 10) in age, possibly early Turolian (MN 11). Kowalskia neimengensis and Kowalskia similis are known from the famous late Late Miocene (late Baodean, NMU 11) local fauna from Ertemte, correlated with the late Turolian (MN 13). In the Late Miocene to Early Pliocene section locality 93002; Lingtai, Gansu, range Kowalskia cf. similis from the middle Baodean (approximately Middle Turolian, MN 12) to the early Yushean (Early Ruscinian, MN 14), and Kowalskia neimengensis from the middle Baodean (Middle Turolian, MN 12) to later Yushean (approximately Late Ruscinian, MN 15) (Zhang and Zheng 2000; Zheng and Zhang 2001). The Early Pliocene (early Yushean, NMU 12) site of Bilike, coeval with the Early Ruscinian (MN 14), has produced Kowalskia zhengi and Kowalskia cf. similis (Qiu and Storch 2000). Kowalskia neimengensis from Harr Obo (Wu 1991; Qiu and Storch 2000) is similar in age (MN 14). Kowalskia sp. has been reported from the late Pliocene (late Yushean, MNU 13) deposits of the Laowogou section near Daodi in northeastern China (Zhang et al. 2003). This site was correlated with the MN 16 zone (Qiu and Li 2003) indicating that Kowalskia survived in China into the Early Villanyian, as in Europe. Kowalskia sp. has recently been reported from the Udunga site (level 2, 3) located in the Transbaikal area and considered Early Villanyian (MN 16a) in age (Erbajeva et al. 2003).

The earliest appearance of the cricetid Copemys in North America coincides with the beginning of the Hemingfordian or somewhat before (Lindsay, 2008), which roughly correlates to the early Orleanian (Prothero 2006) in Europe, and ranged into the early Pliocene. The lower jaw morphology and the tooth crown pattern of Copemys are very similar to that of of the Eurasian Democricetodon. Therefore, Copemys is considered to be very near to Democricetodon (Fahlbusch 1967), but it is not justified to make Democricetodon a junior synonym of Copemys (Freudenthal 2006).

According to Fejfar et al. (1996), Copemys pagei from the Middle Miocene Barstow Formation (California) is one of the earliest North American taxa and could have been derived from Eurasian cricetids (Democricetinae), as formerly suggested by Fahlbusch (1967). Slaughter and Ubelaker (1984) proposed an origin of Neotropical cricetine rodents and North American peromyscines from Copemys. In contrast, Jacobs and Lindsay (1984) considered Copemys to be an immigrant and suggested a non-copemyine ancestor for the Neotropical cricetine rodents: the Late Hemphillian Bensonomys, derived from a separate Miocene Old World lineage of Megacricetodon that would have dispersed to North America. More recently, Lindsay (2008) reports the range of Copemys as Late Arikareean – Late Blancan. Unlike Eurasian Miocene cricetids, Copemys never developed hypsodont molars.

Tooth group B: Brachyodont molars with lophodont dental pattern and planar occlusal surface

Rotundomys bressanus Mein, 1966
Figure 1, Figure 2.8-11, 20

Rotundomys is known only from the Vallesian (late MN 9 - MN 10) of western Europe (Mein 1966; 1975; Kälin 1999; Sesé 2006). Rotundomys bressanus is the last member of a line: R. sabadelliensisR. hartenbergeriR. montisrotundi; the species R. montisrotundi and R. bressanus are equipped with brachyodont molars, but they clearly display further tendencies toward a lophodont tooth crown pattern and a planar occlusal surface without interlocking cusp (Pineda-Muñoz et al. 2010). The absence of both mesolophs and mesolophids is typical, as is the lack of a posterior transverse loph, formed by the "äußeren Quersporn" (sensu Fahlbusch 1964) and the entoconid. The anteroconid is broad and both protoconid and metaconid are longitudinally offset and medially connected with the cusps of the anteroconid. During evolution, the molars became relatively narrower. In R. bressanus both broad anteroconid and anterocone (anteroloph) include symmetrically a circular enamel islet that rapidly disappears with wear. Both species were recovered from paludal lignites at the sites of Montredon, district of Montoulieres, Hérault (R. montisrotundi, lower level of MN 10), and in Soblay, district of Ain, France (R. bressanus, upper level of MN 10) (Mein 1975).

Rotundomys is a derivative of Cricetulodon (Freudenthal 1967; Kälin 1999). The specific molar crown pattern of Rotundomys bressanus shows remarkable similarities to that of early arvicolids (Microtodon, Promimomys). This affinity could favor the idea that Rotundomys bressanus or a still unknown closely related species restricted to a relatively small region in Western Europe during the Vallesian could have given rise to this rodent group. The solution to this question requires further material and study.

Tooth group C: Mesodont molars with tendencies towards a prismatic dental pattern

Microtocricetus Fahlbusch and Mayr, 1975
Figure 1, Figure 3.15-23

Only one species has been reported for this genus: Microtocricetus molassicus (Fahlbusch and Mayr 1975; Bachmayer and Wilson 1984; Wellcomme et al. 1991; Rögel et al. 1993; Kowalski 1993; Fejfar 1999, Hír and Kókay 2010). Sarmatomys podolicus (Topachevskij and Skorik 1988, 1992) is likely a junior synonym of Microtocricetus molassicus. The members of this taxa differ from other groups of microtoid cricetid rodents in having rooted mesodont molars (higher than in Rotundomys bressanus) with flat chewing surfaces and an aberrant occlusal pattern. Behind the anterior cusp of the m1, there are three buccal and four lingual anticlines as well as three buccal and four lingual synclines; the same number of synclines and anticlines is seen in the m2 (Fahlbusch and Mayr 1975; Fejfar 1999; Kretzoi and Fejfar 2005). The upper molars display three buccal and four lingual anticlines as well as three buccal and two lingual synclines (Fahlbusch and Mayr 1975; Fejfar 1999; Kretzoi and Fejfar 2005). The synclines lack cementum. Both narrow triangles (anticlines) and reentrants (synclines) are more or less transverse with an irregular alternation. Roots in molars of adult individuals are well developed. The thickness of the enamel walls increases distinctly with wear. The presence of the posterior transverse ridge ("äußerer Quersporn"; Äqs, fig. 1:17), which is an important feature of advanced species of Democricetodon, suggests a possible ancestry from these Miocene cricetids.

Microtocricetus molassicus is the oldest recorded cricetid rodent with mesodont prismatic molars. It apparently represents an isolated lineage of microtoid cricetids of the Vallesian (MN 9, MN 10) (Fahlbusch and Mayr 1975; Fejfar 1999). Its unique molar structure excludes affinities with other microtoid cricetids that developed similar prismatic molars, such as Pannonicola (= Ischymomys) pontica (MN 11) and Microtoscoptes praetermissus (MN13, MN 14). Most likely, Microtocricetus inhabited moist riparian or paludal environments where it fed on reed leaves, as did Microtoscoptes and Goniodontomys. Microtocricetus molassicus appears to have been distributed throughout southeastern, central and western Europe. It is as old as or older than the records of Goniodontomys and Paramicrotoscoptes from the Early Hemphillian in North America.

Microtocricetus can be considered an "early experiment" of microtoid cricetids to develop high-crowned prismatic molars. However, success was not forthcoming and Microtocricetus became extinct without descendants.

Tooth group D: Mesodont and hypsodont molars with prismatic dental pattern and opposing and slightly alternating triangles

Microtoscoptes Schaub, 1934
Figure 1, Figure 3.1-9, 19, 20

Microtoscoptes is a medium-sized advanced cricetid rodent with rooted hypsodont molars. The occlusal surface is flat, and the reentrants lack cementum. The buccal and lingual reentrants and triangles of the prisms have sharp apices and are opposing, they typically form rhomboid dentine fields, except for M2 and M3 (cf. Figure 4). The thickness of the enamel walls is not differentiated in the prismatic folds of the occlusal surface. The apices of the opposing reentrants touch each other, and the dentine fields of adjacent lophs are not confluent even with well worn molars. Although the opposition of the reentrants and triangles of the teeth is close to being exact in all lower teeth and in the first upper tooth, it is not true of the second and third upper teeth. This feature distinguishes Microtoscoptes from the very similar Goniodontomys, in which the reentrants and triangles of all teeth are directly opposite.

The m1 has 3 buccal and 4 lingual salient angles and 3 buccal and 4 lingual reentrants. The anteroconid complex ("Vorderlobus"), as in arvicolids, has one lingual and one labial reentrant; posterior to these are the most anterior two salient angles; behind the anteroconid complex there are two lingual and one labial triangles exactly analogous to the three basic alternating triangles of the arvicolid m1, the most posterior labial triangle never has an opposing triangle but is confluent with the "posterior loop" of Hinton (1926, p. 106). Triangles T2 and T3 make a rhombus, a typical trait of the Microtoscoptinae not seen in true arvicolids and other branches of arvicolid-like cricetids rodents. The m3 of Microtoscoptes is bilophed with one pair of opposing reentrants and a posterior ridge that is the remnant of the third (posterior) loph in contrast to the two other genera of Microtoscoptinae and to the older species of Paramicrotoscoptes from North America. M2 and M3 have 3 buccal and 2 lingual dentine fields ("triangles") with 2 buccal and 1 lingual reentrants. The M3 and m3 in Microtoscoptes are distinctly reduced.

The Eurasians species Microtoscoptes praetermissus was first described from the sites of Ertemte and Olan Chorea, Inner Mongolia (Nei Mongol), China (Schaub 1934, 1940; Jacobs et al. 1985; Fahlbusch 1987). The Ertemte local fauna is considered Late Miocene (late Baodean, NMU 11) in age correlated with the European mammalian zone MN 13 (Qiu and Qiu 1995; Flynn et al. 1997; Qiu and Wang 1999; Qiu et al. 1999; Qiu et al. 2003; Qiu and Li 2003). Additional remains come from Harr Obo, also Inner Mongolia (Nei Mongol, China) (Fahlbusch 1987), the age of which was previously also equated with that of Ertemte (Late Turolian, MN 13). However, the occurrence of Rhagapodemus and Hypolagus at the site of Harr Obo suggests that this locality is slightly younger than Ertemte, early Yushean (Early Ruscinian, MN 14) (Qiu and Qiu 1995; Qiu and Li 2003; Qiu et al. 2003). Microtoscoptes praetermissus is also known from fluvio-limnic deposits (Sasin Formation, Odonin Member) exposed at Olchon Island, Lake Baikal, Irkutsk (Russia) (Mats et al. 1982). These Microtoscoptes praetermissus-bearing strata are assigned to the late Turolian (MN 13).

There are other imperfectly known Eurasian species of Microtoscoptes, not yet fully described or sufficiently figured in literature, such as as "Microtoscoptes tjuvanensis" (Zazhigin in Gromov and Polyakov 1977) and "Microtoscoptes sibiricus" (Zazhigin and Zykin 1984). The latter is larger and apparently more advanced than the previous species. Both species seem to be Late Turolian (MN 13) in age, like Microtoscoptes praetermissus. At the present stage of knowlege it cannot be excluded that "Microtoscoptes tjuvanensis" and "Microtoscoptes sibiricus" are possibly junior synonyms of Microtoscoptes praetermissus. Note, however, that the known measurements of "Microtoscoptes tjuvanensis" (Zazhigin in Gromov and Polyakov 1977, p. 101) might indicate another Eurasian Microtoscoptes species, that is distinctly larger and slightly younger than Microtoscoptes praetermissus (Fahlbusch 1987). The distribution of Microtoscoptes is confined to Eurasia.

Paramicrotoscoptes Martin, 1975
Figure 1, Figure 4.11-14, 17, 18

The North American Paramicrotoscoptes hibbardi is a medium-sized cricetid species with rooted, hypsodont molars and flat occlusal surface. There is no cementum in the reentrants. Thickness of the enamel walls is not differentiated. Buccal and lingual reentrants have rounded apices (more so in worn molars) that are not exactly opposite each other, showing slight alternation so that the rhomboid dentine areas are less symmetrical, and the apices of the opposing reentrants do not touch each other in some worn teeth as they do in Goniodontomys. Therefore, the dentine fields may be slightly confluent with those anterior or posterior to them. An islet may be present in the anteroconid of the m1.

Paramicrotoscoptes hibbardi is distinguished from Eurasian species of Microtoscoptes and from the North American Goniodontomys disjunctus (see below) by having the posterior loop of M3 reduced (from the condition in Goniodontomys) to an "r", but it is less reduced than in Microtoscoptes praetermissus and other Eurasian species in which it is reduced to a simple oblique oval. In addition, the posterior loop of m3 is reduced to a narrow crest in Microtoscoptes praetermissus.

The pattern of the enamel on the occlusal surface of the M3 of Paramicrotoscoptes hibbardi closely matches that of the very late Miocene prometheomyine North American genus Protopliophenacomys (= Propliophenacomys, see Martin 2003a), which is about the age of Microtoscoptes praetermissus; however the triangles in Propliophenacomys have become completely alternating.

The structure of the mandible of Paramicrotoscoptes, lacking an arvicoline groove, is also primitive and shows the anterior edge of the ascending ramus leading straight to the upper masseteric crest and to its anterior union with the lower masseteric crest, in the pattern that is typical of low-crowned cricetid rodents; apparently propalinal mastication developed before the musculature that aided it did. The temporal muscles with well=developed deep temporal fossa appear to be extremely typical of arvicolines. By comparison with representatives of the subfamily Microtodontinae (Microtodon, Prosomys, Promimomys), Paramicrotoscoptes has twice the dental specialization in hypsodonty and enamel complication, while having half of the specializations in masticator musculature (Repenning 1968).

The genus Paramicrotoscoptes was introduced by Martin (1975). Repenning (1987, in litt.) considered Paramicrotoscoptes a junior synonym of Microtoscoptes and listed Paramicrotoscoptes hibbardi as a species of Microtoscoptes. The palaeogeographic and stratigraphic range of Paramicrotoscoptes hibbardi is confined to North America and the Early Hemphillian (Shotwell 1970; Hibbard 1970; Repenning 1987), beginning at about 9 Ma (Woodburne and Swisher 1995). Records of Paramicrotoscoptes hibbardi are known from Idaho, Oregon, Nevada and Nebraska (Repenning 1987; Bell 2000).

Goniodontomys Wilson, 1937
Figure 4.10, 15, 16

Goniodontomys is known only from a single North American species: Goniodontomys disjunctus (Wilson 1937; Schaub 1940; Hibbard 1959, 1970; Repenning 1987). The hypsodont molars are rooted, with flat occlusal surface, but without cementum. The enamel walls of the prismatic triangles are not differentiated. In Goniodontomys, the anteroconid complex has two very strong and directly opposing wings and a strongly doubled anteroconid, the lingual one more prominent and extending farther anteriorly. Buccal and lingual triangles of the prisms have sharp apices and are exactly opposing. Buccal and lingual reentrants are directly opposing with apices solidly appressed and usually flattened against each other along the midline of the tooth. They are never slightly offset as in Microtoscoptes. The rhomboid dentine areas are well expressed, and the fields of different lophs are not confluent (with the same exception at the posterior loop of m1 as in Microtoscoptes).

As discussed above, the conspicuous differences that separate Goniodontomys disjunctus from the species of Microtoscoptes and Paramicrotoscoptes are: (1) in Goniodontomys the structure of the M2 shows three symmetrical rhomboid dentine fields that are closed by the contact between the opposing reentrants, whereas in Microtoscoptes the lingual half of the middle rhomboid loph has been lost; (2) the structure of M3 of Goniodontomys is most complex, somewhat reduced in Paramicrotoscoptes hibbardi, and is further reduced in Eurasian Microtoscoptes praetermissus where the posterior loop is practically lost (see Shotwell 1970, text-fig. 32 J to O; Martin 1975, Figure 3D and F, and Fahlbusch 1987, text-fig 1 and 28-30). (3.) The m3 is not reduced in Goniodontomys disjunctus or Paramicrotoscoptes hibbardi but was conspicuously reduced in the younger Microtoscoptes praetermissus from Asia.

Goniodontomys occurred in the Hemphillian of North America, probably aequivalent to Early Turolian (MN 11) or possibly closer to MN12 (according to the data given by Agusti et al. 2001). Goniodontomys disjunctus is known from localities in Wyoming and Oregon (Repenning 1987; Bell 2000). Both Goniodontomys and Paramicrotoscoptes are likely North American natives (Repenning 1987; Repenning et al. 1990). As with Paramicrotoscoptes, the distribution of Goniodontomys is restricted to North America. Martin (2008) excludes the north American Microtoscoptes and Goniodontomys from the Arvicoilidae (see also Lindsay 2008).

Tooth group E: Mesodont and hypsodont molars with prismatic dental pattern and opposing triangles; M1, m1, and m2 trilophodont

Trilophomys Depéret, 1892
Figure 1, Figure 5.1-5, 28, 29

Trilophomys is a medium-sized advanced cricetid rodent with simply formed hypsodont prismatic molars. The M1, m1, and m2 are trilophodont with two lingual and two buccal reentrants; the M3 is reduced, bilophodont, with one lingual and one buccal reentrant. The occlusal pattern reveals that the reentrants of the lower cheek teeth are basically opposing, those of the upper molars basically alternating, except for the M2, which is equipped with a pair of opposing triangles that are prominently confluent. All reentrants lack cementum. The relatively thick enamel walls are distinctly differentiated. In lower molars the posterior walls of the triangles or lophs are thicker than the anterior ones, and the reverse true for the upper cheek teeth. The tooth crown base (linea sinuosa) is moderately sinuous. It curves slightly up on the salient angles (triangles) and slopes down below the reentrants. There is a trend in Trilophomys to increase the height of the tooth crowns.

The very massive and short lower jaw of Trilophomys shows a strong lower masseteric crest that terminates far anteriorly beneath the anterior end of the first lower molar; it is, however, positioned very high on the buccal side of the ramus. The massiveness of the jaw results largely from the stout lower incisor, which has a short radius of curvature so that at first glance the jaw appears more sciurid than cricetid. These features, combined with stout and upward-directed incisors, suggest arboreal or fossorial habits of Trilophomys.

As with Baranomys, the temporal and palaeogeographical range of Trilophomys was confined to the Early Ruscinian (MN 14) to Early Villanyian (MN 16) of Europe (Mein 2003) and possibly MN17. Trilophomys pyrenaicus is a large species with relatively broad molars (Schaub 1940; Adrover 1986). A distinct additional lingual reentrant on the anteroconid of the m1 persists during wear. In occlusal view the reentrants and triangles are opposing, not alternating, and the dentine fields were less confluent than in other species of Trilophomys. Trilophomys pyrenaicus was dispersed in Western Europe during the Ruscinian (MN 14, MN 15) (Schaub 1940; Adrover 1986; Kowalski 1990).

Trilophomys schaubi (Fejfar 1964; Adrover 1986) is smaller than Trilophomys pyrenaicus. In addition, it possesses relatively narrow molars. The narrow anteroconid of the m1 has a deep additional lingual reentrant that persists during wear for a long time. The buccal reentrants are shallower than the lingual ones. In occlusal view, the reentrants are slightly alternating, not directly opposing as in Trilophomys pyrenaicus, and the dentine fields are less confluent than those of the adjacent lophs in Trilophomys depereti. Trilophomys schaubi is known from numerous sites in Western and Central Europe. Stratigraphic range: Ruscinian (MN 14, MN 15).

Trilophomys depereti (Fejfar 1961, 1964; Adrover 1986; Popov 2004) also is smaller than Trilophomys pyrenaicus. The molars are relatively narrow, the lower cheek teeth less hypsodont but strongly trilophodont. The m1 (resembling that of Epimeriones) displays a narrow anteroconid, with shallow additional lingual reentrant seen only in juvenile individuals. In occlusal view the reentrants are weakly alternating, particularly in the upper cheek teeth. In contrast to other species of Trilophomys, the dentine fields are distinctly confluent between the lophs. In lower molars, the buccal reentrants are shallower than the lingual ones. Like the two Trilophomys species described above, Trilophomys depereti was distributed in Western and Central Europe during the Ruscinian (MN 14, MN 15).

Trilophomys vandeweerdi (Brandy 1979; Adrover 1986) is the most advanced species of Trilophomys known so far. It differs from other Trilophomys species in having unmistakably greater hypsodonty. Moreover, the m1 possesses especially well-developed dentine tracts, which, however, do not interrupt the enamel walls of the chewing surface before the cheek tooth is about half worn. Trilophomys vandeweerdi ranges in Spain from the Late Ruscinian (MN 15) to the Early Villanyian (MN 16) (Sesé 2006).

The youngest records of Trilophomys are in Rebielice Krolewskie (MN 16b), Poland (Nadachowski 1989) and in Osztramos 3 (MN 17) (Jánossy 1970).

Tooth group F: Mesodont molars with prismatic dental pattern and opposing or alternating triangles

Pannonicola Kretzoi, 1965 (= Ischymomys Zazhigin 1982)
Figure 1, Figure 3.1-14

Pannonicola brevidens is based on two rooted molars, a heavily worn left m2 and a left M3, also deeply worn (Kretzoi 1965). Both molars, possibly representing a single individual, were recovered from a deep core near Jászladány, North Hungary (Kretzoi 1965). The specimens were tentatively assigned to Middle Turolian (MN 12). Although heavily worn, the short and broad molars of Pannonicola brevidens display an arvicolid-like occlusal pattern. The dentine fields are broadly confluent owing to advanced abrasion.

More recently, additional records of Pannonicola have been reported from the Hungarian sites Nyárad and Sümegprága (Kordos 1994), which have not yet been fully described. The available data suggest a Late Vallesian (MN 10) to Early Turolian (MN 11) age for these findings (Kordos 1994). Kordos (1994) argues that, Ischymomys Zazhigin 1982 is a junior synonym of Pannonicola, and the present authors agree.

Outside Central Europe, two additional species of Pannonicola (= Ischymomys) have been reported from Turolian deposits in Eastern Europe and Western Asia. The younger species Pannonicola (= Ischymomys) quadriradicatus was recovered from Middle Turolian (MN 12) Ishim strata exposed on the right bank of the River Ishim, Petropavlovsk, Kazakhstan in Western Asia (Zazhigin 1982, Zazhigin in Gromov and Polyakov 1977); the older one, Pannonicola (= Ischymomys) ponticus, known from Frunzovka 2 near Odessa, Ukraine (Topachevskij et al. 1978) is referred to the EarlyTurolian (MN 11).

Pannonicola (= Ischymomys) ponticus is a medium-sized cricetid with subhypsodont, prismatic, cementless, and rooted molars; similar to the preceding species. Enamel walls of molars are relatively thin and not differentiated; the triangles are slightly alternating with tendency to form a dentine rhomboid across the occlusal surface; the apices of the reentrants touch the opposing apices near the center of the tooth as in the Microtoscoptini. Enamel islets are present in the central anteroconid of m1 and in the posterior lobe of M2 and M3, persisting with deeper wear. One specimen illustrated by Topachevski et al. (1978, text-fig. 1:3) clearly shows that the islet in the anteroconid complex derives from the most anterior lingual reentrant of the tooth. If consistent, this differs with the derivation from the most anterior buccal reentrant in the arvicolines and with derivation from a reentrant at the anterior end of the anteroconid complex, called the "cricetine islet" by Repenning (1968, text-fig. 10) and apparently is characteristic of the prometheomyines and possibly the ondatrines (Hinton 1926, text-figs 58 and 62). The mesial/buccal wall of the anteroconid in M1 is variably undulated but less so than in the younger and more advanced Pannonicola (= Ischymomys) quadriradicatus.

Pannonicola (= Ischymomys) quadriradicatus is a large advanced cricetid with slightly hypsodont, prismatic, cementless, and rooted molars. The M2 and M3 have four roots (hence the specific name). The enamel walls of molars are relatively thin and not differentiated; the reentrants and triangles slightly alternating with tendency to form rhombic, lophate dentine fields across the occlusal surface. The apices of the reentrants touch medially in many individuals, similar to the Microtoscoptini. Prominent enamel islets are present in the anteroconid of m1 in some individuals as well as in the posterior loop of the m3 and M3 (M2?), persisting during considerable wear. The mesial wall of the anteroconid in m1 is variably undulated.

The origin of Pannonicola is unknown. A close affinity with the genus Microtocricetus is excluded because of the non-homologous molar structure. However, all Turolian/Hemphillian microtoid cricetid genera (Goniodontomys, Microtoscoptes, Microtocricetus, and Pannonicola) invariably occur in paludal/fluviatile sediments, which suggest a wet habitat. In many respects (e.g., the tendency to have rhomboid dentine fields showing little or no alternation of the triangles, and contact of opposing reentrant apices near the midline of the tooth) the younger Pannonicola (= Ischymomys) quadriradicatus and the older Pannonicola (= Ischymomys) ponticus both resemble the Microtoscoptini. The species of Pannonicola (= Ischymomys) could represent a lineage of "Old World" microtoscoptines during the Turolian, more advanced in the offset of its dental triangles but less advanced in the development of hypsodonty.

Owing to the advanced molar structure already developed during the Late Vallesian (MN 10) or Early Turolian (MN11) (Kordos 1994), Pannonicola cannot be derived from Microtodon, a younger taxon that displays a more primitive molar pattern, although it appeared distinctly later in the fossil record of Asia for the first time (Late Turolian, MN 13). Most probably, Pannonicola gave rise to Dolomys and Propliomys represented first in the Late Ruscinian record (MN 15) of Europe as well as possibly to Dicrostonyx, which made its first appearance during the Biharian.

Tooth group G: Mesodont molars with prismatic dental pattern and alternating triangles

Anatolomys Schaub, 1934
Figure 1, Figure 2.12-14

Small microtoid cricetid with rooted, cementless, mesodont molars; occlusal outline is longer than wide. Both unworn M1 and m1 display a broad and inflated anteroconid with shallow islet(s) that soon disappear with wear, and with variably short mesoloph/id. The axes of the inner (in upper molars) and the outer (in lower molars) reentrants are transverse, as in Microtodon or Promimomys. The axes of the outer (in upper molars) and inner (in lower molars) reentrants are oblique. The relatively thick enamel walls are not differentiated. There are mainly three roots in upper and two roots in lower molars.

The general pattern of the mandible and molars of Anatolomys suggests similarities with the Ruscinian genus Baranomys described by Kormos (1933). However, the genera probably are unrelated. Anatolomys likely was adapted for a digging lifestyle.

The only known species Anatolomys teilhardi was originally described from the late Late Miocene site of Ertemte, Inner Mongolia (Nei Mongol), China (Schaub 1934; Jacobs et al. 1985; Fahlbusch and Moser 2004), and referred to the late Baodean (NMU 11) (Qiu et al.  2003) that is considered approximately equivalent in age to the European Late Turolian (MN 13) (Qiu and Li 2003). The localities Harr Obo and Bilike, also Inner Mongolia (Nei Mongol), China, that have produced Anatolomys cf. teilhardi (Qiu and Storch 2000; Fahlbusch and Moser 2004) are younger than Ertemte and can be assigned to the early Yushean (NMU 12), regarded equivalent to the Early Ruscinian (MN 14) in Europe (Qiu and Li 2003; Qiu et al. 2003).

Celadensia Mein, Moissenet and Adrover, 1983
Figure 1, Figure 5.17-18, 20, 23

In occlusal view the mesodont, cementless molars have a barrel-like outline, because the cheek teeth become conspicuously wider toward the crown base (Aguilar et al. 1982; Mein et al. 1983). The small anteroconid of the m1 retains a circular islet during wear. Unlike other microtoid cricetids (e.g., Microtoscoptes, Trilophomys, and Baranomys), the M1 retains two buccal reentrants. Only one species has been recognized so far (Mein et al. 1983), Celadensia nicolae. Records were recovered from a few Late Miocene and Early Pliocene sites of the Iberian Peninsula and in France (e.g., Aguilar et al. 1982; Mein et al. 1983). Celadensia nicolae ranges from the Late Turolian (MN 13) to Early Ruscinian (MN 14) (Fejfar 1999; Mein 1999, 2003; Sesé 2006).

Baranomys Kormos, 1933
Figure 1, Figure 5.6, 14-16, 21, 22, 26, 27

The molars of Baranomys are rooted, mesodont, and initially prismatic. The helmet-like anteroconid complex of the m1 is very short, and an enamel islet is present. The molars of Baranomys are, as in Microtodon, very small (length of the m1 about 1.5 mm), and the reentrants (synclines) are distinctly asymmetrical. There is no cementum in the reentrants. The dentine fields are confluent. The number of reentrants in Baranomys corresponds to that in Microtodon and Promimomys (or Prosomys of some authors). There is an important feature in the dentition of Baranomys that is worth mentioning: The anterior edges of the lower molars and the posterior edges of the upper molars are curved backward in Baranomys, a distinct contrast to the condition in arvicolids and arvicolid-like rodents where these edges are always straight.

The chewing surface of the cheek teeth displays some similarities to those of Microtodon (e.g., Kowalski 1960; Kretzoi 1962; Sulimski 1964; Repenning 1968; Fahlbusch and Moser 2004), the taxonomic, and status of Baranomys has been discussed for a long time. Some authors (e.g., Kretzoi 1955a, Fejfar and Storch 1990) accept Baranomys and Microtodon as two distinct genera, whereas others (e.g., Fahlbusch and Moser 2004) regard the separation as unjustified and consider Baranomys and Microtodon as identical taxa.

All fossil remains of Baranomys are from Europe. The three species of Baranomys range from Early Ruscinian (MN 14) to Early Villanyian (MN 16), like Trilophomys. The oldest species, Baranomys kowalskii, was first reported from the Early Ruscinian (MN 14a) locality Podlesice at Kroczyce, south Poland (Kretzoi 1962). Baranomys loczyi (Kowalski 1956; Kretzoi 1962) ranges from the late Ruscinian (MN 15) to the Early Villanyian (MN 16). Baranomys longidens is known from a series of localities in Europe (Kowalski 1960; Sulimski 1964; Repenning 1968; Fejfar and Storch 1990; Fejfar et al. 2006) confined to the late Ruscinian (MN 15).

Bjornkurtenia Kowalski, 1992
Figure 1, Figure 5.7-13

The rooted and cementless molars of Bjornkurtenia are small, brachyodont, and massive. The enamel band is very thick and undifferentiated. The thickness of the enamel walls increases progressively during wear. Dentine tracts are not developed. The short anteroconid complex of the m1 displays a simple design. The reentrants alternate and the dentine fields are typically broadly confluent. On the chewing surface of the m1 there are no islets in the AC, the lingual reentrants of which disappear with wear. A striking feature is the extremely thick enamel band that reaches with Bjornkurtenia its maximum thickness for the Baranomyinae. The molars of Bjornkurtenia recovered from sites in Central Europe are smaller then those from the type locality La Jasse at Terrats in France (Michaux 1976). The single species described so far, Bjornkurtenia canterranensis (Michaux 1976; Kowalski 1992; Popov 2004), appeared in the Early Ruscinian (MN 14) and became extinct during the Late Ruscinan (MN 15). Most likely, the lifestyle of Bjornkurtenia canterranensis was fossorial.

Microtodon Miller, 1927
Figure 1, Figure 6.1-7

The molar crown pattern of Microtodon is rather similar to that of Promimomys. The molars are mesodont and low prismatic-lophodont. They have alternating reentrants and triangles. The apices of the triangles are sharp and V-shaped in moderately worn molars, and U-shaped in heavily worn cheek teeth. The enamel band is of more or less equal width, but distinctly thicker in heavily worn molars than in unworn or moderately worn cheek teeth. The dentine fields are confluent, especially with increasing wear. The relatively deep and long existing enamel islet of the anteroconid complex is a true "cricetine islet", derived from the medial groove of the anteroconid which was originally bilobed (Fahlbusch and Moser 2004).

The mandible shows some arvicoline features. A distinct but nascent "arvicoline groove" (Repenning 1968) is present in some members, but rudimentary (or unknown) in others, there is a prominent and anteriorly placed lower masseteric crest, and a deep internal temporal fossa separates the row of lower molars from the ascending ramus. In all respects the masticatory musculature obviously must have been proto-arvicoline (Repenning 1968).

Microtodon is known from the Late Baodean (MNU 11, Late Turolian, MN 13) Ertemte (Nei Mongol, China) and from early Yushean (MNU 12, Early Ruscinian, MN 14) Harr Obo and Bilke (described as Microtodon cf. atavus), also Inner Mongolia, China (Schlosser 1924; Schaub 1934; Jacobs et al. 1985; Qiu and Storch 2000; Fahlbusch and Moser 2004). Other records of Microtodon that are slightly larger than those from Ertemte have been recovered at the localities of Kirgiz Nur 2 and Pavlodar in western Asia also correlated with the European Late Turolian (MN 13) (Fejfar et al. 1997). Recently, Microtodon was recorded at Komanos 1 in Greece (MN 13-14) (Hordijk and de Bruijn 2009).

An important fauna including Microtodon sp. has been collected more recently from Gaotege, central Nei Mongol, China (Li et al. 2003). The Gaotege mammalian assemblage is thought to be stratigraphically younger than that from Bilike. It was assigned to the Early Pliocene (early Yushean) and tentatively correlated with the Early (MN 14) or early Late Ruscinian (early MN 15) of Europe (Li et al. 2003). If correctly assessed, this Chinese record suggests that Microtodon passed the Miocene/Pliocene (Turolian/Ruscinian) boundary in East Asia.

Baranarviomys Nesin, 1996

Baranarviomys admirabilis was recovered from a sequence of fluvio-limnic deposits exposed at the site Vinogradovka 1, Ukraine (Nesin 1996). It is Late Turolian (MN 13) in age. The lower jaw and the molars of Baranarviomys admirabilis display a primitive arvicolid-like condition. The cheek teeth are mesodont and low prismatic-lophodont. In the m1, the reentrants are alternating, U-shaped in faintly worn molars and sharply V-shaped in heavily worn molars. The broad dentine fields are confluent. In the overall morphology of the molars, Baranarviomys admirabilis is rather similar to Microtodon atavus known from the Late Turolian (MN 13) of Central Asia. Possibly, Baranarviomys admirabilis is a junior synonym Microtodon atavus.

Promimomys Kretzoi, 1955
Figure 1, Figure 6.14-18, 19, 20, 23-28

This genus is a member of the Microtodon-Promimomys-Mimomys-Arvicola lineage and has a far-reaching importance for understanding the evolution of arvicolids. Promimomys entered the fossil record for the first time in Eastern Europe (Mugureny, Vinogradovka: Promimomys sp. 1 and 2) during the late Turolian (MN 13) (Fejfar et al. 1997). In contrast to its dental similarity to Baranomys (Figure 1, Figure 5, and Figure 7), the mandible of Promimomys shows several characteristic arvicoline features. The arvicoline groove is well developed and the lower masseteric crest is prominent, especially anteriorly where it and the higher arvicoline groove meet and run forward to the anterior termination of the insertion of the masseter muscle. The internal temporal fossa (separating the lower molars from the ascending ramus) is deep and broad, and the entire mandiblar ramus is deeper and more robust, like an arvicoline, while that of Baranomys is more slender with a procumbent incisor as in many low-crowned cricetids. The lower jaw has a short diastema and a lower masseteric crest relative to the alveolar margin. The arvicoline groove and lower masseteric crest merge at a low angle rather far behind the anterior termination of the masseteric insertions, a conspicuous character of most arvicolines. The symphyseal insertion for the digastric muscle is moderately strong. In all respects the masticatory musculature was obviously arvicoline (Repenning 1968).

The massive lower incisor passes below the posterior root of m2 and continues on the buccal side of m3. The basal part of the lower incisor ascends far above the level of the occlusal surfaces of the molars and forms a distinct elevation on the lateral side of processus articularis. The anteroconid has a deep, persistent enamel islet. In worn molars the occlusal surface is much broader than in juvenile teeth.

The first lower molar of Promimomys insuliferus, the index fossil of the early Early Ruscinian (MN 14a), is characterised by a short anteroconid complex that is equipped with an oval enamel islet. The M3 also displays an enamel islet. In juvenile indivduals, the anteroconid complex of the m1 has a typical mesial opening and a moderately undulated, arched anterior margin; a lingual indentation is not developed. The enamel walls are rather thick, the enamel band undifferentiated, and the base of the tooth crown, the so-called linea sinuosa, is almost straight. The mesodont, rooted, and cementless molars correspond in size and height of the tooth crown well with those of Promimomys mimus (or Prosomys mimus of other authors). However, the lingual reentrants are narrower and more sharply cornered. In worn molars the occlusal surface is much broader than in unworn juvenile teeth. The base of the enamel crown is nearly straight.

Promimomys insuliferus was originally described from Podlesice, Poland, and other important eastern Europe sites are Novaya Andriashevka, Antipovka, and Chugunovka (Fejfar et al. 1997). The site Novaya Stanica in western Siberia has yielded Promimomys sp. (Zazhigin and Zykin 1984), a form more primitive than Promimomys insuliferus from Europe.

The genus Promimomys is based on an incomplete left lower jaw with a heavily worn m1, which was recovered at the MN 14b site of Csarnóta 2 (Hungary) and later described as Promimomys cor by Kretzoi (1955b). More recently, Promimomys cor was reported from Early Pliocene lignitic mudstone deposits of the Haltipasa Formation in Anatolia, Turkey (Kaya et al. 2004) and at Komanos 1 and Vorio in Greece (Hordijk and de Bruijn 2009). Promimomys cor, the index fossil of the late Early Ruscinian (MN 14b), has mesodont prismatic, cementless, and rootless molars. The enamel band is relatively thin, but not differentiated; it becomes slightly thicker during wear. The reentrants of m1 are distinctly alternating, U-shaped in unworn molars and sharply V-shaped in worn molars. The dentine fields are confluent but usually prominently constricted. There is a high and broad anteroconid in the m1 with oval shaped islet becoming circular and disappearing during wear. The linea sinuosa is almost straight.

Promimomys asiaticus, described from early Yushean (early Pliocene) cave deposits of Dajushan Hill (Hainan, Anhui Province, China) is said to be more primitive than Promimomys cor from western Eurasia (Jin and Zhang 2005).

The stratigraphic range of Promimomys in Eurasia is confined to the Late Turolian (MN 13) and Early Ruscinian (MN 14) (Fejfar et al. 1997; Sesé 2006). Promimomys gave rise to Mimomys. The interval during which Mimomys appeared for the first time is poorly understood and needs further study. According to the fossil arvicolid record in Western and Southern Europe Mimomys vandermeuleni which was reported from the early Late Ruscinian (MN 15a) site Villalba Alta Rio (VAR 2a), Spain, is among the most primitive Mimomys species known so far (Fejfar et al. 1990). An even more primitve species was described from the late Early Ruscinian (MN 14b) in Western Asia as Mimomys antiquus (Zazhigin and Zykin, 1984).

Prosomys Shotwell, 1956
Figure 1, Figure 6.8-13

The North American records of Promimomys were described as Prosomys by Shotwell (1956). According to Repenning (1968, 1987) Prosomys Shotwell, 1956 is a junior synonym of Promimomys Kretzoi, 1955. Records of Promimomys (or Prosomys) approximately equivalent in age to the Late Turolian (MN 13) and Early Ruscinian (MN 14) in Eurasia are reported from the Late Hemphillian in North America (Repenning 1987, 2003; Bell 2000; Martin 2003a). According to Tedford et al. (1987, 2004), the immigration of Promimomys as well of Agriotherium, Felis, Megantereon, Ochotona, Plesiogulo and Odocoileini cervids defines the beginning of the Late Hemphillian. Records of Promimomys (or Prosomys) in North America are known from two sites in Oregon (McKay, Christmas Valley) and from one locality in Nebraska (Mailbox) (Repenning 1987; Bell 2000; Martin 2003a). Promimomys mimus is the index fossil for the first central Great Plains rodent zone (Rz 1, Martin 2003a).

Tooth group H: Hypsodont molars with prismatic dental pattern and alternating triangles

Mimomys Forsyth Major, 1902
Figure 1

Among known early arvicolid taxa, the fossil representatives of the genus Mimomys have been the main focus of research for a long time. Many extant arvicolid taxa, such as Microtus and Arvicola, are derived from Mimomys. Both genera are important side branches of the main stream of Mimomys evolution that can be traced over approximately three million years since the end of the Turolian (MN 13). All species of Mimomys have only three basic alternating triangles on the m1. Diagnostic features are the structure of the anteroconid in m1, the height of the dentine tracts (linea sinuosa), presence or absence of cementum within the reentrants, the thickness of the enamel band (the Schmelzmuster), and others.

Among the oldest and most primitive species are Mimomys antiquus from the late Early Ruscinian (MN 14 b) of Peshniovo in western Asia (Zazhigin and Zykin 1984), and Mimomys vandermeuleni and Mimomys davakosi reported from the early Late Ruscinian (MN 15a) of western and southern Europe (Fejfar et al. 1990; Van de Weerd 1979). Mimomys vandermeuleni and Mimomys davakosi are more advanced than Mimomys antiquus from Siberia. Heavily worn molars recovered from fluvio-limnic deposits at the site of Malusteni, Roumania (Kormos 1932), and previously described as Mimomys moldavicus (or Promimomys moldavicus of some authors) belong most probably to Mimomys davakosi (Fejfar et al. 1998).

All these early species of Mimomys have an m1 with a highly complicated anteroconid. In the m1 of juvenile individuals, the mesial portion of the anteroconid is variably undulated. An enamel islet is positioned within the anteroconid of the m1 and in the posterior portion of the M3 as well. On the buccal side of the anteroconid, a Mimomys ridge (Mimomys-Kante) is developed, which is progressively reduced in later species of the Villanyian and Biharian. Moreover, also in contrast to later species of the Villanyian and Biharian (e.g., Mimomys pliocaenicus, Mimomys savini), (1) the enamel band is thick and undifferentiated, (2) the linea sinuosa (base of the tooth crown) of the m1 is faintly undulated, (3) the dentine fields of the m1 triangles T1 and T4 are still more or less confluent, and (4) the reentrants (synclines) lack cementum, which later appeared in the Mimomys lineage, as in the late Late Ruscinian (MN 15) Mimomys hassiacus from Gundersheim or the earliest Villanyian (MN 16a) Mimomys stehlini.

The early history of Mimomys in eastern Asia is poorly known. Early Pliocene records described as Mimomys sp. were reported from the early Yushean Gaozhuang assemblage (Yushe fauna, NMU 12), China (Lindsay 1994; Flynn et al. 1997; Qiu and Li 2003) that possibly correlates to the Early (MN 14) or Late Ruscinian (MN 15). Mimomys irtyshensis from the late Yushean Mazegou assemblage (Yushe fauna) (Flynn et al. 1997; Qiu and Li 2003) is also notable. It belongs most likely to the Early Villaniyan (MN 16) (Qiu and Li 2003). Mimomys orientalis recovered from the upper part of the Daodi Formation (Nihewan Basin), China, seems to be of the same age (Qiu and Li 2003; Zhang et al. 2003). Late Yushean Mimomys youhenicus also would belong to the Early Villanyian (Qiu and Qiu 1995).

The early Yushean (Early Pliocene, MN 14) Bilike site (Inner Mongolia, China) produced a primitive arvicolid species originally called Mimomys sp. (Qiu 1988; Lindsay 1994; Qiu and Qiu 1995). The size and dental pattern of this species that are not described in Qiu and Storch (2000) displays similarities with Mimomys (Cseria) gracilis from Central Europe (Repenning, in litt.). It is assumed to represent the earliest and most primitive record of Mimomys in China (Qiu and Qiu 1996).

Mimomys has an extensive fossil record in Eurasia, dating back to at least to the Early Pliocene (Mimomys antiquus, Ruscinian, MN 14b; Fejfar et al. 1997; Repenning 2003). However, in contrast to Eurasia, there is no consensus on the origin and dispersal history of Mimomys in North America. Many authors (e.g., Repenning 1968, 1987, 2003; Repenning et al. 1990; Woodburne and Swisher 1995) consider Mimomys to be an immigrant from Eurasia, whereas von Koenigswald and Martin (1984) assumed that it had never reached North America because the Schmelzmuster patterns do not match that of the Eurasian Mimomys (see also Martin 2003b; Martin et al. 2002, 2006). If so, the North American Mimomys must have evolved directly from a Promimomys grade arvicolid in North America (or possibly in Asia?), a conclusion that agrees with earlier suggestions by Lindsay et al. (1984: 483): "If North America species of Mimomys evolved from a different species of Promimomys than did the Eurasian species of Mimomys, as we propose, then separate North American and European lineages of Mimomys should be distinct."

Aratomys Zazhigin, in Gromov, 1972
Figure 1

Aratomys has originally been reported from the early Pliocene of Mongolia (Schaub 1934). The molars are rooted and cementless. The m1 of Aratomys is relatively short and broad, and displays three alternating basic triangles. The stucture of the anteroconid is distinctly more complex than that in Promimomys, which also occured in the Pliocene (Early Ruscinian) of Eurasia.

The species Aratomys multifidus is based on material collected at the site of Chono-Khariak 2 in western Mongolia. The deposits that produced the remains of Aratomys multifidus were referred to the late Early Ruscinian (upper MN 14b) by Zazhigin (Qiu and Storch 2000). Aratomys bilikensis was described from Bilike, Inner Mongolia, China (Qiu and Storch 2000), also early Yushean (Early Ruscinian, MN 14) in age (Qiu and Storch 2000; Qiu and Li 2003). According to Qiu and Storch (2000), Aratomys bilikeensis seems to be slightly more primitive than Aratomys multifidus.

Aratomys is larger than typical Mimomys. The molar pattern of Aratomys multifidus and Aratomys bilikeensis could suggest relations to the Mimomys branch (Fejfar et al. 1998), and comparison with early species of Mimomys (e.g., Mimomys vandermeuleni, Mimomys davakosi) reveals little differences in the occlusal pattern and the developmental stage of the tooth crown base (linea sinuosa).

Aratomys multifidus and Aratomys bilikeensis, as well as Kilarcola kashmirensis from Kilar, Kashmir (Kotlia 1994; Kotlia and von Koenigswald 1992), were assigned to Mimomys (Aratomys) by Repenning (2003), who considered this taxon a possible Oriental faunal region subgenus.

Dolomys Nehring, 1898
Propliomys Kretzoi, 1959
Figure 1

Dolomys is characterised by relatively large rooted hypsodont prismatic molars. As a representative of the tribe Ondatrini, Dolomys has five basic triangles on m1. In lateral view, all molars display a conical shape that resulted in a typical increase of the length and width of the cheek teeth in advanced stages of wear. The enamel base of the tooth crown (linea sinuosa) is faintly undulated. In moderately worn m1 of juvenile individuals, the mesial walls of the anteroconid are variably undulated; in heavily worn m1 of adult to senile individuals, however, the shape of the anteroconid is mushroom-like. The chewing surface displays a striking asymmetry: the lingual triangles and synclines are larger than the buccal ones. The triangles of the m1 are alternating, as they are in M1. The apices of the triangles are sharp in juvenile individuals, but rounded in adult and senile individuals. The enamel band is not differentiated; the reentrants lack cementum. A typical trait of the m1 is the deeply inserted reentrant (syncline) BRA 3 that distinguishes Dolomys from Mimomys where BRA 3 is always distinctly shallower (for details see Maul 1996; Popov 2004).

Dolomys species were present throughout the southern parts of western, central and eastern Europe during the Ruscinian (MN 14b and MN 15); it became extinct in the Villanyian (MN 16a), since the probable latest record is from Beremend 15, where Dolomys milleri co-occurs with Mimomys cf. hajnackensis and Mimomys cf. pitymyoides (Jánossy 1990). Important early species are Dolomys adroveri reported from Orrios 3, 9 and from Villalba Alta Rio 3, Escorihuela B, and Tollo Chiclana (TCH1), Spain (Fejfar et al. 1990; Minwer-Barakat 2004), Dolomys milleri (Nehring 1898) (e.g., Beremend 5, Hungary, Kretzoi 1956), Dolomys nehringi (e.g., Csarnóta 2, Hungary, Kretzoi 1959), as well as Dolomys occitanus from Sète, France, which was first described as Mimomys occitanus by Thaler (1955). Except for Csarnóta 2 (MN 14b), all these localities are Late Ruscinian (MN 15) in age. According to Fejfar et al. (1990), Dolomys milleri and Dolomys nehringi are more derived than Dolomys adroveri from Teruel basin, Spain. In the western Mediterranean Dolomys gave rise to the large rootless genus Kislangia.

It is worth mentioning, that a poorly preserved m1 found in Blancan 2 deposits of the Kettleman Hills of southern California and tentatively assigned to Dolomys indicates that muskrats possibly entered the North American fossil record earlier than previously thought (Repenning et al. 1995; Bell 2000).

The hypsodont prismatic molars of Propliomys are smaller than those of Dolomys. As revealed by Propliomys hungaricus from the late Early Ruscinian (MN 14b) of Csarnóta 2 (Hungary), the anteroconid is elongated, the so-called Vorderkappe sharply edged and inclined medially. Propliomys shares with Dolomys the asymmetry of the alternating synclines and the deep buccal syncline BRA 3. Moreover the lingual triangles are larger than the buccal ones, as in Dolomys.

Cosomys Wilson, 1932
Ophiomys Hibbard and Zakrzewski, 1967
Ogmodontomys Hibbard, 1941

The overall molar morphology of these three taxa shows such striking similarities with the cheek teeth of the Eurasian Mimomys that an allocation to this genus and the subgenus status of Cosomys, Ophiomys, and Ogmodontomys has been widely accepted for many years (e.g., Repenning 1987, 2003). However, this taxonomical assignment has been challenged, based on the molar enamel structure of these taxa (von Koenigswald and Martin 1984; for Ogmodontomys see also Martin et al. 2002; Martin et al. 2006). These authors suggest that Cosomys, Ophiomys, and Ogmodontomys did not evolve from the Mimomys stock (see above) but belong to a North American vole clade with its own history derived from a Promimomys grade arvicolid.

Ophiomys and Cosomys mark the beginning of the Blancan in North America (Repenning 1987). Ophiomys mcknighti, a small form, and Cosomys sawrockensis (or Ogmodontomys sawrockensis, see Hibbard 1957; Martin 2003a) a large form, are the hallmarks of the Blancan I (Repenning 1987). According to Martin (2003a – there referred to as Ogmodontomys), Cosomys sawrockensis is the index fossil for the central Great Plains rodent zone 4. It appeared by about 4.8 Ma for the first time (Repenning 2003). The m1 is characterised by three basic triangles, very low dentine tracts, persistent islet on the anteroconid complex, gradual loss of the Mimomys-Kante, and a primitve Schmelzmuster that consists apparently entirely of radial enamel (Repenning 2003).

Ophiomys mcknighti (Blancan I), known from localities of the western U.S. faunal region (Fejfar and Repenning 1992; Repenning et al. 1990) is also rather primitive because it has only low undulations of its linea sinuosa on the m1, indicating an early stage in the development of dentine tracts. Ophiomys mcknighti evolved through an intermediate form (Ophiomys "mcknighti -taylori") (Blancan II) into Ophiomys taylori (Blancan III) (Repenning 1987).

In the western U.S. faunal region, Cosomys (or Ogmodontomys of some authors) sawrockensis or a closely related species gave rise during the Blancan to (Repenning 1987; Bell 2000) Cosomys primus as well as to Ogmodontomys poaphagus, the index fossil for the central Great Plains rodent zone 6 (Martin 2003a). Cosomys disappears at the middle/ late Bancan boundary at about 3.0 Ma (Martin 2008).

In contrast to early members of Mimomys in Eurasia such as the Early Pliocene (Villanyian, MN 16) Mimomys hassiacus or Mimomys stehlini, the North American members of Cosomys, Ophiomys, and Ogmodontomys never developed cementum. Mimomys (Cromeromys) virginianus from the Cheetah Room fauna of West Virgina represents, according to Repenning (2003, in litt.), an independent immigration of Mimomys to North America, about at the beginning of the Pleistocene.

Protopliophenacomys Martin, 1975 (= Propliophenacomys Martin, 1975)
Pliophenacomys Hibbard, 1937

Protopliophenacomys is an endemic North American arvicolid taxon that appeared with Protopliophenacomys parkeri during the Late Hemphillian for the first time (Repenning 1987; Martin 2003a). It is the index fossil for the central Great Plains rodent zone 2 (Martin 2003a). Pliophenacomys is apparently a derivative of Protopliophenacomys. Protopliophenacomys displays a dental pattern that is intermediate between Promimomys mimus and Pliophenacomys (Martin 2003a). The oldest record of Pliophenacomys was described as Pliophenacomys wilsoni and comes from the Concha fauna of the site of Chihuahua near Yepomera (Lindsay and Jacobs 1985). It is Blancan I in age (Repenning 1987). In the eastern U.S. faunal region, Pliophenacomys wilsoni evolved into Pliophenacomys finneyi during the Blancan II, followed by Pliophenacomys primaevus in the Great Plains of Blancan III (central Great Plains rodent zone 9, Martin 2003a) (Repenning et al. 1990; Bell 2000).

Tobienia Fejfar and Repenning, 1998
Figure 1
Lemmini nov. gen. nov. sp. Fejfar and Repenning, 1998

Two potential ancestors of lemmings with rooted molars (A, B) were recorded in sandy lenses below the lignite seam in the lignite quarry at Wölfersheim near Frankfurt M, . Rheinhessem, FGR; the age of the fossilferous layer is the Latest Ruscinian, MN zone 15b (Fejfar and Repenning, 1998).

A. The first record: the molars of the genus Tobienia are rooted; the occlusal surface is simple „Mimomys-like“, relatively narrow. The longitudinal tooth axis is shifted slightly to the labial side (in lower molars) and to the lingual side side (in upper molars); this shifting of the longitudinal axis is best visible in the first lower and upper molars. Enamel bands are relatively thin; the anterior and posterior enamel edges of the prisms-triangles (lingual of the lower molars; buccal of the upper molars) are mostly straight The cementum in the reentrant folds is present.The nearest comparable taxon to Tobienia is the rootless and about 2,5 - 1,5 Ma years younger Mictomys vetus (Wilson, 1933) from the North American Late Pliocene. Tobienia represents an ancestral „Mimomys-like“ stage when the lemming type of enamel was not yet fully developed. B. The second record: Lemmini nov. gen., nov. sp.. in Wölfersheim: one rooted deeply worn first lower molar - with the occlusal pattern gennerally shaped as a „smaller Mimomys-like“ arvicoline. The anteroconid complex (Acc) is short, narrow and simple without Mimomys-Kante but with a circular, long, persisting islet that is not present in more advanced lemmings without rooted teeth; the interruptions of the buccal enamel wall are due to the distinct high dentine tracts on the posterior loop. The longitudinal occlusal axis of the deep alternating reentrants is slightly buccally shifted. The buccal enamel edges of the triangles are slightly differentiated in thickness: the anterior ones are thinner than the posterior ones; the lingual edges are equally thick. The buccal enamel crown base is strongly undulated similar to Tobienia. Fejfar and Repenning, 1998 stated that Tobienia belongs in the affinity of the tribe Synaptomyini whereas the second record of Lemmings - the Lemmini nov. gen. nov. sp. – represents the Lemmini tribe. Both are the first record of the Lemmings with rooted molars. Both tribes appear to have evolved from early Mimomys and thus belong to the subfamily Arvicolinae of the family Cricetidae; the subfamily Lemminae is no longer recognized.


Next Section

Origin of arvicolids
Plain-Language & Multilingual  Abstracts | Abstract | Introduction
Cheek Tooth Adaptations in Cricetids, Microtoid Cricetids and Arvicolids
Miocene and Pliocene Cricetids, Microtoid Cricetids and Arvicolids | DiscussionAcknowledgments | References
Print article