Since its establishment by Thaler in 1966, over forty species have been included in Eucricetodon (van de Weerd et al. 2021a), with type localities ranging in age from the Late Eocene to the Early Miocene. The species included in the genus occur across an extensive area stretching from the Isle of Wight to the Indian subcontinent and the eastern shores of Asia. Within this vast range, at least four Late Eocene–Oligocene palaeobiogeographic provinces may be recognised, each characterised by distinctive rodent assemblages and apparently distinctive Eucricetodon species.

The West European Province, covering present-day Spain, the United Kingdom, France, Germany, and adjacent regions, was a continuous landmass during the Early Oligocene (Figs 1, 2). Among the rodents, Eocene and Oligocene assemblages of Western Europe are characterised by a diversity of endemic theridomorphs and glirids. At the Eocene–Oligocene transition (EOT), a major faunal turnover named the Grande Coupure (Stehlin 1909) occurred in this province. The turnover is characterised by large-scale immigration of Asiatic taxa and extinction and replacement of endemic European land animals. At and after the EOT, the endemic theridomorphs and glirids of the West European Province were joined by muroids, eomyids, sciurids, and a cohort of large mammals, all of Asian origin, ending the isolation of the province.

The first appearance datum (FAD) of rodents and large mammals at the EOT is only applicable for the continuous landmass of the West European Province. The Dětaň site in Bohemia (West European Province; 21 in Fig. 2) has yielded an assemblage of muroid rodents and large mammals, including Entelodon, Lophiomeryx, Anthracotherium, and Elomeryx. Their first appearance in the West European Province is at the EOT (~33.7 Ma), necessitating allocation of the Dětaň assemblage to a time after this date, in MP21, MP22, or an even later zone. Biotite in tuffaceous material present in one of the fossiliferous beds yielded an age of 37.7 ± 1.5 Ma (early Priabonian; Fejfar 1987; Fejfar and Kaiser 2005), considered to be erroneous, and the site was placed in MP21. A bulk-sample of an overlying basaltic lava flow yielded a K/Ar age of 32.6 ± 1.7 Ma (Mikuláš et al. 2003), in agreement with allocation of the fossils to MP21 (Speijer et al. 2020). However, Maridet et al. (2025) propose that the 37.7 ± 1.5 Ma age (early Priabonian) is correct; consequently, Bohemia with Dětaň cannot be part of the West European Province, and the listed muroids and large mammals cannot be considered as a FAD for Bohemia. If correct, Bohemia-Volhynia is a small sub-province that served as a stepping stone between Asia and West Europe (Fig. 2). This proposal by Maridet et al. (2025) is backed up by a revised biostratigraphy of assemblages from fissure fillings located in Germany, west of Dětaň. For this interesting suggestion to be valid, a firm Late Eocene barrier, difficult for land mammals to cross, is needed between the Bohemian terrane and Western Europe (Fig. 2). Unfortunately, evidence for such a barrier is lacking, as the proposed potential barrier of the Rhine Graben became a marine strait much later, during a short period in the late Rupelian (Popov et al. 2004; Berger et al. 2005).

The West and Central Asian Province extends from the Aral Sea region in Kazakhstan across northeastern Kazakhstan (western Asia), northwestern China, Mongolia, Inner Mongolia (central Asia), and possibly into adjoining parts of southern Siberia (Fig. 1). Its boundaries—poorly defined and not coinciding with present-day coasts—depend largely on the distribution of fossil assemblages (data points) and the limited understanding of contemporaneous climatic zones. This province represents the northern margin of an increasingly arid zone that developed after the Middle Eocene, an area that has been the focus of extensive research (see Abels et al. 2011; Bosboom et al. 2011, 2015; Barbolini et al. 2020; Sun et al. 2022 and references therein). Several authors propose and discuss such a province (Tsubamoto et al. 2004; López-Guerrero et al. 2017; Mennecart et al. 2021; Li et al. 2022). We envision here an Late Eocene–Oligocene bioprovince encompassing the Late Oligocene assemblages from the Aral Sea area (Kazakhstan) in the west (Lopatin 1996, 2004; Bendukidze et al. 2009) and, in the east, the Late Eocene–Oligocene assemblages of the Erlian Basin (Chinese Nei Mongol; Li et al. 2016; Fig. 1). In between are the Mongolian Plateau, the Junggar Basin, the Valley of Lakes, and Ulantatal, with Oligocene assemblages (Vianey-Liaud et al. 2006; Maridet et al. 2009; Daxner-Höck et al. 2010; Gomes Rodrigues et al. 2012 and references therein). Rodent assemblages from this province are distinguished by a great diversity of dipodids and muroids, together with several genera of ctenodactylids, anaplodontids, cylindrodontids, and tsaganomyids—though the latter two groups are absent from the Aral Formation near the province’s western limits (Lopatin 2004). Collinson and Hooker (2003) characterise the Oligocene plant communities of the West and Central Asian Province as Asian Sclerophyllous; for a review of the Eocene–Oligocene history of its biome see Barbolini et al. (2020).

Tsubamoto et al. (2004) and Mennecart et al. (2021) observe that by the end of the Middle Eocene (Sharamurunian), faunal assemblages of northern and southern China began to diverge. Two palaeobiogeographic provinces subsequently emerged: the dry, temperate West and Central Asian Province described above, and the tropical, humid South Asian Province. The Caijiachong Formation in the Yuezhou Basin (Yunnan Province, China; Fig. 1) belongs to the South Asian Province. This formation exceeds 200 m in thickness and includes at least six fossiliferous horizons dated between 40 and 32 Ma (Maridet and Ni 2013; Wang et al. 2019; Lu et al. 2021). Rodent genera known only from this formation are the Early Oligocene muroids Caecocricetodon Lu et al., 2021, Paracricetops Maridet & Ni, 2014, and the baluchimyid (Asiatic hystricognath) Dianomys. Additionally, the muroids Cricetops and Eucricetodon (Li et al. 2017) are reportedly present in both provinces, albeit with different species. Unfortunately, accompanying faunal lists to provide broader contextual information are missing. In addition to rodents, large mammals have been recovered from this formation (Mennecart et al. 2021). Tsubamoto et al. (2004) linked the thermophilic Late Eocene–Early Oligocene mammals of the Caijiachong Formation with those of the Bugti Hills in Pakistan (Flynn et al. 1986; Marivaux and Welcomme 2003) to form a South Asian Province characterised by faunas of humid tropical nature. The Bugti Hills deposits are middle or Late Oligocene; magnetostratigraphy suggests ~27 Ma, or two million years younger. The rodent assemblages from the Bugti Hills are diverse and well documented, being dominated by six genera of baluchimyids, together with several diatomyid species, an anomalomyid, and two muroid genera (Métais et al. 2009). Late Eocene–Early Oligocene land mammals are not known from the Indian subcontinent, but have been recovered from several sites in eastern and southern Asia such as the Priabonian Na Duong site in Vietnam (Böhme et al. 2013), the Late Eocene–Early Oligocene Krabi Basin (Métais et al. 2001), and the Late Oligocene Nong Ya Plong locality (Thailand; de Bonis et al. 2021). Apart from the Bugti Hills, diverse rodent assemblages are not yet known from this province, but are expected to be recovered from the Caijiachong Formation. A baluchimyid has been described from the Krabi Basin (Thailand; Marivaux et al. 2000) and a diatomyid from Nong Ya Plong (Marivaux et al. 2004), while other rodents, if present, remain undescribed.

The Balkan-Anatolian Province is a neck of land formed by terranes now part of Iran, Anatolia, and the Balkans; it emerged in the Eocene and progressively separated the Paratethys and the Mediterranean Tethys. Evidence of South Asian mammals present in Balkan-Anatolia comes from incidental finds of large perissodactyle fossils mainly in Eocene coal pits in the Balkans (Heissig 1979, 1990; Nikolov and Heissig 1985; Hellmund and Heissig 1994; Böhme et al. 2013; Lihoreau et al. 2023). The Late Eocene fauna, typically consisting of Asiatic elements, was preceded by a highly endemic insular fauna occurring on the Pontide tectonic block (near location 16 in Fig. 2), a terrane now incorporated into Anatolia. Licht et al. (2017, 2022), Métais et al. (2018), and Montheil et al. (2025) date the insular fauna as to the Lutetian (~48–41 Ma); it includes embrithopods, large herbivores known from the African plate, up to five species of pleuraspidotheriids, ungulates outside Anatolia only known from the Paleocene of Western Europe, three taxa of metatherians (stem marsupials), a bat, and a primate (Beard et al. 2021 and references therein). The first Asiatic mammals invading Balkan-Anatolia, terminating the endemic Lutetian insular fauna, are dated to the early Bartonian (~ 41 Ma; Licht et al. 2022). Most Late Eocene and Early Oligocene reconstructions of the province envision a series of islands with active seaways separating the terranes (Heissig 1979; Popov et al. 2001; Grandi and Bona 2017; van de Weerd et al. 2023; Maridet et al.. 2025). Palcu and Krijgsman (2023) studied euxinic Paratethys sediments and established that by the end of the Eocene, the Balkan-Anatolia area effectively isolated the Paratethys (Fig. 2), transforming it into an euxinic basin.

The Late Eocene–Early Oligocene small mammal assemblages obtained by screen washing consist predominantly of rodents. Compared to the few incidental finds of Paleogene large mammals, these rodent assemblages are relatively rich. Muroids dominate and are assumed to be related to South Asian taxa—currently poorly known—because most Late Eocene muroids known from the West and Central Asian Province are unrelated. If this is correct, it is not surprising that several as yet undescribed muroid molars from the Early Oligocene of Serbia resemble Caecocricetodon from the Caijiachong Formation (site 12 in Fig. 1). In addition, baluchimyid and diatomyid rodents from Anatolia and the Balkans (Marković et al. 2017; van de Weerd et al. 2021a, 2023), as well as large-sized perissodactyls and several large ruminants (Mennecart et al. 2018, 2021) clearly originate from the South Asian Province.

Late Eocene and Early Oligocene fossil floras from the Rhodopian terrane (Balkans; Bozukov et al. 2009), together with an unpublished fossil flora of similar age collected near the Valniš locality (19 in Fig. 2; Djordjević Milutinović, pers. comm.) indicate a warm and wet climate, probably not very different from Late Eocene–Early Oligocene South Asian climate. Temporal barriers hampering faunal exchange were probably present along the relatively narrow land connection to Asia, but their effect was likely limited. In contrast, the deep Alpine foreland basins remained an effective barrier between Western Europe and the Balkan-Anatolian provinces during the Late Eocene and Early Oligocene. The West and Central Asian and Balkan-Anatolian Provinces were separated by both the Paratethys Sea and arid landscapes (Figs 1, 2). The Paratethys extended far eastward during the Eocene (Bosboom et al. 2011, 2015, 2017) and was connected to the Arctic by the Turgai Strait, which acted as a physical barrier between West and Central Asia and Western Europe throughout much of the Eocene.

The disappearance of the Turgai barrier at ~ 37 Ma (Poblete et al. 2021), together with the sea-level lowering at the Eocene–Oligocene transition (EOT), allowed Asian mammals to migrate into Western Europe via the northern route (Mennecart et al. 2021). This migration resulted in the faunal turnover termed the Grande Coupure by Stehlin (1909). Mennecart et al. (2021) showed that the South Asian ruminants Iberomeryx and Bachitherium reached Western Europe later—at ~31 Ma, via the southern route, that is, through Balkan-Anatolia and crossing the Alpine basins.

Three marsupial species occur in the assemblages from the Late Eocene–Early Oligocene sites of Serbia (Wessels et al. 2024; localities 18 and 19 in Fig. 2). Because these taxa, and unpublished insectivores, resemble those from the West European Province, faunal exchange was likely, already during the Late Eocene. In contrast, there was no faunal exchange of muroids between the two provinces prior to the Oligocene. The Early Oligocene entry of muroids into Western Europe via the southern route was probably limited, as major differences persisted between the Oligocene rodent assemblages of the Balkan-Anatolian and West European provinces. Faunal exchange in the other direction—from Europe to Balkan-Anatolia—is indicated by the presence of West European glirids, eomyids, and sciurids among the dominant non-European muroids in the middle Oligocene assemblages of the Thrace Basin (Ünay-Bayraktar 1989; locality 17 in fig. 20).

Thaler (1966) established several subgenera within Cricetodon Schaub, 1925, one of which was Eucricetodon. The type species, Cricetodon collatus Schaub, 1925, is from the Late Oligocene of Küttigen (Swiss Molasse). Thaler did not provide a formal diagnosis, as Eucricetodon was conceived as a provisional grouping for species that did not fit comfortably within other genera. From the outset, therefore, Eucricetodon represented a heterogeneous taxon. Thaler anticipated that future studies would identify additional genera, thereby resolving this heterogeneity.

Following Thaler’s work, the subgenera within Cricetodon were soon elevated to genera, and Mein and Freudenthal (1971) established the Subfamily Eucricetodontinae, to which numerous species from Europe and Asia have subsequently been referred. Building upon Thaler’s classification, Freudenthal (1996) proposed the genus Atavocricetodon for Early Oligocene (MP21–24) species that had previously been included in Eucricetodon. Freudenthal and Martín-Suárez (2016) further subdivided the remaining West European species into three principal groups, resulting in the following classification:

The “collatus” group corresponds to Eucricetodonsensu stricto, as originally envisioned by Thaler. Freudenthal and Martín-Suárez (2016) suggested that their groups represent distinct genera that separately reached Europe and do not show clear trends in time-successive species. Only the “collatus” group shows some trends in morphotype frequencies in time-successive species (Engesser 1985; Freudenthal and Martín-Suárez 2016).

Kalthoff (2000) examined lower incisors of Eucricetodon species belonging to the “dubius” and “collatus” groups and identified the K4 enamel microstructure (schmelzmuster). Gomes Rodrigues et al. (2013) sectioned lower incisors of Atavocricetodon from two Early Oligocene (MP22–23) localities in the Quercy region, recording the K1 enamel microstructure. A specimen referred to E. huerzeleri from an MP26 Quercy site also showed the K1 microstructure (Gomes Rodrigues et al. 2013), whereas the illustrated enamel of a specimen identified as E. cf. huberi (Belgaric-1, Quercy; MP25) corresponds to K7. If correct, the latter specimen is more appropriately assigned to Pseudocricetodon rather than Eucricetodon.

Thaler’s (1966) concept of Eucricetodon was originally based exclusively on European material. At that time, comparable rodents had not yet been reported from other regions. Today, however, Eucricetodon is recognised as a common element in Oligocene faunas of West and Central Asia (Fig. 1), with at least nine formally described species, several of which are poorly known, with problematic diagnoses. The large middle Oligocene species Eucricetodon asiaticus (Matthew & Granger, 1923) from the Hsanda Gol Formation in Mongolia is comparatively well studied owing to Lindsay’s (1978) comprehensive revision. Lindsay examined both the original type material and specimens collected much later. Not surprisingly, he recognised “two poorly defined populations”.

Eucricetodon caducus and E. deploratus Shevyreva, 1967 were described and poorly illustrated from limited material from Kyzyl-Kak, northeastern Kazakhstan. Russel and Zhai (1987) listed this locality as middle Oligocene in age. Given its size similarity to E. asiaticus, E. deploratus is regarded by many authors as a junior synonym of that species. Eucricetodon youngi Li & Qiu, 1980 is represented by two maxillary fragments from a small Early Miocene assemblage of the Xining Basin. The holotype, bearing an M1, M2 and partial M3, was not illustrated; a second specimen preserves two molars. Eucricetodon sajakensis Bendukidze, 1993 (see Bendukidze et al. 2009) and the slightly smaller E. occasionalis Lopatin, 1996 originate from the Aral Formation near the Aral Sea. Lopatin (2004) considered the formation to be earliest Miocene, though most authors regard it as Late Oligocene. Lopatin (2004) also described Eumyarion tremulus, which co-occurs with Eumyarion. occasionalis in Bonebeds 1 and 4. Given their similar size, morphology and co-occurrence, these may represent different morphotypes of a single species.

Three species have been described from the extensive Eucricetodon assemblages at Ulantatal (Nei Mongol, China): the large E. asiaticus and E. jilantaiensis, and the smaller E. bagus. The two large species are distinguished through Fourier analysis of M1 outlines (Vianey-Liaud et al. 2011) and by certain morphological features of the molars. Intermediate morphotypes—normally expected in such mixed assemblages—have not been reported (Gomes Rodrigues et al. 2012). Successions of Eucricetodon assemblages have been documented from the Junggar Basin (Maridet et al. 2009), Ulantatal (Gomes Rodrigues et al. 2012), and the Valley of Lakes (López-Guerrero et al. 2016). Most sites contain two size-differentiated species. Frequency of morphological traits varies through time in these successions, but clear directional trends are not evident—evolution appears mosaic-like. This is similar to the observations in Europe (Freudenthal and Martín-Suárez (2016). Maridet et al. (2009) also observed a “strong morphological affinity between Asian and European species” in the Junggar Basin material.

Available observations of Eucricetodon incisors of West and Central Asia (Table 1) show a prominent longitudinal ridge, and one or two parallel ridges, indicating presence of a syncline in the microstructure. Presence is confirmed in three species. Current evidence thus indicates the presence of K10 incisors in West and Central Asiatic Eucricetodon species. Further work is required to confirm this pattern, particularly for specimens from the Junggar Basin, to test the suggestion of European ties. Interesting too would be the investigation of the incisor structure in the oldest Asiatic Eucricetodon species, E. wangae Li et al., 2016 from the Late Eocene of Inner Mongolia.

Eucricetodon meridionalis and E. leptaleos were described by Wang and Meng (1986) from strata initially interpreted as Lower Oligocene in the Caijiachong Formation, later reassigned to the Upper Eocene. Wang (2007) transferred the two species to Eocricetodon and Oxynocricetodon respectively. Two muroids are known from the Oligocene assemblage of Paali Nala C2 in the Bugti Hills (Pakistan; Marivaux et al. 1999): Pseudocricetodon nawabi and Atavocricetodon paaliense. The latter, closely related to Oxynocricetodon leptaleos, is probably more appropriately included in Oxynocricetodon, while P. nawabi resembles Oligocene Kerosinia from the Thrace basin (Gomes Rodrigues et al. 2012). The incisor enamel microstructure of these species remains unknown, and molars attributable to Atavocricetodon or Eucricetodon have as yet not been found in Oligocene deposits of this bioprovince.

A small assemblage of Eucricetodon-like molars from Kocayarma (Thrace basin; 18 in Fig. 2) has been identified as E. aff. dubius (Ünay-Bayraktar 1989) and a few morphologically similar molars from the Late Oligocene of Ugljevik (northern Balkans) have been referred to Eucricetodon sp. (van de Weerd et al. 2021b). The Ugljevik Eucricetodon includes an associated lower incisor exhibiting the K10 enamel microstructure.

The four Eurasian palaeobiogeographic provinces outlined above are characterised by distinct rodent assemblages and by the presence of effective dispersal barriers. Species assigned to Eucricetodon occur in three of these provinces. In the Balkan-Anatolian and West and Central Asian Provinces, the incisor enamel exhibits the K10 microstructure, whereas in the West European Province, Eucricetodon possesses the K4 type, indicating different evolutionary trajectories. The oldest occurrence of Eucricetodon with confirmed K10 incisors is from Yeniköy (Sivas basin; 15 in Fig. 2) van de Weerd et al. 2023). Given the phylogenetic significance of enamel microstructure (Kalthoff 2006; van de Weerd et al. 2026) and based on a thorough review of all available evidence, the Gözükızıllı (Çankiri-Çorum basin; 16 in Fig. 2) and Yeniköy species possessing K10 incisors cannot be retained within Eucricetodon and are herein referred to a new genus. Similarly, the Asiatic species are better excluded from Eucricetodon. These taxonomic revisions restrict Eucricetodon to the species of the West European Province belonging to the “collatus” group, as originally intended by Thaler (1966), and probably also the “dubius” group as defined by Freudenthal and Martín-Suárez (2016).

Superfamily Muroidea Illiger, 1811

Molecular studies of extant Muroidea show that the Spalacinae, Rhizomyinae, and Myospalacinae subfamilies are sister groups within the monophyletic Spalacidae (Steppan and Schenk 2017; He et al. 2020). Extant Spalacinae, Myospalacinae, and one of the three extant genera of the Rhizomyinae are characterised by the K10 schmelzmuster of the lower incisor (Kalthoff 2000); two rhizomyine genera possess the K2 incisor, which is clearly a recent acquisition, as will be discussed below. The K10 incisor micropattern is present in many extinct muroids, including Dekamyarion, and these extinct muroids can be phylogenetically linked to one of the three extant subfamilies of the Spalacidae; others represent extinct branches within the family (van de Weerd et al. 2026). The K10 incisor microstructure thus appears to be a stable phylogenetic feature characteristic of the Spalacidae as proposed by van de Weerd et al. (2026). Fig. 5 shows the stratigraphic ranges of extinct genera and subfamilies characterised by the K10 schmelzmuster. It is similar to the chronological scheme for Balkan-Anatolia in van de Weerd et al. (2026), but is here expanded to include regions in Asia, Africa, and North America. Additionally, Fig. 5 shows the phylogeny of extant Spalacidae, based on molecular data from Steppan and Schenk (2017), superposed over the ranges of the extinct taxa. A similar phylogeny established by He et al. (2020) is not shown. Due to space constraints, only selected genera are shown in the Fig. 5; Table 2 lists all genera assigned to the extant and extinct subfamilies of the Spalacidae.

Molecular studies estimate the divergence between Spalacidae and other living muroids at approximately 34.5 Ma (Late Eocene; Steppan et al. 2004; Steppan and Schenk 2017), He et al. (2020) estimate this divergence between 36.6 and 27.3 Ma. This is broadly contemporaneous with the emergence of the Pseudocricetodontinae, Paracricetodontinae and Melissiodontinae, distinguished by the K7, K8 and K1 incisor structures respectively (van de Weerd et al. 2026). The K10 incisor structure, characterising the Spalacidae, may have evolved during this interval, but the oldest evidence currently known of the K10 structure is younger, with ~29 Ma of Dekamyarion in Yeniköy (Fig. 5).

By the Late Oligocene, at least thirteen muroid genera had emerged that display the K10 schmelzmuster in the lower incisor (Fig. 5; van de Weerd et al. 2026). This radiation corresponds temporally with the divergence of the three extant sister subfamilies — Spalacinae, Myospalacinae and Rhizomyinae—as established by the molecular analyses (Steppan and Schenk 2017; He et al. 2020). Figs 3, 4 show the earliest genera that can be linked to the three extant subfamilies and to extinct groups. The Oligocene genera not linked to an extant or extinct group may be considered the belong to the stem group of the Spalacidae. In Balkan-Anatolia these include Dekamyarion, Muhsinia, Enginia and Meteamys. Incisors of only a few Asiatic Oligocene genera have been investigated; Fig. 5 includes three. Aralocricetodon, from West and Central Asia (Lopatin 2004; Bendukidze et al. 2009; López-Guerrero et al. 2017) possesses K10 incisors (in Kalthoff 2006 erroneously identified as Eumysodon, see Marković et al. 2019). Paracricetops from South Asia is included in Fig. 5, because its lower incisor shows a weak but distinct longitudinal rib (Maridet and Ni 2013), suggesting presence of a syncline. Its taxonomic relationship remains unclear (Marković et al. 2019). The Asiatic Eucricetodon of Fig. 5 will be discussed below.

The scheme of Fig. 5 is simplified after the two phylogenetic studies of fossil and extant rhizomyines by López-Antoňanzas et al. (2013, 2015). Both studies differ from each other in the details of the Miocene-Pliocene evolution of this subfamily—details not shown in Fig. 5. The earliest known rhizomyide is Prokanisamys kowalskii, from site Z113 (Zinda Pir, Pakistan) with an Early Miocene age (~23 Ma; Lindsay and Flynn 2016). Shown in Figs 3, 4 is the more recent P. arifi, the type species (from site H-GSP 116), selected for its available illustration. Prokanisamys kowalskii, has slightly lower-crowned molars. This genus gave rise to Kanisamys; four species, ranging in age from ~14 to 5.5 Ma, are described from Siwalik sites. Several genera, closely related to Kanisamys, are known from the Siwalik region: Rhizomyides (five species, ranging from ~9.8 to 2.5 Ma), Eicooryctus (8–7 Ma) and Protachyoryctes (8–7 Ma). Anepsirhizomys, from the late Pliocene, is the last known rhizomyine from the Indian subcontinent (Flynn 1982). Only Rhizomyides is shown in Fig. 5, but all genera are listed in Table 2.

Two migration events from the Indian subcontinent to Africa involving rhizomyines with K10 incisors are proposed. Wessels et al. (2003) described Prokanisamys from a site in Libya dated at ~18 Ma, supporting the first event. An assumed descendant is Pronakalimys from the Old Fort Ternan site in Kenya (~13 Ma; Tong and Jaeger 1993); it gave rise to Nakalimys lavocati from the Nakali site in Kenya (~10 Ma; Flynn and Sabatier 1984). This initial migration at ~18 Ma predates the divergence of African Tachyoryctini (extant Tachyoryctes) and Asiatic Rhizomyini (extant Cannomys and Rhizomys) inferred from molecular data (Fig. 5). Steppan and Schenk (2017) dated the split at ~8 Ma, and He et al. (2020) dated it between 15.3 and 10.6 Ma.

The lower incisors of Rhizomys and Cannomys (from south-east Asia) have the K2 microstructure and lack a longitudinal ridge on the anterior surface (Kalthoff 2000). This contrasts with all other rhizomyines, and the Spalacinae and Myospalacinae, which show the K10 micropattern. Given their phylogenetic relationship with Tachyoryctes (Steppan and Schenk 2017; He et al. 2020), their Hunter-Schreger bands must have rotated 90°, that is, from the longitudinal orientation of the HSB in the K10 microstructure to the transverse orientation in K2. Flynn (2009) placed Miorhizomys—with six species—in the Rhizomyini, alongside Rhizomys and Cannomys. Similar to Rhizomys and Cannomys, its incisor is smooth (Flynn 2009: Khan and Flynn 2017), implying that it does not have the K10, but probably the K2 structure. The first known appearance of Miorhizomys is in the Siwaliks at ~10 Ma (Flynn 2009; Fig. 5), where it coexisted with the K10 taxa Kanisamys and Rhizomyides in the Siwaliks till ~6 Ma (Khan and Flynn 2017). This divergence of Rhizomyini from Tachyoryctini is earlier than proposed by Steppan and Schenk (2017), and more in line with He et al. (2020).

The divergence of the Spalacinae from Myospalacinae and Rhizomyinae was dated to the Late Oligocene (Fig. 5; Steppan and Schenk 2017; He et al. 2020). De Bruijn et al. (2022) proposed a scenario in which Spalacinae represent a distinct lineage deriving from an Eocene muroid ancestor. However, with the additional evidence from He et al. (2020), unbeknownst to the authors at that time, this scenario seems untenable. Four Late Oligocene–Early Miocene genera from the Balkans, Heramys, Vetusspalax, Pannoniamys, and Debruijnia, are classified within the Spalacinae (Fig. 5; Vetusspalax is shown in Figs 3, 4). These exhibit considerable diversity in molar morphology. Lopatin (2004) considers Argyromys (Oligocene Aral Fm.) related to Debruijnia and Heramys and consequently includes it in the Spalacidae. Among the Early Miocene- Oligocene genera, only Debruijnia may be related to Sinapospalax, Pliospalax, and extant Spalax (de Bruijn 2016); Vetusspalax, Pannoniamys, and Heramys, represent extinct spalacine lineages. Extant Spalax possesses the K11 incisor enamel pattern, a slightly more derived variant of K10 (Kalthoff 2000). The fossil genera Pliospalax and Debruijnia exhibit the K10 pattern. The enamel structure of Vetusspalax is identified for this study as K10. The geographical range of the subfamily remained confined to the Balkan-Anatolian Province—Argyromys excepted. Land connections formed later in the Miocene and facilitated the dispersal of Spalax (or its progenitor) into the Middle East and North Africa.

The Cricetodontinae (including DeperetomysMarković et al. 2019) is a very large group containing Cricetodon, with an estimated 30 described species, and some eleven other genera (Table 2). These numbers indicate the importance of this subfamily, ubiquitous in sediments ranging in age from the Late Oligocene to the Pliocene of Europe, Asia Minor (van de Weerd et al. 2026) and West and Central Asia.

Recent phylogenetic studies of the Cricetodontinae are by Martin et al. (2023) and López-Antoñanzas et al. (2024). Although different in content, the latter study is much larger; a remarkable difference in the position of Deperetomys species in the two studies is to be noted. In Martin et al. (2023), Deperetomys species are in a tight cluster in the Cricetodontinae; the second study splits Deperetomys species between two subfamilies. While the result of Martin et al. seems more consistent, the disagreement underscores the importance of replicating studies using independent datasets and approaches.

Martin et al. (2023) described Deperetomys dingusi and Pliotomodon primitivus from the Miocene of western North America (Fig. 5). The first has a distinct central ridge on its lower incisor (Fig. 6), indicative of the K10 structure, but not confirmed by study of its microstructure. Pliotomodon closely resembles the Late Miocene cricetodontine Byzantinia; its incisor is not known. Martin et al. (2023) conclude that Deperetomys and Byzantinia have reached North America in the latest Oligocene and Late Miocene, respectively.

Phylogenetic analyses showed the close relationship of Gobicricetodon to the Cricetodontinae (Martin et al. 2023; López-Antoñanzas et al. 2024). Several authors (Flynn 2009; Li and Wang 2015; Qiu and Li 2016) proposed that the extant Myospalacinae (zokors) of central Asia descended from Gobicricetodon, Li and Wang 2015 reflected this in their classification of the Myospalacinae in the Spalacidae. Cricetodon sonidensis is the first known cricetodontine to appear in West and Central Asia in the Early Miocene (Fig. 5); it may have migrated from the Balkan-Anatolian region to West and Central Asia. If so, the zokors are the last surviving cricetodontines, Figs 7, 8 illustrate the sequential change in molar morphology from Early Miocene Cricetodon to Pliocene Prosiphneus, although evolution may have followed a different pathway. Clearly evident in the illustrations are the change from bunodont to lophodont, from a cuspidate to a flat occlusion surface, the increase in molar height, and the development of enamel free areas on the lateral molar sides.

Eumyarion was described by Thaler in 1966 from the Middle Miocene of Europe; at that time, it was considered an immigrant in the region. Later, it was found to be present in upper Oligocene and lower Miocene strata of the Balkans and Anatolia (Fig. 5), where up to three species are present in Early Miocene assemblages (Bilgin et al. 2023, and references therein). The last appearance of Eumyarion in Anatolia is in local zone G (de Bruijn et al. 2013b). Several species are described from the Middle Miocene of Europe; the last occurrence is in the Late Miocene. De Bruijn et al. (2013a) included Mirrabella in the Eumyarioninae, but specifically excluded Alloeumyarion Qiu, 2010 (Early Miocene, coastal province of Jiangsu, China). However, molars of Metaeucricetodon Qiu & Li, 2016 (Early Miocene, West and Central Asia) resemble those of Eumyarion, and if its lower incisor is type K10, it could be a spalacid.

The oldest known Anomalomys is A. aliverensis Klein Hofmeijer & de Bruijn, 1985 from Aliveri (Greece) of Early Miocene age (MNA-E, MN4). These late-Early Miocene species have derived, compact, lophodont-hypsodont molars with a flat occlusion surface (Figs 3, 4). Striking in M1 are the long anterior arm of the protocone and the mesoloph, both reaching the labial tooth margin. Similar features are well developed in Eumyarion, which is therefore mostly suggested as related to the Anomalomyinae (De Bruijn et al. 2013a).

The remaining genera Dekamyarion, Meteamys, Enginia, Muhsinia, Paracricetops, and Asiatic Eucricetodon (Fig. 5) are not linked to one of the later subfamilies or lineages but are nevertheless included in the Spalacidae because of their K10 incisors.

Investigated Eucricetodon species from the West and Central Asian paleobiogeographic province show the K10 incisor structure only (Table 1, Fig. 5). Additionally, several species exhibit one or two anterior incisor ridges typical of the K10 schmelzmuster. The oldest of these, “E.” wangae, originates from the Eocene Red Beds of Erden Obo. These are shown in Fig. 5 as “Asiatic Eucricetodon”. If incisor sections confirm presence of the K10 pattern, “E.” wangae would represent the earliest known occurrence of the trait. Figs 9, 10 show molars of some selected species, illustrating their relatively unspecialised morphology.

The genera of the North American tribe Leidymini (Fig. 5) were long considered closely related to, and descended from, Eurasian Eucricetodontinae, because of their similar molar morphology (Martin 1980; Lindsay et al. 2016 and references therein). These North American “cricetids” have K10 incisors (Kalthoff 2006; Lindsay et al. 2016), believed to be an independent acquisition, as it was known that Eucricetodon has K4 incisors. The observation that Asiatic “Eucricetodon” has K10 incisors implies that the K10 structure in the Leidymini is not an independent acquisition but inherited from their Asiatic “cricetid” forbears that crossed Beringia and settled in America.