A new species of Megacricetodon from the Early-Middle Miocene of Czech Republic and its importance for the understanding of the earliest evolution and dispersal of the genus in Europe

ABSTRACT The sudden appearance of modern cricetids in Europe during the Early Miocene was one of the most significant mammalian events of European Neogene. A characteristic representative of this mammalian turnover was the genus Megacricetodon that quickly became major components of the rodent faunas across Europe. Here, we review the fossil record of this genus from the Early-Middle Miocene (MN4-MN5) localities of Czech Republic, the key area for the understanding the earliest history of this genus in central Europe. We provide an age refinement of the localities under study, especially in the context of new data from the Northern Alpine Foreland Basins. A detailed analysis of the available material revealed a presence of Megacricetodon grueneri sp. nov., described here from Dolnice 3 (MN4a/b). The oldest record of Megacricetodon comes from Ořechov (~late MN4a) and marks the first occurrence of the genus in central Europe. We discuss and propose the most likely scenarios for the earliest phylogeny and dispersal of Megacricetodon into central and western Europe that were strongly influenced by major tectonic processes in the Pannonian and Pre-Alpine areas having a major impact on the development of terrestrial ecosystems and the composition of faunas. Zoobank: http://zoobank.org/urn:lsid:zoobank.org:pub:D2154E84-BD58-4AF9-80C7-8E8AEED7517D


Introduction
The Early Miocene is a time period in which fundamental palaeogeographical changes occurred in Europe having a major impact on the development of terrestrial ecosystems and the composition of faunas.During the 'Cricetid vacuum' (~MN3) most of the primitive Oligocene cricetid taxa became extinct in central and western Europe, and Melissiodon was the only cricetid recorded during that time.The gradual interconnection between the Balkan Peninsula and Asia Minor during the Burdigalian caused the migration of a number of Asian taxa to Europe via a newly formed landbridge.One of the most significant biotic events was the sudden appearance of modern cricetids, initially represented by the genera Democricetodon and Megacricetodon.This relatively rapid dispersal into central and western Europe had to be strongly influenced by major tectonic processes in the Pannonian and Pre-Alpine areas.
Megacricetodon Fahlbusch, 1964 is a significant rodent genus during most of the European Miocene since it is remarkably diverse and abundant.Its high species diversity with relatively short stratigraphic ranges has turned it into one the most characteristic biostratigraphic markers of the continental Miocene faunas in Europe.It is used to define a number of local biozones for the Early and Middle Miocene in several European regions (Cicha et al. 1972;Daams et al. 1999;Abdul Aziz et al. 2008;Kälin andKempf 2009, 2010;Van der Meulen et al. 2012;Casanovas-Vilar et al. 2016;Prieto and Rummel 2016;Jovells-Vaqué and Casanovas-Vilar 2021), in combination with other rodent taxa, also to higher-ranking biochronological units, such as MN (Mammal Neogene) zones in Europe (Mein 1989(Mein , 1999;;Agusti et al. 2001;Hilgen et al. 2012).
In recent years, Megacricetodon has been studied quite intensively and in a number of regions of Europe such as e.g. the Iberian Peninsula (Oliver and Pelaez-Campomanes 2013;2014, 2016;Jovells-Vaqué and Casanovas-Vilar 2021, and references therein), the taxonomy of many species has often been reasonably well elucidated.Several authors have proposed different lineages of this genus for the Early and Middle Miocene in Europe (Daams and Freudenthal 1988;Heissig 1990;Aguilar 1995Aguilar , 1997;;Lazzari and Aguilar 2007;Abdul Aziz et al. 2010;Oliver and Pelaez-Campomanes 2013, 2014, 2016).However, the phylogenetic and palaeobiogeographic relationships between the taxa, particularly between western and south-eastern Europe have not yet been fully clarified.From this perspective, the territory of the Czech Republic, where it is recorded some of the oldest Megacricetodon samples in Europe, is undoubtedly a territory of crucial importance for understanding the earliest history of the genus.Unfortunately, since the extraordinary pioneering studies of Fejfar (1974), the findings from there have not been reviewed in detail.
Thus, in the above framework, the aim of this article is to clarify the taxonomy of the Early-Middle Miocene records of Megacricetodon from the Czech Republic and then discuss and propose the most likely scenarios for the dispersal and phylogeny of the oldest representatives of the genus within Europe.

The localities
The Megacricetodon material analysed in this paper stems from five Early to Middle Miocene (MN4-5) localities of the Czech Republic (Figure 1).

Location and geological settings
Dolnice 1-3 and Františkovy Lázně are located in the northwestern part of the Bohemian Massif, in the Cheb Basin (western Ohře/Eger Rift).The localities are in a superposition within littoral facies consisting of the greenish calcareous marls and silts.The locality of Dolnice is represented by three fossiliferous horizons (Dolnice 1-3) developed above the Main Coal Seam Formation (~early Burdigalian; cf., Shrbený 1994;Rojík 2004) of the basin.The fossil material was excavated near the surface in a trench for a water pipe nearby the Dolnice farm in northern suburbs of Cheb (Fejfar and Roček 1986).The closely neighboring (~2 km NE) locality of Františkovy Lázně contains deposits of the Cypris Formation (~late Burdigalian; cf., Shrbený 1994;Rojík 2004), where the layers immediately overlying the previous formation.The analysed material was excavated in the foundations of school building in 1957-1958 (Fejfar et al. 1959).
Another locality, originating from Cenozoic relicts in a region of the České Budějovice Basin (Southern Bohemia), is Strakonice.This locality was in an abandoned brickworks pit north of the town of Strakonice (Fejfar 1974).The fossil material was excavated from the calcareous marl lenses belonging to the lower part of the Mydlovary Formation (~early Langhian; cf., Shrbený 1994;Ševčík et al. 2007).
The last two localities lie on the eastern slopes of the Bohemian Massif, near the Brno city in South Moravia Region.The locality Ořechov was in the small outcrop located in the south of the homonymous village.The fossil material originates from the near coast oligohaline deposits of the Carpathian Foredeep consisting of fine to medium grained calcareous-glauconitic sandstones with layers of cross-bedded fine ferruginous sands (Fejfar 1974).The mammal record was located below the Rzehakia (Oncophora) beds (Cicha et al. 1972;Fejfar 1974Fejfar , 1990)); the marine sediments with endemic mollusks, generally correlated with the late Ottnangian (Čtyroký 1968;Piller et al. 2007;Kováč et al. 2018).This made possible a direct parallelisation with the marine layer sequence of the Central Paratethys.The locality of Mokrá Turtle-Fissure was situated in the Western Quarry of the Mokrá open-cast limestone mine a few kilometres northeast from Brno (southeast of the Moravian Karst, close to the margin fault of the West Carpathian Foredeep) (Ivanov and Musil 2004).The fossil material has been found in terrestrial fossiliferous deposits of a karst fissure formed within the massive and biodetritic Vilémovice limestone (Givetian-Frasnian) (Ivanov et al. 2006).

Biochronology
The mammalian assemblages from the localities under study contain the first modern cricetids that clearly indicate an age after the 'Cricetid vacuum' (Daams and Freudenthal 1990), i.e. younger than the MN3.

Material and methods
The Megacricetodon material under study comes from five Early to Middle Miocene localities of the Czech Republic (see Figure 1).Specimens from Dolnice 1-3, Ořechov, Františkovy Lázně, and Strakonice are stored at the Department of Paleontology (National Museum, Prague); specimens from Mokrá -Turtle Fissure are stored at the Department of Geological Sciences (Faculty of Science, Masaryk University).
The nomenclature of the cheek teeth of Megacricetodon is based on Oliver and Pelaez-Campomanes (2013).The description and distribution of character states are listed in comparative tables, using character states modified after Oliver andPelaez-Campomanes (2013, 2016).Data from specimens with moderate to high wear are not considered (Appendix 1, description of character states and their distribution in Tables 1 -24 there).The photographs of the occlusal surface of the cheek teeth were made with a scanning electron microscope model S-3700 N.
Measurements were taken with tooth orientations following the methodology proposed by Daams and Freudenthal (1988), except for the M2, which is after Oliver and Pelaez-Campomanes (2013).Length and width represent the maximun antero-posterior and bucco-lingual distances, taken perpendicular to each other.The measurements have been taken using an Olympus SZX12 and are given in millimetres.
The 3D images were made using a Micro-CT scanner Nikon XTH 160 and processed using the ImageJ software (software Fiji-ImageJ 1.04s

Derivation of name
Named in honour of Joseph Sebastian Grüner (*1780- †1864), the police and municipal councillor of the town Cheb also an amateur mineralogist, naturalist, and collector of Cheb folk traditions, who managed to obtain the first find from Dolnice (a mastodon tooth).
In 1822, he showed it to his friend the German poet Johann Wolfgang Goethe (*1749- †1832) who recognised the importance of this specimen.Their joint efforts subsequently highlighted the palaeontological importance of the Cheb region.(see Fejfar 2011 for details)

Diagnosis
Large-sized species of Megacricetodon characterized by long and slender crests.The lower first molars with rounded and simple anteroconid; the anteroconid cusp is lower than the other four main cusps, and the metalophulid is anteriorly connected.The mesolophids of short to medium length in m1 and m2.The upper molars with well-developed or medium mesolophs in M1 and M2.17-21, Appendix 2).The anteroconid is rounded and simple.The lingual and labial spurs of the anterolophulid is absent.The metalophulid is anteriorly connected.There is a short labial mesocingulid.The lingual mesocingulid is absent.The mesolophid is medium.The ectomesolophid is absent.The hypolophulid is anterior.
M3. (Figure 3j, q; Appendix 1: Table 16).The only teeth available are from Františkovy Lázně.The labial anteroloph is always present, it is extended to the paracone in eight out of nine and it is not connected to the paracone in the remaining one.The lingual anteroloph is absent in five, well developed but not connected in one and connected to the protocone in two specimens.The paracone is well developed.There is a mesostyl in two out of nine specimens.The metacone is absent in two and present in six teeth.The hypocone is always present (9/9).The metalophule is absent in one, it is connected to the neo-entoloph in two, it is connected to the protolophule and the anterior arm of the hypocone in four specimens and it is connected to the posterior arm of the protocone in one (Appendix 1: Table 16).The mesoloph is absent in four out of nine specimens, it is short in one, medium in two and long in the remaining two.The second protolophule is absent in two, it is connected to the metalophule in one, it is connected to the anterior arm of the hypocone in one, and it is connected to the anterior arm of the hypocone and the metacone in four.The posterior arm of the protocone is absent.In two out of seven the neo-entoloph is present.The sinus is absent in one, incipient in another one and it is present in three specimens.The posteroloph is short in one, and it is long and connected to the metacone in six specimens.
m3. (Figure 3r; Appendix 1: Table 24).The labial anterolophid it is short in one, and it is connected to the protoconid in five.The lingual anterolophid is short in two, and medium in the remaining five.The labial mesocingulid is incipient (1/7) or well developed (6/ 7).The mesolophid is always absent.

Remark
Available teeth of Megacricetodon grueneri sp.nov.from the Czech Republic show a conservative pattern, with stability in morphology and size through time.The small differences, in the relative abundance of some character states observed between assemblages, are interpreted as due to the scarce material available from some of the localities (e.g.Mokrá -Turtle Fissure, see Figure 4 and Appendix 1: Tables 1-24).Since its first occurrences in the MN4 (Dolnice 1-3, Ořechov, and Mokrá -Turtle Fissure) to its last occurrences in the early MN5 (Františkovy Lázně) (Figure 6), this species shows no significant evolutionary trend in size (Table 1 and Figures 4, 7) or in morphology (Appendix 1: Tables 1-24).The most characteristic dental pattern is represented by lower first molars with rounded and simple anteroconid; cups with different height, where the anteroconid cusp is lower than the other four main cusps (Figure 5).Lower molars with mesolophids of medium to short length, and upper molars with mesolophs well developed or medium length.Megacricetodon aff.grueneri (Figure 3x, Table 1, Appendix 1: Tables 1-10)

Remark
The single specimen (M1) from Strakonice corresponds to the phenotype of the new species described here both in its dimensions (Table 1) and basic morphological characters, such as the long and slender crests, slightly subdivided anterocone and medium mesoloph.However, the age of the locality is correlated with the late MN5 (see Figure 6 and 'Ages of Megacricetodon bearing localities . ..' below) and is thus distinctly younger than other localities with occurrences of M. grueneri sp.nov.Considering the very limited material from Strakonice and the proposed age of this locality, we tentatively classify the M1 as Megacricetodon aff.grueneri.

Palaeogeographical background
The appearance and subsequent dispersal of Megacricetodon in Europe were linked with important tectonic changes that took place during the Early Miocene.The counterclockwise rotation of the Arabian Plate and its collision with the Eurasian plate system resulted in the gradual closure of the Indopacific/Mediterranean/ Paratethys seaway (Steininger and Wessely 2000).For the first time, the Mediterranean was cut off from the Indian Ocean (Rögl 1999).
During a global sea level drop (cycle TB 2.1 sensu Haq et al. 1988) a continental migration bridge between Eurasia and Africa, known as 'Gomphotherium-landbridge', was formed (Rögl 1999, Agustí et al. 2001) and enabled a remarkable mammal exchange.The Balkan Peninsula was completely connected with Asia Minor except for some lakes or lagoons in the northern Aegean area (Dermitzakis and Papanikolaou 1981).The sea level fall also accentuates the beginning isolation of the Paratethys from the Mediterranean Sea (Piller et al. 2007).The tectonic activities closed off the marine realm of the Eastern Paratethys (Rögl 1999).This causes the formation of the Kotsakhurian Sea with the peculiar endemic semi-marine fauna (Popov et al. 1993).During the Ottnangian, the Central Paratethys was connected across the western Alpine foredeep with the western Mediterranean Tethys by a narrow active seaway, during which a basin-wide transgression flooded the North Alpine Foreland Basin (Allen et al. 1985;Rögl 1998).At this time, the western Mediterranean Tethys was also connected with the North Sea realm through the Rhine Graben (Martini 1990).During the Karpatian, these western marine connections ceased.The entire western Paratethys, the western Molasse Basin and the Rhine Graben dried up (see Rögl 1998).The connection into the Eastern Paratethys was reduced to a few narrow gateways (Piller et al. 2007) and connection to the Mediterranean was open only through the 'Trans-Tethyan-Trench-Corridor' (Bistricic and Jenko 1985) that disappeared in the late Badenian (Rögl 1998).

Weaknesses and limitations
The distribution of localities and mammal taxa is patchy in space and time (Van der Meulen et al. 2011).The chronological correlation of biotic events is complicated by a number of natural phenomena such as variable spatial distributions of taxa, differences in migration rates, local evolutionary trends with presence of anagenetic lineages, or the very limited geographical ranges of some taxa (Van der Meulen et al 2012).This implies that most of the studied biotic events are more or less asynchronous (Lindsay 2001;Van der Meulen et al. 2011, 2012).The treatment for subdivision of biotic events within the evolution of terrestrial mammals differs  Fahlbusch (1964Fahlbusch ( , 1975)); Fahlbusch and Wu (1981); Wu (1982); Ziegler and Fahlbusch (1986); Heissig (1990); Bolliger (1992) conceptually and methodologically from stratigraphic treatment of marine invertebrates (Qiu et al. 2013).The mammal units are primarily biological entities rather than biostratigraphic units that are based on rock sequences.So, the nature of the mammal fossil record generally favours the application of the biochronologic method that lies outside the methods governed by stratigraphic codes (Lindsay and Tedford 1990).It is also important to note that each timescale is based on mammals in different continental landmasses and has its own history of development that reflects the uniqueness of the record and the extent to which faunal succession has been resolved (Rook et al. 2019).
Within the above framework, two major systems, MN -Mammal Neogene and LMA -Land Mammal Ages, have been developed and gradually widely accepted.Although the both systems differ in many respects, their differences -stemming from their different historical developments and purposes for which they were created -have often not been sufficiently taken into account.These misuses and subjective loose interpretations have produced inconsistencies such as various and/or multiple definitions of boundaries or the confusion between local and continental durations of these units.At any rate, the both systems in practical usage are neither in accordance with a proper definition of chronostratigraphic units -neither with the stages, nor with their correlative geochronologic units (Steininger 1999).Historically conditioned definitions and use of ELMA units include not only biological but also non-biological criteria including a combination of chronostratigraphic, bio(/geo)chronological, and biostratigraphic approaches, and are therefore not used in the sense of proper biochronological ages.On the other hand, MN units are time units of biochronology in a pure sense (Qiu et al. 2013).The MN-system as defined by Mein (1975aMein ( , 1975b) ) can never be considered as a biostratigraphical scale, because it is exclusively based on fossil associations and the biological evolution that they reflect (Fahlbusch 1991;Van Dam et al. 2001); in the vast majority of cases they are treated as such.The mammal palaeontologists/stratigraphers working in the Iberian Peninsula and the North Alpine Foreland Molasse area, where terrestrial deposits with rich mammal fossils are widely developed were unsatisfied with the MN system and tended to 'redefine' its units by building up real biostratigraphic biozones based on rock bodies with mammal fossils (for details see Daams et al. 1977;1987;Daams and Freudenthal 1981;De Bruijn et al. 1992;Van der Meulen and Daams 1992;Krijgsman et al. 1994;Bolliger 1997;Heissig 1997;Kälin 1997;Kempf et al. 1997;Daams et al. 1999;Böhme et al. 2002;Van der Meulen et al. 2005;2011, 2012;Van Dam et al. 2006; Van der Meulen and Peláez-Campomanes 2007; Abdul Aziz et al. 2008Aziz et al. , 2010;;Kälin and Kempf 2009;Casanovas-Vilar et al. 2016;Crespo et al. 2021).

Context for biochronologic/biostratigraphic considerations on the localities under study
As stated in the previous chapter, most of the studied biotic events are asynchronous in different parts of Europe (Figure 6).Absolutely crucial to understanding these phenomena and considerations on the age estimations of localities under study are the three best known records from Spain, Switzerland, and Germany.They provide the most complex bio-stratigraphical information on the late Early to early Middle Miocene supported by magnetostratigraphy and radiometric ages.The correlations between Swiss and Bavaria are quite straightforward because the both have an almost identical distribution of taxa and discrepancies can be found particularly in the proposed chronologies (cf.Abdul Aziz et al. 2008, 2010vs Kälin and Kempf 2009;Reichenbacher et al. 2013).However, correlations of central European biozones with Spanish ones are more difficult due to the low number of shared taxa and they can be established particularly based on the proposed chronologies ( Van der Meulen et al. 2011, 2012).
The oldest fauna with Democricetodon as the first modern cricetid (together with the eomyid Ligerimys florancei) in North Alpine Foreland comes from Glovelier (Swiss Molasse Basin), the reference and only locality of the Glovelier Assemblage Zone (= Democricetodon franconicus-Megacricetodon collongensis interval zone sensu Kälin and Kempf 2009), which corresponds to MN4a (Figure 6).In the German Molasse Basin there is no fauna equivalent to that Swiss interval zone, however, this period is very well recorded in localities Petersbuch and Erkertshofen (see Prieto and Rummel 2016 and references therein).In the Aragonian type area, the fauna similar to Glovelier comes from localities of San Roque 4A and 4B (ca 16.99 and 17.0 Ma;Van Dam et al. 2006) belonging to Local Zone A (upper Ramblian), the oldest recognised biozone in that area (Figure 6).The assemblage is characterised by the presence of Ligerimys antiquus, Melissiodon and, although rare, the first record of Democricetodon represented by D. hispanicus.In contrast to Gloverier, L. florancei is not recorded in Spain until the end of Zone B. Depending on the criterion used to define MN 4 (the FO or FCO of Democricetodon), these faunas could be considered as MN3 or MN4.The beginning of the Aragonian (Zone B) is recognised by the FCO of Democricetodon hispanicus and is constrained by the estimated ages of 16.99-16.77Ma (Daams and Freudenthal 1988;Daams et al. 1999;Van der Meulen et al. 2012).This interval, corresponding to the earliest part of the MN4.The mentioned ages indicate that the entry and increase in relative abundance of the genus in the Aragonian area took place in the late Ramblian during chron C5Cr (17.235-16.721 Ma) (Van der Meulen et al. 2011).
The presence of Megacricetodon in combination with the presence of Ligerimys has been used to define local biozones all over Europe.In Germany, it characterises the OSM A biozone, in Switzerland, the Tägernaustrasse biozone, and in Spain, Zone C; all units/zones can roughly be correlated with the MN4b (Figure 6).The OSM A in the Bavarian Molasse Basin is characterised by the FO of Megacricetodon bezianensis (M.cf./aff.collongensis) cooccurring with Democricetodon franconicus.Ligerimys and Melissiodon are still present, but become very rare near the end of the OSM A. The OSM A unit corresponds to the Tägernaustrasse assemblages Zone (= Megacricetodon collongensis-Keramidomys interval zone sensu Kälin and Kempf 2009) of the Swiss Molasse Basin.The presence and abundance of the eomyid Ligerimys ellipticus characterises zone C together with the first appearance of Megacricetodon (M.primitivus) and Eumyarion (Daams and Freudenthal 1988;Daams et al. 1999;Van der Meulen et al. 2012;Crespo et al. 2019).The upper boundaries of these units are generally correlated with the transition of MN4 and MN5.
The beginning of the MN5 is defined mainly on the basis of the LO of Ligerimys and the FO of Keramidomys.The extinction of L. florancei and its replacement by Keramidomys is used as a common criterion to distinguish between MN4 and MN5 faunas in central Europe.This occurs in the North Alpine Foreland around 16.3-16.5Ma (Reichenbacher et al. 2013).However, the biostratigraphic/chronologic use and correlation based on the Ligerimys extinction is somewhat problematic since the last representatives of this genus or its species, are different in central Europe and Spain.The last Spanish representative of Ligerimys is the endemic species L. ellipticus (the LO with an estimated age of 15.97 Ma;Van Dam et al. 2006), characteristic for the local zone C, whereas the last species of the genus in central Europe is L. florancei (cf.Álvarez-Sierra 1987; Daams et al. 1999;Van Dam et al. 2006;Abdul Aziz et al. 2010;Jovells-Vaqué and Casanovas-Vilar 2021).So, if we use the species level instead of the genus one then the upper boundaries Tägernaustrasse biozone and OSM A should be correlated within the lower part of Zone C of Spain (Figure 6).Applying the commonly used criterion -the extinction of the genus Ligerimys then the Spanish faunas of upper Zone Ca to Cb should be placed in MN4, while the approximately coeval Swiss faunas of Overkulm-Sämlen and lower Vermes 1 should be placed in MN5 (Figure 6; Van der Meulen et al. 2012) Kälin and Kempf (2009)), according to Reichenbacher et al. (2013) rather to its lower part.
Development and succession of mammal communities within the following zones (~OSM Swiss zones Vermes 1-Aspitobel 520 m; ~Bavarian zones OSM C + D-E; ~Spain zones Da-Dc-E) corresponding to the rest of MN5 in the both the Swiss and German (Bavarian) Molasse Basins remain nearly identical (cf.Heissig 1997;Abdul Aziz et al. 2008; Figure 6).Minor differences can be observed in the occurrence of Anomalomys minor in the Bavarian biozones OSM C and the occurrence of Cricetodon meini in the Bavarian biozone OSM D (Kälin and Kempf 2009).In contrast, the faunal similarities between central Europe and Spain are very small.Most of the taxa appearing in central Europe with biochronologic value, such as Keramidomys, Megacricetodon, Cricetodon, or Eomyops are in comparison with Spain distinctly diachronous, represented by different species, or absent in one area (cf.Kälin and Kempf 2009;Van der Meulen et al. 2011, 2012;Jovells-Vaqué and Casanovas-Vilar 2021).
The most important Megacricetodon events in this time period are the FO of M. minor (upper OSM E or Aspitobel 520 m in central Europe, transition of zones F/G1 in Spain) and the FO of M. gersii (OSM F/F? transition or Oeschgraben in central Europe, transition of zones E/F in Spain).The FO of M. minor forms a biggest discrepancy among sequences of first entries of biochronologically important taxa (see Van der Meulen et al. 2011 for detail).In Spain, as opposed to Switzerland and Germany, the FO of M. minor occurs after that of Cricetodon and M. gersii.Kälin and Kempf (2009) considering the presence/absence of a small-sized Megacricetodon cf.minor as an important criterion and distinguished two units -Megacricetodon aff.bavaricus-Megacricetodon cf.minor interval zone (reference locality Vermes 1) and Megacricetodon cf.minor-Megacricetodon aff.bavaricus overlap zone (reference locality Tobel Hombrechtikon); analogous to the OSM C + D of the Bavarian Molasse Basin (Prieto and Rummel 2016).However, there is no well-defined sized-based criterion allowing a clear distinction between M. bavaricus and M. aff.bavaricus, and it remains more or less arbitrary.So, the Swiss M. aff.bavaricus-M.cf.minor interval zone is not convincingly recognised in Germany (Prieto and Rummel 2016).In view by a molar size comparable to Megacricetodon minor is not achieved in localities preceding Cricetodon occurrences, i.e.OSM E (Prieto and Rummel 2016).
Another important species, Megacricetodon lappi, is present in central and western Europe, but has never been found in Spain.Its short occurrence in the Swiss OSM, reliably estimated from approximately 15.0 to 14.9 Ma, defines the Megacricetodon lappi taxon range zone; with the reference locality Aspitobel 520 m (Kälin and Kempf 2009).
Cricetodon meini is the first Cricetodontini to migrate to eastern and central Europe at about 15 Ma (Boon 1991;Heissig 1997;Daxner-Höck 2003;Hír 2013).The oldest representative record of this species in Bavarian OSM has been recovered from Burg-Balzhausen.The very small size of this taxon leads Seehuber (2009) to correlate the locality to the upper OSM E. The FO of Cricetodon (C.aff.aureus) in Switzerland comes from Uzwil-Nutzenbuech and is correlated to chron C5Bn.1n with an estimated age of 14.9 Ma (Kälin and Kempf 2009).It slightly postdates the FO of the genus in Bavaria (Abdul Aziz et al. 2010).The FO of the genus in Spain is C. soriae in Zone E in the locality of Las Umbrias 11 correlated to chron C5ACn with an estimated age of 14.06 Ma (Van der Meulen et al. 2011).However, it is worth noting that the first Cricetodon in Spain belongs to a different lineage (De Bruijn et al. 1993;López-Guerrero et al. 2013, 2014).A rapid increase in size of C. meini leading to forms close to C. aureus do not currently allow dating OSM F localities relative to each other (Heissig 1997;Hír 2013;Prieto and Rummel 2016;Hír and Venczel 2018;Prieto et al. 2022).
Eomyops first occurrence in Switzerland is at the base of Aspitobel 520 m (15.0 Ma), while in Spain it is not recorded until Zone G1 (13.5 Ma).The FO of Eomyops is the last considered eomyid event in both areas ( Van der Meulen et al. 2011).
Summed up.Considerable asynchronies among the major rodent events used for the zonation are evident between central Europe and Spanish bioprovince.The numerical ages given to the MN boundaries in central Europe are about 0.5-1.0Myr older than in Spain (Figure 6).The diachrony of the FO/FCO of modern cricetids is probably due to their east-west immigration, while the diachronous extinction of Ligerimys was probably due to local environmental and climatic changes toward a more open and less humid environment (Kälin and Kempf 2009).
Within the very well-studied and well-documented profiles of the Bavarian Molasse Basin (see e.g.Piller et al. 2007;Reichenbacher et al. 2013;Pipperr and Reichenbacher 2017;Sant et al. 2017a, and references therein) this type of fauna/facies occurs there in the Grimmelfingen and Kirchberg Fms.The fossil-rich marls of Kirchberg Fm transgressively overlie the Grimmelfingen Fm, and most probably represent a short-term Karpatian transgression of the Swiss Molasse Sea (Kiderlen 1931;Lemcke 1988;Reichenbacher et al. 1998Reichenbacher et al. , 2013)).The brackish transgression was possibly associated with a global rise in sea level (the beginning of the global third-order sea-level cycle Bur 4; Piller et al. 2007).The co-occurrence of Rzehakia, Mytilopsis, and cardiids in the Grimmelfingen Fm suggests a predominantly upper Ottnangian age (Pippèrr and Reichenbacher 2017).Lemcke (1988) assumed that the uppermost sections of the Rzehakia-beds are contemporaneous with the Kirchberg Fm.This view is supported by recent data of Pipperr and Reichenbacher (2017).Nevertheless, Rzehakia occurs only in the Grimmelfingen Fm, but is absent from the Kirchberg Fm (Pippèrr and Reichenbacher 2017).At any rate, the gastropod and bivalve assemblages of the brackish Kirchberg Fm are similar to, but not entirely congruent with those of the Rzehakia-beds in the Central Paratethys because endemic species occur in each sub-basin (Lower Bavaria: e.g.Schlickum 1971;Reichenbacher 1993;Austria: Mandic and Ćorić 2007;Czech Republic: Čtyroký 1972).
As regards mammal fossil record, the basal layer of the Grimmelfingen Fm are indicative for the mammal unit MN4a (Heizmann 1984;Reichenbacher et al. 1998;Sach and Heizmann 2001).The uppermost part of the Grimmelfingen Fm has yielded a rodent tooth assigned to the genus Megacricetodon, and members of this genus first appear in the NAFB in mammal unit MN4b (Heissig 1997;Prieto 2007;Kälin and Kempf 2009).The sediments of the Kirchberg Fm (Pfaffenhofen T1) contain a small-mammal tooth that was identified as Megacricetodon aff.collongensis, an index species for the mammal unit MN4b in the NAFB (Pippèrr and Reichenbacher 2017).
The Grimmelfingen Fm has a reverse (lower part) -normalreverse (uppermost part) polarity pattern.The lower Kirchberg Fm is consistently reverse and the upper Kirchberg Fm has reverse to normal polarity (Reichenbacher et al. 2013;Pipperr and Reichenbacher 2017).
Thus, all the above suggests that the Grimmelfingen Fm can be correlated with the upper Ottnangian (ca chron C5Dn) to the lowermost Carpathian and the Kirchberg Fm with the lower Carpathian (ca chron C5Cr) (Pippèrr and Reichenbacher 2017).The boundary between the units is thus situated within the lower Carpathian (ca early MN4b) at around 17 Ma (Figure 6).
However, it should be noted that the unequivocal stratigraphic equivalency of the different members of the Rzehakia-beds throughout the Central Paratethys (cf.Čtyroký et al. 1973a, 1973b) has not been fully proven yet (see Reichenbacher et al. 2013 for details).It is possible that facies characterised by Rzehakia were developed locally and differences between Bavaria, Austria, and Moravia (Piller et al. 2007) might indicate a further disintegration of the Paratethys into several isolated brackish lakes (Mandic and Coric 2007).At any rate, existing evidence suggests (cf.Cicha et al. 1972;Lemcke (1988); Pippèrr and Reichenbacher 2017) that the Ořechov fauna was most likely overlain by sediments of an age corresponding to the Kirchberg Fm.
Thus summing up, all of the above suggests that Ořechov is slightly older than the localities of the Bavarian Molasse Basin (Figure 6), from which the oldest occurrences of the genus Megacricetodon, correlated with the lower part of OSM A, are documented (see Prieto and Rummel (2016) for details).The teeth of Megacricetodon in this locality are also smaller than those of the oldest Megacricetodon of the NAFB recorded from Langenau 1 (Figure 7); the locality correlated to the base of the Kirchberg Fm; cf.Reichenbacher et al. 1998.Ořechov therefore most likely corresponds at least to the late MN4a and marks the FO of Megacricetodon in central Europe.As for other localities, faunal composition in Dolnice, Ořechov, and Mokrá-Turtle is very similar and together with the phenotypic closeness of available index taxa, it suggests that the localities are very close in age.The smaller size of the teeth of Megacricetodon from Ořechov (Figure 7) could indicate a greater relative age in relation to other localities (Figure 6).
The MN4/MN5 transition is characterised in the Northern Alpine Foreland Basin (i.e.ca OSM A/B; Tägernau-strasse /Oberkulm-Sämlen transitions); (Figure 6)) by the first appearance datum of Megacricetodon bavaricus (in the size of the type population from Langenmoosen, Fahlbusch 1964) and the FCO of Keramidomys thaleri, whereas Ligerimys florancei and Melissiodon dominans have been already disappeared during the late MN4.The absence of Melissiodon and Ligerimys and at the same time the common occurrence of Keramidomys and Megacricetodon in Františkovy Lázně thus clearly indicates belonging the locality to MN5.Keramidomys is rare in Františkovy Lázně (Cicha et al. 1972).This FO in the Cheb basin probably shortly predates the FO in the North Alpine Foreland recorded in Langenmoosen that is correlated with the OSM B (Abdul Aziz et al. 2010).On the basis of the faunas occurring at these two localities, Cicha et al. (1972) define the Faunal Group Františkovy Lázně -Langenmoosen, following the older one Ořechov -La Romieu.
The fossil record from Strakonice (the lower part of Mydlovary Fm) is quite scanty, nevertheless the presence of Karydomys wigharti and a large cricetid from the proximity of Cricetodon (referred by Fejfar (1990) as C. meini or by Fejfar (1974) as C. cf.meini) suggests a much younger age than Františkovy Lázně (Figure 6).Karydomys wigharti predominantly occurs at localities that are correlated with the late MN5 (Mörs and Kalthoff 2004;Prieto and Scholtz 2013).In the Swiss part of the NAFB the taxon is recorded from the Megacricetodon lappi -Democricetodon gracilis and D. gracilis -M.gersii interval zones, from the Bavarian part from OSM unit F (Prieto and Scholz 2013).In the Swiss Molasse Basin the remains of Karydomys are most often found together with Cricetodon (e.g. in Rümikon, Wielzholz, Uzwil-Nutzenbuech; Bolliger (2000), Kälin and Kempf (2009)).A well-proven representative sample of Cricetodon meini from the Bavarian Molasse Basin is recorded from the late OSM E -ca the Swiss Megacricetodon lappi taxon range zone (Prieto and Scholz 2013;Prieto and Rummel 2016).The FO of Cricetodon (C.aff.aureus) in Switzerland is in Uzwil-Nutzenbuech (Van der Meulen et al. 2011).As already mentioned, this first occurrence is correlated to chron C5Bn.1n, with an estimated age of 14.9 Ma (Kälin and Kempf 2009), and slightly postdates the FO in Bavaria (Abdul Aziz et al. 2010).However, it should be mentioned that the fossil material from Strakonice referred to Cricetodon is represented by only two teeth (Fejfar 1990), thus its generic attribution is not entirely unambiguous.Daxner-Höck (2003) or Prieto et al. (2010) even question its belonging to this genus.

Megacricetodon in the Early-Middle Miocene of Europe: dispersal and palaeobiogeography
In general, we agree with proposals on dispersal scenarios for the early Megacricetodon taxa in Europe provided by Oliver and Peláez-Campomanes (2016).However, based on new data about the Czech record, taking into account palaeogeographic changes during the Early-Middle Miocene, we provide here some precisations and/or corrections in this matter (see Figure 6).
We are still far from understanding the details of migration and/ or speciation events leading to dispersal of Megacricetodon in Europe.But it is apparent that its spread at least into central Europe was relatively fast.The fossil record and its palaeogeographic context imply that distribution of Megacricetodon across Europe was the result of rather gradual, more or less continuous, dispersal than several separate migration waves from Asia, as tentatively hypothesized by Oliver and Pelaez-Campomanes (2016).Within the framework of this gradual dispersal rather individual regionally limited migration/speciation events occurred, which were modulated by the changing landmasses distribution (opening of local corridors or, on the contrary, the isolation of some areas) eventually also by the changing palaeoenvironment.At any rate, it cannot be excluded that the appearance of some new taxa/clades was the result of a migration wave from Asia but this seems less likely, as there is no fossil evidence or other indications of a direct route north of the Kotsakhurian Sea (cf.Van den Hoek Ostende et al. 2015: fig. 3a).
The MN3/4 faunal turnover in Europe, characterised by the end of 'Cricetid vacuum' and followed by a sudden onset of modern cricetids, was undoubtedly caused by the new land layout in the Middle east area.Its impact on the terrestrial ecosystems was a driving factor for Asian taxa to invade Europe via a newly formed landbridge.
The oldest European locality well preserving representatives of an eastern migration wave is Aliveri (early MN4, Greece; Van den Hoek Ostende et al. 2015;Oliver and Pelaez-Campomanes 2016).The locality documents the first migration/speciation event of the genus in Europe (Figure 6).Its fauna is formed by a strong Anatolian influence with cricetids such as Cricetodon, Democricetodon, Megacricetodon, Mirrabella and Eumyarion, and spalacid as Heramys, but also with a lot of European immigrants such as the insectivores Heterosorex, Plesiodimylus, and Myxomygale, and the eomyid Pseudotheridomys.The geographic location of Aliveri during the late Early Miocene cannot be clearly identified ( Van den Hoek Ostende et al. 2015).However, taking into account the opening of the Aegean together with relatively rapid westward movement of the Aegean-Anatolian block in the late Neogene (Ten Veen and Kleinspehn 2002), Aliveri must have originally been located several hundred kilometers to the NE of its present location ( Van den Hoek Ostende et al. 2015); thus, very close to the presumed main dispersal land bridge from Asia Minor.In any case, in our view the Aliveri surroundings were more or less isolated, or may have gradually become isolated, and in some periods peninsular to insular conditions may have prevailed there.
The presence of typical European elements in Aliveri (see Van den Hoek Ostende et al. 2015) shows that faunal interchanges with central European areas already existed at the time of deposition of the site, i.e. shortly after the opening the land bridge from Asia Minor.It can therefore be assumed that Megacricetodon after reaching Aliveri was gradually spreading further towards central Europe.The main geographical barrier at that time was the Central Paratethys, the semi-enclosed basin asynchronously communicating with the Western Mediterranean via several active seaways.This configuration implies two most likely hypotheses/scenarios on the potential migration routes to central Europe/Bohemian Massif: (1) northernly, around the just-forming Alpine Chain through Moesia and the Volhynian High; and/or (2) northwesternly, through the Dinarian and Transdanubian Highs.
The first hypothetical route is clearly apparent from the palaeogeographic configuration of landmasses at the time of late Early Miocene and this one was commonly assumed by many authors (cf.e.g.Koufos et al. 2005;den Hoek Ostende Lw et al. 2015;Oliver and Peláez-Campomanes 2016).It appears to be completely passable during the late Ottnagian -Karpatian (i.e.ca MN4-5).The second hypothetical route is less apparent, but on the other hand, it is supported by the coincidence of palaeogeographic events with the first occurrences of main cricetid taxa in central Europe.This corridor has probably never been fully passable in its entirety; first being a barrier the Alpine gateway (active from the Eggenburgian to late Ottnangian; Sant et al. 2017b;Kováč et al. 2018), later the Dinaride gateway (active from the late Ottnangian to early/late Badenian transition; Sant et al. 2017b;Kováč et al. 2018).During the Ottnangian, the Alpine gateway was gradually closing and Central Paratethys began to communicate with the Mediterranean through the just opening Dinaride gateway; so the route to Bohemian High was gradually, in several dispersal waves, passable.Locally isolated areas providing suitable conditions for an allopatric speciation may have been formed there.Although the second scenario seems to be more likely and is preferred here, given the present state of knowledge, neither one can be conclusively supported or rejected.
In any case, the appearance of modern cricetids in central Europe correlates roughly with the closure of the Alpine gateway that can be dated around 17.5 Ma (Sant et al. 2017a;Kováč et al. 2018) (Figure 6), i.e. it corresponds approximately to the transition of the OMM to the OSM.The first occurrences of Megacricetodon grueneri sp.nov.from the territory of Czech Republic are recorded near the MN4a/b boundary (Figure 6).These can be considered as a consequence of the here postulated second migration/speciation event.
Despite the differences in size between Megacricetodon grueneri sp.nov.and M. hellenicus, these two species show several features in common, such as: (a) a dental pattern with longer and slender crests that give them a lophodont aspect; (b) the entoloph of the M2 is long and angular with the posterior arm of the protocone better developed than the anterior arm of the hypocone; (c) the five main cusps of the m1 have no similar height, the anteroconid is notably lower than the other four cusps (see Figure 5).These characters are regarded here as synapomorphies suggesting a close relationship of the both species.
The relationship of Megacricetodon grueneri sp.nov. to the taxa of the Bavarian and Swiss Molasse Basins (= bavaricus-group) is not yet clear.Their first appearances were almost synchronous.However, the above data suggest that the earliest Czech records are slightly older than the earliest ones from the Bavarian/Swisse Molasse Basin (cf.Langenau 1, etc.).Considering (i) the distinctly different tooth morphology of Megacricetodon grueneri sp.nov.from taxa of the Bavarian-Swisse Molasse and (ii) their different dimensional gradient of tooth sizes over time (Figure 7), which seems not to be as pronounced in the case of the Ořechov/Dolnice-Františkovy Lázně lineage as it is in the case of the bavaricus-group; it seems most likely that the both groups/taxa evolved independently and were part of distinct lineages.This hypothesis (see Figure 6) is supported in this paper.At any rate, the bavaricusgroup, as proposed by Oliver and Peláez-Campomanes (2013), is credibly proven and is most probably the result of another speciation/migration event.The early Megacricetodon material from the Swiss and German Molasse Basins was assigned to Megacricetodon aff.collongensis (Heissing 1990(Heissing , 1997;;Kälin et al. 2001;Kälin and Kempf 2009;Prieto and Rummel 2009).However, we agree with Oliver (2015), who assigned this species to M. bezianensis.This form dispersed and evolved throughout the central and southwestern Europe.While in Switzerland and Germany occurred M. bezianensis during middle MN4, evolving to M. bavaricus in MN5 (OSM B), M. aunayi in OSM C + D (previously assigned to M. aff.bavaricus) (Heissig 1990) and finally M. lappi (OSM E) (Abdul Aziz et al. 2008;Kälin andKempf 2009, 2010).In France, Megacricetodon bezianensis keep relatively without variation during MN4 and MN5.Otherwise, in southwestern Europe, this group evolved to an endemic form in the middle MN5 (local zone Db), Megacricetodon vandermeuleni (Oliver & Peláez-Campomanes 2013).
The final spread to southwestern Europe (France and Spain) was probably linked with the closure of Rhine Graben taken place during the first half of the Karpatian that made its possible way to the Iberian Peninsula.This migration/speciation event corresponds in particular to the Megacricetodon primitivus lineage.This smallsized species appeared during MN4 in Southwestern Europe.The first occurrence was in southern France (Oliver and Pelaez-Campomanes 2016) and reached the Iberian Peninsula at latest MN4 (Ginsburg and Bulot 2000;Oliver and Peláez-Campomanes 2014).Also, the taxa of bavaricus-group were gradually spreading to the west, reaching the Iberian Peninsula in the middle of MN 5 and evolving to M. vandermeuleni (Oliver and Peláez-Campomanes 2013).

Figure 1 .
Figure 1.Geographical localization and age estimation of the localities under study.

Figure 2 .
Figure 2. Megacricetodon grueneri sp.nov.from type locality Dolnice 3. a, NM-Pv 10463 M1 right; b, NM-Pv 10464 M2 left; c, holotype NM-Pv 10462 m1 right; d, NM-Pv 10465 m2 left.All of the teeth are at the same magnification and are figured as left specimens.Right side specimens are underlined.Scale bar equals 1 mm.
hellenicus differs by having more often a labial spur of the anterolophule in the M1, double metalophulid of the m1, and larger mesolophid of the m2.(3) Among similarly sized species, in comparison to Megacricetodon grueneri sp.nov.: • M. aguilari, M. bezianensis, and M. vandermeuleni differ by the robust and swollen cups and the 'crescent' shape anteroconid; • M. andrewsi differs by the never split anteroconid and the longer mesolophids in m1 and m2; • M. beijiangensis differs by having longer mesolophs in M1 and M2 and double protolophule of the M2; • M. crisiensis differs by having a strong divided anteroconid of the m1 and absent mesolophid of the m2; • M. bourgeoisi differs by having M2 with longer mesoloph and anterior metalophule, and m2 with longer mesolophid; • M. gersii differs by having in the M1 the anterocone more subdivided and the furrow usually reaching the crown basis, the anterolophule usually connects to the middle of the two cones forming the anterocone, and the mesoloph is shorter; in the m1 the anteroconid usually subdivided; • M. lalai differs by having shorter mesolophs in the M1, more often connection of the ectoloph with the mesoloph in the M2; more often labial spur of the anterolophulid in the m1; more variable height of the lingual anterolophid and longer mesolophid in the m2; • M. similis differs by having labial spur of the anterolophule of the M1, double or almost double protolophule of the M1 and M2, larger mesoloph of the M2, slightly subdivided anteroconid of the m1 and shorter or absent mesolophid of the m1 and m2.DescriptionHolotype -m1.(Figures2c, 5b; Appendix 1: Tables

Figure 6 .
Figure 6.Bio-chronologic/stratigraphic and biogeographic distribution of early Megacricetodon species in Europe with indication of the proposed migration/speciation events among different parts of Europe.The magnetostratigraphic scale and ages (Ma) are after Hilgen et al. (2012) -black indicate normal magnetic polarity chrons, white indicate inversed magnetic polarity chrons; local bio-stratigraphy/chronology of Spain including local zones and MN units is after Van der Meulen et al. (2012); local biostratigraphy/chronology of Central Europe including Bavarian/Swiss local units/reference faunas, Paratethis stages, and MN units is after Reichenbacher et al. (2013); distribution of Magacricetodon taxa is modified from Oliver and Pelaez-Campomanes (2016), arrows indicate the sense of migration.For locality acronyms see section Abbreviations of the main text.

Figure 7 .
Figure 7. Length-width comparison of the first lower molar (mean values) of Megacricetodon from localities of the Czech Republic with those of the Northern Alpine Foreland Basin (NAFB).Explanatory notes: red symbols represent localities of the Czech Republic, blue symbols localities of the S-German (Bavarian) Molasse Basin, and green symbols localities of the Swiss Molasse Basin; the filling of symbols refers to the age of localities -closed to the late MN4, open to the middle MN5, transient to the early MN5; colored bars bounded by 'L' and 'K' show the relative positions of Ligerimys/Keramidomys transitions in the each area to sizes (length) of Megacricetodon m1, each bar is defined by proven LO of Ligerimys and FO Keramidomys in the concerned area (note that bars do not reflect relative age or stratigraphic spans); localities of the NAFB are related to the Bavarian and Swiss local zones (in the color corresponding to the concerned areas).Data were taken fromFahlbusch (1964Fahlbusch ( , 1975));Fahlbusch and Wu (1981);Wu (1982);Ziegler and Fahlbusch (1986);Heissig (1990);Bolliger (1992);Ziegler (1995);Sach and Heizmann (2001);Reichenbacher et al. (2005Reichenbacher et al. ( , 2013)); Böttcher et al. (2009); Wessels and Reumer (2009); Abdul Aziz et al. (2010); Prieto and Rummel (2016).
Figure 7. Length-width comparison of the first lower molar (mean values) of Megacricetodon from localities of the Czech Republic with those of the Northern Alpine Foreland Basin (NAFB).Explanatory notes: red symbols represent localities of the Czech Republic, blue symbols localities of the S-German (Bavarian) Molasse Basin, and green symbols localities of the Swiss Molasse Basin; the filling of symbols refers to the age of localities -closed to the late MN4, open to the middle MN5, transient to the early MN5; colored bars bounded by 'L' and 'K' show the relative positions of Ligerimys/Keramidomys transitions in the each area to sizes (length) of Megacricetodon m1, each bar is defined by proven LO of Ligerimys and FO Keramidomys in the concerned area (note that bars do not reflect relative age or stratigraphic spans); localities of the NAFB are related to the Bavarian and Swiss local zones (in the color corresponding to the concerned areas).Data were taken fromFahlbusch (1964Fahlbusch ( , 1975));Fahlbusch and Wu (1981);Wu (1982);Ziegler and Fahlbusch (1986);Heissig (1990);Bolliger (1992);Ziegler (1995);Sach and Heizmann (2001);Reichenbacher et al. (2005Reichenbacher et al. ( , 2013)); Böttcher et al. (2009); Wessels and Reumer (2009); Abdul Aziz et al. (2010); Prieto and Rummel (2016).

Table 1 .
Descriptive statistics of the upper and lower molars of Megacricetodon grueneri sp.nov.from Early to Middle Miocene localities of the Czech Republic.