INTRODUCTION

The Late Pliocene to Early Pleistocene (Piacenzian to Gelasian) adaptive radiation of Mimomys is one of the most significant processes in small mammal communities in the middle latitudes of Eurasia. The radiation produced several lineages, one of which is the (sub)genus Cromeromys. The concept of Cromeromys as a separate genus of Pliocene and Pleistocene mimomyoid voles was established by Zazhigin in 1980. According to Zazhigin, Cromeromys includes forms with the characteristic features of external cement in the molar reentrants, the anteroconid of the lower m1 always without an enamel islet, and the posterolingual reentrant of the upper M3 always deep, never forming an enamel islet. The taxon was based on scarce remains of a small Mimomys species (Mimomys irtyshensis) from Late Pliocene deposits located in southern Siberia. Zazhigin (1980) also included in the genus a small Mimomys species from the Early Pleistocene (Gelasian) crags of East Anglia that was referred to as Mimomys newtoni and eventually re-described as Mimomys tigliensis (see Mayhew and Stuart 1986; Tesakov 1998). Zazhigin (1980) also included a large species of the Mimomys savini group (see discussion by Neraudeau et al. 1995) in the genus.Zazhigin (1980) later listed Cromeromys irtyshensis and C. cf. irtyshensis from the Early Pleistocene (early Gelasian through 'Calabrian' equivalents) levels of the Arctic Krestovka locality in northeastern Arctic Siberia (Zazhigin 1997). It is these forms that most closely correspond to the original definition of Cromeromys, showing an early reduction of the enamel islet in the m1, and with the M3 having a hook-like posterior lobe without an enamel islet. Another stable feature of the m1 is the strong development of the Mimomys-ridge (= Mimomys Kante), a peculiar indentation of the fourth dental triangle that is persistently present in many Pliocene archaic rhizodont voles. This morphology is also known in several Early Pleistocene (Irvingtonian = 'Calabrian') molars from the United States assigned to three different species: Mimomys monohani, M. virginianus, and M. dakotaensis (L. Martin 1972; Repenning and Grady 1988; R. Martin 1989; Bell and Barnosky 2000). Repenning (Repenning and Grady 1988; Repenning 2003) considered Cromeromys to be a subgenus of Mimomys that includes all these forms.

The remarkable Zuurland drilling project of Mr. L. Hordijk (Brielle, The Netherlands) in Western Europe produced a rich superposed sequence of arvicoline faunas ranging from Early Pleistocene (Gelasian) to Holocene (Hordijk 1988; van Kolfschoten 1988). A form morphologically similar to Cromeromys was discovered in materials from Early Pleistocene (Gelasian, Late Villanyian, MN17) levels of the Zuurland sequence. Van Kolfschoten and Tesakov (1998) described this vole as Mimomys hordijki based on limited material available at that time. Although the general similarity to Mimomys (Cromeromys) species was obvious to van Kolfschoten and Tesakov (1998), they refrained from attributing the new species to that subgenus because no M3 were known at that time.

C. Repenning (personal commun., 1999) suggested in his comments that the new European vole is conspecific with the late Irvingtonian species from Hamilton Cave, USA and stated, "..... but what makes less sense is to say 'we do not have the last upper molar and therefore are not sure that the specimens from Mr. Hordijk's borehole are not Mimomys virginianus, so we will give it a new name' How did you two bright men miss seeing the error of this action?"

New material from the Zuurland boreholes fortunately included third upper molars, confirmed the attribution of M. hordijki to Mimomys (Cromeromys), and showed that this European vole is a distinct species probably ancestral to North American forms. It reconciles us with our conscience and, furthermore, we hope that Repenning would have been glad to see this contribution. We describe the new M. hordijki material and discuss the possible relation between Mimomys (Cromeromys) and the modern Nearctic sagebrush vole, genus Lemmiscus.

The Subgenus Cromeromys

Study of the Cromeromys record of Eurasia and North America suggests that Cromeromys in its original definition includes different lineages. According to Zazhigin (1980), it is the Cromeromys group of Mimomys-like or mimomyoid voles that gave rise to the modern Microtus and Arvicola, and this affinity is marked in the dental morphology by complex third upper molars and absence of enamel islets in all members of Cromeromys group. In Mimomys, in contrast to the former group, the m1 and M3 have enamel islets. However, a number of the forms included in Cromeromys do not comply with the original diagnosis (see also Repenning and Grady 1988). The common occurrence of an enamel islet in the anteroconid of the m1 of juvenile Mimomys savini (and even its modern descendant Arvicola), and the posterior enamel islet normally occurring on the M3 of Mimomys tigliensis, and, possibly (personal observation) even in the type species, indicates that it is likely that all forms originally included in Cromeromys represent different lineages of the genus Mimomys. The original definition of the (sub)genus is not accurate enough. Nevertheless, we agree with Repenning and Grady (1988) that the name is useful to denote Eurasian and North American mimomyoid voles with a well developed Mimomys-ridge and complex third upper molars, and its use considered here as a subgenus of the genus Mimomys. The authors also agree with R. Martin (2003) that, apart from Microtus s.l., this particular clade of Mimomys (Cromeromys) represents the only known group of the Palearctic Mimomys radiation to ever reach North America. Multiple arvicolines with rooted dentition of Pliocene Blancan age (e.g., Ogmodontomys, Ophiomys) represent native North American vole lineages that are distinct from the Mimomys lineage sensu stricto. Mimomys s.str. includes only post Early Pliocene forms sharing the character syndrome of Mimomys pliocaenicus Forsyth Major 1902, with the important role of external cement and pachyknem Schmelzmuster.