A new genus and species of arvicolid rodent (Mammalia) from the early Pleistocene of Spain

In this paper, a new genus and species of arvicolid rodent is described from the late early Pleistocene levels of the sections of Fuente Nueva 3 (Guadix-Baza Basin, Granada, southern Iberian Pen-insula)


INTRODUCTION
A main event in the early Pleistocene evolution of the Palearctic rodent faunas was the emergence and spread of arhizodont voles (arvicolines with rootless molars or superhypsodont; Martin 1993), most of them included in the genus Allophaiomys Kormos, 1933(Van der Meulen 1973Rabeder 1981;Agustí 1991). In Europe, these early arhizodont voles coexisted with voles with rooted molars of the genus Mimomys Forsyth Major, 1902 approximately until the early-middle Pleistocene transition (Rabeder 1981;Laplana & Cuenca-Bescós 2000). While the replacement of Mimomys by Allophaiomys appears as a clear event in central and eastern Europe, the situation seems to have been more complex in southern Europe, especially in the Iberian Peninsula. In this area, evidence from southern Spain demonstrated a local evolution of endemic Mimomys populations towards root loss. This was the case of Mimomys oswaldoreigi Agustí, Castillo & Galobart, 1993, present at the sites of Barranco de los Conejos (Guadix-Baza Basin) and Gilena 2 (Agustí et al. 1993a), as well as of Orcemys giberti Martin, Tesakov, Agustí & Johnston, 2018, present in the Guadix-Baza Basin at the sites of Barranco de los Conejos and Barranco del Paso (Agustí et al. 2013;Martin et al. 2018). Moreover, in the Guadix-Baza Basin (Barranco de los Conejos), the first evidence of allochthonous arhizodont species cannot be assigned to Allophaiomys but rather to Tibericola Koenigswald, Fejfar & Tchernov, 1992, a genus of eastern Mediterranean affinities also present in Turkey and Israel (Agustí 1991;Agustí et al. 2013).
The first occurrence of true representatives of Allophaiomys in the Guadix-Baza Basin is recorded at the sites of Venta Micena, Fuente Nueva 2 and Orce 7 (Allophaiomys ruffoi Zone; Agustí et al. 2010aAgustí et al. , 2015a, where Orcemys giberti and Mimomys oswaldoreigi are absent. However, it seems that some populations close to Mimomys oswaldoreigi persisted during the time-interval represented by the Allophaiomys ruffoi Zone, since a fully arhizodont species displaying affinities with Mimomys oswaldoreigi is present at the late early Pleistocene sections of Fuente Nueva 3 (FN 3) (Guadix-Baza Basin, c. 1.4-1.2 Ma; Duval et al. 2012;Lozano-Fernández et al. 2015) Piñero et al. 2020Piñero et al. , 2022. This new arvicolid has been variously cited as Mimomys sp. (Sánchez-Bandera et al. 2020) or Allophaiomys sp. (Agustí et al. 2010b;Piñero et al. 2015Piñero et al. , 2020, sharing features common to both genera. However, since it cannot be securely allocated to Allophaiomys or Mimomys, we propose the recognition of the new genus Manchenomys and the new species Ma. orcensis, for some arvicolid specimens from Fuente Nueva 3 and Quibas. We also include within the new genus Manchenomys the species Mimomys oswaldoreigi, closely related to the new species and sharing characters that preclude its inclusion in Mimomys or Allophaiomys. Molar enamel is predominantly negatively differentiated (Mimomys-type). The third upper molar (M3) is also simple, with a short posterior cap. Manchenomys orcensis n. sp. is described from Fuente Nueva 3, and Mimomys oswaldoreigi Agustí, Castillo & Galobart, 1993 from Gilena 2 and Barranco de los Conejos is recombined as Manchenomys oswaldoreigi n. comb. The chronostratigraphic range of Manchenomys n. gen. covers the upper Matuyama geomagnetic chron, between 1.8 Ma (post-Olduvai subchron) and 0.99 Ma (Jaramillo subchron). Manchenomys n. gen. was possibly derived from a local population of Mimomys tornensis Janossy & Meulen, 1975, an arvicolid species present in older early Pleistocene levels of Spain.
A new genus and species of arvicolid rodent (Mammalia) from the early Pleistocene of Spain

Site and inStitutional
The material included in this study comes from the sections of Fuente Nueva 3 (levels FN 3-3, FN 3-4, FN 3-5 and FN 3-6; Guadix-Baza Basin, Granada, Spain) and Quibas (levels QS-1, QS-3, QS-4, QC 4-5 and QG-1; Quibas karstic complex, Murcia, Spain). The material from the Fuente Nueva 3 section includes the 22 first lower molars (m1) documented here and ten third upper molars (M3). The material from the Quibas section is documented by 28 first lower molars (m1) and 18 third upper molars (M3). The distribution of molars in each level of the two sections is documented in Tables 1 and 2. This material is currently housed at the Institut de Paleoecologia Humana i Evolució Social (IPHES-CERCA) in Tarragona (Spain) and final depòsit is going to be at the Museo Arqueológico de Granada (Fuente Nueva 3) and the Museo Arqueológico de Murcia (Quibas).
The nomenclature used for the description of the diagnostic molars, i.e., m1 and M3, follows Van der Meulen (1973) (Fig. 1). The linea sinuosa is defined according to Rabeder (1981). Enamel differentiation is defined as nega-  tive (Mimomys-type), undifferentiated, or positive (Microtustype), according to Martin & Tesakov (1998). Length (L) and width (W) for the m1 have been measured according to Van der Meulen (1973), as well as the standard arvicolid quantities a, b and c (Fig. 1B). Parameters A/L (= a/L × 100), B/W (= b/W × 100) and C/W (= c/W × 100) were calculated according to Van der Meulen (1973). All measurements are expressed in millimetres and were taken with the software DinoCapture 2.0, using photographs from the Digital Microscope AM4115TL Dino-Lite Edge. Some molars represent micrographs taken with Environmental Scanning Electron Microscopy ( Agustí, Castillo & Galobart, 1993, the latter species was formerly assigned to Mimomys. diagnoSiS. -Medium-sized arvicolid with simple dental pattern. Enamel-islet is not present on the m1. Re-entrant angles are filled by abundant cement. In the lower molars the enamel differentiation is negative (Mimomys-type), that is, thicker at the posterior edges, although in some specimens it is undifferentiated. Teeth do not develop roots, with the occasional exception of some m3. However, closure of the enamel at the base of the crown can be observed in some molars, presumably from older individuals. The A/L index ranges between 36 and 40. The B/W index ranges between 25 and 35. The C/W index ranges between 20 and 26. The M3 is simple, with a shallow LRA3 and absence of LRA4.
differential diagnoSiS. -Manchenomys n. gen. presents a typical Mimomys m1 occlusal pattern, with a short, rounded anteroconid, without BSA4 or LSA5. The enamel is also either negatively differentiated or undifferentiated. However, Manchenomys differs from other contemporaneous Mimomys species by absence of roots on all the molars, with the occasional exception of the m3, in which they can appear at a late stage of development. It also differs from most Allophaiomys species in the relatively short anteroconid complex (lower values of A/L index), with the exception of Allophaiomys deucalion (Kretzoi, 1969). The M3 of Manchenomys n. gen. shows a morphology simpler than any Allophaiomys species except A. deucalion, with a shallow LRA3 and absence of LRA4.  (Table 1; Fig. 5A) The AC2 is also more isolated with respect to T4-T5 (lower B/W values; Table 3; Fig. 5B). Roots have not been observed in FN-3 specimens, although a closing of the sinuous line at the base of the crown is present in some specimens. All Allophaiomys species aside from Allophaiomys deucalion differ from Manchenomys orcensis n. gen., n. sp. in their relatively longer  Table 3). Manchenomys orcensis n. gen., n. sp. also differs from A. deucalion in its larger size (L, W) and its more isolated AC2 with respect to T4-T5 (lower B/W values, Table 3). All examples of M3 of Manchenomys orcensis n. gen., n. sp. present a simpler morphology than most Allophaiomys species, with a very shallow LRA3 and absence of LRA4. Only Allophaiomys deucalion presents an M3 with a comparable morphology (Van der Meulen 1974: fig. 3g). However, the M3 of Manchenomys orcensis n. gen., n. sp. never develops a deep LRA3 or LSA4, as is the case for Allophaiomys deucalion M3s (Van der Meulen 1974: fig. 3h).

DESCRIPTION
The m1 of Ma. orcensis n. gen., n. sp. from Fuente Nueva 3 displays an anteroconid cap (AC2), five alternating triangles and a posterior lobe. All the re-entrant angles are filled with abundant cement. The anteroconid is short and wide in 14 m1s. Enamel is always lacking in the anterior half of the wall of the anteroconid complex. Specimens show negative enamel differentiation, with the exception of one specimen with undifferentiated enamel. The T4 is wider and in some cases shorter than the T5. LRA4 and BRA3 are well developed, therefore constraining the connection between AC2 and the T4-T5 dentine fields. The T4 and T5 alternate, although they are usually widely confluent. Dentine channels between the posterior lobe, T1, T2, T3 and T4 are very narrow. Lower first molars from different levels of the Quibas section present the same morphology as those from Fuente Nueva 3. However, in two m1s from Quibas a Mimomyan-ridge is present (levels QS-1 and QS-3; Fig. 3C). In addition, the number of teeth showing undifferentiated enamel is higher (8 out of 28 m1). Examples of M3 from Fuente Nueva 3 show an occlusal pattern composed of a transverse anterior lobe, two alternating triangles (T2-T3) and a posterior cap (PC1). There is always a relatively wide connection between T3 and PC1. The PC1 is simple, in some cases rounded. Some M3s present a very shallow LRA3, while it is lacking in others. No M3 expresses LSA4. Other than FN 3-5, the remaining M3s from the levels FN 3-3 and FN 3-6 of the Fuente Nueva 3 section present a similar dental pattern. This is also the case for the three specimens coming from the levels QS-1, QS-3 and QC 4-5 from the section of Quibas.

remarkS
The first occurrence of Manchenomys n. gen. (Ma. oswaldoreigi n. comb.) is recorded at the post-Olduvai site of Barranco de los Conejos (Guadix-Baza Basin; Agustí et al. 2013). This species is also present at other coeval levels of the Mimomys (now Manchenomys) oswaldoreigi Zone in the Guadix-Baza Basin, such as Cortes de Baza 1 and Fuentecica 5 (Agustí et al. 1999(Agustí et al. , 2015bOms et al. 2000a;Fig. 4). In the Guadix-Baza Basin, Manchenomys n. gen. seems to be absent at the levels of the Allophaiomys ruffoi Zone, such as Venta Micena 1 and 2, Orce 7 and Cañada de Murcia 1 (Agustí et al. 2015b). However, a form close to Manchenomys oswaldoreigi n. comb. appears to be associated with Allophaiomys ruffoi (Pasa, 1947)  . Manchenomys orcensis n. gen., n. sp. is still present in the lower levels of the section of Quibas, covering the Matuyama-Jaramillo transition at 1.07 Ma. However, it is already absent from the post-Jaramillo upper level of this section (QS-7, between 0.99 and 0.78 Ma), the last occurrence of this species being recorded within the Jaramillo geomagnetic subchron (Piñero et al. 2020(Piñero et al. , 2022. Therefore, the stratigraphic range of Manchenomys n. gen. covers the whole upper Matuyama geomagnetic chron between the Olduvai and Jaramillo subchrons (Fig. 6). The persistence of Manchenomys n. gen. in the late early Pleistocene of southern Spain parallels a similar persistence of small-sized Mimomys [Mimomys pusillus (Mehely, 1914), Mimomys blanci Van der Meulen, 1973] in the late early Pleistocene of western (including Italy) and central Europe, in all the cases associated with late representatives of Mimomys savini (Van der Meulen 1973).
Previous to the findings of Fuente Nueva 3 and Quibas, Agustí et al. (1993a) already defined a new species of Mimomys characterized by its arhizodont molars, with the rare exception of the lower m 3, at the site of Gilena 2, again in southern Spain. The inclusion of M. oswaldoreigi within the genus Mimomys was always problematic, provided the practical absence of roots in its molars. The appearence of a more derived arhizodont vole, Manchenomys orcensis, in younger levels enables us to clarify the position of M. oswaldoreigi, as a first member of an independent, endemic lineage of arhizodont voles. Therefore, the new combination Manchenomys oswaldoreigi is presented in this paper.

DISCUSSION
The m1 of Manchenomys n. gen. presents a simple occlusal pattern also being present in some advanced early Pleistocene Mimomys species, such as Mi. pusillus and Mi. tornensis (Janossy & Van der Meulen 1975;Rabeder 1981). There is no record of Mi. tornensis in the Guadix-Baza Basin in southern Spain, although it is present in eastern Spain in the early Pleistocene pre-Olduvai section of Almenara-Casablanca 1 (Castellón, Spain), where it is associated with the arvicolid Kislangia gusii Agustí, Galobart & Martín-Suárez, 1993(Esteban Aenlle & López Martínez 1987Agustí et al. 1993bAgustí et al. , 2011. Nevertheless, provided its proximity to the Guadix-Baza Basin, Mi. tornensis appears as a feasible ancestor for Manchenomys n. gen. Mi. tornensis is similar in size to Manchenomys orcensis n. gen., n. sp., although it is larger than Ma. oswaldoreigi n. comb. from the type-locality of Gilena 2 ( Table 1). The A/L index of Mi. tornensis is similar to the two species of Manchenomys n. gen. (Table 3). The B/W index is similar to Ma. orcensis n. gen., n. sp., although smaller than Ma. oswaldoreigi n. comb. The main difference lies in the C/W index, which is considerably lower in Mi. tornensis. Mi. tornensis has been proposed as the ancestor of the arhizodont arvicolids of the genus Allophaiomys (Rabeder 1986; however, see Garapich & Nadachowski 1996 for a different view), and this may have been also the case for Manchenomys n. gen. However, currently this question remains unanswered.