This is the accepted version of the journal article: Azzarà, Beatrice; Breda, Marzia; Cirilli, Omar; [et al.]. «Vigna Nuova : the first Middle Villafranchian mammal assemblage from the Valdichiana Basin, Perugia (Italy)». Bollettino della Società Paleontologica Italiana, Vol. 61, Issue 2 (September 2022), p. 223-247. DOI 10.4435/BSPI.2022.12 This version is available at https://ddd.uab.cat/record/271465 under the terms of the license 1 Vigna Nuova (Perugia): the first Middle Villafranchian mammal assemblage 2 from the Valdichiana Basin 3 4 Abbreviated title: Mammals from Vigna Nuova 5 6 Beatrice AZZARÀ*, Marzia BREDA, Omar CIRILLI, Joan MADURELL-MALAPEIRA, Federica RUZZA, 7 Leonardo SORBELLI, Domenico TANCREDI & Marco CHERIN 8 9 B. Azzarà, Dipartimento di Fisica e Geologia, Università degli Studi di Perugia, Via A. Pascoli, 06123 Perugia, Italy; 10 beatrice.azzara@studenti.unipg.it *corresponding author 11 M. Breda, Centro di Ateneo per i Musei (CAM), Università di Padova, Via Giotto 1, 35121, Padova, Italy; 12 marzia.breda@unipd.it 13 O. Cirilli, Laboratory of Evolutionary Biology, Department of Anatomy, College of Medicine, Howard University, 20059, 14 Washington DC, USA; Dipartimento di Scienze della Terra, Paleo[Fab]Lab, University of Florence, Via La Pira 15 4, 50121, Firenze, Italy; omar.cirilli@howard.edu 16 J. Madurell-Malapeira, L. Sorbelli, Institut Català de Paleontologia Miquel Crusafont, Universitat Autònoma de 17 Barcelona, 08193 Cerdanyola del Vallès, Spain; joan.madurell@uab.cat, leonardo.sorbelli@icp.cat 18 F. Ruzza, M. Cherin, Dipartimento di Fisica e Geologia, Università degli Studi di Perugia, Via A. Pascoli, 06123 Perugia, 19 Italy; federica.ruzza@studenti.unipg.it, marco.cherin@unipg.it 20 D. Tancredi, Dipartimento di Chimica, Biologia e Biotecnologie, Università degli Studi di Perugia, Via dell’Elce di Sotto 21 8, 06123 Perugia, Italy; domenico.tancredi@studenti.unipg.it 22 23 24 KEY WORDS – Canis, Croizetoceros ramosus, Early Pleistocene, Equus senezensis, Fossil mammals, 25 Umbria. 26 27 ABSTRACT - Umbria (central Italy) represents one of the most interesting areas for the study of 28 Mediterranean Plio-Pleistocene mammal faunas thanks to the abundance of paleontological deposits 29 distributed throughout the territory. Most of the fossils have been discovered in the Tiberino Basin, 30 which extends over ca. 1800 km2 along the middle part of the regional territory, and in the small 31 Tavernelle-Pietrafitta Basin, south of Trasimeno Lake. Instead, in this paper we report on the 32 assemblage from Vigna Nuova (Piegaro, Perugia) in the Valdichiana Basin, whose continental 33 deposits crop out in few areas of western Umbria. Taphonomic and geological evidence, coupled 1 34 with the few available excavation notes, allow to divide the sample into two assemblages. The younger 35 comes from the conglomerates (here called “upper layer”) still visible on the outcrop and is 36 composed only of remains of cf. Leptobos. The data obtained from this fragmentary sample and some 37 sedimentological-stratigraphic considerations suggest that it can be correlated with other well- 38 known Valdichiana assemblages (e.g., Selvella, Farneta), dated at ca. 1.5 Ma. The older sub-sample 39 comes from a putative paleosol (here called “lower layer”) no longer visible in the field. It includes 40 cf. Megantereon, Canis sp., Mustelidae indet., Proboscidea indet., Equus cf. senezensis, cf. Leptobos, 41 Croizetoceros ramosus, ‘Pseudodama’ sp., Cervidae indet. (large size), and Sus cf. strozzii. This 42 assemblage can be referred to the Coste San Giacomo Faunal Unit (late Middle Villafranchian; ca. 43 2.2–2.1 Ma). It represents the first well-preserved sample of this age in Umbria and one of the few in 44 Italy, opening new interesting research perspectives on the distribution of mammals in the Peninsula 45 and also on the stratigraphy and evolution of the Valdichiana Basin. 46 47 Aggiungeva che per scavare le poche ossa che mi avrebbe subito inviate aveva 48 dovuto faticare enormemente e lottare per due giorni contro ogni sorta di difficoltà. 49 Ricevuto il graditissimo invio mi affrettai a consolidare e restaurare con ogni cura quelle 50 primizie ed oggi sono lieto di poter annunziare, in così solenne adunanza, la interessante 51 scoperta… 52 53 He added that in order to excavate the few bones that he would immediately send 54 me he had had to work enormously and fight for two days against all sorts of difficulties. 55 Having received the very welcome shipment, I hastened to consolidate and restore those 56 fruits of the soil with every care and today I am pleased to be able to announce, in such a 57 solemn meeting, the interesting discovery… 58 59 Giovanni Capellini (1890: 275) 2 60 61 62 INTRODUCTION 63 64 In spite of its small geographical extension (ca. 8500 km2), Umbria represents a treasure trove 65 of very interesting discoveries for Plio-Pleistocene vertebrate paleontology. Mentioned since the first 66 half of the 19th century (Cuvier, 1821), in the following decades these discoveries began to draw the 67 attention of the Italian scientific community (e.g., Verri, 1879; Pantanelli, 1886; Tuccimei, 1891, 68 1895; Clerici, 1894, 1895; Meli, 1895), including eminent scholars such as Giovanni Capellini, who 69 published at least three contributions on Umbrian mammals (Capellini, 1888, 1889, 1890). 70 From the paleogeographic point of view, during the Plio-Pleistocene, central Italy was 71 intersected by a set of NW-SE trending grabens related to the eastward migration of the Apennine 72 extensional deformation (Martini & Sagri, 1993; Pascucci et al., 1999; Collettini et al., 2006). These 73 depressions were filled by marine sediments on the west and by continental sediments on the east, 74 with the transition between the two realms (i.e., the Tyrrenian paleocoast) crossing western Umbria 75 from Città della Pieve south-east to Orvieto (Ambrosetti et al., 1987; Bizzarri & Baldanza, 2020). 76 The main intermontane basin in the central-eastern part of the region is the Tiberino Basin (Fig. 1a), 77 which extends for ca. 1800 km2 from Sansepolcro to Terni (southwest) and Spoleto (southeast) 78 (Basilici, 1997). This extensional basin, historically erroneously interpreted as a single huge lake 79 (“Tiberino Lake”; Lotti, 1917), has actually been occupied at least over the last 3 Ma by a complex 80 and dynamic system of lake, marsh, and river environments interconnected each other (Basilici, 81 1997). Among the smaller basins, it is worth noting the Tavernelle-Pietrafitta Basin (Fig. 1a), located 82 south of Trasimeno Lake, characterized alternatively by lacustrine/palustrine and alluvial plain 83 conditions during the Plio-Pleistocene. This basin hosts the most important Umbrian fossiliferous 84 site, and one of the most important in Europe, that is, Pietrafitta (Martinetto et al., 2014; Sorbelli et 85 al., 2021, and references therein). The Tavernelle-Pietrafitta Basin is interposed between the Tiberino 3 86 Basin to the east and the Valdichiana Basin to the west (Fig. 1a). The latter develops from north-west 87 to south-east along the Tuscany-Umbria-Latium boundary and has been filled by shallow marine and 88 continental sediments since the early Pliocene (Bizzarri & Baldanza, 2020). 89 Reports on fossil continental mammals are abundant in the literature for the Tiberino Basin 90 (e.g., localities of Pantalla, Torre Picchio, Podere San Lorenzo, Villa San Faustino, among others) 91 and for the Tavernelle-Pietrafitta Basin (see the aforementioned site of Pietrafitta) and, following the 92 steps traced by the famous 19th century researchers, they have been the subject of many modern 93 studies and revisions (e.g., Sardella et al., 1995, 2018; Girotti et al., 2003; Cherin et al., 2016, 2019a). 94 On the contrary, in the western part of Umbria (i.e., in the Valdichiana Basin), the occurrence of 95 marine vertebrates is known above all (fossil cetaceans from the Allerona area; Baldanza et al., 2018). 96 This knowledge gap is filled by the Vigna Nuova local faunal assemblage (LFA), which 97 represents the only record of continental mammals from this portion of the regional territory. The 98 collection was briefly mentioned by Gentili (1993). A faunal list was published by Petronio et al. 99 (2002) and copied-and-pasted by Argenti (2004), who also stated that the Vigna Nuova LFA should 100 be referred to the Tasso faunal unit (FU) “for its stratigraphic position with respect to Pietrafitta and 101 its paleontological assemblage (mainly for the presence of Megantereon cultridens)” (Argenti, 2004: 102 70). 103 Here we provide the first systematic study of the Vigna Nuova LFA, ascertaining that, in fact, 104 it is made up of two distinct assemblages, one of which offers some surprising biochronological 105 information. 106 107 108 GEOLOGICAL SETTING AND TAPHONOMY 109 110 The Vigna Nuova LFA was unearthed in 1989 during excavation works on the front of a hill 111 (named “Greppo Lidolfo”) for the planting of a vineyard (hence the name, which means “new 4 112 vineyard”). The site is located near Piegaro (Perugia; 42°58'04.2"N, 12°04'04.7"E) at the easternmost 113 border of the Valdichiana Basin (Fig. 1a). 114 Some inconsistencies exist as regards the stratigraphy of the deposits cropping out in the area. 115 In the description of the Geological Map of Italy in 1:50,000 scale (Foglio 310, Passignano sul 116 Trasimeno), Barchi & Marroni (2010) mention the Vigna Nuova assemblage as recovered from the 117 Fosso Cigne Unit (Tiberino Basin, Nestore Synthem). Following Argenti (2004), the assemblage, 118 hence the unit, is referred to the Tasso FU (ca. 1.8 Ma). Conversely, the most recent Umbrian 119 geological mapping projects (Regione Umbria, 2013) include the Vigna Nuova locality in the Meletto 120 Unit (Valdichiana Basin, Sanfatucchio Synthem), also referred to the Tasso FU. The Meletto Unit is 121 very similar to and probably partially heteropic with the nearby Gioiella Unit, which yielded 122 important samples of terrestrial mammals in the surroundings of the homonymous town, in particular 123 at the locality of Selvella (De Giuli, 1987), which is located only 16 km north-west of Vigna Nuova 124 (Fig. 1a). The Selvella LFA, originally dated at ca. 1 Ma (De Giuli, 1987), is today referred to the 125 Farneta FU (ca. 1.5 Ma; Alberdi & Palombo, 2013; Cirilli et al., 2020). In fact, the lithological and 126 sedimentological features of the Fosso Cigne and Meletto Units are quite similar. Both are mainly 127 composed by conglomerates with a reddish medium-coarse sand matrix. Clasts are poorly sorted and 128 predominantly made of sandstone in the first unit and carbonate in the second. The main difference 129 is that the Fosso Cigne conglomerates have channeled geometry, while those of Meletto show no 130 clear structures. Unfortunately, the Vigna Nuova outcrop (about 6 m in height) today is almost 131 completely covered by vegetation, so it is not possible to describe the local stratigraphic succession. 132 The few exposed portions (Fig. 1b) are made of conglomerates with rounded, polygenic clasts, 133 sometimes with imbricate bedding, alternating with thin layers of medium-coarse sand. The current 134 evidence therefore does not allow to resolve the doubts about the stratigraphic attribution of the 135 outcrop, for which further geological studies will be necessary. 136 Nonetheless, the taphonomic features of the Vigna Nuova mammal collection, coupled with 137 excavation notes and labels, offer some hints on the stratigraphic context. Based on the lithology of 5 138 the embedding sediment (when available), color of fossils, and their mineralization, it is possible to 139 clearly separate the collection into two groups. The first includes fossils embedded by or maintaining 140 traces of a whitish-yellowish sand with carbonate matrix, rich in large-sized quartz crystals and spots 141 of black organic matter. Calcite crystals can occur as coating and infills within larger bone cavities 142 (e.g., medullary cavity in long bones). Bones are both mineralized and impregnated (i.e., minerals 143 and secondarily organic matter filled the smaller cavities such as trabeculae; Fig. 2), thus being very 144 heavy. Some show root traces. Few specimens (SABAP_UMB 129847, 129849, 129852, 129853, 145 129869, 129881) are encrusted by a yellow-brown carbonate coating with reticular concretions. This 146 assemblage is characterized by considerable variability in taphonomic attributes, ranging from poorly 147 preserved isolated specimens (very abraded and broken bones) to well-preserved ones, sometimes 148 still in anatomical connection. The weathering stages (Behrensmeyer, 1978) are heterogeneous, 149 ranging from 0 to 4 (Fig. 2). Several taxa are recognized in this first assemblage (Supplementary 150 Online Material), with about half of the identifiable specimens referred to Cervidae (see Systematic 151 paleontology). On the contrary, only a bovid taxon (probably a single individual) is recognized in the 152 second group of bones. These latter show traces of a dark-yellow to reddish sand. Mineralization is 153 much less advanced and the inner cavities of the bones, regardless of their size, are empty. Most of 154 the remains show evidence of abrasion and breakage typical of fluvial transport, whereas weathering 155 is low (Fig. 2) suggesting little time spent in a sub-aerial environment prior to burial. 156 The observations above support the hypothesis that the two assemblages come from different 157 stratigraphic layers. The remains of the second group certainly come from the conglomerates 158 cropping out at Vigna Nuova. A putative fluvial depositional environment for these sediments is 159 compatible with the taphonomic features of bovid bones. On the other side, no traces of the sediments 160 associated to the fossils of the first group are visible today in the field. Based on original notes and 161 labels, we know that the paleontological material was recovered near the bottom of the hill, therefore 162 from a portion of the succession that is no longer exposed today. On the basis of the few lithological 6 163 and taphonomic elements available, we can assume that this lower layer was a paleosol, but only 164 future studies in the area could confirm this hypothesis. 165 The fossils of the two assemblages are treated as separate (lower layer and upper layer) 166 hereinafter. 167 168 169 MATERIALS AND METHODS 170 171 The studied material is curated by SABAP_UMB and was prepared/restored and studied at the 172 Dipartimento di Fisica e Geologia of the Università degli Studi di Perugia (Italy). For preparation 173 details see Cherin et al. (2019a). The identification codes of the SABAP_UMB are of two types: those 174 assigned in the past have six digits (e.g., SABAP_UMB 129842), while those recently assigned are 175 composed of an alphanumeric code such as SABAP_UMB 22.S241-3.263 (the complete list is 176 provided in Supplementary Online Material). In the manuscript the identifier “S241-” is omitted, thus 177 obtaining a code like SABAP_UMB 22.3.263. 178 Identification of paleontological specimens is based on anatomical descriptions and 179 morphological comparative analysis. Biometric/statistical comparisons are performed is some cases. 180 In particular, for equids, bivariate plots and PCA were built in R Environment v. 1.4.1103 (R Core 181 Team, 2013), using the packages prcomp() v. 3.6.2 (Venables & Ripley, 2008) and ggplot2() v. 3.3.3 182 (Wickham, 2016). PCA were performed on third metatarsals, using the parameter scale = T to have 183 a unit variance before starting the analyses in R. The Log10 Ratio diagrams on third metatarsals are 184 based on the log-transformed mean values of the Höwenegg Hippotherium primigenium von Meyer, 185 1829 (Bernor et al., 1997). The Höwenegg H. primigenium sample represents a well-studied, 186 homogeneous sample of a single species of Equinae, for which the full range of descriptive statistics 187 (mean, standard deviation, confidence limits, and minimum, maximum, and median values) are 188 available for each bone, including the cranium, mandible, and teeth. 7 189 All measurements are taken with digital calipers and are given with a precision of 0.1 mm. 190 Measurements are provided in Supplementary Online Material. Sources of data used for comparisons 191 and measurement and anatomical abbreviations are indicated in the relative figure/table captions. 192 Anatomical and/or measurement nomenclature follows Eisenmann et al. (1988) and Bernor et 193 al. (1997) for equids; Sala (1986), Masini (1989), and Duvernois (1990) for bovids; Heintz (1970) 194 for cervids; Fujita et al. (2010) and Cherin et al. (2018) for suids. 195 Institutional abbreviations – AUT, Aristotle University of Thessaloniki, Greece; IGF, Museo 196 di Storia Naturale, Sezione di Geologia e Paleontologia, Università di Firenze, Italy; IPS, Institut 197 Català de Paleontologia Miquel Crusafont, Spain; LAMC, Natural History Museums of Los Angeles 198 County, USA; MCL, Musée des Confluences de Lyon, France; NHMB, Natural History Museum 199 Basel, Switzerland; SABAP_UMB, Soprintendenza Archeologia Belle Arti e Paesaggio dell’Umbria, 200 Italy. 201 202 203 SYSTEMATIC PALEONTOLOGY (LOWER LAYER) 204 205 Class MAMMALIA Linnaeus, 1758 206 Order CARNIVORA Bowdich, 1821 207 Family FELIDAE Fischer von Waldheim, 1817 208 Genus Megantereon Croizet & Jobert, 1828 209 cf. Megantereon 210 (Fig. 3) 211 212 Referred material – Fragmented right upper canine (SABAP_UMB 129842). 8 213 Description – SABAP_UMB 129842 (Fig. 3a) preserves the root and a small part of the crown 214 without appreciable crenulations on the edges. The root is moderately curved in mesiodistal direction. 215 In mesial view, the tooth appears overall stouter and only slightly compressed labiolingually. 216 Comparisons and discussion – Sabertoothed felids (subfamily Machairodontinae) are a 217 widespread group of top predators in the Plio-Pleistocene carnivoran guilds of Europe and several 218 other parts of the world (Antón, 2013). 219 The fragmentary nature of the specimen from Vigna Nuova does not permit detailed 220 comparisons. The absence/presence of crenulations along the mesial and distal edges of the crown is 221 the more reliable criteria to differentiate Smilodontini and Homotherini, with the latter normally 222 showing clear crenulations. However, the absence of crenulations in the small preserved portion of 223 the crown of SABAP_UMB 129842 could be due to preservation issues. The morphology of the 224 crown is another diagnostic character, being normally stouter and less compressed in Homotherium 225 and slenderer and more labiolingually compressed in Megantereon. Unfortunately, this character 226 cannot be assessed too. The curvature of the root in SABAP_UMB 129842 is intermediate between 227 that of Megantereon cultridens Cuvier, 1824 from Villarroya and that of Homotherium crenatidens 228 Fabrini, 1890 from the Incarcal complex as can be seen in Fig. 4a. The morphology of the tip of the 229 root is not rounded as in Smilodontini, but more pointed as in Homotherium (Fig. 4a). However, the 230 taxonomic value of this character is not discussed in the literature and we do not know its variability 231 other than in the samples we have selected for comparison. 232 To solve these issues, we built a violin plot of the ratio between labiolingual with and 233 mesiodistal length of the upper canine of Old World Homotherium and Megantereon (Fig. 4b). Our 234 results show that the studied specimen fully falls into the range if Megantereon species, although it 235 is also close to the earliest representatives of the Homotherium lineage (i.e., early Pliocene of Odessa 236 Catacombs), for which however, biometric variation is virtually unknown due to poor record. 237 Accordingly, we confidently refer the studied specimen to cf. Megantereon. 238 9 239 Family CANIDAE Fischer von Waldheim, 1817 240 Genus Canis Linnaeus, 1758 241 Canis sp. 242 (Fig. 3) 243 244 Referred material – Articulated cervical vertebrae 3-7 (SABAP_UMB 129850); rib 245 (SABAP_UMB 129875); thoracic vertebra 1 and rib fragments (SABAP_UMB 129877); articulated 246 thoracic vertebrae 2-3 (SABAP_UMB 129879). 247 Description – The canid material from Vigna Nuova is represented only by elements of the 248 axial skeleton, all belonging to a single individual. The articulated cervical vertebrae 3-7 (C3-7; 249 SABAP_UMB 129850) have retained their original shape, even if they miss some portions (Fig. 3b). 250 The spinous process is still present only in C3, while the right transverse process is preserved only in 251 C4-6. The vertebrae have a long and stout body, which gradually became shorter and narrower 252 towards C7 (see measurements in Supplementary Online Material). The first thoracic vertebra (T1; 253 SABAP_UMB 129877 partim) preserves only the body and few fragments of the neural arch. T2 and 254 T3 are still articulated (SABAP_UMB 129879) and in the former, a slender, long, and posteriorly 255 directed spinous process is still visible, but half broken (Fig. 3d). The two most complete ribs 256 (SABAP_UMB 129875 and SABAP_UMB 129877 partim; Fig. 3c) probably correspond to the first 257 pair, as they perfectly articulate into the articular facets on T1-T2. 258 Comparisons and discussion – The first occurrence of Canis in Europe has long been regarded 259 as one of the main biochronological markers of the onset of the Late Villafranchian, i.e., the so-called 260 “wolf event” (Azzaroli, 1983; Azzaroli et al., 1988; Torre et al., 1992, 2001). Nevertheless, further 261 discoveries such as those of Canis sp. in the Early Villafranchian of Vialette (France; Lacombat et 262 al., 2008) and in the Middle Villafranchian of Coste San Giacomo (Italy; Bellucci et al., 2012, 2014), 263 point to an earlier arrival of Canis in Western Europe and support a model of diachronous dispersal 264 across the continent (Sotnikova & Rook, 2010). Aside from the aforementioned records of Canis 10 265 sensu stricto in the Early-Middle Villafranchian, several species are recognized in later times, 266 including Canis etruscus Forsyth Major, 1877, Canis arnensis Del Campana, 1913, and Canis 267 mosbachensis Soergel, 1928, among others (Cherin et al., 2014; Bartolini-Lucenti & Rook, 2016; 268 Bartolini-Lucenti et al., 2017, 2020 and references therein). 269 The C3 and C4 from Vigna Nuova show striking morphological resemblance with the same 270 vertebrae of C. etruscus from Olivola (Italy; IGF 4523) figured by Torre (1967: plate VIII, fig. 8). In 271 turn, the same author stated that the postcranial morphology of C. etruscus does not differ 272 significantly from that of the extant wolf Canis lupus Linnaeus, 1758, except in relatively smaller 273 dimensions (Torre 1967: 131). The vertebrae from Vigna Nuova are extremely similar to those of 274 Olivola especially in the relative development of the pre- and postzygapophyses with respect to the 275 neural arch and body, in the inclination (ca. 45° anterodorsally to posteroventrally) of the anterior and 276 posterior articular facets in lateral view, and in the ventral projection of the posterior portion of the 277 body relative to the anterior (i.e., the ventral edge of the body is inclined posteroventrally in lateral 278 view) (Fig. 3e-f). In addition to the morphological similarities, the vertebrae are also very similar in 279 size (e.g., the length of the body is 35.6 versus 34.7 mm in C3 and 31.4 versus 32.1 mm in C4 in 280 SABAP_UMB 129850 and IGF 4523, respectively). However, despite these similarities with C. 281 etruscus, the sample from Vigna Nuova does not offer sufficient diagnostic characters for 282 identification at the species rank and is therefore referred to Canis sp. 283 284 Family MUSTELIDAE Fischer von Waldheim, 1817 285 Mustelidae indet. 286 (Fig. 3) 287 288 Referred material – Crown of left upper canine (SABAP_UMB 22.3.263). 289 Description – An almost complete canine crown from the Vigna Nuova collection can be 290 attributed to a small-sized carnivoran (Fig. 3g). The crown is straight and has a roughly circular 11 291 section along all its height. The lingual wall is smooth, whereas the labial one shows a shallow vertical 292 furrow in distolabial position and another groove, markedly deeper, in mesiolabial position. Neither 293 of them reaches the apex of the crown. The latter is relatively rounded. A well-developed keel is 294 present along the mesial margin of the tooth and a weaker one along the distal one. 295 Comparisons and discussion – The incompleteness of the specimen hinders any attempt at 296 taxonomic attribution. Assuming that it is an upper canine (due to the virtually absent curvature of 297 the crown) of the left side (due to the position of furrows and keels), the field can be restricted to 298 small-sized carnivorans (the preserved portion of the crown is about 1 cm in height). We can exclude 299 felids and canids, characterized by upper canines with more marked curvature, very pointed apexes, 300 and labiolingual flattening (i.e., elliptical section). The greatest similarities are found with the 301 mustelids, which instead have straight upper canines, with an approximately circular section and a 302 complex of furrows corresponding, in relative position and development, to that of SABAP_UMB 303 22.3.263. The overall size of the latter is approximately between that of a marten and that of a badger, 304 i.e., it approaches that of extinct forms such as Pannonictis (see Colombero et al., 2012). 305 306 Order PROBOSCIDEA (Illiger, 1811) 307 Proboscidea indet. 308 309 Referred material – Rib fragment (SABAP_UMB 22.3.264). 310 Description – The only available proboscidean material is a small fragment of a rib. 311 Comparisons and discussion – At least four proboscideans occur in the European 312 Villafranchian mammal assemblages: Mammut borsoni (Hays, 1834), Anancus arvernensis (Croizet 313 & Jobert, 1828), Mammuthus rumanus (Stefanescu, 1924), and Mammuthus meridionalis (Nesti, 314 1825) (approximately in stratigraphic order, from older to younger). Due to the fragmentation of the 315 available specimen, it is not possible to refer it to a specific taxon. 316 12 317 Order PERISSODACTYLA Owen, 1848 318 Family EQUIDAE Gray, 1821 319 Genus Equus Linnaeus, 1758 320 Equus cf. senezensis Prat, 1964 321 (Fig. 5) 322 323 Referred material – Articulated left metacarpals II-III-IV, proximal part (SABAP_UMB 324 129840); left metacarpal III, proximal epiphysis (SABAP_UMB 129857); cervical vertebrae 325 (SABAP_UMB 129867, SABAP_UMB 129874); left astragalus (SABAP_UMB 129869); mandible 326 with left and right c1-i3 and left p2, divided into two pieces (SABAP_UMB 129835+129837); partial 327 left hindlimb in anatomical connection with distal tibia, tarsus, metatarsals II-III-IV (SABAP_UMB 328 129856+129861+129862+129868). 329 Description – The mandible SABAP_UMB 129835+129837 includes a well-preserved left p2 330 with medium-high stage of wear. It shows a short and blunt anterostylid, the metaconid is small and 331 rounded, and the metastylid is pointed distally. The linguaflexid is V-shaped, a feature diagnostic of 332 stenonian horses. The ectoflexid is very shallow. The preflexid is short whereas the postflexid is 333 longer with simple ornamented morphology. This last feature indicates an advanced wearing stage. 334 A small pli caballinid is present on the mesial border of the hypoconid. The anterior portion of the 335 mandible exhibits a slender symphysis with the left mandibular body partially preserved. The lower 336 incisors show a high degree of wear, with the infundibulum not always present (e.g., left i1). The 337 lower canines are visible as small blunt enamel knobs that emerge from the mandible about 1 cm 338 behind the i3s. 339 SABAP_UMB 129840 includes a left third metacarpal (MC3) preserving about half diaphysis 340 and the proximal epiphysis. It has a narrow appearance in anterior view, although not slender. The 341 proximal epiphysis has a flat articular surface for the magnum and large articular facets for the second 13 342 and fourth metacarpals (MC2, MC4). The latter are almost complete and preserved in anatomical 343 connection. 344 The most complete specimen is the articulated left hindlimb SABAP_UMB 345 129856+129861+129862+129868. The tibia preserves the diaphysis and the distal epiphysis, lacking 346 the proximal one. Its general appearance is rather bulky. Much information about the morphology of 347 the distal epiphysis cannot be not observed due the presence of the astragalus and calcaneum in 348 anatomical connection, with some sediment in between. The latter bones are complete. The 349 calcaneum has a short and robust corpus, with a well-preserved massive tuber calcanei. The 350 morphology of the other ankle elements (central, first+second, third, and fourth tarsal bones) cannot 351 be described due to anatomical connection. The second, third, and fourth metatarsals (MT2-4) are 352 articulated and complete. The MT3 appears robust, due the reduced maximal length, the relatively 353 large proximal and distal epiphyses, and the wide circular diaphysis. The distal epiphysis has small 354 medial and lateral protuberances, which are not larger than the medial and lateral trochleae. The latter 355 are round, with a well-developed crista sagittalis in between. In posterior view, the long lines for the 356 attachment of the interosseus ligaments connecting the MT2 and MT4 are partially preserved. 357 SABAP_UMB 129869 is a left astragalus, with incomplete and weathered articular surface for 358 the calcaneum. The astragalus has a short and massive aspect, with small medial and lateral trochleae 359 and a large medial tuberculum tali for the attachment of the ligamentum collaterale. The trochlear 360 groove is large. 361 Comparisons and discussion – In Europe, the Equus Datum is marked by the appearance of the 362 large-sized Equus livenzovensis Bajgusheva, 1978, occurring at the beginning of the Pleistocene (ca. 363 2.6 Ma) in the localities of Liventsovka (Russia), Montopoli (Italy), Roca-Neyra (France), and El- 364 Rincón 1 (Spain) (Alberdi et al., 1997, 1998; Azzaroli, 2000; Bernor et al., 2018, 2019; Rook et al., 365 2019; Cirilli et al., 2021a, 2021b, 2021c). This biochronological event represents the dispersal of the 366 monodactyl horses form North America into the Old World, with the species Equus simplicidens 367 Cope, 1892 as the most likely original pool for the Eurasian species. During the Gelasian, monodactyl 14 368 equids diversified in different species such as Equus stenonis Cocchi, 1867, Equus major Depéret in 369 Delafond & Depéret, 1893 ex Boule, Equus senezensis, and Equus stehlini Azzaroli, 1964. Equus 370 stenonis represents the most widespread species during this time frame, with its first occurrence in 371 the Early Pleistocene site of Saint Vallier (France; 2.45 Ma; dating in Nomade et al., 2014). This 372 species dispersed in western Eurasia, from Caucasus to the Iberian Peninsula, and diversified into 373 different populations/ecomorphotypes (Palombo & Alberdi, 2017; Cherin et al., 2021; Cirilli et al., 374 2021a). The distribution of E. major is still not well defined, but at the present time it seems mostly 375 present in Central, North, and Eastern Europe (Palombo & Alberdi, 2017). Cirilli (2022) reported 376 some new insights on the evolution and dispersals of the Gelasian medium-small sized horses. Recent 377 research on the Middle Villafranchian localities of Senèze (France) and Coste San Giacomo (Italy) 378 have confirmed the presence of a medium-sized horse, intermediate between E. stenonis and E. 379 stehlini (Palombo et al., 2017; Cirilli et al., 2021a; Cirilli, 2022), identified as E. senezensis. The 380 species was originally described as a subspecies of E. stenonis, namely E. stenonis senezensis, by Prat 381 (1964, 1980) from the locality of Senèze (France; 2.2–2.0 Ma; dating in Nomade et al., 2014). 382 Subsequently, Alberdi et al. (1998) questioned its subspecific rank, suggesting a better attribution to 383 the species level. This interpretation was supported by the revision of the European E. stenonis and 384 E. stehlini samples undertaken by Cirilli et al. (2021a) and Cirilli (2022), suggesting that the medium- 385 sized horse from Senèze truly represents a different species during the Early Pleistocene. Beside 386 Senèze, E. senezensis has been reported also in Italy, from the Early Pleistocene localities of Coste 387 San Giacomo and Montecarlo (Upper Valdarno Basin). These sites have been dated at ca. 2.2–2.1 Ma 388 (Ghinassi et al., 2005; Bellucci et al., 2012, 2014; Fidolini et al., 2013; Florindo et al., 2021), hence 389 similar in age with Senèze. The European medium-sized horses of the late Early Pleistocene 390 underwent a turnover at the Gelasian-Calabrian boundary. Indeed, the first occurrence of Equus 391 altidens von Reichenau, 1915 has been recognized in the paleoanthropological site of Dmanisi 392 (Georgia; 1.85–1.76 Ma), predating the first Italian occurrences of Farneta and Pirro Nord (Bernor et 393 al., 2021). Equus altidens would appear to be the most geographically widespread species in the late 15 394 Early Pleistocene, sharing its range with the last representatives of E. stenonis and the earliest 395 representatives of Equus suessenbornensis Wüst, 1900, and being part of an extensive faunal turnover 396 occurring at ca. 1.8 Ma in Eastern Europe and slightly later in Central and Western Europe (Alberdi 397 & Palombo, 2013; Palombo & Alberdi, 2017; Cirilli et al., 2020; 2021a; Bernor et al., 2021; Bartolini- 398 Lucenti et al., 2022). 399 Morphometric comparisons on the Vigna Nuova equid remains are here provided based on the 400 most relevant elements, namely the p2 and the complete MT3. Fig. 6a plots the maximum width (M6) 401 versus maximum length (M1) for p2, comparing the Vigna Nuova specimen with the European Early 402 Pleistocene species. SABAP_UMB 129835 is included in the range of variation of E. senezensis from 403 Senèze, and plots close to the largest E. stehlini (Upper Valdarno Basin) and the smallest E. altidens 404 (Dmanisi and Venta Micena). It is well separated from the European E. stenonis, thus far highlighting 405 its attribution to a medium-small sized species. Equus suessenbornensis represents the largest species 406 considered in the plot, whereas E. stenonis mygdoniensis from Gerakarou-1 (Mygdonia Basin, 407 Greece) overlaps the morphometric range of variation of E. stehlini and E. altidens. 408 More interesting results are depicted in the multivariate analyses on the MT3s. Fig. 6b shows 409 the results for the MT3 PCA, with PC1 and PC2 accounting for 85.4% of the total variance (PC1 = 410 80.4%; PC2 = 5.0%). The loading distribution is reported as an inset in Fig. 6b and in Supplementary 411 Online Material within the variance components. PC1 separates species by maximal length from 412 negative to positive values (less to more elongate), whereas PC2 denoted a slenderer morphology 413 from positive to negative values. The Vigna Nuova specimen is included in the range of variation of 414 E. senezensis, distinct from the E. stehlini sample from Upper Valdarno and close to the smallest E. 415 altidens individuals from Dmanisi. Equus senezensis occupies an intermediate position between E. 416 stenonis and E. stehlini, confirming the previous results obtained by Cirilli et al. (2021b) and Cirilli 417 (2022). Equus stehlini is well separated from the rest of the comparative sample and shows a narrow 418 and massive MT3 morphology, with the largest E. stehlini specimens overlapping the smallest E. 419 senezensis. On the other hand, the medium-sized E. altidens shows a slenderer morphology, with a 16 420 more elongated maximal length a reduced diaphysis. The Dmanisi and Pirro Nord populations overlap 421 their range of variation, whereas the sample from Venta Micena (Spain) is somewhat larger. The 422 Gerakarou-1 medium-sized E. stenonis mygdoniensis is included in the range shown by the European 423 E. altidens samples. The European E. stenonis samples (Saint Vallier, Olivola, Matassino, and Upper 424 Valdarno Basin) overlap in their range of variation, characterized by larger dimensions and more 425 robust morphology than the medium-sized E. senezensis and E. altidens. Equus eisenmannae Qiu et 426 al., 2004 from Longdan (China) and E. livenzovensis exhibit again longer MT3s when compared with 427 E. simplicidens and E. stenonis, whereas E. major and E. suessenbornensis remain the largest equid 428 species from the European Early Pleistocene. 429 These results are also confirmed in the Log10 ratio diagrams on MT3. Fig. 6c plots the mean 430 values of the European Gelasian Equus species, whereas Fig. 6d shows the Calabrian ones. Both 431 diagrams include the complete MT3 from Vigna Nuova (SABAP_UMB 432 129856+129861+129862+129868 partim). As shown in Fig. 6c, the Vigna Nuova MT3 has a pattern 433 close to that of E. senezensis from Senèze, characterized by a reduced maximum length (M1) and 434 robust diaphysis, proximal, and distal epiphyses (M3-M14). It is overall larger than that of E. stehlini 435 (Upper Valdarno), whereas the incomplete MT3 from Montecarlo (Upper Valdarno Basin) plots in 436 between. The Vigna Nuova MT3 is smaller than that of E. stenonis, although showing a similar 437 pattern. Fig. 6d compares the Vigna Nuova MT3 with the E. altidens samples and E. 438 suessenbornensis. Although SABAP_UMB 129856+129861+129862+129868 partim shows similar 439 dimensions in the proximal and distal epiphyses than E. altidens, but its M1-M3 pattern is much 440 different form the E. altidens samples, which show a slender morphology with a more elongated M1 441 and reduced M3. SABAP_UMB 129856+129861+129862+129868 partim exhibits the same M3 442 values of E. altidens from Venta Micena, the largest among the E. altidens samples. As reported in 443 the PCA, E. stenonis mygdoniensis from Gerakarou-1 is included in the European E. altidens samples. 444 Considering this evidence, we ascribe the medium-sized equid from Vigna Nuova to E. cf. 445 senezensis, recognizing their close similarities, especially in the postcranial elements. 17 446 447 Order ARTIODACTYLA Owen, 1848 448 Family BOVIDAE Gray, 1821 449 Genus Leptobos Rütimeyer, 1877-1878 450 cf. Leptobos 451 (Fig. 7) 452 453 Referred material – Horn core fragment (SABAP_UMB 129851). 454 Description – The only bovid fossil from the Vigna Nuova lower layer is a fragment of horn 455 core. Unfortunately, it is so heavily weathered and abraded that the surface is no longer visible. Faint 456 parallel furrows are visible along one of the edges. The horn core is virtually straight and does not 457 show clear spiralization. 458 Comparisons and discussion – Based on the absence of curvature and spiralization, specimen 459 SABAP_UMB 129851 is here referred to cf. Leptobos (see detailed discussion in the section on the 460 bovid remains from the upper layer of Vigna Nuova). 461 462 Family CERVIDAE Goldfuss, 1820 463 Genus Croizetoceros Heintz, 1970 464 Croizetoceros ramosus (Croizet & Jobert, 1828) 465 (Fig. 7) 466 467 Referred material – Left maxilla with P2-M3 (SABAP_UMB 129863). 468 Description – Specimen SABAP_UMB 129863 belongs to an adult individual with teeth in 469 moderate stage of wear. The upper premolars are relatively large and similar in size to each other. In 470 occlusal view, the P3 and especially the P2 stand out for the pointed lingual margin. The P4 has a 471 more squared lingual margin, with a slight hint of molarization (i.e., a faint vertical groove crosses 18 472 the lingual wall of the tooth). On the labial side, the three premolars show pointed parastyle and 473 metastyle and, between them, a strong pillar of the paracone, which is mesially shifted and points 474 forward. The upper molars are similar to each other in overall morphology and exhibit lingual cones 475 with trapezoidal shape. The protoconal fold (pli protoconal) is noticeable in the M3, present but not 476 marked in the M2, and not visible in the M1 due to the more advanced wear. A sharp entostyle is 477 developed in all molars, while no lingual cingula are present. 478 Comparisons and discussion – Croizetoceros is an extinct cervid with uncertain phylogenetic 479 relationships and body size smaller than the Dama-like deer (see section on ‘Pseudodama’ below). 480 The genus is reported in Europe since the late Miocene with scanty remains (Azanza, 2000) and 481 becomes a relatively common element of the Early-Middle Villafranchian assemblages with the 482 species C. ramosus, ranging from Spain to Greece, with different samples sometimes referred to 483 different subspecies (Valli, 2004 and references therein). Croizetoceros sp. is reported based on a 484 single tooth from the Pliocene of Burdur (Turkey; Mayda et al., 2017). 485 The main dental features of C. ramosus are well described by Heintz (1970) on the basis of 486 several samples from France and Spain, and resumed by Valli (2004) with reference to the richest 487 European collection, namely that of Saint Vallier (France). Most of the recognized features can be 488 found also in the Vigna Nuova specimen, including the asymmetrical development of the strong pillar 489 of the paracone in the premolars, the presence of the protoconal fold in the molars, the relative 490 development of the styles on the labial wall of the molars. In particular, the maxilla from Vigna Nuova 491 shows striking morphological similarities with some known specimens of C. ramosus such as NHMB 492 Prr.346 from the type locality of Les Etouaires (France), MCL QSV-766 from Saint Vallier, AUT 493 GER-188 from Gerakarou (Greece), IPS28245/740b from La Puebla de Valverde (Spain) (see Fig. 8 494 for comparisons). Dimensionally, the Vigna Nuova specimen falls into the lower part of the range of 495 C. ramosus, approaching in particular to the material from Gerakarou (measurements in 496 Supplementary Online Material). However, it must be noted that the apparently larger size of the 497 specimens from the French localities of Saint Vallier, Pardines, and Les Etouaires could be due to a 19 498 slightly different measuring technique (Supplementary Online Material). The same applies to the ratio 499 between the premolar and molar length (LP/LM), which is 0.70 in SABAP_UMB 129863 and 500 between 0.69 and 0.79 in C. ramosus from several localities (Heintz, 1970; Kostopoulos, 1996; 501 Kostopoulos & Athanassiou, 2005). 502 The overall size of SABAP_UMB 129863 is also similar to the average values of small-sized 503 cervids of the roe deer lineage (i.e., Procapreolus, Capreolus). However, these latter taxa are well 504 recognizable in the more articulated occlusal pattern of the upper teeth (e.g., numerous enamel folds 505 that develop into the fossae), unlike the relatively simple morphology observed in SABAP_UMB 506 129863. Moreover, roe deer have more elongated premolar series with respect to molars. For instance, 507 the average LP/LM is 0.92 in Procapreolus cusanus from Les Etouaires (Heintz, 1970), 0.89 in 508 Capreolus cusanoides from Untermassfeld (Germany) and 0.84 in Capreolus suessenbornensis from 509 Süssenborn (Germany) (personal data). 510 The above lines of evidence allow us to confidently refer the Vigna Nuova maxilla to C. 511 ramosus, of which it represents one of the most complete and better-preserved specimens in Italy. 512 513 Genus ‘Pseudodama’ Azzaroli, 1992 514 ‘Pseudodama’ sp. 515 (Fig. 7) 516 517 Referred material – Left i2 (SABAP_UMB 129841); right frontal bone fragment with basal 518 antler (SABAP_UMB 129843); left frontal bone fragment with basal antler (SABAP_UMB 129844); 519 left frontal bone fragment with basal antler (SABAP_UMB 129845); right frontal bone fragment with 520 basal antler (SABAP_UMB 129846); right hemimandible with p3-m3 (SABAP_UMB 129848); 521 antler fragments (SABAP_UMB 129849, SABAP_UMB 129853, SABAP_UMB 129881); left 522 hemimandible fragment with dp4 (broken), m1, and m2 (SABAP_UMB 129860+129865), right M1 20 523 (SABAP_UMB 129870); broken left m3 (SABAP_UMB 129871); left i3 (SABAP_UMB 129872); 524 left i1 (SABAP_UMB 129873); right antler fragment (SABAP_UMB 129880). 525 Description – Fossils of a medium-sized deer are relatively common in the Vigna Nuova 526 collection. The antler remains SABAP_UMB 129843, SABAP_UMB 129844, SABAP_UMB 527 129845, and SABAP_UMB 129846 are quite similar to each other both in shape and dimensions 528 (measurements in Supplementary Online Material). All show a basal portion of the beam with an 529 almost circular section and the insertion of the first tine (unfortunately never preserved) at a certain 530 distance above the burr. The pedicles are also roughly circular in section (measurements in 531 Supplementary Online Material). The most complete specimen (SABAP_UMB 129880) is a long 532 fragment of right antler with basis of the basal tine, but missing the burr. The basal tine branches at 533 about 3 cm from the bottom edge of the antler. Above the basal tine, the beam runs roughly vertical, 534 with a slight lateral curvature in the first part and a medial curvature in the second. 535 The only upper tooth is the isolated right M1 SABAP_UMB 129870. It is virtually unworn and 536 shows mesiodistally compressed lingual cones, between which a small entostyle develops. 537 Mesiolingual, lingual, and distolingual cingula are present but weak. The labial styles are instead very 538 prominent, as are the pillars of the labial cones. 539 The right hemimandible SABAP_UMB 129848 bears the complete cheek tooth row, but in a 540 very advanced stage of wear (the m1 is almost completely worn out). This prevents a detailed 541 description of the dental morphology. In occlusal view, the lingual wall of the paraconid is open, that 542 is, the paraconid and metaconid are not fused (i.e., “unmolarized” condition). Lingual cingula and a 543 faint ectostylid are still visible in the m2 and m3, but not in the m1 due to wear. The lower molar 544 morphology is better recognizable in the hemimandible fragment SABAP_UMB 129865+129860, 545 bearing a fragmented dp4 and complete m1 and m2. Both molars show a mesial cingulum and a 546 prominent ectostylid in labial view. 547 The lower incisors SABAP_UMB 129841, SABAP_UMB 129872, and SABAP_UMB 129873 548 may belong to the same individual based on preservation, relative size, and absent wear. 21 549 Comparisons and discussion – Pliocene and Pleistocene cervids with overall morphology and 550 size similar to the extant fallow deer are known as Dama-like deer. Villafranchian and 551 Epivillafranchian forms from Europe are characterized by un-palmated antlers with three or four 552 points. These forms have been allocated into a variety of genera (e.g., Cervus, Dama, Axis, Rusa, 553 Metacervocerus, Praeelaphus, etc.) and species, whose relationships with each other and with 554 present-day species are not clear (see a summary in Cherin et al., 2019a). Azzaroli (1992) proposed 555 to refer all these species to the genus Pseudodama, including the “Italian” lineage - with the species 556 Pseudodama lyra Azzaroli, 1992, Pseudodama nestii (Azzaroli, 1947), and Pseudodama farnetensis 557 Azzaroli, 1992 - and the “French” lineage - with the species Pseudodama pardinensis (Croizet & 558 Jobert, 1828), Pseudodama rhenana (Dubois, 1904), and Pseudodama perolensis (Bout & Azzaroli, 559 1952). Although we believe that some of the aforementioned species could probably be synonymized, 560 we follow Azzaroli’s concept, recognizing significant similarities between them all (Breda & Lister, 561 2013; Breda, 2015), but also including the Epivillafranchian Pseudodama vallonnetensis (de Lumley 562 et al., 1988) (Breda et al., 2020). However, pending a comprehensive phylogenetic analysis, we prefer 563 to indicate the genus ‘Pseudodama’ with inverted commas (following Breda et al., 2015, 2020; Cherin 564 et al., 2019a, among others). 565 The morphology and size of the identifiable material of middle-sized deer from Vigna Nuova 566 are clearly akin to the genus ‘Pseudodama’. On the other hand, the available fossils do not retain 567 characters that allow for an attribution at the species rank, for which more complete craniodental 568 remains (especially antlers) would have been needed. The antlers from Vigna Nuova differ from those 569 of derived species of ‘Pseudodama’ such as ‘P.’ farnetensis and ‘P.’ vallonnetensis, in which the 570 basal tine is normally inserted very close to or even leaning against the burr, as in the extant Dama. 571 Similarly, the sub-vertical morphology of the best-preserved antler (SABAP_UMB 129880) also 572 recalls the earlier forms of ‘Pseudodama’ rather than the later ones, in which the antlers tend to 573 diverge greatly posteriorly and laterally. However, these features must be considered with 574 thoughtfulness because they can also be subject to ontogenetic variation. 22 575 Other primitive characters of the Dama-like deer from Vigna Nuova lie in the teeth and include 576 the development of labial cones and styles and lingual cingulum in the upper molar (SABAP_UMB 577 129870), enlarged i1 (SABAP_UMB 129873), and un-molarized p4 (SABAP_UMB 129848) 578 (Azzaroli, 1947, 1992; Croitor, 2006; Pfeiffer, 2016). Another primitive condition is the high ratio 579 between the lower premolar and molar lengths measured in SABAP_UMB 129848 (Lp/Lm=0.69), 580 which exceeds the average values available for ‘P.’ lyra from Montopoli (Italy; Lp/Lm=0.64, n=2), 581 ‘P.’ rhenana from Saint Vallier (Lp/Lm=0.68, n=18) and Senèze (Lp/Lm=0.65, n=11), ‘P.’ nestii 582 from Pantalla (Italy; Lp/Lm=0.68, n=3), Olivola (Italy; Lp/Lm=0.62, n=5) and Upper Valdarno (Italy; 583 Lp/Lm=0.63, n=16). On the contrary, later forms of ‘Pseudodama’ show relatively shorter premolar 584 rows, as can be seen in ‘P.’ farnetensis from Pietrafitta (Italy; Lp/Lm=0.63; n=5) and Pirro Nord 585 (Italy; Lp/Lm=0.58; n=2) and ‘P.’ vallonnetensis from Untermassfeld (Germany; Lp/Lm=0.59; n=21) 586 and Vallparadís Estació (Spain; Lp/Lm=0.57; n=9) (Breda, 2015; this paper). 587 In summary, the material of ‘Pseudodama’ of Vigna Nuova is suggested to belong to a basal 588 species of the genus (e.g., ‘P.’ lyra, ‘P.’ rhenana), but the incompleteness of the specimens requires 589 a cautious attribution to ‘Pseudodama’ sp. 590 591 Cervidae indet. 592 (Fig. 7) 593 594 Referred material – Left maxilla with dP2-M2 (SABAP_UMB 129866). 595 Description – Specimen SABAP_UMB 129866 is particularly interesting. It is a left maxillary 596 fragment of a large-sized deer with deciduous premolars dP2-dP4, slightly worn M1, and freshly 597 erupted M2. The dP2 is heavily worn and shows a distal lobe which is only slightly wider than the 598 mesial one. Labially, the parastyle, paracone, and metastyle are pointed, while the metacone does not 599 form a pillar on the labial wall. In the dP3, the distal lobe is significantly wider and shorter than the 600 mesial one. On the labial wall, the pillar of the paracone and the mesostyle are particularly prominent 23 601 and are oriented mesially. A similar morphology characterizes the labial wall of the dP4, in which 602 however, the two lingual cones are almost equally developed and constricted mesiodistally. In the 603 M1, a low and rounded entostyle develops between the protocone and hypocone. The parastyle and 604 mesostyle are very strong and stretch in labial direction. The M2 looks like a slightly larger version 605 of the M1, although the entostyle is replaced by a lingual cingulum. In all teeth, the enamel on the 606 lingual walls is rough. 607 Comparisons and discussion – The most common large-sized deer occurring in the 608 Villafranchian is Eucladoceros, characterized by spectacular comb-like antlers (Azzaroli & Mazza, 609 1992; Croitor, 2018). Nevertheless, other groups of cervids of comparable size can be found in the 610 same time interval, although their record is poorer. These include the enigmatic Arvernoceros, whose 611 systematic status and chronological occurrence are debated (Heintz, 1970; Petronio & Pandolfi, 2011; 612 Croitor, 2018). The genus Praemegaceros seems to replace Eucladoceros, at least in southwestern 613 Europe, in the late Late Villafranchian (Abbazzi, 2004). Unfortunately, the diagnostic characters of 614 all these forms reside mainly in the antlers, of which we have no evidence in Vigna Nuova. No 615 peculiar character can be identified in the upper deciduous or permanent dentition, neither in shape 616 nor in size. For this reason, the large-sized deer of Vigna Nuova remains unidentifiable. 617 618 Family SUIDAE Gray, 1821 619 Genus Sus Linnaeus, 1758 620 Sus cf. strozzii Forsyth Major, 1881 621 (Fig. 7) 622 623 Referred material – Fragmented right maxilla with P4-M2 and broken M3 (SABAP_UMB 624 129847) and fragmented left maxilla with broken M3 (SABAP_UMB 129852). 625 Description – SABAP_UMB 129847 shows teeth with an advanced stage of wear that prevents 626 a detailed description of the occlusal pattern. However, wear makes it possible to appreciate the 24 627 considerable thickness of the enamel in all preserved teeth. The P4 has a squarish occlusal outline, 628 but appears slightly wider than long. A mesial cingulum and a distal cingulum are visible. The mesial 629 one has a slight constriction in the middle. A hinted protoprestyle is visible. The protofossa is narrow, 630 long and positioned lingually in the middle-mesial part of the tooth. The M1 is very worn (wear stage: 631 16, wear group: 4; Zeder & Lemoine, 2020), so much that the main cusps are unrecognizable. The 632 tooth has a figure-eight shape in occlusal view, with the mesial lobe smaller than the distal one and a 633 labial constriction between them. The M2 is similar in shape to the M1, but larger. Also in this molar, 634 none of the four main cusps can be described in detail due to the state of wear (wear stage: 12, wear 635 group: 2; Zeder & Lemoine, 2020). However, the difference in width between the mesial and distal 636 lobe is more marked than in the M1. The mesial cingulum is flat and does not seem to bear a 637 protopreconule. The oval-shaped paracone is narrow and in contact with the lingual cingulum. The 638 residue of the protocone is pointed, tiny and slightly visible. The labial cingulum of the mesial lobe 639 is broken but it is possible to recognize that a hypoectoconid was present. The transverse valley 640 between the paracone and metacone is wide lingually and ends exactly halfway through the tooth; no 641 tetrapreconule is visible. The tetracone is central and rectangular. In all the teeth no styles are visible 642 in lingual and labial view. 643 No descriptions can be provided for SABAP_UMB 129852 as the specimen preserves only a 644 distal splinter of the M3, which however is dimensionally comparable to the corresponding tooth of 645 SABAP_UMB 129847 and shows a similarly high enamel thickness. Based on preservation features, 646 it is likely that the two specimens belonged to the same individual. 647 Comparisons and discussion – The genus Sus, of putative Asian origin, spread rapidly towards 648 Europe since the Pliocene and became a common element in the Early Pleistocene faunal assemblages 649 throughout the continent (Cherin et al., 2020). The earliest member of this group is Sus arvernensis 650 (Croizet & Jobert, 1828), occurring in Eurasia during the Ruscinian and Early Villafranchian. At the 651 beginning of the Early Pleistocene, this species was replaced by the larger-sized S. strozzii. The latter 652 survived in Europe at least until the end of the Early Pleistocene, although its presence during the 25 653 latest Villafranchian (i.e., ca. 1.7–1.3 Ma) is yet not confirmed [see the so-called “suid gap” 654 hypothesis by Martinez-Navarro et al. (2015)] (Cherin et al., 2018, 2020). Starting from the Middle 655 Pleistocene, the wild boar Sus scrofa Linnaeus, 1758 replaced S. strozzii rapidly expanding 656 throughout Eurasia up to the present day (Cherin et al., 2020; Iannucci et al., 2020). 657 The distinction between the closely-related S. arvernensis and S. strozzii on one side and S. 658 scrofa on the other, is easy when lower canines are available, even in the absence of complete skulls. 659 This is because the first two species show a “verrucosic” morphology of lower canines, as opposed 660 to the “scrofic” condition of S. scrofa (see Cherin et al., 2020). The lack of lower canines in the Vigna 661 Nuova collection leads to rely only on biometric considerations (also because the advanced state of 662 wear of the teeth in SABAP_UMB 129847 prevents any detailed comparisons in dental morphology). 663 The box-plots in Fig. 9 clearly show that the Pliocene S. arvernensis can be distinguished from S. 664 strozzii and S. scrofa by its smaller size (especially molar length). SABAP_UMB 129847 has 665 relatively large-sized teeth, which fall in the ranges of these two latter species. Although in the 666 literature there are several works in which an attempt has been made to recognize S. strozzii and S. 667 scrofa on a dimensional basis (e.g., Ambrosetti et al., 1979; Faure & Guérin, 1983; Van der Made, 668 1999; Van der Made et al., 2017), Cherin et al. (2020) showed that this approach alone cannot be 669 sufficient, especially when large samples are considered both geographically and chronologically (the 670 wild boar shows enormous dimensional variation over time and space). As far as current knowledge 671 is concerned, only the dimensions of the dp4 can be used for a statistically significant distinction 672 between S. strozzii and S. scrofa (Iannucci et al., 2020). In this context, the length of the teeth of 673 SABAP_UMB 129847 alone is not decisive for the taxonomic identification of the specimen, while 674 considerations on their proportions can be. In particular, the P4 exhibits a relatively low length-to- 675 width ratio (i.e., it is more rectangular in occlusal view; Fig. 9). In this, it resembles S. strozzii rather 676 than S. scrofa, which instead has a squarer P4. This evidence – associated with the very thick enamel 677 possessed by the cheek teeth [i.e., one of the diagnostic characters of S. strozzii; Cherin et al. (2018)] 678 – allows to refer the Vigna Nuova material to S. cf. strozzii. 26 679 680 681 SYSTEMATIC PALEONTOLOGY (UPPER LAYER) 682 683 Order ARTIODACTYLA Owen, 1848 684 Family BOVIDAE Gray, 1821 685 Genus Leptobos Rütimeyer, 1877-1878 686 cf. Leptobos 687 (Fig. 10) 688 689 Referred material – Left hemimandible with p4-m3 and roots of p2-p3 (SABAP_UMB 690 129836); right metacarpal (SABAP_UMB 129838); atlas (SABAP_UMB 129839); intermediate 691 phalanx (SABAP_UMB 129854); left P2 (SABAP_UMB 129855); left tibia, proximal epiphysis 692 (SABAP_UMB 129858); left tibia diaphysis (SABAP_UMB 129876); humerus distal diaphysis 693 (SABAP_UMB 129878); horn core fragments (SABAP_UMB 22.3.248–22.3.252, 22.3.267); distal 694 phalanx fragments (SABAP_UMB 22.3.256, SABAP_UMB 22.3.266); cranium fragment 695 (SABAP_UMB 22.3.265); left medial malleolus fragment (SABAP_UMB 22.3.268). 696 Description – An incomplete hemimandible and an isolated P2 were recovered from the Vigna 697 Nuova upper layer. The mandible (SABAP_UMB 129836) includes the molars and the p4 with root 698 fragments of p2-p3. The teeth are all heavily worn. In the p4, the parastylid and paraconid are 699 separated by a shallow furrow, the protoconid is quite developed. The lingual valley between the 700 paraconid and protoconid strongly indents mesially. The metaconid is large and has the shape of a 701 blade, emerging lingually and strongly curving distally. The furrow separating the metaconid and 702 entoconid is quite deep and narrow due to the distal shift of the former. The entostylid and entoconid 703 are completely fused. The distal labial groove separating the protoconid from the hypoconid is strong. 704 The molars are mesiodistally elongated with well-separated lobes. The parastylid is well developed 27 705 and mesially projecting. The protoconid and hypoconid are morphologically similar except for a 706 slight mesiodistal constriction of the latter. The ectostylid emerges mesially from the distal lobe and 707 lies almost attached to the protoconid. The entostylid is smaller than the parastylid and projects 708 lingually. The inner enamel islets are halfmoon shaped. The mesial one is characterized by a 709 constriction in the mid part. The islet of the distal lobe in m3 is marked by a bubaline fold (sensu 710 Masini, 1989) in the distal portion. The cement is present in all the teeth, on both sides. The P2 711 SABAP_UMB 129855 has an advanced stage of wear. The paracone and parastyle lie almost 712 attached, separated by a shallow furrow that reaches the base of the tooth. The metacone is strongly 713 reduced. The metastyle projects distally. 714 Several horn core fragments are part of the collection. Most of them present nutrient foramina 715 and/or furrows on the outer side (e.g., SABAP_UMB 22.3.248 and SABAP_UMB 22.3.249) and 716 clear sign of pneumatization in the inner portion (e.g., SABAP_UMB 22.3.248 and SABAP_UMB 717 22.3.251). Despite their scanty preservation, it is possible to state that all of the remains are 718 characterized by a quite low degree of curvature and do not present evidences of spiralization. 719 The atlas SABAP_UMB 129839 preserves most of the body and the left wing. Dorsally, the 720 atlas is characterized by two large openings, on each wing, shifted towards the anterior portion of the 721 bone, in which two foramina are located. The left wing projects posteriorly but not far from the 722 posterior zygapophyses. The dorsal tubercle and the dorsal portion of the neural arch are missing. 723 The alar fossae, visible on the ventral side of the wings, are large and relatively deep depressions. 724 The posterior tubercle is partially preserved and not well developed. The anterior articular surface is 725 almost complete and extends on the ventral side. The area connecting the two posterior zygapophyses 726 is broad and flat. 727 The only complete metacarpal (SABAP_UMB 129838) is relatively slender. It has an hourglass 728 shape, with the two epiphyses having approximately the same width, mediolaterally wider than the 729 diaphysis. The proximal epiphysis, seen from above, is D-shaped; the anterior margin is sub-rounded 730 whereas the posterior one is straight. The lateral articular facet is triangular and much smaller than 28 731 the medial, which is sub-quadrangular with rounded anterior margin. The two facets are separated by 732 a small crest perpendicular to the major axis of the articulation. The crest ends posteriorly in a deep, 733 narrow groove which hosts the proximal nutrient foramen. Another foramen is present distally to the 734 proximal end, connected with the former through a very narrow furrow. All the anterior border of the 735 proximal epiphysis is characterized by a wrinkled tuberosity. The diaphysis has its minimum 736 mediolateral width slightly above the mid-point. There are two shallow grooves in the proximo- 737 anterior portion of the diaphysis. The anterior vascular groove at the distal shaft is narrow and 738 shallow, getting deeper and marked by sharp edges distally. The ovoidal foramen is located just above 739 the distal end of the vascular groove, covered by sediment. The mediolateral width at the contact area 740 between the diaphysis and the distal epiphysis is slightly smaller than that across the trochleae. The 741 intertrochlear margins as well as the crests are subparallel. The lateral and medial trochlear pits are 742 deep and marked by faint radial rugosities. The lateral abaxial hemicondyle is anteroposteriorly 743 thinner than the medial one. 744 The tibia remains includes a fragment of the diaphysis (SABAP_UMB 129876) and a partial 745 proximal epiphysis (SABAP_UMB 129858). The articulation surface of the latter is composed by the 746 two large condyles. Neither of these elements are complete. The proximally projected intercondylar 747 eminences are partially preserved. The diaphysis fragment represents the proximal half of the tibia 748 due to the visible crest running along one side of the shaft. 749 The intermediate phalanx SABAP_UMB 129854 is complete and well preserved. The bone is 750 stout, having a short and thick diaphysis. The proximal articulation is divided by the sagittal crest into 751 two facets for the articulation with the proximal phalanx (abaxial and axial glenoid cavities). 752 Posteriorly to the crest there is a deep U-shaped furrow. The abaxial cavity is slightly larger than the 753 axial. The abaxial tubercle and the axial pit are visible although not well developed. The distal 754 articular surface is subtriangular. The abaxial lobe has a pointed tip developing toward the diaphysis 755 of the phalanx. 29 756 The other available specimens are heavily fragmented and do not allow for any detailed 757 description. 758 Comparisons and discussion – The largest forms of bovids are part of the subfamily Bovinae 759 which, during the earliest Pleistocene of Europe, was mainly represented by the genus Leptobos. This 760 extinct taxon was populating most of the subtropical and temperate areas of Eurasia, spanning from 761 the Iberian Peninsula to China, from Northern India to, possibly, British Islands (Cherin et al., 2019b). 762 Although Leptobos has been matter of interest in numerous works (Merla, 1939; Pilgrim 1939; 763 Masini, 1989; Duvernois, 1990; Cherin et al., 2019b, among others), there is still no consensus about 764 its systematics and evolutionary history. Most of the scholars divide European Leptobos into two 765 lineages/groups or subgenera (Masini, 1989; Duvernois, 1990, 1992; Masini et al., 2013; Cherin et 766 al., 2019b; Sorbelli et al., 2021). The first includes the species Leptobos stenometopon (Sismonda, 767 1846), Leptobos elatus (Pomel, 1853), Leptobos merlai De Giuli (1987), and Leptobos furtivus 768 Duvernois (1989) (called Leptobos ex gr. SEM hereinafter) constrained to the Early-Middle 769 Villafranchian of Central and Western Europe; the second includes Leptobos etruscus (Falconer, 770 1868) and Leptobos vallisarni Merla (1949) (called Leptobos ex gr. EV hereinafter), limited to the 771 Late Villafranchian of Europe and, as concerning the latter species, Asia (Masini et al., 2013; Tong 772 et al., 2016; Sorbelli et al., 2021). Species of Leptobos ex gr. SEM are characterized by elongated 773 neurocranium, “taurine” horns (i.e., laterally emerging, with strong anticlockwise spiralization), and 774 slender limbs, whereas species of Leptobos ex gr. EV share shorter postcornual portion of the 775 cranium, posterolaterally emerging horns with single curvature, and stouter limb proportions. During 776 the last stages of the Villafranchian, a new group of large bovids reached Europe from Asia. These 777 bovines represented the first western dispersal of Bison which, most probably, evolved in mainland 778 Asia from a derivate stock of Leptobos at the end of the Pliocene (Sorbelli et al., 2021 and references 779 therein). These forms of large and stout bovids, grouped in the subgenus Bison (Eobison), were well 780 adapted to the harshened environmental condition that started to affect Europe during the Late 781 Villafranchian and rapidly replaced Leptobos in all the herbivore guilds of the continent. With the 30 782 faunal turnover of the Early-Middle Pleistocene Transition (i.e., during the Epivillafranchian), 783 Eobison, in turn, was replaced by the “true” Bison, which will be the dominant large bovids in all the 784 Holarctic realm until the Holocene (Sorbelli et al., 2021). 785 Due to their general morphology and size, the sample from Vigna Nuova is referable to a large 786 bovid. As already remarked, during the Early Pleistocene, the only European representatives of this 787 group are Leptobos and Bison (Eobison). The proportions of metapodials, in particular metacarpals, 788 are a useful taxonomic tool in order to discriminate between the various species belonging to these 789 genera (Maniakas & Kostopoulos, 2017; Sorbelli et al., 2021). However, due to the high intraspecific 790 variation that characterizes bovids, a cautious approach is recommended when dealing with small 791 samples, as in the Vigna Nuova case. The single metacarpal from this locality shows a relatively 792 slender built, with a long and thin diaphysis. In order to assess the robusticity of this element, a biplot 793 diagram comparing the total length of the metacarpal and the ratio between the transversal width and 794 the length is provided (Fig. 11). The Vigna Nuova specimen falls within the range of the largest 795 Leptobos, in the area overlapping with Eobison and the slenderest specimens of Bison schoetensacki 796 Freudenberg, 1914. The most imponent and robust species of Leptobos are those belonging to the EV 797 group. While the postcranial skeleton of L. etruscus is well known, that of L. vallisarni is still poorly 798 represented in the fossil record. Moreover, few limb bones referred to L. vallisarni are quite similar 799 to those of L. etruscus, thus, without complete cranial elements or a large sample of postcranial 800 remains, it is virtually impossible to discern the two taxa. 801 The lower dentition of Leptobos and Bison are quite similar, except for the evident increasing 802 of hypsodonty and larger average size in the latter. The morphology of the p4 has been used as tool 803 in order to differentiate the two groups of Leptobos (Masini, 1989; Duvernois, 1990). The distally 804 projected metaconid, the strong hypoconid, and the deep distolabial groove are characters shared 805 between the Vigna Nuova and Leptobos ex gr. EV. Nonetheless, several scholars underlined the 806 strong variability affecting the tooth morphology of large bovids, remarking how taxonomic 31 807 inferences based on dentognathic remains should be taken cautiously (Skinner & Kaisen, 1945; Olsen, 808 1990; Sher, 1997, among others) 809 The horn core morphology is one of the most useful tools for taxonomic purposes. Leptobos ex 810 gr. SEM features long, spiralized horns emerging posterolaterally. On the contrary, Leptobos ex gr. 811 EV shows crescent moon-shaped cores with no or little torsion, emerging in a more posterior position 812 than in the former group (i.e., smaller angle between the sagittal plane and the core’s mid line) and 813 gently curving inwards. Eobison is characterized by relatively small and stout horns, projecting 814 laterally with strong upward bending and slight anticlockwise torsion. The horn core fragments from 815 Vigna Nuova show a low degree of bending and no sign of torsions, as in Leptobos ex gr. EV. Other 816 skeletal elements from Vigna Nuova do not bear/preserve diagnostic characters and are not discussed 817 furtherly. 818 Despite their scantiness, the analyzed elements evidence that the bovid from the upper layer of 819 Vigna Nuova has affinities with Leptobos ex gr. EV. However, due to the limited number and 820 preservation of the fossils and the strong variation that characterizes this group of large bovids, we 821 prefer to cautiously refer the sample to cf. Leptobos pending for new material from the locality. 822 823 824 CONCLUSIONS 825 826 The fossil mammals from Vigna Nuova, collected more than thirty years ago, have revealed 827 new extremely intriguing data. First of all, the combination of taphonomic, historical, and, to a lesser 828 extent, geological data made it possible to divide the sample into two different assemblages. The one 829 coming from what we have called the “upper layer” of Vigna Nuova (i.e., the only portion still visible 830 in the outcrop), is made up of 18 identifiable remains, plus about 25 unidentifiable ones. The 831 identifiable material is all assigned to cf. Leptobos in the light of morphological affinities with this 832 genus of Villafranchian bovines. In particular, especially as regards the dimensions of the metacarpal 32 833 (Fig. 11), the remains of Vigna Nuova upper layer are similar to those of the most derived forms of 834 the group, namely L. etruscus and L. vallisarni. Remains of the latter species are reported from the 835 nearby locality of Selvella (De Giuli, 1987; Masini et al., 2013) (Fig. 1), whose referral to the Farneta 836 FU (ca. 1.5 Ma) is supported by biochronological data on fossil horse (Alberdi & Palombo, 2013; 837 Cirilli et al., 2020). 838 The assemblage from the “lower layer” (unfortunately not more cropping out at Vigna Nuova) 839 includes at least ten taxa, some of which carrying crucial biochronological information. The co- 840 occurrence of C. ramosus and E. cf. senezensis allows an attribution of the fauna to the second part 841 of the Middle Villafranchian, matching a correlation with the Coste San Giacomo FU in the Italian 842 biochronological framework. This result contrasts with the first hypotheses on a younger age of the 843 fauna (Tasso FU; Argenti, 2004), probably also failing to recognize the division into two assemblages. 844 In Italy, scanty remains of Croizetoceros ramosus are reported only from Montopoli (ca. 2.6. 845 Ma; De Giuli & Heintz, 1974) and Coste San Giacomo (ca. 2.1 Ma; Bellucci et al., 2012, 2014). 846 Therefore, the well-preserved maxilla with upper dentition (SABAP_UMB 129863) from Vigna 847 Nuova allows to confirm the presence of this enigmatic species in the Italian Middle Villafranchian. 848 Similarly, the horse E. senezensis is rare in Italy, having been reported with scanty remains only in 849 Coste San Giacomo and Montecarlo fossil sites, in contrast to the rich French sample from Senèze 850 (Palombo et al., 2017; Cirilli et al., 2021a; Cirilli, 2022). The occurrence of this species in Vigna 851 Nuova provides a new support of the dispersion of this species in the Italian peninsula, which may 852 have been the source for the origin of E. stehlini (Cirilli, 2022). 853 Some primitive characters of the sample of ‘Pseudodama’ from Vigna Nuova suggest that it 854 may be referred to a primitive species of the genus, thus supporting the attribution of the assemblage 855 to the Middle Villafranchian. On the other hand, this hypothesis can only be confirmed by the 856 discovery of more complete fossils. 857 Finally, the discovery of the remains of Canis sp. is noteworthy. Although these predators 858 probably arrived in Europe already in the Early Villafranchian as evidenced by the Vialette record 33 859 (Lacombat et al., 2008; Sotnikova & Rook, 2010), there is no trace of them in Italy before the second 860 part of the Middle Villafranchian. Indeed, the record of Canis sp. from Coste San Giacomo (Bellucci 861 et al., 2021, 2014) marks the first occurrence of the genus Canis in Italy, which is confirmed here by 862 the Vigna Nuova fossils. 863 The recognition of a Middle Villafranchian assemblage in Umbria is novel information, given 864 that the majority of the other LFAs in the region, especially those from the Tiberino Basin, have been 865 attributed to the Late Villafranchian (see Cherin et al., 2019a for a summary). The only exception is 866 represented by the small assemblage from the upper layers of Cava Toppetti near Todi (Perugia), 867 which was referred to the Coste San Giacomo FU (Abbazzi et al., 1997). However, in our opinion, 868 the sample is too small and the fossils too fragmentary to confirm such biochronological conclusion. 869 Our chronological results also open up stimulating research prospects in the stratigraphic field. 870 To date, the continental deposits of the Valdichiana Basin were generically referred to the Calabrian, 871 especially on the basis of biochronological data on mammals (e.g., Selvella, Farneta, Cava Liberatori 872 LFAs, all dated to ca. 1.5. Ma; Masini & Sala, 2007). On the other hand, recent studies focused on 873 the paleoenvironmental history of the marine portion of the Basin, which underwent a complex 874 depositional history at least throughout the Early Pleistocene (Gelasian and Calabrian; Bizzarri & 875 Baldanza, 2020). Further research is needed to identify in the field the deposit that yielded the fossils 876 of the “lower layer” of Vigna Nuova, here dated to the Gelasian (ca. 2.2–2.1 Ma), and to understand 877 its stratigraphic relationships with the surrounding marine and continental units of the Valdichiana 878 Basin. 879 880 881 AUTHORS’ CONTRIBUTIONS 882 M.C. and B.A. conceived the study. F.R., D.T., M.C., and B.A. prepared the fossils. B.A., M.C., 883 and F.R. collected the data. B.A., M.C., J.M.M., L.S., and O.C. made the figures. M.C., B.A., M.B., 884 J.M.M., L.S., and O.C. developed the methods and performed the analyses. B.A. and M.C. wrote the 34 885 final version of the manuscript with input from the other authors. All authors contributed equally to 886 develop the ideas, discussed the results, and reviewed the final version of the manuscript. 887 888 889 ACKNOWLEDGEMENTS 890 891 We are grateful to Giorgio Carnevale and Alberto Collareta for having invited us to join this 892 issue; Angela Baldanza, Roberto Bizzarri, Sergio Gentili, and Fausto Pazzaglia for the useful 893 discussions on the stratigraphy of Vigna Nuova and information on the discovery of the fossils; 894 Leonardo Brustenga, Cristiano Spilinga, and Giovanni Testaferri for their availability and help in 895 accessing the outcrop; Marco Bagli for the translation of the Capellini’s quote; Massimo Delfino, 896 Marco P. Ferretti, Andrea Villa, and Giorgio Carnevale for their opinion on the mustelid tooth; Luca 897 Pulcinelli and Paola Romi (SABAP_UMB) for constant assistance; Dimitris Kostopoulos for the 898 pictures of Greek Croizetoceros; Luca Pandolfi for his thorough review; Luca Bellucci for his review 899 and for kindly providing additional data on canid vertebrae. O.C. gratefully acknowledges Howard 900 University, Washington D.C. for space and resources for his postdoctoral research. L.S. is supported 901 by the FI AGAUR fellowship (ref. 2020 FI_B1 00131) funded by the Secretaria d’Universitats i 902 Recerca de la Generalitat de Catalunya and the European Social Fund. 903 904 905 REFERENCES 906 907 Abbazzi L. (2004). Remarks on the validity of the generic name Praemegaceros Portis 1920, and an 908 overview on Praemegaceros species in Italy. Rendiconti Lincei, 15 (2): 115-132. 909 Abbazzi L., Albianelli A., Ambrosetti P., Argenti P., Basilici G., Gentili S., Masini F. & Pontini 910 M.R. (1997). Paleontological and sedimentological records in Pliocene distal alluvial fan 35 911 deposit at Cava Toppetti (Todi, Central Italy). Bollettino della Società Paleontologica 912 Italiana, 36 (1-2): 5-22. 913 Alberdi M.T. & Palombo M.R. (2013). The late Early to early Middle Pleistocene stenonoid horses 914 from Italy. Quaternary International, 288: 25-44. 915 Alberdi M.T., Cerdeño E., López-Martínez N., Morales J. & Soria M.D. (1997). La fauna 916 villafranquiense de El Rincón-1 (Albacete, Castilla-La Mancha). Estudios Geológicos, 53: 69- 917 93. 918 Alberdi M.T., Ortiz Jaureguízar E. & Prado J.L. (1998). A quantitative review of European 919 stenonoid horses. Journal of Paleontology, 72: 371-387. 920 Ambrosetti P., Bartolomei G., De Giuli C., Ficcarelli G. & Torre D. (1979). La breccia ossifera di 921 Slivia (Aurisina - Sistiana) nel Carso di Trieste. Bollettino della Società Paleontologica 922 Italiana, 18 (2): 207-220. 923 Ambrosetti P., Carboni M.G., Conti M.A., Esu D., Girotti O., La Monica G.B., Landini B. & Parisi 924 G. (1987). Il Pliocene ed il Pleistocene inferiore del bacino del fiume Tevere nell’Umbria 925 meridionale. Geografia Fisica e Dinamica Quaternaria, 10 (1): 10-33. 926 Antón M. (2013). Sabertooth. 268 pp. Indiana University Press, Bloomington. 927 Argenti P. (2004). Plio-Quaternary mammal fossiliferous sites of Umbria (central Italy). Geologica 928 Romana, 37 (2003-2004): 67-78. 929 Azanza B. (2000). Los Cervidae (Artiodactyla, Mammalia) del Mioceno de las Cuencas del Duero, 930 Tajo, Calatayud-Teruel y Levante. Memorias del Museo Paleontologico de la Universidad de 931 Zaragoza, 8: 1-376 932 Azzaroli A. (1947). I cervi fossili della Toscana con particolare riguardo alle specie villafranchiane. 933 Palaeontographia Italica, 43: 45-82. 934 Azzaroli A. (1964). The two Villafranchian horses of the Upper Valdarno. Palaeontographia 935 Italica, 59: 1-12. 36 936 Azzaroli A. (1983). Quaternary mammals and the “end-Villafranchian” dispersal event - a turning 937 point in the history of Eurasia. Palaeogeography, Palaeoclimatology, Palaeoecology, 44: 938 117-139. 939 Azzaroli A. (1992). The cervid genus Pseudodama n. g. in the Villafranchian of Tuscany. 940 Palaeontographia Italica, 79: 1-41. 941 Azzaroli A. (2000). On Equus livenzovensis Baigusheva 1978 and the “stenonid” lineage of equids. 942 Palaeontographia Italica, 87: 1-17. 943 Azzaroli A. & Mazza P. (1992). On the possible origin of the Giant Deer genus Megaceroides. 944 Rendiconti Lincei, 3 (1): 23-32. 945 Azzaroli A., De Giuli C., Ficcarelli G. & Torre D. (1988). Late Pliocene to early mid-Pleistocene 946 mammals in Eurasia: faunal succession and dispersal events. Palaeogeography, 947 Palaeoclimatology, Palaeoecology, 66 (1-2): 77-100. 948 Bajgusheva V.S. (1978). Krupnaja Loshad Khaprovskogo Kompleksa is alluvija 949 SeveroVostochnoto Priazovija. Invesija Servo-Kavkazkogo Nauchnogo Zentra Vychei Shkoly, 950 1: 98-102. 951 Baldanza A., Bizzarri R., Famiani F., Garassino A., Pasin, G., Cherin M. & Rosatini F. (2018). The 952 early Pleistocene whale-fall community of Bargiano (Umbria, Central Italy): Paleoecological 953 insights from benthic foraminifera and brachyuran crabs. Palaeontologia Electronica, 21: 1- 954 27. 955 Ballesio R. (1963). Monographie d’un Machairodus du gisement villafranchien de Senèze: 956 Homotherium crenatidens. Travaux et Documents des Laboratoires de Géologie de Lyon, 9: 957 1-129. 958 Barchi M.R. & Marroni M. (2007). Note illustrative della Carta Geologica d’Italia alla scala 959 1:50.000. Foglio 310-Passignano sul Trasimeno. 196 pp. APAT, Dipartimento Difesa del 960 Suolo, Servizio Geologico d’Italia. 37 961 Bartolini-Lucenti S. & Rook L. (2016). A review on the Late Villafranchian medium-sized canid 962 Canis arnensis based on the evidence from Poggio Rosso (Tuscany, Italy). Quaternary 963 Science Reviews, 151: 58-71. 964 Bartolini-Lucenti S., Alba D.M., Rook L., Moya-Sola S. & Madurell-Malapeira J. (2017). Latest 965 Early Pleistocene wolf-like canids from the Iberian Peninsula. Quaternary Science Reviews, 966 162: 12-25. 967 Bartolini-Lucenti S., Bukhsianidze M., Martínez-Navarro B. & Lordkipanidze D. (2020). The wolf 968 from Dmanisi and augmented reality: review, implications, and opportunities. Frontiers in 969 Earth Science, 8: 131. 970 Bartolini-Lucenti S., Cirilli O., Pandolfi L., Bernor R.L., Bukhsianidze M., Carotenuto F., 971 Lordkipanidze D., Tsikaridze N. & Rook L. (2022). Zoogeographic significance of Dmanisi 972 large mammal assemblage. Journal of Human Evolution, 163: 103125. 973 Basilici G. (1997). Sedimentary facies in an extensional and deep lacustrine depositional system: 974 the Pliocene Tiberino Basin, Central Italy. Sedimentary Geology, 109: 73-94. 975 Behrensmeyer A.K. (1978). Taphonomic and ecologic information from bone weathering. 976 Paleobiology, 4 (2): 150-162. 977 Bellucci L., Mazzini I., Scardia G., Bruni L., Parenti F., Segre A.G., Segre Naldini E. & Sardella R. 978 (2012). The site of Coste San Giacomo (Early Pleistocene, central Italy): palaeoenvironmental 979 analysis and biochronological overview. Quaternary International, 267: 30-39. 980 Bellucci L., Bona F., Corrado P., Magri D., Mazzini I., Parenti F., Scardia G. & Sardella R. (2014). 981 Evidence of late Gelasian dispersal of African fauna at Coste San Giacomo (Anagni Basin, 982 central Italy): Early Pleistocene environments and the background of early human occupation 983 in Europe. Quaternary Science Review, 96: 72-85. 984 Bernor R.L., Tobien H., Hayek L.A.C. & Mittmann H.W. (1997). Hippotherium primigenium 985 (Equidae, Mammalia) from the late Miocene of Höwenegg (Hegau, Germany). Andrias, 10: 986 1-230. 38 987 Bernor R.L., Cirilli O., Wang S.Q. & Rook L. (2018). Equus cf. livenzovenzis from Montopoli, Italy 988 (early Pleistocene; MN16b; ca. 2.6 Ma). Bollettino della Società Paleontologica Italiana, 57 989 (3): 203-216. 990 Bernor R.L., Cirilli O., Jukar A.M., Potts R., Bukhsianidze M. & Rook L. (2019). Evolution of 991 Early Equus in Italy, Georgia, the Indian Subcontinent, East Africa, and the Origins of 992 African Zebras. Frontiers in Ecology and Evolution, 7: 166. 993 Bernor R.L., Cirilli O., Bukhsianidze M., Lordkipanidze D. & Rook L. (2021). The Dmanisi Equus: 994 Systematics, biogeography and paleoecology. Journal of Human Evolution, 158: 103051. 995 Bizzarri R. & Baldanza A. (2020). Integrated stratigraphy of the marine early Pleistocene in 996 Umbria. Geosciences, 10 (9): 371. 997 Bout P. & Azzaroli A. (1952). Stratigraphie et Faune du creux de Peirolles pres Perrier (Puy-de- 998 Dôme). Annales de Paléontologie, 38: 3-22. 999 Bowdich T.E. (1821). An analysis of the Natural Classification of Mammalia for the use of students 1000 and travellers. 184 pp. J. Smith, Paris. 1001 Breda M. (2015). The early Middle Pleistocene fallow deer Dama roberti: new insight on species 1002 morphology from a complete postcranial skeleton from Valdemino (northwestern Italy). 1003 Geological Journal, 50 (3): 257-270. 1004 Breda M. & Lister A.M. (2013). Dama roberti, a new species of deer from the early Middle 1005 Pleistocene of Europe, and the origins of modern fallow deer. Quaternary Science Reviews, 1006 69: 155-167. 1007 Breda M., Peretto C. & Thun Hohenstein U. (2015). The deer from the early Middle Pleistocene site 1008 of Isernia la Pineta (Molise, Italy): revised identifications and new remains from the last 15 1009 years of excavation. Geological Journal, 50 (3): 290-305. 1010 Breda M., Kahlke R.D. & Lister A.M. (2020). New results on cervids from the Early Pleistocene 1011 site of Untermassfeld. In Kahlke R.D. (ed) The Pleistocene of Untermassfeld near Meiningen 1012 (Thüringen, Germany), Part 4: 1197-1249. 39 1013 Burchak-Abramovich N.I. & Vekua A.K. (1994). Novyy pryedstavityel’ lyeystotsyenobykh bykov 1014 izz vostochnoy Gruzii. Voprosy teriologii Rossiyskaya Akademiya Nauk, 1: 253-261 (in 1015 Russian). 1016 Capellini G. (1888). Sui resti di Mastodon arvernensis recentemente scoperti a Spoleto, Pontremoli 1017 e Castrocaro. Memorie della Regia Accademia delle Scienze, 4 (9): 251-258. 1018 Capellini G. (1889). Sulla scoperta di una caverna ossifera a Monte Cucco. Bollettino della Società 1019 Paleontologica Italiana, 8: 274-278. 1020 Capellini G. (1890). Comunicazione sui vertebrati fossili rinvenuti nella grotta di Monte Cucco 1021 presso Fabriano. Bollettino della Società Paleontologica Italiana, 9: 743-744. 1022 Cherin M., Bertè D.F., Rook L. & Sardella R. (2014). Re-defining Canis etruscus (Canidae, 1023 Mammalia): a new look into the evolutionary history of Early Pleistocene dogs resulting from 1024 the outstanding fossil record from Pantalla (Italy). Journal of Mammalian Evolution, 21: 95- 1025 110. 1026 Cherin M., Iurino D.A., Willemsen G. & Carnevale G. (2016). A new otter from the Early 1027 Pleistocene of Pantalla (Italy), with remarks on the evolutionary history of Mediterranean 1028 Quaternary Lutrinae (Carnivora, Mustelidae). Quaternary Science Reviews, 135: 92-102. 1029 Cherin M., Sorbelli L., Crotti M., Iurino D.A., Sardella R. & Souron A. (2018). New material of Sus 1030 strozzii (Suidae, Mammalia) from the Early Pleistocene of Italy and a phylogenetic analysis of 1031 suines. Quaternary Science Reviews, 194: 94-115. 1032 Cherin M., Azzarà B., Breda M., Brobia Ansoleaga A., Buzi C., Pandolfi L. & Pazzaglia F. (2019a). 1033 Large mammal remains from the Early Pleistocene site of Podere San Lorenzo (Perugia, 1034 central Italy). Rivista Italiana di Paleontologia e Stratigrafia, 125 (2): 489-515. 1035 Cherin M., D’Allestro V. & Masini F. (2019b). New bovid remains from the Early Pleistocene of 1036 Umbria (Italy) and a reappraisal of Leptobos merlai. Journal of Mammalian Evolution, 26: 1037 201-224. 40 1038 Cherin M., Alba D.M., Crotti M., Menconero S., Moullé P.É., Sorbelli L. & Madurell-Malapeira J. 1039 (2020). The post-Jaramillo persistence of Sus strozzii (Suidae, Mammalia) in Europe: new 1040 evidence from the Vallparadís Section (NE Iberian Peninsula) and other coeval sites. 1041 Quaternary Science Reviews, 233: 106234. 1042 Cherin M., Cirilli O., Azzarà B. & Bernor R.L. (2021). Equus stenonis (Equidae, Mammalia) from 1043 the Early Pleistocene of Pantalla (Italy) and the dispersion of stenonian horses in Europe. 1044 Bollettino della Società Paleontologica Italiana, 60 (1): 1-18. 1045 Cirilli O. (2022). Equus stehlini Azzaroli, 1964 (Perissodactyla, Equidae). A revision of the most 1046 enigmatic horse from the Early Pleistocene of Europe, with new insights on the evolutionary 1047 history of European medium- and small-sized hoses. Rivista Italiana di Paleontologia e 1048 Stratigrafia, 128 (1): 241-265. 1049 Cirilli O., Pandolfi L. & Bernor R.L. (2020). The Villafranchian perissodactyls of Italy: knowledge 1050 of the fossil record and future research perspectives. Geobios, 63: 1-21. 1051 Cirilli O., Pandolfi L., Rook L. & Bernor R.L. (2021a). Evolution of Old World Equus and origin of 1052 the zebra-ass clade. Scientific Reports, 11 (1): 10156. 1053 Cirilli O, Saarinen J., Pandolfi L., Rook L. & Bernor R.L. (2021b). An updated review on Equus 1054 stenonis (Mammalia, Equidae): new implications for the European Early Pleistocene Equus 1055 taxonomy and paleoecology, and remarks on the Old World Equus evolution. Quaternary 1056 Science Reviews, 269: 107155. 1057 Cirilli O., Bernor R.L. & Rook L. (2021c). New insights on the Early Pleistocene equids from 1058 Roca-Neyra (France, central Europe): implications for the Hipparion LAD and the Equus 1059 FAD in Europe. Journal of Paleontology, 95 (2): 406-425. 1060 Clerici E. (1894). Sul ritrovamento del castoro nelle ligniti di Spoleto. Bollettino della Società 1061 Paleontologica Italiana, 13 (169): 199-202. 1062 Clerici E. (1895). Ritrovamento di resti di tapiro nelle ligniti di Spoleto. Bollettino della Società 1063 Paleontologica Italiana, 14: 296-297. 41 1064 Cocchi I. (1867). L’uomo fossile dell’Italia Centrale. Memorie della Società Italiana di Scienze 1065 Naturali, 2: 1-180. 1066 Collettini C., De Paola N., Holdsworth R.E. & Barchi M.R. (2006). The development and behaviour 1067 of low-angle normal faults during Cenozoic asymmetric extension in the Northern Apennines, 1068 Italy. Journal of Structural Geology, 28 (2): 333-352. 1069 Colombero S., Pavia M. & Rook L. (2012). Pannonictis nestii (Galictinae, Mustelidae), a new 1070 element in the vertebrate association of the human site of Pirro Nord (Italy, Early 1071 Pleistocene). Geodiversitas, 34 (3): 665-681. 1072 Cope E.D. (1892). A contribution to the vertebrate paleontology of Texas. Proceedings of the 1073 American Philosophical Society, 30: 123-131. 1074 Croitor R. (2006). Early Pleistocene small-sized deer of Europe. Hellenic Journal of Geosciences, 1075 41 (1): 89-117. 1076 Croitor R. (2018). Plio-Pleistocene deer of Western Palearctic: Taxonomy, Systematics, Phylogeny. 1077 142 pp. Institute of Zoology of the Academy of Sciences of Moldova, Chişinău. 1078 Croizet J.B. & Jobert A. (1828). Recherches sur les ossemens fossils du Département du Puy-de- 1079 Dôme. 224 pp. Thibaut-Landriot, Paris. 1080 Cuvier G. (1821). Recherches sur les ossemens fossiles, où l’on rétablit les caractères de plusiers 1081 animaux dont les révolutions du globe ont détruit les espéces. Tome 1. 334 pp. Dufour et 1082 d’Ocagne, Paris & Amsterdam. 1083 Cuvier G. (1824). Recherches sur les ossemens fossiles, où l’on rétablit les caractères de plusiers 1084 animaux dont les révolutions du globe ont détruit les espéces. Tome 5, partie 2. 547 pp. 1085 Dufour et d’Ocagne, Paris & Amsterdam. 1086 de Bonis L. (1976). Un félidé á longues canines de la colline de Perrier (Puy-de-Dôme): ses 1087 rapports avec les félinés machairodontes. Annales de Paléontologie, Vertébrés, 62 (2): 159- 1088 198. 42 1089 De Giuli C. (1987). Late Villafranchian faunas in Italy: the Selvella Local Fauna in southern Chiana 1090 Valley - Umbria. Palaeontographia Italica, 74: 11-50. 1091 De Giuli D. & Heintz E. (1974). Croizetoceros ramosus (Cervidae, Artiodactyla, Mammalia) de 1092 Montopoli, nouvel élément de la faune villafranchienne d'Italie. Atti della Società Toscana di 1093 Scienze Naturali, 81: 241-252. 1094 Delafond F. & Depéret C. (1893). Les terrains tertiaires de la Bresse et leurs gites de lignites et de 1095 minerais de fer. Etudes des Gites Minéraux de la France. 332 pp. Imprimerie Nationale, Paris. 1096 Del Campana D. (1913) I cani pliocenici di Toscana. Palaeontographia italica, 19: 189–254. 1097 De Lumley H., Kahlke H.D., Moigne A.M. & Moullé P.E. (1988). Les faunes de grandes 1098 Mammifères de la grotte du Vallonet Roquebrune-Cap Martin, Alpes-Maritimes. 1099 L’Anthropologie, 92: 465-496. 1100 Dubois E. (1904). On an equivalent of the Cromer Forest-bed in the Netherlands. Proceedings 1101 Koniklijke Akademie der Wetenschapen, Section Science, 7 (3): 214-222. 1102 Duvernois M.P. (1989). Les Leptobos (Mammalia, Artiodactyla) du Villafranchien inférieur 1103 d’Europe occidentale. Comptes rendus de l’Académie des sciences. Série 2, 309 (7): 769-775. 1104 Duvernois M.P. (1990). Les Leptobos (Mammalia, Artiodactyla) du Villafranchien d'Europe 1105 occidentale. Systématique, évolution, biostratigraphie, paléoécologie. Travaux et Documents 1106 des Laboratoires de Géologie de Lyon, 113 (1): 3-213. 1107 Duvernois M.P. (1992). Mise au point sur le genre Leptobos (Mammalia, Artiodactyla, Bovidae); 1108 implications biostratigraphiques et phylogénétiques. Geobios, 25 (1): 155-166. 1109 Eisenmann V., Alberdi M.T., De Giuli C. & Staesche U. (1988). Studying Fossil Horses. Volume I 1110 - Methodology. In Woodburne M.O. & Sondaar P.Y. (eds), Collected Papers after the New 1111 York International Hipparion Conference 1981. Brill, Leiden: 1-71. 1112 Fabrini E. (1890). I Machairodus (Megantereon) del Val d’Arno superiore. Bollettino Comitato 1113 Geologico d'Italia, 21: 121-144. 43 1114 Falconer H. (1868). Paleontological memoirs and notes, mastodon, elephant, rhinoceros, ossiferous 1115 caves, primeval man and his contemporaries. Volume 2. 675 pp. Robert Hardwicke, London. 1116 Faure M. & Guérin C. (1983). Le Sus scrofa (Mammalia, Artiodactyla, Suidae) du gisement 1117 pléistocène supérieur de Jaurens à Nespouls, Corrèze, France. Publications du musée des 1118 Confluences, 21 (1): 45-63. 1119 Ficcarelli G. (1979). The Villafranchian machairodonts of Tuscany. Paleontographica Italica, 71: 1120 17-256. 1121 Fidolini F., Ghinassi M., Magi M., Papini M. & Sagri M. (2013). The Plio-Pleistocene fluvio- 1122 lacustrine Upper Valdarno Basin (central Italy): stratigraphy and basin fill evolution. Italian 1123 Journal of Geosciences, 132 (1): 13-32. 1124 Florindo F., Marra F., Angelucci D.E., Biddittu I., Bruni L., Florindo F., Gaeta M., Guillou H., 1125 Jicha B., Macrì P., Morigi C., Nomade S., Parenti F., Pereira A. & Grimaldi S. (2021). 1126 Environmental evolution, faunal and human occupation since 2 Ma in the Anagni basin, 1127 central Italy. Scientific Reports, 11: 7056. 1128 Forsyth Major C.I. (1877). Considerazioni sulla fauna dei Mammiferi pliocenici e postpliocenici 1129 della Toscana. III. Cani fossili del Val d’Arno superiore e della Valle dell’Era. Memorie della 1130 Società Toscana di Scienze Naturali, 3: 207-227. 1131 Forsyth Major C.I. (1881). Studi sugli avanzi pliocenici del genere Sus. Atti della Società Toscana 1132 di Scienze Naturali, Processi Verbali, 2: 227. 1133 Freudenberg W. (1914). Die Säugetiere des älteren Quartars von Mitteleuropa. Geology and 1134 Paleontology. Abh, 12, 533-549. 1135 Fujita M., Kawamur, Y. & Murase N. (2000). Middle Pleistocene wild boar remains from NT cave, 1136 Niimi, Okayama prefecture, west Japan. Journal of Geosciences Osaka City University, 43: 1137 57-95. 1138 Gentili S. (1993). I paleoambienti plio-pleistocenici continentali umbri: caratteri sistematici, 1139 pleoecologici e tafonomici delle mammalofaune fossili. CNR, Comitato 05, Scienze 44 1140 Geologiche e Minerarie, “Sedimentologia e stratigrafia delle formazioni pliopleistoceniche”. 1141 23 pp. Relazione Borsa di Studio 1991/93. 1142 Ghinassi M., Abbazzi L., Esu D., Gaudant J. & Girotti O. (2005). Facies analysis, stratigraphy and 1143 paleontology (molluscs and vertebrates) in the upper Pliocene sandy flood-basin deposits of 1144 the Upper Valdarno basin (Northern Apennines). Rivista Italiana di Paleontologia e 1145 Stratigrafia, 111 (3): 467-487. 1146 Girotti O., Capasso Barbato L., Esu D., Gliozzi E., Kotsakis T., Martinetto E., Petronio C., Sardella 1147 R. & Squazzini E. (2003). The section of Torre Picchio (Terni, Umbria, Central Italy): a 1148 Villafranchian site rich in vertebrates, molluscs, ostracods and plants. Rivista Italiana di 1149 Paleontologia e Stratigrafia, 109 (1): 77-98. 1150 Goldfuss G.A. (1820). Handbuch der Zoologie. Erste Abteilung. 696 pp. Schrag, Nürnberg. 1151 Gray J.E. (1821). On the natural arrangement of vertebrose animals. London Medical Repository, 1152 15: 296-310. 1153 Hays I. (1834). Descriptions of the specimens of inferior maxillary bones of mastodons in the 1154 cabinet of the American Philosophical Society, with remarks on the genus Tetracaulodon 1155 (Godman). Transactions of the American Philosophical Society, 4: 317-339. 1156 Heintz E. (1970). Les cervidés villafranchiens de France et d’Espagne. Memoires du Museum 1157 National d'Histoire Naturelle Nouvelle Serie, Serie C, Sciences de la Terre, 22 (1): 303. 1158 Hemmer H. (2001). Die feliden aus dem Epivillafranchium von Untermassfeld. In Kahlke, R.D. 1159 (ed), Das Pleistozän von Untermassfeld bei Meiningen (Thüringen), Teil 3: 699-782. 1160 Iannucci A., Sardella R., Strani F. & Mecozzi B. (2020). Size shifts in late Middle Pleistocene to 1161 Early Holocene Sus scrofa (Suidae, Mammalia) from Apulia (southern Italy): 1162 ecomorphological adaptations? Hystrix, 31 (1): 10-20. 1163 Illiger C.D. (1811). Prodromus Systematis Mammalium et Avium Additis Terminis Zoographicis 1164 Uttriusque Classis. 301 pp. Salfeld, Berlin. 45 1165 Kostopoulos D.S. (1996). The Plio-Pleistocene Artiodactyls of Macedonia (Greece) – systematics, 1166 palaeoecology, biochronology, biostratigraphy. PhD thesis. 671 pp. Aristotle University of 1167 Thessaloniki, Thessaloniki (in Greek). 1168 Kostopoulos D.S. & Athanassiou A. (2005). In the shadow of bovids: suids, cervids and giraffids 1169 from the Plio-Pleistocene of Greece. Quaternaire, 179-190. 1170 Lacombat F., Abbazzi L., Ferretti M.P., Martínez-Navarro B., Moullé P.E., Palombo M.R., Rook 1171 L., Turner A. & Valli A.M.F. (2008). New data on the Early Villafranchian fauna from 1172 Vialette (Haute-Loire, France) based on the collection of the Crozatier Museum (Le Puy-en- 1173 Velay, Haute-Loire, France). Quaternary International, 179 (1): 64-71. 1174 Leidy J. (1853). Description of an extinct species of American lion: Felis atrox. Transactions of the 1175 American Philosophical Society, 10: 319-321. 1176 Linnaeus C. (1758). Systema naturae per regna tria naturae, secundum Classes, Ordines, Genera, 1177 Species cum characteribus, differentiis, synonimis, locis. 824 pp. Laurentii Salvii, Stockholm. 1178 Lotti B. (1917). L’antico Lago Tiberino e le ligniti dell’Umbria. La Miniera Italiana, 7: 1-8. 1179 Maniakas I. & Kostopoulos D.S. (2017). Morphometric-palaeoecological discrimination between 1180 Bison populations of the western Palaearctic. Geobios, 50 (2): 155-171. 1181 Martinetto E., Bertini A., Basilici G., Baldanza A., Bizzarri R., Cherin M. Gentili S. & Pontini M.R. 1182 (2014). The plant record of the Dunarobba and Pietrafitta sites in the Plio-Pleistocene 1183 palaeoenvironmental context of central Italy. Alpine and Mediterranean Quaternary, 27 (1): 1184 29-72. 1185 Martínez-Navarro B., Madurell-Malapeira J., Ros-Montoya S., Espigares M.P., Medin T., Hortola 1186 P. & Palmqvist P. (2015). The Epivillafranchian and the arrival of pigs into Europe 1187 Quaternary International, 389: 131-138. 1188 Martini I.P. & Sagri M. (1993). Tectono-sedimentary characteristics of Late Miocene-Quaternary 1189 extensional basins of the Northern Apennines, Italy. Earth-Science Reviews, 34 (3): 197-233. 46 1190 Masini F. (1989). I bovini villafranchiani dell’Italia. PhD Thesis. 152 pp. Università di Modena- 1191 Bologna-Firenze-Roma. 1192 Masini F. & Sala, B. (2007). Large-and small-mammal distribution patterns and chronostratigraphic 1193 boundaries from the Late Pliocene to the Middle Pleistocene of the Italian peninsula. 1194 Quaternary International, 160 (1): 43-56. 1195 Masini F., Palombo M.R. & Rozzi R. (2013). A reappraisal of the Early to Middle Pleistocene 1196 Italian Bovidae. Quaternary International, 288: 45-62. 1197 Mayda S., Alçiçek M.C., Demirel F.A. & Alçiçek H. (2017). The first record of a Villafranchian 1198 small cervid “Croizetoceros” from Turkey. 15thRCMNS Congress (Regional Committee on 1199 Mediterranean Neogene Stratigraphy), 3-6 September, Athens, Greece. 1200 Meli R. (1895). Notizie sui resti di Mammiferi fossili rinvenuti in località italiane. Bollettino della 1201 Società Geologica Italiana, 14: 148-164. 1202 Merla G. (1949). I Leptobos Rütim. italiani. Palaeontographica Italica, 46: 41-155. 1203 Nesti F. (1825). Sulla nuova specie di elefante fossile del Valdarno all’Illustrissimo Sig. Dott. Prof. 1204 Ottaviano Targioni Tozzetti (Lettere sopra alcune ossa fossili del Valdarno non per anco 1205 descritte). Nuovo Giornale di Lettere, 11: 195-216. 1206 Nomade S., Pastre J.F., Guillou H., Faure M., Guérin C., Delson E., Debard E., Voinchet P. & 1207 Messager E. (2014). 40Ar/39Ar constraints on some French landmark late Pliocene to early 1208 Pleistocene large mammalian paleofaunas: paleoenvironmental and paleoecological 1209 implications. Quaternary Geochronology, 21: 2-15. 1210 Olsen S.J. (1990). Fossil ancestry of the yak, its cultural significance and domestication in Tibet. 1211 Proceedings of the Academy of Natural Sciences of Philadelphia, 142: 73-100. 1212 Owen R. (1848). Description of the teeth and portions of jaws of two extinct anthracotheroid 1213 quadrupeds (Hyopotamus vectianus and Hyop. bovinus) discovered by the Marchioness of 1214 Hastings in the Eocene deposits of the N.W. coast of the Isle of Wight: with an attempt to 47 1215 develop Cuvier’s idea of the classification of the pachyderms by the number of their toes. 1216 Quarterly Journal of the Geological Society of London, 4: 103-141. 1217 Palombo M.R. & Alberdi M.T. (2017). Light and shadows in the evolution of South European 1218 stenonoid horses. Fossil Imprint, 73 (1-2): 115-140. 1219 Palombo M.R., Alberdi M.T., Bellucci L. & Sardella R. (2017). An intriguing middle-sized horse 1220 from Coste San Giacomo (Anagni Basin, central Italy). Quaternary Research, 87 (2): 347- 1221 362. 1222 Pantanelli D. (1886). Monografia degli strati pontici del Miocene superiore nell’Italia settentrionale 1223 e centrale. Memorie della Reale Accademia di Scienze, Lettere ed Arti di Modena, 4: 127-231. 1224 Pascucci V., Merlini S. & Martini I.P. (1999). Seismic stratigraphy of the Miocene-Pleistocene 1225 sedimentary basins of the Northern Tyrrhenian Sea and western Tuscany (Italy). Basin 1226 Research, 11 (4): 337-356. 1227 Petronio C. & Pandolfi L. (2011). First occurrence of the genus Arvernoceros Heintz 1970 from the 1228 late Early Pleistocene of Italy. Rivista Italiana di Paleontologia e Stratigrafia, 117 (3): 501- 1229 508. 1230 Petronio C., Argenti P., Caloi C., Esu D., Girotti O. & Sardella R. (2003). Updating Villafranchian 1231 mollusc and mammal faunas in Umbria and Latium (Central Italy). Geologica Romana, 36: 1232 369-387. 1233 Pfeiffer T. (2016). Deer from the Pliocene site of Bad Deutsch-Altenburg 26 (Lower Austria, 1234 Leithagebirge) - Conclusions based on skeletal morphology. Annalen des Naturhistorischen 1235 Museums in Wien, Serie A, 118: 133-173. 1236 Pilgrim G.E. (1939). The fossil Bovidae of India. Palæontologia Indica. New series, 26: 1-356. 1237 Pomel A. (1853). Catalogue methodique et descriptif des Vertebres fossiles decouverts dans le 1238 bassin hydrographique superieur de la Loire, et surtout dans la vallee de son affluent 1239 principal, l’Allier. 193 pp. J.B. Bailliere, Paris. 48 1240 Pons Moyà J. (1987). Los carnivoros, Mammalia de Venta Micena, Granada, España. Paleontologia 1241 i Evolució, Memòria especial, 1: 109-128. 1242 Prat F. (1964). Contribution à la classification des équidés Villafranchiens. Procès-verbaux de la 1243 Société Linnéenne de Bordeaux, 101: 14-32. 1244 Prat F. (1980). Les equidés villafranchiens en France, genre Equus. Cahiers du Quaternaire, 2: 1- 1245 290. 1246 Qiu Z.X., Deng T. & Wang B.Y. (2004). Early Pleistocene mammalian fauna from Longdan, 1247 Dongxiang, Gansu, China. Palaeontologia Sinica New Series C, 27: 162-164. 1248 R Core Team. (2013). R: a language and environment for statistical computing. R Foundation for 1249 Statistical Computing, Vienna, Austria. 1250 Regione Umbria (2013). Cartografia geologica informatizzata vettoriale della Regione Umbria in 1251 http://www.territorio.regione.umbria.it/ and in http://dati.umbria.it/dataset/carta-geologica- 1252 dell-umbria. Regione Umbria, Servizio Geologico. 1253 Rook L., Bernor R.L., Avilla L.D., Cirilli O., Flynn L., Jukar A.M., Sanders W., Scott E. & Wang 1254 X. (2019). Mammal biochronology (Land Mammal Ages) around the world from the late 1255 Miocene to middle Pleistocene and major events in horse evolutionary history. Frontiers in 1256 Ecology and Evolution, 7: 278. 1257 Rütimeyer L. (1877-1878). Die Rinder der Tertiär-Epoche nebst vorstudien zu einer natürlichen 1258 Geschichte der Antilopen. Paläontologische Gesellschaft, 4: 1-72. 1259 Sala B. (1986). Bison schoetensacki Freud. from Isernia La Pineta (early mid Pleistocene, Italy) and 1260 revision of the European species of bison. Palaeontographica Italica, 74: 113-170. 1261 Sardella R. & Iurino D.A. (2012). The latest Early Pleistocene sabertoothed cat Homotherium 1262 (Felidae, Mammalia) from Monte Peglia (Umbria, central Italy). Bollettino della Società 1263 Paleontologica Italiana, 51 (1): 15-22. 1264 Sardella R., Di Stefano G. & Petronio C. (1995). The Villafranchian mammal faunas from the Tiber 1265 River Basin (Umbria, Central Italy). Il Quaternario, 8: 509-514. 49 1266 Sardella R., Mazzini I., Giustini F., Mecozzi B., Brilli M., Iurino D.A., Lembo G., Muttillo B., 1267 Massussi M., Sigari D., Tucci S. & Voltaggio M. (2018). Grotta Romanelli (Southern Italy, 1268 Apulia): legacies and issues in excavating a key site for the Pleistocene of the Mediterranean. 1269 Rivista Italiana di Paleontologia e Stratigrafia, 124 (2): 247-264. 1270 Sher A.V. (1997). An Early Quaternary bison population from Untermaßfeld: Bison menneri sp. 1271 nov. In Kahlke R.D. (ed), Das Pleistozan von Untermaßfeld bei Meiningen (Thüringen). Teil 1272 1: 101-180. 1273 Sismonda E. (1846). Descrizione dei pesci e dei crostacei fossili nel Piemonte. 24 pp. Stamperia 1274 Reale, Torino. 1275 Skinner M.R. & Kaisen O.C. (1946). The fossil Bison of Alaska and preliminary revision of the 1276 genus. Bulletin of the American Museum of Natural History, 89 (3): 123-256. 1277 Soergel W. (1928). Ein kleiner Wolf aus den Kiesen von Süßenborn. Zeitschrift der Deutschen 1278 Geologischen Gesellschaft, 80: 227-237. 1279 Sorbelli L., Alba D.M., Cherin M., Moullé P.É., Brugal J.P. & Madurell-Malapeira J. (2021). A 1280 review on Bison schoetensacki and its closest relatives through the early-Middle Pleistocene 1281 transition: Insights from the Vallparadís Section (NE Iberian Peninsula) and other European 1282 localities. Quaternary Science Reviews, 261: 106933. 1283 Sotnikova M. & Rook L. (2010). Dispersal of the Canini (Mammalia, Canidae: Caninae) across 1284 Eurasia during the late Miocene to early Pleistocene. Quaternary International, 212 (2): 86- 1285 97. 1286 Stefanescu S. (1924). Sur la presence de l’Elephas planifrons et de trois mutations de l’Elephas 1287 antiquus dans les couches géologiques de Roumanie. Comptes rendus de l’Académie des 1288 sciences, 179: 1418-1419. 1289 Teilhard de Chardin P. & Piveteau J. (1930). Les mammifères fossiles de Nihowan (Chine). 1290 Annales de Paléontologie, 19: 1-134. 50 1291 Tong H.W., Chen X. & Zhang B. (2016). New fossils of Bison palaeosinensis (Artiodactyla, 1292 Mammalia) from the steppe mammoth site of early Pleistocene in Nihewan basin, China. 1293 Quaternary International, 445: 450-468. 1294 Torre D. (1967). I cani villafranchiani della Toscana. Palaeontographia Italica, 63: 113-138. 1295 Torre D., Ficcarelli G., Masini F., Rook L. & Sala B. (1992). Mammal dispersal events in the early 1296 Pleistocene of Western Europe. Courier Forschungsinst Senckenberg, 153: 51-58. 1297 Torre D., Abbazzi L., Bertini A., Fanfani F., Ficcarelli G., Masini F., Mazza P. & Rook L. (2001). 1298 Structural changes in Italian late Pliocene - Pleistocene large mammal assemblages. Bollettino 1299 della Società Geologica Italiana, 40 (2): 303-306. 1300 Tuccimei G. (1891). Sui mammiferi fossili dell’Umbria e della provincia di Roma. Atti 1301 dell’Accademia Pontificia de’ Nuovi Lincei, 44: 274-275. 1302 Tuccimei G. (1895). Sopra i resti fossili di mammiferi trovati alla Villa Spinola presso Perugia. Atti 1303 dell’Accademia Pontificia de’ Nuovi Lincei, 48: 44-46. 1304 Valli A.M.F. (2004). Les Cervidae du gisement Pliocène supérieur (Villafranchien moyen) de Saint- 1305 Vallier (Drôme, France). Geobios, 37: S191-S232. 1306 Venables W.N. & Ripley B.D. (2008). Modern Applied Statistics with R. 498 pp. Springer-Verlag, 1307 New York. 1308 Van der Made J. (1999). Ungulates from Atapuerca TD6. Journal of Human Evolution, 37 (3-4): 1309 389-413. 1310 Van der Made J., Rosell J. & Blasco R. (2017). Faunas from Atapuerca at the Early–Middle 1311 Pleistocene limit: The ungulates from level TD8 in the context of climatic change. 1312 Quaternary International, 433: 296-346. 1313 Verri A. (1879). Avvenimenti nell’interno del Bacino del Tevere antico durante e dopo il periodo 1314 pliocenico. Atti della Società Italiana di Scienze Naturali, 21: 149-180. 1315 von Meyer H.V. (1829). Taschenbuch für die gesammte Mineralogie. Zeitschrift für Mineralogie, 1316 23 (1): 150-152. 51 1317 von Reichenau W. (1915). Beiträge zur näheren Kenntnis fossiler Pferde aus deutschen Pleistozän. 1318 Abhandlungen der Grossherzoglich Hessischen Geologischen Landesanstalt zu Darmstadt, 7: 1319 1. 1320 Wickham H. (2016). ggplot2: Elegant Graphics for Data Analysis. 213 pp. Springer-Verlag, New 1321 York. 1322 Wüst E. (1900). Das Pliozän und das älteste Pleistozän Thüringens. Nördlich vom Thüringer Walde 1323 und Westlich von der Saale. Abhandlungen der Naturforscgung Gessellschaften zu Halle, 23: 1324 1-352. 1325 Zeder M.A. & Lemoine X. (2020). Factors affecting molar size in Sus scrofa. Journal of 1326 Archaeological Science, 124: 105266. 52 1327 CAPTIONS 1328 1329 Fig. 1 - (color online) Location of the study area. a) The extensional basins cited in the text are 1330 highlighted. Red star: Vigna Nuova; black star: Selvella. b) Detail of the Vigna Nuova outcrop. 1331 1332 Fig. 2 - (color online) Taphonomy of the Vigna Nuova assemblages. a) Detail of the inner part of a 1333 proboscidean bone (SABAP_UMB 22.3.264) showing heavy mineralization and filling of trabeculae 1334 by minerals and organic matter. b) Distribution of weathering stages (following Behrensmeyer, 1978) 1335 in the fossils from the two fossiliferous layers of Vigna Nuova. 1336 1337 Fig 3 - (color online) Carnivorans from Vigna Nuova (Italy), lower layer. a) cf. Megantereon: right 1338 upper canine (SABAP_UMB 129842) in labial (a1) and distal (a2) views. b-d) Canis sp.: cervical 1339 vertebrae 3–7 (SABAP_UMB 129850) in dorsal view (b); left rib (SABAP_UMB 129875) in lateral 1340 view (c); thoracic vertebrae 2–3 (SABAP_UMB 129875) in lateral view (d). Scale bar: 2 cm. e-f) 1341 Comparison between the vertebrae C3-C4 from Vigna Nuova (SABAP_UMB 129850) (e) and those 1342 of Canis etruscus from Olivola (IGF 4523) (f) in left lateral view. Vertebrae are shown in anatomical 1343 connection, as they were discovered; the relative position of the vertebrae in the two specimens is 1344 slightly different (C4 is rotated clockwise with respect to C3 in SABAP_UMB 129850, while the two 1345 vertebrae are almost aligned in IGF 4523). Red asterisks indicate broken portions. The morphology 1346 of the posterior edge of C4 in SABAP_UMB 129850 is estimated (dotted line) because it is partially 1347 covered by C5. Scale bar: 2 cm. g) Mustelidae indet., left upper canine (SABAP_UMB 22.3.263) in 1348 mesial view. Scale bar: 1 cm. 1349 1350 Fig. 4 - (color online) Comparisons between Machairodontinae upper canines. a) From left to right: 1351 Smilodon fatalis Leidy, 1868 from La Brea Tar Pits (LAMC Hc1142); Megantereon cultridens from 1352 Villarroya (IPS36785); M. cultridens from Villarroya (IPS36784); cf. Megantereon from Vigna 53 1353 Nuova (SABAP_UMB 129842; reversed); Homotherium crenatidens from Incarcal-I (IPS IN-I1542); 1354 H. crenatidens from Incarcal-I (IPS IN-I-421). Scale bar: 2 cm. b) Violin plot for the width/length 1355 ratio of the upper canine (measured at the cervix) in Pliocene to Middle Pleistocene Homotherium, 1356 M. cultridens, Megantereon adroveri Pons-Moyà, 1987, and the Vigna Nuova specimen. Legend: 1357 Epi, Epivillafranchian; EVal, Early Villafranchian; Gal, Galerian; LVal, Late Villafranchian; MVal, 1358 Middle Villafranchian. Sources of data: Ballesio (1963); de Bonis (1976); Ficcarelli (1979); Hemmer 1359 (2001); Sardella & Iurino (2012). 1360 1361 Fig. 5 - Equus cf. senezensis from Vigna Nuova (Italy), lower layer. a) Left hindlimb with distal tibia, 1362 tarsus, metatarsals II-III-IV (SABAP_UMB 129856+129861+129862+129868): specimen in 1363 anatomical connection, lateral view (a1); metatarsals in anterior (a2) and posterior (a2) views. b) Left 1364 astragalus (SABAP_UMB 129869) in anterior (b1) and medial (b2) views. c) Left metacarpals II-III- 1365 IV (SABAP_UMB 129840) in anterior (c1) and posterior (c2) views. d) Mandible (SABAP_UMB 1366 129835+129837): incisor arcade in occlusal (d1) and ventral (d2) views; left corpus fragment with 1367 p2 in labial (d3) and occlusal (d4) views. Scale bar: 5 cm. 1368 1369 Fig. 6 - (color online) Biometric comparisons between the horse from Vigna Nuova and several Equus 1370 samples from the Plio-Pleistocene of Eurasia. a) Bivariate plot comparing the maximum width (M6) 1371 versus maximum length (M1) of the p2. b) PCA performed on ten measurements of the third 1372 metatarsal (M1, maximum length; M3, midshaft width; M4, depth of the diaphysis at level of M3; 1373 M5, proximal articular width; M6, proximal articular depth; M10, distal maximal maximum supra- 1374 articular width; M11, distal maximum articular width; M12, distal maximum keel depth; M13, distal 1375 maximum depth of the lateral condyle; M14, distal maximum depth of the medial condyle). c-d) 1376 Log10 ratio diagrams of third metatarsal measurements in European Gelasian (c) and Calabrian (d) 1377 equids. In addition to the measurements used in the PCA, this analysis includes M7 (maximum 54 1378 diameter of the articular facet for the third tarsal) and M8 (diameter for the anterior facet for the fourth 1379 tarsal). 1380 1381 Fig 7 - (color online) Artiodactyls from Vigna Nuova (Italy). a) cf. Leptobos: horn core fragment 1382 (SABAP_UMB 129851). b) Croizetoceros ramosus: maxilla with P2-M3 (SABAP_UMB 129863) 1383 in occlusal (b1) and labial (b2) views. c-h) ‘Pseudodama’ sp.: right hemimandible with p2-m3 1384 (SABAP_UMB 129848) in occlusal (c1) and labial (c2) views; left hemimandible with dp4 1385 (fragmented), m1-m2 (SABAP_UMB 129865+129860) in occlusal (d1) and labial (d2) views; left 1386 frontal fragment with antler basis (SABAP_UMB 129845) in medial (e1) and anterior (e2) views; 1387 right antler fragment (SABAP_UMB 129880) in lateral (f1) and anterior (f2) views; right M1 1388 (SABAP_UMB 129870) in occlusal view (g). h) Cervidae indet. (large size): left maxilla with dP2- 1389 M2 (SABAP_UMB 129866) in occlusal (h1) and labial (h2) views. j) Sus cf. strozzii: maxilla with 1390 P4-M2 and M3 fragment (SABAP_UMB 129847) in occlusal view. Scale bar: 2 cm. 1391 1392 Fig. 8 - Comparisons between Croizetoceros ramosus upper tooth rows. a) SABAP_UMB 129863 1393 from Vigna Nuova. b) NHMB Prr.346 from Les Etouaires (France; reversed). c) MCL QSV-766 from 1394 Saint Vallier (France). d) AUT GER-188 from Gerakarou (Greece). e) IPS28245/740b from La 1395 Puebla de Valverde (Spain; reversed). The fossils are drawn not to scale to highlight morphological 1396 similarities. 1397 1398 Fig. 9 - (color online) Boxplots of upper tooth measurements (mm) comparing the suid from Vigna 1399 Nuova with Sus arvernensis, Sus strozzii, and Sus scrofa. Boxplots show minimum, median, and 1400 maximum values with 25th and 75th percentile of each sample. L, length; W, width. 1401 1402 Fig. 10 - (color online) cf. Leptobos from Vigna Nuova (Italy), upper layer. a) Left hemimandible 1403 with p4-m3 (SABAP_UMB 129836) in labial (a1) and occlusal (a2) views. b) Left P2 (SABAP_UMB 55 1404 129855) in occlusal view; c) Atlas (SABAP_UMB 129839) in ventral (c1) and dorsal (c2) views. d) 1405 Right metacarpal (SABAP_UMB 129838) in proximal (d1) and anterior (d2) views. e) Horn core 1406 fragment (SABAP_UMB 22.3.248). f) Horn core fragment (SABAP_UMB 22.3.249). g) 1407 Intermediate phalanx (SABAP_UMB 129854) in abaxial (g1) and axial (g2) views. Scale bar: 5 cm. 1408 1409 Fig. 11 - (color online) Bivariate plot comparing the maximum length (Lmax; in mm) versus the ratio 1410 between the distal end width (DEW) and maximum length of the metacarpal in several samples of 1411 Eurasian bovines. The sample of Leptobos etruscus includes data from Faella/Matassino and Olivola 1412 (Italy), Senèze (France), and Fonelas P-1 (Spain). The sample of Leptobos vallisarni includes data 1413 from Upper Valdarno and Farneta (Italy). The sample of Leptobos ex gr. SEM includes the following 1414 sub-samples: Leptobos cf. stenometopon from Montopoli (Italy); Leptobos stenometopon/elatus from 1415 Upper Valdarno (Italy; several sites); Leptobos elatus from Les Etouaires (France) and Villarroya 1416 (Spain); Leptobos merlai from Saint Vallier (France); Leptobos cf. furtivus from Senèze and Olivola; 1417 Leptobos aff. furtivus from Upper Valdarno and Carsoli (Italy). The sample of Eobison includes the 1418 following sub-samples: Eobison sp. from Venta Micena (Spain), Eobison degiulii from Pirro Nord 1419 (Italy); Eobison cf. degiulii from Capena (Italy) and Mygdonia Basin (Greece); Eobison georgicus 1420 (Burchak-Abramovich & Vekua, 1994) from Dmanisi (Georgia); Eobison palaeosinensis Teilhard de 1421 Chardin & Piveteau, 1930 from Nihewan Basin (China). The sample of Bison menneri includes data 1422 from Untermassfeld (Germany). The sample of Bison schoetensacki includes data from Süssenborn 1423 and Mauer (Germany), Vallparadís composite section (Spain), Le Vallonnet, Le Vassirie, and Durfort 1424 (France). Source of data: Sorbelli et al. (2021), and references therein. 56