Introduction
Records of the giant short-faced hyena Pachycrocuta Kretzoi, Reference Kretzoi1938 are known from many Early Pleistocene localities across Eurasia, from England to China (Ficcarelli and Torre, Reference Ficcarelli and Torre1970; Howell and Petter, Reference Howell and Petter1980; Qiu, Reference Qiu1987; Kurtén and Garevski Reference Kurtén and Garevski1989; Sotnikova, Reference Sotnikova1989; Werdelin and Solounias, Reference Werdelin and Solounias1991; Turner and Antón, Reference Turner and Antón1996; Turner, Reference Turner, Kahlke and Kahlke2001; Sotnikova et al., Reference Sotnikova, Baigusheva and Titov2002; Qiu et al., Reference Qiu, Deng and Wang2004; Sotnikova and Titov, Reference Sotnikova and Titov2009; Madurell-Malapeira et al., Reference Madurell-Malapeira, Minwer-Barakat, Alba, Garcés, Gómez, Aurell- Garrido, Ros-Montoya, Moyá-Solà and Berástegui2010; Iannucci et al., Reference Iannucci, Mecozzi, Sardella and Iurino2021; Lavrov et al., Reference Lavrov, Gimranov, Startsev and Lopatin2021; Liu et al., Reference Liu, Liu, Zhang, Wagner, Jiangzuo, Song, Liu, Wang and Jin2021; Marciszak et al., Reference Marciszak, Semenov, Portnicki and Derkach2021; Jiangzuo et al., Reference Jiangzuo, Liu, Jiang and Wang2022). Pachycrocuta from the middle Pleistocene are known only from Asia (Owen, Reference Owen1870; Pei, Reference Kretzoi1934; Kretzoi, Reference Pei1938; Kurtén, Reference Kurtén1956; Iannucci et al., Reference Iannucci, Mecozzi, Sardella and Iurino2021; Liu et al., Reference Liu, Liu, Zhang, Wagner, Jiangzuo, Song, Liu, Wang and Jin2021).
The fossil remains of the giant hyenas in Europe were well studied (Kretzoi, Reference Kretzoi1938; Kurtén, Reference Kurtén1956; Ficcarelli and Torre, Reference Ficcarelli and Torre1970; Howell and Petter, Reference Howell and Petter1980; Werdelin and Solounias, Reference Werdelin and Solounias1991; Turner and Antón, Reference Turner and Antón1996; Turner, Reference Turner, Kahlke and Kahlke2001; Turner et al., Reference Turner, Antón and Werdelin2008; Iannucci et al., Reference Iannucci, Mecozzi, Sardella and Iurino2021, Reference Iannucci, Mecozzi, Kahlke, Sardella and Kahlke2022; Marciszak et al., Reference Marciszak, Semenov, Portnicki and Derkach2021; Espigares et al., Reference Espigares, Palmqvist, Rodríguez-Ruiz, Ros-Montoya, Pérez-Ramos, Rodríguez-Gómez and Guerra-Merchán2023). Recently, materials from China have also been published (Liu et al., Reference Liu, Liu, Zhang, Wagner, Jiangzuo, Song, Liu, Wang and Jin2021; Jiangzuo et al., Reference Jiangzuo, Liu, Jiang and Wang2022). However, in Asia, outside China, records of Pachycrocuta are limited. Sporadic finds were reported from Java (Dubois, Reference Dubois1908; Brongersma, Reference Brongersma1937; van den Bergh et al., Reference Van den Bergh, de Vos and Sondaar2001), the Republic of Tajikistan (Sotnikova, Reference Sotnikova1989; Sotnikova and Vislobokova, Reference Sotnikova and Vislobokova1990), the Republic of Kazakhstan (Sotnikova et al., Reference Sotnikova, Dodonov and Pen’kov1997), and Georgia (Vekua, Reference Vekua1986).
Central Asian Pachycrocuta from the Early Pleistocene localities of Russia and Mongolia was only listed, but not studied (Vangengeim and Sotnikova, Reference Vangengeim and Sotnikova1981; Zhegallo et al., Reference Zhegallo, Zazhigin, Kolosova, Malaeva, Murzaeva, Sotnikova, Vislobokova, Dmitrieva and Dubrovo1982; Turner and Antón, Reference Turner and Antón1996). Moreover, these publications are hardly referred to in the international reviews of Hyaenidae. Stratigraphically, Asian Russian and Mongolian finds come from sediments correlated to the paleomagnetic interval between Jaramillo and the Matuyama-Brunhes paleomagnetic transition (0.9–0.78 Ma (Vangengeim and Sotnikova, Reference Vangengeim and Sotnikova1981; Zhegallo et al., Reference Zhegallo, Zazhigin, Kolosova, Malaeva, Murzaeva, Sotnikova, Vislobokova, Dmitrieva and Dubrovo1982; Sotnikova, Reference Sotnikova and Ralf-Dietrich2001; Fig. 1)). These samples are especially important, because Pachycrocuta finds from this time interval in Europe are rare compared with the abundance known from the earlier levels of the Early Pleistocene (2.2–1.4 Ma; e.g., Sainzelles [France]; Gerakarou 1 [Greece]; Fonelas 1, Venta Micena, and Fuente Nueva 3 [Spain]; Olivola, Upper Valdarno [Italy]; Taurida Cave [Crimean Peninsula], Liventsovka [European Russia]) (Weithofer, Reference Weithofer1889; Ficcarelli and Torre, Reference Ficcarelli and Torre1970; Sotnikova et al., Reference Sotnikova, Baigusheva and Titov2002; Arribas et al., Reference Arribas, Baeza, Bermúdez, Blanco, Durá, Garrido, Gumiel, Hernández, Soria and Viseras2004; Martínez- Navarro et al., 2010; Medin et al., Reference Medin, Martínez-Navarro, Rivals, Madurell-Malapeira, Ros-Montoya, Espigares, Figueirido, Rook and Palmqvist2017; Lavrov et al., Reference Lavrov, Gimranov, Startsev and Lopatin2021).
Map indicating the main localities with Pachycrocuta and Pliocrocuta discussed in the text.
In recent years, many studies have been published on the paleoecology and paleobiogeography of giant hyenas. Special attention was given to the analysis of the size, food adaptations, and geographic and geological distribution of this predator (Palmqvist et al., Reference Palmqvist, Martínez-Navarro, Pérez-Claros, Torregrosa, Figueirido, Jiménez-Arenas, Patrocinio Espigares, Ros-Montoya and de Renzi2011; Koufos, Reference Koufos2014; Madurell-Malapeira et al., Reference Madurell-Malapeira, Ros-Montoya, Espigares, Alba and Aurell-Garrido2014; Iannucci et al., Reference Iannucci, Mecozzi, Sardella and Iurino2021; Marciszak et al., Reference Marciszak, Semenov, Portnicki and Derkach2021). Little attention has been paid recently to their dental morphology (but see Liu et al., Reference Liu, Liu, Zhang, Wagner, Jiangzuo, Song, Liu, Wang and Jin2021; Iannucci et al., Reference Iannucci, Mecozzi, Kahlke, Sardella and Kahlke2022). Even though the genus Pachycrocuta existed in Eurasia for about 1.5 Ma, the evolutionary trends of its dental morphology remain inadequately understood and confusing in terms of advanced and primitive characters. The same is true for the chronology of the evolutionary changes. These issues are caused by an incomplete geological record, as well as by insufficient preservation of the materials. The new material discussed here significantly expands knowledge about the Asian Pachycrocuta and contributes to establishing a correlation between European and Asian localities of the Early Pleistocene.
The aim of this work is a detailed description of the materials on P. brevirostris from Zasukhino-3 and Nalaikha (detailed later), as well as a comprehensive morphological analysis of dental characters using materials on Pachycrocuta from other Eurasian regions (e.g., France, Italy, Germany, Slovakia, Ukraine, south of Russia, Tajikistan, and China), in order to identify the main morphological trends and features related to the evolution of this group of carnivores in time and space.
Problems of generic and subspecies diagnostics of crocutoid hyaenids (Pliocrocuta and Pachycrocuta)
The Pliocene–Pleistocene large mammalian assemblages of the Old World are characterized by the presence of the extinct crocutoid hyenas Pliocrocuta Kretzoi, Reference Kretzoi1938 and Pachycrocuta. In Eurasia, Pliocrocuta was recorded from 4.5 to 1.8 Ma, and Pachycrocuta is known from 2.6 to 0.5 Ma (Werdelin and Solounias, Reference Werdelin and Solounias1991; Martínez-Navarro, Reference Martínez-Navarro, Fleagle, Shea, Grine, Baden and Leakey2010; Vinuesa, Reference Vinuesa2018; Iannucci et al., Reference Iannucci, Mecozzi, Sardella and Iurino2021). Their coexistence was noted only in Greece at Gerakarou-1, dated to 1.8 Ma (Koufos, Reference Koufos2014).
The co-occurrence of Pliocrocuta and Pachycrocuta was also noted in the Liventsovka locality, dated from 2.0 to 1.6–1.4 Ma (Fig. 1) in the Rostov region, European Russia (Sotnikova et al., Reference Sotnikova, Baigusheva and Titov2002; Titov, Reference Titov2008; Tesakov, Reference Tesakov2021). However, recent revision of the locality and the fauna suggests that the finds are not contemporary. The individual teeth assigned to Pl. perrieri came from the lower levels of the section, correlative to the middle Villafranchian (2.6–2.0 Ma), while the mandible assigned to early P. brevirostris was associated with upper levels horizons (Fig. 2; late Villafranchian, 2.0–1.1 Ma).
Mandible of Pachycrocuta brevirostris from Liventsovka, RSU-231, (a) occlusal view, (b) right ramus, (c) left ramus. Scale bar: 5 cm.
The genera Pliocrocuta and Pachycrocuta have often been distinguished by the significantly larger size of the latter genus. Consequently, with more records of Pachycrocuta, it became clearer that cheek teeth of these hyena genera may occasionally overlap in size (Howell and Petter, Reference Howell and Petter1980, table 4). Moreover, revised diagnoses for these taxa are rare in the publications of the last decade, and the validity of the genus Pliocrocuta was questioned (Qiu et al., Reference Qiu, Deng and Wang2004; Liu et al., Reference Liu, Liu, Zhang, Wagner, Jiangzuo, Song, Liu, Wang and Jin2021; Pérez-Claros, Reference Pérez-Claros2024).
In the original description, Kretzoi (Reference Kretzoi1938) gave very limited diagnoses for these genera. In the diagnosis of Pliocrocuta, only carnassial m1 was mentioned among the lower teeth. The morphology of the m1 was characterized by an absent metaconid and a wide talonid, with a poorly pronounced hypoconid and entoconid. The diagnosis of Pachycrocuta was more elaborate, pointing to the absence of anterior cusps on p2 and p3, a strong main cusp on p4, a short and thick m1 lacking a metaconid, a centrally placed posterior crest of the protoconid m1, and the reduced talonid m1, bearing a conical hypoconid surrounded by a cingulid.
These generic names of Pliocrocuta and Pachycrocuta were not used for about 30 years, while all the Plio-Pleistocene fossil crocutoid taxa were assigned to the genus Hyaena (Kurtén, Reference Kurtén1956). The taxonomy of Kretzoi (Reference Kretzoi1938) gradually returned to the literature only in the 1970s–1980s. The genus name Pliocrocuta was then synonymized with Pachycrocuta (Ficcarelli and Torre, Reference Ficcarelli and Torre1970; Howell and Petter, Reference Howell and Petter1980).
Subsequently, Werdelin and Solounias (Reference Werdelin and Solounias1991) restored the validity of the genera Pliocrocuta and Pachycrocuta. The latter was recognized as monotypical, with the only species being P. brevirostris. Based on cladistic analysis, these two genera were considered as sister taxa (Werdelin and Solounias, Reference Werdelin and Solounias1991). This concept was also confirmed by the analysis of deciduous teeth of Pachycrocuta, conducted by Baryshnikov and Averianov (Reference Baryshnikov and Averianov1995). This point of view is still adopted by many scientists (e.g., Turner and Antón, Reference Turner and Antón1996; Turner, Reference Turner, Kahlke and Kahlke2001; Vinuesa, Reference Vinuesa2018; Iannucci et al., Reference Iannucci, Mecozzi, Sardella and Iurino2021, Reference Iannucci, Mecozzi, Kahlke, Sardella and Kahlke2022). However, the concept of Werdelin and Solounias (Reference Werdelin and Solounias1991) was not supported in the works of others describing Asian Pachycrocuta (Qiu et al., Reference Qiu, Deng and Wang2004; Liu et al., Reference Liu, Liu, Zhang, Wagner, Jiangzuo, Song, Liu, Wang and Jin2021).
The revised diagnosis by Qiu et al. (Reference Qiu, Deng and Wang2004, p. 178) for the genus Pachycrocuta states: “Metaconid present in m1 in most cases, talonid is large, with 2 or 3 cusplets.” Such modification of the original diagnosis of Kretzoi (Reference Kretzoi1938) allowed the authors to reconsider the taxa Pl. perrieri and Pl. pyrenaica in the genus Pachycrocuta, which automatically placed the genus Pliocrocuta in synonymy with Pachycrocuta (Qiu et al., Reference Qiu, Deng and Wang2004, pp. 69–70). In contrast to Werdelin and Solounias (Reference Werdelin and Solounias1991), Qiu et al. (Reference Qiu, Deng and Wang2004) considered the Pliocene and Pleistocene forms as a single evolutionary lineage formed ca. 5 Ma ago. This wide concept of the genus Pachycrocuta by Qiu et al. (Reference Qiu, Deng and Wang2004) was also used in the recent study by Liu et al. (Reference Liu, Liu, Zhang, Wagner, Jiangzuo, Song, Liu, Wang and Jin2021).
Due to different opinions on the generic and specific taxonomy of European and Asian Plio-Pleistocene crocutoid hyenas, a comparative analysis is difficult. In particular, this concerns the Longdan fauna (China, ca. 2.3–2.2 Ma), where the first appearance of Pachycrocuta is determined to be earlier than in Europe. This fauna includes two groups of crocutoid hyenas, different in size, but assigned to the same species “Pachycrocuta licenti” (Qiu et al., Reference Qiu, Deng and Wang2004, table 16, p. 76). One of them (numbers v 13551, HMV 1199) is closer in size to P. brevirostris, and the second to Pl. perrieri (numbers HMV 1201, HMV 1202) (Fig. 3). Such a size difference suggests the possibility that “P. licenti” from Longdan includes two different taxa of crocutoid hyenas. This possibility is indicated not only by the size of the lower teeth but also by the pliocrocutoid features of the smaller form, such as a three-cusped m1 talonid with small buccal and lingual cusps, and a ridge-shaped hypoconid located centrally (Qiu et al., Reference Qiu, Deng and Wang2004, p. 74). The presence of Pliocrocuta in Longdan is further supported by numerous remains of Pliocrocuta (in particular, Pl. perrieri) previously described from another Asian locality, Kuruksai in Tajikistan (Fig. 1; 2.5–2.2 Ma), which is close in age and has a similar mammalian community (Sotnikova, Reference Sotnikova1989; Sharapov, Reference Sharapov and Abdusalamov2014).
Scatter plot of lower carnassial m1 of Pachycrocuta, Pliocrocuta, and forms attributed to “P. licenti/Pl. licenti” from Longdan and Zilfi (data in Supplementary Table 1).
According to the original description of the holotype “Hyaena licenti” from Nihewan (China) (= Hyaena sinensis in Teilhard de Chardin and Piveteau, Reference Teilhard de Chardin and Piveteau1930), three cusps are present on its m1 talonid. This primitive feature is not characteristic of Pachycrocuta but is a feature of Pliocrocuta. Therefore, it is possible that even the type specimen “P. licenti” from Nihewan may belong to the genus Pliocrocuta, as indicated by its m1 talonid morphology.
The Asian appearance of the form similar to “Pliocrocuta licenti” was also identified at the locality of Zilfi (= Yakkobed; Fig. 1) in Tajikistan (IZIP, Sharapov, Reference Sharapov and Abdusalamov2014). This locality is situated close to Kuruksai, and based on the mammalian community, correlates in age (2.6–2.0 Ma; Sotnikova et al., Reference Sotnikova, Dodonov and Pen’kov1997). According to the description by Sharapov (Reference Sharapov and Abdusalamov2014), “Pliocrocuta licenti” from Zilfi is larger than Pl. perrieri (Fig. 3), but shows pliocrocutoid dental morphology. Thus, the materials from Tajikistan might indicate the presence of pliocrocutoid features in forms classified as “Pl. licenti/P. licenti/P. br. licenti.”
If our assumptions regarding the forms discussed earlier are correct, the attribution of the Early Pleistocene forms with plio- and pachycrocutoid morphology of the m1 to the single taxon “Pl. licenti/P. licenti/P. br. licenti” requires major revision. Taking into account the uncertain taxonomic position of licenti (what is it: Pliocrocuta or Pachycrocuta?), and considering the arguments earlier, we accept the existence of two genera, Pliocrocuta and Pachycrocuta, in the Plio-Pleistocene of Eurasia. We also support the opinion of Werdelin and Solounias (Reference Werdelin and Solounias1991) that Pachycrocuta is a monospecific genus, which includes only P. brevirostris.
The following is a description of the material on P. brevirostris from the localities of Zasukhino-3 (Asian Russia) and Nalaikha (Mongolia).
Geological and faunal settings
Zasukhino-3
The locality of Zasukhino-3 (Fig. 1) is situated in Buryat Western Transbaikalia (Asian Russia), 40 km north of Ulan-Ude, on the right bank of the Itantsa River. The locality was discovered in 1966 by Rezanov, and the large mammal material was excavated in 1968–1970 and 1978 by Dobretsov Geological Institute of Siberian Branch of Russian Academy of Sciences and the Geological Institute of the Russian Academy of Sciences (Vangengeim and Sotnikova, Reference Vangengeim and Sotnikova1981).
In this locality, Neogene and Quaternary deposits fill an ancient gully cut into Mesozoic bedrock. Lens-shaped clusters of large and small mammals were uncovered in the strata of proluvial–deluvial Quaternary sediments. Five lithological strata were distinguished in the section (Vangengeim and Sotnikova, Reference Vangengeim and Sotnikova1981; Vangengejm et al., Reference Vangengejm, Erbajeva and Sotnikova1990).
The material came from the third bottom sandy loam strata (Zasukhino-3), which contains the so-called Zasukhinian fauna. The list of large mammalian fauna is presented in Table 1 (Vangengeim and Sotnikova, Reference Vangengeim and Sotnikova1981; Alexeeva, Reference Alexeeva2005; Erbajeva et al., Reference Erbajeva, Khenzykhenova and Alexeeva2013). The small mammal association includes the rare Prosiphneus and Allophaiomys pliocaenicus, co-occurring with Terricola, Alticola, and Microtus, and is completely devoid of rhizodont vole, suggesting the age of the terminal Early Pleistocene (0.99–0.78 Ma) (Erbajeva and Alexeeva, Reference Erbajeva and Alexeeva2000; Alexeeva, Reference Alexeeva2005; Erbajeva et al., Reference Erbajeva, Khenzykhenova and Alexeeva2013).
Large mammals from the localities of Zasukhino-3 and Nalaikha (data from Vangengejm and Sotnikova, 1981; Zhegallo et al., Reference Zhegallo, Zazhigin, Kolosova, Malaeva, Murzaeva, Sotnikova, Vislobokova, Dmitrieva and Dubrovo1982; Eisenmann and Kuznetsova, Reference Eisenmann and Kuznetsova2004; Alexeeva, Reference Alexeeva2005; Kuznetsova and Zhegallo, Reference Kuznetsova and Zhegallo2009; Erbajeva et al., Reference Erbajeva, Khenzykhenova and Alexeeva2013).
Nalaikha
The locality of Nalaikha (Fig. 1) is situated 30 km southeast of Ulaanbaatar (northern Mongolia), in a quarry on the left bank of the Tola River. The locality was discovered in 1975 and was studied by the Soviet-Mongolian Geological Expedition team.
The exposed section of Pliocene–Pleistocene deposits is 15–17 m thick. The section is composed of four lithological members. The second stratum from the bottom, of alluvial-proluvial genesis, contains rich bone-bearing lenses (Zhegallo et al., Reference Zhegallo, Zazhigin, Kolosova, Malaeva, Murzaeva, Sotnikova, Vislobokova, Dmitrieva and Dubrovo1982; Table 1).
The age of this fauna was initially dated broadly from the second half of the Early to the early middle Pleistocene (Zhegallo et al., Reference Zhegallo, Zazhigin, Kolosova, Malaeva, Murzaeva, Sotnikova, Vislobokova, Dmitrieva and Dubrovo1982). Subsequently, a detailed study of equids suggested a late Early Pleistocene age (Eisenmann and Kuznetsova, Reference Eisenmann and Kuznetsova2004; Kuznetsova and Zhegallo, Reference Kuznetsova and Zhegallo2009). In addition, the similarity in the evolution stage of the lower carnassial of Хеnосуоn lycaonoides from Nalaikha and Untermassfeld (Germany, ca. 1.1 Ma) was noted by Sotnikova (Reference Sotnikova and Ralf-Dietrich2001, Taf. 112, 4). All of this suggests an age of ca. 0.9 Ma for the Nalaikha fauna.
Material and methods
Abbreviations
Institutional and collection abbreviations
BDEU, Department of Biology, Ege University (Izmir, Turkey); GIN RAS, Geological Institute of the Russian Academy of Sciences (Moscow); GIN SB RAS, Dobretsov Geological Institute of Siberian Branch of Russian Academy of Sciences (Ulan-Ude); IPAE, Institute of Plant and Animal Ecology of the Ural Branch of the Russian Academy of Sciences (Yekaterinburg); IQW, Senckenberg Research Station of Quaternary Palaeontology (Weimar, Germany); IVPP, Institute of Vertebrate Paleontology and Paleoanthropology, Chinese Academy of Sciences (Beijing); IZIP, Institute of Zoology and Parasitology of the Academy of Sciences of the Republic of Tajikistan (Dushanbe); PIN RAS, Borissiak Paleontological Institute of the Russian Academy of Sciences (Moscow); RSU, Rostov State University, (Rostov-on-Don, Russia); UrFU, Ural Federal University (Yekaterinburg, Russia); ZIN RAS, Zoological Institute of the Russian Academy of Sciences (Saint Petersburg).
Collection and locality abbreviations
IGF, Museum of Natural History, Geological and Palaeontological section, the University of Florence (Italy); MCP, Musèe Crozatier, Le Puy-en-Velay (France); NMB, Naturhistorisches Museum Basel (Switzerland); RRM, Rostov Regional Museum (Rostov-on-Don, Russia); SAI, Sainzelles collection (France); V.A., collections of Upper Valdarno (Tuscany, Italy) from NMB; ZKD-1, Loc.1 of Zhoukoudian (China); ZMMU, Zoological Museum of Moscow University (Russia); JYC, Jinyuan Cave (China).
Tooth measurements and abbreviations
Teeth are abbreviated to their first letter, with numbers referring to the corresponding tooth (e.g., p2, lower second premolar; m1, lower first molar).
Dental measurements: L, maximum length; W, maximum width; Wa, anterior width; Wp, posterior width; Lp2–p4, length of the tooth row from p2 to p4.
Mandibular measurements: Tmp2, Tmp3, Tmp4, thickness under respective premolar; Tmm1, thickness under the m1; Hmp2, height under p2.
Other measurements: D, diameter.
All the measurements are given in millimeters. The dental and mandibular measurements are presented in Supplementary Figure 1.
Material
The studied material from the localities of Zasukhino-3 (Fig. 4) and Nalaikha (Fig. 5) is housed at the GIN RAS (collection number is GIN 4370) and PIN RAS (collection number is PIN 3747). Specimens include fragments of mandibles and isolated lower teeth of seven individuals:
Zasukhino-3 (Fig. 4): a fragment of the right hemimandible GIN 4370/93, a fragment of m1 GIN 4370/99, right isolated check teeth GIN 4370/98;
Nalaikha (Fig. 5): a fragment of the right hemimandible PIN 3747-242, a fragment of the left hemimandible GIN 4370/120, a right m1 PIN 3747-122, an isolated canine PIN 3747-240.
Mandibular and dental material of Pachycrocuta brevirostris from Zasukhino-3 (Asian Russia): (a) GIN 4370/93, right hemimandible, (a) buccal, (a’) lingual, (a’’) occlusal view; (b–d) GIN 4370/98, right isolated teeth of the one individual, (b) p4, (c) p3, (d) p2, (b–d) occlusal view, (b’–d’) buccal view; (e) GIN 4370/99, isolated fragment of the left m1, buccal view. Scale bar: 2 cm.
Mandibular and dental material of Pachycrocuta brevirostris from Nalaikha (Mongolia): (a) PIN 3747-242, right hemimandible, (a) buccal, (a’) lingual, (a’’) occlusal view; (b–d) GIN 4370/120, left hemimandible, (b) lingual, (b’) buccal, (b’’) occlusal view, (c) isolated incisor, (d) isolated canine; (e) PIN 3747-122, isolated right m1, (e) buccal, (e’) lingual, (e’’) occlusal view; (f) PIN 3747-240, isolated canine. Scale bar: 3 cm.
Сomparative material
Because P. brevirostris had a wide stratigraphic and geographic distribution, our comparison included Eurasian material from selected sites of a broad stratigraphic range from 2.2 to 0.5 Ma to identify features related to geological age or area. The geographic distribution of the sites mentioned in the text appears in Figure 1. A direct study was made on the material on P. brevirostris from the Early Pleistocene localities of Lakhuti-2 (southern Tajikistan, GIN 3848/282-67; Fig. 6), Liventsovka (Rostov region, Russia, RSU-231; Fig. 2), Akhtanizovskaya (Taman Peninsula, Russia, GIN 1126/1); middle Pleistocene locality of ZKD-1 (China, PIN 538-531); on Pliocrocuta from the early Pliocene locality of Odessa Catacombs (Ukraine, PIN 390-5(9)) and Early Pleistocene locality of Kuruksai (Tajikistan, PIN 3120-289, PIN 3120-351, GIN 3848/57-97).
One of the authors (MS) studied the holotype of P. brevirostris from Sainzelles (France, SAI-2003-15-92), stored at MCP, as well as samples from Upper Valdarno (Italy, IGF847, IGF848), and Untermassfeld (Germany, IQW 1999/26624 (Mei. 26 153), IQW 1988/26601 (Mei. 22 120)). These materials had been described previously (Gervais, Reference Gervais1850; Weithofer, Reference Weithofer1889; Ficcarelli and Torre, Reference Ficcarelli and Torre1970; Turner, Reference Turner, Kahlke and Kahlke2001; Iannucci et al., Reference Iannucci, Mecozzi, Sardella and Iurino2021). The extant specimens of Hyaena hyaena, Parahyaena brunnea, and Crocuta crocuta housed in ZIN RAS and ZMMU were studied as well.
Mandibular and dental material of Pachycrocuta brevirostris from Lakhuti-2 (Tajikistan): (a) GIN-3848/282-67, left hemimandible, (a) buccal, (a’) lingual, (a’’) occlusal view; (b) GIN 3848/248-67, right isolated m1, (b) lingual, (b’) buccal, (b’’) occlusal view. Scale bar: 3 cm.
Photographic and graphic materials were used from: Weithofer (Reference Weithofer1889), Pei (Reference Pei1934), Kretzoi (Reference Kretzoi1938), Turner (Reference Turner, Kahlke and Kahlke2001), Qiu et al. (Reference Qiu, Deng and Wang2004), Madurell-Malapeira et al. (Reference Madurell-Malapeira, Minwer-Barakat, Alba, Garcés, Gómez, Aurell- Garrido, Ros-Montoya, Moyá-Solà and Berástegui2010), Koufos (Reference Koufos2014), Liu et al. (Reference Liu, Liu, Zhang, Wagner, Jiangzuo, Song, Liu, Wang and Jin2021), and Iannucci et al. (2019, Reference Iannucci, Mecozzi, Kahlke, Sardella and Kahlke2022). In addition, photographs of crocutoid hyenas were used from the collections of NMB and IPAE, which were kindly shared by S. Mayda, D. Gimranov, and D. Khantemirov.
Methods
The lower carnassial m1 and the other cheek teeth (p3 and p4) were studied. For illustrative purposes, the basal morphotype typical for the Pliocene Pliocrocuta sp. was presented. In this work, we mostly use the dental terminology of Werdelin and Solounias (Reference Werdelin and Solounias1991). The lower carnassial m1 consists of trigonid and talonid. The trigonid is composed of paraconid (anterior cusp), protoconid (posterior cusp), and metaconid (posterolingual cusp). The talonid is composed of the main cusp (hypoconid), lingual cusp (entoconid), buccal elevation or cusp, and sometimes posterior cusp (hypoconulid). The premolars consist of the main cusp, anterior and posterior cusps, and cingulid.
The material was measured using a digital caliper with an accuracy of 0.1 mm. Measurements are mainly following Werdelin and Solounias (Reference Werdelin and Solounias1991) with the addition of anterior and posterior widths of the p3 measured separately. The quantitative analyses were based on bivariate diagrams made in Microsoft Excel (Table 1, Supplementary Table 1).
The analysis of morphotypes was carried out from the point of view of the analysis of evolutionary trends in hyaenids (e.g., Pei, Reference Pei1934; Kretzoi, Reference Kretzoi1938; Kurtén, Reference Kurtén1956, Reference Kurtén1957, Reference Kurtén1972; Ficcarelli and Torre, Reference Ficcarelli and Torre1970; Howell and Petter, Reference Howell and Petter1980; Qiu, Reference Qiu1987; Sotnikova, Reference Sotnikova1989; Werdelin and Solounias, Reference Werdelin and Solounias1991; Liu et al., Reference Liu, Liu, Zhang, Wagner, Jiangzuo, Song, Liu, Wang and Jin2021; Iannucci et al., Reference Iannucci, Mecozzi, Kahlke, Sardella and Kahlke2022), and with the addition of our observations (see “Discussion”). The features include: for m1: (1) position and quantity of the posterior crests of the protoconid; (2) position and size of the hypoconid; (3) presence/absence, position, and size of the entoconid; and (4) presence/absence and size of the buccal elevation or cusp on the talonid; (5) size of the buccal and lingual talonid space; for premolars: (1) presence/absence of the anterior cusp; (2) relative size of the cusps; and (3) presence/absence of the cingulid.
The photos of the material were taken using a Canon EOS R camera with an adapter and an older version of the Canon EF 100 mm/f/2.8 Macro lens. The photos were processed in Digital Photo Professional 4 and Adobe Photoshop CS6.
Results
Description of P. brevirostris from Zasukhino-3 and Nalaikha
Zasukhino-3
The fragment of the right hemimandible with p2–m1 and a posterior part of the canine alveolus (GIN 4370/93; Fig. 4a) lacks the ventral part, symphysis, and ascending branch, which are broken off. It belongs to a young adult with slightly worn teeth. The mandibular body is large and robust. The single mental foramen is located beneath the middle of the p2 and is oval (H = 8.2/L = 6.4). The mandible thickness under the cheek teeth decreases from the p2 (Tmp2 = 26.94) to the m1 (Tmm1 = 19.09). The tooth row is strongly curved. Premolars are set without diastemas. In buccal view, only the base of the p2 crown is at the same level as the basal line of the m1. The p4 and p3 are set below this level. The tooth sizes are within the variability of the Eurasian Early Pleistocene hyenas P. brevirostris (Fig. 7a, Table 2).
Dental and mandibular measurements of Pachycrocuta brevirostris from Zasukhino-3 and Nalaikha.
Scatter plot of lower teeth measurements (a, p3; b, m1) of Pachycrocuta brevirostris from the current study and European localities (data in Supplementary Table 1).
The p2 is massive, two-roоted, and low-crowned. Anterior and posterior low crests are well developed and directed buccally and lingually, respectively. Weak basal cingulid surrounds the tooth anteriorly and posteriorly.
The p3 is high and massive. The tooth has a subquadrate outline, with the anterior width being greater than the posterior width. Its greatest width exceeds the width of the other cheek teeth (Table 2). The anterior cusp is absent, but there is a distinct anterior crest running up to one-half tooth height. Unlike the other premolars, the main cusp of the p3 is large and high, occupying approximately four-fifths of the dental length. The posterior cusp is strong and closely appressed to the main cusp; the cusps are separated from each other by a distinct notch well marked on the buccal and lingual sides. A weak anterior cingulid is present; the posterior cingulid is almost completely merged with the posterior cusp.
The p4 is the only tricuspid among the premolars. Its outline is narrower and more elongated than that of the p3. The main cusp is high and conical, leans backward, and has a distinct posterior crest. The anterior cusp is well developed and shifted lingually. The large and wide posterior cusp is well separated from the main cusp. The cingulid is almost reduced; in the posterior side, it is merged with the posterior cusp.
The m1 (lower carnassial) is only 1.5 mm longer than the p4 (Table 2). The paraconid is 1.22 mm longer than the protoconid. There is no trace of a metaconid. The posterior crest of the protoconid is directed lingually, but in the lower third of its high, it turns buccally. A weak cingulid surrounds the anterior part of the m1. The talonid is broken off.
The m1 talonid can be studied only on a fragment of the protoconid and talonid part of the left m1 (GIN 4370/99; Fig. 4e). The posterior crest of the protoconid (postprotocristid), similar to the mandible described earlier (GIN 4370/93), runs down the lingual side of the tooth. In the lower quarter of its high, this crest turns buccally and connects with the posterior crest of the single central talonid cusp (hypoconid). The talonid is surrounded by a cingulid, which is well separated posteriorly from the hypoconid. The cingulid forms a small elevation lingually and buccally from the talonid main cusp, with the buccal elevation being better pronounced than the lingual one.
The right isolated p2, p3, and p4 belong to the one individual (GIN 4370/98; Fig. 4b–d). The crowns of the teeth are partially destroyed. The p2 differs from GIN 4370/93 by a separated posterior cusp. The rest of the premolars have shape and morphological features similar to GIN 4370/93, but they are slightly larger (Table 2). In contrast to the previously described samples, the root structure is clearly observable. The massive anterior and posterior roots have nearly the same diameter. The root length of the p2 (L = 29.5) is almost twice the height of its crown height (H = 17.6). The roots are closer together on the p4 than on the p3.
Nalaikha
The material from Nalaikha is within the size variability range of P. brevirostris (Fig. 7).
The anterior fragment of the right hemimandible with p2–p4 and canine alveolus (PIN 3747-242; Fig. 5a) belongs to a very old individual with heavily worn teeth. The mandibular body is high; its height beyond the p2 exceeds all measurements known for European P. brevirostris (Table 2). The upper and lower mandible borders are almost parallel. The symphysis pattern is wide; in the central part it reaches the level of the anterior root of the p2. The symphysis narrows downward and forms a weak chin protrusion. There are two mental foramina under the p2 merged into a single channel inside the mandibular body. The anterior foramen is oval (H = 5.2/L = 2.0), and the posterior one is round (D = 4.1) and located 2 mm above the anterior one. In occlusal view, the mandible is extremely robust (Wmp2 = 31.2). The premolar row is curved buccally. There is a diastema (Lalv = 8) between the canine and the p2. The rest of the cheek teeth sit closely together. The p3 crown strongly overlaps the crown of p4. The canine alveolus is very wide and deep.
The p2 differs from the samples described earlier in having a larger and more pronounced posterior cusp well separated from the cingulid, which surrounds the tooth and also forms a small anterior bulge. The rest of the teeth are almost completely worn, but they show similar proportions to those from the samples from Zasukhino-3, although they are larger (Table 2).
The fragment of the left hemimandible with p2-p4 (GIN 4370/120; Fig. 5b–d) lacks the symphysis, ventral part, and ascending branch, which are broken off. Isolated lower incisor and lower canine are present. The hemimandible is much thinner than the hemimandible PIN 3747-242. It shows the greatest thickness under the p2 (Wmp2 = 26.7). A feature of this sample is also the presence of two mental foramina, but in contrast to PIN 3747-242, they are not connected inside the mandible. In occlusal view, the cheek tooth row is curved. The isolated incisor (apparently i2) is strongly worn and has a very long root (with L = 8.92 and W = 7.06). The isolated canine is robust and oval in cross section, with a long and strong root, and has the greatest width in the middle third of its length (W = 19.21).
The p2 has the posterior cusp well separated from the main cusp by a shallow groove. The cingulid is developed from the anterior and lingual sides; the posterior cingulid is absent. In contrast to GIN 4370/93, the p3 has an anterior crown width 0.4 mm smaller than the posterior width. Otherwise, in size and morphology, the teeth are similar to those previously described (Table 2).
The right m1 with a small fragment of the mandible (PIN 3747-122; Fig. 5e) is similar to the lower carnassial (GIN 4370/93) from Zasukhino-3 in size and structure of the trigonid without a metaconid, but differs by the presence of a weak anterior crest on the paraconid. Similar to the samples from Zasukhino-3, the posterior crest of the protoconid of PIN 3747-122 runs along the lingual margin of the protoconid toward the missing metaconid, and in the lower third of its height turns buccally, connecting with the base of the central talonid cusp (hypoconid). The cingulid surrounds the tooth from the lingual and anterior sides and forms a distinct protrusion from the anterobuccal side of the paraconid. The cingulid forms very small bulges from the lingual and posterior sides of the talonid, but from the buccal side of the talonid, the cingulid is not preserved.
The isolated canine (PIN 3747-240; Fig. 5f) is better preserved than GIN 4370/120 and shows similar morphology. Its greatest width is in the middle third of its length.
Remarks and comparison
The lower dentition provides the most comprehensive information for identifying crocutoid forms. Based on their large size, the absence of anterior cusps on the p2 and p3, the absence of the m1 metaconid, and the reduced m1 talonid, the described material from Zasukhino-3 and Nalaikha undoubtedly belongs to P. brevirostris.
An unusual feature of our material is the presence of two mental foramina per hemimandible in the forms from Nalaikha. This feature is a basal trait for hyaenids and is very rare in P. brevirostris. A similar feature was found only in the form from Upper Valdarno (Weithofer, Reference Weithofer1889). Another feature is the large size of the hemimandible (PIN 3747-242), which even exceeds that of the holotype from Sainzelles (Fig. 7, Table 2). The analysis of the available comparative material allowed us to identify two morphotypes for lower cheek teeth (m1, p4, and p3) of P. brevirostris, replacing each other in time.
Modifications of m1
The earliest morphotype А (Fig. 8) was identified on the lower carnassial m1 of the earliest finds (i.e., Sainzelles, Upper Valdarno, Liventsovka; Fig. 9g–j). This morphotype shows the posterior crest of the protoconid directed posterolingual, to the place where the metaconid would have been located. In most cases (Fig. 9h–j), this crest does not directly connect to any of the talonid cusps. Then an additional small crest derives from the main crest on the basal part of the protoconid. This additional crest is directed to the talonid. The talonid is wide and bears two cusps. The buccal non-cingular cusp (hypoconid) connects to the additional crest of the protoconid. In case of the absence of the additional crest, the hypoconid connects to the main posterior crest of the protoconid (Fig. 9g). The lingual cusp (entoconid) is present and is formed by the cingulid (Figs. 8 and 9g–j).
Morphotypes A and B of m1 of Pachycrocuta and basal morphotype of the Pliocene Pliocrocuta sp. in accordance with the stratigraphic positions of the finds.
Basal morphotype of the m1 of the Pliocene Pliocrocuta sp. and morphotypes A and B of m1 of Pachycrocuta brevirostris from the Early and early middle Pleistocene localities: (a) Nalaikha (PIN 3747-122); (b) Zasukhino-3 (GIN 4370/99, reversed); (c) Lakhuti-2 (GIN 3848/248-67); (d) Lakhuti-2 (GIN-3848/282-67, reversed); (e) ZKD-1 (PIN 538-531, reversed); (f) Odessa Catacombs (PIN-390-5); (g) Upper Valdarno (IGF 847, reversed); (h) Upper Valdarno (V.A. 1191, reversed); (i) Liventsovka (RSU-231); (j) Sainzelles (SAI-2003-15-92, reversed). Scale bar: 1 cm.
The derived morphotype B (Fig. 8) is observed on the m1 from Zasukhino-3 and Nalaikha, as well as on the specimens from Lakhuti-2 and ZKD-1, which are of younger age (Fig. 9a–e). This type of lower carnassial has the protoconid reduced in size relative to the paraconid. In this case, the protoconid always bears only one posterior crest. The crest in the lower third of its height turns buccally to the hypoconid and connects with its anterior ridge (Fig. 10). The talonid is narrow. Its lingual area and the entoconid are reduced. Meanwhile, the buccal cingulid of the talonid increases, forming a small buccal area (space) on which, in most cases, a buccal cingular elevation or а cusp forms.
Position of the posterior crest of m1 protoconid in: (a) Pliocrocuta sp., Odessa Catacombs (PIN-390-5); (b–e) Pachycrocuta brevirostris: (b) Zasukhino (GIN 4370/99, reversed); (c) Lakhuti-2 (GIN 3848/248-67); (d) Nalaikha (PIN 3747-122); (e) ZKD-1 (PIN 538-531, reversed). Posterior view. Not to scale.
In conclusion, we would like to emphasize that according to our analysis, the hypoconid is always the main talonid cusp that connects with the posterior crest of the protoconid (with the main crest or the additional one), while the rest of the talonid cusps have a cingular origin and never connect with the protoconid (Fig. 8).
Premolar modifications
In the р3, the more primitive morphotype A (Fig. 11d and e) was found in the oldest samples ranging from 2.0 to 1.4–1.3 Ma, in particular, of the holotype from Sainzelles, as well as materials from Upper Valdarno and Liventsovka. In this morphotype, the posterior cusp is large, and the posterior cingulid is separated from the posterior cusp.
Basal morphotype of the p3 of the Pliocene Pliocrocuta sp. and morphotypes A and B of p3 of Pachycrocuta brevirostris from the Early and early middle Pleistocene localities: (a) Lakhuti-2 (GIN-3848/282-67); (b) Zasukhino-3 (GIN 4370/9359); (c) ZKD-1 (PIN 538-531); (d) Liventsovka (RSU-231); (e) Upper Valdarno (NMB.V.A.1191); (f) Odessa Catacombs (PIN-390-5). mc, main cusp; pc, posterior cusp; cing, cingulum; r, reversed. Scale bars: 1 cm.
In the more derived Pachycrocuta from Zasukhino-3, Nalaikha, Lakhuti-2, and ZKD-1 (morphotype B; Fig.11a–c), the posterior cusp is pressed to the main cusp, and the posterior cingulid almost completely merges with the posterior cusp. The area occupied by the main cusp in this morphotype is larger. The same morphology was noted in the material from other Eurasian late Early Pleistocene localities, for example, Yunxian, China (Échassoux, Reference Échassoux, A.-M, P.-E, Li, Feng, Li, Wu, Lumley and Li Tianyuan2008).
A similar picture is observed in the p4 of the same samples. The earlier morphotype A (Fig. 12d–f), shows the posterior cingulid well pronounced and pulled back from the posterior cusp. In occlusive view, an anterobuccal cingulid is strong. In the more derived morphotype B (Fig. 12a–c), the cingulid is reduced, the posterior cingulid partially merges with the posterior cusp, and the anterobuccal cingulid is weak.
Basal morphotype of the p4 of the Pliocene Pliocrocuta sp. and morphotypes A and B of the p4 of Pachycrocuta brevirostris from the Early and early middle Pleistocene localities: (a–f) P. brevirostris, (a) Lakhuti-2 (GIN-3848/282-67); (b) Zasukhino-3 (GIN 4370/93); (c) ZKD-1 (PIN 538-531); (d) Liventsovka (RSU-231); (e) Upper Valdarno (e) NMB.V.A.1191, (e’) NMB.V.A.1673; (f) Sainzelles (SAI-2003-15-92); (g) Pliocrocuta sp., Odessa Catacombs (PIN-390-5); mc, main cusp; ac, anterior cusp; pc, posterior cusp; cing, cingulum; r, reversed. Scale bars: 1 cm.
Discussion
Historical review on cusp homology and talonid modifications
The process of development of the cusps on the m1 talonid and its significance for understanding the evolution of some hyaenid lineages have been repeatedly discussed in the literature. Some researchers are of the opinion that the change in the number of cusps on the Pachycrocuta talonid is not coordinated with the geological age of the finds (Kurtén, Reference Kurtén1972; Werdelin and Solounias, Reference Werdelin and Solounias1991). Other scholars use a count of the number of cusps and their relative sizes on the talonid to divide P. brevirostris into different subspecies (Liu et al., Reference Liu, Liu, Zhang, Wagner, Jiangzuo, Song, Liu, Wang and Jin2021). At the same time, the homology/nature of these cusps and the processes occurring during the restructuring of the talonid are usually overlooked. Thus, usually only the cusps are used as morphological characters, and other morphological structures (e.g., the position of the posterior crest of the protoconid) are not mentioned. However, when initially describing the genus Pachycrocuta (Kretzoi, Reference Kretzoi1938) and discussing the taxa (later assigned to this genus) by Pei [Reference Pei1934] and Kurtén [Reference Kurtén1956], the authors addressed a wider range of features important for understanding the evolutionary processes of the lower carnassial tooth.
Initially, Pei (Reference Pei1934, p. 91) noted that during the evolution of hyaenids, the transition from a tricuspid talonid to a unicuspid occurred. This process led to the “disposition of the cutting edge” of the talonid from the buccal to the lingual side. According to Pei’s theory (Reference Pei1934), this shift was “not due to an abnormal connection of the protoconid with the endoconid” but “to a shifting of the hypoconid towards the lingual part of the heel.” In other words, Pei (Reference Pei1934) noticed that during the reconstruction of the talonid, the hypoconid, as the main cusp, must always remain connected to the protoconid. Therefore, the cusp appearing on the lingual part of the talonid is not an entoconid, but a lingually displaced hypoconid.
In contrast to Pei (Reference Pei1934), Kurtén (Reference Kurtén1956) considered the remaining single cusp of the m1 talonid to be entoconid, rather than hypoconid. However, Kurtén (Reference Kurtén1956) did not give any explanation for this theory, although it is used to explain the homology of the cusps in recent works (Liu et al., Reference Liu, Liu, Zhang, Wagner, Jiangzuo, Song, Liu, Wang and Jin2021) and this will be discussed later.
Kretzoi (Reference Kretzoi1938), when describing the form from Gombaszög (“Pachycrocuta robusta progressa”), noted that the posterior crest of the protoconid on the m1 has a posterior position, instead of a lingual position. Thus, Kretzoi (Reference Kretzoi1938) was the first to note the difference in the position of the posterior crest of the protoconid in crocutoid hyenas and the fact that the crest always tends toward the metaconid. Unfortunately, this feature was subsequently hardly used in the description and analysis of Pachycrocuta.
The changes concerning the position of the hypoconid noted by Pei (Reference Pei1934) and the posterior position of the posterior crest of the protoconid in forms synchronous with Gombaszög were confirmed during the study of the materials from Zasukhino-3, Nalaikha, Lakhuti-2, and ZKD-1. Moreover, in the current study, these features were expanded and modified.
An unusual feature of the lower carnassial tooth of Pachycrocuta is the presence of a cusp on the posterior crest of the protoconid (Fig. 13). This phenomenon is absent in our material, but it presents on the finds from Stránská skála (1.0–0.9 Ma; Kurten, Reference Kurtén1972) and Taurida cave (1.8–1.5 Ma, Crimea; Khantemirov et al., Reference Khantemirov, Gimranov and Lavrov2021; Lavrov et al., Reference Lavrov, Gimranov, Startsev and Lopatin2021; Fig. 13). It also is periodically encountered in the description of other crocutoid hyenas (Pei, Reference Pei1934; Kretzoi, Reference Kretzoi1938; Kurtén, Reference Kurtén1972). Pei (Reference Pei1934) was the first to note that it may be a rudimentary metaconid included in the posterior crest of the protoconid. Kretzoi (Reference Kretzoi1938) argued that the term “metaconid” for the cusp lying on the talonid immediately behind the posterior crest of the protoconid is incorrect and that it is simply a hypoconid, which in extreme cases may resemble a metaconid.
Lower molar m1 of Pachycrocuta brevirostris from the locality of Taurida Cave showing the shifted metaconid to the posterior crest of the protoconid; IPAE 727/81, (a) buccal, (b) lingual, (c) occlusal, (d) postero-lingual view. Scale bar: 1 cm.
Apparently, Pei’s opinion about the inclusion of a rudimentary metaconid in the posterior crest of the protoconid is more correct, because this character was common for another member of Feliformia, Felidae. In this group, a metaconid–talonid complex was formed on the posterior crest of the protoconid, with two cusps placed one behind the other or one above the other. A similar structure is also known in Percrocutidae, and exactly because of the presence of this feature, the taxonomy of this group of carnivorans was revised (Chen and Schmidt-Kittler, Reference Chen and Schmidt-Kittler1983).
Evolutionary trends in the lower cheek teeth of crocutoid hyenas
Teeth of extant hyenas display both generalized (Hyaena) and more advanced (Crocuta) features associated with the development of their crocutoid specialization. The extinct Pleistocene Pachycrocuta shows a transitional intermediate morphology (Werdelin and Solounias, Reference Werdelin and Solounias1991).
Lower carnassial m1
Our study reveals that Pachycrocuta features are best expressed in the modification of the lower carnassial (m1). Compared with the highly specialized m1 (a low disproportionately elongated tooth with a strongly reduced talonid) of the genus Crocuta, the m1 of Pachycrocuta preserved the typical hyaenid proportions (Fig. 14), and its specialization is mainly associated with the loss of the metaconid. We established that metaconid reduction is directly related to the restructuring of the posterior part of the m1, in particular to changes in the location of the posterior crest of the protoconid and simplification of the m1 talonid.
Comparison of the proportions of m1 of Crocuta (a), Pachycrocuta (b), and Hyaena (c), buccal view. Not to scale.
Usually, hyaenids retaining the metaconid on the m1 (basal morphotype, in particular, more primitive Pliocrocuta), show the posterior crest of the protoconid descending lingually to the developed or reduced metaconid. There is also a small additional basal crest of the protoconid connected to the hypoconid (Figs. 8 and 9f). In turn, although in Pachycrocuta the metaconid is completely reduced, in the earlier forms (early stage, morphotype A; Figs. 8 and 9g–j) the posterior crest of the protoconid is located on the lingual side of the protoconid. Similar to the basal morphotype (Figs. 8 and 9f), a small additional basal crest of the protoconid servs to connect the main crest with the hypoconid. However, in later forms of Pachycrocuta (later stage, morphotype B; Figs. 8, 9a–e, and 10), the posterior crest of the protoconid shifts from the lingual to the posterior side. This process occurred synchronously with a reduction of the lingual cusp (entoconid). The loss of the entoconid was accompanied by a reduction of the lingual space of the talonid. As a result, the hypoconid becomes a single central cusp, directly connected to the posterior crest of the protoconid, while the additional basal crest of the protoconid disappears. Similar morphology appears in some present forms of C. crocuta, although the talonid is more reduced (Fig. 15).
Lower carnassial tooth m1 of the resent Crocuta crocuta; S-152715, occlusal view. 1, paraconid; 2, protoconid; 3, posterior crest of the protoconid; 4, hypoconid; 5, buccal talonid space. Not to scale.
However, the current analysis of the m1 showed that the retention of only one cusp was not optimal for the Pachycrocuta talonid. Its further more advanced modifications are observed on the later forms from Zasukhino-3, Nalaikha, Lakhuti-2, and ZKD-1 from the late Early Pleistocene and early middle Pleistocene (Figs. 8, 9a–e, and 10). The talonid m1 of these forms is characterized by a reduced lingual cusp and lingual talonid space. The loss of this space is compensated by an increased buccal area of the talonid due to the development of the buccal cingulid. On the resulting buccal space of the talonid, a ridge-like elevation or even a cusp is formed (Figs. 8 and 9a–e), which is often mistakenly identified as a hypoconid. But in fact, the main cusp (hypoconid), is displaced lingually, thereby creating the impression of the presence of a large entoconid on the lingual side of the talonid. However, the hypoconid is easily distinguished from the other cusps by its connection to the posterior crest of the protoconid.
A similar confusion apparently occurred with the interpretation of the homology of the talonid cusps when describing the material from JYC (China) (Liu et al., Reference Liu, Liu, Zhang, Wagner, Jiangzuo, Song, Liu, Wang and Jin2021). Accepting the opinion of Kurtén (Reference Kurtén1956), these authors noted that some samples of P. brevirostris (JYC, ZKD-1) with bicuspid m1 talonid, have the lingual cusp (that they considered to be entoconid) larger than the buccal one, in contrast to the earlier samples of P. brevirostris, where the hypoconid is the dominant cusp. Based on this observation, Liu et al. (Reference Liu, Liu, Zhang, Wagner, Jiangzuo, Song, Liu, Wang and Jin2021, p. 12) supposed “a tendency for the entoconid to eventually replace the hypoconid as the only single cuspid.” However, when we look carefully at the talonid of m1 of specimen IVPP V26310 from JYC (Liu et al., Reference Liu, Liu, Zhang, Wagner, Jiangzuo, Song, Liu, Wang and Jin2021, Fig. 5C), we can clearly see that the lingual cusp is connected to the single posterior crest of the protoconid. This fact makes it clear that the lingual cusp is, in fact, a hypoconid, rather than an entoconid, and the buccal cusp is a cingular formation, similar to those of the derived morphotype B. Thus, the unusual talonid structure with a “large lingual entoconid” and a “small hypoconid located buccally” might be explained by the presence of the m1 morphotype B (specimen IVPP V26310 from JYC). This leads us to consider the possibility of an earlier appearance of Pachycrocuta with the lower carnassial of morphotype B in the southern regions of Asia (China). However, this issue is beyond the scope of this work.
Thus, the modifications of m1 discussed above appear in two stages, characterized by successive, stratigraphically significant morphotypes A (first stage) and B (second stage) (Figs. 8 and 9). The transition to the late morphotype B can be most clearly traced in the m1 in the fossil population (five specimens) from the locality of Untermassfeld, Germany, 1.1 Ma (Iannucci et al., Reference Iannucci, Mecozzi, Kahlke, Sardella and Kahlke2022, Fig. 6). In contrast to the earlier forms with morphotype A from Upper Valdarno, Sainzelles, and Liventsovka, the specimens from Untermassfeld have a rather narrow talonid with a reduced lingual area devoid of an entoconid. Moreover, on the m1 talonid (IQW 1988/ 22 601 (Mei. 22 120)) there is a distinct cingulate elevation on the buccal side. The buccal elevation is also traced in another m1 (IQW 1995/25 336 (Mei. 24 865)) (Iannucci et al., Reference Iannucci, Mecozzi, Kahlke, Sardella and Kahlke2022, Fig. 6). These features are characteristic of the more advanced morphotype B (Figs. 8 and 9a–e).
However, the population from Untermassfeld shows peculiarities. On m1 of IQW 1995/25 336 (Mei. 24 865), IQW 1986/21 365 (Mei. 20 888), and IQW 1990/23 457 (Mei. 22 976), a second talonid cusp occurred, located postero-buccally with regard to the hypoconid and connected with it. The nature of this cusp is unclear, but most likely it is a bifurcated hypoconid or hypoconulid.
Premolars
The current study of the dental characters confirmed that Pachycrocuta premolars improved their crushing function. Representatives of the genus Hyaena have the most generalized type of premolars, which increase in size uniformly from p2 to p4. In contrast, in Pachycrocuta, there was a functional differentiation of these teeth, expressed primarily in the sharp increase of p3. The proportions of this tooth changed relative to the other premolars due to the increase in its width and height, as well as in the notable increase in the area occupied by its main cusp (Fig. 11, Table 2).
In Pachycrocuta premolars, changes can be traced through modifications of the cingulids. These structures merged with the posterior cusps at later stages of the genus’s evolution (Figs. 11 and 12). This evolutionary trend in premolars led to a strengthening of the cracking function, which aligns with the recent work based on tooth proportions (Pérez-Claros, Reference Pérez-Claros2024).
Distribution and ecology
The appearance of Pachycrocuta in Europe is an important biochronological event. Recently, it was dubbed as the “Pachycrocuta-event” that marks the beginning of the late Villafranchian. These hyenas spread widely in Europe and were an essential component of the carnivoran guild until the end of the Early Pleistocene (Azzaroli, Reference Azzaroli1983; Martínez-Navarro, Reference Martínez-Navarro, Fleagle, Shea, Grine, Baden and Leakey2010; Martínez-Navarro et al., Reference Martinez-Navarro, Bartolini Lucenti, Palmqvist, Ros-Montoya, Madurell-Malapeira and Espigares2021).
In Europe, the earliest reliable occurrence of P. brevirostris is recorded in the fauna of Olivola in Italy (Weithofer, Reference Weithofer1889; Iannucci et al., Reference Iannucci, Mecozzi, Sardella and Iurino2021). The sediments containing this fauna have reverse magnetism and are correlated with the pre-Olduvian part of the Matuyama chron, dated to 2.1 Ma (Napoleone et al., Reference Napoleone, Albianelli, Azzaroli, Bertini, Magi and Mazzini2003). Similar in age finds of P. brevirostris were also reported from Spain, Fonelas P-1, ∼2.0 Ma (Madurell-Malapeira et al., Reference Madurell-Malapeira, Ros-Montoya, Espigares, Alba and Aurell-Garrido2014); Italy, Poggio Rosso, ∼1.92–1.77 Ma (Mazza et al., Reference Mazza, Bertini and Magi2004); Greece, Gerakarou-1, 1.8 Ma (Koufos, Reference Koufos2014); Georgia, Dmanisi, 1.8 Ma (Gabunia and Vekua, Reference Vekua1995; Vekua, Reference Vekua1995), and Russia, Liventsovka, 2.0 to 1.6–1.4 Ma (Azov region; see “Problems of Generic and Subspecies Diagnostics of Crocutoid Hyaenids (Pliocrocuta and Pachycrocuta)”) (Sotnikova et al., Reference Sotnikova, Baigusheva and Titov2002; Tesakov, Reference Tesakov2021). The holotype of P. brevirostris comes from the locality in France, also dating to the mid-Early Pleistocene (Sainzelles, 1.67–1.3/1.4 Ma; Gervais, Reference Gervais1850; Thouveny and Bonifay, Reference Thouveny and Bonifay1984). In addition, new material on P. brevirostris was recently described from Taurida Cave, the Black Sea region, 1.8–1.5 Ma (Lavrov et al., Reference Lavrov, Gimranov, Startsev and Lopatin2021). Thus, in the first half of the Early Pleistocene of Europe, P. brevirostris was widespread in the Mediterranean and Black Sea regions.
Approximately from 1.25 to 0.7 Ma, Eurasia experienced the so-called Mid-Pleistocene Revolution, a period of climatic instability characterized by several intense glacial periods. During this time, the glacial cycles became gradually longer as a response to changes in the Earth’s orbital eccentricity, reaching up to 100 ka. This marked a shift toward a highly nonlinear climate system and was accompanied by a notable increase in the global ice volume around 0.94 Ma. These climatic changes had a significant impact on both the biota and the physical landscape, leading to gradual environmental changes and latitudinal shifts in European biomes, influencing mammalian dispersal (Clark et al., Reference Clark, Archer, Pollard, Blum, Rial, Brovkin, Mix, Pisias and Roy2006; Palombo, Reference Palombo2016; Madurell-Malapeira et al., Reference Madurell-Malapeira, Alba, Espigares, Vinuesa, Palmqvist, Martínez-Navarro and Moya-Sola2017).
During this period, the range of P. brevirostris expanded northward in Europe. The more northern records are known from Mosbach, Untermassfeld (Germany; von Koenigswald and Tobien, Reference Von Koenigswald and Tobien1987; Iannucci et al., Reference Iannucci, Mecozzi, Kahlke, Sardella and Kahlke2022), Stránská skála (Czech Republic; Kurtén, Reference Kurtén1972), and Gombaszög (Slovak Republic; Kretzoi, Reference Kretzoi1938). The northernmost European finds of P. brevirostris come from the second part of the Early Pleistocene of England, specifically from the locality of Westbury-sub-Mendip (Bishop, Reference Bishop1982; Lewis et al., Reference Lewis, Pacher and Turner2010). Its records are also known from the English sites of Overstrand, Sidestrand, and Bacton; however, their exact age is questionable (Lewis et al., Reference Lewis, Pacher and Turner2010; Iannucci et al., Reference Iannucci, Mecozzi, Sardella and Iurino2021). The available record shows that P. brevirostris still inhabited the southern Mediterranean and Black Sea regions in the second half of the Early Pleistocene, as evidenced by the faunal data from Ceyssaguet, Pirro Nord, Nogaisk, Vallonnet Cave, Grosse Marguerite Cave, Akhakalaki 1, and elsewhere (Vekua, Reference Vekua1986; Moullé, Reference Moullé1992; Tsoukala and Bonifay, Reference Tsoukala and Bonifay2004; Petrucci et al., Reference Petrucci, Cipullo, Martínez-Navarro, Rook and Sardella2013; Fourvel and Lateur, Reference Fourvel and Lateur2015; Iannucci et al., Reference Iannucci, Mecozzi, Sardella and Iurino2021; Marciszak et al., Reference Marciszak, Semenov, Portnicki and Derkach2021).
According to a recent review of Chinese finds by Liu et al. (Reference Liu, Liu, Zhang, Wagner, Jiangzuo, Song, Liu, Wang and Jin2021), the stratigraphic range of P. brevirostris in Asia is much broader than in Europe. The first appearance is registered in the faunas of China, particularly in the Longdan fauna (2.3–2.2 Ма; Qui et al., Reference Qiu, Deng and Wang2004). The latest middle Pleistocene presence of P. brevirostris was noted in China as well (in localities of ZKD-1, Hulu Cave, etc.; Pei, Reference Pei1934; Liu et al., Reference Liu, Liu, Zhang, Wagner, Jiangzuo, Song, Liu, Wang and Jin2021). Asian records of these hyenas are also known from the Early Pleistocene localities of Pakistan (Turner and Antón, Reference Turner and Antón1996), Java (Brongersma, Reference Brongersma1937), Kazakhstan (Sotnikova et al., Reference Sotnikova, Dodonov and Pen’kov1997), and Tajikistan (Sotnikova and Vislobokova, Reference Sotnikova and Vislobokova1990).
Finds of P. brevirostris in Asian Russia (Zasukhino-3) and Mongolia (Nalaikha) have expanded our understanding of the distribution of these hyenas at the end of the Early Pleistocene. The material described herein represents the northernmost record of the distribution of P. brevirostris in Asia. The latitudinal position of these sites (correspondingly ca. 52°N and 47°N) is close to the northernmost European records in England (ca. 55°N).
In Asian Russia (in particular in western Transbaikalia), as well as in northern Mongolia, cooling and aridization during the terminal Early Pleistocene led to the formation of open landscapes (Alexeeva, Reference Alexeeva2005; Erbajeva et al., Reference Erbajeva, Khenzykhenova and Alexeeva2013; Kahlke, Reference Kahlke2014). Carnivoran taxa associated with P. brevirostris from the studied Asian localities are also present in other Eurasian localities of temperate latitudes and of a close age (from 1.1 to 0.78 Ma) (Untermassfeld, Vallonnet Cave, Manastirec, Westbury-sub-Mendip, Gombaszög, Vallparadis, Lakhuti-2, ZKD-9 and 13, etc.; Kretzoi, Reference Kretzoi1938; Kurtén and Garevski, Reference Kurtén and Garevski1989; Sotnikova, Reference Sotnikova1989; Moullé, Reference Moullé1992; Qiu, Reference Qiu2006; Lewis et al., Reference Lewis, Pacher and Turner2010; Madurell-Malapeira et al., Reference Madurell-Malapeira, Morales, Vinuesa, Boscaini, Gibert and Ferràndez-Cañadell2015; Iannucci et al., Reference Iannucci, Mecozzi, Kahlke, Sardella and Kahlke2022). The following taxa are mainly associated with open-space habitats and are represented by Canidae (Хеnосуоn lycaonoides, Canis mosbachensis/variabilis) and Felidae (Panthera gombaszoegensis, Homotherium latidens). The same taxa are present in the fauna of Zasukhino-3 and Nalaikha (Table 1) (Vangengeim and Sotnikova, Reference Vangengeim and Sotnikova1981; Zhegallo et al., Reference Zhegallo, Zazhigin, Kolosova, Malaeva, Murzaeva, Sotnikova, Vislobokova, Dmitrieva and Dubrovo1982; Sotnikova, Reference Sotnikova1988, Reference Sotnikova1989, Reference Palombo2016). Pachycrocuta brevirostris was the only representative of the family Hyaenidae among the carnivorans. The feeding habits of the giant hyena were shown to be hunting and scavenging, and among this carnivoran assemblage, it was most likely kleptoparasitic to the large hypercarnivores H. latidens and Х. lycaonoides, subsisting on carcasses of large ungulates (Sotnikova, Reference Sotnikova and Ralf-Dietrich2001; Palmqvist et al., Reference Palmqvist, Martínez-Navarro, Pérez-Claros, Torregrosa, Figueirido, Jiménez-Arenas, Patrocinio Espigares, Ros-Montoya and de Renzi2011; Madurell-Malapeira et al., Reference Madurell-Malapeira, Bartolini-Lucenti, Prat-Vericat, Costeur, Rook, Cherin and Collareta2023).
It is noteworthy that this list of carnivoran taxa associated with Pachycrocuta were part of the Early Pleistocene Olyorian fauna, which existed within the Arctic Circle in northern Siberia (northeastern Russia) (Sher, Reference Sher1987). The ice sheet in this region formed at the beginning of the Pleistocene, resulting in the development of a harsh climatic environment characteristic of the tundra-steppe zone (Sher et al., Reference Sher, Kaplina, Giterman, Lozhkin, Arkhangelov, Kiselyov, Kouznetsov, Virina and Zazhigin1979; Kahlke, Reference Kahlke2014). Some of these taxa (e.g., Хеnосуоn, jaguar-sized Panthera, Homotherium) have also been recorded in North America (Antón et al., Reference Antón, Salesa, Galobart and Tseng2014). The exception is P. brevirostris, which is absent from the Olyorian fauna of northern Siberia (Sotnikova, Reference Sotnikova2018). From this, it can be inferred that Pachycrocuta was more dependent on climatic factors—most likely temperature—than the other carnivorans, which hindered its northward expansion. As a result, the giant hyaenas were unable to cross the Bering Land Bridge and reach the North American continent.
Conclusions
In this work, the material on P. brevirostris from the localities of Zasukhino-3 and Nalaikha was described for the first time. The significance of this material lies in its in situ origin and well-defined stratigraphic position. A detailed comparison was made with both European and Asian crocutoid hyenas from Plio-Pleistocene localities.
Comparative analysis revealed evolutionary trends in the lower teeth of P. brevirostris during their existence in Eurasia for more than 1.5 Ma. The features we could trace allowed us to identify two evolutionary stages characterized by morphotypes A and B in the dental features of the lower molar and premolars of P. brevirostris. The earlier morphotype A is characteristic of the late Villafranchian, while the later morphotype B is found in forms of the terminal Early Pleistocene through the beginning of the middle Pleistocene. The transition between the morphotypes occurred at ca. 1.1 Ma.
Advanced modifications of the premolars aimed at differentiating their function, appeared in Pachycrocuta for the first time in the Pliocene–Pleistocene crocutoid hyenas. . They included a strong increase of the p3 size in the tooth row, an increase of the massiveness of its anterior part, a merging of cingulids in the p3 and p4 with their posterior cusps, and a reduction of the antero-lingual cingulid of the p4.
In the lower carnassial m1, modifications were aimed at enhancing the cutting function of the m1 of P. brevirostris. It was found that the changes in talonid m1 are not due to variations in the number of cusps, as is commonly believed, but rather in the modification of their position on the talonid. In this process, the posterior crest of the protoconid plays a key role, as it changes its direction synchronously with the relocation of the hypoconid and alterations of the talonid space.
During the evolution of P. brevirostris, the hypoconid of m1 was displaced toward the center of the talonid due to a reduction in both the lingual space and the lingual cusp (entoconid). At the same time, the posterior crest of the protoconid shifted from the lingual to the posterior side. At this stage, a single cusp remains on the talonid, directly connected to the main posterior crest of the protoconid. However, to strengthen the talonid that had lost the entoconid, a cingular elevation (sometimes in a cusp-like shape) was formed on the buccal side of the talonid. Therefore, in the late stages of development of the talonid of P. brevirostris, the large lingual cusp is often mistaken for an entoconid. It was found that these changes in the lower teeth can be used as a biostratigraphic proxy.
The described finds of P. brevirostris from Asian Russia (Zasukhino-3) and Mongolia (Nalaikha) are the northernmost known records of P. brevirostris in Asia, which align with the northern boundary of this species’ range in Europe.
The carnivoran taxa from the studied Asian localities are also present in other late Early Pleistocene Eurasian localities of similar age, as well as in the Olyorian fauna of northern Siberia. However, P. brevirostris was absent from higher latitudes, likely due to its temperature sensitivity, which prevented its expansion across the Bering Land Bridge to North America.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/qua.2025.10027.
Acknowledgments
We are grateful to Serdar Mayda (BDEU) for sharing the photos of hemimandibles NMB.V.A.1191 and NMB.V.A.1673; to Dmitriy Gimranov (IPAE) and Daniyar Khantemirov (UrFU) for sharing the photo of m1 IPAE 727/81 from Taurida Cave; to Frédéric Lacombat (MCP) for the opportunity to study the holotype of P. brevirostris; to Alexander Lavrov (PIN RAS) for sharing the photo of the C. crocuta tooth S-152715; to collection curators, for the opportunity to study fossil materials (PIN RAS, RRM) and extant hyaenids (ZIN RAS, ZMMU). We would like to thank Alex Tesakov (GIN RAS) for the critical reading of the manuscript. We are grateful to our reviewers, Paul Palmqvist (University of Malaga) and Margaret Lewis (Stockton University), for their constructive comments and suggestions, which greatly improved the quality of this article. The research was carried out with the assistance of the Center for Integration in Science, Ministry of Aliyah and Integration, State of Israel (PN). It was conducted as part of the scientific work plan of GIN RAS (PN and MS).
Competing interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Open access
