This study examines the ecomorphological characteristics of two Asian rhinoceros species: the critically endangered Sundaic rhinoceros and the vulnerable Indian rhinoceros. Among the five living rhinoceros taxa, the three Asian species are notable for their tusked incisors. Fossil evidence highlights the divergence between Rhinoceros and Eurhinoceros in cheek tooth morphology, linked to different dietary specialisations. The Sundaic rhinoceros, a generalist browser restricted to the Ujung Kulon peninsula of Java, exhibits distinctive features such as a grey hide with polygonal patterns, a typical 'saddle' on the nape, a slender head shape and a protrusion instead of a horn in females. The latter is a unique trait among Rhinocerotini species. In contrast, the Indian rhinoceros, a variable grazer, inhabits riverine grasslands in northern India and southern Nepal, displaying deep skin folds and tubercles. Ecological behaviours differ significantly, with the Sundaic rhinoceros being solitary wanderers and Indian rhinoceros forming temporary crashes. Both species possess unique adaptations for survival, emphasising the importance of understanding their systematics for effective conservation. The study further examines the interrelationships among the one-horned Asian species of the Rhinocerotidae family, highlighting their distinct features. The revision delves into skull morphology, dentition, and ecological dynamics, revealing evolutionary patterns and ancestral traits. Both single horned rhinoceroses went a separate and diverging way of evolution that was not triggered by geographical separation but by niche partitioning. Comparative analyses shed light on the evolutionary trajectory and ecological adaptations of each species. The fossils, the ecological and morphological adaptations of both species, suggest designating 'Rhinoceros' sondaicus as distinct from Rhinoceros unicornis, under the one-horned rhinoceros Eurhinoceros, as proposed by Gray (1868). Eurhinoceros sondaicus emerges as a persistently more primitive form.

Ecology, Eurhinoceros sondaicus, Indian rhinoceros, Javan, morphology, palaeontology, Rhinoceros unicornis, Sundaic rhinoceros, systematics

“The forehead and the nose behind the base of the horn flat, both in the living animal and skull. Eurhinoceros.” With these words, John Edward Gray (1868: 1009) described a new one-horned rhinoceros to classify the Sundaic rhinoceros. A century later Heißig (1972) considered valid the genus Eurhinoceros to significantly distinguish E. sondaicus (Desmarest, 1822), a separate one-horned rhinoceros from Rhinoceros to which belongs the Indian rhinoceros Rhinoceros unicornis Linnaeus, 1758. Nardelli (1988: 41) foresaw that 'Rhinoceros' sondaicus could be classified in a separate genus from Rhinoceros unicornis, suggesting that differences in their morphology and behaviour – such as the distinct external features of 'R.' sondaicus’ head, especially its lips, and the fact that the former is a browser and the latter a grazer – support this distinction.

This work aims to examine the differing ecomorphological characters of Rhinoceros and Eurhinoceros. The results suggest that the single-horned rhinoceroses, Rhinoceros unicornis and Eurhinoceros sondaicus, followed separate evolutionary paths not due to geographical isolation but rather as a result of niche partitioning. This can be followed by a number of stepping stones from the Middle Miocene onwards.

Heißig (1972: 29) reported the parallel presence of Gaindatherium and Eurhinoceros aff. sondaicus in the Middle Miocene Nagri formation of the Siwalik region of Pakistan. Fossil records of ancient rhinoceros species provide insights into their evolutionary history and the environmental conditions they inhabited (MacFadden 1998). By examining the morphology and geographic distribution of fossils, researchers can uncover patterns of niche partitioning and evolutionary divergence. Extant examples include the browser Diceros bicornis Linnaeus, 1758 and the exclusive grazer Ceratotherium simum Burchell, 1817, which are sympatric today (MacFadden 1998: 282), as well as the variable grazer Rhinoceros unicornis and the generalist browser Eurhinoceros sondaicus (Pandolfi et al. 2021: 9 and references therein as Rhinoceros), which had some overlapping ranges.

Pandolfi and Maiorino (2016) provide data on the coexistence of ‘Rhinoceros’ sondaicus and Rhinoceros unicornis in certain parts of their former ranges, with dietary specialisation playing a key role. The major distinctions between these two species are found in their dental morphology, indicating adaptation to different feeding strategies. Eurhinoceros sondaicus is more adapted to browsing on softer, leafy vegetation, while R. unicornis is more inclined towards grazing on tougher, more abrasive plants material. This divergence in feeding habits minimised direct competition for resources, thus facilitating their coexistence.

Despite some similarities in cranial features, their teeth reflect evolutionary adaptations to different ecological niches, with R. unicornis showing adaptations for grazing-related diets and E. sondaicus for browsing. This separation allowed these species to share habitats without exhausting common food sources (Pandolfi and Maiorino 2016).

Extant species of Rhinocerotini Gray, 1821 (sensu Pandolfi 2015) exhibit diverse feeding habits that include grazing, mixed feeding, and browsing. Likewise, the diets of Pleistocene species were varied. In their study, Hernesniemi et al. (2011) analysed mesowear patterns in both extant and Pleistocene species, comparing Diceros bicornis, Ceratotherium simum, Dicerorhinus sumatrensis (Fischer, 1814), ‘Rhinoceros’ sondaicus, and Rhinoceros unicornis to fossil samples from Pleistocene Rhinocerotini, such as Stephanorhinus kirchbergensis (Jäger, 1839), S. hemitoechus (Falconer, 1859), S. hundsheimensis (Toula, 1902), and Coelodonta antiquitatis Blumenbach, 1799. When clustered by mesowear scores from the first and second molars, Stephanorhinus kirchbergensis grouped closely with Dicerorhinus sumatrensis, while S. hundsheimensis showed affinity with ‘Rhinoceros’ sondaicus. These browsing species, together with the somewhat separate Stephanorhinus hemitoechus, were distinctly separated from Rhinoceros unicornis, a variable grazer, which clustered near the grazer Coelodonta antiquitatis and the exclusive grazer Ceratotherium simum.

In the case of Rhinoceros unicornis and Eurhinoceros sondaicus, the results of the present study imply that environmental pressures led to the development of differences in dental morphology, specialised feeding strategies, habitat adaptations and behavioural traits. Despite inhabited some overlapping areas, the two species evolved independently.

“The key differences between R. unicornis and R. sondaicus are most apparent in their teeth, whereas the morphology of the skull is rather similar” (Pandolfi and Maiorino 2016: 10). Among the five extant species, the Asian ones stand out for possessing tusk-like incisors, a feature absent in African species. The synapomorphy of a sloped orbital floor connects these species without incisors. The three Asian forms can be categorised on a scale of increasing specialisation, ranging from Dicerorhinus sumatrensis with two horns to Eurhinoceros sondaicus and Rhinoceros unicornis with a single horn. The shape and posture of the skull align with this scale, but the same cannot be said for the morphology of the molars. While the molar and premolar type of R. unicornis can be understood as a high-crowned specialisation form of the primitive D. sumatrensis, the type of E. sondaicus diverges, showing similarities to African forms in the weakening of the metaconus rib, molarisation of the premolars, and strengthening of the inner cingulum (Heißig 1981).

The significance of the distinctions in cheek tooth morphology (Table 1) can be estimated only with a look at the paleontological evidence. Nearly all paleontological knowledge depends on the tooth morphology because skulls are rarely preserved. Cheek teeth with similarities to Rhinoceros and Eurhinoceros are known from Lower Miocene deposits in the Himalayan foreland basin (Forster-Cooper 1934). Other less complete specimens from the Lower Miocene Bugti beds, stored in the Natural History Museum, London show the fusion of lingual cusps more or less high above the enamel basis, the presence of a narrow metacone ridge in the upper premolars, and of both crista and crochet as in Rhinoceros (Forster-Cooper 1934: fig. 36). The first skulls with single horns and comparable tooth morphology are known from the Middle Miocene of Europe with Lartetotherium Ginsburg, 1974 and South Asia with Gaindatherium browni Colbert, 1934, the latter with somewhat higher tooth crowns.

The series with increasing tooth height continues to Rhinoceros sivalensis Falconer & Cautley, 1847, which already has the typical skull profile of Rhinoceros. Differences in the combination of crown height, secondary folds, and cingulum formation show that this phylogenetic lineage was more complexly branched so that it is not easy to find the direct line to the living Rhinoceros. It makes no sense to distinguish primitive and progressive morphologies, as these are due to diverging dietary specialisations. Whereas Eurhinoceros was constantly a non-selective browser, the early relatives of Rhinoceros managed the transition from selective browsing to grazing, following the increasing dominance of grasses in the vegetation (Barry et al. 2002). As proposed by Singh et al. (2012), several sub-basins located south of the Himalayas experienced a gradual increase in aridity from approximately 12 million years ago to the present. This environmental shift likely has influenced the dietary habits of the rhinoceroses.

Owing to its confined habitat and the openness of its environment, as well as ex situ breeding programs, the ecological dynamics of R. unicornis are closely monitored and well documented. Extensive studies have been conducted on this species, yielding substantial data and allowing Laurie (1978, 1982), Laurie et al. (1983), Dinerstein (2003), and Bhattacharya (2020a) to provide comprehensive insights into the physical characteristics, socialisation, mating, feeding habits, and other patterns. Despite a fairly extensive body of literature on E. sondaicus, this species has not been given sufficient attention in investigations on the phylogenetic development of the Rhinocerotidae family. As a result, there is presently no consensus on their phylogenetic attributes. Several comparisons bring forth specific facts that would emerge from these studies. As shown in Tables 1, 2, E. sondaicus can be differentiated from R. unicornis based on several key characteristics of the teeth and the skulls as these features are more primitive in E. sondaicus, suggesting a harmonic developmental growth that uniformly impacts nearly all aspects of cranial and dental anatomy. According to Colbert (1942), this clear progression in the skull and tooth characters from E. sondaicus to R. unicornis is unique, allowing certain Pleistocene species of the one-horned rhinoceroses to assume intermediary positions between these two extant forms.

Based on skeletal morphology highlighted by current evidence in this study, E. sondaicus is considered the most primitive member of the one-horned rhinoceroses, exhibiting structural features that, through subsequent developments and adaptations became distinctive for Pleistocene species such as Rhinoceros sinensis, Rhinoceros sivalensis, and the contemporaneous Rhinoceros unicornis (Colbert 1942; Hooijer 1946: 675). Eurhinoceros sondaicus is the sole Asian rhinoceros species well represented by a substantial number of specimens from Pakistan, Myanmar, Thailand, Cambodia, and Vietnam fossil deposits, allowing us to trace its evolutionary history from the Miocene (Heißig 1972) through the early Late Miocene (Longuet et al. 2024) to Plio-Pleistocene (Beden and Guérin 1973; Bacon et al. 2004; Zin-Maung-Maung-Thein et al. 2006, 2010; Khan 2009; Abdul et al. 2014; Suraprasit et al. 2016; Siddiq et al. 2016). In its dentition and skeletal characteristics, the representatives of Pleistocene E. sondaicus mirror the extant Sundaic rhinoceros.

Heißig’s (1981) cladistic analysis suggests that R. unicornis and E. sondaicus share a common ancestral lineage. However, the considerable diversity of fossil Rhinocerotini in South Asia, particularly in the Siwalik beds, presents challenges in attributing isolated remains to specific taxa, obscuring evolutionary inferences. Therefore, the relationship of fossil Rhinocerotini to extant species must be based primarily on dental characters. In most systematics studies, diet and cheek tooth morphology are often thought to be of less importance than other characters, even though dental microwear texture analysis can be employed to deduce ancient diets (Hullot et al. 2019: 398).

Following Heißig’s initial discovery (1972), more findings emerged to confirm the ancient presence of Eurhinoceros in the Siwalik beds of Pakistan. Khan (2009) examined several fossil rhinoceros species from the Pleistocene Siwalik beds, identifying the dental characteristics of some specimens as belonging to Eurhinoceros sondaicus (as R. sondaicus and R. aff. sondaicus). Abdul et al. (2014) recorded the finding of Pleistocene rhinocerotid fossils from localities of Siwalik beds in the Gujrat district at Sar Dhok in Pabbi hills, where specimens belonging to Rhinoceros sivalensis, ‘Rhinoceros’ sondaicus, and Rhinoceros unicornis were recovered as well as finds from Jari Kas in Mirpur district which belong to Eurhinoceros sondaicus and Rhinoceros platyrhinus Falconer & Cautley, 1847. These fossils, dating from the Tatrot and Pinjor stages of the Soan Formation ~ 3.5–0.9 million years ago include well-preserved maxillary and mandibular fragments, in addition to isolated teeth (Abdul et al. 2014). Siddiq et al. (2016) described fossil Rhinoceros aff. ‘R.’ sondaicus collected from various Siwalik bed localities, including the areas of Sar Dhok in Gujrat district, Tatrot in Jhelum district, Jari Kas in Mirpur district of Azad Jammu and Kashmir regions. These samples include maxillary and mandibular fragments, as well as isolated teeth.

Fossils of Eurhinoceros have been also found in Myanmar, Thailand, Cambodia, and Vietnam. Zin-Maung-Maung-Thein et al. (2006) identified right and left maxillae of ‘Rhinoceros’ sondaicus from the upper part of the Irrawaddy (Ayeyarwady) Formation. Additional cranial remains, including upper teeth, were recovered along the Irrawaddy sediments in central Myanmar (Zin-Maung-Maung-Thein et al. 2010). Fossil specimens of Rhinoceros cf. ‘R.’ sondaicus consisting of post-cranial remains were collected from the Tebingan area in the Magway Region of Myanmar (Longuet et al. 2024). The mammalian fauna from the Irrawaddy Formation is estimated based on the presence of several genera with well-established chronological distributions in the Siwalik deposits (Longuet et al. 2023, 2024). Deposits at Khok Sung in the Nakhon Ratchasima province of Thailand have preserved ‘Rhinoceros’ sondaicus cranial, mandibular, and dental fossils (Suraprasit et al. 2016). At the Phnom Loang, Middle Pleistocene to the Holocene site in Cambodia, Guérin (1973) identified ‘Rhinoceros’ sondaicus guthi Guérin, 1973 fossil specimens. Molar fossils provisionally referred to ‘Rhinoceros ’ cf. sondaicus, were identified at the Pleistocene Ma U'Oi cave, northern Vietnam (Bacon et al. 2004). In the above-mentioned studies four taxa were identified based on dental morphology and dimensions. Fossils of Rhinocerotini occur as early as in the Lower Miocene of the Bugti beds. Some of these single finds, stored in the Museum of Natural History, London are still undescribed.

The differences, especially in the premolars, can be associated with the dissimilar diets of R. unicornis and E. sondaicus (Hullot et al. 2019). Rhinoceros unicornis is a variable grass eater and has high-crowned, block-like upper cheek teeth that can withstand, for a lifetime, the abrasion caused by fibres and silicic acid. On the other hand, E. sondaicus is a generalist browser, exposing its premolars to twig fragments, requiring cingula to prevent lesions of the gums. Without the stress of strong abrasion, the teeth could stay mesodont. The premolars possess a narrow metacone rib and a high lingual wall, even if the lingual cusps are separated by lingual and labial furrows. A lingual cingulum may be present or absent, and this stage of molarisation was defined as 'paramolariform' by Heißig (1969: 16). It is rather widespread in the tribe Rhinocerotini and still present in the living R. unicornis.

Members of the Rhinoceros clade are well represented in the Siwalik beds, starting with the genus Gaindatherium which is characterised by the absence of a lingual cingulum and a distinct metacone rib. This lineage appears to transition into the higher-crowned Rhinoceros sivalensis and eventually leads to the even more high-crowned extant species, although earlier and contemporaneous specimens of R. sivalensis appear to be too high-crowned to be considered direct ancestors of Eurhinoceros. Eurhinoceros seems to have arrived relatively early in Java during the Pleistocene, potentially with an ancestor referred to as ‘Rhinoceros’ sivasondaicus Dubois, 1908. Evidence of this evolutionary pathway can be traced through fossil specimens found at Ngandong, Sangiran, Djetis, and Trinil Pleistocene localities in Java (Koenigswald 1935; Hooijer 1964; Antoine 2012).

Looking for earlier members of the Eurhinoceros lineage in the Siwalik collections, Heißig (1972: 29) described two premolars from Sethi Nagri and an upper molar and some lowers from two localities of the Chinji Formation in Pakistan. The specimens come from Dehm’s expeditions to the Siwalik Hills in Pakistan (winter 1955/1956), housed in the Bavarian State Collection of Paleontology and Geology, Munich and the University of Utrecht Collection, Netherlands. They all differ from Gaindatherium by the formation of cingula. The complete set of characters foreshadowing the morphology of the recent Eurhinoceros shows the upper premolar described by Heißig (1972: 18) from the Dhok Pathan Formation of Parlewali. It differs from the premolar figured by Matthew (1929: figs 33–34) with the probable provenance Dhok Pathan by the loss of the metacone rib, the deeper separation of the lingual hills and the formation of a lingual cingulum around the protocone base.

The two premolars were identified as Eurhinoceros aff. sondaicus due to their resemblance to the extant species. Heißig (1972: 30) detailed a M2 and two m3, one of them on a mandible fragment from the Middle Miocene of the Chinji Formation as a probable Eurhinoceros sp. as illustrated by Heißig (1972: pl. 5, figs 8, 9; pl. 7, figs 1, 2). The upper molar, albeit somewhat fragmentary, aligns with the features of the extant species in the faint antecrochet and the subtle protocone constriction, typical traits of most Rhinocerotini. Additionally, it shares the development of a lingual cingulum and the absence of a metacone rib. The absence of the cingulum in the external groove of the lower molars is also not observed in all living Rhinocerotini e.g., Diceros and Dicerorhinus (Heißig 1972: 30). If this determination is correct, these specimens would prove a separation of the one-horned rhinoceroses from the main stock of Rhinocerotini around the Middle Miocene.

At first sight, R. unicornis and E. sondaicus display similarity in traits such as the head position and the partly or nearly complete armour of the skin, although both are distinctly different when analysed in depth. The relatively high position of the head in the variable grazer R. unicornis, compared to the exclusive grazer C. simum, as shown by the forward inclination of the occipital plane, corresponds to its feeding habits, which include consuming very tall elephant grass (Pennisetum purpureum) and reaching for high-hanging twigs in dense forest vegetation (Bales 1996: 274). The armour, however, is not easily understood as convergence and may have been a characteristic of shared ancestral lineage. According to Endo et al. (2009), the development of skin folds in R. unicornis is considered a thermoregulatory adaptation, serving to protect the rhinoceroses from heat.

The morphology of the skull shows a strong correlation with the hypsodont index (Piras et al. 2010). This indicates that adaptations to different feeding habits occur deeper within the rhinoceros evolutionary tree, rather than limited to the species level. These adaptations include distinct modifications to the posterior part of the skull, which is known to vary with feeding behaviour: while the mandible and skull are more constrained by phylogeny and their developmental integration, the upper tooth row is less influenced by this relationship and teeth are the most adaptable cranial structure in evolutionary terms (Piras et al. 2010).

The skull dimensions in adults of the two living species are as follows: the occipito-nasal distance range is 613–694 mm in unicornis and 567–669 mm in sondaicus; the maximum width at the zygomatic arches is 355–435 mm in unicornis and 324–365 mm in sondaicus; unicornis has a mandibular length of 526–600 mm and a condylar height of 277–309 mm, while sondaicus has a mandibular length of 467–518 mm and a condylar height of 208–247 mm (Guérin 1980). At this point, it is practical to delineate the disparities in the skull between the two extant species (Table 2). These distinctions carry significant importance for prospective discussions regarding the phylogenetic position of each taxon in relation to the other (Colbert 1942).

The skull morphology of the two genera differs, with R. unicornis having, in proportion, a larger and heavier skull compared to E. sondaicus. Additionally, the occiput of R. unicornis is higher and narrower, making the dorsal outline of the skull very concave. Overall, while both genera share certain anatomical features, they also exhibit distinct differences in skull morphology (Fig. 7). The analysis of one ‘Rhinoceros’ sondaicus specimen from the Manchester Museum, UK by Cave (1985) reveals several key features that are characteristic of E. sondaicus cranial morphology. The occipital plane is inclined forward. The orbital-aural length is longer than the orbito-nasal dimension. This ratio of orbital-aural length to orbito-nasal dimension is characteristic of E. sondaicus and contributes to its unique cranial proportions. The infra-orbital foramen is positioned above the second premolar tooth, which is a characteristic feature of its cranial anatomy. The mode of anterior attachment of the petro-sphenoidal ligament serves as a distinguishing feature between E. sondaicus and R. unicornis. In E. sondaicus, the anterior attachment of the petro-sphenoidal ligament is characterised by a small tubercle located at the termination of the Eustachian crest. This tubercle serves as the point of connection for the petro-sphenoidal ligament. On the other hand, in R. unicornis, there is no corresponding tubercle at the termination of the Eustachian crest. Instead, the petro-sphenoidal ligament attaches subtly to the posterior free margin of the alisphenoid bone. This difference in the anterior attachment of the petro-sphenoidal ligament between the two taxa is a significant anatomical distinction that aids in the identification and classification of their respective crania (Cave 1985). By examining these features, it is possible to differentiate between their specimens and gain insights into the unique cranial morphology of each species.

Figure 7. 

A Left side view of the skull of Eurhinoceros sondaicus (juv.), specimen NHM-DMA-26801/1, collected in Java, 1838. Credit: Natural History Museum, Oslo. Adapted from Matschiner (2021: 4842), with permission from Elsevier Eds. Photograph by Lars Erik Johannessen B left side view of the skull of Rhinoceros unicornis (juv.), Smithsonian NMAH. Licensed under CC BY-SA 4.0. The image has been modified by isolating the skull and inverting its position horizontally. Photograph by David J. Stang C partial ventral view of the cranium of Eurhinoceros sondaicus D partial ventral view of the cranium of Rhinoceros unicornis. Adapted from Flower (1876: 446, 447). Main cranial differences outlined following Colbert (1942). Refer to Table 2 for numbered annotations.

“The (skulls) shape in the lateral view reflects ecological niche, in particular feeding type from browsing to grazing, and it also represent taxonomic discrimination” (Pandolfi et al. 2021: 1). The lateral shape of rhinoceros skulls plays a crucial role in reflecting their ecological niches, particularly distinguishing feeding habits ranging from browsing to grazing. Browsing species, typically exhibit different skull structures compared to grazing species. In the lateral view, the shape helps identify these feeding behaviours by highlighting morphological traits adapted to their diets, such as the orientation of the occipital area and the development of other cranial features. This differentiation in skull forms also serves as an important indicator of taxonomic distinctions among rhinoceros species, aiding in the classification and understanding of their evolutionary adaptations. These anatomical differences are consistent with the feeding habits, as grazing rhinoceroses like the white and Indian rhinoceroses feed closer to the ground, while browsing species such as the Sundaic and the Asiatic two-horned rhinoceroses feed on leaves and saplings. This connection between skull morphology and dietary preferences can also be traced through fossil records, demonstrating how rhinoceroses adapted to varying environments and food sources over time (Pandolfi et al. 2021). As shown in Fig. 8, a smaller angle o suggests a posteriorly extended occipital crest, which corresponds to a downward-oriented skull posture. When the lateral semicircular canals (LSC) are aligned horizontally, the palatal plane in E. sondaicus is nearly horizontal (Fig. 8A), whereas in R. unicornis, the rostrum and palatal plane tilt slightly downward towards the ground (Fig. 8B). The angle between the plane defined by the LSC of both inner ears and the palatal plane is also smaller in E. sondaicus.

Figure 8. 

Polygonal surface model of A Eurhinoceros sondaicus skull (in yellow) and B Rhinoceros unicornis skull (in blue) with measured angles. Abbreviations: Iscp–angle between plane spanned between the lateral semicircular canals (LSC) of both inner ears and the palatal plane; o–angle of the occipital crest between the occipital plane and the parietal plane; po–angle between occipital plane and palatal plane. Skull drawings and colours have been maintained from the original. Adapted with permission from Schellhorn (2018: 53).

Schellhorn (2018) highlights that the connection between the shape of the occiput and head posture has been recognised for a long time (Zeuner 1934; Bales 1996). Zeuner (1934) was the first to quantitatively analyse the shape of the occiput, specifically the angles o and po, and found that the angle o that is formed between the parietal and occipital planes is smaller in grazing rhinoceroses compared to browsing species. Schellhorn (2018: 58) also affirms: “These results also support the hypothesis that rhinoceroses with large lower second incisors but one small nasal horn, like R. sondaicus, have a horizontal head posture”.

Heißig (1972, 1989) suggested that fossil rhinoceroses with strong lower tusks likely used them as defensive weapons against predators, similar to the behaviour of extant Asian rhinoceroses. This inference is based on the well-documented behaviour of the living species, which use their tusks in combat (Sody 1959; Owen-Smith 1988). In contrast, the African species, which have completely reduced incisors, rely on their large horns for fighting (Dinerstein 2011). Dinerstein (2011) further proposed a potential link between these behaviours and typical head posture, noting that rhinoceroses with strong lower tusks might favour a horizontal head posture, which would facilitate the use of their tusks as weapons.

This idea is reinforced by the observation that a downward head posture is predominantly seen in both living and fossil rhinoceroses with large horns and without lower second incisors (Heißig 1973). Schellhorn (2018) and Benoit et al. (2020) emphasise the intricate relationship between anatomical features, such as the LSC orientation and occipital shape, and behavioural aspects like feeding preferences and head posture. While Benoit et al. (2020) underline the complexity and potential limitations in reconstructing these features accurately in extinct species, Schellhorn (2018) demonstrates a practical application in extant rhinoceroses, providing new insights that could potentially be adapted to fossil records. A pivotal insight from the work of Benoit et al. (2020) is the pronounced influence of phylogenetics on all examined variables in ungulates. Their findings propose that, in ungulates, LSC orientation predominantly corresponds to evolutionary relationships. For a more precise understanding, upcoming research endeavours should scrutinise each subclade independently, demanding a broader sample size for each (Pandolfi et al. 2021).

The one-horned Asiatic rhinoceroses are examples of evolutionary histories driven by ecological pressures. Adaptations of large terrestrial mammals to various environments are linked to the diversity of food items they can consume, which is reflected in the variation of their dental and cranial morphologies. In rhinoceroses, these adaptations are identified in their teeth structure and head posture. Evidence on feeding habits aligns with the positioning of skulls within the morphospace, allowing us to infer the existence of distinct feeding types or ecomorphotypes.

Ecological niche modelling studies have demonstrated differences in habitat preferences and ecological requirements, suggesting niche partitioning as a mechanism for coexistence and evolutionary divergence. These models predict distinct distributions and habitat suitability for each taxon also within overlapping geographical ranges.

Behavioural observations of R. unicornis and E. sondaicus in their natural habitats have provided insights into their substantial dissimilar dietary intake, marked activity patterns and distinct habitat utilisation. These observations indicate specific behaviours supporting the notion of niche partitioning to reduce competition. This separation of ecological niches not only prevented direct competition but also contributed to their distinct evolutionary trajectories. The fossils support this divergence, with evidence indicating that E. sondaicus evolved to exploit a browsing niche, while R. unicornis became increasingly specialised as a grazer. The significant distinctions in tooth morphology, including the variation in wear patterns, reflect their adaptation to different diets and ecological settings over extended geological timescales.

Genetic assessments were not feasible in this study due to the lack of comparative genetic research. Studies on population dynamics and demographic history reveal patterns of genetic diversity, gene flow, and population structure within R. unicornis and E. sondaicus populations.

The morphological and ecological differences between the two taxa are not merely superficial adaptations to different dietary intakes, but significant structural changes that have evolved over paleontological epochs. The distinctions reflect deep evolutionary adaptations, not short-term ecological plasticity. The study identifies key divergences, not minor traits but fundamental anatomical features tied to their evolutionary adaptations. These diversities are expected to be satisfactory in taxonomy to justify genus-level distinctions, as seen in other genera like Ceratotherium and Diceros. Since the extant African rhinoceroses are classified into different genera, it is reasonable to separate Rhinoceros and Eurhinoceros due to their similar dietary adaptations, among a number of other characteristics. By integrating evidence from paleontological records, ecological niche modelling, morphology, behavioural observations and population dynamics, this assessment substantially supports the idea of distinct evolutionary trajectories for Rhinoceros unicornis and Eurhinoceros sondaicus.

Available data do not provide justification for classifying Eurhinoceros sondaicus as a congeneric species with Rhinoceros unicornis or as a subgenus within Rhinoceros. In our view, the phenotypic and adaptive differences observed between the two lineages warrant a reassessment of its taxonomic status at the genus level. This approach not only reflects their evolutionary separation but also provides a clearer framework for better understanding of their distinct characteristics within a phylogenetic context.

We extend our sincere gratitude to the editors and anonymous peer reviewers for their significant and constructive feedback, which greatly improved the manuscript, in particular Nathalie Yonow for skillfully polishing the final text. We are grateful to Robert and Harsanti Morley and Rico Schellhorn for their generous authorisation to use Figs 3 and 8, respectively. We especially thank Luca Pandolfi (University of Basilicata) and Giuliano Doria (Museum of Natural History 'Giacomo Doria' of Genoa) for providing original photographs of specimens featured in Figs 4(A), 5(A, B), and 6(A), and for kindly granting permission for their inclusion in this work. We received extraordinary support from Phaedra Kokkini (British Museum of Natural History) for taking photographs in lingual view of the Indian rhinoceros maxilla, possibly the first ever published for this species, in Fig. 6(B), as well as the image in Fig. 4(B). Spartaco Gippoliti and Tommaso De Francesco are sincerely acknowledged for their insightful suggestions. The Rhino Resource Centre (RRC) website is recognized as the most valuable repository, providing access to the majority of the papers cited below: http://www.rhinoresourcecenter.com/.