The long-trunked elephantids underwent a significant evolutionary stage characterized by an exceptionally elongated mandible. The initial elongation and subsequent regression of the long mandible, along with its co-evolution with the trunk, present an intriguing issue that remains incompletely understood. Through comparative functional and eco-morphological investigations, as well as feeding preference analysis, we reconstructed the feeding behavior of major groups of longirostrine elephantiforms. In the Platybelodon clade, the rapid evolutionary changes observed in the narial region, strongly correlated with mandible and tusk characteristics, suggest a crucial evolutionary transition where feeding function shifted from the mandible to the trunk, allowing proboscideans to expand their niches to more open regions. This functional shift further resulted in elephantids relying solely on their trunks for feeding. Our research provides insights into how unique environmental pressures shape the extreme evolution of organs, particularly in large mammals that developed various peculiar adaptations during the late Cenozoic global cooling trends.
This study presents fundamental findings on the evolution of extremely elongated mandibular symphysis and tusks in longirostrine gomphotheres from the Early and Middle Miocene of northern China. The integration of multiple methods provides compelling results in the eco-morphology, behavioral ecology, and co-evolutionary biology of these taxa. In doing so, the authors elucidate the diversification of fossil proboscideans and their likely evolutionary responses to late Cenozoic global climatic changes.
Proboscideans are known for their exceptionally elongated and versatile trunks (Shoshani, 1998). However, unlike modern elephants, proboscideans underwent a prolonged evolutionary phase characterized by the presence of greatly elongated mandibular symphysis and mandibular tusks (Cuvier and Laurillard, 1850; Schulz et al., 2022; Shoshani, 2000). This elongation can be traced back to the Late Oligocene species Palaeomastodon and Phiomia, which are among the earliest elephantiforms, and continued through to the Late Miocene Stegotetrabelodon, a stem elephantid (Andrews, 1906). Extreme longirostriny, a feature observed in fossil and modern fishes, reptiles, and birds, was relatively rare among terrestrial mammals and its occurrence in large-bodied proboscideans is particularly intriguing. Particularly, during the Early and Middle Miocene (approximately 20–11 Ma), the morphology of mandibular symphysis and tusks exhibited remarkable diversity, with over 20 genera from six families (Deinotheriidae, Mammutidae, Stegodontidae, Gomphotheriidae, Amebelodontidae, and Choerolophodontidae) displaying variations (Tassy, 1996). Why did proboscideans have evolved such a long mandible of so diversified morphology? How did fossil proboscideans use their strange mandibular symphysis and tusks, and what was the role of trunk in their feeding behavior? Finally, what was the environmental background for the co-evolution of their mandible and trunks, and why did proboscideans finally lose their long mandible? These important issuers on proboscideans evolution and adaptation remain poorly understood. Addressing these significant aspects of proboscidean evolution and adaptation requires comprehensive investigations into the functional and eco-morphology of longirostrine proboscideans.
Mandible and skull morphology of gomphotheres
During the Early and Middle Miocene, gomphotheres, as the ancestral group of living elephants, flourished (Osborn, 1936; Lambert, 1992). Gomphotheres include Amebelodontidae, Choerolophodontidae and “Gomphotheriidae”; all possess a greatly elongated mandibular symphysis (Gheerbrant, 2009; Cantalapiedra et al., 2021). Our comprehensive phylogenetic reconstructions contained the majority of longirostrine elephantiform taxa at the species level and strongly supported this taxonomic scheme (Figure 1A, S1, Appendix S1, Data S1, Code S1–S3). The three families were characterised by their distinctive mandible and mandibular tusk morphology (Tassy, 1996; Shoshani, 1996). The paraphyletic “Gomphotheriidae” have clubbed lower tusks (Figures 1B, K, S2B); their mandibular symphysis is relatively narrow. This morphology is rather unspecialised, and the extant elephants are derived from “Gomphotheriidae” (Tassy, 1996; Cantalapiedra et al., 2021). The mandibular symphysis of Amebelodontidae is generally shovel-like and the mandibular tusks are usually flattened and wide. Platybelodon is the most specialised genus within this family; it possesses extremely flattened and widened mandibular tusks with a sharp distal cutting edge (Osborn, 1936) (Figures 1C, I, J, S2A). Choerolophodontidae is unique because it completely lacks mandibular tusks and it has a long trough-like mandibular symphysis (Figures 1D, L, M, S2C–F). A very deep slit is present on each side of the distal alveolar crest (distal mandibular trough edge) (Figure 1L), which is presumably for holding a keratinous cutting plate (Figure 1M), similar to the slits for holding large claws in felids and some burrowing mammals (e.g., anteaters). The anterior mandibular foramen of Choerolophodon is extremely large and tube-like (Figure S2D), which indicates a very developed mental nerve and eponymous artery (Eales, 1926), for the nutrition of tissues (i.e., keratins) grown from symphyseal dermis. The three lineages exhibit different evolutionary states of food acquisition organs (PC1 scores from mandible and tusk characters; see Supplementary Methods) (Figure S4A, B). Different food acquisition organs morphology strongly indicates different methods of food acquisition among the three gomphothere families.
Evolutionary dynamics and ecological niches of three longirostrine gomphotheres families
The three longirostrine gomphotheres families were widely distributed in the Early–Middle Miocene in northern China (Figure S4), ranging from ∼19 to 11.5 Ma (most of the Shanwangian and Tunggurian stages of the Chinese Land Mammal Age) (Deng et al., 2018). However, the relative abundances of these three families were very different and varied through time. We investigated fossil gomphotheres from four regions: Linxia Basin (LX), Tongxin region (TX), Junggar Basin (JG), and Tunggur region (TG) (Figures 2, S4, Data S2). We focused on different fossil assemblages in different ages. We evaluated the amounts of fossils of each elephantiform taxon from major fossil assemblages in the above regions from three museums (IVPP, HPM and AMNH, for full names, please see the Material and Methods) and calculated their proportions among all gomphotheres fossils (Figure 2). During ∼19–17 Ma, the quantities of both Choerolophodon (Choerolophodontidae) and Protanancus (Amebelodontidae) were larger than those of Gomphotherium (“Gomphotheriidae”) despite the high diversity of Gomphotherium in species level. During ∼17–15 Ma (the Mid–Miocene Climate Optimum [MMCO]), the relative fossil abundances of the three gomphothere families were similar. However, within Amebelodontidae, Platybelodon appeared and rapidly replaced the primitive Protanancus. After ∼14.5 Ma (i.e., the beginning of the Mid–Miocene Climate Transition [MMCT]) (Westerhold et al., 2020), Choerolophodon suddenly experienced regional extinction, and Gomphotherium also gradually declined. Within Amebelodontidae, Aphanobelodon, which is morphologically similar to Platybelodon, only occurred in one fossil assemblage, and the Platybelodon population greatly expanded. After ∼13 Ma, Gomphotherium were regionally extinct, and Platybelodon became the only genus among gomphotheres.
We analyzed stable isotope values of the tooth enamel for each gomphothere taxon in different fossil assemblages (Data S3). Overall, the δ13C values indicate a relatively open environment that consisted of a diverse range of habitats (including grasslands, wooded grasslands, and forests) dominated by C3 plants in northern China from ∼19 to 11.5 Ma (Zachos et al., 2001) (Figures 2, S5). In LX1 (Figure 2, Data S2), at approximately ∼18.5 Ma (see Table S1), both δ13C and δ13O values of Choerolophodon and Protanancus are similar and have a wide range of overlap, and those of Gomphotherium are in the middle of the range.
This indicates that the niches of these three groups overlapped without obvious differentiation. In TX1,2 (∼17.3 Ma), during which Platybelodon first appeared, the δ13C values of Platybelodon are lower than those of Protanancus. The δ13C values of Choerolophodon and Gomphotherium, however, are all within the δ13C range of Amebelodontidae. In JG2 (∼16 Ma), the δ13C value of Platybelodon shows a distinct positive shift, which indicates expansion into more open habitats (Figure 2B). However, Choerolophodon may have persisted in a relatively closed environment (Figure 2B). Moreover, the ecological niche of Gomphotherium appeared to be in between those of Platybelodon and Choerolophodon, and they potentially lived in the boundary area of open and closed habitats (Figure 2B). This isotopic niche pattern is also observed in JG3 (∼14.3 Ma), with the rise of Platybelodon and decline of Choerolophodon and Gomphotherium. In TX3, LX4, and TG2,3 (after ∼14 Ma), only Platybelodon were examined (as Gomphotherium specimens are too rare to sample); its ecological niche was similar to that previously occupied by Choerolophodon after the latter went extinct.
Specialized feeding behaviors of three gomphothere families
To reconstruct the feeding behaviors of the three gomphothere families, we performed Finite Element (FE) analyses on three models as representatives of each family: Choerolophodon, Gomphotherium, and Platybelodon (for model settings see Figures S6–S9, Tables S2–S4, 3D models S1–S6). We conducted two kinds of tests: the distal forces test (Code S4) and the twig-cutting test (Code S5). In the distal forces test, after full muscle forces were exerted, a 5000 N vertical force is loaded on the distal end of each mandible (Figures 3A–C, Movies S1–S3). The mandible strain energy curve (MSEC) of Platybelodon suddenly reaches a very high value, while the MSEC increases of Gomphotherium and Choerolophodon are far less than that of Platybelodon (Figure 3D). Then, keeping the magnitude of the external force unchanged, with change of the external force from vertical to horizontal direction, the MSEC of Platybelodon decreases even lower than that of Gomphotherium and was similar to that of Choerolophodon. This result indicates that the mechanical performance of the Platybelodon mandible is disadvantageous under distal vertical external forces, but greatly improved under horizontal external forces (Figure 3A–C, Movies S1–S3).
In twig-cutting tests, a cylindrical twig model of orthotropic elastoplasticity was posed in three directions to the distal end of the mandibular tusks of Platybelodon and Gomphotherium, and to the keratinous cutting plate of Choerolophodon. The sum of the equivalent plastic strain (SEPS) from total twig elements was calculated (equivalent plastic strain represents the irreversible deformation of an element, and the sum from all twig elements can reflect the cutting effects) (Figure 3E), and the cutting movies are provided (Movies S4–S19). When the twig was placed horizontally, the SEPS of the Choerolophodon model is the largest, which means that Choerolophodon has the highest twig-cutting efficiency, followed by Gomphotherium; while Platybelodon exhibits the much lower efficiency in cutting horizontal twigs than the other two models. When the twig was placed obliquely (45° orientation), the Platybelodon model still shows the smallest SEPS value, although it was nearly one order of magnitude higher than that of the horizontal twig. Finally, when the twig itself was in a vertical direction, the growth direction of the keratinous cutting plate determines that Choerolophodon cannot cut in this condition. The SEPS of the Gomphotherium model also decreases and shows lower cutting efficiency. In contrast, the SEPS of the Platybelodon model increases another order of magnitude, which is substantially larger than that of any other taxa in any cutting state. These data strongly indicate that Platybelodon mandible is specialized for cutting vertically growing plants. However, the Choerolophodon mandible is specialized for cutting horizontally or obliquely growing plants, this explains the absence of mandibular tusks, but they are likely not able to feed on vertically growing plants. The cutting effect of the Gomphotherium mandible is relatively even for all directions.
Co-evolution of narial morphology and characters of horizontal cutting among gomphothere families
The three lineages with different mandibular morphology also exhibited different stages of trunk evolution, which can be inferred from the narial region morphology (Wang and Li, 2022). Among the three groups, Gomphotherium has similar narial morphology to living elephantids (Figures 1E, G, S3D, E), with comparable size and morphology of the nasal bone process, insertion for mesethmoid cartilage, and perinasal fossae. Alternatively, the narial region of Choerolophodon shows a relatively primitive evolutionary stage (Wang and Li, 2022) (Figure 1H, S3A–C). It has a very wide and large nasal bone process for an elephantiform, a fairly wide and long groove for mesethmoid cartilage insertion, and a pair of somewhat incipient (or even absent) perinasal fossae. Amebelodontids usually possess common narial morphology similar to that of Gomphotherium (Wang and Li, 2022). However, Platybelodon has noteworthy differences in narial morphology (Boas and Paulli, 1908) (Figure 1F, S3G–L). Its nasal aperture is greatly enlarged, which results in a very broad area for attaching maxillo-labialis (Eales, 1926; Wang et al., 2017), the “core” muscle of the proboscis. The nasal bone process is very short and stout, with a slight dorsal bulge. The slit for mesethmoid cartilage insertion is very tiny. Beside the well-developed perinasal fossa, a vast inclined region is positioned rostral to the perinasal fossa, this is hereafter referred to as the prenasal slope (Boas and Paulli, 1908), and it potentially provides additional attachment for the nasialis (Wang et al., 2017).
The narial morphology of Platybelodon (and the closely related genus Aphanobelodon (Tassy, 1994) (Figure S3F), is unique in elephantiforms, and shows even more derived characters than living elephants. On the phylogenetic tree (Figure 4A), the Platybelodon lineage (the dark purple star in Figure 4) displays the most derived narial morphological combination (PC1 scores from characters in relation to the narial region), while Choerolophodontidae (the dark purple circle in Figure 4) show the least specialized narial morphology, which is close to the stem taxon Phiomia. Interestedly, the evolutionary level of the character-combine in relation to horizontal cutting (in terms of PC2) is highly correlated with that of narial region (Figure 4B). In the Platybelodon clade, this character-combine also shows high evolutionary level; while in the Choerolophodontidae clade, the evolutionary level of horizontal cutting is rather low, comparable with that of narial region. The result strongly suggests a highly co-evolution between narial region and horizontal cutting behavior in the trilophodont longirostrine gomphotheres.
In several fossil and living terrestrial mammalian groups, including living Tapirus (Moyano and Giannini, 2017) and extinct Proboscidipparion (Ma et al., 2023), Astrapotherium (Alejandro et al., 2019), and Macrauchenia (Blanco et al., 2021), elongated noses have evolved as food procuring organs. However, none of these groups possess a long and dexterous trunk like elephants, as they have not lost their mandibular incisors. Only living elephants are capable of solely accomplishing food procurement with their trunks. Coincidentally, longirostrine proboscideans, characterized by their extremely elongated mandibular symphysis and tusks, are the only group where these highly developed organs may have co-evolve (Tassy, 1996). It is evident that the great elongation of the nose is well-matched with the extreme elongation of the mandibular symphysis. However, different lineages exhibit different evolutionary strategies strongly influenced by their ecological adaptations.
Our focus is on the ancestral elephantids, longirostrine gomphotheres, which include three families (Amebelodontidae, Choerolophodontidae, and “Gomphotheriidae”). These groups flourished during the MMCO, a period of global warmth from 17 to 15 Ma (Tassy, 1983; Sanders et al., 2010; Lambert and Shoshani, 1998). With the climatic shift during the MMCT (after ∼14.5 Ma), Choerolophodon sharply declined and experienced regional extinction in northern China. Gomphotherium also declined but persisted until around 13 Ma, while Platybelodon became predominant until the end of the Middle Miocene (Figure 2, S5). Worldwide, Choerolophodon and Gomphotherium continued to flourish during the MMCT and survived to the early Tortonian in other regions (Göhlich, 1999; Wang et al., 2013; An, et al., 2001). However, Platybelodon was mostly restricted to Central Asia, especially northern China, with only a few records found in other regions (Wu et al., 2022; Guo et al., 2002) (Figure S4). These uneven distributions were likely influenced by climatic and ecological factors, such as the relatively drier climate and more open ecosystems in northern China, and Central Asia, strongly affected by the elevation of the Tibetan Plateau (Tang and Ding, 2013; Miao et al., 2012). Extensive studied, using evidence from various disciplines including geochemical and geomagnetic proxies, and fossil records, has been conducted to explore this issue (Jiang et al., 2008; Wang et al., 2022). As discussed, the three groups can be distinguished by different mandibular symphysis and tusk morphologies (Gheerbrant and Tassy, 2009), which indicate different feeding behaviors. Their distinct narial regions further reflect different evolutionary stages of their trunks (Wang and Li, 2022). Choerolophodon possesses a highly specialized mandibular symphysis for horizontally cutting growing plants and was confined to relatively close habitats. The low evolutionary level of the narial region suggests a relatively primitive or clumsy trunk in Choerolophodon. The feeding strategy of Gomphotherium was unspecialized and flexible, relying on the coordination of the enamel band of an upper tusk and the corresponding lower tusk. Previous research also suggested that Gomphotherium steinheimense, the sister taxon of elephantoids (Figures 1A, S1) from the Halamagai Fauna, fed on grasses (Wu et al., 2018). While FE analysis could not provide a clear suggestion for the trunk function of Gomphotherium, the narial evolutionary stage, which is close to living elephants, suggests a relatively flexible trunk in Gomphotherium, as it is phylogenetically closer to living elephants than the other two groups.
Enamel isotope results also support the idea that Platybelodon expanded its living habitats to more open environments, such as grasslands, more than any other gomphotheres, likely due to its distinct feeding strategy. Our FE analyses strongly indicate that the Platybelodon mandible is specifically suited for cutting vertically growing plants. In open environments, there are vertically growing plants, such as soft-stemmed herbs. Platybelodon did not survive to the Late Miocene in northern China, possibly due to a mass extinction caused by the global climatic shift known as the Tortonian Thermal Maximum. However, from another perspective, if the Platybelodon trunk functioned similarly to that of living elephants, such as pulling out herbs from the earth (McKay, 1973), the greatly enlarged mandibular symphysis and tusks would have become redundant.
The reduction of the mandibular symphysis and tusks in the Late Miocene occurred in every lineage of elephantiform, coinciding with the large-scale expansion of C4 grasses in the middle and low latitudes (Göhlich, 1999). This may reflect the functional evolution of trunk grasping and manipulation in all elephantiforms lineages (Jiangzuo et al., 2023). From around 8 Ma to 5 Ma in Africa, some derived members of “Gomphotheriidae” (e.g., Anancus) and stem taxa of Elephantidae (e.g., Stegotetrabelodon, Primelephas) showed a strong inclination towards grazing on C4-grasses, even though their cheek tooth morphology remained relatively primitive (Jiangzuo et al., 2023). Thus, open environments might be a key factor in both trunk development and the evolution of modern elephants.
As we have discussed, mandibular elongation was a prerequisite for the extremely long trunk of proboscideans, and open-land grazing further promoted the evolution of trunks with complex manipulative functions. This may explain why tapirs never developed a trunk as dexterous as that of elephants, as tapirs never shifted their adaptation zones to open lands. Living in dense forests, foods were easily accessible and procured through the mouth. Furthermore, the living mammals of the Late Cenozoic, in various open areas, have undergone specific organ evolution, such as the elongated necks of giraffes (Wang et al., 2022), the extravagate saber-tooth evolution in carnivores (Janis, 1982), and the development of various strange cranial appendages in different ruminants (Purkart et al., 2022). It is possible that the highly evolved trunks of elephants evolved somewhat accidentally, under the pressure of ecological changes from closed to open environments.
In this study, we have examined the functional and eco-morphology, as well as the feeding behaviors, of longirostrine gomphotheres. Our findings demonstrate that multiple eco-adaptations have contributed to the diverse mandibular morphology observed in proboscideans, while open-land grazing has driven the development of trunk coiling and grasping functions and ultimately led to the loss of the long mandible. Specifically, the longirostrine elephantiform, Platybelodon, represents the first known proboscidean to have evolved both grazing behavior and trunk coiling and grasping functions. We have arrived at this conclusion through three lines of evidence, including the palaeoecological reconstruction based on tooth enamel stable isotope data, the reconstruction of feeding behaviors through finite element analyses, and the examination of mandibular and narial region morphology correlated with characteristics associated with horizontal cutting behavior. The coiling and grasping ability of the trunk in Platybelodon evolved within the ecological context of Central Asia, which experienced regional drying and the expansion of open ecosystems following the MMCT (Jiang et al., 2008; Wang et al., 2022; Wu et al., 2018). As a result, Platybelodon outcompeted other longirostrine gomphotheres and flourished in the open environment of northern China until the end of the MMCT. This scenario sheds light on how proboscideans overcame an evolutionary bottleneck. Initially, the elongation of mandibular symphysis and tusks served as the primary feeding organs, with the trunk being used as an auxiliary tool, However, through some necessary modifications such as tactile specialization and water intake adaptations (Wang and Deng, 2005), the feeding function gradually shifted entirely to the trunk, which offered advantages in terms of flexibility and lighter weight of the feeding organs. Consequently, elephantiforms rapidly reduced the length of their mandibular symphysis and tusks (Gheerbrant and Tassy, 2009). Similar stories of open-land adaptation and the acquisition of iconic characteristics have been observed in various mega-mammalian lineages, highlighting the crucial role of open-land adaptation for successful survival in modern ecosystems (Janis, 1982; Purkart et al., 2022; Wang and Deng, 2005).
All examined crania and mandibles of the three longirostrine gomphothere families from the Early and Middle Miocene of northern China were housed in three museums: the Institute of Vertebrate Paleontology and Paleoanthropology (IVPP), American Museum of Natural History (AMNH), and Hezheng Paleozoological Museum (HPM, HMV is the specimen prefix). For FE analyses, surface 3D digital models of representative specimens were generated using a handheld Artec Spider 3-dimensional scanner, and the 3D models were reconstructed using Artec Studio 14 Professional. The geometrical models were further loaded into Abaqus CAE (V 6.14) for FE analyses. Phylogenetic reconstructions of longirostrine elephantiforms were performed using Bayesian total-evidence dating and maximum parsimonious analyses, which were done by MrBayes 3.2.7 and TNT 1.1, respectively. Principal component analyses were conducted using Past 4.04. Enamel isotopes analyses were conducted at the National High Magnetic Field Laboratory, USA. For detailed information, please see Supplementary Information.
We appreciate Jie Ye, Xiaoxiao Zhang for his great guidance and suggestions on taxonomy and stratigraphy; Chi Zhang for assistance in phylogeny; Sijian Xu for assistance in fossil preparation; Jin Meng of AMNH, Rong Yang and Shanqin Chen of HPM for access to specimens in these two museums for comparison and statistics; Xiaocong Guo and Yu Wang for reconstruction. We thank Mallory Eckstut, PhD, from Liwen Bianji (Edanz) (www.liwenbianji.cn) for editing the English text of a draft of this manuscript. This work was supported by the Second Tibetan Plateau Scientific Expedition (2019QZKK0705), National Natural Science Foundation of China (52178141, 41625005), National Science Foundation Cooperative Agreement No. DMR-1157490, the State of Florida.
Supplementary Information is available for this paper, including material and methods, additional results and discussion, as well as the source data (Figures S1–S9, Tables S1–3, Appendix S1, Data S1–S4, Movies S1–S19, Codes S1–S5, and 3D Models S1–S6).
Reprints and permission information is available at www.nature.com/reprints
Data availability statement
All the raw data for statistics and analyses are available in the Supplementary Information.
Code availability statement
All the codes for analyses are available in the Supplementary Information.
- 1.Astrapotherium from the Middle Miocene Collón Cura Formation and the Decline of Astrapotheres in Southern South AmericaAmeghiniana 56:290–306
- 2.Evolution of Asian monsoons and phased up lift of the Himalaya–Tibetan plateau since Late Miocene timesNature 411:62–66
- 3.A Descriptive Catalogue of the Tertiary Vertebrata of the Fayûm, Egypt:1–324
- 4.Macrauchenia patachonica Owen, 1838: Limb bones morphology, locomotory biomechanics, and paleobiological inferencesGeobios 68:61–70
- 5.The Elephant’s Head: Studies in the Comparative Anatomy of the Organs of the Head of the Indian Elephant and Other Mammals:1–80
- 6.The rise and fall of proboscidean ecological diversityNat Ecol. Evol 5:1266–1272
- 7.Anatomie Comparée: Recueil des Planches de Myologie:1–654
- 8.Neogene integrative stratigraphy and timescale of ChinaSci. China Earth Sci 62:310–323
- 9.The anatomy of the head of a foetal African elephantElephas africanus (Loxodonta africana). Earth Environ. Sci. Trans. R. Soc. Edinb 54:491–551
- 10.L’origine et l’évolution des éléphantsComptes Rendus Palevol 8:281–294
- 11.Order ProboscideaThe Miocene land mammals of Europe :157–168
- 12.Onset of Asian desertification by 22 Myr ago inferred from loess deposits in ChinaNature 416:159–163
- 13.Evolution of horns in ungulates: Ecology and PaleoecologyBiol. Rev 57:261–318
- 14.Cooling-driven climate change at 12–11 Ma: multiproxy records from a long fluviolacustrine sequence at Guyuan, Ningxia, ChinaPalaeogeogr. Palaeoclimatol. Palaeoecol 265:148–158
- 15.Origin of adaptations to open environments and social behaviour in sabretoothed cats from the northeastern border of the Tibetan PlateauProc. Royal Soc. B 290
- 16.The feeding habits of the shovel-tusked gomphotheres: evidence from tusk wear patternsPaleobiology 18
- 17.ProboscideaEvolution of Tertiary Mammals of North America. Volume 1: Terrestrial Carnivores, Ungulates, and Ungulatelike Mammals :606–621
- 18.The role of behaviour in adaptive morphological evolution of African proboscideansNature 500:331–334
- 19.Dietary niche reconstruction of Pliocene and Pleistocene Equidae from the Linxia Basin of northwestern China based on stable isotope analysisPalaeogeogr. Palaeoclimatol. Palaeoecol 111416
- 20.Behavior and ecology of the Asiatic elephant in southeastern CeylonSmithson. Contrib. Zool 125:1–113
- 21.What controlled Mid–Late Miocene long-term aridification in Central Asia? — global cooling or Tibetan Plateau uplift: a reviewEarth Sci Rev 112:155–172
- 22.Comparative cranial ontogeny of Tapirus (Mammalia: Perissodactyla: Tapiridae)J. Anat 231:665–682
- 23.Proboscidea: a Monograph of the Discovery, Evolution, Migration and Extinction of the Mastodonts and Elephants of the World:1–802
- 24.Trigeminal ganglion and sensory nerves suggest tactile specialization of elephantsCurr. Biol 32:904–910
- 25.ProboscideaCenozoic Mammals of Africa :161–251
- 26.Skin wrinkles and folds enable asymmetric stretch in the elephant trunkProc. Natl Acad. Sci. USA 119
- 27.Para- or monophyly of the gomphotheres and their position within ProboscideaThe Proboscidea: Evolution and Palaeoecology of Elephants and Their Relatives :149–177
- 28.Understanding proboscidean evolution: a formidable taskTrend Ecol. Evol 13:480–487
- 29.Anatomy and physiologyElephants: Majestic Creatures of the Wild :66–81
- 30.Advances in proboscidean taxonomy & classification, anatomy & physiology, and ecology & behaviorQuat. Int 126:5–20
- 31.A palynological insight into the Miocene aridification in the Eurasian interiorPalaeoworld 22:77–85
- 32.Les Elephantoidea Miocènes du Plateau du Potwar, Groups de Siwalik, Pakistan. IIe partie: choerolophodontes et gomphothèresAnn. Paleontol 69:235–297
- 33.Who is who among the ProboscideaThe proboscidea: Evolution and Palaeoecology of Elephants and Their Relatives :39–48
- 34.Gaps, parsimony, and early Miocene elephantoids (Mammalia), with a re-evaluation of Gomphotherium annectensZool. J. Linn.
- 35.Carbon isotope ratio of Cenozoic CO2: a comparative evaluation of geochemical proxiesPaleoceanography 25:1–11
- 36.A 25 m.y. isotopic record of paleodiet and environmental change from fossil mammals and paleosols from the NE margin of the Tibetan PlateauEarth Planet. Sci. Lett 236:322–338
- 37.Gomphotheriid mammal Platybelodon from the Middle Miocene of Linxia Basin, Gansu, ChinaActa Palaeontol. Pol 58:221–240
- 38.Morphological and ecological diversity of Amebelodontidae (Proboscidea, Mammalia) revealed by a Miocene fossil accumulation of an upper-tuskless proboscideanJ. Syst. Palaeontol 15:601–615
- 39.Attributing “Gomphotherium shensiense” to Platybelodon tongxinensis, and a new species of Platybelodon from the latest Middle MioceneVert. PalAsiat 60:117–133
- 40.Sexual selection promotes giraffoid head-neck evolution and ecological adaptationScience 376
- 41.An astronomically dated record of Earth’s climate and its predictability over the last 66 million yearsScience 369:1383–1387
- 42.Reorganization of Asian climate in relation to Tibetan Plateau upliftNat. Rev. Earth Environ 3:684–700
- 43.A grazing Gomphotherium in Middle Miocene Central Asia, 10 million years prior to the origin of the ElephantidaeSci. Rep. 8
- 44.Trends, rhythms, and aberrations in global climate 65 Ma to presentScience 292:686–693
- 45.C4 expansion in the central Inner Mongolia during the latest Miocene and early PlioceneEarth Planet. Sci. Lett 287:311–319