Evidence for the latest fossil Pongo in southern China

Journal of Human Evolution 170 (2022) 103233 Contents lists available at ScienceDirect Journal of Human Evolution journal homepage: www.elsevier.com/locate/jhevol Evidence for the latest fossil Pongo in southern China Wei Liao a, Terry Harrison b, Yanyan Yao a, c, Hua Liang a, Chun Tian a, Yuexing Feng d, Sheng Li e, Christopher J. Bae f, *, Wei Wang a, * a Institute of Cultural Heritage, Shandong University, 72 Jimo-Binhai Road, Qingdao 266237, China Center for the Study of Human Origins, Department of Anthropology, New York University, New York, NY, 10003, USA c Anthropology Museum of Guangxi, Nanning, 530012, China d Radiogenic Isotope Facility, School of Earth and Environmental Sciences, University of Queensland, Brisbane, QLD 4072, Australia e No.3 Institute of Geological & Mineral Resources Survey of Henan Geological Bureau, Zhengzhou 450000, China f Department of Anthropology, University of Hawaii at Manoa, 2424 Maile Way, 346 Saunders Hall, Honolulu, HI, 96822, USA b a r t i c l e i n f o a b s t r a c t Article history: Received 30 June 2021 Accepted 1 July 2022 Available online 26 August 2022 Pongo fossils with precise absolute age brackets are rare, limiting our understanding of their taxonomy and spatiotemporal distribution in southern China during the Late Pleistocene. Twenty-four isolated teeth of fossil orangutans were recently discovered during excavations at Yicun Cave in Guangxi Zhuang Autonomous Region, southern China. Here, we dated the fossil-bearing layer using Uranium-series dating of the associated flowstone and soda straw stalactites. Our results date the Yicun orangutan fossils to between 66 ± 0.32 ka and 57 ± 0.26 ka; thus, these fossils currently represent the last appearance datum of Pongo in southern China. We further conducted a detailed morphological comparison of the Yicun fossil teeth with large samples of fossil (n ¼ 2454) and extant (n ¼ 441) orangutans from mainland and island Southeast Asia to determine their taxonomic position. Compared to other fossil and extant orangutan samples, the Yicun Pongo assemblage has larger teeth and displays greater variation in occlusal structure. Based on the high frequency of cingular remnants and light to moderate enamel wrinkling of the molars, we assigned the Yicun fossils to Pongo weidenreichi, a species that was widespread in southern China throughout the Pleistocene. Lastly, we used published stable carbon isotope data of Early to Late Pleistocene mammalian fossil teeth from mainland Southeast Asia to reconstruct changes in the paleoenvironment and to interpret dental size variation of Pongo assemblages in a broader temporal and environmental context. The carbon isotope data show that dental size reduction in Pongo is associated with environmental changes. These morphological changes in Pongo appear to coincide with the expansion of savannah biomes and the contraction of forest habitats from the Middle Pleistocene onward. The variation in dental size of forest-dwelling Pongo in mainland Southeast Asia may have resulted from habitat differentiation during the Pleistocene. © 2022 Elsevier Ltd. All rights reserved. Keywords: Orangutans Late Pleistocene Yicun Cave Guangxi Stable carbon isotope Habitat 1. Introduction Orangutans are the only extant great apes in Asia, where today they are restricted to Borneo (Pongo pygmaeus) and northern Sumatra (Pongo abelii and Pongo tapanuliensis; Groves, 2001; Nater et al., 2017). During the Pleistocene, however, they were widely distributed across mainland Southeast Asia, including southern China, and Southeast Asia during the Pleistocene (Fig. 1; Hooijer, 1948; Kahlke, 1972; Bacon and Long, 2001, 2002; * Corresponding authors. E-mail addresses: [email protected] (C.J. Bae), [email protected] (W. Wang). https://doi.org/10.1016/j.jhevol.2022.103233 0047-2484/© 2022 Elsevier Ltd. All rights reserved. Zhao et al., 2009a; Ibrahim et al., 2013). Genomic data for extant Pongo reveal a complex evolutionary history (Locke et al., 2011; Nater et al., 2015). Based on current genomic evidence, the ancestor of extant orangutans colonized Sundaland from the Southeast Asian mainland during the Pliocene (MattleGreminger et al., 2018). The subsequent divergence between P. abelii and P. tapanuliensis occurred in Sumatra at ~3.4 Ma, representing the oldest split on Sundaland (Nater et al., 2017). Pongo pygmaeus separated from P. abelii much later, at about 674 ka (Nater et al., 2017). Although orangutans are not present on mainland Asia today, the region is important for documenting the evolutionary history and distribution of Pongo during the Pleistocene, particularly in W. Liao, T. Harrison, Y. Yao et al. Journal of Human Evolution 170 (2022) 103233 Figure 1. Location of Yicun Cave and the geographical distribution of the Pleistocene (Zhao et al., 2009a; Ibrahim et al., 2013) and extant Pongo sites (Nater et al., 2017). 1, Shiziyan, China; 2, Luokeng, China; 3, Shuangyan Cave, China; 4, Qixingyan, China; 5, Xiashan Cave, China; 6, Shanbeiyan, China; 7, Dushizai, China; 8, Guantang, China; 9, Feishu Cave, China; 10, Huangyan Cave, China; 11, Yanhui Cave, China; 12, Mawokou Cave, China; 13, Heshang Cave, China; 14, Xianren Cave, China; 15, Xianren Cave, China; 16, Jiulong Cave, China; 17, Mopanshan, China; 18, Gigantopithecus Cave (Lengzhaishan Cave), China; 19, Tongtianyan, China; 20, Ganqianyan, China; 21, Mujishan, China; 22, Gigantopithecus Cave (Nongmoshan), China; 23, Ganxian Cave, China; 24, Wuyun Cave, China; 25, Dingmo Cave, China; 26, Daxinhei Cave, China; 27, Baxian Cave, China; 28, Baikong Cave, China; 29, Juyuan Cave, China; 30, Sanhe Cave, China; 31, Queque Cave, China; 32, Yixiantian Cave, China; 33, Zhiren Cave, China; 34, Shuangtan Cave, China; 35, Quzai Cave, China; 36, Xinchong Cave, China; 37, Coc Muoi, Vietnam; 38, Tham Khuyen, Vietnam; 39, Keo Leng, Vietnam; 40, Hang Hum, Vietnam; 41, Nguom, Vietnam; 42, Duoi U’Oi, Vietnam; 43, Lang Trang, Vietnam; 44, Tham Om, Vietnam; 45, Nam Lot, Laos; 46, Ban Fa Suai I, Thailand; 47, Tham Prakai Phet, Thailand; 48, Thum Wiman Nakin, Thailand; 49, Boh Dambang, Thailand; 50, Badak Cave C, Malaysia; 51, Batu Caves, Malaysia; 52, Padang Highland Caves, Sumatra; 53, Niah Cave, Borneo; 54, Madai Caves, Borneo; 55, Sangiran, Java; 56, Trinil, Java; 57, Punung, Java. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) 2 W. Liao, T. Harrison, Y. Yao et al. Journal of Human Evolution 170 (2022) 103233 strategies, and life history (Wich et al., 2004; van Noordwijk and van Schaik, 2005; Wich et al., 2009). However, no study has examined the relationship between morphological variation in fossil orangutan assemblages and environmental change during the Pleistocene. A welldocumented feature of Pongo evolution during the Pleistocene is a general reduction in overall crown size (Hooijer, 1948; Gu et al., 1987; Ho et al., 1995; Schwartz et al., 1995; Harrison et al., 2021). Against this backdrop of a general reduction in tooth size, Pongo assemblages from mainland Southeast Asia exhibit significant regional variation in dental size during the Middle and Late Pleistocene. At the same time, stable carbon isotope data derived from fossil mammals from mainland Southeast Asia reveal substantial environmental changes through time as represented by variation in the forest and woodland-grassland distribution during the Pleistocene (Louys and Roberts, 2020). Thus, correlations between orangutan dental size and stable carbon isotope composition of associated fossil mammals may provide evidence to assess the impact of habitat change on Pongo during the Pleistocene. The present study has three aims. First, we use U-series dating to constrain the age of a Pongo assemblage from Yincun Cave in Guangxi Zhuang Autonomous Region, southern China. Second, we present a detailed description of the Yincun Pongo teeth and conduct metric and morphological comparisons with extant and fossil Pongo assemblages from other sites in mainland and island Southeast Asia. Lastly, we attempt to correlate the dental size of fossil Pongo assemblages to a large-scale data set of stable carbon isotope data of fossil mammals from mainland Southeast Asia to explore Pongo dental size variation and the potential impact of environmental change on Pongo assemblages through time (i.e., from Early to Late Pleistocene). The results of our study contribute to a better understanding of Late Pleistocene Pongo assemblages in southern China and the relationship between environmental change and dental size variation of Pongo from mainland Southeast Asian sites during the Quaternary. terms of first and last appearance datums. Although the Nanling Mountains in northern Guangxi appear to have served as a geographical barrier to the northern extent of Pongo weidenreichi (Liang et al., 2020; also known as the ‘Liang Hua Line’, Demeter and Bae, 2020), Pongo fossils are known from more than 30 cave sites throughout southern China, including those in Guangxi, Guizhou, Yunnan, Guangdong, and Hainan (Zhao et al., 2009a; Ibrahim et al., 2013; Harrison et al., 2014). As such, these new fossil discoveries are providing important evidence to document the early evolution of Pongo during the Pleistocene (Zhao et al., 2009a; Harrison et al., 2014; Wang et al., 2014). The Pongo fossils from Baikong Cave in Guangxi have been dated to ~2.0 Ma based on biostratigraphy and paleomagnetic dating (Jin et al., 2014; Sun et al., 2014), and thus currently represent the earliest appearance of Pongo. The Pongo fossils from Baikong Cave, Juyuan Cave (~1.8 Ma), Sanhe Cave (~1.2 Ma), and Queque Cave (~1.0 Ma) have been attributed to P. weidenreichi (Harrison et al., 2014; Wang et al., 2014). In general, P. weidenreichi has been distinguished from laterappearing Pongo taxa by its overall larger teeth, lower crowned canines, and considerably larger second and third molars relative to the first molars (Hooijer, 1948; Harrison et al., 2014). Pongo specimens from Ganxian Cave, dated to between 168.9 ± 2.4 ka and 362 ± 78 ka by Uranium-series (U-series) and electron spin resonance dating, have also been classified as P. weidenreichi (Liang et al., submitted). Pongo weidenreichi from the Middle Pleistocene of Yixiantian Cave and Daxin Hei Cave, retain the same morphological features of P. weidenreichi from the Early Pleistocene in southern China, but have undergone a major reduction in dental size (Harrison et al., 2021). A recent study has shown that P. weidenreichi may have survived to the early Late Pleistocene (Harrison et al., 2021), as suggested by evidence from Zhiren Cave that dates to ~116e110 ka (Jin et al., 2009; Liu et al., 2010; Cai et al., 2017). Based on mammalian faunal comparisons, the ages of Pongo fossils from Shuangtan Cave, Baxian Cave, and Quzai Cave in southern China are considered comparable to those of Zhiren Cave. However, more recent geochronological research at Zhiren Cave indicates that the deposits may date to MIS 6, around 190e130 ka (Ge et al., 2020). In sum, Pongo fossils with definitive dates from the Late Pleistocene of southern China are relatively scarce. Additional Pongo fossils with more precise absolute age determinations are needed to document the taxonomic and phylogenetic status of Pongo assemblages in southern China during the Late Pleistocene, and in particular, the later period of the Late Pleistocene. Long-term field studies reveal that extant orangutan populations on Sumatra and Borneo are heavily influenced by habitat and ecological conditions (Wich et al., 2009, 2016; Manduell et al., 2012; Vogel et al., 2015; Pandong et al., 2019; Seaman et al., 2019; Meijaard et al., 2021). Sumatran forests are of higher productivity than Bornean forests, as represented by a higher proportion of fruit-bearing trees and shorter periods of low-fruit availability (Wich et al., 2009, 2011; Pandong et al., 2019; Seaman et al., 2019). Because orangutan diets consist primarily of fruits supplemented by leaves and young shoots, bark, flowers, and insects, habitat productivity and fruit availability directly impact orangutan ecology (Wich et al., 2009). Habitat variation is thus likely to have shaped differences between extant Sumatran and Bornean orangutans in terms of population densities (van Schaik et al., 1995; Russon et al., 2001; Morrogh-Bernard et al., 2003; Husson et al., 2009; Vogel et al., 2015; Seaman et al., 2019), morphology (Taylor, 2006, 2009; Taylor and van Schaik, 2007), reproductive 2. Materials and methods 2.1. Geological setting Yicun Cave (107160 8.4300 E, 22 27048.6400 N) is located in Yicun Village in Jiangzhou District, Chongzuo City, Guangxi, southern China (Fig. 1). The site was discovered and excavated in 2017 by a research team from the Anthropology Museum of Guangxi. Yicun is located in an area of a karst peak cluster formed by the dissolution of Middle Permian limestone beds. The cave entrance is located on the northwest side of the mountain about 4 m above present-day ground level (Fig. 2A). Upon entering the cave, it branches about 11 m from the entrance into left, central, and right chambers. Sediments are preserved in all three chambers (Fig. 2BeE). Three test pits were excavated, with test pit 1 (T1) and test pit 2 (T2) located in the left chamber and test pit 3 (T3) in the central chamber (Fig. 2F). The location of test pits was determined based on the distribution of the fossiliferous sediments. Test pits were excavated in 10 cm intervals, down to a depth of 70 cm in T1, and a depth of 50 cm in T2 and T3. During the excavation, fossil mammals and soda straw stalactites were mapped, recorded, and collected from each layer. Sediments from each layer of all three test pits were dry-sieved through 4 mm mesh screens outside the cave to recover smaller fossils. Based on the lithology, the stratigraphic profiles for the three different test pits are similar. The section is divided into three layers 3 W. Liao, T. Harrison, Y. Yao et al. Journal of Human Evolution 170 (2022) 103233 Figure 2. Yicun Cave. A) Landscape of the Yicun Cave. B) Well preserved sediment of T2 before excavation. C) T2 after excavation. D) Stratigraphic profile of the east side of T2. E) Flowstone above the sediment of T2. F) Plan of the Yicun Cave showing the location of the three test pits. G) Stratigraphic profile of T2 with chronological framework. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) 4 W. Liao, T. Harrison, Y. Yao et al. Journal of Human Evolution 170 (2022) 103233 Means and standard deviations (SD) of the occlusal areas (MD  BL) for different tooth types of extant orangutans were adopted for reference (Harrison, 2000). All of the Yicun fossil teeth were separated into different size classes based on the number of SD intervals ( 4 SD to þ6 SD) of each tooth from the means of the corresponding tooth type of extant orangutans. Based on the percentage of teeth falling into different intervals, the dental size of the Yicun Pongo assemblage was compared with other Pongo assemblages from the Early Pleistocene to the early Late Pleistocene of mainland Southeast Asia. Because tooth size is highly correlated with body weight in nonhuman primates (Gingerich et al., 1982; Conroy, 1987), the overall dental size distribution can be used to reflect the overall body weight for the fossil orangutan assemblage. Although ape body weight regression formulae derived from individual teeth are ideal (Conroy, 1987; Harrison et al., 2021), because we sometimes used published data that were aggregated, in certain cases, we had to rely on overall dental size from these samples. To reconstruct paleoenvironmental change in mainland Southeast Asia, we collected published stable carbon isotope (d13C) values of mammalian fossil teeth from Early to Late Pleistocene sites (Supplementary Online Material [SOM] Table S1). Because we are only interested in the relationship between Pongo and the paleoenvironment, we restricted the comparative data to sites where fossil orangutans have been reported. For consistency and comparability of the data, all d13C values were corrected to the d13C value of the diet of the mammal (d13Cdiet) following the method of Louys and Roberts (2020). The d13Cdiet data set provides evidence for large-scale variation in vegetation cover, from forested environments dominated by C3 plants (d13Cdiet < 23‰) to savannah environments dominated by C4 plants (d13Cdiet > 15‰) in mainland Southeast Asia. Correlations between the overall dental size variation of Pongo assemblages and the d13Cdiet data of the Early Pleistocene, Middle Pleistocene, and Late Pleistocene enabled us to qualitatively assess the overall body weight variation of Pongo assemblages in a broader temporal and environmental context. from top to bottom (Fig. 2D, G). Layer 1 is an unconsolidated greybrown sandy clay, 5e15 cm in thickness, and contains only a few fossil teeth. Layer 2 is a slightly cemented yellowish-brown sandy clay, 10e50 cm in thickness, which yielded the highest density of fossils. Layer 3, a calcified variegated sand, with a thickness of more than 10 cm, is a sterile layer with no fossils. 2.2. Uranium-series dating In the southeast part of T2, the sediment from Layer 1 was covered by a thin layer of flowstone (Fig. 2D, E). In addition, two soda straw stalactites were collected from Layer 2 at depths of 20 cm and 40 cm. The flowstone and soda straw stalactites were selected for U-series dating to constrain the age of the fossil teeth (Fig. 2G). For each sample, 1e2 g of calcite material were carefully cleaned using a diamond-tipped burr attached to a dental drill. The cleaned samples were then crushed to an approximately 1 mm size fraction using an agate mortar and pestle and then ultrasonically washed in Milli-Q water four times and dried on a hotplate at 40  C. Finally, only the cleanest calcite pieces free from any visible forms of alteration were handpicked under a binocular microscope and used for U-series dating. UraniumeThorium dating for these samples was carried out using a Nu Plasma HR multicollector inductively coupled plasma mass spectrometer (MC-ICP-MS) in the Radiogenic Isotope Facility at the School of Earth and Environmental Sciences, The University of Queensland, following chemical treatment procedures and MC-ICPMS analytical protocols described in Zhao et al. (2001, 2009b), Clark et al. (2014), and Liu et al. (2014). The 230Th/238U and 234U/238U activity ratios of the samples were calculated using the decay constants given in Cheng et al. (2000). The nonradiogenic 230Th was corrected using an assumed bulk-Earth atomic 230Th/232Th ratio of 4.4 ± 2.2  10 6. UraniumeThorium ages were calculated using Isoplot/Ex v. 3.0 (Ludwig, 2003). 2.3. Morphological descriptions and linear metric analysis of the Yicun Pongo fossils 3. Results The mammal fossils including the Pongo teeth unearthed from Yicun are housed in the Anthropology Museum of Guangxi. Twenty-four isolated teeth of fossil orangutans are described and analyzed in this study (Fig. 3; Table 1). Metric comparisons were made between the Yicun Pongo teeth and a large comparative sample of fossil and extant orangutans from mainland and island Southeast Asia (Table 2). These included fossil Pongo from the Early to Late Pleistocene (i.e., P. weidenreichi, Pongo sp., Pongo javensis, Pongo palaeosumatrensis, Pongo duboisi; n ¼ 2454) and extant orangutans from Borneo (P. pygmaeus) and Sumatra (P. abelii; n ¼ 441). The comparative dental metric data were compiled from Hooijer (1948), Drawhorn (1995), Zhao et al. (2009a), Ibrahim et al. (2013), Harrison et al. (2014, 2021), Wang et al. (2014), Tshen (2016), and Filoux and Wattanapituksakul (2019). The morphological descriptions and comparisons of the fossil teeth employ standard anatomical terms following Hooijer (1948), Schwartz et al. (1994, 1995), Drawhorn (1995), Zhao et al. (2009a), Ibrahim et al. (2013), Harrison et al. (2014), and Wang et al. (2014). Mesiodistal (MD) and buccolingual (BL) dimensions of the Yicun orangutan fossil teeth were measured with a hand-held caliper to the nearest 0.1 mm. 3.1. Uranium-series dating results Three samplesdone flowstone and two soda straw stalactites from T2dwere used in the U-series analysis to determine the age of the Yicun fossils (Fig. 2G; Table 3). The U-series age of the flowstone above Layer 1 is 57.8 ± 0.3 ka, indicating that the Yicun fauna is older than this age. The two soda straw stalactites found at depths of 20 cm and 40 cm from Layer 2 of T2 were dated to 307.1 ± 5.5 ka and 66.3 ± 0.3 ka, respectively. Because the U-series ages of the soda straw stalactites should be older than the host sediments (St Pierre et al., 2009, 2012, 2013), fossils from the depths of 20 cm and 40 cm should be younger than 307.1 ± 5.5 ka and 66.3 ± 0.3 ka, respectively. When combining the U-series ages of the flowstone and the soda straw stalactites, the Yicun fauna can be confidently constrained between 57.8 ± 0.3 ka and 66.3 ± 0.3 ka. 3.2. Morphological descriptions of the Yicun orangutan fossils Dental measurements for the 24 isolated orangutan teeth from Yicun are presented in Table 2. Most of the crowns of these teeth are well preserved (Fig. 3). 2.4. Overall dental size of the Yicun Pongo assemblage and paleoenvironmental reconstruction Left upper lateral incisor The Yicun I2 (YC-1) is generally well preserved, although the enamel surface of the lingual side and the middle parts of the labial side are lightly weathered (Fig. 3A; SOM Fig. S1). The crown is moderately worn. The labial surface of the To compare dental size variation of the Yicun orangutan fossils with extinct and extant Pongo assemblages from mainland Southeast Asia, we used the method developed by Harrison (2000). 5 W. Liao, T. Harrison, Y. Yao et al. Journal of Human Evolution 170 (2022) 103233 Figure 3. Pongo fossil teeth from Yicun Cave. A) YC-1, left I2; B) YCT2-5, left C1; C) YCT2-2, right P3; D) YCT2-3, right P3; E) YCT2-10, left P3; F) YCT3-1, left P3; G) YCT1-5, right P4; H) YCT2-12, right P4; I) YCT1-4, left P4; J) YCT2-8, left P4; K) YCT2-6, right M1; L) YCT2-7, right M1; M) YCT2-11, right M1; N) YCT1-8, left M1; O) YCT2-13, left M1; P) YC-2, right M2; Q) YCT1-7, right M2; R) YCT1-6, right M2; S) YCT1-3, right M3; T) YCT1-2, left M3; U) YCT2-9, left M3; V) YCT1-1, left P3; W) YCT2-1, right P4; X) YCT2-4, right M2. buccolingually, with shallow mesial and distal longitudinal grooves, and its apex curves mesially. crown is convex apicobasally and mesiodistally. The apex is situated nearer to the mesial side of the crown. The apical margin is curved in the mesiodistal plane. The distal crest of the apex is longer than the mesial crest and it slopes away at a steeper angle. In lingual view, the crown has an irregular pentagonal shape. The different heights of the mesial and distal crests give the crown a bilaterally asymmetrical appearance. The lateral incisal angle is closer to the gingival line than the medial-incisal angle. A welldeveloped lingual cingulum slopes toward the middle part of the lingual face to join the mesial crest; the right portion of the lingual cingulum shows evidence of rodent gnawing. The lingual surface above the cingulum is concave vertically and convex transversely. In mesial view, the crown is cone shaped. The mesial face is convex apicobasally and mesiodistally. The gingival line presents a concavity toward the root. In distal view, the crown is cone shaped, and convex apicobasally and buccolingually. The root is flattened Left upper canine The Yicun C1 crown (YCT2-5) is well preserved and moderately worn but the tip of the crown and the root are missing (Fig. 3A; SOM Fig. S1). Based on the size and morphology of the crown, the canine can be inferred to have belonged to a male individual. The crown has a large conical shape which is elongated mesiodistally. The anterior border is rounded and slightly convex vertically. Near the cervix of the crown, the anterior border passes into a triangular prominence. The posterior border forms a sharp edge and is slightly concave. The distolingual side displays an extensive wear facet that runs from the apex to the cervix. In lingual view, there is a vertical groove that closely follows the curved contour of the anterior border. This anterior groove is well developed in the middle of its course and fades away toward the apex 6 W. Liao, T. Harrison, Y. Yao et al. Journal of Human Evolution 170 (2022) 103233 the buccal side (the paracone) and a smaller blunt lingual cusp (the protocone). The paracone and protocone are separated by a deep anteroposterior valley on the occlusal surface. The buccal part of the crown is wider and less convex than the lingual side and separated from the flattened anterior and posterior surfaces by obtuse but well-defined angles, whereas on the lingual part, there are no angular prominences. A notable feature of the P3s is that the anteroexternal angle of the crown projects anteriorly. This feature easily distinguishes the P3s from the P4s. There are two marginal ridges (anterior and posterior) that originate from the apex of the paracone. The anterior ridge of the paracone is steeper than its posterior ridge and descends lower on the anterior side of the crown. The marginal ridge ascends inward from the anteroexternal angle along the anterior surface of the crown which is depressed in the middle at the cervix, and culminates in the apex of the protocone. The posterior ridge of the paracone descends backward and then inward from the postero-external angle along the posterior surface of the crown and ends at the apex of the protocone. From the apex of the paracone, an anterior transverse ridge runs inward and joins the marginal ridge in front of the protocone. The anterior transverse ridge, the anterior ridge of the paracone, and the anterior marginal ridge all border a triangular depression, the anterior fovea. From the apex of the paracone, or somewhat behind it, there is another oblique ridge that runs inward parallel to the anterior transverse ridge and joins the marginal ridge behind the protocone. There is a posterior fossa that is bordered by an oblique ridge and a posterior marginal ridge. A central fossa is present between the oblique ridge and the anterior transverse ridge. There is a great deal of variation in the transverse ridge locations and secondary cusp number on the P3s and P4s. In the P3 specimen YCT3-1, the oblique ridge is divided into three minor ridges. The P4 specimen YCT2-12 has a prominent hypocone behind the protocone and its anterior and latter transverse ridges form a ‘V’ shape that encloses the paracone. In the P4 specimen YCT1-5, the hypocone and metacone are separated from the protocone and paracone by a cleft. An anterior transverse ridge passes down from the anterior part of the paracone and runs inward and forward to join a minor ridge in front of the protocone. An almost parallel posterior ridge connects the apex of the metacone to the distal aspect of the protocone. There are two distinct accessory cusps on the posterior marginal ridge. The P3s and P4s have three roots, two buccal and one lingual. Table 1 Dimensions of orangutan teeth from the Yicun Cave. Specimen Test pit Depth (cm) Tooth category MD (mm) BL (mm) YC-1 YC-2 YCT1-1 YCT1-2 YCT1-3 YCT1-4 YCT1-5 YCT1-6 YCT1-7 YCT1-8 YCT2-1 YCT2-2 YCT2-3 YCT2-4 YCT2-5 YCT2-6 YCT2-7 YCT2-8 YCT2-9 YCT2-10 YCT2-11 YCT2-12 YCT2-13 YCT3-1 T1 T1 T1 T1 T1 T1 T1 T1 T1 T1 T2 T2 T2 T2 T2 T2 T2 T2 T2 T2 T2 T2 T2 T3 0e10 0e10 20e30 40e50 40e50 40e50 40e50 40e50 50e60 50e60 0e10 10e20 10e20 10e20 20e30 20e30 20e30 20e30 20e30 20e30 30e40 30e40 30e40 10e20 Left I2 Right M2 Left P3 Left M3 Right M3 Left P4 Right P4 Right M2 Right M2 Left M1 Right P4 Right P3 Right P3 Right M2 Left C1 Right M1 Right M1 Left P4 Left M3 Left P3 Right M1 Right P4 Left M1 Left P3 8.3 15.2 16.9 12.3 16.0 11.1 12.5 14.1 15.5 14.2 13.1 12.4 10.9 16.9 20.2 15.2 14.0 10.3 15.4 12.2 12.7 12.6 14.8 11.0 9.2 15.6 11.0 14.6 17.9 15.4 15.7 15.1 17.2 15.3 11.7 14.9 12.0 14.0 13.3 15.7 14.9 13.1 15.8 14.0 14.2 16.6 15.5 13.1 Abbreviations: ML ¼ mediolateral, BL ¼ buccolingual. and cervix. Posterior to the anterior groove, the lingual surface forms a ridge that is defined posteriorly by a groove, so that the lingual surface is trilobate. In labial view, the crown is trapezoidal in shape. The labial surface is convex mesiodistally and apicobasally. The anterior part of the labial surface possesses a number of minor grooves and ridges. Upper third and fourth premolars The crowns of the right P3s (YCT2-2 and YCT2-3; Fig. 3C, D; SOM Fig. S2), left P3s (YCT2-10 and YCT3-1; Fig. 3E, F; SOM Fig. S2), and the right P4s (YCT1-5 and YCT212; Fig. 3G, H; SOM Fig. S3) are generally well preserved. The enamel surfaces of the left P4s (YCT1-4 and YCT2-8, Fig. 3I, J; SOM Fig. S3) are heavily weathered, but the structure of their occlusal surfaces is preserved. The crowns are slightly to moderately worn. All of the P3s and P4s have two main cusps, a large pointed cusp on Table 2 Pongo specimens used in the morphological comparisons and linear metric analyses. Fossil site Species Age Region Source Baikong Cave Juyuan Cave Sanhe Cave Queque Cave Yixiantian Cave Daxinhei Cave Ganxian Cave Badak Cave C Punung I, II, III Pongo Pongo Pongo Pongo Pongo Pongo Pongo Pongo Pongo weidenreichi weidenreichi weidenreichi weidenreichi weidenreichi weidenreichi weidenreichi sp. javensis Early Pleistocene Early Pleistocene Early Pleistocene Early Pleistocene Middle Pleistocene Middle Pleistocene Middle Pleistocene Middle Pleistocene Middle to Late Pleistocene Southern China Southern China Southern China Southern China Southern China Southern China Southern China Peninsular Malaysia Java Shuangtan Cave Baxian Cave Zhiren Cave Quzai Cave Batu Caves Tham Prakai Phet Sibrambang Lida Ajer and Djamboe Niah Cave Pongo Pongo Pongo Pongo Pongo Pongo Pongo Pongo Pongo Pongo Pongo weidenreichi weidenreichi weidenreichi weidenreichi sp. sp. palaeosumatrensis duboisi pygmaeus pygmaeus abelii Late Pleistocene Late Pleistocene Late Pleistocene Late Pleistocene Late Pleistocene Late Pleistocene Late Pleistocene Late Pleistocene Late Pleistocene Extant Extant Southern China Southern China Southern China Southern China Peninsular Malaysia Thailand central Sumatra central Sumatra Borneo Borneo North Sumatra Harrison et al. (2014, 2021) Harrison et al. (2014, 2021) Wang et al. (2014); Harrison et al. (2021) Harrison et al. (2014, 2021) Harrison et al. (2021) Harrison et al. (2021) Liang et al. (submitted) Ibrahim et al. (2013) Badoux (1959); Drawhorn (1995); Harrison et al. (2014) Harrison et al. (2021) Harrison et al. (2021) Zhao et al. (2009a); Harrison et al. (2021) Harrison et al. (2021) Ibrahim et al. (2013) Filoux and Wattanapituksakul (2019) Hooijer (1948); Harrison et al. (2014) Hooijer (1948); Harrison et al. (2014) Hooijer (1960) Hooijer (1948) Hooijer (1948) 7 Upper first molars The crowns of the right (YCT2-6, YCT2-7, and YCT2-11; Fig. 3KeM; SOM Fig. S4) and left (YCT1-8 and YCT2-13; Fig. 3N, O; SOM Fig. S4) M1s are well preserved, except for YCT211, which is missing the posterointernal part of its crown. All are missing their roots. The crowns are slightly to moderately worn. The M1s have four main cusps. The paracone and metacone on the buccal side are higher and more closely approximated and extend down to the cervix of the crown more steeply than the protocone and hypocone of the lingual side. The protocone, which is the largest cusp, is positioned less anteriorly than the paracone. The crista obliqua, connecting the protocone and metacone, is interrupted by a groove in the middle of its course. An anterior transverse ridge runs inward from the anteroexternal edge in front of the paracone to the anterior marginal ridge close to the protocone. A linear anterior fovea is surrounded by an anterior transverse ridge and an anterior marginal ridge. A posterior transverse ridge passes inward from the apex of the metacone to the apex of the hypocone. Only in YCT2-6 does this posterior transverse ridge originate from the apex of the accessory cusp behind the metacone. The anterior and posterior transverse ridges define the central fossa of the crown, which is divided by the crista obliqua into a trigon fossa anteriorly and a hypocone fossa posteriorly. The trigon fossa is generally wider and deeper than the hypocone fossa. Only in YCT2-13 is the trigon fossa shallower than the hypocone fossa. The posterior fossa is bordered by a posterior transverse ridge and a posterior marginal ridge. Except for YCT2-7, an accessory cusp is separated from the metacone by a cleft. A lingual cingulum is represented by a short ledge that passes from the mesiolingual angle to the inner part of the anterior marginal ridge except in YCT1-8. There is evident variation in the degree of wrinkle development on the occlusal surface. For example, in YCT2-7, the wrinkles are coarse and patent, whereas in other M1s, the wrinkles are represented by fine lines on the slightly worn occlusal surface. There are elliptical interproximal facets on the mesial and distal faces of the crowns. The 234U/238U and 230Th/238U activity ratios of the samples were calculated using the decay constants given in Cheng et al. (2000). 230Th/238U ages were calculated using the Isoplot/Ex 3.0 Program v. 3.0 (Ludwig, 2003). The nonradiogenic 230Th was corrected using an assumed bulk-earth atomic 230Th/232Th ratio of 4.4 ± 2.2  10 6. Corrected Initial (234U/238U) was calculated iteratively using the measured (234U/238U) and the age of flowstone. All errors reported in this table are quoted as 2s. 57.8 ± 0.3 307.1 ± 5.5 66.3 ± 0.3 58.2 ± 0.2 308.2 ± 5.3 66.7 ± 0.3 1.1458 ± 0.0011 1.1229 ± 0.0015 1.9684 ± 0.0016 0.33493 ± 0.00013 0.09976 ± 0.00003 1.64805 ± 0.00173 YC-1 YC-2 YC-3 0 20 40 Flowstone Soda straw stalactite Soda straw stalactite 4.909 ± 0.006 4.782 ± 0.006 41.458 ± 0.136 98.94 ± 0.27 68.18 ± 0.22 113.29 ± 0.49 0.4779 ± 0.0012 1.0930 ± 0.0032 0.9393 ± 0.0029 Corrected230 Th Age (ka) Uncorrected230 Th Age (ka) (230Th/238U) (234U/238U) (230Th/232Th) Th (ppb) 232 U (ppm) Material Sample depth (cm) Sample name Table 3 Results of U-series dating from the Yicun Cave. 1.1724 ± 0.0013 1.2956 ± 0.0042 2.1761 ± 0.0040 Journal of Human Evolution 170 (2022) 103233 Corrected Initial (234U/238U) W. Liao, T. Harrison, Y. Yao et al. Upper second molars All three M2s are from the right side. The outermost layer of the enamel surface of YC-2 (Fig. 3P; SOM Fig. S5) is heavily weathered, although the structure of its occlusal surface is discernible. The crowns of YCT1-7 and YCT1-6 (Fig. 3Q, R; SOM Fig. S5) are well preserved with only lightly weathered enamel surfaces. The morphology of the M2s is similar to that of the M1s, except for their obvious smaller hypocone. For YC-2, a linear anterior fovea is surrounded by the anterior transverse ridge and the anterior marginal ridge (Fig. 3P). The posterior transverse ridge of YC-2 originates from the apex of the accessory cusp behind the metacone. In YCT1-6 (Fig. 3R), the posterior transverse ridge passes inward from the apex of the metacone to the apex of the hypocone. The lingual cingulum is represented by a short ledge that passes from the mesiolingual angle to the inner part of the anterior marginal ridge. There are elliptical interproximal facets on the mesial and distal sides of the crown. Upper third molars The crown of YCT2-9 (left M3; Fig. 3U; SOM Fig. S5) is well preserved. However, the enamel surfaces of YCT13 (right M3; Fig. 3S; SOM Fig. S5) and YCT1-2 (left M3; Fig. 3T; SOM Fig. S5) are heavily weathered. The morphology of the M3s is similar to M1s and M2s, except for the absence of an elliptical interproximal facet on the distal side of the crowns. A linear anterior fovea is surrounded by the anterior transverse ridge and the anterior marginal ridge of the crown. No posterior transverse ridge is present. A lingual cingulum is represented by a short ledge that passes from the mesiolingual angle to the inner part of the anterior marginal ridge. Lower third premolar The enamel surface of YCT1-1 (left P3; Fig. 3V; SOM Fig. S6) is heavily weathered, although the main features of its 8 W. Liao, T. Harrison, Y. Yao et al. Journal of Human Evolution 170 (2022) 103233 3.3. Size and morphological comparisons of the Yicun Pongo fossils occlusal surface are preserved. The crown is moderately worn. In occlusal view, YCT1-1 has a subtriangular outline, and the cervix of the crown projects mesially and buccally. The long axis runs from mesiobuccal to distolingual. The buccal side of the crown is more convex apicobasally and mesiodistally than that of the lingual side. There are three ridges (mesial, transverse, and distal) that originate from the apex of the protoconid. The mesial ridge is sharp and straight and extends basally to meet the lingual cingulum. The transverse ridge descends internally and distally from the apex of the protoconid to join into a prominent cervix which represents the metaconid. The mesial ridge, the lingual cingulum, and the transverse ridge enclose an anterior fovea or trigonid basin that is triangular in shape. The distal ridge descends distally from the apex of the protoconid to join the distal marginal ridge. The distal ridge, the distal marginal ridge, and the transverse ridge make up the defining borders of the talonid fossa, which is circular in outline. The roots of YCT1-1 are missing. Metric comparisons of the Yicun Pongo teeth are presented in SOM Table S2. In general, the Yicun fossil orangutan teeth fall within the general broad range of metric variation of fossil and extant Pongo (Fig. 4). However, the comparison reveals that the Yicun teeth display a unique combination of features, including a relatively small I2, extremely large C1 and M3, and a stronger cingulum development on the upper and lower molars (Fig. 4). The left I2 specimen from Yicun (YC-1) is much smaller than the 2 I s (879 and 2508) of P. weidenreichi from the Early Pleistocene Baikong Cave located in the same region of Guangxi. YC-1 is also smaller than the I2s of P. duboisi from the Late Pleistocene of Lida Ajer and Djamboe, and of P. palaeosumatrensis from the Late Pleistocene of Sibrambang, all located in the Padang Highlands of west central Sumatra. The dimensions of YC-1 are within the range of variation of the I2s of P. javensis from the Middle-Late Pleistocene fissure-fillings from Punung in Java, P. pygmaeus from the Late Pleistocene of Niah Cave in Borneo, and extant Pongo spp. The male C1 specimen (YCT2-5) from Yicun is exceptionally large, but is similar in size to the C1 (GCSD0410-178) of P. weidenreichi from the Early Pleistocene site of Sanhe Cave in the same Chongzuo area as Yicun. YCT2-5 is larger than the C1s from Middle Pleistocene Ganxian Cave of Guangxi, southern China, the Late Pleistocene site of Tham Prakai Phet in northeast Thailand, P. palaeosumatrensis from the Late Pleistocene of Sibrambang, and P. pygmaeus from the Late Pleistocene of Niah Cave. The dimensions of YCT2-5 fall within the upper end of the range of variation of the C1s of P. duboisi from the Late Pleistocene of Lida Ajer and Djamboe, and of extant Pongo spp. The four P3s from Yicun can be divided into two size groups: large (YCT2-2 and YCT2-10) and small (YCT2-3 and YCT3-1). A similar distribution occurs for the Yicun P4s, M1s, M2s, and M3s (see below). Except for the extremely large P3 (1089) from the Early Pleistocene site of Baikong, the dimensions of the Yicun P3s fall within the size range of all fossil and extant orangutan species included in our study. The dimensions of two P4s (YCT1-5 and YCT2-12) are particularly large. Except for two specimens of P. palaeosumatrensis from the Late Pleistocene of Sibrambang, their sizes, at least in their mesiodistal lengths, are larger than the rest of the P4s of fossil and extant orangutans included in this study. The two smaller P4s (YCT1-4 and YCT2-8) fall within the size distribution of the orangutan assemblages from the Early Pleistocene of Sanhe and Queque in Guangxi, the Middle Pleistocene of Tham Khuyen in Vietnam, Badak in Peninsular Malaysia, and Ganxian in Guangxi, and the Late Pleistocene of Tham Prakai Phet in northeast Thailand, Batu in Peninsular Malaysia, and Sibrambang, Lida Ajer, and Djamboe in Sumatra. They also fall within the upper range of size variation of P. javensis from the Middle-Late Pleistocene Punung in Java and of the extant orangutan samples. All of the Yicun P4s are larger than the P. pygmaeus P4s from the Late Pleistocene of Niah Cave in Borneo. Similar to the P4s, the dimensions of the Yicun M1s are also relatively large. Except for three specimens of P. weidenreichi from Middle Pleistocene of Ganxian, two specimens of P. palaeosumatrensis from the Late Pleistocene of Sibrambang, and one specimen of P. duboisi from the Late Pleistocene of Lida Ajer and Djamboe, the M1 (YCT2-6) of Yicun is larger than all other M1s included in this study. The remaining four M1s (YCT1-8, YCT2-7, YCT2-11, and YCT2-13) of Yicun are within the size range of fossil and extant orangutan assemblages. The Yicun M2s are smaller than the M2 (1091) from the Early Pleistocene of Baikong, but larger than those of P. pygmaeus from the Late Pleistocene of Niah in Borneo, and within the size range of other fossil and extant orangutan samples. Except for one specimen (1118) of P. weidenreichi from the Early Pleistocene of Baikong, one specimen of P. palaeosumatrensis from the Late Pleistocene of Sibrambang, and two specimens of P. duboisi Lower fourth premolar The enamel surface of YCT2-1 (right P4; Fig. 3W; SOM Fig. S6) is lightly weathered, but the structure of its occlusal surface is well-preserved. YCT2-1 is moderately worn. In occlusal view, the crown is subrectangular with rounded angles, being broadest anteriorly and narrowing posteriorly. The buccal side of the crown is sloping and long while the lingual side is straight and short. On the buccal surface, the enamel border descends anteriorly toward the roots. There are two cusps, buccal (protoconid) and lingual (metaconid), arranged in a transverse line and positioned much nearer to the anterior than to the posterior border of the crown. The protoconid and metaconid are connected by a transverse ridge which is bisected by an anteroposterior sulcus in the midline of the crown. A small and pit-like trigonid fossa is enclosed by the transverse ridge and anterior marginal ridge. The posterior ridges of the protoconid and metaconid descend backward from the cusp apices and are continuous with the posterior marginal ridge. They enclose a spacious and relatively deep talonid fossa. Most of the enamel wrinkles have been obliterated by wear. There is a carious cavity on the posterior marginal ridge. Only the upper parts of the roots are preserved, and these display evidence of rodent gnawing. Lower second molar The enamel surface of YCT2-4 (right M2; Fig. 3X; SOM Fig. S6) is lightly weathered; but otherwise, the occlusal surface is well preserved and only moderately worn. The crown is trapezoidal in outline with rounded corners. The anterior surface is perpendicular to the long axis of the crown. In occlusal view, the five cusps are arranged around the periphery of the crown. The tooth has little occlusal relief with low cusps and shallow basins. The protoconid and metaconid, which are similar in size, are the largest cusps. The latter is more elevated than the former. The entoconid is lower than the metaconid. The hypoconid and the hypoconulid are the lowest cusps, and both are worn with small areas of dentine exposed. The hypoconulid is positioned toward the buccal side of the central axis. The entoconid, hypoconid, and hypoconulid are all similar in size. A low and indistinct anterior transverse ridge links the protoconid and metaconid to form the distal border of the trigonid fossa. Posterior to the trigonid fossa is the spacious and relatively deep talonid fossa, which is bordered by the anterior and posterior transverse ridges. The linear, shallow posterior fossa is bordered by the posterior transverse and posterior marginal ridges. The main cusps and grooves are arranged in a typical Y-5 pattern: the metaconid comes into contact with the hypoconid on the floor of the talonid basin. Most of the enamel wrinkles on the crown are obscured by wear. Interproximal facets are present on the mesial and distal faces. A cingular remnant is represented by a ledge on the buccal side between the protoconid and the hypoconid. The roots are mostly missing. 9 W. Liao, T. Harrison, Y. Yao et al. Journal of Human Evolution 170 (2022) 103233 Figure 4. Metric comparisons of Pongo teeth from the Yicun Cave. Comparative data are from Hooijer (1948), Drawhorn (1995), Zhao et al. (2009a), Ibrahim et al. (2013), Harrison et al. (2014, 2021), Wang et al. (2014), Filoux and Wattanapituksakul (2019), and A.-M. Bacon (unpub. data). Tooth sizes of the Yicun Pongo assemblage are mostly within the size range of P. weidenreichi. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Peninsular Malaysia, the Middle-Late Pleistocene of Punung in Java, and the Late Pleistocene of Niah in Borneo. At the same time, its dimensions are within the size range of several other fossil orangutan assemblages, including the Early Pleistocene of Sanhe in southern China, the Middle Pleistocene of Tham Khuyen in Vietnam, the Late Pleistocene of Sibrambang, Lida Ajer and Djamboe in Sumatra, and extant orangutan assemblages. Although the Yicun right P4 (YCT2-1) is smaller than the P4s from the Early Pleistocene of Baikong and Sanhe in southern China, from the Late Pleistocene of Lida Ajer and Djamboe, two of the Yicun M3s (YCT1-3 and YCT2-9) are larger than all other M3s included in this study. The remaining Yicun M3 (YCT1-2) is within the size range of fossil and extant orangutan assemblages. Although the Yicun left P3 (YCT1-1) is smaller than the P3s from the Early Pleistocene of Baikong, Juyuan, and Queque in southern China and the Late Pleistocene of Tham Prakai Phet in Thailand, its dimensions are larger than those from the Middle Pleistocene of Ganxian in southern China, the Middle Pleistocene of Badak in 10 W. Liao, T. Harrison, Y. Yao et al. Journal of Human Evolution 170 (2022) 103233 By contrast, the Middle and Late Pleistocene Pongo assemblages are predominately þ1 and þ2 SDs (54.2% and 64.3%, respectively). Thus, the Early Pleistocene Pongo assemblages from mainland Southeast Asia are characterized by relatively larger overall teeth than Pongo assemblages from the Middle and Late Pleistocene. Furthermore, the Middle Pleistocene Pongo assemblages have some teeth that can be considered exceptionally small. Previously published d13Cdiet values for mammalian fossil teeth attest to the large-scale variation in vegetation cover in mainland Southeast Asia from the Early to the Late Pleistocene (Fig. 5). The d13Cdiet data indicate that during the Early Pleistocene, mainland Southeast Asia was predominantly a forest environment. By the Middle Pleistocene, this region was dominated by savannahs, but some forests persisted. The second half of the Late Pleistocene saw a slight shift toward woodland and forest environments while grasslands persisted. it is larger than those from the Middle Pleistocene of Badak, the Late Pleistocene of Batu in Peninsular Malaysia, and the Middle/Late Pleistocene of Zhiren in southern China, and falls within the size range of the rest of the comparative samples. Except for three specimens (P041, P042, and P043) of P. weidenreichi from the Early Pleistocene of Baikong, the Yicun M2 (YCT2-4) is larger than the M2s from the Middle/Late Pleistocene of Zhiren in southern China, the Middle-Late Pleistocene of Punung in Java, and the Late Pleistocene of Niah in Borneo, and falls within the size range of the remaining comparative samples included in this study. 3.4. Overall dental size of Pongo assemblages and paleoenvironmental reconstruction The overall dental size distribution of fossil Pongo assemblages from the Early to Late Pleistocene are presented in terms of SD units of the occlusal areas of the fossil Pongo teeth relative to the extant Pongo reference standard from Harrison (2000). Forty-six percent of the teeth of the Yicun Pongo assemblage are 3 SD or greater than the mean of the extant orangutan samples (Table 4). The Yicun Pongo teeth are relatively larger overall than other Late Pleistocene Pongo assemblages in southern China. The Pongo assemblages from Shuangtan Cave, Zhiren Cave, Baxian Cave, and Quzai Cave have, respectively, 8.0%, 15.4%, 17.2%, and 36.6% of their teeth that are 3 SD or greater than the mean of extant Pongo. The overall dental size distribution of the Early Pleistocene Pongo assemblages from mainland Southeast Asia is predominately þ2 and þ3 SDs (58.9%). 4. Discussion and conclusions Although Pongo fossils have been found in many cave sites across southern China (Gu et al., 1987; Zhao et al., 2009a; Ibrahim et al., 2013), few sites have been subject to detailed chronological work until recently (i.e., Harrison et al., 2021). Pongo fossils from Baikong, Juyuan, Sanhe, and Queque Caves from the Chongzuo area in Guangxi, provide evidence that documents orangutan evolution during the Early Pleistocene, from ~2.0 Ma to ~1.0 Ma (Harrison et al., 2014; Wang et al., 2014). Pongo fossils from Ganxian in Bubing, Guangxi, provide evidence of tooth size and morphological Table 4 Dental size distribution of the Yicun and comparative fossil orangutan populations. Standard deviation intervals from the mean values of extant orangutans Late Pleistocene Yicun Cave (southern China) Shuangtan (southern China)a Zhiren Cave (southern China)b Baxian (southern China)a Quzai (southern China)a Keo Leng (Vietnam)c Nguom rock shelter (Vietnam)c Duoi U’Oi (Vietnam)c Sang Cave (Vietnam)c Tham Prakai Phet (Thailand)d Total n Middle Pleistocene Yixiantian Cave (southern China)a Daxin Hei (southern China)a Ganxian Cave (southern China)e Tham Khuyen (Vietnam)c Lang Trang (Vietnam)c Tham Om (Vietnam)c Thum Wiman Nakin (Thailand)c Total n Early Pleistocene Queque Cave (southern China)f Sanhe Cave (southern China)g Juyuan Cave (southern China)f Baikong Cave (southern China)f Total n a b c d e f g Data Data Data Data Data Data Data from from from from from from from n 4 SD 3 SD 24 25 13 232 41 26 7 16 16 14 414 100.0% 53 137 105 29 122 31 15 492 100.0% 29 26 10 25 90 100.0% 2 SD þ1 SD þ2 SD þ3 SD þ4 SD þ5 SD 5 6 5 74 12 12 6 2 2 31 11 7 3 2 8 4 5 1 1 43 10.4% 8 12 4 92 12 1 3 5 3 2 142 34.3% 1 6 3 124 30.0% 4 7 5 75 18.1% 14 45 32 4 31 10 4 140 28.5% 15 40 22 9 36 7 2 131 26.6% 12 25 17 12 18 8 1 93 18.9% 5 6 1 3 15 16.7% 10 11 2 3 26 28.9% 9 6 3 9 27 30.0% 5 2 25 2 1 1 3 5 2 0.5% 2 1 1 3 0.6% 1 7 1.4% 4 13 20 4 14 6 4 65 13.2% 1 1 3 2 1 1 0.2% 1 SD 1 1.1% 1 1.1% Harrison et al. (2021). Zhao et al. (2009a). Tshen (2016). Filoux and Wattanapituksakul (2019). Liang et al. (submitted). Harrison et al. (2014). Wang et al. (2014). 11 þ6 SD 1 1 2 22 5.3% 1 6 1.4% 6 10 10 2 1 2 15 2 3 41 8.3% 1 8 1.6% 3 0.6% 4 3 3 10 11.1% 1 1 5 7 7.8% 1 2 3 3.3% W. Liao, T. Harrison, Y. Yao et al. Journal of Human Evolution 170 (2022) 103233 Bourgon et al., 2020), Cambodia (Bacon et al., 2018), Thailand (Zeitoun et al., 2010; Filoux and Wattanapituksakul, 2019), and Peninsular Malaysia (Ibrahim et al., 2013), and some of these fossil assemblages are dated to as recently as ~20 ka (Bacon et al., 2015). The Yicun Pongo fossils, dated to between 66.3 ± 0.3 ka and 57.8 ± 0.3 ka, represent the latest orangutan occurrence to date in southern China. The Yicun Pongo fossil assemblage thus helps fill a gap in our knowledge of orangutan evolution during the Late Pleistocene in Mainland Southeast Asia. The fossil orangutans from Yicun expand the temporal range of Pongo in southern China and show that they survived in the region until at least ~60 ka. The major cause(s) for the disappearance of Pongo from southern China apparently only emerged after ~60 ka. The taxonomic and phylogenetic status of the Late Pleistocene Pongo assemblage in southern China has been a source of debate. Initially, the Pongo fossil teeth from Zhiren Cave were classified as Pongo pygmaeus weidenreichi based on their larger cheek teeth with fewer and simpler occlusal enamel wrinkles than both fossil and modern orangutans from Indonesia (Zhao et al., 2009a). Later, the Zhiren Pongo fossil teeth were identified as a separate species, P. devosi, based on their significantly smaller size compared with the Early Pleistocene P. weidenreichi in southern China (Harrison et al., 2014). It was argued that the more diminutive P. devosi replaced P. weidenreichi during the Middle to Late Pleistocene (Harrison et al., 2014). More recently, based on an expanded sample of fossil teeth (>800) from Chongzuo in southern China, Harrison et al. (2021) showed that there was an overall temporal trend of dental size reduction while the distinguishing morphological features of orangutans remained unchanged from the Early to Late Pleistocene. Thus, the Pongo teeth fossils from Zhiren Cave and other Late Pleistocene cave sites have now been classified as P. weidenreichi (Harrison et al., 2021). The Yicun Pongo fossil teeth provide new evidence to document the taxonomic and phylogenetic status of the Late Pleistocene Pongo assemblage in southern China. In general, the Yicun fossil teeth largely plot at the upper end of the size range of extant Pongo teeth, with some specimens falling outside the upper end of this size range. All of the Yicun specimens are larger than the mean values of extant Pongo species, and fall in the range of variation of the Early Pleistocene orangutan assemblages of Baikong, Juyuan, and Queque in the Chongzuo region. The Yicun fossil teeth are larger than those of Late Pleistocene Shuangtan, Zhiren, Baxian, and Quzai of the same region (Table 4). Thus, the Yicun fossil teeth represent a Pongo assemblage relatively larger in dental size than other Late Pleistocene assemblages from the region. In addition, a distinctive feature of the Yicun orangutan assemblage is its high frequency of cingular remnants on the upper and lower molars (Fig. 3). Except for one M1 (YCT1-8), a cingular remnant is observed on all of the remaining 11 M. In addition, the upper and lower molars of the Yicun orangutan assemblage have light to moderate enamel wrinkling on the occlusal surface (Fig. 3). Overall, the dental size and occlusal structure of the Yicun orangutan samples generally fall within the range of P. weidenreichi and can be assigned to this species. These results confirm previous findings that P. weidenreichi survived into the Late Pleistocene in southern China (Harrison et al., 2021). Tooth size dimensions are among the most easily measured variables used to document the evolutionary history of the genus Pongo during the Pleistocene (Hooijer, 1948; Gu et al., 1987; Harrison et al., 2014; Wang et al., 2014). In particular, we were able to compare overall dental size variation among the different orangutan assemblages using the extant Pongo reference standard from Harrison (2000). Previous research has shown that there was a progressive reduction in dental size in orangutans from southern China during the Pleistocene (Harrison et al., 2014, 2021; Wang Figure 5. Comparisons of the dental size distribution of the orangutan assemblage and d13Cdiet data in mainland Southeast Asia from different periods of the Pleistocene. A) Late Pleistocene; B) Middle Pleistocene; and C) Early Pleistocene. Dental size distribution of the orangutan assemblages in mainland Southeast Asia from different periods of the Pleistocene are presented in Table 4. The d13Cdiet data are presented in SOM Table S2. variation of orangutan assemblages during the late Middle Pleistocene (Liang et al., submitted). The orangutan assemblage from the Late Pleistocene site of Zhiren in Chongzuo, Guangxi, has been dated to >106.2 ± 6.7 ka by U-series (Liu et al., 2010) and to 116e106 ka by combined paleomagnetic and optically stimulated luminescence dating methods (Cai et al., 2017). More recent chronological work suggests the Zhiren deposits may actually be older, dating to 190e130 ka (Ge et al., 2020). Late Pleistocene Pongo fossil assemblages have been recovered from Vietnam (Bacon and Long, 2001, 2002; Bacon et al., 2008, 2015), Laos (Bacon et al., 2015; 12 W. Liao, T. Harrison, Y. Yao et al. Journal of Human Evolution 170 (2022) 103233 on fossil orangutan assemblages are consistent with these previous findings for extant orangutan assemblages. The relationship between overall body weight and habitat variation as discussed earlier suggests that arboreal orangutan assemblages in Mainland Southeast Asia during the Pleistocene may have been heavily influenced by their habitat ecology, as represented by the relative proportion of available forest (Fig. 5). Similar to what has been observed for extant orangutans, forest habitat degradation may also have resulted in reduced body mass of the orangutan assemblage during the Pleistocene. Nonetheless, we note that our conclusion about the relationship between overall body weight and habitat change must be viewed as provisional. Future research focused on evaluating the relationship between environmental change and Pongo body weight from a broader range of sites across time and space should help to further clarify this potentially important point. et al., 2014). The latest Pongo assemblage from Yicun has a relatively larger overall dental size than the earlier Pongo assemblages from Shuangtan, Zhiren, Baxian, and Quzai Caves in southern China (Table 4). The Yicun Pongo assemblage does not conform to the pattern of progressive reduction in overall dental size, and may indicate a Late Pleistocene dental size differentiation between the Pongo assemblages in southern China. From a region-wide perspective, the overall dental size reduction of Pongo in mainland Southeast Asia occurred principally during the Early and Middle Pleistocene (Fig. 5). In addition, the Middle Pleistocene Pongo assemblages unexpectedly have some teeth of relatively small size. Tooth size has been shown to be highly correlated with body weight in nonhuman primates, with correlation coefficients for upper and lower teeth with body weight ranging between 0.84e0.95 and 0.85e0.97, respectively (Gingerich et al., 1982; Conroy, 1987). However, it is currently unknown whether tooth size and body weight are highly and significantly correlated in orangutans, and/or whether this relationship scales isometrically. Assuming that tooth size provides a reasonable guide to body weight in orangutan assemblages over time scales, it can be inferred that Early Pleistocene Pongo had a mean body weight that was much larger than those from the Middle and Late Pleistocene. Compared with the Late Pleistocene Pongo assemblages, the Middle Pleistocene Pongo assemblages are distinguished by some individuals with what appear to be the smallest body weights. We must acknowledge, however, that the current approach has unknown errors associated with it, and that pooling across tooth positions in the present study introduces more noise so that some degree of caution in interpreting these results is warranted. When combined with the stable carbon isotope data from fossil mammals (SOM Table S1), this study suggests that habitat change may have impacted the body weight of Pongo assemblages in mainland Southeast Asia during the Pleistocene (Fig. 5). In forest environments dominated by C3 plants, Pongo assemblages from the Early Pleistocene are characterized by larger overall dental sizes, and, assuming dental size accurately tracks body mass, relatively larger body weight. By the Middle Pleistocene, the widespread open savannahs in Mainland Southeast Asia are associated with Pongo assemblages that have undergone a reduction in body weight, with some individuals having the smallest body weights. Sustainable forest environments in parts of Mainland Southeast Asia from the Middle to Late Pleistocene provide support for some Pongo assemblages of overall larger body weight. The relationship between body weight and habitat variation suggests that arboreal orangutan assemblages in the Pleistocene may have been heavily influenced by their habitat and ecology as represented by the relative proportion of available forests. The results presented here are consistent with field studies of extant orangutan assemblages on Sumatra and Borneo. As habitat quality decreases from Sumatra to Borneo, cranial capacity of adult Pongo species has been shown to decrease from west to east, with a significantly smaller cranial capacity of the easternmost Pongo pygmaeus morio compared to most other orangutan groups (Taylor and van Schaik 2007). Along with the cranial capacity decrease, average postcanine dental sizes of P. pygmaeus from Borneo have been reported to be smaller than those of P. abelii from Sumatra (Uchida, 1996; Cameron, 2001). Studies focused on the body mass of Bornean orangutans also demonstrate that habitat degradation may reduce the body mass of the adult assemblages (Rayadin and Spehar, 2015). A recent study has further shown that the estimated lean body mass of P. pygmaeus was lower during low-fruit periods in all age-sex classes examined, including flanged males, unflanged males, adult females, and adolescents (O'Connell et al., 2021). Our preliminary findings of the impact of habitat variation Declaration of competing interest 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. Acknowledgments We thank S. Xing from the Institute of Vertebrate Paleontology and Paleoanthropology for inviting us to contribute to this special  de Paris for sharing issue. We thank A.-M. Bacon from Universite the metrical data of extant Pongo species, G. F. Lu and X. H. Wu from the Cultural Relics Bureau of Chongzuo City, J.Y. Li and F. Tian from the Tiandong County Museum, C. L. Huang from the Natural History Museum of Guangxi, and S. S. Wei, J. P. Li, X. Fu, and B. Xiao from the Anthropology Museum of Guangxi for their assistance with field work. Critical comments and thoughtful suggestions from Andrea Taylor (co-Editor-in-Chief), the Guest Editor, and three anonymous reviewers helped us improve the manuscript. This work was supported by the National Natural Science Foundation of China (42002025), Major Program of National Social Science Foundation of China (20&ZD246), and the BaGui Scholars Project of the Guangxi Zhuang Autonomous Region. Supplementary Online Material Supplementary online material related to this article can be found at https://doi.org/10.1016/j.jhevol.2022.103233. References Bacon, A.-M., Long, V.T., 2001. The first discovery of a complete skeleton of a fossil orang-utan in a cave of the Hoa Binh Province, Vietnam. J. Hum. Evol. 41, 227e241. Bacon, A.-M., Long, V.T., 2002. Erratum: The first discovery of a complete skeleton of a fossil orang-utan in a cave of the Hoa Binh Province, Vietnam. J. Hum. Evol. 42, 505. Bacon, A.-M., Demeter, F., Duringer, P., Helm, C., Bano, M., Long, Vu The, Thuy, Nguyen Kim, Antoine, P.-O., Mai, Bui Thi, Huong, Nguyen Thi Mai, Dodo, Y., Chabaux, F., Rihs, S., 2008. The Late Pleistocene Duoi U’Oi cave in northern Vietnam: Palaeontology, sedimentology, taphonomy, palaeoenvironments. Quat. Sci. Rev. 27, 1627e1654. Bacon, A.-M., Westaway, K., Antoine, P.-O., Duringer, P., Blin, A., Demeter, F., Ponche, J.L., Zhao, J.x., Barnes, L., Sayavongkhamdy, T., Nguyen, T.K.T., PatoleEdoumba, E., Vu, T.L., Shackelford, L., 2015. Late Pleistocene mammalian assemblages of Southeast Asia: New dating, mortality profiles and evolution of the predator-prey relationships in an environmental context. Palaeogeogr. Palaeoclimatol. Palaeoecol. 422, 101e127. Bacon, A.-M., Duringer, P., Westaway, K., Joannes-Boyau, R., Zhao, J.-x., Bourgon, N., Dufour, E., Pheng, S., Tep, S., Ponche, J.-L., Barnes, L., Blin, A., Patole-Edoumba, E., Demeter, F., 2018. Testing the savannah corridor hypothesis during MIS2: The Boh Dambang hyena site in southern Cambodia. Quat. Int. 464, 417e439. Badoux, D.M., 1959. Fossil Mammals From Two Fissure Deposits at Punung (Java). Kemink en Zoon, Utrecht. 13 W. Liao, T. Harrison, Y. Yao et al. Journal of Human Evolution 170 (2022) 103233 Delehaunty, K.D., Fronick, C., Schmidt, H., Fulton, L.A., Fulton, R.S., Nelson, J.O., Magrini, V., Pohl, C., Graves, T.A., Markovic, C., Cree, A., Dinh, H.H., Hume, J., Kovar, C.L., Fowler, G.R., Lunter, G., Meader, S., Heger, A., Ponting, C.P., MarquesBonet, T., Alkan, C., Chen, L., Cheng, Z., Kidd, J.M., Eichler, E.E., White, S., Searle, S., Vilella, A.J., Chen, Y., Flicek, P., Ma, J., Raney, B., Suh, B., Burhans, R., Herrero, J., Haussler, D., Faria, R., Fernando, O., Darre, F., Farre, D., Gazave, E., Oliva, M., Navarro, A., Roberto, R., Capozzi, O., Archidiacono, N., Della Valle, G., Purgato, S., Rocchi, M., Konkel, M.K., Walker, J.A., Ullmer, B., Batzer, M.A., Smit, A.F.A., Hubley, R., Casola, C., Schrider, D.R., Hahn, M.W., Quesada, V., Puente, X.S., Ordonez, G.R., Lopez-Otin, C., Vinar, T., Brejova, B., Ratan, A., Harris, R.S., Miller, W., Kosiol, C., Lawson, H.A., Taliwal, V., Martins, A.L., Siepel, A., RoyChoudhury, A., Ma, X., Degenhardt, J., Bustamante, C.D., Gutenkunst, R.N., Mailund, T., Dutheil, J.Y., Hobolth, A., Schierup, M.H., Ryder, O.A., Yoshinaga, Y., de Jong, P.J., Weinstock, G.M., Rogers, J., Mardis, E.R., Gibbs, R.A., Wilson, R.K., 2011. Comparative and demographic analysis of orangutan genomes. Nature 469, 529e533. Louys, J., Roberts, P., 2020. Environmental drivers of megafauna and hominin extinction in Southeast Asia. Nature 586, 402e406. Ludwig, K.R., 2003. Users manual for isoplot/ex version 3.0: A geochronological toolkit for microsoft excel. Berkeley Geochronology Center Special Publication No.3, Berkeley. Manduell, K.L., Harrison, M.E., Thorpe, S.K., 2012. Forest structure and support availability influence orangutan locomotion in Sumatra and Borneo. Am. J. Primatol. 74, 1128e1142. Mattle-Greminger, M.P., Sonay, T.B., Nater, A., Pybus, M., Desai, T., de Valles, G., Casals, F., Scally, A., Bertranpetit, J., Marques-Bonet, T., van Schaik, C.P., Anisimova, M., Kruetzen, M., 2018. Genomes reveal marked differences in the adaptive evolution between orangutan species. Genome Biol. 19, 193. Meijaard, E., Ni’matullah, S., Dennis, R., Sherman, J., Wich, S.A., 2021. The historical range and drivers of decline of the Tapanuli orangutan. PLoS One 16, e0238087. Morrogh-Bernard, H., Husson, S., Page, S.E., Rieley, J.O., 2003. Population status of the Bornean orang-utan (Pongo pygmaeus) in the Sebangau peat swamp forest, Central Kalimantan, Indonesia. Biol. Conserv. 110, 141e152. Nater, A., Greminger, M.P., Arora, N., van Schaik, C.P., Goossens, B., Singleton, I., Verschoor, E.J., Warren, K.S., Kruetzen, M., 2015. Reconstructing the demographic history of orang-utans using Approximate Bayesian Computation. Mol. Ecol. 24, 310e327. Nater, A., Mattle-Greminger, M.P., Nurcahyo, A., Nowak, M.G., de Manuel, M., Desai, T., Groves, C., Pybus, M., Sonay, T.B., Roos, C., Lameira, A.R., Wich, S.A., Askew, J., Davila-Ross, M., Fredriksson, G., de Valles, G., Casals, F., PradoMartinez, J., Goossens, B., Verschoor, E.J., Warren, K.S., Singleton, I., Marques, D.A., Pamungkas, J., Perwitasari-Farajallah, D., Rianti, P., Tuuga, A., Gut, I.G., Gut, M., Orozco-terWengel, P., van Schaik, C.P., Bertranpetit, J., Anisimova, M., Scally, A., Marques-Bonet, T., Meijaard, E., Krützen, M., 2017. Morphometric, behavioral, and genomic evidence for a new orangutan species. Curr. Biol. 27, 3487e3498. O'Connell, C.A., DiGiorgio, A.L., Ugarte, A.D., Brittain, R.S., Naumenko, D.J., Utami Atmoko, S.S., Vogel, E.R., 2021. Wild Bornean orangutans experience muscle catabolism during episodes of fruit scarcity. Sci. Rep. 11, 10185. Pandong, J., Gumal, M., Aton, Z.M., Sabki, M.S., Koh, L.P., 2019. Threats and lessons learned from past orangutan conservation strategies in Sarawak, Malaysia. Biol. Conserv. 234, 56e63. Rayadin, Y., Spehar, S.N., 2015. Body mass of wild Bornean orangutans living in human-dominated landscapes: Implications for understanding their ecology and conservation. Am. J. Phys. Anthropol. 157, 339e346. Russon, A.E., Erman, A., Dennis, R., 2001. The population and distribution of orangutans (Pongo pygmaeus pygmaeus) in and around the Danau Sentarum Wildlife Reserve, West Kalimantan, Indonesia. Biol. Conserv. 97, 21e28. Schwartz, J.H., Long, V.T., Cuong, N.L., Kha, L.T., Tattersall, I., 1994. A diverse hominoid fauna from the Late Middle Pleistocene breccia cave of Tham Khuyen, Socialist Republic of Vietnam. Anthropol. Pap. Am. Mus. Nat. Hist. 73, 1e11. Schwartz, J.H., Long, V.T., Cuong, N.L., Kha, L.T., Tattersall, I., 1995. A review of the Pleistocene hominoid fauna of the Socialist Republic of Vietnam (excluding Hylobatidae). Anthropol. Pap. Am. Mus. Nat. Hist. 76, 1e24. Seaman, D.J., Bernard, H., Ancrenaz, M., Coomes, D., Swinfield, T., Milodowski, D.T., Humle, T., Struebig, M.J., 2019. Densities of Bornean orang-utans (Pongo pygmaeus morio) in heavily degraded forest and oil palm plantations in Sabah, Borneo. Am. J. Primatol. 81, e23030. St Pierre, E., Zhao, J.-x., Reed, E., 2009. Expanding the utility of Uranium-series dating of speleothems for archaeological and palaeontological applications. J. Archaeol. Sci. 36, 1416e1423. St Pierre, E., Zhao, J.-x., Feng, Y.-x., Reed, E., 2012. U-series dating of soda straw stalactites from excavated deposits: Method development and application to Blanche Cave, Naracoorte, South Australia. J. Archaeol. Sci. 39, 922e930. St Pierre, E.J., Westaway, K.E., Zhao, J.-x., Gagan, M.K., Lentfer, C., Due, R.A., Morwood, M.J., Hantoro, W.S., Djubiantono, T., Suwargadi, B.W., 2013. Preliminary U-series and thermoluminescence dating of excavated deposits in Liang Bua sub-chamber, Flores, Indonesia. J. Archaeol. Sci. 40, 148e155. Sun, L., Wang, Y., Liu, C., Zuo, T., Ge, J., Zhu, M., Jin, C., Deng, C., Zhu, R., 2014. Magnetochronological sequence of the Early Pleistocene Gigantopithecus faunas in Chongzuo, Guangxi, southern China. Quat. Int. 354, 15e23. Taylor, A.B., 2006. Feeding behavior, diet, and the functional consequences of jaw form in orangutans, with implications for the evolution of Pongo. J. Hum. Evol. 60, 377e393. Bourgon, N., Jaouen, K., Bacon, A.M., Jochum, K.P., Dufour, E., Duringer, P., Poncheh, J.-L., Joannes-Boyaui, R., Boesch, Q., Antoine, P.-O., Hullot, M., Weis, U., Schulz-Kornas, E., Trost, M., Fiorillo, D., Demeter, F., Patole-Edoumba, E., Shackelford, L.L., Dunn, T.E., Zachwieja, A., Duangthongchit, S., Sayavonkhamdy, T., Sichanthongtip, P., Sihanam, D., Souksavatdy, V., Hublin, J.J., Tütken, T., 2020. Zinc isotopes in Late Pleistocene fossil teeth from a Southeast Asian cave setting preserve paleodietary information. Proc. Natl. Acad. Sci. USA 117, 4675e4681. Cai, Y., Qiang, X., Wang, X., Jin, C., Wang, Y., Zhang, Y., Trinkaus, E., An, Z., 2017. The age of human remains and associated fauna from Zhiren Cave in Guangxi, southern China. Quat. Int. 434, 84e91. Cameron, D.W., 2001. Morphometric evolutionary trends in the dental complex of Pongo. Primates 42, 253e266. Cheng, H., Edwards, R.L., Hoff, J., Gallup, C.D., Richards, D., Asmerom, Y., 2000. The half-lives of uranium-234 and thorium-230. Chem. Geol. 169, 17e33. Clark, T.R., Roff, G., Zhao, J.-x., Feng, Y.-x., Done, T.J., Pandolfi, J.M., 2014. Testing the precision and accuracy of the UeTh chronometer for dating coral mortality events in the last 100 years. Quat. Geochronol. 23, 35e45. Conroy, G.C., 1987. Problems of body-weight estimation in fossil primates. Int. J. Primatol. 8, 115e137. Demeter, F., Bae, C.J., 2020. Dispersal barriers into Southeast Asia during the Late Pleistocene: An introduction to Special Issue No. 563. Quat. Int. 563, 1e4. Drawhorn, G.M., 1995. The systematics and paleodemography of fossil orangutans (genus Pongo). Ph.D. Dissertation, University of California. Filoux, A., Wattanapituksakul, A., 2019. The Late Pleistocene Orangutan from Tham Prakai Phet: New discoveries. Ann. Paleontol. 105, 287e293. Ge, J., Deng, C., Wang, Y., Shao, Q., Zhou, X., Xing, S., Pang, H., Jin, C., 2020. Climateinfluenced cave deposition and human occupation during the Pleistocene in Zhiren Cave, southwest China. Quat. Int. 559, 14e23. Gingerich, P.D., Smith, B.H., Rosenberg, K., 1982. Allometric scaling in the dentition of primates and prediction of body weight from tooth size in fossils. Am. J. Phys. Anthropol. 58, 81e100. Groves, C.P., 2001. Primate Taxonomy. Smithsonian Institution Press, Washington. Gu, Y., Huang, W., Song, F., Guo, X., Chen, D., 1987. The study of some fossil orangutan teeth from Guangdong and Guangxi. Acta Anthropol. Sin. 6, 272e283. Harrison, T., 2000. Archaeological and ecological implications of the primate fauna from prehistoric sites in Borneo. Bull. Indo-Pacific Prehist. Assoc. 20, 133e146. Harrison, T., Jin, C., Zhang, Y., Wang, Y., Zhu, M., 2014. Fossil Pongo from the Early Pleistocene Gigantopithecus fauna of Chongzuo, Guangxi, southern China. Quat. Int. 354, 59e67. Harrison, T., Zhang, Y., Yang, L., Yuan, Z., 2021. Evolutionary trend in dental size in fossil orangutans from the Pleistocene of Chongzuo, Guangxi, southern China. J. Hum. Evol. 161, 103090. Ho, C.K., Zhou, G.X., Swindler, D.R., 1995. Dental evolution of the orang-utan in China. Hum. Evol. 10, 249e264. Hooijer, D.A., 1948. Prehistoric Teeth of Man and Orang-Utan from Central Sumatra, with Notes on the Fossil Orang-Utan from Java and Southern China, vol. 29. Zool. Meded., Leiden, pp. 175e301. Hooijer, D.A., 1960. The orang-utan in Niah Cave prehistory. Sarawak Mus. J. 9, 408e421. Husson, S.J., Wich, S.A., Marshall, A.J., Dennis, R.D., Ancrenaz, M., Brassey, R., Gumal, M., Hearn, A.J., Meijaard, E., Simorangkir, T., Singleton, I., 2009. Orangutan distribution, density, abundance and impacts of disturbance. In: Wich, S.A., Atmoko, S.S.U., Setia, T.M., van Schaik, C.P. (Eds.), Orangutans: Geographic Variation in Behavioral Ecology and Conservation. Oxford University Press, New York, pp. 77e96. Ibrahim, Y.K., Tshen, L.T., Westaway, K.E., Cranbrook, E.O., Humphrey, L., Muhammad, R.F., Zhao, J.-x., Peng, L.C., 2013. First discovery of Pleistocene orangutan (Pongo sp.) fossils in Peninsular Malaysia: Biogeographic and paleoenvironmental implications. J. Hum. Evol. 65, 770e797. Jin, C., Pan, W., Zhang, Y., Cai, Y., Xu, Q., Tang, Z., Wang, W., Wang, Y., Liu, J., Qin, D., 2009. The Homo sapiens Cave hominin site of Mulan Mountain, Jiangzhou District, Chongzuo, Guangxi with emphasis on its age. Chinese Sci. Bull. 54, 3848e3856. Jin, C., Harrison, T., Dong, W., Bae, C.J., 2014. Multidisciplinary perspectives on the Gigantopithecus fauna and Quaternary biostratigraphy in eastern Asia. Quat. Int. 354, 1e3. Kahlke, H., 1972. A review of the Pleistocene history of the orang-utan (Pongo pede 1799). Asian Perspect. 15, 5e14. Lace Liang, H., Liao, W., Yao, Y., Bae, C.J., Wang, W., 2020. A late Middle Pleistocene mammalian fauna recovered in northeast Guangxi, southern China: Implications for regional biogeography. Quat. Int. 563, 29e37. Liang, H., Harrison, T., Shao, Q.F., Bahain, J.J., Zhao, J.X., Bae, C.J., Liao, W., Wang, W., New Middle Pleistocene Pongo fossils from Ganxian Cave in southern China with implications for understanding the mode and tempo of dental size evolution in orangutans. J. Hum. Evol. submitted. Liu, W., Jin, C.-Z., Zhang, Y.-Q., Cai, Y.-J., Xing, S., Wu, X.-J., Cheng, H., Edwards, R.L., Pan, W.S., Qin, D.-G., 2010. Human remains from Zhirendong, South China, and modern human emergence in East Asia. Proc. Natl. Acad. Sci. USA 107, 19201e19206. Liu, E.-t., Zhao, J.-x., Clark, T.R., Feng, Y.-x., Leonard, N.D., Markham, H.L., Pandolfi, J.M., 2014. High-precision UeTh dating of storm-transported coral blocks on Frankland Islands, northern Great Barrier Reef, Australia. Palaeogeogr. Palaeoclimatol. Palaeoecol. 414, 68e78. Locke, D.P., Hillier, L.W., Warren, W.C., Worley, K.C., Nazareth, L.V., Muzny, D.M., Yang, S.-P., Wang, Z., Chinwalla, A.T., Minx, P., Mitreva, M., Cook, L., 14 W. Liao, T. Harrison, Y. Yao et al. Journal of Human Evolution 170 (2022) 103233 Wich, S.A., Utami-Atmoko, S.S., Setia, T.M., Rijksen, H.D., Schürmann, C., Van Hooff, J., van Schaik, C.P., 2004. Life history of wild Sumatran orangutans (Pongo abelii). J. Hum. Evol. 47, 385e398. Wich, S.A., Setia, T.M., van Schaik, C.P., 2009. Orangutans: Geographic Variation in Behavioral Ecology and Conservation. Oxford University Press, New York. Wich, S.A., Vogel, E.R., Larsen, M.D., Fredriksson, G., Leighton, M., Yeager, C.P., Brearley, F.Q., van Schaik, C.P., Marshall, A.J., 2011. Forest fruit production is higher on Sumatra than on Borneo. PLoS One 6, e21278. Taylor, A., 2009. The functional significance of variation in jaw form in orangutans: The African apes as an ecogeographic model. In: Wich, S.A., Atmoko, S.S.U., Setia, T.M., van Schaik, C.P. (Eds.), Orangutans: Geographic Variation in Behavioral Ecology and Conservation. Oxford University Press, New York, pp. 15e32. Taylor, A.B., van Schaik, C.P., 2007. Variation in brain size and ecology in Pongo. J. Hum. Evol. 52, 59e71. Tshen, L.T., 2016. Biogeographic distribution and metric dental variation of fossil and living orangutans (Pongo spp.). Primates 57, 39e50. Uchida, A., 1996. Craniodental variation among the great apes. Peabody Museum Bulletin 4, Peabody Museum of Archaeology and Ethnology, Harvard University, Cambridge, MA. van Noordwijk, M.A., van Schaik, C.P., 2005. Development of ecological competence in Sumatran orangutans. Am. J. Phys. Anthropol. 127, 79e94. van Schaik, C., Priatna, A., Priatna, D., 1995. Population estimates and habitat preferences of orangutans based on line transects of nests. In: Nadler, R.D., Galdikas, B.F.M., Sheeran, L.K., Rosen, N. (Eds.), The Neglected Ape. Plenum Press, New York, pp. 129e147. Vogel, E.R., Harrison, M.E., Zulfa, A., Bransford, T.D., Alavi, S.E., Husson, S., MorroghBernard, H., Firtsman, T., Utami-Atmoko, S.S., Van Noordwijk, M.A., 2015. Nutritional differences between two orangutan habitats: Implications for population density. PLoS One 10, e0138612. Wang, C.-B., Zhao, L.-X., Jin, C.-Z., Wang, Y., Qin, D.-G., Pan, W.-S., 2014. New discovery of Early Pleistocene orangutan fossils from Sanhe Cave in Chongzuo, Guangxi, southern China. Quat. Int. 354, 68e74. Wich, S.A., Singleton, I., Nowak, M.G., Atmoko, S.S.U., Nisam, G., Arif, S.M., Putra, R.H., Ardi, R., Fredriksson, G., Usher, G., 2016. Land-cover changes predict steep declines for the Sumatran orangutan (Pongo abelii). Sci. Adv. 2, e1500789. Zeitoun, V., Lenoble, A., Laudet, F., Thompson, J., Rink, W.J., Mallye, J.-B., Chinnawut, W., 2010. The Cave of the Monk (Ban Fa Suai, Chiang Dao wildlife sanctuary, northern Thailand). Quat. Int. 220, 160e173. Zhao, J.-x., Hu, K., Collerson, K.D., Xu, H.-k., 2001. Thermal ionization mass spectrometry U-series dating of a hominid site near Nanjing, China. Geology 29, 27e30. Zhao, L., Wang, C., Jin, C., Qin, D., Pan, W., 2009a. Fossil orangutan-like hominoid teeth from Late Pleistocene human site of Mulanshan Cave in Chongzuo of Guangxi and implications on taxonomy and evolution of orangutan. Chinese Sci. Bull. 54, 3924e3930. Zhao, J.-x., Yu, K.-f., Feng, Y.-x., 2009b. High-precision 238Ue234Ue230Th disequilibrium dating of the recent past: A review. Quat. Geochronol. 4, 423e433. 15