New genus of extinct Holocene gibbon associated with humans in Imperial China

See allHide authors and affiliations

Science  22 Jun 2018:
Vol. 360, Issue 6395, pp. 1346-1349
DOI: 10.1126/science.aao4903

The noblewoman's ape

Human activities are causing extinctions across a wide array of taxa. Yet there has been no evidence of humans directly causing extinction among our relatives, the apes. Turvey et al. describe a species of gibbon found in a 2200- to 2300-year-old tomb ascribed to a Chinese noblewoman. This previously unknown species was likely widespread, may have persisted until the 18th century, and may be the first ape species to have perished as a direct result of human activities. This discovery may also indicate the existence of unrecognized primate diversity across Asia.

Science, this issue p. 1346


Although all extant apes are threatened with extinction, there is no evidence for human-caused extinctions of apes or other primates in postglacial continental ecosystems, despite intensive anthropogenic pressures associated with biodiversity loss for millennia in many regions. Here, we report a new, globally extinct genus and species of gibbon, Junzi imperialis, described from a partial cranium and mandible from a ~2200- to 2300-year-old tomb from Shaanxi, China. Junzi can be differentiated from extant hylobatid genera and the extinct Quaternary gibbon Bunopithecus by using univariate and multivariate analyses of craniodental morphometric data. Primates are poorly represented in the Chinese Quaternary fossil record, but historical accounts suggest that China may have contained an endemic ape radiation that has only recently disappeared.

A Warring States–period tomb excavated in 2004 at Shenheyuan, Xi’an (formerly the ancient capital Chang’an), Shaanxi—possibly attributable to Lady Xia, grandmother of China’s first emperor Qin Shihuang (259–210 BCE)—contains 12 pits with animal remains (Fig. 1) (1, 2). Similar tomb menageries are known from other Chinese high-status burials of comparable age (3). Pit K12 contains skeletons of leopard (Panthera pardus), lynx (Lynx lynx), Asiatic black bear (Ursus thibetanus), crane (Grus sp.), domestic mammals and birds (1), and a gibbon (Shaanxi Provincial Institute of Archaeology, Shenheyuan M1K12:3).

Fig. 1 Cranium and mandible of Junzi imperialis holotype (M1K12:3).

(A) Cranium, anterior view. (B) Mandible, lateral view. (C) Upper dentition, occlusal view. (D) Lower dentition, occlusal view. (E) Right M3. Scale bar, 10 mm. (Inset) Modern distribution of hylobatids [dark gray; modified from (18)] and historical distribution across China [light gray; modified from (20)], showing Chang’an (star), Bunopithecus sericus collection locality (solid circle), and major rivers.

Gibbons and siamangs (Hylobatidae) include four living genera (Hoolock, Hylobates, Nomascus, and Symphalangus) consisting of 20 species (4, 5). Six extant species are known historically from China (5, 6). Gibbons were considered culturally important throughout Chinese history; their perceived “noble” characteristics made them symbols of scholar-officials (junzi), and they became high-status pets from the Zhou Dynasty (1046–256 BCE) (7). They are extremely scarce in China’s Pleistocene-Holocene record, and most premodern remains are isolated teeth or postcrania insufficiently diagnostic for species-level or genus-level identification (8, 9). The most complete Quaternary Chinese hylobatid is a left mandibular fragment from Chongqing (AMNH-18534), probably early-middle Pleistocene in age, described in 1923 as an extinct genus and species, Bunopithecus sericus (10). By contrast, M1K12:3 includes a partial facial skeleton (missing the posterior neurocranium) with complete anterior dentition, left-right PM3-4, and right M1-2; an associated right M3; a partial mandible with almost complete anterior dentition (missing left I2), left-right pm3-4, and right m1-2; and nondiagnostic right distal forelimb elements (Fig. 1).

Destructive sampling of M1K12:3 was not possible because of the single specimen’s protected archaeological status, and previous attempts to amplify DNA from Chinese Holocene samples have often proved unsuccessful because of poor biomolecule preservation under subtropical conditions (11), so we conducted multivariate and univariate morphometric analyses to determine its affinities to other hylobatids. First, we conducted canonical variate analyses (CVAs) using 16 cranial landmarks shared between M1K12:3 and a dataset including all extant hylobatid genera (Hoolock, n = 53; Hylobates, n = 327; Nomascus, n = 34; and Symphalangus, n = 63) (12, 13). We partially restored a three-dimensional scan of the M1K12:3 cranium before analysis through mirror-imaging and reference-based reconstruction of the zygomatic bone, zygomatic arch, posterior maxilla, and posterior frontal (13). Landmarks are distributed across nearly the entire remaining or restored cranial surface. All CVAs were performed by using genus as the classifying variable when assessing the position of M1K12:3 in morphospace.

Permutation tests (10,000 rounds) for between-group Procrustes and Mahalanobis distances show significant differentiation between all extant genera (P < 0.0001, all comparisons) and between M1K12:3 and extant genera (Fig. 2, Table 1, and table S1). CV1 (60.90% variation) is associated with expansion of the facial region and primarily separates Symphalangus, the largest, most morphologically distinct hylobatid. CV2 (23.04% variation) represents shape changes to the frontal, orbit, and infraorbital region and strongly differentiates M1K12:3 from extant genera owing to its expanded upper anterior neurocranium: M1K12:3 exhibits a more superior position of the frontal posterior margin (bregma, stephanion), the anterior margin (glabella, upper orbital rim) has undergone an inferior shift, the zygomaxillary suture is shortened to give a narrower cheekbone, and molar dentition is more widely set together with an inferior shift (table S2). Posterior probabilities indicate extremely high classification accuracy (96 to 97%) (table S3), with M1K12:3 consistently classified as a separate group.

Fig. 2 Plots of first two canonical variates (CV1 and CV2) of hylobatid cranial and molar analyses.

M1K12:3, red; Bunopithecus, black (m2 only); Hoolock, green; Hylobates, orange; Nomascus, purple; Symphalangus, blue. Cranial plot includes both reconstructions of M1K12:3.

Table 1

Comparisons between M1K12:3 and extant hylobatids for permutation tests (10,000 rounds) of cranial and molar Procrustes and Mahalanobis distances.

View this table:

We collected molar (M1-3, m1-2) landmark data [homologous landmarks at main cusp tips, 20 (upper) or 22 (lower) semilandmarks along outline], tooth crown areas (maximum occlusal area), polygon areas (ratio from lines connecting cusps relative to total occlusal area), and cusp angles (calculated from homologous landmark coordinates) from M1K12:3 (13). We compared these data with a new dataset containing morphometric data for 789 hylobatid molars representing 279 individuals (Hoolock, n = 77; Hylobates, n = 129; Nomascus, n = 41; and Symphalangus, n = 32), including all extant Chinese species and AMNH-18534 (13).

Permutation tests (10,000 rounds) for Mahalanobis distances again show significant differentiation between all extant genera (P < 0.001, all comparisons), although Procrustes distances do not consistently differentiate extant genera, especially Nomascus (table S5). M1K12:3 is statistically differentiated from extant genera in several features, including occlusal area (significantly larger M2, M3, and m2 than Hylobates; significantly smaller M1, m1, and m2 than Symphalangus); larger M3 paracone angle than that of Nomascus; and smaller protoconid, metaconid, entoconid, and/or hypoconid angles than those of all genera (fig. S4 and tables S6 and S7). CVAs derived from semilandmark data demonstrate the distinctive molar shape of M1K12:3. M3, m1, and m2 all fall outside the range of extant hylobatid variation (Fig. 2) and have high CVA classification accuracy (76 to 86%) (table S3). Permutation tests for Procrustes and Mahalanobis distances show significant differentiation between M1K12:3 and extant hylobatids for upper and/or lower molar outline; pairwise distances are greater than between extant genera (Table 1 and table S1). CVA for m2, the only tooth shared by M1K12:3 and AMNH-18534, demonstrates that these specimens are also morphologically distinct (Fig. 2); Bunopithecus shows a very different relationship to extant hylobatids compared with M1K12:3, with a likely close relationship to Hoolock (10).

Although these analyses cannot reconstruct M1K12:3’s phylogenetic affinities, even genomic analyses have proved unable to clarify higher-order hylobatid relationships, possibly because living genera diverged through near-instantaneous radiation ~5 million years ago (14). However, cranial and molar data clearly differentiate M1K12:3 from living hylobatids and the only other Chinese Quaternary hylobatid. We therefore describe M1K12:3 as a new extinct genus and species, Junzi imperialis (15). Although other Holocene primate losses are known (21 extinctions in “ecologically naïve” Madagascan and Caribbean island faunas, with two species persisting beyond 1500 CE) (16, 17), the disappearance of J. imperialis constitutes the first documented postglacial extinction of an ape or of any continental primate.

Gibbons are today restricted to southwestern China (6), with the closest populations >1200 km from Chang’an and separated by major drainages (Fig. 1). Large rivers can represent barriers to gene flow in hylobatids (18), providing biogeographic support for evolutionary differentiation of central Chinese gibbons. Chang’an was an important regional power center under the Qin State and became China’s political and economic center during the Han Dynasty (19); gibbons could therefore have been transported to Chang’an as trade items or tributes. However, other mammals from the Shenheyuan tomb still occur in Shaanxi (6), suggesting a similar local origin for M1K12:3. Contemporary accounts describe gibbons being caught near Chang’an into the 10th century (7), and gibbon survival in Shaanxi until the 18th century (20). Southern Shaanxi represents the northern limit of China’s subtropical forest ecoregion and retains remnant populations of primates and other mammals (such as giant pandas) that co-occurred with gibbons in Quaternary assemblages (6, 21).

Global ecosystems have experienced extreme human-caused biodiversity loss in recent centuries, with extinction rates elevated by several orders of magnitude; it is increasingly accepted that a mass extinction is underway (22, 23). Eastern and southeast Asian biotas have been disrupted disproportionately; this region contains the most threatened mammals (4), and 73% of Asian primates are threatened compared with 60% globally (24). In China, two gibbon species (Hylobates lar and Nomascus leucogenys) have recently disappeared, surviving species are all critically endangered, and the Hainan gibbon (Nomascus hainanus) may be the world’s rarest mammal with ~26 surviving individuals (4).

The background mammalian extinction rate is estimated at 1.8 extinctions/million species/year (22). Because 525 Holocene-Recent primates, including 27 apes, are recognized (5, 24, 25), expected background extinction rates are 9.45 × 10−4/year for primates, and 4.86 × 10−5/year for apes. We could therefore expect 11.1 background primate extinctions and 57% probability of background ape extinction across the 11,700-year Holocene [although only 45% probability of primate extinction and 2% probability of ape extinction since 1500 CE, the International Union for Conservation of Nature (IUCN) threshold used to assess human-caused extinctions (4), and the period into which J. imperialis likely persisted]. A hypothesis of “natural” rather than anthropogenically mediated extinction of J. imperialis therefore cannot be discarded completely. However, few extinctions across the climatically stable Holocene can even questionably be interpreted as nonanthropogenic (16). Central Chinese landscapes have supported among the world’s highest human densities for millennia (19) and experienced extensive Holocene mammal extinctions (21). The discovery of M1K12:3 in a tomb provides direct evidence of human exploitation, and extensive deforestation occurred near Chang’an during the late Imperial period, with remaining high-elevation forests representing suboptimal gibbon habitat (26). Analysis of predictors of Chinese Holocene mammal range loss has shown that best-supported models include an index of anthropogenic impact (21), and reconstruction of historical gibbon decline across China demonstrates extinction following a wavefront of directional pressures that matches known human population expansion (20).

Although primates are disproportionately threatened today (24), previous studies suggest that they have not experienced elevated levels of past extinction (27). However, they are underrepresented in Quaternary archives, which remain understudied across most areas of primate distribution (8, 21). Our description of J. imperialis suggests that past human-caused primate diversity loss may be underestimated, with important implications for understanding extinction vulnerability and informing conservation (24). Our findings also emphasize the extreme vulnerability of hylobatids even compared with other primates. Historical records document former gibbon occurrence across central and southern China (7, 20), in areas separated from distributions of extant species and J. imperialis by major drainages (Fig. 1). These populations may represent undescribed extinct species, suggesting a much greater historical loss of global ape diversity. We encourage further investigation of Asian environmental archives to reconstruct past human-caused biodiversity loss in this global conservation hotspot and provide new insights for understanding faunal vulnerability and resilience in order to help prevent future extinctions.

Supplementary Materials

Materials and Methods

Supplementary Text

Figs. S1 to S4

Tables S1 to S7

References (2844)

References and Notes

  1. Materials and methods are available as supplementary materials.
  2. Systematic paleontology is available as supplementary materials.
Acknowledgments: Access to collections was provided by the Institute of Zoology, Beijing; Kunming Institute of Zoology; South China Institute of Endangered Animals; Natural History Museum, London; American Museum of Natural History; Center for the Study of Human Origins; National Museum of Natural History; Museum of Comparative Zoology; and the Zoological Museum, Vietnam National University. We thank J. Korsgaard, C. Kimock, H. Ma, and B. Garrod for assistance. Funding: Research was supported by a Royal Society University Research Fellowship (UF080320/130573). Author contributions: S.T.T. and H.J.C. designed the research; J.H., A.O., S.H., Y.D., T.Z., S.T.T., and K.B. collected data; K.B., A.O., and H.J.C. analyzed data; and S.T.T., A.O., and H.J.C. wrote the paper. Competing interests: None declared. Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper or the supplementary materials.
View Abstract

Stay Connected to Science

Navigate This Article