Morphology of the thoracolumbar spine of the middle Miocene hominoid Nacholapithecus kerioi from northern Kenya

Journal of Human Evolution 88 (2015) 25e42

Contents lists available at ScienceDirect

Journal of Human Evolution journal homepage: www.elsevier.com/locate/jhevol

Morphology of the thoracolumbar spine of the middle Miocene hominoid Nacholapithecus kerioi from northern Kenya Yasuhiro Kikuchi a, *, Masato Nakatsukasa b, Yoshihiko Nakano c, Yutaka Kunimatsu d, Daisuke Shimizu b, Naomichi Ogihara e, Hiroshi Tsujikawa f, Tomo Takano g, Hidemi Ishida h a

Division of Human Anatomy and Biological Anthropology, Department of Anatomy and Physiology, Faculty of Medicine, Saga University, Saga, 849-8501, Japan Laboratory of Physical Anthropology, Graduate School of Science, Kyoto University, Kyoto, 606-8502, Japan c Laboratory of Biological Anthropology, Graduate School of Human Sciences, Osaka University, Osaka, 565-0871, Japan d Department of Business Administration, Faculty of Business Administration, Ryukoku University, Kyoto, 612-8577, Japan e Laboratory of Evolutionary Biomechanics, Department of Mechanical Engineering, Faculty of Science and Technology, Keio University, Kanagawa, 223-8522, Japan f Department of Rehabilitation, Faculty of Medical Science and Welfare, Tohoku Bunka Gakuen University, Miyagi, 981-8551, Japan g Japan Monkey Centre, Aichi, 484-0081, Japan h Kyoto University, Kyoto, 606-8502, Japan b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 22 April 2015 Accepted 1 September 2015 Available online xxx

A new caudal thoracic and a new lumbar vertebra of Nacholapithecus kerioi, a middle Miocene hominoid from northern Kenya, are reported. The caudal thoracic vertebral body of N. kerioi has a rounded median ventral keel and its lateral sides are moderately concave. The lumbar vertebral body has an obvious median ventral keel. Based on a comparison of vertebral body cranial articular surface size between the caudal thoracic vertebrae in the present study and one discussed in a previous study (KNM-BG 35250BO, a diaphragmatic vertebra), N. kerioi has at least two post-diaphragmatic vertebrae (rib-bearing lumbartype thoracic vertebrae), unlike extant hominoids. It also has thick, rounded, and moderately long metapophyses on the lumbar vertebra that project dorsolaterally. The spinous process bases of its caudal thoracic and lumbar vertebrae originate caudally between the postzygapophyses, as described previously in the KNM-BG 35250 holotype specimen. In other words, the postzygapophyses of N. kerioi do not project below the caudal border of the spinous processes, similar to those of extant great apes, and unlike small apes and monkeys, which have more caudally projecting postzygapophyses. Nacholapithecus kerioi has a craniocaudally expanded spinous process in relation to vertebral body length, also similar to extant great apes. Both these spinous process features of N. kerioi differ from those of Proconsul nyanzae. The caudal thoracic vertebra of N. kerioi has a caudally-directed spinous process, whose tip is tear-drop shaped. These features resemble those of extant apes. The morphology of the spinous process tips presumably helps vertebral stability by closely stacking adjacent spinous process tips as seen in extant hominoids. The morphology of the spinous process and postzygapophyses limits the intervertebral space and contributes to the stability of the functional lumbar region as seen in extant great apes, suggesting that antipronograde activity was included in the positional behavior of N. kerioi. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Vertebra Fossil apes Spinous process Dorsostability Evolution

1. Introduction

* Corresponding author. E-mail addresses: [email protected] (Y. Kikuchi), [email protected]. kyoto-u.ac.jp (M. Nakatsukasa), [email protected] (Y. Nakano), kunimats@ biz.ryukoku.ac.jp (Y. Kunimatsu), [email protected] (H. Ishida). http://dx.doi.org/10.1016/j.jhevol.2015.09.003 0047-2484/© 2015 Elsevier Ltd. All rights reserved.

Studying the anatomy of the lumbar spine (including the functional lumbar vertebrae, i.e., the caudal thoracic vertebrae, the diaphragmatic and, when present, post-diaphragmatic vertebrae) in Miocene apes compared to extant primates is essential in understanding the evolution of hominoid posture and locomotion

26

Y. Kikuchi et al. / Journal of Human Evolution 88 (2015) 25e42

(Walker and Rose, 1968; Ward, 1991, 1993, 2007; Sanders and Bodenbender, 1994; Harrison, 1999; MacLatchy et al., 2000;
-Sola
et al., 2004; Nakatsukasa et al., MacLatchy, 2004; Moya 2007; Nakatsukasa, 2008; McCollum et al., 2010). Lumbar spine anatomy has been described for the early Miocene Proconsul heseloni (Walker and Pickford, 1983; Kelley, 1986; Walker et al., 1993) and Proconsul nyanzae (Ward, 1991, 1993; Ward et al., 1993), both dated to 17e18.5 Ma (millions of years ago) (Drake et al., 1988), as well as for Morotopithecus bishopi (Walker and Rose, 1968; Sanders and Bodenbender, 1994; MacLatchy et al., 2000; Nakatsukasa, 2008), dated to either 20 Ma (Gebo et al., 1997) or 17 Ma (Pickford, 1998; Pickford et al., 1999, 2003). Studies have also been undertaken on the lumbar anatomy of the middle Miocene (Feibel and Brown, 1991) Equatorius africanus (Ward et al., 1999), as well as on the species that is the focus of the present study, Nacholapithecus kerioi (Ishida et al., 2004; Nakatsukasa et al., 2003, 2007), from northern Kenya, dated to around 15 Ma (Sawada et al., 1998). Nacholapithecus kerioi specimens have provided much information about the postcranial anatomy and diversity of middle Miocene hominoids (Nakatsukasa and Kunimatsu, 2009), and the new specimens we describe and analyze here shed further light not only on adaptations in N. kerioi itself but also on hominoid locomotor and postural evolution more generally. In cercopithecoids, ventroflexion of the lower vertebrae displaces the application point of the compression from the center of the spine centra towards their ventral margins (Sanders, 1995), with the vertebral body primarily transmitting compression and extension forces, through the ventral centra (Rockwell et al., 1938; Young, 1962). If the center of the vertebral body moves the point of compression even a short distance ventrally, the vertebral body experiences longitudinal bending stress along its ventral face (Badoux, 1974). To resist such stresses, the ventral keel may be well developed (Sanders and Bodenbender, 1994). In extant hominoids, there is not the same range of flexion (Ward, 1991), and extant hominoids probably do not experience the same stresses as are found in the vertebral body of cercopithecoids (Sanders, 1995). Instead, the vertebral body is better adapted to bear loads in orthograde postures, and the lower vertebral centra of extant hominoids consequently have a more columnar appearance without hollowing, pronounced spooling, or ventral keels (Sanders and Bodenbender, 1994). The spinous process morphology of the lumbar vertebrae discriminates extant apes from other primate species. The spinous process is one of the main subjects of the present study and has been recently discussed by Williams and Russo (2015). If the spinous process of a lumbar vertebra is oriented caudally, as seen in extant hominoids but not in cercopithecoids (Ward, 1993; Shapiro, 1993), it may be strongly related to the action of the multifidus muscle in stabilizing the lower back (Slijper, 1946; Shapiro, 1993). Moreover, this orientation may also give stability by closely approximating the adjacent vertebrae, creating bony blocks to extension, particularly when the spinous processes are tall craniocaudally (Slijper, 1946; Shapiro, 1993). The position of the lumbar transverse processes also distinguishes extant apes from other primate species. The transverse processes in extant great apes arise from the neural arch (Ward, 1991, 1993). On the other hand, the transverse processes originate from the vertebral body in monkeys, with small apes having a state intermediate between extant great apes and monkeys (Ward, 1991, 1993). However, even in the last lumber vertebrae of cercopithecoids, the transverse processes arise from the centro-pedicular junction or begin on the pedicles (Sanders, 1995). Diaphragmatic vertebrae are defined as rib-bearing thoracic vertebrae that have thoracic-type prezygapophyses

(prezygapophyseal articular facets directed in the coronal plane) and lumbar-type postzygapophyses (postzygapophyseal articular facets directed in the sagittal plane) (Clauser, 1980). Extant apes exhibit a reduced lumbar spinal length as a result of a decreased number of lumbar vertebrae and a reduction in the craniocaudal length of each vertebral body (Schultz, 1938, 1961; Benton, 1967; Rose, 1975; Ward, 1991, 1993; Shapiro, 1993; Haeusler et al., 2002; McCollum et al., 2010; Williams, 2012a). In extant great apes, the diaphragmatic vertebral level is positioned more caudally (by one to three vertebral levels) than in cercopithecoids and platyrrhines, and is typically also the last rib-bearing vertebra (Washburn and Buettner-Janusch, 1952; Shapiro, 1993; Haeusler et al., 2002; Williams, 2011, 2012b). A reduction in the lumbar spinal length provides a greater degree of lower back stability, and reduces dorsoventral flexibility (Ward, 1993; Sanders and Bodenbender, 1994; Sanders, 1995; Johnson and Shapiro, 1998). Concerning Miocene hominoid lumbar spine anatomy, the stem hominoid Proconsul has a relatively primitive lumbar anatomy for a catarrhine, for example having six or seven lumbar vertebrae (defined not by how the zygapophyses articulate but by possession of rib facets), a long vertebral body with a ventral median keel, and a more ventral position of the transverse processes relative to extant apes (Ward, 1991, 1993). These primitive characters are retained in the lumbar spine of N. kerioi although the transverse process origin is slightly more dorsally situated (Nakatsukasa et al., 2007). The corresponding features in a specimen of Equatorius (KNM-TH 28860Y caudal thoracic vertebra; Sherwood et al., 2002) are unclear because of deformation, but its relatively small and keeled vertebral body recalls those of Proconsul and N. kerioi. The lumbar vertebra of M. bishopi (UMP 67.28) has several derived features, such as a more dorsal position of the transverse process and a craniocaudally short vertebral body with no ventral keel formation (Walker and Rose, 1968; Sanders and Bodenbender, 1994; MacLatchy et al., 2000; Nakatsukasa, 2008). Those lumbar features could be interpreted as shared derived states of Morotopithecus and the extant great apes (MacLatchy et al., 2000; Young and MacLatchy, 2004). From Europe, lumbar vertebral specimens of Pierolapithecus
-Sola
et al., 2004), Hispanopithecus laiecatalaunicus (12 Ma; Moya € hler, 1995, 1996; Ko €hler et al., tanus (9.5e10 Ma; Moy
a-Sol
a and Ko 1999; Susanna et al., 2014), and Oreopithecus bambolii (8 Ma; € hler and Moya
-Sola
, 1997; Harrison and Rook, Harrison, 1986; Ko 1997) are known. These later European apes exhibit a more, but not yet fully, modern ape-like lumbar anatomy in terms of dorsostability and/or invagination of the lumbar spine when compared with earlier African fossil apes such as Proconsul and N. kerioi. For example, the lumbar vertebrae of P. catalaunicus and H. laietanus have more dorsally positioned transverse processes located on the pedicle (pedicular/body-junction in P. catalaunicus), more elliptical
-Sola
body articular surfaces, and no apparent ventral keel (Moya et al., 2004; Susanna et al., 2014). Oreopithecus bambolii also has lumbar vertebrae with dorsally oriented transverse processes located on the pedicle (Harrison and Rook, 1997). These traits are not observed in Proconsul or N. kerioi. As the only derived lumbar vertebral trait in Proconsul, both P. nyanzae and P. heseloni possess caudally oriented spinous processes, shared with extant hominoids, and not found in cercopithecoids (Ward, 1993). In N. kerioi, the base of the spinous process is positioned more caudally between the postzygapophyses than in cercopithecoids, suggesting that this fossil hominoid also had a caudally inclined spinous process (Nakatsukasa et al., 2007). In P. catalaunicus, the spinous process is nearly horizontal and slightly caudally inclined (Susanna et al., 2010). Hispanopithecus laietanus may have a caudally oriented spinous process, as indicated by two

Y. Kikuchi et al. / Journal of Human Evolution 88 (2015) 25e42

neural arch fragments that preserve the base of the spinous process (Susanna et al., 2014). Due to temporal and spatial biases in the fossil record, the evolution of the caudal thoracic and lumbar anatomy of hominoids is not well understood, and, accordingly, details about hominoid locomotor and postural evolution themselves are unclear. The derived lumbar anatomy of extant great apes might have a common evolutionary origin from any fossil hominoid allied with Morotopithecus (MacLatchy et al., 2000; Young and MacLatchy, 2004). Alternatively, the modern-looking lumbar anatomy of Eurasian fossil apes and/or the orangutan might have evolved independently from extant African apes (Larson, 1988; Ward, 2007; Moy
a-Sol
a et al., 2009). Otherwise, all the extant great apes would have needed to acquire their dorsostable lumbar spines independently (Lovejoy et al., 2009). To resolve the question of the origin of the great ape lumbar anatomy, clarification of the character states and distinction of synapomorphic and homoplastic features of the fossil ape vertebral anatomy are critically important. Characterizing the N. kerioi caudal thoracic and lumbar spine contributes to the debate over evolutionary aspects of the lumbar anatomy, with the interpretation of derived characters such as the morphology of the spinous processes being especially important to this (Nakatsukasa et al., 2007). Although there are published samples of N. kerioi caudal thoracic and lumbar vertebrae, most of the material has been deformed and weathered. In addition, most information has come from a single male individual, KNM-BG 35250, the holotype of N. kerioi. Here, we describe new caudal thoracic and lumbar vertebrae of N. kerioi. Although the lumbar specimen is craniocaudally compressed, the new caudal thoracic specimen is mostly free from plastic deformation/fragmentation, which helps to refine the character description of the functional lumbar spine of N. kerioi. Our specific aims are to elucidate the number of the post-diaphragmatic vertebrae in N. kerioi by estimating with precision the level of this caudal thoracic specimen, and to clarify the spinous process morphology of these two specimens in N. kerioi. Moreover, the degree of similarity of the caudal thoracic specimen to the post-diaphragmatic vertebrae in Old World monkeys and New World monkeys or to the diaphragmatic vertebrae in extant apes should be clarified. 2. Materials and methods 2.1. Sample and data collection The caudal thoracic vertebra KNM-BG 42810B and the lumbar vertebra KNM-KB 42763C that are the focus of this study were excavated from the horizon of the BG-K fossil site in 2002 and are held at the National Museums of Kenya. The exact vertebral level of KNM-BG 42810B is not clear, because, although the specimen is mostly in good condition, the prezygapophyses are deformed. On the other hand, it is clear that the previously published N. kerioi caudal thoracic vertebral specimen KNM-BG 35250BO is a diaphragmatic vertebra, because typical thoracic prezygapophyseal articular facets and typical lumbar postzygapophyseal articular facets are both preserved (Nakatsukasa et al., 2007). The size difference between KNM-BG 35250BO and KNM-BG 42810B is large, with CrA values of 176 mm2 (Nakatsukasa et al., 2007) and 264.5 mm2 (this study, see measurement methods below) respectively. Sexing was relatively easy for the N. kerioi vertebra described and analyzed here, as body mass (BM) sexual dimorphism in N. kerioi is thought to be very large, with males being nearly twice the size of females (Ishida et al., 2004); the BM of an adult male (KNM-BG 35250) was estimated at ~22 kg (Ishida et al., 2004; Nakatsukasa et al., 2007; Kikuchi et al., 2012), and male lumbar vertebral size is known from the vertebral series of KNM-BG 35250.

27

KNM-BG 42810B is a male specimen since it is larger than the postdiaphragmatic thoracic vertebra of KNM-BG 35250BP, and KNM-KB 42763C is also a male specimen because it is similar in size to the ultimate lumbar vertebra of KNM-BG 35250BT. The extant comparative sample used in this study, drawn from collections at the Anthropological Institute and Museum, University of Zürich (Switzerland), Museum für Naturkunde, Berlin (Germany), and the Royal Museum for Central Africa (Belgium), is detailed in Table 1. All extant samples were adults (maturation state was noted in the specimen box and checked by the dental eruption or epiphyseal union of limb bones), and none of the subjects exhibited any bone disease or deformation. The measurement set used is summarized in Table 2 and illustrated in Figure 1. The six vertebral body linear measurements (VL, ventral body length of the vertebra; DL, dorsal body length of the vertebra; CrH, cranial body height of the vertebra; CrW, cranial body width of the vertebra; CaH, caudal body height of the vertebra; and CaW, caudal body width of the vertebra) were taken using digital calipers (BestoolKANON Nakamura-Seisakusho Inc., AND Inc., Tokyo, Japan). Measurement error using these calipers were within 2% (CV, coefficient of variation) as estimated by 10 replicate measurements. The cranial articular surface area (CrA) of the vertebral body and four linear and angle measurements (PL, right pedicular basal length; SH, spinous process basal height; SL, spinous process length nearly in the midline; and SPA, spinous process angle) were measured from photographs (Canon Eos 5D digital single lens reflex camera with standard 50-mm macro lens, Canon Inc., Tokyo) using ImageJ 1.48v

Table 1 Extant primate species for metric comparison with Nacholapithecus kerioi.a Taxon

Group1

Group2

Pan paniscus Pan troglodytes Gorilla gorilla Pongo pygmaeus Symphalangus syndactylus Hylobates sp.

P. paniscus P. troglodytes Gorilla Pongo Symphalangus Hylobates

GA GA GA GA SA SA

Cercopithecus albogularis Cercopithecus ascanius Cercopithecus diana Cercopithecus hamlyni Cercopithecus l'hoesti Cercopithecus mitis Cercopithecus mona Chlorocebus aethiops Chlorocebus pygerythrus Chlorocebus tantalus Lophocebus albigena Mandrillus sphinx Papio anubis Papio hamadryas Theropithecus gelada Colobus angolensis Colobus guereza Nasalis larvatus Piliocolobus foai Presbytis sp. Semnopithecus entellus Trachypithecus cristatus Trachypithecus obscurus

Cercopithecus Cercopithecus Cercopithecus Cercopithecus Cercopithecus Cercopithecus Cercopithecus Chlorocebus Chlorocebus Chlorocebus Lophocebus Mandrillus Papio Papio Theropithecus Colobus Colobus Nasalis Colobus Presbytis Semnopithecus Trachypithecus Trachypithecus

OWM OWM OWM OWM OWM OWM OWM OWM OWM OWM OWM OWM OWM OWM OWM OWM OWM OWM OWM OWM OWM OWM OWM

Alouatta sp. Ateles geoffroyi Brachyteles arachnoides Cacajao sp. Cebus sp. Lagothrix sp. Pithecia sp.

Alouatta Ateles Brachyteles Cacajao Cebus Lagothrix Pithecia

NWM NWM NWM NWM NWM NWM NWM

a Abbreviations: GA, extant great apes; SA, small apes; OWM, Old World monkeys; NWM, New World monkeys.

28

Y. Kikuchi et al. / Journal of Human Evolution 88 (2015) 25e42

Table 2 Measurements of the caudal thoracic vertebrae with abbreviations. Abbreviation CrA VL DL PL CrH CrW CaH CaW SHa SLb SPA INW DPP VAP

Label in Figure 1

Measurement

a b c d e f g h i j k l m

Cranial articular surface area of the vertebral body (not including costal facets) Ventral body length of the vertebra Dorsal body length of the vertebra Right pedicular basal length (just posterior to the vertebral body) Cranial body height of the vertebra (median sagittal plane) Cranial body width of the vertebra (maximum) Caudal body height of the vertebra (median sagittal plane) Caudal body width of the vertebra (maximum) Spinous process basal height craniocaudally Spinous process length nearly in the midline dorsoventrally Spinous process angle (between caudal border line of spinous process and dorsal border of vertebral body) Inferior vertebral notch width at the level of the caudal vertebral body surface Degree of projection of the postzygapophysis in relation to the level of the caudal vertebral body surface Ventral angle of the postzygapophysis in relation to the dorsal vertebral body surface

a

Maximum height just posterior to the postzygapophysis. If the accessory process obscures the dorsal parts of the inferior vertebral notch, the ventral border of the middle line of the spinous process is defined as cranially extended line of the ventral border of the postzygapophysis. b

software (http://imagej.nih.gov/ij/index.html). Using a spirit level to keep the camera lens horizontal, a clear scale was photographed next to the samples. Both sample placement during photography and measuring the resulting image may cause measurement errors. For this reason, measurement error was checked and calculated by

Figure 1. Measurements used in this study, as illustrated on the diaphragmatic vertebra of a Pan troglodytes specimen. Labels as detailed in Table 2.

resetting the vertebrae followed by measuring the photo using ImageJ 1.48v software 10 times. These measurement errors were within 2.2% (CV). Photographs of the vertebral specimens at various distances between the samples and lens were taken to assess measurement error due to parallax. These errors were below 1.8% (CV) when measured 17 times at 30 mm increments between 200 and 680 mm from the specimen to the lens surface, using an ultimate thoracic vertebra of a female Pan troglodytes (average CrA, 441 mm2; full length, 56.0 mm dorsoventrally). Because the maximum value (CV) due to these two errors is 4.0%, the measurement errors can be discounted. Diaphragmatic vertebrae in extant great apes, small apes, Old World monkeys, and New World monkeys, and the first and second post-diaphragmatic vertebrae in Old World monkeys and New World monkeys were measured. In this study, the vertebra caudal to the diaphragmatic vertebra was defined as the first postdiaphragmatic vertebra, and the next caudal one as the second post-diaphragmatic vertebra (these are rib-bearing). Five specimens of New World monkeys did not have second postdiaphragmatic vertebrae and so data could not be obtained. Data for DL, PL, CrH, CrW, CaW, SH, and SL in P. nyanzae were obtained from Ward et al. (1993). Although the vertebral levels of the diaphragmatic vertebrae (extant great apes, small apes, Old World monkeys, and New World monkeys), the first post-diaphragmatic vertebrae (Old World monkeys and New World monkeys), and the second post-diaphragmatic vertebrae (Old World monkeys and New World monkeys) differ among the primate species, the functional characters at each level are considered to be mostly similar (Clauser, 1980; Sanders, 1995; Williams, 2011). This is because each vertebra has the same pre- and post-zygapophyseal articular facet orientation and shape. For this reason, comparisons using these vertebrae are considered to be valid. The BM of extant primate species was estimated from the proximal tibial widths and the femoral head craniocaudal widths using two sex-pooled formulas (proximal tibial width versus BM, and femoral head craniocaudal width versus BM; Ruff, 2003), and the averages of the estimated BMs from the two formulae were used. The BMs of extant primate species were used for the spinous process restoration of KNM-BG 42810B, and clarifying the relationships among extant and extinct samples between BM and the measurements (CrA, VL, DL, PL, CrH, CrW, CaH, CaW, SH, and SL) with scatter plot diagrams. In previous studies, male N. kerioi BM was estimated to be approximately 22 kg (Rose et al., 1996; Ishida et al., 2004; Nakatsukasa et al., 2007; Kikuchi et al., 2012). However, since that value was preliminary, 22 ± 5 kg (17e27 kg) was

Y. Kikuchi et al. / Journal of Human Evolution 88 (2015) 25e42

used for male N. kerioi (±5 kg: based on the standard deviation of male BMs in the similarly-sized Papio spp. and Mandrillus sphinx; Delson et al., 2000). The BM of P. nyanzae estimated from the KNMMW 13142A specimen was 34.3 kg (Ruff et al., 1989). 2.2. Vertebral level estimation for N. kerioi KNM-BG 42810B To examine whether the CrA size difference between the N. kerioi specimens KNM-BG 35250 and KNM-BG 42810B is due to variation within the same species or due to vertebral level differences, the following analysis was conducted. Using four female Cercopithecus mitis, two male Chlorocebus aethiops, two female Chlorocebus pygerythrus, four male Lophocebus albigena, seven male Papio hamadryas, two female Theropithecus gelada, and two female Colobus guereza, the minimum CrA of the diaphragmatic vertebra, the maximum CrA of the diaphragmatic vertebra, the maximum CrA of the first post-diaphragmatic vertebra, and the maximum CrA of the second post-diaphragmatic vertebra were examined within the same sex and species. These data were natural log (Ln) transformed and a least square regression (LSR) was applied with the minimum CrA of the diaphragmatic vertebra as the independent variable, and the maximum CrA of the diaphragmatic vertebrae, the maximum CrA of the first post-diaphragmatic vertebrae, or the maximum CrA of the second post-diaphragmatic vertebrae as the dependent variables. In each of these comparisons, the position of the N. kerioi sample was examined, assuming that KNM-BG 35250BO represented the minimum CrA of the diaphragmatic vertebra, and that KNM-BG 42810B represented either the maximum CrA of the diaphragmatic vertebra, the maximum CrA of the first post-diaphragmatic vertebra, or the maximum CrA of the second post-diaphragmatic vertebra. Using these procedures, the vertebral level of KNM-BG 42810B was estimated by comparison with KNM-BG 35250BO and the CrA size of Old World monkeys. 2.3. Spinous process restoration for N. kerioi KNM-BG 42810B Since the spinous process of KNM-BG 42810B is thought to be oriented more caudoventrally than it was in life, the original orientation was estimated and the spinous process was restored in a right-hand view. The inferior vertebral notch width at the level of the caudal vertebral body surface (INW, k in Fig. 1), the degree of projection of the postzygapophysis in relation to the level of the caudal vertebral body surface (DPP, l in Fig. 1), and the ventral angle of the postzygapophysis in relation to the dorsal vertebral body surface (VAP, m in Fig. 1) were measured from photographs using ImageJ 1.48v software for the diaphragmatic vertebrae of extant great apes, Old World monkeys, and New World monkeys, and for the continuing two caudal thoracic vertebrae (first and second post-diaphragmatic vertebrae) of Old World monkeys and New World monkeys (Table 2). The maximum value (CV) due to combining measurement errors (for DPP and VAP) due to parallax with those created by resetting the vertebra followed by measuring the resulting image using ImageJ 1.48v software ten times was 7.0% (the CV value for only the INW is below 4.0%). This maximum CV

29

value needs to be considered when interpreting results involving DPP and VAP. The INW and DPP data were Ln transformed, and the interrelationships between BM and these measurements were examined by LSR of extant great apes, Old World monkeys, and New World monkeys. Using the LSR of extant great apes, Old World monkeys, and New World monkeys, the predicted INW and DPP of N. kerioi were calculated. The predicted INW and DPP values, and the average VAP, were then used for virtual restoration of the spinous process of N. kerioi.

2.4. Metric comparison between extant primate species and N. kerioi KNM-BG 42810B Comparisons were made between KNM-BG 42810B, and the first and second post-diaphragmatic vertebrae in Old World monkeys and New World monkeys, as well as with the diaphragmatic vertebrae in extant great apes and small apes. First, the Ln transformed CrA and raw SPA (measured from the restored spinous process) of N. kerioi were compared with those of extant primate species by box plot diagrams to assess size and spinous process orientation, respectively. For the CrA, analyses were performed on single sex samples for extant apes, and males in monkeys, in order to compare the size of the N. kerioi specimen, a male, with other males of similar size. The relationship between the Ln transformed BM and the Ln transformed CrA was also examined by scatter plot diagrams to clarify the similarity of the KNM-BG 42810B specimen to both sexes in the extant samples with an equivalent BM. The scatter plot diagrams were then used to evaluate the relationships between Ln BM and the nine linear measurements to elucidate any similarities between the extant species, and N. kerioi and P. nyanzae (ultimate/penultimate thoracic vertebral specimen, KNM-MW 13142H). A principal component analysis (PCA, correlation matrix) was conducted to summarize the overall vertebral shape affinities of N. kerioi and the extant species. To eliminate the effect of different body sizes among anthropoids, the PCA was based on logtransformed Mosimann shape variables (Mosimann, 1970; Jungers et al., 1995). In the PCA of the diaphragmatic vertebrae in apes and the first post-diaphragmatic vertebrae in monkeys, each linear measurement (VL, DL, PL, CrH, CrW, CaH, CaW, SH, and SL) was divided by the geometric mean (GM) of the nine measurements and then Ln transformed. In the PCA of the diaphragmatic vertebrae in apes and the second post-diaphragmatic vertebrae in monkeys, the caudal thoracic vertebral data of P. nyanzae KNM-MW 13142H (Table 3) was also compared with extant species. Because the VL and CaH could not be obtained for KNM-MW 13142H, seven variables were examined. The GM in this case was calculated using seven measurements (DL, PL, CrH, CrW, CaW, SH, and SL). These measurements were divided by the GM and then Ln transformed. All analyses were performed using Microsoft Excel 2013 and JMP 9.0.0 software. Averages and standard deviations of the extant primate samples are listed in the Supplementary Online Material [SOM] Tables 1e3.

Table 3 Measurements (in mm2 for area and mm for linear measurements) of the fossil samples; Nacholapithecus kerioi (KNM-BG 42810B) and Proconsul nyanzae (KNM-MW 13142H).a Accession No. KNM-BG 42810B KNM-MW 13142H

CrA

CaA

VL

DL

PL

CrH

CrW

CaH

CaW

SH

SL

SPA

264.5 e

295.4 e

16.3 e

18.3 23.5

14.7 17.7

15.9 20.0

23.9 27.8

16.4 e

24.9 27.6

16.6 10.0

12.7 24.0

[116] e

The linear data for KNM-MW 13142H were obtained from Ward (1993). Value in square brackets [] measured at modified spinous process. a Abbreviations: CrA, Cranial articular surface area of the vertebral body; CaA, Caudal articular surface area of the vertebral body; VL, Ventral body length of the vertebra; DL, Dorsal body length of the vertebra; PL, Right pedicular basal length; CrH, Cranial body height of the vertebra; CrW, Cranial body width of the vertebra; CaH, Caudal body height of the vertebra; CaW, Caudal body width of the vertebra; SH, Spinous process basal height; SL, Spinous process length in the midline; SPA, Spinous process angle.

30

Y. Kikuchi et al. / Journal of Human Evolution 88 (2015) 25e42

3. Results 3.1. Description 3.1.1. KNM-BG 42810B This fossil is a caudal thoracic vertebra (Fig. 2). The metric data for this specimen are presented in Table 3. A relatively large oval costal facet is positioned on the craniodorsal

part of the vertebral body on each side. This facet is separated from the rim of the cranial articular surface and differs from the demifacet on the diaphragmatic thoracic vertebra KNM-BG 35250BO (Nakatsukasa et al., 2007). The rib facet is relatively large with respect to vertebral body size (Fig. 2) when compared with the rib facet on the last thoracic vertebra of P. nyanzae (KNM-MW 13142H). The prezygapophyses are broken on both sides, but the

Figure 2. Above the dashed line: Nacholapithecus kerioi, KNM-BG 42810B caudal thoracic vertebra, in cranial (a), caudal (b), ventral (c), dorsal (d), right-hand original (e) and restored (g), and left-hand (f) views. Nacholapithecus kerioi, KNM-BG 35250BO, and Proconsul nyanzae, KNM-MW 13142H, in right-hand view are also shown. Please note that the photograph (g) of the restored KNM-BG-42810B is to a different scale to the other photographs. Dotted lines in dorsal (d) and right-hand (e) views of KNM-BG-42810B indicate the fracture line. In the right-hand view (e) of KNM-BG-42810B, the costal facet is bordered (indicated by the black dotted circle). Note the relatively caudal position (indicated by the white arrows) of the spinous process in the right-hand view of KNM-BG-42810B after restoration (g). The arch of the cranial border in the inferior vertebral notch is quite smooth (white arrows in [g]) in restored KNM-BG-42810B. Below the dashed line: Nacholapithecus kerioi, KNM-BG 42763C lumbar vertebrae of Nacholapithecus in cranial (h), ventral (i), right-hand (j) and dorsal views (k). Position of the transverse process base is indicated by brackets in the cranial view (h) and by a white dotted line (t) in the right-hand view (j) in KNM-BG 42763C. Note the well-developed metapophysis (m) bordered with a white dotted line in the dorsal view of KNM-BG 42763C.

Y. Kikuchi et al. / Journal of Human Evolution 88 (2015) 25e42

postzygapophyses are well preserved. The existence of anapophyses and metapophyses is not clear. The vertebral body is relatively well preserved, though it might have been weakly compressed in the dorsoventral direction. A piece of eroded bone aggregate adheres to the vertebral body and closes the left half of the cranial opening of the vertebral canal (Fig. 2). A rounded median ventral keel is present, and the lateral sides of the vertebral body are moderately concave. The vertebral body is waisted in the middle. The cranial vertebral body height (CrH) and width (CrW), the area of the cranial surface (CrA), and the equivalent measurements (CaH, CaW, and CaA) on the caudal surface, along with the dorsal vertebral body length (DL), are reported in Table 3. The caudal half of the dorsal elements is relatively well preserved. The lamina is transversely cracked at the mid-level (Fig. 2). The break runs transversely through the most cranial level of the inferior vertebral notch. Despite the caudal portion of the lamina (below the break line) being pushed in the ventral direction, it has not suffered from plastic deformation. The cranial part of the base of the spinous process is just barely preserved. The cranial border of the spinous process is entirely eroded, and the caudal and dorsal borders

31

are preserved. In dorsal view, the caudal base of the spinous process occupies the notch between the postzygapophyses. A similar feature was observed in the previously described caudal thoracic and lumbar vertebral series (KNM-BG 35250; Nakatsukasa et al., 2007). Despite the cracked and missing cranial part of the spinous process, its caudal tip is well preserved. It may originally have been tear-drop shaped (Fig. 2). The spinous process tip of N. kerioi resembles those of both extant great apes and small apes (Fig. 3). A tear-drop shaped spinous process tip is also seen in Old World monkeys. However, these tips almost always have a very small process in the middle of the spinous process tip caudally (Fig. 3) that is not seen in most of the extant hominoids. 3.1.2. KNM-BG 42763C This specimen is a nearly complete lumbar vertebra that has suffered from craniocaudal compression (Fig. 2). This compression resulted in exaggerated bevelling of the cranial articular surface hanging over the ventral surface. Thus, the information that can be derived from the morphology of the vertebral body is limited. The dorsal elements have shifted ventrally. Nonetheless, the ventral keel is present, and the preand post-zygapophyses and the base of the spinous process retain some of the original characters.

Figure 3. Extracted spinous process tips in extant primate samples (taken from diaphragmatic vertebrae in extant apes, and second post-diaphragmatic vertebrae in monkeys) used in the present study, with the dorsal view of KNM-BG 42810B. The dashed white circle highlights the spinous process tip of the KNM-BG 42810B specimen. White arrows highlight the small process found caudally in the middle of the spinous process tips in Old World monkeys.

32

Y. Kikuchi et al. / Journal of Human Evolution 88 (2015) 25e42

On the prezygapophyses, a thick, rounded, moderately long metapophysis projects dorsolaterally from the lateral side of the right articular facet (Fig. 3). The extant great apes have thicker, more rounded, and more caudally positioned metapophyses, while the small apes and monkeys have cranially oriented and thinner metapophyses on the cranial side of the articular facet (Sanders, 1995; Fig. 4). The condition seen in N. kerioi is intermediate between small apes and monkeys, and extant great apes. Even though the metopophyses are longer in P. nyanzae (KNM-MW 13142K), they are similar in overall shape to those of N. kerioi (Fig. 4). The existence of anapophyses is not clear due to deformation. The transverse processes on both sides have been eroded. These bases are thick (cranial view in Fig. 2). The caudal ends of the process bases lie above the inferior vertebral notch, and the cranial ends lie at the most dorsal part of the body (cranial and right-hand views in Fig. 2). The thick spinous process originating from the pedicular base suggests that this specimen is a penultimate or ultimate lumbar vertebra as seen in cercopithecoids (Sanders, 1995). The postzygapophyseal articular surface is craniocaudally long (Fig. 2). The postzygapophyses project prominently in the caudal direction, but this condition may be exaggerated because of a caudal-ward slip of the lamina during fossilization.

The spinous process is eroded and lost except for a few millimeters near its origin on the lamina (Fig. 2). The caudal base of the spinous process is positioned remarkably caudally between the postzygapophyses. This morphology is similar to that of Gorilla and Pongo, accompanied with a strongly caudally directed spinous process (Nakatsukasa et al., 2007). 3.2. Vertebral level estimation for N. kerioi KNM-BG 42810B The results of LSR analyses, between the minimum CrA of the diaphragmatic vertebra on one hand, and the maximum CrA of the diaphragmatic, the maximum CrA of first post-diaphragmatic, or the maximum CrA of second post-diaphragmatic vertebrae on the other hand, are shown in Table 4 and Figure 5. All LSR lines indicated high regression coefficients (r > 0.98, p < 0.05) and isometry. When KNM-BG 42810B is assumed to be the maximum CrA of the diaphragmatic vertebra, N. kerioi is far from the LSR line of Old World monkeys (Fig. 5). The residual is quite large and this fossil is out of the maximum range of Old World monkeys. This means that KNM-BG 42810B is not the same level as KNM-BG 35250BO and should not be considered to be a diaphragmatic vertebra. Similarly, in the case where KNM-BG 42810B is assumed to be the maximum CrA of the first post-diaphragmatic vertebra,

Figure 4. Extracted metapophyses (left and right sides) in Nacholapithecus kerioi KNM-BG 42763C and Proconsul nyanzae KNM-MW 13142K, and right sides in extant primate samples with the superior articular process evident in ultimate lumbar vertebrae (dorsal views).

Y. Kikuchi et al. / Journal of Human Evolution 88 (2015) 25e42

the residual of N. kerioi is the largest, and this fossil is out of the maximum range of Old World monkeys (Fig. 5). In the case where KNM-BG 42810B is assumed to be the maximum CrA of the second post-diaphragmatic vertebra, the residual of N. kerioi is within the range of Old World monkeys. It is therefore assumed that KNM-BG 42810B is a post-diaphragmatic vertebra and most likely a second post-diaphragmatic vertebra given the CrA size comparison. 3.3. Spinous process restoration for N. kerioi KNM-BG 42810B The results of the LSR analyses between BM and INW or DPP are shown in Table 5 and Figure 6. Regression coefficients of LSR lines are highest in INW of New World monkeys (0.87) and lowest in DPP of extant great apes (0.39). Some of the regression coefficients in LSRs are low and the measurement errors (CV) in DPP are over 5.0%. However, since the LSRs are statistically highly significant (p < 0.005) and there are no other means of estimating the INW and DPP of N. kerioi, we thus used these models for estimation, albeit with caution. Using each LSR formula, individual predicted values for N. kerioi (BM of 17 and 27 kg, respectively) were calculated and estimated for the INW and DPP (Table 5). Predicted values of INW for N. kerioi ranged from 6.6 to 8.7 mm (average 7.6 mm), and predicted values of DPP ranged from 2.7 to 4.8 mm (average 3.7 mm), using each LSR formula of extant great apes and Old World monkeys. Since the predicted values of DPP for N. kerioi using the LSR formula of New World monkeys are quite large (Table 5), they were not suitable for spinous process restoration (when using these values for restoration in N. kerioi, the spinous process position in relation to the vertebral body is too caudal and unlike any extant species). The ventral angle of the postzygapophysis in relation to the dorsal vertebral body surface (VAP) ranged from 8.8
to 10.0
both for the diaphragmatic vertebrae of extant great apes and Old World monkeys, and for the first and second post-diaphragmatic vertebrae of Old World monkeys (Fig. 7, total average of 9.6
). The spinous process for N. kerioi was restored in a right-hand view so that INW was held at 7.6 mm, DPP at 3.7 mm, and VAP at 9.6
. This restoration is adequate because the arch of the cranial border in the inferior vertebral notch is quite smooth (Fig. 2). After this restoration, the spinous process is still caudally projecting (Figs. 2 and 8). The morphology of the spinous process in N. kerioi is similar to that of extant great apes, especially Pan and Gorilla (Figs. 2 and 8). 3.4. Metric comparison between extant primate species and N. kerioi KNM-BG 42810B The Ln transformed CrA value in N. kerioi is smaller than those for the first and second post-diaphragmatic vertebrae in largebodied Old World monkeys (Mandrillus, Papio, Semnopithecus, Theropithecus, and Nasalis, where BM > ca. 10 kg in both sexes; Smith and Jungers, 1968), and within the range of Symphalangus (Fig. 9), which has mean body masses of 11.9 kg (male) and 10.7 kg (female) (Smith and Jungers, 1968). Since the BM of N. kerioi was

33

estimated to be approximately 22 kg (Rose et al., 1996; Ishida et al., 2004; Nakatsukasa et al., 2007; Kikuchi et al., 2012), the CrA of N. kerioi is quite small. The scatter plot diagrams of the relationships between BM and CrA show that N. kerioi falls below largebodied Old World monkeys (Fig. 9). The spinous process angle of N. kerioi after restoration is similar to extant apes and greater than those of Old World monkeys and most New World monkeys (Fig. 10). Scatter plots of the relationships between BM and the nine linear measurements are shown in Figure 11. Nacholapithecus kerioi is positioned near the large-bodied Old World monkeys in VL, DL, PL, CrW, CaW, and SL. However, N. kerioi is below the range of the large-bodied Old World monkeys in CrH and CaH, indicating that the cranial and caudal heights of the vertebral body are smaller than those of large-bodied Old World monkeys of equivalent BM. When the second post-diaphragmatic vertebrae in monkeys is included, the upper limit of SH in large-bodied Old World monkeys overlaps with the value for N. kerioi, even though it is relatively higher in N. kerioi than in large-bodied Old World monkeys of equivalent BM. Compared with N. kerioi, P. nyanzae, with an equivalent BM, has a relatively long DL, PL, and SL, although SH is quite low in comparison with extant primate species. The results of the PCA of overall caudal thoracic shape are shown in Table 6 and Figure 12. In the PCA, including the diaphragmatic vertebrae of extant apes and the first post-diaphragmatic vertebrae of monkeys, PC1 accounted for 66% of the variance. This shows that the extant great apes, small apes, and large-bodied Old World monkeys differ from smaller-bodied Old World monkeys and New World monkeys despite low loadings of the vertebral body variable. Principal component 2 accounted for 15% of the variance, with variable loadings that were relatively high and >0.5 for variables, attributed to spinous process height and length. It is indicated by PC2 that many extant great apes and New World monkeys generally, but not universally, have smaller PC2 values than those of small apes and Old World monkeys. This means that extant great apes and New World monkeys have large spinous processes (in both SH and SL) in relation to small apes and Old World monkeys. Nacholapithecus kerioi was positioned among Pan, Old World monkeys, and New World monkeys. The size of the spinous process in N. kerioi is relatively large, and it is positioned just below the minimum range of Old World monkeys on PC2. Principal component 3, despite only accounting for 9% of the variance, exhibited high variable loadings on Ln (SH/GM) and Ln (SL/GM), which are related to the height and length of the spinous process. Although taxonomic group (extant great apes, small apes, Old World monkeys, and New World monkeys) was not differentiated by this axis, N. kerioi has a relatively craniocaudally tall and dorsoventrally short spinous process, similar to some species of extant great apes, small apes, and New World monkeys. In the PCA that included the diaphragmatic vertebrae of extant apes and the second post-diaphragmatic vertebrae of monkeys, PC1 accounted for 65% of the variance. This PC divided extant great apes, small apes, and large-bodied Old World monkeys from smaller-bodied Old World monkeys and New World monkeys,

Table 4 Regression coefficients (r), and slopes and intercepts with 95% confidence intervals (CI) of the least square regression (LSR) between the minimum CrA of the diaphragmatic vertebra on one hand, and the maximum CrA of the diaphragmatic, the maximum CrA of first post-diaphragmatic, or the maximum CrA of second post-diaphragmatic vertebrae on the other hand in Old World monkeys (OWM).a

Ln (Min CrA of Dia) vs. Ln (Max CrA of Dia) Ln (Min CrA of Dia) vs. Ln (Max CrA of 1st Post) Ln (Min CrA of Dia) vs. Ln (Max CrA of 2nd Post)

r

Slope

95% CI

Intercept

95% CI

0.98 0.98 0.99

1.00 0.94 0.94

0.82e1.18 0.77e1.12 0.83e1.05

0.20 0.34 0.72

0.66 to 1.07 0.23 to 1.46 0.18e1.25

a Abbreviations: CrA, cranial articular surface area of the vertebral body. Dia, diaphragmatic vertebrae. 1st Post, first post-diaphragmatic vertebrae. 2nd Post, second post-diaphragmatic vertebrae. All p-values for regressions were below 0.05. Data for LSR were natural log transformed.

34

Y. Kikuchi et al. / Journal of Human Evolution 88 (2015) 25e42

of the variance, with variable loadings being relatively high and >0.7 for variables associated with spinous process length. Principal component 2 differentiated extant great apes and New World monkeys from small apes and Old World monkeys somewhat but not consistently. Nacholapithecus kerioi fell among extant great apes, Old World monkeys, and New World monkeys, but closer to Old World monkeys. Proconsul nyanzae is positioned between extant great apes, and small apes and Old World monkeys. However, definitive differences between N. kerioi and P. nyanzae could not be detected in PC1 or PC2. Principal component 3, despite accounting for only 10% of the variance, exhibited relatively high variable loadings on Ln (SH/GM) and Ln (SL/GM), which were related to the height and length of the spinous process. This is similar to the PCA including the diaphragmatic vertebrae of extant apes and the first post-diaphragmatic vertebrae of monkeys. Although taxonomic group (extant great apes, small apes, Old World monkeys, and New World monkeys) was not differentiated by this axis, N. kerioi has a relatively craniocaudally tall and dorsoventrally short spinous process, similar to some species of extant great apes, small apes, and New World monkeys. Proconsul nyanzae has a quite distinctive spinous process, out of the range of extant primates, which may be due to its incomplete preservation (Figs. 2 and 8; Ward, 1991). 4. Discussion Figure 5. Scatter plots showing LSR for Old World monkeys between the minimum (min) cranial articular surface of vertebrae (CrA) of the diaphragmatic vertebrae on one hand, and the maximum (max) CrA of the diaphragmatic, the maximum (max) CrA of first (1st) post-diaphragmatic, or the maximum (max) CrA of second (2nd) postdiaphragmatic vertebrae on the other hand. The data were natural log (Ln) transformed. Diagram of residuals from the LSR line beside each LSR diagram. Abbreviations: Dia, diaphragmatic vertebrae; Post, post-diaphragmatic vertebrae. Point abbreviations: N, Nacholapithecus kerioi (KNM-BG-42810B); C, Cercopithecus mitis; c, Chlorocebus aethiops and Chlorocebus pygerythrus; L, Lophocebus albigena; P, Papio hamadryas; T, Theropithecus gelada; G, Colobus guereza.

resembling the PCA that included the diaphragmatic vertebrae of extant apes and the first post-diaphragmatic vertebrae in monkeys. Larger PC1 values were associated with the articular size of the vertebral body, and smaller PC1 values with the height of the vertebral body (Table 6). Principal component 2 accounted for 18%

The two newly-described vertebral specimens, KNM-BG 42810B and KNM-BG 42763C, analyzed here provide important information about the anatomy of the post-diaphragmatic and lumbar spine of N. kerioi. The caudal thoracic vertebral specimen, KNM-BG 42810B, is most likely to be a second post-diaphragmatic vertebra as judged by the cranial articular size of the vertebra in comparison with the size variation in Old World monkeys, as well as when compared to the size of KNM-BG 35250BO, which is certainly a diaphragmatic vertebra (Nakatsukasa et al., 2007). In extant hominoids, the diaphragmatic vertebra (which is usually the last ribbearing vertebra) is positioned caudally (Washburn and BuettnerJanusch, 1952; Erikson, 1963; Shapiro, 1993; Haeusler et al., 2002; Williams, 2011, 2012b). This is in contrast to the primitive condition evident in non-hominoid primates, which have level differences between the diaphragmatic vertebra and the last rib-bearing

Table 5 Regression coefficients (r), and slopes and intercepts with 95% confidence intervals (CI) of the least square regression (LSR) of the measurements (INW and DPP) on body mass (BM) in extant great apes (GA), Old World monkeys (OWM), and New World monkeys (NWM).a Vertebral level

Group

r

Slope

95% CI

Intercept

GA OWM NWM OWM NWM OWM NWM

0.71 0.79 0.77 0.77 0.87 0.84 0.87

0.25 0.28 0.35 0.25 0.38 0.26 0.40

0.19e0.32 0.21e0.35 0.23e0.47 0.19e0.32 0.29e0.47 0.21e0.32 0.30e0.51

1.24 1.10 0.89 1.27 0.94 1.29 1.03

GA OWM NWM OWM NWM OWM NWM

0.39 0.46 0.83 0.45 0.85 0.53 0.81

0.34 0.51 0.92 0.75 1.18 0.59 1.14

0.13e0.56 0.21e0.81 0.66e1.18 0.29e1.20 0.88e1.49 0.30e0.89 0.76e1.53

0.03 0.24 0.54 0.91 1.05 0.38 0.87

95% CI

Predicted value (mm) 17 kg

INW

Dia

INW

1st post-dia

INW

2nd post-dia

DPP

Dia

DPP

1st post-dia

DPP

2nd post-dia

0.96e1.52 0.95e1.25 0.69e1.09 1.13e1.42 0.79e1.09 1.18e1.41 0.86e1.20 0.88 0.91 0.96 1.93 1.54 1.03 1.47

to to to to to to to

0.95 0.42 0.11 0.11 0.55 0.26 0.27

27 kg

7.1 6.6 6.6 7.3 7.6 7.7 8.8

7.9 7.6 7.7 8.2 9.1 8.7 10.6

2.7 3.3 7.9 3.3 10.1 3.7 10.7

3.2 4.2 12.1 4.7 17.4 4.8 18.1

a Abbreviations: INW, inferior vertebral notch width at the level of the caudal vertebral body surface. DPP, degree of projection of the postzygapophysis in relation to the level of the caudal vertebral body surface. Dia, diaphragmatic vertebrae.1st post-dia, first post-diaphragmatic vertebrae. 2nd post-dia, second post-diaphragmatic vertebrae. All p-values for regressions were below 0.005. Data for LSR were natural log transformed. Predicted values in mm were calculated with body masses of both 17 and 27 kg for N. kerioi for each model.

Y. Kikuchi et al. / Journal of Human Evolution 88 (2015) 25e42

35

Figure 6. Scatter plots showing LSR between body mass (BM), and inferior vertebral notch width at the level of the caudal vertebral body surface (INW) or degree of projection of the postzygapophysis in relation to the level of the caudal vertebral body surface (DPP) in the diaphragmatic (extant apes, Old World monkeys, New World monkeys), first (1st) post-diaphragmatic (Old World monkeys, New World monkeys), and second (2nd) post-diaphragmatic vertebrae (Old World monkeys, New World monkeys). The data were natural log (Ln) transformed and LSRs were performed on samples of extant great apes, Old World monkeys, and New World monkeys. Marks: red line, LSR for GA (extant great apes); black line, LSR for OWM (Old World monkeys); blue line, LSR for NWM (New World monkeys); red, Pan paniscus; yellow, Pan troglodytes; green, Gorilla; aqua, Pongo; purple, SA (small apes); black, OWM (Old World monkeys); blue, NWM (New World monkeys). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Figure 7. Box plot of ventral angle in degrees of the postzygapophysis in relation to the dorsal vertebral body line (VAP). Abbreviations: Dia, diaphragmatic vertebrae; Postdia, post-diaphragmatic vertebrae; Averages of each taxonomic group are indicated by dots. The bars in the boxes indicate the median values.

vertebra, and usually two or three post-diaphragmatic vertebrae with rib-bearing characters (Erikson, 1963; Clauser, 1975, 1980; Kelley, 1986; Sanders, 1995; Williams, 2011, 2012b). Since the post-diaphragmatic thoracic vertebrae and lumbar vertebrae are similarly shaped and have similarly directed zygapophyseal articular facets, possessing post-diaphragmatic vertebrae results in the functional equivalent of a long lumbar vertebral column. This facilitates a longer stride length during quadrupedal walking and running and/or increases leaping distance, made possible by increasing the distance through which force is applied during acceleration for take-off (Jenkins, 1974; Preuschoft et al., 1979; Jungers, 1984; Shapiro, 1993; Ward, 1993; Sanders and Bodenbender, 1994; Johnson and Shapiro, 1998). On the other hand, the decrease in functional lumbar length in extant apes and humans (Washburn and Buettner-Janusch, 1952; Shapiro, 1993; Williams, 2011, 2012b) facilitates enhanced lumbar stability and is an adaptation for climbing and suspension (Erikson, 1963; Cartmill and Milton, 1977; Haeusler et al., 2002). As N. kerioi possesses at least two post-diaphragmatic thoracic vertebrae, it is unlikely to have employed a very similar positional behavior as extant hominoids and was likely to be engaged in monkey-like arboreal quadrupedalism (Rose et al., 1996; Nakatsukasa et al., 2003; Ishida et al., 2004). The cranial articular surface area of the post-diaphragmatic N. kerioi vertebra, KNM-BG 42810B, is comparable to that of Symphalangus, which has a surface area that is smaller than that of large-bodied Old World monkeys. However, the peculiarly narrow vertebral morphology in KNM-BG 42810B may be due to a slight

36

Y. Kikuchi et al. / Journal of Human Evolution 88 (2015) 25e42

Figure 8. Right-hand views of Nacholapithecus kerioi (KNM-BG 42810B), P. nyanzae (KNM-MW 13142H), diaphragmatic vertebrae in extant hominoids, second post-diaphragmatic vertebrae in Old World monkeys and New World monkeys, The spinous process of Nacholapithecus kerioi was restored. Scale bars ¼ 1 cm.

dorsoventral compression deformation, because the cranial and caudal vertebral body heights are very small compared with those of extant primate taxa of an equivalent BM (Fig. 11). Accordingly, it is probable that the vertebral body in N. kerioi is at least as large as in Symphalangus and may be quite close in size to that of largebodied Old World monkeys. The narrow vertebral body of N. kerioi is thought to be an extant non-hominoid trait; that is, a primitive rather than a derived condition (Schultz, 1953; Benton, 1967; Rose, 1975; Ward, 1991; Shapiro, 1993). The rounded median keel, which is present ventrally on the KNM-BG 42810B post-diaphragmatic thoracic vertebrae, accords with the clear ventral keel observed, despite mediolateral compressive deformation, in another N. kerioi post-diaphragmatic vertebra, KNM-BG 35250BP (Nakatsukasa et al., 2007). A ventral keel is also present on ultimate/penultimate lumbar vertebra KNMBG 42763C (this study) and in the lumbar vertebral series of KNMBG 35250 (Nakatsukasa et al., 2007). However, based on KNM-BG 35250BO, there is no median ventral keel on the N. kerioi diaphragmatic vertebra (Nakatsukasa et al., 2007). Given this pattern of presence/absence in different parts of the vertebral column, it is difficult to assess whether N. kerioi has greater affinity to cercopithecoids (that have ventral keels) or to hominoids (that lack true ventral keels). Although ‘mosaic’ patterns of vertebral body shape are not uncommon in other Miocene apes (Williams and Russo, 2015), future studies could usefully examine the functional and/or evolutionary reasons for the inconsistency in ventral keeling in N. kerioi. Based on observation of the morphology of the spinous process base, Nakatsukasa et al. (2007) suggested that N. kerioi had caudal thoracic and lumbar vertebrae with more caudally positioned spinous process bases between the postzygapophyses than did P. nyanzae. In other words, the postzygapophyses of N. kerioi are positioned cranially and do not project below the caudal border of the spinous process as they do in extant great apes. In contrast, the caudally projecting postzygapophyses are seen in small apes and monkeys. This condition presumably allows sagittal flexibility between the adjacent vertebral bodies because these postzygapophyses create a wide intervertebral space. The cranially positioned postzygapophyses in N. kerioi reinforce the stability of the caudal thoracic region, indicating that this fossil hominoid exhibited some similarities in positional behavior to those seen in the extant

great apes. The spinous process bases of the N. kerioi postdiaphragmatic and lumbar vertebra in the present study are also caudally originated, supporting Nakatsukasa et al.'s (2007) inference, and suggesting very strongly that the difference between N. kerioi and P. nyanzae is sufficiently large to be considered an inter-taxon difference rather than the product of intraspecific variation. Although cercopithecoids have cranially angled spinous processes at the anticlinal element (around T10e11), extant hominoids lack the anticlinal vertebra and these spinous processes of the caudal thoracic and lumbar vertebrae are angled caudally (Sanders, 1995). Slijper (1946) proposed that if the multiple muscles attach on a spinous process, the strongest muscle acting on it affects the spinous process angle in relation to the vertebral body, making it perpendicular to this muscle fiber to obtain the greatest possible mechanical advantage. The more caudally oriented spinous processes in extant apes indicates that the strongest and most important muscles in the lower back of hominoids are the multifidus muscles (Slijper, 1946; Ward, 1991). Slijper (1946) also suggests that these muscles contribute to the maintenance of upright posture in these animals. Thus, the more caudally directed spinous processes of the N. kerioi post-diaphragmatic vertebrae described in the present study suggest a more enhanced antipronogrady. Craniocaudally expanded and short spinous processes in caudal thoracic and lumbar vertebrae can passively limit axial extension by providing a bony stop between adjacent vertebrae (Granatosky et al., 2014), and this morphology also limits the intervertebral space and overall trunk extension (Shapiro, 1993, 1995, 2007; Sargis, 2001; Shapiro and Simons, 2002; Shapiro et al., 2005). Moreover, as the craniocaudal height of the spinous process is tall, the interspinous distance available for the interspinous ligaments or muscles presumably decreases and the rigidity of the lumbar region increases (Shapiro, 1993). In contrast, craniocaudally narrowed spinous processes are associated with an increase in the sagittal flexibility of the lumbar region because of the expansion of the ligaments or muscles (Shapiro, 1993). The craniocaudally wide spinous processes of the post-diaphragmatic vertebra of N. kerioi found in the present study may contribute to a decreased dorsomobility in the functional lumbar region, which was strongly confirmed by the ratio of the spinous process basal height (SH) to the dorsal length of the vertebral body (DL), as indicated in Figure 13. In these diagrams, N. kerioi is positioned at the upper

Y. Kikuchi et al. / Journal of Human Evolution 88 (2015) 25e42

Figure 9. Box plot (a) of the cranial articular surface of the vertebrae (CrA) in both sexes in extant hominoids and males in monkeys. The data were natural log (Ln) transformed. Dotted red line represents the value of Nacholapithecus kerioi (KNM-BG 42810B). The scatter plots are of body mass (BM) and CrA in both sexes in extant hominoids and monkeys. BM of Nacholapithecus kerioi is assumed to be 17e27 kg. (b), the scatter plots for the diaphragmatic (extant apes) and first (1st) post-diaphragmatic (Old World monkeys, New World monkeys). (c), the scatter plots for the diaphragmatic (extant apes) and second (2nd) post-diaphragmatic vertebrae (Old World monkeys, New World monkeys). These data also were natural log (ln) transformed. Abbreviations: Dia, diaphragmatic vertebrae; Post-dia, post-diaphragmatic vertebrae. Point abbreviations: N, Nacholapithecus kerioi (KNM-BG 42810B); P, P. nyanzae (KNM-MW 13142H); red, Pan paniscus; yellow, Pan troglodytes; green, Gorilla; aqua, Pongo; purple, SA (small apes); black, smaller-bodied OWM (Old World monkeys); gray, large-bodied OWM (Old World monkeys); blue, NWM (New World monkeys). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

37

Figure 10. Box plot of spinous process angle (SPA) in degrees. Dotted red line represents the value of N. kerioi (KNM-BG-42810B). Abbreviations: Dia, diaphragmatic vertebrae; Post-dia, post-diaphragmatic vertebrae. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

38

Y. Kikuchi et al. / Journal of Human Evolution 88 (2015) 25e42

Figure 11. Scatter plot of body mass (x axis) and the linear measurements (y axis). The data were natural log (Ln) transformed. Body mass of Nacholapithecus kerioi is assumed to be 17e27 kg. Abbreviations for measurements; BM, body mass; VL, ventral body length of the vertebra: DL, dorsal body length of the vertebra: PL, right pedicular basal length: CrH, cranial body height of the vertebra: CrW, cranial body width of the vertebra: CaH, caudal body height of the vertebra: CaW, caudal body width of the vertebra: SH, spinous process basal height craniocaudally: SL, spinous process length nearly in the midline dorsoventrally. Marks are the same as in Figure 9.

Y. Kikuchi et al. / Journal of Human Evolution 88 (2015) 25e42 Table 6 Results of the Principal Components Analysis (PCA) summarizing diaphragmatic (apes), and first and second post-diaphragmatic (monkeys) vertebral shape in extant and fossil primate samples.a 1st % of Variance Cumulative % Variable loadings Ln (VL/GM) Ln (DL/GM) Ln (PL/GM) Ln (CrH/GM) Ln (CrW/GM) Ln (CaH/GM) Ln (CaW/GM) Ln (SH/GM) Ln (SL/GM) 2nd % of Variance Cumulative % Variable loadings Ln (DL/GM) Ln (PL/GM) Ln (CrH/GM) Ln (CrW/GM) Ln (CaW/GM) Ln (SH/GM) Ln (SL/GM)

PC1

PC2

PC3

65.520 65.520

14.952 80.472

9.130 89.602

0.372 0.353 0.370 0.358 0.363 0.380 0.350 0.213 0.164

0.165 0.367 0.225 0.260 0.052 0.167 0.261 ¡0.502 ¡0.608

0.176 0.190 0.018 0.120 0.138 0.136 0.058 0.681 ¡0.643

PC1

PC2

PC3

65.335 65.335

17.982 83.317

9.555 92.872

¡0.408 ¡0.419 0.402 0.428 0.416 0.314 0.203

0.315 0.295 0.292 0.192 0.242 0.319 ¡0.729

0.283 0.232 0.094 0.129 0.086 0.788 ¡0.461

a Only the three first axes (percentages of cumulative proportion containing 90% and 93% of total vertebral shape variation, respectively) provided a meaningful discrimination. Each original measurement was size-adjusted by the geometric mean (GM). In the case of second post-diaphragmatic vertebrae in monkeys, only variables for which measurements could be obtained from P. nyanzae, are listed. Variables with absolute loading >0.4 are marked in bold.

range of extant great apes, while, by contrast, P. nyanzae exhibits a very low value. This means that, similar to extant great apes, N. kerioi had a craniocaudally expanded spinous process in relation to vertebral body length, and this contributed to the stability of the caudal thoracic region. This condition is quite unlikely to have been present in P. nyanzae. In conjunction with the wide craniocaudal spinous process, the spinous process tip of the N. kerioi post-diaphragmatic vertebra is tear-drop shaped and resembles some of the extant great apes and small apes as indicated in the present study. This morphology may contribute to vertebral stability in orthograde extant hominoids by ‘stacking’ or closely packing the adjacent spinous process (Fig. 14). This is because the tear-drop shape of the caudal part of the spinous process fits exactly with the caudally adjacent cranial part of the process. In some Old World monkeys, tear-drop shaped spinous process tips are also present. However, in these tips, there is a very small, caudally projecting process in the middle of the spinous process tip. This character is absent in most of the extant hominoids and may contribute in cercopithecoids to the attachment of interspinous ligaments or muscles. Of course, although the interspinous ligaments or muscles do exist in extant hominoids, the dorsal extension of the caudal thoracic and lumbar vertebrae in the orthograde position (upright posture, as seen in extant apes) presumably influences spinous process tip morphology, such as a teardrop shape, by ‘stacking’ the adjacent spinous process. A tear-drop shaped spinous process tip in N. kerioi may also relate to the antipronograde activity common in extant apes. The extant great apes have caudally positioned thick and round metapophyses on the lumbar vertebrae, whereas cercopithecoids have more cranially oriented and thinner metapophyses. The morphology of the metapophysis of N. kerioi lumbar vertebra

39

(KNM-BG 42763C) is intermediate between these two. The transversospinal muscles originate on the lumbar metapophyses. Lateral flaring of the metapophyses alters the line of action of the lumbar transversospinal muscles so that they are more effective for resisting lateral flexion or rotation of the lower spine in extant great apes (Sanders, 1995). The transversospinal muscles are also important in controlling the movements of individual vertebrae in the sagittal plane against ventrally flexing torque (Shapiro and Jungers, 1988; Sanders, 1995). Lumbar transversospinal musculature has a proportionately greater mass relative to the erector spinae muscles in extant great apes (and small apes) than in extant non-hominoid primates (Benton, 1974; Sanders, 1995), suggesting a relatively greater emphasis on the functions of those muscles in extant great apes (and small apes). The metapophyses in P. nyanzae are similar in shape to those of N. kerioi, although they are longer. Length difference is putatively size-related, although specimens of the smaller P. heseloni are not available to be studied. A comparison between N. kerioi and the contemporaneous E. africanus (Ward et al., 1999) would be interesting. Unfortunately, the post-diaphragmatic vertebra (KNM-TH 28860Y) of E. africanus is not well preserved because this specimen was subject principally to bilateral compression; however, we can briefly comment on the morphology of the spinous process for which the basal part is present (Fig. 1 of Sherwood et al., 2002). This spinous process base is craniocaudally wide and resembles the condition in N. kerioi; it is unlike the narrow process in P. nyanzae. This may raise a possibility that at 15e16 Ma, apes with think dental enamel experienced a common positional behavioral change from a Proconsul-like precursor, a change that would be secondarily diversified in positional behaviors such as suspensory locomotion or knuckle walking later on. An Equatorius-like type might have been ancestral or it might have been derived from a N. kerioi-like morphology. However, in order to substantiate such a possibility, it is necessary to have better represented materials that provide a clearer picture of similarity/ diversity of the lumbar anatomy in 15e16 Ma African/western Eurasian apes. 5. Conclusion This paper reported a new caudal thoracic and a new lumbar vertebra of N. kerioi. In the caudal thoracic vertebral specimen, a rounded median ventral keel is present and lateral sides of its vertebra are moderately concave. In contrast, an obvious medial ventral keel is seen in the vertebral body of lumbar specimen. The present study revealed that N. kerioi has at least two postdiaphragmatic vertebrae. The lumber vertebral specimen has thick, rounded, and moderately long metapophyses, projecting dorsolaterally. The spinous process bases of both N. kerioi specimens originate caudally between the postzygapophyses. In other words, the postzygapophyses in N. kerioi do not project below the caudal border of spinous processes. This condition is similar to that in extant great apes and in contrast to small apes and monkeys, and was described previously in the N. kerioi holotype specimens. Combined with a craniocaudally expanded spinous process in relation to the vertebral body length of N. kerioi, the spinous process morphology of N. kerioi differs from those of P. nyanzae. A caudally directed spinous process and tear-drop shaped tip of its process is seen in the caudal thoracic vertebral specimen and these features resembles those of extant apes. The morphology of the spinous process and postzygapophyses in N. kerioi limits the intervertebral space and contributes to the stability of the functional lumbar region as seen in extant great apes. This suggests that the positional behaviors of N. kerioi include antipronograde activity.

40

Y. Kikuchi et al. / Journal of Human Evolution 88 (2015) 25e42

Figure 12. Principal Component Analysis (PCA; Table 6) summarizing overall caudal thoracic vertebral shape. The analysis was conducted on the covariance matrix of nine shape variables in first (1st) post-diaphragmatic vertebrae and seven shape variables in second (2nd) post-diaphragmatic vertebrae in monkeys with those of the diaphragmatic vertebrae in extant apes. a, PC1 and PC2 in diaphragmatic vertebrae in extant apes and first (1st) post-diaphragmatic vertebrae in monkeys. b, PC1 and PC3 in diaphragmatic vertebrae in extant apes and first (1st) post-diaphragmatic vertebrae in monkeys. c, PC1 and PC2 in diaphragmatic vertebrae in extant apes and second (2nd) post-diaphragmatic vertebrae in monkeys. d, PC1 and PC3 in diaphragmatic vertebrae in extant apes and second (2nd) post-diaphragmatic vertebrae in monkeys. Point abbreviations are the same as in Figure 9.

Figure 13. Scatter plots of dorsal length of the vertebral body (DL) and spinous process basal height craniocaudally (SH). The data were natural log (Ln) transformed. Point abbreviations are the same as in Figure 9.

Y. Kikuchi et al. / Journal of Human Evolution 88 (2015) 25e42

Figure 14. Dorsal views of the articulated last five thoracic vertebrae in an extant great ape (Pan troglodytes) and an Old World monkey (Colobus guereza).

Acknowledgments We thank Sarah Elton (Editor), an Associate Editor, and two anonymous referees for reviewing our paper, using their precious time and giving us helpful comments. We extend our gratitude to the Office of the President of the Republic of Kenya, especially, Ahmed Yassin, Stephen Rucina Mathai, Idle Farah, and Emma Mbua (now of Mount Kenya University), and the staff of the National Museums of Kenya for permission to carry out research in Kenya. We thank the curators of the Division of Osteology of the National n Museums of Kenya; Christoph Zollikofer, and Marcia Ponce de Leo €t Zürichof the Anthropological Institute and Museum, Universita Irchel; Christiane Funk of the Museum für Naturkunde Berlin; and Emmanuel Gilissen and Wim Wendelen of the Department of African Zoology at the Royal Museum for Central Africa for access to collections under their care. The Japan Society for Promotion of Science, Nairobi Research Station provided us with support to conduct research in Kenya. This study was supported by JSPS KAKENHI Grant Numbers 20247033, 24000015, and 15K14621, and by a fiscal 2014 Grant-in-Aid for Scientific Research and a Grant for Basic Science Research Projects from the Sumitomo Foundation, grant number 140080. Supplementary Online Material Supplementary online material related to this article can be found at http://dx.doi.org/10.1016/j.jhevol.2015.09.003. References Badoux, D.M., 1974. An introduction to biomechanical principles in primate locomotion and structure. In: Jenkins Jr., F.A. (Ed.), Primate Locomotion. Academic Press, New York, pp. 1e44. Benton, R.S., 1967. Morphological evidence for adaptations within the epaxial region of the primates. In: Vagtborg, H. (Ed.), The Baboon in Medical Research, Volume 2. University of Texas Press, Austin, pp. 201e216. Benton, R.S., 1974. Structural patterns in the Pongidae and Cercopithecidae. Yearb. Phys. Anthropol. 18, 65e88. Cartmill, M., Milton, K., 1977. The lorisiform wrist joint and the evolution of ‘‘brachiating’’ adaptations in the Hominoidea. Am. J. Phys. Anthropol. 47, 249e272. Clauser, D.A., 1975. The numbers of vertebrae in three African cercopithecine species. Folia Primatol. 23, 308e319. Clauser, D.A., 1980. Functional and comparative anatomy of the primate spinal column: some locomotor and postural adaptations. Ph.D. Dissertation. The University of Wisconsin-Milwaukee.

41

Delson, E., Terranova, C.J., Jungers, W.L., Sargis, E.J., Jablonski, N.G., Dechow, P.C., 2000. Body mass in Cercopithecidae (Primates, Mammalia): estimation and scaling in extinct and extant taxa. Am. Mus. Nat. Hist., Anthropol. Pap. 83, 1e159. Drake, R.E., Van Couvering, J.A., Pickford, M.H., Curtis, G.H., Harris, J.A., 1988. New chronology for the Early Miocene mammalian faunas of Kisingiri, Western Kenya. J. Geo. Soc. 145, 479e491. Erikson, G.E., 1963. Brachiation in New World monkeys and in anthropoid apes. Symp. Zool. Soc. Lond. 10, 135e164. Feibel, C.S., Brown, F.H., 1991. Age of the primate-bearing deposits on Maboko Island, Kenya. J. Hum. Evol. 21, 221e225. Gebo, D.L., MacLatchy, L., Kityo, R., Deino, A., Kingston, J., Pilbeam, D., 1997. A hominoid genus from the early Miocene of Uganda. Science 276, 401e404. Granatosky, M.C., Lemelin, P., Chester, S.G.B., Pampush, J.D., Schmitt, D., 2014. Functional and evolutionary aspects of axial stability in euarchontans and other mammals. J. Morphol. 275, 313e327. Haeusler, M., Martelli, S.A., Boeni, T., 2002. Vertebrae numbers of the early hominid lumbar spine. J. Hum. Evol. 43, 621e643. Harrison, T., 1986. A reassessment of the phylogenetic relationships of Oreopithecus bambolii Gervais. J. Hum. Evol. 15, 541e583. Harrison, T., 1999. Scaling of lumbar vertebrae in anthropoid primates: its implications for the positional behavior and phylogenetic affinities of Proconsul. Am. J. Phys. Anthropol. 28 (Suppl.), 146. Harrison, T., Rook, L., 1997. Enigmatic anthropoid or misunderstood ape? The phylogenetic status of Oreopithecus bambolii reconsidered. In: Begun, D.R., Ward, C.V., Rose, M.D. (Eds.), Function, Phylogeny and Fossils: Miocene Hominoid Evolution and Adaptation. Plenum Press, New York, pp. 327e362. Ishida, H., Kunimatsu, Y., Takano, T., Nakano, Y., Nakatsukasa, M., 2004. Nacholapithecus skeleton from the Middle Miocene of Kenya. J. Hum. Evol. 46, 67e101. Jenkins Jr., F.A., 1974. Tree shrew locomotion and the origins of primate arborealism. In: Jenkins Jr., F.A. (Ed.), Primate Locomotion. Academic Press, New York, pp. 85e115. Johnson, S.E., Shapiro, L.J., 1998. Positional behavior and vertebral morphology in atelines and cebines. Am. J. Phys. Anthropol. 105, 333e354. Jungers, W.L., 1984. Scaling of the hominoid locomotor skeleton with special reference to lesser apes. In: Preuschoft, H., Chivers, D.J., Brockelman, W.Y., Creel, N. (Eds.), The Lesser Apes: Evolutionary and Behavioural Biology. Edinburgh University Press, Edinburgh, pp. 146e169. Jungers, W.L., Falsetti, A.B., Wall, C.E., 1995. Shape, relative size, and size adjustments in morphometrics. Yearb. Phys. Anthropol. 38, 137e161. Kelley, J., 1986. Paleobiology of Miocene hominoids. Ph.D. Dissertation. Yale University. Kikuchi, Y., Nakano, Y., Nakatsukasa, M., Kunimatsu, Y., Shimizu, D., Ogihara, N., Tsujikawa, H., Takano, T., Ishida, H., 2012. Functional morphology and anatomy of cervical vertebrae in Nacholapithecus kerioi, a middle Miocene hominoid from Kenya. J. Hum. Evol. 62, 677e695. €hler, M., Moya
-Sol
Ko a, S., 1997. Ape-like or hominid-like? The positional behavior of Oreopithecus bambolii reconsidered. Proc. Natl. Acad. Sci. 94, 11747e11750. €hler, M., Moya
-Sola
, S., Andrews, P., 1999. Order Primates. In: Ro €ssner, G., Ko Heissig, K. (Eds.), The Miocene Land Mammals of Europe. Dr Friedrich Pfeil, München, pp. 91e104. Larson, S.G., 1988. Subscapularis function in gibbons and chimpanzees: implications for interpretation of humeral head torsion in hominoids. Am. J. Phys. Anthropol. 76, 449e462. Lovejoy, C.O., Suwa, G., Simpson, S.W., Matternes, J.H., White, T.D., 2009. The great divides: Ardipithecus ramidus reveals the postcrania of our last common ancestors with African apes. Science 326, 100e106. MacLatchy, L., 2004. The oldest ape. Evol. Anthrop. 13, 90e103. MacLatchy, L., Gebo, D., Kityo, R., Pilbeam, D., 2000. Postcranial functional morphology of Morotopithecus bishopi, with implications for the evolution of modern ape locomotion. J. Hum. Evol. 39, 159e183. McCollum, M.A., Rosenman, B.A., Suwa, G., Meindl, R.S., Lovejoy, C.O., 2010. The vertebral formula of the last common ancestor of African apes and humans. J. Exp. Zool. B. Mol. Dev. Evol. 314, 123e134. Mosimann, J.E., 1970. Size allometry: size and shape variables with characterizations of the lognormal and generalized gamma distributions. J. Am. Stat. Assoc. 65, 930e945.
-Sol
€hler, M., 1995. New partial cranium of Dryopithecus lartet, 1863 Moya a, S., Ko (Hominoidea, Primates) from the upper Miocene of Can Llobateres, Barcelona, Spain. J. Hum. Evol. 29, 101e139.
-Sol
€hler, M., 1996. A Dryopithecus skeleton and the origins of great-ape Moya a, S., Ko locomotion. Nature 379, 156e159.
-Sol
€hler, M., Alba, D.M., Casanovas-Vilar, I., Galindo, J., 2004. PierMoya a, S., Ko olapithecus catalaunicus, a new Middle Miocene great ape from Spain. Science 306, 1339e1344.
-Sol
cija, S., Casanovas-Vilar, I., Ko €hler, M., De EstebanMoya a, S., Alba, D.M., Alme Trivigno, S., Robles, J.M., Galindo, J., Fortuny, J., 2009. A unique Middle Miocene European hominoid and the origins of the great ape and human clade. Pro. Natl. Acad. Sci. 106, 9601e9606. Nakatsukasa, M., 2008. Comparative study of Moroto vertebral specimens. J. Hum. Evol. 55, 581e588. Nakatsukasa, M., Kunimatsu, Y., 2009. Nacholapithecus and its importance for understanding hominoid evolution. Evol. Anthropol. Issues, News, and Reviews 18, 103e119.

42

Y. Kikuchi et al. / Journal of Human Evolution 88 (2015) 25e42

Nakatsukasa, M., Kunimatsu, Y., Nakano, Y., Takano, T., Ishida, H., 2003. Comparative and functional anatomy of phalanges in Nacholapithecus kerioi, a Middle Miocene hominoid from northern Kenya. Primates 44, 371e412. Nakatsukasa, M., Kunimatsu, Y., Nakano, Y., Ishida, H., 2007. Vertebral morphology of Nacholapithecus kerioi based on KNM-BG 35250. J. Hum. Evol. 52, 347e369. Pickford, M., 1998. A new genus of Tayassuidae (Mammalia) from the middle ontol. 84, 275e285. Miocene of Uganda and Kenya. Ann. Pale Pickford, M., Senut, B., Gommery, D., 1999. Sexual dimorphism in Morotopithecus bishopi, an early Middle Miocene hominoid from Uganda and a reassessment of its geological and biological contexts. In: Andrews, P., Banham, P. (Eds.), Late Cenozoic Environments and Hominid Evolution: a Tribute to Bill Bishop. Geolog. Soc., London, pp. 27e38. Pickford, M., Senut, B., Gommery, D., Musiime, E., 2003. New catarrhine fossils from Moroto II, Early Middle Miocene (ca 17.5 Ma) Uganda. Comptes Rend. Palevol. 2, 649e662. Preuschoft, H., Fritz, M., Niemitz, C., 1979. Biomechanics of the trunk in primates and problems of leaping in Tarsius. In: Morbeck, M.E., Preuschoft, H., Gomberg, N. (Eds.), Environment, Behavior, and Morphology: Dynamic Interactions in Primates. Gustav Fischer, New York, pp. 327e345. Rockwell, H.F., Evans, E., Pheasant, H.C., 1938. The comparative morphology of the vertebral spinal column; its form related to function. J. Morphol. 63, 87e117. Rose, M.D., 1975. Functional proportions of primate lumbar vertebral bodies. J. Hum. Evol. 4, 21e38. Rose, M.D., Nakano, Y., Ishida, H., 1996. Kenyapithecus postcranial specimens from Nachola, Kenya. Afr. Study Monogr. 24 (Suppl.), 3e56. Ruff, C.B., 2003. Long bone articular and diaphyseal structure in Old World monkeys and apes II: estimation of body mass. Am. J. Phys. Anthropol. 120, 16e37. Ruff, C.B., Walker, A., Teaford, M.F., 1989. Body mass, sexual dimorphism and femoral proportions of Proconsul from Rusinga and Mfangano Islands, Kenya. J. Hum. Evol. 18, 515e536. Sanders, W.J., 1995. Function, allometry and evolution of the australopithecine lower precaudal spine. Ph.D. Dissertation. New York University. Sanders, W.J., Bodenbender, B.E., 1994. Morphometric analysis of lumbar vertebra UMP 67-28: implications for spinal function and phylogeny of the Miocene Moroto hominoid. J. Hum. Evol. 26, 203e237. Sargis, E.J., 2001. A preliminary qualitative analysis of the axial skeleton of tupaiids (Mammalia, Scandentia): functional morphology and phylogenetic implications. J. Zool. Lond. 253, 473e483. Sawada, Y., Pickford, M., Itaya, T., Makinouchi, T., Kabeto, K., Ishida, S., Ishida, H., 1998. K-Ar ages of Miocene Hominoidea (Kenyapithecus and Samburupithecus) from Samburu Hills, Northern Kenya. C. R. Acad. Sci. Paris 326, 445e451. Schultz, A.H., 1938. The relative length of the regions of the spinal column in Old World primates. Am. J. Phys. Anthrop. 24, 1e22. Schultz, A.H., 1953. The relative thickness of the long bones and the vertebrae in primates. Am. J. Phys. Anthropol. 11, 277e311. Schultz, A.H., 1961. Vertebral column and thorax. In: Hofer, H., Schultz, A.H., Starck, D. (Eds.), Primatologia, Handbook of Primatology, Vol. 4, Part 5. Karger, Basel, pp. 1e66. Shapiro, L.J., 1993. Functional morphology of the vertebral column in primates. In: Gebo, D.L. (Ed.), Postcranial Adaptation in Nonhuman Primates. Northern Illinois University Press, Dekalb, pp. 121e149. Shapiro, L.J., 1995. Functional morphology of indrid lumbar vertebrae. Am. J. Phys. Anthropol. 98, 323e342. Shapiro, L.J., 2007. Morphological and functional differentiation in the lumbar spine of lorisids and galagids. Am. J. Primatol. 69, 86e102.

Shapiro, L.J., Jungers, W.L., 1988. Back muscle function during bipedal walking in chimpanzee and gibbon: Implication for the evolution of human locomotion. Am. J. Phys. Anthropol. 77, 201e212. Shapiro, L.J., Simons, C.V.M., 2002. Functional aspects of strepsirrhine lumbar vertebral bodies and spinous processes. J. Hum. Evol. 42, 753e783. Shapiro, L.J., Seiffert, C.V., Godfrey, L.R., Jungers, W.L., Simons, E.L., Randria, G.F., 2005. Morphometric analysis of lumbar vertebrae in extinct Malagasy strepsirrhines. Am. J. Phys. Anthropol. 128, 823e839. Sherwood, R.J., Ward, S., Hill, A., Duren, D.L., Brown, B., Downs, W., 2002. Preliminary description of the Equatorius africanus partial skeleton (KNM-TH 28860) from Kipsaramon, Tugen Hills, Baringo District, Kenya. J. Hum. Evol. 42, 63e73. Slijper, E., 1946. Comparative biologic-anatomical investigations on the vertebral column and spinal musculature of mammals. Verh. K. Ned. Akad. Wet. 42, 1e128. Smith, R.J., Jungers, W.L., 1968. Body mass in comparative primatology. J. Hum. Evol. 32, 523e559.
cija, S., Moya
-Sola
, S., 2010. The lumbar vertebrae of the Susanna, I., Alba, D.M., Alme Middle Miocene stem great ape Pierolapithecus catalaunicus (Primates: Hominoidae). Cidaris 30, 311e316.
cija, S., Moya
-Sol
Susanna, I., Alba, D.M., Alme a, S., 2014. The vertebral remains of the
slate Miocene great ape Hispanopithecus laietanus from Can Llobateres 2 (Valle
s Basin, NE Iberian Peninsula). J. Hum. Evol. 73, 15e34. Penede Walker, A.C., Rose, M.D., 1968. Fossil hominoid vertebrae from the Miocene of Uganda. Nature 217, 980e981. Walker, A.C., Pickford, M., 1983. New postcranial fossils of Proconsul africanus and Proconsul nyanzae. In: Ciochon, R.I., Corruccini, R.S. (Eds.), New Interpretations of Ape and Human Ancestry. Plenum Press, New York, pp. 325e351. Walker, A.C., Teaford, M.F., Martin, L., Andrews, P., 1993. A new species of Proconsul from the early Miocene of Rusinga/Mfangano Islands, Kenya. J. Hum. Evol. 25, 43e56. Ward, C.V., 1991. The functional anatomy of the lower back and pelvis of the Miocene hominoid Proconsul nyanzae from the Miocene of Mfangano Island, Kenya. Ph.D. Dissertation. Johns Hopkins University. Ward, C.V., 1993. Torso morphology and locomotion in Proconsul nyanzae. Am. J. Phys. Anthropol. 92, 291e328. Ward, C.V., 2007. Postcranial and locomotor adaptations of hominoids. In: Henke, W., Tattersall, I. (Eds.), Handbook of Paleoanthropology, Vol II: Primate Evolution and Human Origins. Springer, New York, pp. 1011e1030. Ward, C.V., Walker, A., Teaford, M.F., Odhiambo, I., 1993. Partial skeleton of Proconsul nyanzae from Mfangano Island. Am. J. Phys. Anthropol. 90, 77e111. Ward, S., Brown, B., Hill, A., Kelley, J., Downs, W., 1999. Equatorius: A new hominoid genus from the middle Miocene of Kenya. Science 285, 1382e1386. Washburn, S., Buettner-Janusch, J., 1952. The definition of thoracic and lumbar vertebrae. Am. J. Phys. Anthropol. 10, 251. Williams, S.A., 2011. Evolution of the hominoid vertebral column. Ph.D. Dissertation. University of Illinois. Williams, S.A., 2012a. Variation in anthropoid vertebral formulae: implications for homology and homoplasy in hominoid evolution. J. Exp. Zool. B. Mol. Dev. Evol. 318, 134e147. Williams, S.A., 2012b. Placement of the diaphragmatic vertebra in catarrhines: implications for the evolution of dorsostability in hominoids and bipedalism in hominins. Am. J. Phys. Anthropol. 148, 111e122. Williams, S.A., Russo, G.A., 2015. Evolution of the hominoid vertebral column: the long and the short of it. Evol. Anthropol. 24, 15e32. Young, J.Z., 1962. The Life of Vertebrates. Oxford University Press, London. Young, N.M., MacLatchy, L., 2004. The phylogenetic position of Morotopithecus. J. Hum. Evol. 46, 163e184.