A small catarrhine talus from the middle Miocene Nachola, northern Kenya

Abstract

We describe a small catarrhine talus discovered in Nachola, northern Kenya. Although Nacholapithecus, Nyanzapithecus, and Victoriapithecus are known from Nachola, this talus is too tiny to be assigned to a female Nacholapithecus. It is also different in shape from both Nacholapithecus and Victoriapithecus. Hence, it is best to assign it to Nyanzapithecus, whose tarsal features were not known before. We compared this talus to tali of other small Miocene catarrhines from Kenya and several extant catarrhines. The estimated body mass of this specimen is 5.2–5.5 kg, depending on the regression formula. It is distinguished from comparative taxa by the feeble development of the malleolar cup, which is a stabilizing mechanism of the talocrural joint at the full dorsiflexion, and a low trochlea, suggesting that the talocrural joint was not specialized for rapid extension-flexion. These features suggest that Nyanzapithecus walked and ran in the trees with less agility and did not frequently engage in leaping. Additionally, it has a wide posteroinferior surface of the trochlea with a straight lateral border, probably enhancing the talocrural joint’s full plantarflexion. This might be suggestive of hindlimb suspension.

Introduction

Nachola (15–16 Ma) is one of the richest primate fossil sites in the Miocene of Africa. The great majority of the primate fossils that Nachola has yielded belong to Nacholapithecus kerioi, a baboon-sized hominoid. The excavation work at Site BG-K, which began in 1996, has greatly increased the number of hominoid fossils at Nachola. However, a considerable number of fossils were collected previously in the 1980s. Rose et al. (1996) first reported the postcranial anatomy of Na. kerioi based on the 1980s Nachola collection. However, they did not include one primate talus in their monograph. This specimen (KNM-BG 15542: Figure 1) is rather small compared to other postcranial specimens from Nachola, which included another talar specimen (Figure 2). Rose et al. (1996) tentatively estimated body mass sexual dimorphism of Na. kerioi as 2:1. Later, Kikuchi et al. (2018) and Kikuchi (2023) more precisely estimated the degree of sexual dimorphism using expanded samples of Na. kerioi and confirmed the initial estimation by Rose et al. (1996). According to this body mass (BM) sexual ratio and the range of the intrasexual size variation (Kikuchi et al., 2018), KNM-BG 15542 is too small to be attributed to a female Na. kerioi. Besides Na. kerioi, two catarrhine taxa are represented from Nachola by dental materials: Nyanzapithecus harrisoni and Victoriapithecus macinnesi. As we detail below, we think that KNM-BG 15542 is distinct from tali of Victoriapithecus and best attributable to Ny. harrisoni.

Figure 1.

KNM-BG 15542 from six aspects. (a) Anterior view. (b) Posterior view. An asterisk indicates an impingement of the medial trochlear margin by an encroachment of the deltoid ligament attachment. An arrow indicates the lateral trochlear margin running straight inferiorly. (c) Superior view. Note a very shallow anterior part of the malleolar surface (asterisk). An arrow indicates a pathological bony growth on the neck (demarcated by a dotted line). (d) Inferior view. (e) Lateral view. An arrow indicates a pathological bony growth on the neck. (f) Medial view. The boundary of the abraded area is shown by a broken line.

Figure 2.

Fossil primate talar specimens examined in this study. (a) KMN-BG 35250I (Nacholapithecus), (b) KNM-BG 15529 (Nacholapithecus), (c) KNM-BG 15542 (Nyanzapithecus), (d) KNM-MB 9422 (Victoriapithecus), (e) KNM-WK 17171A (Simiolus), (f) KNM-RU 1748 (?Dendropithecus), (g) KNM-SO 392 (?Limnopithecus evansi), (h) KNM-SO 967 (?Kalepithecus), (i) KNM-SO 968 (?Rangwapithecus), (j) KNM-RU 2036CO (Ekembo heseloni), (k) KNM-LG 621 (?Limnopithecus legetet), (l) KNM-WK 16952 (Turkanapithecus), (m) KNM-SO 1402 (Proconsul africanus).

Nyanzapithecus is a small-sized extinct catarrhine, and four species (Ny. vancouveringorum, Ny. pickfordi, Ny. harrisoni, Ny. alesi) have been erected based on specimens discovered from fossil sites in Kenya such as Mfwangano, Rusinga, Maboko, Kipsaraman, Nachola, and Napudet. Nyanzapithecus is included in the subfamily Nyanzapithecinae along with Turkanapithecus, Rangwapithecus, and some other genera, namely, Mabokopithecus and Xenopithecus in Harrison (2010) and Rukwapithecus and Oreopithecus in Nengo et al. (2017). The higher taxonomy of the Nyanzapithecinae differs by authors. Harrison (2010) proposes placing the Nyanzapithecinae in the family Proconsulidae (the superfamily Proconsuloidea) together with the Proconsulinae and Afropithecinae, whereas Nengo et al. (2017) suggest Nyanzapithecinae and Afropithecinae form a clade of stem hominoids. Despite disagreement with its taxonomy, there is a consensus that the Nyanzapithecinae are a monophyletic group basal to crown hominoids.

Nyanzapithecus is poorly known postcranially, and only two postcranial specimens have been assigned to this genus. McCrossin (1992) described a small hominoid proximal humerus discovered in Maboko (15 Ma) and provisionally attributed it to Nyanzapithecus (Ny. pickfordi) on a size basis. Harrison (2002) proposed that a proximal humerus (KNM-RU 17376) from Rusinga Island (17–18 Ma) was best assigned to Ny. vancouveringorum based on its size, although Gebo et al. (1988) attributed this specimen to either Dendropithecus or Ekembo heseloni. Thus, this talus is one of a few Nyanzapithecus postcranial elements and the first foot element to be reported. We describe this specimen and infer the positional behavior of Ny. harrisoni.

Materials and Methods

KNM-BG 15542 was found at Site BG-O in 1984 and is currently stored at the National Museums of Kenya (NMK). We compared KNM-BG 15542 to fossil tali of early and middle Miocene small-sized catarrhines housed at the NMK (Table 1). We also included one talus of male Nacholapithecus (KNM-BG 35250I) in the comparison. We used published data in Rose et al. (1992) for quantitative comparisons with extant primate taxa. The extant taxa are Pan, hylobatids, Cercopithecus, Colobus, Ateles, and Cebus. To eliminate inter-observer errors, one of the authors (L.K.) took linear measurements from fossil tali examined by Rose et al. (1992), compared them to the published datasets, and confirmed that the inter-observer error does not affect the result significantly (SOM Table S1). Conventional bivariate analyses (indices) were taken for quantitative comparison (Table 2). For qualitative analysis, we utilized a laser scanner (Transcan C 3D Scanner, Shining3D) to produce 3D rendering models of the fossil tali, which helped to eliminate any illusory effects caused by surface color and texture (Figure 3).

Table 1.

Fossil measurements¹

	a	b	c	d	e	f	g	h	i	j
	Trochlea length	Trochlea width anterior	Trochlea height lateral	Total length	Trochlea width middle	Trochlea height medial	Trochlea width posterior	Head & neck length	Neck width	Neck angle
KNM-WK 17171A Simiolus	—	11.0	8.3	—	10.2	9.3	—	12.8	9.0	33
KNM-RU 1748 ?Dendropithecus	15.0	10.9	9.5	23.1	10.6	10.4	9.0	13.4	9.0	35
KNM-SO 392 ?Limnopithecus evansi	12.2	9.7.0	7.7	20.7	8.7	8.4	6.3	12.8	7.6	34
KNM-SO 967 ?Kalepithecus	14.8	11.1	8.4	25.3	10.7	9.1	9.2	14.6	8.5	26
KNM-SO 968 ?Rangwapithecus	—	12.8	—	25.2	12.0	10.1	—	12.0	11.3	42
KNM-RU 2036CO Ekembo heseloni	16.5	12.8	10.5	27.1	12.6	10.4	9.1	15.4	12.1	38
KNM-LG 621 ?Limnopithecus legetet	—	10.0	7.5	—	9.8	9.2	—	13.6	8.8	31
KNM-WK 16952 Turkanapithecus	15.7	11.2	8.8	24.7	11.1	9.9	—	14.8	8.5	23
KNM-SO 1402 Proconsul africanus	20.5	14.1	13.7	31.4	13.1	12.7	10.5	15.5	13.5	24
KNM-BG 35250I Nacholapithecus	23.0	16.8	14.1	33.8	14.8	12.6	12.3	20.2	13.8	20
KNM-BG 15542 Nyanzapithecus	13.7	10.5	7.3	22.5	9.7	7.7	7.3	12.5	9.5	16

¹ Fossil measurement definitions following Rose et al. (1992). Length measurements are given in mm and angles in degrees. We followed their taxonomic assignment for non-Nachola specimens, but KNM-LG 621 and KNM-SO 392, which were assigned to cf. L. legetet and L. evansi, respectively by Harrison (1982). Measurements are given in mm except for Neck angle (°).

Table 2.

Talar indices of extant and fossil taxa (%)¹

	Index 1 Trochlear length a/d	Index 2 Trochlear width e/b	Index 3 Trochlea wedging g/b	Index 4 Medial trochlear height f/e	Index 5 Lateral trochlear height c/e	Index 6 Neck width i/h
Pan	59	96.5	84.4	84.3	92.7	84.7
	48–73	90–103	76–94	76–92	84–101	72–105
	7.4	3.9	5.4	5.1	6.9	10.0
Hylobatids	61.6	91.6	80.5	94	91.2	79.7
	57–68	85–100	67–92	83–109	84–96	69–104
	4.0	4.9	7.8	8.6	4.1	10.5
Cercopithecus	63.8	96.8	74.8	84.9	82.5	65.9
	59–70	93–100	70–79	77–95	79–89	53–75
	3.2	2.2	3.0	5.8	3.0	6.2
Colobus	57.3	93.7	75.6	86.5	76.2	72.2
	51–65	87–97	68–83	81–93	69–84	62–80
	4.4	3.5	4.6	4.0	5.6	6.7
Ateles	66	81.7	68.6	70.2	62.5	53.6
	60–73	77–88	58–80	64–77	56–69	47–72
	4.6	3.4	7.1	4.5	3.6	7.3
Cebus	57.9	92.6	78	80.4	85.3	56.6
	55–62	89–96	74–82	69–95	75–92	47–68
	2.5	2.4	2.6	8.2	6.0	6.9
KNM-WK 17171A Simiolus	—	92.7	—	91.2	81.4	70.3
KNM-RU 1748 ?Dendropithecus	64.9	97.2	82.6	98.1	89.6	67.2
KNM-SO 392 ?Limnopithecus evansi	58.9	89.7	64.9	96.6	88.5	59.4
KNM-SO 967 ?Kalepithecus	58.5	96.4	82.9	85.0	78.5	58.2
KNM-SO 968 ?Rangwapithecus	—	93.8	—	84.2	—	94.2
KNM-RU 2036CO Ekembo heseloni	60.9	98.4	71.1	82.5	83.3	78.6
KNM-LG 621 ?Limnopithecus legetet	—	98	—	93.9	76.5	64.7
KNM-WK 16952 Turkanapithecus	63.6	99.1	—	89.2	79.3	57.4
KNM-SO 1402 Proconsul africanus	65.3	92.9	74.5	96.9	104.6	87.1
KNM-BG-35250I Nacholapithecus	68.0	88.1	73.2	85.1	95.3	68.3
KNM-BG-15542 Nyanzapithecus	60.9	92.4	69.5	79.4	75.3	76.6

¹ Data for extant primates data adopted from Rose et al. (1992). Values are average, range, and standard deviation.

Figure 3.

Three-dimensional renderings of the tali of Miocene catarrhines and some extant catarrhines examined in this study. Right tali are shown in the mirror image for comparative purposes. Not to scale.

We used formulae developed by Tsubamoto et al. (2016) to estimate the BM of the individual to which KNM-BG 15542 belonged. The trochlear width, Li1 in Tsubamoto et al. (2016), was chosen as the independent variable because we believe this measurement is least susceptible to inter-observer error. We used three models: the cercopithecid model, the monkey (= cercopithecids and platyrrhines) model, and the monkey and ape model (SOM Table S2). Although KNM-BG 15542 falls in the size range of the cercopithecid and platyrrhine samples, their feet might scale differently from those of stem catarrhines or hominoids. Along with the predicted values, the approximated 95% prediction intervals were calculated from the standard error of the estimate (Ruff, 2003). However, because the 95% intervals were wide, we also calculated the ranges of ±20% prediction values following the recommendation by Delson et al. (2000).

Results and Discussion

Description (Figure 1, SOM digital 3D object 1)

KNM-BG 15542 is 22.5 mm long proximodistally (Table 1). This specimen is relatively well preserved and does not exhibit severe deformation or cracking, unlike specimens recovered from Site BG-K, including the Na. kerioi holotype. Its mature status is supported by the sharpness of the trochlear rims and other articular surface borders.

The dorsomedial side of the distal part is sheared off. However, the damage is restricted to the medial surface of the head and does not extend to the neck (Figure 1f, broken line). A small piece of surface bone (approximately 3 mm in diameter) is cut off on the lateral side of the head articular surface. The surface of the inferomedial portion of the talar body is eroded (demarcated by a dotted line in Figure 1b, d). This erosion extends posteriorly to the medial side of the posteromedial tubercle. Otherwise, no significant damage is present.

Looking at the dorsal view, the trochlea appears moderately long and anteriorly wide. Both the relative trochlear length (index 1 in Table 2) and relative trochlear width (index 2) fall within the ranges of variation found in many comparative extant taxa. However, the anterior trochlea shows strong wedging (index 3, 69.5%), which is similar to the degree of wedging observed in Ateles (index 3, 68.6% ± 7.1%), but not in other compared extant primates (Figure 4a). The lateral trochlear margin does not extend anteriorly further than the centre part of the trochlea due to the shallowness of the trochlear groove in the anterior trochlea (Figure 1c). Thus, no notch is formed between the lateral trochlear margin and the talar neck. This condition is not common in extant catarrhines.

Figure 4.

Plots of indices: (a) index 3, (b) index 4, and (c) index 5. Average ±1 standard deviation for extant taxa. Data for extant primates were adopted from Rose et al. (1992).

In posterior view, the trochlear groove is modestly deep. The lateral trochlear margin is higher than the medial one, but the height difference is moderate (Figure 1a, b). Unlike the anterior part of the trochlea, the wedging is weak in the posterior part of the trochlea (Figure 1c). In most catarrhine tali, the lateral trochlear surface margin changes course medially as it approaches the posterolateral tubercle. However, this specimen’s lateral margin runs almost straight inferiorly (Figure 1b, arrow). Therefore, the trochlear surface is wide towards the inferior end, although the medial border is partially impinged by an encroachment of the deltoid ligament attachment on the medial side of the talar body (Figure 1b, asterisk). The posterolateral tubercle of the talar body is sharp and pronounced. The flexor hallucis groove is wide. This groove starts below the centre of the trochlea surface and changes its course medially on the plantar side.

The trochlea is relatively low, especially on the medial side. The medial trochlear height index (index 4, 79.4%) is lower than the averages of extant catarrhines but is similar to (or higher than) those of Ateles (70.2% ± 4.5%) and Cebus (80.4% ± 8.2%) (Figure 4b). A similar pattern is observed for the lateral trochlear height (index 5: Figure 4c). Therefore, the curvatures of the medial and lateral margins are not strong (Figure 1e, f). The articular surface for the tibial malleolus is only weakly concave in its anterior part, and the malleolar cup is almost absent (Figure 1c, asterisk: see also Figure 3). The anterior border of the malleolar facet is obtuse and raised only faintly in KNM-BG 15542. The facet is shallow and faces medially and slightly dorsally. The posterior-facing component in the malleolar facet is almost absent. This is unlike the condition in extant catarrhines. Laterally, the articular surface for the fibular malleolus is flat and vertical except for the triangular area on the lateral process.

The posterior calcaneal articular surface is elongated and narrow on the plantar aspect (Figure 1d). Its shape resembles a rectangle, with a slight narrowing in the middle. The surface has a slight spiral shape. The middle calcaneal articular surface is moderately wide/long and becomes narrower towards the posterior end, which is situated at the midpoint of the posterior calcaneal articular surface.

The angle of the neck is 16° (Table 1). Because the medial portion of the head is damaged, this value should be regarded as underestimated. Nonetheless, it is relatively low among the examined fossil specimens.

A mild pathological symptom is present on the dorsal side of the neck. Adjacent to the dorsolateral border of the articular surface for the navicular, there is a rounded bony growth on the neck cortex (Figure 1c, e, arrow). The surface of this tubercle is smooth. It projects dorsally and laterally. The lateral projection narrows the non-articular hollow on the lateral side of the talar neck. This tubercle is homologous with the ‘dorsal tubercle’ (Harrison, 1982, 1989), which provides the site of anchorage for the superior talonavicular ligament. Although the aetiology is not certain, an excessive osteogenic response at the attachment of the dorsal talonavicular ligament probably caused this bony projection. Its growth provides another line of evidence about the mature status of this individual.

The dorsoventral diameter of the head is 8.7 mm. Although the erosion on its medial side hampers precise measurement of the head width, the head appears moderately tall (Figure 1a). The neck is not elongated and appears wide, although the pathological bony growth may emphasize the thickness of the neck (index 6). In this regard, it differs from Ateles and Cebus, whose talar neck is well constricted mediolaterally.

The BM of the individual that KNM-BG 15542 belonged to was calculated using three formulae devised by Tsubamoto et al. (2016). The predicted BM (±20% interval) ·was 5.51 (4.41–6.61) kg, 5.45 (4.36–6.54) kg, and 5.22 (4.19–6.26) kg by the monkey, anthropoid, and primate models, respectively. The ±20% intervals are shown here because the 95% intervals are very wide (SOM Table S2).

Comparison with Victoriapithecus macinessi and Nacholapithecus kerioi

The estimated BM of Ny. harrisoni is close to that of male V. macinessi, which is 3.5–5.5 kg, according to Harrison (1989). However, KNM-BG 15542 and tali of V. macinessi are morphologically distinct. KNM-BG 15542 is unique in its lack of a malleolar cup (Figure 2, Figure 3). While there are variations in the development of the malleolar cup among cercopithecoid taxa, the malleolar cup is identified by a raised anterior articular border on the talar neck in all of them (Strasser, 1988; see also Harrison, 1989). In general morphology, the medial side of the trochlea is less slanted, and the lateral trochlear margin is more tightly curved in the posterior part in V. macinessi (Figure 2).

The BM estimated from KNM-BG 15542 is small for a female individual of Na. kerioi. According to Kikuchi et al. (2018), BM estimates based on female femoral specimens range from 7.7 to 10.1 kg (total range of the lower and upper 20% intervals for these estimated values = 6.16–12.1 kg). The upper 20% ranges of the estimated BMs of KNM-BG 15542 marginally overlap with this range (SOM Table S2). However, besides the size difference, KNM-BG 15542 differs from Na. kerioi talus (KNM-BG 35250I), lacking a malleolar cup and having a lower lateral trochlear margin and the straight lateral margin of the posterior trochlear surface (Figure 2, Figure 3).

Comparison with other fossil catarrhines

Early Miocene non-cercopithecoid catarrhine tali from East Africa are morphologically similar except for size differentiation, showing characteristics that are moderate in their expression among a range of various living primates: for example, depth of the trochlear groove, height of the lateral trochlear margin relative to the medial one, posterior wedging of the trochlea, size of the medial and lateral tubercles of the posterior talar body, lateral projection of the malleolar facet, malleolar cup on the dorsomedial aspect of the neck, and medial angulation of the neck (Harrison, 1982). This generalization also holds for more recently discovered specimens, including those from the early and middle Miocene (Rose et al., 1992). However, there is some morphological variation that has not been discussed.

The absence of the malleolar cup is a distinguishing characteristic of the Ny. harrisoni talus. Although other fossil specimens examined here showed varied levels of development of the malleolar cup, none of them displayed a comparable morphology as that of Ny. harrisoni talus (see Figure 3). It is possible that the development of the malleolar cup is related to body size. However, even the talus of ?Limnopithecus evansi (KNM-SO 392), which is smaller than KNM-BG 15542 (Table 1), exhibits a distinct malleolar cup (Figure 3).

Another notable feature of KNM-BG 15542 is the weak anterior projection of its lateral trochlear margin (in other words, the shallowness of the trochlear groove in the anterior part). Tali of many fossil catarrhines (especially large forms) from the East African Miocene have lateral margins that protrude further distally than the depressed central part of the trochlear surface. Therefore, a notch is formed between the lateral border and talar neck in dorsal view. In KNM-BG 15542, the anterior border of the trochlea surface is straight and lacks a notch. There are a few small-sized tali that are similar in this regard. For example, KNM-RU 1748 (?Dendropithecus), KNM-SO 967 (?Kalepithecus), KNM-LG 621 (?Limnopithecus legetet), and KNM-WK 16952 (Turkanapithecus) exhibit no or slight projection (Figure 2, Figure 3). However, in the case of KNM-WK 16952, the projection may be underestimated due to fracture damage in this specimen.

Regarding general trochlear morphology, anterior trochlear wedging of KNM-BG 15542 is relatively strong. Among fossil specimens, KNM-SO 392 (?L. evansi) and KNM-RU 2036CO (Ekembo heseloni) exhibit a similar degree of wedging (Table 2 index 3; Figure 4a). However, the trochlea of KNM-BG 15542 differs from their trochlea at lower heights (indices 4, 5; Figure 4b, c). This also holds for distinction from other nyanzapithecines (i.e. KNM-SO 968 ?Rangwapithecus, KNM-WK 16952 Turkanapithecus). KNM-BG 15542’s trochlear is also unique in having a broad posteroinferior surface with an inferiorly running lateral margin, unlike any other fossil tali.

Functional adaptation and locomotor behavior

The talus of Ny. harrisoni shows a unique combination of characters not observed in extant and fossil catarrhines. Its trochlea is low. A high trochlea increases lever arm for the talocrural joint and is suitable for leaping as seen in tarsiers and galagos, though this is not always the case in leaping strepsirrhines (Gebo, 1988). The opposite condition in Ny. harrisoni suggests that the talocrural joint was not specialized for rapid extension-flexion (dorsiflexion-planterflexion). A deep and well-defined malleolar cup provides increased stability of the talocrural joint during rapid movements of the upper ankle joint (Harrison, 1989). A deep trochlear groove prevents side-to-side movements and is adaptive for leaping (Gebo, 1988). Very weak development of such stabilizing mechanism of the talocrural joint in Ny. harrisoni suggests that dorsiflexion-extension occurred with relatively low velocity compared to other similar-sized catarrhines, providing further evidence to exclude leaping as an important locomotor mode in Ny. harrisoni. In the trochlear morphology, Ny. harrisoni shows some similarity to Ateles, although the trochlea of the latter is lower and has a deeper groove. This may endorse the above interpretation. This morphology contrasts with that observed in KNM-SO 392 (?L. evansi), which exhibits a high trochlea and a deep trochlear groove, suggesting committed leaping behaviour.

The wide posteroinferior surface of the trochlea, along with the straight lateral border in Ny. harrisoni, will facilitate the full plantarflexion of the talocrural joint. The ankle joint is maximally plantarflexed at the initial and end stages of the stance phase in primate quadrupedal walk (Higurashi et al., 2019; Druelle et al., 2021). Although greater plantarflexion of the upper ankle joint may contribute to a compliant gait by increasing step length, contact time, and joint yield, the same effects can also be produced by knee extension (see Larney and Larson, 2004). Unfortunately, the kinematics of the ankle joint is not well studied in primate quadrupedalism, and such an advantage is inconclusive. On the other hand, it is known that extreme plantarflexion of the upper ankle joint, subtalar inversion, and transverse tarsal supination is required in hindlimb suspension (Meldrum et al., 1997). Although KNM-BG 15542 alone cannot attest to the possibility of hindlimb suspensory behaviour, additional tarsal specimens may help to find such a functional complex.

According to McCrossin (1992), Nyanzapithecus in Maboko (Ny. pickfordi) was an active tree climber and did not have adaptations for forelimb suspension as evidenced by a proximal humerus discovered from Maboko. The present study first revealed the functional adaptation of the Nyanzapithecus ankle. Nyanzapithecus probably walked and ran in the trees with less agility compared to similar-sized extant arboreal cercopithecids and did not engage in leaping frequently.

Fossil evidence such as the dominance of browsers among herbivore mammals or the presence of a lorisid Mioeuticus suggests forested palaeoenvironments of Nachola (Tsujikawa and Nakaya, 2005; Kunimatsu et al., 2017). The predominantly abundant primate Na. kerioi had elbow features that implicate arboreal generalized quadrupedalism and some antipronograde behaviours such as vertical climbing or clambering (Ishida et al., 2004; Nakatsukasa and Kunimatsu, 2009; Takano et al., 2018, 2020; Nishimura et al., 2022; Pina and Nakatsukasa, 2024). Na. kerioi’s total morphological pattern is consistent with living in the forest. Because using the understorey or forest floor increases predation risk for small and less agile primates, Ny. harrisoni likely used a continuous pathway to travel through the arboreal milieu. This suggests the canopy was closed in the forests in Nachola.

Acknowledgments

We thank the NMK management for allowing us to carry out this study. Special gratitude goes to Dr E. Ndiema, the Head of the Earth Sciences Department, for his kind support and for facilitating access to the fossil specimens, to Dr F. K. Manthi for moral support, to the Osteology Section of the NMK for providing the extant specimens, and to editors and anonymous reviewers for thoughtful and constructive comments. The 3D data of Ateles used in Figure 3 were downloaded from Morphosource (https://www.morphosource.org/): Ateles-fusciceps-f-AMNH-188135-talus-L-000085553.ply (doi 10.17602/M2/M85553). This study was partly supported by Basic Science Research Projects from the Sumitomo Foundation, Japan (grant number 2300366, to Y.K.), Kakenhi 23K27253 (to M.N.), and 20K06835 and JP 24K02108 (to Y.K.) from the JSPS.

References

•  Delson  E.,  Terranova  C.J.,  Jungers  W.L.,  Sargis  E.J.,  Jablonski  N.G., and  Dechow  P.C. (2000) Body mass in Cercopithecidae (Primates, Mammalia): estimation and scaling in extinct and extant taxa. Anthropological Papers of the American Museum of Natural History 83, American Museum of Natural History, New York.
•  Druelle  F.,  Supiot  A.,  Meulemans  S.,  Schouteden  N.,  Molina-Vila  P.,  Rimbaud  B., et al (2021) The quadrupedal walking gait of the olive baboon, Papio anubis: an exploratory study integrating kinematics and EMG. Journal of Experimental Biology, 224: jeb242587. doi: 10.1242/jeb.242587
•  Gebo  D.L. (1988) Foot morphology and locomotor adaptation in Eocene primates. Folia Primatologica, 50: 3–41. doi: 10.1159/000156332
•  Gebo  D.L.,  Beard  K.C.,  Teaford  M.F.,  Larson  S.G.,  Jungers  W.L., and  Fleagle  J.G. (1988) A hominoid proximal humerus from the early Miocene of Rusinga Island, Kenya. Journal of Human Evolution, 17: 393–401. doi: 10.1016/0047-2484(88) 90028-0
•  Harrison  T. (1982) Small-bodied apes from the Miocene of East Africa. Ph.D. dissertation, University College London. https://www.academia.edu/12515686/1982_Harrison_Small_bodied_apes_from_the_Miocene_of_East_Africa
•  Harrison  T. (1989) New postcranial remains of Victoriapithecus from the middle Miocene of Kenya. Journal of Human Evolution, 18: 3–54. doi: 10.1016/0047-2484(89)90022-5
•  Harrison  T. (2002) Late Oligocene to middle Miocene catarrhines from Afro-Arabia. In: W.C. Hartwig (ed.), The Primate Fossil Record. Cambridge University Press, Cambridge, pp. 311–338.
•  Harrison  T. (2010) Dendropithecoidea, Proconsuloidea and Hominoidea (Catarrhini, Primates). In: L. Werdelin and W.J. Sanders (eds.), Cenozoic Mammals of Africa. University of California Press, Berkeley, pp. 429–469.
•  Higurashi  Y.,  Maier  M.A.,  Nakajima  K.,  Morita  K.,  Fujiki  S.,  Aoi  S.,  Mori  F.,  Murata  A., and  Inase  M. (2019) Locomotor kinematics and EMG activity during quadrupedal versus bipedal gait in the Japanese macaque. Journal of Neurophysiology, 122: 398–412. doi: 10.1152/jn.00803.2018
•  Ishida  H.,  Kunimatsu  Y.,  Takano  T.,  Nakano  Y., and  Nakatsukasa  M. (2004) Nacholapithecus skeleton from the Middle Miocene of Kenya. Journal of Human Evolution, 46: 67–101. doi: 10.1016/j.jhevol.2003.10.001
•  Kikuchi  Y. (2023) Body mass estimates from postcranial skeletons and implication for positional behavior in Nacholapithecus kerioi: Evolutionary scenarios of modern apes. The Anatomical Record, 306: 2466–2483. doi: 10.1002/ar.25173
•  Kikuchi  Y.,  Nakatsukasa  M.,  Tsujikawa  H.,  Nakano  Y.,  Kunimatsu  Y.,  Ogihara  N., et al (2018) Sexual dimorphism of body size in an African fossil ape, Nacholapithecus kerioi. Journal of Human Evolution, 123: 129–140. doi: 10.1016/j.jhevol.2018.07.003
•  Kunimatsu  Y.,  Tsujikawa  H.,  Nakatsukasa  M.,  Shimizu  D.,  Ogihara  N.,  Kikuchi  Y., et al (2017) A new species of Mioeuoticus (Lorisiformes, Primates) from the early Middle Miocene of Kenya. Anthropological Science, 125: 59–65. doi: 10.1537/ase.170322
•  Larney  E. and  Larson  S.G. (2004) Compliant walking in primates: elbow and knee yield in primates compared to other mammals. American Journal of Physical Anthropology, 125: 42–50. doi: 10.1002/ajpa.10366
•  McCrossin  M.L. (1992) An oreopithecid proximal humerus from the middle Miocene of Maboko Island, Kenya. International Journal of Primatology, 13: 659–677. doi: 10.1007/BF02551258
•  Meldrum  D.J.,  Dagosto  M., and  White  J. (1997) Hindlimb suspension and hind foot reversal in Varecia variegata and other arboreal mammals. American Journal of Physical Anthropology, 103: 85–102. doi: 10.1002/(SICI)1096-8644 (199705)103:1<85::AID-AJPA6>3.0.CO;2-C
•  Nakatsukasa  M., and  Kunimatsu  Y. (2009) Nacholapithecus and its importance for understanding hominoid evolution. Evolutionary Anthropology, 18: 103–119. doi: 10.1002/evan. 20208
•  Nengo  I.,  Tafforeau  P.,  Gilbert  C.C.,  Fleagle  J.G.,  Miller  E.R.,  Feibel  C., et al (2017) New infant cranium from the African Miocene sheds light on ape evolution. Nature, 548: 169–174. doi: 10.1038/nature23456
•  Nishimura  A.C.,  Russo  G.A.,  Nengo  I.O., and  Miller  E.R. (2022) Morphological affinities of a fossil ulna (KNM-WS 65401) from Buluk, Kenya. Journal of Human Evolution, 166: 103177. doi: 10.1016/j.jhevol.2022.103177
•  Pina  M. and  Nakatsukasa  M. (2024) New quantitative analyses of the Nacholapithecus kerioi proximal ulna confirm morphological affinities with Equatorius and large papionins. American Journal of Biological Anthropology, 185: e25000. doi: 10.1002/ajpa.25000
•  Rose  M.D.,  Leaky  M.G.,  Leaky  R.E.F., and  Walker  A.C. (1992) Postcranial specimens of Simiolus enjiessi and other primitive catarrhines from the early Miocene of Lake Turkana, Kenya. Journal of Human Evolution, 22: 171–237. doi: 10.1016/S0047-2484(05)80006-5
•  Rose  M.D.,  Nakano  Y., and  Ishida  H. (1996) Kenyapithecus postcranial specimens from Nachola, Kenya. African Study Monographs (Supplementary Issue), 24: 3–56. doi: 10.14989/68384
•  Ruff  C.B. (2003) Long bone articular and diaphyseal structure in Old World monkeys and apes. II: Estimation of body mass. American Journal of Physical Anthropology, 120: 16–37. doi: 10.1002/ajpa.10118
•  Strasser  E. (1988) Pedal evidence for the origin and diversification of cercopithecid clades. Journal of Human Evolution, 17: 225–245. doi: 10.1016/0047-2484(88)90055-3
•  Takano  T.,  Nakatsukasa  M.,  Kunimatsu  Y.,  Nakano  Y.,  Ogihara  N., and  Ishida  H. (2018) Forelimb long bones of Nacholapithecus (KNM-BG 35250) from the middle Miocene in Nachola, northern Kenya. Anthropological Science, 126: 135–149.
•  Takano  T.,  Nakatsukasa  M.,  Pina  M.,  Kunimatsu  Y.,  Nakano  Y.,  Morimoto  N., et al (2020) New forelimb long bone specimens of Nacholapithecus kerioi from the Middle Miocene of northern Kenya. Anthropological Science, 128: 27–40. doi: 10.1537/ase.181022
•  Tsubamoto  T.,  Egi  E.,  Takai  M.,  Thaung-Htikec, and  Zin-Maung-Maung-Theind (2016) Body mass estimation from the talus in primates and its application to the Pondaung fossil amphipithecid primates. Historical Biology, 28: 27–34. doi: 10.1080/08912963.2014.971783
•  Tsujikawa  H. and  Nakaya  H. (2005) Geologic age and paleoenvironments of mammalian faunas from Samburu Hills, northern Kenya. Chikyu Monthly, 27: 603–611. (in Japanese with English title)