Immature remains and the first partial skeleton of a juvenile Homo naledi, a late Middle Pleistocene hominin from South Africa

Immature remains are critical for understanding maturational processes in hominin species as well as for interpreting changes in ontogenetic development in hominin evolution. The study of these subjects is hindered by the fact that associated juvenile remains are extremely rare in the hominin fossil record. Here we describe an assemblage of immature remains of Homo naledi recovered from the 2013–2014 excavation season. From this assemblage, we attribute 16 postcranial elements and a partial mandible with some dentition to a single juvenile Homo naledi individual. The find includes postcranial elements never before discovered as immature elements in the sub-equatorial early hominin fossil record, and contributes new data to the field of hominin ontogeny.


Introduction
Our knowledge of maturational processes and life history stages in human evolution is hindered by how extremely rare immature remains are in the hominin fossil record. Further complicating our understanding is that hominin fossil remains rarely combine teeth, cranial and associated postcranial bones, and only a handful of such partial skeletons represent immature individuals. Yet such remains provide important evidence about the maturation of extinct hominins. Apart from recent modern humans and Neanderthals (e.g., [1]), only three hominin species are represented by immature partial skeletons at this time: Australopithecus afarensis [2,3], Australopithecus sediba [4,5] and Homo erectus [6][7][8][9]. Augmenting this limited dataset is of great significance in trying to understand the evolution of human growth and development.
The immature individuals from the Dinaledi and Lesedi assemblages of Homo naledi represent individuals of many juvenile and adult life stages, presenting a unique opportunity to identify and document in particular a range of immature individuals of an extinct species, and contribute data to the study of comparative hominin ontogeny [10][11][12][13]. The Lesedi Chamber  in three pieces, with heavily eroded surface morphology [23]; a left calcaneus (U.W. 101-907) (Fig 3); and a talus and navicular (recovered together as U.W. 101-910) (Fig 4). The three tarsals were originally designated "Foot 2" [24]. The preservation of the complete left tibia is noteworthy as we are familiar with no other specimen in the hominin fossil record that preserves epiphyses in place with this degree of surface detail. An immature right tibia, U.W. 101-996 ( Fig 5) appears to be the antimere of the DH7 left tibia, U.W. 101-1070 [23]. This specimen is

PLOS ONE
Considering the number of immature skeletal material recovered from the Dinaledi hominin deposit (Table 1), we set out to test the hypothesis that additional elements of this partial skeleton were present in the sample.

Materials and methods
We assess 70 immature skeletal remains from the 2013-2104 season to identify materials that may belong to the proposed DH7 individual comprised of the six lower limb elements. All https://doi.org/10.1371/journal.pone.0230440.g005 Table 2. Homo naledi age classes for 13 individuals represented in 2013-2014 excavation season, based on dentition. Two additional individuals are present based on MNI, but only approximate age can be determined: One is an adult, and one is an immature [13].

PLOS ONE
Fossil and Rock Collections. Additionally, 3D surface models and images of DH7 elements are available for public access and download at Morphosource.org (See Table 3). The South African Heritage Resource Agency and Cradle of Humankind UNESCO World Heritage Site Management Authority granted the permits to work on the Rising Star site, excavation permit ID: 952. All necessary permissions were obtained for the described study, which complied with all relevant regulations. A minimum number of individuals (MNI) was previously assessed from the 2013-2104 season based on dental remains: nine immature individuals, and six adult individuals [10,13] ( Table 2). Six of the dental individuals are either 'Infants' (deciduous dentition only) or 'Early Juveniles' (with first molars erupted). Only a single dental specimen represents a 'Late Juvenile' with second molars erupted [13] (Table 2). This 'Late Juvenile' is represented by a right mandibular fragment, U.W. 101-377, preserving permanent teeth C, P3, P4, M1 and M2 (canine U.W. 101-1014 refit in lab) [10,11,13,20] (Fig 6). The partial mandible was recovered from depth 1 in the excavation unit.

U.W. 101-1070
Left tibia, with proximal and distal epiphyses. The proximal epiphysis of this specimen appears almost fully formed in size and shape, and is well-fitted to the diaphysis, although unfused. The distal epiphysis appears less formed in adult morphology, but is also less well preserved [23].

U.W. 101-1088
Right ischium, partial. The posterior acetabular epiphysis is fully fused to form a continuous lunate surface. The acetabular surfaces and the ischial tuberosity epiphysis are unfused, evident by beveling and rough surfaces. The pubic ramus fusion information is not preserved [36].

U.W. 101-1098
Right femur epiphysis, proximal. The partial femoral head is unfused but mature in size and appearance, with a well-formed fovea, a smooth surface, and an angulated margin with a "beak-like" projection similar to pre-adolescent human head epiphyses [23].
U.W. 101-1271 Unsided metacarpal, shaft fragment. Metacarpal shaft fragment is pre-adult in size, but age-consistent with the other hand materials combined into "Hand 5" [15]. Shaft exhibit defined adult morphology of a distinctive palmar curvature.

U.W. 101-1536
Unsided metacarpal, shaft fragment only [17]. Shaft fragment is pre-adult in size, but exhibits defined features of metacarpal, with palmar curvature and a slight ridging on the palmar aspect.  Criteria for including or excluding elements from association with DH7 include: duplication of element; developmental stage; size difference; spatial proximity (depth and span); refitting of specimens into a single element; and articulation/anatomical position of elements (Table 3). Skeletal maturation continuity was assessed using established standards from the literature [30][31][32][33]; cf [5]. Evaluations were made using the original fossils, original excavation notes and photographs, and published descriptions of H. naledi anatomy [17,23,24,34,35,36]. The vertical and horizontal context of immature material within the excavation unit was also used to help identify potentially associated elements [19]. If depth location was not available from the previously collected 3D data [18,19], depth was estimated based on excavation photographs and comparisons with known-depth specimens.
In this commingled sample we set a very high evidentiary standard to accept the association of different elements. Anatomical in situ articulation of the left lower limb elements provides certain evidence that these elements represent a single individual, deposited with soft tissue still intact. The U.W. 101-996 tibia meets this standard on the evidence of its mirror-image anatomy to U.W. 101-1070 [23], its identical state of epiphyseal development, and its very close spatial and depth position to the articulated left lower limb elements of DH7. The discovery that all fragments of two antimere elements are very near each other may suggest a low degree of disaggregation of skeletal elements in this part of the deposit.
The evidence for skeletal association here meets the standards linking immature partial skeletons at other hominin sites with commingled individuals [1,4,8,9]. El Sidrón J1 is an Early Juvenile partial skeleton (H. neanderthalensis) recovered mostly from a 1 m 2 unit, together with multiple elements from at least one younger and one older immature individual. The consistent developmental stage and some direct articulations support the interpretation that the J1

PLOS ONE
material represents a single individual [1]. The MH1 immature partial skeleton from Malapa also includes some elements in direct anatomical articulation [4], and of consistent developmental stage, although one adult partial skeleton and another immature individual are known to be present in this assemblage. A sub-adult partial skeleton from Dmanisi, Georgia is evidenced based upon stratigraphic association, with all elements found within a 1 m stratum across 7 m 2 , along with developmental consistency and lack of element duplication [8,9]. As in the Dinaledi example, the recovery of more skeletal material from any of these active sites may necessitate revisions to the original hypothesis of association.
The 2013-2014 Dinaledi excavations involved a single 0.8 m 2 unit, in addition to a small amount of material collected from the chamber surface. While the surface area of the chamber consists of approximately 20 m 2 , the depth of the sediments within is currently unknown. Consequently, we cannot predict what other remains of this individual may yet be recovered. We further caution that many elements in the collection are too fragmentary to accurately assess developmental stage.

Results
We identified additional skeletal specimens that are developmentally consistent in maturity indicators with the remains of DH7, recovered within 2-20 cm of the articulated lower limb (Figs 7 and 8; Table 4). Consistent with the dental evidence from the Dinaledi Chamber of the 'Late Juvenile' age class, postcranial elements that demonstrate a 'Late Juvenile' stage of maturation are quite rare within the sample. A right proximal femur fragment (U.W. 101-1000) was recovered from depth 1 (Fig 9A). The unfused head of this proximal femur fragment refits a right femoral head epiphysis (U.W. 101-1098), located at depth 2 ( Fig 9B). A nearly complete right ischium (U.W. 101-1088) was found at the distal end of the articulated lower left limb in the excavation unit, at depth 2 ( Fig  10). The distal segment of the left immature fibula (U.W. 101-817) was found adjacent to the articulated in situ ankle, at depth 2 (Figs 11 and 12). Two right humeral shaft pieces were found adjacent to the left tibia, and refit in the lab for a nearly complete humerus (U.W. 101-948) (Fig 13).
Additional fragmentary hand bones were found near the articulated lower limb and hemimandible, and have provisionally been associated with DH7. Despite being small in size, these hand bones are consistent with each other in their developmental morphology and age indicators [17]. Two unsided immature metacarpal shaft fragments (U.W. 101-1633 and U.W. 101-1636), and two immature fragments from a proximal and intermediate phalanx (U.W. 101-1635 and U.W. 101-1664), provisionally attributed to "Hand 5" [17], were recovered en bloc at depth 2, and processed together in the lab (Fig 14A-14D). Two additional metacarpal shafts (U.W. 101-1536 and U.W. 101-1271), also attributed to "Hand 5", were recovered from depth 2, and in spatial proximity to the partial mandible U.W. 101-377 (Fig 14E and 14F).
Although dental MNI indicated only one 'Late Juvenile' [10,13], there is postcranial evidence of a second 'Late Juvenile' individual in the Dinaledi Chamber deposit [37]. This second 'Late Juvenile' is present in the northeastern portion of the excavation pit at depth 3, is not as complete as DH7, and duplicates elements of DH7. The material includes a right femur refit from four pieces (U.W. 101-938a), an epiphyseal head (U.W. 101-938b) and two right pelvic  (Fig 15). This partial lower limb orients towards the periphery of the unit, and the more fragmentary state may be a result of the position of the bones in the commingled pit. Bodies in the more compacted center of commingled deposits from forensic and archaeological mass graves undergo differential taphonomic factors, and may result in differential preservation from those on the outer edges [cf. 38].
Considering the entire fossil collection, we hypothesize that the 'Late Juvenile' material from the southwestern portion, and from depths 1-3 of the excavation pit, derive from the same individual as the associated postcranial material. We therefore provisionally attribute all this material to DH7. However, we note that the abundance of material in the present collection may still represent only a small fraction of the total amount of hominin skeletal material in the Dinaledi Chamber as a whole.

Discussion
The conditions inside the Dinaledi Chamber have yielded an unprecedented number of Homo naledi fossils. In this case, the number of immature remains in the Dinaledi assemblage offers the possibility to expand our knowledge of stages of maturational processes in H. naledi. The

PLOS ONE
identification of a partial older juvenile skeleton of H. naledi substantially adds to the database of associated immature partial skeletons in the human evolutionary fossil record.
The skeletal maturation of DH7 places it near the same maturational stage as partial skeletons of two other species, the MH1 holotype of Au. sediba, dated to 1.98 million years old [4,5,39], and the KNM-WT 15000 skeleton attributed to H. erectus, dated to 1.6 million years old [6,[40][41][42][43]. All three partial skeletons share a combination of unfused long bone epiphyses and pelvic elements, indicating that growth was still occurring at the shoulder, hip, knee and ankle regions. Dentally, all three individuals lack erupted third molars, although the stage of premolar and canine eruptions varies across specimens. The age at death of KNM-WT 15000 is estimated between 8.3 to 8.8 years based on dental microhistology [42]; MH1 has been estimated to have been between 9-11 years old at death [5].
The DH7 partial skeleton (Fig 4) from the Late Middle Pleistocene contributes important new data to the field of hominin ontogeny. As designated, DH7 is consistent in skeletal maturity with MH1 and KNM-WT 15000, both of which exhibit an ape-like (non-human) maturational timing [5,29,41,42]. However, other body systems complicate the characterization of somatic maturity in DH7. Dentally, H. naledi appears to have a pattern of dental eruption which suggests an affinity with H. sapiens, with premolars fully emerged by the time the second molars are fully erupted; however, the pattern of dental root formation is more ape-like [44]. H. naledi has a unique dental pattern of surface enamel deposition unlike any other hominin, and unlike humans [45]. The mosaic of brain size and body size characters in H. naledi further complicates interpretations of the maturation pattern. At 480-610cc, H. naledi has an intermediate brain size between Au. sediba and H. erectus [12,46]. H. naledi has a stature more similar to Au. sediba, estimated at 143.5 cm for adults [5,46]. If Au. sediba, H. erectus and H. naledi shared a similar mode of maturity, this would suggest an age at death for DH7 between~8-11 years old.
The phylogenetic arrangement of these three species relative to modern humans is uncertain [47,48]. The contrast in brain size among them is striking, raising the possibility of testing whether the maturation pattern in these immature skeletons relates to brain development. The maturational consistency between MH1 and KNM-WT 15000 has been used to support the hypothesis that Early Pleistocene hominins had not experienced the temporal shift in ontogeny that characterizes H. sapiens, irrespective of their brain or body size [5,29,43,49,50].

PLOS ONE
It remains possible that H. naledi had a slower maturation schedule than earlier hominins, similar to modern humans and Neanderthals. The species did exist contemporaneously with both early modern humans and Neanderthals in the Late Middle Pleistocene, and some evidence suggests that H. naledi may be a phylogenetic sister to these large-brained forms [47]. If so, the DH7 individual might be older than MH1 or KNM-WT 15000,~11-15 years based on human maturational standards (Table 5). An intermediate pattern of somatic maturity is also possible.

Description:
A right humerus, no fusion of the proximal epiphysis. There is possibly a minute metaphyseal surface on the most lateral distal olecranon region.

�12-15 male
H. sapiens: The humeral head unites as early as 13 in females and 15 in males. The possibility of a metaphyseal surface on the distal region, in conjunction with the overall size, would suggest a younger age of 11-12 years [32].

Description:
A proximal femur and partial femoral head that is unfused but mature in size and appearance.
�12-13 female/ male H. sapiens: The femoral head unites between 14-19 years in modern humans, and although unfused, the mature appearance of the U.W. 101-1000 specimen suggests an upper age limit [32].

Right ischium (U.W.101-1088)
Description: The posterior acetabular epiphysis of this specimen is fully fused to form a continuous lunate surface. The acetabular surfaces and the ischial tuberosity epiphysis are unfused.

11-12 female/ male
H. sapiens: Union of the posterior acetabular epiphysis begins between 10-11 years. The acetabula in modern humans begin fusing between 11-15 years in girls, and 14-17 years in boys. The ischial tuberosity initiates fusion between 13-16 years, leading to an estimated age at death of <13 years for U.W.101-1088 [36].

Right tibial shaft and left tibia with epiphyses (U.W.101-996; U.W.101-1070)
Description: For U.W.101-1070, the proximal epiphysis appears almost fully formed in size and shape, and is well-fitted to the diaphysis, although unfused. The distal epiphysis appears less well formed, but is also less well preserved.

�14-15 male
H. sapiens: The proximal tibial epiphysis initiates fusion as early as 13 years in females and 15 years in males. However, reported timings for the beginning of distal epiphysis fusion vary from as early as 11 years in girls and 14 years in boys [31], to as late as 14 years in girls and 15 years in boys [33]. combination of primitive and derived traits, and its relatively recent existence, may shed light on the evolutionary adaptations that drove changes in the tempo of maturity, and possibly life history, in other late members of the Homo genus. The unprecedented assemblage of immature remains, plus the rare juvenile H. naledi partial skeleton, provide new opportunities to further the field of hominin ontogeny, and to explore the factors that shaped the evolution of extended maturity and life history in our own species.