When Did the Modern Human Pattern of Childbirth Arise? New Insights from an Old Neandertal Pelvis

COMMENTARY

When did the modern human pattern of childbirth arise? New insights from an old Neandertal pelvis

Robert G. Franciscus1 Department of Anthropology, 114 Macbride Hall, University of Iowa, Iowa City, IA 52242

uman birthing is difficult ow- complete 1.4–0.9 million-year-old fe- ing to a tradeoff between male Homo erectus pelvis from Gona, large neonatal brain size and Ethiopia (6), with the primitive nonrota- maternal pelvic dimensions, tional pattern persisting until the Middle Hwhich are constrained by aspects of bi- Pleistocene (5). Resolving the appear- pedal biomechanics (1, 2). The net ef- ance of rotational birth more precisely fect is that human neonatal head size has been hindered by the paucity of fos- closely matches maternal pelvic dimen- sil pelvic remains, especially from fe- sions, unlike in our closest living rela- males, because of the fragility of pelvic tives, the great apes, whose pelvic bones. In a recent issue of PNAS, dimensions are larger than neonatal Weaver and Hublin (7) report on a vir- head sizes. This size relationship, along tual reconstruction of the pelvis of the with a twisted birth canal shape, makes Tabun C1 adult female Neandertal human parturition mechanically difficult which indicates, surprisingly, that Nean- and results in a unique pattern of ‘‘rota- dertals retained the primitive birth pat- tional’’ birth (Fig. 1). As in humans, and tern into the Late Pleistocene, and that unlike great apes, monkey neonatal the shift to a rotational birth pattern head size closely matches the mother’s occurred later in human evolution than pelvic dimensions. Although birth rota- previously thought. tion occurs in some monkeys, the rota- Over the past 25 years, rotational tion pattern is different from that in birth has been inferred for archaic modern humans, with monkey neonates Homo sapiens based largely on 3 speci- exiting the birth canal facing forward mens (3): Sima 1, a 600,000-year-old (1). Therefore, in both apes and mon- male pelvis from the Sima de los Hue- keys, it is relatively easy for a mother to sos, Spain (8, 9); a 260,000-year-old left guide her infant out of the birth canal, os coxa belonging to a partial female keep the umbilical cord from wrapping skeleton from Jinniushan, northeastern around the neck, and extract mucous China (10, 11); and a 60,000-year-old from the nose and mouth to facilitate partial pelvis from a male Levantine breathing. All of these critical activities Neandertal, Ke´bara2 (12). Nonetheless, are much more difficult for a human it has always been recognized that ex- mother, whose infant emerges facing trapolating female birth canal shape backwards, and pulling on the infant Fig. 1. The human birth canal inlet (a) is larger transversely (side to side) than it is anteroposteri- from male pelvic remains is uncertain. from this position also risks serious neck orly (front to back) because bipedal efficiency fa- Moreover, whereas the inlet anatomy of injury. For these reasons, humans vors a shorter anteroposterior distance between a the female Jinniushan specimen can be uniquely engage in assisted birth (obli- line that passes through both hip joints (d) and the estimated (11), the midplane and outlet gate midwifery). Although unassisted sacrum (e). The human neonatal head is longer dimensions are unknown. birth, the norm for nonhuman primates, from front to back (Lower) and therefore enters For these reasons, Weaver and Hublin does occasionally occur in humans, the transversely wider birth inlet (a) facing side- have returned to a fossil discovered assisted birth and its myriad social im- ways (Upper). As the neonate enters the pelvic nearly 80 years ago. The Tabun C1 pel- plications are the human norm cross- midplane (b), which is anteroposteriorly wider, its head rotates 90° to accommodate the twisted birth vis belongs to a 60,000- to 100,000-year- culturally (1). canal, and faces backwards away from the pelvic old female Neandertal skeleton and was Obstetrical difficulties and some form pubic bones (Lower) while its transversely wide and partially reconstructed and described in of birthing assistance may have arisen at relatively rigid shoulders are accommodated by the 1939 (13). Her pelvic remains consist of the outset of hominin evolution in con- wider inlet. The baby then exits its mother’s body portions of the left pubis and ilium, as cert with bipedalism (3), yet it seems facing backwards, turning slightly sideways again well as the right pubis, ischium, and il- likely based on 2 preserved Pliocene as its shoulders align to pass through the antero- ium. The advent of computerized virtual female pelves, A.L. 288-1 (Australopithe- posteriorly wider midplane and outlet (c). This fig- reconstruction allows highly accurate cus afarensis) and Sts 14 (Australopithe- ure was prepared by Nathan Holton, University of Iowa. [Upper reproduced with permission (Copy- restorations of fragmented and variably cus africanus), that birth in australopiths right 2009, Corbis). Lower reproduced with permis- distorted fossil elements, including mir- was nonrotational, with the long axis of sion (Copyright 2009, Nucleus Medical Art).] ror imaging of damaged or missing ele- the neonate’s head oriented transversely ments. The Weaver and Hublin recon- as it moved through the 3 birth canal planes (2). The extant human pattern of broad, rigid shoulders (4). But when did rotational birth, along with intensified this happen? Author contributions: R.G.F. wrote the paper. and obligate midwifery, most likely Until now, the consensus has been The author declares no conflict of interest. emerged later when a critical threshold that rotational birth was not present in See companion article on page 8151 in issue 20 of volume of brain size attainment in Homo was Early Pleistocene Homo (5), recently 106. added to the shared hominoid pattern of confirmed by the discovery of a largely 1E-mail: [email protected].

www.pnas.org͞cgi͞doi͞10.1073͞pnas.0903384106 PNAS ͉ June 9, 2009 ͉ vol. 106 ͉ no. 23 ͉ 9125–9126 Downloaded by guest on October 1, 2021 struction is actually the second modern- cial missing points such as those from hemipelvis by using a variety of compel- day effort. In a 2008 study in PNAS the nonexistent sacrum. Finally, the ling corroboratory techniques. assessing Neandertal neonatal brain size more complete right side of the pelvis What do the divergent Neandertal and subsequent brain growth relative to was mirrored to create the left side. and extant modern human birthing pat- recent modern humans to compare life The resulting birth inlet and outlet terns imply? Despite its male attribu- history rates, Ponce de León and col- areas of Tabun C1’s reconstructed pelvis tion, Sima 1 best represents pelvic shape leagues (14) produced a virtual recon- are very similar to the mean area value in the last common ancestor of Nean- struction of the Tabun C1 pelvis and for a geographically diverse sample of dertals and humans (9), and it has been used it to simulate the birth process in recent modern female pelves compiled used to infer neonatal midplane rotation Neandertals by using a complete neona- by the authors. The surprising result, (8). Weaver and Hublin raise caveats tal skeleton from Mezmaiskaya Cave, about that interpretation, but note that Russia (15). In their reconstruction of all agree on a transversely oval outlet in the Tabun C1 pelvis, Ponce de León Neandertals had a Sima 1. They argue that Middle Pleisto- and colleagues argued that the midplane cene encephalization trends (16) in- and outlet configuration could not be different birthing creased obstetric constraints in both the established from the fossil elements Neandertal and modern human lineages alone because of an inability to accu- pattern from modern with divergent results. Neandertals con- rately infer the ischial and sacral orien- tinued to expand transverse pelvic outlet tations. Consequently, they assumed the humans. dimensions (the primitive condition), modern human rotational pattern for commensurate with large bi-iliac the Mezmaiskaya neonate and modern breadths and cold adaptation, with sec- human cephalopelvic proportions to however, was that for birth canal shape. ondary consequences such as more or parameterize the reconstructed birth Tabun C1’s inlet index (anteroposterior less equal pubic bone lengths in both dimensions. breadth/transverse breadth) at 0.79 is Neandertal sexes (10). Modern humans, Weaver and Hublin (7) focus on par- low compared with recent human fe- in contrast, underwent anteroposterior turition itself, and they ask whether Ne- males, and her outlet index at 0.70 is expansion of the birth outlet, since the andertal females, represented by the extremely low, in fact, falling entirely African ancestors of modern humans Tabun C1 pelvis, actually had the same out of the range of Weaver and Hublin’s were built on a transversely narrow pel- birth canal size and shape proportions extensive comparative sample. Com- vis bauplan related to warmer climate. as recent humans, and thus rotational pared to the characteristically antero- The new wide-hipped female Homo birth, which was assumed in the Ponce posteriorly oval outlet in modern hu- erectus pelvis from Gona (6) complicates de Leónet al. (14) study. After com- mans (index values Ͼ1.0, Fig. 1 Lower), this climatic explanation, but does pro- puted tomography scanning the original Tabun C1’s outlet is transversely oval. vide evidence for continuity in wide bi- pelvic fragments, Weaver and Hublin Although poor preservation of the is- iliac breadths from Australopithecus to first virtually ‘‘disassembled’’ portions of chial spine precludes definitive assess- Neandertals. New interpretations of the the original 1939 reconstruction and ment of the midplane, the results for the thorax in Neandertals mirrors this conti- separated the femoral head from the outlet shape in Tabun C1 indicate that nuity by postulating that a wide body right acetabulum; the heads of both the Neandertals had a different birthing pat- with high body mass represents the right and left femora were cemented so tern from modern humans. Rather than primitive hominin condition retained in firmly within the hip joints that the orig- rotating so that the head is anteropost- Neandertals, which, if related to cold inal describers (13) never attempted to eriorly oriented (facing away from the adaptation, represented an exaptation separate them. After this step, Weaver mother, Fig. 1 Lower), Neandertals rather than cold adaptation per se (17). and Hublin mirrored missing elements would have aligned their heads trans- Finally, as Weaver and Hublin (7) ac- on the left side from the more complete versely exiting through the birth outlet. knowledge, the differences in the right side elements, and then fit and The authors test the veracity of their precise birthing pattern between Nean- aligned the elements from each side results against 2 key potential sources of dertals and modern humans should not of the pelvis by using both manual reconstruction error: the estimation of let us lose sight of the fact that both and computer algorithm techniques. the sacral landmarks, and the orienta- lineages would have had difficult births, They then used an E-M (expectation- tion of the pubis fragment relative to obligate midwifery, and all of the atten- maximization) program to estimate cru- the hip joint and other elements of the dant social implications.

1. Rosenberg KR, Trevathan W (2002) Birth, obstetrics and 8. Arsuaga J-L, et al. (1999) A complete human pelvis from 13. McCown TD, Keith A (1939) The Stone Age of Mount human evolution. BJOG 109:1199–1206. the Middle Pleistocene of Spain. Nature 399:255–258. Carmel: The Fossil Human Remains from the Levalloiso- 2. Tague RG, Lovejoy CO (1986) The obstetric pelvis of A.L. 9. Bischoff JL, et al. (2007) High-resolution U-series dates Mousterian (Clarendon, Oxford), Vol 2. 288-1 (Lucy). J Hum Evol 15:237–255. from the Sima de los Huesos hominids yields 600 kyrs: 14. Ponce de Le´on MS, et al. (2008) Neanderthal brain size 3. Rosenberg KR, Trevathan W (2001) The evolution of Implications for the evolution of the early Neanderthal at birth provides insights into the evolution of human human birth. Sci Am 285:72–77. lineage. J Archaeol Sci 34:763–770. life history. Proc Natl Acad Sci USA 105:13764–13768. 4. Trevathan W, Rosenberg KR (2000) The shoulders fol- 10. Rosenberg KR (1998) in Neandertals and Modern Hu- 15. Golovanova LV, et al. (1999) Mezmaiskaya cave: A Ne- low the head: Postcranial constraints on human child- mans in Western Asia, eds Akazawa T, Aoki K, Bar- anderthal occupation in the Northern Caucasus. Curr birth. J Hum Evol 39:583–585. Yosef O (Plenum, New York), pp 367–379. Anthropol 40:77–89. 5. Ruff CB (1995) Biomechanics of the hip and birth in 11. Rosenberg KR, Zune´L, Ruff CB (2006) Body size, body 16. Bruner E, Manzi G, Arsuaga J-L (2003) Encephalization early Homo. Am J Phys Anthropol 98:527–574. proportions, and encephalization in a Middle Pleisto- and allometric trajectories in the genus Homo: Evi- 6. Simpson SW, et al. (2008) A female Homo erectus pelvis cene archaic human from northern China. Proc Natl dence from the Neandertal and modern lineages. Proc from Gona, Ethiopia. Science 322:1089–1092. Acad Sci USA 103:3552–3556. Natl Acad Sci USA 100:15335–15340. 7. Weaver TD, Hublin J-J (2009) Neandertal birth canal 12. Rak Y (1991) in Le squelette Mouste´rien de Ke´bara, eds 17. Go´mez-Olivencia A, et al. (2009) Ke´bara 2: New insights shape and the evolution of human childbirth. Proc Natl Bar-Yosef O, Vandermeersch B (CNRS, Paris), pp 147– regarding the most complete Neandertal thorax. J Acad Sci USA 106:8151–8156. 166. Hum Evol, in press.

9126 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0903384106 Franciscus Downloaded by guest on October 1, 2021