Elsevier

HOMO

Volume 61, Issue 1, February 2010, Pages 3-15
HOMO

Geometric morphometric analyses of hominid proximal femora: Taxonomic and phylogenetic considerations

https://doi.org/10.1016/j.jchb.2010.01.001Get rights and content

Abstract

The proximal femur has long been used to distinguish fossil hominin taxa. Specifically, the genus Homo is said to be characterized by larger femoral heads, shorter femoral necks, and more lateral flare of the greater trochanter than are members of the genera Australopithecus or Paranthropus. Here, a digitizing arm was used to collect landmark data on recent human (n=82), chimpanzee (n=16), and gorilla (n=20) femora and casts of six fossil hominin femora in order to test whether one can discriminate extant and fossil hominid (sensu lato) femora into different taxa using three-dimensional (3D) geometric morphometric analyses. Twenty proximal femoral landmarks were chosen to best quantify the shape differences between hominin genera. These data were first subjected to Procrustes analysis. The resultant fitted coordinate values were then subjected to PCA. PC scores were used to compute a dissimilarity matrix that was subjected to cluster analyses. Results indicate that one can easily distinguish Homo, Pan, and Gorilla from each other based on proximal femur shape, and one can distinguish Pliocene and Early Pleistocene hominin femora from those of recent Homo. It is more difficult to distinguish Early Pleistocene Homo proximal femora from those of Australopithecus or Paranthropus, but cluster analyses appear to separate the fossil hominins into four groups: an early australopith cluster that is an outlier from other fossil hominins; and two clusters that are sister taxa to each other: a late australopith/Paranthropus group and an early Homo group.

Introduction

The proximal femur is an anatomical complex that has been argued to exhibit shape differences between the genus Homo and other hominin genera (e.g., Australopithecus, Paranthropus, Orrorin; Aiello and Dean, 1990; Galik et al., 2004; Harmon, 2009; Lovejoy, 1975, Lovejoy, 1988; Lovejoy et al., 1973; Napier, 1964; Pickford et al., 2002; Richmond and Jungers, 2008; Robinson, 1972; Senut, 2006; Walker, 1973). First, as seen in Fig. 1, it has been demonstrated that the femoral heads of the australopiths (including members of the genera Orrorin and Paranthropus) tend to be smaller in absolute size than those of Homo (Harmon, 2009; Jungers, 1988, Jungers, 1991; Kennedy, 1983a; Lovejoy et al., 1973; Napier, 1964; Pickford et al., 2002; Richmond and Jungers, 2008; Robinson, 1972; Ruff, 1988; Senut et al., 2001; Walker, 1973). Second, Orrorin, Australopithecus, and Paranthropus evince longer femoral necks than Homo, even when corrected for the size of the femoral head (Galik et al., 2004; Harmon, 2009; Lovejoy et al., 1973; Napier, 1964; Pickford et al., 2002; Richmond and Jungers, 2008; Robinson, 1972; Senut, 2006; Senut et al., 2001; Walker, 1973). Third, it has been noted that there is a more lateral flare to the greater trochanter in the genus Homo (Kennedy, 1983a; Lovejoy, 1975, Lovejoy, 1988; Lovejoy et al., 1973; Walker, 1973). Based on external dimensions, australopith femora are also said to exhibit anteroposteriorly compressed, or “narrow” femoral necks (Harmon, 2009; Reed et al., 1993; Richmond and Jungers, 2008; Walker, 1973). Early (non-Homo) hominin femora have also been shown to be characterized by more medio-laterally expanded proximal shafts than those of Early Pleistocene or recent Homo specimens (Richmond and Jungers, 2008). Finally, the australopiths are said to have on average lower femoral neck–shaft angles than H. sapiens (Harmon, 2009; Lovejoy, 1975; Walker, 1973). Most, if not all of these differences are likely to have functional and/or locomotor implications (Harmon, 2009; Lovejoy, 1975, Lovejoy, 1988; Lovejoy et al., 1973, Lovejoy et al., 2002; Richmond and Jungers, 2008), and while these features are macroscopically evident, how might they best be quantified? Standard one- and two-dimensional osteometrics may be of little use in quantifying the morphological relationship between femoral neck length and lateral flare, since (depending on how each is measured) these measurements could overlap significantly – a fact that would not become readily apparent from an analysis of these measurements themselves. Fortunately, an alternative to standard osteometrics for examining shape exists in those methods that are collectively known as geometric morphometrics. In fact, in recent years, the application of geometric morphometrics to paleoanthropological questions has become much more prevalent (e.g., Bookstein et al., 1999; Dean et al., 1998; Delson et al., 2001; Gunz and Harvati, 2007; Harmon, 2007, Harmon, 2009; Harvati, 2003a, Harvati, 2003b; Lague, 2002; Ponce de León and Zollikofer, 2001; Yaroch, 1996). The field of morphometrics itself is widely regarded as the study of variation and change in biological form (Bookstein, 1991). From its infancy, geometric morphometrics has been used to quantify biological form via the use of landmark data (Thompson, 1917), and since its data are three-dimensional coordinate points that capture the shape of the object of study, it is arguably the best means by which to quantify morphological shape (Rohlf and Marcus, 1993).

The current study uses 3D morphometrics to quantify proximal femoral shape differences among extant African hominids (sensu lato) and fossil hominins in light of the following hypotheses: (1) extant African hominid (sensu lato) taxa are distinguishable from each other based on proximal femoral shape, and (2) the proximal femora of Homo will be distinguishable in shape from those of Australopithecus or Paranthropus. It is important to note that there has been some variation in findings with regard to these hypotheses; a recent 3-D morphometric studies of proximal femoral data (Harmon, 2007, Harmon, 2009) found considerable support for the first hypothesis and qualified support for the second, while a recent multivariate analysis of 2-D proximal femoral data (Richmond and Jungers, 2008) failed to find support for the first hypothesis, but reported considerable support for the second. Here, we use a different set of landmarks than those used by Harmon, 2007, Harmon, 2009 in an attempt to further examine morphological patterning in the proximal femora of hominids.

Section snippets

Materials and methods

Twenty proximal femoral anatomical landmarks and semi-landmarks (Table 1, Fig. 2) were digitized using a digitizing arm (Microscribe® 3D, Immersion Technologies, San Jose, CA, USA) on a sample of femora that included those of 82 recent humans, 16 chimpanzees, 20 gorillas, and casts of six fossil hominins (A.L. 288-1; KNM-ER 1472; KNM-ER 1481; KNM-WT 15000; SK 82; SK 97). These landmarks and semi-landmarks were chosen to elucidate the overall shape of the proximal femur, including the head,

Results

Fig. 3 is a scatter plot of principal component 1 (PC 1) and principal component 2 (PC 2) scores calculated from the Procrustes shape data. PC 1 accounts for 20.4% of the total variance, and separates those individuals on the left, who have larger femoral heads, longer femoral necks, less cranially projecting greater trochanters, and less posteriorly-positioned lesser trochanters, from those individuals on the right, who have smaller femoral heads, shorter femoral necks, more cranially

Discussion and conclusions

The current study investigated two main hypotheses. The first hypothesis, that proximal femoral shape could be used to distinguish extant African hominid (sensu lato) taxa from each other, is confirmed by the current data. As was found by Harmon (2007), we were able to discriminate Pan, Gorilla, and recent Homo from each other in multivariate Procrustes shape space. Specifically, Homo differed from Pan and Gorilla in that it was characterized by a larger femoral head and longer femoral necks,

Acknowledgments

We thank Prof. R. Eckhardt and Prof. M. Henneberg for helpful comments on this manuscript.

References (60)

  • B. Senut et al.

    First hominid from the Miocene (Lukeino Formation, Kenya)

    C. R. Acad. Sci. Paris, Sci. Terre Plan.

    (2001)
  • D.S. Strait et al.

    A reappraisal of early hominid phylogeny

    J. Hum. Evol.

    (1997)
  • E. Trinkaus

    Femoral neck–shaft angles of the Qafzeh-Skhul early modern humans, and activity levels among immature Near Eastern Middle Paleolithic hominids

    J. Hum. Evol.

    (1993)
  • A.C. Walker

    New Australopithecus femora from East Rudolf, Kenya

    J. Hum. Evol.

    (1973)
  • L. Aiello et al.

    An Introduction to Human Evolutionary Anatomy

    (1990)
  • S.C. Antón

    Natural history of Homo erectus

    Yrbk Phys. Anthropol.

    (2003)
  • F.L. Bookstein

    Morphometric Tools for Landmark Data

    (1991)
  • F.L. Bookstein et al.

    Comparing frontal cranial profiles in archaic and modern Homo by morphometric analysis

    Anat. Rec.

    (1999)
  • C.K. Brain

    Structure and stratigraphy of the Swartkrans cave in the light of new excavations

  • F.H. Brown et al.

    Geologic setting and age

  • E. Delson et al.

    Sambungmachan 3 Homo erectus calvaria: a comparative morphometric and morphological analysis

    Anat. Rec.

    (2001)
  • K. Galik et al.

    External and internal morphology of the BAR 1002′00 Orrorin tugenensis femur

    Science

    (2004)
  • R.L. Gorsuch

    Factor Analysis

    (1983)
  • J.C. Gower

    Generalized Procrustes analysis

    Psychometrika

    (1975)
  • F.E. Grine

    Description and preliminary analysis of new hominid craniodental fossils from the Swartkrans Formation

  • E.H. Harmon

    The shape of the hominoid proximal femur: a geometric morphometric analysis

    J. Anat.

    (2007)
  • E.H. Harmon

    The shape of the early hominin femur

    Am. J. Phys. Anthropol.

    (2009)
  • K. Harvati

    The Neanderthal taxonomic position: models of intra- and inter-specific craniofacial variation

    J. Hum. Evol.

    (2003)
  • K. Harvati

    Quantitative analysis of Neanderthal temporal bone morphology using three-dimensional geometric morphometrics

    Am. J. Phys. Anthropol.

    (2003)
  • J.L. Hintze

    Number Cruncher Statistical System Version 6.0

    (1996)
  • Cited by (0)

    View full text