Conference PaperPDF Available

Non-metric Cranial and Pelvic Traits as a Measure of Sexual Dimorphism in a Modern South African Population

  • Department of Defense
Non-metric Cranial and Pelvic Traits as a Measure of Sexual
Dimorphism in a Modern South African Population
Kyra E. Stull1, Michael W. Kenyhercz2, and Ericka L’Abbe1
1 Department of Anatomy, University of Pretoria, Private Bag x323, 0007, Arcadia, South Africa
2 Department of Anthropology, University of Alaska, Fairbanks
Even though the overall size and shape of
the South African crania is less dimorphic than
other populations, morphological cranial traits
have moderate to high classication accuracies
when population specic formulae are applied
(Kruger et al, 2013; Messer, 2013; L’Abbé et al,
2013). Researchers expect the pelvis to exhibit
the largest levels of sexual dimorphism in the
human body and similarly, previous studies have
demonstrated that evaluation of morphological
pelvic traits result in high classication accuracies
in South African populations (Kenyhercz, 2012).
However, none of the previous studies have
looked at corresponding pelves and crania to
evaluate variable importance in terms of the
degree of sexual dimorphism. The current study
investigated the variable trends and classication
accuracies when morphological variables of the
cranium and pelvis were evaluated from the
same individuals.
The sample consists of 112 individuals from the
University of Pretoria Bone Collection. Individuals
represent modern South Africans from black,
coloured, and white population groups (B=98,
C=1, W=13). The Walker (2008) morphological
traits were used to score the crania (Figure 1)
and the Klales et al. (2012) method, based on
the Phenice (1969) traits, was used to score the
pelves (Figure 2). The mental eminence showed
no signicant relationship to sex; therefore, the
trait was removed from further analyses.
Statistical analyses included discriminant
function analysis (DFA), exible discriminant
analysis (FDA), logistic regression (LG),
classication trees (CT), and random forest
models (RFM). Each analysis was completed
with the innominate variables, the cranial
variables, and then a combination of both
innominate and cranial. All statistical analyses
were completed in R Statistical Software (R
version 2.15.2 (2012-10-26).
Table 1. Frequency distribution of cranial trait scores for females and males.
Glabella 1 2 3 4 5
F72.4 17.2 6.9 0.0 3.5
M38.9 30.6 8.3 11.1 11.1
F69.0 17.2 10.3 3.5 0.0
M22.2 11.1 33.3 22.2 11.1
F35.0 50.0 15.0 0.0 0.0
M34.6 26.9 15.4 19.2 3.6
F37.9 48.3 6.9 3.5 3.5
M22.2 19.4 16.7 25 16.7
F44.8 31.0 20.7 3.5 0.0
M8.3 13.9 16.7 30.6 30.6
Figure 1. Cranial traits and their associated scores (taken from Buikstra and
Ubelaker 1994 and used by Walker 2008)
Figure 2. Pubic bone traits and their associated scores (taken from Klales et al.
Table 2. Frequency distribution of pelvic trait scores for females and males.
SPC 1 2 3 4 5
F9.1 70.5 20.5 0.0 0.0
M0.0 4.9 29.5 50.8 13.1
F29.6 43.2 22.7 4.6 0.0
M0.0 1.6 9.8 44.3 44.3
F38.6 40.9 20.5 0.0 0.0
M0.0 8.2 45.9 37.7 8.2
Table 3. Percent correct (%) classications by statistical method for cranial traits
Method Male Female Combined Sex Bias
DFA 77 84 80.5 7
FDA 86 79 82.5 7
LG 87.7 86.1 86.9 1.6
CT 75 92.9 84 17.9
RFM 79 77.2 78.1 1.8
Table 4. Percent correct (%) classications by statistical method for pelvic traits.
Method Male Female Combined Sex Bias
DFA 97 96 96.5 1
FDA 97 96 96.5 1
LG 98.4 100 99.2 1.6
CT 97.5 86.8 92.2 10.8
RFM 95 97 96 2
Table 5. Percent correct (%) classications by statistical method for all traits.
Method Male Female Combined Sex Bias
DFA 100 94 97 6
FDA 98 100 99 2
LG 100 100 100 0
CT 97.5 86.8 92.2 10.7
RFM 95 97 96 2
Males demonstrated low cranial scores (3 and
below) in all traits except orbit (Table 1)
Males demonstrated high pelvic scores (3 and
above) and females demonstrated low pelvic
scores (< 3) (Table 2)
Total correct classications range from 80.5-
100% (Tables 3, 4, and 5)
Greatest sex bias favored females in cranial
traits (17.9)
Logistic regression consistently performed best
Combined trait accuracies showed the highest
Discussion & Conclusion
Pelvic traits demonstrated greater sexual
dimorphism than cranial traits in the modern South
African sample under study, as demonstrated
by the trait frequency distributions and overall
correct classications
Use of pelvic traits only resulted in greatest
correct classications and least sex bias
Sexually dimorphic expressions are reduced
in the cranial traits of South Africans. Similarly,
craniometric analyses also show lower rates of
sexual dimorphism when compared to worldwide
populations (L’Abbé et al. 2013; Messer and
Ousley, 2013). Differential genetic backgrounds
or assortative mating patterns may inuence
the reduced pattern expression
*For complete list of references email:
... While these population differences in patterns of cranial and postcranial sexual dimorphism have received considerable attention in the contexts of bioarchaeology and paleodemography, they have only recently come to the forefront within forensic anthropology [14,[24][25][26][27][28][29][30]. Initially, this attention was largely raised by forensic practitioners working in human rights contexts and other international settings. ...
The increasing recognition in forensic anthropology of variation in sexual dimorphism of skull size and shape among and within human populations has increased motivation among researchers to collect craniometric data from regional populations. Modern high levels of migration on an international scale has contributed to this endeavour, as has increasing understanding of historic population histories across the world. Mainland Southeast Asia is a region that has a complex population history. A large sample of male and female skulls (n=320) selected from a collection of donated modern skeletons from northeast Thailand is used in this study to provide reliable data on the pattern of sexual dimorphism in the region. The data from these skulls is analysed using both univariate and multivariate statistics to derive discriminant functions that can be applied in sex estimation during personal identification. The results from the analyses are compared with those derived from other modern skeletal collections from northern and central Thailand. This shows that variation across the country is minimal. An on-line method of easily applying the results of the analysis to data from unknown skeletons is presented and freely available for use by practitioners of forensic anthropology.
... Additionally, classification accuracy in the present research was higher (100%) than in the Hamann-Todd sample (94.5%) used in the original study byKlales et al. (2012). Higher classification accuracy in the present research is consistent with tests in a modern South African sample (99.2%;Kenyhercz 2012;Stull, Kenyhercz, & L'Abbe, 2013). Despite the same combined correct classification accuracy, females classified higher in the South ...
This article reports on an exhumation of an individual identified to be a victim of albinism from a trench dug for a well on 16 April 2018, Ukonga, Dar es Salaam, Tanzania. The skeletal remains were sent to the Forensic Bureau of the Tanzania Police Force for further investigation. Ante-mortem and postmortem information, the deoxyribonucleic acid (DNA) and skeletal diagnosis were used in the identification. Skeletal diagnosis suggests that the victim was a young African male aged between 19-26 years old with stature of 157-166 cm. Ante-mortem trauma suggest that the individual was wounded with a sharp object on the left shoulder resulting in severe infection. The victim had a problem with acute dental caries suggesting the existence of biochemical reaction. He also had porosity lesions on cranium and vertebral column which is indicative of chronic malaria infection or anemic condition. The distal end of the tibia exhibited a peri-mortem pilon fracture on the articular surface of the malleolus. This signifies he fell into that long edge trench either intentionally or accidentally or pushed by people and that is the circumstances of death. This article also provides a general picture of challenges facing people with albinism in Africa. People with albinism have been victimized in many ways such as restricted right of self-expression, deficiency of freedom, being hunted or repudiated to death, and deficiency of medical treatment. They also are suffering from dermal diseases, injuries and infections that can easily lead to skin cancer and skeletal diseases due to their disability. In fact, this problem requires urgent resolution for health departments and those involved in the protection of civilians, especially the security agencies. It is an intolerable life; therefore, African governments should take serious measures to curtail atrocities committed to people with albinism.
Determination of sex in a burial or comingled assemblage is an essential element in forensic human skeletal identification. Establishing the sex of individual skeletal remains using skull landmarks is one of the traditional human osteological methods. This study sampled 86 skeletonized skulls of contemporary Tanzanians to test a popular sex assessment technique developed by Walker's cranial nonmetric approach. The sex estimate was scored from Walker's Log Regression equations involving nuchal crest, mastoid, glabella, mental eminence and supra orbit ridge characters. Basing on the formula, females display feature scores of 1 to 3, and males typically display individual feature scores of 3 to 5. At the same time, mastoid and glabella were the best features of sex determination after the evaluation. Also, the test displayed a high overlap between males' and females' characteristics of mental eminence, nuchal crest, and supraorbital ridge. Generally, the Walker sex estimation method using cranial features on current skeletonized Tanzania population failed to provide concrete sex assessment results, thus justifying the suggestion that; we need a modified population-specific approach.
The pelvis has long been considered the best indicator of sex due to sexual dimorphism in morphology related to parturition. Over 20 different traits have been described for their utility in sex estimation in the literature; however, many of these traits currently used for sex estimation from the pelvis have not been adequately tested for validity and reliability. Despite the multitude of morphological pelvic traits and methods developed for sex estimation, the three traits originally described by Phenice (1969) continue to be the most utilized and most accurate morphological traits for pelvic sex estimation: ventral arc, subpubic contour, and the medial aspect of the ischiopubic ramus. Other pelvic traits to be consistently utilized in sex estimation include parturition markers, the greater sciatic notch, the obturator foreman shape, pelvic inlet/outlet, and sacral morphology. Several of these popular morphological traits have been tested and shown to be more objective with metric analyses or geometric morphometric approaches. A brief history of each trait and recommendations for its use or disuse for pelvic sex estimation are presented within.
When faced with commingled remains, it might be assumed that a more “masculine” pelvis is associated with a more “masculine” cranium, but this relationship has not been specifically tested. This study uses geometric morphometric analyses of pelvic and cranial landmarks to assess whether there is an intra-individual relationship between the degrees of sexual expression in these two skeletal regions. Principal component and discriminant function scores were used to assess sexual dimorphism in 113 U.S. Black individuals. Correlation values and partial least squares regression (PLS) were used to evaluate intra-individual relationships. Results indicate that the os coxae is more sexually dimorphic than the cranium, with element shape being more sexually dimorphic than size. PLS and correlation results suggest no significant intra-individual relationship between pelvic and cranial sexual size or shape expression. Thus, in commingled situations, associations between these skeletal elements cannot be inferred based on degree of “masculinity.”
Klales et al. (2012) devised an ordinal scoring system for the morphoscopic pelvic traits described by Phenice (1969) and used for sex estimation of skeletal remains. The aim of this study was to test the accuracy and reliability of the Klales method using a large sample from the Hamann-Todd collection (n = 279). Two observers were blinded to sex, ancestry, and age and used the Klales et al. method to estimate the sex of each individual. Sex was correctly estimated for females with over 95% accuracy; however, the male allocation accuracy was approximately 50%. Weighted Cohen's kappa and intraclass correlation coefficient analysis for evaluating intra-and interobserver error showed moderate to substantial agreement for all traits. Although each trait can be reliably scored using the Klales method, low accuracy rates and high sex bias indicate better trait descriptions and visual guides are necessary to more accurately reflect the range of morphological variation.
This research evaluated whether adult morphological sex estimation methods of the innominate could be adapted and applied to subadults. The subpubic concavity, described by Phenice (1969) and revised by Klales et al. (2012), was modified for use with subadults. Two observers scored radiographic images from the PATRICIA database of 334 individuals of both sexes aged between 1.19 and 20.47 years. Score frequencies shifted from score 2 (straight) to higher frequencies of score 3 (convexity) in males and score 1 (concavity) in females with increasing age. Using ordinal logistic regression, sex classification was highest for the oldest age cohort at 97.2% and then decreased by age cohort. Interobserver error rendered a high level of agreement (0.806) using the intraclass correlation coefficient. Results indicate that the Klales method can be modified and applied to subadults to accurately estimate sex following the onset of puberty with a high degree of reliability and validity.
This research evaluates secular change in Phenice's (Am J Phys Anthropol, 30, 1969 and 297) three morphological traits of the pubis, as described by Klales et al. (Am J Phys Anthropol, 149, 2012 and 104): medial aspect of the ischio-pubic ramus, subpubic contour, and ventral arc. Ordinal scores were collected for these traits and compared between a sample of innominates from the historical Hamann–Todd Collection (n = 170) and modern Bass Donated Collection (n = 129). Using the Freeman–Halton test, significant differences between temporal sample score frequencies were found for all traits in females and for the subpubic contour and ventral arc in males. Despite these findings, classification accuracy using logistic regression between the temporal periods remained low (68.7%). These results suggest that secular changes in trait expression are occurring; however, sex estimation methods using these traits and created with historical samples are still applicable to modern forensic cases. In fact, the secular changes occurring in these traits contribute to better classification accuracy between sexes in modern populations.
ResearchGate has not been able to resolve any references for this publication.