A morphometric analysis of maxillary molar crowns of Middle-Late Pleistocene hominins

Journal of Human Evolution 47 (2004) 183e198

A morphometric analysis of maxillary molar crowns of Middle-Late Pleistocene hominins Shara E. Bailey* CASHP, Department of Anthropology, The George Washington University, 2110 G Street NW, Washington DC 20052 Received 29 April 2004; accepted 12 July 2004

Abstract This study explores the significance of shape differences in the maxillary first molar crowns of Neandertals and anatomically modern humans. It uses morphometric analysis to quantify these differences and to investigate how the orientation of major cusps, relative cusp base areas and occlusal polygon area influence crown shape. The aims of this study were to 1) quantify these data to test whether the tooth shapes of Neandertals and anatomically modern humans differ significantly and 2) to explore if either of the shapes is derived relative to earlier fossil hominins. Data were collected from digital occlusal photographs using image-processing software. Cusp angles, relative cusp base areas and occlusal polygon areas were measured on Neandertals (n Z 15), contemporary modern humans (n Z 62), Upper Paleolithic humans (n Z 6), early anatomically modern humans (n Z 3) and Homo erectus (n Z 3). Univariate and multivariate statistical tests were used to evaluate the differences between contemporary modern humans and Neandertals, while the much sparser data sets from the other fossil samples were included primarily for comparison. Statistically significant differences reflecting overall crown shape and internal placement of the crown apices were found. Neandertals are distinguished from contemporary humans by possessing maxillary first molars that 1) are markedly skewed; 2) possess a narrower distal segment of the occlusal polygon compared to the mesial segment; 3) possess a significantly smaller metacone and a significantly larger hypocone; and 4) possess a significantly smaller relative occlusal polygon area reflecting internally placed cusps. Differences in relative cusp base areas of the hypocone and metacone may contribute to the shape differences observed in Neandertals. However, early anatomically modern humans possessing a pattern of relative cusp base areas similar to Neandertals lack their unusual shape. That the morphology observed in non-Neandertal fossil hominins is more anatomically modern human-like than Neandertallike, suggests that this distinctive morphology may be derived in Neandertals. Ó 2004 Elsevier Ltd. All rights reserved. Keywords: postcanine dental morphology; morphometrics; maxillary molars; anatomically modern humans; Neandertals; teeth

* Send Correspondence To: Shara E. Bailey, Max-Planck-Institute for Evolutionary Anthropology, Department of Human Evolution, Deutscher Platz 6, D-04103 Leipzig, Germany. Tel: C49 341 3550 356; fax: C49 341 3550 399. E-mail address: [email protected]. 0047-2484/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.jhevol.2004.07.001

184

S.E. Bailey / Journal of Human Evolution 47 (2004) 183e198

Introduction Teeth have long been an integral part of taxonomic studies describing and interpreting early hominin fossils. In addition to preserving well, teeth provide myriad morphological information, from simple measurements of length and breadth to details of crown and root morphology. It is surprising, therefore, to discover that dental morphology has played a less significant role in the interpretation of later (MiddleeLate Pleistocene) hominin fossils. Until recently, dental morphology has rarely been used to address questions of alpha taxonomy and phylogeny during this time period. This most likely results from the assumption that Neandertal teeth differ little from those of modern humans, a view that can be traced back to (among others) Boule and Vallois (1957) during the middle part of the last century. Recent analyses of Neandertal dental morphology suggests that there are a number of important and diagnostic ways in which Neandertal teeth differ from those of anatomically modern humans and other fossil hominins (Bailey, 2000, 2002a, 2002b; Bailey and Lynch, in press). Identifying and describing these differences is important for a number of reasons. First, doing so may help in identifying fossil human remains that are represented only by teeth. Although relative tooth size (i.e., anterior relative to posterior teeth) has been shown to distinguish Neandertals from anatomically modern humans (Brace, 1967; Wolpoff, 1971; Trinkaus, 1978; Bytnar et al., 1994) simple tooth measurements, alone, are of limited use because the postcanine teeth of these two groups broadly overlap in the size and shape (Smith, 1982; Manzi and Passarello, 1995). Identifying morphometric or non-metric characters that are unique (in presence or frequency) to particular taxonomic groups will result in more certain identification of isolated teeth in the fossil record (e.g., Harvati, 2003a). Second, new dental traits identified in the process can be used to assess phylogenetic relationships among Middle-Late Pleistocene hominins. Most of the recent work on Neandertal dental morphology has been based on modern

human standards (e.g., the Arizona State University Dental Anthropology System or ASUDAS: Turner et al., 1991). This line of research has been useful as a baseline for identifying phenetic relationships (Bailey, 2000). However, interpretation is limited because dental traits that are absent or rare in modern humans are not included in the standard scoring procedure. Developing a scoring system that is applicable to both fossil hominins and anatomically modern humans is important for accurately assessing biological and evolutionary relationships among hominins. The first step in devising such a system is to identify relevant morphological traits and determine how they vary among groups. This study focuses on the maxillary molars, specifically the M1. Morphologically speaking, human maxillary molars are relatively simple. They usually present four cusps and sometimes possess additional features such as a distal cusplet (metaconule or Cusp 5), mesiolingual accessory cusp (Carabelli’s cusp), mesiobuccal style (parastyle, or paramolar tubercle) and/or mesial marginal tubercles (Fig. 1). Each of these traits can be found in anatomically modern humans, Neandertals and other fossil

Fig. 1. A left maxillary molar illustrating common accessory features: a: distal cusplet (metaconule or Cusp 5), b: Carabelli’s cusp, c: Mesial marginal tubercle. Orientation for all figures, M: mesial, D: distal, B: buccal, L: lingual. Abbreviations: Pr: protocone, Pa: paracone, Me: Metacone, Hy: hypocone.

S.E. Bailey / Journal of Human Evolution 47 (2004) 183e198

hominins. The frequencies for most maxillary molar traits in Neandertals fall within the ranges observed in contemporary modern human populations, albeit often near the limit of those ranges (see Scott and Turner, 1997; Bailey, 2002b). In certain cases (e.g., hypocone size) trait expression may exceed the highest grade on the modern human scale (e.g., the ASUDAS: Turner et al., 1991). Based on ASUDAS standards alone, the maxillary molars of Neandertals are not particularly diagnostic. However, they are unusual in a number of other ways. In occlusal view, Neandertal maxillary molars are strongly skewed. The buccal cusps (paracone and metacone) appear to be more mesially placed relative to the lingual cusps (protocone and hypocone). In addition, the cusps appear to be internally compressed, the cusp apices being oriented more towards the occlusal basin than is observed in contemporary modern humans (Fig. 2). Finally, the relative contribution of individual cusps to the total crown base area appears to differ from that observed in anatomically modern humans. One of the goals of this study was to quantify this variation in tooth shape. More often than not,

185

simple tooth crown measurements, such as buccolingual and mesiodistal breadths are used to estimate tooth shape (e.g., shape Z BL/ MD*100). These measurements, as well as the indices and crown area estimates derived from them, are sometimes used to assess relationships of fossil hominins (e.g., Bermu´dez de Castro, 1993). Although they are useful ways of quantifying the shape of rectangular teeth, mesiodistal and buccolingual breadths are insufficient to accurately describe irregularly shaped teeth. Instead of using simple measurements of length and width, this study utilizes a number of different approaches. First, measurements of cusp angles were used to quantify overall tooth shape. Morris’s (1986) study of contemporary human maxillary molars showed that tooth shape can be evaluated using angles and linear measurements of an occlusal polygon described by cusp apices (Fig. 3). This occlusal polygon was found to have similar buccolingual and mesiodistal relationships as the crown itself. In his work on contemporary modern human teeth, Morris found significant differences among populations. Until now, these methods have not been applied to the study of fossil hominins.

Fig. 2. A comparison M1s from a Neandertal (A) and a contemporary modern human (B). Note the skew of the two teeth, the relative contribution of each cusp and the placement of the cusp apices relative to the tooth’s perimeter. Abbreviations as for Fig. 1.

186

S.E. Bailey / Journal of Human Evolution 47 (2004) 183e198

Fig. 3. A M1 showing the occlusal polygon, drawn by connecting the cusp apices of the protocone (A), paracone (B), metacone (C) and hypocone (D). The angles of the occlusal polygon are used as one method to quantify tooth shape in this study. The area enclosed by the segments AB, BC, CD, AD represents the occlusal polygon area (OPA). This area relative to the total crown base area (outlined) reflects the orientation of cusp apices.

Second, relative cusp base areas were examined relative cusp base areas to determine if there were significant differences among taxa and to assess the extent to which different cusps contribute to the perceived shape differences. A detailed study of relative cusp base areas in Middle-Late Pleistocene humans has not been undertaken, although the groundwork has been laid by studies of earlier hominins (Wood et al., 1983; Wood and Uytterschaut, 1987; Wood and Engleman, 1988) and non-human primates (Uchida, 1998a, 1998b) . Wood and colleagues (Wood et al., 1983; Wood and Engleman, 1988) concluded from their detailed studies of maxillary and mandibular postcanine dental morphology in autralopiths and early Homo that relative cusp base areas of certain teeth may be useful taxonomic discriminators (especially in distinguishing robust from nonrobust Australopithecines). Uchida (1998a, 1998b) found significant differences in maxillary and mandibular molar relative cusp base areas among subspecies of Pongo pygmaeus and Gorilla gorilla. Based on these studies of early hominins

and hominoids it is reasonable to hypothesize that the contribution of different cusps to the molar crown base area may also differ significantly among Middle-Late Pleistocene humans. The final assessment of maxillary molar shape involved an analysis of the degree to which cusp apices are internally compressed. Tattersall and Schwartz (1999, p 7119) noted that compressed and internally placed cusps distinguish both deciduous and permanent mandibular first molars of Neandertals from those of anatomically modern humans. My observations indicate that this is true of Neandertal permanent maxillary molars as well. The decision to focus on maxillary molars is based on the fact that mandibular molars are subject to greater variability in cusp number than are maxillary molars (e.g., mandibular molars may have anywhere between 4 to 7 cusps). This fact confounds attempts to accurately quantify cusp placement. Maxillary molars (especially M1), on the other hand, present simpler and more stable morphology. The aim of this study is to quantify shape differences observed in Neandertal and anatomically modern maxillary molars and, in so doing, provide new information useful to developing standards for measuring variation in fossil hominins. It is hoped that this information will have utility for taxonomic identification as well as phenetic and phylogenetic studies. In the process I address the following questions: 1) Are there quantifiable differences between Neandertals and anatomically modern humans in M1 shape? 2) Do differences in cusp base areas account for any shape differences observed between the two groups? 3) Is the internal placement and orientation of Neandertal maxillary molar cusps a statistical reality when compared to anatomically modern humans and other fossil hominins?

Materials The M1 was chosen as the focus of this study because it is the least variable of the three

S.E. Bailey / Journal of Human Evolution 47 (2004) 183e198

maxillary molars (Butler, 1963) and nearly always possesses all four major cusps (protocone, paracone, metacone and hypocone e see Fig. 1). For measurements that were based on cusp apices, only those teeth that exhibited minimal wear (dentin exposed on not more than one cusp) were used. Where dentin was exposed on a particular cusp, the cusp apex was estimated to have been in the center of the exposed dentin. In order to measure relative cusp base areas, the teeth needed only to preserve the major fissures separating cusps to be included in the study. The differences between Neandertals and contemporary modern humans were of primary interest to this study. However, in order to assess the phylogenetic significance of these differences data from small samples of other fossil hominins (Homo erectus, early anatomically modern humans and Upper Paleolithic modern humans) were also included. Table 1 provides a list of the fossil and contemporary modern human samples used in this study. For the purpose of determining trait polarity Homo erectus is assumed to represent the primitive condition. Although some scholars believe the African and Asian Homo erectus specimens represent two distinct species (Homo ergaster and Homo erectus, respectively, e.g. Wood, 1994), not everyone agrees (see Turner and Chamberlain, 1989; Bra¨uer and Mbua, 1992; Harrison, 1993). African and Asian Homo erectus specimens are pooled together (as Homo erectus sensu lato) in this study to provide a larger sample size for comparative purposes.

187

the tooth was in its approximate anatomical position. Loose teeth were mounted on modeling clay, and for teeth in situ the cranium or mandible was manipulated so that the cervical line(s) of the particular tooth was in the proper position. Cusp apices were marked with a soft pencil. A millimeter scale, placed at the same horizontal plane as the buccal cusp apices, was included in each photograph for calibration. SigmaScanÒ Pro 5.0 (ÓSPSS, Inc.) imaging software was used to take linear, angular and area measurements from the occlusal photographs. Each image was calibrated three times: along the mesial, central, and distal portions of the tooth (Fig. 4). The final calibration used for measurements was based on the average of the three values. Each tooth was then measured twice and the average of the two measurements was used in the analysis. Overall, measurement error was approximately 3%.1 Cusp angles were measured by connecting cusp the apices of the four major cusps (see Figure 3). SigmaScan ProÓ automatically calculated angles and distances between cusps. Angles were converted into radians for statistical analyses. Individual cusp base areas were measured by tracing along the tooth perimeter and the major fissures separating the cusps. In a case where the fissure did not extend to the tooth margin, I projected its course by following the direction of the fissure before it became obscured. Most teeth were minimally worn but, where necessary, I made corrections to interproximal wear in a manner similar to that of Wood and Engleman (1988) by estimating the original mesial or distal borders

Methods High-resolution images of the occlusal surface of M1s were taken with a Nikon CoolPix 950 digital camera using the macro setting. A small aperture was used to provide the greatest depth of field. The camera was mounted on a tripod and a leveling device was used to maintain a consistent camera angle. The camera’s LCD monitor was used to compose images and control for camera parallax. Each tooth was positioned so that the buccal and, where possible, distal cervical line were perpendicular to the camera’s focal point and

1 This represents the amount of measurement and calibration error only. An earlier analysis used 2-D and 3-D methods (photograph vs. point digitizer) to assess measurement error related to cusp tip identification (Bailey, 2002b). On average, measurement error was about 4%. More recently, a study of interobserver error in measuring cusp base areas indicates that measurement error due to tooth positioning and measuring technique is comparable to intraobserver error, or about 2% (Bailey et al, accepted for publication). Because tooth wear can make it difficult to identify cusp tips, only teeth that exhibited very little wear were used in the analysis of cusp angles and occlusal polygon area.

188

S.E. Bailey / Journal of Human Evolution 47 (2004) 183e198

Table 1 Samples used in this study and their composition* Group Homo erectus (n Z 3) Neandertals (n Z 15) (NEA)

Early anatomically modern humans (n Z 3) (EAMH) Upper Paleolithic modern humans (n Z 6) (UPMH) Contemporary modern humans (N Z 62) (CMH)

*

Individuals/Population Zhoukoudian L140, KNM-WT15000, KNM-ER 807 Krapina D100, D101, D171, D134, MxC MxD, MxA, Le Moustier, Pontnewydd 1216, 4, La Quina 18, Ku˚lna, Tabun C1, Taddeo, Saccopastore 2 Qafzeh 11,9,5

Dolnı´ Veˇstonice 14, Gough’s Cave 1, La Madeleine, Mladecˇ 1 & 2, Vachons North Africa (n Z 6) West Africa (n Z 13) Europe (n Z 8) Northeast Asia (n Z 10) India (n Z 3) Australasia (n Z 22)

Not all individuals were used in all analyses

marginal tubercles) the primary fissure was extended to the tooth margin and the appropriate proportions of the area of any additional cusps were added to the areas of the adjacent main cusps. In the case of Cusp 5, for example, its area was divided between the hypocone and metacone. The total crown base area was calculated by summing individual cusp base areas. Relative cusp base areas were calculated by dividing the actual cusp base area by the total crown base area. All measurements were rounded to the nearest tenth of a millimeter Estimating the degree to which tooth cusps were internally compressed involved calculating an occlusal polygon area (defined by lines connecting cusp apices) and dividing it by the total crown base area (see Fig. 3). Teeth that have more internally placed cusps will have smaller relative occlusal polygon areas than those with more externally placed cusps. This is the first time this observation has been quantified in this way.

based on the buccolingual extent of the wear facet and shape of the tooth (Fig. 5). Following the protocol of Wood and Engleman (1988), where additional cusps were present (e.g., Cusp 5 and

Fig. 4. The SigmaScan ProÓ measuring tool was calibrated three times along the mesial, central and distal aspects of the tooth. The average of the three calibrations was used to measure each tooth.

Fig. 5. Saccopastore M1 (left) illustrating the most worn of the molars used in the analysis, which still maintains fissures between major cusps. Outlines illustrate individual cusp base areas and how worn teeth were corrected for interproximal wear. (Note: This tooth was not used in the analysis of cusp angles and occlusal polygon area).

189

S.E. Bailey / Journal of Human Evolution 47 (2004) 183e198

differed significantly from the Qafzeh sample (representing early anatomically modern humans) for angle C only (p ! .05). Angle A, represented by the cusp apices of the paracone, protocone and hypocone, is the most conservative and differences among groups are statistically insignificant. Small sample sizes precluded including other fossil hominin groups in the significance tests. A principal component analysis (PCA) was undertaken to assess variation within samples. The PCA of cusp angles indicates that the first two components account for approximately 90% of the total variance (Table 4). Angles A and C contribute positively to the first principal component (PCI) while angles B and D contribute negatively. This indicates that PCI describes the buccolingual skew of the rhomboid. For the second principal component (PCII), angles A and B contribute positively while C and D contribute negatively. Since angles A and B represent the mesial portion of the tooth and angles C and D represent the distal portion, PCII represents the degree to which the shape of the occlusal polygon is narrower distally and wider mesially. When the individual specimens are projected on the factor plane for PCI and PCII it is apparent that teeth at the opposite poles of PCI and PCII have very different shapes (Figure 6). Teeth toward the negative pole of PCI are less skewed in outline and teeth toward the positive pole are more skewed. Along PCII, teeth that have a larger distal breadth occupy a positive position and teeth that have a narrower distal breadth occupy a negative position. The groups separate out fairly well along PCI with Neandertals falling predominantly

Statistical analysis Kruskal Wallis (a non-parametric alternative to the ANOVA) and multiple Mann-Whitney-U (a non-parametric alternative to the Student’s t-test) tests were used to examine the significance of the differences in cusp angles and relative cusp base areas among groups. Because multiple comparisons were made, the Bonferroni Correction was used to protect against Type 1 errors and maintain a table-wide alpha level of .05. However, some researchers feel that the Bonferroni Correction is an inappropriate method for this kind of data (e.g., Lockwood, pers. comm. 2002) or that the costs (increase in Type II error, for example) do not outweigh the benefits (e.g., Perneger, 1998). Therefore, tables also indicate where values are significant at uncorrected .05 and .01 levels.

Results Cusp angles All fossil hominins and contemporary modern humans are characterized by rhomboid-shaped M1s (i.e., angles A and C are larger than angles B and D). However, angle C is relatively larger and angles B and D are relatively smaller in Neandertals than all other groups (Table 2). A MannWhitney-U test confirms significant differences between Neandertals and contemporary modern humans for three of the four angles (Table 3). These two groups are significantly different in their mean values for angles B, C and D. Neandertals

Table 2 Descriptive statistics for M1 cusp angles in fossil and contemporary modern humans M1 Angle A B C D

EAMH

Homo erectus*

NEA

UPMH

CMH

n

X

SD

n

X

SD

n

X

SD

n

X

SD

n

X

SD

2 2 2 2

105.3 75.2 96.8 82.0

e e e e

3 3 3 3

106.0 72.6 104.9 76.5

9.7 1.4 3.2 8.9

10 10 10 10

106.4 65.1 120.9 67.7

5.0 6.9 10.1 7.1

2 2 2 2

105.8 70.6 110.3 73.3

e e e e

24 24 24 24

101.3 74.2 106.1 78.4

10.1 4.0 5.5 7.7

Abbreviations: EAMH e Early anatomically modern humans, NEA e Neandertals, UPMH e Upper Paleolithic modern humans, CMH e Contemporary modern humans Dashes indicate sample sizes were too small to calculate standard deviation. * Includes Zhoukoudian L140 and KNM-WT 15000.

190

S.E. Bailey / Journal of Human Evolution 47 (2004) 183e198

Table 3 Significance of between group comparisons of M1 cusp angles based on Mann Whitney U non-parametric tests !A EAMH EAMH NEA CMH

e N.S. N.S.

NEA e N.S.

!B CMH

EAMH

e

e N.S. N.S.

NEA e ***

!C CMH

EAMH

e

e * N.S.

NEA e ***

!D CMH

EAMH

NEA

CMH

e

e N.S. N.S.

e ***

e

N.S., not significant; *significant at p ! .05; **significant at p ! .01; ***significant at p ! .003 (Bonferroni correction) Abbreviations: EAMH e Early anatomically modern humans, NEA e Neandertals, UPMH e Upper Paleolithic modern humans, CMH e Contemporary modern humans

toward the positive pole and contemporary modern humans falling toward the negative pole (albeit with some overlap). An examination of the distribution of individuals along PCI indicates that Neandertal M1s are more skewed while contemporary modern humans are more rectangular (angles are closer to 90 degrees). Along PCII Neandertals tend to fall toward the negative pole indicating that the M1 occlusal polygon is more narrow distally than it is mesially. Contemporary modern humans, on the other hand, are scattered along PCII with little pattern to their distribution. Two of the three early anatomically modern humans fall closer to the contemporary modern human specimens than they do to Neandertals along PCI and the third is quite unlike Neandertals along PCII, possessing an occlusal polygon that has subequal mesial and distal components. The sample sizes for Homo erectus and Upper Paleolithic modern humans were too small to be included in the statistical analysis; however, Figure 6 shows individuals plotted on the PCI and PCII axes. The two Homo erectus specimens fall closer Table 4 Principal component analysis of cusp angles M1 PCI

PCII

Eigenvalues % variance

2.6 63.9

1.08 27.1

Eigenvectors Angle A Angle B Angle C Angle D

0.64 -0.86 0.76 -0.90

0.76 0.26 -0.62 -0.24

PCI and PCII account for 90% of the total variance.

to the contemporary modern human specimens than they do to Neandertals, while the single Upper Paleolithic specimen falls in the area of overlap between contemporary modern humans and Neandertals. From this analysis it appears that the occlusal polygon shape is a useful tool for discriminating between Neandertals and contemporary modern humans. To test this hypothesis, a discriminant analysis of the four angles was undertaken. This analysis shows a good separation of groups (Table 5). Correct allocation to groups was obtained 92% of the time, with the classification of contemporary modern humans somewhat higher (96% correct) than that of Neandertals (83% correct). When analyzed individually, Angle C appears to correctly classifying individuals most often (83% of Neandertals and 100% of contemporary modern humans), while Angle A provides no discrimination between groups. Cusp base areas The pattern of relative cusp sizes for contemporary modern humans found in this study (protocone O paracone O metacone O hypocone) confirm that of other researchers (Carlsen, 1987). This pattern is found in each of the contemporary modern human samples used in this study, regardless of geographic origin. In Neandertals, however, the modal pattern (14/16) of relative cusp size is slightly different (protocone O paracone R hypocone O metacone). This pattern reflects the relatively large hypocone and relatively small metacone in Neandertals (Table 6). A MannWhitney-U test confirms that these differences

191

S.E. Bailey / Journal of Human Evolution 47 (2004) 183e198 3

H. erectus Early H. sapiens Upper Paleolithic Neandertal Contemporary H. sapiens

2

PCII: 32.9%

1

0

-1

-2

-3 -0.2

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

PCI: 57.1% Fig. 6. Graphic results of a principal component analysis of cusp angles. Individuals at the extremes are plotted to illustrate differences in tooth shapes for PCI and PCII.

between Neandertals and contemporary modern humans are significant for these two cusps (Table 7). A Principal component analysis was undertaken to assess variation within the samples. When actual cusp size is used, PCI accounts for approximately 76% of the total variance. The eigenvectors have the same sign and approximately equal weight is given to each of the cusps suggesting that the major separating effect is size. When relative cusp size is used, eigenvectors for PCI and PCII have both high and low and positive and negative scores indicating that they contain information about tooth shape. The first two

Table 5 Result of discriminant function analysis using cusp angles

Contemporary modern humans Neandertals Total

Percent correct

Contemporary modern humans

Neandertals

96

23

1

83 92

2 25

10 11

components account for approximately 74% of the total variance (Table 8). An examination of eigenvectors indicates that the cusp base areas that make up the trigone (protocone, paracone, metacone) contribute positively to PCI, while the hypocone contributes negatively. For PCII both the metacone and paracone contribute negatively (the metacone much more so than the paracone), while the protocone contributes positively to the variance. These differences correspond to differences in the buccal and lingual portions of the tooth respectively. When specimens are plotted along PCI and PCII (Fig. 7) the distribution is somewhat scattered. However, there are some trends of note. Neandertals tend to fall toward the negative pole of PCI, while contemporary modern humans tend to fall toward the positive pole. This reflects the relatively larger contribution the hypocone makes to total crown base area in Neandertals and the relatively larger contribution the trigone (especially the paracone) makes to total crown base area in contemporary modern humans. Early

192

S.E. Bailey / Journal of Human Evolution 47 (2004) 183e198

Table 6 Descriptive statistics for relative cusp base areas in M1 in fossil and contemporary modern humans Protocone

Homo erectus Early anatomically modern humans Neandertal Upper Paleolithic modern humans Contemporary modern humans

Paracone

Metacone

Hypocone

n

X

SD

n

X

SD

n

X

SD

n

X

SD

3 3

31.4 29.5

1.1 3.1

3 3

26.0 23.5

2.3 0.8

3 3

21.4 19.6

3.3 0.1

3 3

21.2 27.4

2.4 2.9

15 6

29.6 30.7

2.6 1.7

15 6

25.4 25.9

2.3 3.9

15 6

21.1 23.5

1.7 2.2

15 6

23.9 20.0

2.2 4.0

62

31.0

2.1

62

25.8

2.1

62

22.9

1.9

62

20.3

2.4

these differences. Information on the relative cusp base areas of other cusps is less useful for discriminating among samples.

anatomically modern humans, who possess the largest hypocone of all groups, occupy a position opposite that of contemporary modern humans along PCI and fall with Neandertals toward the negative pole. Homo erectus and Upper Paleolithic modern human specimens are encompassed within the variation of Neandertals and contemporary modern humans. Along PCII there is a complete intermixing of groups. Differences in relative cusp base area among fossil hominin and modern groups can be largely attributed to differences in the relative size of the distal portion of the tooth. The hypocone and metacone appear to be the primary contributor to

The relationship between cusp angles and cusp base areas Because both the relative cusp size and the cusp angle of the hypocone and metacone were found to differ significantly between Neandertals and contemporary modern humans it was of interest to investigate the relationship of these two factors. Low to moderate negative correlations were found between cusp base areas and cusp angles. In

Table 7 Significance of between group comparisons of M1 relative cusp base areas based on Mann-Whitney-U non-parametric tests Protocone HE HE EAMH NEA UPMH CMH

e N.S. N.S. N.S. N.S.

EAMH e N.S. N.S. N.S.

NEA

e N.S. N.S.

Paracone UPMH

e N.S.

CMH

HE

EAMH

NEA

UPMH

CMH

e

e * N.S. N.S. N.S.

e N.S. N.S. *

e N.S. N.S.

e N.S.

e

Metacone HE HE EAMH NEA UPMH CMH

e N.S. N.S. N.S. N.S.

EAMH e * * **

NEA

e * ***

Hypocone UPMH

e N.S.

CMH

HE

EAMH

NEA

UPMH

CMH

e

e * N.S. N.S. N.S.

e N.S. * **

e * ***

e N.S.

e

N.S., not significant; * significant at p ! .05; ** significant at p ! .01; *** significant at p ! .005 (Bonferroni correction). Abbreviations: EAMH e Early anatomically modern humans, NEA e Neandertals, UPMH e Upper Paleolithic modern humans, CMH e Contemporary modern humans

193

S.E. Bailey / Journal of Human Evolution 47 (2004) 183e198 Table 8 Principal components analysis of actual and relative cusp base areas in M1 Actual Cusp base area

Eigenvalues % variance Eigenvectors Protocone Paracone Metacone Hypocone

PCI

PCII

PCI

PCII

3.0 75.6

0.47 11.8

1.77 44.2

1.18 29.6

-0.52 -0.51 -0.51 -0.46

0.12 0.30 0.35 -0.88

0.32 0.48 0.27 -0.77

0.73 -0.12 -0.67 -0.00

Neandertals these were much lower than in contemporary modern humans: -0.24 vs. -0.44 for hypocone and -0.06 vs. -0.35 in metacone. This suggests that the relationship of these two factors is different in the two groups. A linear regression was applied to the data to test the hypothesis that cusp angle can be predicted from cusp size. The jackknife procedure was used

Relatively larger lingually

0

Relatively larger buccally

2

PCII: 29.6%

3

Relative Cusp base area

to test the significance of the regression coefficients. This procedure involves resampling from the original sample(s) by eliminating one of the values or cases each time. The test statistic is recalculated each time and the mean value is used. This resampling technique showed that, although cusp sizes and cusp angles are negatively correlated, within-taxon correlations were insignificant

H. erectus Early H. sapiens Upper Paleolithic Neandertal Contemporary H. sapiens

1

-1

-2

-3 -4

-3

Relatively larger hypocone

-2

-1

PCl: 44.2%

0

1

2

3

Relatively larger trigone

Fig. 7. Graphic results of a principal component analysis of relative cusp base areas. Individuals to the right of the graph have a relatively larger paracone and relatively smaller hypocone compared to those on the left.

194

S.E. Bailey / Journal of Human Evolution 47 (2004) 183e198

and cusp size cannot be used to accurately predict cusp angle. Relative occlusal polygon area The relative sizes of the occlusal polygon area are presented in Table 9. It is clear that the M1 occlusal polygon area in the Neandertal sample is considerably smaller than it is in all other groups. A Kruskal-Wallis test revealed significant differences among groups. Subsequent Mann-Whitney-U tests of pairs showed that the occlusal polygon area of the Neandertal M1 is significantly different from that of both early anatomically modern humans and contemporary modern humans (p ! .02). Due to small sample sizes (n Z 2), significance tests of the differences between Neandertals and Homo erectus and Upper Paleolithic modern humans could not be performed; however, inspection of their mean values shows that they are more similar to contemporary and early anatomically modern humans then they are to Neandertals (Table 9, Fig. 8). Discussion Although the groundwork for studying tooth shape using crown base areas and/or cusp angles has been around since the 1960s (Erdbrink,

Table 9 Relative occlusal polygon areas among human groups M1

Homo erectus Early anatomically modern humans Neandertal Upper Paleolithic modern humans Contemporary modern humans *

No

Mean

SD

Range

2 3

32.9 33.1

e 3.5

30.8,35.0 29.6-36.6

12 2

26.8* 34.3

1.8 e

24.5-30.5 31.8,36.8

24

37.5

5.4

27.0-50.4

This value is significantly different from that of early anatomically modern humans and contemporary modern humans at p ! .02. Homo erectus and Upper Paleolithic specimens could not be assessed for statistical significance; however, their means most closely approximate that of early anatomically modern humans

1967; Biggerstaff, 1969; Corruccini, 1977; Wood and Abbott, 1983; Wood et al., 1983; Morris, 1986; Wood and Engleman, 1988; Mayhall, 1991) little has been done to apply these methods to Middle to Late Pleistocene hominin fossils. This study quantified M1 crown shape using cusp angles, cusp base areas and placement of cusp apices. As a result, significant shape differences between Neandertals and contemporary modern humans were revealed. Compared to the M1 of contemporary modern humans, the Neandertal tooth is more skewed. It appears that the position of one cusp e the metacone e is likely driving the shape differences. The angle that shows the greatest disparity among samples (angle C) reflects the relationship of this cusp to the paracone and hypocone. In Neandertals the metacone is shifted lingually and mesially relative to that in contemporary modern humans. This shift acts to increase this angle and decrease the angles of the hypocone (D) and paracone (B). Significant differences in relative cusp base areas were also revealed. In Neandertals the M1 possesses a very large hypocone and relatively small metacone compared to contemporary modern humans. While it is tempting to suggest that the differences in relative cusp base area are responsible for the differences in tooth shape, a regression analysis indicates that no significant relationship exists between cusp base areas and cusp angles. Consistent with the hypothesis of no relationship is the fact that specimens representing early anatomically modern humans and Homo erectus, who also possess a relatively large hypocone and relatively small metacone, do not exhibit the distinctively skewed shape of Neandertals. In addition, the two principal components analyses show that early anatomically modern humans fall within the Neandertal variation for relative cusp size, but fall closer to contemporary modern humans for cusp angles. This suggests that the distinctive Neandertal tooth shape cannot be attributed to the differences in relative size of the main cusps. The final measurement of crown shape involved quantifying cusp tip placement relative to the crown perimeter. Here, measurements revealed

195

S.E. Bailey / Journal of Human Evolution 47 (2004) 183e198 46 44 42 40 38 36 34 32 30 28 Mean

26

_ SE + _ SD +

24

Outliers

22

NEA

EAMH

H.erectus

UPMH

CMH

Fig. 8. Box and whisker plot showing relative occlusal polygon areas in contemporary modern humans and Middle-Late Pleistocene fossil hominins (see Table 9 for statistical analysis).

that the relative occlusal polygon area is smaller in Neandertals than in all other groups and that the differences between Neandertals and both early anatomically modern humans and contemporary modern humans are significant. This small occlusal polygon area reflects the degree to which the cusp apices are internally compressed (more so in Neandertals than other groups). The Neandertal problem, once a question of whether Neandertals evolved into (e.g., Hrdlicka, 1911; Weidenreich, 1943) or along side of (e.g., Howell, 1951; Sergi, 1958) modern humans, has shifted to be a question of the degree to which Neandertals contributed to modern human morphology. Supporters of the Recent African Origin hypothesis (Cann et al., 1987; Stringer and Andrews, 1988), who posit that anatomically modern humans evolved in Africa about 200 Kya, dispersed and replaced archaic populations elsewhere about 100 Kya, believe the genetic contribution of Neandertals (and other archaic populations) to anatomically modern humans is trivial at best (e.g. Stringer et al., 1984; Rak, 1986; Stringer and Bra¨uer, 1994; Schwartz and Tattersall, 1996; Stringer, 1996). On the other hand,

supporters of Multiregional Evolution, who posit that certain modern features evolved in different geographical regions and together contributed to what we view as ‘‘anatomically modern’’, believe that archaic humans in a particular region contributed significantly to the modern populations of that region (Wolpoff et al., 1994). Under this model Neandertals, the archaic predecessors to anatomically modern Europeans, are likely to have contributed significantly to modern human morphology in that region (Wolpoff et al., 1981; Smith, 1982; Frayer et al., 1993; Wolpoff et al., 2000). A central point of contention between competing hypotheses for modern human origins is whether or not Neandertals and anatomically modern humans were distinct species. The argument for specific distinction between the groups is based on evidence for a long separate evolutionary history (Krings et al., 1997), the degree of their morphological (Harvati, 2003b) and molecular (Krings et al., 1999, 2000) distinctiveness and the Neandertals’ possession of uniquely derived traits (autapomorphies) (Stringer et al., 1984; Tillier, 1989; Stringer, 1993; Rak et al., 1994; Schwartz and Tattersall, 1996).

196

S.E. Bailey / Journal of Human Evolution 47 (2004) 183e198

This examination of maxillary molar differences shows that, at minimum, Neandertal maxillary first molar morphology differs significantly from that of contemporary modern humans. The comparative analysis with other (albeit small) samples of fossil hominins suggests that beyond being distinctive, the cusp relationships seen in Neandertal maxillary molars (e.g., their internal placement and placement relative to each other) may be uniquely derived. If additional data gathered from larger fossil samples confirm this, we may consider adding these characters to the growing number of dental characters (e.g., Bailey, 2002a; Bailey and Lynch, in press) that support the uniquely derived nature of the Neandertal dentition and their designation as species in their own right (Homo neanderthalensis).

Acknowledgements I am grateful to all the people who provided access to, and assistance with, the fossils and contemporary specimens examined for the Ph.D. dissertation on which this paper is based: I. Tattersall, K. Mowbray and G. Schwartz at the American Museum of Natural History, New York; P. Ungar at the University of Arkansas, Fayetteville; C. Stringer and R. Kruzinski at the Natural History Museum, London; Y. Rak of the Sackler School of Medicine, Tel Aviv; G. Manzi at the University of Rome; G. Giacabini of the University of Turin; F. Mallegni of the University of Pisa; G. Koufos, of the Aristotle University of Thessaloniki; H. de Lumley at the Institute of Human Paleontology, Paris, Y. Coppens of the College of France, Paris; P. Tassy at the National Museum of Natural History, Paris; J. Leopold of the Museum of National Antiquities, St. Germainen-Laye; M. Tavaso and F. Marchal at the Laboratory of Historical Geology, Marseille; J-J. Cleyet-Merle and A. Morala at the National Museum of Prehistory, Les Eyzies; V. MerlinAnglade and G. Marchesseau at the Museum of Perigord; E. Ladier at the Museum of Natural History, Montauban; R. Ziegler at the National Museum of Natural History, Stuttgart; S. Dusek at the Museum of Prehistory, Weimar; H-E.

Joachim at the State Museum of the Rhine, Bonn; W. Menghin at the Museum of Prehistory, Berlin; N. Farsan of the University of Heidelberg; M. Teschler-Nicola, at the Natural History Museum, Vienna; R. Orban and P. Semal of the Royal Institute of Natural Sciences of Belgium, Brussells; J. Radovcˇic´ at the Croatian Natural History Museum, Zagreb; M. Paunovicˇ of the Institute for Quaternary Geology and Paleontology, Zagreb; J. Svoboda, of the Institute of Archaeology - Paleolithic and Paleoethnology Research Center, Dolnı´ Veˇstonice; and M. Dockalova at the Moravian Museum, Brno. I would also like to thank B.A. Wood, G.R. Scott and three anonymous reviewers for their helpful comments on the original manuscript. The data presented in this study were collected while the author was a PhD student at Arizona State University. Data collection and research were supported by grants from the National Science Foundation (BCS-0002481), LSB Leakey foundation and the Philanthropic Educational Organization (PEO).

References Bailey, S.E., 2000. Dental morphological affinities among late Pleistocene and recent humans. Dent. Anthropol. 14, 1e8. Bailey, S.E., 2002a. A closer look at Neanderthal postcanine dental morphology: I. The mandibular dentition. New Anat. 269, 148e156. Bailey, S.E., 2002b. Neandertal Dental Morphology: Implications for Modern Human Origins. Ph.D. Dissertation, Arizona State University, Tempe. Bailey, S.E., Lynch, J.M. Diagnostic differences in mandibular P4 shape between Neandertals and anatomically modern humans. Am. J. Phys. Anthropol., doi:10.1022/ajpa.20037, in press. Bailey, S.E., Pilbrow, V., Wood, B.A. Interobserver error involved in independent attempt to measure cusp base areas of Pan M1s. J. Anat., accepted for publication. Bermu´dez de Castro, J.M., 1993. The Atapuerca dental remains. New evidence (1987-1991 excavations) and interpretations. J. Hum. Evol. 24, 339e371. Biggerstaff, R.H.L., 1969. The basal area of posterior tooth crown components: The assessment of within tooth variations of premolars and molars. Am. J. Phys. Anthropol. 31, 163e170. Boule, M., Vallois, H.V., 1957. Fossil Men. Dryden Press, New York.

S.E. Bailey / Journal of Human Evolution 47 (2004) 183e198 Brace, C.L., 1967. Environment, tooth form and size in the Pleistocene. J. Dent. Res. 46, 809e816. Bra¨uer, G., Mbua, E., 1992. Homo erectus features used in cladistics and their variability in Asian and African hominids. J. Hum. Evol. 22, 79e108. Butler, P.M., 1963. Tooth morphology and primate evolution. In: Brothwell, D.R. (Ed.), Dental Anthropology. Pergamon Press, New York, pp. 1e13. Bytnar, J.A., Trinkaus, E., Falsetti, A.B., 1994. A dental comparison of Middle Paleolithic near eastern hominids. Am. J. Phys. Anthropol. (Suppl. 19), 63. Cann, R., Stoneking, M., Wilson, A., 1987. Mitochondrial DNA and human evolution. Nature 325, 31e36. Carlsen, O., 1987. Dental morpholopgy. Munksgaard, Copenhagen. Corruccini, R.S., 1977. Crown component variation in the hominoid second premolar. J. Dent. Res. 56, 1093e 1096. Erdbrink, D.P., 1967. A quantification of lower molar patterns in deutero-Malayans. Z. Morph. Anthrop. 59, 40e56. Frayer, D., Wolpoff, M., Thorne, A., Smith, F., Pope, G., 1993. Theories of modern human origins: the paleontological test. Am. Anthropol. 95, 14e50. Harrison, T., 1993. Cladistic concepts and the species problem in hominoid evolution. In: Kimbel, W.H., Martin, L.B. (Eds.), Species, Species Concepts and Primate Evolution. Plenum Press, New York, pp. 345e372. Harvati, K., 2003a. First Neanderthal remains from Greece: the evidence from Lakonis. J. Hum. Evol. 45, 465e473. Harvati, K., 2003b. The Neanderthal taxonomic position: models of intra- and inter-specific craniofacial variation. J. Hum. Evol. 44, 107e132. Howell, F., 1951. The place of Neanderthal man in human evolution. Am. J. Phys. Anthropol. 9, 379e416. Hrdlicˇka, A., 1911. Human dentition and teeth from the evolutionary and racial standpoint. Dom. Dent. J. 23, 403e417. Krings, M., Capelli, C., Tschentscher, F., Geisert, H., Meyer, S., von Haeseler, A., Grossschmidt, K., Possnert, G., Paunovic, M., Pa¨a¨bo, S., 2000. A view of Neandertal genetic diversity. Nat. Genet. 26, 144e146. Krings, M., Geisert, H., Schmitz, R., Krainitzki, H., Pa¨a¨bo, S., 1999. DNA sequence of the mitochondrial hypervariable region II from the Neandertal type specimen. Proc. Nat. Acad. Sci. 96, 5581e5585. Krings, M., Stone, A., Schmitz, R.W., Krainitzki, H., Stoneking, M., Pa¨a¨bo, S., 1997. Neandertal DNA sequences and the origin of modern humans. Cell 90, 19e30. Manzi, G., Passarello, P., 1995. At the Archaic/Modern Boundary of the genus Homo: The Neandertals from Grotta Breuil. Curr. Anthropol. 35, 355e366. Mayhall, J.T., 1991. Tooth size and the Carabelli Trait. Am. J. Phys. Anthropol. 84, 427e432. Morris, D.H., 1986. Maxillary molar occlusal polygons in five human samples. Am. J. Phys. Anthropol. 70, 333e338. Perneger, T.V., 1998. What’s wrong with Bonferroni adjustments. Br. Med. J. 316, 1236e1238.

197

Rak, Y., 1986. The Neandertal: a new look at an old face. J. Hum. Evol. 15, 151e164. Rak, Y., Kimbel, W., Hovers, E., 1994. A Neandertal infant from Amud Cave, Israel. J. Hum. Evol. 26, 313e324. Schwartz, J., Tattersall, I., 1996. Significance of some previously unrecognized apomorphies in the nasal region of Homo neanderthalensis. Proc. Nat. Acad. Sci. 93, 10852e10854. Scott, G.R., Turner II, C.G., 1997. The Anthropology of Modern Human Teeth. Dental Morphology and its Variation in Recent Human Populations. Cambridge University Press, Cambridge. Sergi, S., 1958. Die Neandertalischen Paleanthropen in Italien. In: von Koenigswalde, G. (Ed.), Hundert Jahre Neanderthaler. Utrecht, pp. 38e51. Smith, F.H., 1982. Upper Pleistocene hominid evolution in south-central Europe: a review of the evidence and analysis of trends. Curr. Anthropol. 23, 667e703. Stringer, C., 1993. Understanding the fossil human record: past, present and future. Rev. di Anthropol. (Roma) 71, 91e100. Stringer, C., 1996. Current issues in modern human origins. In: Meikle, W., Howell, F., Jablonski, N. (Eds.), Contemporary Issues in Human Evolution. California Academy of Sciences, San Francisco, pp. 116e134. Stringer, C., Andrews, P., 1988. Genetic and fossil evidence for the origin of modern humans. Science 239, 1263e1268. Stringer, C., Bra¨uer, G., 1994. Methods, misreading and bias. Am. Anthropol. 96, 416e424. Stringer, C., Hublin, J., Vandermeersch, B., 1984. The origin of anatomically modern humans in Western Europe. In: Smith, F., Spencer, F. (Eds.), The Origins of Modern Humans: A World Survey of the Fossil Evidence. Alan, R. Liss, New York, pp. 51e135. Tattersall, I., Schwartz, J.H., 1999. Hominids and hybrids: The place of Neanderthals in human evolution. Proc. Nat. Acad. Sci. 96, 7117e7119. Tillier, A.-m., 1989. The evolution of modern humans: evidence from young Mousterian individuals. In: Mellars, P., Stringer, C. (Eds.), The Human Revolution: Behavioral and Biological Perspectives in the Origins of Modern Humans. Edinburgh University Press, Edinburgh, pp. 286e297. Trinkaus, E., 1978. Dental remains from the Shanidar Adult Neanderthals. J. Hum. Evol. 7, 369e382. Turner, A., Chamberlain, A., 1989. Speciation morphological change, and the status of African Homo erectus. J. Hum. Evol. 18, 115e130. Turner II, C.G., Nichol, C.R., Scott, G.R., 1991. Scoring procedures for key morphological traits of the permanent dentition: The Arizona State University Dental Anthropology System. In: Kelley, M., Larsen, C. (Eds.), Advances in Dental Anthropology. Wiley Liss, New York, pp. 13e31. Uchida, A., 1998a. Variation in tooth morphology of Gorilla gorilla. J. Hum. Evol. 34, 55e70. Uchida, A., 1998b. Variation in tooth morphology of Pongo pygmaeus. J. Hum. Evol. 34, 71e79.

198

S.E. Bailey / Journal of Human Evolution 47 (2004) 183e198

Weidenreich, F., 1943. The ‘‘Neanderthal Man’’ and the ancestors of Homo sapiens. Am. Anthropol. 45, 39e48. Wolpoff, M., Frayer, D., Oliva, M., Jelinek, J., 2000. The Mladecˇ males: Aurignacian crania from Moravia. Paper presented at the annual meetings of the Paleoanthropology Society, Philadelphia, PA. Wolpoff, M., Smith, F., Malez, M., Radovcic, J., Rukavina, D., 1981. Upper Pleistocene hominid remains from Vindija Cave, Croatia, Yugoslavia. Am. J. Phys. Anthropol. 54, 499e545. Wolpoff, M., Thorne, A., Smith, F., Frayer, D., Pope, G., 1994. Multiregional Evolution: A World-wide source for modern human populations. In: Nitecki, M., Nitecki, D. (Eds.), Origins of Anatomically Modern Humans. Plenum Press, New York, pp. 176e199. Wolpoff, M.H., 1971. Metric Trends in Hominid Dental Evolution. Press of Case Western Reserve University, Cleveland.

Wood, B.A., 1994. Taxonomy and evolutionary relationships of Homo ergaster. Cour. Forschungstinst. Senckenb. 171, 159e165. Wood, B.A., Abbott, S.A., 1983. Analysis of the dental morphology of Plio-Pleistocene hominids: I. Mandibular molars: crown area measurements and morphological traits. J. Anat. 136, 197e219. Wood, B.A., Abbott, S.A., Graham, S.H., 1983. Analysis of the dental morphology of Plio-Pleistocene hominids: II. Mandibular molars e study of cusp areas, fissure pattern and cross sectional shape of the crown. J. Anat. 137, 287e 314. Wood, B.A., Engleman, C.A., 1988. Analysis of the dental morphology of Plio-Pleistocene hominids: V. Maxillary postcanine tooth morphology. J. Anat. 161, 1e35. Wood, B.A., Uytterschaut, H., 1987. Analysis of the dental morphology of Plio-Pleistocene hominids: III. Mandibular premolar crowns. J. Anat. 154, 121e156.