AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 58:75-79 (1982) Discrete Dental Variations and Biological Distances of Nubian Populations DAVID L E E GREENE Department of Anthropology, university of Colorado, Boulder. Colorado 80309 KEY WORDS Discrete dental variations, biological distance, divergence, Smith-Grewal statistic, Nubia, Wadi Halfa, Kulubnarti ABSTRACT During the 1963-1964 field season, the University of Colorado’s Nubian Expedition excavated a series of Meroitic, X-Group, and Christian cemeteries from Wadi Halfa Sudan. Recently a joint expedition sponsored by the Universities of Colorado and Kentucky excavated two additional Christian cemeteries from Kulubnarti, some 80 miles south of Wadi Halfa. Earlier analysis of discrete dental variations demonstrated that the Wadi Halfa populations did not differ significantly from one another. Application of the Smith-Grewal multivariate measure of biological divergence, as modified by Sjovold, and Green and Suchey, corroborates the original conclusion of biological stability and continuity for the Wadi Halfa populations, as well as demonstrating that the Kulubnarti populations are part of that pattern. None of the populations are significantly different from one another. In 1979 the University of Colorado, in conjunction with the University of Kentucky, excavated two Christian cemeteries located near the Medieval Christian stronghold of Kulubnarti, some 80 miles south of Wadi Halfa along the Nile in the region of Nubia known as the Batn-el-Hajr, or Belly of Stones (Van Gerven, 1981).Skeletons from these cemeteries (Table 1) add to the population samples from Nubia previously collected by the University of Colorado’s Nubian Expedition in the Wadi Halfa region of the Sudan (Armelagos et al., 1965). Analyses of the skeletal remains from the Wadi Halfa region have contributed to the understanding of the skeletal biology of Nubia as well as addressing major questions in paleodemography, paleoepidemiology, paleopathology, microevolution, and population relationships (e.g., Van Gerven et al., 1973; Armelagos, 1969; Swedlund and Armelagos, 1969; Greene, 1973, 1981). The majority of these investigations have depended upon the proposition that there was a considerable degree of genetic stability and continuity in the Wadi Halfa area through the sequential cultural phases labeled by archaeologists as Meroitic, X-Group, and Christian periods. However, prior to the late 1960s most 0002-948318215801-0075$02.000 1982 ALAN R. LISS, INC. specialists had argued that instead of genetic stability and continuity, the area was, to the contrary, characterized by major population change brought about by the movement of the X-Group people into the area as the preceding Meroitic culture declined (Elliot-Smith and Wood-Jones, 1910; Batrawi, 1935). The X-Group people were thought to be more Negroid than the Meroitics and thus introduced different genetic material through hybridization with the remnant Meroitics. Therefore, following this older perspective, one would expect to find differences in the expression of traits with significant genetic components when contrasting Meroitics with X-Group people or X-Group people with those of the Christian era. This older perspective, genetic change produced by major migration and hybridization, was initially tested by Greene (1966, 1967a) in his analysis of discrete dental morphological traits from populations in the Wadi Halfa area. Using a series of univariate comparisons, he demonstrated that there were no significant differences in the occurrences of 16 permanent discrete dental morphological traits between Received May 8. 1981. accepted September 28.1981. 76 D.L. GREENE TABLE 1. Cemetery samples and their cultural affiliations Area Wadi Halfa Cemetery 6-B-16 NAX 6-K-3 Affiliation Meroitic X-Group Christian Dates' ? - 350 AD ca. ca. 350- 550 AD ca. 1050-1150 AD Kulubnarti 21-$46 21-R-2 Christian Christian ca. 550- 750 AD ca. 750-1450 AD 'Dates are for the approximate durations of the Meroitic and X-Group periods in the Wadi Halfa area (Greene. 1967a). The Christian cemetery 6-K-3 is dated following Adams (1967). The dates for the Kulubnarti cemeteries are from Van Gerven (1981). the populations. From this, he concluded that the Meroitic and X-Group populations are not representatives of different racial groups, but instead represent a continuum of successive local Mendelian breeding populations in the Wadi Halfa area (Greene, 1973, 1981). Greene's conclusions were reaffirmed by Van Gerven et al.'s (1976) discriminant function and principle components analysis of cranial metrics from the same Wadi Halfa populations. Also, other investigators such as Nielson (1970), using different population samples from Lower Nubia than those used by Greene and Van Gerven et al., argue as well for population continuity. Some investigators, though, still argue for discontinuity through racial displacement in Lower Nubia (Strouhal, 1971; Strouhal and Jungwirth, 1980). Thus the question of continuity versus discontinuity has not been completely resolved among those physical anthropologists working on skeletal materials from Nubia. Consequently, the analysis of new population data and a re-analysis of older data using more powerful techniques will help clarify the controversy. The purpose of this paper is twofold. First, in that a newer and potentially more powerful multivariate method, the Smith-Grewal statistic (Grewal, 1962),than the series of univariate comparisons originally used by Greene has been developed and tested with considerable success in assessing biological distance or divergence of populations outside of Nubia, it will be used to further test and corroborate Greene's original conclusion that the Wadi Halfa populations show a high degree of population stability and continuity. Second, the Smith-Grewal statistic will also be used to evaluate the degree of biological divergence between the Wadi Halfa populations and the new population samples from Kulubnarti, some 80 miles to the south. How similar are these populations? The answer to this question will help determine the direction of further research into the nature of biocultural evolution in Nubia. If the Kulubnarti populations diverge significantly from the Wadi Halfa populations, then researchers will have to evaluate the potential effects of population genetic differences on disease rates, mortality, and microevolutionary process. On the other hand, if the Kulubnarti populations do not diverge significantly from those of Wadi Halfa, then researchers will be able to assume genetic stability and continuity, not only for the Wadi Halfa area, but over a much larger area extending at least 80 miles further to the south. METHODS In 1967 Berry and Berry first applied the Smith-Grewal statistic to the determination of biological distance in human populations using discrete skeletal variants. Since that time it has been widely used as a non-metrical multivariate measure of distance. Most applications in physical anthropology have focused upon cranial variations (Finnegan and Cooprider, 1978),though at least one has utilized discrete morphological variants in the dentition (Berry, 1976). Sjgvold in 1973 reviewed the applicability of the Smith-Grewal statistic and further developed its theoretical rationale. Green and Suchey, in 1976, added additional theoretical understanding, and in particular demonstrated that the Freeman-Tukey transformation should be used to stabilize variances in small samples. Consequently in this study, each discrete trait frequency of occurrence was transformed using 8 = %sin-'(l -2k/(n+1)) + ?hsin-'(l-2(k+l)/(n+l)), where k equals the occurrence of a trait and n equals the total number of observations. Mean measures of divergence (or distance) were then found: 77 BIOLOGICAL DISTANCES OF NUBIAN POPULATIONS where r equals the number of pairs of traits considered, Oli is transformed frequency of the ith trait in the first population, nIi the total number of observations for the ith trait in the first population, O,, the transformed frequency of the ith trait in the second population, and nZithe total number of observations for the ith trait in the second population. The variance of a mean measure of divergence then equals: and the standard deviation equals the square root of the variance. The statistic D and its variance are based upon the assumption of statistical independence between all traits considered (Sjqvold, 1973). care must be taken when considering discrete traits which are bilaterally expressed to control for the potential confounding effects of symmetrical or asymmetrical expression (Sjqvold, 1973; Greene, 1967b). If there are significantly different expressions of a trait between the sexes in a population, then sexspecific comparisons should be made in calculating D. Otherwise the frequencies can be combined. In the earlier study of the Wadi Halfa populations (Greene 1967a), 16 dental traits were found to be statistically independent of one another, not significantly asymmetrical in bilateral expression, and not significantly different in distribution by sex. Due to small sample sizes, it is not possible to determine whether or not the same pattern of bilateral symmetry and the lack of differential expression by sex prevails in the Kulubnarti populations. Furthermore, even though some vari- ants are individually observable in fairly large numbers (Table 2), due to tooth loss and heavy wear the number of cases showing the simultaneous Occurrence of several different traits is not large enough to test statistically for independence. However, for purposes of this paper, it is assumed that the Wadi Halfa pattern of statistical independence, bilateral symmetry, and no difference in expression by sex also prevails in the Kulubnarti populations. Not all of the original 16 traits are appropriate for the current analysis. In the earlier study, all teeth were extracted and direct observations were made on molar root fusions and third molar agenesis as well as maxillary premolar root fusions. Similar extractions are not feasible at the current time for the Kulubnarti samples; and even if they were, there is some doubt that root fusions in particular can be scored objectively as discrete variations in a way comparable to the categories used in the earlier study since fusions do vary in a continuous way. The loss of third molar agenesis information is probably not serious in that even when it was used in the original study it added very little to the crude measure of distance then used. In addition, even though the same investigator (Greene ) made all of the observations in the Wadi Halfa and Kulubnarti samples, experience has shown that mandibular molar fissure patterns are most difficult to score consistently and therefore have been eliminated. Berry (1976), in comparing fissure observations made by different observers, came to the same conclusion. Cusp number on the first maxillary molars was not used since almost all of the Nubian populations have four cusps at frequencies near 100%. Therefore, this would have added little discriminatory power to the D estimates. Finally, for some of the traits that were in the early study categorized into more than present or absent classes, the finer categories were combined as present. Doing this, as well TABLE 2. Trait frequencies and Freeman-Tukey tramformations NAX 6-B-16 1) 2) 3) 4) 5) 6) 7) Wn 11/48 36157 28/36 32/36 41/48 16/34 22/48 e 0.5660 -0.2616 -0.5675 -0.8314 -0.7511 0.0572 0.0817 Wn 19/64 42163 38/47 29137 52161 29/44 24142 6-K-3 8 0.4108 -0.3341 -0.6417 -0.5814 -0.7914 -0.3163 -0.1401 Wn 20148 14/28 10113 58161 24131 8/18 43/62 2 1-S-46 8 0.1631 0.0000 -0.5224 -1.0083 -0.5568 0.1056 -0.3903 Wn 26/52 14/24 13/15 38/38 17/17 317 10121 8 0.0000 -0.1608 -0.7442 -1.1364 -1.0929 0.1263 0.0455 21-R-2 Wn 21/58 32149 34139 34/37 20124 9117 10126 8 0.2746 -0.3045 -0.7905 -0.9096 -0.6851 -0.0557 0.2242 78 D.L. GREENE TABLE 3. Distances (0) 6-B-16 6-B-16 NAX -0.0028 (0.0236) NAX 6-K-3 6-K-3 0.0155 (0.1103) 21-S46 0.0574 (0.0466) 21-R-2 -0.0299 (0.0307) 0.0212 (0.0311) 0.0108 (0.0447) 0.0061 (0.0291) -0.0053 (0.0054) 0.0192 (0.0388) -0.0422 (0.05941 21-546 21-R-2 Standard deviations are in parentheses. as making the observations between Wadi Halfa and Kulubnarti more readily comparable, also insures the independence of the traits. Otherwise, for example, if maxillary second molars were scored as 4, 4 -, 3 + , or 3, then several, depending upon views of degrees of freedom, statistically dependent variables would be included in D; pesence or absence of 4, presence or absence of 4 - , presence or absence of 3 + , and presence or absence of 3. Therefore, the traits used in the present study are as follows: 1)Carabelli’s Trait on Maxillary First Molar: Cusp, pit, and groove combined as present. Genetic studies indicate that cusps, pits, and grooves may be variable manifestations of the same genotype (Kraus, 1951; Tsuji, 1958).This is further supported in Nubian populations by bilateral concordance studies (Greene, 1967a).2) Four cusps on Maxillary Second Molar: The 4 and 4- categories used in the earlier study are combined since the distinction between the two can be obscured by relative tooth size (Greene, 1967a).3) Three cusps on Maxillary Third Molar: Both 3 and 3 categories scored as present. Occasionally 3 + + teeth (Greene, 1967a) occur here. The supernumerary cuspule in this type may be the metaconule recently reintroduced into the literature by Harris and Bailit (1980). 4) Five cusps on Mandibular First Molar: This also includes a very small number of six cusped molars (C6or C7).5) Four cusps on Mandibular Second Molar. 6) Four cusps on Mandibular Third Molar. 7)Shoveling on Central Maxillary Incisors: Hrdlicka’s trace or moderate shoveling are combined. Shoveling in Nubian populations rarely exceeds these levels. If central incisors were missing and lateral incisors could be observed, then the shoveling trait manifestation on the laterals was used for this observation, since the shoveling field in the incisors always produces shoveling in the later- + als if it exits in the centrals (Carbonell, 1963; Greene, 1967a). When both right and left teeth were present and observable, observations were made from the left teeth. When only one or the other was present and observable, it was used. Data from both sexes were combined. All observations were made on permanent teeth. And, finally, observations were restricted to those teeth with minimum wear, no greater than Brothwell’s Stage 3 (Brothwell, 1963). RESULTS The three largest cemetery samples from the Wadi Halfa area (6-B-16,NAX, and 6-K-3)were used in the earlier study to evaluate biological distance and are used here to confirm the results of that earlier study and to compare with the recently excavated cemetery samples from Kulubnarti (214-46 and 21-R-2).) As Sjqivold (1973) points out, the SmithGrewel measure of biological divergence (or distance) is a multivariate statistic that tests the hypothesis that the difference between two populations is no greater than zero. Consequently, D values can be negative as well as positive (Table 3). If the absolute value of a D value exceeds twice its standard deviation, then the difference is statistically significant at least at the 0.05 level (SjBvold, 1973).When the Freeman-Tukey transformation is used to calculate 0 as in this analysis, then if the D value exceeds twice its standard deviation the difference is statistically significant at somewhere between 0.05 and 0.03 (Green and Suchey, 1976). All of the D values are extremely low and none are even close to being significant. DISCUSSION Prior to Greene’s analysis of the Wadi Halfa populations, many archaeologists and physical BIOLOGICAL DISTANCES OF NUBIAN POPULATIONS anthropologists believed that the transition from the Meroitic cultural phase to the X-Group in Nubia was accompanied by a major infusion of people who were biologically different from the Meroitic people. Indeed, some (i.e. Strouhal, 1971)continue to argue that the X-Group people are far more negroid than are the Meroitics. On the other hand, none of the recent proponents of that opinion have incorporated the Wadi Halfa populations, or those from Kulubnarti, into their analysis, nor have their population samples been incorporated in this study. It is possible, however, that there may have been some infusion of more Negroid peoples into other parts of Nubia during the X-Group period. The present analysis does clearly corroborate the earlier study of the Wadi Halfa populations and, furthermore, demonstrates that the pattern of biological stability and continuity extends from Wadi Halfa to Kulubnarti. In this area of Nubia there is no evidence for the infusion of people with radically different biological backgrounds during the X-Group period. ACKNOWLEDGMENTS Research support for this paper was made possible by National Science Foundation grant no. BNS-78-00255,Co-principle Investigators W.Y. Adams, D.L. Greene, and D.P. Van Gerven. LITERATURE CITED Adams, WY (1967)Continuity and change in Nubian culture industry. Sudan Notes and Records 4O:l-32. Armelagos, G J (1969) Disease in ancient Nubia. Science 163:255-259. Armelagos, GJ, Ewing, GH. Greene. DL and Greene, KK (1965) Report of the physical anthropology section, University of Colorado Nubian Expedition. Kush 13:29-28. Batrawi, AM 1193.5) Report on the human remains. In Mission Archeologique de Nubie 1929-1934. Cairo: Govt. Press. Berry, AC (1976) The anthropological value of minor variants of the dental crown. Am. J. Phys. Anthropol. 45:257-267. Berry, AC and Berry, RJ (1967) Epigenetic variation in the human cranium J. Anat. 101:361-379. Brothwell, DR (1963) “Digging Up Bones.” London: British Museum (Natural History). Carbonell, VM (1963) Variations in the frequency of shovelshaped incisors in different populations. In “Dental 79 Anthropology,”Brothwell, DR ed., New York Pergamon, pp. 211-234. Elliot-Smith, G. and Wood-Jones, G (1910) Report of the human remains. In Archeological Survey of Nubia, Report for 1907-1908, 11. Cairo: Govt. Press. Finnegan, M and Cooprider, K (1978)Empirical comparison of distance equations using discrete traits. Am. J. Phys. Anthropol. 49:39-46. Green, RF and Suchey, JM (1976) The use of inverse sine transformations in the analysis of non-metric cranial data. Am. J. Phys. Anthropol. 4561-68. Greene, DL (1966) Dentition and the biological relationships of some Meroitic, X-Group and Christian populations from Wadi Halfa, Sudan. Kush 142285-288. Greene, DL (1967a) Dentition of Meroitic. X-Group and Christian Populations from Wadi Halfa, Sudan. Salt Lake City: University of Utah Anthropological Papers, No. 85. 71 pps. Greene, DL (1967b) “Genetic’s, Dentition and Taxonomy.” Laramie: University of Wyoming Publications, vol. 33, No. 2. 75 pps. Greene, DL (1973) Dental anthropology of early Egypt and Nubia. In Browthwell, DR and Chiarelli, BA. eds.. “Population Biology of the Ancient Egyptians.” London and New York Academic Press, pp. 315-324. Greene, DL (1981)A critique of methods used to reconstruct racial and population affinity in the Nile Valley. Bulletins et Memoires, Societe d’Anthropologic de Paris, in press. Grewal. MS (1962)The rate of genetic divergence of sublines in the C57BL strain of mice. Genet. Res. 3:226-237. Harris, FH and Bailit. HL (1980) The metaconule: a morphologic and familial analysis of a molar cusp in humans. Am. J. Phys. Anthropol. 53:349-358. Kraus, BS (1951) Carabelli’s anomaly of the maxillary molar teeth. Am. J. of Hum. Genet. 3:348-355. Nielson, OV (1970) “The Nubian Skeleton Through 4000 Years.” Denmark: Andelsbortrykkeriet i Odense. Sjqvold. T (1973) The occurence of minor non-metrical variants in the skeleton and their quantitative teatment for population comparisons. Homo 24:204-233. Strouhal. E (19711A contribution to the anthropologyof the Nubian X-Group. In “Anthropologial Congress Dedicated to Ales Hrdlicka,” Praha: Czechoslovak Academy of Sciences, pp. 541-547. Strouhal, E and Jungwirth J (1980) Paleopathology of the late Roman-early Byzantine cemeteries at Sayala, Egyptian Nubia. J. Hum. Evol. 9:61-70. Swedlund, AC and Armelagos. G J (1969) Une recherche en paleodemographie: La Nubie Soudanaise. Annales: Economie, Sociologie, Civilization 6:1287-1298. Tsuji, T (1958) Incidence and inheritance of the Carabelli’s cusp in a Japanese population. Jpn. J. Hum. Genet. 3.21. Van Gerven, DP (1981) Nubia’s last Christians: The cemeteries of Kulubnarti. Archaeology 3422-30. Van Gerven. DP, Carlson, DS and Armelagos, GJ (1973) Racial history and bio-cultural adaptation of Nubian archaeological populations. J. Afr. Hist. 14:555-564. Van Gerven, DP, Armelagos, G J and Rohr, A (1976) Continuity and change in cranial morphology of three Nubian archaeological populations. Man 12.270-277.