Population variation in tooth jaw and root size A radiographic study of two populations in a high-attrition environment.код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 79:197-206 (1989) Population Variation in Tooth, Jaw, and Root Size: A Radiographic Study of Two Populations in a High-Att rition Environment PATRICIA SMITH, YOCHANAN WAX, AND FANNY ADLER Department of Anatomy and Embryology (P.S.), and Department of Statistics (Y,W., F.A.), Hebrew University, Jerusalem, POB 11 72 Israel KEY WORDS: Teeth, Jaws, Roots, Mandible, Australia, New Zealand alveolar bone height ABSTRACT Radiographs were taken of the jaws of skeletal remains of two populations of different-phenotype Prehistoric Australians from Roonka and Early New Zealanders (Maoris). On these radiographs crown, root, and corpus size were measured. Corpus height was subdivided into alveolar bone height, defined a s the bone superior to the mandibular canal, and basal bone height, defined as that inferior to the mandibular canal. Both between and within the two populations there was a significant and negative correlation between crown size and corpus height. The differences between the two populations in corpus height were associated with differences in alveolar bone height rather than basal bone height and support hypotheses associating continued eruption of adult teeth with growth of the alveolar bone. The findings also support previous studies that have shown only a low correlation between crown size, root size, and corpus height. In studies of tooth-jaw relations in human populations, tooth size is usually defined in terms of crown size, with no reference to root size or form, although the two show only a low correlation (Garn et al., 1978; Rosing, 1983; Smith et al., 1986). Moreover, neither root size nor crown size appears to be highly correlated with jaw size in recent populations (Anderson et al., 1977; Moore et al., 1968; Lavelle, 1962). In investigations of the functional significance of evolutionary trends in the hominid dentofacial complex, more attention then needs to be paid to the relation between the size and proportions of the crown, roots, and jaws. It must also be recognized that the relationship between these three variables is dynamic and changes throughout life. With function, the crowns of the teeth reduce in size, through occlusal and interproximal attrition, and the amount of tooth material so lost varies with the diet and other activities that cause dental attrition (Molnar, 1971; Smith, 1984). In contrast the roots normally show little change in size with age or function, although in rare cases root resorp- 0 1989 ALAN R. LISS, INC. tion or excessive deposition of cementum may occur (Gustafson, 1950). The latter has been described a s relatively common in Eskimos, where it has been attributed to excessive forces exerted on the affected teeth (Pedersen, 1949).Corpus height may increase or decrease throughout adult life, depending on the extent of increase in alveolar bone deposition with continued eruption of the teeth (Murphy, 1959; Ainamo and Talari, 1975), or alveolar resorption in the presence of periodontal disease (Loe et al., 1979). Data on Near Eastern populations suggest that crown size, root size, and corpus height show only a low correlation (Smith et al., 1986). However, it may be argued that the majority of the populations examined in this study were utilizing relatively soft and nonabrasive foods, so that the findings are not typical of populations in whom functional demands on the dentition are high. The present study was designed to examine these assumptions, by investigating the associaReceived August 7, 1987; revision accepted August 3, 1988. 198 P. SMITH ET AL. tion between crown, root, and corpus height dimensions in prehistoric Australians (Aboriginals from Roonka) and New Zealanders (Maoris). Both suffered rapid and severe attrition but differed in dietary base and food preparation techniques. They provide a good case study of the extent and pattern of phenotypic variation in the teeth, roots, and jaws and the effects of severe attrition on corpus height. MATERIALS AND METHODS This study was carried out on the mandibles of skeletal samples of prehistoric Australians and New Zealand Maori. The Australian sample was excavated from Roonka, a site on the Murray river in southeastern Australia and dates from 7000 BP to 1800 AD (Pretty, 1977). It is housed in the South Australian Museum, Adelaide, Australia. The Australians were hunters and gatherers, exploiting riverine resources that included fish, shellfish, and birds as well as marsupials and plants (Pretty, 1977). The individuals from Roonka were of moderate stature (males 167 cm) with dolichocranic heads and relatively short faces and small jaws (Prokapec, 1979). Their teeth were exceptionally large, with severe attrition and little caries or periodontal disease (Smith et al., 1988).Although the sample covers a period of some 7000 years, it shows no unidirectional diachronic trends in skeletal morphometrics or pathology, suggesting little change in the gene pool or environmental stress over the entire period. The Maori sample, housed at the University of Otago, Duneidin, New Zealand, covers a shorter time span (circa 1100-1900 AD) but is more heterogeneous since it is largely composed of individual specimens from different sites within New Zealand. The population sampled was tall with exceptionally large faces, large “rocker” mandibles, characteristic of Polynesians, and teeth of moderate size. They were horticulturalists, but their diet included seals, fish, and shellfish, as well as wild birds and plants, and especially fern root, which caused rapid and severe dental attrition (Taylor, 1963; Shawcross, 1967; Houghton, 1978a, 1980). Criteria for inclusion of a specimen in the study was the presence of a t least three teeth (premolars or permanent molars) with fully formed roots. Age and sex were determined by examination of the skeleton (Krogman, 1962) and compared with that recorded by Prokapec (1979) for the Australians and by Houghton (unpublished catalogue) for the New Zealanders. Only a few female specimens fitted the criteria for inclusion in this study, so analysis was restricted to male specimens. Corpus height was measured directly on the mandibles, between M1 and M2, and compared with measurements taken on the radiographs to assess distortion and magnification. Radiographs and measurements were taken as described in Smith and coworkers (1986), with the addition of two new measurements of alveolar and basal bone, as described below. Radiographs were taken of the premolar and molar teeth, using dental occlusal films taped parallel to the dental arch in the premolar-molar region (Fig. 1). Where both sides were intact, radiographs were taken of both sides. The mandible and film were positioned such that the film was a t right angles to the X-ray beam and at a distance of 1.5 m from the energy source. Along each mandible a 4-cm aluminium wedge was placed to standardize density and provide a scale. On the radiographs, digitized tracings were recorded of crown outline, root outline, and corpus height. The crown and root were demarcated by a line joining the cervical margin of the enamel on the mesial and distal aspect of each tooth. Mesiodistal crown length (CW) was traced parallel to this line, and the maximum value was recorded. For the premolar roots the width midway along the root length was traced parallel to this line (RW1 and RW2), and root height (RH) was traced parallel to the long axis of the tooth to the root tip. For the molars, lines were constructed joining the root apices and used to measure the distance between them (RW3, RW4, and RW5). Root height was measured from these lines to the line demarcating the crown and root. Corpus height was traced interproximally from the interdental crest to the base of the mandible, mesial to each of the molars and the first and second premolar. From the tip of the premolar root and mesial roots of the molars a line was drawn parallel with the line defining corpus height crossing the inferior dental canal. The distance from the root apex to the lower border of the inferior dental canal was then traced and used as a measure of “alveolar bone” (IDC), while the distance from the inferior dental canal to the lower border of the mandible was used as a measure of “basal bone” (LB). From these computerized tracings, we obtained crown height, TOOTH AND JAW VARIATION 199 Fig. 1. Radiograph of molar region of the mandible. Arrow shows upper limit of basal bone. crown length, cross-sectional root area, root height, root width, corpus height, alveolar bone height, and basal bone height for each tooth. Measurements of corpus height taken on the mandibles were compared with those taken from the radiographs, to assess possible distortion and/or magnification on the radiographs. No significant differences were found. A further 15 radiographs were redigitized, then compared with the original measurements to determine experimental error related to location of landmarks. This was calculated after Dahlberg (1940) and found to average 2-396 for parameters examined. It was greatest for measurement of basal bone and alveolar bone, because of difficulties in locating the lower border of the inferior dental canal. Statistical analysis On the specimens from Roonka, radiographs were taken of both left and right sides, and the paired Wilcoxon and student’s t-tests were used to check for right-left asymmetry. The correlation between the two sides was also examined to evaluate the effect of averaging the measurements of both sides. Because of the relatively small number of observations, the differences in crown root and jaw proportions between the two populations was assessed through several statistical procedures. Scattergrams were plotted to show the marginal distribution of the parameters for each population as given by the corresponding distribution points on each axis. The projections of the points on the axes represent the measurements of the variable associated with the axis. Univariate comparisons on each parameter and tooth were performed using the t-test and Mann Whitney test. The KolmgorovSmirmov test was applied when the distribution seemed to have a different shape in the two populations. Spearman’s rank correlation coefficients were calculated between variables in each population. Multivariate characterization of the differences between Roonka and Maoris was obtained using the linear discriminant technique and analysis of covariance. Discriminant analysis defined the linear function of the variables measured in each tooth that best discriminates between the populations. The efficiency of the discrimination is estimated by the misclassification error rates (Lachenbruch, 1975). I n order to determine if the differentiation between the two populations is expressed similarly in each tooth, the correlation between the discriminant analysis scores for adjacent teeth was examined. The contribution of each parameter to the difference between the two populations was also derived from the analysis of covariance technique. Through this method, the independent contribution of any 200 P. SMITH ET AL specific parameter to the difference between the two populations is assessed after adjusting for the effect of the other variables. The relationship between covariance adjustment and discriminant analysis in choosing discriminating variables is discussed in Rao (1966), and the use of covariance analysis is discussed in Grizzle and Sen (1983). RESULTS A relatively high correlation was present between the right and left sides of the same individual, while no consistent statistical differences were found between them. Accordingly, for those individuals with both sides present, averaged measurements were used in all further analyses. The difference between the two populations in any one parameter a s well as the difference in the relationship of any two parameters in the second molar can be deduced graphically from Figures 2-5. The other teeth studied showed a similar relationship, with maximum crown height and width and root measurements larger in Roonka and corpus height greater in Maoris. The marginal distribution of the parameters for each population is shown by the corresponding distribution of the points on each axis. The projections of the points on the axes represent the measurements of the variable associated with the axis. The range of values recorded for crown height and mesiodistal diameter clearly differs in the two populations (Fig. 3). The maximum values are those of unknown teeth, and these are consistently larger in Roonka. The univariate analysis clearly shows that the two populations differed in all parameters (Table 1). Mean values for crown height, width, and cross-sectional area were consistently higher in Roonka, a s was root size for all teeth except the third molar. Corpus height was greater in the Maoris, and this was mainly due to differences in alveolar bone height (Table 1). Statistically significant differences between populations were most marked in the second molar and least marked in the first premolar. In both populations correlations between the various crown parameters and those of the bone tended to be negative, and this applied especially to the correlation between crown height and corpus height (Table 2). Correlations between root height and area and corpus height were, however, positive. Population correlation coefficients and the bivariate scattergrams further indicate that correlations among crown and root parameters tended to be positive, while correlations between alveolar bone height and basal bone height tended to be negative. The multivariate analyses (discriminant and covariance analyses) were based on crown height, crown width, root area, and corpus height. The contribution of each specific variable to the difference between the populations is summarized in Table 3. The results indicate that not all the variables included in the analysis contributed significantly to the relatively good separation between the two groups. Crown width showed the highest discriminatory power in every tooth. Corpus height contributed to the differentiation in the second premolar and third molar, and crown height contributed to the differentiation in the second premolar. Root area gave poor discrimination in all teeth. The high correlation between the discriminant analysis scores indicate that the overall difference between the two populations is preserved in adjacent teeth (Figs. 6, 7). Discriminant analysis carried out with basal bone and alveolar bone entered as separate variables showed that alveolar bone made the main contribution to differences in bone height. Covariance analysis supports the above results and those implied by the signs obtained from the standardized coefficients (Table 4): namely, significantly higher adjusted mean mesiodistal diameters in Roonka and significantly higher adjusted means for bone height in the Maoris in the second premolar and third molar. Controlling for the relatively higher mesiodistal diameter and lower bone height in Roonka yielded a lower adjusted crown height in this group a s compared to the Maoris. The higher values obtained for unadjusted crown height in the Roonka teeth seem to be due to the observed differences in mesiodistal diameter and bone height. DISCUSSION The discriminant analysis demonstrates the marked differences present between the two populations (Figs. 8-11). Crown height and mesiodistal diameter gave good discrimination between them. However, while corpus height was greater in the Maoris, it seems that this was mainly due to differences in alveolar bone height. The contribution of basal bone height was nil for the second premolar and first two molars, while root 201 TOOTH AND JAW VARIATION cw4 c 144 cn4 I I* . 120 108 RW4 0 LB4 96 cn4 0 21 . 0 O .. 35 . . * O 0 49 D . 63 77 CH4 Figs. 2-7. Scattergrams showing distribution of parameters studied for the second molar. Solid circles represent values recorded for the Australians, and open starred circles represent values recorded for the New Zealanders. Fig. 2. Values plotted for alveolar bone height (IDC4) against crown height (CH4). Fig. 3. Values plotted for crown width (CW4) against crown height (CH4). Fig. 4. Values plotted for root width (RW4) against crown height (CH4). Fig. 5. Values plotted for basal bone height (LB4) against crown height (CH4). dimensions were similar in the two groups. The discriminant analysis provides further evidence of the independent contribution of tooth size and corpus height to the differences between the two populations. The analysis of covariance provides complementary information with respect to the differences between the two populations and assesses the contribution of each parameter. The Maori face and mandible is large, and the ascending ramus is high (Houghton, 1978b; Pietrusewsky, 1984). The mandible, like that of other Polynesians, is characterized by a large “rocker” form that is due to a convex inferior border. Houghton (1978a,b) attributed this to buckling of the basal bone during growth. Darling and Levers (1975) have described the important role of alveolar 202 P. SMITH ET AL 1 I - ... . . .. 1471 1 7 0 I RA4 .. .0.. . 0 9 133 GO.. 0 " cw4 . * 0 . : b b 119 0 D . 1100 1 3 0 1 I 0 0 ti* ... 00 v * Q * O 0 105 0 0 0 0 5 0 9oot ci u 260 I o 300 340 380 1 220 0 I 260 300 340 ~ n 4 B H ~ Fig. 2-7. (Continued) Fig. 7. Values plotted for crown width (CW4) against Fig. 6. Values plotted for root area (RA4) against corcorpus height (BH4). pus height (BH4). TABLE 1 . Two-tailed P-values for differences between Roonka and Maoris in tooth parameters, using t-test or the Kolmagorou-Smirnov test when the distribution in the two s a m d e s differed1 Variable/tooth Pml Pm2 M1 M2 M3 CH CW CA RH 1DC LB RW RA BH .32 .06 .13 .97 .15 ,000 ,005 .0001 .10 ,000 .03 .35 .002 ,000 .18 .352 ,043 .033 .79 .10 .I8 .607 .I1 .I2 .I0 .05 ,033 .02 ,0053 ,563 .002 .003 .013 .78 .09 .80 .7Ei2 ,005 .03 .21" ,573 .0033 ,223 .573 ,003" 1 Crown size is larger in Roonka. 2 Root size is larger in Roonka for all teeth except M3. 3 Total bone height and alveolar bone height values are lower m Roonka 2Mean values are similar. 3Mean values in Roonka are smaller than those of Maoris. For all other values, mean values in Roonka are greater than those of the Maoris. bone growth in maintaining the teeth in occlusion after eruption. The exceptionally large height of the corpus in the molar region of the Maoris a s compared to Roonka may then be due to compensatory growth of the alveolar bone as the ramus increases in height and the basal bone bends during growth, since it is not related to tooth or root size. With similar root area but larger teeth, the potential force exerted per unit area to the bone in Roonka is greater than in the Maoris. This study is two-dimensional only and does not discuss buccolingual diameters of the teeth or roots. Corpus width of the two groups examined here is, however, similar. Mean values are 15.7 mm in Roonka and 15.5 mm in the Maoris in the first molar region. The Roonka sample represents one of the most robust, largest-toothed Holocene populations known from Australia (Brown, 1982; Smith, 1982).The severe attrition of the teeth and prominent areas of muscle insertion on the bone attest to powerful and severe masticatory stress experienced by this population and borne out by ethnographic accounts of eating habits. The condition of the teeth and jaws of earlier Aboriginal populations from Cobool Creek and Kow Swamp (Brown, 1982; Thorne, 1976) shows very clearly that heavy dental function was the norm throughout Australian prehistory. We find, however, in the Roonka population, as indeed in other Australian aboriginals, a combination of large teeth and small corpus height (Solow et al., 1982). Indeed molar corpus height in the Roonka sample is similar to that of a recent Near Eastern population, although crown and root size in them is some 30% smaller than that found in the Roonka population (Smith et al., 1986). Prehistoric Polynesian populations from Hawai and Tonga have comparatively little 203 TOOTH AND JAW VARIATION TAB1,E 2. Within-populationassociation of tooth parameters: Spearman correlation coefficients for the second molar in Roonka and Maori samdes C€[4 0.5551 1 CH4 2 cw4 3 RA4 4 CA4 5RH4 6 IDC4 7 LB4 8 RW4 9 BH4 cw4 0.5540 -0.C683 0.E 099 -0.2486 -0.5093 -0.5816 -0.C444 -0.4885 0.3726 0.7471 -0.1121 -0.3446 0.0099 0.4236 -0.3132 RA4 CA4 -0.2634 -0.1154 0.9396 0.6546 -0.2334 0.0711 0.5948 -0.0124 0.0124 0.6360 0.2539 -0.1697 -0.5286 -0.2556 0.1222 -0.4420 Maori RH4 -0.6579 -0.3997 0.6346 -0.6539 0.1082 -0.0861 0.1497 0.4729 IDC4 LB4 RW4 BH4 -0.062 -0.1940 -0.4060 -0.0423 -0.4006 -0.4060 -0.1883 0.2779 -0.3469 0.3892 -0.2510 0.2057 0.1667 0.5127 0.2944 -0.1631 -0.1948 0.2930 -0.3732 -0.0651 -0.3089 0.0999 0.4904 0.3118 -0.3010 0.1974 0.5357 0.0319 0.0572 -0.3091 Roonka TABLE 3. Discriminant analysis results: standardized coefficients and classification dates for each tooth and variable VariabWtooth CH cw RA BH Constant Correct classification rates (W) Pml Pm2 M1 M2 M3 ,515 -1.705 -.01068 22199 2.68589 1.19 -3.00 .00051 .382 3.67 -.0196 -1.170 ,017 ,053 10.79 .28 -1.17 -.a26 .09 10.83 .59 -34 .O1 .28 2.8039 59.4 85.4 70.0 78.6 TABLE 4. Aiialvsis of covariance-summary table Adjusted means Dependent variable Col ariates Tooth Maori Roonka P-value 1 CH CW, BH, RA BH CH, CW, RA 2 3 4 5 1 2 3 4 CW RA CH. RA.BH BH, CH,CW 5 1 2 3 4 5 1 2 3 4 5 45.93 51.57 45.86 52.05 51.79 342.39 347.66 311.08 291.76 305.47 72.65 71.16 114.55 40.76 40.16 46.02 47.54 47.01 330.35 316.38 307.73 282.80 281.99 80.54 80.14 119.56 118.51 129.24 112.48 178.03 716.27 723.54 743.12 742.28 1,066.42 1,104.58 1,128.75 1,209.18 1,105.51 1,063.85 * ,0017 ,9621 ,3296 .1753 ,2900 * .7275* ,4785 ,0356 ,0128 ,0000 ,0395 -0011 .0784 ,858 (*) ,492 (*) ,470 *The difference be ween the populations depends on the covanate levels. P-values are for equality of adjusted means. dental attrition (Smith unpublished data). The Maoris are Polynesians and in contrast to the Austr.dian aboriginals were horticulturalists w i t h similar traditions to those of other Pacific: Polynesians; presumably like 69.2 them they were accustomed to a low-abrasive diet. Houghton (1978a, 1980) has emphasised the seventy of the New Zealand climate, especially in the South Island. He suggested that the climate deteriorated over time, reducing the yields of horticulture and resulting in increasing dependence on abrasive fern roots as a dietary staple. The Maori and Roonka samples were then descended from populations that had experienced different selective pressures on the teeth and jaws in the past, and this presumably accounts for the small teeth of the Maoris despite their recent exposure to a high-attrition diet (Smith, 1982). However, both populations showed a similar response to attrition, in that corpus height increased a s crown height and width decreased. As the correlation coefficients show (Table 2), the increase in corpus height was mainly due to a n increase in the height of the alveolar bone and may be attributed to alveolar growth associated with continued eruption of the teeth as described by Ainamo and Talari (1975). The correlation coefficient between loss of crown height and increased corpus height was, however, low, suggesting that the two events were not highly correlated. This assumption receives some support from the 204 P. SMITH ET AL. 2 0 3.2 V 0 1 4 off 0 V ' 0 V 1.8 0 V 0 . . s1 1 -2 . . . . . . .... . . 0 52 ao 0 . - 0 O 0 V * 0 . 0. . - 1.6 .v -3 - *5 --5o -50 15 53 SZ 0 V 32 1. 0 0 00 9 0 0 V Q 0 s3 0 . .. a. .. ... .. s4 * Q(1 O - .9 .1.f -2.5 . . -. .... .. 0 0 0 0 0 m - 0 0 1.6 . 0 0.1 0 0 0 0 0. - - 0 -16 v . 0 -.50 0 . 1.5 a5 54 -12 0.0 1.2 2-4 $55 Figs. 8-11. Plot of the discriminant analysis scores for adjacent teeth. S1 = scores for the first premolar; S2 = scores for the second premolar; 53 = scores for the first molar; S4 = scores for the second molar; and S5 = scores for the third molar. Symbols as in Figures 2-7. findings of Murphy (1959) and Tallgren (1957). They found that in populations with severe attrition, lower facial height (which is largely governed by the combined height of alveolar bone and crown height) decreases, while it increases in populations with little attrition. Murphy (1959) also reported that the observed variation in facial height is associated with growth of alveolar bone as well as active eruption and attrition of the teeth. This means that while continued eruption in adult life is a n ongoing process that compensates for loss of crown height through attrition, it does not seem to be a very finely tuned process. This is achieved by continued remodeling of the mandibular condyle in older individuals to compensate for changes in the level of the occlusal plane, as described by Mongini (1975) and Richards (1984). The pros and cons of a physiologic as opposed to a pathologic basis for alveolar resorption have been discussed a t great length over the years, with some workers claiming that under normal conditions, there is physiological migration of the gingival attachment to maintain constant clinical crown height following attrition. However, such claims are invariably derived from populations with a 205 TOOTH AND JAW VARIATION high-cereal diet, which is conducive to plaque accumulation and so to periodontal disease and alveolar bone loss. This probably accounts for the marked differences in alveolar bone loss reported by Lavelle and Moore (1969) between 8th and 17th century British populations. Moreover, close examination of the frequently quoted study by Philippas (1952) on Greek skeletal remains shows that in only 50% of the attrited teeth was clinical crown height equivalent to that of unworn teeth. In the other 50% clinical height was considerably greater, and this Philippas acknowledged as due to the pathological resorption of alveolar bone from periodontal disease. We would argue that in the other 50% periodontal disease was also present, but in a less advanced form that approximated the height of unworn teeth. Since both attrition and periodontal disease are age related, the rae of both must occasionally coincide, a s found by Newman and Levers (1979) for Anglo-Saxons and by Whittaker et a]. (1982; 1985) for a Romano-British sample. That the association is accidental we infer from the fact that there is no correlation found between the two in the majority of populations studied. This is especially true when the populations are eating a n abrasive but lowcarbohydrate diet as i n the Australians studied by Murphy (1959), Amerindians studied by Goldberg et al. (1964), and the Australians and New Zealanders studied here. Clinical studies have shown that active eruption, even under extreme conditions, is usually associated with growth of alveolar bone (Ainamo and Talari, 1975;Ingber, 1974). Mechanically this is important, since reduction of crown height, without accompanying reduction of the area of root attached to the bone, is mechanically advantageous. If on the contrary, crown height is maintained at a constant level, through apical migration, the clinical crown-root ratio gradually increases. This produces a n unfavorable crown-root ratio that cannot withstand lateral forces and contributes to the rapid progress of periodontal disease. In conclusion, this study provides further evidence of the low correlation present between crown size, root size, and corpus height both within and between populations. The results also show that for all variables studied, the differences are maximal in the second molar region. These results conform to those previously obtained for Near Eastern populations of different phenotype and den- tal function. They suggest that factors other than crown and root size must be considered in the analysis of evolutionary trends in the hominid mandible. ACKNOWLEDGMENTS We are indebted to the Departments of Dental Radiography of the Faculties of Dentistry at the Universities of Adelaide and Otago for use of their equipment and assistance in taking the radiographs on which this study was based. We would also like to thank G. Pretty, South Australian Museum, and P. Houghton, University of Otago, for permission to examine the specimens described here and T. Brown for his hospitality and assistance with all phases of the study. LITERATURE CITED Ainamo J, and Talari A (1975) Eruptive mbvements of teeth in human adults. In DEG Poole and MV Stack (eds.): The Eruption and Occlusion of Teeth. London: Sutterworth, pp. 97-107. Anderson D, Thompson L, and Popovich F (1977) Tooth, chin, bone and body correlations. Am. J. Phys. Anthropol. 46:7-12. Brown P (1982) Cobool Creek A Prehistoric Australian Hominid Population. Ph.D. thesis. Canberra: Austral. Nat. University. Dahlberg G (1940) Statistical Methods for Medical and Biological Students London: George Allen and Unwin Ltd. Darling AI, and Levers BG (1975) The pattern of eruption of some human teeth. Arch. Oral Biol. 20:89-96. Garn SM, Van Abstine WL, Jr., and Cole P E (1978) Root length and crown size correlations in the mandible. J. Dent. Res. 57:114. Goldberg H, Weintraub J,Roghmann K, and Cornwell W (1976) Measuring periodontal disease in ancient populations: Root and war indices in study of American Indian skulls. J. Periodontol. 47r348-351. Grizzle J E , and Sen PK (1983) Selection of information variables in multi-variable analysis: An analysis of covariance approach. I n PK Sen (ed.): Contribution to Statistics. Essays in Honor of Normal L. Johnson. Elsevier, pp. 195-204. Gustafson G (1950) Age determination of teeth. J. Am. Dent. Assoc. 41:45-54. Houghton P (1978a) Dental evidence for dietary variation in prehistoric New Zealand. J. Polynes. SOC. 87357-263, Houghton P (1978b) Polynesian mandibles. J. Anat. 127:251-260. Houghton P (1980) The First New Zealanders. Aukland: Stoughton Ltd. Ingber J (1974) Forced eruption: Part I. A method of treating isolated one and two wall infrabony osseous defectsrationale and case report. J . Periodontol. 45199-206. Krogman WM (1962) The Human Skeleton in Forensic Medicine. Springfield, I L Charles C. Thomas. Lachenbruch PA (1975) Discriminate Analysis. New York: Hafner Press. Lavelle CLB (1962) A comparison between the mandibles of Romano-British and 19th century periods. Am. J. Phys. Anthropol. 36213-220. 206 P. SMITH ET AL. Lavelle C, and Moore W (1969) Alveolar bone resorption in Anglo-Saxon and seventeenth century mandibles. J. Periodont. Res. 4:70-73. Loe H, Anerud A, Boysen H, and Smith M (1979) The natural history of periodontal disease in man. J . Periodontol. 49t607-620. Molnar S (1971) Sex, age and tooth position a s factors in the production of tooth wear. Am. Antiquity 36:182-188. Mongini F (1975) Dental abrasion a s a factor in remodelling of the mandibular condyle Acta Anat. (Basel) 92.992-300. Moore WJ, Lavelle CLB, and Sperce T F (1968) Changes in the size and shape of the human mandible in Britain. Br. Dent. J. 125:163-169. Murphy T (1959) Compensatory mechanisms in facial height adjustment to functional tooth attrition. Aust. Dent. J. 4:312-323. Newman H, and Levers B (1979)Tooth eruption and function in an early Anglo-Saxon population. J . R. SOC.Med. 72341-350. Pederson PO (1949) The East Greenland Eskimo Dentition. Kobenhaven: CA Reitzels Forlag. Philippas G (1952)Effects of function of healthy teeth: The evidence of ancient Athenian remains. J. Am. Dent. ASSOC.45:443-453. Pietrusewsky M (1984) Metric and non-metric cranial variation in Australian Aboriginal populations compared with populations from the Pacific and Asia. Occ. Papers in Human Biology. Aust. Inst. Aboriginal Studies 3:l-113. Pretty GL (1977) The cultural chronology of Roonka Flat-a preliminary consideration. In RVS Wright (ed.): Stone Tools a s Cultural Markers: Change, Evolution and Complexity. Canberra: Australian Inst. of Aboriginal Studies, pp. 288-331. Prokopec M (1979) Demographical and morphological aspects of the Roonka population. I n AP Elkin (ed.): Oceania Monograph, Sydney No. 22. pp. 161-176. Rao CR (1966) Covariance adjustment and related problems in multivariate analysis. In PR Krishnaiah (ed.): Multivariate Analysis. New York Academic Press, pp. 87-103. Richards LC (1984) Principal axis analysis of dental attrition data from two Australian Aboriginal Populations Am. J. Phys. Anthropol. 655-13. Rosing FW (1983) Sexing immature human skeletons. J. Hum. Evol. 12:149-155. Shawcross K (1967) Fern root and 18th century Maori food production in the agricultural areas. J . Polynes. SOC. 76:330-357. Smith HB (1984) Patterns of molar wear in huntergatherers and agriculturalists. Am. J. Phys. Anthropol. 63.39-56. Smith P, Wax Y, Adler R, Silberman U, and Heinic G (1986) Post Pleistocene changes in tooth, root and jaw relationships. Am. J . Phys. Anthropol. 70:339-348. Smith P, Prokopec M, and Pretty G (1988) The dentition of a prehistoric population from Roonka Flat. S. Aust. Archeol. Oceania 23:31-36. Solow B, Barret MJ, and Brown T (1982) Craniocervical morphology and posture in Australian Aboriginals. Am. J. Phys. Anthropol. 59~33-46. Tallgren A (1957) Changes in adult face height due to aging, wear, and loss of teeth and prosthetic treatment. Acta Odont. Scandinavia 15:l-22 (Supp. 24). Taylor (1963) RMS Cause and effect of wear of teeth Acta Anat. (Basel) 53:97-157. Thorne AG (1976) Morphological contrasts in Pleistocene Australians. I n RL Kirk and AG Thome (eds.): Origin of the Australians. Canberra: Australian Inst. Aboriginal Studies, pp. 95-1 12. Whittaker D, Parker J, and Jenkins C (1982) Tooth attrition and continuing eruption in a Romano-British population. Arch. Oral Biol. 27:405-409. Whittaker D, Molleson T, Daniel A, Williams J, Rose P, and Resteghnini R (1985) Quantitative assessment and tooth wear, alveolar-crest height and continuing eruption in a Romano-British population. Arch. Oral Biol. 30:493-501.