Patterns of size variation and correlation in the dentition of the red colobus monkey (Colobus badius).код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 54:139- 146 (1981) Patterns of Size Variation and Correlation in the Dentition of the Red Colobus Monkey (Colobus badius) LARRY R. COCHARD Department of Anthropology, University of Wisconsin, Madison, Wisconsin 53 706 KEY WORDS Growth fields, Tooth-crown size, Colobus badius ABSTRACT There have been numerous studies on variability and correlation in dental crown size, but the significance of the resulting patterns remains unclear. Regions of low variation and high correlation have been hypothesized to represent the poles of Butler’s morphological fields, to be related to absolute tooth size, or to be related to morphological complexity of the teeth and functional efficiency. Variation and correlation of tooth lengths and breadths were investigated in 138 red colobus monkeys to further assess the relations among size associations, crown morphology, and absolute tooth size. In the maxilla and mandible, the postcanine teeth are the most highly correlated and least variable, followed by the incisors, then the canines. There are also lower correlations between premolars and molars than within either group. While there appears to be a relation between degree of morphological differentiation and levels of correlation and variation, there are no notable differences in the correlation of opponents along the dental arcade, which is the most important functional consideration. This suggests that different levels of correlation and variation within upper or lower teeth are “artifacts” of tooth dimensions that contribute to different geometric designs in different tooth groups as the germs develop. This morphological effect is coupled with the influence of integration fields, indicated by higher variability and lower correlations of the third molar, the largest or most molarized tooth. It is concluded that there are wide functional tolerances in occlusion with respect to the gross dimensions of dental crowns and their interrelationships. Teeth are one of the most important components of an animal’s adaptation to its environment. Because of this and because of their abundance in the fossil record, their high genetic component of variation, and their organization as serially homologous organs, variation and covariation in tooth size have received much attention. Such studies are important for diagnosing sexual differences or species differences in the fossil record (Gingerich, 1974; Gingerich and Schoeninger, 1979) and for making inferences about developmental processes (Van Valen, 1962). Patterns of variability and correlation have also been described in terms of dental growth fields (Dahlberg, 1945; Henderson, 1975;Suarez and Bernor, 1972;Van Valen, 1962), primarily to provide a framework for understanding evolutionary changes in dentitions. Butler (1939) first applied the concept of 0002-948318115401-0139$01.70 0 1981 ALAN R. LISS, INC. a developmental field to teeth to explain the detailed similarities of adjacent teeth, differences in form in the incisor, canine, and postcanine regions of the dentition, and evolutionary changes in gradients of size and form in the three different regions. While much attention has been devoted to patterns of variation and correlation in tooth size, their developmental and/or functional significance is not fully understood. The traditional view is that regions of low variation are the poles, or areas of strongest effect, of Butler’s morphological fields (Dahlberg, 1945; Schuman and Brace, 1954). Others consider the highest correlations between adjacent teeth to represent the polar areas of morphological fields (Suarez and Bernor, 1972; Henderson, Received November 1, 1977; accepted July 9.1980. 140 L.R. COCHARD 1975). Olson and Miller (1958) and Kurten (1953)consider integration (correlation)within a dentition to be largely a function of relative tooth size, which is related to Butler’sconcept of functional fields. Gingerich and Schoeninger (1979) recently advanced the hypothesis that patterns of variation in size are related to occlusal complexity of the teeth. Among a large series of primate species, they found the lowest coefficients of variation located in the first or second molar. They concluded that because of the more complex morphology of the postcanine teeth, they must be less variable to occlude with the same functional efficiency as canines or incisors. In the present study, patterns of variation and correlation in tooth-crown size were investigated in a n arboreal Old World monkey, CoZobus badius, to further our understanding of factors t h a t influence tooth-crown size. By comparing these patterns to the fields responsible for each morphological class, i.e., incisors, canines, premolars, and molars (Dahlberg, 1945), and to Butler’s morphological fields, the hypothesis was tested that the fields influencing tooth-crown size are closely related to the fields determining tooth morphology. The red colobus monkey is suitable for such a study because of an enlarged third molar, which will aid in determining the effects of absolute size. Hominoids, with their reduced third molars, have usually been the object of study for crown-size fields. In this paper, “growth field” is used only in a descriptive sense, since t h e specific developmental mechanisms for the patterns are unknown. MATERIALS AND METHODS The sample ofCoZobus badius consisted of 85 female and 53 male a d u l t monkeys from Liberia. They were killed for food by natives occupying the right bank of the Cess River between the villages of Potogle in the north and Kpeaple in the south. The teeth were measured by Leutenegger (1971, 1976) with a sliding caliper, calibrated to 0.1 mm and with arms ground to fine points. An average error of 0 . 2 m m was obtained on repeated measurements. The data consist of mesiodistal and buccolingual diameters of all teeth on the right side of the maxillary and mandibular dentitions. Only right teeth were measured because there are no significant differences between the size of right and left teeth (Ashton and Zuckerman, 1950). If a right tooth was missing, its antimere was substituted. Mesiodistal diame- ter (length) was defined as the maximum distance between the mesial and distal contact points, parallel to the tooth row. Buccolingual diameter (breadth) was measured as the maximum distance between the buccal and lingual surfaces, perpendicular to the mesiodistal diameter. Both dimensions of the canines and breadth of the incisors were taken at the base of the crowns. Matrices of correlation coefficients ( r )for all possible pairs of teeth were computed for lengths and breadths. The correlation matrices for each variable were compiled separately for males and females and for maxillary and mandibular teeth. For each matrix, the hypothesis tested was that correlations for all pairs of adjacent teeth are samples of a common parameter (Snedecor and Cochran, 1967:186). Fisher’s z scores were used for tests of significance in comparing individual correlation coefficients. The patterns of correlation and the distribution of the coefficients of variation were compared with each other, with the posterior shift in molarization due to a n enlarged third molar, and with the division of the dental arcade into regions corresponding to the four morphological classes of teeth. The results were compared with other studies of variation and/or correlation of tooth size in mammals. All data were processed on a Univac 1110 computer using programs from the Statjob Series of the madison Academic Computing Center, University of Wisconsin. Since some teeth were missing or damaged, a bivariate subsample method was employed for treating missing data; each correlation coefficient was computed for data pairs in which values for both variables in the pair are present. Thus, individual sample sizes for the correlations range from 76 to 85 for females and from 40 to 51for males. Wherever male and female data have been combined, a weighted average for the means of each sex was computed. RESULTS Correlation matrices for the buccolingual breadths of maxillary and mandibular teeth are given in Table 1and for mesiodistal lengths in Table 2. Values for males are found above the diagonals and for females below. Correlations greater than r = 0.27 (n = 51) and r = 0.21 (n = 85) are significantly different from zero (P < .05) for males and females, respectively. Although the correlations tend to be higher for females, of 112 male-female comparisons for individual correlation coefficients, eight are significantly different a t the 5% level, which 141 SIZE ASSOCIATIONS IN THE DENTITION OF COLOBUS BADIUS T A B M 1 . Correlation matrices for the buccolingual widths of the maxillary and mandibular dentitions of Colobus badiuq intradentition correlations for males ( n = 46-51) are above the diagonals, for females f n = 78-85], below Maxilla 1' 11 I' C' p:' p4 M' M' M" .73* .50 .48 .55 .50 52 .50 I' C' Pt P MI M' MJ .49 .45 .52 .36 .32 .49 .34 58 .66 .71 .44 .67 .61 .57 .71 .37 .60 .63 .59 .73 .86 .40 .60 .69 56 .75 .83* .90 .54 .50 .49 .46 .45 .48 .45 .45 .46 .38 .39 .I4 .49 .49 .46 .65 .68 .80 .67 .64* .84 P4 M, MI M3 .46 .29 .20 .33 52 56 .29 .29 .36 .52 .89 .59 .39 .37 .40 .49 .79 .85 Mandible 1, .74 11 12 C, P, P4 MI MI M3 12 .79 53 .27 .43 .48 .44 .44 .59 .31 .44 .46 .45 .51 C, P:, .30 .18* .26 .47 .39 .40 .40 .35 .28 .35 .57 .32 .27 .24 *Significantly different from the opposite sex a t P = .29 .35 .40 55 57 .58 52 .80 ,653 .78 .05. TABLE 2. Correlation matrices for the mesiodistal lengths of the maxillary and mnndibular dentitions of Colobus badius; intradentition correlations for males f n = 41-51) are aboue the diagnoals, for females f n = 78-85), below I' I' 12 C1 PJ P M' M' MJ 12 50 Maxilla C' .48 .32 .65 .52 .38 .47 .55 .45 .25 .55 .38 .42 .30 .27 .20 11 1, C, ,563 23 .24 p:' P4 M' M' M I .25 .43 .46 .43 .28* .68* .26 .29 26 .48 .47 .26 .41 .35 .57 .60 .66 .18 .30 .25 .35 24 .53 .I2 .20* .61* .84* .43 .44 .43 .68* 50 .45 .42 .48 .49 .46 .73 .43 .51 p:, P4 Mi M:! M3 .31 .31 .35 .29 .42 .26 .26 .38 .22 .08* -.08* .53 .48 .32 .I3 .30 .72 .33 .17 .17 .05 .44 .24 .37 Mandible 1, 2, p, P4 M, M, M3 .69 .01 .I4 .24 .45 .32 .14 .12 .27 .30 .28 26 .15 .37 .22 .31 .25 .32 'Significantly different from the opposite sex at P .53* .28* .40 .31 = -05 .28 .20 .20 .30 .66 .34 .37 142 L.R. COCHARD 0.9 0.8 0.8 0.7 0.7 0.6 0.6 0.5 0.5 0.4 0.4 0.3 0.2 L' h 0.6 0.5 0.4 0.3 0.2 I I' ' l2 ' ' ' ' ' ' C P3 P4 M' M2 0 .1 M3 Fig. 1. Correlations (rj for the mesiodistal length (L) and buccolingual breadth (B) of adjacent maxillary teeth. Fig. 2. Correlations (r)for the mesiodistal length (L) and buccolingual breadth (B) of adjacent mandibular teeth. approximates the expected sampling error. In each table, the highest correlations are found between adjacent teeth and generally decrease inversely with the number of intervening teeth. But in Table 2, a n irregularity is evident in the matrix for the lengths of male mandibular teeth. The correlations between the third mandibular premolar (PJ and all teeth distal to its hover around zero, including a value of r = 0.08 for P, and P,. This is probably an artifact of the rotation of the sectorial PB,since its mesiodistal length was measured and not its maximum length. The spurious correlations of P:, were not included in any calculations nor considered in discussion of the results. Figures 1and 2 plot correlations for adjacent teeth and indicate two general patterns for the distribution of the values. Though the correlations are generally high between adjacent teeth, the hypothesis that they are identical is rejected at the 5% level for both variables in both jaws and both sexes. The correlations are not equally distributed along the tooth row, but generally increase both mesially and distally from the canine region. This is also evident in comparing the lowest correlation, involving the canine and either the second incisor or the third premolar, with the highest value among the molars. The difference is significant (P < .05) in seven of eight cases. Superimposed upon this trichotomy of the anterior, canine, and postcanine teeth, there appears to be a relation between the patterns of correlations and each morphological class, i.e., there is a further distinction between the premolars and molars. Figures 1 and 2 indicate that in six of seven plots (the spurious correlation of male P:,length disregarded), the peaks fall within the morphological classes while the lows occur between them. The correlations of P3-P4 and the molars Ml-M2 are both significantly higher (P < ,051 than for P 4 M 1 in two cases: female maxillary and female mandibular lengths. Similar to the general patterns for correlations between adjacent teeth, the coefficients of variation for length and breadth (Fig. 3 and 4) distinguish the canine region from the incisors and postcanine teeth. The coefficients gradually increase from the first or second molar to the central incisor, but in each case the gradation is interrupted by elevated values for the canine and/or first premolar. The patterns of variation and correlation in this sample are not only concordant in their distributions, but also in magnitude. Those teeth that are most highly integrated in size are least variable in size; thus the sequence of increasing variability is postcanine teeth, incisors, and canine region. Correlations between the size of occluding teeth were also calculated (Table 3). They are similar in magnitude to the correlations between adjacent teeth but reveal no major differences between the three regions of the dentition. While the values for the length of the 143 SIZE ASSOCIATIONS IN THE DENTITION OF COLOBUS BADZUS i 4 .\‘ ___ d -0 .-- -. -c, c. v. c. v. 3 i 1 I’ I* C P3 P4 M’ M2 M3 Fig. 3. Coefficientsofvariation (C.V.) for the mesicdistal length (L) and buccolingual breadth (B) of maxillary teeth. Fig. 4. Coefficients of variation (C.V.) for the mesicdistal length (L)and buccolingual breadth (B)ofmandibularteeth. upper and lower canines in both sexes are low, the coefficients for the canine breadths are comparable to those of the other teeth. For lengths, the values in the center of the molar region are slightly higher than for incisors, but the reverse is true for breadths. One important point is the difference in values between the second and third molars. For breadths, the values are a little higher for the third molars, but in both males and females, the correlations between lengths are significantly lower (P < ,051 for the third molars. The distribution of the correlation coefficients and coefficients of variation relate well to the regions of morphological differentiation but not to the absolute size of the postcanine teeth. The third molar is larger than the second, with 3% more surface area in the upper and 28% more in the lower tooth (sexes combined). But in general, the size of the third molars are more variable, more weakly intercorrelated than both the first upper and lower molars (M,-M1) and the second pair (M2-MS), and more weakly correlated with M2 than M2 is with M1. pared schematically in Figure 5 with Butler’s (1939) fields of incisiform, caniniform, and molariform differentiation and with the fields responsible for each morphological class (Dahlberg, 1945). The darkest areas are the poles of the fields, indicated here as teeth with the lowest amount of size variation or the highest intercorrelation in size of adjacent and occluding teeth. The third molar is the polar tooth of Butler’s postcanine field, since the concept of molarization includes absolute size in addition to the number of cusps on a tooth. The fields for each morphological class have no poles because they are qualitatively different. While the incisor, canine, and postcanine regions appear independent with respect to the patterns, the poles within the postcanine teeth are not concordant, suggesting that fields influencing variation and correlation are somewhat independent of the relative size, or degree of molarization, of t h e poscanine teeth. Gingerich and Schoeninger (1979) also concluded that patterns of variation do not represent the stable poles of Butler’s fields and found no relation between size variation and time of eruption. They proposed their occlusal complexity-functional efficiency theory, that teeth with more complex morphology must be less variable than more simple teeth to occlude a t the same functional level. While it seems likely that there is some relation among occlusal complexity, size restraints, and functional efficiency, covariation in the size of occludingteeth DISCUSSION AND CONCLUSIONS The continuous distribution of correlations and coefficients of variation along the dental arcade suggests some type of common influence regulating the size of developing tooth crowns. To help determine the significance of these patterns, the field effects of the results are com- 144 L.R. COCHARD would seem to be a more important limiting factor in occlusion than absolute variation. A greater amount of variability in any region of the dentition could be tolerated a s long as small or large teeth occlude with teeth that deviate from the mean in the same direction. While the incisor, canine, and postcanine teeth differ in levels of variation in both jaws of the red colobus, there are no outstanding differences in the covariation of occluding teeth along the dental arcade. The correlations between opposing molars are generally no higher than for opposing incisors, even though the latter are more variable in size and have a more simple edge to edge occlusion. Size correlations between the lengths of upper and lower canines are low, but this may be a sampling artifact. In other mammals, including t h e orangutan, brown bear, red fox, and bobcat, the corre- TABLE 3. Correlations between homologous dimensions of occluding teeth Length 11-1, 1'-12 C'-C, PI-P,, P-P, MI-M, M'-M2 M'-M,, Breadth males females males females .55 .23* .23 .45 .50 .67 .63 .37 .64 .57* .36 .52 .46 .67 .70 .50 .59 .63 .59 .44 .49 .54 .51 .67 .61 .67 .49 *Significantly different from the opposite sex at P l n c i s vizat i o n , c a n i i ization, and nolariza t ion: Morp rological c l a s s : = .40 .50 .52 .59 .64 .05 I1 ... j -1 12 lations for the length of opposing canines are among the highest in the dentitions (Kurten, 1953). The same is true for the canines of the gorilla, even though they are also among the most variable teeth in length and breadth within either sex (Suarez and Bernor, 1972). The occlusal complexity-functional efficiency theory also implies t h a t t h e same size restraints that influence the first and second molars should influence the third molar in the red colobus, because it is the largest tooth; but it is more variable, and the correlations between opponents or with M2 are generally lower for the third molar than for the other molars. At least two factors may account for the metric patterns in this study: (1)size relations are "artifacts" of the action of morphological fields, and (2)a field for integration of the teeth is acting independently of all other fields. Although the metric patterns do not correlate well with the absolute size of the postcanine teeth, they do correlate well with the degree of morphological differentiation among the four classes of teeth. This is particularly evident in the patterns of correlations. The hypothesis t h a t tooth germs a r e controlled by developmental regimes that influence both morphology of the crowns and size relations with other teeth may be described in relation to growth and development as follows: Whether or not a tooth germ differentiates into an incisor, canine, premolar, or molar depends on the amount of growth of the germ in three dimensions. In this sense, the concept of form involves a n element of size in each dimension. The C P3 1- P4 M1 I.'..:'.:;.;:*.;: M2 M3 ::I Co r r e l a t i o n (adj.) : Correlation (opp.): Size variation: Fig. 5. Crown-size fields in Colobus badius compared with the fields for the morphological classes (Dahlberg,'45)and for incisivization, caninization, and molarization (Butler, '39).Darkened areas represent the poles of the fields. Size correlations of adjacent (adj.)and occluding (opp.)teeth are shown separately. SIZE ASSOCIATIONS IN THE DENTITION OF COLOBUS EADIUS growth of a tooth germ in breadth, for example, will be coordinated with growth in other dimensions to produce a specific geometric shape. It would seem likely, then, t h a t the final breadths of a molar series would share an element in common that is not present in the breadths of teeth genetically programmed to participate in a different geometric design. The common element may be manifest in size correlations, canalization (and the resultant variability), or any process that is responsible for the final size of the teeth. Of course, this morphological effect is operating concurrently with a number of other factors that may influence size relations, such as proximity of tooth germs, integration fields, or factors that affect the size of an entire dentition. This morphological hypothesis is supported by factor analyses of tooth-crown size. In a n analysis of tooth-crown size in eight primate species, the projecting canines of the nonhuman species were isolated in the terminal factor solutions,while the incisiform cenines of the human samples were grouped with the incisors (Henderson, 1975). Premolars and molars show some independence in factor analyses of crown size in human populations (Lombardi, 1975; Suarez and Williams, 19731,even though both are grinding teeth. Even the shape of tooth crowns is not independent of morphology. In human populations, correlations between the shape (length/breadth) of adjacent dental crowns are 1% higher within tooth groups than between (Garn et al., 1967). The higher variability and lower correlations of M3 may be due to its position at the periphery of the tooth row, and hence, its lower integration, even though it is a larger tooth. Data from Swindler (1976) indicate that many macaque and baboon species have enlarged third molars that are more variable in size than the second molars. In addition to fields responsible for tooth form, there may be a field for integration operating independently and wholly within the constraints for minimal functional efficiency. Integration fields may also be responsible for the higher coefficients of variation in the canines. A small canine morphological field wedged between two much larger fields may result in some developmental instability. But as mentioned before, the magnitude of variation along the tooth row is not critical. Developmental influences or disturbances that vary by region in the dentition can be tolerated since “. . . most of the factors influencing dentition size operate equally on both the maxillary and mandibular dentitions” (Lavelle, 1976:122). 145 Even though distinct patterns of correlation and variation emerge, the relationship between functional efficiency, size associations, and the complex interplay of the teeth in mastication remains enigmatic. Explaining patterns of tooth size as artifacts of morphological fields or integration fields implies that they have little functional significance or that there is a wide functional tolerance i n relation to the gross dimensions of dental crowns. Olson and Miller (1958:199) applied a n elaborate correlation technique to 83 measurements of the upper and lower molars and P4 of the South American owl monkey, Aotus triuirgatus,and found that integration among the teeth is “. . . low when compared to integration found in such functional systems as the limbs or axial structures of various vertebrates.” Even integration between the parts of upper and lower teeth that meet in occlusion is low in the owl monkey. In the present study, none of the correlations, either between adjacent or occluding teeth, are particularly high, indicating that little of their variance is shared. Much more work on mechanisms controlling tooth development and occlusal relations is necessary. ACKNOWLEDGMENTS I would like to thank Dr. Walter Leutenegger for permission to use the data and for commenting on the manuscript, Mr. James Kelly for assistance in preparing the data, and Ms Susan Larson and Dr. James Cheverude for their helpful suggestions. I would also like to thank Dr. Richard Osborne, Dr. Kenneth Bennett, and a n anonymous reviewer for their comments on an earlier draft of the paper. LITERATURE CITED Ashton, EH, and Zuckerman, S (1950)The influence of geographic isolation on the skull of the green monkey (Cercopithecus aethiops sabueusi: I. A comparison between the teeth of the St. Kitts and the African green monkey. Proc. Roy. SOC.,137B:212238. Butler, PM 11939)Studies of the mammalian dentition: Differentiation of the post-canine dentition. Proc. Zool. SOC., London, 109E: 1-36. Dahlberg, AA (1945) The changing dentition of man. J. Am. Dent. Assoc., 32:67&690. Garn, SM, Lewis, AB, and Kerewsky, RS (1967) Shape similarities throughout the dentition. J. Dent. Res.,461481. Gingerich, PD (1974) Size variability of the teeth in living mammals and the diagnosis of closely related sympatric fossil species. J. Paleontol., 48:89&903. Gingerich, PD, and Schoeninger, MJ (1979)Patterns of tooth size variability in the dentition of primates. Am. J. Phys. Anthropol., 51r457-466. Henderson, AM (1975) Dental Field Theory: An Application to Primate Dental Evolution. Ph.D. dissertation, Univ. of Colorado. 146 L.R. COCHARD Kurt&, B (1953)On the variation and population dynamics of fossil and recent mammal populations. Acta 2001.Fennica, 76:l-122. Lavelle, CLB (1976)Odontometric study of African monkey teeth. Acta anat., 96:115-127. Leutenegger, W (1971) Metric variability of the postcanine dentition in colobus monkeys. Am. J. Phys. Anthropol., 35: 91- 100. Leutenegger, W (1976) Metric variability in the anterior dentition of African colobines. Am. J. Phys. Anthropol., 4545-52. Lombardi, AV (1975) A factor analysis of morphogenetic fields in the human dentition. Am. J. Phys. Anthropol., 42~99-104. Olson, EC, and Miller, RL (1958)Morphological Integration. Chicago: University of Chicago Press. Schuman, EL, and Brace, CL (1954)Metric and morphologic variations in the dentition of the Liberian chimpanzee; comparison with anthropoid a nd human dentitions. Human Biol., 26:239-268. Snedecor, GW, and Cochran, WG (1967)Statistical Methods. Ames, Iowa: Iowa State University Press. Suarez, BK, and Bernor, R (1972)Growth fields in the dentition of the Gorilla. Folia primat., 18:35&367. Suarez, BK, and Williams, BJ (1973) Dental growth fields and premolar morphology. J. Dent. Res., 52:632. Swindler, DR (1976) Dentition of Living Primates. New York: Academic Press. Van Valen, L (1962) Growth fields in the dentition ofPeromyscus. Evolution, 16272-277.