Preferential Expression of Linear Enamel Hypoplasia on theSectorial Premolars of Rhesus Monkeys (Macaca mulatta)

DEBBIE GUATELLI-STEINBERG* AND JOHN R. LUKACSDepartment of Anthropology, University of Oregon, Eugene, Oregon 97403

Key words: primate dentition; enamel hypoplasia; dental enameldefects; stress indicators; crown height; enamel thickness;perikymata

ABSTRACT Three hundred and sixty rhesus macaque specimens at theCaribbean Primate Research Center were examined for evidence of linearenamel hypoplasia (LEH). A previously unreported intertooth pattern in LEHwas observed. Defects occur preferentially on the sectorial premolar of bothmales and females. Relative to other teeth, the sectorial premolar exhibitsmore prominent defects and is more likely to exhibit multiple defects. Thispattern is unlike the human intertooth LEH pattern and unlike patternspreviously reported for monkeys and apes. These observations are discussedin the context of factors thought to influence the intertooth distribution ofLEH in humans and in nonhuman primates. The authors reject crown height,the timing of crown development, and the duration of crown formation asfactors contributing to the observed pattern and favor an explanationinvolving enamel thickness, perikymata spacing, and/or prism orientation.Am J Phys Anthropol 107:179–186, 1998. r 1998 Wiley-Liss, Inc.

Linear enamel hypoplasia (LEH) is a de-velopmental defect of enamel appearing asone or more horizontal lines or grooves onthe surface of a tooth crown (Goodman andRose, 1990). The defect forms when physi-ological stress (such as febrile disease orpoor nutrition) disturbs enamel matrix for-mation, resulting in a deficiency of enamelthickness (Goodman and Rose, 1990; TenCate, 1994). Although many anthropologicalstudies have used LEH as a tool for assess-ing levels of stress in human populations,the use of this defect as a stress indicator inextant and extinct nonhuman primates isnot widespread.

In order to make use of LEH as a stressindicator in nonhuman primates, basic de-scriptive research in the patterning of thisdefect across and within species, popula-tions, and the tooth crowns themselves isneeded. Most previous research in nonhu-man primate LEH has focused either oninterspecific variation in LEH incidence

(Colyer, 1936, 1947; Moggi-Cecchi andCrovella, 1991; Schuman and Sognnaes,1956; Skinner and Guatelli-Steinberg, 1997;Vitzthum and Wikander, 1988) or on LEH inthe Hominoidea (Eckhardt, 1992; Jones andCave, 1960; Moggi-Cecchi and Crovella,1991; Skinner, 1986; Skinner and Roksan-dic, 1995; Skinner et al., 1995; Zhang, 1987).Less attention has been focused on LEH inthe Cercopithecoidea, especially with re-spect to intertooth variation in LEH expres-sion.

Moggi-Cecchi and Crovella (1991) foundthat within the Cercopithecoidea, enamelhypoplasia (defined as pits, lines, andgrooves) reaches its greatest expression on

Contract grant sponsor: National Center for Research Re-sources; contract grant number RR-03640; Contract grant spon-sor: NSF; contract grant number SBR 9615006.

*Correspondence to: Debbie Guatelli-Steinberg, Departmentof Anthropology, University of Oregon, Eugene, Oregon 97403.E-mail: guatelli@darkwing.uoregon.edu

Received 5 December 1997; accepted 4 July 1998.

AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 107:179–186 (1998)

r 1998 WILEY-LISS, INC.

the maxillary incisors and mandibular ca-nines. Similarly, Vitzthum and Wikander(1988) observed that, in cercopithecoids, mostdefects occur on the anterior teeth. Thesepatterns are like those observed in humans(Goodman and Armelagos, 1985) and Afri-can apes (Skinner, 1986; Vitzthum and Wi-kander, 1988), in which maxillary incisorsand mandibular canines are most often af-fected. LEH variation among teeth presum-ably arises as a result of differences indevelopmental timing, crown formation time,crown height, and/or developmental canali-zation (Goodman and Armelagos, 1985).

Both Moggi-Cecchi and Crovella (1991)and Vitzthum and Wikander (1988) com-bined observations from several species intheir cercopithecoid samples, making spe-cies-specific and population-specific pat-terns unobservable. By contrast, in this in-vestigation, intertooth variation in theexpression of LEH is studied in a singlelarge population of rhesus macaques. A clearand unexpected pattern is observed: LEH isexpressed preferentially on the sectorial pre-molar. This pattern is discussed in the con-text of factors thought to influence the distri-bution of LEH across the dentition.

MATERIALS AND METHODS

The Caribbean Primate Research Centerin Puerto Rico maintains the large collectionof Cayo Santiago rhesus macaque skeletonsthat formed the basis of this study (for adescription of the Cayo Santiago rhesusmacaque colony, see Rawlins and Kessler,1986). Permanent teeth of 360 specimens(179 males, 181 females) were examined forevidence of LEH. Teeth were observed underdiffuse lighting with a second, direct lightsource oriented obliquely to the specimen(see Goodman and Rose, 1990). A 310 handlens aided in defect identification. If one-halfor more of a tooth was not visible (due towear, breakage, or partial eruption), it wasconsidered to be unobservable and was notscored. Because LEH was observed alongthe honing surface of sectorial premolars,wear was an obvious problem. Sectorial pre-molars were not scored if one-half or more oftheir honing surfaces were covered withwear striae. The threshold for scoring LEHpresence was set deliberately low to include

mild stress events occurring during toothdevelopment (Skinner et al., 1995). Hillsonand Bond (1997:98) recently found that evensmall, microscopic furrows are ‘‘just as muchindicators of disturbance to growth as ...large defects.’’

Defects were rated as either mild or pro-nounced based on their width and depth. Itis not clear to what extent defect dimensionsreflect the severity of stress events. Sucklinget al. (1986) showed that when sheep areexperimentally infected with nematodes, theseverity of the systemic reaction influencesthe amount of missing enamel in the result-ing hypoplastic lesions. On the other hand,Hillson and Bond (1997) recently demon-strated that defect width and depth arestrongly influenced by the position of thedefect on the crown (as a consequence ofcrown growth geometry). Thus, no asser-tions will be made here about the meaning ofdefect dimensions. Instead, the defect ratingsystem employed is meant simply to give anindication of defect prominence.

Because there are no side differences inLEH expression (df 5 1, chi-square 5 0.0135,P , 0.9075), left and right teeth are pooledin the analysis. An individual is scored aspositive for the defect if LEH occurs on atleast one antimeric pair. This conventionmaximizes the possibility that the stressevents precipitating defect formation aresystemic rather than local (Goodman andRose, 1990). Intertooth variation in LEHexpression is analyzed only for individualsscoring positive for LEH, making the effec-tive sample size 61: 43 females, 18 males.There are fewer males than females scoringpositive for the defect because more malesthan females had worn sectorial premolarsthat were scored as unobservable. LEH pres-ence on a tooth is defined as the occurrenceof one or more defects and absence as a lackof any defects.

Both chi-square tests and generalized esti-mating equations (GEEs) are used to ana-lyze the results. The starting point for theGEE analysis is a generalized linear model,using logistic regression to model LEH pres-ence vs. absence (a binary response vari-able) and Poisson regression to model thenumber of hypoplastic defects (a count re-sponse variable) (SAS Institute Inc., 1996).

180 D. GUATELLI-STEINBERG AND J.R. LUKACS

The regression procedure fits the linearmodel to the data by maximum likelihoodestimation of the model’s parameters (SASInstitute, Inc., 1996).

To the generalized linear model, the GEEmethodology adds the ability to analyzecorrelated data sets, such as the presence orabsence of LEH on the teeth of a singleindividual. The GEE procedure makes weakassumptions about the actual correlation; itemploys a working correlation matrix thatapproximates the average dependenceamong repeated (or clustered) observationsover all subjects (Stokes et al., 1995). Bytaking the correlation into account, GEEsare sensitive to detecting differences in theLEH expression of different teeth within thesame specimen. The GEE procedure pro-vides an odds ratio that specifies the chancesof obtaining defects on one tooth vs. another.GEEs have become ‘‘an important strategyfor the analysis of correlated data whichmay arise from longitudinal studies...or clus-tering in which measurements are taken onsubjects who share a common characteris-tic’’ (SAS Institute Inc., 1998). In the presentanalysis, the correlated data are LEH counts(or presence/absence) on an individual’steeth; their shared characteristic or clusteris the individual specimen to which theteeth belong.

RESULTSPresence/absence

Table 1 shows, for the combined sample(males plus females), the percent of teethwith LEH for each tooth type in individualsscoring positive for the defect. Figure 1 is ahistogram of these frequencies.

A presence vs. absence chi-square analy-sis for the combined sample reveals highlysignificant differences among all teeth (df 511, chi-square 5 741.3, P , 0.001). Uppermolars are grouped into one category andlower molars into another in order to main-tain adequate cell size. Of all the presencecells in the chi-square table, the lower P3has the highest actual count relative to itsexpected count; molars have the lowest.

GEE analysis of presence/absence is con-ducted on the lower canine, lower incisors,lower premolars, and upper central incisor(other teeth had too few defects to be in-cluded). Relative to these teeth, the lower P3shows a significantly greater occurrence ofLEH. Table 2 shows the results of the GEEanalysis. The likelihood of defect presenceon the lower P3 relative to other teethaffected with LEH is given in the secondcolumn. A GEE analysis conducted on malesand females separately reveals similar rela-tionships between the lower P3 and otheraffected teeth (see Tables 3, 4;) (lower P4scould not be included because they had toofew defects.) The GEE analyses of intertoothdifferences in the presence of LEH revealthat within an individual the lower P3 has asignificantly greater likelihood of having adefect than any other tooth.

Defect counts

Table 5 displays the raw data for thenumber of teeth showing zero, one, or mul-tiple defects in the combined sample. Thelower premolar is often affected with mul-tiple defects.

The number of defects exhibited by lowercanines and maxillary central incisors iscompared to that of lower P3s. GEE analysisof the combined sample shows that lowerP3s are 6.23 times more likely to have agreater number of defects than upper cen-tral incisors (df 5 1, chi-square 5 63.1,standard error 5 0.2303, P , 0.0001) and3.96 times more likely than the lower ca-nines (df 5 1, chi-square 5 50.0, standarderror 5 0.1946, P , 0.0001).

Separate analyses of females and malesreveal a similar relationship. For females,the lower P3 is 14.60 times more likely tohave a higher number of defects than theupper central incisor (df 5 1, chi-square 5

TABLE 1. Percent of tooth class in each jaw exhibitingdefects in 61 individuals scoring positive for LEH

Toothtype

Totalteeth in

lower jawsample

Percentof teeth

affected inlower jaw

Totalteeth in

upper jawsample

Percent ofteeth

affected inupper jaw

I1 111 11.7 109 15.6I2 108 3.7 112 3.6C 104 25.0 108 0.9P3 105 80.0 120 0.3P4 119 5.0 119 0.0M1 116 0.0 122 0.0M2 122 1.6 119 0.8M3 98 0.0 88 0.2

181LEH IN RHESUS MONKEYS

40.6, standard error 5 0.4208, P , 0.0001)and 3.29 times more likely than the lowercanine (df 5 1, chi-square 5 29.9, standarderror 5 0.2178, P , 0.0001). For males, thelower P3 is 3.24 times more likely to have ahigher number of defects than the uppercentral incisor (df 5 1, chi-square 5 15.3,standard error 5 0.3005, P , 0.0001) and7.20 times more likely than the lower canine(df 5 1, chi-square 5 20.0, standard error 50.4410, P ,0.0001). Thus, the GEE analysesfor defect counts demonstrate that within anindividual the sectorial premolar is morelikely than any other tooth to exhibit mul-tiple defects.

Defect prominence

The greatest defect category assigned toany of the defects on a tooth is the tooth’sLEH prominence value. Therefore, a toothitself is characterized as having pronouncedLEH if one or more of the tooth’s defects areclassified as pronounced.

The LEH prominence values of the uppercentral incisors, lower canines, and lowerP3s are compared with each other in achi-square test. There are significantly differ-ent LEH prominence values for these teeth(df 5 4, chi-square 5 117.3, P , 0.001). Inthe pronounced category, the lower P3 alonehas a higher actual cell count (37) relative to

Fig. 1. Percent of tooth type exhibiting one or more defects in individuals scoring positive for LEH(actual percents given in Table 1).

TABLE 2. Presence of LEH1 on the lower P3 relative toother teeth

Tooth

Rate atwhich LEH

is morelikely tooccur on

P3 relativeto tooth

listed at left dfStandard

error

Chi-squarevalue

Pvalue

Upper I1 21.65 1 0.3595 73.2 0.0001Lower I1 30.15 1 0.3830 79.1 0.0001Lower I2 104.00 1 0.5649 67.6 0.0001Lower C 12.00 1 0.3329 55.7 0.0001Lower P4 75.33 1 0.4848 79.5 0.00011 Sexes combined (61 individuals scoring positive for LEH).

TABLE 3. Presence of LEH on the lower P3 relative toother teeth in females1 only

Tooth

Rate atwhich LEH

is morelikely tooccur on

P3 relativeto tooth

listed at left dfStandard

error

Chi-squarevalue

Pvalue

Upper I1 63.92 1 0.5824 51.0 0.0001Lower I1 22.90 1 0.4246 54.4 0.0001Lower I2 139.00 1 0.7668 41.4 0.0001Lower C 10.00 1 0.3784 37.0 0.00011 Forty-three individuals scoring positive for LEH.

182 D. GUATELLI-STEINBERG AND J.R. LUKACS

its expected cell count (15). The chi-squareanalysis could not be conducted for males(because of low cell counts), but the compari-son for females is similar to that for thecombined sample. For females, differencesin prominence are significant (df 5 4, chi-square 5114.0, P , 0.001), and the lower P3has a higher actual pronounced cell count(31) than is expected (12).

Figure 2 is a photograph of a pronounceddefect on a left lower P3 of a male rhesusmacaque. No LEH was observed on the leftlower canine. Figure 3 is a closeup of apronounced defect on a left lower P3 of asecond rhesus male.

DISCUSSION AND CONCLUSIONS

The data presented here demonstrate apattern of LEH distribution within the den-tal arcade that has not previously beennoted in a nonhuman primate species. Thesectorial P3 exhibits LEH preferentially, ismore likely to have multiple defects, and hasdefects that appear more pronounced in thatthey are deeper and wider than those ofother teeth. Moggi-Cecchi and Crovella(1992) noted that the P3s of cercopithecoidsoften showed LEH. However, these authorsreported that cercopithecoid maxillary inci-sors and mandibular canines were moreoften affected than lower P3s. Previous re-searchers may not have noted the patterndescribed in this paper because this study isthe first to have examined intertooth LEHvariability in such large numbers of rhesusmonkeys.

Moggi-Cecchi and Crovella (1991) andGoodman and Rose (1990) consider canineteeth to be more vulnerable to LEH as a

result of their crown height. The depositionof large amounts of enamel on the highcanine crown is thought by these authors tomake ameloblasts more vulnerable to disrup-tion. In the view of Skinner et al. (1995),canine teeth simply have a greater opportu-nity to record disruptions since they formover longer time periods than other teeth. Inlight of these explanations, the preferentialexpression of LEH on the mandibular P3 ofrhesus monkeys is interesting. There arecurrently no published data on the calcifica-tion times of lower P3s and canines inrhesus monkeys. However, mandibular P3sand canines take approximately the sameamount of time to calcify in pig-tailed ma-caque (M. nemestrina) females; in M. nemes-trina males, lower P3s form over a shortertime period than lower canines (Sirianniand Swindler, 1985). If tooth development inM. mulatta is similar to that of M. nemes-trina, then the duration of calcification doesnot offer a clue as to preferential P3 LEHexpression in rhesus monkeys. An explana-tion in terms of crown height is also unten-able. In the females of this rhesus sample,the mandibular P3 has a slightly shortercrown height (an average of 8.6 mm over 21specimens) than the mandibular canine (anaverage of 9.6 mm over 18 specimens). Inmales, the average mandibular canine crownheight is 15.4 mm (19 specimens), and theaverage lower P3 crown height is 13.2 mm.(These are unpublished measurements takenby the first author on this sample.)

Although it is possible that lower P3s arecalcifying at times when other teeth are notand perhaps when stress events are occur-ring more frequently, this does not seemlikely for the following reason. In M. nemes-trina females, the lower P3 and mandibularcanine form simultaneously (Sirianni andSwindler, 1985). Assuming that female rhe-sus monkeys have a calcification schedulesimilar to that of female pig-tailed ma-caques, one would expect that the lower P3and mandibular canine would exhibit de-fects equally if they were equally suscep-tible. However, in the female rhesus samplestudied here, the lower P3 is ten times morelikely to exhibit defects than the lower ca-nine. These findings are exactly oppositethose reported by Condon (1981) for hu-

TABLE 4. Presence of LEH on the Lower P3 relative toother teeth in males1 only

Tooth

Rate atwhich LEH

is morelikely tooccur on

P3 relativeto tooth

listed at left dfStandard

error

Chi-squarevalue

Pvalue

Upper I1 9.29 1 0.6465 11.9 0.0006Lower I1 78.75 1 0.9116 22.9 0.0001Lower I2 78.13 1 0.9112 22.6 0.0001Lower C 21 1 0.7113 18.3 0.00011 Eighteen individuals scoring positive for LEH.

183LEH IN RHESUS MONKEYS

mans. In comparing 30 matched mandibularcanines and lower P3s, Condon (1981) founddefects on the canine that could not bematched to any defects on the premolar,

despite that fact that these teeth developsimultaneously in humans.

If the duration of crown formation, crownheight, and the timing of crown develop-ment are not involved, then what factorsmay explain the rhesus pattern? A numberof characteristics of the sectorial premolarmay make it more susceptible to defectexpression. The tooth is highly angled be-cause of its functional relationship as a‘‘whetstone’’ for the upper canine. Perhapsthe enamel prisms are oriented to the enamelsurface in such a way as to make defectsmore likely to appear. Goodman and Armela-gos (1985) explain human LEH intratoothpatterns in these terms. Alternatively, peri-kymata spacing may differ between man-dibular canines and P3s, perhaps influenc-ing defect definition. Hillson and Bond (1997)have shown that defect dimensions varywithin tooth crowns mainly because of varia-tion in the spacing of perikymata grooves. Inaddition, Goodman and Armelagos (1985)suggest that ameloblasts forming enamel ata high rate may be more vulnerable todisruption than ameloblasts with less tax-ing enamel production schedules. Recently,Washburn (1997) demonstrated that the sec-torial premolars of both male and female oldworld anthropoids are covered with a thicklayer of enamel (presumably related to thehoning function of this tooth). If ameloblastsare secreting a thicker layer of enamel onthe sectorial P3 than they are on the man-dibular canine during the same span of time,

TABLE 5. Number of teeth exhibiting 0, 1, or multiple defects1

Tooth

Number of defects in lower jaw Number of defects in upper jaw

0 1 2 3 4 0 1 2

I1 98 9 4 — — 92 12 5(88.3) (8.1) (3.6) (84.4) (11.0) (4.6)

I2 104 2 2 — — 108 4 —(96.2) (1.9) (1.9) (96.4) (3.6)

C 78 21 3 2 107 — 1(75.0) (20.2) (2.8) (2.0) (99.1) (0.9)

P3 21 49 25 7 3 117 3 —(20.0) (46.7) (23.8) (6.7) (2.8) (97.5) (2.5)

P4 113 6 — — — 119 — —(95.0) (5.0) (0.0)

M1 116 — — — — 122 — —(100) (0.0)

M2 120 2 — — — 118 1 —(98.4) (1.6) (99.2) (0.8)

M3 98 — — — — 86 2 —(0.0) (97.7) (2.3)

1 Percents in parentheses.

Fig. 2. Pronounced defect on the left mandibular P3;defects absent on the lower left canine.

Fig. 3. Pronounced defect on the left mandibular P3.

184 D. GUATELLI-STEINBERG AND J.R. LUKACS

then ameloblasts of the lower P3 might bemore easily disrupted.

Goodman and Rose (1990) argue that de-velopmental canalization may explain thehigh rate of enamel defects on functionallykey teeth such as the teeth of the sectorialcomplex (maxillary canine, sectorial premo-lar). If development is highly canalized infunctionally key teeth, the sectorial complexmay be unable to alter its developmentalpath in response to systemic stress. Garn etal. (1966) noted that the second incisor,canine, and first premolar show similar pat-terns of sexual dimorphism, indicating thatthese teeth develop as part of a caninegrowth field. The canine growth field couldhelp explain why the lower canine and secto-rial premolar both have elevated defect fre-quencies, but it does not help elucidate thecause(s) of preferential LEH expression onthe sectorial premolar.

It is not clear to what extent the observedpattern characterizes other populations ofrhesus monkeys or other species of cercopi-thecines. This population differs from feralpopulations of rhesus monkeys in that it isprovisioned, but we can see no reason whyprovisioning would influence the intertoothdistribution of defects. Peak infection by thenematode parasite Strongyloides occurs inyearling and 2-year-old Cayo Santiago mon-keys (Kessler et al., 1984). While this age ofinfection overlaps with the time span forsectorial premolar calcification, it also over-laps with that for the lower canine, based oncalcification schedules in M. nemestrina (Si-rianni and Swindler, 1985). As emphasizedearlier, differences in calcification timingcannot explain the greater expression ofLEH on the sectorial premolar relative tothe lower canine.

Unpublished data by the first author onsmall samples of other LEH-positive cercopi-thecines indicates that similar patterns canbe found in M. fascicularis (N 5 8) and inPapio (N 5 13). In the M. fascicularis sam-ple, the mandibular first incisor is followedby the sectorial premolar in having thegreatest incidence of LEH. In the Papiosample, the sectorial premolar is second tothe maxillary central incisor in being mostoften affected. Interestingly, in several of thePapio specimens, pronounced defects oc-

curred on the sectorial premolar, while moremild defects (or no defects at all) were foundon the lower canine. The greater expressionof LEH on the sectorial premolar relative tothe mandibular canine in these samples ofM. fascicularis and Papio requires furtherinvestigation in much larger samples (and aspecies-by-species examination in Papio).

An investigation into the causes of secto-rial premolar LEH susceptibility in rhesusmonkeys may be able to further our under-standing of the reasons teeth vary in theirability to reflect systemic stress events. Forreasons explained above, we believe thatdifferences in crown height, the duration ofcrown formation, and developmental timingare not likely explanations of the heightenedexpression of LEH on the sectorial premolarrelative to other teeth. Instead, we favor anexplanation involving prism orientation,enamel thickness, perikymata spacing, andto a lesser extent developmental canaliza-tion. Currently we cannot discern the rela-tive contribution of these factors, but futureresearch could help illuminate these poten-tial causes. Prism orientation and enamelthickness could be examined by comparativehistological analysis of mandibular caninesand P3s. Comparison of perikymata groovespacing on the mandibular canine and P3would illuminate this potential explanation.

Whatever the cause(s), these data haveimportant practical applications to the studyof LEH in nonhuman primates. It is commonpractice for LEH researchers to focus onlyon teeth most likely to exhibit the defect.Therefore, when time is limited, researchersoften examine only the mandibular caninesand maxillary incisors for LEH (as recom-mended by Goodman and Rose, 1990). Be-cause large samples of each species of nonho-minoid primates have yet to be studied forLEH, there is a possibility that intertoothpatterns will vary from species to species.Thus, as is evidenced by this study, if onewere to examine only maxillary incisors andmandibular canines in nonhuman primates,a substantial number of defects could bemissed, resulting in an underestimation ofLEH frequencies. Until more basic researchon the intertooth distribution of LEH innonhuman primates is conducted, it will beimportant for researchers using LEH as a

185LEH IN RHESUS MONKEYS

stress indicator to examine all teeth of thepermanent dentition.

ACKNOWLEDGMENTS

This investigation was supported in partby an animal resources branch programaward, RR-03640, from the National Centerfor Research Resources, National Institutesof Health, and the University of Puerto Rico,Medical Sciences Campus. Funding was alsoprovided by an NSF dissertation improve-ment grant, SBR 9615006. The authorsthank the following people for their helpfulsuggestions in the preparation of this re-search: Jean Turnquist, Nancy Hong, JohnBerard, Daris Swindler, Mark Skinner, Ja-copo Moggi-Cecchi, and Bruce Floyd. Thanksare also due the following people for theirsupport: Dan Steinberg, Roslyn Steinberg,Rose Guatelli, Benedikt Hallgrimsson, andMyrna and Fedelia Reyes. Special thanks toRobin High for his statistical guidance andto Doug Ferguson for his assistance withgraphics.

LITERATURE CITED

Colyer F (1936) Variations and Diseases of the Teeth ofAnimals. London: John Bale and Sons and Danielson.

Colyer F (1947) Dental disease in animals. Br. Dent. J.82:1–10.

Condon KW (1981) Intratooth associations of develop-mental enamel defects. Am. J. Phys. Anthropol. 63:147.

Eckhardt RB (1992) Tooth crown development: Non-human primate perspectives on the interpretation oflinear enamel hypoplasia frequencies in present andpast hominid populations. In AH Goodman and LLCapasso (eds.): Recent Contributions to the Study ofEnamel Developmental Defects. J. Paleopathol., Mono-graphic Publications 2. Teramo, Italy: Edigrafital, pp.293–306.

Garn SM, Kerewsky RS, and Swindler DR (1966) Ca-nine ‘‘field’’ in sexual dimorphism of tooth size. Nature212:1501–1502.

Goodman AH, and Armelagos GJ (1985) Factors affect-ing the distribution of enamel hypoplasias within thehuman permanent dentition. Am. J. Phys. Anthropol.68:479–493.

Goodman AH, and Rose JC (1990) Assessment of physi-ological perturbations from dental enamel hypoplasiaand associated histological structures. Yrbk. Phys.Anthropol. 33:59–110.

Hillson S, and Bond S (1997) Relationship of enamelhypoplasia to the pattern of tooth crown growth: Adiscussion. Am. J. Phys. Anthropol. 104:89–103.

Jones TS, and Cave AJE (1960) Diet, longevity, anddental disease in the Sierra Leone chimpanzee. Proc.Zool. Soc. Lond. 135:147–155.

Kessler MJ, Yarbrough B, Rawlins RG, and Berard J(1984) Intestinal parasites of the free-ranging CayoSantiago rhesus monkeys (Macaca mulatta). J. Med.Primatol. 13:57–66.

Moggi-Cecchi J, and Crovella S (1991) Occurrence ofenamel hypoplasia in the dentitions of simian pri-mates. Folia Primatol. 57:106–110.

Moggi-Cecchi J, and Crovella S (1992) Dental develop-mental defects in the dentitions of Dryopithecus fromSpain. Antropologia Contemporanea 15:47–50.

Rawlins RG, and Kessler MJ (eds.) (1986) The CayoSantiago Macaques. Albany: State University of NewYork Press.

SAS Institute Inc. (1996) SAS/STATt Software: Changesand Enhancements through Release 6.11. Cary: SASInstitute, Inc.

SAS Institute Inc. (1998) URL: http://www.sas.com/rnd/app/stat/gee612.html

Schuman EL, and Sognnaes RF (1956) Developmentalmicroscopic defects in the teeth of subhuman pri-mates. Am. J. Phys. Anthropol. 14:193–214.

Sirianni JE, and Swindler D (1985) Growth and Develop-ment of the Pigtailed Macaque. Boca Raton: CRCPress.

Skinner MF (1986) Enamel hypoplasia in sympatricchimpanzee and gorilla. Hum. Evol. 1:289–312.

Skinner M, and Guatelli-Steinberg D (1997) Distribu-tion of linear enamel hypoplasia in sympatric mon-keys and apes. Am. J. Phys. Anthropol. 24(Suppl.):213(abstract).

Skinner MF, and Roksandic M (1995) Malaria in Chi-nese fossil apes. In XIA Wuping and Z Yongzu (eds.):Primate Research and Conservation. Beijing: ChinaForestry Publishing House, pp. 92–97.

Skinner MF, Dupras TL, and Moya-Sola S (1995) Period-icity of linear enamel hypoplasia among MioceneDryopithecus from Spain. J. Paleopathol. 7:195–222.

Stokes MG, Davis CS, and Koch GC (1995) CategoricalData Analysis Using the SAS System. Cary: SASInstitute.

Suckling G, Elliot DC, and Thurley DC (1986) Themacroscopic appearance and associated histologicalchanges in the enamel organ of hypoplastic lesions ofsheep incisor teeth resulting from induced parasitism.Arch. Oral Biol. 31:427–439.

Ten Cate AR (1994) Oral Histology: Development, Struc-ture, and Function. St. Louis: CV Mosby.

Vitzthum VJ, and Wikander R (1988) Incidence andcorrelates of enamel hypoplasia in non-human pri-mates. Am. J. Phys. Anthropol. (Suppl.) 75:284(abstract).

Washburn A (1997) Redundant P3 enamel among someOld World anthropoids: A facet to be considered. Am.J. Phys. Anthropol. 24(Suppl.):236 (abstract).

Zhang Y (1987) Enamel hypoplasia of Gigantopithecusblacki. Acta Anthropologica Sinica 6:175–179.

186 D. GUATELLI-STEINBERG AND J.R. LUKACS