6. 3-D interferometric microscopy applied to the study of buccalenamel microwear

F. ESTEBARANZSecc. Antropologia, Dept. Biologia AnimalFac. Biologia, Universitat de BarcelonaAvgda. Diagonal 645, 08028 Barcelona. Spainestebaranz@ub.edu

J. GALBANYSecc. Antropologia, Dept. Biologia AnimalFac. Biologia, Universitat de BarcelonaAvgda. Diagonal 645, 08028 Barcelona. Spainjgalbany@ub.edu

L.M. MARTÍNEZSecc. Antropologia, Dept. Biologia AnimalFac. Biologia, Universitat de BarcelonaAvgda. Diagonal 645, 08028 Barcelona. Spainlmartinez@ub.edu

A. PÉREZ-PÉREZSecc. Antropologia, Dept. Biologia AnimalFac. Biologia, Universitat de BarcelonaAvgda. Diagonal 645, 08028 Barcelona. Spainmartinez.perez-perez@ub.edu

Keywords: SEM, interferometry, microwear, enamel, hominoid

Abstract

Dental microwear analysis is based on the assumption that a correlation exists between ingested diet andmicrowear patterns on the enamel surface of teeth, such that diet can be reconstructed by quantifying enamelmicrowear. Abrasive particles, such as plant phytoliths or silica-based sands incorporated into food items, alongwith food processing techniques and tooth morphology, are responsible for the microwear features observed.Dental microwear has been extensively studied in both extant and extinct primates, including human popula-tions. The dietary and ecological information that can be derived from dental microwear analyses makes ita technique useful for analyzing non-primate species, such as muskrats, sheep, bats, moles, antelopes, pigsand even dinosaurs. In the attempt to reconstruct species’ ecology and diet, microwear research has becomea successful procedure. The proliferation and persistence of different methods to quantify microwear patterns

391

S.E. Bailey and J.-J. Hublin (Eds.), Dental Perspectives on Human Evolution, 391–403.© 2007 Springer.

392 Estebaranz et al.

require very accurate definitions of microwear variables, since inter-observer error rates cannot be neglected.The use of semiautomatic methods to quantify microwear features does not guarantee low inter-observer erroraffecting dental microwear results. Error can be caused by taphonomy, microscopy drawbacks of back-scatteredelectrons, or differences in SEM reproducibility depending on sample shape and orientation. However, fullyautomatic procedures lack discrimination between ante-mortem and post-mortem wear processes that affect toothenamel at various degrees, and their application requires experienced control and evaluation.

Introduction

Plant foods contain significant amounts ofphytoliths in their tissues, such that dentalmicrowear is directly related to ecologicalconditions and diet composition (Teaford,1994; Ungar and Teaford, 1996; Ungar, 1998).Seeds, shoots, and inflorescence (Ball et al.,1996) are among the main food items that havean effect on enamel microwear (Danielsonand Reinhard, 1998; Gügel et al., 2001),although food processing techniques canalso influence enamel microwear by incorpo-rating dust and ashes into the ingested foods(Teaford and Glander, 1991, 1996; Ungar,1995; Daegling and Grine, 1999). In addition,tool technology plays also an important role(Teaford, 1991; Pérez-Pérez et al., 1994). Itshould be noted that diet-related variables donot affect dental enamel in isolation, but thatgnathic morphology must also be taken intoaccount (Gordon, 1982).

Dental microwear of both extinct and extantprimates has been widely studied (Ryan,1979; Gordon, 1982; Kay, 1987; Teaford andRunestad, 1992; Ungar, 1992, 1994, 1996;Ungar et al., 1995; King et al., 1999b; Galbany,2004; Godfrey et al., 2004; Nystrom et al.,2004; Galbany et al., 2005b), including humanpopulations (Grine, 1986; Lalueza and Pérez-Pérez, 1993; Lalueza et al., 1996; Martínezet al., 2004; Pérez-Pérez et al., 1999, 2003).Enamel microwear analysis has proven to behighly informative regarding dietary habits andpaleoecology, and has been applied to a widerange of taxa, including muskrats (Lewis et al.,2000), sheep (Mainland, 2003), bats (Strait,1993), moles (Silcox and Teaford, 2002),

antelopes (Solounias and Hayek, 1993), pigs(Ward and Mainland, 1999), and suids (Hunterand Fortelius, 1994).

When attempting to reconstruct species’ecology and diet, dental microwear researchhas thus become a successful line of research.The first dental microwear papers werepublished in the 1950’s (Butler, 1952; Mills,1955), although no quantitative results weregiven. It was not until the 1980s thatseveral authors proposed alternative methodsto quantify microwear features (Gordon, 1982,1984; Grine, 1986; Ungar et al., 1991, 1995).However, the persistence of an abundanceof different methods to quantify microwearpatterns greatly limits the comparison ofresults among researchers (Grine et al.,2002; Galbany, 2005a). Due to high inter-observer error rates, Grine et al. (2002)proposed the adoption of Microware 4.0 (byP. Ungar) as standard software for semi-automatic microwear analysis. Nevertheless,Galbany (2005a), using Sigma Scan Pro 5.0(by SPSS) have shown that error rates areindependent of the software used, but arehighly dependent on how variables are definedand the researcher’s expertise (Pérez-Pérezet al., 1999; Galbany et al., 2004b).

The use of semiautomatic methods doesnot guarantee reliability of results becausevarious sources of measurement error persist,such as using back-scattered or secondaryelectrons in SEM observation (Pérez-Pérezet al., 2001; Galbany, et al., 2004b), varyingthe working distance, or any surface tilt.SEM images, which depend on the shapeand orientation of the sample (Gordon, 1982;King et al., 1999b; Ungar, 2003), do not

Dental Enamel Roughness and Microwear 393

generally reproduce the exact tooth surface.In addition, recent studies have demonstratedthe existence of important inter-observer andintra-observer error rates due to subjectivityof criteria used to measure microwear features(Grine et al., 2002; Galbany, 2005a). Finally,taphonomic analyses on dental microwearshow that post-mortem wear produces ageneralized polish of dental enamel, ratherthan the addition of microwear features,both in occlusal (King et al., 1999a) andbuccal tooth surfaces (Martínez and Pérez-Pérez, 2004).

Although the generalized use of a fullyautomatic quantification procedure might beuseful to prevent subjectivity and inter-observer error (Ungar, 2003; Galbany, 2005a),an understanding of the influence of post-depositional processes on enamel microwearis required. The aim of the present contri-bution is to test whether or not post-mortemenamel microwear can be distinguished fromdiet-related microwear patterns on buccaltooth surfaces. Dental microwear on thebuccal enamel surface is not influenced bytooth-to-tooth contact during food chewingand, thus, reflects only diet (Galbany, 2004;Pérez-Pérez, 2004). In addition, on the buccalsurfaces pits are absent, and researchers haveto deal only with scratches. Pits are frequentlyvery difficult to characterize due to overlapand their high variability in shape and size(Pérez-Pérez, 2004). Automatic measuringprocedures must demonstrate that cleartopographic differences exist between post-mortem eroded surfaces and well-preservedenamel. Topographic 3-D techniques havebeen successfully applied to the analysis oftooth morphology and wear (Mayhall andKageyama, 1997; Reed, 1997; Zuccotti et al.,1998; Jernvall and Selänne, 1999; Ungar andWilliamson, 2000; Kaiser and Katterwe, 2001;Ungar and M’Keirera, 2003; Dennis et al.,2004; Ungar, 2004). However, all method-ological procedures based on topographicanalysis, at least for tooth surfaces, are bound

to be highly sensitive to enamel preservation.An indiscriminate analysis of enamel surfaceswithout considering preservation is fated tobe meaningless in terms of interpretationof dietary adaptations and ecology of fossilpopulations.

Materials and Methods

The study teeth came from a wide collectionof tooth casts curated at the University ofBarcelona. The molds were obtained duringthe course of an international collaborativeproject on dental microwear (Galbany et al.,2004b). Tooth crown molds were obtained withPresident MicroSystem Regular Body (Coltène®) polyvinylsiloxane. This impression materialis widely used in dental microwear research(Ungar, 1996; Ungar and Spencer, 1999).It reproduces features with resolutions to afraction of a micron (Teaford and Oyen,1989), maintains the resolution for many years(Beynon, 1987), and shows an excellent dimen-sional stability and reproduction detail (Andrit-sakis and Vlamis, 1986). Resin positive replicaswere obtained from the molds using epoxy resinEpo-Tek #301 as well as polyurethane FeropurPR-55, both showing the same resolution as theoriginal tooth (Galbany et al., 2004b). Once theresin or polyurethane casts were dry, they weremounted on aluminum stubs with term fusiblegum. An argent belt (Electrodag 1415M-Acheson Colloiden) was applied between theplastic cast and the aluminum stub, and allcasts were sputter-coated with a 40A goldlayer (see Galbany et al, 2004b for a moredetailed description of these methodologicalprocedures).

The sample included 57 teeth (one tooth perindividual was studied) from three hominoidgenera (Gorilla N = 13, Pan N = 4; PongoN = 8) and three hominid species (Australop-ithecus anamensis N = 2, A. afarensis n = 22,and A. africanus N = 8). Whenever available,the left M2 was selected. Otherwise, the rightM2 was chosen. If neither was available,

394 Estebaranz et al.

then the left M1 or, in its absence, the rightM1 was studied. Despite some studies inwhich upper teeth were chosen when thelower ones were was missing (King et al.,1999b), in the present analysis only thelower dentition was studied. The lower M2

has been extensively studied and the liter-ature on the M2 is quite rich (Gordon, 1982;King et al, 1999b; Nystrom et al., 2004;Ungar, 2004). No differences were expectedin enamel roughness and microwear betweenM1 and M2 teeth, although no evidence forthis is currently available for the samplestudied.

For each tooth, a SEM micrograph ofthe enamel surface was obtained with aCambridge Stereoscan-120 scanning electronmicroscope following usual procedures formicrowear research (Lalueza et al., 1996;Ungar and Teaford, 1996; Pérez-Pérez et al.,1999; Pérez-Pérez et al., 2003; Pérez-Pérez,2004). SEM working parameters includedacceleration voltage of 15 kV, with workingdistance ranging between 18 and 25 mm,and 100X magnification. All images weretaken at the medial third of the buccalsurface, avoiding both cervical and occlusalthirds (Pérez-Pérez et al., 1999). Images werethen processed with Adobe ® Photoshop ®6.0. A semiautomatic measure of striationdensity (total number of striations over theanalysed surface of 0.56 mm2) was obtainedusing SigmaScan Pro ® 5.0 (SPSS). Striationdensity on the buccal surface of the humandentition has been shown to depend on theabrasiveness of food, and it appears to bea reliable measure of dietary habits in bothancient and modern human hunter-gathererpopulations (Pérez-Pérez et al., 1994; Laluezaet al., 1996; Pérez-Pérez et al., 1999, 2003;Pérez-Pérez, 2004).

At the same time, 3-D topographic imagesof the same buccal tooth surfaces wereobtained using a Veeco NT 1100 interphero-metric microscope housed at the Nanotech-nology Platform at the “Parc Científic” of

the University of Barcelona. Each sample wasplaced perpendicular to the objective axis.All of the 3-D surfaces were analysed at50X magnification in VSI mode to control z-axis movements, and with the quality levelset to ‘full’ in order to achieve the highestresolution possible in pixels per area. A totalof 199 3-D topographies (Gorilla N = 47,Pan N = 13, Pongo N = 20, A. anamensisN = 8, A. afarensis n = 79, and A. africanusN = 32) were analyzed on the 57 teeth in thestudy sample. All roughness measures wereobtained for a 124.4 × 94.6 �m surface topog-raphy. Some image treatment was performedbefore an automatic measure of surfaceroughness was derived for each image. First,a histogram of height measures (z score)was drawn to assess the distribution ofindividual surface parameters. Backgroundnoise was reduced by applying a maskfilter that eliminated low frequency z scorescorresponding to false peaks or valleys.A median smoothing pass filter was thenapplied to filter out ‘noisy and spiky’ data(as indicated in the Veeco reference package).This median filter is particularly effective forpreserving edges and steps in the data. Finally,several measures of surface roughness werecalculated (Rp, Rv, Rt, Rq, and Ra). Rp isthe maximum profile peak height, measuredas the height difference between the meanline and the highest point over the evalu-ation length; Rv is the maximum profile valleydepth, measured as the height differencebetween the mean line and the lowest pointover the evaluation length; Rt is the peak-to-valley difference calculated over the entiremeasured array; Rq is the root-mean-squaredroughness calculated over the entire measuredarray; and Ra is the average roughnesscalculated over the entire measured array,calculated according to the ANSI B46.1standard.

The independent measures of striationdensity and surface roughness obtained onthe buccal enamel surfaces of the same

Dental Enamel Roughness and Microwear 395

teeth were analyzed to test the alternativehypotheses that: (1) a significant corre-lation between microwear striation density andsurface roughness can be observed (indicativeof a significant association between automaticmeasures and dietary related habits), andthat (2) automatic measures of surfaceroughness can be used to discriminate betweendifferent types of enamel surfaces, includingwell preserved enamel, post-mortem erodedsurfaces, or even inter-proximal enamel facets.

In order to test the first hypothesis,SEM pictures were classified into oneof four categories of degree of post-mortem damage (Figure 1): nil, no clearsigns of erosion, numerous dietary stria-tions present (1); moderate, some evidenceof erosion, though numerous striations stillvisible (2); marked, clear signs of erosion,low number of striations visible (3); andintense, generalized erosion, striations hardlyvisible (4).

Results

All measures of surface roughness were highlycorrelated. In fact, they are all estimationsof the overall difference between minimumand maximum z scores. High peak-to-valleydistances are indicative of an uneven processof enamel wear affecting certain patchesof enamel more intensively than others.Homogeneous wear processes would erodeaway the enamel without increasing theroughness values, whereas uneven wear mightincrease roughness throughout the analyzedenamel surfaces. The overall paired linearcorrelation between the roughness measuresand the degree of post-mortem wear wasnegative and highly significant. In all species,high post-mortem enamel damage was corre-lated with low surface roughness (Figure 2).Consequently, taphonomic, post-depositionalprocesses may be significant factors toconsider when recording automatic measures

Figure 1. Sample images showing the four categories post-mortem damage: (1) nil, no post-mortemabrasion, (2) low, (3) moderate and (4) high.

396 Estebaranz et al.

Figure 2. Linear regression between post-mortem damage and buccal striation density (NT) byspecies. A clear reduction of striation density with increasing post-mortem damage is seen. Sample

SEM images of some analyzed teeth are shown in the plot.

of surface relief. However, there is no clearassociation between striation densities andsurface roughness. While striation densitytends to decrease with increasing enameldamage, roughness increases in slightlydamaged surfaces (compared to non-damagedenamel), and decreases in medium and highlydamaged enamel (Figure 3).

The relationship between striation densityand enamel roughness depends on post-mortem damage of the enamel surface.When we controlled for this factor, signif-icant quadratic regression coefficients wereobtained, demonstrating that roughness tendsto increase if enamel damage is slight, butsharply decreases with higher degrees of post-mortem damage (Figure 4). Even by species,those teeth heavily affected by enamel erosionshowed a clear tendency toward reduced

numbers of striations, whereas moderate ratesof erosion showed higher roughness values,and highly striated surfaces showed reducedroughness (Figure 5).

Discussion

Despite the initial goal of this research –to infer the usefulness of automatic proce-dures to characterize diet-related behavior bycomparing roughness and striation density –it became apparent that taphonomic processesare of major concern to topographic analyses.As post-mortem abrasion smoothes the tooth,the microwear density pattern decreases.Intensive erosion on enamel surfaces iscertain to cause a progressive obliterationof microwear features that would imply aninitial increase in surface roughness, followed

Dental Enamel Roughness and Microwear 397

Figure 3. Box plots of striation density NT (left) and roughness values RT (right) for the four degreesof enamel damage (nil, low, moderate and high). The background category includes casts of enamelsurfaces that could not be classified into one of the previous groups because of bad preservation or

poor replication. The IP group included well-preserved surfaces of inter-proximal facets. No commonpattern of variation of NT and RT by enamel damage categories was observed.

by complete polishing of the enamel, with areduction in overall roughness.

Although it is clear that highly scratchedenamel surfaces show higher values ofenamel roughness than do smooth or polishedsurfaces, this relationship is not indicative ofa meaningful association between automaticmeasures of roughness and dietary habits. Theprimary reason is that roughness is highlysensitive to post-mortem damage. At very lowlevels of erosion, an increase in roughnesscould be indicative of higher microwearfeature densities, which is a variable clearlydependent on dietary habits. However, furthereffects by erosive factors rapidly evensurface topography, dramatically reducingroughness.

The second hypothesis to be tested ismore complicated. In an optimal situation,automatic measures of 3-D topographic relief

are required to distinguish not only wellpreserved from eroded surfaces, but alsofrom other surfaces not informative forinferring diet-related behavior. Roughnessmeasures of surface topography do not appearto be useful for this purpose, as similarmeasures of roughness can be derived fromcompletely different enamel surfaces, suchas well preserved surfaces, enamel withexposed enamel prisms or inter-proximalenamel facets not exposed to food itemsduring food chewing (Figure 6). Automaticprocedures still require a thorough processof image selection by the researcher priorto analysis, but a certain degree of inter-observer subjectivity still remains in deter-mining the frontier between no erosion andsome incipient erosion.

However, surface analysis is not limited toroughness variables. Measures of anisotropy

398 Estebaranz et al.

Figure 4. Quadratic regressions between striation density (NT) and average roughness (RA) for allthe post-mortem damage categories. Two distinct patterns were noted: nil (1) and low (moderate in

the plot) damaged enamel surfaces showed higher variability of striation densities (ranging from 100to 300), whereas moderate (marked in the plot) and marked (intense in the plot) showed reduced

values of striation densities.

over entire enamel surfaces are capable ofcharacterizing heterogeneity of surface reliefon an array of orientations, and scale-sensitivefractal analyses of surface topography caneasily depict differences between smooth andrough enamel surfaces. However, anisotropyand fractal measurements, as well as roughness,are quantifications of the same underlyingphenomenon: the heterogeneity of a peak-to-valley score throughout a three dimen-sional space. Ideally, unscratched and unerodedenamelsurfaces,correctedforspatialcurvature,would show no orientation anisotropy, nofractal variations to scale, and no surfaceroughness. In contrast, enamel surfaces highlyaffected by uneven processes of either erosion,scratching, pitting, cracking, plucking or prismexposure would show high anisotropy, fractalsensitivity and roughness. Automatic proce-dures themselves do not account for all thepossible causes of surface heterogeneity. Yet,the researcher’s expertise is required, and

measures of inter-observer error in decidingwhich surfaces might be indicative of dietaryhabits and which are not still relevant. Semiau-tomatic procedures of microwear analysismay be highly subjective. However, automaticprocedures require the same type of subjectivedecision levels during the analysis as thesemiautomatic methods do. Further researchis still needed to demonstrate that automaticimage processing procedures can act as asubstitute for the human eye-to-brain decisioncapabilities. For instance, it is interesting tonote that the patterns observed for enamelroughness (Figure 5) vary depending on thespecies studied: A. afarensis and Pan sp. sharethe same patterns, as do Pongo and Gorilla,whereasA.africanusandA.anamensisconformto somewhat different models. Whether thesedifferences among species have a biologicalexplanation is something that further investiga-tions will elucidate.

Dental Enamel Roughness and Microwear 399

Figure 5. Quadratic regressions between striation density (NT) and average roughness (RA) byspecies. R2 values for each species are shown with sample topography images. The regression patternof Australopithecus afarensis and Pan troglodytes differed in the variation range of NT from those of

Gorilla, Pongo and A. africanus.

Conclusion

The present study assessed the ability of fullyautomatic procedures to analyze microwearpatterns on primate tooth surfaces throughestimation of roughness. Our aim was todetermine if roughness is a reliable indicatorof dietary related habits rather than a measureof enamel preservation. The results indicatethat, in fact, it is true that roughness is ahighly sensitive parameter of enamel preser-vation, and it was not impossible to distin-guishdifferent typesofenamelpreservationandrelief. Low degrees of post-mortem damagewere characterized by an increase in surfaceroughness, but as erosion increases roughnesstends to decrease, as does striation density.The density of diet-related scratches was highlysensitive to abrasion, as was the roughness

variable. However, teeth affected by abrasionare not informative of dietary habits and, inthe same way, roughness appears to be moreinformative of taphonomic processes than ofdiet. Future research requires that only well-preserved enamel surfaces be studied. Theresults obtained indicate that there exists aclear and statistically significant correlationbetween some measures of enamel surfaceroughnessandthemicrowearpattern.However,this relationship varies and is dependant onthe species studied. In addition, the analysisof roughness seems to be scale-sensitive and,thus, the magnification needs to be considered(Estebaranz et al., 2005). Finally, the appli-cation of automatic procedures for microwearresearch will still require a subjective interpre-tation of what factors are relevant to interpreta-tions of diet and what factors are not.

400 Estebaranz et al.

Figure 6. Topographic images of four distinct types of enamel surfaces: intensive enamel damage,well-preserved enamel surface, enamel prisms on a severely damaged surface, and an inter-proximalfacet. Although the intensively damaged surface showed a high roughness value, the other surfaces

had similar Ra values despite clearly different types of surfaces.

Acknowledgments

The Spanish project CGL2004-00775/BTEof the MEC funded this work. All SEMmicrographs were obtained at the ServeisCientíficoTècnics (SCT) of the Univer-sitat de Barcelona, and the interferometricimages were performed at the Plataformade Nanotecnologia of the Parc Científic deBarcelona (PCB).

References

Andritsakis, D.P., Vlamis, K.F., 1986. A new gener-ation of the elastomeric impression materials.Odontostomatol ogike Proodos 40(3), 133–142.

Ball, T., Gardner, J.S., Brotherson, J.D., 1996. Identi-fying phytoliths produced by the inflorescencebracts of three species of wheat (Tricutum

monococcum L., T. dicoccon, Schrank., andT. aestivum L.) using computer-assisted imageand statistical analyses. Journal of Archaeo-logical Science 23, 619–632.

Beynon, A.D., 1987. Replication technique for studyingmicrostructure in fossil enamel. ScanningMicroscopy 1, 663–669.

Butler, P.M., 1952. The milk molars of perissodactylawith remarks on molar occlusion. Proceedingsof the Zoological Society of London 121,777–817.

Daegling, D.J., Grine, F.E., 1999. Terrestrial foragingand dental microwear in Papio ursinus.Primates 40(4), 559–572.

Danielson, D.R., Reinhard K.J., 1998. Human dentalmicrowear caused by calcium oxalate phytolithsin prehistoric diet of the lower Los PecosRegion, Texas. American Journal of PhysicalAnthropology 107, 297–304.

Dennis, J.C., Ungar, P.S., Teaford, M.F., Glander, K.E.,2004. Dental topography and molar wear

Dental Enamel Roughness and Microwear 401

in Alouatta palliata from Costa Rica.American Journal of Physical Anthropology125, 152–161.

Estebaranz, F., Losada, M.J., Galbany, J.,Martínez, L.M., Pérez-Pérez, A., 2005.Tafonomía y microdesgaste: análisistopográfico de superficies de esmalte dentario.Revista española de antropología 25, 75.

Galbany, J., Martínez, L.M., Hiraldo, O., Espurz, V.,Estebaranz, F., Sousa, M., Martínez-López-Amor, H., Medina, A.M., Farrés, M., Bonnin,A., Bernis, C., Turbon, D., Pérez-Pérez, A.,2004a. Teeth: catálogo de los moldes de dientesde homínidos de la Universitat de Barcelona.Universitat de Barcelona, Barcelona.

Galbany J., Martinez, L.M., López-Amor, H.M.,Espurz, V., Romero, A., De Juan, J., Pérez-Pérez, A., 2005a. Error rates in buccal-dental mircrowear quantification using scanningelectron microscopy. Scanning 27, 23–29.

Galbany, J., Martínez, L.M., Pérez-Pérez, A., 2004b.Tooth replication techniques, SEM imaging andmicrowear analysis in primates: methodologicalobstacles. Anthropologie XLII/1, 5–12.

Galbany, J., Pérez-Pérez, A., 2004. Buccal enamelmicrowear variability in Cercopithecoideaprimates as a reflection of dietary habitsin forested and open savanna environments.Anthropologie XLII/1, 13–19.

Galbany, J., Pérez-Pérez, A., Moyà-Solà, S.,2005b. Dental microwear variability on buccaltooth enamel surfaces of extant Catar-rhini and the Miocene fossil Dryopithecuslaietanus (Hominoidea). Folia Primatologica76, 325–341.

Godfrey, L.R., Semprebon, G.M., Jungers, W.L.,Sutherland, M.R., Simons, E.L., Solounias, N.,2004. Dental use wear in extinct lemurs:evidence of diet and niche differentiation.Journal of Human Evolution 47, 145–167.

Gordon, K.D., 1982. A study of microwear onchimpanzee molars: implications of dentalmicrowear analysis. American Journal ofPhysical Anthropology 59, 195–215.

Gordon, K.D., 1984. Hominoid dental microwear:complications in the use of microwear analysisto detect diet. Journal of Dental Research 63,1043–1046.

Grine, F.E., 1986. Dental evidence for dietary differ-ences in Australopithecus and Paranthropus.Journal of Human Evolution 15, 783–822.

Grine, F.E., Ungar, P.S., Teaford, M.F., 2002.Error rates in dental microwear quantifi-

cation using scanning electron microscopy.Scanning 24, 144–153.

Gügel, I.L., Grupe, G., Kunzelmann, K-H., 2001.Simulation of dental microwear: characteristicstraces by opal phytoliths give clue to ancientdietary behavior. American Journal of PhysicalAnthropology 114, 124—138.

Hunter, J.P, Fortelius, M., 1994. Comparative dentalocclusal morphology, facet development, andmicrowear in two sympatric species of Listridon(Mammalia: Suidae) from the Middle Mioceneof Western Anatolia (Turkey). Journal of Verte-brate Paleontology 14, 105–126.

Jernvall, J., Selänne, L., 1999. Laser confocalmicroscopy and geographic informationsystems in the study of dental anthropology.Palaeontologia Electronica 2(1), 1–17.

Kaiser, T.M., Katterwe, H., 2001. The application of3D-microprofilometry as a tool in the surfacediagnosis of fossil and sub-fossil vertebrate hardtissue. An example from the Pliocene UpperLaetoli Beds, Tanzania. International Journal ofOsteoarchaeology 11, 350–356.

Kay, R.F., 1987. Analysis of primate dental microwearusing image processing techniques. ScanningMicroscopy 1(2), 657–662.

King, T., Aiello, L.A., Andrews, P., 1999b. Dentalmicrowear of Griphopithecus alpani. Journal ofHuman Evolution 36, 3–31.

King, T., Andrews, P., Boz, B., 1999a. Effect oftaphonomic processes on dental microwear.American Journal of Physical Anthropology108, 359–373.

Lalueza, C., Pérez-Pérez, A., 1993. The diet of theNeanderthal Child Gibraltar 2 (Devil’s Tower)through the study of the vestibular striationpattern. Journal of Human Evolution 24,29–41.

Lalueza, C., Pérez-Pérez, A., Turbón, D.M., 1996.Dietary inferences through buccal microwearanalysis of Middle and Upper Pleistocenehuman fossils. American Journal of PhysicalAnthropology 100, 367–387.

Lewis, P.J., Gutierrez, M., Johnson, E., 2000.Ondatra zibethicus (Arvicolinae, Rodentia)dental microwear patterns as a potential tool forpalaeoenvironmental reconstruction. Journal ofArchaeological Research 27, 789–798.

Mainland, I.L., 2003. Dental microwear in grazingand browsing Gotland sheep Ovis aries and itsimplications for dietary reconstruction. Journalof Archaeological Science 30, 1513–1527.

402 Estebaranz et al.

Martínez, L.M., Galbany, J., Pérez-Pérez, A., 2004.Paleodemography and dental microwear ofHomo habilis from East Africa. AnthropologieXLII/1, 53–58.

Martínez, L.M., Pérez-Pérez, A. 2004. Post-mortemwear as indicator of taphonomic processesaffecting enamel surfaces of hominin teeth fromLaetoli and Olduvai (Tanzania): implications todietary interpretations. Anthropologie XLII/1,37–42.

Mayhall, J.T., Kageyama, I., 1997. A new three-dimensional method for determining tooth wear.American Journal of Physical Anthropology103, 463–469.

Mills, J.R.E., 1955. Ideal dental occlusion in primates.Dental Practitioner 6, 47–51.

Nystrom, P., Phillips-Conroy, J.E., Jolly, C.J., 2004.Dental microwear in anubis and hybrid baboons(Papio hamdryas, Sensu Lato) living in AwashNational Park, Ethiopia. American Journal ofPhysical Anthropology 125, 279–291.

Reed, D.N.O., 1997. Contour mapping as a new methodfor interpreting diet from tooth morphology.American Journal of Physical AnthropologySuppl. 24, 194.

Pérez-Pérez, A., 2004. Why buccal microwear? Anthro-pologie XLII/1, 1–3.

Pérez-Pérez, A., Bermúdez de Castro, J.M.,Arsuaga, J.L., 1999. Nonocclusal dentalmicrowear analysis of 300,000-year-oldHomo heidelbergensis teeth from Sima delos Huesos (Sierra de Atapuerca, Spain).American Journal of Physical Anthropology108(4), 433–457.

Pérez-Pérez, A., Espurz, V., Bermúdez de Castro, J.M.,de Lumley, M.A., Turbón, D., 2003. Non-occlusal dental microwear variability in asample of Middle and Late Pleistocene humanpopulations from Europe and the Near East.Journal of Human Evolution 44, 497–513.

Pérez-Pérez, A., Galbany, J., Fontarnau R., 2001.Feature extinction in back-scattered SEM. In:Universitat de Barcelona (Eds.), AbstractsMicroscopy. Universitat de Barcelona.Barcelona, pp. 41–42

Pérez-Pérez, A., Lalueza, C., Turbón, D., 1994. Intradi-vidual and intragroup variability of buccal toothstriation pattern. American Journal of PhysicalAnthropology 94, 175–187.

Ryan, A.S., 1979. A preliminary scanning electronmicroscope examination of wear striationdirection on primate teeth. Journal of DentalResearch 58, 525–530.

Silcox, M., Teaford, M.F., 2002. The diet of worms:an analysis of mole dental microwear andits relevance to dietary inference in fossilmammals. Journal of Mammalogy 83, 804–814

Solounias, N., Hayek, L.A.C., 1993. New methodsof tooth microwear analysis and application todietary determination of two extinct antelopes.Journal Zoology London 229, 421–445.

Strait, D.S., 1993. Differences in occlusal morphologyand molar size in frugivores and faunivores.Journal of Human Evolution 25, 471–484.

Teaford, M.F., 1991. Dental microwear: what can it tellus about diet and dental function? In: Kelley,M.A., Larsen, C.S. (Eds.), Advances in DentalAnthropology. Wiley-Liss, Inc., New York, pp.341–356.

Teaford, M.F., 1994. Dental microwear anddental function. Evolutionary Anthropology3(1),17–30

Teaford, M.F., Glander, K.E., 1991. Dental microwearin wild-trapped Alouata pallaia from CostaRica. American Journal of Physical Anthro-pology 85(3), 313–320.

Teaford, M.R., Glander, K.E., 1996. Dental microwearand diet in a wild population of mantled howlers(Alouatta palliata). In: Norconk, M.A., Rosen-berger, A.L., Garber, P.A. (Eds.), AdaptiveRadiations of Neotropical Primates. PlenumPress, New York, pp. 433–449.

Teaford, M.F., Oyen, O.J., 1989. Live primatesand dental replication: new problems andnew techniques. American Journal of PhysicalAnthropology 80, 73–81.

Teaford, M.F., Runestad, J.A., 1992. Dental microwearand diet in Venezuelan primates. AmericanJournal of Physical Anthropology 94, 339–363.

Ungar, P.S., 1992. Feeding behaviour and dentalmicrowear in Sumatran anthropoids. AmericanJournal of Physical Anthropology 88, 347–364.

Ungar, P.S., 1994. Incisor behaviour and dentalmicrowear of Sumatran anthropoid primates.American Journal of Physical Anthropology 94,339–363.

Ungar, P.S., 1995. A semiautomated image analysisprocedure for the quantification of dentalmicrowear II. Scanning 17, 57–59.

Ungar, P.S., 1996. Dental microwear of EuropeanMiocene catarrhines: evidence for diets andtooth use. Journal of Human Evolution 31,335–366.

Ungar, P.S., 1998. Dental allometry, morphology, andwear as evidence for diet in fossil primates.Evolutionary Anthropology 6(6), 205–217.

Dental Enamel Roughness and Microwear 403

Ungar, P.S., 2004. Dental topography and dietsof Australopithecus afarensis and earlyHomo. Journal of Human Evolution 46,605–622.

Ungar, P.S., Brown, C.A., Bergstrom, T.S., Walker, A.,2003. Quantification of dental microwear bytandem scanning confocal microscopy andscale-sensitive fractal analyses. Scanning 25,185–193.

Ungar, P.S., M’Kirera, F., 2003. A solution to theworn tooth conundrum in primate functionalanatomy. Proceedings of the National Academyof Sciences of the USA 10(7), 3874–3877.

Ungar, P.S., Simon, J.C., Cooper, J.W., 1991.A semiautomated image analysis procedurefor the quantification of dental microwear.Scanning 13, 31–36.

Ungar, P.S., Spencer, M.A., 1999. Incisor microwear,diet, and tooth use in three Amerindian popula-tions. American Journal of Physical Anthro-pology 109, 387–396.

Ungar, P.S., Teaford, M.F., 1996. Preliminary exami-nation of non-occlusal dental micro-wear inanthropoids: implications for the study of fossilprimates. American Journal of Physical Anthro-pology 100, 101–113.

Ungar, P.S., Teaford, M.F., Glander, K.E., Pastor, R.F.,1995. Dust accumulation in the canopy: apotential cause of dental microwear in primates.American Journal of Physical Anthropology 97,93–99.

Ungar, P.S., Williamson, M., 2000. Exploringthe effects of tooth wear on functionalmorphology: a preliminary study usingdental topographic analyses. PalaeontologiaElectronica 3(1), 1–18. http://palaeo-electronica.org/2000_1/gorilla/issue1_00.htm

Ungar, P.S., 2001. Microware Software, Version 4.0A semiautomated image analysis system for thequantification of dental microwear. Fayetteville,AR, U.S.A.

Ward, J., Mainland, I.L., 1999. Microwear in modernfree-ranging and stalled pigs. The potential ofdental microwear analysis for exploring pig dietand management in the past. EnvironmentalArchaeology 4, 25–32.

Zuccotti, L.F., Williamson, M.D., Limp, F.E.,Ungar, P.S., 1998. Technical note: modellingprimate occlusal topography using geographicalinformation systems technology. AmericanJournal of Physical Anthropology 107,137–142.