Enamel thickness, microstructure and development in ... · Enamel thickness, microstructure and development in Afropithecus turkanensis Tanya M. Smith1*, Lawrence B. Martin2, Meave

Enamel thickness, microstructure and development inAfropithecus turkanensis

Tanya M. Smith1*, Lawrence B. Martin2, Meave G. Leakey3

1Interdepartmental Doctoral Program in Anthropological Sciences, Department of Anthropology, Stony Brook University,

Stony Brook, NY 11794-4364, USA2Departments of Anthropology and of Anatomical Sciences, Stony Brook University, Stony Brook, NY 11794-4364, USA

3Department of Palaeontology, National Museums of Kenya, P.O. Box 40658, Nairobi, Kenya

Received 12 June 2002; accepted 14 November 2002

Abstract

Afropithecus turkanensis, a 17–17.5 million year old large-bodied hominoid from Kenya, has previously beenreported to be the oldest known thick-enamelled Miocene ape. Most investigations of enamel thickness in Miocene apeshave been limited to opportunistic or destructive studies of small samples. Recently, more comprehensive studies ofenamel thickness and microstructure in Proconsul, Lufengpithecus, and Dryopithecus, as well as extant apes and fossilhumans, have provided information on rates and patterns of dental development, including crown formation time, andhave begun to provide a comparative context for interpretation of the evolution of these characters throughout the past20 million years of hominoid evolution. In this study, enamel thickness and aspects of the enamel microstructure in twoA. turkanensis second molars were quantified and provide insight into rates of enamel apposition, numbers of cellsactively secreting enamel, and the time required to form regions of the crown. The average value for relative enamelthickness in the two molars is 21.4, which is a lower value than a previous analysis of this species, but which is stillrelatively thick compared to extant apes. This value is similar to those of several Miocene hominoids, a fossil hominid,and modern humans. Certain aspects of the enamel microstructure are similar to Proconsul nyanzae, Dryopithecuslaietanus, Lufengpithecus lufengensis, Graecopithecus freybergi and Pongo pygmaeus, while other features di!er fromextant and fossil hominoids. Crown formation times for the two teeth are 2.4–2.6 years and 2.9–3.1 years respectively.These times are similar to a number of extant and fossil hominoids, some of which appear to show additionaldevelopmental similarities, including thick enamel. Although thick enamel may be formed through several develop-mental pathways, most Miocene hominoids and fossil hominids with relatively thick enamel are characterized by arelatively long period of cuspal enamel formation and a rapid rate of enamel secretion throughout the whole cusp, buta shorter total crown formation time than thinner-enamelled extant apes.! 2003 Elsevier Science Ltd. All rights reserved.

Keywords: Afropithecus turkanensis; Miocene hominoid; enamel thickness; enamel microstructure; crown formation time; dailysecretion rate; Retzius line; cross-striation; lamination; intradian line

* Corresponding author. Tel.: +1-(631)-632-1364; fax: +1-(631)-632-9165E-mail addresses: [email protected] (T.M. Smith), [email protected] (L.B. Martin).

Journal of Human Evolution 44 (2003) 283–306

0047-2484/03/$ - see front matter ! 2003 Elsevier Science Ltd. All rights reserved.doi:10.1016/S0047-2484(03)00006-X

Introduction

Miocene hominoids and enamel development

Leakey and Leakey (1986) first describedAfropithecus turkanensis as a large-bodied earlyMiocene hominoid from northern Kenya datedbetween 17 to 17.5 mya, which appeared to haverelatively thick enamel. Leakey et al. (1988)suggested that A. turkanensis shared postcranialfeatures with Proconsul nyanzae, cranial featureswith Aegyptopithecus zeuxis, and dental featureswith Heliopithecus leakeyi, Kenyapithecus sp., andthe large hominoid material from Moroto andNapak. Leakey and Walker (1997) recentlyreviewed this evidence and concluded that A.turkanensis was a primitive, arboreal quadrupedsimilar to P. nyanzae, with a primitive facialmorphology and derived dental characteristicsspecialized for a diet of hard fruits, which wassimilar to fossils from Maboko Island attributedto Kenyapithecus (since assigned to Equatoriusafricanus by Ward et al., 1999). Sherwood et al.(2002) suggested, based on an analysis of thepartial skeleton of E. africanus from the TugenHills, that postcranial similarities between earlyMiocene apes Proconsul and A. turkanensis and themore terrestrial Equatorius are shared primitivefeatures that do not indicate a close taxonomica"liation.

The thickness of molar enamel is often reportedin descriptions of fossil material. Enamel thicknessis commonly assessed as a linear measurement ofenamel visible in worn or naturally fractured teeth,and is generally characterized as ‘thick’ or ‘thin’.Martin (1983) demonstrated that it was di"cult toassess enamel thickness from exposed enamel ac-curately, and measured thickness from a buccal-lingual section cut across the mesial cusp tips,which could be scaled with a surrogate for bodysize to make comparisons across taxa (Martin,1983, 1985; Grine and Martin, 1988; Andrews andMartin, 1991). In the first comparative synthesis ofrelative enamel thickness among several Miocenehominoids, Andrews and Martin (1991) reportedvalues ranging from 8.5 for P. africanus (cat-egorized as having thin enamel) to 28.3 forGraecopithecus freybergi (considered thick-hyper

thick enamel). They suggested that thin enamelwas the primitive condition for Miocene homi-noids, which was supported by data from twospecies of Proconsul, as well as data on Hylobatesand cercopithecoids (Martin, 1983, 1985).

Martin (1995) suggested that A. turkanensis isthe oldest known thick-enamelled hominoid,although he did not provide a value for relativeenamel thickness. Leakey and Walker (1997) alsocite personal communication with Martin that thisspecies had extremely thick molar enamel, whichthey suggested distinguished it from Kenya-pithecus. Although it was not the intent of thisstudy to examine the phylogenetic utility of enamelthickness, information on relative enamel thick-ness in fossil and living hominoids is reviewed andrevised below to provide a comparative frameworkfor the interpretation of the two Afropithecusmolars examined in this study, which is thenrelated to information on enamel microstructure.Together, these variables may provide a morecomplete picture of the dental development of thisearly Miocene ape, which may shed light onadditional aspects of its paleobiology (Kelley andSmith, 2003).

Since the publication of the first few studies ofincremental enamel microstructure in extant apes(e.g., Fukuhara, 1959; Shellis and Poole, 1977;Dean and Wood, 1981), there has been a dramaticincrease in the number of studies on hominoidsduring the past two decades. Martin (1983) under-took the first examination of fossil hominoids,including Sivapithecus and some material fromPasalar, Turkey that is currently attributed toGriphopithecus. Bromage and Dean (1985)reassessed the age at death of fossil hominids usingincremental features, and concluded that earlyhumans were characterized by ape-like dentaldevelopment, in contrast to the previouslyassumed modern human pattern. These were fol-lowed by many more studies of fossil hominidsthat examined the nature of enamel developmentand enamel thickness as revealed by microstruc-ture (Beynon and Wood, 1986; Dean et al., 1986;Beynon and Dean, 1987; Beynon and Wood, 1987;Dean, 1987a,b; Beynon and Dean, 1988; Grine andMartin, 1988; Mann et al., 1991; Beynon, 1992;Dean et al., 1993a; Ramirez Rozzi, 1993a,b; Mann

T.M. Smith et al. / Journal of Human Evolution 44 (2003) 283–306284

et al., 1994; Ramirez Rozzi, 1994; Dean, 1995;Ramirez Rozzi, 1997; Ramirez Rozzi et al., 1997;Ramirez Rozzi, 1998; Dean et al., 2001; Dean andReid, 2001). These investigations have providedinformation on age at death in individuals withdeveloping dentitions, the absolute and relativetiming of dental development, di!erences in thedevelopmental pathways of enamel formation, andlife history characteristics.

Recent work on extant and fossil apes hasbegun to provide an improved comparative frame-work to complement the results of studies of fossilhominoids (Beynon et al., 1991a,b; Beynon andReid, 1995; Reid and Beynon, 1995; Kelley, 1997;Beynon et al., 1998; Dean, 1998; Dean and Shellis,1998; Dirks, 1998; Reid et al., 1998a; Kelley, 1999;Schwartz et al., 1999; Dean, 2000; Kelley andBulicek, 2000; Smith and Martin, 2000; Kelleyet al., 2001; Schwartz and Dean, 2001; Schwartzet al., 2001a,b; Smith et al., 2001; Dirks, 2002;Kelley and Smith, 2003; Schwartz et al 2003; Smithet al., 2003). Beynon et al. (1998) examined theenamel and dentine microstructure of two speciesof Proconsul from Rusinga Island. They foundsome features to be most similar to Pongo, whileother similarities were found with Pan and Homo.Kelley et al. (2001) reported developmental simi-larities among Dryopithecus laietanus, Pan, andP. nyanzae. Schwartz et al. (2003) reported di!er-ences in both microstructure and enamel thicknessbetween Lufengpithecus lufengensis and L. hudien-ensis, with the latter showing some similarities tospecies of Proconsul and Pongo. Additional on-going work on other Miocene taxa, as well as on alarge sample of extant apes, may permit morerefined assessment of the significance of compari-sons of limited fossil material, as well as onchanges in hominoid dental development throughtime (Smith et al., 2003).

The present study quantifies the molar relativeenamel thickness of A. turkanensis, provides infor-mation on several classes of incremental featuresof enamel microstructure, and provides estimatesof crown formation time derived from two lowersecond molars. Polarized light microscopy, scan-ning electron microscopy and tandem scanningreflected (confocal) light microscopy are used togenerate complementary data on aspects of enamel

development in A. turkanensis, as well as insightinto specific methodological issues. These data arecompared with those of other hominoids in anattempt to understand the developmental basis ofvarious patterns of enamel thickness.

Development of enamel microstructure

Enamel development is characterized by theproduction of long- and short-period incrementallines that are formed in the enamel prisms, repre-senting rhythmic changes or disturbances inenamel secretion (Fig. 1). Long-period lines,known as striae of Retzius in cross section, repre-sent the two dimensional position of the entireenamel forming front at a given point in time.Crown growth in height occurs in two ways: as acervical extension of newly di!erentiated enamelforming cells (ameloblasts) along the enameldentine junction (EDJ), and in an appositionalmanner as the ameloblasts move away from theEDJ. Boyde (1964) suggested that the angles whereRetzius lines intersect the EDJ provide evidence ofthe di!erentiation rate of new cells, with a smallerangle indicating a faster rate. The first formedlong-period lines over the dentine horns do notmeet the enamel surface, creating appositional orcuspal enamel. Later-formed Retzius lines extendto the surface of the tooth and form perikymata,which are circumferential imbrications in thelateral enamel surface. This region is referred to asthe lateral or imbricational enamel.

Short-period lines, known as cross-striations,are the result of rhythmic changes in enamelproduction that manifest in prisms perpendicularto the long axis. Numerous studies have demon-strated that cross-striations show a circadianrepeat interval (see FitzGerald, 1998 for a review).As the time of formation between regularly spacedRetzius lines is uniform within all teeth of the sameindividual, many studies have used the number ofcross-striations between adjacent Retzius lines,termed the periodicity, in conjunction with thenumber of Retzius lines, to reconstruct an indi-vidual’s age at death and/or the duration of crownformation. Recent work on several extant homi-noids has demonstrated that this rhythm may

T.M. Smith et al. / Journal of Human Evolution 44 (2003) 283–306 285

range from 6 to 12 days among taxa (Dean andReid, 2001; Schwartz et al., 2001a; Reid et al.,2002). Additional extra-short-period featuresbetween cross-striations, known as ultradian orintradian lines, have been described as fine bandsthat divide cross-striations into two or three seg-ments (Gustafson and Gustafson, 1967; Dean,1995; Dean and Scandrett, 1996; FitzGerald,1996). There are currently no empirical data con-cerning the periodicity or causation of intradianlines in enamel, although there is evidence of an8–12 h secretion rhythm in dentine (Rosenbergand Simmons, 1980; Ohtsuka and Shinoda, 1995).More detailed reviews of enamel microstructuremay be found in Boyde (1989), Risnes (1998) andstandard texts such as Aiello and Dean (1990),Hillson (1996), and Ten Cate (1998).

Materials and methods

Two teeth attributed to A. turkanensis wereexamined in this study: an isolated LM2(KNM-WK 17024) and a single RM2 (KNM-WK24300) that is associated with other mandibularand maxillary teeth from Kalodirr. Both teeth weresectioned in a buccal-lingual plane through themesial cusps. Prior to this study, the talonid basinof KNM-WK 24300 was accidentally fracturedduring transport, also in a buccal-lingual plane,providing a further, naturally fractured surfacefor study. Prior to sectioning, molds of both teethwere created with Coltene President impressionmaterial, which will permit accurate reconstruction.

The teeth were first refluxed and embeddedin methyl methacrylate according to procedures

Fig. 1. Schematic molar cross-section showing incremental features and regions of the enamel microstructure in Afropithecusturkanensis (KNM-WK 24300). Enamel prisms extend from the enamel dentine junction (EDJ) to the tooth surface and preserve long-and short-period features. Striae of Retzius are long-period lines that run obliquely from the EDJ to the surface and represent theregular position of the enamel forming front at one moment in time. Cross-striations are short-period lines that cross enamel prismsperpendicularly and represent the daily secretion of the enamel forming cells. The box in the upper right shows the periodicity, orconsistent relationship of cross-striations between pairs of Retzius lines. The crown may be divided into cuspal and imbricationalenamel regions, which are distinguished from one another by the first stria to reach the surface of the tooth. Separate methods are usedto determine the crown formation time in these two regions (see text). (Modified from Ramirez Rozzi, 1994.)


described by Boyde (1989). Longitudinal cutsacross the mesial cusps were made using aUnipress wire saw, which destroys approximately50 µm of enamel per cut. Thin sections approxi-mately 100 µm thick were cut from each tooth andglued onto microscope slides. These were groundand polished to a 40–60 µm thickness using aBuehler Metaserv grinder-polisher and finishedwith a 0.3 µm alumina suspension on a polishingcloth. Sections were then ultrasonicated, cleanedwith alcohol, cleared in xylene, and cover slipswere mounted with DPX mounting media (FlukaChemicals). Photomontages of both sections weregenerated with 6.3", 12.5" and 25" objectiveson a Zeiss polarizing light microscope (PLM).Images were also digitally captured with a HitachiKP-C553 CCD color camera, and NIH Imagesoftware was used for all measurements.

The remaining blocks, which represented thefaces immediately anterior and posterior to thethin section taken from the mesial cusps of eachtooth, were briefly polished with a series of finergrades of diamond abrasive (6 µm, 1 µm, 0.25 µm),lightly etched with dilute phosphoric acid, driedand sputter coated with silver. The fractured distalhalf of the talonid basin of KNM-WK 24300 wasnot embedded, polished, or etched, but was gluedto a stub and coated with a thin layer of gold. Theblocks were examined using back-scattered elec-tron and secondary electron imaging in an Amray1810 scanning electron microscope (SEM). Over-views of each block face were generated at lowmagnification, and montages of portions of bothcrowns were also generated at 250" and 350".

Thin sections and block faces were also exam-ined and photographed under a tandem scanningreflected light microscope (TSRLM), built byMilan Hadravsky. The TSRLM is a confocalimaging system that provides focal plane specificinformation about subsurface structures to a depthof approximately 100–150 µm (Petran et al., 1985).Digital overviews of all blocks and thin sectionswere also generated with a Nikon Coolpix 995.

Relative enamel thickness

Enamel thickness was measured on each toothfrom enlarged overviews of the two exposed block

faces of the mesial cusps and the thin section, inaccordance with methods developed by Martin(1983) (Fig. 2). Measurements were made on adigitizing tablet with SigmaScan Pro software.Relative enamel thickness was calculated by divid-ing the area of the enamel cap (c) by the length ofthe enamel dentine junction (EDJ) (e), and thisquantity was then divided by the square root of thearea of the dentine (b) and finally multiplied by100. This provides a dimensionless index of enamelthickness that is suitable for comparisons acrosstaxa. The section of each tooth that showed thelowest relative enamel thickness was consideredclosest to the ‘ideal plane of section’ which passesthrough the tip of the dentine horns (minimizingobliquity). These two sections were used to calcu-late the average relative enamel thickness. Asshown in Fig. 2, the enamel outline of the brokencervices of KNM-WK 17024 and the tip of theprotoconid of KNM-WK 24300 were reconstructeddue to fractured enamel and light attrition. Tocompensate for a fracture through the occlusalbasin of KNM-WK 17024, the area and lengthrepresented by the gap between the two halves wassubtracted from the total measurements.

Enamel and dentine microstructure

Several aspects of enamel microstructure werequantified in both teeth: periodicity, daily secretionrate, and angle of Retzius lines at the EDJ. Inaddition, imbricational, cuspal, and total crownformation times were determined (discussed in thefollowing section). The periodicity (number ofcross-striations between Retzius lines) was deter-mined using two methods (Dean et al., 1993b;Swindler and Beynon, 1993). Where possible, di-rect counts were made of cross-striations betweenRetzius lines that clearly met the surface of theenamel (in contrast to other variable parallelaccentuated lines). Additionally, the average spac-ing of the Retzius lines was divided by the averagespacing of cross-striations measured from prismsin that immediate area. These two methods werealso applied where possible to incremental lines inthe dentine of KNM-WK 17024, as the relation-ship between long-period (Andresen) and short-period (von Ebner) lines in the dentine is


equivalent to those in the enamel (Dean et al.,1993b; Dean, 1995).

Daily secretion rate (DSR) was measured fromthe spacing of thousands of cross-striationsthroughout the enamel of both thin sections with a50" objective under the PLM. Cross-striationswere defined as light/dark band units rather thanas prism varicosities/constrictions. Measurementsare reported according to the region within theenamel cap: cuspal inner, middle, and outer thirds;lateral inner, middle, and outer thirds; and cervicalinner and outer halves. Intradian lines wereintentionally avoided when determining the DSR,particularly in certain subsurface regions thatshowed highly variable, closely spaced features.They were recognized as fine bands that appearedbetween square/slightly rectangular light and darkbands (cross-striations). In an attempt to clarifythese poorly studied features, the position and

nature of intradian lines was noted and photo-graphed. In addition, general observations weremade on the appearance of Retzius lines and thepresence of aprismatic enamel. The angle of inter-section between the striae of Retzius and the EDJwas also measured in cuspal, lateral, and cervicalregions of the protoconid and metaconid in eachthin section. Repeated measurements were takenvery close to the point of contact between theRetzius line and the EDJ, and averaged values andranges are reported.

Crown formation time

Histological determination of crown formationtime is most commonly accomplished by combin-ing the time of imbricational enamel formationwith the time of cuspal enamel formation. Deter-mination of imbricational enamel formation time

Fig. 2. (a–d). Two Afropithecus turkanensis teeth showing reconstructions of the enamel cap and measurements used to determine therelative enamel thickness. The upper two sections represent the anterior (a) and posterior (b) faces of the mesial cusps of KNM-WK17024 (LM2). The lower two sections represent the anterior (c) and posterior (d) faces of the mesial cusps of KNM-WK 24300 (RM2).Reconstructions of the original enamel cap were made at the cervices of KNM-WK 17024 and the tip of the protoconid of KNM-WK24300. In (d) the area of the enamel cap is represented as c, the area of the dentine under the enamel cap is represented as b, and thelength of the enamel dentine junction is represented as e. The relative enamel thickness may be calculated as ((c/e)/#b)*100. (SEMimages with 1 mm scale bar shown beneath each section.)


involves counting the number of Retzius lines thatreach the surface as perikymata, and multiplyingthis number by the periodicity. In the thin sectionof KNM-WK 17024, it was only possible to countthe number of Retzius lines in the protoconid, as alarge portion of the cervix of the metaconid wasmissing. Two methods were employed to deter-mine the amount of lost enamel. First, an accen-tuated line was found in the dentine that mostlikely intersected the EDJ at or near the point ofcrown completion, indicating the original posi-tion of the end of the cervical enamel (Reid,pers. comm.). The length of dentine tubules wasmeasured from the EDJ at the position of thebreak in the enamel to the accentuated dentineline, and then divided by the average local dentineDSR to yield a time in days. To verify thisestimate, the number of long-period lines withinthis interval was counted and multiplied by theperiodicity. Both methods yielded similar results.In the enamel of KNM-WK 24300, the tip of thecervix of the metaconid was broken o!, which wasestimated to contain five to six striae that wereadded to the number observed. This estimate wasderived from studies of the identical region ofmore complete hominoid material. Informationfrom the dentine was not available in this toothdue to poor dentine preservation. The results fortotal number of Retzius lines are presented as anarrow range, due to the di"culty of determiningthe first stria to reach the surface (which delineatesthe boundary between imbricational and cuspalenamel), and the estimation of the tip of eachmetaconid.

The issue of determining the cuspal enamelformation time has historically been the mostdi"cult aspect of histological reconstruction, andseveral researchers have employed a number ofmethods. Estimates derived from di!erentmethods may vary by a few weeks to severalmonths. Massler and Schour (1946) proposed thatthe duration of formation could be determined bydividing the length of the enamel prism from theEDJ to the tooth surface by the ‘characteristic rateof apposition’, or DSR. Risnes (1986) demon-strated that, because the prisms follow a sinusoidalcourse through the enamel, the linear enamelthickness must be corrected by a factor to yield the

length of enamel prisms. He proposed a correctionfactor of 1.15 for multidimensional movement,derived from an empirical model of prism path inthe lateral enamel of human premolars. In thisstudy, cuspal enamel formation was determined bytwo similar methods, and compared to valuesobtained by applying Risnes’ correction factor.

The first method involves the generation of aspecies-specific correction factor derived from eachcusp. SEM montages (at 350") were made of theless oblique section of both cusps from each tooth.Tracings of several cohorts of prisms were madefrom the EDJ to the tooth surface, and the averagelength of three ‘path length’ tracings was dividedby the linear cuspal enamel thickness, producing aratio of two-dimensional prism deviation (orsinuosity). This deviation ratio was then doubledto compensate for the actual course of prismsthree-dimensionally, which results in a factor thatis slightly more conservative than that of Risnes’(1986) calculations. Cuspal enamel formation timewas calculated by multiplying the linear cuspalenamel thickness by this factor, which was thendivided by the average cuspal DSR to produce avalue for days of formation.

A second method, based on work by FitzGeraldet al. (1999), was employed for each protoconid,as this cusp is more relevant for the determina-tion of total crown formation time and cross-striations were more visible than in the respectivemetaconid. FitzGerald et al. (1999) demonstratedthat, in teeth that do not show a continuousseries of cross-striations from the EDJ to thetooth surface, approximations of the number ofcross-striations could be made using intervalsalong this distance. This involved dividing up theprism path length into 20 successive intervals(each representing 5% of the total length), deter-mining the average DSR from as many cross-striations as may be measured within eachinterval, and then dividing the interval length byeach average DSR, which yielded a time in daysthat was summed for the total time of cuspalenamel formation. Because it was di"cult toassess the DSR in the final region of cuspalenamel due to light attrition and additionalshort-period features, the DSR from the adjacentzone was used as the divisor for this region.


Due to the fact that all cusps within an indi-vidual tooth do not begin and end enamel forma-tion at the same time, the order of initiation andcompletion must be considered when reporting thetotal crown formation time (Beynon et al., 1998).Histological studies of humans and chimpanzeeshave shown that the protoconid initiates first andtakes longer to form than the metaconid, and thatthe last mandibular cusp to complete calcificationis the hypoconid (Reid et al., 1998a,b). Totalcrown formation times in both A. turkanensis M2swere determined by adding the time of protoconidformation (imbricational plus cuspal enamel) toan estimate of 0.1 years for the final (non-overlapping) period of hypoconid formation. Thisestimate is derived from histological studies ofhumans and chimpanzees (Reid et al., 1998a,b),and is assumed to apply to other large bodiedhominoids. It is critical that di!erences betweencusps are considered when total crown formationtimes are reported, as a number of previous studiesdo not specify which cusps were used to determinethese values and may not be directly comparable.

Results

Relative enamel thickness

The relative enamel thickness values for the twoteeth are reported in Table 1. The average valuefor A. turkanensis is 1.38. Table 2 shows that A.

turkanensis is one of the oldest known hominoidswith relatively thick enamel; however, it is not asthick as was suggested by Martin (1995). Reexami-nation of the original figures and data has shownthat Martin miscalculated relative enamel thick-ness, leading him to conclude that A. turkanensishad very thick enamel. The present results indicatethat enamel thickness in A. turkanensis is similar tothat of several other hominoids such as Gripho-pithecus sp., Sivapithecus sivalensis, P. nyanzae,and Australopithecus africanus, as well as tomodern Homo sapiens (Martin, 1983; Grine andMartin, 1988; Andrews and Martin, 1991; Beynonet al., 1998).

Enamel and dentine microstructure

Counts and measurements made in both enameland dentine show a periodicity of seven inKNM-WK 17024 and eight in KNM-WK 24300(Fig. 3). The range of periodicities in A. turkanen-sis overlaps with those of most hominoids andsome large cercopithecoids (Table 3).

Table 4 shows the average DSR in the inner,middle, and outer zones of the cuspal, lateral, andcervical regions. The average DSR in these zonesranges from 3.46–5.54 µm and 3.27–5.26 µm forKNM-WK 17024 and KNM-WK 24300 respec-tively. The combined average DSR range of bothteeth is 3.38–5.12 µm. Rates generally increasefrom inner to outer zones and from cervical to

Table 1Relative enamel thickness (RET) values for Afropithecus turkanensis

KNM-WK Face b c e c/e ((c/e)/#b)*100=RET

17024-LM2 anterior 24.57 16.21 15.35 1.06 21.37thin section 26.32 16.52 16.14 1.02 19.88posterior 27.71 17.99 15.93 1.13 21.48

24300-RM2 anterior 29.65 20.54 15.66 1.31 24.04thin section 30.83 20.18 15.89 1.27 22.88posterior 30.58 19.73 15.43 1.28 23.15

Mean 21.38

b=area of dentine under the enamel cap, c=area of enamel, e=length of enamel dentine junction. Values of b and c are in mm2,e is in mm. The final column, relative enamel thickness, is dimensionless. The mean reported was derived from averaging the twobold values as these sections showed a minimum relative enamel thickness and were considered closest to the ideal plane ofsection.


cuspal regions save for three exceptions. The DSRin the inner lateral enamel of KNM-WK 17024 isgreater than in the inner cuspal enamel. In themiddle lateral enamel of both teeth, the DSR isgreater than in the middle cuspal enamel. InKNM-WK 24300, the middle cuspal enamel DSRis greater than that in the outer cuspal enamelDSR. The zones and regions that showed thegreatest variation in DSR (outer lateral and outercuspal) also showed the greatest frequency ofintradian lines. A comparison between the two

teeth shows that rates are fairly similar in equiva-lent zones and regions. Regional di!erencesgreater than one standard deviation are observedonly in the inner and outer cuspal enamel.

Fine bands, which we interpret as intradianlines, are sometimes visible between square orslightly rectangular cross-striations (light and darkbands) in both teeth under PLM, SEM andTSRLM (Fig. 4). As noted above, they are mostcommon in the outer cuspal enamel, although theyare also observed in the outer cervical and lateral

Table 2Average relative molar enamel thickness (RET) in extant and fossil hominoids

Taxon RET Range n Category*

Proconsul africanus3 8.5 – 1 thinGorilla gorilla1 10.0 6.8–13.4 17 thinPan troglodytes1 10.1 7.0–13.3 14 thinHylobates lar1 11.0 – 1 thinDryopithecus laietanus3,10 12.7 – 1 intermediate thinOreopithecus bambolii3,11 13.0 – 1 intermediate thinPan paniscus8 13.6 – 1 intermediate thinProconsul major3,11 13.7 – 1 intermediate thinLufengpithecus hudienensis5 14.1 – 1 intermediate thinRangwapithecus gordoni5 14.9 – 1 intermediate thickPongo pygmaeus1 15.9 11.3–20.5 17 intermediate thickProconsul heseloni4 17.0 – 1 intermediate thickSivapithecus sivalensis1,11 19.2 16.3–20.9 3 thickGriphopithecus sp.1,11 19.3 16.5–23.0 8 thickAfropithecus turkanensis6 21.4 19.9–22.9 2 thickAustralopithecus africanus2,11 21.4 21.3–21.6 2 thickHomo sapiens1 22.4 13.8–32.3 13 thickProconsul nyanzae4 22.4 – 1 thickLufengpithecus lufengensis7 24.1 – 1 thickGraecopithecus freybergi3,11 25.9 – 1 thickParanthropus robustus2,9 29.6 – 1 thick-hyper thick

*Categories are defined by 95% confidence limits of species’ means reported by Martin (1985), and individual specimens may befound whose RET is not in the range that defines the species’ mean value.

1Martin (1985).2Grine and Martin (1988).3Andrews and Martin (1991).4Beynon et al. (1998) mesial sections only.5Schwartz et al. (2003).6This study.7Reid and Schwartz (unpublished).8Olejniczak and Martin (in prep).9Grine (pers. comm.) attributes SKX 21841 to Paranthropus robustus rather than to Paranthropus crassidens as reported in

Grine and Martin (1988).10Martin attributes this material to Dryopithecus laietanus rather than to D. fontani as reported in Andrews and Martin (1991).11Material has been reevaluated from additional images during the course of this study and original reported values have been

updated. Due to methodological considerations that may a!ect accuracy, some data from Grine and Martin (1988) and Andrewsand Martin (1991) on P. robustus, Paranthropus boisei, and Heliopithecus leakeyi have been omitted.


enamel in association with Retzius lines. Addi-tional evidence of sub-daily lines is apparentbetween daily lines in the coronal dentine ofKNM-WK 17024. Observations of these regionsunder the SEM and TSRLM confirm that intra-dian lines are neither the result of light microscopyinterference patterns from other layers, nor anartifact of preparation.

‘S-shaped striae’ as described and reported byBeynon et al. (1991b) and Dean and Shellis (1998)for P. nyanzae, Proconsul heseloni, Pongo pyg-maeus, and Hylobates (Symphalangus) syndactylus

are not evident in either of the two A. turkanensisteeth examined. In addition, the SEM shows a fewareas of aprismatic (prism-free) enamel, particu-larly in the outer cuspal enamel of KNM-WK24300. Where present, subsurface aprismaticenamel ranges in maximum thickness from 21–63 µm.

Angular measurements of Retzius lines at theEDJ show that angles are low in cuspal enamel,highest in lateral enamel, and intermediate incervical enamel (Table 5). An examination of theangular variation between the protoconid andmetaconid of each tooth shows substantial di!er-ences in the cuspal and lateral regions ofKNM-WK 24300, but not in KNM-WK 17024. Itis not clear if di!erences between cusps are due tobiological variability in this feature, or if theyrelate to di!erences in the timing of initiation andcompletion. When the two teeth are compared, therange of variation seen within KNM-WK 24300encompasses the values of both cusps ofKNM-WK 17024, with the exception of thecervical region.

Crown formation time

Table 6 shows the crown formation times of themesial cusps of both teeth. The two-dimensionaltracings of the cuspal enamel of KNM-WK 17024yield a ratio of cohort path length to enamelthickness of 1.05 for both cusps. To compensatefor prism undulation, a doubled ratio of 1.111 wasused in calculations of the range of cuspal forma-tion time (3-D column in Table 6), yielding a valueof 354 days for the protoconid and 317 days forthe metaconid based on average DSRs of 4.55 and4.06 µm respectively. In KNM-WK 24300, thedoubled ratio is 1.06 for the protoconid, and 1.11for the metaconid, yielding 416 and 411 daysrespectively (4.38 and 4.29 µm DSRs). The meth-odology of FitzGerald et al. (1999) produced simi-lar results of 387 and 383 days for the protoconidof each tooth. Risnes’ (1986) method produced

1 The value for two dimensional deviation was 1.055, whichwas doubled before rounding to yield a correction factor of1.11.

Fig. 3. Polarized light micrograph of the outer cervical enamelof Afropithecus turkanensis (KNM-WK 24300). The surface ofthe tooth is at the right and the cervix is to the bottom. Enamelprisms are shown running from left to the right, with striae ofRetzius (white arrows) and cross-striations crossing prisms. Therelationship between Retzius lines and cross-striations, orperiodicity, is shown (bracketed) as eight cross-striations andare clearly seen between pairs of Retzius lines. The scale bar inthe lower right represents 50 µm.


results generally within a month of the other twoestimates.

The number of Retzius lines in the imbrica-tional enamel that reach the surface is 71–75 and52–59 in the protoconid and metaconid ofKNM-WK 17024. This range is multiplied by aperiodicity of seven, yielding a range of 497–525and 364–413 days of imbricational enamel forma-tion respectively. In KNM-WK 24300, 81–85 and68–71 total striae are evident at the surface of theprotoconid and metaconid. This is multiplied byan eight-day periodicity, resulting in a range of648–680 and 544–568 days of imbricational enamelformation for each cusp.

Summing the ranges of cuspal and imbrica-tional enamel formation yields crown formationtimes of 2.33–2.50 years (851–912 days) and 1.87–2.00 years (681–730 days) for the protoconid andmetaconid of KNM-WK 17024, and 2.82–3.00years (1031–1096 days) and 2.62–2.67 years (955–

979 days) for the protoconid and metaconid ofKNM-WK 24300. Total crown formation time forA. turkanensis second molars is estimated to rangefrom 2.43–3.10 years, which are the combinedminimum and maximum protoconid formationtimes added to 0.1 years of final hypoconid forma-tion. Table 7 shows that these values for A. turka-nensis are similar to crown formation timesreported for second molars of P. pygmaeus, Gorillagorilla, and modern Homo. Values for A. turka-nensis are greater than estimated crown formationtimes of second molars reported for Hylobates lar,P. heseloni, D. laietanus, and are less than that ofP. troglodytes.

Discussion

Enamel thickness

Histological analysis of dental material facili-tates understanding of the final functional product

Table 3Periodicity (number of cross-striations between Retzius lines) in extant and fossil primates

Taxon Periodicity n Source

Macaca nemestrina 4 1 Smith and Sirianni (unpublished)Macaca mulatta 4 3 Bowman (1991)Hylobates lar 4 1 Dirks (1998)Proconsul heseloni 5 2 Beynon et al. (1998)Theropithecus oswaldi 6 unpublished Macho et al. (1996)Proconsul nyanzae 6 2 Beynon et al. (1998)Dryopithecus laietanus 6–7 3 Kelley et al. (2001)Pan paniscus 6–7 2 Reid (unpublished)Papio hamadryas 7 2 Reid and Dirks (1997)Theropithecus gelada 7 4 Swindler and Beynon (1993)Lufengpithecus hudienensis 7 2 Schwartz et al. (2003)Paranthropus boisei 7 1 Dean (1987a)Afropithecus turkanensis 7–8 2 This studyPan troglodytes 6–9 20 Schwartz et al. (2001a)Graecopithecus freybergi 8 1 Smith et al. (in prep)Gorilla gorilla 7–10 36 Schwartz et al. (2001a)Homo sapiens 6–12 82 Reid et al. (2002)

7–11 28 Schwartz et al. (2001a)7–11 115 Reid and Dean (2000)8–12 96 FitzGerald (1996)

Lufengpithecus lufengensis 8–9 2 Reid and Schwartz (unpublished)Paranthropus robustus 9 1 Dean et al. (1993a)Pongo pygmaeus 8–11 24 Schwartz et al. (2001a)

Values for n represent number of individuals sampled. Values for extant primates from Schour and Ho!man (1939), Okada(1943) and Fukuhara (1959) were not included because methods could not be verified. For extant hominoids, results from studieswith larger samples were included over those based on few individuals.


of the processes of development and growth, whichmay be understood in terms of enamel thick-ness (macrostructure) and enamel microstructure.Several issues regarding enamel thickness warrantdiscussion, including the degree of variation withinand between species, as well as the polarity of thistrait. Knowledge of the variation of this feature isparticularly important when comparing valuesderived from small samples, such as the two A.turkanensis teeth. It has been shown that relativeenamel thickness within extant great apes andhumans varies by 100%, however 95% confidenceintervals for species’ means define discrete taxo-nomic groupings with minimal overlap exceptwhen Pan and Gorilla are compared (Martin, 1983,1985). Values of relative enamel thickness for thetwo teeth examined in this study di!er by about13%, which is within the variation seen in largersamples of Griphopithecus sp. and S. sivalensis.

Previous studies have documented substan-tial intrageneric variation in small samples ofProconsul and Lufengpithecus (Beynon et al.,1998; Schwartz et al., 2003). Beynon et al. (1998)presented data that demonstrated that not allspecies of Proconsul have enamel as thin as Pro-consul major and P. africanus (Andrews andMartin, 1991). Furthermore, recent work onLufengpithecus also shows variation in enamelthickness between closely related species or sub-

species (Schwartz et al., 2003). These studies haveserious implications for the determination of thepolarity of this trait. Martin (1983, 1985) firstsuggested that thin enamel is the ancestral condi-tion for early Miocene apes based on the posses-sion of thin enamel in non-hominoid anthropoidsand Hylobates, which was supported by the initialdiscovery of thin enamel in Proconsul (Andrewsand Martin, 1991). However, evidence of thickenamel in A. turkanensis and P. nyanzae dating to17–18.5 million years ago shows a greater earlydiversity than was previously assumed. It nowappears that patterns of enamel thickness in homi-noids throughout the last 20 million years are morecomplicated than previously suggested. In certaininstances, this trait may have undergone a reversalto the ancestral condition, or evolved in parallelbetween taxa. Recent and ongoing work on thefunctional significance of enamel thickness mayeventually facilitate an integration of traditionalphylogenetic and functional perspectives of thistrait (Andrews and Martin, 1991; Macho andBerner, 1993, 1994; Macho and Spears, 1999;Schwartz, 2000a,b; Olejniczak and Martin, 2002).

Enamel microstructure

The periodicity, numbers of cells secretingenamel, rate of secretion, and duration of secretion

Table 4Average daily secretion rate (DSR) in Afropithecus turkanensis

KNM-WK Region Inner Zone Middle Zone Outer Zone

17024 Cuspal 3.56#0.50 (40) 4.22#0.75 (32) 5.54#1.12 (28)Lateral 3.72#0.48 (38) 4.25#0.67 (35) 4.99#0.91 (23)Cervical 3.46#0.49 (25) – 4.04#0.59 (21)

24300 Cuspal 4.31#0.50 (49) 4.43#0.68 (42) 4.29#0.63 (35)Lateral 4.16#0.48 (17) 4.59#0.53 (18) 5.26#1.35 (24)Cervical 3.27#0.62 (19) – 3.50#0.62 (26)

Overall Cuspal 3.97#0.62 (89) 4.34#0.71 (74) 4.85#1.08 (63)Lateral 3.85#0.52 (55) 4.36#0.64 (53) 5.12#1.14 (47)Cervical 3.38#0.55 (44) – 3.74#0.66 (47)

Data derived from polarized light microscopy. Three equal divisions made along the EDJ from dentine horn to cervix defineregions. Zones represent enamel depth within a particular region. In the cervical enamel, zones were divided into inner and outeronly. Values in µm represent the average of all prisms measured within each region, with number of prisms reported inparentheses. Standard deviation is reported as#one deviation.


have been shown to vary within primates, leadingto di!erences in enamel thickness, enamel distribu-tion and crown formation time. As noted above, ameaningful examination of these traits requiresinformation on a number of growth parametersdetermined from histological sections.

PeriodicityA recent study on a large combined sex sample

of extant great ape and human canines demon-strated significant taxonomic di!erences in period-icity values for all taxa, except between Gorilla andHomo (Schwartz et al., 2001a). These authorssuggested that a relationship might exist betweenperiodicity and body size or a related correlatesuch as brain size or metabolic rate. When extantand fossil hominoids are considered exclusively,periodicity is found to be significantly correlatedwith estimated body mass (Smith et al., 2003).Afropithecus has an estimated body mass ofapproximately 34–35 kg (Leakey and Walker,1997), which is similar to P. troglodytes, andperiodicity ranges from seven to eight in theformer and six to nine in the latter (Table 3).Although this relationship appears to hold truewithin catarrhines and/or hominoids, the inclusionof other primates complicates this hypothesis.Schwartz et al. (2002) recently described themicrostructure of Palaeopropithecus ingens, achimpanzee-sized subfossil Malagasy lemur thathas a periodicity of two. In this instance, period-icity does not appear to correlate with body size,and may be influenced by phylogeny as well as lifehistory (Schwartz, pers. comm.).

Daily secretion rateDSR in A. turkanensis generally increased from

inner to outer enamel within each region, and fromcervical to cuspal regions, which is in keeping withthe results of most studies of other hominoids(Beynon et al., 1991a; Beynon et al., 1998; Dean,1998; Reid et al., 1998a; Kelley et al., 2001;Schwartz et al., 2001a). Exceptions to this patternof consistent increase have been found in someprimates (Beynon et al., 1998; Dean, 1998; Dirks,1998; Schwartz et al., 2001a), as well as in thisstudy. The most marked deviation in the two teethexamined here was seen in the middle and outer

cuspal and lateral enamel, and may have beeninfluenced by additional short-period features(discussed below).

Dean (1998) noted that defining these rates indiscrete inner, middle, and outer zones may maskvariation and lead to a simplified categorization ofthis complex feature. He presented data in agraphical format with a monthly series of box plotsof DSR from the beginning to the end of cuspalenamel formation. Inspection of this type of dataillustrates that rates may level o! or decreaseslightly in the outer enamel, especially in Pongoand Hylobates (e.g., Beynon et al., 1998). This issimilar to the decrease in rate seen in KNM-WK24300. This trend may have been even more evi-dent had DSR been measured and reported inconsecutive monthly zones, but was not possibledue to the variable visibility of cross-striationsthroughout the cuspal enamel.

The mean values of DSR in the cuspal enamelof A. turkanensis are similar to values reported forP. nyanzae, excluding the final month of formation(Beynon et al., 1998). Cuspal DSRs for A. turka-nensis are also similar to those of G. freybergi,L. hudienensis, P. pygmaeus, and D. laietanus,although A. turkanensis does not appear to be-gin enamel secretion as slowly as the three lattertaxa (Beynon et al., 1998; Kelley et al., 2001;Smith et al., 2001; Schwartz et al., 2003). In someMiocene hominoid molars, cuspal enamelsecretion begins at approximately 3.5–4 µm perday, while extant apes and humans begin at 2.5–3µm per day (Beynon et al., 1998; Dean, 1998). Itmay be the case that Miocene hominoids share acommon pattern of higher initial and overall aver-age cuspal rates relative to extant apes, althoughadditional work is necessary to address the issue ofvariation within and between individuals beforedefinitive conclusions are reached.

Ameloblast extensionIt is commonly held that the angle of inter-

section between Retzius lines and the EDJ may beregarded as a proxy for the extension rate ofameloblast activation in a cervical direction(Boyde, 1964; Shellis, 1984). Provided that DSRremains constant, lower angles represent morerapid extension as a larger number of cells are


activated to secrete enamel in a given time. In A.turkanensis, these angles seem to indicate thatextension begins at a high rate, slows from cuspalto lateral enamel, and then increases from lateralto cervical enamel. When compared to values fromother hominoids, the cuspal enamel of A. turka-nensis does not have angles as low as extant apesand humans, nor do they increase as dramaticallyfrom cuspal to cervical enamel as in the otherspecies (Beynon and Reid, 1995; Beynon et al.,1998). This may also be explained by the higherDSR in the inner enamel regions, and the moreuniform pattern of secretion rate from cuspal tocervical inner enamel in A. turkanensis when com-

pared to other extant hominoids (Beynon et al.,1991a, 1998; Dean, 1998; Dirks, 1998; Reid et al.,1998a). Two species of Proconsul display variableangular patterns, which di!er from that of A.turkanensis. Proconsul nyanzae shows a pro-nounced cervical slowing with high angles, whileP. heseloni shows a gradual slowing similar to thatof extant hominoids with moderate angles (Beynonet al., 1998). Schwartz et al. (2003) reported thatL. hudienensis is similar to P. heseloni and extantapes, with a moderate cervical slowing. Afro-pithecus shares the condition of low cervical angleswith some fossil hominids (Dean, 2000). At thispoint, it is not clear how the relative contributions

Fig. 4. (a–c). Intradian lines as seen under multiple forms of microscopy in Afropithecus turkanensis (KNM-WK 24300). In thesefigures, large arrows indicate cross-striations, and potential intradian lines are indicated by thin white lines. (a) Polarized lightmicrograph of the occlusal enamel showing enamel prisms running from the lower right (direction of EDJ) to the upper left (directionof tooth surface). Cross-striations are light and dark bands crossing enamel prisms in a perpendicular manner. Intradian lines are finelight and dark subdivisions of the cross-striations. The scale bar in the lower right represents 50 µm. (b) Split-image scanning electronmicrograph of the outer and middle cervical enamel of a naturally fractured, unetched surface. The image on the left is an overviewof the region, with the box outlined in white indicating an area of enlargement shown on the right. The scale bar in the lower leftrepresents 50 µm. On the right, cross-striations and intradian lines are visible in the enamel prisms, which run from the bottom (EDJ)to the top of the image (tooth surface). The scale bar in the bottom right represents 10 µm. (c) Tandem scanning reflected lightmicrograph of the outer cuspal enamel. Enamel prisms run from the bottom (EDJ) to the top of the image (tooth surface) showingvariable cross-striations and intradian lines. The scale bar in the lower right represents 25 µm. Faint, slightly curved lines run from theupper right corner to the bottom of the image are artifacts of microscopy (scanning lines) due to the holes in the Nipkow disc ofthe TSRLM.


Fig. 4. (b and c).


of changes in extension rate or local DSR impactdi!erences in angles between A. turkanensis andother primates, or if these di!erences are due to acombination of both factors.

Crown formation time and patterns of development

Histological estimates of molar crown forma-tion time in most fossil and some extant hominoidsrange from two to three years. H. lar and P.

heseloni display more rapid dental developmentalschedules, while P. troglodytes appears to have thelongest period of crown formation. Althoughcrown formation time in the two A. turkanensisteeth di!ers by approximately 0.6 years, this vari-ation is not as great as that seen in a recentstudy of four medieval human dentitions (Reidet al., 1998b). Few published data exist on crownformation time in large samples of extant apemolars, but di!erences of 0.5–0.7 years within

Table 5Angle of intersection of the striae of Retzius with the enamel dentine junction (EDJ)

KNM-WK Cusp Cuspal Lateral Cervical

17024 Metaconid 20.3#5.8 (12) 38.6#2.8 (5) –Protoconid 18.8#7.9 (5) 39.1#8.0 (13) 35.0#5.9 (7)

24300 Metaconid 24.5#3.7 (4) 33.7#5.8 (9) 29.7#2.7 (9)Protoconid 15.0#3.6 (3) 42.9#4.0 (10) 31.8#6.2 (9)

Overall 20.0#6.1 (24) 38.7#6.7 (37) 31.9#5.4 (25)

Angles are presented as degrees#one standard deviation. The number of Retzius lines measured is in parentheses. Due to thefractured cervix of the metaconid of KNM-WK 17024, it was not possible to measure angles in this region. Overall valuesrepresent the average of weighted means of both cusps of each tooth.

Table 6Crown formation time in Afropithecus turkanensis

Cusp Cuspal Imbricational

Thick 3-D DSR Path 5% 1.15 Ret Per Imb Total

KNM-WK 17024Protoconid 1450 1.11 4.55 354 387 366 71–75 7 497–525 851–912Metaconid 1160 1.11 4.06 317 – 329 52–59‡ 7 364–413 681–730KNM-WK 24300Protoconid 1720w 1.06 4.38 416 383 452 81–85 8 648–680 1031–1096Metaconid 1590 1.11 4.29 411 – 426 68–71 8 544–568 955–979

Cusp=Protoconid or metaconid. Thick=Maximum cuspal enamel thickness in µm. 3-D=Cusp specific correction factordetermined from path length montages. DSR=Average daily secretion rate in µm determined from measurements in cuspal enamel.Path=Number of days of cuspal enamel formation determined from the enamel thickness multiplied by the 3-D corrected pathlength and divided by the DSR. 5%=Number of days of cuspal enamel formation determined in 5% intervals from the dentinehorn to the tooth surface. The interval length was divided by the local DSR and the days were summed. Determined for theprotoconid in each tooth only. 1.15=Number of days of cuspal enamel formation determined from the thickness multiplied by astandard correction factor (1.15) and divided by the DSR. Provided for comparison only and not used in calculation of total.Ret=Number of imbricational Retzius lines. Per=Periodicity (number of cross-striations between Retzius lines). Imb=Number ofdays of formation of imbricational enamel, determined by the number of imbricational Retzius lines multiplied by the periodicity.Total=Total number of days of enamel formation. Range determined by combining the minimum and maximum of the Path and5% cuspal estimates with the minimum and maximum Imb values.

‡Metaconid cervical enamel missing for KNM-WK 17024, some Retzius lines estimated from dentine. wProtoconid enamelthickness for KNM-WK 24300 was slightly reconstructed, as a small amount was lost to attrition.


small samples of the same tooth type are notuncommon.

Di!erences in crown formation times betweenthe mesial cusps of A. turkanensis support theresults of previous studies, as the protoconid is thefirst mesial mandibular cusp to calcify (Butler,1956; Oka and Kraus, 1969; Reid et al., 1998a,b;Kelley et al., 2001). Subsequent formation pro-ceeds over a longer amount of time compared tothe metaconid, making the protoconid (or theprotocone in the maxillary dentition) the moreappropriate mesial cusp for estimations of totalcrown formation time. In a thorough study ofseveral human dentitions, Reid et al. (1998b) dem-onstrated that the metaconid represented onlyapproximately 80% of the total crown formationtime, while the protoconid represented more than

95% of this period in second molars. This di!er-ence may relate to di!erences in cuspal enamelthickness, as the protoconid is the thicker mesialmandibular cusp (Khera et al., 1990; Schwartz,2000b).

Recent histological studies of chimpanzee andhuman molars have suggested that crown forma-tion time increases in the same cusps from first tothird molars (Reid et al., 1998a,b), which may berelated to a trend of increased enamel thickness,particularly in humans. Although a direct relation-ship between formation time and enamel thicknessmay exist between cusps within a tooth or alongthe molar row within a taxon, a comparison ofextant and fossil hominoids with varying degreesof enamel thickness showed no statistical relation-ship between total crown formation and relative

Table 7Histological crown formation times of hominoid molars (in years)

Taxon/Tooth Cuspal Imbricational n Total

Hylobates lar1 0.67 0.65 1-M1 1.320.68 0.48 1-M2 1.16

Proconsul heseloni2 0.45 0.70–0.90 1-M1 1.0–1.20.50 0.9–1.05 1-M2 1.3–1.60.78 0.84–1.12 1-M3 1.7

Dryopithecus laietanus3 0.73 1.23 1-M1 1.960.64–0.67 1.57–1.58 2-M2 2.22–2.25

Proconsul nyanzae2 0.90 0.92–1.13 1-M1 1.8–2.00.90–0.96 1.04–1.58 1-M2 2.1–2.5

Lufengpithecus hudienensis4 0.73–0.80 1.38 1-M1 2.11–2.18Paranthropus robustus5 1.20 1.23 1-M1 2.43Paranthropus boisei6 1.4–1.55 0.72–1.04 2-(M1/M3) 2.12–2.59Early Homo6 1.01 1.41–1.60 2-M1 2.42–2.62Afropithecus turkanensis7 0.97–1.14 1.36–1.86 2-M2 2.43–3.10Pongo pygmaeus8 0.40 2.33 2-M1 2.7–3.0

unpublished unpublished 2-M2 2.8–3.3Gorilla gorilla8 0.5 2.21–2.39 2-M1 2.7–3.1

0.5–0.6 2.02–2.70 2-M2 2.6–3.3Homo sapiens9 0.79–1.14 1.67–2.72 8-M1 2.61–3.59

1.10–1.25 1.58–2.52 7-M2 2.83–3.621.10–1.30 1.78–2.27 5-M3 3.08–3.37

Lufengpithecus lufengensis11 1.00–1.32 2.89 2-M2 2.83–3.89Pan troglodytes10 0.53–0.65 2.20 5-M1 2.73–2.85

0.60–0.92 1.80–2.96 3-M2 3.56–3.690.70–0.90 2.78–3.14 3-M3 3.48–4.04

1Dirks (1998). 2Beynon et al. (1998). 3Kelley et al. (2001). 4Schwartz et al. (2003). 5Dean et al. (1993a). 6Beynon and Wood(1987). These authors did not specify the tooth from which specific values were measured. 7This study. Total crown formationtime includes a correction of 0.1 years added to the protoconid formation time of each tooth to compensate for final hypoconidformation. 8Beynon et al. (1991b). 9Reid et al. (1998b). 10Reid et al. (1998a). 11Reid and Schwartz (unpublished data). The totalvalue of one of these two teeth was determined using information from the dentine.


enamel thickness (Smith et al., 2003). Secondmolar crown formation times of thinner-enamelledgibbons, gorillas, and chimpanzees range fromthe shortest to the longest reported values.Hominoids with relatively thick enamel, such as A.turkanensis, Paranthropus and P. nyanzae, showintermediate molar formation times.

It appears that cuspal crown formation timemay show a more direct relationship with relativeenamel thickness than total crown formation time.Compared to extant apes, the development of

thick enamel in A. turkanensis results from arelatively greater ratio of cuspal enamel to imbri-cational enamel formation time (Table 7). Extantgreat ape cuspal enamel formation represents onlyapproximately 15–20% of total crown formationtime, compared to approximately 40% in A.turkanensis. The A. turkanensis ratio is similarto the pattern seen in Paranthropus robustus andP. nyanzae, while an even higher ratio is seen inG. freybergi and Paranthropus boisei, which haveextremely thick enamel (Beynon and Wood, 1987;Grine and Martin, 1988; Beynon et al., 1998;Smith et al., 2001).

Grine and Martin (1988) considered three vari-ables that relate to enamel thickness: ameloblastsecretion rate, the total period of secretion, and thenumber of cells that are active at any given time.They hypothesized that similar degrees of enamelthickness could result from multiple develop-mental pathways. Dean (2000) substantiated theirhypothesis with a comparison of cuspal enamelformation in second molars of P. nyanzae and amodern human. He showed that equally thickcuspal enamel resulted from a higher secretion ratein the former and a longer period of secretion inthe latter. Like P. nyanzae, A. turkanensis formedthick cuspal enamel at a high overall secretion rateand completed cuspal enamel formation severalmonths earlier than do modern humans.

Methodological considerations

During the course of this study, it became clearthat a number of factors may potentially influencehistological assessments of enamel microstructure,which have received little attention in previousreports. It is not our intent to provide a systematicmethodological revision here, rather we illustratesome challenges that future work must not onlytake into consideration, but should attempt toimprove upon.

PeriodicitySeveral factors may complicate the accurate

determination of periodicity. It is well known that,in an ideal section, counts should be made onentire series of well-defined cross-striationsbetween adjacent Retzius lines that clearly meet

Fig. 5. Polarized light image of several classes of incrementalfeatures in Afropithecus turkanensis (KNM-WK 24300).Enamel prisms run horizontally to the tooth surface at theright, four striae of Retzius (large arrows) run from the lowerleft to the upper right and meet the enamel surface, four to fivelaminations (small arrows) run parallel between each stria, andnumerous faint bands can be seen crossing the enamel prismsperpendicularly. The periodicity, or number of cross-striationsbetween Retzius lines, was determined in other regions to beeight. This micrograph illustrates the di"culty of determiningthe periodicity in regions where laminations and intradian linesare evident. In addition, this image suggests that laminationsare not equivalent to daily lines.


the surface and form perikymata. In the teethexamined in this study, however, cross-striationsoften appear immediately adjacent to a Retziusline, but become less defined within the intervalbetween two Retzius lines as prisms coursetowards the tooth surface. Risnes (1998) hasspeculated that ameloblasts may lose and reformtheir secretory Tomes’ processes between the pro-duction of successive Retzius lines, which mayresult in poorly defined or absent cross-striations.Both SEM and PLM observations made on sub-surface enamel during this study support thishypothesis, demonstrating well-defined cross-striations frequently near Retzius lines with a‘stair-step’ appearance, but we feel that thiswarrants further study.

In addition, calculated values of periodicity arenot always consistent with direct counts of cross-striations, as the DSR does not always remainconstant between pairs of Retzius lines. This maybe due to local variation in secretion rate, oftencausing the convergence of striae at the toothsurface. Even in regions with clear Retzius linesand cross-striations, additional features such asintradian lines and laminations may complicatedetermination of the periodicity (Fig. 5).

Daily secretion rateIn order to collect data that are methodologi-

cally consistent with other published results onhominoids (e.g., Beynon et al., 1991a, 1998;Dean, 1998; Schwartz et al., 2001a), cross-striations were recorded wherever possible in allregions of the enamel. However, Dirks (1998)and Reid et al. (1998a,b) have suggested thatmeasurements should not be made in the first 100µm of enamel at the EDJ or the last 100 µm atthe surface, because aprismatic enamel in theseregions and the convergence of striae at the toothsurface may obscure or complicate measurementsof daily lines. This disparity may be seen whencomparing data from Reid et al. (1998a) andDean (1998), as reported cuspal DSR in chim-panzees in the former do not show as wide arange as values reported by the latter. Dean’sinclusion of the first and last month of cuspalenamel formation (corresponding approximatelyto the first and last 100 µm) often increases the

minimum and maximum range of DSR values(Dean, 1998; Beynon et al., 1998). In addition,categorizing and averaging data in three discreteenamel depths (inner, middle, and outer) tends tominimize the range of values. It is critical thatthis is considered when comparing results be-tween studies. We would also like to suggest that,in the final month of cuspal enamel formation,the DSR in fossil material may be di"cult toassess due to a high proportion of additionalshort-period features of uncertain periodicity,which may be di"cult at times to distinguishfrom cross-striations (e.g., Fig. 4c).

Additional short-period featuresGustafson and Gustafson (1967) appear to have

been the first to observe and report on intradianlines, which they referred to as a “double bande!ect” (in relation to typical cross-striations). Theydismissed the possibility that these lines were opti-cal artifacts, a conclusion that has recently beenadvocated (Boyde, 1989; Shellis, 1998; Antoineet al., 1999). Others researchers have not specifi-cally addressed them or may have regarded themas equivalent to daily lines (e.g., Whittaker, 1982;Martin, 1983). Whittaker’s and Martin’s micro-graphs likely represented the first published SEMevidence of these features (Whittaker: Figure 7,p.392; Martin: Figure 5.4e, p.332, Figure 5.7a,p.341), although they were not identified assuch. This may have led Martin to concludeerroneously that certain regions of ape and hu-man enamel have extremely slow DSRs, whichhas not been substantiated by other studies. Inthis study, SEM and TSRLM documentation ofintradian lines that are distinct from cross-striations provides additional direct evidence thatthese lines are not the result of interference fromunderlying layers, as both of these forms ofmicroscopy present information from a singleplane of section only.

Another class of incremental feature, knownas laminations, was described by Ripa et al.(1966) and Whittaker (1982) in association withaprismatic enamel. These features, recognized hereas closely spaced incremental lines that run parallelto regular Retzius lines, were observed in thetwo teeth in this study using various forms of


microscopy, particularly in the lateral and cuspalenamel near the EDJ and the tooth surface(Fig. 5). It is not yet known how these featuresdevelop, though it appears they are sometimesassociated with the production of aprismatic orPattern 1 enamel as they are variably visible nearthe EDJ and tooth surface.

We find the presence of additional short-periodfeatures to be one of the most challenging compli-cations of this type of study, particularly as thepositive identification of known period incrementsis a fundamental requirement for analyses of incre-mental development. Experimental work on ratdentine by Rosenberg and Simmons (1980) andOhtsuka and Shinoda (1995) demonstrated thepresence of 2–3 intradian increments per day.FitzGerald (1996) reported a similar observationin human enamel, but his conclusions are di"cultto verify without experimental evidence. In regardto laminations, Risnes (1998) reported that thisfeature shows ‘a relationship’ to cross-striations;however, we present evidence from KNM-WK24300 that this feature does not have a dailyperiodicity (Fig. 5). If laminations are a circadianfeature, one would expect seven lines (eight withone Retzius line included) between and parallel toeach Retzius line, as there is an eight day period-icity in this individual. It is clear that only four tofive laminations are present between the twoRetzius lines, which imply a repeat interval that isgreater than that of circadian cross-striations.Work in progress on fluorescent-labelled macaqueteeth may provide more rigorous empirical data onthe specific periodicity of both intradian lines andlaminations (Smith et al., 2002).

Ameloblast extensionIt appears that published reports of this

angle may not always be comparable due tomethodological di!erences. In certain studies, it isnot clear if reported values represent the angle ofthe overall trend of the Retzius line or the portionimmediately adjacent to the EDJ. Striae may beginat a relatively low angle, which quickly becomesmore dramatic as the prisms or forming frontcourses cervically (see images of Proconsul teeth inBeynon et al., 1998). Additionally, the possibilityof measurement error must be considered when

interpreting or comparing results, as it is di"cultto determine this curvilinear relationship withprecision and accuracy (Shellis, 1984; Grine andMartin, 1988).

Crown formation timeThe most direct method of determining the

cuspal enamel formation time involves countingcross-striations from the beginning of enamel for-mation at the dentine horn to the end of formationat the tooth surface (Gysi, 1931; Boyde, 1963).However, histological material of a quality thatpermits this is quite rare. An alternative to themethods described here and above was proposedby Dean (2000), who suggested that cuspal enamelformation time may be predicted by taxon specificregression equations that require knowledge of theenamel thickness only. The derivation of theseequations, however, requires information aboutDSR in su"ciently large samples, which limits theapplication of this approach to taxa in which teethhave previously been sectioned. In addition, Reidet al. (1998a) suggested that section obliquity mayhave a significant influence on the accuracy ofcertain methods, and advocated methods otherthan those that involve utilizing the linear cuspalenamel thickness. Although Dean’s regressionequations may provide the simplest method ofestimating this growth parameter in non-obliquesections where the DSR is well sampled, additionalwork on fossil and living apes is necessary beforethese regressions can by extended to otherMiocene hominoids such as A. turkanensis.

Summary and Conclusions

Molar enamel thickness in Afropithecus turka-nensis is not as thick as was previously suggested(Martin, 1995). Relative enamel thickness in thetwo second molars examined here is 21.4, which issimilar to values for Griphopithecus sp., Siva-pithecus sivalensis, Australopithecus africanus,modern Homo, and Proconsul nyanzae, and is un-like extant apes. Features of the microstructure,such as the periodicity and daily secretion rate, donot conclusively distinguish A. turkanensis fromProconsul nyanzae, Lufengpithecus, Dryopithecus


laietanus, or extant apes and humans, althoughsome variation exists in the patterning of thesefeatures among hominoids. The periodicity isfound to overlap with some cercopithecoids, extantgreat apes and humans, D. laietanus, Graeco-pithecus freybergi, Lufengpithecus, and Paranthro-pus boisei. The cuspal enamel daily secretion ratesare most similar to those of G. freybergi, L. hudie-nensis, Pongo pygmaeus, P. nyanzae and D.laietanus. Angles of Retzius lines at the EDJ aremost similar to those of fossil hominids and di!erfrom living and fossil apes. Evidence was alsofound of additional incremental features known asintradian lines and laminations, which may a!ectdetermination of the periodicity and daily secretionrate if not properly identified. The developmentalbasis and periodicity of these two features are notwell understood, and warrant further study.

Crown formation times of the two A. turka-nensis M2s are 2.4–2.6 and 2.9–3.1 years. Thesevalues are most similar to those of second molarsof P. pygmaeus, Gorilla gorilla and modern Homo,and overlap slightly with P. nyanzae and L.lufengensis. Afropithecus shows a relatively longerperiod of cuspal enamel formation than thinner-enamelled African apes, although the latter have alonger period of total crown formation. Teeth withrelatively thick enamel may result from a longerperiod of cuspal enamel formation, higher cuspaldaily secretion rate, and/or a greater numberof active cells than in thin-enamelled extanthominoids. In Miocene hominoids, thick enamelappears to result from higher daily cuspal enamelsecretion rates and/or longer cuspal formationtimes relative to extant apes and humans.Additional studies are needed to document thevariation in these parameters in larger samples ofextant species, which will lead to more informedinterpretation of limited fossil material.

Afropithecus turkanensis may be categorized asa primitive hominoid with relatively thick enamel,which formed in a manner similar to otherMiocene hominoids. The possession of thickenamel may have permitted the exploitation ofadditional food resources unavailable to earliercatarrhines, which are presumed to have relied onsoft-fruits that required little processing (Andrewsand Martin, 1991). Aspects of the developmental

biology of A. turkanensis, such as crown formationtime and age at M1 emergence (Kelley and Smith,2003), indicate that by 17 million years ago,hominoids had begun to show a more prolongedperiod of dental development relative to species ofProconsul, Hylobates or extant cercopithecines.Evidence for even greater periods of dental devel-opment characteristic of great apes and humans isnot apparent until the late Miocene.

Acknowledgements

We are grateful to the Government of Kenyaand the National Museums of Kenya for permis-sion to study these valuable teeth. We wouldespecially like to thank Don Reid for his invalu-able advice and assistance with this work. ChrisDean, Gary Schwartz, Wendy Dirks, and JayKelley provided generous technical advice and/orhelpful discussions. We would also like to thankJay Kelley, Anthony Olejniczak, Don Reid, ChrisDean, Fred Grine, Gabriele Macho, SummerArrigo-Nelson, Allison Cleveland and two anony-mous reviewers for helpful comments on thismanuscript. Milan Hadravsky provided essentialtechnical support with the TSRLM. RandySusman provided access to the Pan paniscus tooth.This study was supported by the National ScienceFoundation, SBR 8918695.

References

Aiello, L., Dean, C., 1990. An Introduction to HumanEvolutionary Anatomy. Academic Press, London.

Andrews, P., Martin, L., 1991. Hominoid dietary evolution.Phil. Trans. R. Soc. 334, 199–209.

Antoine, D., Dean, C., Hillson, S., 1999. The periodicityof incremental structures in dental enamel based on thedeveloping dentition of post-medieval known-age children,in: Mayhall, J.T., Heikkinen, T. (Eds.), Dental Morphology1998. Oulu University Press, Oulu, pp. 48–55.

Beynon, A.D., 1992. Circaseptan rhythms in enamel develop-ment in modern humans and Plio-Pleistocene hominids, in:Smith, P., Tchernov, E. (Eds.), Structure, Function andEvolution of Teeth. Freund, London, pp. 295–309.

Beynon, A.D., Dean, M.C., 1987. Crown formation time of afossil hominid premolar tooth. Arch. Oral Biol. 32,773–780.


Beynon, A.D., Dean, M.C., 1988. Distinct dental developmentpatterns in early fossil hominids. Nature 335, 509–514.

Beynon, A.D., Dean, M.C., Reid, D.J., 1991a. On thick andthin enamel in hominoids. Am. J. phys. Anthrop. 86,295–309.

Beynon, A.D., Dean, M.C., Reid, D.J., 1991b. Histologicalstudy on the chronology of the developing dentitionin Gorilla and Orangutan. Am. J. phys. Anthrop. 86,189–203.

Beynon, A.D., Dean, M.C., Leakey, M.G., Reid, D.J., Walker,A., 1998. Comparative dental development and microstruc-ture of Proconsul teeth from Rusinga Island, Kenya. J. hum.Evol. 35, 163–209.

Beynon, A.D., Reid, D.J., 1995. Comparative studies onenamel structure and development in modern hominoids, in:Radlanski, R.J., Renz, H. (Eds.), Proceedings of the 10thInternational Symposium on Dental Morphology. “M”Marketing Services, Berlin.

Beynon, A.D., Wood, B.A., 1986. Variations in enamel thick-ness and structure in East African hominids. Am. J. phys.Anthrop. 70, 177–193.

Beynon, A.D., Wood, B.A., 1987. Patterns and rates of enamelgrowth in the molar teeth of early hominids. Nature 326,493–496.

Bowman, J.E., 1991. Life history, growth and dental develop-ment in young primates: a study using captive rhesusmacaques. Ph.D. Dissertation, University of Cambridge.

Boyde, A., 1963. Estimation of age at death of young humanskeletal remains from incremental lines in the dental enamel.Excerpta medica Int. Cong. Ser. 80, 36. Proc. 3rd Int. Meet.Forensic Immunol. Med. Pathol. and Toxicol. Lond.

Boyde, A., 1964. The structure and development of mammalianenamel. Ph.D. Dissertation, University of London.

Boyde, A., 1989. Enamel, in: Oksche, A., Vollrath, L. (Eds.),Handbook of Microscopic Anatomy, Volume V/6: Teeth.Springer-Verlag, Berlin, pp. 309–473.

Bromage, T.G., Dean, M.C., 1985. Re-evaluation of the age atdeath of immature fossil hominids. Nature 317, 525–527.

Butler, P.M., 1956. The ontogeny of molar pattern. Biol. Rev.31, 30–70.

Dean, M.C., 1987a. The dental development status of six EastAfrican juvenile fossil hominids. J. hum. Evol. 16, 197–213.

Dean, M.C., 1987b. Growth layers and incremental markingsin hard tissues; a review of the literature and some prelimi-nary observations about enamel structure in Paranthropusboisei. J. hum. Evol. 16, 157–172.

Dean, M.C., 1995. The nature and periodicity of incrementallines in primate dentine and their relationship to periradicu-lar bands in OH 16 (Homo habilis), in: Moggi-Cecchi, J.(Ed.), Aspects of Dental Biology: Palaeontology, Anthro-pology and Evolution. International Institute for the Studyof Man, Florence, pp. 239–265.

Dean, M.C., 1998. A comparative study of cross striationspacings in cuspal enamel and of four methods of estimatingthe time taken to grow molar cuspal enamel in Pan, Pongoand Homo. J. hum. Evol. 35, 449–462.

Dean, C., 2000. Progress in understanding hominoid dentaldevelopment. J. Anat. 197, 77–101.

Dean, M.C., Beynon, A.D., Reid, D.J., Whittaker, D.K.,1993b. A longitudinal study of tooth growth in a singleindividual based on long- and short-period incrementalmarkings in dentine and enamel. Int. J. Osteoarchaeol. 3,249–264.

Dean, M.C., Beynon, A.D., Thackeray, J.F., Macho, G.A.,1993a. Histological reconstruction of dental developmentand age at death of a juvenile Paranthropus robustus speci-men, SK 63, from Swartkrans, South Africa. Am. J. phys.Anthrop. 91, 401–419.

Dean, M.C., Leakey, M.G., Reid, D., Schrenk, F., Schwartz,G.T., Stringer, C., Walker, A., 2001. Growth processes inteeth distinguish modern humans from Homo erectus andearlier hominins. Nature 414, 628–631.

Dean, M.C., Reid, D.J., 2001. Perikymata spacing and distri-bution on hominid anterior teeth. Am. J. phys. Anthrop.116, 209–215.

Dean, M.C., Scandrett, A.E., 1996. The relation betweenlong-period incremental markings in dentine and dailycross-striations in enamel in human teeth. Arch. Oral Biol.41, 233–241.

Dean, M.C., Shellis, R.P., 1998. Observations on stria mor-phology in the lateral enamel of Pongo, Hylobates, andProconsul teeth. J. hum. Evol. 35, 401–410.

Dean, M.C., Stringer, C.B., Bromage, T.G., 1986. Age at deathof the Neanderthal child from Devil’s Tower, Gibraltar andthe implications for studies of general growth and develop-ment in Neanderthals. Am. J. phys. Anthrop. 70, 301–309.

Dean, M.C., Wood, B.A., 1981. Developing pongid dentitionand its use for ageing individual crania in comparativecross-sectional growth studies. Folia primatol. 36, 111–127.

Dirks, W., 1998. Histological reconstruction of dental develop-ment and age at death in a juvenile gibbon (Hylobates lar).J. hum. Evol. 35, 411–425.

Dirks, W., 2002. Dental development in hylobatids, or how toget to the same place in the same time on a di!erent road.Am. J. phys. Anthrop. 34 Suppl., 63.

FitzGerald, C.M. 1996. Tooth crown formation and the varia-tion of enamel microstructural growth markers in modernhumans. Ph.D. Dissertation, University of Cambridge.

FitzGerald, C.M., 1998. Do enamel microstructures have regu-lar time dependency? Conclusions from the literature and alarge-scale study. J. hum. Evol. 35, 371–386.

FitzGerald, C.M., Saunders, S.R., Macchiarelli, R., Bondioli,L., 1999. Large scale histological assessment of deciduouscrown formation, in: Mayhall, J.T., Heikkinen, T. (Eds.),Dental Morphology 1998. Oulu University Press, Oulu, pp.92–101.

Fukuhara, T., 1959. Comparative-anatomical studies of thegrowth lines in the enamel of mammalian teeth. Acta Anat.Nipp. 34, 322–332.

Grine, F.E., Martin, L.B., 1988. Enamel thickness and develop-ment in Australopithecus and Paranthropus, in: Grine, F.E.(Ed.), Evolutionary History of the “Robust”Australopithecines. Aldine de Gruyter, New York, pp. 3–42.


Gustafson, G., Gustafson, A-G., 1967. Microanatomy andhistochemistry of enamel, in: Miles, A.E.W. (Ed.), Struc-tural and Chemical Organization of Teeth. Academic Press,London, pp. 75–134.

Gysi, A., 1931. Metabolism in adult enamel. Dent. Digest 37,661–668.

Hillson, S., 1996. Dental Anthropology. Cambridge UniversityPress, Cambridge.

Kelley, J., 1997. Paleobiological and phylogenetic significanceof life history in Miocene hominoids, in: Begun, D.R.,Ward, C.V., Rose, M.D. (Eds.), Function, Phylogeny, andFossils: Miocene Hominoid Evolution and Adaptations.Plenum Press, New York, pp. 173–208.

Kelley, J., 1999. Age at first molar emergence in Afropithecusturkanensis. Am. J. phys. Anthrop. 28 Suppl., 167.

Kelley, J., Bulicek, C., 2000. Identification of a birth cohort inthe Miocene hominoid fossil record. Am. J. phys. Anthrop.30 Suppl., 195.

Kelley, J., Dean, M.C., Reid, D.J., 2001. Molar growth in thelate Miocene hominoid. Dryopithecus laietanus, in: Brook,A. (Ed.), Dental Morphology 2001. She"eld AcademicPress, She"eld, pp. 123–134.

Kelley, J., Smith, T.M., 2003. Age at first molar emergencein early Miocene Afropithecus turkanensis and life-historyevolution in the Hominoidea, J. hum. Evol. 44.

Khera, S.C., Carpenter, C.W., Vetter, J.D., Staley, R.N., 1990.Anatomy of cusps of posterior teeth and their fracturepotential. J. Prosthet. Dent. 64, 139–147.

Leakey, R.E., Leakey, M.G., 1986. A new Miocene hominoidfrom Kenya. Nature 324, 143–148.

Leakey, R.E., Leakey, M.G., Walker, A.C., 1988. Morphologyof Afropithecus turkanensis from Kenya. Am. J. phys.Anthrop. 76, 289–307.

Leakey, M., Walker, A., 1997. Afropithecus function andphylogeny, in: Begun, D.R., Ward, C.V., Rose, M.D.(Eds.), Function, Phylogeny, and Fossils: Miocene Homi-noid Evolution and Adaptations. Plenum Press, New York,pp. 225–239.

Macho, G.A., Berner, M.E., 1993. Enamel thickness of humanmaxillary molars reconsidered. Am. J. phys. Anthrop. 92,189–200.

Macho, G.A., Berner, M.E., 1994. Enamel thickness and thehelicoidal occlusal plane. Am. J. phys. Anthrop. 94,327–337.

Macho, G.A., Spears, I.R., 1999. E!ects of loading on thebiochemical behavior of molars of Homo, Pan, and Pongo.Am. J. phys. Anthrop. 109, 211–227.

Macho, G.A., Reid, D.J., Leakey, M.G., Jablonski, N.,Beynon, A.D., 1996. Climatic e!ects on dental developmentof Theropithecus oswaldi from Koobi Fora and Olorgesailie.J. hum. Evol. 30, 57–70.

Mann, A.E., Lampl, M., Monge, J.M., 1994. Comparativestudies of enamel structure in living and extinct populationsof Homo sapiens. Am. J. phys. Anthrop. 18 Suppl., 135.

Mann, A.E., Monge, J.M., Lampl, M., 1991. Investigation intothe relationship between perikymata counts and crownformation times. Am. J. phys. Anthrop. 86, 175–188.

Martin, L.B., 1983. The relationships of the later MioceneHominoidea. Ph.D. Dissertation, University CollegeLondon.

Martin, L., 1985. Significance of enamel thickness in hominoidevolution. Nature 314, 260–263.

Martin, L.B., 1995. Oldest known thick-enamelled ape. Am. J.phys. Anthrop. 20 Suppl., 144.

Massler, M., Schour, I., 1946. The appositional life span of theenamel and dentin-forming cells. J. Dent. Res. 25, 145–150.

Ohtsuka, M., Shinoda, H., 1995. Ontogeny of circadian den-tinogenesis in the rat incisor. Arch. Oral Biol. 40, 481–485.

Oka, S.W., Kraus, B.S., 1969. The circumnatal status of molarcrown maturation among the Hominoidea. Arch. Oral Biol.14, 639–659.

Okada, M., 1943. Hard tissues of animal body; highly interest-ing details of Nippon studies in periodic patterns of hardtissues are described. The Shanghai Evening Post, Spec. Ed.;Health, Recreation and Medical Progress. September 1943,pp. 15–31.

Olejniczak, A., Martin, L., 2002. A comparison of relativeenamel thickness of deciduous and permanent teeth in Pantroglodytes. Am. J. phys. Anthrop. 34 Suppl., 120.

Petran, M., Hadravsky, M., Benes, J., Kucera, R., Boyde, A.,1985. The tandem scanning reflected light microscope. Part1—the principle, and its design. Proc. Roy. Micro. Soc. 20,125–129.

Ramirez Rozzi, F.V., 1993a. Tooth development in EastAfrican Paranthropus. J. hum. Evol. 24, 429–454.

Ramirez Rozzi, F.V., 1993b. Microstructure and developmentof the enamel tooth of the neandertal from Zafarraya,Spain. C. r. Acad. Sci., Paris 316, 1635–1642.

Ramirez Rozzi, F.V., 1994. Enamel growth markers in hominiddentition. Mic. & Anal. July, 21–23.

Ramirez Rozzi, F.V., 1997. Dental development in Plio-Pleistocene hominids. Enamel extension rate in fossilhominids. C. r. Acad. Sci., Paris 325, 293–296.

Ramirez Rozzi, F.V., 1998. Can enamel microstructure be usedto establish the presence of di!erent species of Plio-Pleistocene hominids from Omo, Ethiopia? J. hum. Evol.35, 543–576.

Ramirez Rozzi, F.V., Bromage, T., Schrenk, F., 1997. UR 501,the Plio-Pleistocene hominid from Malawi. Analysis of themicroanatomy of the enamel. C. r. Acad. Sci., Paris 325,231–234.

Reid, D.J., Beynon, A.D., 1995. Stria slope and spacing inhuman and great ape molar teeth. J. Dent. Res. 74, 829.

Reid, D.J., Beynon, A.D., Ramirez Rozzi, F.V., 1998b. Histo-logical reconstruction of dental development in four indi-viduals from a medieval site in Picardie, France. J. hum.Evol. 35, 463–477.

Reid, D.J., Dean, M.C., 2000. Brief communication: Thetiming of linear hypoplasias on human anterior teeth. Am.J. phys. Anthrop. 113, 135–139.

Reid, D.J., Dirks, W., 1997. Histological reconstruction ofdental development and its relationship to periodic stress inthe Awash baboons. Am. J. phys. Anthrop. 24 Suppl.,195–196.


Reid, D., Ferrell, R., Walton, P., 2002. Histologically derivedcanine crown formation times from a medieval Danishsample. Am. J. phys. Anthrop. 34 Suppl., 129.

Reid, D.J., Schwartz, G.T., Dean, C., Chandrasekera, M.S.,1998a. A histological reconstruction of dental developmentin the common chimpanzee, Pan troglodytes. J. hum. Evol.35, 427–448.

Ripa, L.W., Gwinnett, A.J., Buonocore, M.G., 1966. The“prismless” outer layer of deciduous and permanentenamel. Arch. Oral Biol. 11, 41–48.

Risnes, S., 1986. Enamel apposition rate and the prism period-icity in human teeth. Scand. J. Dent. Res. 94, 394–404.

Risnes, S., 1998. Growth tracks in dental enamel. J. hum. Evol.35, 331–350.

Rosenberg, G.D., Simmons, D.J., 1980. Rhythmic dentino-genesis in the rabbit incisor: circadian, ultradian, andinfradian periods. Calc. Tiss. Inter. 32, 29–44.

Schour, I., Ho!man, M.M., 1939. The 16 microns calcificationrhythm in the enamel and dentine from fish to man. J. Dent.Res. 18, 91–102.

Schwartz, G.T., 2000a. Taxonomic and functional aspects ofthe patterning of enamel thickness distribution in extantlarge-bodied hominoids. Am. J. phys. Anthrop. 111,221–244.

Schwartz, G.T., 2000b. Enamel thickness and the helicoidalwear plane in modern human mandibular molars. Arch.Oral Biol. 45, 401–409.

Schwartz, G.T., Dean, C., 2001. Ontogeny of canine dimor-phism in extant hominoids. Am. J. phys. Anthrop. 115,269–283.

Schwartz, G.T., Dean, M.C., Reid, D.J., Zihlman, A.L., 2001b.Histological analysis of dental development in Gorilla. Am.J. phys. Anthrop. 32 Suppl., 134.

Schwartz, G.T., Liu, W., Zheng, L., 2003. Preliminary investi-gation of dental microstructure in the Yuanmou hominoid,Yunnan Province, China, J. hum. Evol. 44.

Schwartz, G.T., Reid, D.J., Dean, C., 1999. Dental develop-ment and canine dimorphism in the common chimpanzee,Pan troglodytes. Am. J. phys. Anthrop. 28 Suppl., 246.

Schwartz, G.T., Reid, D.J., Dean, C., 2001a. Developmentalaspects of sexual dimorphism in hominoid canines. Int. J.Primatol. 22, 837–860.

Schwartz, G.T., Samonds, K.E., Godfrey, L.R., Jungers, W.L.,Simons, E.L., 2002. Dental microstructure and life historyin subfossil Malagasy lemurs. Proc. natl. Acad. Sci. 99,6124–6129.

Shellis, R.P., 1984. Variations in growth of the enamel crown inhuman teeth and a possible relationship between growthand enamel structure. Arch. Oral Biol. 29, 697–705.

Shellis, R.P., 1998. Utilization of periodic markings in enamelto obtain information on tooth growth. J. hum. Evol. 35,387–400.

Shellis, R.P., Poole, D.F.G., 1977. The calcified dental tissuesof primates, in: Lavelle, C.L.B., Shellis, R.P., Poole, D.F.G.(Eds.), Evolutionary Changes to the Primate Skull andDentition. Charles C. Thomas, Springfield, pp. 197–279.

Sherwood, R.J., Ward, S., Hill, A., Duren, D.L., Brown, B.,Downs, W., 2002. Preliminary description of the Equatoriusafricanus partial skeleton (KNM-TH 28860) fromKipsaramon, Tugen Hills, Baringo District, Kenya. J. hum.Evol. 42, 63–73.

Smith, T.M., Martin, L.B., 2000. A re-evalution of enamelthickness and development in Afropithecus turkanensis. Am.J. phys. Anthrop. 30 Suppl., 285–286.

Smith, T.M., Martin, L.B., de Bonis, L., Koufos, G.D., 2001.Dental development in Graecopithecus freybergi (Ourano-pithecus macedoniensis). Am. J. phys. Anthrop. 32 Suppl.,140.

Smith, T.M., Martin, L.B., Sirianni, J.E., 2002. Incrementalfeatures of pre- and postnatal enamel microstructure inpigtailed macaques (Macaca nemestrina). Am. J. phys.Anthrop. 34 Suppl., 144.

Smith, T.M., Dean, M.C., Kelley, J., Martin, L.B., Reid, D.J.,Schwartz, G.T., 2003. Molar crown formation time inMiocene hominoids: a preliminary synthesis, Am. J. phys.Anthrop. 36 Suppl., 196.

Swindler, D.A., Beynon, A.D., 1993. The development andmicrostructure of the dentition of Theropithecus, in:Jablonski, N.G. (Ed.), Theropithecus: The Rise and Fall ofa Primate Genus. Cambridge University Press, Cambridge,pp. 351–381.

Ten Cate, A.R., 1998. Oral Histology—Development, Struc-ture, and Function. Mosby, St. Louis.

Ward, S., Brown, B., Hill, A., Kellley, J., Downs, W., 1999.Equatorius: a new hominoid genus from the MiddleMiocene of Kenya. Nature 285, 1382–1386.

Whittaker, D.K., 1982. Structural variations in the surface zoneof human tooth enamel observed by scanning electronmicroscopy. Arch. Oral Biol. 27, 383–392.


Documents

Enamel thickness, microstructure and development in ... · Enamel thickness, microstructure and development in Afropithecus turkanensis Tanya M. Smith1*, Lawrence B. Martin2, Meave