Investigation into the relationship between perikymata counts and crown formation times.код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 86:175-188 (1991) Investigation Into the Relationship Between Perikymata Counts and Crown Formation Times ALAN E. MA", JANET M. MONGE, AND MICHELLE LAMPL Department of Anthropology, University of Pennsylvania, Philadelphia, Pennsylvania 19104 KEY WORDS Paleoanthropology, Dental development, Fossil hominines, Hominid ABSTRACT This study reports on a sample of 12 modern human incisors (from two archaeological sites) that were viewed with a scanning electron microscope and whose perikymata were counted. These 12 incisors more than doubles the previously published sample size of modern human incisors that have served as the published standard for perikymata number in human incisors and have been employed to define taxonomic relationships in fossil hominids. All previously published fossil specimens fall within the expanded range of modern human perikymata counts and can no longer be considered distinctively nonhuman in dental formation time. Five neandertal incisors from the Krapina site in Croatia, Yugoslavia, were also examined. These incisors substantially expand the previous data base for counts of perikymata in Homo sapiens neanderthalensis, likewise overlapping the previously published modern human range. Finally, the validity of methods that have been employed for deriving crown formation times from perikyrnata counts in fossil hominines is called into question. Utilizing the presently known perikymata ranges for modern humans, these methods do not predict the range of known crown formation times of modern humans as assessed from studies of living children. For almost half a century there have been discussions of the importance of reconstructing the pattern of maturation in earlier hominines' (e.g., Dart, 1948; Dobzhansky, 1962;Mann, 1972,1975; a u l d , 1976; Lovejoy, 1981; Holloway, 1983). In these discussions, there was a recognition that the prolongation of childhood dependency represented a crucial shiR in human evolution; one that provided the time necessary for complex behaviors and detailed information about the environment to be learned and internalized by a new generation. In 1975, Mann presented data based on gross details of dental eruption and development in South African australopithecines, in particular, Australopithecus robustus from Swartkrans, that suggested these early hominines conformed t o the modern human 'In this paper, we follow Weiss (1987) and Weiss and Mann (1990)in employing the subfamily Homininae rather than the family Hominidae in describinghumans and our close immediate ancestors. @ 1991 WILEY-LISS, INC. pattern of a prolonged childhood (Mann, 1972). At the time that this evidence was presented, there was an accepted view that the australopithecines' had crossed an evolutionary threshold and had become more human than apelike. This view was based on studies of endocranial details, as well as on archaeological traces from East African sites (e.g., Holloway, 1972; Isaac, 1972, 1978). In that climate of opinion, prolongation of australopithecine dependency periods was not unexpected and tended to confirm the prevailing perspective. More recently, there has been a critical reconsideration of these early hominines, one including a reassessment of their behavior and cultural repertoire and their overall similarity to modern humans. Reevaluations 'The term australopithecine is used here in an entirely descriptive manner to referto all members of the genus Australopithecus (which, for us, continues to included. mbustus and A. boisei),as well as specimens referred to as Homo habilis or described generally as "earlyHomo." Received March 5, 1990; accepted January 15,1991. 176 A.E. MA” ET AL (Fig. 1). These authors argued that each perikymata represented approximately 7 days of enamel deposition, a conclusion based on previously published observations of dental microstructure that reported seven to eight regularly spaced cross striations between striae of Retzius and thus, by association of these structures to their surface manifestations, perikymata. The cross striations, it was maintained, represented the daily activity of a single enamel-forming cell (Boyde, 1976). STUDIES OF DENTAL MICROSTRUCTURE AND On this foundation, Bromage and Dean THEIR IMPLICATIONS FOR PAmERNS OF (1985) counted perikymata on a sample of HOMINID GROWTH AND DEVELOPMENT nine australopithecine juvenile incisors, a In 1985, Bromage and Dean published combined sample of maxillary and mandibuobservations on the microstructure of dental lar, central and lateral incisors. Bromage enamel that seriously challenged the notion and Dean found a range of between 57 and of a modern humanlike prolongation of 135 perikymata on these fossil incisors and growth in australopithecines. They pre- contrasted these results with a sample of 10 sented evidence suggesting that enamel sur- modern human incisor teeth (specific idenface manifestations, termed perikymata, of tity not noted), with perikymata counts internal enamel bands, termed striae of Ret- ranging between 165 and 202 (Bromage and zius, could be employed to determine the Dean, 1985). These differences in perikylength of time that a tooth crown took to form mata counts have been interpreted to indi- of earlier hominine brain endocasts (e.g., Falk, 1987) and archaeological evidence and inference (e.g., Potts, 1984a,b; Potts and Shipman, 1981; Shipman, 1983, 1986; Binford, 1981, 1984, 1985) have led some to suggest that the essential core of humanness was not to evolve until the evolutionary origin of modern humans. In this climate of opinion, the patterns of growth and maturation in early hominines are also being reexamined. Fig. 1. The labial surface of an incisor from the site of Hasanlu at x20. Perikymata are visible on the surfaces ofthe incisors as bands. While many of the perikymata are quite distinct, others are difficult to distinguish even at the higher magnifications at which these structures were counted. 177 PERIKYMATA AND CROWN FORMATION TIMES cate different total incisor crown formation times between australopithecines and modern humans (Bromage and Dean, 1985; Beynon and Dean, 1988). Bromage and Dean (1985) based their conclusions on their calculation of total crown formation time in the fossil incisors: This was taken as the time of‘enamel formation, documented by the visible increments of growth (the perikymata) plus an additional estimation of the time recorded in hidden enamel increments and not expressed on the surface. Bromage and Dean (1985) estimated that there were 20-30 hidden increments not observable as surface perikymata, which were said to represent 6 months of additional enamel formation unrecorded by observable surface manifestations. Therefore, the total time of incisor enamel crown formation was derived from the total number of perikymata observable on the tooth surface, each perikymata accounting for 1 week‘s growth, added to 6 months of hidden growth. The perikymata counts and derived crown formation times originally published by Bromage and Dean (1985) are provided in Table 1. Bromage and Dean (1985) concluded that the total time for enamel formation in all the australopithecines including early Homo was significantly shorter than that in modern humans, and, based on the ages at death derived from these crown completion times, “the growth periods for Plio-Pleistocene hominids were similar to the modern great apes” (Bromage and Dean, 1985526). These original conclusions are interesting to review. The important aspect of these conclusions is that early hominines had a foreshortened and apelike growth period. This idea was based on the perikymata data, which were said to document fossil hominines’ shorter incisor crown formation times relative to modern human dental development: “Earlyhominids evidently had shorter periods of dental development than modern humans and therefore a less prolonged infancy” (Beynon and Dean, 1988). At the time Bromage and Dean (1985) published their work, no enamel histological data for modern great ape incisor formation comparable to that available for modern humans were known. This made direct comparisons of enamel formation times between apes and humans difficult. Likewise, it was difficult to evaluate the inferential interpretation of an apelike growth period based on incisor crown formation times. The reference TABLEI. Perikymatacountsandcrownformationtimes from Bromage and Dean (1985) Taxonomic ~ O U D SamDle no. A. afarensis A. africanus A. robustus 5 Early Homo 2 H. s. sapiens 10 1 1 Range of crown formation times 3 years 3 years 1.1 months 1 year 7.2 months 2 years 1.8 months 2 years 6.2 months 2 years 7.4 months 3 years 7.9 months 4 vears 4.4 months cited by Bromage and Dean (1985) for data on ape incisor crown formation times was a previous study by Dean and Wood (1981) in which radiographs of unaged ape skeletal specimens had been used to predict great ape crown formation times. However, in this 1981 reference, Dean and Wood had published the observation that the extant evidence “suggests that there is no difference in the basic rate of enamel formation between humans and pongids” (Dean and Wood, 1981:123). In fact, the earlier reference of Dean and Wood (cited by Bromage and Dean, 1985) actually shows only a 0.25 year differende in incisor development time between humans and apes (Dean and Wood, 1981: 116). Thus it is unclear how the conclusions of an apelike growth period for fossil hominines could have been distinguished on the basis of comparative incisor crown formation times between the taxa. Since the original Bromage and Dean (1985) study was published, additional fossil specimens have been analyzed that require modification of the original conclusions. Additional perikymata counts from fossil hominine incisors (three A. afarensis, four A. boisei, and two early Homo) have been published that expand the hominine incisor perikymata count range from 57 to 180, as well as extend the respective crown formation times (Dean, 1987a; Beynon and Dean, 1988).For example, among the australopithecine specimens are the incisors from Laetoli Hominid 3 (LH 3). The maxillary central incisor has a perikymata count of 170 and the maxillary lateral incisor has a perikymata count of 180, translating to crown formation times of 3.8 and 4.0 years, respec- 178 A.E. MA” tively (by the method of Bromage and Dean, 1985). Both the perikymata counts and the derived crown formation times for the LH 3 specimen are within the originally published range of modern human incisor crown formation times reported by Bromage and Dean in their 1985 paper, although no explanation has ever been published for this contradiction of their conclusions (Dean, 1989). Furthermore, in 1986, Dean et al. reported on the analysis of an upper right central incisor from the neandertal youngster from Devil’s Tower, Gibraltar. They found 119 perikymata on this tooth, or an estimated 2.8 years for total crown formation. This total of 119 perikymata is notable, for it is significantly below the range of 165-202 perikymata that is the modern human incisor perikymata range given by Bromage and Dean (1985) and is also below the perikymata results obtained for australopithecines such as LH 2 (with 130 perikymata on the mandibular right central incisor), Sts 24a (135 perikymata on the maxillary right central incisor) (Bromage and Dean, 19851,LH 3 (170 and 180, as noted above), and very close to the early Homo specimen from East Turkana, ER 808 (123 perikymata on the maxillary lateral incisor [side unspecified]) (Beynon and Dean, 1988).Dean et al. (19861, however, concluded that, by reference to data on modern human crown formation times, “the time of crown completion derived from the perikymata counts in this study for the Gibraltar child are likely to be accurate, as the figure obtained falls within the known ranges of crown completion in Homo sapiens” (Dean et al., 1986:306). Why were 119 perikymata deriving from a Homo sapiens neanderthalensis child accepted as within the modern human range of crown formation times when as many as 124-180 perikymata deriving from australopithecines were excluded and determined specifically apelike? There is an inconsistency in the acceptance of 119 perikymata and a 2.8 year crown formation time as humanlike for the Devil’s Tower neandertal child (Dean et al., 1986) while rejecting australopithecine incisors including early Homo specimens as humanlike, even when several of these, e.g., Sts 24a, possess comparable numbers of perikymata (Bromage and Dean, 1985; see Table 1). Review of these previous studies identifies methodological differences that lead to these conflicting interpretations. A source of these discrepancies is found in the selective use of ET AL data from modern humans: In the work of Bromage and Dean (1985) on the australopithecines, only an estimation of a mean was employed to represent incisor crown formation times in modern H . sapiens while in the analysis of Dean et al., (1986) on the Neandertal juvenile data on the ranges of crown formation times for modern humans were employed. The debate concerning human and ape dental maturation has been recently extended by Dean, who notes that the “great apes may take up t o 8 or 9 years to form enamel on their incisors and canines” (1989:171). Dean (1989:171) continues by observing that perikymata counts of australopithecine anterior tooth crowns do not take “as long as great ape anterior crowns take to form.” Thus it is now reasonable to examine critically the claim, based on incisor crown formation times, presented in the original 1985 paper that “Plio-Pleistocene hominids had markedly abbreviated growth periods relative to modern man, similar to those of the modern great apes” (Bromage and Dean, 1985525). PRESENT STUDY The difficulties in the interpretation of enamel surface structures, and the often conflicting claims made about their significance in earlier hominines, suggest that an important current objective ought to be the enlargement of the reference sample of enamel surface structure counts for living and extinct members of Homininae. Materials and methods Twelve modern human incisor teeth and six Upper Pleistocene hominine teeth were chosen for this study. All 18 teeth were unworn and unerupted. Fourteen teeth were crown complete with little o r no root development. Four teeth, all fossil specimens, possessed not quite complete crowns. Because fossil specimens cannot themselves undergo the preparation necessary for electron microscopy, which is the preferred technique for viewing enamel surface details, replicas of the fossils must be produced. The six fossil hominine teeth were reproduced using a dental silicone impression material, and the resultant epoxy casts were imaged by the scanning electron microscope (SEM). A critical determinant in the accuracy of reproduction, and thus in the quality of detail obtained in the casts, is the choice of molding and casting media. The advantages PERIKYMATA AND CROWN FORMATION TIMES and problems with various materials has been frequently discussed in the literature (Gordon, 1984; Barnes, 1978, 1979; Pameijer, 1978; Scott, 1982). As the data base in studies of fossil dental microstructure have been replicas, a control study for protocol was undertaken prior to the collection of data in the present study. The techniques used for replication of fossil specimens used by Bromage and Dean (1985) and by Dean et al. (1986) were outlined by Bromage (1985). The use of the materials suggested by Bromage (19851, especially the use of hardware epoxies, met with little success in the present study. As a control, prior to data collection in the present study, two of the modern human incisors were employed t o examine the amount of detail captured by the replication process. A Reprosil Light Body mold was made of these teeth. One mold was cast in low viscosity (2,000 centipoise), slow-setting Hysol-brand epoxy (TE 6345 with HD 0111); the other mold was cast in a medium viscosity, but fast-cure epoxy, the sort that can be purchased in hardware stores. Both of the original specimens, and both replicas, were sputter-coated with gold palladium at the same time and mounted in the SEM chamber as matched pairs to be viewed simultaneously at magnifications of x20, x40, and ~ 8 0 These . were chosen because the only previously published magnification level cited for perikymata counting was x50 (Dean et al., 1986). Perikymata counts were made on the matched pairs. The differences between the detail quality of the two types of casting resins were dramatic. The use of Reprosil as a molding media, with Hysol epoxy as the casting resin, was an excellent combination for the production of highly accurate replicas of perikymata. Perikymata numbers on the replica were within +5 of the original specimen. The castings were not effected by pitting or bubbling a t these magnifications as outlined by Gordon (1984). In contrast, the use of Reprosil with the fast-cure hardware epoxy produced inaccurate perikymata counts on the replica. The number of perikymata observed on the replica were one-third less than on the original tooth. These results are similar to those described by Hillson and Jones (1989) and indicate that the techniques used to produce replicas for study is crucial in the analysis of enamel surface microstructure in fossils. Although Bromage (1985) tested these materi- 179 als on various structures within dentin, it is possible that the porosity differences between dentin and enamel is critical here. Gordon (1984) noted that the production of replicas of very smooth surfaces, such as the enamel surface, can create a distinct series of problems not encountered in the reproduction of other hard tissues. Samples Modern human incisors. The samples of 12 modern human incisor teeth came from two sites. Ten specimens came from the Iranian site of Hasanlu, a walled city sacked a t about 3000 B.C. These Hasanlu specimens were discovered in the citadel, described by the excavators as the location of a mass murder of the city’s inhabitants by invaders (Muscarella, 1991). None of these children, therefore, died as the direct result of illlness. The other two modern human specimens came from the 1200-year-old Island Field site in the State of Delaware. These 12 modern human incisors were chosen because each tooth was unerupted and crown complete with a portion of root formation. Perikymata could be viewed over the entire crown surface at low magnifications using a binocular microscope. Neandertal incisors. Krapina 90, 91, 93, 94, and 95 (RadovEiC et al., 1988) were molded as described above with the permission of Dr. Jakov RadovEid of the Croatian Natural History Museum, Yugoslavia, in the summer of 1988. An additional mold was made of one unerupted incisor in the crypt of Krapina mandible A. To prevent any effects from mold shrinkage, positive replicas were cast using Hysol-brand epoxy within 48 hours of making the molds. Perikymata could be counted on the entire labial surfaces of all of these dental casts except the incisor of mandible A. Two of the Krapina specimens, 91 and 93, were less than two-thirds crown complete, and Krapina 95 was only one-third crown complete. Krapina 94 was assessed as three-fourths crown complete, and Krapina 90 had about one-third root development. Thus only Krapina 90 is precisely comparable with previously collected data. Data from the other Krapina teeth are, however, presented (see Table 2). Microscope protocol All 12 modern human specimens, as well as the epoxy casts of the fossils, were sputter- 180 A.E. MAXN ET AL TABLE 2. Perikymata counts and estimated crown formation times in early hominines Taxonomic group specimen A. afarensis LH2l LH32 LH32 LH62 A. africanus Sts 24a' A. robustus SK 62l SK 62' SK 63l SK 71' SK 73l A. boisei OH 302 ER 8122 ER 14772 ER 18202 Early Homo SK 74b' ER 820' OH 62 ER 80@ H. sapiens neanderthalensis Gibraltar3 Krapina 90 Krapina 9 l t Krapina 93t Krapina 94t Krapina 95' Homo sapiens sapiens Hasanlu Perikymata counts 130 170 180 116 3.0 3.8 4.0 2.7 135 3.1 57 64 86 >6 1 >79 1.6 1.7 2.1 >1.7 >2.0 101 86 92 82 2.4 2.2 2.3 2.1 110 105 95 123 2.6 2.5 2.3 2.9 119 205 f 10* >loo f 4 >lo7 2 >144 f 7 >50 2.8 4.4 124 f 1 134 f 4 99 f 5* 128 f 4 157 f 12* 9 0 f 11 75 7* 134 2 103 t_ 1 93 f 1 148 f 7 113 3 2.9 3.1 2.4 2.9 3.5 2.2 1.9 3.1 2.5 2.3 3.3 2.7 + * + Island Field Estimated crown formation time + The present study data are listed as the mean and i 1 SD of the counts: all teeth were counted by three individuals with the exception of Krapina 90 and three Hasanluspecimens (noted by *), which were counted by four observers. The Krapinaspecimens (noted by +)are crown incomplete.Thereferencesforperikymatacountdatainclude1BromageandDean(1985),2BeynonandDean(1988),and3Deanetal.(1986). coated with gold palladium and viewed directly by SEM. All teeth were viewed a t x20, ~ 4 0and , x80 magnifications at 6,12 and 20 KeV. The labial surface of the entire crown surface was recorded on Polaroid film. A montage made up of the several pictures that cover the entire crown surface was produced. At least three observers counted perikymata on each tooth, and the results were averaged. It should be pointed out that rarely did the counters record exactly the same number of perikymata on any one tooth. The reasons for this are obvious: Perikymata are found on the enamel surface on many of the specimens in a very complicated pattern. Some perikymata are clear and distinct, others are packed together so densely that counting individual structures is extremely difficult. Occasionally, perikymata are so indistinct that identification becomes subjective. The location on the labial surface where perikymata were counted PERIKYMATA AND CROIWN FORMATION TIMES 181 RESULTS Modern human incisors Each of the 12 immature modern human specimens used in this study showed perikymata over the entire crown surface. Additionally, they showed the pattern of compactness of perikymata similar to that previously reported in the literature for normal human teeth (Dean, 1987b), specifically, a gradual increase in the density of perikymata towards the CEJ. Table 2 details the perikymata counts on the 12 modern human incisor teeth. The range of perikymata counts in this sample is 75-157, with a mean of 116, a median of 118, and an SD of 25. Neandertal incisors Table 2 also includes the individual perikymata counts from the prepared fossil epoxy casts. Only Krapina 90 has a fully formed crown and therefore is the only suitable specimen for direct comparison with previous perikymata counts on crown complete incisors. Krapina 90, a lower right lateral incisor, has 205 10 perikymata. Figure 2, at x20, shows that perikymata are present over the entire crown surface with progressive compactness of perikymata a t the CEJ. Perikymata were counted at x40. Krapina 94, an upper right central incisor about three-fourths crown complete, and therefore without the area of greatest compactness of perikymata, possesses 144 2 7 perikymata. * Fig. 2. Montage ( ~ 2 0of ) the entire labial surface of Krapina 90. Even though the tooth was excavated nearly 100 years ago and shows some abrasion, perikymata could be viewed over the entire tooth crown. Perikymata counts were performed at ~ 4 0 . DISCUSSION Neandertals The data from Krapina 90 begin to establish a range of perikymata counts for H.S. neanderthalensis incisors, one that must extend at least from 119 (Devil’s Tower, Gibraltar) to 205, the present specimen. Crown formation times calculated from these perikymata numbers range from 2.8 to 4.4 years (according to the method of Bromage and Dean ). These data provide an interesting perspective for interpretations of also effected the number of observable neandertal dental development. Krapina 90 has over 80 more perikymata perikymata. Whereas previous researchers have reported only a single perikymata than the Gibraltar child and falls at the high count number on fossil specimens, we be- end of the previously published range of lived it is appropriate that perikymata perikymata counts for modern human inciranges, means, and standard deviations be sor teeth. The Krapina 90 specimen reprereported for each of the specimens. As in all sents an estimated crown formation time of scientific work, some error is to be expected 4.4 years, well within the modern human range for mandibular incisors by comparison on the basis of chance alone. 182 A.E. MA" ET AL. normal modern humans from 75 to 157 and expand the known range of perikymata numbers in modern human incisors from 75 to 202 (with estimated crown formation times of 1.9 to 4.4 years, according to the methods of Bromage and Dean [19851)). The data from this study are notable for their lack of correspondence to the previously published range for modern humans165-202 (Bromage and Dean, 1985twhile including most of the published fossil hominines-57-180 perikymata (Bromage and Dean, 1985; Dean, 1987a; Beynon and Dean, 1988). The perikymata numbers we report here for human incisors are similar to those found in an independent study of modern human incisors (i.e., a range of 111-179, N = 23) (Bacon, 1987). To understand the sampling differences in perikymata counts from modern humans, it is necessary to consider the effects of the very small sample sizes in both studies, as well as the potential variability in the biological feature being examined. The small samples (N = 10, Bromage and Dean [19851;N = 12, the present study) likely come from genetically distinct populations, given the specific archaeological context of those used in the present study (Bromage and Dean [19851do not specify the provenience of their sample). The features under comparative investigation, perikymata, are said to represent a biological feature (crown formation time) that is used to document one of the most genetically (and environmentally) variable processes known in humans: growth and maturation. Sampling effects are to be expected and narrow ranges to be suspect. The limitation of the previously established modern human perikymata range was already implicit in the published analysis of Dean et al. (1986) on the Gibraltar specimen: 119 perikymata, a number clearly outside of the previously established human perikymata range, was found by Dean et al. to be compatible with the range of crown formation times in modern humans. Thus the perikymata range found in the present study, 75-157, while previously regarded as nonhuman or apelike, actually overlaps with other published analyses of perikymata Modern humans counts leading to human crown formation The data presented here also substantially times. This line of reasoning leads to a reasenlarge the published range of perikymata sessment of the estimates of Bromage and numbers in modern human incisors. These Dean (1985) regarding what it is to be huresults document a range of perikymata man, at least in terms of perikymata numnumber in a small sample of apparently bers and crown formation times. with radiographic studies from modern children. These radiographic studies have published mean and median ages of dental developmental stages for samples of modern human children: The Krapina 90 specimen is considerable above the median age of 3.6 years for lateral incisor crown completion published by Fass (1969) for a sample of modern human children and is likewise above the means of Nolla (1960) for both the males in her sample, 4.3 years of age for completion of lateral incisors, and the females, 4.0 years. Krapina 94, while not a complete crown, is nonetheless informative. An incomplete crown, without the area of greatest compactness of perikymata, it contains 25 more perikymata than the fully formed incisor crown of the Devil's Tower, Gibraltar neandertal child specimen (with 119 perikymata). Thus the Krapina 94 specimen, a right upper central incisor, like the Devil's Tower specimen, has 21% more perikymata on a tooth that is 25% less complete. Krapina 94 (a maxillary central incisor a t about three-fourths crown complete), with 3.4 years already documented according t o perikymata assessment, is already within the modern human range for total crown completion. For example, the earliest maxillary central incisor crown completion time found by Fass (1969)in his sample of modern human children was 2.5 years, while the same sample had a median of 3.85 years. A number of other radiographic studies further support the human developmental time period of this Krapina tooth, with maxillary central incisor crown completion ages from different samples of children that include a mean of 3 years (Haavikko, 1970),a mean of 4.25 years (Nolla, 1960) and ranges of 3.56.25 years at the 2 SD level (Moorrees et al., 1963). These data do not support the generalization made by Dean et al. (1986) regarding relative dental developmental acceleration in incisor crown formation times in neandertals in general. Perikymata counts on Krapina 91, 93, and 95, all with only partially formed enamel, are presented in Table 2. PERIKYMATA AND CROWN FORMATION TIMES 183 TABLE 3. Comparative perikymata number and perikymata-derived crown formation times for hominines' Taxonomic group A. afarensis A . afrieanus A. robustus A . boisei Early Homo H. s. neanderthalensis H. s. sapiens Perikymata no. Crown formation (years) 116-180 135 57-86 82-101 95-123 119-205 75-202 2.7-4.0 3.1 1.6-2.1 2.1-2.4 2.3-2.9 2.8-4.4 1.9-4.4 'These data invalidate previous claims of meaningful distinction between modern humans and all early hominines. The sample sizes within each of these groupings make statistical tests of intergroup significant differences meaningless: All groups are significantly different from one another (ANOVA P < 0.01). While the previous human data of Bromage and Dean (1985)cannot be tested against the present data (only mean and range were published), they would show significant mean differences with the present human group, thus making interhominine comparisons on these small samples questionable. Crown formation times: Perikymata us. radiographic evidence Perikymata are said to represent temporal markers of crown formation time and to gain importance as they accurately reflect rates of growth and maturation. The significance of perikymata counts as data bases for the derivation of crown fonntion times and age at death have been recently reviewed (Beynon and Dean, 1988; Dean, 1989). Neither the validity of perikymata counts, and the associated estimates of both the hidden increments (said to document time before the perikymata reach the surface) and the time before the onset of calcification (3 months) as predictors of crown formation time, nor the claims that crown formation times thus derived in the fossil hominines could provide "good evidence of a great apelike developmental growth period" (Dean et al., 1986:308) have ever been seriously questioned. The perikymata counts documented in the present study, taken together with the inconsistencies in the previously published analyses, make it necessary to reestablish the range of perikymata counts in modern humans and to clarify the implications this has for the interpretation of data from fossil hominines. Implicit in this analysis is an evaluation of the presently postulated relationships between perikymata counts and crown formation times. The crown formation times derived from the perikymata counts as outlined by Bromage and Dean (1985) and subsequently employed (Beynon and Dean, 1988) must be recognized as hypothetical, based on theoretical ideas regarding dental development. As hypotheses, they are amenable to testing. If the proposed method for calculating crown formation times from perikymata is accurate, the range of perikymata postulated to represent all modern humans, together with the perikymata-derived crown formation times, will predict the known times of crown formation radiographically documented for modern humans. An overlap between perikymata-derived crown formation times and radiographically documented crown formation times was proposed by Bromage and Dean (1985) to verify their methods. Their claim of validation between their perikymata numbers and modern human incisor crown formation times deriving from radiographic studies has two complications: 1) radiography tends to lag behind actual calcification time by some months (e.g., Hess et al., 1932); and 2) an age-independent factor, crown formation time as derived here, is being compared with age-dependent data (crown formation times from radiologic inspecition derive from the chronological ages of living children). Bromage and Dean (1985) did not initially comment on the radiographic problem and postulated that the gap between crown calci- 184 A.E. MA” ET AL. fication time and chronological age could be resolved by estimating the age of initial crown calcification. They stated that central and lower lateral incisors are “well known t o begin calcification about 3 4 months after birth” (1985526).Thus, to derive chronological age, 3 months were added t o the perikymata-derived crown formation times. Employing this method, Bromage and Dean (1985526) stated, referring to Nolla (1960) and Fass (19691, respectively, that “estimates for the duration of crown formation in the modern human sample (mean number of perikymata 188, range 165-202); mean crown formation times (4.2 years) agree with estimates of crown formation times calculated from longitudinal and cross-sectional studies.” A review of this aspect of their methodology is important in evaluating the valid relationship between perikymata and crown formation times. While Bromage and Dean (1985) are not incorrect that their human perikymata range agrees with estimates of modern human crown formation times, it should be noted that their figures substantially underrepresent both the modern human range as well as the mean ages for incisor crown completion that they cite. Data from Fass (1969)document a range of crown formation times from 2.5 years (or 2.75-2.8 chronological years, when the 3 months between birth and predicted onset of crown calcification are added on), the earliest observed crown completion of maxillary and mandibular central as well as mandibular lateral incisors, to medians of 3 . 0 5 4 . 5 years (or 3 . 3 4 . 7 5 chronological years). These crown completion times translate to theoretical perikymata counts ranging from 104 to 209, with medians of 133-209 perikymata. These ranges far exceed those found by Bromage and Dean (1985) for humans and in fact encompass most of the fossil hominines they investigated. The earliest reported incisor crown formation times in modern humans are those from this sample reported by Fass (1969), which predicts a perikymata-count of 104, for the children whose dentition was observed to be completed by this age. Such a perikymata count is 61 perikymata below the human range according to the original propositions of Bromage and Dean (19851, directly in the range of most of the fossil hominines. As it is well known that radiography underestimates actual calcification by several months, the actual human ranges must be even broader on the lower end than calculated above and theoretically substantially lower perikymata counts in modern humans than proposed by Bromage and Dean in 1985. The addition of the information from the present study to these previously published perikymata data assists in bridging this lower range gap. Alternatively, data from Nolla (1960) give mean times for incisor crown completion ranging from 3.25 to 5.25 years (chronological ages of 3.5-5.5 years) or theoretical perikymata counts of 143-248. This translates to perikymata counts 46 greater than the previously published upper range. The data from the present study add no information to this upper range gap. Again, as radiographic studies tend to lag behind actual calcification, it is difficult to predict theoretically the extent of divergence between histological and x-ray data. In addition t o the studies cited by Bromage and Dean (1985) as providing data to validate their propositions of the relationship between perikymata counts and crown formation times, published data on the ages of incisor crown completion in samples of modern humans include the maxillary data of Moorrees et al. (1963). Mean crown formation times range from 4.5 to 5.7 years (4.95-6 chronological years), with the 2 SD range extending from 3.5 years (3.8 chronological years) for the maxillary central incisor in females to 7.15 (7.4 chronological years) for the lateral incisors in males, or theoretical perikymata counts ranging from 159-347. This is again far beyond the range of perikymata numbers proposed as the modern human range by Bromage and Dean (1985);the present data add no information to the upper end of this range. The data of Moorrees et al. (1963) emphasize a previously ignored aspect of incisor development: All of the earlier perikymata studies have treated central and lateral (maxillary and mandibular) incisors as identical teeth. This is entirely inappropriate in terms of the actual time of formation according to the data collected by Moorrees et al. (1963) and deserves careful attention in future analyses. It is unlikely that all of these teeth initiate calcification simultaneously at 3 postnatal months of age, and it is clear that they do not complete calcification in the same amount of time. Thus a review of some of the studies documenting ages of incisor crown completion in modern humans amply illustrates the lim- * PERIKYMATA AND CROWN FORMATION TIMES ited nature of the human perikymata range as originally published. This original perikymata range, translating to crown formation times of 3 . 7 4 . 4years represents only 15%of the known human incisor crown formation range: The documented range for completion of modern human incisors, both central and lateral, is 2.5-7.2 years. The original proposition of Bromage and Dean (1985), that their perikymata range (165-202) and the crown formation times derived from them (3.7-4.4years), were good estimates for the known crown formation times of modern humans is a considerable generalization and overstatement. These figures actually represent only 15% of the documented time required for incisor crown formation by even a small sample of modern human children. The remaining 85%variation, previously unaccounted for, is an unacceptably high oversight. Based on radiographic evidence, the theoretical range of perikymata counts for modern human incisors must be at least 104347, using the method outlined by Bromage and Dean (1985). Thus the observed human incisor perikymata counts found in the present study are within theoretically predicted limits. However, the new perikymata range, 75-202, still fails to accommodate the known range of incisor crown completion times. Only 61% of the upper range is estimated, or 4.4 vs. 7.2 years; and a 1.9 year crown formation time is predicted on the lower end, or 0.6 years below the earliest documented crown formation times by radiographic data. Some of the latter may be explained by the lag in radiographic evidence of actual calcification. But a 1.9-4.4 year range of crown formation time as derived from perikymata counts does not overlap with the known range of 2.5-7.2 years documented radiographically. 185 formation times in humans by as much as 3 years. Both the original study and the present one are based on very small sample sizes. To accommodate all humans, a much larger sample size of a number of populations is required; however, this would have to expand the perikymata number nearly 150%. Second, perikymata do not regularly record exactly 7 days of growth. Crown formation times as derived above assume a 7 day periodicity (a circaseptan rhythm; Dean, 1987a). However, as was early noted by Asper (1916) and reported by Bromage and Dean (19851, the number of cross striations repeatedly found between striae of Retzius range from 6 to 10 in different individuals. Thus the derivation of crown formation times from perikymata counts will vary in accordance with the number of days (or cross striations) between each striae. Correspondingly, significant changes in crown formation times would occur from calculations based on 6 striaelperikymata vs. 10 striae1 perikymata, on the order of 150%. When the fossil hominines are reassessed by these cross-striation intervals, the crown formation times range from 1.75 years (6 cross striations between striae) to 5.4 years (10 cross striations between striae), for a perikymata range of 57-180. This compares with a modern human range of 1.7-6.1 years, when the human perikymata range thus far documented (75-202) is subjected to the same recalculations. These ranges are notable for their close correspondence. However, the lower range of the human crown formation time still underestimates the radiographic data and thus calls into question these methods. A similar check on fossil hominines is obviously unavailable, but the human data should caution against precocious assumptions regarding these specimens. Perikymata as predictors of human Third, chronological age derived from crown formation perikymata counts cannot be as simple as Based on this lack of correspondence be- outlined. It is highly probably (if not certain) tween predicted and observed crown forma- that there is an error in the temporal estimation times, a reassessment of the proposed tion factors in the method as originally properikymatdcrown formation time method as posed, that is, with the 6 months of hidden proposed is in order. A number of sources of increments andlor the time of onset of calcierror may account for this failure of periky- fication. For example, the proposed method mata-derived crown formation times t o pre- oversimplifies dental growth and development in failing to account for variability in dict known human crown formation times. First, it may be that the range of human the time of onset of incisor calcification. The perikymata is still too narrow. The data from data of Moorrees et al. (1963) cited above the present study extend the previously pub- document the later develoment of the lateral lished range, but still underestimate crown incisors in the children in their sample. Thus 186 A.E. MA” ET AL. a distinction in these two teeth should be made in the timing of their respective development. Furthermore, based on the well-documented interindividual variability in maturational rate, it is not possible that all humans begin calcification of their central and lateral incisors a t precisely 3 postnatal months.of age, as this method assumes. This faulty generalization would lead to an error in the estimation of chronological age, which would over- and, more importantly, underestimate both crown formation times and age a t death by a substantial amount. This may be an important source of crown formation time underestimates calculated from perikymata counts. Finally, perikymata, striae of Retzius, and cross striations may not represent an accurate maturation clock as proposed. Enamel matrix formation and calcification is an extremely complex biological process resulting in a structure, dental enamel, that itself is three-dimensionally complex. This threedimensional aspect of enamel is thus far unappreciated in the “counting” approaches presently employed in anthropology. Cross striations may, or may not, be daily growth increments. In contrast to previous publications, this is not yet an experimental fact. While it is assumed to be the case (e.g., Osborn and Ten Cate, 19831, data for this derive from inferential analysis of related structures (e.g., Rimes, 1986) and from calculations of observable cross striations between tetracycline markers in developing teeth, the tetracycline following calcification, not enamel rod growth (e.g., Okada, 1943; Bromage, 1989). In this context, it should be noted that some recent histological research has suggested that these observable phenomena may be morphological structures, artifacts of specimen preparation, or other epiphenomena not formed in anytimeordered predictable sequence (Weber and Ashrafi, 1979; Weber et al., 1974; Warshawsky and Bai, 1983; Warshawsky, 1989). A consideration of the data presented here, and a review of the methods surrounding the use of perikymata, require a reassessment of the crown formation times and ages at death of the fossil hominines published in recent years, as well as the conclusions drawn from these data. Whether viewed from the perspective of perikymata counts or the crown formation times derived from them, there is no valid evidence to distinguish total incisor crown formation times in modern humans from those in the fossil hominine sample (Table 3). Taken together with the recently published data from Dean (1989) on the timing of ape incisor crown calcification, previous conclusions regarding early hominines as “ape-like in growth and development” must be reconsidered. Moreover, the mismatch between incisor crown formation times derived from radiographic data and the present data base on perikymata raises questions on the validity of the method described and on the accuracy of chronological age assessment. CONCLUSION The data presented here document a range of perikymata number in H . sapiens that includes those observed in many australopithecines, including early Homo, and the single neandertal previously reported. Employing these perikymata numbers as originally proposed (Bromage and Dean, 1985), incisor crown formation times in early hominines and modern humans cannot be reliably distinguished. Taken together with recent reports of evidence on ape incisor crown formation times, we conclude that previous claims based on these data for “good evidence of a great-ape-like developmental growth period (Dean et al., 1986:308) in early hominines are invalid. The data presented here support a more critical evaluation of these methods and suggest that it is premature to base major biological reconstructions on these approaches prior t o substantial further research. Furthermore, we find that calculation of incisor crown formation times from the present modern human perikymata range does not predict the known range of modern human incisor calcification times, but consistently underestimates modern humans. It is likely to do the same for the fossil specimens. Potential error in this method leads t o the conclusion that it is inappropriate with the present methodology to age fossil specimens using perikymata counts. With attention to the problems noted, the use of perikymata, striae of Retzius, and cross striation data may, assuming they are shown to be time-related structures, or may not be able to offer us a range of ages forindividual teeth in the future. The original precision of the method that published ages at death to the tenths of a year in the absence of any standard deviation or range is invalid. PERIKYMATA AND CROIWN FORMATION TIMES LITERATURE CITED Asper von H (1916)Uber die 'Braune Retzius'sche Parallelstreifung' im Schmelz der Menschlichen Zahne. Schweiz. Vierteljahrschr. Zahnheilk. 26.275-214. Bacon A-M (1987) Estimation de I'age des individus a partir du comptage des perikymaties. D.E.A. de Paleontologie, Universite Paris VI. Barnes IE (1978) Replication techniques for the scanning electron microscope. I. History, materials and techniques. J. Dent. 6t327-341. Barnes IE (1979) Replication techniques for the scanning electron microscope. 2. Clinical and laboratory procedures: Interpretations.J. Dent. 75'5-37. Beynon AD, and Dean MC (1988) Distinct dental development patterns in early fossil hominids. Nature 335:509-5 14. Binford LR (1981) Bones: Ancient Men and Modern Myths. New York: Academic Press. Binford LR (1984) Bones of contention: A reply to Glynn Isaac. Am. Antiquity 49:164-167. Binford LR (1985) Human ancestors: Changing views of their behaviors. J. Anthropol. Arch. 4.292-327. Boyde A (1976) Amelogenesis and the structure of enamel. In B Cohen and IRH Kramer (eds.):Scientific Foundation of Denistry. London: Heinemann, pp. 335352. Bromaee TG (1985) Svstematic inauirv in tests of negaiive/positive replica combinations for SEM. J. Microsc. 137209-216. Bromage TG (1989) Experimental confirmation of enamel incremental periodicity in the pitail macaque. Am. J. Phys. Anthropol. 78.197. (Abstract) Bromage TG, and Dean MC (1985) Re-evaluation of the age a t death of Plio-Pleistocene fossil hominids. Nature 31 7t525-528. Dart RA (1948) The infancy of Australopithecus. In Robert Broom Commemorative Volume. Special Publication of the Royal Society of South Africa, pp. 143152. Dean MC 11987a) Growth layers and incremental markings in hard tissues: A review of the literature and some preliminary observations about enamel structure inPuranthropus. J. Hum. Evol. 16t157-172. Dean MC (1987b3 The dental development status of six juvenile fossil hominids from Koobi Fora and Olduvai Gorge. J. Hum. Evol. 16.197-213. Dean MC (1989) The developing dentition and tooth structure in hominoids. Folia Primatol. 53t160-176. Dean MC, Stringer CB, and Bromage TG (1986) A new age a t death for the neanderthal child from Devil's Tower, Gibraltar and the implications for studies of general growth and development in neanderthals.Am. J. Phys. Anthropol. 70t301-309. Dean MC, and Wood BA (1981) Developing pongid dentition and its use for ageing individual crania in comparative cross-sectional growth studies. Folia Primatol. 36t111-127. Dobzhansky T (1962) Mankind Evolving: The Evolution of the Human Species. New Haven: Yale University Press. Falk D (1987) Hominid paleoneurology. Ann. Rev. Anthropol. 16:13-30. Fass EN (1969) A chronology of growth of the human dentition. J. Dent. Child. 36t391-401. Gordon KD (1984) Pitting and bubbling artefacts in surface replicas made with silicone elastomers. J. Microsc. 134:183-1 88. Gould SJ (1976) Ontogeny and Phylogeny. Cambridge: Harvard University Press. 187 Haavikko K (1970) The formation and the alveolar and clinical eruption of the permanent dentition. Suomen Hammaslaakariseuran Toimituksia 65t103-170. Hess AF, Lewis, JM, and Roman B (1932) A radiographic study of calcification of the teeth from birth to adolescence. Dent. Cosmos Z.miu:1053-1061. Hillson SW, and Jones BK (1989) Instruments for measuring surface profiles: An application in the study of ancient human tooth crown surfaces. J Arch Sci. 16t95-105. Holloway RL (1972) Australopithecine endocasts, brain evolution in the Hominoidea and a model of hominid evolution. In RTuttle (ed.): The Functional and Evolutionary Biology of Primates. Chicago: Aldine, pp. 185204. Holloway RL (1983) Human paleontological evidence relevant to language behavior. Hum. Neurobiol. 2:105-114. Isaac GL (1972) Chronology and the tempo of cultural change during the Pleistocene. In WW Bishop and J A Miller (eds.): Calibration of Hominid Evolution. Edinburgh: Scottish Academic Press, pp. 3 8 1 4 3 0 . Isaac GL (1978) Foodsharing behavior of protohuman hominids. Sci. Am. 238:99-108. Lovejoy CO 11981) The origin of man. Science 211:341350. Mann AE (1972) Hominid and cultural origins. Man ns. 7.379-386. Mann AE (1975) Some Paleodemographic Aspects of the South African Australopithecines. Philadelphia: University of Pennsylvania Press. Moorrees CRA, Fanning EA, and Hunt EE (1963) Age variation of formation stage for ten permanent teeth. J. Dent. Res. 42t1490-1502. Muscarella OW (1991) Warfare at Hasanlu in the late 9th Century B.C. Expedition (in press). Nolla CM (1960) The development of the permanent teeth. J. Dent. Child. 27r254-266. Okada M (1943) Hard tissues of animal body: Highly interesting details of Nippon studies in periodic patterns of hard tissues are described. Shanghai Evening Post 15-31. Osborn JW,and Ten Cate AR (1983) Advanced Dental Histology. London: Wright. Pameijer CH (1978) Replica techniques for scanning electron microscopy-A review. Scan. Electron Microsc. 2:8314336. Potts R(1984a) Home bases and early hominids. Sci. Am. 72t338-347. Potts R (198413) Hominid ecology? Problems of identifying the earliest hunterlgatherers. In R Foley (ed.): Hominid Evolution and Community Ecology. New York: Academic Press, pp. 129-166. Potts R, and Shipman P (1981) Cutmarks made by stone tools on bones from Olduvai Gorge, Tanzania. Nature 291:577-580. Radovcic J, Smith FH, Trinkaus E, and Wolpoff MH (1988) The Krapina Hominids: An Illustrated Catalog of Skeletal Collections. Zagreb, Yugoslavia: Mladost. Risnes S (1986) Enamel apposition rate and the prism periodicity in human teeth. Scand. J. Dent. Res. 94t394-404. Scott EC (1982) Replica production for scanning electron microscopy: A test of materials suitable for use in field settings. J. Microsc. 225:337-341. Shipman P (1983) Early hominid lifestyle: hunting and gathering or foraging and scavenging. In J CluttonBrock and C Grigson (eds.):Animals and Archaeology: Hunters and Their Prey. Oxford: British Archaeology Rep. No. 163, pp. 31-49. 188 A.E. MA" ET AL. Shipman P (1986) Scavenging or hunting in early hominids: Theoretical framework and tests. Am. Anthropol. 88:27-43, Warshawsky H (1989) Are linear markings on dental enamel valid indicators of time? Paper presented at the annual meetings, Canadian Association for Physical Anthropology, Vancouver, B.C. Warshawsky H, and Bai P (1983) Knife chatter during thin sectioning of rat incisor enamel can cause periodicities resembling cross-striations. Anat. Rec. 207t533-538. Weber DF, and Ashrafi SH (1979) Structure of Retzius lines in partially demineralised enamel. Anat. Rec. 194:563-570. Weber DF, Eisenmann D, and Glick PL (1974)Light and electron microscope studies of Retzius lines in human cervical enamel. Am. J. Anat. 141:91-104. Weiss ML (1987) Nucleic acid evidence bearing on hominoid relationships. Yrbk. Phys. Anthropol. 30:41-73. Weiss ML, and Mann AE 11990) Human Biology and Behavior, 5th ed. Chicago: Scott, Foresmafiittle, Brown.