Age of Death from Microanatomy

Microanatomical studies of enamel and dentine are beginning to yield real estimates of the age of attainment of marker events in the life history of Homo erectus and other hominin species (Bromage and Dean, 1985; Dean et al., 1993b, 2001; Smith et al., 2007a). In the best cases, thin sections of teeth can be analyzed to determine age of death, as well as the timing of certain stressful life history events, with an astonishing accuracy. Schwartz et al. (2006; Schwartz and Dean, 2008) were able to date some life events to within a day by counting growth increments preserved in dentine and enamel of immature teeth of a juvenile captive gorilla. Accuracy on such a scale demands that material is well preserved, teeth can be thin sectioned, and juveniles are young (older juveniles may require cross-matching across two or more teeth). To date, studies with careful methods have produced estimates within 2-5% of the known age at death in modern human samples (Antoine, 2001; Antoine et al., 2000, 2009). Thus, when well preserved teeth can be sectioned, we may hope to achieve comparable accuracy for fossils. Nariokotome, however, remains too complete and too precious to section. Thus, although maximum precision is out of our reach at present, microanatomy can still be used to provide a good estimate of his age of death.

The basis for estimating the age at death of an individual from histological sections of teeth is the record of daily incremental markings that exists in enamel. Other coarser increments also exist several days growth apart called striae of Retzius. A notable feature of stria of Retzius, however, is that each emerges at the tooth surface in the form of a perikyma (plural perikymata; Gk. = waves around a tooth) creating an alternating pattern of troughs and grooves over the surface of the crown (Hillson, 1996; Hillson and Bond, 1997). Perikymata, often visible to the naked eye, can be counted on the tooth surface. Within all the teeth of an individual, the striae of Retzius and periky-mata are set the same number of days apart (Smith, 2008). As first shown by Bromage and Dean (1985) and then subsequently by others (Beynon and Wood, 1987; Dean et al., 1993a; Moggi-Cecchi et al., 1998), perikymata visible on tooth surfaces of fossil hominins can be used to count time elapsed in tooth development without the need to make a histological section.

The catch is that the number of daily increments between adjacent perikymata has a range across individuals known as the periodicity. In modern humans this is between 6 and 12 days but more usually 7, 8, or 9 days with a mean and mode of 8 days (Dean, 1987b; FitzGerald, 1998; Smith et al., 2007b). While sectioning is required to determine the periodicity of each individual's striae of Retzius or perikymata, the range of expected values is relatively small in closely related groups. In many hominins, striae of Retzius are formed each week: Lacruz et al. (2008) reported that 59% of 29 australopiths examined showed a mean and mode periodicity of 7 days. To date, only seven specimens of early Homo fossils have been examined microscopically; two of these had a periodicity of 7 days, four of 8 days, and one of 9 days (Lacruz et al., 2008).

On the face of it, a modal value of 8 days seems the most likely choice for KNM-WT 15000, but a higher periodicity is probable for several reasons. First, the total perikymata counts on the surfaces of each of the Nariokotome anterior teeth is comparatively low (Dean and Reid, 2001a, b). Reid and colleagues have shown that teeth with widely spaced perikymata that are few in number tend to have high periodicities (Reid and Ferrell, 2006; Reid et al., 2008), whereas those with many more tightly spaced perikymata tend to have low periodicities. Secondly, an 8-day periodicity leads to unrealistic estimates of root growth, exceeding rates seen in Gorilla and Pongo (see Dean and Vesey, 2008 and below). Indeed, a periodicity of 10 days for KNM-WT 15000 gives the most parsimonious estimates for both molar and anterior crown formation times (see below). In the following analysis, therefore, periodicities of both 8 and 10 days have been used to provide what are more likely to be reasonable upper and lower estimates of the age at death of Nariokotome.

Because Nariokotome is an older juvenile, no single tooth crown records time over his entire life. It is therefore necessary to begin counts on an early forming tooth, then cross match to a tooth which overlapped, but continued forming later. We must also extend these counts beyond crown completion of the later forming tooth and make an estimate for the period of subsequent root growth up until the time of death (see below and Table 10.4).

The upper right canine tooth, unerupted in life and now isolated from the skull, provides an ideal starting point. In all primates, this tooth begins to mineralize within a few months of birth (we here used the estimate for initiation of 274 days given in Dean et al., 2001; and in Dean and Reid, 2001a, b). Deep within the enamel under the cusp of the tooth, a small thickness of enamel forms prior to that visible on the tooth surface (we here used the estimate of 266 days for this as calculated in Dean et al., 2001). We can count 100 periky-mata between the cusp tip and the cervix on this canine tooth,

Table 10.4 Estimating age of death of KNM-WT 15000 as the sum of five segments of time represented in microanatomy of the upper canine and second molar

Days

Table 10.4 Estimating age of death of KNM-WT 15000 as the sum of five segments of time represented in microanatomy of the upper canine and second molar

Days

Perikymata

8-day

10-day

Time segment

count

periodicity

periodicity

1. Birth to initiation

274

274

of UCa

2. Formation of

266

266

hidden UC

cusp enamela

3. Formation of

100

800

1,000

visible UC crown

to second

hypoplastic line

4. M2 formed from

25

200

250

second hypoplastic

line to crown

completion

5. Time to form

1,241-1,424

1,241-1,424

9.3 mm of M2 rootab

Total days

2,781-2,964

3,031-3,214

Total years

7.6-8.1 years

8.3-8.8 years

a Estimated from comparative studies of humans and apes (Dean and Reid, 2001a).

b Range based on time to form the first 10 mm of M2 roots form in great apes (6.5 |im/day) and in humans (7.6 |im/day) (Dean and Vesey, 2008).

a Estimated from comparative studies of humans and apes (Dean and Reid, 2001a).

b Range based on time to form the first 10 mm of M2 roots form in great apes (6.5 |im/day) and in humans (7.6 |im/day) (Dean and Vesey, 2008).

which with periodicities of either 8 or 10 days equals 800 or 1,000 days. The sum of these intervals equals the age at canine crown completion and Table 10.4 shows this would have been 1,340 days (3.67 years) with an 8 day periodicity or 1,540 days (4.22 years) with a 10 day periodicity.

Several of the teeth of KNM-WT 15000 show evidence of two linear hypoplastic bands, or accentuated lines, across their crowns or roots (Fig. 10.4). These represent periods of slowed enamel or dentine formation that correspond with an illness or physiological upset that interrupted tooth growth (Hillson and Bond, 1997; Ritzman et al., 2008). In the upper right canine, the last of these bands coincides exactly with the end of enamel formation at the cervix (both the buccal and lingual), and the first occurred a mere 15 perikymata earlier. Both linear hypoplastic bands extend from enamel onto root dentine on the mesial (interproximal) surface (Fig. 10.4). In addition, they are preserved on the roots of most of the lower incisors and on the lateral enamel of some of the premolar and second permanent molar crowns that were growing at the same time (Figs. 10.5 and 10.6). The position of these bands on each of these teeth clinches the fact that they were caused by two consecutive interruptions to tooth formation (Dean et al., 1993a).

Using SEM, 15 perikymata can be counted between the start of the first band and the end of the second on the right upper canine crown as well as the left P4 and right M2 of KNM WT-15000 (Figs. 10.4 and 10.6). The severity of this kind of disturbance is expressed differently on each tooth type and may be influenced by the rate of enamel secretion at the time (Hillson and Bond, 1997). The fact that at least three crowns (the upper right canine, left P4 and right M2) all show these bands to be 15 perikymata apart is again a clear indication that they record the same two growth disturbances. Thus, by cross-matching the twin disturbances, we can continue the perikymata counts from the canine onto the right M2 crown. Here we see an additional 25 perikymata beyond the second linear hypoplastic band on both the P4 and M2 to the end of enamel formation, which puts M2 and P4 crown completion at 4.2 years (8 day periodicity) or as much as 4.9 years (10 day periodicity) (see Table 10.4).

To estimate an age at death for KNM WT-15000 it becomes necessary to calculate the root formation time for one of the developing teeth. Several teeth in KNM-WT 15000 have roots that were still growing at the time of death. Of these, the isolated upper right canine root and the exposed distal aspect of the right M2 root are most accessible, but the M2 root is shorter (9-10 mm depending on where it is measured) and also represents a smaller interval of time. For these reasons alone it makes sense to use the M2 root to estimate the time beyond enamel completion up to death in this specimen.

There are now a number of approaches to estimating rates of root growth and the time taken to grow roots in humans and great apes (Dean, 1995; Smith et al., 2007a, c). As with many fossil hominin teeth, periradicular ("around the root") bands are clearly visible on some tooth roots of KNM WT-15000. More than 80 can be counted on the distal aspect of the palatal root of the right M2 for example. In places, as on other roots of KNM-WT 15000, the bands are spaced roughly 5-6 per millimetre. However, there are regions where finer bands (as many as 15 per millimetre) can be counted (Fig. 10.7). This raises problems about defining exactly which periradicular root bands are equivalent to perikymata on the enamel. Some

Fig. 10.4 a) A low power SEM image shows two hypo-plastic bands (arrows) on the isolated and unerupted upper right permanent canine of KNM-WT 15000. The first band formed runs over the cervical enamel and interproximal root dentine; the second also runs over the interproxi-mal root dentine, coinciding with the last formed increment of enamel at the cervix. b) A higher power SEM image of the cervical enamel of the same tooth in the mid buccal cervical region. Two white arrows indicate the position of the hypoplastic bands, which have been traced across from the low power image. Perikymata are just visible across the tooth surface.

Fig. 10.4 a) A low power SEM image shows two hypo-plastic bands (arrows) on the isolated and unerupted upper right permanent canine of KNM-WT 15000. The first band formed runs over the cervical enamel and interproximal root dentine; the second also runs over the interproxi-mal root dentine, coinciding with the last formed increment of enamel at the cervix. b) A higher power SEM image of the cervical enamel of the same tooth in the mid buccal cervical region. Two white arrows indicate the position of the hypoplastic bands, which have been traced across from the low power image. Perikymata are just visible across the tooth surface.

Fig. 10.5 Low power SEM images showing the twin hypoplastic bands identifiable on four teeth of KNM-WT 15000 (arrows). From left to right: interproximal root dentine of the permanent upper right canine, the lower right lateral incisor, lower right central incisor and lower left lateral incisor. These bands were not visible on the lower left central incisor, which is poorly preserved. Expoy resin casts, sputter coated with gold, were prepared for SEM from Coltene moulds made from the original specimen.

Fig. 10.6 Recurrence of twin hypoplastic bands (arrows) on a) the distopalatal aspect of the upper left P4 and b) the upper right M2. Both teeth were oriented with occlusal surfaces towards the upper right of the images. Epoxy resin casts of both teeth were made from Coltene moulds of the original teeth and sputter coated with gold before being photographed under a binocular microscope in oblique incident light.

Fig. 10.6 Recurrence of twin hypoplastic bands (arrows) on a) the distopalatal aspect of the upper left P4 and b) the upper right M2. Both teeth were oriented with occlusal surfaces towards the upper right of the images. Epoxy resin casts of both teeth were made from Coltene moulds of the original teeth and sputter coated with gold before being photographed under a binocular microscope in oblique incident light.

of the best-preserved periradicular bands in KNM WT-15000 occur at the cervix of the left I2 where ten clear widely spaced bands (each ~100 ||m wide) exist between the enamel cervix and the first linear hypoplastic band 1,100 |im beyond this (Fig. 10.8).

Immediately beneath this, between the two hypoplastic bands on the left I2 root, one may also count ten similarly-spaced periradicular bands (Fig. 10.8). But the expectation here is 15, based on 15 perikymata counted between the equivalent hypoplastic bands on the enamel surfaces of the canine, P4 and M2. One can then make two estimates for the rate of root growth in the I2 root, one based on a periodicity of

10 days between periradicular bands (10.0 |im/day) and a second based on 15 perikymata being equivalent to the time between the twin hypoplastic root bands (7.5 |im/day). Thus, while it is tempting to use periradicular bands in the same way as perikymata to estimate M2 root formation time in KNM-WT 15000, there are clearly problems in defining which bands to count as well as in confirming the periodicity of the bands.

Another approach to estimating duration of root formation is to make use of root extension rates determined from histology. We know from Dean and Vesey (2008) that the first 10 mm of M2 root forms at a rate of 7.5 |im/day in Pan and at 6.5 ||m/day in H. sapiens. These very similar rates in close

Fig. 10.7 Periradicular bands on the distopalatal root of the upper right M2 visible on an epoxy resin cast sputter coated with gold. The enamel cervix is visible in the lower part of the image and 9 mm of root extends upward (an apical 1 mm not shown). White arrows indicate two of the regions where periradicular bands are either very widely spaced (upper arrow) or very narrowly spaced (lower arrow). In excess of 80 bands (perhaps 100) can be estimated over most of the root length but these are indistinct and uncountable at both the root cervix and root apex.

Fig. 10.7 Periradicular bands on the distopalatal root of the upper right M2 visible on an epoxy resin cast sputter coated with gold. The enamel cervix is visible in the lower part of the image and 9 mm of root extends upward (an apical 1 mm not shown). White arrows indicate two of the regions where periradicular bands are either very widely spaced (upper arrow) or very narrowly spaced (lower arrow). In excess of 80 bands (perhaps 100) can be estimated over most of the root length but these are indistinct and uncountable at both the root cervix and root apex.

relatives of H. erectus are reasonable to use to estimate the duration of formation of the M2 root in KNM-WT 15000. Here we can measure 9.3 mm of root between the last formed enamel and the incomplete apex on the distobuccal aspect of the M2. At 6.5 |m/day, 9,300 |m of root would form in 1,430 days (3.9 years) and at 7.6 |im/day, would form in 1,240 days (3.4 years). Adding together these estimates for root formation time with those made to the end of enamel completion on the M2 provides the widest range of estimates for the age at death of KNM-WT 15000. As Table 10.4 shows, these extend from 7.6-8.8 years. In other words, using an 8 day periodicity

Fig. 10.8 Above, an SEM micrograph and below, an epoxy cast of the enamel cervix and cervical portion of the root of the lower left lateral incisor. In the SEM micrograph the root is tilted to the left of the image to cast a shadow of the widely spaced periradicular bands. The expoxy resin cast, sputter coated with gold, is tilted to the right and illuminated with oblique incident light. The upper white arrow indicates the enamel cervix. The middle and lower white arrows indicate the twin hypoplastic root bands approximately 1 mm apart. Ten periradicular bands can be counted between the cervix and the first formed band and a further ten bands between the twin hypoplastic bands.

Fig. 10.8 Above, an SEM micrograph and below, an epoxy cast of the enamel cervix and cervical portion of the root of the lower left lateral incisor. In the SEM micrograph the root is tilted to the left of the image to cast a shadow of the widely spaced periradicular bands. The expoxy resin cast, sputter coated with gold, is tilted to the right and illuminated with oblique incident light. The upper white arrow indicates the enamel cervix. The middle and lower white arrows indicate the twin hypoplastic root bands approximately 1 mm apart. Ten periradicular bands can be counted between the cervix and the first formed band and a further ten bands between the twin hypoplastic bands.

(for the enamel portion of the estimate) gives estimates that center around 7.9 years of age, whereas using a 10 day periodicity gives estimates that center around 8.5 years.

Was this article helpful?

0 0

Post a comment