The purpose of this edited volume was to bring together specialists from various fields to present all the information needed to understand and implement cementochronology, the analysis of cementum growth. This interdisciplinary “Cementum Research Program” was initiated in 2010/2011 at a Paris workshop with Jean-Pierre Bocquet-Appel, Joël Blondiaux, Thomas Colard, and me. Initially, the goal was to develop a standardized cementum protocol for age estimation in anthropology (Reference Colard, Bertrand, Naji, Delannoy and BécartColard et al. 2015). However, the program’s scope expanded rapidly into an first poster symposium held at the 2012 annual meeting of the American Association of Physical Anthropology (AAPA) in Knoxville (USA). With the invaluable support of Jane Buikstra, selected presentations were published in a special issue of the International Journal of Paleopathology (Reference Naji, Colard, Blondiaux, Bertrand, d’Incau and Bocquet-AppelNaji et al. 2016).
This event’s positive feedback prompted our group to reach out even more broadly to the international community. Our purpose was to connect with zooarchaeologists who were implementing cementum analyses routinely and paleoanthropologists interested in evolutionary processes using dental remains. In collaboration with William Rendu and Lionel Gourichon, the second phase of our research program started in 2015 (Reference Naji, Rendu, Gourichon, Balasse, Brugal, Dauphin, Geigl and OberlinNaji et al. 2015) funded by a sizable French grant (ANR CemeNTAA) dedicated to cementum analysis. The results were presented as a second poster symposium at the 2017 annual meeting of the AAPA in New Orleans (USA). This event was an ideal opportunity to meet new colleagues and motivated students. Consequently, prompted by many colleagues’ support, in particular Daniel Antoine, we decided to publish our collaborative efforts in this edited volume.
This publication’s premise is to reply to one of the first questions anthropologists ask when considering dental cementum as an age indicator: Why should cementum growth follow an annual/seasonal deposition pattern in distinct layers on the roots’ surface?
This chapter is a short introduction to the field of chronobiology to present current theories explaining and interpreting cyclic growth marks observed in skeletal and dental tissues, specifically in cementum. The references are by no means comprehensive, but the reader should find adequate primary sources to explore the topic in greater detail in (Reference Dunlap, Loros and DeCourseyDunlap et al. 2003; Reference FosterFoster 2005; Reference KumarKumar 2017; Reference LemmerLemmer 2009; Reference LincolnLincoln 2019).
Cementochronogly Nomenclature
First, I will briefly discuss the various labels used to name cementum growth deposits to ensure that all scientists, regardless of their field, are talking about the same histological structures. For a detailed history of cementum discovery in biology, see Reference FosterFoster (2017) and Buikstra (Chapter 1). According to the following definitions, we have tried to homogenize the nomenclature relative to cementum growth analysis throughout this volume.
As early as 1887, Black summarized contemporary cementum knowledge, describing its structure as individual “lamellae, layers, or strata” divided by distinct lines called “incremental lines of the cementum” and observing its chronological deposition pattern as “each lamellae being the results of a single period of activity […] each successive lamella is younger than the preceding one” (Reference BlackBlack 1887,105–6). Sixty years later, these incremental structures started to be tested for aging animals and humans under several different names describing the contrasting optical layers visible in histological sections.
For animals, Reference SchefferScheffer (1950) and Laws (Reference Laws1952) defined external “growth ridges” and internal “annual growth zones or rings,” respectively, for dentin growth in sea mammals, but not cementum. Reference Sergeant and PimlottSergeant and Pimlott (1959) were the first to investigate the principle of growth layers for age estimation in cementum using moose as their study sample. They referred to “cement growth layers” to characterize the annual histological structure composed of the two seasonal “growth zones,” one opaque and one translucent. Klevezal’ and Kleinberg, in their first seminal review (Reference Klevezal’ and KleinenbergKlevezal’ & Kleinenberg 1967, 67), referenced “annual layers” composed of “bands” and “stripes.”
In 1978, the conference on odontocete age estimation (Reference Perrin and MyrickPerrin et al. 1980) proposed standardizing dental growth markers’ terminology. Every layer parallel to a tissues’ formative surface, contrasting with the adjacent one, was defined as an “incremental growth layer.” A repeatable pattern of growth layers counting as a time unit was then termed a “growth layer group (GLG).”
In terrestrial mammals, however, cementum growth layers have a slightly different interpretation: The “growth zones” (Reference Baglinière, Castanet, Conand and MeunierBaglinière et al. 1992) represent layers interpreted as a rapid deposition due to increased metabolic activity during the “favorable” season (Reference Demars, Le Gall, Martin and BeauneDemars, Le Gall & Martin 2007, 109). Conversely, annuli (singular, annulus) are slow growth layers formed during a decrease of osteogenic activity. Annuli are thus thinner than zones. Also, in French terminology, “growth rest lines” (ligne d’arrêt de croissance) (Reference CastanetCastanet 1980) were defined as a very thin structure, highly birefringent and often hypermineralized, that can be found within an annulus or alone and alternating with rapid growth zones (Reference Baglinière, Castanet, Conand and MeunierBaglinière et al. 1992, 444).
Finally, in her second seminal review, Klevezal’ (Reference Klevezal’1996) proposed a new definition for the cementum growth unit as a “growth layer of the first order” composed of an incremental cementum line (principal element counted for age estimation) and a cementum band (intermediate element).
In humans, the first use of cementum for age estimation specifically was published by Gustafson in 1950 with a nonspecific “cementum apposition” component, representing cementum width, not incremental counts, in his multicriteria dental method (Reference GustafsonGustafson 1950). Three decades later, the pivotal publication of Reference Stott, Sis and LevyStott and colleagues (1982) used the incremental count of cementum deposits for the first time and described them as “cemental annulations.” Since then, cementum annuli (Reference Stein and CorcoranStein & Corcoran 1994) was used preferentially.
In the influential first large-scale controlled study, Reference Wittwer-Backofen, Buba, Hoppa and VaupelWittwer-Backofen and Buba (2002) labeled the use of cementum as the teeth cementum annulations (TCA) method. The acronym seems to be the most preferred among anthropologists today. However, this name can create confusion with the mammal’s slow growth layer’s “annulus.” In forensic anthropology, Wedel proposed another descriptor, dental cementum increment analysis (DCIA) (Reference WedelWedel 2007), which is entirely accurate.
Finally, following the 1992 publication of the symposium on vertebrate age estimation using hard tissues (Reference Baglinière, Castanet, Conand and MeunierBaglinière et al. 1992), another term, the French “cémentochronologie” was proposed to follow the larger context of chronobiology (Reference GrosskopfGrosskopf 1996; Reference MartinMartin 1995). In the English literature, “cementochronology” was introduced during the 82nd annual conference of the American Association of Physical Anthropology in 2013 in a contributed poster symposium titled “Cementochronology,” organized by Naji, Colard, and Bertrand (Reference Naji, Colard and BertrandNaji et al. 2013). The purpose of this nomenclature shift was to reflect the broader multidisciplinary approach to a common biological growth process and move away from potentially confusing descriptors.
Chronobiology, the Cycles of Life
The periodicity of growth processes has a deep history, from the ancient Greeks’ understanding of daily leaf movement to the first published demonstration of endogenous plants’ periodicity in 1832 (cited in Reference Schwartz, Daan and KumarSchwartz & Daan 2017). Formally, the field of chronobiology, the study of biological rhythms, can be traced to the 1960 edition of the Cold Spring Harbor Biological Laboratory’s annual symposium in Long Island, New York, titled “Biological Clocks” (Reference LemmerLemmer 2009).
Chronobiology rests on the premise that the regular rotation of the earth around its central axis and around the sun produces two fundamental periodicities to which all life, from unicellular organisms to primates, has become adapted. The hypothesis is that circadian clocks govern daily rhythmicity, and circannual clocks provide a seasonal endogenous calendar (Reference LincolnLincoln 2019). The various external stimuli (e.g., light cycle, food availability, temperature variation) provide a template for living organisms to anticipate cyclic environmental events by periodic and predictable internal adjustments in physiology and behavior, even where standard environmental cues are weak or ambiguous (Reference Piccione, Giannetto, Casella and CaolaPiccione et al. 2009). The frequencies of these rhythms have evolved to cover nearly every division of time (Reference LemmerLemmer 2009), from intradian – less than a day – oscillations of one per second (e.g., brain waves), or one per several seconds (e.g., heart rate); circadian, one within twenty-four hours (e.g., enamel cross-striations, dentine’s von Ebner’s lines); multidien, five-day rhythm (e.g., enamel of domestic pig, Reference Bromage, Idaghdour, Lacruz and SchrenkBromage et al. 2016); circaseptan, near-seven-day periodicity (e.g., heart rate and pressure, Reference Reinberg, Dejardin, Smolensky and TouitouReinberg et al. 2017); lunar, once a month (e.g., ovulation); to circannual, one per year (e.g., reproduction, molt, migration, and cementum).
Today, chronobiology is incorporated into practically all fields of human and nonhuman endeavors, including ecology, biology, sociology, and psychology (Reference Reinberg, Dejardin, Smolensky and TouitouReinberg et al. 2017) to optimize sleep, diet, immune system response, or performances, among other factors.
Mechanisms Responsible for Annual Cycles
We will focus here on the mechanisms and consequences of the annual cycles (For a review of circadian cycles, see Reference Panda, Hogenesch and KayPanda et al. 2002; Reference Weinert, Waterhouse and KumarWeinert & Waterhouse 2017). Annual rhythms can be classified into three types (Reference LincolnLincoln 2019; Reference Zucker, Lee, Dark, Klein, Reppert and MooreZucker et al. 1991).
Type 1 trans-generational annual rhythms are mostly observed in short-lived species with multiple generations throughout the year. The innate circannual timing mechanism passes from individual to offspring across the year and may be expressed at only one phase of the life cycle (Reference LincolnLincoln 2019). The presence of annual environmental cues such as annual variations in temperature and photoperiod influences the annual neuroendocrine rhythms via hormonal maturation, for example, puberty and reproduction (Reference Zucker, Lee, Dark, Klein, Reppert and MooreZucker et al. 1991). This is a population/cohort response because an individual cyst only hatches once. A variant is a modular rhythm for longer-lived insects that transform from the egg, through larval instars, to pupation and hatching of the sexually mature adult (Reference LincolnLincoln 2019). Again, this rhythm is a population/cohort event because each animal pupates only once.
Type 2 seasonal (circannual) progressive rhythms recur under constant conditions with a period between ten and twelve months, more typical of long-lived species, including primates. The progressive development from juvenile to adulthood to old age and circannual timing generates cycles in multiple aspects of physiology and behavior (e.g., gonad size, body weight, food intake, gut morphology, immune function, molt, thermoregulation, hibernation, and migration) (Reference LincolnLincoln 2019). Type 2 persists in the absence of periodic light input (Reference Zucker, Lee, Dark, Klein, Reppert and MooreZucker et al. 1991) but varies among and within individuals (Reference Piccione, Giannetto, Casella and CaolaPiccione et al. 2009). Animals have an individual annual chronotype (e.g., early rutting/late rutting). This circannual chronotype has been observed in humans with a seasonal affective disorder that could be interpreted as a natural adaptation to winter, where a change in appetite and increased body weight in autumn and the development of withdrawal behaviors in winter was once an advantage for our hunter‐gatherer ancestors (Reference LincolnLincoln 2019). Also, mRNA expression levels indicated in one study that 23 percent of the genome showed significant seasonal differences with two distinct antiphase patterns: One set of genes up‐regulated in summer and the other, approximately equal, up‐regulated in winter (Reference Dopico, Evangelou, Ferreira and ToddDopico et al. 2015).
Type 3 annual rhythms are found in animals living in unpredictable environments (desert/equator) where the cue is rainfall and plant growth. In other words, type I rhythms are evoked by environmental cycles, while types II and III are synchronized to environmental cycles (Reference Piccione, Giannetto, Casella and CaolaPiccione et al. 2009).
The current working hypothesis to explain circannual cycles assumes that this timing first evolved in free-living eukaryote cells as an adaptation to survive the winter (Reference LincolnLincoln 2019). These organisms alternate between seasonal growth and dormancy across their life-history, which requires a genetically regulated, cell-autonomous, and transgenerational mechanism. Switching between growth in the circannual “summer” and dormancy in the circannual “winter” is a highly adaptive strategy that has thus been conserved in our evolution and is observable today in five of the eight eukaryotic kingdoms (Reference Helm, Lincoln and KumarHelm & Lincoln 2017).
The Clock‐Shop Model: Combining plant models with animal models provides a general theoretical mechanism linking circadian (daily) modification to circannual rhythm timescale (Reference Schwartz, Daan and KumarSchwartz & Daan 2017). Knowledge of the physiological pathways governing time measurement mechanisms has been unfolding rapidly in the past decades through comparative studies that have uncovered the photoperiodic signal transduction cascades in birds, fish, and mammals (Reference FosterFoster 2005). These studies revealed the universality and diversity of photoperiodic mechanisms, such as the fact that molecules involved are conserved while the tissues responsible for these mechanisms are species-specific (Reference 15Ikegami, Yoshimura and KumarIkegami & Yoshimura 2017). The highly adaptive strategy of annual growth, irrespective of an organism’s size and longevity, has determined the evolution of innate, genetically regulated timing processes, broadly encompassed by the clock‐shop model (Reference LincolnLincoln 2019). In mammals, in particular, we now understand that the cyclic rhythms are controlled by the body’s “central clock” located in the brain at the suprachiasmatic nuclei and supported by “peripheral clocks” located in several other tissues (Reference Liu, Panda and KumarLiu & Panda 2017; Reference Zheng, Ehardt, McAlpin and PapagerakisZheng et al. 2014).
The clock‐shop model proposes that environmental signals, notably photoperiod, are relayed by the sensory systems to the central pacemakers to synchronize physiology with the seasons using, among other mechanisms, melatonin‐responsive thyrotropic cells in the pars tuberalis of the mammalian pituitary gland (Reference Ganguly, Klein and KumarGanguly & Klein 2017). The conjecture is that long‐term timing mechanisms reside in all tissues but with dominant pacemaker systems in the brain and pituitary gland orchestrating the circannual phenotype (Reference LincolnLincoln 2019). Circadian and circannual timing systems thus share formal properties: ancestry, cell autonomy, innateness, entrainment, temperature compensation, and ubiquity (Reference LincolnLincoln 2019).
At a molecular level, the favored model for circannual timekeeping proposes that the long‐time domain is generated by the cyclical epigenetic regulation of chromatin structure (DNA and histone proteins), determining whether specific circannual timer genes are transcriptionally active or not. This regulation drives the oscillation between the two stable, operational states of subjective summer and subjective winter (Reference LincolnLincoln 2019; Reference Stevenson, Lincoln and KumarStevenson & Lincoln 2017).
The consequence is that circannual timing mechanisms are more flexible and reprogrammable in the long term, which means if a consistent seasonal change is observed in any organism, it is most likely to be regulated by an endogenous timing mechanism rather than by a passive response to the environment.
Several cyclic mineralization pathways have been identified for bones, suggesting that bone deposition and mineralization are under direct circadian controls (Reference Zheng, Ehardt, McAlpin and PapagerakisZheng et al. 2014). For example, the diurnal variation in the synthesis of type I collagen and osteocalcin is under a local circadian oscillator mechanism. Also, analyses in sheep and humans suggest that a single biological rhythm governs all lamellar bone formation within a given taxon (Reference Bromage, Lacruz, Hogg and BoydeBromage et al. 2009).
For teeth, there are daily variations in the rate of production and secretion of enamel proteins between early morning and late afternoon, suggesting that enamel protein secretion is under circadian control and that enamel matrix production and maturation are closely controlled by selectively regulating key enamel matrix proteins encoding genes (Reference Zheng, Ehardt, McAlpin and PapagerakisZheng et al. 2014). Like enamel, dentin is formed incrementally, indicating the involvement of a circadian clock mechanism during dentinogenesis. This cyclic growth has been demonstrated using proline tracers that labeled collagen during dentin formation and showed that twice as much collagen is secreted during the daylight twelve hours as during the nocturnal twelve hours. These studies suggest that dentin, similar to bone and enamel, is controlled by a circadian clock mechanism (Reference Zheng, Ehardt, McAlpin and PapagerakisZheng et al. 2014).
Unlike the other dental hard tissues, cementum does not seem to be controlled by circadian mechanisms but more likely by circannual ones. Reference Stock, Finney, Telser, Maxey, Vogt and OkasinskiStock and colleagues (2017) first identified second-order lines in Beluga whales between first-order (annual) increments in contrasting Ca and Zn variations. Reference Dean, Le Cabec, Spiers, Zhang and GarrevoetDean and colleagues (2018) have also identified an average of twelve second-order increments (monthly), interpreted as menstrual within chimpanzee samples. Recent work on cementum presented in this volume (Chapter 1, Chapter 6, Chapter 7, Chapter 14) is starting to document specific pathways (e.g., pyrophosphate regulation, vitamin D absorption, hormonal variations) that might be involved in mammals’ annual cycles, specifically.
Linking circadian mechanisms to annual ones across mammalian species of various body mass is a complex question. Lincoln argues that circannual rhythms are independent and cannot be explained by frequency demultiplication of circadian rhythms, although changes in the circadian system occur in parallel with the circannual cycle and are the basis of photoperiod entrainment of the circannual clock (Reference LincolnLincoln 2019, 4).
However, Reference Bromage, Idaghdour, Lacruz and SchrenkBromage and colleagues (2016) have proposed a hypothesis that “a periodic rhythm longer than the daily biological clock regulates some aspects of metabolic variability that contribute to variability in body size and the pace and pattern of life” (Reference Bromage, Idaghdour, Lacruz and SchrenkBromage 2016, 19). Their metabolome and genome analyses from blood plasma in thirty-three domestic pigs revealed that blood plasma metabolites and small noncoding RNA (sncRNA) strongly oscillate on a five-day multidien rhythm, as does the pig enamel.
The adaptive benefits for circannual timekeeping, especially for larger vertebrates, include two critical features. The first is its predictive power to anticipate and prepare for upcoming seasonal changes in the environment (Reference Ball, Alward, Balthazart and KumarBall et al. 2017). In a highly seasonal habitat, where changes in food supply and other selective pressures can be predicted through photoperiod, the timing mechanism allows for precise regulation of the timing of cycles in physiology and behavior of fundamental significance in evolution (Reference Helm, Lincoln and KumarHelm & Lincoln 2017). The second is organisms’ ability to express robust annual cycles, which at specific phases override the effects of proximate cues, including photoperiods, such as in cross-equatorial migratory birds or hibernating species (Reference Helm, Lincoln and KumarHelm & Lincoln 2017).
Circannual Rhythms Validation Studies
Even though, in theory, circadian and circannual rhythms are relatively well understood from an evolutionary perspective, seasonal rhythms are expressed at the individual level. Therefore, these cyclic growth markers still need to be identified by repeated measurements on documented individuals to understand the mechanism’s geographical and temporal variability. Circannual rhythms can be observed in many biological or behavioral mechanisms (Reference Ball, Alward, Balthazart and KumarBall et al. 2017). In wildlife biology or anthropology, identifying a seasonal pattern in hard tissues to define a precise and accurate marker to estimate age and season at death or any life event represents a powerful tool to explore topics such as demography and mobility patterns robustly.
To demonstrate hard tissue seasonal growth rate in animals, including humans, three lines of evidence can be sought from documented subjects: (1) Long term capture–recaptures and sampling; (2) chemical labeling; and (3) empirical identification (Reference Baglinière, Castanet, Conand and MeunierBaglinière et al. 1992). We will see that all three have been successfully used in skeletal and dental tissues to support the hypothesis of circannual incremental deposits.
Bone and Other Hard Tissues
Sclerochronology is the method that describes elapsed time from recorded hard tissue (Reference Baglinière, Castanet, Conand and MeunierBaglinière et al. 1992). Otholometry studies fish’s otolith (inner ear bone), composed of accretion of calcium layers with two periodicities: a daily cycle influenced by water temperature and an annual period (Reference Baillon, Baglinière, Castanet, Conand and MeunierBaillon 1992; Reference KimuraKimura 1977; Reference Kimura and ChikuniKimura & Chikuni 1987). The annual periodicity of growth has been validated using both injection of fluorochrome calcein and empirical observations (Reference Baillon, Baglinière, Castanet, Conand and MeunierBaillon 1992; Reference Mounaix, Baglinière, Castanet, Conand and MeunierMounaix 1992). Similarly, fish scale growth scalimetry has been empirically observed to have an annual deposition rhythm in various species (Reference Mounaix, Baglinière, Castanet, Conand and MeunierMounaix 1992).
Skeletochronology is the use of periodic incremental growth structure in bones for age estimation studies. It has been applied in dinosaurs, reptiles, and mammals (Reference Castanet, Meunier and RicqlèsCastanet et al. 1977; Reference Woodward, Padian and AndrewWoodward et al. 2013) as early as the 1930s (Reference ClercClerc 1927). Long-term capture–recaptures have demonstrated the annual growth of one layer of new bone every year through successive amputations of the same limb in amphibians such as toads (Reference HemelaarHemelaar 1985). The annual growth pattern has also been tracked in vivo bone labeling of fluorescent marker calcein or tetracycline and recaptures in crocodiles (Reference de Buffrénil and Castanetde Buffrénil & Castanet 2000). For example, in the dermal scutes of Nile crocodiles, the laminae (the bone tissue between successive growth marks) are deposited by accretion/resorption phases so that the presence of one zone and one annulus marks the passage of one year (Reference Woodward, Padian and AndrewWoodward et al. 2013). Also related, cartilaginous shark bones present comparable circannual growth layers (Reference Baglinière, Castanet, Conand and MeunierBaglinière et al. 1992).
Mammals: The annual periodicity of bone growth has also been tested in mammals. An extensive study of the mouse lemur Microcebus assessed the number of annual growth layers in captive individuals of known age across the skeleton (Reference Castanet, Francillon-Vieillot, Meunier, Ricqlès and HallCastanet et al. 1993). More recently, circannual cycles have also been empirically observed in forty species of ungulates of varied size, diet, and habitat from the Equator to near the Poles environments (Reference Köhler, Marín-Moratalla, Jordana and AanesKöhler et al. 2012). Results showed that lamellar bone growth is arrested during the unfavorable season and accompanied by decreases in body temperature, metabolic rate, and bone-growth-mediating plasma insulin-like growth factor-1 levels. This “growth arrest” forms part of a plesiomorphic metabolic strategy for energy conservation (Reference Köhler, Marín-Moratalla, Jordana and AanesKöhler et al. 2012). Conversely, at the beginning of the favorable season, phases of intense tissue growth coincide with peak metabolic rates and correlated hormonal changes, indicating an increased efficiency in acquiring and using seasonal resources (Reference Köhler, Marín-Moratalla, Jordana and AanesKöhler et al. 2012). These results from tachymetabolic mammals show unequivocally that annual growth layers are a universal pattern of homoeothermic endotherms and should not be regarded as anything more than endogenous markers of annual rhythms (Reference Padian and LammPadian & Lamm 2013).
Human Studies: In humans, the use of bone remodeling patterns in age estimation is limited to growth stages of life because intracortical remodeling usually prevents any meaningful analysis in senescing individuals. In forensic contexts, however, the histological determination of adult age in cortical bone employs remodeling activity through the evaluation of osteon density, the number of primary vascular canals, the amount of unremodeled lamellar bone, the percent remodeled bone, and the average size of secondary osteons or Haversian canals (Reference Streeter, Crowder and StoutStreeter 2011). However, the correlation between observed age and estimated age in adults is poor at best since not all growth marks reflect environmental cues on growth. Some marks reflect temporary realignments of internal bone structure such as cortical drift; others may directly reflect environmental stresses (Reference Woodward, Padian and AndrewWoodward et al. 2013).
The timing of lamellar growth rates was measured using fluorescent labeling in the bones of rats, monkeys, sheep, and humans and revealed that the number of days needed to form one lamella is species-dependent: seven days for rats, twenty-eight days for macaques, thirty-five days for sheep, and fifty-six days for humans (Reference Bromage, Lacruz, Hogg and BoydeBromage et al. 2009). Further research based on the histological analyses of skeletal remains of twelve Bantu individuals of known sex and life-history discovered that incremental lamellar bone is deposited with long-period (five to six weeks) growth rate variability previously unobserved in humans (Reference Bromage, Juwayeyi, Smolyar, Hu, Gomez, Scaringi, Bondalapati, Kaur and ChisiBromage et al. 2011). Of greater interest to us, potential annual growth deposits have also been observed in some of the individuals (Reference Bromage, Juwayeyi, Smolyar, Hu, Gomez, Scaringi, Bondalapati, Kaur and ChisiBromage et al. 2011, 505).
Overall, there is ample evidence that skeletal growth markers that are periodic often occur annually (Reference Woodward, Padian and AndrewWoodward et al. 2013). It now also appears that in most vertebrates, including dinosaurs and mammals, cyclical growth markers resulting from the temporary cessation of growth simply reflect internal hormonal cues rather than direct environmental influence (Reference Woodward, Padian and AndrewWoodward et al. 2013).
Teeth
Odontochronology: With the advent of more powerful microscopes, dentin and enamel have also been explored to identify growth cycles (Reference Hogg, Croft, Su and SimpsonHogg 2018) and, ultimately, cementum during the fifties (Chapter 1).
Enamel and Dentine: Mammalian teeth exhibit microanatomical incremental features representing successive forming fronts of enamel and dentine at varying timescales (Reference Bromage, Lacruz, Hogg and BoydeBromage et al. 2009; Reference DeanDean 2006). The outcome is visible in light microscopy as a daily “cross-striation” and as a long period, “stria of Retzius,” measured as the number of cross-striations between adjacent striae and is thus reported in units of whole days. The number of daily increments between striae is identical for all teeth of an individual yet variable between and occasionally within a species that reflects a positive relationship with body size (Reference Padian and LammPadian & Lamm 2013).
Incremental dentin lines are termed von Ebner’s lines, which delineate the amount of mineral deposited in a single day. The circadian mineralizing lines in dentine are distinguished by their characteristic appearance, where small spheres of mineralizing dentine increase in size until they eventually coalesce (Reference DeanDean 2006). Using a biomarker on macaques, Reference BromageBromage (1991) confirmed the daily rate of dentine von Ebner’s lines formation. Later, Reference Dean and ScandrettDean and Scandrett (1996) also used biomarkers on humans to correlate dentin and enamel formation. However, enamel and dentine have not been linked to any circannual growth patterns in mammals but have been in some species of toothed fish, where the annual growth lines were demonstrated by tetracycline labeling (Reference Day and CarrelDay et al. 1986).
Cementum: The benchmark of experimental studies involves using fluorochrome dyes to record cementum growth at precise intervals. In bears, epi-fluorescent photomicrography was used to date a chemical biomarker’s exact position in cementum increments to the nearest year (Reference Matson and KerrMatson & Kerr 1998). Reference Bosshardt, Luder and SchroederBosshardt and colleagues (1989) used tetracycline labeling and fluorescence mapping of acellular and cellular cementum in one macaque to demonstrate that acellular cementum formation is a tightly controlled biological phenomenon that occurs with the same regularity and speed wherever this type of tissue is needed.
Empirically, the presence of an endogenous growth rhythm based on growth mark formation in vertebrate skeletal tissues has been proposed by many researchers in a diversity of environments (Reference Grue and JensenGrue & Jensen 1979; Reference Klevezal’ and KleinenbergKlevezal’ & Kleinenberg 1967). For example, hibernating mammals form an annual increment despite being prevented from hibernating (i.e., in the absence of typical environmental influences) (Reference Perrin and MyrickPerrin et al. 1980). Also, Reference GrueGrue (1976) noted that mink raised on farms where they were fed a steady diet reduced their food intake during winter months. Likewise, cyclical cementum incremental growth is visible in tropical vertebrates and cannot be ascribed to marked seasonal fluctuations (Reference Klevezal’Klevezal’ 1996, 93). Logically, the early hypotheses formulated to interpret these observations included the presence of a genetic component for increment formation (Reference Grue and JensenGrue & Jensen 1979).
Finally, age and seasonal data obtained through the analysis of incremental growth structures corroborate similar data obtained through various other morphometric methods (e.g., teeth attrition) in living specimens. In many cases, these provide more accurate and precise age estimates (Reference Matson, Van Daele, Goodwin, Aumiller, Reynolds and HristienkoMatson et al. 1993; Reference MillerMiller 1974; Reference Stallibrass, Wilson, Grigson and PayneStallibrass 1982).
Overall, the annual cementum growth cycle has been observed in control groups of known age and season of death of terrestrial (Reference Klevezal’ and KleinenbergKlevezal’ & Kleinenberg 1967; Reference LiebermanLieberman 1993) and marine mammals (Reference Klevezal’ and MyrickKlevezal’ & Myrick 1984; Reference Perrin and MyrickPerrin et al. 1980; Reference von Biela, Testa, Gill and Burnsvon Biela et al. 2008). Additionally, cementum has been found in fossils and extant animals outside of the mammalian phylogeny, including reptiles (Reference Enax, Fabritius, Rack, Prymak, Raabe and EppleEnax et al. 2013; Reference Luan, Walker, Dangaria and RieppelLuan et al. 2009), ichthyosaurs (Reference Maxwell, Caldwell and LamoureuxMaxwell et al. 2011), mosasaurs (Reference LeBlanc, Brink, Cullen and ReiszLeBlanc et al. 2017), dinosaurs (Reference García and ZurriaguzGarcía & Zurriaguz 2016), and toothed fossil birds (Reference Dumont, Tafforeau, Bertin and LouchartDumont et al. 2016). Similarly, in humans, more than thirty-five studies of known-age individuals have been repeatedly successful in demonstrating the strongest correlation between acellular cementum increment number and chronological age (for a full summary, see Chapter 1; Reference Naji, Colard, Blondiaux, Bertrand, d’Incau and Bocquet-AppelNaji et al. 2016; Reference Naji and Koel-AbtNaji & Koel-Abt 2017).
Biological Aging and the Tempo of Senescence
Today, biologists have a strong hypothesis for interpreting cyclic growth patterns. The evolutionary clock-shop model is gaining rapid empirical validation in chronobiology, even though tissue-specific molecular/physiological pathways are not fully understood yet, and intermediate mechanisms between circadian and circannual patterns may still need exploring (Reference Bromage, Idaghdour, Lacruz and SchrenkBromage et al. 2016).
For cementum, the annual/seasonal periodicity has been demonstrated repeatedly in documented collections by chemical labeling studies and the most extensive cross-species empirical validation tests probably ever produced for an age indicator (Chapter 1). Also, there is some evidence that the antiquity of the thecodont tooth attachment system that includes the alveolar bone, a periodontal ligament, and cementum has been hypothesized to be a plesiomorphic shared feature of all amniotes for the past 290 million years (Reference LeBlanc, Brink, Cullen and ReiszLeBlanc et al. 2017; Reference Newham, Gill, Brewer, Benton, Fernandez, Gostling and HaberthürNewham et al. 2020) and is thus probably under tight genetic control (Chapter 3, Chapter 4, Chapter 16).
Not knowing the precise molecular and physiological mechanisms for cementogenesis should not lead anthropologists to dismiss cementochronology entirely. This is not a failure of the method; this is simply the state of our current knowledge. Today, we cannot explain most of the adult standard anthropological aging methods we have been using for more than a century. Fundamental research and validation studies have allowed us to tease out some variables responsible for the skeletal and dental changes we observe and use to model biological age.
For example, in skeletal estimators based on the ilium’s auricular surface, the sternal rib end or the cranial sutures, the “age” component is the weakest to explain observed changes. Biomechanical wear and tear are the principal components of change for the first two, and we still do not fully understand the processes for the latter. Infections (e.g., septic arthritis), mineralization defects (e.g., osteoporosis), or genetic syndromes (e.g., craniosynostosis) will also have a compounding or leading influence in any degenerative age indicator. Of course, time is correlated to these changes because it creates a longer exposure to external stimuli.
In healthy individuals, mechanical loading is the primary driver. Theoretically, an anthropologist analyzing the remains of someone who has been immobilized for several decades (e.g., in a coma) using these standard age indicators should observe no or limited articular degeneration and conclude to a very young skeletal age, whereas we have ample evidence of young adult weight lifters with significant joint damage that an anthropologist could (mis)interpret as belonging to an elderly individual. Using these degenerative changes should thus be restricted to relative comparisons in similar environmental, social, or behavioral contexts and individual frailty. Therefore, this is not a promising area of research for precise and accurate age estimations. However, we are looking forward to recent developments of the transition analysis (www.statsmachine.net/software/TA3).
Conversely, growth mechanisms are part of a strong selective process. They are reliably used for age estimations in developing humans and animals, with indicators for the youngest age categories being the most accurate and precise since the exposure to environmental insults is the shortest. Of course, external (e.g., diet, pathogens, pollutants) variables can influence growth and maturation and, ultimately, adult height, but because we observe a reasonably stable species-wide mechanism, the principal component of growth is undoubtedly genetic, not environmental.
Cementum deposition is akin to, if not entirely, a growth process. As we will see in Part I, cementum is continuously produced throughout life to ensure proper dental attachment in the alveolar bone and constant occlusion. Moreover, cementum is avascular and is not remodeled by cementoblast activity once deposited, contrary to bone.
Chapters Presentation
In this volume, we argue that cementochronology should enjoy a privileged place among hard tissue age indicators, as it is the only one identified so far with such well-defined and permanent growth characteristics. To present our arguments, we have divided the chapters into three parts that cover (i) cementum biology, (ii) protocols, and (iii) applications.
In Part I, Buikstra contextualizes the discovery of cementum (Chapter 1) and our current understanding of cementum biology (Chapters 2–5). Chapters 6 and 7 present state-of-the-art analyses of cementum ultrastructure using synchrotron x-ray fluorescence and diffraction mapping. Today, synchrotron-level energy is the only available technique to analyze the composition of individual cementum increments. Finally, in Chapter 8, we summarize the current literature on one of the most exciting research avenues in cementochronology: identifying life-history events, specifically pregnancies.
Part II presents recent advances in cementum analysis protocols and starts with a proposal for a standardized method for ungulates and new options for human samples (Chapter 9). Chapters 10 and 11 complement this approach by testing specific steps to optimize published protocols. Wedel proposes expanding cementochronology to decidual teeth (Chapter 12), and Wilson expands season-at-death estimation to human remains (Chapter 13), specifically in a forensic context, a method pioneered by Wedel in 2007. Colard (Chapter 14) offers a robust yet straightforward multi-analyses protocol to identify cementum growth variations resulting from pregnancies and potentially other life-history events or stressors. This is a welcome contribution that reinforces other potential methods (see Chapter 8 for a review). The innovative breakthrough in cementum protocols comes with the use of 3D synchrotron micro-CT reconstruction of cementum. Two independent teams are leading the way. First, Le Cabec (Chapter 15) demonstrates the potential of nondestructive imaging in human samples. Then, Newham (Chapter 16) proposes for the first time a fully automated 3D analysis package tested on various species with unparalleled precision and accuracy to estimate the age-at-death noninvasively.
Finally, Part III offers a range of cementochronology applications. Using faunal remains to estimate season at death, Rendu (Chapter 17) discusses his work on Neandertal’s residential mobility patterns in southern France. In contrast, Jimenez (Chapter 18) presents original data on human and hyaena seasonal settlements in Belgium during the Late Pleistocene. Cormier (Chapter 19) offers a robust osteobiography analysis of a Middle Woodland woman with a combined skeletal dysplasia and illustrates how cementochronology is effectively used for individual age at death while severe pathologies compromise other skeletal components. Naji (Chapter 20) publishes for the first time the age-at-death distribution of a hunter-fisherman shell-midden sample in Brazil, while Lanteri (Chapter 21), using a similar methodology, tests the accuracy of cementochronology death distribution with archival data of a sixteenth- to eighteenth-century French cemetery. Finally, the last two chapters offer new options for demographic studies using cementochronology. Propst (Chapter 22) illustrates the effectiveness of cementochronology in a complex Nabataean sample from Petra, Jordan, using a Gompertz-Makeham hazard model to contrast age-specific mortality risk. Wittwer-Backofen (Chapter 23) finishes the range of application by showing how cementochronology can help shape an early medieval age-at-death distribution based on standard skeletal age indicators. In conclusion, Naji and Rendu (Chapter 24) summarize the advances made in cementum studies in the last decade or so and address some of the remaining challenges to moving forward.
We hope that this collection of chapters, written by most of the leading authors in cementochronology today, will provide the necessary context to evaluate this method and stimulate future collaborative research.
Acknowledgments
I would like to thank Jane Buikstra, Brian Foster, Adeline Le Cabec, and Aaron Stutz for their helpful comments on earlier drafts. Unless otherwise stated, all chapters have been peer-reviewed anonymously by two specialists. Supplemental online materials are available at www.cambridge.org/naji.
