Teeth,prenatal growth rates,and the evolution of human-like pregnancy in later Homo

Evidence of how gestational parameters evolved is essential to understanding this fundamental stage of human life. Until now, these data seemed elusive given the skeletal bias of the fossil record. We demonstrate that dentition provides a window into the life of neonates. Teeth begin to form in utero and are intimately associated with gestational development. We measured the molar dentition for 608 catarrhine primates and collected data on prenatal growth rate (PGR) and endocranial volume (ECV) for 19 primate genera from the literature. We found that PGR and ECV are highly correlated (R² = 0.93, P < 0.001). Additionally, we demonstrated that molar proportions are significantly correlated with PGR (P = 0.004) and log-transformed ECV (P = 0.001). From these correlations, we developed two methods for reconstructing PGR in the fossil record, one using ECV and one using molar proportions. Dental proportions reconstruct hominid ECV (R² = 0.81, P < 0.001), a result that can be extrapolated to PGR. As teeth dominate fossil assemblages, our findings greatly expand our ability to investigate life history in the fossil record. Fossil ECVs and dental measurements from 13 hominid species both support significantly increasing PGR throughout the terminal Miocene and Plio-Pleistocene, reflecting known evolutionary changes. Together with pelvic and endocranial morphology, reconstructed PGRs indicate the need for increasing maternal energetics during pregnancy over the last 6 million years, reaching a human-like PGR (i.e., more similar to humans than to other extant apes) and ECV in later Homo less than 1 million years ago.

Life history describes the schedule and process of growth and development, otherwise known as ontogeny (1, 2). Key milestones in mammalian life history begin early in development with conception and gestation, and proceed throughout the lifespan, ending in death. The earliest stages of primate life history are intricately balanced through the maternal–infant relationship (3). Developing human embryos, like those of other primates, are entirely reliant on their gestating parent throughout prenatal development; it is the only stage of primate life history where a gestating female cannot pass on or distribute any of the metabolic or physiological burden of growing and raising offspring to other members of the social group (3–5). However, a gestating parent can receive support through the community, including through a partner, whether or not they have contributed genetic material to the offspring. This support is almost always linked to resource provisioning and protection from predators (6–8).The growing body of research on primate maternal energetics, infant growth and development, and the evolution of cognition revolves around a complex network of resource provisioning, maternal health, trophic status, locomotion, body size, infant dependency (sometimes referred to as altriciality), and social network (5, 9–17). The importance of each of these factors in the evolution of the human species is still very much under debate, largely complicated by the difficulty of assessing many of these physiological and behavioral traits in the fossil record. While a large body of research has focused on postnatal growth rates and duration in humans and other hominids (18–20), almost no research has investigated prenatal growth rates (PGRs) in the fossil record. This imbalance is due in part to the prior inability to infer such a complex life history trait from skeletal remains.Prenatal growth, the rate of embryonic and fetal growth in utero, plays a key role in establishing the trajectory of an individual’s metabolism, neurological development, and ultimate growth during their lifetime (21). Calculated as the ratio of birth weight (mass in grams) to gestation length (days) (22), mammalian species that have high rates of prenatal growth have infants that are larger at birth relative to other species with comparable lengths of gestation. In primates, gestational length is relatively similar across the phylogeny, while PGR can vary quite significantly (23, 24), suggesting that growth rate provides a key source of variation upon which primate evolution occurs.Despite the oft-repeated statement that human infants are secondarily altricial, meaning that they have most of their growing to do after they are born, humans have the highest PGR among primates (25–27). This life history distinguishes humans from even their closest living relatives (26, 28, 29). This high PGR results in human infants that are quite large at birth relative to their time of gestation, with both large neonatal body mass and brain mass compared to other primates (25, 30). Yet, despite this large body and brain mass, at birth the human infant brain is still only 30% of the size of the adult brain, a developmental characteristic that leads to a helpless infant that is highly reliant on parents and other social group members for survival (19, 31).Cooperative breeding, pair bonding, and group care have been hypothesized as critical factors in the evolution of the human brain and the helpless infant (11, 17, 21, 32–37). The helpless infant requires constant care and attention, which is provided only by the birthing parent unless they can receive some assistance. Research in nonhuman primates shows that mothers who receive assistance are better able to secure resources for themselves and their infants (30, 36, 38). As the brain is a metabolically expensive organ, resource intake is essential for infant brain growth and development (32, 39, 40). This is also true at earlier life history stages, when the fetus is developing in utero (5, 15, 28). Research in cetaceans has demonstrated that brain size in these large and highly social animals is related to PGR as well as gestational length, with growth rate dependent on maternal energetics, including metabolism, resource intake, and other physiological traits involved in gestation (41). This supports the hypothesis that PGRs are also significantly associated with adult endocranial volume (ECV) in primates, which would not be a surprising result since adult ECV is strongly correlated with neonatal ECV in primates (4). Across eutherians, large neonatal and adult relative brain size do not appear to be related to periods of accelerated brain growth during fetal development, as higher encephalization seems to be a product of slower body growth rather than an increased rate of brain growth (42). However, to our knowledge the direct correlation between PGR and adult ECV has not been previously quantified in primates, and so we test this hypothesis here.Like the brain, teeth begin to form in utero and are thus intimately associated with the developmental processes in mammalian early life history (43, 44). After birth, teeth remain linked to crucial life history landmarks, including weaning, menarche, and sexual activity through mate competition, and have thus been used extensively to reconstruct primate life histories in the fossil record (1, 29, 45–55). Considerable pleiotropy underlies dental development, with genetic and phenotypic variation in individual tooth morphology being shared with body size, craniofacial structures, and other teeth across the dental arch (43, 44, 56, 57). More recent work has demonstrated that these pleiotropic effects extend even to traits in other parts of the body, including those associated with the maternal–infant relationship, such as a lactation (58).Our previous work in callitrichids linked slow PGRs to third molar loss, providing one of the first lines of evidence that dental morphology may be directly associated with PGR in primates (59). This insight has direct bearing on humans, as third molar reduction remains one of the most distinct dental features of humans compared to other primates (60–62). A large body of work has generated hypotheses for the drivers of third molar reduction and loss in humans, which range from craniofacial reduction, to brain size increases, to changes in diet and technology, and may in fact be related to all of these morphological and behavioral evolutionary shifts; these relationships have yet to be resolved (60, 63, 64). Based on our previous work, we hypothesized that PGR has a direct quantitative relationship with both ECV and third molar reduction in humans. One of the clearest ways to investigate third molar reduction is through quantification of molar proportions (60–63).In order to test our hypothesis, we collected dental data on extant and fossil primates and calculated the molar module component (MMC), a genetically patterned trait that captures the ratio of the third to first molar length and thus third molar relative size. MMC has been genetically and phenotypically assessed in a range of catarrhine primates (57, 65), including fossil hominids (66, 67), as well as more broadly across mammals (68, 69). All of these studies demonstrate that MMC carries a strong phylogenetic signal and is not driven by dietary adaptation or body size. Given that disruptions to intrauterine growth in humans result in reduced cranial and tooth size (43, 70), the genetic mechanism underlying variation in MMC may be shared between these traits. We present here an investigation of the relationship between MMC and PGR in extant and fossil primates. Our work aims to develop methods for estimating the evolution of PGR in the hominid fossil record and thereby open a window into the evolutionary history of human gestation.One of the great challenges to human evolutionary research lies in the limitations of the fossil record. Although fossils are an essential resource that provide the only direct line into the past, the factors at play in fossilization result in a vertebrate record that is comprised almost exclusively of fossilized teeth and bones (71). Thus, in order to elucidate the evolution of human life history, researchers are tasked with the critical job of developing new methods for deducing life history details from hard tissues. Reconstructing life history in the fossil record remains a primary focus of human evolutionary studies (49, 72–74), and a large body of research has focused on the evolutionary interplay between brain size and craniodental and pelvic morphology in fossil hominids (75, 76). Considering that PGR is linked to maternal investment and ECV, a clearer understanding of PGRs in fossil hominids is invaluable to the study of the evolution of human gestation and brain size, crucial elements in life history theory. Using extant and fossil primate data, including dental metrics, PGR, and ECV, we tested the following hypotheses:

• H1)PGR is highly correlated with ECV in extant catarrhine primates.
• H2)Third molar proportions (as captured by MMC) are significantly associated with PGR and ECV in extant catarrhine primates.
• H3)Third molar proportions (as captured by MMC) can accurately predict ECV, and thus PGR, in the hominid fossil record.

By testing these hypotheses, we aimed to address two larger outstanding questions about the evolution of life history in humans:

• Q1)Have hominoid PGRs increased significantly throughout the Miocene and Plio-Pleistocene?
• Q2)When did human-like (i.e., more similar to humans than to other extant apes) PGRs first evolve?

In order to test our hypotheses and answer these essential questions about life history, we developed two models for predicting PGR (infant body mass in grams divided by gestation length in days) in the fossil record by using statistical relationships between PGR, ECV, and molar proportions (third molar length relative to first molar length, MMC) in extant catarrhines (n = 608). We present evidence that fossil PGRs can be predicted from both ECV and MMC and thus provide two methods for investigating life history in the fossil record. The relationship between ECV and PGR is extremely tight, while the dental model has more variable predictions, probably due to the impacts of derived craniofacial morphology (e.g., extreme prognathism). To check that dental proportions accurately reconstruct PGR and ECV, we validated our dental model through a quantitative comparison of predicted and observed ECV in the hominid fossil record (n = 13 species). Additionally, we statistically compared the results from both models and assessed the evolution of PGR over the last 6 million years of the hominid fossil record.