Exploring the use of tomography for the quantification of cementum growth patterns across the mammal phylogeny

This project is focused on the application of state-of-the-art imaging technologies and novel analytical techniques in order to improve understanding of the growth and structure of cementum; the mineralised tissue that connects teeth to the peridontium in mammals. Like all vertebrate hard tissues, the growth of cementum is controlled by a series of natural rhythms that control its pattern and pace. This is recorded in circum-annual increments (similar to tree rings) hypothetically created by dietary differences in the production rate of its two key components; hydroxyapatite matrix and collagen fibres. As cementum is only rarely resorbed, these increments can be counted to provide a direct estimate of chronological age, a technique known as cementochronology. Further, the circum-annual rhythm followed by cementum increments may also allow them to record discrete life history events that create severe strain on body metabolism over the course of several months, such as pregnancy. However, the exact causation and rhythm of cementum increments is still poorly understood due to a paucity of direct experimental study. The majority of previous studies of cementochronology have been based on thin-section histology, and several caveats of this approach have undermined confidence in the ability to quantitatively analyse cementum increments. Overall, current study of cementum lacks the application of modern technology and analytical methods that have revolutionised interpretation of microstructures and ultrastructures of other hard tissues such as bone, and increased their use as hallmarks of disease or as records of life history. The application of synchrotron X-ray tomography (SR CT) has here been used to improve understanding of the structure and count of cementum increments in mammals ranging from our oldest ancestors, to some of our closest relatives. The non-destructive nature and high throughput of SR CT has allowed study of large samples of Morganucodon and Kuehneotherium, two of the oldest known fossil mammals from the early Jurassic (~200 million years ago). The three-dimensional perspective and ultra-high resolutions (<500nm voxel size) of SR CT data provided an unprecedented level of detail to study and more precise counts of increments than previous thin-section-based techniques, which in-turn provided minimum estimates of maximum lifespan for both animals. Further, as the lifespans of modern mammals are inversely related to their basal metabolic rate and post-natal growth rate, these estimates have been used to improve understanding of the physiology of early mammals and the evolution of the sophisticated endothermic (warm-blooded) physiology of modern mammals. From this study the potential of SR CT for analyzing cementum growth, and the relationship between cementum growth and life history variables, was made clear. Subsequent SR CT imaging was conducted on a sample of rhesus macaque monkeys (Macaca mulatta), raised in laboratory conditions. One sub-sample consisted of breeding females, another of non-breeding females, as well as a juvenile female and a male. This sample was used to study the optimal experimental settings for SR CT imaging of cementum, and generate an image processing and analysis workflow to automatically count cementum increments and study their shape and texture using computer vision. This workflow was then used to study sexual dimorphism in increment structure, and to investigate the potential for pregnancy events to be recorded in SR CT cementum data in both the M. mulatta sample and samples of C12th and C19th century archaeological humans of known sex (including three C19th individuals of known age and reproductive history). Significant dimorphism in increment shape and texture was found in both taxa. Female increments were found to be significantly more tortuous than males, with lower contrast and less well defined boundaries. Further, increments formed during pregnancy were found to be more chaotic in shape and texture than surrounding increments. Finally, SR CT was used to image the cementum increments of a diverse fauna of fossil mammals from the Middle Jurassic of the UK (Bathonian: ~168-166 million years ago). This fauna can be split into stem mammals (mammaliaforms alongside Morganucodon and Kuehneotherium), and crown mammals (bracketed evolutionarily by living mammals). Using the automative counting technique developed for counting primate cementum increments, it was found that mammaliaforms lived significantly longer maximum lifespans than contemporary crown mammals. This in-turn suggests a proportional disparity in metabolic potential, and that Mid Jurassic crown mammals had developed a similar basal metabolic rate to living mammals of comparable body mass. In summary, the work presented in this thesis has shown that the application of state-of-the-art analysis techniques has the ability to maximise the potential of cementum as a recording structure of disparate elements of life history among mammals. The workflow for imaging, processing and study developed here can be applied to a wide range of life history variables, for both fossil and extant taxa. The automated nature of the image analysis techniques presented overcomes many of the major caveats highlighted in previous thin-section based studies of cementum and their applicability, robusticity and accuracy can only be improved by their continued development in the wider cementochronology community outside of this project