Bone Histology

The origins of bone histology in paleontology trace back to the development of petrographic thin-section techniques in the early 19th century. In 1828, British scientists Henry Witham and William Nicol pioneered the method of grinding petrified wood into translucent sheets for microscopic examination. Louis Agassiz subsequently applied these techniques to fossil vertebrates. In 1849, John Thomas Quekett published a seminal comparative study of bone histological structure across mammals, birds, reptiles, and fish, systematically describing vascular canals, lacunae, canaliculi, and trabecular endosteal bone across the four vertebrate classes. This work is widely regarded as the foundation of comparative bone histology.

In 1850, British paleontologist Gideon Mantell produced the first unambiguous thin sections and microscopic descriptions of dinosaur bone, from the ankylosaur Hylaeosaurus and the titanosaur Pelorosaurus. Mantell described Haversian canals and bone cells preserved with a fidelity comparable to modern bone, demonstrating that original histological structures could survive the fossilization process—an observation confirmed repeatedly over the following 170 years.

Despite this early promise, systematic study of fossil archosaur bone microstructure did not begin in earnest until Enlow and Brown (1956, 1957, 1958) described the histology of limb and skull bones from extant and fossil birds, crocodilians, and several iconic dinosaurs including Allosaurus, Brachiosaurus, and Iguanodon. Their work confirmed that a wealth of biological information was preserved in fossil bone and laid the groundwork for subsequent advances.

The most consequential figure in the development of paleohistology as a tool for understanding dinosaur biology was Armand de Ricqlès, who from the 1960s through the 1980s produced an extensive body of work examining fossil bone microstructure across numerous vertebrate groups. De Ricqlès initially sought phylogenetically diagnostic characters in bone tissue but soon recognized that bone microstructure more reliably reflected growth dynamics than taxonomic affinity.

Building on Amprino's (1947) observation that local bone tissue organization correlates with its rate of deposition—now known as 'Amprino's rule'—de Ricqlès was the first to apply this principle systematically to fossil bone. He deduced that organisms with highly vascularized cortical bone grew faster than those with fewer vascular canals, and that woven bone (the least organized collagen fiber arrangement) was deposited most rapidly, while lamellar bone (the most organized) formed more slowly. His examinations of dinosaur bones revealed that their cortical bone was highly vascularized and predominantly fibrolamellar in character, resembling that of extant endothermic birds far more than the poorly vascularized, lamellar-zonal bone of extant ectothermic reptiles such as crocodilians. He concluded that dinosaurs possessed a physiology closer to that of modern birds than to living reptiles, challenging the long-standing perception of dinosaurs as sluggish, cold-blooded animals.

This qualitative and semi-quantitative relationship between bone tissue type, vascularization, growth rate, and physiology remains a foundational principle in paleohistology and was statistically validated by de Margerie, Cubo, and Castanet (2002) and by Montes et al. (2007).

Fibrolamellar bone (FLB) is a composite tissue consisting of woven bone matrix richly supplied with primary osteons. It is the hallmark of rapid periosteal growth and is typical of juvenile mammals, birds, and most non-avian dinosaurs. Depending on the orientation and density of vascular canals, FLB is further classified into laminar, plexiform, reticular, and radial sub-types, with radial FLB reflecting the highest deposition rates.

Lamellar-zonal bone consists of parallel-fibered or lamellar bone organized into zones separated by lines of arrested growth (LAGs). It reflects slow, cyclical growth and is characteristic of extant ectothermic reptiles. In dinosaurs, lamellar-zonal bone is typically observed in the outermost cortex as part of the external fundamental system (EFS), indicating the approach to skeletal maturity.

Haversian (secondary osteonal) bone results from remodeling of primary tissue by secondary osteons. Extensive Haversian remodeling is common in ontogenetically older individuals and can obscure primary growth marks, complicating age estimation.

Lines of arrested growth (LAGs) are hyper-mineralized lines approximately 10 μm thick within the bone cortex that record annual cessation or significant deceleration of periosteal bone deposition. Analogous to tree rings, they form during the harshest season of each year and their periodicity has been well documented in extant vertebrates. Counting LAGs provides a minimum estimate of individual age at death, while the spacing between successive LAGs reflects annual growth increments.

Early in the history of paleohistology, the presence of LAGs was correlated with ectothermy, and their absence was taken to suggest endothermy. However, Köhler et al. (2012) demonstrated that LAGs also form in some extant endothermic mammals as part of a plesiomorphic thermometabolic strategy to conserve energy during harsh seasons. Consequently, the presence or absence of LAGs alone cannot be used to infer metabolic status.

As an individual approaches skeletal maturity, LAG spacing progressively decreases until an external fundamental system (EFS) or outer circumferential layer (OCL) is deposited at the periosteal surface. The EFS has been documented in birds, crocodilians, and several dinosaur groups, and its presence indicates the attainment of skeletal maturity.

Annuli represent a related but distinct type of cortical growth mark—diffuse bands of more organized bone tissue reflecting periods of reduced (but not completely arrested) growth. Unlike LAGs, annuli are sometimes visible only under cross-polarized or circularly polarized light, a finding that has significant implications for age estimation, as demonstrated by Woodward et al. (2026).

Tyrannosaurus rex growth dynamics: Erickson et al. (2004) published a foundational study in Nature analyzing seven T. rex specimens to construct a growth curve. They reported a maximum growth rate exceeding 767 kg per year (approximately 2 kg per day) during a rapid growth phase between ages 14–18, with individuals reaching adult body masses above 8,000 kg within approximately 20 years and a lifespan approaching 30 years. However, Woodward et al. (2026) presented the most comprehensive histological analysis of Tyrannosaurus ontogeny to date, based on transverse diaphyseal sections of femora and tibiae from 17 individuals. By incorporating annulus-like birefringent bands visible only in circularly polarized light, they found that the best-supported growth model indicated lower maximum growth rates and a delayed attainment of asymptotic body size at approximately 35–40 years—some 15 years later than previously estimated. This study suggests that T. rex grew more gradually and over a longer lifespan than earlier models indicated.

Pachycephalosaur ontogeny: Horner and Goodwin (2009) used cranial histology to propose that Dracorex hogwartsia, Stygimoloch spinifer, and Pachycephalosaurus wyomingensis represent a single species at different ontogenetic stages ('ontogimorphs'), united by cranial bone metaplasia that dramatically transforms dome and ornament morphology during growth. Similarly, Scannella and Horner (2010) used frill histology to suggest that Triceratops and Torosaurus might represent different growth stages of a single taxon.

Maiasaura population growth: Woodward et al. (2015) analyzed 50 tibiae from the hadrosaur Maiasaura, constructing the most data-rich growth curve for any fossil vertebrate at the time. They determined that Maiasaura reached sexual maturity in approximately 3 years and skeletal maturity after 8 years.

Medullary bone discovery: Schweitzer, Wittmeyer, and Horner (2005) first reported medullary bone—a female-specific reproductive tissue previously known only in extant egg-laying birds—in the femur of a T. rex specimen. This discovery opened avenues for sex determination and reproductive biology studies in non-avian dinosaurs and suggested that this reproductive strategy may have evolved within the theropod lineage before the origin of birds.

Since de Ricqlès first proposed in the 1960s–1970s that dinosaur bone histology indicated endothermy, the metabolic status of dinosaurs has remained one of the most actively debated topics in paleontology. Legendre et al. (2016) used paleohistological models to estimate bone growth rates and resting metabolic rates, concluding that Mesozoic theropod dinosaurs exhibited metabolic rates very close to those of modern birds and that archosaurs share an ancestrally high metabolic rate, with the ectothermic physiology of extant crocodilians being a derived condition.

In contrast, Grady et al. (2014) analyzed growth rate allometry across a broad phylogenetic sample and proposed that dinosaurs were 'mesotherms'—organisms with metabolic rates intermediate between ectotherms and endotherms, most closely resembling some extant fish such as tuna and certain sharks. Werner and Griebeler (2014) suggested yet another possibility: that dinosaurs were fast-growing ectotherms.

The current consensus recognizes that the simplistic endotherm-versus-ectotherm dichotomy is inadequate for characterizing dinosaur physiology. Given the enormous taxonomic, ecological, and size diversity within the Dinosauria, a range of metabolic strategies likely existed across different lineages and even varied ontogenetically within species.

Thin-section preparation: A sample is removed from the diaphysis (shaft) of a long bone using a diamond-bladed saw. The sample is embedded in polyester resin for stabilization, then a wafer approximately 3 mm thick is cut and glued to a frosted plastic or glass slide. The mounted section is then ground and polished to a final thickness between 80 and 250 μm until optically translucent.

Microscopic examination: Sections are examined under transmitted light using multiple illumination modes. Plane-polarized light (PPL) reveals vascular canals, osteocyte lacunae, and LAGs. Cross-polarized light (XPL) highlights collagen fiber orientation through birefringence patterns. Circularly polarized light (CPL), produced by adding quarter-wave plates, eliminates periodic extinction and reveals annuli that may be invisible in other modes—a technique that proved critical in the 2026 Woodward et al. study.

Digital quantification: High-resolution composite images are assembled from motorized-stage micrographs. Software packages such as ImageJ/Fiji with the BoneJ plugin are used to quantify geometric centroids, cortical growth mark circumferences and enclosed areas, and periosteal and endosteal surface dimensions.

Destructive nature: Because bone must be physically sectioned, application to holotype specimens or exceptionally rare material is often restricted. Core sampling reduces specimen damage but provides less spatial information than complete transverse sections.

Histovariability: Different skeletal elements from the same individual can preserve different numbers of LAGs and different tissue types. Woodward et al. (2026) demonstrated that in a single Tyrannosaurus specimen (BDM 050), secondary osteon frequency in the femur was 10–15% higher than in the tibia, while the fibula was almost entirely remodeled. Element selection therefore significantly influences results, and standardization of sampling protocols (preferring femora or tibiae from the mid-diaphysis) has been advocated.

Statistical growth modeling advances: Earlier studies relied on subjective methods to estimate starting ages and construct composite growth curves. Myhrvold (2013) introduced the first algorithmic approach using least squares clustering. Woodward et al. (2026) further refined this by treating starting ages as independent variables estimated simultaneously with sigmoid function parameters using nonlinear optimization, producing more objective and statistically rigorous growth curves with simultaneous confidence bands.

Integration with molecular paleontology: The field is increasingly moving beyond traditional thin-section analysis to incorporate immunohistochemistry, molecular paleontology, and paleohistochemistry. These methods enable investigation of originally soft tissues, protein preservation, and chemical verification of structures such as medullary bone (Schweitzer et al., 2016), representing a frontier of paleohistological research that promises to extract even more biological information from fossil skeletal tissues.