Developmental palaeontology of Reptilia as revealed by histological studies

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright

Author's personal copy Seminars in Cell & Developmental Biology 21 (2010) 462–470

Contents lists available at ScienceDirect

Seminars in Cell & Developmental Biology journal homepage: www.elsevier.com/locate/semcdb

Review

Developmental palaeontology of Reptilia as revealed by histological studies Torsten M. Scheyer a,∗ , Nicole Klein b , P. Martin Sander b a b

Paläontologisches Institut und Museum, Universität Zürich, Karl Schmid-Strasse 4, CH-8006 Zürich, Switzerland Steinmann-Institut für Geologie, Mineralogie und Paläontologie, Universität Bonn, Bonn, Germany

a r t i c l e

i n f o

Article history: Available online 11 November 2009 Keywords: Bone histology Microstructure Fossil record Life history Growth record

a b s t r a c t Among the fossilised ontogenetic series known for tetrapods, only more basal groups like temnospondyl amphibians have been used extensively in developmental studies, whereas reptilian and synapsid data have been largely neglected so far. However, before such ontogenetic series can be subject to study, the relative age and affiliation of putative specimens within a series has to be verified. Bone histology has a long-standing tradition as being a source of palaeobiological and growth history data in fossil amniotes and indeed, the analysis of bone microstructures still remains the most important and most reliable tool for determining the absolute ontogenetic age of fossil vertebrates. It is also the only direct way to reconstruct life histories and growth strategies for extinct animals. Herein the record of bone histology among Reptilia and its application to elucidate and expand fossilised ontogenies as a source of developmental data are reviewed. © 2009 Elsevier Ltd. All rights reserved.

Contents 1. 2. 3. 4.

5.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bone histology applied to fossilised ontogenies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Parareptilia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eureptilia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Basal Eureptilia and Basal Diapsida . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Ichthyopterygia and Thalattosauriformes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3. Testudinata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4. Lepidosauromorpha . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.1. Sauropterygia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.2. Squamata and Rhynchocephalia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5. Archosauromorpha . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5.1. Basal Archosauromorpha and Crurotarsi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5.2. Ornithodira . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion and conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix A. Supplementary data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction To understand modern life, it is essential to study its evolutionary past. Fossils allow us to assess the breadth of bone biology and development in evolutionary history, because many

∗ Corresponding author. Tel.: +41 044 63 423 22; fax: +41 044 63 449 23. E-mail addresses: [email protected], [email protected] (T.M. Scheyer), [email protected] (N. Klein), [email protected] (P.M. Sander). 1084-9521/$ – see front matter © 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.semcdb.2009.11.005

462 463 463 463 463 463 464 464 464 464 465 465 465 468 468 468 468

fossil bones present tissue structures and developmental pathways not known in extant species. A fossil specimen however represents only a glimpse of the complete ontogenesis from fertilization to death. In order to better understand the evolution of extinct life history in a fossil species, it is important to analyse as much ontogenetic data as possible. There is continued and heightened interest to incorporate ontogenetic aspects in palaeobiological but also in phylogenetic studies, and vice versa, to also include fossils in studies of evolutionary developmental biology.

Author's personal copy T.M. Scheyer et al. / Seminars in Cell & Developmental Biology 21 (2010) 462–470

During the past two decades, major advances in the study and phylogenetic analyses of skeletal developmental events were achieved [1]. Studying fossil growth series has largely been restricted to basal groups among Vertebrata (e.g., tetrapodomorphs, temnospondyl amphibians), basically because of the abundance and exceptional preservation of fossils in these groups ([2,3] this issue). Fossil embryos and juvenile specimens, although rare faunal elements in the fossil record, are known from a variety of amniote lineages ([4] this issue), however, because of preservation and abundance of fossils, dinosaurs and marine reptiles from the Triassic strata of Europe and China are especially well suited to increase the scope of these studies into the amniote record. However, even in the UNESCO site of Monte San Giorgio, one of the most important fossil Lagerstätten of marine reptiles from the Middle Triassic, the discovery of well based developmental series including embryos, juveniles and adults, is a rare circumstance. Ichthyosaurs and the small sauropterygian pachypleurosaurs yielded so far the best growth series, but additional taxa, including for example other marine sauropterygians or archosauromorphs might also be suited to provide valuable data in the future. With the exponential output of descriptions of closely related and well preserved fossils from China [5–10], developmental data for comparison are steadily increasing. For over a century, comparative histological research of fossil vertebrate bones has been established as a complementary venue of research next to gross morphology and osteology [11–13]. In contrast to the developmental studies which were dominated so far by fossil non-amniote taxa, the field of bone histology of fossil tetrapods was, until recently [14,15], dominated by amniote lineages [16–20]. Bone histology permits to access data about palaeoecology and phylogeny [21–23], as well as growth and life history (Fig. 1) and individual age of extinct taxa [24–27], the latter being important in appraising the ontogenetic age of individuals within a fossil series. In this regard, it was aptly stated that “bone tissues throughout the skeleton should be described at as many stages of growth as can be made available in taxa whose phylogenetic relationships are established on the basis of other characters” ([28]: p. 351). Herein, the skeletochronological aspect of bone histology in Reptilia (sensu [29]) is summarised, and the importance of bone histology to potentially reveal and expand fossil ontogenetic series is outlined. Bone histological terminology and classification of bone tissues follows Francillon-Vieillot et al. [25] and Sander et al. [30]. Fossil reptilian groups for which life history data based on bone microstructures are available are shown in Fig. 2. Note that the taxonomic position of some taxa, i.e., turtles and ichthyosaurs, is still under debate. A comprehensive list of studies and sampled taxa using bone histology in amniotes is available under http://www.developmental-palaeontology.net.

2. Bone histology applied to fossilised ontogenies The assessment of individual age in extant and fossil vertebrates is for a large part based on the counting of periodically deposited growth marks, i.e., annuli or lines of arrested growth (LAGs) or growth cycles in primary bone tissue, a method known as skeletochronology [27,31–37]. In poikilotherms, LAGs mainly occur due to annual cessation of bone growth, although they are also known from bones of homoeothermic animals (as summarised in Refs. [36–38]). The annularity of these cyclical growth marks has been validated for several vertebrate groups using extant examples [39–43], however, the mechanism of LAG deposition and identification of LAGs in fossils are still debated [44–47].

463

Possible reasons discussed for annual growth marks in bone are environmental changes (seasonal changes in light intensity, temperatures, or related to wet and dry seasonal patterns), nutrition and diet, disease, as well as migratory and reproductive cyclicity [27,32,48]. Where the growth record was lost because of internal remodelling processes of the bone, i.e., through the centrifugal expansion of the marrow cavity, potential growth cycles have to be retrocalculated. Given the intrinsic connection of bone structures and rate of bone deposition known as Amprino’s rule [46,49–51], bone deposition rates can corroborate age estimates based on cyclical growth marks [52–54]. Similarly, the analysis of the isotope composition, i.e., changing ␦18 O values within the cortical regions of the bone, may contribute to infer cyclical or seasonal growth patterns [55]. Many of these methods have been recently developed, and thus their assets to the growing field of skeletochronology still have to be critically evaluated. Several studies on dinosaurs [30,54,56–58] have demonstrated that histological sections of fossil bone samples can be used to establish the relative age/ontogenetic stage of specimens. Because bones are usually affected by remodelling processes a single bone does not reveal the complete growth of an individual, and overlapping histological records of younger and older individuals must be combined to elucidate the full ontogenetic growth trajectory for the species. Once we have the necessary sufficient phylogenetic and ontogenetic control for the interpretation of histological data, “it is becoming possible to see how growth strategies change throughout the evolution of an extinct clade, and how they allow animals to exploit new evolutionary opportunities” ([59]: p. 144). 3. Parareptilia Among Parareptilia, a diverse group of Palaeozoic and Early Mesozoic reptiles [60], descriptions of bone histology are available basically only for mesosaurs, pareiasaurs and procolophonids [17,61], with newer studies focusing more on functional and phylogenetic aspects of bone microstructures than on aspects related to growth and aging [62,63]. Skeletochronological or life history studies using several specimens of different ontogenetic stages have, up to our knowledge, never been attempted in any parareptile. 4. Eureptilia 4.1. Basal Eureptilia and Basal Diapsida With the exception of the more diverse and widespread basal eureptilian Captorhinidae, and the diapsid Younginiformes and Claudiosaurus germaini (Late Permian of Madagascar), for which bone histological data are partly also available [17,21,32,61,64], the fossil record of basal eureptiles and basal diapsids (small to medium sized terrestrial lizard-like animals) is poor. Those specimens sectioned for histology usually exhibit periosteal lamellar-zonal bone suitable for growth studies [64]. Good growth series are also known from some of the marine Younginiformes, e.g., Hovasaurus and Tangasaurus [65,66]. 4.2. Ichthyopterygia and Thalattosauriformes Growth series including embryos were mainly known for ichthyosaurs and thalattosaurs from the Lower Jurassic Posidonia Shale [67] and Middle Triassic Besano Formation, Monte San Giorgio [68], respectively, but exceptional growth series as evidenced by abundant embryonic to adult specimens were recently reported also for the small sized ichthyosaur Qianichthyosaurus and the tha-

Author's personal copy 464

T.M. Scheyer et al. / Seminars in Cell & Developmental Biology 21 (2010) 462–470

Fig. 1. Acquisition of bone histological data for life history analyses. (A) Sketch of prosauropod Plateosaurus engelhardti in which the position of humerus and femur, two of the skeletal elements usually used for microstructural analysis is marked. Sampling is usually done by either complete cross-sectioning (see separate sketch of femur) or core-drilling of the bone at mid-diaphysis. (B–D) Histology of P. engelhardti. (B) Polished section of humerus (NAA F88/B640, 44.5 cm long, Frick). (C) Close-up of the outer region of the cortex of same specimen. (D) Interpretative drawing of histology seen in C. Note change from fibrolamellar bone tissue (light grey) with growth cycles to lamellar-zonal bone (dark grey) with more closely spaced lines of arrested growth (LAGs).

lattosaur Xinpusaurus from the Upper Triassic Xiaowa Formation, southwestern China [10]. The bone histology has been studied only in few ichthyosaurs, representing mostly medium to large sized taxa [69], but data on smaller, basal forms will soon become available [70,71]. Among these larger forms, the primary periosteal bone consists of very fast deposited woven bone tissue [69]. Growth marks were found to be present only in Omphalosaurus, an enigmatic durophagous ichthyosaur from the Early Triassic of uncertain systematic position [72], whereas growth marks were absent in the continuously growing bones of the Jurassic taxa Ichthyosaurus and Stenopterygius [69]. Similar to Omphalosaurus, LAGs are preserved in smaller and more basally positioned ichthyosaurs such as Utatsusaurus (Lower Triassic Osamu Formation, Japan [70]), and Mixosaurus (Middle Triassic strata of Monte San Giorgio, Switzerland, and surrounding localities [71]). 4.3. Testudinata Several works on the skeletochronology of bones of extant, mainly marine, turtles (e.g., [33,43,73]) are available now, but there are only few results on fossil turtle taxa. Growth data and age estimates were reported for two species of fossil tortoises [74], Gopherus laticuneus and Stylemys nebrascensis (Eocene-Oligocene White River Group, northwestern Nebraska, USA). LAGs were encountered in both the fossil and living species, thus allowing age estimates for the fossils between hatchlings in their first year of life and adults that lived more than 40 years [74]. 4.4. Lepidosauromorpha 4.4.1. Sauropterygia The bone microstructures of the marine Sauropterygia from the Triassic are well studied, largely because of the abundance and fossil preservation. The primary cortical bone of ribs and appendicular bones of Neusticosaurus and Nothosaurus is usually composed of (the typically reptilian) lamellar-zonal bone tissue [19], whereas the cortex in the humerus of the placodont Placodus gigas was reported to comprise well vascularised ‘periosteal “woven fibred” bone tissue’, indicating higher growth

rates for this species [75]. An ontogenetic shift from pachyosteosclerotic to osteoporotic-like patterns of bone growth was reported in two plesiosaur taxa from the Upper Cretaceous of New Zealand [76], but further skeletochronological data were not given. 4.4.1.1. Pachypleurosauria. A skeletochronological analysis using polished sections of humeri has been conducted for two species of pachypleurosaurs, Neusticosaurus peyeri and Neusticosaurus pusillus from the Middle Triassic of Monte San Giorgio, Switzerland [77,78]. Growth marks, i.e., annuli, were also encountered in several other bones of these taxa, including femora, vertebral centra and ribs. The first growth curve for a fossil amniote was published for these two species of Neusticosaurus [78], and it was speculated that nine or ten years were the maximum age in N. peyeri, whereas the smaller N. pusillus seemed to have a life span of approximately eight years. Because Neusticosaurus is exceptional among fossil reptiles in displaying unequivocal secondary sex characters in its skeleton, the onset of sexual maturity is easily determined from morphology and was found to lie between two and three years of life in both species The Neusticosaurus species studied thus represent the typical reptilian life history/growth pattern. 4.4.2. Squamata and Rhynchocephalia Among fossil squamates, the bone histology of varanids, snakes and Upper Cretaceous marine mosasaurs was extensively studied [79]. Based on growth of varanid vermiform osteoderms, Varanus (Megalania) priscus, which inhabited Australia during the Pleistocene, was reported to reach its enormous size (twice the length of Varanus komodoensis) by prolongation of the growth period with juvenile growth rates similar to those of the largest extant varanids [80]. Rhynchocephalia (sensu [81]), a group of lizard-like diapsids that originated during the Triassic, are represented today only by Sphenodon punctatus and Sphenodon guentheri on islands near mainland New Zealand [82]. Bone cortices of Lacertilia and S. punctatus consist of avascular or very poorly vascularised lamellar or pseudolamellar tissue, and there is a positive correlation of growth rate with rising temperatures in S. punctatus [83].

Author's personal copy T.M. Scheyer et al. / Seminars in Cell & Developmental Biology 21 (2010) 462–470

465

Fig. 2. Phylogenetic framework of Reptilia (sensu [29,60]). Lineages subjected to skeletochronological study are indicated by bold names and representative sketches (redrawn and modified from various sources, not to scale).

4.4.2.1. Mosasauridae. Whereas the bone histology of mosasaurs has been sampled in various elements of the skeleton (e.g., [12,17,19,84,85]), skeletochronology has been analysed only in the limbs of Tylosaurus, Platecarpus and Clidastes [86]. These exhibit cortical structures similar to those of extant varanids, i.e., lamellar-zonal bone with LAGs and zones, but also interspersed supernumerary LAGs, whose meaning and development is still poorly understood [86]. Growth rates seem to have been raised in these marine forms compared to extant terrestrial relatives, and sexual maturity was reported to occur between five and seven years of age. 4.5. Archosauromorpha 4.5.1. Basal Archosauromorpha and Crurotarsi Crurotarsi are often interpreted to be typical poikilothermic tetrapods, whose cortical bone is usually composed of parallel-fibred or lamellar-zonal bone (e.g., [87–91]); tissues generally associated with reduced deposition rates. However, most crurotarsan lineages potentially reveal higher growth rates in the form of depositing fibrolamellar tissue, especially during early development [89]. By including more basally positioned archosauromorphs, it was further indicated that the “possibility of reaching and maintaining very high growth rates through ontogeny could have been a basal characteristic of archosauriforms” ([90]: p. 57–58). These finds indicate that, e.g., crocodylomorphs might have secondarily reverted to “more generalised reptilian growth

strategies” ([89]: p. 58). For example, analyses of the growth strategy of the giant North American genus Deinosuchus from the Late Cretaceous reveal that giant body size is associated with a simple prolongation of growth with juvenile growth rates, compared to smaller crocodylian species [91]. 4.5.2. Ornithodira 4.5.2.1. Pterosauria. The genera Rhamphorhychus and Pterodactylus have been indicated to have the best, although not complete, ontogenetic record in Pterosauria [28], however, bones of juveniles have been used sparsely for histological study [28,92,93]. Comparative bone histology of the extremely thin pterosaurs bones so far revealed structures indicative of rapid growth rates, i.e., highly vascularised fibrolamellar bone tissue, similar to that found in birds and mammals, but LAGs were found to occur occasionally, especially in smaller, less fast growing pterosaurs [88,94,95]. Because of the typically hollow nature of pterosaur bones, the large scale skeletochronological applicability in this clade was doubted [95]; however, previous studies are still too preliminary in terms of taxon and element sampling to allow such an assessment. Life history data for a growth series of Pterodaustro gui˜ nazui (Early Cretaceous Largacito Formation of Argentina), a filterfeeding pterosaur, were recently described [86,96]. In this pterodactyloid rapid growth seems to have taken place until sexual maturity is reached after two years, followed by three to four years of growth with slightly lower growth rates, before skeletal maturity marked by a change from zonal fibrolamellar bone to

Author's personal copy 466

T.M. Scheyer et al. / Seminars in Cell & Developmental Biology 21 (2010) 462–470

cycles of parallel-fibred and lamellar bone tissue ([96]: p. 1474) sets in at about five to six years of age. However, in these and other archosaurs for which life history was studied from bone histology, sexual maturity could not be detected from the appearance of primary or secondary sexual characters but instead was hypothesised to correlate with a sudden decrease in growth rate, marking the shift in resource allocation from growth to reproduction.

4.5.2.2. Dinosauria (including birds). Recent reviews of dinosaur growth and life history data were presented [51,97] and, in some aspects, controversially discussed [45], but much new and exciting research has appeared since then. Numerous ontogenetic series [51,54,56–58,97–104] as well as skeletochronology have been studied (i.e., [56,57,103,105,106]). Dinosaur taxonomic diversity is mirrored by diverse growth patterns and strategies, often related to their maximum adult size and mass which range from approximately 1 kg to 75 t. All dinosaurs grow principally with the fibrolamellar bone tissue type which represents always fast growth rates; differences concern mainly the density and organization of vascular systems. In spite of their fast growth in many dinosaurs, growth is regularly interrupted by LAGs. The deposition of an external fundamental system [107] is common in fully grown individuals, showing that growth rate clearly had ceased and was determinate. 4.5.2.2.1. Ornithischia. Among ornithischian dinosaurs, primarily Ornithopoda has been studied with regard to bone histology, while very little is known about Thyreophora and Marginocephalia. Recently, data have become available on the histology of the most basal ornithischian, Lesothosaurus diagnosticus [108] which reached body length of about 1.8 m. The juveniles of Lesothosaurus grew very fast with highly vascularised fibrolamellar bone while the adults show laminar fibrolamellar bone with growth marks. This suggests that growth rates in ornithischians were plesiomorphically high. Thyreophora: Among Thyreophora, long bone histology of the small basal form Scutellosaurus [94] indicates slower growth than in Lesothosaurus. This is in accordance with results on comparative long bone histology of Stegosaurus [109]. While of fibrolamellar architecture, the primary bone of Stegosaurus, grew more slowly than in other large dinosaurs as evidenced by low vascularity, primarily longitudinal canals, and much parallel-fibred bone matrix. The Thyreophora as a lineage thus may have been characterised by a secondary reduction in growth rate. The histology of the peculiar plates of Stegosaurus had received attention early on [110], but mainly from a physiological perspective. This has changed with the of a growth series of Stegosaurus plates and some long bones [111], noting the presence of fibrolamellar bone indicative of high growth rates and a delay in the maturation of the plates compared to the long bones, suggesting a function in intraspecific display of the plates. Nothing has been published on ontogenetic histology of the other major group of Thyreophora, the Ankylosauria, although the histogenesis and possible function of armour plates of thyreophoran dinosaurs were addressed before [112,113]. Marginocephalia: Among Marginocephalia, the histology of Pachycephalosauria and Neoceratopsia remains largely undescribed with few exceptions [114], despite the abundant and diverse material of especially the latter taxon. One of the most in-depth ontogenetic histologic studies was done on the basal ceratopsian Psittacosaurus [56]. The study represents a milestone in that it uses a good skeletochronologic record in long bones to derive the first robust mass-based growth curve for a dinosaur, earlier studies having used some linear measurement or crude estimator of body mass. Despite being a small dinosaur, Psittacosaurus shows abundant fibrolamellar bone and grew at a rate comparable to a similar-sized marsupial mammal.

Ornithopoda: Despite the great diversity and excellent fossil record of Ornithopoda, not many studies have investigated their bone histology, the focus mainly being on the most derived members of the clade, the Hadrosauridae. Work on the Late Jurassic ornithopod Dryosaurus lettowvorbecki [115] suggested fast and uninterrupted growth, but the sample did not contain fully grown individuals. The latter part of the ontogeny of D. lettowvorbecki thus remains unknown. A recent study on ontogenetic series of Dryosaurus altus and other “hypsilophodontid” euornithopods confirmed the lack of adults in Dryosaurus based on comparisons with adults known from other ornithopods (i.e., Orodromeus, Tenontosaurus and hadrosaurs) that experienced determinate growth [116]. Hadrosaurs, e.g., Maiasaura [57] and Hypacrosaurus [117], are similar in histology to Dryosaurus, also depositing fibrolamellar bone in their long bone cortex, with LAGs only appearing in the subadult stage. Ornithopod histology and growth strategies thus appear little changed from the ancestral condition, although it is not known how the body size increase from the small basal forms to large hadrosaurs and iguanodontians came about. To decide between an increase in growth rate and a prolongation of the growth phase, growth curves would be required which so far have not been published for any ornithopod. The only absolute data on time required to grow to full size are those for Maiasaura with estimated six to eight years [57]. Nevertheless, hadrosaur ontogenetic histology offers interesting perspectives on the relationship between evolution and development because of their record of embryonic, perinatal, and early juvenile histologies which are lacking for most other dinosaurs. Both Maiasaura and Hypacrosaurus [107] are similar to birds in this regard but differ from turtles and crocodiles, indicating that high growth rates were typical throughout ontogeny. 4.5.2.2.2. Saurischia. Sauropodomorpha: Basal sauropodomorphs generally have fibrolamellar bone interrupted by regularly spaced growth marks, and termination of growth is recorded in an external fundamental system (EFS, [57]). The prosauropod Plateosaurus engelhardti is the most common dinosaur from the Upper Triassic of Central Europe and is also one of the best studied dinosaurs, including its long bone histology [105,106]. The primary bone which consists mainly of fibrolamellar bone is regularly stratified by growth marks, which allows skeletochronology. Although early ontogenetic stages are yet not known for Plateosaurus, a growth curve was reconstructed for the available growth series which span several phases of its late ontogeny. A sudden slow-down in growth, presumably recording sexual maturity occurred well before full size was reached in Plateosaurus. After sexual maturity, growth continued on a lower rate for several years. Finally growth ceased which is recorded by the deposition of an EFS. Qualitative (growth stop, EFS) and quantitative (growth-mark counts) features of the bone histology of Plateosaurus are poorly correlated with body size, indicating developmental plasticity as well as a variability of life history and dependence by environmental factors, as is typical for modern ectothermic reptiles, but not for mammals, birds, or other dinosaurs [105]. Chinsamy was one of the first to study bone histology in dinosaurs from an ontogenetic perspective (summarised in Ref. [97]). The bone histological study on a partial growth series of the prosauropod Massospondylus [99] was the first growth curve to be published for a dinosaur. Massospondylus also shows fibrolamellar bone which is regularly interrupted by growth marks, but unlike Plateosaurus, Massospondylus shows a close correlation between age and body size. This may reflect a rather different growth strategy or, more likely, that no fully grown individuals were sampled. This is suggested by the lack of EFS in any sample of Massospondylus [97].

Author's personal copy T.M. Scheyer et al. / Seminars in Cell & Developmental Biology 21 (2010) 462–470

Sauropoda. Because of their giant body size and the various questions raised by gigantism [118], Sauropoda is a histologically well studied group (summarised in Ref. [30]). However, because of the rareness of taxonomical assignable juveniles in the fossil record, studying sauropod growth series remains difficult [58]. Additionally, standard skeletochronology failed in most sauropod taxa, due to the lack of growth marks in the primary bone tissue for most of their ontogeny. Only single specimens do show a relatively good growth record, which always starts in the adult stage ([54,101,119], also [120]). Nevertheless, different histologic ontogenetic stages, representing animals from hatchlings to senescent individuals, are easy to identify on the basis of long bone tissue types [30,54]. The problem remains, however, how to quantify time for these histologic ontogenetic stages. Bone histology and histologic ontogenetic stages also contribute to taxonomical questions, e.g., by distinguishing juvenile specimen from dwarf forms [104,121]. Growth of the giant sauropods differs from that of most other dinosaurs. All sauropod long bones grew fairly uniformly, laying down continuously deposited large amounts of laminar fibrolamellar bone tissue in their cortex. Growth marks of any kind are rare and not consistently present in any taxon sampled from growth series. Differences in primary bone tissue types mainly concern the organization of the vascular system, the degree of vascularisation and the presence and degree of development of primary and secondary osteons, all of which are mainly related to ontogeny. Growth rate decreases gradually from very young to fully grown individuals. Growth rate and final size are taxon-specific in sauropods, not variable, and genetically predetermined. The few quantitative growth records available (e.g., [53,101]) suggest that sauropods reached their maximum adult size (∼15–75 t) in no more than two or three decades. However, the fast growth rates of sauropods of over 5 t per year previously cited [122] appear to be great overestimates [30,54,119]. Unfortunately, the use of these growth rates compromises the results of other studies, e.g., the overhearing problem postulated for sauropods [123] may not existed if more realistic growth rates of