Liebe, L. & Hurum, J.H.:
Gross internal structure and microstructure of plesiosaur limb bones from the Late Jurassic, central Spitsbergen.
NorwegianJournal of Geology, Vol 92, pp. 285-309. Trondheim 2012. ISSN 029-196X.
Plesiosaur limb bones of four specimens from the Late Jurassic of Svalbard have been studied to map gross internal structure and microstructure and compare this to extant marine reptiles and mammals. Two specimens; one juvenile (Djupedalia engeri) and one subadult (Colymbosaurus svalbardensis) are from the Janusfjellet locality in the Adventdalen Group, Janusfjellet Subgroup, Agardhfjellet Formation, Slottsmøya Member, dated as MidVolgian; one adult (Colymbosaurus svalbardensis) from the Agardhfjellet locality, also dated to MidVolgian; and one juvenile (sp.indet.) of unknown stratigraphical age. The bones examined are propodials, phalanges, mesopodials and metapodials. This study is the first to describe the microstructure of the latter two in plesiosaurs. The inner bone structure fits that of an active marine animal. Many of the features in the present material are often found in animals with rapid growth and high metabolism, including secondary osteons, high vascularisation, pits on the outside of the epiphysis, and woven and possibly fibro-lamellar bone in some areas. The long bones have two endochondral cones in a periosteal sheath, with a small medullary cavity. The propodials have a defined and quite compact cortex in the subadult, a finding that rejects the view that all plesiosaur bones became more porous through ontogeny. There is a microstructural difference between bones from different ontogenetic stages: juvenile bones lack remodelling and completely ossified endochondral cones. One of the bones has a circumferential vascular orientation with rings made of trabeculae, maybe resulting from cyclic growth caused by seasonality, migration or ontogeny. Lene Liebe & Jørn Harald Hurum, Natural History Museum, University of Oslo. Postboks 1172 Blindern, 0318 Oslo, Norway. E-mail corresponding author: firstname.lastname@example.org
Plesiosaurs (Sauropterygia, Plesiosauria) were secondarily aquatic marine reptiles, up to 15 metres in length and one of the main marine animal groups globally distributed in the Mesozoic. The first plesiosaurs originated in the Late Triassic, had a large diversification during the Jurassic and the Cretaceous, and became extinct in the Late Cretaceous (Brown 1981; Mazin 2001). About 43 genera and 65 species are recognised (Mazin 2001), but there are major gaps both geographically and stratigraphically in the plesiosaur fossil record (Druckenmiller 2006). There were two main groups of plesiosaurs: the long-necked plesiosaurs with small heads (Plesiosauroidea) and the short-necked pliosaurs with larger heads (Pliosauroidea) (Brown 1981). Both types are found in the Upper Jurassic deposits of Spitsbergen. Sauropterygiahas an uncertain position within the Reptilia, and the origin and ancestry of sauropterygians is unknown (Klein 2010). Plesiosaurs had a euryapsid skull, meaning only a singletemporal fenestra dorsal to the postorbital and the squamosal, a condition derived from a diapsid ancestor (Liem et al. 2001). The plesiosaur limb was very derived and important for propulsion, with fore- and hindlimbs of almost the same size and shape, evolved into flippers (Liem et al. 2001; but see discussion in Munthe-Kaas 2011). The propodials were large, while other limb bone elements were reduced in size and became more similar in shape through evolution. Hyperphalangy elongated the distal part of the limb. The limb is thought to have been a stiff unit, but there is an ongoing debate about how it was used and how the animal propelled itself through water (Thewissen & Taylor 2007; Carpenter et al. 2010; Munthe-Kaas 2011). Plesiosaurs were carnivorous (Druckenmiller & Russell 2008) and maybe endothermic (Bernard et al. 2010). A pregnant Late Cretaceous plesiosaur has recently been found, indicating that plesiosaurs evolved viviparity. This has for a long time been an unanswered question (Mazin 2001; O’Keefe & Chiappe 2011). However, the mode of reproduction in plesiosaurs was not unexpected since the pachypleurosaur Keichousaurus from the Middle Triassic has been proved to have given live birth (Cheng et al. 2004).
Fossil bones are subjected to many factors and shape, size, mineral composition and colour have often changed with time; but the integrity of the histology is often preserved and can be studied in thin-sections. This provides a considerable amount of information and is used for studies of animals from the earliest to extant vertebrates (Chinsamy 1997). Common structures to describe and use for interpretation are the degree and system of vascularisation, porosity, fibrillar organisation and bone tissue type, relationship between cortex and medullary region, presence and orientation of osteocyte lacunae and canaliculi, and degree of remodelling. Four principal factors determine the type and form of hard tissues in vertebrates: phylogeny, ontogeny, mechanics and the environment (Ricqlès 1977; Horner et al. 2000), but it is unclear to what degree each of these factors results in the observable structures. There has been a debate about whether bone microstructure mainly reflects phylogeny or ontogenetic and functional factors
(including growth rate and biomechanical constraints), but it seems that “microstructural characters may have both a functional significance and a systematic value at some level of the phylogeny” (Cubo et al. 2005: p.562). Between the observable characters and the factors that influence it, there are several interrelationships (Montes et al. 2010). Caution should therefore be taken in interpretation. Using microstructural characters for inferring phylogeny has been discussed for sauropterygians( Klein 2010) and correlation between microstructure and ecology has been used for mosasaurs (Sheldon 1997) and plesiosaurs (Wiffen et al. 1995).
Previous work on plesiosaur bone microstructureThe first article on bone microstructure of plesiosaurs described a vertebra and was published in 1878 (Hasse 1878). Around year 1900 (Kiprijanoff 1883a, 1883b;
Figure 1. Early illustrations of the inner structure in plesiosaur propodials mentioned in the introduction. Not to scale. A. Thaumatosaurus mosquensis, right humerus. Cretaceous. Kiprijanoff 1883a, plate 13, figure 2. B. Juvenile Plesiosaurus neocomiensis, femur. Longitudinal section. The earliest illustration of the inner cones in plesiosaur long bones. Kiprijanoff 1883b, plate 2, figure 5. C. Lütkesaurus, right humerus, proximal half. Longitudinal dorsoventral section. Cretaceous. Kiprijanoff 1883a, plate 20, figure 2. D. Plesiosaur, propodial. Jurassic specimen from the Kimmeridge clay. Longitudinal section. The earliest interpretation of the components of the inner structure. Lydekker 1889, figure 46. E. Juvenile plesiosaur propodial. Williston 1903, plate 22, figure 2. F. Juvenile plesiosaur propodial, the same bone as in E, longitudinal section. Moodie 1908, figure 4. G. Subadult plesiosaur propodial. Longitudinal section. Moodie 1908, figure 8. H. Adult plesiosaur, humerus. Longitudinal section. Moodie 1908, figure 9. I. Plesiosaur propodial. Diagram for longitudinal section based on analysed material. Moodie 1916, figure 6.
Lydekker 1889; Moodie 1908), the strange inner structure of the plesiosaur limb was discussed: it looked like an hourglass broken in the middle, or two large cones, which are also known to field palaeontologists, because the cones are sometimes freed from the surrounding sheath or have collapsed. Kiprijanoff (1883a, 1883b) first produced illustrations of the cones and did thin-section studies of the plesiosaur limb (Figs. 1A, B, C). Lydekker (1889; Fig. 1D) called the cones a “remarkable peculiarity” and thought they were epiphyses, the latter a mistake corrected by Moodie (1908) who also proposed a theory on how the plesiosaur limb grew (1916; Figs. 1F-I). Other early work on the structure and growth of plesiosaurian propodials was done by Williston (1903; Fig. 1E) and Woodward (1898). After this, hundreds of papers have been written about plesiosaurs (e.g., see references in Druckenmiller & Russell 2008), but only a few are concerned with the inside of the bone, and no one has described the growth process. It has been suggested that the inner structure results from ossification of cartilaginous cones, “but this is uncertain and ground sections need to be investigated” (Haines 1938; 1969: p. 92). In the early 20th century, survey studies of bone microstructure in different animal groups included plesiosaur limb bones (Seitz 1907; Gross 1934; Nopcsa & Heidsieck 1934). More recently, a few studies have been done on
the microstructure of plesiosaur limb bones (propodials and phalanges) (Wiffen et al. 1995; Fostowicz-Frelik & Gazdzicki 2001; Salgado et al. 2007; Gren 2010), focusing mainly on ballasting or lightening of bones as an adaptive mechanism, and the degree of internal remodelling. Most of this work is on Late Cretaceous species. Other plesiosaurian bones that have been studied through their microstructure are ribs, vertebrae, girdle bones and gastralia (Enlow & Brown 1957; Cruickshank et al. 1996; Street & O’Keefe 2010). These studies were also mainly focusing on fibrillar organisation, bone tissue type and ballasting, while plesiosaur bones also have been thin sectioned to study diagenesis and preservation (Martill 1991; Pewkliang et al. 2008; Kihle et al. 2012). Other sauropterygians for which bone microstructure has been studied are Placodus (Buffrénil & Mazin 1992), Pachypleurosauria (Buffrénil & Mazin 1989; Sander 1990; Klein 2010, Hugi et al. 2011), Nothosaurus, ?Cymatosaurus and Pistosaurus (Krahl et al. 2009; Klein 2010). In addition, the marine reptiles ichthyosaurs (Buffrénil& Mazin 1990) and mosasaurs (Sheldon 1997) have been studied in this way.
Fossil material cannot be tested experimentally. It is necessary with comparisons to extant animals to uncover the links between growth rate, other factors and bone microstructure (Starck & Chinsamy 2002). Finding the best taxa for comparison to plesiosaurs is not easy, since the closest extant and extinct relatives of Sauropterygia are not known. Extant reptiles can be used, for instance marine turtles (Rhodin 1985; Fig. 2A) and crocodiles (Lee 2004; Klein et al. 2009). But there is no real extant marine counterpart to the plesiosaurs. Both ecologically and possibly physiologically they resemble more extant marine mammals than reptiles (Fig. 2B). Bone growth and body growth The patterns and sequence of ossification in different bones have been investigated in different reptiles and can give valuable information on the biology of the animal studied (e.g., neonatal squamates in Maisano 2001; plesiosaurs, ichthyosaurs and mosasaurs in Caldwell 2002). Being hard, bone cannot grow by inner expansion, only by deposition of new bone to so-called appositional surfaces: outside and inside of the bone, inside canals and in cancellous bone (Enlow 1963). Bone is deposited on the outside from a tissue called the periosteum (derived from the embryonic perichondrium), producing periosteal bone, and on inner surfaces from the endosteum. Endochondral bone is endosteal bone that replaces cartilage (Reid 1996). Two processes act at the same time during the life span of a bone: deposition and resorption, the latter being a process adapted to hard bone tissue that maintains the shape of the bone and remodels it from the inner by removing bone in some areas whilst bone is deposited in others (Enlow 1963). The literature is conflicting about the level of internal remodelling in plesiosaur bone. This information may be important in order to understand how much the plesiosaur limbs changed during ontogeny and to compare them to extant reptiles. In most limb bones, the periosteal tissue is a sheath that constrains the shape of the bone and thus makes the classic dumb-bell shape of long bones (Farnum 2007), while the endochondral forms the inner matrix and the articular faces at the ends, beneath a cartilage cap (Caldwell 1997a). There are, however, large differences as to how the two components work together. In many animals, growth in length is accomplished by endochondral bone growth beneath the epiphyseal cartilage, with resorption from the periosteal side happening at the same time (Enlow 1969), whereas in others the periosteal also plays a role in bone elongation (Farnum 2007). In nonlong bones, endochondral bone is the main component, and through evolution some long bones, for instance in plesiosaurs, completely lose the periosteal component (Caldwell 2002). Determinate growth is growth that ends at a given time and cannot resume, while indeterminate growth is when growth has the potential to continue for the entire life span of the animal. In both cases, rapid growth is common at first followed by a plateau (Farnum 2007). Reasons for reaching a specific size can be mechanical constraints, a need to be a certain size to eat certain prey, or
an energetic limit to growth (Sebens 1987). Indeterminate growth has also been defined as cases where size is correlated with age, but this definition includes invertebrateswhich have more flexible growth patterns (Sebens 1987). In this work the first definition is used. Determinate growth is common in mammals and birds, but exceptions include elephants and rodents (Sebens 1987). Indeterminate growth is by far the most common for fish, non-avian reptiles and amphibians, but signs indicating determinate growth have been found in pteranodons (Bennett 1993), dinosaurs (Curry 1999; Sander 2000) and among extant reptiles such as varanids (Buffrénil et al. 2005) and alligators (Woodward et al. 2011). There has been no study devoted to find out whether plesiosaurs had determinate growth. Another feature of bone growth is whether it is constant or proceeds in cycles. Cyclic growth can be caused by seasonal interruption, most commonly alternating coldhot or wet-dry seasons, cycles of sexual activity, migration or endogenous factors. Such changes in growth can be recorded in the bones, for vertebrates in dermal bones of the head region, vertebral centra and limb bones (Peabody 1961). When cyclic growth gives structural results in the bone, it is often as relatively uniform bands, termed zones or “growth rings”, in the periosteal cortex. This is common in reptiles (Enlow 1969) and such bone is called zonal bone. The zones are bounded by rest lines (often termed lines of arrested growth) if the growth stops and starts, or annuli if they result from periods of faster and slower growth (Reid 1996; Chinsamy-Turan 2005). Zonal patterns are seen mainly in ectotherms, whilst continuous growth is known mainly from endotherms, though exceptions do occur (Reid 1996). In extant reptiles a single zone is usually formed per year (Chinsamy-Turan 2005). “Growth rings” have been used in many cases to establish the age of a single individual (e.g., Bryuzgin 1939), but this has also been contested since rings can have different origins, and the earliest formed rings might disappear during remodelling of the bone. Zonation is reported for plesiosaurs in most summaries on their microstructure (Ricqlès 1976; Wiffen et al. 1995).
The plesiosaur fossils in this study from the 2009 field season (PMO 216.839 and PMO 216.838) were found in the Upper Jurassic deposits from the Sassenfjorden area of central Spitsbergen, Svalbard. They were found in the Adventdalen Group, Janusfjellet Subgroup, Agardhfjellet Formation, Slottsmøya Member (Dypvik et al. 1991), which is dated as MidVolgian (Fig. 3). Two specimens found in 1976 are also included in the study (juvenile PMO 220.401 and adult PMO 218.377). The adult was found at Agardhfjellet, also dated as MidVolgian, while the juvenile has been claimed to be Cretaceous, but both this and its locality is uncertain.
The geology and paleontology of Svalbard, and of the Agardhfjellet Formation, have been subject to intensive research through many years (e.g., Nagy et al. 1988; Buffrénil & Mazin 1990; Dypvik et al. 1991; Angst et al. 2010; Fig. 3). Extensive, exposed, Jurassic deposits on Spitsbergen, Wilhelmøya and Kong Karls Land are several hundreds metres in thickness (Vajda & Wigforss- Lange 2009). The Jurassic was a warm and humid period globally, with relatively little variation across the latitudes, and ice-free polar regions. The CO2 concentrations in the air were several times higher than they are today. Pangaea became fragmented and this led to a general sea level rise. Sedimentary successions from Scandinavia, Greenland and Svalbard show several transgressive and regressive episodes, which are probably linked to local tectonism (Vajda & Wigforss-Lange 2009). This can be seen in studies of the Janusfjellet Subgroup, also showing that Slottsmøya Member has a maximum flooding interval in the lower half of the member (Reolid et al. 2010). The Barents Shelf was for a long time a stable platform, only to a small degree affected by the rifting farther south. Sea-floor spreading in the MidAtlantic, connected to the break-up of Pangaea, progressed northeastwards during the Jurassic. In the Late Jurassic, tectonism caused rifting between East Greenland and Norway and the development of a continuous rift from the North Sea to the Barents Sea (Vajda & Wigforss-Lange 2009). In the Mid to Late Mesozoic, Svalbard was situated at a latitude of around 70 degrees north (Ditchfield 1997). There was an open sea between Northwest Europe and the Arctic in the Late Jurassic, after it had been closed and reopened again in the Callovian, but tectonic activity and a sealevel fall resulted in faunal provinciality in today’s northwestern Europe and the Arctic (Doré 1991). In the Late Jurassic, several anoxic events occurred in the marine environment of northwestern Europe and the Arctic, resulting in the deposition of organic-rich shales, which are found today throughout the area, from West Siberia to North Alaska (Doré 1991). The Slottsmøya Member consists mainly of dark-grey to black shales (commonly paper shales), with some discontinuous silty, sideritic beds and sideritic and dolomitic concretions (Fig. 3). The organic carbon content varies between 1.5% and 12% (Nagy et al. 1988). Bioturbation is common higher up in the section, and terrigenous remains are preserved in several horizons. Some beds represent possible storm deposits (Collignon & Hammer, 2012). The Agardhfjellet Formation has been interpreted as having been deposited on an open marine shelf with low oxygen conditions, the Slottsmøya Member as a deep shelf with offshore bars (Nagy et al. 1988; Dypvik et al. 1991). The area is argued to have been some hundred kilometres east or south of the shoreline (Dypvik et al. 2002). The sea temperature at this time is not fully known, and estimates from places in Svalbard and Siberia vary between 2°C and 21°C (see discussion in Hammer et al. 2011). The ecosystem seems to have had a relatively low faunal diversity, deduced from ammonite diversity (Wierzbowski et al. 2011). Methane seeps, found by the Spitsbergen Jurassic Research Group, might have influenced the ecosystem at the sea-bottom occasionally in the upper part of the Slottsmøya Member (Hammer et al. 2011). Since 2004, a high-latitude marine fauna has been discovered in this locality and in surrounding areas, consisting of ichthyosaurs, pliosaurs, plesiosaurs, invertebrates in a specialised seep fauna (Hammer et al. 2011), and invertebrates in general (e.g., an interesting echinoderm fauna (Rousseau et al. 2012) and ammonites (Wierzbowski et al. 2011)). Seven field seasons have revealed one of the richest sites for Jurassic marine reptilesworldwide (Hurum et al. 2008).
The first aim of this study has been to describe the inner bone structures of plesiosaurs from Svalbard. Secondly, Wiffen et al. (1995) proposed a hypothesis on ballasting and lightening in plesiosaurian limbs based on microstructure, and wrote that testing of this hypothesis is possible “since paleontological material from the Upper Jurassic or Lower Cretaceous would allow a reconstruction of the ontogenetic trajectory of the closer ancestors of our Upper Cretaceous fossils” (Wiffen et al. 1995: p. 637). The Jurassic plesiosaurs in this study are most suitable. The second aim of this study has thus been to test Wiffen et al.’s conclusions about a possible relationship between ontogeny and a microstructural adaptation for differential buoyancy. Thirdly, even though bone growth is important, little is known of the postnatal limb development in reptiles (Farnum 2007), and very few studies have focused on sauropterygian bone histology (Wiffen et al. 1995; Klein 2010), especially the perichondral bone (Caldwell 1997b). That the propodials of plesiosaurs have an inner cone structure is known, but to gain a clearer evolutionary explanation of the distribution of these structures, and of the distribution of secondary ossification centres in fossil tetrapods, more work is needed (Carter et al. 1998). The third aim of this study has therefore been to reveal whether an investigation of the periosteal and endochondral bone of the Svalbard plesiosaurs can provide more information about the postnatal limb development in plesiosaurs. Material and methods Material In this study, specimens of different ontogenetic stages have been investigated and compared. Since the specimens are not all of the same taxa, and as every bone in a skeleton shows a different growth record, comparison must be treated with caution (Starck & Chinsamy 2002; Chinsamy-Turan 2005). The sample for this study consists of four plesiosaurs from Svalbard. Two were collected in the 2009 field season, PMO 216.839, which is a juvenile Djupedalia engeri (Knutsen et al. 2012b) (Figs. 4A, B and 5B1-3) and PMO 216.838 (Figs. 5A1-3), which is a subadult Colymbosaurus svalbardensis (Knutsen et al. (2012c). The two others were collected in 1976, PMO 220.401, a juvenile sp. indet. (Figs. 4E, F) and PMO 218.377, an adult Colymbosaurus svalbardensis (Figs. 4C, D). The assignment of the bones to taxa and ontogenetic level is based on Knutsen (2012). PMO 216.839 is an almost complete, partly articulated plesiosaur, lacking only the skull, the left hind flipper and the tail. PMO 216.838 was also found partly articulated. The bones from PMO 220.401 were found disarticulated, but associated with several bones from the same specimen, but it is not known whether the bones belong to a fore- or hind limb. The bones from PMO 218.377 originate from an articulated limb that is probably a hind limb. These specimens, together with the thin- sections, are kept at the Natural History Museum, University of Oslo. Propodials and phalanges were used in this study because they are long bones and therefore best retain the record of the bone’s history due to little remodeling. Growth marks might also be easier to observe (Botha & Chinsamy 2000). Mesopodials and the metapodium were included in the study because their microstructure has not been previously studied for plesiosaurs, and to map possible differences to the phalanges. The metapodium from PMO 220.401 showed a ring-like structure when broken in two while being excavated in the field. This feature was not observed in the other bones studied, and because of this, bones from this specimen were included in the study. Comparison between the different specimens must be done with care, since each varies according to both ontogeny and phylogeny. There are also many different factors influencing the visible bone structure in any animal. As in all such palaeontological studies, the limited amount of material makes it impossible to make definite conclusions about any larger group, such as all Jurassic plesiosaurs. For comparison to gross internal structure, a radius from an extant undetermined mysticete (PMO 220.788) and the right humerus from an extant leatherback turtle (PMO 220.789) were used (Fig. 2). These are also kept at the Natural History Museum, University of Oslo. Methods The bones were freed of all matrix and glued together with epoxy glue (Araldite). Casts were made in order to study shapes and provide possible exhibition material before the originals were sectioned. In the two propodials, a longitudinal pre-postaxial cut was made using a water-lubricated slab saw (Diamant Boart) with a diamond blade. One of the halves from each of the bones was then cut longitudinally, but this time dorsoventrally. The result was one half and two quarters of each propodium. The half and one of the quarters were used for a gross internal structure study. This was done by grinding the longitudinal pre-postaxial, inner side of the half, the dorsoventral inner side of the quarter and a transverse section from the last quarter on a lapidary wheel (Struers RotoPol-35) with 120 and 600 grit diamond abrasive discs (Struers MDPiano). The ground sides were then scanned (CanoScan 5600F) using rectified spirit between the fossil bone and the glass. Together the three scans provide a 3D understanding of
The first the gross internal structure in these fossil bones (Fig. 5). The whale and the turtle bones were cut on a band saw (Aigner) in the same way as the plesiosaur bones, resulting in one half and two quarters, together giving a 3D picture. The turtle bone pieces were ground on a wet grinder using 120 grit (Struers Knuth Rotor) before being scanned. The whale radius was photographed (Canon Digital Ixus 95 IS) (Fig. 2). The last quarter from each of the two plesiosaur propodials was used for making thin-sections. Six pieces from different locations in the bone were taken out
(5A1, B1). Transverse sections from the midshaft (neutral) region and longitudinal sections from the articular regions were included to ensure the best available information about both ontogeny, growth in length and structural orientation of trabeculae (Felts & Spurrell 1965; Chinsamy-Turan 2005). Transverse sections might be misinterpreted if seen isolated, due to the gross internal structure. Therefore, thin-sections were made in all three directions (Figs. 5A1, B1). The pieces were cut, ground on a lapidary wheel (Struers Discoplan cutting and grinding machine), impregnated with epoxy (Epo- Fix) and mounted on glass before being ground to the right thickness (Thorlag thin-section machine).
(Figs. Figure 4. Plesiosaur limb bones thin-sectioned in this study. Drawing of the entire limb shows the position of the bones in the limb. Red = PMO 216.839: carpal and phalanx. Blue = PMO 218.377: tarsal and phalanx. Yellow = PMO 220.401: mesopodium and metapodium (shown on hind limb, but this is not certain). Drawing to the right of each photograph shows positions of thin-sections, one transverse and one longitudinal for each bone (Roman numbers and the PMO number together make up the names of the thin-sections). All thin-sections are dorsoventral, except from PMO 220.401V, which is longitudinal pre-postaxial. A. Juvenile plesiosaur phalanx. PMO 216.839. Directions unknown, not found articulated. B. Juvenile plesiosaur carpal. PMO 216.839. Directions unknown, not found articulated. C. Adult plesiosaur phalanx (cast). PMO 218.377. D. Adult plesiosaur tarsal. PMO 218.377. E. Juvenile plesiosaur metapodium. PMO 220.401. F. Juvenile plesiosaur mesopodium. PMO 220.401. Scale bar= 25 mm. 4D courtesy of Espen Madsen Knutsen.
Figure 5. Propodials, gross internal structure. Drawing of entire skeleton show the position of the propodials, red = PMO 216.839 humerus and green= PMO 216.838 femur. Small drawings show position of thin-sections (Roman numbers and the PMO number together make up the names of the thin-sections, for directions see frame to the right in the figure) and interpretation of the photos. A1-3. Subadult plesiosaur femur. PMO 216.838. A1. Longitudinal pre-postaxial cut. Thin-section I and VII are from the same spot, but of different thickness. A2. Dorsoventral cut, half section, dorsal to middle. Distal part mirrored. A3. Transverse view, quarter of section, outer to middle. Section taken from the proximal part of diaphysis. B1-3. Juvenile plesiosaur humerus. PMO 216.839. B1. Longitudinal prepostaxial cut. Thin-section VI and VII are from the same spot, but of different thickness. B2. Dorsoventral cut, half section, dorsal to middle. B3. Transverse view, quarter of section, dorsal to middle. Endochondral part is missing in the material. Section taken from distal part of diaphysis. Scale bar= 25 mm. Abbreviations c=cortex, m=medullary region, n= nutrient artery, y = medullary cavity, z= cartilage cap (estimated size).
Figure 6. Thin-sections from plesiosaur limb bones. Gross internal structure. All sections except A, C, E and K are sections from the middle of the bone to the outer edge. Longitudinal cut have epiphysis pointing upwards. Roman numbers and the PMO number together make up the names of the thin-sections. The pictures are coloured digitally, corresponding to the letter abbreviations for different areas of the bones. A-B. Juvenile plesiosaur PMO 216.839. A1. Carpal transverse section (X). A2. Carpal longitudinal dorsoventral section (XI). B1. Phalanx transverse section (VIII). B2. Phalanx longitudinal dorsoventral section (IX). Note the cone-shape in longitudinal section. C-D. Juvenile plesiosaur 220.401. C1. Mesopodium transverse section (III). C2. Mesopodium longitudinal section (IV). D1. Metapodium transverse section (I). D2. Metapodium longitudinal section (II). E-F. Adult plesiosaur PMO 218.377. E1. Tarsal transverse section (III). E2. Tarsal longitudinal dorsoventral section (IV). F1. Phalanx transverse section (I). F2. Phalanx longitudinal dorsoventral section (II), note the large vascular canals in longitudinal direction from the middle to the epiphysis. Scale bar = 5mm. Abbreviations: c=cortex, m=medullary region, n= nutrient artery, t= area with collapsed trabeculae.
thin-sections made were 30 - 35 μm in thickness and the later ones 50-60 μm. The increase in thickness was done to investigate whether more structures could be detected. From the phalanges, the mesopodial and the metapodium thin-sections were made. A transverse cut was made through the bone, followed by a longitudinaldorsoventral cut through one of the resulting halves. Thin-sections were then made, following the procedure described for the propodials, one from a transverse sectionand one from a longitudinal section for each of the bones, and in addition a longitudinal pre-postaxial section for PMO 220.401 (Fig. 4). The thin-sections were scanned to show the overall structure, using a Nikon Super coolscan 4000 ED scanning device and Nikon Scan 3.1 software (Fig. 6). The thin-sections were then studied under a microscope (LeicaDMLP) on 2-100x magnification and photos were taken using a LeicaDC300 camera connected to the microscope and computer software (ACDSee 6.0) (Figs. 7-10). Articulated limbs from two of the plesiosaur specimens were used for taking measurements of the distance between the propodials, epipodials and mesopodials to obtain information of the possible existence and size of cartilage caps. For PMO 216.839 and PMO 216.838 these measurements were taken on hind limbs. For PMO 218.377 and PMO 220.401 such measurements were not possible.
In bone microstructure research the terminology used by different workers is conflicting and not always well defined. Many studies have used the terminology of Francillon-Viellot and Buffrénil (2001). In this study the microstructure was described using the terminology of Reid (1996) except that “longitudinal canals” is used here for describing what is called “parallel canals” in that work. The term “dumb-bell shape” is used as in Caldwell (2002), meaning bones that are narrowest in the middle part. A dumb-bell shape indicates that the bone has a perichondral tissue (P. Druckenmiller, pers.comm.; Caldwell 2002, Farnum 2007). Institutional abbreviations PMO: Palaeontological Museum Oslo (now a part of the Natural History Museum, University of Oslo). SVB: Svalbard Museum, Longyearbyen. Abbreviations used in the figures c = cortex, d = decomposition, l = lamellar bone, m = medullary region, n = nutrient artery, o = osteocyte lacunae, po = primary osteon, r = resorption cavity, so = secondary osteon, t = area with collapsed trabeculae, v = simple vascular canals, vs = vascular system (unspecified), w = woven bone, y = medullary cavity, z = cartilage cap.
PMO 216.839 Gross internal structure
PMO 216.839 humerus (Figures 5B1-B3) The juvenile humerus has an inner structure of two cones that is visible due to a different overall bone structure. The apices of the cones do not meet, being separated by a small, probably free, medullary cavity. The bones were made up of endochondral bone when the animal was alive. Now, the cone-shaped areas are mainly infilled with clay mineral and have only small isolated patches of collapsed trabeculae left. It is thus clear that the cones consisted of a much weaker building material than the rest of the bone. The distal part is more cancellous than the proximal end. In the epiphysis there is no sign of a secondary ossification centre. The surface of the proximal epiphysis shows several small pits, each a few millimetres in diameter, while the surface of the distal end is not preserved. Outside the cones is a sheath of periosteal bone. It makes a bridge through the bone at around the medullary cavity, separating the two cones completely. The periosteal sheath does not cover the whole length of the bone and covers different amounts on different sides of the bone. The cortex is overall slightly denser than the endochondral area, with the most cancellous parts being the areas distally and laterally. The vascular canal density is quite high. In the endochondral cones, many large canals lead from the medullary cavity longitudinally towards the distal end of the bone. The vascular canals in the proximal part have no such pattern. In the periosteal, the canals run longitudinally in the distal part and with no pattern in the proximal part. In the middle of the anterior periosteal, a large canal intrudes the bone towards the area around the medullary cavity. This is probably the nutrient artery. Microstructure PMO 216.839 humerus. (Thin-sections 216.839I-VII. Bone figure 5A1, microstructure figures 7A, B and E-G) The bone is divided in two: the cortex, made from periosteal bone (Fig. 7A), and the medullary region, where the bone is now almost completely replaced by clay minerals in all the thin-sections, mixed with smaller parts of collapsed trabeculae (Fig. 7E), probably from endochondral bone. The periosteal parts are ossified completely. The cortex consists of woven bone and does not contain any growth marks (Fig. 7B). Distally, the cortex is more compact in the outer parts, but proximally the inner parts are more compact. Outermost there is a band containing collapsed trabeculae. The cortex is well vascularised. Close to the proximal
Figure 7. Microstructure PMO 216.839 humerus and carpal. All pictures are taken with plane polarized light unless otherwise specified. A. Humerus cortex. Longitudinal canals, mostly simple vascular. Section 216.839II, 5x. B. Same as A, but with crosspolarised light and gypsum filter. Note woven bone. C. Carpal cancellous cortex of complete trabeculae to the right, collapsed to the left. Section 216.839X, 5x. D. Same as C, but with crosspolarised light and gypsum filter. Section 216839X, 5x. E. Humerus medullary region. Note decomposition on the right side of picture and the large canal leading towards epiphysis at the left. Section 216.839VI, 5x. F. Humerus cortex in dorsoventral cut showing longitudinal canals. Section 216.839V, 2,5x. G. Same as F, but with crosspolarised light and gypsum filter. Section 216.839V, 2,5x. Abbreviations: c = cortex, d = decomposition, po = primary osteon, t = area with collapsed trabeculae, v = simple vascular canals, w = woven bone.
epiphysis the periosteal is more cancellous and and many vascular canals are not oriented although there is a dominance of longitudinal vascular canals. The canals are mostly primary vascular canals not yet surrounded by lamellar bone, in addition to some primary osteons. There is a large size variation among the canals (Fig. 7G). No remodelling seems to have taken place. The bone shows several signs of decomposition, both a narrow band of small dots a few millimetres from the edge and what might be traces of burrowing bacteria or fungi. Microstructure PMO 216.839 carpal. (Thin-sections 216.839X and XI. Bone figure 4B, gross structure figures 6A1 and A2, microstructure figures 7C and D) The carpal does not have a real dumb-bell shape, but one of the lateral sides has a shape resembling those of bones with a periosteal sheath (Fig. 4B). In transverse section no cortex can be seen, but in longitudinal section, the bone shows an inner cone-like structure comparable to the one in the propodial and in the phalanx (Fig. 6A2). In transverse section, the trabecular orientation is very different in the dorsal and ventral part, and between these two parts is a band consisting of collapsed trabeculae and clay mineral infill (Fig. 6A1). The assumed dorsal part has radiating trabeculae, especially on one lateral side (Fig. 7C), while the assumed ventral side shows no orientation of the trabeculae. The dominant bone tissue is woven bone (Fig. 7D). The vascular system is consists mostly of simple vascular canals in a reticular pattern. A few primary osteons are visible in the inner parts of the bone. There are no secondary osteons or other traces of remodelling. Microstructure PMO 216.839 phalanx. (Thin-sections 216.839VIII and IX. Bone figure 4A, gross structure figures 6B1 and B2) The phalanx has a typical dumb-bell shape (Fig. 4A), which means that it has a periosteal sheath around the middle. It does not show a compact cortex, the outer parts are instead cancellous. When seen in longitudinal section, a cone similar to the one in the carpal is visible, but not as well defined. The inner structure of the cone is collapsed. In transverse section a large area consists of collapsed trabeculae and shale as in the longitudinal section (Figs. 6B1, B2). Around the outer edges of the bone, the trabeculae are more complete and the bone is cancellous. Some primary osteons can be seen in the outer parts, together with simple vascular canals and some possible resorption cavities. The main vascular direction seems to be longitudinal. In the cone one primary osteon appears. There are no signs of remodelling or growth marks. As for the carpal and the propodial from this specimen, the preservation is not very good and the sections show several signs of decomposition.
PMO 220.401 MicrostructurePMO 220.401 mesopodium. (Thin-sections 220.401III-V. Bone figure 4F, gross structure figures 6C1 and C2, microstructure figures 8B-D) This bone does not have a dumb-bell shape, indicating that there is no periosteal sheath (Fig. 4F). There is no compact cortex and the bone is porous throughout. The inner structure is three-layered and quite similar to the adult mesopodium (PMO 218.377) with a band in the middle containing small pieces of collapsed trabeculae, and cancellous bone on both sides, but here there are cone shapes in longitudinal section (Figs. 6C1, C2). The inner part of the bone consists of collapsed trabeculae, with clay mineral infill in the epiphysis (Fig. 8B). In one outer area there is an extremely high density of osteocyte lacunae with visible canaliculi (Fig. 8C). The longitudinal pre-postaxial section shows three areas: the outermost, partly decomposed with clay mineral infill, the middle area that is very porous and with no apparent organization, and the innermost area which is more compact with numerous vascular canals (Fig. 8D). The bone also has two large cavities, one probably leading into a medullary cavity. The vascular system consists of simple vascular canals that are not organised in “rings” as observed in the metapodium. There are no traces of remodelling. As for several other bones, degradation of some kind has taken place in the middle of the trabeculae. Microstructure PMO 220.401 metapodium. (Thin-sections 220.401I and II. Bone figure 4E, gross structure figures 6D1 and D2, microstructure figure 8A) The bone has a dumb-bell shape, and thus a periosteal component (Fig. 4E). There is no compact cortex, and the bone is porous throughout (Fig. 8A). In longitudinal section, the bone has a cone inside, consisting of collapsed trabeculae, while the areas around it are almost complete (Fig. 6D2). The epiphysis of the bone has clay mineral infill, like several of the others, with an increasing amount of collapsed bony trabeculae inwards. In the middle part and in five “stripes” leading into it, the trabeculae are collapsed (Fig. 6D1). The most striking feature is that the trabeculae in the outer parts are organised in “rings” (Figs. 6D1, D2 and 8A). In the outer parts in longitudinal section, many osteocyte lacunae are visible. The vascular canals are mostly simple canals and numerous, with a few primary osteons in transverse section closer to the middle. One large canal leads from the outside and into the middle of the bone. There are no signs of remodelling. The trabeculae show the same degradation in the middle (Fig. 8A) as in many other specimens in this study.
Figure 8. Microstructure PMO 220.401 metapodium, PMO 220.401 mesopodium and PMO 218.377 tarsal. All pictures are taken with plane polarized light unless otherwise specified. A. Metapodium cortex. Note the trabecular rings, and that there are smaller cavities furthest out, the large canal and brecciated trabeculae to the left. Section 220.401I. 2,5x. B. Mesopodium medullary region. Epiphysis with collapsed trabeculae and clay mineral infilling. Section 220.401III, 2,5x. C. Mesopodium. One area with especially high density of osteocyte lacunae. Section 220401III, 40x. D. Mesopodium, prepostaxial section. Note small vessels in inner area, higher porosity further out and degradation. Section 220.401V, 2,5x. E. Tarsal. Bone wall of nutrient artery. Section 218.377III, 2,5x. F. Same as E, but with higher magnification. Note remodeling over lamellar bone. Section 218.377III, 10x. Abbreviations: c = cortex, d = decomposition, l = lamellar bone, n = nutrient artery, o = osteocyte lacunae, po = primary osteon, so = secondary osteon, t = area with collapsed trabeculae, v = simple vascular canals.
PMO 216.838 Gross internal structurePMO 216.838 femur. (Figures 5A1-A3) The two endochondral cones are clearly visible, and occupy a relatively larger part of the bone than in the juvenile. The medullary cavity appears between the cones. The preservation is better than in the juvenile, and there are complete trabeculae throughout the endochondral cones to the epiphyses, indicating that ossification is completed. The periosteal sheath does not go all the way to the epiphyses, but covers only the middle third. The periosteal tissue is more compact than in the juvenile and much more compact than the endochondral bone. In the endochondral region, some of the larger vascular canals, especially distally, have a longitudinal direction, but most trabeculae do not define any consistent pattern. There might be a nutrient artery, similar to the one in the juvenile, on the posterior side. Microstructure PMO 216.838 femur. (Thin-sections 216.838I-VII. Bone figure 5A1, microstructure figure 9A-G) Two distinct parts of the bone are visible: the outer, very compact cortex from periosteal bone, and the inner medullary region consisting of cancellous endochondral bone (Figs. 9B, C, D, G). Both of the epiphyses have larger cavities and are more cancellous than sections taken from the shaft. In the proximal epiphysis area, the endochondral cancellous medullary region extends to the articular surface. The outermost millimetres are more compact than the inner, and the most proximal trabeculae all lie parallel to the proximal epiphysis and have experienced very little remodelling (Fig. 9F). The distal epiphysis also has similar trabeculae extending to the epiphysis. It is more compact than the proximal epiphysis. A few larger canals appear in the medullary region, more commonly in the distal half of the bone. One of these has lamellar bone on one side and woven bone on the other. The cavities in the medullary region are large and the trabeculae show clear signs that remodelling has occurred, being reconstructed cancellous bone (Figs. 9C, G). Secondary osteons are abundant, some of them wholly filled, and osteocyte lacunae are very abundant compared to the cortex (Fig. 9C). In the outermost part some decomposition has occurred (Fig. 9E). The cortex varies in thickness depending on where on the propodium the thin-section is taken, due to the inner cone structure. The cortex itself has two regions (Figs. 5A1, A2 and c1 and c2 in Fig. 9A). Woven bone is present in some places in the cortex (Fig. 9D), in addition to larger fibres that in many places form a regular network in the outer parts. The vascularisation is mostly longitudinal, with a few short canals running radially (Figs. 9A, G). Proximal to the medullary cavity the cortex is remarkably more compact and less vacularised than sections distal to the mid-shaft. The outermost part contains several resorption cavities, and a few primary osteons (Fig. 9A). In the inner part of the cortex there are no resorption cavities, but abundant secondary osteons that increase in density inwards, though not making up compact Haversian bone. Some of the secondary osteons are completely infilled, and between them are a few primary osteons (Figs. 9A, B). The outer edge shows signs of decomposition, probably made by bacteria or fungi. There are no growth marks, and the bone is azonal. There is no calcified cartilage in any part of the bone. In the middle of many trabeculae, both in the medullary region and in the cortex, some form of decomposition has taken place , in the medullary region appearing as small bubbles (Fig. 9D). PMO 218.377 Microstructure PMO 218.377 tarsal. (Thin-sections 218.377III and IV. Bone figure 4D, gross structure figures 6E1 and E2, microstructure figures 8E and F) The external shape of the bone might indicate that there is some periosteal bone on one side (Fig. 4D). On the epiphysis, pits where inner vascular canals came out are observed. The inner structure is a middle band with collapsed trabeculae making up approximately half of the width, with cancellous bone on both sides consisting of very remodelled trabeculae (Figs. 6E1,E2). A few osteocyte lacunae are scattered randomly. There is no compact cortex. On the sides, close to the middle of the bone, are some more compact areas. There is no organisation of the vascular direction apart from a few larger longitudinal vessels. Only secondary osteons are visible throughout the bone (Figs. 8E, F). One very large canal with a triangular shape comes in from the side almost in the middle of the bone. This is probably the nutrient artery (Fig. 8E). The walls of this canal are lamellar, originally avascular, but later they have been reconstructed into densely spaced secondary osteons (Fig. 8F). On the other side of the bone than the blood vessel, there is a small, more compact area where the outside shape indicates that there might be periosteal bone. The bone is azonal. In the outer parts some resorption cavities are observed. The longitudinal section also shows some signs of decomposition. Microstructure PMO 218.377 phalanx. (Thin-sections I and II. Bone figure 4C, gross structure figures 6F1 and F2, microstructure figure 10A-I) The dumb-bell shape indicates the presence of periosteal bone tissue (Fig. 4C). The bone has an inner cancellous cone-shaped medullary region and a slightly more
Figure 9. Microstructure PMO 216.838 femur. All pictures are taken with plane polarized light unless otherwise specified. A. Cortex. c1 and c2 shows the two cortex types, c1 more compact and with resorption cavities and c2 further in with osteons. Section 216.838I, 2,5x. B. Transition between cortex and medullary region. Note the remodelling that has been happening in the medullary region and the high density of secondary osteons in the innermost part of the cortex. 216.838II, 2,5x. C. Same as B, but with higher magnification. Note numerous osteocytes in medullary region trabeculae. Section 216.838VII, 5x. D. Same as C, but with crosspolarised light and gypsum filter. Note decomposition in the middle of several trabeculae. Section 216.838VII, 5x. E. Medullary region, distal epiphysis. Note the decomposition. Section 216.838VI, 5x. F. Medullary region, proximal epiphysis. Section 216.838III, 2,5x. G. Transition between cortex and medullary region, longitudinal dorsoventral cut. Note compact cortex and the vascular system. Section 216.838IV, 2,5x. Abbreviations: c = cortex, d = decomposition, m = medullary region, o = osteocyte lacunae, po = primary osteon, r = resorption cavity, so = secondary osteon, vs = vascular system (unspecified), w = woven bone.
Figure 10. Microstructure PMO 218.377 phalanx. All pictures are taken with plane polarized light unless otherwise specified. A. Cortex and medullary region. Note remodelling in the medullary region. Section 218.377I, 2,5x. B. Same as A, but with crosspolarised light. Section 218.377I, 2,5x. C. Medullary region. Note the secondary osteons. Section 218.277I, 10x. D. Same as C, but with crosspolarised light. Section 218.377I, 10x. E. Cortex in longitudinal cut. Section 218.377II. 2,5x. F. Same as E, but with crosspolarised light. Section 218.377II, 2,5x. G. Cortex with possible rest lines, higher magnification. Section 218.377I, 10x. Abbreviations: c = cortex, d = decomposition, m = medullary region, so = secondary osteon.
compact, narrow outer part which probably consists of periosteal bone (Fig. 6F1). The inner medullary region has continuous trabeculae throughout, and turns more cancellous towards the centre. It is strongly remodelled, and the trabeculae have lamellae that surround cavities of different size, of which some are wholly infilled (Figs. 10B, C, D). In longitudinal section, three large longitudinal canals are visible, reaching from the cancellous middle part and all the way to the epiphysis (Fig. 6F2), but apart from these large canals, there is no apparent vascular organisation. On the outside of the bone, small pits can be seen where the large vessels terminate. In transition to the cortex there are several small secondary osteons, commonly infilled. Some of the trabeculae in the medullary region have a core where decomposition has taken place, as seen in several other thin-sections (Fig. 10D). The outer part is like the PMO 218.838 femur containing some resorption cavities, secondary osteons and large, visible fibres (Figs. 10A, F). In the periosteal bone, small secondary osteons are visible, several of them infilled. There are also several larger fibres, without an apparent pattern (Fig. 10F). A few osteocyte lacunae are visible. The epiphysis has in some way been degraded: the outer parts are infilled and partly replaced by clay minerals. Distance between limb bones in articulated specimens The measurements of the distance between the bones in articulated limbs from the same specimens as the thinsectioned bones, show a difference between the juvenile (PMO 216.839) and the subadult (PMO 216.838). The juvenile measures 10-15 mm from the distal epiphysis of the femur to the epipodials, whereas the subadult bones almost touch each other (1-3 mm distance). From the epipodials to the first row of tarsals the distance is 10 mm in the juvenile and 1 mm in the subadult, while there is a greater distance from the first row of tarsals to the next in the subadult than in the juvenile. Measurements of the distance between the different limb bones in articulated specimens show a larger available space for this cartilage in the juvenile PMO 216.839 than for the subadult PMO 216.838, the latter having virtually no space for this between the propodials and the epipodials, but more space between other bones. This observation might be used to determine whether the plesiosaurs had a cartilage cap on the bones and how this was related to age and growth.
Ontogenetic statusThe specimens in this study have been argued to represent two juveniles (PMO 216.839 and 220.401), one subadult (PMO 216.838) and one adult (PMO 218.377). Study of the inner features of the bones confirms this: there are no traces of remodelling in PMO 216.839 and 220.401 (Figs. 7A and 8A), whilst PMO 216.838 and PMO 218.377 both show remodelling by secondary osteons (Figs. 8E and 9B).
Large diversity in inner structure
The bones in this study originate from different species, specimens and position in the skeleton, and thus the gross internal structure is very varied (e.g., Figs. 5, 6 and 11). Some general observations may, however, be stated. In all the long bones, an inner cone structure can be seen, most pronounced in the propodials. The propodials also seem to have a tiny medullary cavity, whereas the rest of the bones are most porous in the medullary region, but with no open cavity. None of the bones has a secondary centre of ossification or any other special structure at the epiphysis. In the propodials there is a clearly defined cortex that is more compact than the medullary region. This is also the case for all the other plesiosaur propodials from Svalbard where this can be observed, but that are not included in this study. This is in contrast to many other marine animals, such as cetaceans, which show a regression or lack compact cortices (Figs. 2B1, B2; Buffrénil & Schoevaert 1988). Many of the bones do show large zones of collapsed trabeculae (e.g., Figs. 7E and 8A). Seen in transverse section, the zone of collapsed trabeculae might result from pressure during fossilisation (Hua & Buffrénil 1996), which is probably the case for some of the three-layered mesopodials. Differences in compaction are not uncommon: among plesiosaur remains in the Oxford Clay, highly compacted and uncompacted bones from the same specimen are found close to each other (Martill 1991). Compaction and deformation are also known from the Slottsmøya member (see e.g. Knutsen et al. 2012a). The structures might, however, reflect the growth pattern, thus fitting the explanation of the long bones with the cone structure.
Inner endochondral conesThe two components of the bone that determine the growth and shape are the periosteal and the endochondral. In particular, bones in some groups, among them several plesiosaur limb bones, the periosteal component is lost altogether (Caldwell 2002). The propodials, the phalanges and the metapodium in this study have a
Figure 11. Phalanges showing circumferential vascular orientation. Phalanges from PMO 219.718. A1. Larger phalanx showing weak "rings". A2. Smaller phalanx with clearly defined "rings" of trabeculae and vascular canals. Scale bar = 5mm. Courtesy of Espen Madsen Knutsen.
periosteal component. Internally, it is clearer in the propodials, but also in the phalanges there is periosteal bone in the outer parts, being more compact than the inner part and containing osteons (Figs. 5 and 6). The propodials, phalanges and metapodium are dumb-bell shaped (Figs. 5 and 4A, C, E), which means that they had periosteal bone (Farnum 2007). It seems that the periosteal bone is almost completely lost in the mesopodials (Figs. 4B, D, F), apart from a small patch on one margin on some of the observed specimens. This is the same as found in another Late Jurassic plesiosaur, Cryptoclidus (Caldwell 1997a). In the long bones, the relationship between the periosteal and the endochondral components make the gross internal structure of two cones (e.g., Figs. 5A1, B1 and 6B2, D2). These cones are important to map in order to understand how such a bone grows, the evolution of growth of long bones (Carter et al. 1998), and a prerequisite for understanding other features of the plesiosaurian limb. Since the cones are shaped as pyramids and the periosteal sheath as a cylinder-shaped box around them, transverse sections taken from different places along the shaft will vary greatly in appearance, unlike crosssections from mammalian tubular bone. This is important when measuring cortex thickness or porosity. With this in mind, surprisingly little work has been done on the gross internal structure of plesiosaur bones, and the inner cones have not been described since the work of Moodie (1916; Figs. 1F-I). The present study describes
the relationship between periosteal and endochondral bone and the possible process behind the cones in some plesiosaur specimens, but further studies are still needed. The structure of two endochondral cones in propodials inside an outer sheath of periosteal bone has been described in several marine animal groups, among them extinct and extant marine mammals (the manateeTrichechuslatiostris, Fawcett 1942; cetaceans, Felts & Spurrell 1965; the dolphin Delphinus delphis, Buffrénil& Schoevaert 1988; early archaeocetes, Madar 1998) and extinct and extant marine reptiles (the leatherback turtle Dermochelys coriacea, Rhodin 1985; the sauropterygians Pachypleurosaurus, Buffrénil & Mazin 1989; and Placodus, Buffrénil & Mazin 1992) (Fig. 2). It has also been observed in plesiosaurs (Kiprijanoff 1883b; Lydekker 1889; Moodie 1908; 1916; Figs. 1B, D, F, G). The cones are interpreted as endochondral bone (Moodie 1908, 1916). In the evolution of the ossified endoskeleton in vertebrates, perichondral ossification evolved before endochondral ossification (Carter et al. 1998). There are two main types of endochondral ossification processes: the well-organised, efficient process in mammals and many extant lizards, and the process in basal tetrapods, dinosaurs, extant crocodiles, turtles and birds. In the last-mentioned group, during prenatal development, perichondral ossification advances towards the bone ends faster than endochondral ossification, unless bone growth is relatively slow (Haines 1969; Carter et al. 1998). This leads to large "cartilage cones" being trapped inside the bone shaft, before erosion of the cartilage and endochondral ossification happens later in development, often quite slowly. It is common that calcified cartilage fragments are "trapped" in trabeculae for a while (Haines 1969). Chondrification and ossification are often decoupled, so that selection can act on them independently, a feature that has led to many different inner bone designs in reptiles (Farnum 2007). Several workers found calcified cartilage in trabeculae in propodials of plesiosaurs. In a juvenile Late Cretaceous elasmosaur, globular calcified cartilage is “the main component” in the medullary region in both humerus and phalanx (Wiffen et al. 1995), and in a subadult elasmosaur and Lütkesaurus (not valid taxa) from the Late Cretaceous, small areas were found in trabeculae in the medullary region (Kiprijanoff 1883a; Fostowicz-Frelik & Gazdzicki 2001). For adult specimens, small remnants have been found in a Late Cretaceous elasmosaur (Wiffen et al. 1995). These observations, even though limited in number and in age, confirm a decreasing amount of calcified cartilage in the endochondral medullary region through ontogeny due to increasing ossification of the bone. The middle of the trabeculae show some form of change in many of the studied bones (Figs. 7E, 8A, 9D and 10D). This is not preserved calcified cartilage, but probably signs of local decomposition in the middle of the trabeculae, proving that they have contained calcified cartilage. In the Svalbard specimens, cones are observed in juvenile, subadult and adult specimens, indicating that they persist through the life of the animal. The only adult plesiosaur for which this has been discussed previously (Moodie 1908) apparently had lost the cones, but this material needs to be reinvestigated. The propodial cones observed in the present study differ between the ontogenetic stages. In the juvenile specimens the cones seem to consist of a weaker material than the surrounding bone. They are often collapsed when found and in thin-sections they consist of scattered pieces of collapsed trabeculae and clay mineral infill (Figs. 7E and 8B). This could be diagenetic, but it is unlikely that diagenesis would give this particular structure of two cones if the bone was subject to a random process. How well vertebrate bones resist compaction during fossilisation depends on internal composition, thickness of the cortical bone and the sediments surrounding it (Martill 1991). This means that areas which are not completely ossified collapse easier than areas consisting of bone. Cartilage is more flexible and elastic than bone, but is less resitant to compression, tension and shear (Liem et al. 2001). The subadult femur has complete trabeculae throughout the medullary region and is not collapsed, probably because it has undergone complete ossification. For the plesiosaurs in this study, the “cartilage cone process” seems very plausible, but with a twist: there are cones in all the long bones, and no free medullary cavity, meaning that the erosion did not happen as described above. This is probably because plesiosaurs were marine animals and as in the case of the leatherback turtle (Fig. 2A1; Rhodin 1985). The propodials both have a small medullary cavity restricted to the middle of the bone, which is relatively smaller in the subadult (Fig. 5A1 compared to 5B1), whilst the other bones seem to have trabeculae throughout (Fig. 6). This is the same as has been reported by other workers . Moodie (1908, 1916) reported that the medullary cavity is small in an embryonic propodium, remaining only partially in subadults and disappearing completely in adults. Some Cretaceous species have medullary cavities filled with trabeculae in their propodials (Kiprijanoff 1883b; Salgado et al. 2007). The ichthyosaur Omphalosaurus had a small medullary cavity, whereas Stenopterygius and Ichthyosaurus did not (Buffrénil & Mazin 1990). Cetaceans and large marine turtles also lack an open medullary cavity, and cancellous bone is widely distributed in most medullary regions of turtle and crocodilian long bones (Figs. 2A1, B1; Enlow 1969). The altered buoyancy in water suppresses the weight-bearing role of the skeleton, and this will alter the histology. The lack of a free medullary cavity might be an epigenetic reaction to an aquatic environment and the mechanism behind it an imbalance between resorption and reconstruction (Buffrénil & Schoevaert 1988; Wiffen et al. 1995). The internal resorption in the dolphinhumerus creates and maintains a complex network of trabeculae, whereas it completely removes them in the tubular bones of terrestrial mammals (Buffrénil & Schoevaert 1988). See further discussion on remodelling below.
Vascularization is the means of transport of blood and nutrients to the bone, and it changes throughout the life of an animal. Embryonic limbs are supplied by a complex network of many blood vessels. During growth, one of these becomes the main artery leading into the limb in the adult, called the brachial (to humerus) or femoral (to femur). It is not derived from the same vessel in every case (Wake 1979). In turtles the brachial splits into the radial and the ulnar artery and then again into digital arteries (Wyneken 2001). In the Svalbard propodials, the main organisation of the vascular system in the cortex is longitudinal, with a few additional canals connecting them (e.g., Fig. 7A, F and 9A, G). The longitudinal pattern is more pronounced distally to the mid-shaft than in the proximal end. In other bones than propodials, the vascular system is less organised. In the literature the descriptions of plesiosaur vascular networks include all possible organisations except from laminar, and it is often unclear which is actually found and which definitions are used. A longitudinal pattern is found in other plesiosaurs, sometimes in the cortexbut more often in the medullary region
(Kiprijanoff1883b; Gross 1934; Wiffen et al. 1995). To overcome the use of subjective and confusing terminology, methods for quantifying bone vascular orientation mathematically can be used (Boef & Larsson 2007). It is easier to reach a conclusion about the amount of vascularisation. All workers in the field have reported on a moderate to high vascularisation of the cortex, which fits with the present material. Without it being calcuated, the juvenile propodial seems more heavily vascularised than the subadult (Figs. 5, 7A and 9A). The abundant vascularisation, also in subadult and adult specimens, is contrary to many reptiles, which often have avascular bone. Dinosaurs are an exception (Currey 2002). In addition to internal vascularisation in the limb bones, the relationship with the rest of the vascular system is important. A large nutrient artery can be seen in the juvenile humerus, probably in the subadult femur, and in the adult tarsal (Figs. 5B1, 6E1 and 8E). It has also been observed in the subadult Cretaceous plesiosaur Ogmodirus martinii (Moodie 1916). Some workers found what they described as four large vascular canals leading into the middle of the bone, in juvenile specimens, a feature that was not observed in adult specimens (Williston 1903; Moodie 1908). In the specimens in this study, some larger canals are visible in the medullary region in the subadult propodial, but not in the same pattern as described by the other workers. In the present material, many of the bones, both propodials, mesopodials and phalanges, and both juvenile and adult bones, show small pits on the external surface of the epiphysis. On the phalanx of PMO 218.377, pits correspond to large vascular canals inside that run longitudinally from the middle of the bone. This has been observed before, in subadult plesiosaurs (Moodie 1916; Fostowicz-Frelik & Gazdzicki 2001), and also in ichthyosaurs from Svalbard (Buffrénil & Mazin 1990). Going through the epiphysis, these blood vessels would enter the cartilage cap, which the plesiosaurs probably had (Kiprijanoff 1883b; Brown 1981). Leatherback turtles show a similar feature. They have a different bone growth in the physeal area from other turtles, and one of the main differences is that there are large vascularised cartilage canals traversing the growth plate and entering the cartilage of the epiphysis (Rhodin 1985). Leatherback turtles are the only turtles and the only extant reptile which show this feature. In the fossil record some protostegid fossil turtles show the structure. Other turtles do not have transphyseal vascularisation in either the juveniles or in adults (A. Rhodin, pers.comm.). Rhodin (1985) hypothesised that the so-called transphyseal vascularisation is not due to large size, but to rapid growth to a large size, where the vascularisation sustained the high metabolic requirements of fast-growing cartilage. The small pits might be an indication of rapid growth in plesiosaurs.
Fibrillar organisation and tissue type
Amprino’s rule states that growth rate influences fibrillar organisation and tissue type, and this has been used to deduce growth rate from bone microstructure. However, this is contested and there is uncertainty about the relationships between growth rate, fibrillar organisation and tissue type (e.g., see review in Cubo et al. 2005; Montes et al. 2010). In areas where it is possible to observe fibrillar organisation in the Svalbard material, there is woven bone (e.g., Figs. 9D and 9B). This has also been found in other plesiosaur propodium and phalanx material (Wiffen et al. 1995; Fostowicz-Frelik & Gazdzicki 2001). The only lamellar bone found is in the adult tarsal (PMO 218.377) around the nutrient artery, and possibly indicates slow growth in that area. Woven bone might indicate fast growth and is often used as proof of relatively rapid growth, in contrast to lamellar bone that grows slower. Not many workers have described the tissue type in plesiosaur bone. Wiffen et al. (1995) found dense Haversian bone in an adult Cretaceous elasmosaur, and Salgado et al. (2007) found even more of it. With the present material this cannot be confirmed. It seems not to have Haversian bone, either because the specimen is a subadult, or because it had yet to evolve. Haversian bone is recognised as an advanced bone tissue (Ricqlès 1976). An interesting question is whether plesiosaurs had fibrolamellar bone, which is considered to be a fast-growing tissue compared to lamellated and zonal bone. A comparison between different humeri morphotypes and corresponding bone histology of Lower Muschelkalk Sauropterygia indicated that plesiosaurs should have fibrolamellar bone (Klein 2010). The material from Svalbard is not really suited to solve the problem, but it seems to have woven bone and primary osteons, as well as having many osteocytes, being well vascularised and mostly azonal, all being characteristics of fibro-lamellar bone. Remodelling In the studied material, the subadult and the adult specimens are clearly more remodelled than the juvenile, which shows no remodelling (Figs. 7A and 8A in contrast to Fig. 8E and 9B). There are secondary osteons in the cortex of the subadult and adult, which is not common in crocodiles and many other reptiles (Chinsamy & Dodson 1995). Ricqlès (1976) stated that osteons found in nothosaurs and plesiosaurs often are only primary even though they look like secondary osteons because the whole cortex is not reconstructed. In the present material (PMO 216.838 and PMO 218.377) there are real secondary osteons, especially in the inner part of the cortex. Secondary osteons in plesiosaurs have also been found in other studies (Kiprijanoff 1883a; Seitz 1907; Wiffen et al. 1995; Fostowicz-Frelik & Gazdzicki 2001; Salgado et al. 2007).
In leatherback turtles there is a lack of internal remodelling allowing the cones to remain well differentiated throughout life (Fig. 2A1; Rhodin 1985). In the Svalbard plesiosaurs, remodelling has happened in the subadult and adult specimens, but not very extensively, so that both the periosteal and the endochondral parts have been kept more or less in what appears to be their original position. Growth pattern An important feature in the present material is the absence of zonation in most of the bones (Figs. 5, 9A, and 10A). This is in contrast to very many other reptile groups and the usual summary of plesiosaur bone microstructure (e.g., Ricqlès 1976; Wiffen et al. 1995). Zonation in plesiosaurs has been reported in Late Cretaceous propodials, ribs (Seitz 1907; Enlow & Brown 1957) and vertebrae (Wiffen et al. 1995; Fostowicz-Frelik & Gazdzicki 2001), whilst other studies report on no cyclic growth in vertebrae and girdle bone (Wiffen et al. 1995; Salgado et al. 2007) and unclear zonal formation in femur (Gross 1934). One juvenile metapodium in this study (PMO 220.401) shows a circumferential vascular orientation with very visible rings made of trabeculae (Figs. 6D1, D2 and 8A). This is observed in two other plesiosaur skeletons from Svalbard: the juvenile Macrotrhachelos larseni SVB 1450 and the juvenile Macrotrhachelos wensaasi PMO 219.718 (Knutsen et al. 2012a) (Fig. 11). It is remarkable that only these three specimens show this feature, whereas the others have unorganized trabeculae throughout. The “rings” might indicate that there is a difference between these specimens and the others, either adaptive or for reasons due to ontogeny, phylogeny or other factors. It is tempting to call these features “growth rings”. But what is usually referred to as growth rings, or more correctly, zonation, appears in a compact periosteal cortex as zones, annuli and lines of arrested growth. Such a cortex can be avascular or vascular, the latter a situation encountered in several dinosaurs (Reid 1996). In the Svalbard specimens, the feature appears in a highly vascularised and porous cortex and resembles more the situation encountered in the leatherback turtle (Rhodin 1985), the placodont Placodus (Buffrénil & Mazin 1992) and the ichthyosaur Omphalosaurus (Buffrénil & Mazin 1990) than zonation in crocodiles, most turtles and mammals such as the manatee (Fawcett 1942; Rhodin 1985). In an adult Placodus specimen (Buffrénil & Mazin 1992) up to eight layers of compact bone without vascular canals were found in a well vascularised periosteal cortex. This is quite similar to the situation in this study but the rings are less well defined due to the presence of more trabeculae in the areas between the marks. It is also similar to the PMO 220.401 in having very small canals in the outermost row. In a Triassic Omphalosaurus from Svalbard (Buffrénil & Mazin 1990), the outer part of the periosteal cortex has a similar stratified appearance. A studied leatherback turtle (Rhodin 1985, fig. 10) had two “growth rings” in a well vascularised cortex and from studying the figure, also an additional circumferential vascular organisation like the previous examples. In these three studies, cyclic growth has been suggested as the reason for the “rings”, but the reason for and the frequency of these cycles are unknown. For all three taxa, rapid growth has been interpreted from other microstructural characters such as woven bone. This need not be a contradiction to cyclic growth, which is known from mammals as well as reptiles. Dinosaurs had fibro-lamellar bone, and some show growth rings, indicating that some dinosaurs had both cyclic and rapid growth (Chinsamy& Dodson 1995). The cyclic growth might, however, be less pronounced than in other animals, and not include complete halts in the growth, only periods of slower growth, where the compact lines are made (Buffrénil& Mazin 1990, 1992). The frequency is suggested to be annual, but this is not certain (Rhodin 1985; Buffrénil& Mazin 1992). The reason for the apparent cyclic growth might be related to the environment that the plesiosaurs encountered in the high-boreal regions around Jurassic Svalbard. Two other ichthyosaurs in the study of Omphalosaurus (Stenopterygius from the Lias of Holzmaden and Ichthyosaurus from the Kimmeridgian of France) do not have any striations and the reason might be the “peculiar conditions in the northern habitat of Omphalosaurus” (Buffrénil & Mazin 1990: p.445). In mammals, annual zonation is known mostly from marine animals that encounter cold conditions and small terrestrial forms from cold regions (e.g., the porpoise Phocoena phocoena, Buffrénil 1982). They typically show closely spaced rest lines in avascular bone. In the beluga whale, there are striations in the periosteal of the radius that resemble the situation in the plesiosaurs in this study, and workers have suggested that water temperature might influence the growth of the bone, but that this needs further investigation (Felts & Spurrell 1966). Migration could also be a possible reason for cyclic growth (Rhodin 1985), and it is possible that some of the plesiosaurs migrated in and out of the sea areas with temperature variation, resulting in the “rings” in some species. Sex might also be a reason for cyclic growth. In the leatherback turtle the growth might slow down due to reproductive effort by females (Rhodin 1985). This is not likely in the case of the Svalbard plesiosaurs, since all the three specimens with “rings” are juveniles. From this, it could seem as if the “rings” were a feature for all juvenile plesiosaurs. There is, however, another juvenile in this study (PMO 216.839) that does not have “rings”, so this is not the case. For PMO 220.401, PMO 219.718 and SVB1450, adult conspecifics are not yet known, so whether these
species had cyclic growth in contrast to the other Svalbard plesiosaurs, or changed their growth pattern during life, cannot be investigated at the moment. The “rings” are found only in small long bones, and propodials from the same specimens would need to be investigated to reveal the real distribution of this feature. In some animals, growth marks are only visible in smaller bones because they have a lower rate of bone deposition (Sander 2000). The rings will also only be visible in bones that have a periosteal component. Another possible reason for the difference in presence of cyclic growth might be endothermy, even though cyclicity is not necessarily connected to ectothermy. For the Svalbard ichthyosaurs, the absence of rings in Stenopterygius and Ichthyosaurus may be because they are younger than the Triassic Omphalosaurus and thus were more advanced towards endothermy (Buffrénil & Mazin 1990). Bernard et al. (2010) carried out geochemical analyses of oxygen isotopes from plesiosaur, ichthyosaur and mosasaur teeth to find out whether these animals could regulate their body temperature. The conclusion was that they were able to maintain a stable, high temperature. This has also been suggested by others because plesiosaurs were large and active predators. The leatherback turtle, which has an inner structure quite similar to plesiosaurs (Fig. 2A1), is unique among marine turtlesbecause it has mechanisms for keeping the body temperature well above that of the sea temperature (James & Mrosovsky 2004). High vascularisation, fibro-lamellar bone, dense Haversian tissue and absence of cyclic marks have been seen as signs of sustained growth, intensive bone-body fluid exchange, high metabolic rates and suggestive of endothermy (see summary in Chinsamy-Turan 2005). Plesiosaurs have several of these chracteristics. This could be seen as signs of increased body temperature, but the discussion around extinct reptile physiology is not solved, and many of the microstructural characters are contested (see summary in Reid 1996; Starck & Chinsamy 2002). Bone structure in aquatic animals There are striking similarities in bone structure in animals secondarily adapted to aquatic life. Many well-documented examples of convergent evolution are known, both for the gross anatomy and the inner organisation of the bone structure (Fig. 2). Two main patterns exist, basically very heavy and very light bones. The inner trabecular architecture of a bone is a highly adaptive structure, and is interpreted as having a clear ecophysiological significance (e.g., dolphins, Buffrénil & Schoevaert 1988; early tetrapods secondary adapted to aquatic life Buffrénil & Mazin 1989). Diaphyseal architecture of archaeocetes and aquatic mammals suggests that diving depth, speed and metabolic rate may influence the structures more than transportation costs, but this needs to be investigated further (Madar 1998). Absence of a medullary cavity, as found in the whale, is thought to strengthen the resistance of the flipper during fast swimming (Cozzi et al. 2009). This might be the reason for the plesiosaurs in this study having either a small or no medullary cavity. Heavy or light bones in tetrapods secondarily adapted to life in water is often thought to contribute to the buoyancy of the animal as a consequence of adaptation to different ecological niches. Animals that live close to the shore and that do not depend on swimming fast commonly show a local or general increase in skeletal mass, by processes called pachyostosis, osteosclerosis or pachyosteosclerosis. This is known for several groups of amphibians, reptiles, birds and mammals such as the manatee, and is often connected to a heterochronic development where calcified cartilage persists and remodelling is inhibited. Very light bones occur in animals that live in the open sea and use rapid and sustained swimming. Their bones are very porous, because compact bone is replaced by cancellous tissue in a process called osteoporosis. This specialisation is found in many cetaceans (Figs. 2B1, B2), crocodiles, turtles (Figs. 2A1, A2), frogs and ichthyosaurs (Hua & Buffrénil 1996; summary in Ricqlès & Buffrénil 2001). Wiffen et al. (1995) concluded that plesiosaur bones change porosity during life, from a ballasting adaptation in juveniles to very porous bones in adults, a view first emphasised by Williston (1903). One of the findings from the present work is that this conclusion is not valid for all plesiosaurs. Since all of the specimens in this study are not conspecific, it is not possible to be sure that the juvenile PMO 216.839 will not become more porous through ontogeny, but the PMO 216.838 material proves that not all adult plesiosaurs have bone that is entirely porous (Figs. 5 and 9A, B, G). Caution has to be taken if one intends to calculate porosity from the type or amount of cavities in an aquatic animal bone, and from that conclude on buoyancy or the amount of blood supply. The cavities not only contained blood vessels in the life of the animal, but also nerves, the lymph system and other connective tissue (Starck& Chinsamy 2002), and marrow cavities in the medullar region contain water, ash, protein and lipids in marine animals (Higgs et al. 2011). Marine reptiles must have had marrow in the long bones, and possibly lipids for buoyancy (Kiprijanoff 1883b; Sheldon 1997; Kaim et al. 2008). Pyrite framboids found in fossil plesiosaur and whale bones have been seen as evidence for lipid content, but this is contested (see summary in Kiel 2008). Kihle et al. (2012) describe possible lipids in pliosaur bones from the Slottsmøya member.
From this study it can be concluded that:
• Many features in the present material are often found in animals with rapid growth and high metabolism, including secondary osteons, high vascularisation, pits on the outside of the epiphysis, and woven and possibly fibro-lamellar bone.
• Plesiosaur inner bone structure is most similar to extant marine reptiles that have elevated body temperature and are active, such as the leatherback turtles; to large marine mammals, also secondarily adapted to life in water; and to advanced reptiles. • The long bones have two inner endochondral cones because the endochondral ossification proceeded slower than the periosteal. In juvenile bones the endochondral ossification is not completed, and the trabeculae in the cones thus collapsed easier during fossilisation.
• The juvenile cortex has mostly simple primary vascular canals and some primary osteons, whereas the subadult and adult have secondary osteons in the inner cortex. This shows that the bones were well vascularised throughout life, and that remodelling happened while the bones were growing, even though the remodelling was not enough to destroy the cone structure.
• In some of the plesiosaur bones there are “rings” of trabeculae in the porous periosteal cortex. This is not what is normally called zonation, but might well reflect cyclic growth due to either temperature, migration, or a combination of these. It might also reflect different growth patterns in juveniles and adults.
• Whether plesiosaurs had determinate growth or not cannot be concluded on the basis of this material.
• Not all plesiosaurs underwent a transition to a very porous bone structure through ontogeny. More research is needed to understand the mechanical properties of plesiosaur limbs, both in terms of porosity and buoyancy caused by the possible presence of lipids.
Acknowledgements – Salahalldin Akhavan is warmly thanked for making the thin-sections used in this study. Preparation and casting of the fossils was done by May-Liss Knudsen Funke, Lena Kristiansen and Bjørn Lund. David Bruton gave valuable comments on an earlier version of the manuscript. Patrick Druckenmiller and Espen Madsen Knutsen are thanked for their endless knowledge of plesiosaurs; Anusuya Chinsamy-Turan for tutoring LL into the world of histology; And Hans Arne Nakrem for lots of positive feedback and stratigraphic knowledge. Nicole Klein and Eric Buffetaut are kindly acknowledged for their constructive comments, which considerably improved the quality of this paper.
Angst, D., Buffetaut, E., Tong, H. & Tabouelle, J. 2010: An ichthyosaur skull from the Late Jurassic of Svalbard. Bulletin de la Société géologique de France 181, 453-458. Bennett, S.C. 1993: The ontogeny of Pteranodon and other pterosaurs.Paleobiology 19, 92-106. Bernard, A., Lécuyer, C., Vincent, P., Amiot, R., Bardet, N., Buffetaut, E., Cuny, G., Fourel, F., Martineau, F., Mazin, J.-M. & Prieur, A. 2010: Regulation of body temperature by some Mesozoic marine reptiles. Science 328, 1379-1382. Botha, J. & Chinsamy, A. 2000: Growth patterns deduced fromt the bone histology of the cynodonts Diademodon and Cynognathus. Journal of Vertebrate Paleontology 20, 705-711. Boef, M. & Larsson, H.C.E. 2007: Bone microstructure: quantifying bone vascular orientation. Canadian Journal of Zoology 85, 63-70. Brown, D.S. 1981: The English Upper Jurassic Plesiosauroidea (Reptilia) and a review of the phylogeny and classification of the Plesiosauria. Bulletin of the British Museum (Natural History) 35, 253- 347. Bryuzgin, V.L. 1939: A procedure for investigating age and growth in Reptilia. Comptes Rendus (Doklady) de l’Academie des Sciences de l’URSS 23, 403-405. Buffrénil, V. 1982: Données préliminaires sur la présence de lignes d’arrêt de croissance périostiques dans la mandibule du marouin commun, Phocoena phocoena (L.), et leur utilisation comme indicateur de l’âge. Canadian Journal of Zoology 60, 2557-2567. Buffrénil, V., Ineich, I. & Bohme, W. 2005: Comparative data on epiphyseal development in the family Varanidae. Journal of Herpetology 39, 328-335. Buffrénil, V. & Mazin, J.-M. 1989: Bone histology of Claudiosaurus germaini (Reptilia, Claudiosauridae) and the problem of pachyostosis in aquatic tetrapods. Historical Biology 2, 311-322. Buffrénil, V. & Mazin, J.-M. 1990: Bone histology of the ichthyosaurs: comparative data and functional interpretation. Paleobiology 16, 435-447. Buffrénil, V. & Mazin, J.-M. 1992: Contribution de l’histologie osseuse a l’interpretation paléobiologique du genre Placodus Agassiz, 1833 (Reptilia, Placodontia). Revue de Paléobiologie 11, 397-407. Buffrénil, V. & Schoevaert, D. 1988: On how the periosteal bone of the delphinid humerus becomes cancellous: Ontogeny of a histological specialization. Journal of Morphology 198, 149-164. Buffrénil V. & Francillon-Vieillot H. 2001: Ontogenetic changes in bone compactness in male and female Nile monitors (Varanus niloticus). Journal of Zoology 254, 539-546. Caldwell, M.W. 1997a: Limb osteology and ossification patterns in Cryptoclidus (Reptilia: Plesiosauroidea) with a review of Sauropterygian limbs. Journal of Vertebrate Paleontology 17, 295-307. Caldwell, M.W. 1997b: Modified perichondral ossification and the evolution of paddle-like limbs in ichthyosaurs and plesiosaurs. Journal of Vertebrate Paleontology 17, 534-547. Caldwell, M.W. 2002: From fins to limbs to fins: Limb evolution in fossil marine reptiles. American Journal of Medical Genetics 112, 236- 249. Carpenter, K., Sanders, F., Reed, B., Reed, J. & Larson, P. 2010: Plesiosaur swimming as interpreted from skeletal analysis and experimental results. Transactions of the Kansas Academy of Science 113, 1-34. Carter, D.R., Mikic, B. & Padian, K. 1998: Epigenetic mechanical factorsin the evolution of long bone epiphyses. Zoological Journal of the Linnean Society 123, 163-178. Cheng, Y-N, X.-c. Wu, & Q. Ji. 2004. Triassic marine reptiles gave birth to live young. Nature 432, 383–386. Chinsamy-Turan, A. 2005: The microstructure of dinosaur bone. The Johns Hopkins University Press, Baltimore, 195 pp. Chinsamy, A. 1990: Physiological implications of the bone histology of Syntarsus rhodensiensis (Saurischia: Theropoda). Palaeontologia Africana 27, 77-82. Chinsamy, A. 1997: Assessing the biology of fossil vertebrates through bone histology. Palaeontologia Africana 33, 29-35. Chinsamy, A. & Dodson, P. 1995: Inside a dinosaur bone. American Scientist 83, 174-180. Collignon, M. & Hammer, Ø. 2012: Lithostratigraphy and sedimentology of the Slottsmøya Member at Janusfjellet, Spitsbergen: evi- 308 L. Liebe & J. H. Hurum, A.d.A. & Parsons, T.S. (eds.): Morphology A, 81-116. Biology of the Reptilia, 1. Academic Press, London. Hammer, Ø., Nakrem, H.A., Little, C.T.S., Hryniewicz, K., Sandy, M.R., Hurum, J.H., Druckenmiller, P., Knutsen, E.M. & Høyberget, M. 2011: Hydrocarbon seeps from close to the Jurassic-Cretaceous boundary, Svalbard. Palaeogeography, Palaeoclimatology, Palaeoecology 306, 15-26. Hasse, C. 1878: Die fossilen Wirbel. Morphologische Studien. Die Histologie fossiler Reptilwirbel. Morphologisches Jahrbuch 4, 480-502. Higgs, N.D., Little, C.T.S. & Glover, A.G. 2011: Bones as biofuel: a review of whale bone composition with implications for deep-sea biology and palaeoanthropology. Proceedings of the Royal Society B 278, 9-17. Horner, J.R., de Ricqlès, A. & Padian, K. 2000: Long bone histology of the hadrosaurid dinosaur Maiasaura peeblesorum: Growth dynamics and physiology based on an ontogenetic series of skeletal elements. Journal of Vertebrate Paleontology 20, 115-129. Hua, S. & de Buffrénil, V. 1996: Bone histology as a clue in the interpretation of functional adaptations in the Thalattosuchia (Reptilia, Crocodylia). Journal of Vertebrate Paleontology 16, 703-717. Hugi J., Scheyer T.M.,Sander P. M., Klein, N. & Sanchez-Villagra, M.R. 2011: Long bone microstructure gives new insights into the life of pachypleurosaurids from the Middle Triassic of Monte San Giorgio, Switzerland/Italy. Comptes Rendus Palevol 10, 413-426. Hurum, J.H.H., Druckenmiller, P., Knutsen, E.M. & Nakrem, H.A. 2008: A preliminary report on a new Late Jurassic marine vertebrate Lagerstätte from Svalbard, Norway (abstract). 33rd International Gelological Congress, Oslo, HPF-01. James, M.C. & Mrosovsky, N. 2004: Body temperatures of leatherback turtles (Dermochelys coriacea) in temperate waters off Nova Scotia, Canada. Canadian Journal of Zoology 82, 1302-1306. Kaim, A., Kobayashi, Y., Echizenya, H., Jenkins, R.G. & Tanabe, K. 2008: Chemosynthesis-based associations on Cretaceous plesiosaurid carcasses. Acta Palaeontologica Polonica 53, 97-104. Kiel, S. 2008: Fossil evidence for micro- and macrofaunal utilization of large nekton-falls: Examples from early Cenozoic deep-water sediments in Washington State, USA. Palaeogeography, Palaeoclimatology, Palaeoecology 267, 161-174. Kihle, J., Hurum, J.H. & Liebe, L. 2012: Preliminary results on liquid petroleum occurring as fluid inclusions in intracellular mineral precipitates in the vertebraa of Pliosaurs funkei. Norwegian Journal of Geology 92, 341-352. Kiprijanoff, W. 1883a: Studien über die fossilen Reptilien Russlands. III. Theil. Gruppe Thaumatosauria N. aus der Kreide-Formation und dem Moskauer Jura. Mémoires de L’académie impériale des Sciencesde St-Petersbourg 31, 1-57. Kiprijanoff, W. 1883b: Studien über die fossilen Reptilien Russlands. IV. Theil. Ordnung Crocodilina Oppel. Indeterminirte fossile Reptilien. Mémoires de L’académie impériale des Sciences de St- Petersbourg 31, 1-29. Klein, N. 2010: Long bone histology of Sauropterygia from the Lower Muschelkalk of the Germanic Basin provides unexpected implications for phylogeny. PLoS ONE 5, 1-25. Klein, N., Scheyer, T. & Tütken, T. 2009: Skeletochronology and isotopicanalysis of a captive individual of Alligator mississippiensis Daudin, 1802. Fossil Record 12, 121-131. Knutsen, E.M., Druckenmiller, P.S. & Hurum, J.H. 2012a: Two new species of long-necked plesiosaurians (Reptilia: Sauropterygia) from the Upper Jurassic (Middle Volgian) Agardhfjellet Formation of central Spitsbergen. Norwegian Journal of Geology 92, 187-212. Knutsen, E.M., Druckenmiller, P.S. & Hurum, J.H. 2012b: A new plesiosauroid (Reptilia: Sauropterygia) from the Agardhfjellet Formation (Middle Volgian) of central Spitsbergen. Norwegian Journal of Geology 92, 213-234. Knutsen, E.M., Druckenmiller, P.S. & Hurum, J.H. 2012c: Redescription and taxonomic clarification of “Tricleidus” svalbardensis based on new material from the Agardhfjellet Formation (Middle Volgian). Norwegian Journal of Geology 92, 175-186. dence for a condensed section. Norwegian Journal of Geology 92, 89-101. Cozzi, B., Mazzariol, S., Podestá, M. & Zotti, A. 2009: Diving adaptations of the cetacean skeleton. The Open Zoology Journal 2, 24-32. Cruickshank, A.R.I., Martill, D.M. & Noè, L.F. 1996: A pliosaur (Reptilia, Sauropterygia) exhibiting pachyostosis from the Middle Jurassic of England. Journal of the Geological Society 153, 873-879. Cubo, J., Ponton, F., Laurin, M., de Margerie, E. & Castanet, J. 2005: Phylogenetic signal in bone microstructure of sauropsids. Systematic Biology 54, 562-574. Currey, J.D. 2002: Bones: structure and mechanics. 2nd edition. Princeton University Press, Princeton, 436 pp. Curry, K.A. 1999: Ontogenetic histology of Apatosaurus (Dinosauria: Sauropoda): New insights on growth rates and longevity. Journal of Vertebrate Paleontology 19, 654-665. Ditchfield, P.W. 1997: High northern palaeolatitude Jurassic- Cretaceouspalaeotemperature variation: new data from Kong Karls Land, Svalbard. Palaeogeography, Palaeoclimatology, Palaeoecology 130, 163-175. Doré, A.G. 1991: The structural foundation and evolution of Mesozoic seaways between Europe and the Arctic. Palaeogeography, Palaeoclimatology, Palaeoecology 87, 441-492. Druckenmiller, P.S. 2006: Early Cretaceous plesiosaurs (Sauropterygia: Plesiosauria) from northern Alberta: palaeoenvironmental and systematic implications. Unpublished Ph.D. thesis, University of Calgary, 558 pp. Druckenmiller, P.S. & Russell, A.P. 2008: A phylogeny of Plesiosauria (Sauropterygia) and its bearing on the systematic status of Leptocleidus Andrews, 1922. Zootaxa 3, 1-120. Dypvik, H., Hakansson, E. & Heinberg, C. 2002: Jurassic and Cretaceous palaeogeography and stratigraphic comparisons in the North Greenland-Svalbard region. Polar Research 21, 91-108. Dypvik, H., Nagy, J., Eikeland, T.A., Backerowe, K., Andersen, A., Haremo, P., Bjaerke, T., Johansen, H. & Elverhoi, A. 1991: The Janusfjellet subgroup (Bathonian to Hauterivian) in central Spitsbergen - a revised lithostratigraphy. Polar Research 9, 21-43. Enlow, D.H. 1963: Principles of bone remodeling. Evans, F.G. (ed.). American Lecture Series, 531. Charles C Thomas, Springfield, 131 pp. Enlow, D.H. 1969: The bone of reptiles. In Gans, C., Bellairs, A.d.A. & Parsons, T.S. (eds.): Morphology A, 45-80. Biology of the Reptilia, 1. Academic Press, London. Enlow, D.H. & Brown, S.O. 1957: A comparative histological study of fossil and recent bone tissues. Part II. Texas Journal of Science 9, 186-214. Farnum, C.E. 2007: Postnatal growth of fins and limbs through endochondral ossification. In Hall, B.K. (ed.): Fins into limbs Evolution, development, and transformation, 118-151. The University of Chicago Press, Chicago. Fawcett, D.W. 1942: The amedullary bones of the Florida manatee (Trichechus latirostris). American Journal of Anatomy 71, 271-309. Felts, W.J.L. & Spurrell, F.A. 1965: Structural orientation and density in cetacean humeri. American Journal of Anatomy 116, 171-204. Felts, W.J.L. & Spurrell, F.A. 1966: Some structural and developmental characteristics of cetacean (Odontocete) radii. A study of adaptive osteogenesis. American Journal of Anatomy 118, 103-134. Fostowicz-Frelik, L. & Gazdzicki, A. 2001: Anatomy and histology of plesiosaur bones from the late Cretaceous of Seymour Island, Antarctic Peninsula. Palaeontologia Polonica 60, 7-32. Gren, J. 2010: Osteo-histology of Mesozoic marine tetrapods – implications for longevity, growth strategies and growth rates. Unpublished bachelor thesis, Lund University, 14 pp. Gross, W. 1934: Die Typen des mikroskopischen Knochenbaues bei fossilen Stegocephalen und Reptilien. Anatomy and Embryology 103, 731-764. Haines, R.W. 1938: The primitive form of epiphysis in the long bones of tetrapods. Journal of Anatomy 72, 323-343. Haines, R.W. 1969: Epiphyses and sesamoids. In Gans, C., Bellairs, Gross internal structure and microstructure of plesiosaur limb bones from the Late Jurassic 309 Tétrapodes. VII. - Sur la classification, la signification fonctionelle et l’histoire des tissus osseux des Tétrapodes. (Deuxième partie, fin.). Annales de Paleontologie Vertebres 63, 133-160. Ricqlès, A. & Buffrénil, V. 2001: Bone histology, heterochronies and the return of tetrapods to life in water: were are we? In Mazin, J.-M. & de Buffrénil, V. (eds.): Secondary Adaptation of Tetrapods to Life in Water, 289-310. Verlag Dr. Friedrich Pfeil, München. Rousseau, J., Nakrem, H.A. & Hammer, Ø. 2012: A Late Jurassic boreal echinoderm lagerstätte from Janusfjellet, central Spitsbergen, arctic Norway. Norwegian Journal of Geology 92, 132-148. Salgado, L., Fernández, M. & Talevi, M. 2007: Observaciones histológicas en reptiles marinos (Elasmosauridae y Mosasauridae) del Cretácio Tardío de Ptatgonia y Antártida. Ameghiniana 44, 513-523. Sander, P.M. 1990: Skeletochronology in the small Triassic reptile Neusticosaurus. Annales des Sciences Naturelles-Zoologie 11, 213- 217. Sander, P.M. 2000: Longbone histology of the Tendaguru sauropods: implications for growth and biology. Paleobiology 26, 466-488. Sebens, K.P. 1987: The ecology of indeterminate growth in animals. Annual Review of Ecology and Systematics 18, 371-407. Seitz, A.L.L. 1907: Vergleichende Studien über den mikroskopischen Knochenbau fossiler und rezenter Reptilien und dessen Bedeutung für das Wachstum und Umbildung des Knochengewebes im Allgemeinen. Nova Acta Abh der Kaiserl Leop-Carol Deutschen Akademie der Naturforscher 87, 229-370. Sheldon, A. 1997: Ecological implications of mosasaur bone microstructure. In Callaway, J.M. & Nicholls, E.L. (eds.): Ancient marine reptiles, 333-354. Academic Press, San Diego. Starck, J.M. & Chinsamy, A. 2002: Bone microstructure and developmental plasticity in birds and other dinosaurs. Journal of Morphology 254, 232-246. Street, H.P. & O’Keefe, F.R. 2010: Evidence of pachyostosis in the cryptocleidoid plesiosaur Tatenectes laramiensis from the Sundance Formation of Wyoming. Journal of Vertebrate Paleontology 30, 1279-1282. Thewissen, J.G.M. & Taylor, M.A. 2007: Aquatic adaptations in the limbs of amniotes. In Hall, B.K. (ed.): Fins into limbs Evolution, development, and transformation, 310-322. The University of Chicago Press, Chicago. Vajda, V. & Wigforss-Lange, J. 2009: Onshore Jurassic of Scandinavia and related areas. Gff 131, 5-23. Wake, M.H. (ed.). 1979: Hyman’s comparative vertebrate anatomy. 3rd edition. The University of Chicago Press, Chicago, 788 pp. Wierzbowski, A., Hryniewicz, K., Hammer, Ø., Nakrem. H.A. & Little, C.T.S. 2011: Ammonites from hydrocarbon seep carbonate bodies from the uppermost Jurassic -lowermost Cretaceous of Spitsbergen and their biostratigraphical importance. Neues Jarbuch Geologie und Paläontologie Abhandlungen 262, 267-288. Wiffen, J., de Buffrénil, V., de Ricqlès, A. & Mazin, J.-M. 1995: Ontogenetic evolution of bone structure in Late Cretaceous Plesiosauria from New Zealand. Geobios 28, 625-640. Williston, S.W. 1903: North American plesiosaurs. Field Columbian Museum Geological Series 2, 1-177. Woodward, A.S. 1898: Outlines of vertebrate paleontology: For students of zoology. Cambridge University Press, Cambridge, 494 pp. Wyneken, J. 2001: The anatomy of sea turtles. Unpublished Ph.D. thesis, U.S. Department of commerce, 172 pp. Knutsen, E.M. 2012: Late Jurassic plesiosaurians (Sauropterygia – Plesiosauria) from the Agardhfjellet Formation on Svalbard. Unpublished Ph.D. thesis, Faculty of Mathematics and Natural Sciences and Natural History Muesum, University of Oslo, 250 pp. Krahl, A., Sander, P. & Klein, N. 2009: Long bone histology of middle Triassic eusauropterygians (Nothosauria and Pistosauria) and its implications for paraxial swimming. Journal of Vertebrate Paleontology 29, 129A. Lee, A.H. 2004: Histological organization and its relationship to function in the femur of Alligator mississippiensis. Journal of Anatomy 204, 197-207. Liem, K.F., Bemis, W.E., Walker, W.F., Jr. & Grande, L. 2001: Functional anatomy of the vertebrates. An evolutionary perspective. Thomson Learning, Belmont, 703 pp. Lydekker, R. 1889: The orders Ichthyopterygia and Sauropterygia. Catalogue of the fossil Reptilia and Amphibia in the British Museum, 2. 118-149. Britsh Museum, London. Madar, S.I. 1998: Structural adaptations of early archaeocete long bones. In Thewissen, J.G.M. (ed.): The emergence of whales: Evolutionary patterns in the origin of Cetacea, 353-378. Plenum Press, New York. Maisano, J.A. 2001: A survey of state of ossification in neonatal squamates. Herpetological Monographs 15, 135-157. Martill, D.M. 1991: Bones as stones: The contribution of vertebrate remains to the lithologic record. In Donovan, S.K. (ed.): The Processes of Fossilization, 270-292. Belhaven Press, London. Mazin, J.-M. 2001: Mesozoic marine reptiles: an overview. In Mazin, J.-M. & de Buffrénil, V. (eds.): Secondary Adaptation of Tetrapods to Life in Water, 95-117. Verlag Dr Friedrich Pfeil, München. Montes, L., Castanet, J. & Cubo, J. 2010: Relationship between bone growth rate and bone tissue organization in amniotes: first test of Amprino’s rule in a phylogenetic context. Animal Biology 60, 25-41. Moodie, R.L. 1908: Reptilian epiphyses. American Journal of Anatomy 7, 443-467. Moodie, R.L. 1916: The structure and growth of the Plesiosaurian propodial. Journal of Morphology 27, 401-411. Munthe-Kaas, N.R.S. 2011: A study of the propodial morphology on Late Jurassic plesiosaurs from Spitsbergen. Unpublished master thesis, University of Oslo, 57 pp. Nagy, J., Løfaldli, M. & Bäckström, S.A. 1988: Aspects of foraminiferal distribution and depositional conditions in Middle Jurassic to Early Cretaceous shales in eastern Spitsbergen. Abhandlungen der Geologischen Bundesanstalt 41, 287-300. Nopcsa, F. & Heidsieck, E. 1934: Über eine pachyostotische Rippe aus der Kreide Rügens. Acta Zoologica 15, 431-455. O’Keefe, F.R. & Chiappe, L.M. 2011: Viviparity and K-selected life historyin a Mesozoic marine plesiosaur (Reptilia, Sauropterygia). Science 333, 870-873. Peabody, F.E. 1961: Annual growth zones in living and fossil vertebrates. Journal of Morphology 108, 11-62. Pewkliang, B., Pring, A. & Brugger, J. 2008: The formation of precious opal: Clues from the opalization of bone. Canadian Mineralogist 46, 139-149. Reid, R.E.H. 1996: Bone histology of the Cleveland-Lloyd dinosaurs and of dinosaurs in general, Part I: Introduction: Introduction to bone tissues. Brigham Young University Geology Studies 41, 25-71. Reolid, M., Nagy, J. & Rodríguez-Tovar, F.J. 2010: Ecostratigraphictrends of Jurassic agglutinated foraminiferal assemblages as a response to sea-level changes in shelf deposits of Svalbard (Norway). Palaeogeography, Palaeoclimatology, Palaeoecology 293, 184-196. Rhodin, A.G.J. 1985: Comparative chondro-osseous development and growth of marine turtles. Copeia 3, 752-771. Ricqlès, A. 1976: On bone histology of fossil and living reptiles, with comments on its functional and evolutionary significance. In Bellairs, A.d.A. & Cox, C.B. (eds.): Morphology and Biology of Reptiles, 123-150. Academic Press, London. Ricqlès, A. 1977: Recherces Paléohistologiques sur les os longs des
NORWEGIAN JOURNAL OF GEOLOGY
[I apologise for the Bibliography not being better organized but Blogger was being troublesome-DD]