1. Introduction
In the Mesozoic, marine ecosystems were dominated by aquatic reptiles that were descended from terrestrial ancestors, among them plesiosaurs and ichthyosaurs (Kelley & Pyenson, Reference Kelley and Pyenson2015). Ichthyosaur and plesiosaur fossils from the Late Jurassic – Early Cretaceous have been discovered and recovered from localities more or less all around the globe, and have attracted increasing interest in recent years, but several questions remain about their evolution, ecology and geographical distribution (Fischer et al. Reference Fischer, Bardet, Benson, Arkhangelsky and Friedman2016; Foffa et al. Reference Foffa, Young and Brusatte2018).
An increasing number of marine reptiles have been described from high-latitude localities (e.g. Fernández & Talevi, Reference Fernández and Talevi2014; Delsett et al. Reference Delsett, Novis, Roberts, Koevoets, Hammer, Druckenmiller, Hurum, Kear, Lindgren, Hurum, Milàn and Vaida2016; Zverkov & Efimov, Reference Zverkov and Efimov2019). Although Greenland is a huge island with widespread Jurassic and Cretaceous deposits in the eastern part that have been the subject of research for decades, records of marine reptiles are relatively few (Marzola et al. Reference Marzola, Mateus, Milàn and Clemmensen2018). The study of the Jurassic succession and its fossils in Greenland (Fig. 1) goes back to the report of Madsen (Reference Madsen1904), and has since received a lot of attention, so far culminating in the publication of The Jurassic of Denmark and Greenland (Ineson & Surlyk, Reference Ineson and Surlyk2003). The first mention of marine reptile fossils was from Milne Land (Aldinger, Reference Aldinger1935) (Fig. 1b), of fossils that were later formally described as a plesiosaur (Huene, Reference Huene1935; redescribed by Smith, Reference Smith2007). From Milne Land a single Late Jurassic ichthyosaur was collected at Pernaryggen in 1970 (NHMD 74798 at the Natural History Museum of Denmark (Marzola et al. Reference Marzola, Mateus, Milàn and Clemmensen2018)) (Fig. 1b). A single plesiosaur propodial was found in 1998 at Kilen in eastern North Greenland (Fig. 1) (Dypvik et al. Reference Dypvik, Hakansson and Heinberg2002; Milàn, Reference Milàn2009).
(a, b) Simplified geological maps of (a) the Wollston Forland – Kuhn Ø area and (b) Jameson Land – Milne Land area in eastern Greenland. (c) Detailed geological map of the Payer Dal area with position of the vertebrate fossil locality at Kingofjeld (modified from Alsgaard et al. Reference Alsgaard, Felt, Vosgerau, Surlyk, Ineson and Surlyk2003, fig. 4).
Data on marine reptiles from the east coast of Greenland provide significant information from a key area situated between the Boreal areas and the Tethys Sea (present-day Great Britain). Oxfordian marine reptiles are known from the UK, and connected localities in the Tethys, such as Greenland, are important to gain a clearer picture of biogeography and evolution at the time, especially in relation to a suggested climate control over marine reptile distribution (Foffa et al. Reference Foffa, Young and Brusatte2018). The objective of this work is to present a short report of Late Jurassic marine reptiles recently discovered in the Ugpik Ravine Member (Bernbjerg Formation) in NE Greenland. The fossils were found on Kuhn Ø during reconnaissance in 2017 (Figs 1, 2). The assemblage includes plesiosauroid, ichthyosaurian and unidentified fossils. Skeletal remains from Wollaston Forland reported by Maync (Reference Maync1947) are possibly from a correlative stratigraphical level as the new material from Kingofjeld (Fig. 1c).
Oblique view from helicopter towards the east and the eastern flank of Payer Dal and Kingofjeld. Encircled the location of the vertebrate locality.
2. Geological setting
The fossil assemblage described in this paper was found in the Upper Jurassic on Kingofjeld mountain, Kuhn Ø, NE Greenland (Figs 1, 2). The Jurassic succession in the area was deposited in the Wollaston Forland Basin (Surlyk, Reference Surlyk1977, Reference Surlyk1978, Reference Surlyk, Ineson and Surlyk2003). The basin formed during the initial phase of rifting in the Middle Jurassic that reached a maximum in the Volgian. In the Middle Jurassic a peneplaned basement was transgressed by the sea and the first marine deposition took place since the Permian. The westward-tilted fault block controlled the palaeogeographic setting. Peninsulas or islands formed at the emerged block crest towards the east and deposition took place in an embayment with sediment input mainly from the embayment head in the north (Surlyk, Reference Surlyk, Ineson and Surlyk2003). Shallow marine sandstones of the Middle Jurassic Pelion Formation were deposited in high-energy environments directly on the crystalline basement (Maync, Reference Maync1947; Surlyk, Reference Surlyk1977, Reference Surlyk, Ineson and Surlyk2003), though fluvial deposition locally took place in incised valleys (Bastians Dal Formation, Kuhn Ø; Alsgaard et al. Reference Alsgaard, Felt, Vosgerau, Surlyk, Ineson and Surlyk2003). During rising sea level in the Middle and Late Jurassic, deposition forms an overall backstepping pattern.
The Pelion Formation is overlain by the heterolithic Jakobsstigen Formation in central Wollaston Forland (Surlyk, Reference Surlyk1977; Vosgerau et al. Reference Vosgerau, Bojesen-Koefoed, Petersen and Surlyk2000; Surlyk, Reference Surlyk, Ineson and Surlyk2003) and by the Payer Dal Formation in northern Wollaston Forland and on Kuhn Ø (Alsgaard et al. Reference Alsgaard, Felt, Vosgerau, Surlyk, Ineson and Surlyk2003). The Payer Dal Formation is an overall coarsening-upwards unit with heterolithic offshore transition deposits and shoreface sandstone deposits. Both formations are overlain by the Ugpik Ravine Member of the Bernbjerg Formation (Fig. 3). The Ugpik Ravine Member forms a transitional unit from the underlying sand succession to the black, laminated shales of the Bernbjerg Formation deposited in an offshore environment when the depositional system was totally drowned (Surlyk, Reference Surlyk, Ineson and Surlyk2003).
Sedimentological log of the Jurassic succession at Kingofjeld (measured by H Vosgerau in 1994).
The age of the assemblage is early late Oxfordian provided by the associated ammonites. Only a few metres above the base of the Ugpik Ravine Member, an imprint of Decipia lintonensis indicates the Ilovaiskii subzone of the Glosense Zone (Callomon & Birkelund, Reference Callomon and Birkelund1980). An ammonite from the vertebrate-bearing level, and also contained within the sample number 590347, is Amoeboceras transitorium, also indicative of the Ilovaiskii subzone.
3. Material and methods
Fieldwork on Kuhn Ø in July 2017 included the finding of numerous skeletal elements of marine vertebrates within a limited area on the western slope of Kingofjeld towards Payer Dal in southern Kuhn Ø (N 74°44′46.7″, W 20°12′14.3″), NE Greenland (Figs 1, 2). Disarticulated vertebrae have been found by previous workers at the same locality on at least two separate occasions (J Therkelsen and F Surlyk, respectively, pers. comm. 2017). Maync (Reference Maync1947) described the succession at Kingofjeld, but even though his observations are rather detailed he did not find, or at least did not report, any vertebrate fossils.
The assemblage described here was associated with the occurrence of numerous eye-catching, large, belemnite rostra, whereas the vertebrae were less conspicuous. More than 100 skeletal fragments were collected. Most of the material is contained within one sample number (GEUS 590347), and subsequently specimens within that sample were assigned separate subnumbers (e.g. 590347-1). Figured specimens (Fig. 4) are stored in the Natural History Museum of Denmark (NHMD) type collection and have thus been assigned with NHMD numbers.
Selected illustrated plesiosaur (a, b) and ichthyosaur (c–f) specimens. Cervical vertebra (NHMD 608568) in (a1) anterior or posterior and (a2) ventral view. Distal limb element (NHMD 608570) in (b). Cervical vertebrae (NHMD 608564) in (c1) lateral, (c2) anterior and (c3) dorsal view, and (NHMD 608565) in (d1) lateral, (d2) anterior and (d3) dorsal view. Anterior caudal vertebra (NHMD 608566) in (e1) dorsal, (e2) anterior and (e3) lateral view. Posterior caudal vertebra (NHMD 608567) in (f1) lateral, (f2) anterior and (f3) dorsal view.
All specimens were found disarticulated on the surface on a plateau formed by the recessive, relatively fine-grained sandstones of the Ugpik Ravine Member overlying the coarser-grained sandstones of the Payer Dal Formation (Figs 2, 3). Pebbles at the fossiliferous belemnite-vertebrate level indicate some degree of reworking processes during deposition of the Ugpik Ravine Member (Fig. 3). This is also indicated by the fossil assemblage, which contains fragments from numerous specimens suggesting deposition after transport. The material has been subject to long-term weathering, generally resulting in relatively poor preservation. However, most of the skeletal elements are three-dimensional and do not appear compressed. Many elements are too worn or fragmentary to establish any referral, even to animal group. Two of these (GEUS 590347-100 and -101) clearly differ in morphology and microstructure (visible in cross-section) from ichthyosaurs and plesiosaurs, and might represent metacarpals or phalanges from a turtle. However, as there is little and fragmentary material, a formal description and referral is not conducted here. If the fragments originate from a turtle, it might have biogeographical implications as turtles are found in Late Jurassic localities with ichthyosaurs and/or plesiosaurs in, for example, Poland, Argentina and Mexico (Alvarado-Ortega et al. Reference Alvarado-Ortega, Barrientos-Lara, Espinosa-Arrubarrena and Melgarejo-Damián2014; Gasparini et al. Reference Gasparini, Fernández, De La Fuente, Herrera, Codorniu and Garrido2015; Blazejowski et al. Reference Blazejowski, Giesszcz and Tyborowski2016), but not in the Late Jurassic – Early Cretaceous Slottsmøya Member Lagerstätte at Spitsbergen (Delsett et al. Reference Delsett, Novis, Roberts, Koevoets, Hammer, Druckenmiller, Hurum, Kear, Lindgren, Hurum, Milàn and Vaida2016; Fernández et al. Reference Fernández, Vennari, Herrera, Campos, Talevi and de la Fuente2018).
4. Systematic palaeontology
Reptilia Laurenti, Reference Laurenti1768
Ichthyopterygia Owen, Reference Owen1840
Ichthyosauria de Blainville, Reference de Blainville1835
Note: All ichthyosaurs known from the Late Jurassic belong to the family Ophthalmosauridae, except the poorly known lineage of Malawania anachronus (Fischer et al. Reference Fischer, Appleby, Naish, Liston, Riding, Brindley and Godefroit2013; Fernández & Campos, Reference Fernández and Campos2015). The ichthyosaurian remains from Kingofjeld very likely belong to the former family, but cannot be referred with certainty because skull, fin and girdle material is lacking.
4.a. Referred specimens
Twenty-five complete and 44 incomplete vertebrae (GEUS 590347-1–72, except -26, -27 and -28), rib fragments (GEUS 590347 84–98) and a single distal limb element (GEUS 590347-27) (Fig. 4c-f).
4.b. Description
4.b.1. Distal limb element
The element (GEUS 590347-27) might represent a carpal, distal carpal metacarpal or phalanx (McGowan & Motani, Reference McGowan and Motani2003). It is oval as in some ophthalmosaurids such as Ophthalmosaurus icenicus (Moon & Kirton, Reference Moon and Kirton2016) but in contrast to others, e.g. Undorosaurus? kristiansenae (Druckenmiller et al. Reference Druckenmiller, Hurum, Knutsen and Nakrem2012; Zverkov & Efimov, Reference Zverkov and Efimov2019), where distal limb elements are more rectangular. The element is missing parts of its surface and no further description is possible.
4.b.2. Vertebrae (Fig. 4c–f)
The vertebrae are referred to Ichthyosauria because they are deeply biconcave, anteroposteriorly shorter than dorsoventrally tall, with neural arches not fused to the centra. The dorsal surface has facets for the neural arches on each side of the flattened area for the neural canal. The centra bear one or two rib facets on the lateral surface. The morphology of the vertebrae is typical of that seen in Ophthalmosauridae, but no formal criteria are in place for referring single vertebrae to the family. Some of the vertebrae are well preserved, whereas many possess a partly or completely eroded surface and/or margins. Many are broken into two or more pieces, but generally show few signs of compression.
Based on the position of apophyses on the lateral surfaces and the overall shape, 18 of the most complete vertebrae can be assigned to specific regions of the vertebral column (Buchholtz, Reference Buchholtz2001; McGowan & Motani, Reference McGowan and Motani2003) (Table 1). The assemblage includes centra from all regions: cervical (Fig. 4c, d), dorsal (not figured), anterior caudal (Fig. 4e) posterior caudal from the preflexural (Fig. 4f) and anterior fluke regions (not figured). Cervical vertebrae bear two facets on the lateral surface; the dorsalmost one is confluent with the neural arch facet. Two of these centra (GEUS 590347-03 and -06) have a more angled ventral margin in anterior view and appear pentagonal compared to some of the others (GEUS 590347-01 and -08), where the ventral margin is rounded. The vertebra from the dorsal region (GEUS 590347-15) possesses two facets for articulation with the ribs on the lateral surface, in a more ventral position than for the more anteriorly placed vertebrae. Vertebrae interpreted to belong to the anterior caudal region (five centra) only have one apophysis, positioned in the ventral portion of the lateral surface, whereas the posterior caudal and fluke centra are relatively small. Vertebrae that are mediolaterally wider than tall (GEUS 590347-14, -16 and -23) are interpreted to belong to the posterior portion of preflexural stock whereas two centra that are dorsoventrally taller than mediolaterally wide probably belong to the anterior portion of the fluke (Buchholtz, Reference Buchholtz2001).
The 18 ichthyosaur vertebral centra that could be referred to a section of the vertebral column
The largest vertebra (diameter 94 mm) is a dorsal half of a centrum. It is interpreted to be a posterior dorsal or an anterior caudal as no facets are preserved in the dorsal half of the lateral surface. The largest centra in the column are found in the anterior caudal region (Buchholtz, Reference Buchholtz2001), and this specimen is similar in size to large Ophthalmosaurus icenicus specimens (Buchholtz, Reference Buchholtz2001; Moon & Kirton, Reference Moon and Kirton2016).
4.b.3. Ribs
There are no complete ribs, and none of the fragments preserve the medial head or the distal tip. Several of the rib fragments (e.g. GEUS 590347-85, -88, -89 and -91) display an eight-shaped cross-section typical of many thunnosaurians (Sander, Reference Sander2000; McGowan & Motani, Reference McGowan and Motani2003). Exceptions to this include ophthalmosaurid specimens from the Slottsmøya Member Lagerstätte on Spitsbergen (Delsett et al. Reference Delsett, Roberts, Druckenmiller and Hurum2017, Reference Delsett, Druckenmiller, Roberts and Hurum2018), the Russian Acamptonectes densus and the North American Platypterygius americanus (Maxwell & Kear, Reference Maxwell and Kear2010; Fischer et al. Reference Fischer, Maisch, Naish, Kosma, Liston, Joger, Krüger, Pardo-Pérez, Tainsh and Appleby2012). Two of the rib fragments, where the inner structure is visible in cross-section, appear to have trabeculae arranged into rings.
Reptilia Laurenti, Reference Laurenti1768
Sauropterygia Owen, Reference Owen1860
Plesiosauria de Blainville, Reference de Blainville1835
4.c. Referred specimen
GEUS 590347-99, distal limb element (Fig. 4b).
4.d. Description
The surface of the element appears eroded, whereas the element is not compressed to any degree. The element is interpreted to be a metapodial or phalangeal element based on its shape. Its proximal and distal ends are approximately equally anteroposteriorly wide and the middle is constricted. The articular surfaces are convex.
Reptilia Laurenti, Reference Laurenti1768
Sauropterygia Owen, Reference Owen1860
Plesiosauria de Blainville, Reference de Blainville1835
Plesiosauroidea Welles, Reference Welles1943
4.e. Referred specimens
GEUS 590347-83, partial propodial (Fig. 5).
Partial plesiosaur propodial (NHMD 608569) in (a) anterior and (b) dorsal view.
GEUS 590347-26, cervical vertebra (Fig. 4a). GEUS 590347-28, caudal vertebra.
4.f. Description
4.f.1. Propodial
GEUS 590347-83 (Fig. 5) represents the distalmost portion of a pleiosauroid propodial and has three distal facets, as in e.g. Colymbosaurus svalbardensis (Knutsen et al. Reference Knutsen, Druckenmiller and Hurum2012). In contrast, Pliosauridae propodials are more rounded. The dorsoventral height of the element is approximately similar throughout the element, and it has no dorsal or ventral crests. The element is incomplete and severely eroded. The anteriormost and posteriormost facets are directed anterodistally and posterodistally, respectively, while the middle facet is significantly larger than the two others, both in anteroposterior length and in dorsoventral height.
4.f.2. Vertebrae
The two vertebrae are assigned to Plesiosauroidea based on their relative dimensions and shape (Knutsen et al. Reference Knutsen, Druckenmiller and Hurum2012; Roberts et al. Reference Roberts, Druckenmiller, Delsett and Hurum2017). The surface of the pleiosauroid vertebrae is eroded, most severely the surface of the caudal centrum. The cervical vertebra (Fig. 4a) is biconcave and mediolaterally widest in the dorsal half. It bears the broken remains of two neural arches, directed dorsolaterally. The ventral surface shows two small foramina. The caudal vertebra is mediolaterally wider than dorsoventrally tall.
5. Discussion
The marine reptile assemblage found at Kingofjeld occurs at a narrow interval close to the base of the Ugpik Ravine Member of the Bernbjerg Formation. It thus appear to be from the same or approximately the same stratigraphic level as the marine vertebrate fossils reported from northern Wollaston Forland, which gave rise to the place name Sauruspasset (Maync, Reference Maync1947). The fossiliferous level was described as a 0.2 m thick ‘bone bed’ with vertebrae, bones, petrified logs, plant remains and ammonites and bivalves (Maync, Reference Maync1947). It would be desirable to locate Maync’s locality in Sauruspasset, and an attempt was made in 2018. It was unsuccessful due to an unusual extensive snow cover (~90–95%) even in the late summer. It is obvious, however, that the Sauruspasset is positioned in an area where the sand-dominated Payer Dal Formation is overlain by the Bernbjerg Formation, similar to the situation at Kingofjeld.
The sand-dominated Middle to lowermost Upper Jurassic Pelion and Payer Dal Formations are relatively rich in fossils, especially in bivalves, fossil wood and occasional ammonites, whereas no vertebrates have previously been reported. The occurrence of vertebrates at two localities at apparently the same stratigraphic level in the overlying Ugpik Ravine Member (Bernbjerg Formation) suggests a possible enrichment of marine reptiles in the upper Oxfordian. This can be due to generally favourable palaeoenvironmental conditions where marine reptiles thrived, or favourable preservation potential of vertebrate fossils in this narrow interval.
Sedimentologically, the reptile finds from Kingofjeld are associated with a conglomerate with rather large pebbles. The conglomerate represents winnowing during transgression subsequent to a sequence boundary. The conglomerate and enrichment of fossils thus apparently represent a transgressive lag removing finer-grained particles and concentrating larger grain sizes including fossils.
The ichthyosaur vertebrae described here are interpreted as the remains of several specimens. Based on their morphology, 18 of them can be placed with confidence in a specific region of the vertebral column. Within the anterior dorsal and anterior caudal regions, the size variation between the single vertebrae is significantly larger than what is observed in other ophthalmosaurids (Buchholtz, Reference Buchholtz2001; Moon & Kirton, Reference Moon and Kirton2016; Delsett et al. Reference Delsett, Roberts, Druckenmiller and Hurum2017), and larger than possible for one individual. Future exploration for further discoveries of marine reptiles can probably benefit from specifically searching in strata correlative to the assemblages at Kingofjeld and Sauruspasset.
A Late Oxfordian age makes the plesiosaurian fossils contemporaneous with some cryptoclidid plesioaurioids with small skulls and long necks, taxa mainly from the northern hemisphere, but older than the Tithonian specimens from the Slottsmøya Member Lagerstätte on Spitsbergen (Knutsen et al. Reference Knutsen, Druckenmiller and Hurum2012; Roberts et al. Reference Roberts, Druckenmiller, Delsett and Hurum2017). Plesiosaur fossils from Ellesmere Island mentioned by Troelsen (Reference Troelsen1952) are not formally described, but based on his description of the locality, they probably originate from the Tithonian – early Valanginian Deer Bay Formation (S Schneider, pers. comm. 2019).
The ichthyosaur specimens are of the same age as occurrences of the very long-lived and abundant Ophthalmosaurus icencus as well as the earliest records of Arthropterygius chrisorum and an Ophthalmosauria indet. specimen from Cuba (Fernández & Iturralde-Vinent, Reference Fernández and Iturralde-Vinent2000; Maxwell, Reference Maxwell2010; Moon & Kirton, Reference Moon and Kirton2016), whereas they are younger than the Bajocian species Mollesaurus periallus from Argentina (Fernández, Reference Fernández1999). The ichthyosaur specimens from Kingofjeld are older than the Jurassic ophthalmosaurids found in Solnhofen, Slottsmøya Member Lagerstätte, and the area around Undory in Russia, as well as Caypullisaurus bonapartei from the Vaca Muerta Formation in Argentina (Bardet & Fernández, Reference Bardet and Fernández2000; Fernández, Reference Fernández2007; Zverkov et al. Reference Zverkov, Arkhangelsky, Pardo-Pérez and Beznosov2015; Delsett et al. Reference Delsett, Novis, Roberts, Koevoets, Hammer, Druckenmiller, Hurum, Kear, Lindgren, Hurum, Milàn and Vaida2016).
A narrow seaway (NE Atlantic seaway / Greenland–Norwegian Seaway) connected the Boreal areas to the Tethys Sea for parts of the Late Jurassic (Mutterlose et al. Reference Mutterlose, Brumsack, Flögel, Hay, Klein, Langrock, Lipinski, Ricken, Söding, Stein and Swientek2003) (Fig. 6). Present-day Greenland and Norway were partly land, and their main separation did not occur until 58 Ma. Tectonic movement and a sea level fall caused a characteristic provincialism and endemism for many marine taxa, with a clear separation of the Tethys and Boreal fauna, whereas sea level rise in the Valanginian gave more cosmopolitan distributions (Mutterlose et al. Reference Mutterlose, Brumsack, Flögel, Hay, Klein, Langrock, Lipinski, Ricken, Söding, Stein and Swientek2003; Hammer et al. Reference Hammer, Hryniewicz, Hurum, Hoyberget, Knutsen and Nakrem2013). Close skeletal similarities have been observed between Boreal ichthyosaur taxa, and to some extent South American taxa, and a possible differentiation between low- and high-latitude taxa has been suggested (Maxwell, Reference Maxwell2010; Roberts et al. Reference Roberts, Druckenmiller, Sætre and Hurum2014; Zverkov et al. Reference Zverkov, Arkhangelsky, Pardo-Pérez and Beznosov2015). Two Late Jurassic seaways might have connected these localities: an Oxfordian–Kimmeridgian route along the Eastern Pacific / Western America (Fernández & Maxwell, Reference Fernández and Maxwell2012) and a Tithonian route through the Russian areas and the newly established Hispanic corridor between the Eastern Pacific and Tethys (Zverkov et al. Reference Zverkov, Arkhangelsky, Pardo-Pérez and Beznosov2015).
Late Jurassic palaeogeographic map showing localities with Ophthalmosauridae specimens. 1. Canada; 2. Spitsbergen; 3. Greenland; 4. Norway; 5. UK; 6. Russia (Volga river); 7. Poland; 8. Germany; 9. France; 10. Italy; 11. India; 12. Madagascar; 13. Argentina; 14. Mexico; 15. Cuba; 16. USA. Modified from Rees et al. (Reference Rees, Ziegler, Valdes, Huber, Macleod and Wing2000).
The now-abandoned hypothesis that the Late Jurassic saw a decline in Ichthyosauria (Bardet, Reference Bardet1994; Sander, Reference Sander2000; Lingham-Soliar, Reference Lingham-Soliar2003; Benson et al. Reference Benson, Butler, Lindgren and Smith2010) was probably a result of a small number of fossiliferous locations yielding ichthyosaurs, few articulated and/or complete specimens and small datasets for phylogenetic analyses (Maxwell, Reference Maxwell2010). Since 2000, ichthyosaurs from the Late Jurassic have been reported from many new localities in France, Germany, Argentina, Russia, UK, Colombia, Italy, Poland, Mexico, Cuba, Canada, India and Spitsbergen (Buffetaut et al. Reference Buffetaut, Romasson and Tong2003; Buchy, Reference Buchy2010; Gasparini et al. Reference Gasparini, Fernández, De La Fuente, Herrera, Codorniu and Garrido2015; Delsett et al. Reference Delsett, Novis, Roberts, Koevoets, Hammer, Druckenmiller, Hurum, Kear, Lindgren, Hurum, Milàn and Vaida2016; Paparella et al. Reference Paparella, Maxwell, Cipriani, Roncacè and Caldwell2016; Tyborowski, Reference Tyborowski2016; Prasad et al. Reference Prasad, Pandey, Alberti, Fürsich, Thakkar and Chauhan2017) (Fig. 6). Raw numbers and diversity analyses indicate that Late Jurassic ichthyosaur diversity was high, and that the clade was not severely affected by an extinction event at the Jurassic–Cretaceous boundary (Zammit, Reference Zammit2012; Fischer et al. Reference Fischer, Bardet, Benson, Arkhangelsky and Friedman2016).
Author ORCIDs
Lene Delsett 0000-0002-6806-1411, Peter Alsen 0000-0001-6218-9054
Acknowledgements
Jette Halskov is warmly thanked for drafting Figures 1 and 3, and Henrik Vosgerau for use of data from his PhD work. Special thanks to Bent Lindow at NHMD, for collection assistance. Aubrey Roberts and Serjoscha Evers are thanked for discussions on plesiosaurs and turtles. The reviewers Judy Massare and Judith Pardo-Pérez are thanked for comments that helped improve the manuscript.
Conflict of interest
The authors declare that there is no conflict of interest.
