An earliest Triassic age for Tasmaniolimulus and comments on synchrotron tomography of Gondwanan horseshoe crabs

Geological context

The oldest Australian xiphosurid, Tasmaniolimulus patersoni, was found in the Jackey Shale of the Upper Parmeener Supergroup, Tasmania (Bicknell, 2019). This formation is largely composed of cross-bedded quartz and feldspathic sandstones, laminated dark grey shales and thin coal lenses (Pike, 1973). Stratigraphically, the fossil was located near the very top of the formation, ∼3 m below the base of the overlying Ross Formation, exposed alongside a cliff on the Poatina Highway (41°48′05″S, 146°53′06″E; Ewington, Clarke & Banks, 1989; Bicknell, 2019). Based on the lithology, the unit likely represents deposition of lake and river sediments in a non-marine swamp with limited coastal influence (Banks, 1973; Ewington, Clarke & Banks, 1989). While the Jackey Shale at the stratigraphic level of the collection locality lacks age-diagnostic fossils, palynomorphs from other, temporally contiguous sites can be assigned to the Protohaploxypinus microcorpus Zone, equivalent to upper APP6 (see Price, 1997) and restricted to the Griesbachian substage, early Induan (Early Triassic) based on previous studies in the Sydney Basin (Laurie et al., 2016; Mays et al., 2020). This contradicts previous interpretations of latest Permian that used now outdated chronostratigraphic ages for this palynomorph zone. An Early Triassic age is further supported by the vertebrate fauna and macro- and microflora of the Protohaploxypinus samoilovichii Zone from the overlying Ross Formation which pertains to the younger Smithian substage of the Olenekian (Early Triassic; Forsyth, 1984). The presence of abundant latest Permian macroflora at stratigraphic levels below the level of T. patersoni in the Jackey Shale does suggest that, at least at some locations, the formation does extend into the latest Permian (Ewington, Clarke & Banks, 1989). Nonetheless, given the high stratigraphic position of T. patersoni, it appears more likely that this specimen is of Early Triassic age.

Slightly younger is Dubbolimulus peetae, which was collected from the Napperby Formation (previously the “Ballimore Formation”) of the Gunnedah Basin in central New South Wales (Pickett, 1984). The only known specimen, with an associated counterpart, was found just south of Western Plains Zoo, Dubbo (at approximately 32°17′30.8″S 148°34′35.8″E). The Napperby Formation consists of white, fine–medium grain, quartz-rich, ferruginous sandstone with occasional cross bedding. Thin horizons of grey to red-brown shale and minor conglomerate lenses are interbedded with this sandstone. The stratigraphic horizon within which the specimen was found is a red-brown, slightly micaceous shale. This lithology indicates a high-energy braided river system or lacustrine deposit (Tadros, 1993), possibly part of the same Triassic delta system that continues into the Sydney Basin to the east. The finer grained shale horizons likely represent lower-energy conditions which presumably occurred in quiet, cut-off river channels or small ponds. The possible presence of acritarchs (McMinn, unpublished data, 1982; Early Permian-Early Jurassic palynology of DM Mirrie DDH 1, northwest of Dunedoo. Geological Survey of New South Wales, Report GS1982/289) suggest the unit may have experienced a slight coastal influence occasionally. A diverse macroflora assemblage has been described from both the fossil site itself (Pickett, 1984) and a nearby locality (Holmes, 1982) which broadly correlate to the Dicroidium zuberi Zone (Helby, 1973; Helby, Morgan & Partridge, 1987; Retallack, 1977; Retallack, 1980; Helby, Morgan & Partridge, 1987) of the Anisian (earliest Middle Triassic) in the Sydney Basin. Palynomorphs from core within the Dubbo area, at Mirrie DOH I (McMinn, unpublished data, 1982; Early Permian-Early Jurassic palynology of DM Mirrie DDH 1, northwest of Dunedoo. Geological Survey of New South Wales, Report GS1982/289) and Pibbon DOH 1 (McMinn, unpublished data, 1984; Palynology of DM Pibbon DDH 1, Goulburn River-Binnaway area. Geological Survey of New South Wales, Report 84/4, GS1984/052), support this age interpretation with placement in the Aratrisporites parvispinosus Zone which correlates to the middle to upper Dicroidium zuberi Zone (Young & Laurie, 1966). A middle D. zuberi Zone stratigraphic position, which indicates an earliest Anisian age, is most likely given palynomorphs from other locations in the Gunnedah Basin, which suggest an age range between the upper Aratrisporites tenuispinosus Zone and lower Aratrisporites parvispinosus Zone.

Of a similar age is Austrolimulus fletcheri, from Beacon Hill Quarry, near the suburb of Brookvale, Sydney, New South Wales (Riek, 1955). The exact co-ordinates of the original collection site are unknown, but are considered to be 33°45′11.2″S, 151°15′55.5″E; the location of the original quarry. The specimen originates from a 8 m thick shale lens in the Hawkesbury Sandstone. This lens mostly consists of numerous thin, recessive, grey-red mudrock laminations with little bioturbation (Webby, 1970) and small amounts of rippling (Herbert, 1983). Overall, the Hawkesbury Sandstone was likely formed in a vast coastal floodplain made up of high energy braided rivers, scour channels, lakes, and sand dunes (Conaghan, 1980 and references therein). Shale lenses, like those at the A. fletcheri site, likely represent lower-energy regimes consisting of shallow water bodies disconnected from a main river channel as isolated shallow pools of water (Herbert, 1980; Herbert, 1997; Rust & Jones, 1987). None of the diverse fossil fauna and flora found at Brookvale (see Bicknell & Smith, 2021 for a recent overview) are diagnostic for relative age estimation. However, the Hawkesbury Sandstone is well constrained within the Aratrisporites parvispinosus Zone and upper Dicroidium zuberi Zone based on palynomorphs and macroflora (Helby, 1973; Retallack, 1977; Retallack, 1980; Helby, Morgan & Partridge, 1987). Similar to the Napperby Formation, this places it within the Anisian (earliest Middle Triassic), likely the earliest Anisian. Recent high-precision U-Pb CA-TIMS obtained from the Garie Formation, which underlies the Newport Formation and succeeding Hawkesbury Sandstone, is dated to the latest Olenekian (248.23 ± 0.13 Ma and 247.87 ± 0.11 Ma; Metcalfe et al., 2015). This further supports an Anisian age for the Hawkesbury Sandstone as there is an unconformity in the Sydney Basin between Newport Formation and Hawkesbury Sandstone (Helby, 1973; Herbert, 1980).

Victalimulus mcqueeni from Koonwarra Fossil Bed of the Strzelecki Group (Riek & Gill, 1971), is the youngest xiphosurid known from Australia. A single partial specimen was found at a road cutting along the South Gippsland Highway, approximately 2.4 km east of Koonwarra (38°33′48.9″S 145°57′33.9″E). The unit at this location consists of a thick (∼7–8 m) lower and upper feldspathic sandstone bracketing a grey-green, fossiliferous mudstone (Waldman, 1971; Jell & Roberts, 1986). The mudstone is made up of extremely fine alternating layers of a clay- and silt-dominated matrix. A freshwater lacustrine environment was originally suggested for the Koonwarra Fossil Bed, with the finely laminated mudstones representing a rhythmic varve formed under freezing conditions (Waldman, 1971; Waldman, 1973; Waldman, 1984). However, the highly diverse fossil fauna and flora (see overview in Poropat et al., 2018), instead suggests a cold, but not freezing, swamp or a lacustrine environment with seasonal flooding causing overbank-type deposits (Douglas & Williams, 1982; Jell & Roberts, 1986). Presence of the palynomorphs Clavatipollenite hughesii Couper, 1957 and Foraminisporis asymmetricus Dettmann, 1963 from the Koonwarra Fossil Bed, and absence of other palynomorphs from younger zones, indicate an age within the Upper Cyclosporites hughesii subzone (Jell & Roberts, 1986; Seegets-Villiers & Wagstaff, 2016; Korasidis & Wagstaff, 2020; Wagstaff et al., 2020). This places the unit entirely within the Aptian Stage (Early Cretaceous). Fission track dating of volcanoclastic sediments in the Koonwarra Fossil Beds suggests an age of 118 ± 5–115 ± 6 Ma, which correlates to the mid-Aptian (Gleadow & Duddy, 1980; Lindsay, 1982).

Age of Tasmaniolimulus patersoni

The revised earliest Triassic age of Tasmaniolimulus patersoni has important implications for the timing of morphological innovation within Austrolimulidae. Tasmaniolimulus patersoni was originally considered to be of latest Permian age (Ewington, Clarke & Banks, 1989; Lerner, Lucas & Lockley, 2017; Bicknell, 2019; Lamsdell, 2020). This age indicated that the first appearance of hypertrophied genal spines in Austrolimulidae was before the end-Permian extinction (Bicknell, Naugolnykh & Brougham, 2020). However, the revised date shifts the first appearance of this trait to the earliest Triassic. Furthermore, T. patersoni is now either the oldest Triassic austrolimulid, or contemporaneous with Vaderlimulus tricki Lerner, Lucas & Lockley, 2017 and Psammolimulus gottingensis Lange, 1923—taxa that all have overdeveloped genal spine morphologies (Meischner, 1962; Lerner, Lucas & Lockley, 2017; Bicknell, Hecker & Heyng, 2021).

Comments on application of synchrotron tomography to the study of fossil xiphosurids

The SRXT examination of the Australian xiphosurid fossils did not reveal extensive novel anatomy, nor traces of soft tissues. The aforementioned specimens were preserved primarily in sand- and siltstones which limits the preservation potential of fine, delicate structures. This is in contrast to the tomographic and laminographic reconstructions of the xiphosurid described by Zuber et al. (2017) and which was preserved in fine grained, Muschelkalk-type limestones. These sediments tend to preserve soft-bodied anatomical details in exceptional detail (Vía, De Villata & Esteban Cerdá, 1977; Briggs & Gall, 1990; Cartañài Martí, 1994; Klug, Hagdorn & Montenari, 2005). Nonetheless, non-destructive three-dimensional imaging using SRXT will likely continue to play a role in anatomical studies of fossil xiphosurids, following the rapid adoption of this imaging modality across palaeontology. Furthermore, NCT is being used more commonly in palaeontology, owing to the ability of neutrons to penetrate through typically radiopaque minerals such as iron pyrite, a high sensitivity to hydrogenous material, and thus to residual organic remains, (Gee et al., 2019; Gee, Bevitt & Reisz, 2019; Na et al., 2021; Smith et al., 2021; Bazanna et al., 2021), and to increasing availability of high-quality neutron imaging facilities at nuclear research reactors and spallation neutron sources around the world (see https://www.isnr.de/index.php/facilities/user-facilities). Finally, techniques that can more readily distinguish areas with very small differences in radiopacity, such as phase-contrast enhanced imaging, show promise for more detailed examination of muscles and other internal structures in suitably well-preserved specimens. Any, or all of these approaches could be applied to the study of specimens of Mesolimulus walchi (Desmarest, 1822) from the Nusplingen Lithographic Limestone (Upper Jurassic, Kimmeridgian), Germany, that have muscle traces preserved under the prosoma (Briggs et al., 2005). Muscle traces have also been described from specimens of Euproops danae from the Upper Pennsylvanian (Virgilian) Lawrence Formation, Kansas (Feldman et al., 1993; Babcock & Merriam, 2000; Bicknell et al., 2022b). Further examination of the Lawrence Formation specimens would determine if the muscles exhibit moldic preservation—as is common for Mazon Creek fossils (Clements, Purnell & Gabbott, 2019; Bicknell et al., 2021c)—or if there are additional, unexpressed anatomical features. The collection of novel soft anatomy from these and other fossil xiphosurids are vitally important in presenting and revising hypotheses regarding homology with extant xiphosurids (sensu Briggs et al., 2005; Bicknell et al., 2022b) and resolving conflicts between phylogenetic hypotheses (e.g., Ballesteros & Sharma, 2019; Bicknell, Lustri & Brougham, 2019; Bicknell, Naugolnykh & Brougham, 2020; Lamsdell, 2020). More broadly, this same approach can be applied to the as-of-yet unnamed xiphosuran specimens from the Fezouata Shale Konservat-Lagerstätte (Lower Ordovician, Morocco; Van Roy et al., 2010), as previous micro-CT imagery of material from this deposits has yielded useful results and allowed for specimens to be differentiated in 3D (Kouraiss et al., 2019).

Three-dimensional reconstructions are increasingly used in computational fluid dynamics (CFD) to study the hydrodynamic properties of extinct aquatic taxa (Rahman et al., 2015a; Darroch et al., 2017; Rahman, 2017; Gibson et al., 2019; Ferrón et al., 2020; Hebdon, Ritterbush & Choi, 2020; Gibson et al., 2021; Song et al., 2021). The majority of CFD studies have focused on enigmatic Ediacaran taxa (Rahman et al., 2015a; Rahman, 2017; Gibson et al., 2019), echinoderms (Rahman et al., 2015b; Rahman et al., 2020; Waters et al., 2017), ammonoids (Hebdon, Ritterbush & Choi, 2020), and vertebrate groups (Dec, 2019; Troelsen et al., 2019; Ferrón et al., 2020; Ferrón et al., 2021). While fossil arthropods have received comparatively less attention than the aforementioned groups (e.g., Pates et al., 2021; Song et al., 2021), CFD studies have modelled lift and drag experienced by modern xiphosurids (Bicknell & Pates, 2019; Davis, Hoover & Miller, 2019). Extending CFD studies to fossil xiphosurids will facilitate comparative studies of the hydrodynamic properties of the carapaces of extinct members of the clade, in addition to elucidating the effects of bizarre morphologies, such as the hypertrophied genal spines, on fluid flow. Such spines have been hypothesised to represent an adaptation to movement through unidirectional fluid flow in primarily freshwater or marginal marine environments (Lamsdell, 2016; Lamsdell, 2021; Bicknell & Pates, 2019; Bicknell & Shcherbakov, 2021; Bicknell et al., 2022a); CFD provides the most compelling method for evaluating the likelihood of this hypothesis. Due to compression of the fossils (consider Dubbolimulus peetae) CFD models of compressed xiphosurids would need to be retro-deformed, likely using modern forms as a proxy for inflation, to account for taphonomic alteration. However, there are specimens, such as Crenatolimulus paluxyensis Feldmann et al., 2011 and Tachypleus decheni (Zincken, 1862), that have maintained their three-dimensionality (Bicknell et al., 2021a). Such specimens may be ideal for scanning and immediate CFD analysis.

Palaeontological and biological collections house a wealth of specimens with academic and historic value. Digitisation of holotype specimens is a salient direction for recording and transferring fundamental anatomical information. These records are traditionally conducted by taking photographs or making line drawings. However, two-dimensional data and views cannot (by definition) display all characteristics needed for modern taxonomic and phylogenetic studies (Mathys et al., 2015; Bicknell et al., 2018a). As such, researchers often need to visit collections to examine specimens in person. This process can be prohibitive for logistic, cost, and policy reasons, to name a few. This complication can be circumvented by producing scans of taxonomically important and unique specimens. Such data is becoming a means of transferring important anatomical data to researchers across the globe and provide interested individuals with another medium with which to examine unique material (Hühne, 2018; Shi, Westeen & Rabosky, 2018; Kouraiss et al., 2019).