A trace fossil made by a walking crayfish or crayfish-like arthropod from the Lower Jurassic Moenave Formation of southwestern Utah, USA

New invertebrate trace fossils from the Lower Jurassic Moenave Formation at the St. George Dinosaur Discovery Site at Johnson Farm (SGDS) continue to expand the ichnofauna at the site. A previously unstudied arthropod locomotory trace, SGDS 1290, comprises two widely spaced, thick, gently undulating paramedial impressions flanked externally by small, tapered to elongate tracks with a staggered to alternating arrangement. The specimen is not a variant of any existing ichnospecies, but bears a striking resemblance to modern, experimentally generated crayfish walking traces, suggesting a crayfish or crayfish-like maker for the fossil. Because of its uniqueness, we place it in a new ichnospecies, Siskemia eurypyge. It is the first fossil crayfish or crayfish-like locomotion trace ever recorded.


INTRODUCTION
Paleoichnology, the study of ichnofossils (fossil tracks and traces), contributes a substantial body of paleobiological information to the understanding of extinct organisms. This is because ichnofossils are direct results of ancient animal behavior (Osgood, 1975) that could only otherwise be inferred from body fossils. Furthermore, the global commonness of ichnofossils compared to body fossils means that the ichnological record often can preserve evidence of the presence of organisms not otherwise or poorly represented in the body fossil record (Osgood, 1975), especially of invertebrates that lack hard parts and therefore fossilize only under exceptional conditions. Except for conchostracans (sensu Kozur & Weems, 2010) and ostracods, which have biomineralized carapaces, arthropods, when compared to their evolutionary diversity, are among the less commonly preserved body-fossil components of terrestrial (including freshwater) paleoecosystems except in various Konservat-Lagerstätten (fossiliferous sites of exceptional preservational quality) (Charbonnier et al., 2010;Luque et al., 2019;Selden & Nudds, 2012;Smith, 2012). Yet from the mid-Paleozoic through the Cenozoic, arthropods-especially insects and arachnidswere certainly the most populous and diverse metazoan components of most terrestrial heretofore unrecognized component of the SGDS ichnofauna. SGDS 1290 is an arthropod locomotory trace because it includes distinct footprints in a discernible cycle, but lacks any features of vertebrate tracks, such as distinct toes (sensu Seilacher, 2007). The trace thus resembles numerous other fossil traces attributed to arthropods, as well as those generated experimentally. The fossil was discovered and collected 11 March 2010 by SGDS volunteer Jon Cross.

GEOLOGICAL SETTING
Most of the fossils preserved at the SGDS, including the ichnofossil described below, occur in the Whitmore Point Member of the Moenave Formation Kirkland et al., 2014), which conformably overlies the Dinosaur Canyon Member of the Moenave Formation and disconformably underlies the Springdale Sandstone Member, which itself has been assigned as both the lowest member of the Kayenta Formation (Lucas & Tanner, 2006) and the uppermost member of the Moenave Formation (Steiner, 2014a). The richest source of the ichnofossils at the SGDS, again including the trace described below, occurs within a fine-grained sandstone near the base of the Whitmore Point Member initially called the "Main Track Layer" Milner, Lockley & Johnson, 2006; and, later and more formally,  (Fig. 2). SGDS 1290 comes from the uppermost strata of this unit, specifically one of several thinly bedded, apparently conformable, fine-grained-sandstone "Top Surface Tracksite" horizons (sensu Kirkland et al., 2014;Milner, Lockley & Johnson, 2006;. The Whitmore Point Member preserves sediments deposited in and around the large, freshwater Lake Whitmore (formerly called Lake Dixie) , 2014; a shoreline paleoenvironment . Ichnologically, invertebrate ichnofossils in this paleoenvironment pertain to the Scoyenia ichnofacies (Buatois & Mángano, 2004;Lucas et al., 2006a), while the associated vertebrate ichnofauna pertains to the Eubrontes ichnocoenosis of the Grallator ichnofacies , 2006c.

MATERIALS AND METHODS
Ichnological terminology for arthropod locomotory traces used herein follows Minter, Braddy & Davis (2007) and Genise (2017). Minter, Braddy & Davis (2007) defined "tracks" as discrete marks made by locomotory appendages, "impressions" as continuous traces made by another portion of the anatomy of a trace maker, and "imprints" as discontinuous such traces; they also provided terms for trackway arrangement and measurements. Genise (2017) outlined various descriptive terms for individual track morphologies. Measurements of SGDS 1290 ( Fig. 3C; Table 1) were taken using digital calipers. The measurements were: track length and width, internal and external widths between paramedial impressions, distances between tracks and adjacent paramedial impressions, and widths of left and right paramedial impressions. Measurements pertaining to the paramedial impressions were taken adjacent to individual tracks/ track sets.
The electronic version of this article in Portable Document Format will represent a published work according to the International Commission on Zoological Nomenclature (ICZN), and hence the new names contained in the electronic version are effectively published under that Code from the electronic edition alone. This published work and the nomenclatural acts it contains have been registered in ZooBank, the online registration system for the ICZN. The ZooBank Life Science Identifiers can be resolved and the associated information viewed through any standard web browser by appending the LSID to the prefix http://zoobank.org/. The LSID for this publication is: urn:lsid: zoobank.org:pub:D78963CE-11C8-4447-8E26-BBCCF0E37143. The LSID for the herein described Siskemia eurypyge isp. nov. is: urn:lsid:zoobank.org:act:769B0815-8991-4F0E-  (2011), the tapering ends of the tapered tracks are presumed to be anterior reflectures, indicating the direction of movement. Most tracks have long axes oriented parallel to the trackway axis; a few (such as in sets L1, L4 and R2) are oblique to the axis. Track R1 is markedly elongate rather than tapered, but also parallel to the trackway axis. Track sets L1, L3 and R4 consist of three distinct but appressed tracks; sets L2, L4 and R2 consist of pairs of appressed tracks, and R1 and R3 appear to consist of single tracks, although the possibility that each comprises multiple, conjoined tracks cannot be ruled out.
The paramedial impressions typically are thick mediolaterally, though they vary and taper briefly to nothingness in a few places (being more continuous than repeating, we consider them "impressions" and not "imprints"). The impressions follow gently undulating (non-linear and low amplitude) pathways. They span approximately 7.5 cm along the slab of rock. Overprints of short segments of the paramedial impressions that are not accompanied by tracks are visible behind the main trace segment on a slightly higher stratum. The impressions vary in width along their lengths, ranging from 0 to 3.1 mm (mean = 2.0 mm) for the left impressions and 0-3.0 mm (mean = 1.6 mm) for the right ( Table 1). The width of the trace from left impression to right impression averages 9.4 mm when measured from the lateral (external) edges and 5.6 mm when measured between the medial (internal) edges (Table 1). The distances between the impressions thus are greater than the distances between the impressions and their flanking tracks (mean = 1.6 mm); the ratio of the distance between a paramedial impression and its flanking track to the distance between the medial edges of the paramedial impressions ranges from 0 to 0.78 (mean = 0.31; see Supplemental Material), so on average, the paramedial impressions are roughly three times farther apart from each other than either is from its flanking tracks. The impressions taper slightly in cross sectional view: they are wider at their bases and narrower at their rounded apices.
Acanthichnus tracks (Fig. 4A), attributed to a chelicerate such as a solifugid, are oppositely arranged, short, elongate impressions in two (or four, per Dalman & Lucas, 2015) parallel rows; tracks either are parallel to or angle slightly away from the trackway axis (Dalman & Lucas, 2015;Hitchcock, 1858). This morphology and organization are both unlike those of SGDS 1290.
Asaphoidichnus tracks (Fig. 4B), attributed to trilobites, are elongate to crescentic, possess 3-4 crescentic branches at one end, and are oriented oblique to the trackway axis (Miller, 1880). They are far more complex in structure than the tracks of SGDS 1290.
Bifurculapes (Fig. 4C), attributed to an insect, possibly a beetle (Getty, 2016), comprises adjacent pairs (rarely triplets) of slightly staggered, elongate, straight to crescentic tracks that lie parallel or slightly oblique to the trackway axis, unlike the tracks of SGDS 1290. Tracks in each pair sometimes converge toward one end in Bifurculapes. This ichnotaxon typically does not possess paramedial impressions, but a specimen described by Getty (2016: fig. 1) possess two such traces, albeit faintly, that lie closer to the tracks than to the trackway axis, as in SGDS 1290. These impressions are far less pronounced than their associated tracks, unlike those of SGDS 1290. Getty (2020) ascertained that Bifurculapes traces were made subaqueously and may have been made by a terrestrial insect that would have left different tracks subaerially.
Coenobichnus tracks (Fig. 4D), attributed to a hermit crab, are thick, crescentic to ellipsoidal, roughly parallel and closely appressed to the trackway axis, and asymmetrical, with the left tracks larger than the right tracks (Walker, Holland & Gardiner, 2003), all of which differentiate this ichnotaxon from SGDS 1290.
Copeza (possibly a variant and synonym of Lithographus (Lull, 1953;Rainforth, 2005; Fig. 4E)) consists of triplets of roughly oppositely arranged, linear, elongate tracks in which the anteriormost lies roughly perpendicular to the trackway axis while the posteriormost pairs lie parallel or oblique to the trackway axis (Lull, 1953). This rare ichnotaxon is thus unlike SGDS 1290.
Danstairia congesta (Fig. 4F) comprises circular to crescentic tracks in closely appressed sets of up to six that are oriented oblique to the trackway axis; tracks often overlap to form V-shaped structures (Walker, 1985), unlike in SGDS 1290.  (2018); also see below) tracks span a range of morphologies. D. aenigma (Fig. 4G), ostensibly the ichnospecies lectotype except that no specimen was designated as such (Stimson et al., 2018), typically comprises elongate, closely packed tracks in parallel rows on either side of the trackway axis; the tracks lie perpendicular to the trackway axis (Dawson, 1873). D. gouldi Type A (Fig. 4H) comprises parallel rows of closely spaced, oppositely arranged, simple, roughly circular to oblong to comma-shaped or irregular tracks with varying orientations to the trackway axis; D. gouldi Type B comprises closely spaced, elongate impressions oriented perpendicular, or nearly so, to the trackway axis, matching the general description of D. aenigma; D. gouldi Type C is similar to Type B, but the tracks are oriented oblique (~45 ) to the trackway axis (Trewin & McNamara, 1994). D. cuithensis, attributed to large arthropleurid myriapods, is similar to both D. aenigma and D. gouldi Type B, but is very large and has widely spaced rows of tracks (Briggs, Rolfe & Brannan, 1979). D. binatus tracks often occur as closely appressed pairs of imprints (Webby, 1983). D. triassicus tracks are small and circular to ovoid rather than elongate, and frequently paired on either side of the trackway axis (Pollard, Selden & Watts, 2008); this ichnospecies has been alternately suggested to be a synonym of D. gouldi  or to pertain to Acripes (Machalski & Machalska, 1994;Pollard, 1985). D. metzi possesses a midline impression that is sometimes interrupted by connections between tracks in the closely appressed rows (Fillmore et al., 2017). The ichnogenus is in need of thorough review (Smith et al., 2003), but in all cases, the tracks are of different morphologies and arrangements than those of SGDS 1290.
Hamipes tracks (Fig. 4K) consist of closely spaced, paired, elongate to crescentic impressions oriented parallel to the trackway axis; the outer tracks are longer than their accompanying inner tracks, and the tracks are staggered or alternately arranged (Getty, 2018;Hitchcock, 1858). Track morphology readily differentiates Hamipes from SGDS 1290.
Laterigradus tracks (Fig. 4L), attributed to sideways-walking crabs, comprises asymmetrical trackways consisting of sets of up to four tracks (De Carvalho et al., 2016). Individual tracks vary widely in shape, ranging from elongate to tapered to comma-shaped to roughly circular. Track sets fall within a narrow trackway width and exhibit different stride lengths along the course of a trackway. While some individual track shapes resemble those of SGDS 1290, the overall arrangement and layout of the tracks is distinct.
Lithographus (including Permichnium sensu Minter & Braddy, 2009) tracks ( Fig. 4M), which match those made by pterygote insects, especially extant cockroaches (Davis, , comprise trios (or pairs, in the case of the Permichnium variant) of elongate to comma-shaped, rather than circular or tapering, tracks that are arranged at different angles to one another, some of which are oriented perpendicular to the trackway axis, and others of which are oblique to the trackway axis (Guthörl, 1934;Hitchcock, 1858;Minter & Braddy, 2009). These track morphologies and arrangements are distinct from those of SGDS 1290. Hornburgichnium reportedly is similar to Permichnium, but has three tracks on either side of the midline instead of two, and at least one of each set is oriented parallel to the trackway axis (Kozur, 1989); it may also be a variant of Lithographus (Lucas et al., 2005b). Trackways of Lithographus can transition into trackways that Hitchcock (1858) called Hexapodichnus (Davis, Minter, Braddy & Davis, 2007), so the latter may be considered a behavioral and/or substrate-consistency variant of the former, and also unlike SGDS 1290.
Maculichna (including Guandacolichnus and Paganzichnus of Pazos (2000) per Buatois & Mángano (2003)) tracks (Fig. 4N) comprise pairs (sometimes more) of small, circular to slightly elongate tracks arranged in closely appressed, slightly staggered rows. Pairings are oriented virtually parallel to the trackway axis (Anderson, 1975a); occasionally, short segments of linear, singular medial or closely spaced, paired paramedial imprints are also preserved that can be offset to one side of the trackway axis (Anderson, 1975a: figs. 8b, 8d and 8e). The pairing of Maculichna tracks differs from that of SGDS 1290. Aceñolaza & Buatois (1991, 1993 and Archer & Maples (1984) described Maculichna traces that exhibit the pairing of classic Maculichna from South Africa, but in which tracks are more ellipsoidal to shaped like slightly inflated isosceles triangles; the long axes of the triangles are oriented close to perpendicular to the trackway axis. Pazos (2000) recognized this morphology as the separate ichnotaxon Paganzichnus. This morphology is also unlike that of SGDS 1290.
The ichnospecies holotype of Merostomichnites narragansettensis (Fig. 4O) consists of parallel rows of roughly oppositely arranged circular to elongate to comma-shaped tracks, the long axes of which are perpendicular to the trackway axis (Packard, 1900). Merostomichnites beecheri tracks are circular and connected across the trackway axis by curvilinear, shallow, M-shaped imprints, creating a sort of segmented midline impression (Packard, 1900 : fig. 4). These track and trace morphologies do not match those of SGDS 1290.
Mirandaichnium (Fig. 4P) consists of two rows of elongate, linear tracks that terminate laterally in small, circular impressions. Tracks are oriented perpendicular or oblique to the trackway axis, oppositely situated, and often grouped into series of eight (Aceñolaza & Buatois, 1993;Buatois et al., 1998), unlike those of SGDS 1290.
Octopodichnus (Fig. 4Q) ichnospecies, attributed to arachnids, have different morphologies. O. didactylus tracks are circular to crescentic to bifurcate or trifurcate oriented parallel to the trackway axis in alternating, arcuate sets of four (Sadler, 1993). O. minor tracks have a similar organization, but the tracks are more amorphous; O. raymondi tracks consist of clusters of four circular to crescentic marks arranged in checkmark-like patterns (Sadler, 1993). These track morphologies and distinctive arrangements are substantially unlike those of SGDS 1290.
Pterichnus tracks (Fig. 4S), attributed to isopods, frequently are segmented and are more linear and elongate than any in SGDS 1290. Tracks (or series of segments) are oriented oblique (Types 1 and 2 of Gaillard et al., 2005) or parallel (Types 3 and 4 of Gaillard et al., 2005) to the trackway axis, and approximately symmetrically arranged in two parallel rows (Gaillard et al., 2005;Hitchcock, 1865) that are somewhat closer together than are those of the morphologically similar Diplichnites. Types 3 and 4 of Gaillard et al. (2005) morphologically grade into Diplopodichnus (Uchman et al., 2011). Hammersburg, Hasiotis & Robison (2018) suggested that Pterichnus comprises undertracks of Lithographus and is thus a junior synonym of that ichnotaxon. In any case, Pterichnus tracks are readily distinguished from SGDS 1290.
Tasmanadia traces (Fig. 4T) consist of two rows of closely packed, elongate, linear tracks oriented generally perpendicular to the trackway axis; occasionally, tracks overlap at one end, creating narrow, V-shaped structures (Chapman, 1929;Glaessner, 1957). Morphologically, this ichnotaxon resembles Diplichnites gouldi Type B and Umfolozia (but lacks the organization of the latter), and differs from SGDS 1290 for the same reasons as those ichnotaxa.
In summary, SGDS 1290 is not a variant of any of these ichnotaxa.
Kouphichnium traces (Fig. 4V) are attributed to limulids and occur in a variety of configurations. Tracks in clear Kouphichnium walking traces that possess singular medial impressions (e.g., many K. lithographicum, but not, for example, K. minkinensis (King, Stimson & Lucas, 2019;q.v. Gaillard, 2011a;Shu et al., 2018)) typically occur in sets of up to five in rows oriented oblique to the medial impression and trackway axis; individual tracks range from circular and ellipsoidal to elongate, and can split into anywhere from 2 to 5 branches at their ends (Caster, 1938(Caster, , 1944King, Stimson & Lucas, 2019;Shu et al., 2018). Well-preserved Kouphichnium tracks are dissimilar to those of SGDS 1290.
Traces referred to as "Merostomichnites" (Fig. 4W) and attributed to the eurypterid Mixopterus by Hanken & Stormer (1975) consist of three elongate and crescentic tracks in oblique rows on either side of an intermittent medial impression; the tracks increase in size laterally, and some split into two or more branches on one end. In any of these cases, however, the tracks are substantially more complex than those of SGDS 1290.
Oniscoidichnus tracks (Fig. 4X) are elongate to crescentic, oriented perpendicular or oblique to the trackway axis, closely packed and closely appressed to the single midline impression (Brady, 1947;Davies, Sansom & Turner, 2006). In all these details, Oniscoidichnus traces differ markedly from SGDS 1290.
Ichnospecies of Palmichnium (Fig. 4Y), also attributed to eurypterids, vary in morphology. Generally, they comprise complex sets of tracks lying lateral to a medial impression that can be either continuous or discontinuous. Tracks range in shape from elongate to crescentic to ovoid to chevron shaped, and they generally parallel the trackway axis. The tracks occur in oblique rows in sets of up to four; in some traces, the lateralmost tracks are elongate and curved, while the more medial tracks are linear and oriented parallel to the trackway axis (Braddy & Milner, 1998;Minter & Braddy, 2009;Poschmann & Braddy, 2010;Richter, 1954). Tracks are both more numerous and differently shaped than those of SGDS 1290.
(2020) attributed traces lacking a medial impression and comprising closely appressed pairs or triplets of mostly elliptical tracks from the Upper Jurassic or Lower Cretaceous of Brazil to a new ichnospecies of Paleohelcura and attributed them to a pterygote insect track maker. Tracks in this ichnospecies are arranged in rows oriented oblique to the trackway axis, and track sets in this ichnospecies lie close to the midline. This morphology is also unlike that of SGDS 1290.
Robledoichnus tracks (Fig. 4BB), attributed to flying insects, resemble tracks of Eisenachichnus but possess a discontinuous, faint medial trace consisting entirely of periodic, V-shaped marks flanked by asymmetrical pairs of tracks. On one side, the tracks are short, tapered, and oriented oblique to the trackway axis; on the other side, the tracks are longer and crescentic, oriented closer to perpendicular to the trackway axis (Kozur & Lemone, 1995). Lucas et al. (2005a) considered Robledoichnus a probable junior synonym of Paleohelcura or Stiaria, and the ichnotaxon differs from SGDS 1290 for similar reasons as those ichnotaxa, in addition to the asymmetry.
Shalemichnus traces (Fig. 4CC), for which only half a trackway is known, include a straight medial impression punctuated at intervals by V-shaped marks. This impression is flanked by sets of three tapered tracks in straight rows oriented perpendicular to the trackway axis; individual tracks have their long axes parallel to the trackway axis (Kozur & Lemone, 1995). Minter & Braddy (2009) considered Shalemichnus a junior synonym of Stiaria. The tracks of Shalemichnus bear some similarity to those of SGDS 1290, but the paramedial impressions of SGDS 1290 lack the V-shaped markings of the medial impression of Shalemichnus.
Stiallia traces (Fig. 4DD) consist of paired rows of long, linear impressions parallel or slightly oblique to the trackway axis and that frequently overlap. Stiallia pilosa lacks any medial or paramedial impressions, but Stiallia (Carrickia of Smith (1909)) berriana possesses a medial row of crescentic to chevron-shaped marks (Smith, 1909;Walker, 1985). Pollard (1995) suggested that Stiallia could be an arthropod swimming, rather than a walking, trace, though it also resembles traces made by bristletail insects walking in highly saturated mud (Getty et al., 2013: figs. 6F and 6G). Stiallia tracks are markedly unlike those of SGDS 1290.
As with locomotory traces lacking medial impressions, SGDS 1290 is not a variant of any of these ichnotaxa.

Arthropod repichnia possessing three or more medial and paramedial impressions
Mitchellichnus (Fig. 4FF), attributed to archaeognathan insects (Getty et al., 2013), is distinguished from SGDS 1290 by possessing three medial impressions (Walker, 1985). Mitchellichnus tracks are complex, comprising two distinct types and arrangements. An inner set, lying close to the medial impressions, consists of apparently elongate tracks in sets of up to six that lie parallel to slightly oblique to the trackway axis; an outer set consists of larger, amorphous impressions (Walker, 1985). Tracks are thus more numerous in Mitchellichnus than in SGDS 1290, and the tracks differ in arrangement and morphology. Like Stiaria, Genise (2017) asserted that Mitchellichnus should be considered a junior synonym of Siskemia.
Keircalia (Fig. 4GG) is distinguished from SGDS 1290 by possessing four medial impressions (Smith, 1909;Walker, 1985). Keircalia tracks are crescentic to irregularly shaped, generally are oriented perpendicular to the trackway axis, and have no discernible arrangement (Walker, 1985). Both track morphology and organization are unlike those of SGDS 1290.
Danstairia vagusa (Fig. 4HH) possesses intermittent, thin, linear imprints that do not always parallel their accompanying trackways. Tracks are circular to triangular, generally have their long axes perpendicular to the trackway axis, and lack any coherent layout (Walker, 1985), unlike those of SGDS 1290. D. vagusa somewhat resembles Keircalia traces, but its tracks are spaced more widely apart.
Warvichnium traces (Fig. 4JJ) are complex, comprising pairs to multiple sets of linear, discontinuous medial and paramedial imprints flanked by varying numbers of linear to crescentic tracks in two or more sets: an inner set, close to the medial imprints, that are oriented slightly oblique to the trackway axis, and an outer set oriented closer to perpendicular to the trackway axis (Walter, 1985), quite unlike SGDS 1290. Getty (2020) noted similarities between Warvichnium and subaqueous Bifurculapes.
Among described arthropod repichnia, SGDS 1290 architecturally most closely resembles ichnospecies of Siskemia by possessing discreet, compact (not linear) tracks and track sets flanking paired paramedial impressions. Three ichnospecies of Siskemia are presently recognized (Walker, 1985): S. bipediculus (Fig. 4KK) comprises small, circular tracks in closely appressed pairs (occasionally trios) in rows oriented perpendicular or slightly oblique to the trackway axis; the pairs are spaced apart at approximately regular intervals and evenly distant from the uniformly straight and narrow paramedial impressions (Walker, 1985). The paramedial impressions lie close to the midline axis, well away from their adjacent tracks (the average ratio of the distance between a paramedial impression and its flanking track to the distance between the medial edges of the paramedial impressions is 1.34; see Supplemental Material). S. elegans (Fig. 4LL) has similarly shaped tracks in sets of up to four; the sets similarly lie well away from the likewise straight, narrow, and closely appressed paramedial impressions (the average ratio of the distance between a paramedial impression and its flanking track to the distance between the medial edges of the paramedial impressions is 1.75; see Supplemental Material). S. bipediculus and S. elegans differ primarily in the orientations of their track rows to the trackway axis and the continuities and thicknesses of their paramedial impressions (Walker, 1985), though these could be behavioral and/or substrate-driven variants. Siskemia latavia (Fig. 4MM; called "lata-via" by Smith (1909) and Walker (1985), but the ICZN does not permit hyphens in genus or species names) tracks comprise tapered or ovoid tracks arranged in roughly triangular sets of three. Most individual tracks have their long axes oriented oblique to the trackway axis; track sets have varying orientations to the trackway axis. Tracks in individual S. latavia sets usually are spaced well apart from each other; rarely do two individual tracks in a set contact one another. S. latavia tracks lack the regular spacing and arrangements of those of S. elegans and S. bipediculus, and occasional individual tracks lie close to the paramedial impressions, farther medially than in either of the other two Siskemia ichnospecies (the average ratio of the distance between a paramedial impression and its flanking track to the distance between the medial edges of the paramedial impressions is 0.80; see Supplemental Material). Each paramedial impression of S. latavia is slightly wider than those of the other two Siskemia ichnospecies (probably a function of the larger overall size of specimens attributed to this ichnospecies), but retain the close appression to the trackway midline and the uniform straightness.
The tracks of SGDS 1290 vary more in morphology than those of any known Siskemia ichnospecies, but grossly share their layout. Tracks in all three Siskemia ichnospecies have a staggered distribution, similar to, but less pronounced than, that of SGDS 1290. SGDS 1290 differs most markedly from any of the three Siskemia ichnospecies in the morphology and positions of the paramedial impressions: in SGDS 1290, the impressions vary in thickness along their lengths and undulate, in contrast to the thin, straight impressions of all three Siskemia ichnospecies. Additionally, the impressions in SGDS 1290 lie farther apart than those of the three Siskemia ichnospecies. In fact, all of the ichnospecies of Siskemia erected by Smith (1909), as well as both specimens later attributed to this ichnogenus (Getty et al., 2013;McNamara, 2014;Pollard, Steel & Undersrud, 1982) and Siskemia-like traces made by extant, archaeognathan insects (Getty et al., 2013), have such thin, linear, closely appressed paramedial impressions (sometimes offset toward one side of the trackway). The only time when archaeognathan traces approach the paramedial impression spacing of SGDS 1290 is when both abdominal styli (laterally) and gonostyli (medially) of the trace makers register impressions and imprints in wet mud, producing two sets of paramedial traces (Getty et al., 2013: figs. 6K and 6L), but even then the linear, lateralmost paramedial impressions still do not resemble the thick, undulating impressions of SGDS 1290. Simultaneously, in such wet mud, archaeognathan tracks are elongate and oriented oblique to the trackway axis, unlike those in SGDS 1290. In total, SGDS 1290 does not fit neatly into any known Siskemia ichnospecies and does not seem to be an archaeognathan insect trace.
Among traces made by extant arthropods, SGDS 1290 bears similarities to traces made by both notostracans and crayfish. Interpretive drawings of experimental traces made by notostracans figured by Trusheim (1931) depict elongate, crescentic, or tapered tracks oriented perpendicular to paramedial impressions; the tracks are arranged oppositely, unlike those of SGDS 1290. Additionally, the thin, linear paramedial impressions figured by Trusheim (1931) lie so far from the trackway axis that they often contact their accompanying tracks, a phenomenon that only occurs in SGDS 1290 near L4 and R4, where the lateral margins of its undulating paramedial impressions meander particularly far laterally. Tasch (1969) noted, though, that the drawings presented by Trusheim (1931) were misleading compared to his own experimentally generated notostracan traces. However, he described the morphologies of his notostracan tracks only as "minute en echelon stripes" (Tasch, 1969: 327), which does not adequately specify how they differed from those of Trusheim (1931); track details are impossible to discern in his lone photographic figure (Tasch, 1969: pl. 1.2). Gand et al. (2008) also conducted neoichnological experiments with notostracans, recovering locomotory traces that were less orderly than those illustrated by Trusheim (1931) (Gand et al., 2008: figs. 16.1, 16.2 and 17.1). Their extant notostracan tracks comprised multiple tracks with rather chaotic distributions lateral to their paramedial impressions, unlike the regular distribution seen in SGDS 1290. Gand et al. (2008) found their notostracan traces to fall within the "etho-morphotype" of Acripes, as exemplified by their novel ichnospecies A. multiformis from the Permian of France. (Linck (1943) and Pollard (1985) also referred Acripes (Merostomichnites of Linck (1943)) tracks to notostracans, but not based on neoichnological experiments.) A. multiformis traces, unlike classic Acripes (Matthew, 1910), possess paramedial imprints, albeit inconsistently. Hammersburg, Hasiotis & Robison (2018), Häntzschel (1975), Miller (1996) and Pemberton, MacEachern & Gingras (2007) all supported classic Acripes as a junior synonym of Diplichnites; the issue of synonymy is beyond the scope of this paper, but we note at least that the tracks in fossils that Gand et al. (2008) called A. multiformis differ from SGDS 1290 in the same ways that Diplichnites tracks do (see above). Lastly, Knecht et al. (2009: figs. 5 and 6) also illustrated traces made by extant notostracans (Fig. 4NN), which are "tidier" than those of Gand et al. (2008) and resemble those of classic Acripes and Diplichnites, albeit with discontinuous paramedial and curvilinear lateral (external) imprints. The tracks in these traces comprise irregular, ellipsoidal sets oriented oblique to the trackway axis, unlike those of SGDS 1290. In total, SGDS 1290 is unlikely to be a notostracan trace. Fairchild & Hasiotis (2011) conducted neoichnological experiments with crayfish to examine their locomotory traces. These traces varied in morphology depending on substrate conditions (sediment grain size and saturation) and slope; in general, when clearest, they consist of sets of 1-4 circular, tapering, ellipsoidal, or elongate tracks, occasionally of different sizes, that are oriented parallel to the trackway axis and that lie lateral to a pair of undulating, variably thick paramedial impressions that lie closer to their flanking tracks than to each other (Figs. 3D and 4OO). Morphologically, the tracks and impressions match those of SGDS 1290, although the tracks made by extant crayfish often are larger than those of SGDS 1290 when produced in dry substrate (Fairchild & Hasiotis, 2011: fig. 9). Track sets in extant crayfish traces have complex arrangements: when comprised of multiple traces, they frequently are in rows oriented perpendicular to the trackway axis, but sometimes rows are oblique to the trackway axis. When fewer tracks are preserved, sets can appear to lie in single, parallel rows on either side of the paramedial impressions. Track positions can be opposite to staggered to alternating, also as in SGDS 1290. In both track and paramedial impression morphology, as well as in overall trace architecture, SGDS 1290 strongly resembles crayfish traces made in damp silt and clay (Fairchild & Hasiotis, 2011: figs. 2e and 2f), dry and saturated, very fine-grained sand ( Fig. 3D; Fairchild & Hasiotis, 2011: figs. 4a, 4b, 4e and 4f), dry and damp, fine-grained sand (Fairchild & Hasiotis, 2011: figs. 5a-5d), and saturated medium sand (Fairchild & Hasiotis, 2011: figs. 6e and 6f). SGDS 1290 is preserved in, and was presumably registered in, a fine-grained sand, lithologically matching one set of experimental conditions in Fairchild & Hasiotis (2011). However, SGDS 1290 is not as detailed as many of the experimentally generated crayfish traces in comparable sediments. This could indicate one or more things: the fossil could be a slight overtrack (sensu Bertling et al., 2006) rather than a direct natural cast; the trace maker may have been partly buoyant; and/or trace-maker behavior and sediment consistency combined such that the lighter limbs did not impress as deeply as the heavier tail.
Several Late Triassic body-fossil specimens also have been reported as crayfish (Hasiotis, 1995;Hasiotis & Mitchell, 1993;Miller & Ash, 1988;Olsen & Huber, 1997;Santucci & Kirkland, 2010), but the identities of these specimens as true astacidans has not been established. Miller & Ash (1988) placed a Late Triassic specimen from Petrified Forest National Park, Arizona in Enoploclytia, which is an erymid lobster, not an astacidan, genus. That generic placement subsequently has been contested (Amati, Feldmann & Zonneveld, 2004;Schweitzer et al., 2010;Urreta, 1989), so the specimen needs detailed restudy, but if it pertains to Erymidae rather than Astacida, then it indicates that a lineage of erymid lobsters colonized terrestrial environments, possibly before true (monophyletic) crayfish. Some older analyses (reviewed in Rode & Babcock, 2003) postulated crayfish origins within Erymidae, but more recent phylogenetic analyses (Devillez, Charbonnier & Barreil, 2019;Karasawa, Schweitzer & Feldmann, 2013;Rode & Babcock, 2003;Schram & Dixon, 2004;Stern & Crandall, 2015) have recovered (a frequently paraphyletic) Erymidae with members at varying distances from Astacida. If those hypothesized phylogenetic relationships are correct, then no erymid can, in a monophyletic sense, be considered a crayfish, even if it was a freshwater taxon. But regardless of semantics or phylogenetic relationships, crayfish and erymid morphological similarities suggest that their locomotory traces might be indistinguishable, making a definitive attribution of SGDS 1290 impossible.
A crayfish or crayfish-like trace maker for SGDS 1290 is tenable both chronologically and ecologically. As mentioned above, multiple crayfish-like morphotypes have been found in the Upper Triassic Chinle Formation of Arizona and Utah (Hasiotis, 1995;Miller & Ash, 1988;Santucci & Kirkland, 2010). The Moenave Formation overlies the Chinle Formation in southwestern Utah, so crayfish or crayfish-like decapods plausibly could have been present in and around freshwater Lake Whitmore both geographically and stratigraphically. As-yet-undescribed, crayfish or crayfish-like body fossils also have been recovered from lacustrine sediments of the uppermost Triassic Chatham Group (Newark Supergroup) in North Carolina (Olsen & Huber, 1997), attesting to how widespread such arthropods were in terrestrial environments in North America even prior to the Jurassic.
Only three locomotory ichnotaxa have been attributed specifically to crayfish. First, Heidtke (1990) erected Pollichianum repichnum for Early Permian ichnofossils from Germany that he attributed to the "crawfish" (in the English abstract; "Krebses" in the German abstract) Uronectes fimbriatus, also from the Early Permian of Germany. However, Uronectes has long been recognized as a syncarid (Brooks, 1962;Calman, 1934;Perrier et al., 2006), not an astacidan, or even a decapod, so the term appearing in the English abstract likely is a simple translation error. Furthermore, however, P. repichnum is not differentiable from the resting trace (cubichnion) P. cubichnum (O'Brien, Braddy & Radley, 2009) and therefore is a junior synonym and is not a locomotory trace. In any case, Pollichianum is morphologically quite unlike both SGDS 1290 and experimentally generated crayfish traces (Fairchild & Hasiotis, 2011). Second, Bolliger & Gubler (1997 hypothesized that their novel, early Miocene ichnospecies Hamipes molassicus was made by a buoyed (presumably swimming) crayfish. Getty (2018) referred these specimens to Conopsoides; later, Getty & Burnett (2019) suggested that at least some of the specimens may pertain to Acanthichnus, and they differ from SGDS 1290 for the same reasons outlined above for Acanthichnus. Third, De Gibert et al. (2000) attributed Early Cretaceous, Spanish specimens that they assigned to Hamipes didactylus to crayfish. Getty (2018) attributed these tracks to Bifurculapes and maintained a crustacean track maker for H. didactylus, but was not more specific. However, neither Bifurculapes nor Hamipes resemble experimentally generated crayfish traces (Fairchild & Hasiotis, 2011), or any of the mortichnial decapod traces, and thus are unlikely to have been made by a crayfish-like decapod, at least while walking. Lastly, we also note that unnamed trackways attributed to crayfish from the Upper Triassic Chinle Formation of Utah were mentioned, but not described, by Hasiotis (1991); Fairchild & Hasiotis (2011) did not note whether or not these were similar to their experimentally generated traces. Additionally, an unnamed "crayfish locomotion trace" was figured, but not described, by Rainforth & Lockley (1996: fig. 1b); it does not resemble either SGDS 1290 or experimentally generated crayfish traces (Fairchild & Hasiotis, 2011).
As detailed above, SGDS 1290 does not fit neatly into any existing ichnospecies of Siskemia. Whether or not to place it in a novel ichnospecies, or even ichnogenus, is, therefore, an open question. Bertling et al. (2006), Gaillard (2011b) and Minter, Braddy & Davis (2007) provided solid criteria for the erection of new ichnotaxa, the latter particularly for arthropods. One criterion is that a new ichnotaxon ideally should be represented by a substantial number of specimens that demonstrate behavioral and substrate-based morphological variation; this prevents erecting several ichnotaxa for minor, readily explained variations in trace morphology. SGDS 1290, as a singular specimen, certainly does not meet that criterion, but Minter, Braddy & Davis (2007) also allowed that truly unique morphologies exhibited by singular specimens can support an ichnotaxon. In terms of uniqueness, another criterion is whether or not a new morphotype falls onto a continuum, established or hypothetical, of morphologies within an established ichnotaxon. SGDS 1290 is closest morphologically to ichnospecies of Siskemia, but has several distinctions from any established ichnospecies therein, particularly the thick and undulating paramedial impressions and the wider spacing between the paramedial impressions and consequent closer appression of the paramedial impressions to the tracks: the average ratios of the distance between a paramedial impression and its flanking track to the distance between the medial edges of the paramedial impressions are 1.34 for S. bipediculus, 1.75 for S. elegans, and 0.80 for S. latavia compared to 0.31 for SGDS 1290 (see Supplemental Material). No published specimen of Siskemia demonstrates the features of SGDS 1290; nor do Siskemia-like traces made by archaeognathan insects in experimental conditions (Getty et al., 2013). Thus, SGDS 1290 does not appear to fall within the established Siskemia continuum. The greater prominence (depth) of the paramedial impressions of SGDS 1290 than their associated tracks suggests either a trace maker with heavier tail elements than the gonostyli of an archaeognathan insect or an archaeognathan trace maker with unusually large styli adopting an unusual posture (possibly partly buoyant), flexing its caudal region downward to create deep styli impressions but not deep track impressions. We consider the latter unlikely; thus, SGDS 1290 does not appear to fall within a hypothetical Siskemia continuum, either. However, SGDS 1290 falls within the continuum of trace morphologies made by extant crayfish in experimental conditions (Fairchild & Hasiotis, 2011). No philosophical basis has been established for the recognition of novel ichnotaxa on the basis of comparison with traces made by extant organisms; only by comparison with fossil ichnotaxa because extant traces cannot be the basis for an ichnotaxon (Bertling et al., 2006;International Commission on Zoological Nomenclature, 1999).
SGDS 1290 clearly is morphologically distinctive. Lacking a sufficient number of specimens with which to determine ranges of morphological variation, however, erecting a new ichnogenus for it clearly is unwarranted. Yet we feel that its unique morphology warrants ichnotaxonomic distinction. Given its distant similarity to Siskemia ichnospecies, we therefore herein place it in a new ichnospecies of that ichnogenus.

Ichnofamily Protichnidae Uchman, Gaździcki & Błażejowski, 2018
Ichnogenus Siskemia Smith, 1909 Type Ichnospecies Siskemia elegans Smith, 1909 Diagnosis. Trace consisting of parallel rows of grouped tracks on either side of two parallel, paramedial impressions. Each group of tracks consists of up to four imprints arranged in series, transversely or obliquely to the midline of the trackway (following Walker, 1985). Walker (1985) further specified that Siskemia was diagnosed by paramedial impressions with maximum widths of 0.5 mm, but following Bertling et al. (2006), size is not a suitable ichnotaxobase.
Ichnospecies Siskemia eurypyge isp. nov. Figures 3A and 3B Diagnosis. Two parallel, undulating, paramedial impressions flanked externally by closely appressed sets of 1-3 small, ovoid to tapered to elongate tracks; tapered and elongate tracks have long axes parallel or oblique to the trackway axis. Track sets are oriented perpendicular to the trackway axis. Left and right tracks are arranged in a staggered to alternating pattern. Paramedial impressions are mediolaterally thick, but discontinuous, tapering out of existence briefly in some places. Impressions are gently undulating (low amplitude). The paramedial impressions lie far from the trackway axis, generally closer to (and sometimes in contact with) the tracks than to the midline axis or each other.
Derivation of name. From the Greek εyρyς (eurys), meaning "broad" or "wide," and πyγή (pyge), meaning "rump." The ichnospecies name refers to the wider spacing between the paramedial impressions, made by the rear end of the trace maker, than those of other Siskemia ichnospecies.

CONCLUSIONS
SGDS 1290, from the Lower Jurassic (Hettangian) Whitmore Point Member of the Moenave Formation, consists of two paramedial impressions that are flanked by staggered to alternating sets of tapered or elongate tracks. The traces closely resemble those made by extant crayfish (Fairchild & Hasiotis, 2011) and are similar in components to traces placed in the ichnogenus Siskemia (Smith, 1909;Walker, 1985). In previously recognized Siskemia ichnospecies, the paramedial impressions are thin, relatively straight, and closely appressed to the trackway axis. But in SGDS 1290, paramedial impressions have the opposite morphology and arrangement: they are thick and lie closer to their tracks than the medial axis of the trackway. SGDS 1290 paramedial impressions also undulate, which is not seen in any previously known Siskemia ichnospecies. Thus, we erect a new ichnospecies, Siskemia eurypyge, to house SGDS 1290 and as-yet undiscovered traces with this morphology and arrangement.
Placing SGDS 1290 in Siskemia extends the known range of the ichnogenus into the Early Mesozoic. All other reported occurrences of the ichnogenus are Paleozoic in age: Early Silurian (McNamara, 2014;Trewin & McNamara, 1994), Late Silurian (Davies, Sansom & Turner, 2006), Early Devonian (Pollard, Steel & Undersrud, 1982;Pollard & Walker, 1984;Smith, 1909;Walker, 1985), and Pennsylvanian (Getty et al., 2013). However, age should not be a factor in ichnotaxonomy (Bertling et al., 2006). At least some Paleozoic Siskemia traces likely were made by archaeognathan insects (Getty et al., 2013), which are extant and for which body fossils are known as early as the Devonian. Based on their similarity to traces made by extant crayfish (Fairchild & Hasiotis, 2011), S. eurypyge likely was made by a crayfish or crayfish-like crustacean, for which body fossils are known as early as the Late Triassic and which also are extant. Thus, Siskemia ispp. traces would be expected to occur from Early Silurian to Recent, but thus far have not been documented except for the occurrences noted above.
SGDS 1290 expands the ichnological record of crayfish and crayfish-like animals to include repichnia in addition to domichnia. Fossil burrows (Camborygma ispp.), usually attributed to crayfish, are well known at some sites and in some formations (Hasiotis, 1995;Hasiotis & Bown, 1996;Hasiotis & Honey, 1995;Hasiotis & Mitchell, 1993;Hasiotis, Kirkland & Callison, 1998; see Schram & Dixon (2004) concerning pre-Cretaceous examples), attesting to the presences-and, in some places, abundances-of crayfish and/or crayfish-like taxa in Mesozoic-Cenozoic freshwater paleoecosystems. Yet locomotion traces made by these burrowers oddly have never before been documented as ichnofossils, possibly because they infrequently venture far from their burrows in substrates suitable for registering locomotory traces, as with modern crayfish (Martin, 2013). SGDS 1290 is the first documented locomotory ichnofossil made by a freshwater crayfish or crayfish-like organism, as well as the first fossil evidence of such a taxon in the Lower Jurassic Moenave Formation and indeed the Early Jurassic of the southwestern US. The absence of Camborygma burrows in the Moenave Formation that would have been made by the SGDS 1290 trace maker is puzzling, and may be a consequence of a lack of recognition; alternatively, the producer of SGDS 1290 was not a burrower.
Andrew R.C. Milner conceived and designed the experiments, performed the experiments, analyzed the data, prepared figures and/or tables, authored or reviewed drafts of the paper, and approved the final draft.

Data Availability
The following information was supplied regarding data availability: Specimen SGDS 1290, the holotype of Siskemia eurypyge isp. nov., is reposited at the St. George Dinosaur Discovery Site at Johnson Farm (SGDS) in St. George, Washington County, Utah, USA. The specimen is a natural cast from the Bug Crossing Quarry (SGDS locality 87) at the SGDS; stratigraphically, it is from the Top Surface of the Johnson Farm Sandstone Bed within the Whitmore Point Member of the Moenave Formation (age: Hettangian, Early Jurassic).

Supplemental Information
Supplemental information for this article can be found online at http://dx.doi.org/10.7717/ peerj.10640#supplemental-information.