Geologic setting

The fossil spinicaudatans were collected from the early Oligocene Medicine Lodge Formation in Beaverhead County, Montana (Fig. 1) by J.B. Orr during the 1960s. The specimens described herein were collected from a single locality along a small bluff on Anton Creek (Sec. 27, 28, T11S, R12), which the authors georeferenced to 44.8527 latitude, -113.0239 longitude, uncertainly 210 m.

Figure 1 Locality map of the Beaverhead Basin in southwestern Montana. All spinicaudatan specimens were collected from the Anton Creek locality along a tributary of the Medicine Lodge Creek. Correlative layers at a locality along Schwartz Creek produced fossil ostracodes, plants, and insects, but not spinicaudatans.

The Medicine Lodge Formation and associated sedimentary units within the Beaverhead Basin include interbedded thin-to-thick shales and sandstones, including a basal conglomerate and a localized coal layer near the top, with noticeable vertical and lateral variations (Scholten et al., Reference Scholten, Keenmon and Kupsch1955; Becker, Reference Becker1969; DeVore and Pigg, Reference DeVore and Pigg2010; Lielke et al., Reference Lielke, Manchester and Meyer2012). These facies represent fluvial, lacustrine, fan-delta, and paludal facies (M’Gonigle and Darymple, 1996) that were deposited within a structural basin bounded by the hanging wall of a Paleogene low-angle fault on the west and the low-angle normal fault to the east (Scholten et al., Reference Scholten, Keenmon and Kupsch1955; Becker, Reference Becker1969; DeVore and Pigg, Reference DeVore and Pigg2010; Lielke et al., Reference Lielke, Manchester and Meyer2012). Based on paleobotanical data (e.g., Scholten et al., Reference Scholten, Keenmon and Kupsch1955; Becker, Reference Becker1969; DeVore and Pigg, Reference DeVore and Pigg2010; Lielke et al., Reference Lielke, Manchester and Meyer2012), the basin was estimated to have an elevation of ~1600 m above sea level.

Becker (Reference Becker1969) described three florules from Paleogene basin-lake deposits in Beaverhead County, Montana: the Christensen, Horse Prairie, and Medicine Lodge florules. The Medicine Lodge Florule is the youngest of these and crops out along several adjacent tributaries of Medicine Lodge Creek. The age of the Medicine Lodge Formation is constrained by radiometric dates and refined using biotic data. Tuffs at the base of the sedimentary sequence at Medicine Lodge and Horse Prairie Basins exhibit ages of 46.04±0.15 Ma to 45.0±0.13 Ma, indicating the initiation of basin fill during the middle Eocene (M’Gonigle and Dalrymple, Reference M’Gonigle and Dalrymple1993, Reference M’Gonigle and Dalrymple1996), whereas overlying basalts have been dated at 27.5 Ma (late Oligocene; Janecke et al., Reference Janecke, Mcintosh and Good1999). Within this constraint, Lielke et al. (Reference Lielke, Manchester and Meyer2012) determined that the preponderance of the evidence, including floral similarity with other sites, was consistent with an earliest Oligocene age (Rupelian Stage, Orellan NALMA) for the three florules described by Becker (Reference Becker1969).

The Medicine Lodge Formation preserves a diverse biotic assemblage. The described flora contains 94 species belonging to 66 genera, including both dicots and conifers, representing a mixed forest biome (Wing, Reference Wing1987; Graham, Reference Graham1999; DeVore and Pigg, Reference DeVore and Pigg2010). In addition to the spinicaudatans, the invertebrate component of the fauna includes insects (including caddisfly larval cases), freshwater gastropods tentatively assigned to Planorbidae, the bivalve Sphaerium sp., and several species of ostracodes (Becker, Reference Becker1969; Lewis, Reference Lewis1972). Vertebrates include relatively common fish scales and rare articulated specimens provisionally assigned to Hiodontidae and Catostomidae. With the exception of an abstract by Cavender (Reference Cavender1977), the fish fauna remains undescribed. Becker (Reference Becker1969) also mentioned the presence of large, unidentified mammal bones.

Systematic paleontology

Phylum Arthropoda von Siebold, 1848

Subphylum Eucrustacea Walossek, Reference Walossek1999

Class Branchiopoda Latreille, Reference Latreille1817

Order Diplostraca Gerstaecker, Reference Gerstaecker1866

Remarks

Spinicaudata, Laevicaudata, and Cyclestherida were once united in the paraphyletic “order Conchostraca.” Recent phylogenetic analyses (e.g., Richter et al., Reference Richter, Olesen and Wheeler2007; Olesen, Reference Olesen2009) have cogently demonstrated that this grouping is paraphyletic. However, the monophyly of the Spinicaudata, which includes the vast majority of “clam shrimp,” is firmly established (e.g., Olesen, Reference Olesen2000, Reference Olesen2009; Stenderup et al., Reference Stenderup, Olesen and Glenner2006; Richter et al., Reference Richter, Olesen and Wheeler2007; Schwentner et al., Reference Schwentner, Timms, Bastrop and Richter2009). A recent review by Astrop and Hegna (Reference Astrop and Hegna2015) examined the relationships among living and fossil spinicaudatans, and their recommendations are followed herein. The systematic terminology used herein follows that of Olesen (Reference Olesen2000) and Martin and Davis (Reference Martin and Davis2001) for higher-level classification and follows Chen and Shen (Reference Chen and Shen1985) for superfamilial and lower levels. Measurements and terminology of carapace features follow Tasch (Reference Tasch1987).

As implied by common name “clam shrimp,” conchostracans are sometimes mistaken for bivalves and vice versa due to their similar gross morphologies of bivalved shells with concentric growth lines. The lightly biomineralized chitinous carapace, ornamentation and growth patterns, and life habits of conchostracans produce a suite of features that allow confident differentiation, many of which are described by Tasch (Reference Tasch1987). Conchostracan valves typically do not disarticulate post-mortem and frequently are preserved overlapping on bedding planes with conspicuous wrinkles due to desiccation and compaction (e.g., Figs. 2.5, 3.1, 5.1; Li and Batten, Reference Li and Batten2005, figs. 3, 4; Shen et al., Reference Shen, Gallego, Buchheim and Biaggi2006, fig. 4; Stigall et al., Reference Stigall, Hembree, Gierlowski-Kordesch and Weismiller2014, fig. 5). Conchostracan carapaces range from elongate to ovate, but in general are characterized by a narrower anterior and a broader posterior angle (see Chen and Shen, Reference Chen and Shen1985; Tasch, Reference Tasch1987; Scholze and Schneider, Reference Scholze and Schneider2015). Conchostracan growth bands are thickest and often more strongly preserved at the distal margin (e.g., Fig. 5.1, 5.2, 5.5; Orr et al., Reference Orr, Briggs and Kearns2008, fig. 5; Astrop and Hegna, Reference Astrop and Hegna2015, fig. 4). Outer layers of carapace growth bands are often replaced or remineralized in patches or as overgrowths (e.g., Fig. 3.1, 3.3; Shen et al., Reference Shen, Gallego, Buchheim and Biaggi2006, fig. 4; Stigall et al., Reference Stigall, Babcock, Briggs and Leslie2008, figs. 6, 12; Gallego et al., Reference Gallego, Cabaleri, Armella, Volkheimer, Ballent, Martínez, Monferran, Silva Nieto and Páez2011, fig. 5). Growth bands frequently appear to be shingled along the distal carapace margin (e.g., Fig. 4.1; Stigall et al., Reference Stigall, Hembree, Gierlowski-Kordesch and Weismiller2014, figs. 5, 6) and sometimes include radial canals (lirae) or serrations for setal exit (e.g., Fig. 4.7; Li and Batten, Reference Li and Batten2005, fig. 3; Boukhalfa et al., Reference Boukhalfa, Li, Ben Ali and Soussi2015, fig. 4). Finally, conchostracans rarely occur in isolation; they more typically occur in massive bedding-plane concentrations (e.g., Fig. 5.3; Monferran et al., Reference Monferran, Gallego, Astrop and Cabaleri2013, fig. 3). The specimens described herein meet each of these criteria for inclusion within the “Conchostraca.”

Figure 2 Part and counterpart specimens of the type series of Estherites? jocelynae n. sp. (1, 2) Part and counterpart slab with holotype specimen (YPM 531187) in lower center and two paratype specimens (YPM 531188, 531189), scale=1mm, white boxes indicate boundaries of Figure 5.10 and Figure 4.1–4.4; (3, 6, 9, 12) photomicrograph and SEM images of the holotype part and counterpart (YPM 531187); distal edges of growth bands marked by depression in molds, whereas surface ornament preserved best near distal margins; (4, 7, 10, 12) photomicrograph and SEM images of the paratype YPM 531188 part and counterpart, scale=1 mm; (5, 8, 11) photomicrograph and SEM images of paratype YPM 531189 part and counterpart, scale=1 mm; (13) close up of distal growth bands of the holotype (YPM 531187) outlined in box in Figure 2.6, scale bar=100 μm; (14) close up of centrally located growth bands of the paratype YPM 531189 outlined in box in Figure 2.8, scale bar=50 μm.

Figure 3 Images and energy dispersive x-ray spectral data for paratype YPM 531190 of Estherites? jocelynae n. sp. (1, 2) Photomicrograph and SEM image of entire left carapace valve, scale bar=1 mm; (3) close up of central carapace outlined with white box in Figure 3.2, note lack of detailed ornamentation on the preserved growth band, scale bar=50 μm; (4) spectral data for centromedial location on carapace where growth bands have been exfoliated, marked as + in Figure 3.2, composition is dominated by silicates; (5) spectral data for posterodistal location on carapace where growth bands are preserved as a thick layer, marked as + on lower right of Figure 3.2, composition is dominated by calcium ions; (6) spectral data for central carapace where growth bands are preserved as a thick layer, marked as + on lower right of Figure 3.3, composition is dominantly calcium carbonate.

Figure 4 SEM images and energy dispersive x-ray spectral data for holotype counterpart (YPM 531187) and paratype counterpart (YPM 531188) of Estherites? jocelynae n. sp. (1) Image of the carapace margins of YPM 531188, left, and YPM 531187, right, as indicated by white box in Figure 2.2, scale bar=500 μm; (2–4) EDAX map with same dimensions as Figure 4.1, scale bars=500 μm; elemental maps of illustrate the relative concentrations of Ca, Si, and Al, respectively. Areas with preserved growth bands have high calcium content but lack silica and aluminum and vice versa; (5) view of distal growth bands of paratype YPM 531188 outlined in the upper left corner white box in Figure 4.1, radial structures indicated by arrow, scale bar=200 μm; (6) close up of distal margin of a growth band of the holotype (YPM 531187) outlined by the lower right white box in Figure 4.1, scale bar=25 μm, notice the thickened distal edge, which corresponds with the stronger depressions marking the distal edge of the growth bands in moldic specimens; (7) spectral data for the growth band illustrated in Figure 4.6, location of scan marked as + in 4.6, composition is dominantly calcium carbonate.

Figure 5 Images of additional spinicaudatan and associated fossils. (1–7) Additional specimens of E.? jocelynae n. sp.: (1) complete right carapace valve (YPM 531191) and partial right? valve preserved with growth bands marked by calcium carbonate replacement (YPM 531192), scale bar=1 mm; (2) posterior portion of a left carapace valve (YPM 531198) underneath the distal margin of another valve (YPM 531199), scale bar=1 mm; (3) dense assemblage of iron-stained carapace molds (YPM 531205–531214) on the underside of slab containing the holotype specimen illustrated in Figure 2.1, scale bar=5 mm; (4) anterior portion of a carapace (YPM 531203), scale bar=1 mm; (5) two carapaces replaced with carbonate (YPM 531200, upper; 531201, lower), scale bar=1 mm, margins of these specimens appear warped, probably from desiccation before burial; (6) laterally compressed specimen (YPM 531202), scale bar=1 mm; (7) E.? jocelynae n. sp. specimen YPM 531218 intersected by an iron-rich Planolites burrow, scale bar=5 mm; (8) gastropod, YPM 531204, scale bar=1 mm; (9) bedding-plane assemblage of associated fauna including ostracodes and macerated insect exoskeleton (YPM 531216), scale bar=1 mm; (10) SEM image of iron-rich burrow, field of view indicated by white box in Figure 2.1, scale bar=250 mm; (11), assemblage of Planolites burrows (YPM 531215) on the holotype slab in Figure 2.1, scale bar=5 μm.

Suborder Spinicaudata Linder, Reference Linder1945

Superfamily Estheriteoidea Zhang and Chen in Zhang et al., Reference Zhang, Chen and Shen1976

Family Estheriteidae Zhang and Chen in Zhang et al.,Reference Zhang, Chen and Shen1976

Genus Estherites? Kobayashi and Huzita, Reference Kobayashi and Huzita1943

Description

Carapace valves of moderate size and elliptical to oval in outline. Mean carapace length and height are 3.89 mm (range: 1.94–5.87 mm) and 3.13 mm (range: 1.53–5.06 mm), respectively (Table 1). The L/H ratio averages 0.81 (range: 0.70–0.90). Carapace growth is isometric (Fig. 6). Dorsal margin is weakly curved and relatively short (typically less than half of the valve length) with umbo centrally located. Anterior margin shorter and more tightly rounded than posterior. Maximum anterior bulge is approximately 0.38x distance from hinge line to ventral margin, and maximum posterior bulge located approximately 0.43x distance from hinge line to ventral margin. Maximum ventral bulge is located slightly anterior of center, approximately 0.38x distance from anterior to posterior of carapace.

Figure 6 Growth plot of E.? jocelynae n. sp. specimens showing isometric growth; Height and Length in mm.

Table 1 Summary of measured data for Estherites? jocelynae n. sp. specimens. Length=maximum carapace length; Height=maximum distance from the dorsal margin to the hinge line; Hinge=length of the hinge line; A=distance from the maximum anterior bulge to the dorsal margin; B=distance from the maximum posterior bulge to the dorsal margin; C=distance from the maximum ventral bulge to the most anterior part of the valve; D=distance from center of the umbo to the anterior margin; units=mm; N=number of specimens measured.

Moldic specimens and those with preserved carapace valves exteriors record an average of eight (range: 7–14) flat growth bands ~0.4 mm wide (e.g., Figs. 2.3, 3.1, 5.1, 5.5). The distal margins of these growth bands are demarcated by a topographic impression within the mold and/or a thickened region of preserved carapace. Many specimens preserve an internal mold of the carapace valve (e.g., Figs. 2.4, 2.7, 5.2, 5.3, 5.4, 5.6) and, consequently, do not preserve any external ornament. One specimen (YPM 531189, Fig. 2.5, 2.8, 2.14) represents a carapace exterior and preserves a series of narrow concentric bands throughout the carapace, which suggests development of narrower growth bands mediodistally. Limited regions around the extreme distal margins of some specimens, notably specimens of the type series, preserve sets of narrow growth bands, ~40 μm wide, as siliceous molds (Figs. 2.13, 2.14, 3.5). Weakly preserved ornamentation patterns of radial lirae and concentric wrinkles are present on some of these narrow, extreme-distal growth bands (Fig. 4.5). There is no evidence for radial ornamentation in the umbo or central region of specimens where that region has been preserved as a siliceous mold. The preservation of a consistent number of wide growth bands among internal molds coupled with the occasionally preserved narrow bands suggests that two stages of concentric band development occurred during ontogeny.

Preservation

Specimens of Estherites? jocelynae n. sp. are preserved as combinations of fossilized carapace material and corresponding molds of the carapace interior and exterior surfaces (e.g., Fig. 2.1). Where present, the carapace material is preserved as a white opaque mineral that may be tinted orange (e.g., Fig. 2.2). Spectra obtained from EDAX analyses indicate that the white mineral is composed primarily of calcium ions with minor peaks at carbon and oxygen and accessory peaks for iron, silica, and aluminum (Figs. 3, 4). This carbonate-dominated composition differs from the elemental composition of the sedimentary background, in which calcium is negligible and aluminum and silicon ions are abundant (Fig. 4). The carbonate preservation also differs from the presumed chitin-calcium-phosphate composition of the original carapace material (for discussion of modern spinicaudatan carapace composition see Stigall et al., Reference Stigall, Babcock, Briggs and Leslie2008; Astrop et al., Reference Astrop, Sahni, Blackledge and Stark2015). This indicates the fossils represent replaced rather than recrystallized carapace material. In regions where a detailed mold of the carapace exterior is preserved (e.g., Figs. 2.13, 4.4–4.6), silicon is the primary ion (Fig. 3.3). Similar high-fidelity preservation via silica molds has been documented in other spinicaudatans, such as Carapacestheria disgregaris from the Jurassic of Antarctica (Stigall et al., Reference Stigall, Babcock, Briggs and Leslie2008).

Compared to other fossil spinicaudatan carapaces, the preservation of the Beaverhead Basin specimens is unusual. Notably, although hundreds of fossil spinicaudatan species have been described, the elemental composition of carapaces has been explicitly tested for fewer than ten species. Thus sets of the comparative data are extremely limited. Most fossil species for which carapace compositions have been analyzed via EDAX are within the ‘Estheriteoidea-Eosestheriodidea’ group, and their carapaces have reported to be preserved primarily as recrystallized calcium phosphate or silica molds. This group includes Triassic specimens from Poland (Olempska, Reference Olempska2004), Jurassic specimens from Antarctica (Stigall et al., Reference Stigall, Babcock, Briggs and Leslie2008), Namibia (Stigall et al., Reference Stigall, Hembree, Gierlowski-Kordesch and Weismiller2014), and China (Hethke et al., Reference Hethke, Fürsich, Jiang and Pan2013), and Cretaceous specimens of Madagascar (Stigall and Hartman, Reference Stigall and Hartman2008). Another identified mode of spinicaudatan preservation, specifically related to exceptional preservation of soft tissues and appendages, is as organic-carbon residues associated with phyllosilicates (Orr and Briggs, Reference Orr and Briggs1999). Carbonate preservation of spinicaudatans has only previously been reported from the Jurassic specimens of western North America (Lucas and Milner, Reference Lucas and Milner2006). Notably, both carbonate and organic-carbon preservation result in lower resolution of carapace details than preservation as recrystallized calcium phosphate or silicic molds.

Although studies of spinicaudatan carapace taphonomy are limited, carbonate replacement or preservation has been studied within both arthropods and vertebrates preserved in freshwater environments. Calcium carbonate precursors have been observed to form from the breakdown of organic matter in aqueous solution (Berner, Reference Berner1968) and have been implicated in the development of concretions as well as exoskeleton and soft-tissue replacement in insects and shrimp (McCobb et al., Reference McCobb, Duncan, Jarzembowski, Stankiewicz, Wills and Briggs1998; Sageman et al., 1999; Park and Downing, Reference Park and Downing2001; Martínez-Delclòs et al., Reference Martínez-Delclòs, Briggs and Peñalver2004). The degree that phosphate is replaced by carbonate within vertebrate bioapatite has been shown to vary among depositional environments due to local geochemical conditions (Keenan et al., Reference Keenan, Engel, Roy and Bovenkamp-Langlois2015), variations in pH (Sagemann et al., Reference Sagemann, Bale, Briggs and Parkes1999), or as a byproduct of bioerosion by aquatic microorganisms (Fernández-Jalvo et al., Reference Fernández-Jalvo, Pesquero and Tormo2016). In some instances (e.g., Fernández-Jalvo et al., Reference Fernández-Jalvo, Pesquero and Tormo2016), bones have been replaced with a chalky version of calcium cabonate that closely resembles the preservation style of E.? jocelynae n. sp. Because spinicaudatan carapaces are primarily chitin interlayered with limited bioapatite (~5% phosphate compared to ~40% in vertebrates), it is likely that low pH lake waters or early diagenetic processes, including bioerosion, could have removed the phosphate and facilitated later carbonate precipitation within voids held open by the chitin remnants. The atypical preservation exhibited by E.? jocelynae n. sp., therefore, may have been facilitated by specific geochemical or microbial conditions within the depositional environment. Additional analyses on preservation of the spinicaudatan and associated fauna of the Beaverhead Basin are in progress to investigate the relationship between preservation and geochemistry in more detail.

Evolutionary and environmental significance

Clam shrimp are common in lacustrine deposits of late Paleozoic and Mesozoic age, but rarely have been reported from Cenozoic strata (see Zhang et al., Reference Zhang, Chen and Shen1976; Chen and Shen, Reference Chen and Shen1981; Shen et al., Reference Shen, Gallego, Buchheim and Biaggi2006; Gallego and Mesquita, Reference Gallego and Mesquita2010). The only previously described Cenozoic clam shrimp from North American are two species from the Eocene Green River Formation (Shen et al., Reference Shen, Gallego, Buchheim and Biaggi2006). These species, Cyclesteriodes wyomingensis and Prolynceus laneynsis, are members of the orders Cyclestheriida and Laevicaudata, respectively. Thus, Estherites? jocelynae n. sp. is the first spinicaudatan “conchostracan” described from the Cenozoic of North America. Estherites? jocelynae n. sp. is also the second youngest clam shrimp to have been described globally—only Paraleptestheria menglaensis (Gallego and Mesquita, Reference Gallego and Mesquita2010) from the late Oligocene of Brazil is younger.

All other species of Estherites are known from older deposits of northeast Asia, thus the presence of E.? jocelynae n. sp. in the Oligocene deposits of the Beaverhead basin suggests that the ancestors of this species dispersed to North America from Asia. Lielke et al. (Reference Lielke, Manchester and Meyer2012) documented an increase in the number of plant species of East Asian affinity within the Medicine Lodge Formation, and the spinicaudatans may have utilized the same dispersal pathway. Estherites? jocelynae n. sp. is also the youngest species of this genus and demonstrates the persistence of this clade beyond the Cretaceous-Paleogene extinction.

Spinicaudatans are explicit environmental indicators because they are restricted to freshwater habitats that exhibit seasonal fluctuations in water availability (Thiéry, Reference Thiéry1996; Vannier et al., Reference Vannier, Thiéry and Racheboeuf2003). As part of their reproductive pathway, many spinicaudatans species produce cysts that must undergo drying before hatching can occur (Thiéry, Reference Thiéry1996; Vannier et al., Reference Vannier, Thiéry and Racheboeuf2003). Thus the basic life-history attributes of spinicaudatans match the highly seasonal, summer dry climate with a mean summer temperate of 30.5°C reconstructed by Lielke et al. (Reference Lielke, Manchester and Meyer2012) for the Beaverhead Basin quite well.

Additional fauna from the spinicaudatan-bearing locality provides further insight to the depositional environment. The presence of fish remains, ostracodes (Fig. 5.5, 5.9), insects—including caddisfly cases (Fig. 5.9), bivalves, and freshwater gastropods (Fig. 5.8) indicates that portions of the lake retained water year-round, although water levels likely fluctuated, allowing marginal regions to dry enough for spinicaudatan cyst development. The presence of a limited number of Planolites burrows (Fig. 5.7, 5.11) indicates that deposit feeders were present, but uncommon on the lake bottom. A fairly limited infaunal population would have facilitated the preservation of the relatively delicate spinicaudatan carapaces.