A Laurentian record of the earliest fossil eukaryotes

The fossil record of eukaryotes extends back to at least the late Paleoproterozoic (Butterfield, 2015), and a recurrent assemblage of distinctive microfossils from China (Ruyang Group; Yin, 1997; Yin et al., 2005), Australia (Roper Group; Javaux et al., 2001; Javaux and Knoll, 2016), India (Semri Group; Prasad et al., 2005), and Siberia (Kamo Group; Nagovitsin, 2009) points to the presence of an ecologically coherent biota extending from ca. 1650 Ma to ca. 1400 Ma. Typified by process-bearing Tappania, reticulate-walled Dictyosphaera/Shuiyoushaeridium, and concentrically striated Valeria, these earliest fossil eukaryotes offer key insights into the early evolution of the clade, and enable practical applications to paleobiogeography and biostratigraphic correlation.

Proterozoic eukaryotes are also richly represented on Laurentia; however, their oldest confirmed records on this craton are of late Mesoproterozoic age (Butterfield and Chandler, 1992; Butterfield, 2000; Hofmann and Jackson, 1994). Older fossils are known from North America, notably Grypania and Horodyskia from the early Mesoproterozoic Belt Supergroup (Walter et al., 1976; Fedonkin and Yochelson, 2002), but neither of these macroscopic forms is unambiguously eukaryotic (Butterfield, 2009). Apart from our report of long-ranging Valeria in the Chamberlain Formation (Adam et al., 2016), the same is true for spheroidal and filamentous microfossils that have been recovered from Belt Supergroup strata (Horodyski, 1980). Here we report a full Tappania-Dictyosphaera-Valeria assemblage from the Greyson Formation of the lower Belt Supergroup, filling a major gap in the early eukaryotic record.

The Belt Supergroup contains as much as 15 km of strata, extending from British Columbia to central Montana, including an eastward-extending limb termed the Helena embayment (Winston and Link, 1993) (Fig. 1A). This intracratonic sedimentary succession has been interpreted as a restricted marine (Horodyski, 1993; Lyons et al., 2000), possibly even a lacustrine (Winston, 1993), deposit. The paleogeographic relationships of the Belt Basin are poorly constrained, with different reconstructions identifying Siberia, Australia, or Antarctica as the conjugate rift margin (Sears and Price, 2003).

The Helena embayment preserves the lowermost strata of the Belt Supergroup, representing a broadly transgressive-regressive sedimentary succession. The fossils described here come from unoxidized shales of the Greyson Formation (Montana, USA) that overlie subtidal carbonate-shale deposits of the Newland Formation, and are overlain in turn by mudcracked redbeds of the Spokane Formation (Fig. 1B). The Greyson Formation crops out extensively in the Little and Big Belt Mountains, including the type section of Grypania at Deep Creek (Walter et al., 1976), and is broadly correlative with the Appekunny Formation in Glacier National Park (Fig. 1A; Horodyski, 1993; Slotznick et al., 2016), including the type section of Horodyskia (Fedonkin and Yochelson, 2002). Our study is based on a roadcut section of the Greyson Formation ∼50 km northeast of Deep Creek, along Newlan Creek Road near White Sulphur Springs, Montana (46.668°N, 110.884°W) (Fig. 1C), where it comprises ∼1100 m of finely laminated mudstones with isolated sandstone layers. We recognize three informal constituent units: (1) a lower siltstone member (∼350 m); (2) a middle shale member (∼435 m); and (3) an upper siltstone-sandstone member (∼315 m). Multiple small-scale shallowing-upward sequences in the two upper members are interpreted to reflect sedimentation in shallow-shelf to subshoreface environments.

The age of the Greyson Formation is constrained by a 1454 ± 9 Ma U-Pb date from a bentonite layer in the overlying Helena Formation of the Middle Belt Carbonate Group (Evans et al., 2000), and is corroborated by 1470–1460 Ma sills intruding correlative Lower Belt Group strata of the Prichard Formation (Sears et al., 1998). The maximum age is set as 1710 Ma by the youngest age of a suite of zircons recovered from the underlying Neihart Quartzite (Mueller et al., 2016), although basement rocks in northern Idaho dated at 1576 ± 13 Ma underlie a unit thought to be correlative to the Neihart Quartzite (Evans and Fischer, 1986; see the GSA Data Repository¹).

Nine unoxidized mudstone horizons of the Greyson Formation at Newlan Creek were collected for microfossil processing. Samples were broken into 1–3 cm fragments, allowed to disaggregate with minimal agitation in 48% hydrofluoric acid, and fossils were individually picked from rinsed slurries.

Microfossils were recovered from five horizons, all within or immediately below the shale member (Fig. 1B). The assemblages are dominated by simple spheroidal microfossils up to 300 µm in diameter (Leiosphaeridia spp.) regularly accompanied by a variety of paired, colonial, and budding forms, including Satka favosa and Gemmuloides doncookii; medial-split release structures are also common (Figs. 2C–2F, 2H). Such morphologies are common among extant protistan-grade eukaryotes, but the absence of any demonstrably crown-group features leaves these fossils taxonomically unresolved. Filamentous microfossils (Siphonophycus, Oscillatoriopsis; Figs. 2A, 2B) are most likely benthic photosynthesizers (Butterfield and Chandler, 1992), providing an independent measure of shallow-water deposition in the Greyson Formation. The longitudinally striated filament Lineaforma elongata (Figs. 2G, 2I) may prove to be biostratigraphically useful with all currently reported occurrences limited to the early Mesoproterozoic (Vorob’eva et al., 2015; Adam et al., 2016; Javaux and Knoll, 2016).

The most significant fossils in the Greyson microbiota are morphologically differentiated vesicles assignable to Tappania plana, Dictyosphaera macroreticulata, and Valeria lophostriata. Only Valeria (Figs. 3D, 3E) has been previously documented on Laurentia: in the middle Neoproterozoic Chuar Group of Arizona (Porter and Riedman, 2016), the late Mesoproterozoic Bylot basin system of Canada (Butterfield and Chandler, 1992; Hofmann and Jackson, 1994), and the Lower Belt Group Chamberlain Formation (Adam et al., 2016). With a total age range exceeding 700 m.y., it offers no useful stratigraphic resolution (Hofmann, 1999). There are, however, good grounds for recognizing its fine concentric microstructure as diagnostically eukaryotic (Javaux et al., 2001); modern analogs are found in the microfibrillar cell-wall construction of glaucophyte algae (cf. Willison and Brown, 1978).

Dictyosphaera is distinguished by its relatively large vesicles (50–300 µm) and a complex polygonal wall structure well beyond the limits of prokaryotic morphogenesis (Yin et al., 2005; Agić et al., 2015). Although the inner surface of the wall is not seen in the Greyson population (Figs. 3A–3C), the specimens correspond closely with the Ruyang Group type material in terms of outer wall ornamentation, size, medial split structures, and the presence of a large intracellular inclusion in some unsplit specimens (Fig. 3A; cf. Pang et al., 2013). The absence of an acanthomorphic (Shuiyousphaeridum) phase seen in the late Paleoproterozoic Ruyang Group and Semri Group biotas (Pang et al., 2013; Singh and Sharma, 2014; Agić et al., 2015) is most likely due to small sample size and local taphonomy.

The most conspicuous and common eukaryote in the Greyson Formation biota is Tappania plana (Fig. 4), a microfossil distinguished by tubular processes and broader neck-like extensions (Yin, 1997). Most Greyson Formation Tappania specimens correspond to previously described populations, but the large sample size reveals a substantially new account of its anatomy. Vesicles range from 30 to 150 µm ( = 89 µm, standard deviation = 23 µm, N = 53) with shapes varying from equidimensional to elongate (Fig. 4C). Both the processes (N ≤ 20) and neck-like extensions (N ≤ 4) vary in number and are randomly distributed on vesicles, although many specimens show clear evidence of hemispherical polarization (Figs. 4B, 4D, 4E, 4H–4J). Process branching shows a broad range of expression, including equidimensional (Figs. 4B,4E, 4H, 4I) and size-reductive bifurcations (Figs. 4B, 4I, 4J), second-order branching (Figs. 4B, 4H), and vesicles bearing multiple branched process (Figs. 4B, 4I). As in the Kamo Group and Roper Group populations, there are rare instances of processes with internal partitions (Fig. 4F). Some specimens also bear outgrowths directly comparable to T. tubata (Fig. 4J), a putative second species.

Greyson Formation Tappania specimens exhibit a number of features not seen in other populations, most notably the differentiation of an outer wall (Figs. 4A, 4B, 4G). Although often expressed as a simple enveloping sheath, this relatively translucent layer can also be seen forming processes that are not present, or only partially developed, on the more robust inner wall; such independent morphogenesis demonstrates a particularly sophisticated level of cytoskeletal control. A further subset of specimens has terminally flared processes (Figs. 4A, 4B, 4D, 4G, 4H), suggesting suspension within a third (unpreserved) outer envelope or adventitious substrate attachment.

In a conventional form-taxonomic context, the novel features and variation exhibited by Greyson Formation Tappania specimens would warrant the establishment of multiple new form genera and form species. Even within this relatively large population, however, there are no obvious modes in size or form. Such morphological continuity, anchored to the presence of diagnostic neck-like extensions, points to pronounced ontogenetic and/or ecophenotypic variation (cf. Yin et al., 2005; Javaux and Knoll, 2017). On the basis of current evidence, Tappania represents a single biological species, T. plana.

The complex wall morphologies and microstructures exhibited by Greyson Formation Tappania, Dictyosphaera, and Valeria specimens are diagnostically eukaryotic. Such features are entirely outside the capacity of any organisms lacking a fully motorized cytoskeleton and endomembrane system (Javaux et al., 2001; Cavalier-Smith, 2002; Pang et al., 2013; Agić et al., 2015; Butterfield, 2015). The sophistication of this cytological machinery is particularly apparent in Tappania, where the inducible development of at least three separate components (neck-like extensions, processes, and independently deployed outer wall) demonstrates a level of morphogenetic control comparable to that of all but the most derived living protists. In this light, there is an increasingly secure case for placing Tappania within, or immediately adjacent to, crown-group Eukarya, despite its unresolved affiliations (Javaux and Knoll, 2016). This same conclusion may be drawn for Dictyosphaera and Shuiyousphaeridium (Pang et al., 2013; Agić et al., 2015), although not for fundamentally simpler Valeria.

All previous records of Tappania and Dictyosphaera came from a relatively narrow range of late Paleoproterozoic to early Mesoproterozoic strata, pointing to their utility as biostratigraphic markers (Yin et al., 2005; Agić et al., 2015; Javaux and Knoll, 2017). This potential is bolstered by the presence of two distinct forms exhibiting the same temporal distribution, along with a third (Valeria) that is conspicuously different. Extension of a Tappania-Dictyosphaera-Valeria assemblage zone into Laurentia now confers a corroborating global signal. Notably, the ca. 1450 Ma minimum age for the Greyson Formation biota overlaps the age of Roper Group fossils (cf. Javaux et al., 2001), which are bracketed in turn by the marginally younger Kamo Group and older Ruyang Group and Semri Group biotas.

A Tappania-Dictyosphaera-Valeria biozone is further supported by the accompanying occurrence of Grypania and Horodyskia. Despite the fundamentally unresolved affiliations of these oldest-known macrofossils, both contribute a possible biostratigraphic signal. In addition to its type occurrence in the Greyson Formation, Grypania spiralis is known from the early Mesoproterozoic Changchengian System of north China (Du et al., 1986) and the late Paleoproterozoic Semri Group of India (Sharma and Shukla, 2009), along with a less resolved population in the 1.87 Ga Negaunee Iron-Formation of Michigan, USA (Han and Runnegar, 1992). Horodyskia is biologically more problematic and extends from the early Mesoproterozoic Appekunny Formation in Glacier National Park (Montana) through Mesoproterozoic–Neoproterozoic strata in Australia (Grey et al., 2010; Calver et al., 2010) and the Ediacaran of south China (Dong et al., 2008). Even so, its first appearance notably is within the narrow time frame delineated by the cooccurrence of Tappania plana, Dictyosphaera macroreticulata, Valeria lophostriata, and Grypania spiralis.

Insofar as the Greyson Formation in the Helena embayment is stratigraphically contiguous with the Appekunny Formation in Glacier National Park, this shallow shelf succession is unique in preserving a complete roster of all known early eukaryotic and macroscopic fossil taxa, along with a comprehensive record of simple spheroidal and filamentous form taxa. The succession demonstrates a clear connection between the Helena embayment and the global oceans (cf. Winston, 1993), with important implications for early Proterozoic paleogeography and the global distribution of the planet’s oldest known eukaryotes.

Z. Adam was supported by the National Science Foundation Graduate Research Fellowship Program, the Lewis and Clark Fund for Exploration Research (NASA Astrobiology Institute and American Philosophical Society), the Tobacco Root Geological Society, the Belt Association, the D.L. Smith Fund, and a Geobiology Postdoctoral Fellowship from the Agouron Institute. We thank D. Winston, G. Zieg, J. Schmitt, and D. Lageson for useful discussions. We thank S. Xiao, E. Javaux, and S. Porter for insightful manuscript review comments.