We report a diverse assemblage of trace fossils from the Lower(?) Cambrian High Head member of the Church Point Formation near Yarmouth, Nova Scotia. Based on the presence of sharp-based beds, load casts, and flute casts, along with other sedimentological criteria, the strata are interpreted to have a turbidite origin. Other characteristics, including wrinkle structures and high fidelity of trace-fossil preservation, suggest that the inter-episode biotope was influenced by the presence of biomats. The trace-fossil assemblage includes (in approximate order of abundance) Planolites, Helminthopsis, Oldhamia, Chondrites, Gordia, Cladichnus, Psammichnites gigas, Treptichnus, Phycodes, Lorenzinia, Palaeophycus, and Teichichnus. Trace fossils are increasingly abundant upwards in the section, but they are ultimately mitigated by an increase in (inferred) sedimentation rates. The trace-fossil assemblage, which bears many similarities to other Lower Cambrian locales (e.g., Puncoviscana Formation of northwest Argentina, Chapel Island Formation of southeastern Newfoundland), and the distribution of ichnofossils strongly support an earliest Cambrian age for the studied strata. More importantly, evolutionary trends can be interpreted from the trace-fossil assemblage, namely (1) an increase in size and diversity of animals in deep water; (2) an expansion of motile strategies used to coexist with biomats; and, (3) increasingly complex mining strategies below the sediment-biomat interface.
Because of the paucity of body fossils in the Lower Cambrian, the biostratigraphic indicator of the start of the Lower Cambrian System is placed at the base of the Treptichnus pedum ichnozone (Brasier et al. 1994). In fact, at several locales trace-fossil occurrences precede preserved body fossils, and the documentation of Ediacaran–Cambrian trace fossils provides a crucial element in interpreting the early Paleozoic rock record (e.g., Conway Morris 1989; Seilacher 1999; McIlroy and Logan 1999; Buatois and Mangano 2004). Just as importantly, evolutionary trends are interpreted from trace fossils and trace-fossil assemblages, and those trends are useful in establishing the relative ages of otherwise fossil-barren strata (e.g., Narbonne et al. 1987; Crimes 1992; Seilacher 1994).
The Meguma terrane of southern Nova Scotia is dominated by the generally unfossiliferous sandstone-dominated Goldenville Group (Fig. 1). However, within the Goldenville Group in the High Head area north of Yarmouth (Fig. 1), White et al. (2001) recognized a distinctive fine-grained unit, 845 m in thickness, which is extremely rich in trace fossils. This unit was subsequently termed the High Head member (White et al. 2005). Age relationships in the Goldenville Group are poorly constrained by a single trilobite faunule containing Middle Cambrian taxa on Tancook Island (near the Goldenville top; Pratt and Waldron 1991) and by detrital zircon data (Waldron et al. 2009). The trace-fossil assemblage in the High Head member, however, provides a means to establish a biostratigraphic and paleoenvironmental context, which is the primary aim of this paper.
Study area and geological setting
Meguma is the most outboard of the Appalachian terranes in Atlantic Canada (e.g., Hibbard et al. 2006). The terrane takes its name from the Meguma Supergroup (White 2008) (formerly Meguma Series or Group; Woodman 1904; Stevenson 1959), a thick Cambrian–Ordovician deep-water succession without recognized correlatives elsewhere in the orogen. The Meguma terrane is interpreted to have originated on the margin of Gondwana in Late Proterozoic time (e.g., Waldron et al. 2009), but its history of subsequent interaction with Gondwana, and with Avalonia (which now lies to the north and west; Fig. 1), is not well constrained. Collision of the Meguma terrane with Laurentia may have been responsible for the Middle Devonian to Early Carboniferous Neoacadian orogeny in the Meguma terrane and possibly elsewhere in the northern Appalachians (e.g., van Staal 2007).
The Meguma Supergroup is divided into a lower, sandstone-dominated Goldenville Group and an upper unit (Halifax Group) dominated by mudrocks (i.e., variably cleaved, fine-grained, low-grade sedimentary and low-grade metasedimentary rocks whose protoliths were siltstone and claystone). The base of the Goldenville Group is not seen in outcrop, but the exposed section is at least 8 km thick in southwestern Nova Scotia (White et al. 2001). The overlying Halifax Group is represented by approximately 5 km of section in this area. Both groups are folded and generally cleaved; strain ratios on bedding planes range up to about 2:1 (Waldron 1988), suggesting that the strata might have been tectonically thickened to a maximum of double their original thickness. Both groups are interpreted as predominantly turbiditic in origin (e.g., Phinney 1961; Schenk 1970; Waldron and Jensen 1985; Waldron 1992). Evidence for the age of the two groups is sparse. The Goldenville Group contains the aforementioned (stratigraphically higher) trilobite faunule of Middle Cambrian age (Pratt and Waldron 1991). Lower parts of the overlying Halifax Group are devoid of fossils, but towards the top of the group, the graptolite Rhadinopora (formerly Dictyonema) has been identified from several localities (e.g., Cumming 1985; Smitheringale 1973; White et al. 1999), indicating Tremadocian (Early Ordovician) age. In addition, Waldron et al. (2009) documented a detrital zircon population from the lowest exposed sandstone in the Goldenville Group, with a youngest concordant grain at 544 ± 18 Ma, providing a maximum depositional age for the formation. A sample at High Head, the location of this study (Fig. 2), approximately 4.5 km above the base of the group, contained a slightly younger grain at 537 ± 15 Ma, again constraining the maximum age of deposition to Cambrian (or, conceivably, within error, latest Proterozoic).
Meguma terrane successions differ slightly in character on either side of the Chebogue Point Shear Zone (CPSZ, Fig. 1). For example, to the east the Goldenville–Halifax transition is notably manganese-enriched, whereas to the west it is not; sections to the west show numerous mafic sills, which are generally absent to the east (Waldron et al. 2009).
The section described here is located to the west of the CPSZ, where sandstone of the Goldenville Group has been assigned to the Church Point Formation (White 2008). The High Head member is a distinctive fine-grained unit within the Church Point Formation and is continuously exposed on the coast on either side of High Head, Yarmouth County. Inland exposure is poor, but the unit can be traced along strike for about 48 km on the basis of its aeromagnetic signature (White et al. 2001, 2005).
At the study location, a continuous section of ∼930 m was logged in detail (Fig. 3). Physical sedimentary structures and trace fossils were recorded at the levels of their occurrence. Trace-fossil diversity, size, distribution and mode of occurrence were also recorded.
The sedimentary strata of the High Head section are dominated by thinly bedded, planar-laminated mudrock (Figs. 4A, 4B). The mudrock shows a strong planar fabric defined by the alignment of fine-grained sheet silicates parallel to sedimentary lamination and bedding that dip ∼55° to the southeast. Locally, a separate cleavage is discernible, dipping slightly more steeply than bedding. In some parts of the section, a later crenulation cleavage is imposed on the bed-parallel fabric, axial planar to small-scale folds. These structures appear to be associated with rare steep faults that cut the section; correlation of strata across these faults suggests that they mainly extend bedding and, therefore, result in small stratigraphic omissions. Based on the general continuity of the section and the small width (<<1 m) of the associated damage zones, it is likely that no more than a few metres of section are omitted.
As a result of their tectonic fabric, together with fine-scale lithological lamination, the mudrock is mainly pinstriped slate (Figs. 4A, 4B). Most commonly this characteristic reflects variation between clay- and silt-dominated fractions; however, very fine-grained sandstone is locally interlaminated or interbedded with the slatey mudstone. Very fine-scale, low-angle cross-lamination is present locally, and rare bedding surfaces show low-relief, short-wavelength (10–20 cm) undulations interpreted to represent ripple marks; however, post-depositional strain has modified their geometry. Where present, the thin (1–3 cm), graded sandstone beds are sharp-based with rare, small load casts (Fig. 4B).
In the lower part of the section thicker, more abundant sand beds are typically decimetre scale, show planar and cross-laminae organized in partial Bouma (1962) sequences, and are interbedded with mudstone in up to metre-scale bedsets dominated by fine sandstone. These thicker sandstone beds also are sharply based and may display flute casts (Fig. 4C), convolute lamination, and diffuse dewatering pipes composed of slightly better sorted sandstone. One interval at the base of the section displays metre-scale irregularly trough-cross-bedded and internally scoured sandstone beds (Fig. 4D), similar to those described by Waldron (1992).
Similarly, in the medial part of the section (450 to 460 m) the mudrock-dominated succession is interrupted by metre-scale bedsets of medium- to fine-grained, graded sandstone beds that display complete and partial Bouma (1962) sequences of sedimentary structures. Rare water-escape pipes are present.
At the top of section (910 to 930 m), very fine- to fine-grained sandstone is the dominant lithology and occurs as very thin- to medium-graded (1–30 cm) beds with planar and cross-laminae in partial Bouma sequences, organized in bedsets 2–5 m thick. Mudrock intervals towards the top and base of the section contain diagenetic pyrite cubes up to 2 cm in diameter. Concretionary carbonate also occurs in these parts of the section, mainly in centimetre-scale, cross-laminated siltstone beds. At several points in the section, greenish-black mafic dykes cut the sedimentary rocks. Based on the geometry of xenoliths and contacts in these dykes, White and Barr (2004) suggested that they were intruded soon after deposition, while the sediments were still wet.
All the reported trace fossils remain in the studied outcrop, so repository numbers were not assigned. Trace fossils occur sporadically throughout the section. The nature of the exposure and the style of trace-fossil preservation dictated that most of the trace-fossil occurrences were observed on bedding planes. Although some penetrative forms were observed, this aspect was generally inferred from studying the bedding planes.
The distribution of the various ichnospecies observed, and their modes of occurrence, are discussed in detail as follows. Several ichnogenera were encountered, including (in approximate order of abundance) Planolites (e.g., Fig. 5A), Helminthopsis (Fig. 5C), Oldhamia (Figs. 5D, 5E), Chondrites (Figs. 5F, 5G), Gordia (Figs. 5B, 5G, 5H), Cladichnus (Fig. 5I), Palaeophycus (Fig. 5J), Psammichnites (Figs. 5K, 5P), Treptichnus (Figs. 5L, 5M), Phycodes (Fig. 5N), Lorenzinia (Fig. 5O), and Teichichnus (Fig. 5Q). Although the trace fossils are distributed throughout the section, apparently barren beds are nevertheless dominant. Trace-fossil occurrence and preservation are largely restricted to isolated beds; however, some intervals reveal diverse assemblages locally. No relationship was established that fully explains the distribution of bioturbation — i.e., although the burrowing dominantly occurs in thinly bedded parts of the succession, many thin beds are apparently barren — and we cannot discount the negative influence of bedding-parallel strain on the sporadic distribution of trace fossils. The distribution of ichnogenera is briefly summarized by stratigraphic interval, as follows.
0 to 445 m: Bioturbation in the lower part of the section occurs above 65 m, marking the approximate base of the High Head member. In the lower half of the studied exposures, Planolites, Helminthopsis, and Gordia, with rare Oldhamia, are most abundant. Although some beds are dominated by only one of these ichnogenera, several beds were identified that contain the first three trace-fossil types. Treptichnus pedum — occurs sporadically at 73, 95, 145, and 410 m.
445 to 580 m: A more diverse trace-fossil assemblage comprising Planolites, Chondrites, Cladichnus, Helminthopsis, Oldhamia, Phycodes, Treptichnus, and Lorenzina is present. Ichnological diversity decreases upwards, such that Planolites and Oldhamia are the dominant ichnofossil forms.
580 to 911 m (the top of the High Head member): Low-diversity assemblages dominated by Planolites occur to 700 m. Above that level, Psammichnites and Phycodes are sporadically present, with rare Helminthopsis. This assemblage grades up into strata dominated by Planolites, with rare Chondrites. The low diversity recorded in the uppermost part of the section, from ∼830 to 911 m, may reflect the character of the outcrop, in a wave-cut platform with few exposed bedding surfaces.
Detailed ichnospecies description and occurrence
Chondrites isp. (Figs. 5F, 5G)
Chondrites is observed only on bedding planes. The trace fossil comprises branching 1–5 mm diameter tubes preserved in convex hyporelief. Branches are only locally observed and many are short (<1 cm), isolated straight segments. As with Planolites, this suggests that the trace fossils possessed a third dimension rendered cryptic by compaction, bedding-perpendicular shortening, and the character of the outcrop surfaces. Chondrites most commonly occurs as isolated decimetre-scale patches on the bedding planes, but larger, metre-scale surfaces are observed that are entirely bioturbated with Chondrites. Chondrites is sporadically observed in the section between 445 and 580 m, and it was documented at 790 m. Chondrites is most commonly observed in association with Planolites.
Poorly preserved Chondrites has been reported in Ediacaran strata (e.g., Jenkins 1995). However, the interpretation of these structures is considered dubious (Seilacher et al. 2005). Chondrites is considered a recurring element of trace-fossil assemblages from the Early Cambrian onwards (Crimes 1992).
Cladichnus fischeri (Fig. 5I)
The ichnofossils are preserved in irregular, convex hyporelief. Cladichnus occurs as meniscate back-filled trace fossils 5 mm in width and more than 1 m in length. Meniscae are slightly lunate. Branches at 20° to 90° are common with meniscae concavity generally directed towards acute intersections. More rarely, branches of C. fischeri crosscut other branches. Some occurrences radiate outwards from a central location. Segments of the back-filled tunnels are straight to curvilinear with rare 60° curves of 30 cm radius. The trace fossil is observed as solitary examples on bedding planes that also contain Chondrites and Planolites. C. fischeri occurs in ichnologically diverse parts of the section with the aforementioned trace fossils and Treptichnus.
Cladichnus fischeri was erected from various branching versions of Muensteria and Taenidium (D’Alessandro and Bromley 1987). The diagnostic criteria includes annulated or monilliform burrow fills that are slightly crescentic. C. fischeri possesses a very thin or no burrow wall. C. fischeri branch successively or radiate outwards. These examples bear similarity to Muensteria of Ksiazkiewicz (1977), which represents part of the taxabase for the erection of Cladichnus.
Gordia marina (Figs. 5B, 5G, 5H)
In the studied section, Gordia marina are 1–4 mm in width. The burrows are smooth to locally bumpy and irregular. The ichnofossil is observed as several spirals that crosscut the trace maker’s previous passages, generating repetitive, cursive, looping spirals (Figs. 5G, 5H), or as larger self-crossing loops intergraded with loose meanders (Fig. 5B) gradational with occurrences of Helminthopsis. In some instances, the burrow is initiated with a loosely meandering structure that abruptly translates into the idiomorphic form reported herein. Patches formed by single occurrences of G. marina are generally 3 to 5 cm across. The ichnofossils are distributed such that patches are at least 20 cm apart. G. marina was observed primarily in the lower half of the section at 100 to 445 m. Gordia marina is observed on bedding planes with Planolites, Helminthopsis, and (or) Oldhamia.
Gordia isp. is discerned from Circulichnus isp. (e.g., Fillion and Pickerill 1990) on the basis of iterative self-crossing observed in the former. The High Head examples differ from the type material figured in Ksiazkiewicz (1977) in that they exhibit increased meandering and crossing. Lower Cambrian Gordia isp. from the UK displays looser meanders (Brasier and Hewitt 1979). Crimes and Anderson (1985) reported G. marina within the lowermost (i.e., Vendian) strata of the Chapel Island Formation (Newfoundland), and Gordia isp. is reported from several other localities in Lower Cambrian and younger strata (e.g., Crimes 1992; Buatois and Mangano 2003).
Helminthopsis isp. (Fig. 5C)
Meandering horizontal, subcylindrical trail (3 mm wide with length up to 40 cm). The width of trace fossil is constant. No tube lining is preserved. Observed material displays loosely developed meanders (Helminthopsis tenuis) through to regularly spaced, non-self-crossing, horseshoe-shaped forms (H. regularis). Helminthopsis is one of the dominant trace-fossil forms in the lower half of the section.
Helminthopsis-like ichnofossils rarely occur in Upper Ediacaran strata. Their taxonomy is difficult because of various modes of preservation (e.g., Ediacaran Heminthoidichnites of South China; Weber et al. 2007). Several studies suggest that Helminthopsis is a common constituent of “deep-water” (i.e., low-energy) ichnofauna from the Lower Cambrian and onwards through to inner shelf settings (e.g., Crimes and Anderson 1985; Tanoli and Pickerill 1989; Hagadorn and Waggoner 2000; Buatois and Mangano 2003; Tangri et al. 2003; Weber et al. 2007).
Lorenzinia apenninica (Fig. 5O)
Stellate burrows that comprise 8 to 16 rays radiating from an unburrowed central area that is 12–20 cm in diameter. The entire trace fossil is 18–32 cm in diameter and has an overall elliptical shape consistent with the tectonic strain. Rays narrow outwards, but in one case the rays are clavate. Rarely occurs in the section; observed in the ichnologically diverse medial part of the section.
Lorenzinia has been interpreted as a radiating graphoglyptid (nearly identical to the morphotype shown in Seilacher 2007), with branching probably occurring at higher intrastratal levels. Younger occurrences of Lorenzinia are well known (e.g., Cenozoic: Monaco 2008; Heard and Pickering 2008; Mesozoic: Yang et al. 1986), and Paleozoic examples are rare (Crimes 1992).
Oldhamia radiata (Figs. 5D, 5E)
Occurrences of Oldhamia radiata are preserved in convex hyporelief on bedding planes. The trace fossils consist of <1 mm diameter unbranching tunnels radiating from a central point. As many as 24 evenly spaced tunnels are observed, forming circular patterns, deformed into ellipses as a result of tectonic strain, which are as much as 2 cm in diameter. In some instances, the radiating pattern is incomplete, and fan-shaped patterns are preserved. The lowermost occurrence of O. radiata is near 90 m. It is increasingly abundant in the medial part of the section (480 to 570 m), with the uppermost occurrence at the top of that interval. O. radiata most commonly occurs on monospecific bedding planes in the most ichnologically diverse parts of the section. Additionally, a poorly preserved example of ?O. recta – ?O. flabellata was observed at 94 m. The poor preservation and rareness of the ichnofossil limit discussion.
Almost identical to Lower Cambrian occurrences reported by Buatois and Mangano (2003) and the Cambrian–Ordovician type locality in Ireland (see Crimes and Crossley 1968). Although Oldhamia is strongly associated with undermat mining (Narbonne and Aitken 1990; Vidal et al. 1994; Gehling 1999; MacNaughton et al. 2000), other than the unusually good preservation of trace fossils along bedding planes, no strong evidence of biomat stabilization is observed in the High Head strata: this is attributed to post-depositional strain, eradicating microbially induced sedimentary structures, such as micro-wrinkles (although rare potential wrinkle structures were observed). O. radiata is associated with Cambrian strata elsewhere (e.g., Hofmann et al. 1994; Buatois and Mangano 2003; Seilacher et al. 2005).
Palaeophycus tubularis (Fig. 5J)
Bedding plane concordant, 2–4 mm diameter, straight to sinuous burrows with similar-diameter branches. Apparently, massive burrow fill; because of the nature of preservation, lining was not observable. Paleophycus was observed sporadically and rarely throughout the section, except for the lowermost 75 m. This burrow diagnosis is only tentative, as the presence or absence of a lining could not be assessed.
Palaeophycus has been putatively reported from Ediacaran strata in Mexico (McMenamin 1996). However, an algal or even inorganic origin for the figured examples in McMenamin (1996) cannot be discounted (Jensen et al. 2006). Narbonne and Aitken (1990) reported similarly sized Upper Ediacaran Paleophycus tubularis, exhibiting more sinuous plan-view morphologies. Cambrian Palaeophycus reported by Buatois and Mangano (2003) are similar in size and form to the High Head occurrences.
Phycodes palmatum (Fig. 5N)
Splayed array of burrows originating from a common burrow origin. Discriminated from Chondrites on the basis of intense burrow crosscuting near the point of branching. Fossil burrows are sand-filled and exposed in convex hyporelief. Burrows mostly 8–10 mm in diameter, with the complete trace being 7–10 cm long. Phycodes palmatum is generally rare, but it is locally abundant on those surfaces between 320 and 550 m in the section.
These burrow systems are similar in size to examples described by Crimes and Anderson (1985) from Lower Cambrian sequences of southeastern Newfoundland and Seilacher (1955) from the Lower Cambrian of the Salt Range, Pakistan. The form studied herein differs in that the burrows branch from the causative burrow at more obtuse angles.
Planolites beverelyensis (Fig. 5A)
Planolites occurs on bedding planes as generally straight or sinuous to locally meandering, unbranched tubes. They are most commonly preserved in convex hyporelief. The trace fossils are displayed in tube lengths of 5–20 cm: the “short” mode of occurrence may suggest that the trace fossils moved in and out of the bedding plane and were not entirely restricted to a single bedding plane. Tube diameters range from 2 to 4 mm with diameter irregular along some examples. Occurrences of Planolites in the High Head section crosscut other Planolites, but rarely loop back and cross themselves. No evidence of meniscate backfill is observed. Planolites is observed throughout the section, except for the lowermost 75 m, which is apparently devoid of trace fossils.
Planolites montanus is distinguished from Planolites beverlyensis by the smaller size and and straight curvilinear appearance of the former (Billings 1862; Pemberton and Frey 1982). The diameter of Pl. beverlyensis also is constant. Planolites occurrences are common in Upper Ediacaran and Lower Cambrian strata (Crimes 1992). Planolites ispp. were taxonomically assessed and discussed in detail in Pemberton and Frey 1982).
Psammichnites isp. (Fig. 5P)
Sinuous to crudely meandering bifid trace fossil, 1–2 cm width. Convex upwards on either side of a depressed medial furrow. Turns can be tight, showing a 2 cm radius to the medial furrow.
Although most commonly associated with later Paleozoic rocks, Psammichnites is also reported from Cambrian rocks (e.g., Seilacher and Gámez-Vintaned 1995; Álvaro and Vizcaïno 1999; Seilacher et al. 2005). Psammichnites is consistent with proposed morphologies for many early bilateria, and the inferred mode of locomotion represented (mucociliary creeping) is consistent with the trace-fossil form (Collins et al. 2000).
Psammichnites gigas (Fig. 5K)
Slightly curvilinear path, 2–3 cm wide, with raised levees associated with an unexposed under track. A regularly sinusoidal trace (wavelength approximately 4 cm; amplitude 2–3 cm) meanders back and forth between the disturbed trace-fossil margins. The trace fossil exceeds 1 m in length. Psammichnites is rare in the High Head section, observed only on three surfaces between 710 and 720 m.
Psammichnites gigas is normally identified on the bases of a lasso-like plan-view pattern resulting from the animal circling repeatedly. Other characters include transversal sculpture on the ventral surface and a sinusoidal line on the dorsal surface (Seilacher-Drexler and Seilacher 1999; Seilacher et al. 2005). The High Head Ps. gigas displays only the sinusoidal dorsal trace, which has been interpreted as the work of a shortened vermiform animal that plowed within the sediment and fed from the surface by moving the siphon laterally to and fro as the animal progressed forward (Seilacher-Drexler and Seilacher 1999; Seilacher 2007). Given the unperturbed motion of the inferred siphon, it is possible that the meandering motion of the siphon served to triangulate geochemical signals while the animal fed at depth. Idiomorphic examples of Ps. gigas have been reported from the lowermost Cambrian in Sweden (Torell 1870), Canada (Hofmann and Patel 1989), Australia (Walter et al. 1989), Greenland (Pickerill and Peel 1990), and the United States (Jensen et al. 2002).
Teichichnus isp. (Fig. 5Q)
Horizontal burrow (two specimens), slightly bow-shaped, with protrusive spreite below the causative burrow. The causative burrow is 4 mm diameter, sand-filled and within sandstone beds.
Teichichnus ranges from the Cambrian to the Holocene and occurs widely in the Lower Cambrian — dominantly in shallow water. In the Tommotian Rovno of the Baltic region, Teichichnus occurs with Treptichnus (Phycodes) pedum and Treptichnus isp. (Fedonkin 1977, 1978). Similar occurrences of Teichichnus are shown in Fritz and Crimes (1983). Teichichnus is an important fabric-forming ichnofossil from the Lower Cambrian and onwards (McIlroy and Logan 1999). Waldron (1988) reported a spectacular occurrence of generally larger, likely deep-water, Teichichnus from probable Middle Cambrian Meguma strata on the eastern coast of Nova Scotia.
Treptichnus pedum (tapho-type 1) (Fig. 5L)
Iteratively branched tunnels that follow a zig-zag horizontal axis up to 1 m long. The trace fossil uses a horizontal Y as its unit structure. The Y-unit is propagated by extending the trace from the left branch and then the right branch in an alternating fashion. Branching angles are affected by tectonic strain; this was corrected using the strain-ratio (∼2.1) derived from Gordia isp. traces, yielding an average angle of 67° between tangents to the burrows at the branch points. Tunnel diameters are approximately 10 mm, and no lining is preserved. The trace fossil is interpreted to represent an intrastratal occurrence: connecting shafts to the sediment–water interface were not observed. Treptichnus pedum occurs as solitary examples on bedding planes associated with Planolites and Cladichnus.
Treptichnus is diagnostically Phanerozoic (Brasier et al. 1994; Narbonne et al. 1987). The iteratively branching planiform-Y morphology is the intrastratal taphomorph observed at other Cambrian localities (Gehling et al. 2001; Droser et al. 2002; Seilacher 2007; Sour-Tovar 2007). The High Head examples are large compared with other locales, but the staggered branching and the angles formed between branches are similar to examples referenced earlier in the text and are typical of the form. Also, post-depositional strain has bent the branches, lending a slight arcuate shape. We were able to render the branches straight by removing the strain from digital images (i.e., spatially transforming the image using the inverse strain ellipse, which was assessed from various observations of strain on the outcrop). The upper part of the trace fossil connects to the sediment–water interface in a uni- or biserial fashion. Biserial examples are referred to as Saerchnites abruptus (Seilacher et al. 2005, fig. 11c).
Treptichnus pedum (tapho-type 2) (Fig. 5M)
Consists of 15–20 mm diameter shafts regularly spaced at 4 to 5 cm. The shafts are arranged in straight to gently curved rows up to 1 m in length. The elliptical shape of the shafts is consistent with tectonic deformation of originally circular structures. Some occurrences show concentric deformation of the sediment around the shafts, with other variants (interpreted to represent deeper intrastratal exposure) showing continuous sediment disruption along the rows of basally connected shafts. The overall structure is interpreted to represent a burrow that was propagated by establishing a J-shaped shaft and aperture at the sediment–water interface and then emplacing an adjacent J-shaft from the horizontal part of the initial J. Tapho-type 2 occurs in the lower section (especially the 75–100 m level) in strata associated with Oldhamia, Planolites, and Helminthopsis.
Successive burrow emplacement is symmetrical about a vertical plane and all vertical and oblique shafts are bisected by that plane. The higher intrastratal preservation precludes certain reconstruction of the subapertural architecture.
The deep-water nature of the High Head member
All previous workers since Phinney (1961) have interpreted the Cambrian Goldenville Group as turbidites formed in a deep-water environment in a rift or continental margin (e.g., Schenk 1970, 1975, 1991, 1997; Waldron and Jensen 1985; Waldron 1992; Waldron et al. 2009). Several sedimentary and stratigraphic aspects of the High Head member also strongly support a deep-water sedimentary environment. Sedimentation by turbidite emplacement is supported by sharply based, tabular, laterally continuous beds, normal grading, complete and partial Bouma (1962) sequences of sedimentary structures, including unidirectional ripple cross-laminae, and by unimodal distributions of current-generated sole markings, including flute and groove casts (Waldron et al. 2009). Water-escape pipes and convolute lamination, indicators of rapid deposition without bedload reworking, are also common. In addition, an important ichnological observation is the presence of Lorenzinia, which is ostensibly a graphoglyptid trace fossil: the mode of preservation for graphoglyptids tends to be erosional exhumation; sediment from a turbidity current casts a mold of the exposed trace. Moreover, the High Head section displays a range of dominantly intrastratal ichnofossils. The paucity of surface tracks may be caused by erosional truncation at the base of turbidity currents. Although turbidity currents are not in themselves restricted to deep-water settings, thick and monotonous units ascribed to turbidite accumulation are more commonly associated with “deep” settings.
Stratigraphically, the High Head member is 845 m thick and within the entire interval, no evidence of shoaling (e.g., wave-generated sedimentary features, coarsening-upwards sandstone bedsets) was observed. The absence of these features preclude sedimentation of tempestites by storm events, diagnostic criteria for which would include bimodal or polymodal paleocurrent distributions, and wave-generated sedimentary structures, such as symmetric ripples and hummocky cross-strata. Although the absence of observable symmetric ripples and wave-generated cross-laminae in the cleaved mudstone might be attributed to the strained nature of the rock, the fine preservation of trace fossils in the same rocks, and of current-generated sedimentary structures in the sandstone beds, which are less highly strained, suggest that the absence of wave-generated structures is a primary characteristic of the succession.
The presence of biomats
Biomat stabilization is inferable by the presence of potential wrinkle structures, abundant Oldhamia, the excellent preservation of trace fossils (i.e., “death-mask” assemblages, Gehling 1999), and the pervasive absence of surface tracks and markings (biomats do not necessarily preclude surface tracks but certainly attenuate their impression). The widespread distribution of biomat stabilization in the Ediacaran and lower Paleozoic strata is an important consideration. On deep-sea bottoms, burrowing animals have been present since the Ediacaran (e.g., Seilacher et al. 2003; Jensen et al. 2005; Liu et al. 2010). However, associated ichnofauna are dominated by burrows attributable to undermat activities and that display impoverished ichnogeneric diversities. Although these lifestyles are important into the Early Cambrian, by the Middle Cambrian, surface grazing by metazoans begins to limit the occurrence of biomats and undermat lifestyles (Bottjer et al. 2000; Seilacher et al. 2005).
Age significance of the ichnological data
The lowermost 75 m of the section, which contains normally graded sandstone and mudstone beds, is unburrowed. The absence of ichnofossils is ascribed to higher sedimentation rates and greater amounts of erosional amalgamation in sand-dominated parts of the turbidite complex. The trace-fossil assemblage from 75 to 445 m is Planolites–Helminthopsis–Gordia–Chondrites dominated, with rare Oldhamia and Treptichnus pedum and very rare Phycodes and T. pedum. Above 445 m, the entire reported assemblage (i.e., Planolites, Helminthopsis, Oldhamia, Chondrites, Gordia, Cladichnus, Psammichnites, Treptichnus, Phycodes, and Lorenzinia) occurs sporadically over 150 m of section. Ichnological occurrences and diversity ultimately wane upwards with increased sand content and limited bedding plane exposure.
The first occurrence of Treptichnus and Oldhamia at 95 m provides the lowermost biostratigraphic constraint — i.e., everything above that level is Cambrian in age. No decisive rationale for inferring the Ediacaran–Cambrian boundary within the studied section can be proposed from the ichnological and sedimentological data reported herein. However, the trace-fossil assemblage is strikingly similar to Lower Cambrian units elsewhere. Ediacaran–Cambrian transitional strata in the Baltic region (Fedonkin 1976, 1977, 1978) also show a rapid increase in the diversity of trace fossils into the Tommotian, with Phycodes (Treptichnus) pedum, Teichichnus isp., and Chondrites isp., appearing in the Lower Cambrian strata. However, these trends have to be viewed cautiously as later workers suggest that the diversity trends are largely ascribable to sedimentation rates (Crimes and Anderson 1985). Buatois and Mangano (2004) reported a lowermost Cambrian assemblage from the Puncoviscana Formation of northwest Argentina. Many elements of the Puncoviscana trace-fossil assemblage are present in the High Head strata, including Helminthopsis ispp., Circulichnus isp. (ethologically similar to Gordia marina), Oldamia radiata (and other Oldhamia ispp.), and Treptichnus ispp., and otherwise it is a somewhat more diverse trace-fossil assemblage than is conveyed in this study. However, the Puncoviscana displays evidence of storm-wave reworking, and parts of it likely were deposited in demonstrably shallower waters than were the High Head strata (Buatois and Mangano 2004).
Crimes and Anderson (1985) reported a diverse ichnofauna from Early Cambrian sequences of southeastern Newfoundland (Chapel Island Formation). Their shallower water Lower Cambrian assemblage — although more diverse than reported herein — has in common Buthotrephis isp. (synonymized with Phycodes palmatum), Planolites isp., Gordia isp., Helminthoida isp., Phycodes isp., Treptichnus isp., and the graphoglyptid Protopaleodictyon. Crimes and Anderson (1985) proposed that the lowermost part of the studied section is Vendian in age: that part of the section contains only Buthotrephis isp., Planolites isp., and Gordia isp.
Based on the similarity of diversity trends and overall ichnological content between the High Head member and high-certainty Lower Cambrian units, it is likely that the lower part of the studied High Head section is Early Cambrian in age. This interpretation is consistent with other data, such as detrital zircon ages and a stratigraphically much higher Middle Cambrian trilobite faunule.
Evolutionary significance of the ichnological data
In terms of evolution, the High Head trace-fossil assemblage has two aspects of particular interest: (1) the section displays a surprisingly high diversity for a Cambrian deep-water setting; and, (2) the High Head section represents an important time in the evolution of seafloor biotopes. These issues are related. Some workers have insightfully suggested that early Cambrian deep-water assemblages are more Ediacaran in aspect than Cambrian, in that overarching ethological strategies were still largely based on the exploitation of microbial mats (e.g., Conway Morris 1989; Seilacher 1999; Buatois and Mangano 2003). Likewise, Seilacher (1999) further suggested that microbial mat-dominated ecologies were restricted to “less favorable” environments (i.e., pushed into deeper water) after the Cambrian explosion. In fact, this idea ties with the idea of deep-water settings as housing relic species (Conway Morris 1989). A moderately diverse, Early Cambrian, deep-water, trace-fossil assemblage does not fit these interpretations perfectly. We interpret the High Head trace-fossil assemblage to primarily indicate one of two scenarios: (1) in the Early Cambrian, animal strategies were becoming robust enough to locally exploit diminished-oxygen, biomat-dominated settings; and (or) (2) though pervasively distributed, biomat development was kept at bay by persistent sedimentation events that left colonization windows open for bioturbating organisms before biomat development minimized access to the sediment. An overprinting consideration may be the down-slope transportation and short-term recolonozation of shelf fauna during and after low-energy turbidite events (cf. “doomed pioneers” of Föllmi and Grimm 1990). In fact, given the putative age of the High Head member, and the nature of the sedimentary environment, it is likely multiple mechanisms are required to explain our observations.
Nevertheless, the reported trace-fossil assemblage represents the leading edge of the agronomic revolution (sensu Seilacher 1999) that would end many millions of years later with motile, grazing animals, greatly diminishing the governing role and wide distribution enjoyed by biomats for the previous 3 billion years (McIlroy and Logan 1999; Seilacher 1999; Buatois and Mangano 2004; Seilacher et al. 2005; Buatois et al. 2009). Before the agronomic revolution, aerobic and anaerobic biomats had penetrated nearly every oceanic niche, and they neatly minimized interactions between the overlaying oceanic water and the underlying marine sediments. The High Head trace fossils reveal an evolutionary strategy of mat exploitation through various means, including employing undermat and substratal farming and feeding strategies, while using the biomat as a stabilizing vector at the sediment–water interface (e.g., Chondrites, Oldhamia, Phycodes, Treptichnus, and Lorenzinia) and developing motile habits that would lead to wholesale grazing on the biomats (e.g., Helminthopsis, Psammichnites, and Cladichnus).
Moreover, the deep-marine deposits of the High Head member display evolutionary trends that largely contribute to the demise of the biomat biome, namely (1) an increase in size and diversity of metazoans in deep-water biomes; (2) a range of motile ethologies used to coexist with microbial stabilization of the substrate; and, (3) increasingly complex mining strategies aimed below the sediment–biomat interface. All of these trends have been observed in earlier studies (e.g., McIlroy and Logan 1999; Seilacher 1999; Buatois and Mangano 2004; Seilacher et al. 2005; Weber et al. 2007) and are consistent with the initiation of the agronomic revolution.
The High Head member contains an important assemblage of trace fossils preserved in a deep-marine environment influenced by turbidite deposition. However, the depositional environment was also stabilized by biomats between turbidite-depositing events. The nature of the sedimentation (i.e., largely episodic) and the characteristics of the sedimentary environment (i.e., influenced by biomats) have led to excellent preservation of the trace-fossil assemblage. Preserved are intrastratal and undermat trace fossils: sediment–water interface trace fossils are not present. The composition of the trace-fossil assemblage is consistent with other Lower Cambrian deposits, and — perhaps more importantly — the distribution of ichnofossils (i.e., increasing in abundance upwards to the first occurrence of Treptichnus pedum) in combination with other data (detrital zircon ages and a stratigraphically higher trilobite faunule) indicate an early Cambrian age.
That the trace fossils occur within putatively biostabilized sediment is consistent with this age. Evolutionary trends that became important later in the Cambrian and that led to the agronomic revolution are evident: animals were becoming larger, motile ethologies were increasingly complex, and mat-mining behaviours were increasingly complex. More generally, the High Head section captures an important transition in the diversification of animal behavior, that by the Ordovician would reshape ocean bottoms to be akin to modern oceans rather than the seafloors of the Cambrian and Precambrian (e.g., McIlroy and Logan 1999; Seilacher 1999; Buatois and Mangano 2004; Seilacher et al. 2005; Weber et al. 2007; Buatois et al. 2009).
MG, JW, and SB acknowledge Natural Sciences and Engineering Research Council Discovery Grants A9222, A8508, and A4230, respectively. We thank Soren Jensen for helpful discussion on Early Cambrian and Ediacaran trace fossils. The reviews of Drs. Luis Buatois and Duncan McIlroy led to substantial improvements in this paper. Alex Kaul, David Dockman, and Jenn Noade assisted in measuring and drafting the section.
- Received January 30, 2010.
- Accepted October 26, 2010.
- Published on the NRC Research Press Web site at http://cjes.nrc.ca on December 10, 2010.