Abstract
Large un-walled backfilled burrows of the Taenidium type are known from Paleozoic deltaic marine environments worldwide where they are often associated with Diplichnites trackways. The latter are generally attributed to arthropleurid myriapods and it may be that the burrows were also made by these animals. Here we describe a Taenidium burrow from the Limestone Coal Formation of the Isle of Arran, a formation that also hosts a well-known example of Diplichnites, supporting the association of the two types of trace fossil and extending their known co-occurrence upward into the Upper Carboniferous.
The Paleozoic succession exposed at Corrie on the Isle of Arran's east coast is visited by hundreds of geology students every year (MacDonald & Herriot 1983; McKerrow & Atkins 1985; Nicholas 2000; MacDonald 2015; McKirdy 2016). It is also expected to attract increasing numbers of interested members of the public following the launch of the Arran Geopark in April 2019 (www.arrangeopark.co.uk). While leading a student excursion we noticed a large Taenidium-type burrow on a bedding plane in the lower part of the Limestone Coal Formation (Clackmannan Group, Bashkirian Stage, lower part of the Upper Carboniferous). Although the burrow has no doubt been observed many times over the years we were unable to find a reference to it in the literature. Here we record its presence for future visitors and provide a detailed sedimentary log and environmental interpretation of the succession in which it is found. The same sedimentary formation on Arran is well known for hosting a spectacular example of a Diplichnites trackway, albeit at a different location (Briggs et al. 1979). We discuss the association of Taenidium and Diplichnites and the fossil record of arthropleurid myriapods, the animals that may have been responsible for both trace fossils.
Geological setting
The Carboniferous succession at Corrie was first described in outline by Sedgwick & Murchison (1829) and in more detail by Gunn (1903), Tyrell (1928), Leitch (1942; although only the upper part) and MacDonald (2015). Overall the succession on Arran is very thin in comparison to the mainland Midland Valley, especially at Corrie where there are no doubt many disconformities (Dean et al. 2011). The succession begins with the red beds and cornstones of the Kinnesswood Formation (Inverclyde Group), followed by volcaniclastic sedimentary rocks and lava flows of the Clyde Plateau Volcanic Formation (Strathclyde Group). Above this, the succession is dominated by siltstones and sandstones with subsidiary limestones. The most notable of the latter is the c. 10 m Corrie Limestone with its spectacular Gigantoproductus assemblages and other shallow-marine fossils which marks the base of the Lower Limestone Formation (Clackmannan Group) and is considered as the local base of the Upper Carboniferous (Dean et al. 2011). The Corrie Limestone is overlain by a clastic succession which is assigned to the Lower Limestone Formation and overlying Limestone Coal and Upper Limestone formations of the Clackmannan Group, although the boundaries between these formations are difficult to place with precision. It is within the part of the succession attributable to the Limestone Coal Formation that the structure described herein is located.
We present a sedimentary log and environmental interpretation of a c. 40 m stratigraphic section exposed on the foreshore, beginning at NS 0249 4354, c. 10 m stratigraphically above the top of the Corrie Limestone (which has been quarried away in Old Corrie Harbour) and ending at NS 0253 4347, c. 35 m below a prominent thick sandstone bed known as Ferry Rock (Figs 1 and 2). The base of our logged section approximates to the base of the Limestone Coal Formation although we note that these clastic beds are difficult to correlate precisely even very locally, presumably due to intraformational unconformities and lateral facies variations.
Sketch map showing the position of the logged section with the Taenidium burrow at Corrie and the previously described Diplichnites trail at Laggan (Briggs et al. 1979). Shaded areas represent large areas of exposed bedrock.
Sedimentary facies log and environmental interpretation of part of the Limestone Coal Formation at Corrie. S, Stigmaria root impressions.
The succession has been interpreted as representing a series of coarsening-upward cycles formed by ‘microdeltas’ prograding into coastal lagoons (McKerrow & Atkins 1985, p. 49). The more distal lagoonal facies consists predominantly of purple and cream-coloured fissile siltstones with very occasional centimetre-scale burrows, grading into more proximal flaggy siltstone and thin (1–10 cm) sandstone interbeds. Thicker (1–5 m), cross-bedded, white, medium to coarse sandstone units with erosive bases and marked lateral thickness variations are interpreted as distributary channel fills. More homogeneous, massive, fine, yellow sandstones with frequent current ripples on their tops and plant impressions (including rootlet horizons and Stigmaria casts) are interpreted as delta-top overbank deposits. These are associated with silt- and sandstone interbeds with abundant plant impressions. There are no organic-rich siltstones or coal horizons at Corrie, although these occur at correlative levels in the Laggan section of north Arran, ∼10 km to the north-west (MacDonald & Herriot 1983; see Fig. 1). The logged succession, in general, coarsens upwards through several cycles, such that delta-top facies predominate in the upper part. This trend continues above the logged section, culminating in the thick, pure, sandstone of Ferry Rock. Just one bed in the logged succession does not fit the general facies pattern, namely Bed 10 at c. 15–16 m, which is a homogeneous, brick-red sandstone with a coarse lag at its base. We interpret this tentatively as aeolian sand which interrupts the generally deltaic succession, perhaps due to upstream delta-switching. The beds immediately above it are not exposed even at low tide and are probably soft distal mudstones and siltstones that begin the next coarsening-upward cycle.
Description
The structure described here occurs at NS 0250 4349, at the high tide level on the weathered top of a decimetre-scale bed of medium sandstone (Bed 12b), directly above a c. 1.5 m thick bed of fine to coarse sand with prominent trough cross-bedding (Bed 12a; see Fig. 2). This bed is the first in the logged succession to show abundant plant impressions, including a poorly preserved metre-long Stigmaria trace on the same surface.
The structure is illustrated with photographs (Fig. 3) and an interpretative field drawing (Fig. 4). It extends for 1.5 m across the top of Bed 12b, eroded through at the top end (where an oval cross-section is revealed) and running beneath another bed at the bottom end. The width is generally 10–15 cm, although at its eroded top end it appears to widen to c. 20 cm but the border of the structure is not as clear, possibly because of the superposition of two traces. Towards the bottom end the structure cuts across what appears to be another more obscure structure, which although difficult to make out may be of similar type but curving and 5 cm wide.
Field photographs: (a) structure from above (scale = 1 m); (b) close-up of burrow indicating sharp margin and lack of distinct wall; (c) detail of curving trace at bottom end; (d) cross-section view from top end.
Field sketch of the structure. Inset: cross-section view from the top end (asterisk).
The fill is poorly sorted, sub-rounded, fine and medium quartz sand with very occasional flakes of white mica. There is no marked difference between the fill and the sand that makes up the bed in general, suggesting that the fill material is derived from the same bed, although the margin of the structure is generally well defined. The fill tends to be less well cemented than the rest of the bed and appears to be eroding slightly faster such that the margins stand proud, in places up to 15 mm high. This may be because the sand is more tightly packed within the structure than outside. There is no obvious ‘wall’ to the structure in the sense of Keighley & Pickerill (1994), i.e. no discrete physical lining of compacted sediment or other material. In places the margin is marked by a centimetre-wide external ‘diagenetic halo’ (Keighley & Pickerill 1994), possibly due to subtle cementation or redox effects caused by fluid flow along the margin. Curving meniscus-like laminations are visible along the length of the structure and are picked out in several places as better-cemented ridges. In the eroded cross-section the curved laminations appear to be approximately concentric.
Interpretation
We interpret the structure as a burrow, on account of the well-defined margin, meniscus fill laminations and oval cross-section. We assign it to Taenidium barretti (Bradshaw) because there is no evidence of a maintained wall structure and because of similarities with the type description and other examples of which we are aware from the Lower Carboniferous (Pearson 1992) and Devonian (Morrissey & Braddy 2004; Morrissey et al. 2012) of Britain. Its full length is not known, nor is the horizon from which it originally penetrated. Above it lies 1.2 m of interbedded thin siltstones and fine sandstones (Bed 13) from which a sand-filled cast of a Stigmaria root currently protrudes. Above that (Bed 14) is a 1.5 m thick fine sandstone bed with four prominent rootlet horizons (which are best seen at low tide). From this we infer that the animal is likely to have been living in a vegetated delta-top environment through which it burrowed until encountering soft homogenous sand and progressed horizontally. The burrow was backfilled with sand by the hind legs as the animal moved forward.
Taxonomic discussion
Taenidium barretti was initially described as a species of the ichnogenus Beaconites. The genera Taenidium and Beaconites were confounded in much of the older literature until clarified by Keighley & Pickerill (1994). According to those authors, both ichnogenera are unbranched meniscate trace fossils but Taenidium comprises backfilled structures with a simple margin, whereas Beaconites burrows possess a distinct wall. We apply that distinction in this study.
The ichnogenus and -species Beaconites antarcticus was erected by Vialov (1962) for mostly horizontal tubular backfilled burrows that are common in the Devonian Beacon Supergroup of South Victoria Land, Antarctica. Bradshaw (1981) noted that there are two distinct types of ‘Beaconites’ burrows in the type area: relatively small, straight and vertical burrows usually between 1 and 3 cm wide, filled with straight, discrete packets of sediment (including the type specimen of Vialov's B. antarcticus); and larger (up to 15 cm in width and up to 2 m long), more curving and generally horizontal burrows with meniscate backfills. She called the latter ‘Beaconites barretti’ and noted the strong similarity of these specimens to burrows that had been described (mostly under Vialov's original ichnospecies name) from the Lower Devonian (Allen & Williams 1981) and Lower Permian (Ridgway 1974) of the UK and the Carboniferous (Graham & Pollard 1982) of Ireland. She noted that the largest of these burrows was 25 cm wide, implying that ‘B. barretti’ could be larger than the maximum 15 cm in the type locality. These ‘B. barretti’ burrows are now recognized as un-walled Taenidium barretti (see synonymy list in Keighley & Pickerill 1994, pp. 327–328).
In their taxonomic revision, Keighley & Pickerill (1994, p. 328) noted that ‘many specimens from the Upper Paleozoic are of giant size … although recorded widths … may be as small as 5 mm'. The post-Paleozoic examples recorded in their synonymy list (Bown 1982; Bown & Kraus 1983; Bracken & Picard 1984) and subsequent reports from post-Paleozoic strata (e.g. Rossetti & Netto 2006; Rodriguez-Tovar et al. 2016; Santos Mineiro & Miloni Santucci 2018; Zhou et al. 2019) are apparently all narrow structures (less than 3 cm and usually c. 1 cm wide or less), suggesting that a minimum width requirement could form part of a revised diagnosis in accordance with Bradshaw's (1981) original description.
The trace maker
Various suggestions have been made as to the animals likely responsible for meniscate backfilled burrows, including molluscs and polychaete worms (Vialov 1962; Gevers et al. 1971) and vertebrates (amphibians or reptiles) (Ridgway 1974). Narrow (up to 3 cm wide) Beaconites and Taenidium burrows occur throughout the Phanerozoic, both in marine and freshwater facies, and were presumably made by a variety of organisms, whereas large (>3 cm, up to c. 45 cm wide) Taenidium barretti burrows are, as far as we are aware, known only from Silurian to Carboniferous fluvial and deltaic facies. Rolfe (1980) suggested that these large burrows were most likely the construction of an arthropod on account of the regular structure which is consistent with a relatively stereotyped behaviour pattern. Bradshaw (1981, p. 633) was apparently the first to suggest that they could have been made by large terrestrial myriapods. One of us (Pearson 1992) described a large ‘Beaconites’ (=Taenidium) burrow (5.25 m long and 45 cm wide) from the Strathclyde Group (Lower Carboniferous) of Fife, Scotland, and suggested that it may have been constructed by Arthropleura. Morrissey & Braddy (2004, p. 325) proposed that abundant Lower Devonian ‘Beaconites’ (=Taenidium) barretti burrows from Pembrokeshire, Wales, were the work of ‘a myriapod (eoarthropleurid) burrowing to the water table in search of moisture to aestivate and/or moult during the onset of the dry season'. The example described here may have had a similar origin although the equatorial coastal delta environment of the Carboniferous probably did not require dry-season aestivation and the burrows may simply have been for habitation and predator avoidance.
Pearson (1992) noted an association between large ‘Beaconites’ (=Taenidium) burrows and Diplichnites trails (morphologically simple trails consisting of two parallel trackways with footfall marks that are approximately normal to the trail and no median tail-scrape; Briggs et al. 1979, p. 289). These often occur in the same sedimentary formations, including the examples from Fife described in that paper. The co-occurrence of the ichnogenera does not demonstrate that they were made by the same type of animal, but it is suggestive because the options are fairly limited for large terrestrial animals of the appropriate geological periods. Morrissey & Braddy (2004) recorded eight associations of Diplichnites and Beaconites ranging from the Upper Silurian to Lower Carboniferous and noted similarities in the reported widths of the trackways and burrows. To these examples can be added more recently described co-occurrences from the Silurian–Devonian of Northern Territories, Australia (Gouramanis & McLoughlin 2016) and the Lower Carboniferous of New Brunswick, Canada (Falcon-Lang & Miller 2007). The burrow described here is another instance of this association because a spectacular Diplichnites trackway has been described from approximately the same stratigraphic level (and similar sedimentary environment) in the Limestone Coal Formation at Laggan on the NE shore section of Arran (Briggs et al. 1979; location shown in Fig. 1).
The combined stratigraphic range of these occurrences corresponds well with the known body fossils of large arthropleurid myriapods assigned to the genera Eoarthropleura and Arthropleura (Shear & Edgecombe 2010), although other large terrestrial arthropods cannot be ruled out (Fayers et al. 2010). The Arran examples extend the known association of Taenidium barretti and Diplichnites into the Upper Carboniferous where most of the rare body fossils of Arthropleura are known. Arthropleura body fossils are of various sizes, the largest reasonably complete specimen being 90 cm in length, but fragments and isolated segments suggest a maximum adult length of c. 2 m and width of c. 42 cm (Hahn et al. 1986, who reconstruct a length to width ratio of 4.78). A selection of Upper Carboniferous body fossils of Arthropleura armata from the Coal Measures of Somerset, England, indicate lengths of 46–66 cm and widths of 10–14 cm (Proctor 1998), consistent with the Taenidium burrow described here. The Arran trackway is considerably wider (36 cm; Briggs et al. 1979) which could, of course, represent a more mature individual or a different species. There are likely to be many currently undescribed large Taenidium burrows exposed. If they prove to be the work of arthropleurids then further documentation may help define the habitat, geographical distribution and stratigraphic range of these animals and their role in Paleozoic terrestrial ecosystems.
Acknowledgements
We thank Chris Boyd and Nicholas Minter for very helpful reviews.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Author contributions
PNP: Conceptualization (Lead), Investigation (Lead), Writing – Original Draft (Lead); RJG: Investigation (Supporting), Visualization (Lead), Writing – Review & Editing (Supporting)
Scientific editing by Tom Challands
- © 2019 The Author(s). Published by The Geological Society of London for EGS and GSG. All rights reserved
References
A large Taenidium burrow from the Upper Carboniferous of Corrie, Isle of Arran, and remarks on the association of Taenidium burrows and Diplichnites trails
