Palaeoenvironments and Palaeoecology of the Early Pennsylvanian Lancaster Formation (‘Fern Ledges’) of Saint John, New Brunswick, Canada

By Falcon-Lang, Howard J Miller, Randall F

Abstract: The Pennsylvanian Lancaster Formation (‘Fern Ledges’) of Saint John, New Brunswick, Canada contains a world-famous fossil biota. Largely unstudied since the works of Hartt, Dawson, Matthew, and Stopes in the mid-nineteenth to early twentieth century, we present new data concerning biostratigraphy, taxonomy, palaeoenvironments, and palaeoecology. Megafloral assemblages suggest a mid- to late Langsettian age for the succession, making it approximately coeval with the classic Joggins Formation of nearby Nova Scotia. Facies analysis suggests deposition on a tectonically influenced coastal plain whose braided channels drained into a shallow brackish gulf. Most of the historical fossil collections are derived from flooding surfaces formed by abrupt subsidence events along the coastline. Three communities are recognized. Phoronids, crustaceans, and xiphosurans lived in brackish coastal waters. A lowland community of gastropods, insects, arachnids, and myriapods inhabited coastal forests. Coastal vegetation was dominated by shrubby cordaitaleans and pteridosperms whereas ferns, sphenopsids, and lycopsids were rare. An upland or dryland community, discernible from allochthonous assemblages, comprised forests of giant cordaitaleans, archaic pteridosperms, and plants of uncertain affinity.

The Pennsylvanian Lancaster Formation of New Brunswick (which includes the famous ‘Fern Ledges’ locality) has an important place in the annals of palaeontology (Falcon-Lang & Miller 2007). Initial discovery of fossil plants (Gesner 1840) was followed, in 1860- 1863, by the documentation of a rich biota by C. F. Hartt and G. F. Matthew (Dawson 1861, 1862, 1868), and other members of the Natural History Society of New Brunswick. Of special interest was the occurrence of insects, one specimen preserving a putative stridulating organ (Scudder 1868a, b). This fact allowed Darwin (1871) and others to imagine primeval forests alive with chirping (Miller & Buhay 1988).

The age of the Lancaster Formation was, at first, controversial, being assigned to the Devonian Period (Dawson 1861, 1862, 1871), and subsequently to the Silurian Period (Matthew 1910a, b). However, repeated protests that plant remains were clearly of Pennsylvanian age (Hagen 1881; Ami 1900a, b; White 1902, 1911) eventually led the Geological Survey of Canada to approach Marie Stopes to resolve the debate. Her classic monograph (Stopes 1914) confirmed the Pennsylvanian age of the succession once and for all (Falcon-Lang & Miller 2007).

Subsequently, the fossil biota received almost no scientific study, although the systematic position of a few taxa has been revised (e.g. Bell 1944; Copeland 1957; Hantzschel 1975; Robbins 1994; Miller 1995a, b; Wagner 2001, 2005a, b). The aim of this paper is to improve knowledge of the geological setting, taxonomy, biostratigraphy, palaeoenvironment and palaeoecology of the site.

Stratigraphie nomenclature

The Pennsylvanian rocks of Saint John were first mapped as Little River Group, and subdivided into the Dadoxylon Sandstone and Cordaites Shale members (Dawson 1868). Later attempts to map these members regionally (Ells 1907) proved unsuccessful and their usage was abandoned. The Little River Group was renamed the Lancaster Formation by Ami (1900a, b), and Alcock (1938) extended the concept to include beds near Saint Martins and Musquash Harbour, southern New Brunswick. However, the Saint Martins sections were subsequently placed in the Tynemouth Creek Formation (Hayes & Howell 1937; Plint & van der Poll 1982) whereas the Musquash Harbour sections were recognized as belonging to a younger, as yet, unnamed unit (van der Poll 1995). In this paper, the term Lancaster Formation is restricted to Pennsylvanian rocks near Saint John. This c. 400 m thick unit conformably overlies the Mississippian(?) Balls Lake Formation in its central area (Currie & Nance 1983), and unconformably oversteps Neoproterozoic rocks towards the SE and NW (McLeod et al. 1994a, b).

Geological setting

Understanding tectonic context is essential for interpreting the palaeoenvironment of the Lancaster Formation. Basement rocks in this part of eastern Canada include the Avalon and Meguma terranes, which accreted during the mid-Devonian Acadian Orogeny (Gibling 1995). The terrane boundary is defined by the Fundy-Cobequid Fault, which was reactivated during Late Devonian times as a result of the far-field effects of Gondwanan convergence. The palaeotropical Maritimes Basin developed along this fault axis in response to dextral strike-slip, and comprised several sub-basins (Calder 1998).

The Lancaster Formation accumulated within the small (55 km by 220 km) Cumberland sub-basin, which straddles northern Nova Scotia and SE New Brunswick, bounded along its SE margin by the Fundy- Cobequid Fault (Fig. la-c). Alluvial fan deposits dominate basin- fill adjacent to this fault-line. Closer to the basin centre, alluvial plain, coastal plain, and brackish bay facies are common (Davies et al. 2005). These latter sediments mostly formed in a paralic setting with fluvial systems draining northeastwards (Gibling et al. 1992) into a mid-European sea (Calder 1998), the inferred source of the brackish transgressions (Falcon-Lang 2005).

The Lancaster Formation was deposited near the western end of the Cumberland sub-basin, where a restraining bend of the Fundy- Cobequid Fault inhibited dextral offset (Yeo & Gao 1986). In this compressional setting (Fig. Ic), the formation became locally overturned within a fold and thrust belt (Rast et al. 1984). Chlorite spots (Nance 1987) and black palynomorphs (spore alternation index of 10; Dolby 1989) imply upper greenschist-facies metamorphism (van der Poll 1995) consistent with fossil plant preservation as highly altered graphitic streaks lacking cuticles (Stopes 1914).

Study material

More than 8000 fossil specimens were collected from the Lancaster Formation by the Natural History Society of New Brunswick (Miller & Buhay 1988). These are almost entirely plants (>99%), with fewer than 70 specimens of fauna known. The historical collections derived from coastal outcrops at Duck Cove [UTM 19T 728325 5014150], Lancaster Shore, also known as ‘Fern Ledges’ [UTM 19T 728963 5014433], Barrack Shore [UTM 19T 731 157 5016428], and Little River [UTM 19T 733202 5017677; Map Sheets NTS 21 G/01E, NTS 21 H/05]. We studied sedimentary facies and fossil taphonomy at each of these sites, in addition to outcrops at Courtney Bay and Cape Spencer (Fig. Id).

During his work on the most prolific fossil site at Lancaster Shore, Hartt (in Dawson 1868) produced a bed-by-bed record of fossil assemblages, recognizing eight principal fossil beds (LS 1-8). As well as recording alpha diversify, Hartt noted the relative abundance of each taxon using a semi-quantitative scheme (simplified in Table 1 as rare, occasional, common, and abundant). We identified five of Hartt’s fossil beds (LSI, 2, 3, 4? and 8) at Lancaster Shore. As a consequence, most fossil collections (reviewed below) can be assigned to our facies scheme.

Megafioral remains

Stopes (1914) was the last worker to fully revise fossil plant systematics, although a few taxa have subsequently been restudied (Bell 1944; Wagner 2001, 2005a, b). A primary revision of the plant assemblage is beyond the scope of this paper (this work is being undertaken by R. H. Wagner). Here we merely collate an updated taxonomic list, and give relative abundance data for each taxon at Lancaster Shore (Table 1). Taxa were fully illustrated by Stopes (1914).

Dominance -diversity characteristics

Cordaitaleans overwhelmingly dominate the assemblage, and Cordaites leaves are especially abundant. The true diversity of Cordaites is difficult to assess because of their conservative morphology, but at least two species are present: C. principalis and C. borassifolius. A third type of cordaitalean leaf, Poacordaites, is rare. Other cordaitalean remains include Dadoxylon ouangondianum tree-trunks (Dawson 1868; Matthew 1910a) with Artisia-type piths, Cardiacarpus (four species) ovules and Cordaianthus fertile axes.

Of the remaining gymnosperms, medullosan pteridosperms are also abundant and include abundant Alethopteris (two species) and Laveineopteris loshii, together with rare Megalopteris dawsonii, Neuralethopteris schlehanii, Paripteris gigantea, and various fertile organs. Archaic pteridosperms include Rhacopteris. The coniferopsid, Dicranophyllum is occasionally present.

Of the pteridophytes, sphenopsids are the most common plants, represented by Calamites axes (two species) and foliage of Annularia (two species) and Asterophyllites (two species) types. Sphenophyllum is very rare. Ferns are also rare although they comprise a diversity of arborescent and herbaceous forms including Lobatopteris, Oligocarpia, Renaultia, Senftenbergia, and Sphenopteris. Remains of arborescent lycopsids and putative progymnosperms (Pseudadiantites) are extremely rare.

Biostratigraphic implications

Comparison of our updated megafiorai list (Table 1 ) with range charts for European taxa allows for biostratigraphic inferences. In key reference sections in central England (Cleal 2006) and Wales (Cleal 2007), several taxa in our assemblage either have a first appearance datum in the early or late part of the Langsettian substage of the Westphalian or a last appearance datum in the late Langsettian (Fig. 2); however, a few of these taxa originate earlier in central Europe. Ranges suggest a mid- to late Langsettian age (Laveineopteris loshii subzone of Cleal 1991), although taxonomic revision (R. H. Wagner, pers. comm.) may alter these findings. The Lancaster Formation is therefore coeval, in part, with the mid- Langsettian Joggins Formation at the eastern end of the Cumberland sub-basin in Nova Scotia (Falcon-Lang 2006a; Falcon-Lang et al. 2006). Fauna remains

In contrast to the abundant plants, fewer than 70 specimens of fauna are known. Many specimens, once thought lost, have now been rediscovered, and we provide a list of museum numbers (Table 2). Fossils are generally poorly preserved, and many considered dubious taxa (Hantzschel 1975). As the fauna has never been properly monographed, we provide a photographic record of the best specimens (Fig. 3).

Aquatic fauna

The aquatic fauna includes phoronids, crustaceans and xiphosurans (Table 2). One of the most common taxa is Sptrorbis erianus (Matthew 1895a), a probable synonym of Spirorbis carbonarius (Fig. 3a), encrusted on Cordaites leaves. Recent work has shown they are ‘spirorbiform microconchids’ related to phoronid worms (Taylor & Vinn 2006).

Crustaceans are represented by Amphipeltis paradoxus (Salter 1863) (Fig. 3b), described from a single specimen with a second found later (Matthew 18956), and Leaia silurica (Matthew 19106) (Fig. 3c), a branchiopod (Copeland 1957). A third taxon, Asmussia (Copeland 1957), has been excluded as locality data confirm that specimens are of Mississippian age.

Xiphosurans were recognized based on Belinuropsis wigudensis (Matthew 19106). Re-examination of type material shows that this genus was erected from two specimens that cannot be reunited as illustrated by Matthew (Miller 19956). One is indeterminate and the other is arthropod cuticle (Fig. 3d; S. J. Braddy, pers. comm.) of possible, but uncertain, affinity to eurypterids (W. Shear, pers. comm.). Despite the dubious nature of Belinuropsis, trackways of Kouphichnium showing ‘pusher prints (Fig. 3e) provide independent support for the existence of xiphosurans in the Lancaster Formation and are known from equivalent facies elsewhere in the Cumberland sub- basin (Archer et al. 1995; Falcon-Lang 2005).

Although eurypterids were reported we consider these occurrences as questionable. Their record is based, in part, on Eurypterella ornata (Matthew 1889). Hantzschel (1975) thought this was a dubious taxon, whereas Tetlie (2004) considered it to be a trigonotarbid. The eurypterid status of a second Eurypterus pulicaris (Salter 1863), has also been questioned (Matthew 1896, 1897), and it has been considered a possible arthropleurid limb (Briggs et al. 1979). In our view, there are no unequivocal eurypterids in the assemblage.

Matthew (19106) described Bipezia bilobata as a vertebrate trace. Glaessner (1957) considered it to be an invertebrate resting trace synonymous with Isopodichnus (= Oniscoidichnus). Examination of the type specimens casts doubt that either is a fossil, although NBMG 3038 (Matthew 19106, plate IV, fig. 2) might be an arthropod resting trace.

Terrestrial fauna

The terrestrial fauna comprises gastropods, insects, arachnids, and myriapods (Table 2). Two taxa of terrestrial gastropods, initially described as Strophites grandaevus (Dawson 1880) and Pupa primaeva (Matthew 1895a) (Fig. 3f), were later transferred to Dendropupa; they are suspected synonyms of D. vestusta (Solem & Yochelson 1 979).

Matthew (1895c) recognized myriapod specimens including the millipede, Euphoberia atava (Fig. 3g) and the centipede, Chilopus dubius. The latter is poorly preserved but exhibits 11-12 segments with vague indications of legs. Some tergites appear to be posteriorly notched, possible evidence of scutigeromorph centipede affinities (W. Shear, pers. coram.). Other putative myriapods are poorly preserved; the placement of Eileticus antiquus and Palaeocampa is uncertain, although the latter has not been re- examined (Hoffman 1969) (W. Shear, pers. comm.). Eurypterus pulicaris (Salter 1863), as noted above, may represent an arthropleurid limb (Briggs et al. 1979), and Beaconites isp., discovered by us, may be an arthropleurid aestivation burrow (Morrissey & Braddy 2004).

Trackways earlier named as Acripes incertipes, A. leavltti and A. minor (Matthew 19106) or Diplichnites (Hantzschel 1975; Robbins 1994) are here referred to as probable Diplichnites gouldi Type A (Fig. 3h and i; Smith et al. 2003). Some specimens demonstrate en echelon sets (

Putative arachnids include Eurymartus spinulosus (Matthew 1 8950), which comprises a poorly preserved abdomen, and E. latus (Matthew 18956), a more or less complete abdomen, too poorly preserved to name (W. Shear, pers. comm.). Matthew placed Eurymartus in the Order Eurymarti, not recognized by later workers. As noted above, Eurypterella ornata, originally described as a eurypterid, may be a trigonotarbid arachnid (Tetlie 2004).

Scudder ( 1 868a, b, 1 880, 1 885) described insect remains, with later additions by Matthew (1889, 1895a, b, 191Oe). Carpenter (1992) described each as a ‘little-known genus’ and included a taxonomic placement made by Handlirsch (1906) and others. Geroneura (Fig. 3j) was placed in the Neoptera, Order uncertain. Aedaeophasma, Gerephemera, Lithentomum, Pseudohomothetus (Fig. 3k), and Xenoneura were placed in the Pterygota, Order uncertain. Platephemera was placed with reservation in the Order Palaeodictyoptera, Family uncertain. Archaeoscolex corneus was first considered to be a probable myriapod (Handlirsch 1906), but later reinstated as an insect (Hantzschel 1975). Taxonomic positions in Table 2 are based, in part, on an unpublished reexamination of type material by J. Kukalova-Peck (pers. comm.) and Rowland (1997).

The putative occurrence of tetrapode is inferred from obscure footprints. Matthew (19106) described (l)Hylopus variabilus and (?)Nanopus vetustus (Fig. 31), but both were subsequently considered as doubtful ‘footprints’ (Abel 1935; Hantzschel 1975). Reexamination of specimens suggests that they may represent poorly preserved undertracks emplaced in soft sediment, but that assignment to Hylopus or Nanopus is questionable.

Sedimentology and fossil assemblages

We studied the sedimentology and facies distribution of fossils at all coastal exposures of Lancaster Formation (Fig. Id). As a result of the meticulous work of Hartt (in Dawson 1 868) most of the historical fossil collections, together with new collections, could be placed into a facies context. Four fossiliferous facies were recognized, organized into two associations (Fig. 4).

Facies 1

This facies, best seen at Lancaster Shore, comprises multi- storey channel sandstone bodies. The full thickness of bodies is never exposed, but partial thicknesses are / = 1 8 mm in diameter (quartz pebbles predominate). The remaining channel-fill comprises medium-grained (locally coarse- to very coarse-grained), trough cross-bedded sandstone. Troughs are 0.2-0.5 m (locally 0.7 m) high and indicate ENE palaeoflow (mean 068[degrees] +- 33.8[degrees]; n = 35). Rarely,

Plant remains are present within channel-base conglomerate, or more rarely in trough cross-bedded sandstone. Fossils include calcite- or silica-permineralized tree-trunks, up to 35 cm in diameter and >2 m long. Trunks preserve Dadoxylon wood anatomy, and in a few specimens show an Artista pith and lateral branching. Charred Dadoxylon fragments, compressed Cordaites leaves, and rare Calamites and cf. Sigillarla are also present. Other fossil assemblages occur in mudstone lenses within channel-fills and include compressions of Rhacopteris, Cordaites, Cordaicarpus, Alethopteris, Calamites, and Pinnularia, amongst other indeterminate material (Matthew 1906). Diplichnites myriapod trackways occur in fine-grained sandstone showing current lineation at Barrack Shore, and probably derived from Facies 1 .

Facies 2

This facies comprises 1.45-7.45 m thick erosive-based channel bodies, 10-28 m wide, and with a basal relief of <<1 m. Basal channel-fill is composed of medium- to coarse-grained (rarely pebbly) sandstone. This grades up into fine-grained sandstone with 0.2-0.4 m high trough cross-beds, or ripple cross-lamination, which show divergent palaeoflow vectors (NW to SE; ? = 5), where measurable. Uppermost channel-fills include

Facies 3

This facies comprises 0.12-2.97 m thick sheets of sediment, which may be lenticular over tens of metres, and show upward coarsening (Fig. 5a). Lowermost mudstone beds are 0.02-0.71 m thick, dark grey (locally red), finely laminated, sharp-based, and may show rhythmic siltstone laminae and lamina-siderite. They grade upwards into fine- to medium-grained sandstone beds, 0.10-2.26 m (typically 0.3-0.7 m) thick, which are horizontally bedded and show symmetrical ripple marks (Fig. 5b). The upper part of some sandstone beds may show low- angle, mutually erosive cross-beds, up to 0.10-0.15 m high, or localized channels,

Facies 4

This facies comprises palaeosols developed within Facies 1-3. Palaeosols show irregularly pitted surfaces with

Fossils are common in the palaeosols, although generally poorly preserved. Rare tree stumps, typically 21-98 mm in diameter (two are c. 0.4 m), occur in growth position as obscure sandstone or mudstone casts, or coalified stumps. Where present, localized tree densities may be as high as 16 trees per square metre on a few surfaces, but distribution is extremely patchy. Nineteen stumps are recognizable as Calamites. Where present, these form monotypic stands, and are associated with Asterophyllites and Paracalamostachys compressions (Fig. 5d). Two additional tree stumps with Stigmaria rhizophores are lycopsids. However, identifiable sphenopsid and lycopsid stumps together represent a very small proportion (<2%) of the total number of autochthonous plants observed. Much more abundant and widespread are small stumps, which show a vertical tap root, and a radiating pattern of subhorizontal roots that is traceable for only a few tens of centimetres. Primary lateral roots are

Palaeoenvironmental context

The widespread occurrence of sandstone units containing symmetrical ripple marks (Facies 1-3) suggests that the Lancaster Formation was mostly deposited in, or directly adjacent to, a standing body of water subject to wave activity. The occurrence of xiphosurans, represented by Kouphichnium trackways, and the annelid trace maker of Planolites montanus suggest a brackish salinity (Calver 1968; Anderson & Shuster 2003; Buatois et al. 2005). Pennsylvanian brackish deposits occur widely across eastern Canada (Calder 1998) and represent the sediments of a large epicontinental, microtidal embayment that opened eastwards into the mid-European sea and was perhaps analogous to the presentday Baltic Sea (Falcon-Lang 2005; Wells et al. 2005).

Facies interpretation

Facies 1 is interpreted as the deposits of large, eastward- flowing braided fluvial channels. Sandstone body dimensions indicate that channels were hundreds of metres across and many metres deep. Channels filled by the accretion of downstream-migrating subaqueous dunes (trough cross-beds) with minor deposits from laterally accreting bars (large-scale tabular cross-stratification) as seen in modern sandy braided systems (Cant & Walker 1978). Fine-grained lenses may represent temporary abandonment facies. Metre-thick successions of symmetrically rippled sandstone in some uppermost channel-fills suggest localized wave action. This observation suggests that channels progradcd out into the brackish water body described above as deltaic distributaries (Elliot 1986), although direct evidence for either a distributary form or large-scale progradation is lacking.

Facies 2 is interpreted as the deposits of smaller fluvial channels, at most just a few tens of metres wide. The dominance of symmetrically rippled sandstone within the channel-fill suggests strong wave action throughout deposition, and implies that channels fed small delta lobes. Channels are exclusively found within thick successions of symmetrically rippled sandstone sheets and lenses of Facies 3. This latter facies shows upward coarsening and may either represent wave-reworked mouth bar deposits associated with the small channels, or, more likely given their widespread occurrence and extent, the product of linear clastic shorelines (Elliot 1986). Rhythmic bedding in some mudstone beds may record weak tidal influence, and small-scale, mutually erosive, lowangle stratification formed in the shoaling wave zone. Red laminated mudstone at some intervals in Facies 3 may record denudation of oxidized soils upstream, and subsequent redeposition in brackish bays; there is no evidence for in situ pedogenesis.

Facies 4 is interpreted as the product of soil formation. Soils developed on emergent substrates, including abandoned channels of various sizes (Facies 1-2) and on mouth bar and shoreface sands after they built up above base level (Facies 3). Shallow penetration of root systems (=le;81 mm deep) and local preservation of a thin O horizon indicate permanently waterlogged soil conditions. The lack of prominent horizonation, the generally weak alteration of parent material, and the existence of dull plasma fabrics indicate that these were immature soils (inceptisols). Soil immaturity was probably related to repeated flooding by brackish coastal waters.

Fades associations

Two facies associations are recognized at Lancaster Shore where strata are least deformed. Facies Association 1 comprises the lower c. 90m of the succession (mostly covered by glacial till) and is dominated by the deposits of braided fluvial channels (Facies 1; 95% exposed stratal thickness) with minor mouth bar deposits (Facies 3; 5%). This association corresponds to the Dadoxylon Sandstone of Dawson (1868). Facies Association 2 comprises the upper c. 45 m of the succession exposed and corresponds to Dawson’s Cordaites Shale. It mainly consists of mouth bar and shoreface deposits (Facies 3; 85%), locally cut by small channels (Facies 2; 15%).

Although sediments record interaction between fluvial channels (of various scales) and a standing water body, the absence of a large-scale coarsening-upward succession argues against an overall deltaic model. Facies Association I was formed in the wave- influenced mouth of a braided river system. Facies Association 2 was formed by small delta lobes, which coalesced to form a coastal plain, whose shoreline was wave-dominated. The sharp transition from the fluvial-dominated Facies Association 1 and the wave-dominated Facies Association 2 implies abrupt and major base-level rise. In this syntectonic setting (Nance & Warner 1986), large-scale accommodation was probably created by tectonism, although custastic effects cannot be entirely excluded. During the deposition of Facies Association 2, accommodation and sedimentation generally remained closely balanced, the entire succession accumulating at, or near, base level.

Palaeoecological synthesis

Analysis of fossils in their palaeoenvironmental context, together with historical dominance-diversity data (Hartt, in Dawson 1868), allows insights into the palaeoecology of the Lancaster Formation. Three communities are recognized (Fig. 6).

Brackish bay communities

Aquatic elements of the fauna are presumed to be autochthonous within the shallow brackish waters that flooded the coastal plain following minor subsidence events (Facies 3). This community was dominated by arthropods as reported for similar facies in Western Europe (Calver 1968). The Planolites trace-maker was an infaunal deposit-feeding annelid (Buatois et al. 2005). Xiphosurans probably fed on benthic invertebrates, and may have made brief amphibious excursions up the sandy beaches to reproduce (Andcrson & Sinister 2003). Spirorbiform microconchs (Taylor & Vinn 2006) were filter feeders that encrusted floating plant debris. Conchostracans, which occur in beds isolated from the other taxa, may have inhabited near- freshwater shallow pools along the coastline; they were non- selective filter feeders living off algal and bacterial detritus (Webb 1979).

Coastal forest communities

(Par)autochthonous plant assemblages preserved in inceptisols (Facies 4) developed on top of channel-fill (Facies 2), mouth bars and shoreline deposit (Facies 3) indicate a vegetated coastline on this brackish embaymcnt (Fig. 6). Stump diameter (typically <0. 1 m) indicates that most trees were small shrubs, and grew in dense, but generally patchy thickets. Although a few sphcnopsids and lycopsids are recognizable in growth position, the majority of stumps are indeterminate. However, compressed megafloral assemblages in overlying mudstone beds (Facies 3) shed light on their identity, and overall coastal forest composition. The laminated mudstone beds of Facies 3 represent brackish flooding surfaces formed when coastal soils were abruptly drowned by minor subsidence events. Given the small magnitude of subsidence envisaged, the concentration of plant remains in the basal 30-35 mm of Facies 3, and the occasional preservation of intact pteridosperm fronds and Conlaitex leaf arrays (Slopes 1914), plant fragments are probably locally derived from coastal vegetation. Some remains may be parautochthonous having floated off the palaeosol surface during base-level rise, whereas others may have been washed from flooded coastal forests up to a few kilometres distant. Probably only a small amount of material is distally derived, having been transported via major distributary channels (Facies 1) from inland communities (see below).

Given the overwhelming dominance of cordiataleans in the flooding surfaces, we infer that most of the indeterminate stumps in underlying palacosols belong to these plants. Similar cordaitalean coastal forests have been identified along brackish coastlines in the coeval Joggins Formation (Falcon-Lang 2005). Some coastal cordaitaleans have been reconstructed as small trees with stilt roots for growth in standing water (Cridland 1964), and a similar structure is envisaged here. Medullosans, sphenopsids, and a few tree ferns were interspersed within this cordaitalean framework, but herbaceous ferns were rare, perhaps because of standing water.

A subcommunity within this mosaic is represented by Hartt’s fossil bed 1 (Fig. 6). This assemblage is exclusively sphcnopsid, and includes some Calamites in growth position. Fossil bed 1 is associated with the major flooding surface that separates Facies Associations 1 and 2. It is probable the remains are of a monotypic stand of Calamites that coloni/ed rapidly aggrading substrates at the mouth of a major distributary channel. Calamites were well adapted to such a setting by virtue of their ability to rcsprout from underground rhizomes (Gastaldo 1992).

Terrestrial fauna, including gastropods, insects, arachnids and myriapods, also occur in the flooding surfaces (Facies 3); they arc similarly envisaged as being of local derivation, and probably inhabited the coastal forests. The base of the food chain was dominated by detritivores, with most animals living off decaying vegetation in or on the litter layer (DiMichele & Hook 1992). As today, arachnids were probably the main invertebrate predators (Rolfe 19X0). Some Knplioheria myriapods developed spines and repugnatorial glands to ward off attack from these carnivores (Huffman 1969). Diverse winged insects were probably detritivores and herbivores living in the forest canopy (DiMichele & Hook 1992). Their abundance and diversity in the Lancaster Formation compared with coeval sites such as the Joggins Formation (Falcon-Lang et al. 2006) may be a taphonomic artefact. Perhaps insects became trapped on the surface of the brackish sea. and concentrated in strandline accumulations. Tetrapode probably mostly fed on terrestrial arthropods, and footprints in Facies 3 suggest that some animals paddled out into coastal shallows to feed on aquatic fauna.

Alluvial-plain and upland communities

Evidence for communities living on inland alluvial plains and uplands comes from assemblages in Facies 1 (Fig. 6). The riparian fauna included myriapods, if Diplichnites tracks of uncertain provenance are derived from these beds. Fossil plants in abandonment facies of braided fluvial channels imply that riparian gallery vegetation was dominated by cordaitaleans. rhacoptcrid pteridosperms, and sphenopsids. Based on the diameter of trunks found within the base of braided channels (<35 cm), these cordaitaleans were very large trees (for a review, see Falcon-Lang & Bashforth 2005), and probably differed from the small shrubby forms inferred to have grown along the coast (Facies 3 and 4). Tree- trunks may be transported great distances in large fluvial channels, and it is possible that some of these remains represent not only riparian vegetation, but also that of uplands further inland. This is supported by the fact that trunks show calcite- and silica- pcrminerali/alion suggestive of a complex taphonomic history prior to linai deposition.

A number of fossil plant taxa found in very low numbers in the coastal deposits (Facies 3) might also represent plants transported down distributary channels from dryland or uplands. These unusual floral elements include the primitive pteridosperms, Megalopteris (Lcary & Pfefferkorn 1977) and Rliacopteris (Galtier et al. 1997), the progymnosperm, Pseiuinailiiiiantites (Wagner 2001), and the eoniferopsid, Dicranophylliim (Wagner 2005b). Their dryland or upland interpretation is not based on taphonomic data, but inferred from the typical facies and stratigraphie distribution of these fossils at other sites, where they are characteristic of seasonally dry environments of Mississippian or Late Pennsylvania!! age. Their occurrence in lowland deposits in F.arly Pennsylvania)! humid tropical environments implies a dryland or upland source, the only local environment where seasonal conditions might exist. Slopes (1914) was the first to propose this probable intermixture of upland and coastal plants in the Lancaster Formation; this further highlights the quality and far-reaching implications of her classic monograph (Falcon-Lang & Miller 2007).

Conclusions

(1) The Pennsylvanian fossil assemblage in the Lancaster Formation (‘Fern Ledges’) is reviewed for the first time since the classic works of Matthew, Hartt, Dawson, and Stopes in the late nineteenth century and early twentieth century.

(2) Megafloral assemblages indicate a mid- to late Langsettian age when ranges are compared with those in European reference sections. The Lancaster Formation is therefore coeval with the famous Joggins Formation of Nova Scotia.

(3) Facies analysis implies deposition on a tectonically influenced coastal plain, whose braided channels drained into a shallow brackish embayment.

(4) Xiphosurans, crustaceans, and phoronids lived in shallow brackish waters. Cordaitalean forests inhabited by gastropods, insects, arachnids, and myriapods existed on coastal sandbars, and were repeatedly drowned by subsidence events. Forests of giant cordaitaleans, archaic pteridosperms, and plants of uncertain affinity existed in upland environments.

H.J.F.-L. gratefully acknowledges receipt of an NERC Post- doctoral Fellowship held at the University of Bristol, UK (NER/I/S/ 2001/00738), the Matthew Fellowship (2005) from the New Brunswick Museum, and the Tyrell Fund (2006) from the Geological Society of London. R.F.M. gratefully acknowledges receipt of an SSHRC-CURA grant (833-2003-1015) and the support of the New Brunswick Environmental Trust Fund. We thank A. R. Bashforth, S. J. Braddy, J. Kukalova-Peck, R. K. Pickerill, and W. Shear for a critical assessment of the some of the fossil remains. We thank C. J. Cleal for sharing unpublished biostratigraphic data. He and J. Dunlop provided helpful reviews of an earlier version of this manuscript. S. Powell prepared the image in Figure 3e.

References

ABEL, O. 1935. Vorzeitliche Lebensspuren. Gustav Fischer, Jena.

ALCOCK, FJ. 1938. Geology of Saint John Region, New Brunswick. Geological Survey of Canada Memoir, 216.

AMI, H. M. 1900a. Notes bearing on the Devono-Carboniferous problem in Nova Scotia and New Brunswick. Ottawa Naturalist, 14, 121- 127.

AMI, H.M. 1900b. On the sub-divisions of the Carboniferous System in eastern Canada, with special reference to the position of the Union and Riversdale formations of Nova Scotia, referred to the Devonian System by some Canadian geologists. Transactions of the Nova Scolimi Institute of Science, 10, 162-178.

ANDERSON, L.I. & SHUSTER, CN. 2003. Throughout geological time: where have they lived? In: Shuster, C. N., Barlow, R. B. & Brockmann. H.J. (eds) The American Horseshoe Crab. Harvard University Press, Cambridge, MA, 189223.

ARCHER, A.W., CALDER, J.H., GIBLING, M.R., NAYLOR, R.D., REID, D.R. & WIGHTMAN, W.G. 1 995. Invertebrate trace fossils and agglutinated foraminilera as indicators of marine influences within the classic Carboniferous section at Joggins, Nova Scotia, Canada. Canadian Journal of Earth Sciences, 32. 2027-2039.

BASHFORTH, A.R. 2005. Late Carboniferous (Bolsovian) macroflora from the Barachois Group, Bay St. George Basin, southwestern Newfoundland, Canada. Palaeontographica Canadiana. Canadian Society of Petroleum Geologists, 24.

BELL, W. A. 1944. Carboniferous Rocks and Fossil Floras of Northern Nova Scotia. Geological Survey of Canada Memoir, 238.

BREWER, R. 1976. Fabric and Mineral Analysis of Soils. Krieger, Huntington, New York.

BRIGGS. D.E.G., ROLFE, W.D.I. & BRANNAN, J. 1979. A giant myriapod trail from the Namurian of Arran, Scotland. Palaeontology, 22, 273-291.

BUATOIS, L.A., GINIIRAS. M. K. & MACEACHERN, J. et al. 2005. Colonisation of brackish-water systems through time: evidence from the trace-fossil record. Palaios, 20, 321-347.

CALDER. J.H. 1998. The Carboniferous evolution of Nova Scotia. In: Blundell, D.W. & Scott, A.C. (eds) Lyell: The Past is the Key to the Present. Geological Society, London, Special Publications, 143, 261-302.

CALVER, M.A. 1968. Invertebrate faunas. In: Murchison, D. & Westoll, T.S. (eds) Coal and Coal-bearing Strata. Oliver & Boyd, Edinburgh, 147-177.

Cant, D.J. & Walker, RG. 1978. Fluvial processes and fades sequences in the sandy braided South Saskatchewan River, Canada. Sedimentology, 25, 625648. CARPENTER, F.M. 1992. Arthropoda Part R. Superclass Hexapoda. In: Moore. R.C. & Kaesler. R. L. (eds) Treatise on Invertebrate Paleontology. Geological Society of America and the University of Kansas Press, 1-655.

CLEAL, CJ. 1991. Carboniferous and Permian biostratigraphy. In: Cleal, CJ. (ed.) Plant Fossils in Geological Investigation: the Palaeozoic. Ellis Horwood, Chichester, 182-215.

CLEAL. CJ. 2006. The Westphalian macrofloral record from the cratonic central Pennines Basin, UK. Zeitschrift fur Geowissenschaflen, 156, 387-401.

CLEAL, CJ. 2007. The Westphalian Stcphanian macrofloral record from the South Wales Coalfield, UK. Geological Magazine, 144, in press.

COPELAND, M.J. 1957. The Arthropod Fauna of the Upper Carboniferous Rocks of the Maritime Provinces. Geological Survey of Canada, Memoir, 286.

CRIDLAND, A.A. 1964. Ameyion in American coal balls. Palaeontology, 7, 186-209.

CuRRiE, K.L. & Nance, R.D. 1983. A reconsideration of the Carboniferous rocks of Saint John, New Brunswick. Current Research, Part A, Geological Survey of Canada, Paper 83-1A, 23-30.

DARWIN. C. 1871. The Descent of Man and Selection in Relation to Sex. John Murray, London.

DAVIES, S.J., GIBLING, M.R., RYGEL, M.C. & CALDER, J.H. 2005. The Joggins Formation: stratigraphie framework and sedimentological log of the historic fossil cliffs. Atlantic Geology, 41, 115-141.

DAWSON, J. W. 1861. On the Pre-Carboniferous flora of New Brunswick, Maine and Eastern Canada. Canadian Naturalist, 6, 161- 180.

DAWSON, J.W. 1 862. On the flora of the Devonian Period in Northeastern America. Quarterly Journal of the Geological Society of London, 18, 296-330.

DAWSON, J.W. 1868. Acadian Geology or The Geological Structure. Organic Remains, and Mineral Resources of Nova Scotia, New Brunswick and Prince Edward Island, 2nd. MacMillan, London.

DAWSON. J.W. 1871. The Fossil Land Plants of the Devonian and Upper Silurian Formations of Canada. Geological Survey of Canada. Dawson Brothers, Montreal.

DAWSON. J.W. 1880. Revision of the land snails of the Palaeozoic era, with description of new species. American Journal of Science, 20, 403-415.

DIMICHELE, W.A. & HOOK. R.W. 1992. Palaeozoic terrestrial ecosystems. In: Behrensmeyer, A.K., Damuth, J.d., DiMichele, W.A., Hans-Dieter, S. & Wing, S. L. (eds) Terrestrial Ecosystems through Time. University of Chicago Press, Chicago, IL, 205-326.

DOLBY, G. 1989. Palynological analysis of samples from the Cumberland Basin. Nova Scotia Department of Mines and Energy, Open File Report, 89-033.

Elliot, T. 1986. Deltas. In: Reading, H.G. (ed.) Sedimentary Environments and Facies. Blackwell Scientific, Oxford, 1 13-154.

ELLS, R.W. 1907. The Geology and Mineral Resources of New Brunswick. Dawson Brothers, Ottawa.

FALCON-LANG, H.J. 2005. Small cordaitalean trees in a marine- influenced coastal habitat in the Pennsylvanian Joggins Formation, Nova Scotia, Canada. Journal of the Geological Society. London, 162, 485-500.

FALCON-LANG, H.J. 2006a. A history of research at the Joggins Fossil Cliffs, Nova Scotia, Canada, the world’s finest Pennsylvanian section. Proceedings of the Geologists’ Association, 117, 377-392.

FALCON-LANG, H.J. 2006b. Vegetation ecology of Early Pennsylvanian alluvial fan and piedmont environments in southern New Brunswick. Palaeogeography, Palaeoclimatology. Palaeoecology, 233, 34-50.

FALCON-LANG, HJ. & BASHFORTH, A.R. 2005. Morphology, anatomy, and upland ecology of large cordaitalean trees from the Middle Pennsylvanian of Newfoundland. Review of Palaeobotany and Palynology, 135, 223-243.

FALCON-LANG, HJ. SL MILLER, R. F. 2007. Marie Stopes and the Fern Ledges of Saint John, New Brunswick. In: Buri ?. C. & Higgs, B. (eds) The Role of Women in the History of Geology. Geological Society, London, Special Publications, 281, 227-245.

FALCON-LANG, HJ., BENTON, MJ., BRADDY. SJ. & DAVIES, SJ. 2006. The Pennsylvanian tropical biome reconstructed from the Joggins Formation of Canada. Journal of the Geological Society, London, 163, 561-576.

GALTIER, J., MEYER-BERTHAUD, B. & BROWN, R. 1997. The anatomy and seed plant affinities of Rhacopteris and Spathulopteris from the Dinantian (Lower Carboniferous) of Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences, 88, 197-208.

GASTALDO, R.A. 1992. Regenerative growth in fossil horse-tails following burial by alluvium. Historical Biology, 6, 203-219.

GESNER, A. 1840. Second Report on the Geological Survey of the Province of New Brunswick. Henry Chubb, Saint John.

GIBLING. M.R. 1995. Upper Paleozoic rocks. Nova Scotia. In: Williams, H. (ed.) Geology of the Appalachian-Caledonian Orogen in Canada and Greenland. Geological Society of America, The Geology of North America, K-I. 493-523.

GIBLING, M. R.. CALDRR, J. H.. RVAN. R.. VAN DE POLL. H.W. & YEH. G. M. 1992. Late Carboniferous and Early Permian drainage patterns in Atlantic Canada. Canadian Journal of Earth Sciences. 29. 338- 352.

GLAESSNER. M. F. 1957. Palaeozoic arthropod trails from Australia. Palaeontologische Zeitschrift. 31. 103-109.

HAGEN, H.A. 1881. The Devonian insects of New Brunswick, with additional remarks upon a fern in the same slab with Platcphcmcra. Bulletin of the Han’ard Museum of Comparative Zoology. 8. 275 284.

HANDLIRSCH, A. 1906. Revision of American Paleozoic insects. Proceedings of the United Stales National Museum. 29. 661 820.

HANIZSCHEL. W. 1975. Miscellanea. Part W. Supplement I. Trace fossils and problematica. In: TEICHERT. C. (ed.) Treatise on Invertebrate Paleontology. Geological Society of America and University of Kansas Press, 1-269.

HAYES. A.O. & HOWKLI.. B. F. 1937. Geology of Saint John. New Brunswick. Geological Society of America. Special Papers, 5.

H(IIiMAN. 1969. Myriapoda, Exclusive of lnsecta. In: MOORL. R. C. (ed.) Treatise on Invertebrate Paleontology. Geological Society of America and University of Kansas Press, 572-606.

LEARV. R. E. & PEEEEERKORN. H.W. 1977. An Early Pennsylvania>> flora with Megalopteris and Noeggera/hia/es from Itesi Centra/ Illinois. Illinois State Geological Survey Circular, 500.

McLEOD. M.J., JOHNSON. S.C. & RUITENBERCI. A.A. 1994a. Geological map of southwestern New Brunswick. New Brunswick Department of Natural Resources and Energy. Mineral Resources. Map. NR-5.

Mc LLOD. MJ. , JOHNSON. S.C. & RUITENHEKG. A.A. I994b. Geological map of southeastern New Brunswick. New Brunswick Department of Natural Resources and Energy. Mineral Resources. Map, NR-6.

MATTHEW. G. F. 1889. On some remarkable organisms of the Silurian and Devonian rocks in southern New Brunswick. Transactions of the Rityal Society a/Canada. 6. 49-62.

MATTHEW. G. F. 1895a. On the organic remains of the Little River Group, No. II. Transactions of the Royal Society of Canada, 12. 89- 100.

MAIIHLW. G. F. 1895b. Organic remains of the Little River Group, No. IV. Transactions of the Royal Society of Canada. Second Series. 1, 273 -279.

MAIIHEW, G. F. 1895c. On the organic remains of the Little River Group. No. III. Transactions of the Royal Society of Canada. 12. 101- 1 1 1.

MAITHEW. G. F. !896. Ancient Myriapods. Canadian Record of Science, 6, 93-99.

MAI IHEW. O.P. 1897. Description of an extinct Palaeozoic insect, and a review of the fauna with which it occurs. Bulletin of the Natural History Society New Brunswick. 15. 49-60.

MAI IHEW, G-F. 1906. A review of the flora of the Little River Group. Transactions of the Royal Society of Canada. Second Series, 12, 99-149.

MATIHEW, G. F. 191Oa. The geologic age of the Little River Group. Transactions of the Royal Society of Canada. Third Series, 3, 67- 75.

MAIIHEW, G. F. 191Ob. Remarkable forms of the Little River Group. Transactions of the Royal Society of Canada. Third Series. 3. 1 15- 133.

MILLER R.F. I995a. The holotype of Leaia silurica Matthew. 1910. Pennsylvanian. Lancaster Formation. New Brunswick. Atlantic Geology. 31. 23-24.

MILLER. R. F. 1995b. The status of Belinuropsis wigudensis Matthew, 1910, a Pennsylvanian merostome from New Brunswick, Canada. Atlantic Geology. 31, 117-?8.

MILLER. R. F. & BUHAV, D.N. 1988. The Stcinhammer Club: geology and a foundation for a natural history society in New Brunswick. Geoscience Canada. 15, 221-226.

MOKRISSEV. L. B. & BRADDV. S. J. 2004. Terrestrial trace fossils from the Lower Old Red Sandstone, southwest Wales. Geological Journal. 39, 315-336.

NANC-I:. R. D. 1987. Dextral transpression and Late Carboniferous sedimentation in the Fundy coastal zone of southern New Brunswick. In: BEAUMCJNT. C. & TANKARD. A. J. (eds) Sedimentary Basins and Basin-forming Mechanisms. Canadian Society of Petroleum Geologists. Memoirs. 12, 363-377.

NANCE, R. D. & WAKNER, J. B. 1986. Variscan tectonostratigraphy of the Mispec Group, southern New Brunswick: structural geometry and deformational history. Current Research. Part A. Geological Stin’ey of Canada. Paper 86-IA, 351 358.

FLINT, A.C. & VAN DLR Poll.. H.W. 1982. Alluvial fan and piedmont sedimentation in the Tynemouth C’rcek Formation (Lower Pennsylvanian) of southern New- Brunswick. Maritime Sediments and Atlantic Geology, 18, 104-128.

RAST. N.. GRANI. R. H.. PAKKER, J.S.D. & TENG. H.C. 1984. The Carboniferous succession in southern New- Brunswick and its state of deformation. In: GELDSETZER, H.H.J. (ed.) Atlantic Coast Basins. 9th International Congress on Carboniferous Stratigraphy. Comptes Rendus, 3. 13-22.

ROBBINS. D.J.C. 1994. Dr. G. F Matthew’s contributions to lchnology and a review of Acripes Matthew IVlO. BSc thesis. University of New Brunswick, Fredericton.

ROLEE, W.D.I. 1980. Early invertebrate terrestrial faunas. In: PANCHEN. A. L. (ed.) The Terrestrial Environment and the Origin of Land \ertebrates. Systematics Association Special Volume. 15. 117- 157.

ROWLAND. J. M. 1997. The Late Paleozoic insect assemblage at Carrizo Arroyo, New- Mexico. Bulletin of the New Mexico Museum of Natural History and Science. Il, 1-7.

SALTER. J. W. 1863. On some fossil Crustacea from the C’oal Measures and Devonian rocks of the British North America. Quarterly Journal of the Geological Society of London. 19, 519-521. SCUDDER, S. H. I868a. The fossil insects of North America. Geological Magazine, 5, 172-177.

SCUDDER, S. H. I868h. The insects of ancient America. American Naturalist. 1, 625-631.

SCUDDER. S. H. 1880. The Devonian Insects of New Brunswick. Anniversary Memoir of the Boston Society of Natural History.

SCUDDER. S. H. 1885. The Earliest Winged Insects of America. A Reexamination of the Devonian Insects of New Brunswick in the Light of Criticisms and of New Studies of Other Paleozoic Types. Cambridge, MA (published by the author).

SMITH. A.. BRADDV. S.J.. MARRIOTT. S. B. & BRIGGS. D.E.G. 2003. Athropod trackways from the Early Devonian of South Wales: a functional analysis of producers and their behavior. Geological Magazine. 140. 63-72.

SOLEM. A. & Yoc HELSON. E. L. 1979. North American Palaeozoic land snails, with a summary of other Palaeozoic non-marine snails. US Geological Survey, Professional Papers, 1072.

SLOPES, M. C. 1914. The ‘Fern Ledges’ Carboniferous flora of Si. John. New Brunswick. Geological Survey of Canada, Memoirs. 41.

TAYLOR, P.D. & VINN, O. 2006. Convergent morphology in small spiral worm tubes (‘Spirorhis’) and its palaeoenvironmental implications. Journal of the Geological Society, London. 163, 225- 228.

TETLIE. O.E. 2004. Euryplerid phytogeny ami the origin of arachnids. PhD thesis. University of Bristol.

VAN DER POLL. H.W. 1995. Upper Palaeozoic rocks. New Brunswick, Prince Edward Island, and lies dc Ia Madeleine. In: WILLIAMS, H. (ed.) Geology of the Appalachian Caledonian Orogen of Canada and Greenland. Geological Survey of Canada. Geology of Canada, 6. 455 492.

WAGNER. R. H. 2001. The extrabasinal elements in Lower Pennsylvanian floras of the Maritime Provinces. Canada: description of Ailiuntitex. Psciidadidiun/iles. and Rhacoptcrium. Revista Espanola de Paleontologia. 16. 187-207.

WAGNER. R. H. 2005a. Aleihopicris lanci/olia Wagner, a rare element of the Lower Westphalian ‘Fern Ledges’ of Atlantic C’anada. Revista Espanola de Paleontologia, 20. 15-19.

WAGNER. R. H. 2005K Dicranophyllum glahnim (Dawson) Slopes, an unusual element of Lower Westphalian floras in Atlantic C’anada. Revista Espanola de Paleontologia. 20. 7-13.

WEBB, J. A. 1979. A reappraisal of the palaeoecology of conchostracans (Crustacea: Branchiopoda). Neues Jahrbuch fur Geologie unii Paleontologie. Abhandlungen. 158. 259-273.

Wi ? is. M. R.. ALI.ISON. P.A.. PIGGOTT. M. D., PAIN. C.C., HAMI- SON. G.J. & DE OLIVIERA. C. R. E. 2005. Large sea, small tides: the Late C’arboniferous seaway of NW Europe. Journal o/ ‘the Geological Society. London. 162,417-420.

WHITE. D. 1902. Stratigraphy versus paleontology in Nova Scotia. Science, 16. 232 235.

WHITE. D. 1911. Discussion and correspondence: ? Carboniferous flora in the Silurian’. Science. 34. 440 442.

YLO, G-M. & GAO. R. 1986. Late Carboniferous dextral movement on the Caheutnd-llolloH’ fault system. Nova Scotia: evidence and implications. Current Research. Part A, Geological Survey of Canada, Paper 86-1A, 399-410.

Received 7 December 2006; revised typescript accepted 12 January 2007.

Scientific editing by John Marshall

HOWARD J. FALCON-LANG1 & RANDALL F. MILLER2

1 Department of Earth Sciences, University of Bristol, Bristol BS8 IRJ, UK (e-mail: howard.falcon-lang@bris.ac.uk)

2Natural Science Department, New Brunswick Museum, 277 Douglas Avenue, Saint John, N.B., E2K IE5, Canada

Copyright Geological Society Publishing House Sep 2007

(c) 2007 Journal of the Geological Society. Provided by ProQuest Information and Learning. All rights Reserved.