Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242 www.elsevier.nl/locate/palaeo
Upland ecology of some Late Carboniferous cordaitalean trees from Nova Scotia and England H.J. Falcon-Lang a, *, A.C. Scott b a Department of Geology, Royal Holloway, University of London, Egham, Surrey TW20 OEX, UK b British Antarctic Survey, High Cross, Madingley Road, Cambridge CB3 OET, UK Received 29 December 1998; received in revised form 27 July 1999; accepted for publication 30 August 1999
Abstract Permineralised logs and charcoal fragments composed of Dadoxylon-type wood and derived from cordaitalean trees are described from the Upper Carboniferous ( Westphalian) of Nova Scotia and England. They exclusively occur in sandstone bodies interpreted as the deposits of braided, meandering and anastomosing river channels. The architecture and ecology of this type of cordaitalean has not yet been satisfactorily described. Estimates of tree height using an allometric approach suggest that the largest trees attained an approximate height of ca 45 m. The absence of true growth rings in the logs indicate that they grew under non-seasonal humid tropical conditions. However, the presence of charred wood fragments implies that droughts of sufficient intensity to permit wildfire did occasionally occur. The charcoal fragments are usually associated with incised, poorly-sorted, chaotically-bedded channel facies, interpreted as flood deposits. Dramatic increases in runoff and sediment discharge are known to follow the destruction of vegetation in modern wildfire events and the high energy charcoal-bearing channel facies described may represent a similar geomorphic response to cordaitalean vegetation fires. Geomorphic sensitivity to fire is greatest for steeply sloping terrains and it is therefore probable that the cordaitaleans described in this paper occupied upland settings. This palaeoecological interpretation is supported by occurrence of the cordaitalean remains in river channel deposits immediately downstream of regions of palaeo-upland and by the palynological record. The presence of upland forests in Late Carboniferous times has important implications for numerical modelling of global geochemical cycles. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Carboniferous; charcoal; cordaite; fire; palaeoecology; upland
1. Introduction The Cordaitales are an extinct order of gymnosperms (mid Carboniferous–Permian), closely related and possibly ancestral to the conifers (Rothwell, 1988). This study considers the ecology of some Late Carboniferous ( Westphalian) cor* Corresponding author. Fax: +44-1223-362616. E-mail address:
[email protected] (H. Falcon-Lang)
daites. Historically, Westphalian cordaites have been examined in greatest detail from permineralised peats (coal balls) where they occur in an anatomically-preserved state. Two stem genera Mesoxylon and Pennsylvanioxylon, consisting of small (<10 cm diameter) axes have been recognised and are characterised by large, septate piths (termed Artisia when preserved as a sediment cast), Dadoxylon-type pycnoxylic wood and long, helically-arranged, strap-shaped leaves (termed Cordaites). Associated rooting systems have been
0031-0182/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S0 0 3 1 -0 1 8 2 ( 9 9 ) 0 0 14 2 - X
226
H.J. Falcon-Lang, A.C. Scott / Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242
termed Ameylon (Cridland, 1964; Trivett and Rothwell, 1985). During the middle Westphalian these cordaites appear to have formed a significant element of some lowland peat-forming communities (Raymond, 1988; DiMichele and Phillips, 1994). In addition to their coal ball occurrence, permineralised cordaitalean remains have also been encountered in channelised sandstone bodies (Falcon-Lang, 1999a). These cordaitaleans consist of large trunks sometimes more than one metre in diameter and >20 m in length (Lindley and Hutton, 1831). In the past, there has been a tendency to combine these large cordaite logs with the smaller stem remains found in the coal balls in order to create a composite whole-plant reconstruction of the cordaite tree (Grand’Eury, 1877; Trivett and Rothwell, 1985). However, it is now emerging that the Cordaitales were a structurally and ecologically diverse order and that each facies association of plant parts may have been derived from a different cordaite genera, each with a distinct architecture and ecology (Scott, 1979). For example, Mesoxylon was probably derived from a tree of small stature which grew on welldrained peat substrates (Costanza, 1985). In contrast, Pennsylvanioxylon birame has been reconstructed as a small (3–4 m high) tree with stilt roots which grew on peaty soils in a brackish coastal mangrove setting (Cridland, 1964; Costanza, 1985; DiMichele and Phillips, 1994). Another peat mire cordaite, Pennsylvanioxylon nauertianum (=Cordaixylon dumusum) may have been a scrambling shrub (Rothwell and Warner, 1984; Costanza, 1985; DiMichele and Phillips, 1994). The largest cordaitalean specimen so far found in the peat mire facies, a complete permineralised tree, was only 12 m high (Brauer, 1986). No palaeoecological analysis, however, has yet been attempted for the very large cordaitalean logs encountered in channelised sandstone bodies. The aim of this paper is to describe the facies distribution, preservation, and anatomy of cordaitalean remains in channel sandstone facies and to attempt to understand their architecture and palaeoecology. Detailed data is presented from the classic Joggins section, Nova Scotia (Langsettian– Duckmantian) and in addition, supporting infor-
Fig. 1. Sites with cordaitalean remains in Carboniferous channelised sandstone bodies in (a) Nova Scotia and (b) England. (c) Global palaeogeography during the Westphalian showing the position of studied sites (after Golonka et al., 1994). Black areas are mountainous terrains. Stippled areas are lowlands.
mation is given in brief from a further eight Westphalian (Langsettian–Westphalian D) sites in Nova Scotia and England (Fig. 1a and b; Table 1). All these sites lay close to equator during the Westphalian on the eastern margins of the Pangean supercontinent ( Fig. 1c; Golonka et al., 1994).
2. Joggins, Nova Scotia 2.1. Geological setting The Joggins section is found on the southeastern shores of Chignecto Bay (Bay of Fundy),
H.J. Falcon-Lang, A.C. Scott / Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242
227
Table 1 Inventory of cordaitalean logs in channel sandstone facies, Nova Scotia and England Locality
Stratigraphy (age)
Facies interpretation
Number of logs
Log dimensions
Preservation
Source
Joggins, NS
Springhill Mines Formation (Duckmantian) Boss Point Formation (Langsettian)
Meandering river channel
3
34–49 cm diameter, >2.88 m long
Calcareous permineralisation
This study
Large braided river channels
Abundant
50 cm diameter, >5 m long
Calcareous permineralisation
Boss Point, NS
Boss Point Formation (Langsettian)
Large braided river channels
>50
7–30 cm diameter, 1–2 m long
Calcareous permineralisation
Kempt Head, NS
South Bar Formation (Duckmantian)
Large braided river channels
Abundant
>30 cm diameter, length unknown
Calcareous permineralisation
Braided river channels
1
2 cm diameter, 7 cm long
Siliceous permineralisation
G. Browne (written communication, 1998) G. Browne and J.H. Calder (written communication, 1998); this study J.H. Calder (written communication, 1999) This study
Large braided river channels
1
57 cm diameter, 3.12 m long
Calcareous permineralisation
This study
Large meandering river channels Braided river channel
Abundant
20–30 cm diameter, 6–50 cm long
Calcareous permineralisation
This study
24
40–60 cm diameter, >2 m long
Calcareous permineralisation
This study
1
1.46 m diameter, 22.15 m long
Siliceous permineralisation
Lindley and Hutton (1831)
Slacks Cove West, NS
Mabou, NS
Port Hood Formation (Duckmantian) Mabou, Inverness NS Formation (Bolsovian– Westphalian D) Table Sydney Mines Head, NS Formation ( Westphalian D) Priory Above High North, Main Coal, Coal UK Measures (Duckmantian) Wideopen, Above High UK Main Coal, Coal Measures (Duckmantian)
Braided river channel
Cumberland County, Nova Scotia ( Fig. 1a; Gibling, 1987). Between Lower Cove and Ragged Reef Point, a 6 km long coastal outcrop exposes a 2000 m thick succession of Upper Carboniferous strata which dip at 100°/21°SW. These rocks are assigned to the Joggins and Springhill Mines Formations of the Cumberland Group and span the Langsettian–Duckmantian (Gibling, 1995; Fig. 2a). The sequence represents the deposits of a tropical floodplain/mire system which developed within the rapidly subsiding, strike–slip Maritimes Basin ( Fig. 1c; Gibling, 1987; Calder and Gibling, 1994; Golonka et al., 1994; Calder, 1998).
Sediment was largely sourced from the Central Appalachian Mountains 900 km to the south-east, and from local upland terrains within the Maritimes Basin itself such as the Caledonia Highlands Massif ca 22 km to the north and northwest (Gibling et al., 1992). Sandstone body architecture indicates that the Joggins floodplain was dominated by anastomosing and meandering channel geometries ( Way, 1968; Rust et al., 1984; Calder, 1994). Palaeoclimate was predominantly humid and non-seasonal, although during times of glacial maxima more seasonally dry conditions prevailed (Calder, 1994; Falcon-Lang, 1998a).
228
H.J. Falcon-Lang, A.C. Scott / Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242
Fig. 2. (a) Map of Joggins Section, Nova Scotia showing where cordaitalean logs and charcoal fragments occur. Black circles indicate charred fragments and white circles indicate permineralised logs. (b) Summary log of Joggins section (after Gibling, 1995; M.R. Gibling, written communication, 1997) showing cordaite-bearing intervals (c) Foreshore map of MacCarron’s Creek showing the architecture of sandstone body at 1518 m in the Joggins Section and the position of the logs.
Stratigraphic intervals in the Joggins section given below refer to the height in metres above the base of the Joggins Formation ( Fig. 2b; after Gibling, 1995; M.R. Gibling, unpublished sedimentary log, 1997). 2.2. Permineralised logs 2.2.1. Facies analysis Three gymnospermous logs occur together with large (25 cm long), unfragmented Cordaites leaves
near the top of a 3.2 m thick fine-grained sandstone body at 1518 m in the Joggins section, 50 m northwest of McCarron’s Creek (Table 1; Fig. 2b and c; Facies 1b of Falcon-Lang, 1999a). This sandstone body is exposed both in plan view on the wave-cut platform and also in vertical section in the sea-cliffs. On its upper surface it exhibits large scale Ridge and Swale Topography (RST; relief ca 20–30 cm) with a gently arcuate crest orientation trending NW–SE (Fig. 3a). Lower horizons in the sandstone body contain trough cross-bed-
H.J. Falcon-Lang, A.C. Scott / Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242
229
Fig. 3. Sandstone body at 1518 m in Joggins Section. Hammer is 35 cm long. Camera case is 15 cm long. (a) Curvilinear feature (RST ) on top of the sandstone body interpreted as a scroll bar. Arrow points to hammer. (b) Trough cross-beds exposed in plan view as ‘rib and furrow’ structure from near the base of the sandstone body. Arrow points to camera case. (c) Large cordaite log near top of sandstone body. The upper surface is eroded and the central longitudinal depression may represent the position of the pith. (d ) Underside of same cordaite log as (c) showing external surface of the original trunk. Oval mounds on this surface may represent branch scars. (e) Transverse section of trunk showing absence of growth rings. Scale bar is 500 mm.
ding (cross-sets up to 30 cm high) exposed in plan view as ‘rib and furrow structure’ (Fig. 3b). These exhibit south-east palaeocurrents, parallel to the
crest orientation of the RST. The long axes of the cordaite logs are also aligned parallel to the crest orientation of the RST (Fig. 2c). At several hori-
230
H.J. Falcon-Lang, A.C. Scott / Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242
zons, closely spaced in situ casts of Calamites stems occur (up to 13 stems m−2) associated with highly fragmented Asterophyllites. The sandstone unit was formed by a large meandering river channel. RST on the upper surface of the sandstone represents the deposits of scroll bars, abandoned on the point bar platform during fluctuating discharge conditions (Hickin, 1974; Nanson, 1980). Trough cross-bedding in the lower part of the sandstone body was formed by the downstream migration of sinuous-crested megaripples along the channel floor ( Walker and Cant, 1984). The large cordaite logs associated with the RST were probably stranded on the top of the point bar platform during falling stage (cf Nanson, 1981). Dense Calamites stands grew on the unstable channel margins (cf Scott, 1978).
2.2.2. Preservation and identity of logs The largest gymnospermous log is 0.49 m in diameter and >2.88 m in length, its full extent being impossible to ascertain due to partial exposure ( Fig. 3c). The outer surface is characterised by irregularly distributed oval mounds up to 28 cm long, 9 cm wide and 4 cm high which are interpreted to be branch scars (Fig. 3b). All the logs are permineralised by non-ferroan calcite and organic cell walls are preserved intact. Locally, calcite has undergone recrystallisation such that cellular preservation is patchy. Anatomically, the logs are entirely composed of pycnoxylic coniferopsid wood; neither the pith nor the periderm is preserved in any of the specimens. Tracheids, typically 60–80 mm in radial diameter, are characterised by three to five seriate, touching, alternately-arranged, circular/hexagonal bordered pitting on the radial walls and unornamented tangential walls. Pit apertures are oval and oblique. In tangential section, parenchymous rays are common, uniseriate or biseriate and 4–42 cells high. Cross-field pitting is araucaroid. This wood closely resembles Dadoxylon acadianum Dawson (Penhallow, 1900). The absence of a diagnostic pith, however, prevents the attribution of these gymnospermous logs with certainty to the cordaites. Nevertheless, both wood anatomy, which is highly characteristic of cordaites, and the association in the sandstone body
with Cordaites leaves strongly suggest a cordaitalean affinity. 2.2.3. Growth rings Acetate peels were taken from the entire crosssection of the three logs at McCarron’s Creek. An examination of these peels demonstrated that true growth rings are entirely absent despite the preservation of files of cells up to 29 cm long ( Fig. 3e). However, all the logs exhibit sub-concentric zones of crushed cells (up to 1 mm wide), spaced 3– 13 mm apart which superficially resemble growth rings. The crush zones may be traced laterally around the circumference of the log to regions where cellular structure is intact. These regions lateral to the crush zones are in some cases characterised by a marked increase in the radial cell diameter from ~75 to ~100 mm (Fig. 4). Increased cell size persists radially for ca five to ten cells before returning to normal. Similar subtle interruptions to wood growth have been illustrated by Jeffrey (1917) in British Upper Carboniferous cordaitaleans. Such zones (being structurally heterogeneous) would have been weaker and the wood has evidently preferentially concertinaed
Fig. 4. The variation in radial transverse cell diameter across a section of cordaitalean wood laterally adjacent to a crush zone. Five adjacent files of cells were measured (after Creber and Chaloner, 1984). Note that the position of the crush zone is associated with a sudden increase in radial cell diameter. Specimen from Joggins.
H.J. Falcon-Lang, A.C. Scott / Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242
along them in response to burial pressure to produce the sub-concentric crush zones observed. 2.3. Charred wood fragments 2.3.1. Facies analysis Abundant gymnospermous wood fragments, preserved as charcoal (=fusain), the product of wildfire (Scott, 1989), occur commonly in a channel sandstone body at 477 m in the Joggins section near Lower Cove (Fig. 2a; Facies 1a of FalconLang, 1999a). This unit is a 4–7 m thick, fine to medium-grained sandstone composed of a series of small (1–2.5 m deep, 10–20 m wide), U-shaped, erosionally-based channel bodies which possess basal groove casts. Three types of deposit infill the U-shaped channels and graphic logs of these may be found in Fig. 2b of Falcon-Lang (1999a). The most common variety ( Type A) consists of massive, fine-grained sandstone units which exhibit a slight upward fining. Locally, low-angle erosive surfaces, rare cross-beds and mudstone/calcrete intraclasts occur. Width:depth ratios for this channel type range from 4.9–15 (mean: 9.9; N=11). The second type of channel fill ( Type B) is composed of laminated green-red mudstone/siltstone beds which locally contain cross-bedded, finegrained sandstone units. Width:depth ratios for this channel type range from 7.5–36 (mean: 22.5; N=3). The third type of channel fill ( Type C ) is composed of a multi-storey sequence of heterolithic, erosionally-based packages. These packages include cross-bedded medium-grained sandstone intervals containing abundant mudstone clasts, sulphur-rich, chaotically bedded, medium-grained sandstone intervals containing abundant plant fragments (1 m long coalified logs, Cordaites, Artisia and Calamites) and calcrete breccia intervals containing terrestrial gastropods (Pupa). The width:depth ratio of the only channel of this type measured was 3.7 ( Falcon-Lang, 1998a). The nature of Type A channels, characterised by an initial phase of down-cutting followed by the deposition of an unstructured fill, implies that they were formed in vertically-accreting fluvial channels with stable banks. In these respects they most closely resemble the deposits of anastomosing bedload rivers (cf Smith, 1986; Stanistreet et al.,
231
1993). Although, it is impossible to demonstrate conclusively that multiple channels were active at a given time (the defining feature of anastomosing fluvial systems), the absence of very large channels in the 2 km stratigraphic thickness of the Joggins section is consistent with the interpretation of these units being the product of a large anastomosing, network of small channels. Mudstone-dominated deposits ( Type B) were most likely formed during abandonment of individual channels as is common in anastomosing systems (Stanistreet et al., 1993). In contrast, heterolithic channel deposits ( Type C ) were formed during irregular pulses of high discharge. The occurrence of beds of calcrete breccia in these units imply both high energy conditions and that sediment was primarily sourced from well-drained, seasonally-dry regions upstream (Goudie, 1983). In addition, the occurrence of large volumes of fossil plant material indicates that vegetation adjacent to the channel was also greatly disturbed. Other charcoal-bearing channel sandstone bodies not described in detail in this paper are present at 462, 562, 1165 and 1520 m. 2.3.2. Preservation and identity of wood fragments Charred gymnospermous wood fragments occur exclusively within the heterolithic channel bodies ( Type C ), covering up to 5% of the surface area of bedding planes (5% cover; Fig. 5a). They are 1–2 cm in size, cubic, sub-rounded and exhibit exquisite anatomical preservation, a consequence of the charring process (Scott, 1989). The charcoal fragments entirely consist of pycnoxylic coniferopsid wood. Tracheids (45–50 mm in diameter) are characterised by three to four seriate, alternatelyarranged, bordered pits on the radial walls and unornamented tangential walls ( Fig. 5b). Pit apertures are oval and oblique. Parenchymous rays are common, uniseriate or biseriate and 2–55 cells high ( Fig. 5c). Cross-field pitting is araucaroid. Growth rings are absent. The wood fragments are assigned to Dadoxylon materiarium Dawson (Penhallow, 1900) and are attributed to the cordaitaleans. Tracheid diameters are considerably smaller in these specimens compared with those encountered in permineralised cordaitalean wood elsewhere in the Joggins section. This reflects the charring process during which an ca 30% reduction
232
H.J. Falcon-Lang, A.C. Scott / Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242
in volume occurs (Jones, 1991) and has no taxonomic significance. The charcoal fragments have a mean standard reflectance under oil of 5.01% when
measured with the method of Jones et al. (1991). Locally, 10 cm long, partially-charred fragments of Artisia, the pith cast of cordaitalean trees, also occur ( Fig. 5a).
3. Other localities 3.1. North-west Nova Scotia Cordaite remains are highly abundant in the 750 m thick Boss Point Formation (Langsettian) which crops out along the margins of Chignecto Bay, north-west Nova Scotia ( Fig. 1a). This formation is characterised by coarse-grained, erosionally-based sandstone packages up to 90 m thick, interpreted as the deposits of large elongate deltas (Plint and Browne, 1994). Cordaitalean remains occur within fining-upwards, trough cross-bedded, channelised sandstone units (1–5.5 m thick) which were formed in major braided distributary channels on the delta top (Browne and Plint, 1994). In a 3 km long coastal section at Boss Point >50 gymnospermous logs (7–30 cm diameter, 1–2 m long) occur within this braided distributary channel facies (Dawson, 1868; Calder, 1998, written communication, 1998). The logs, which are preserved as calcareous permineralisations, consist of the pycnoxylic coniferopsid wood, Dadoxylon acadianum Dawson. In addition, some examples exhibit large (48 mm diameter), septate piths confirming their cordaitalean identity (Fig. 6a). Charred coniferopsid wood of Dadoxylon-type is also found in the braided distributary channel facies as occasional, small (2–5 mm) fragments (<1% cover; Falcon-Lang, 2000a). At another site, Mary’s Point near Slacks Cove West (Plint and Browne, 1994) very large calcareously permineralised cordaitalean logs are also present. They are
Fig. 5. Charcoal from sandstone body at 1520 m in the Joggins Section. (a) Bedding plane scattered with cubic blocks of black charcoal. Coin for scale (15 mm diameter). Arrow on left points to partially-charred Artisia pith cast. Arrow on right points to cubic Dadoxylon charcoal fragment (b) SEM of radial view of Dadoxylon-type wood showing 3-seriate, alternate pitting. (c) SEM of tangential view of Dadoxylon-type wood with one to two seriate rays up to 55 cells high.
H.J. Falcon-Lang, A.C. Scott / Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242
233
ca 0.5 m in diameter and >5 m long ( Fig. 1a; G. Browne, written communication, 1998). 3.2. Cape Breton, Nova Scotia Cordaitaleans are also found in Cape Breton Island, north-eastern Nova Scotia. At Table Head, near Sydney, Eastern Cape Breton, calcified gymnosperm stems are found in the Sydney Mines Formation ( Westphalian D) where they occur in an erosionally-based, coarse-grained, 4 m thick channelised sandstone body interpreted as the deposit of a large meandering river channel (Calder et al., 1996; Fig. 1a). These stems, which consist of Dadoxylon-type wood, are most commonly washed out of the cliffs, accumulating on the foreshore. At Sydney Mines, the type section of the Sydney Mines Formation, also in Eastern Cape Breton, charred coniferopsid wood fragments occur within large, multi-storey channel sandstone bodies which infill a palaeo-valley (Gibling and Bird, 1994; Tandon and Gibling, 1994; Fig. 1a). The fragments are small, highly crushed and poorly preserved but probably have cordaitalean affinities ( Falcon-Lang, 2000a). At Kempt Head, Central Cape Breton, large permineralised cordaitalean logs several decimetres in diameter occur in the braided fluvial sandstone deposits of the Duckmantian South Bar Formation (Rust and Gibling, 1990; J. H. Calder, written communication, 1999). Finally, at Mabou in Western Cape Breton, small silicified and calcified wood fragments and logs occur in the Port Hood Formation (Calder, 1998; Langsettian; Fig. 1a). They occur in sandstone bodies produced by low sinuosity, braided river systems ( Keighley and Pickerill, 1996). A large calcified log also occur in the
Fig. 6. Material from other localities. (a) Transverse section of small calcified cordaite log from Boss Point, Nova Scotia showing a large septate pith (p) surrounded by a thick cylinder of wood. Much of the wood anatomy has been destroyed by recrystallisation (r). Arrow points to a thin (dark) layer of wood where anatomy is intact. Field of view is 5 cm wide. (b) Large cordaite log in a channel sandstone body at Priory North, near Durham, UK (photograph from Geoff Willis). (c) Some log fragments extracted from Priory North Opencast Site. Scale is 10 cm. The log in the foreground is ca 50 cm in diameter.
234
H.J. Falcon-Lang, A.C. Scott / Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242
Inverness Formation (Bolsovian–Westphalian D) at Mabou associated with a very thick (93 m), multi-storey channel sandstone unit formed by a braided river (Gibling, 1995). In both cases, wood is of Dadoxylon-type and is attributed with reservation to cordaites.
3.3. Northern England Large cordaitalean trunks are known from two sites in England (Fig. 1a). The first and as yet largest example was recorded by Lindley and Hutton (1831; plate 1) from a coarse-grained sandstone body above the High Main Coal at Wideopen coal mine (now disused) near Newcastle (Duckmantian). This ‘Wideopen Log’ was 22.15 m long, 1.46 m in diameter (tapering to 0.46 m) and was characterised by numerous, randomly distributed branch scars. The trunk consisted entirely of the pycnoxylic coniferopsid wood Dadoxylon. No palaeoevironmental analysis of the coarse-grained sandstone body was given in Lindley and Hutton’s description of the trunk but sandstone units of the same age and geographic region have been interpreted as the deposits of large, braided distributary channels in a delta top setting (Fielding, 1984). Close to the same stratigraphic interval as Wideopen, at Priory North Opencast Coal Mine, near Durham (Duckmantian), 24 large (40–60 cm diameter, >2 m long) logs occur in a channel sandstone body above the High Main Coal (Fig. 6b and c). This sandstone body has a deeply erosional base, in places cutting down into an underlying coal seam, and exhibits large scale (1 m high) cross-bedding. This unit is also interpreted as a braided river channel deposit. The Priory North logs are generally aligned parallel to palaeocurrent and their concentration in this opencast mine suggests that they may have become stranded on sandbars during times of low discharge. Anatomically, the trunks consist of Dadoxylontype wood and in transverse section, although they do not possess true growth rings, they do exhibit regularly-spaced, sub-concentric crush zones. Trunks from both Wideopen and Priory North probably have cordaitalean affinities.
4. Palaeoecological interpretation 4.1. Architecture and growth An estimate of the height attained by the cordaitalean trees described in this paper may be gained by projecting the taper of the logs to their vanishing point. For example, the 22.15 m long Wideopen log tapered from a diameter of 1.46 m at one end to 0.46 m at the other. Projecting this taper to the vanishing point would imply a minimum height for the living tree of ca 32 m. Only minimum estimates can be made as it is impossible to determine how high the thick end of the log was above the trunk base. Using another method of estimating tree height from log diameter, the allometric approach of Niklas (1994), a value of 44.59 m was attained for the Wideopen tree. These trees were much taller than cordaitaleans previously described from peat mire facies (cf Cridland, 1964; Brauer, 1986). Of the modern tropical conifers, the araucarians probably possess the greatest structural similarity to the cordaites and mature trees of this family range in height from 30-89 m ( Enright, 1995). In contrast to many modern araucarians, which bear branches in whorls, branches appear to have been borne irregularly on cordaitalean trunks as evidenced by the position of the branch scars (Lindley and Hutton, 1831). The complete absence of true growth rings in all the trunks studied makes it tremendously difficult to estimate the age of these large trees. Cordaitaleans are believed to have been amongst the slowest growing trees of the Late Carboniferous forests (Rothwell, 1993). A maximum diametric growth rate of 6.3 mm year−1 has been measured for modern araucarian conifers growing in the tropical zone ( Enright, 1995). Applying this value to the cordaitaleans, the 1.46 m diameter trunk from Wideopen may have been >230 years old when it died. Undisturbed stands of modern tropical araucarian conifers may attain ages of >1000 years ( Enright, 1995). The presence of only very subtle, irregularlyspaced growth interruptions in the woods of some of the cordaitalean trunks studied ( Fig. 4) probably indicate tree growth under a non-seasonal humid tropical climate (Creber and Chaloner,
H.J. Falcon-Lang, A.C. Scott / Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242
1984). However, Falcon-Lang (2000b) has recently shown that the degree to which growth rings are formed in conifer woods is inversely related to the leaf life-span of the tree. For example, conifers with very high leaf longevities (e.g. araucarians) produce only very subtle growth rings even under intensely seasonal monsoonal climates (FalconLang, 1999b, c). It is evident from the length of leaf traces in some permineralised stems, that cordaites were evergreen and maintained large, photosynthetically active leaves for long periods (Rothwell, 1993). It follows, therefore, that cordaites possessed wood with a low sensitivity to climatic seasonality like modern araucarian conifers. Furthermore, the incorporation of cordaite logs into fluvial sandstone units suggest that the trees were growing in streamside niches. In such environments, adequate water supply may exist even during periods of seasonal drought and therefore trees may not always produce growth rings in their woods (Demko et al., 1998; Falcon-Lang, 1999b). Taking into account both these considerations, it is possible that the cordaites described here may have grown under a slightly seasonal climate but not expressed this in their wood anatomy. Indeed, the presence of calcrete fragments associated with cordaite material within some channel units at Joggins, implies that climate was periodically seasonal (Goudie, 1983). Nevertheless, the absence of growth rings in cordaite woods is striking, even when compared to the woods of araucarians growing in seasonally humid tropical environments (Ash, 1983). Therefore, we conclude that the most likely environment of growth for the large cordaites was a non-seasonal humid tropical setting with little or no intra-annual variation in rainfall or temperature. 4.2. Fire ecology At several localities, cordaite wood commonly occurs as charcoal indicating that these coniferopsid communities were occasionally prone to wildfire, probably initiated by lightning ( Falcon-Lang, 2000a). Jones et al. (1991) demonstrated that a linear relationship exists between mean charcoal reflectance and charring temperature and therefore
235
temperatures attained in ancient fires may be crudely estimated. Reflectance values derived from Joggins charcoal (5.01%) imply that cordaite fires exceeded temperatures of 760°C as is typical of some modern conifer forest fires (Scott and Jones, 1994). It is more difficult to gauge the frequency of cordaite fires, however. Geochemical models have suggested that atmospheric oxygen levels were very high in Late Carboniferous times (up to 35%; Berner and Canfield, 1989). Beerling et al. (1998) numerically modelled the impact of such an oxygen-rich atmosphere on Carboniferous vegetation dynamics and predicted that fires would occur about every 3–6 years on well-drained soils under such conditions. Given that cordaitaleans were very long-lived trees, it is hard to believe that they could have tolerated such high fire frequencies. Modern environments with high fire frequencies, such as tropical savannahs, favour the growth of plants with rather short life-spans ( Walter, 1973). It is significant that when viewed in transverse section, the large permineralised cordaite logs at Joggins completely lack fire-scars. Fire scars are wounds in the wood formed when an individual tree is scorched during a fire event but its vascular cambium subsequently recovers to continue growth (Gutsell and Johnson, 1996). The high longevity of cordaites and the absence of fire-scars in their woods suggest that Late Carboniferous fire frequency was probably not as high as Beerling et al. (1998) have predicted.
4.3. Taphonomy of cordaitalean material The remains of the large cordaitalean described here are not found in situ at Joggins or sites elsewhere. Therefore, speculations concerning the transport history of this plant material must be made in order to ascertain the site of growth of these trees. Actualistic studies suggest that most plant debris in modern river channel deposits is derived from the adjacent channel leve´e (Scheihing and Pfefferkorn, 1984). Whilst this is certainly true of leaves which are rather fragile and therefore rapidly destroyed by fluvial processes, tree trunks being both highly buoyant and robust have the
236
H.J. Falcon-Lang, A.C. Scott / Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242
potential to undergo many hundreds of kilometres of fluvial transport intact. Many references concerning the transport potential of tree trunks exist. For example, Eggertsson (1994) studied the taphonomy of tree trunks in the MacKenzie Delta of northern Canada, first noted by Sir Charles Lyell (Scott, 1998). He matched the dendrochronology of 290 allochthonous conifer logs with the tree-ring chronologies of known geographic regions and demonstrated that although 29% of the logs were derived from the immediate MacKenzie Delta vicinity and had been transported <200 km, the larger proportion of logs had been sourced from the Laird River tributary, some 1300 km upstream (cf Kindle, 1921). Logs apparently had entered the Laird river system during phases of lateral channel migration or following environmental disturbances such as storms or fires. During their long passage downstream, the trunks periodically became stranded on sand-bars before being flushed out into the deltaic region by flood events. Most logs reached the MacKenzie Delta within 25 years of death ( Kindle, 1921; Eggertsson, 1994). A little is also known about the transport potential of tree trunks in Late Carboniferous ( Westphalian) river channels. For example, Liu and Gastaldo (1992) described some partially hollow lycopsid logs from the Warrior Basin of the USA which were filled with out-sized clasts (gravel, pebbles and one boulder). Provenance studies showed that these clasts had been derived from the Appalachian mountains >50 km upstream. Liu and Gastaldo (1992) concluded that the hollow logs had accumulated the coarse detritus while they lay stranded on an extra-basinal floodplain before being remobilised during a flood and transported >50 km downstream into the depocentre. In addition, it is clear than wood charcoal fragments also potentially may be transported for very long distances in fluvial systems. Experiments using wave tanks have shown that 1–2 mm charcoal fragment remain in suspension for several days and that settling times increase with increasing particle sizes ( Vaughn and Nichols, 1995). In addition, studies have shown that charcoal fragments incorporated into the bedload of braided
fluvial systems in Australia are several hundred years old, implying that they had been transported for long distances (Blong and Gillespie, 1978). Many of the charcoal fragments encountered in the river channel facies at Joggins exhibit a degree of roundness consistent with long episodes of fluvial transport (cf Cripps et al., 1997). Although, it is not possible to determine with certainty the site of growth of the cordaitalean trees described here from taphonomy, the trees must have been growing closely adjacent to the river channels either on the channel levee´s in the basins or up to several hundred kilometres upstream.
4.4. Geomorphic impact of Cordaitalean vegetation fires Cordaite charcoal is not randomly distributed at Joggins but is most commonly found in channel sandstone bodies that were deposited under particularly high energy conditions (Type C ) whilst it is generally lacking in the other channel sandstone bodies ( Types A and B). Charcoal-bearing channel deposits typically contain abundant mudstone intraclasts indicating high levels of bank erosion and very large quantities of cordaitalean debris implying upstream disturbance to vegetation. In addition, they are poorly sorted and chaoticallybedded indicating rapid deposition and have lower width:depth ratios indicating channel incision under conditions of high discharge. It is well known that vegetation plays a very important role in influencing cycles of erosion and deposition (Langbein and Schumm, 1958). For example, vegetation cover cushions the impact effects of rain droplets, limits frost-shattering and protects soils from wind erosion whilst roots bind the soils together (Swanson, 1981). Furthermore, in the humid to seasonally humid tropics vegetation may intercept as much as 88% of rainfall through evapotranspiration thus substantially reducing catchment runoff (Salati, 1987). The sudden removal of vegetation and litter during fires can therefore cause rates of runoff and sediment discharge to dramatically increase (Barro and Conrad, 1991). For example, debris flows were common in steep catchments following the 1988
H.J. Falcon-Lang, A.C. Scott / Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242
Yellowstone Fires in the USA (Meyer et al., 1992; Meyer and Wells, 1997). In addition, following some Californian fires, Wells (1981) recorded postfire floods where run-off increased by nine-fold and sediment discharge by 200-fold, 1–2 km downstream of the burned slopes. It is suggested that high energy fluvial facies associated with charcoal at Joggins may represent a similar geomorphic response to cordaitalean fires. Such fire–flood sequences have also been identified in the Lower Carboniferous rocks of Ireland (Nichols and Jones, 1992; Falcon-Lang, 1998b).
5. Discussion: a Carboniferous upland flora In summary, it is known that the large cordaitalean logs found in Late Carboniferous channel sandstone units were derived from tall trees of up to 45 m in height, many times larger than cordaitaleans previously described from peat mire facies. These large cordaite trees lived for several hundred years and appear to have grown under a nonseasonal, humid, tropical palaeoclimate. Despite this climatic humidity, fires occurred in these cordaitalean communities. Destruction of this vegetation cover may have caused soils to become destabilised, leading to pulses of sediment being introducing into downstream fluvial systems. The key outstanding question surrounding the ecology of these trees concerns their site of growth. In the following discussion, a wide body of circumstantial evidence is summarised which implies that the large cordaitaleans trees may have grown in ‘upland’ environments. The term ‘upland’ is defined in this paper as any extrabasinal region of net erosion containing or bounded by slopes of >10°. The term includes both area of small elevation (100–200 m) such as local faulted highs within depositional basins in addition to true mountainous terrains (after Pfefferkorn, 1980; FalconLang, 1998a). 5.1. Evidence for upland ecology Given that upland terrains are never incorporated into the long term (Carboniferous) geological record, evidence for plants growing in such envi-
237
ronments is by nature circumstantial and tentative. Three lines of evidence may suggest that the tall cordaite trees grew in extrabasinal uplands rather than on fluvial levee´s in basinal regions. First, all the cordaitalean fossils described in this paper are always found close to regions of palaeo-upland. For example, the specimens from Joggins, Nova Scotia occur immediately downstream of two minor upland regions. The Lower Palaeozoic rocks of the Caledonia Highland Massif, generated in Late Namurian times due to transpression along the Harvey–Hopewell Fault lay 22 km north-west of Joggins and a smaller upland terrain composed of Vise´an rocks, the Minudie Anticline, lay 8 km north of Joggins (Calder, 1994, 1998). The cordaite trunks at Boss Point and Slacks Cove West, are found even closer to these same palaeo-uplands, occurring in river channel deposits only 2–20 km downstream (Calder, 1994). At Table Head, Kempt Head, Sydney and Mabou in Cape Breton Island log-bearing sandstone units all occur within 15 km of the basin margins (Calder, 1998) and in northern England trunks occur in the deposits of large braided rivers, <100 km downstream of the Southern Uplands topographic high (Cope et al., 1992). In strong support of this argument for an upland ecology for the large cordaites, Calder (1998) observed that cordaite trunks in Nova Scotia are most abundant immediately overlying unconformable surfaces. These data indicate that incursions of cordaitalean logs into Nova Scotian depocentres were usually coincident with major phases of uplift and hinterland rejuvenation and that cordaites were therefore most likely stripped from rising upland areas. A second line of evidence supporting an upland ecology concerns the geomorphic impact of cordaitalean vegetation fires. Geomorphic-sensitivity to fire increases with increasing slope. For example, Menaut et al. (1993) recorded sediment yields 100 times greater following savannah fires on 15° slopes compared with fires on the plains. In burned steepland settings, annual runoff may increase up to 10-fold and sediment discharge up to 1000-fold above background values (papers summarised in Falcon-Lang, 1998a). If we accept that the chaotic channel deposits bearing cordaite logs and charcoal represent fire-flood sequences and that this
238
H.J. Falcon-Lang, A.C. Scott / Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242
phenomenon is most pronounced in burned steepland terrains then this implies that the charred cordaitaleans found in the channel facies at Joggins were playing an important role in slope stabilisation and therefore probably occupied extra-basinal upland niches (Falcon-Lang, 1999a). The third piece of data supporting the existence of upland cordaites, relates to the sequence stratigraphic distribution of cordaite pollen. Neves (1958) noted that cordaite pollen grains (Florinites) were most abundant in a Namurian sequence in Staffordshire ( UK ) during the time of maximum marine transgression. Chaloner (1958), reinterpreting Neves’s data, suggested that this phenomenon could best be explained in terms of pollen input from surviving upland cordaitalean communities following the drowning of lowland swamp communities during the transgression. Since then Florinites peaks have been noted in other maximum transgression deposits ( Turner et al., 1994). Additional studies have indicated that Florinites and Potonieisporites (possible conifer pollen) is particularly abundant in some basins immediately adjacent to known areas of palaeoupland (Scott, 1979; Scott and Chaloner, 1983). Macrofloral remains in some maximum transgression deposits are also cordaite dominated and these too have been interpreted as upland floras (Scott et al., 1997).
5.2. Implications of upland ecology All these data imply, but do not prove beyond doubt, that the tall cordaitalean trees found in Late Carboniferous Euramerican channel facies grew in upland terrains. Because of the difficulties in identifying upland floras discussed in this paper, the timing of upland colonisation is poorly constrained ( Frederiksen, 1972; Scott, 1980; Mapes and Gastaldo, 1986). Nevertheless this is an extremely important question to address, because through influencing the rate of silicate weathering, the evolution of upland floras almost certainly exerted a profound impact on atmospheric CO 2 levels (Algeo and Sheckler, 1998). For example, the dramatic decline in atmospheric CO estimated 2 at the Late Devonian–Early Carboniferous bound-
ary, which culminated in the Carboniferous– Permian Gondwanan glaciation (Berner, 1998), may have been linked to the establishment of upland floras. It is generally accepted that by the end of the Carboniferous upland floras existed and were dominated by conifers, cordaites, seed ferns and cycads (Frederiksen, 1972; Scott and Chaloner, 1983; Rothwell and Mapes, 1988; Lyons and Darrah, 1989). This hypothesis is based on abrupt the appearance of morphologically advanced walchian conifers in the Late Westphalian– Stephanian coincident with an increase in climatic aridity (Moore et al., 1936; Cridland and Morris, 1963; Leary, 1975, 1981; Winston, 1983). This event is interpreted to represent the migration of conifers from the seasonally-dry upland terrains where are they had evolved, to the lowland basins ( Frederiksen, 1972; Scott, 1980; DiMichele and Aronson, 1992). It is evident that Late Westphalian–Stephanian conifers were adapted to water-stress; foliage has a xeric character and woods exhibit growth rings (Baxter and Hartman, 1954; Rothwell and Mapes, 1988). This study presents evidence for the existence of upland floras some 5–10 million years earlier in the Langsettian–Duckmantian. These floras were dominated by cordaitaleans as indicated by permineralised logs and charcoal in river channel deposits. Some early conifers were also present; putative conifer charcoal is associated with cordaite remains at Joggins ( Falcon-Lang, 1999a). Biological innovations such as easily wind-dispersed pollen and the production of droughttolerant seeds (DiMichele and Phillips, 1994) may have facilitated the colonisation of uplands by these cordaitalean trees. In contrast to those upland communities which followed in the Late Westphalian–Stephanian, Early Westphalian upland cordaite forests had a distinctly mesic character; growth rings are entirely absent in all cordaitalean logs studied, and leaves were very long and broad. Such mesic cordaitalean upland forests probably evolved towards the beginning of the Namurian as evidenced by the palynological record (Chaloner, 1958; Scott, 1979; Scott and Chaloner, 1983) and were gradually replaced by more xeric conifer forests during the Late
H.J. Falcon-Lang, A.C. Scott / Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242
Westphalian in response to increasing climatic aridity.
Acknowledgements The authors thank John Calder, Greg Browne, Randy Miller and Martin Gibling for generous field support in Nova Scotia, for use of unpublished data and for stimulating discussion. The extremely thoughtful and thorough comments of Chris Cleal and an anonymous reviewer greatly improved this paper. HJFL acknowledges the receipt of a Royal Holloway, University of London grant and Geology Department Research Committee funding to undertake fieldwork in Nova Scotia. ACS acknowledges NATO Grant 940559. The authors also thank Tony Johnson, Brian Young and Geoff Willis for field support at Priory North ( UK ). Photographs were kindly printed by Kevin D’Souza and the staff of the Electron Microscopy Unit (Royal Holloway, University of London).
References Algeo, T.J., Scheckler, S.E., 1998. Terrestrial–marine telecommunications in the Devonian: links between the evolution of land plants, weathering processes, and marine anoxic events. Phil. Trans. R. Soc. London Series B 353, 113–130. Ash, J., 1983. Growth rings in Agathis robusta and Araucaria cunninghami from tropical Australia. Aust. J. Bot. 31, 269–275. Barro, S.C., Conrad, S.G., 1991. Fire effects on Californian chaparral systems: an overview. Environment Int. 17, 135–149. Baxter, R.W., Hartman, E.L., 1954. The coal-age flora of Kansas V. A fossil coniferophyte wood. Phytomorphology 4, 316–325. Beerling, D.J., Woodward, F.I., Lomas, M.R., Wills, M.A., Quick, W.P., Valdes, P.J., 1998. The influence of Carboniferous palaeoatmospheres on plant function: an experimental and modelling assessment. Phil. Trans. R. Soc. London B 353, 131–140. Berner, R.A., 1998. The carbon cycle and CO over Phanerozoic 2 time: the role of land plants. Phil. Trans. R. Soc. London B 353, 75–82. Berner, R.A., Canfield, D.E., 1989. A new model for atmospheric oxygen over Phanerozoic time. Am. J. Sci. 289, 333–361.
239
Blong, R.J., Gillespie, R., 1978. Fluvial transported charcoal gives erroneous 14C ages for recent deposits. Nature 271, 739–741. Brauer, D.F., 1986. A permineralised cordaitalean tree from the middle Pennsylvanian of Iowa. Am. J. Bot. 73, 716 Browne, G.H., Plint, A.G., 1994. Alternating braidplain and lacustrine deposition in a strike–slip setting: the Pennsylvanian Boss Point Formation, Maritime Canada. J. Sediment. Res. B 64, 40–59. Calder, J.H., 1994. The impact of climate change, tectonism and hydrology on the formation of Carboniferous tropical intermontane mires: the Springhill coalfield, Cumberland Basin, Nova Scotia. Palaeogeogr. Palaeoclimatol. Palaeoecol. 106, 323–351. Calder, J.H., 1998. The Carboniferous evolution of Nova Scotia. In: Blundell, D.J., Scott, A.C. ( Eds.), Lyell: The Past is the Key to the Present, Geol. Soc. London Spec. Publ. 143, 261–302. Calder, J.H., Gibling, M.R., 1994. The Euramerican Coal Province: controls on Late Paleozoic peat accumulation. Palaeogeogr. Palaeoclimatol. Palaeoecol. 106, 1–21. Calder, J.H., Gibling, M.R., Eble, C.F., Scott, A.C., MacNeil, D.J., 1996. The Westphalian D fossil forest at Table Head, Sydney Basin, Nova Scotia: sedimentology, paleoecology and floral response to changing edaphic conditions. Int. J. Coal Geol. 31, 277–313. Chaloner, W.G., 1958. The Carboniferous upland flora. Geol. Mag. 95, 261–262. Cope, J.C.W., Guion, P.D., Sevastopulo, G.D., Swan, A.R.H., 1992. Carboniferous, in: Atlas of Palaeogeography and Lithofacies, Cope, J.C.W., Ingham, J.K., Rawson, P.F. ( Eds.), Geol. Soc. Lond. Mem. 13, 67–84. Costanza, S.H., 1985. Pennsylvanioxylon of middle and upper Pennsylvanian coals from Illinois Basin and its comparison with Mesoxylon. Palaeontographica B 197, 81–121. Creber, G.T., Chaloner, W.G., 1984. Influence of environmental factors on the wood structure of living and fossil trees. Bot. Rev. 50, 357–448. Cridland, A.A., 1964. Ameylon in American coal balls. Palaeontology 7, 186–209. Cridland, A.A., Morris, J.E., 1963. Taeniopteris, Walchia and Dichophyllum in the Pennsylvanian system of Kansas. Univ. Kansas Sci. Bull. 44, 71–82. Cripps, J.A., Nichols, G.J., Scott, A.C., Collinson, M.E., 1997. Modelling charcoal deposition in sandy fluvial systems. Unpublished abstract, Plant Taphonomy Meeting, Cardiff, 4. Dawson, J.W., 1868. Acadian Geology. Macmillan, London. Demko, T.M, Dubiel, R.F., Parrish, J.T., 1998. Plant taphonomy in incised valleys: implications for interpreting paleoclimate from fossil plants. Geology 26, 1119–1122. DiMichele, W.A., Aronson, R.B., 1992. The Pennsylvanian–Permian vegetational transition: a terrestrial analogue to the onshore–offshore hypothesis. Evolution 46, 807–824. DiMichele, W.A., Phillips, T.L., 1994. Palaeobotanical and palaeoecological constraints on models of peat formation in
240
H.J. Falcon-Lang, A.C. Scott / Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242
the Late Carboniferous of Euramerica. Palaeogeogr. Palaeoclimatol. Palaeoecol. 106, 39–90. Eggertsson, O., 1994. MacKenzie River driftwood — a dendrochronological study. Arctic 47, 128–136. Enright, N.L., 1995. Conifers of tropical Australasia. In: Enright, N.L., Hill, R.S. ( Eds.), Ecology of Southern Conifers. Melbourne University Press, Victoria, pp. 197–222. Falcon-Lang, H.J., 1998a. Carboniferous wildfires: their impact on vegetation, ecology and depositional environments. PhD Thesis, University of London. Falcon-Lang, H.J., 1998b. The impact of wildfire on an Early Carboniferous coastal system. Palaeogeogr. Palaeoclimatol. Palaeoecol. 139, 121–138. Falcon-Lang, H.J., 1999a. Fire ecology of a Late Carboniferous floodplain: Joggins, Nova Scotia. J. Geol. Soc. London 156, 137–148. Falcon-Lang, H.J., 1999b. The Early Carboniferous (Asbian– Brigantian) seasonal tropical climate of northern Britain. PALAIOS 14, 116–126. Falcon-Lang, H.J., 1999c. The Early Carboniferous (Courceyan–Arundian) monsoonal climate of the British Isles: evidence from growth rings in fossil woods. Geol. Mag. 136, 177–187. Falcon-Lang, H.J., 2000a. Fire ecology of the Carboniferous tropical zone. In: Scott, A.C., Moore, J., Brayshay, B. ( Eds.), Fire and the Palaeoenvironment. Palaeogeogr. Palaeoclimat. Palaeoecol. (submitted for publication). Falcon-Lang, H.J., 2000b. The relationship between leaf longevity and growth ring anatomy in modern conifers and its implications for palaeoclimatic studies. Palaeogr. Palaeoclimat. Palaeoecol. (submitted for publication). Fielding, C.R., 1984. Upper delta plain lacustrine and fluviolacustrine facies from the Westphalian of the Durham coalfield, NE England. Sedimentology 31, 547–567. Frederiksen, N.O., 1972. The rise of the Mesophytic flora. Geoscience and Man 4, 17–28. Gibling, M.R., 1987. A classic Carboniferous section: Joggins: Nova Scotia Centennial Field Guide — NE section. Geological Society of America, pp. 409–414. Gibling, M.R., 1995. Upper Palaeozoic rocks: Nova Scotia, Chapter 5: Geology of the Appalachian–Caledonian Orogen of Canada and Greenland, Williams, H. ( Ed.), Geol. Surv., Canada, Geol. Canada 6, 493–523. Gibling, M.R., Bird, D.J., 1994. Late Carboniferous cyclothems and alluvial paleovalleys in the Sydney Basin, Nova Scotia. Bull. Geol. Soc. Am. 106, 105–117. Gibling, M.R., Calder, J.H., Ryan, R., Van de Poll, H.W., Yeo, G.M., 1992. Late Carboniferous and Early Permian drainage patterns in Atlantic Canada. Can. J. Earth Sci. 29, 338–352. Golonka, J., Ross, M.I., Scotese, C.R., 1994. Phanerozoic palaeogeographic and palaeoclimatic modeling maps — Pangea. In: Embry, A.F., Beauchamp, B. ( Eds.), Global Environments and Resources, Can. Soc. Petrol. Geol. Mem. 17, 1–48. Goudie, A.S., 1983. Calcrete. In: Goudie, A.S., Pye, K. ( Eds.),
Chemical Sediments and Geomorphology. Academic Press, London, pp. 93–131. Grand’Eury, C.F., 1887. Memoire sur la flore Carbonifere du department de la Loire et du centre de la France. Mem. Acad. Sci. Inst. Nat. France 24, 1–624. Gutsell, S.L., Johnson, E.A., 1996. How fire scars are formed: coupling a disturbance process to its ecological effect. Can. J. For. Res. 26, 166–174. Hickin, E.J., 1974. The development of meanders in natural river channels. Am. J. Sci. 274, 414–442. Jeffrey, E.C., 1917. The Anatomy of Woody Plants. Chicago. Jones, T.P., 1991. The Nature, Origin and Recognition of Fusain. Unpublished PhD Thesis, University of London. Jones, T.P., Scott, A.C., Cope, M.J., 1991. Reflectance measurements and the temperature of formation of modern charcoal and implications for studies of fusain. Bull. Soc. Geol. Fr. 162, 193–200. Keighley, D.G., Pickerill, R.K., 1996. The evolution of fluvial systems in the Port Hood Formation ( Upper Carboniferous), western Cape Breton Island, eastern Canada. Sediment. Geol. 106, 97–144. Kindle, E.M., 1921. MacKenzie River driftwood. Geogr. Rev. 11, 50–53. Langbein, W.D., Schumm, S.A., chumm. Yield of sediment in relation to mean annual precipitation. Trans. Am. Geophys. Union 39, 1076–1084. Leary, R.L., 1975. Early Pennsylvanian palaeogeography of an upland area, Western Illinois, USA. Bull. Belg. Ver. Geol. 84, 19–31. Leary, R.L., 1981. Early Pennsylvanian geology and palaeobotany of the Rock Island County, Illinois Area. Part 1: Geology. Illinois State Museum 37, 88. Lindley, J., Hutton, W., 1831. In: The Fossil Flora of Great Britain vol. 1. J. Ridgeway, London. Liu, Y., Gastaldo, R.A., 1992. Characteristics and provenance of log-transported gravels in a Carboniferous channel deposit. J. Sediment. Petrol 62, 1072–1083. Lyons, P.C., Darrah, W.C., 1989. Earliest conifers of North America: upland and/or paleoclimatic indicators? PALAIOS 4, 480–486. Mapes, G., Gastaldo, R.A., 1986. Late Paleozoic non-peat accumulating floras. In: Land Plants: Notes for a Short Course, Broadhead, T.W. (Ed.), Studies in Geology 15. University of Tennessee, Tennessee, USA, pp. 115–127. Menaut, J.C., Abbadie, L., Vitousek, P.M., 1993. Nutrient and organic matter dynamics in tropical ecosystems. In: Crutzen, P.J., Goldammer, J.G. ( Eds.), Fire in the Environment: The Ecological, Atmospheric and Climatic Importance of Vegetation Fires. Wiley, New York, pp. 215–232. Meyer, G.A., Wells, S.G., 1997. Fire-related sedimentation events on alluvial fans, Yellowstone National Park, USA. J. Sediment. Res. 67, 776–791. Meyer, G.A., Wells, S.G., Balling, R.C., Timothy Jull, A.T., 1992. Response of alluvial systems to fire and climate change in Yellowstone National Park. Nature 357, 147–150. Moore, R.C., Elias, M.K., Newell, N.D., 1936. A ‘‘Permian’’
H.J. Falcon-Lang, A.C. Scott / Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242 flora from the Pennsylvanian rocks of Kansas. J. Geol. 44, 1–31. Nanson, G.C., 1980. Point bar and floodplain formation of the meandering Beatton River, northeastern British Columbia, Canada. Sedimentology 27, 3–29. Nanson, G.C., 1981. New evidence of scroll-bar formation on the Beatton River. Sedimentology 28, 889–891. Neves, R., 1958. Upper Carboniferous plant spore assemblages from the Gastrioceras subcrematum horizon, North Staffordshire. Geol. Mag. 95, 1–19. Nichols, G.J., Jones, T.P., 1992. Fusain in Carboniferous shallow marine sediments, Donegal, Ireland: the sedimentological effects of wildfire. Sedimentology 39, 487–502. Niklas, K.J., 1994. Predicting the height of fossil plant remains: an allometric approach to an old problem. Am. J. Bot. 81, 1235–1243. Penhallow, D.P., 1900. Notes on the North American Species of Dadoxylon, with special reference to type material in the collections of the Peter Redpath Museum, McGill College. Trans. R. Soc. Canada 4, 51–97. Pfefferkorn, H.W., 1980. A note on the term ‘‘upland flora’’. Rev. Palaeobot. Palynol. 30, 157–158. Plint, A.G., Browne, G.H., 1994. Tectonic event stratigraphy in a fluvio-lacustrine, strike–slip setting: the Boss Point Formation ( Westphalian A), Cumberland Basin, Maritime Canada. J. Sediment. Res. B 64, 341–364. Raymond, A., 1988. The palaeoecology of a coal-ball deposit from the middle Pennsylvanian of Iowa dominated by cordaitean gymnosperms. Rev. Palaeobot. Palynol. 53, 233–250. Rothwell, G.W., 1988. Cordaitales. In: Beck, C.B. ( Ed.), Origin and Evolution of Gymnosperms. Columbia University Press, New York, pp. 273–297. Rothwell, G.W., 1993. Cordaixylon dumusum (Cordaitales). 2. Reproductive biology, phenology and growth ecology. Int. J. Plant. Sci. 154, 572–586. Rothwell, G.W., Mapes, G., 1988. Vegetation of a Palaeozoic conifer community. In: Mapes, G., Mapes, R.H. ( Eds.), Regional geology and palaeontology of upper Paleozoic Hamilton quarry area in southeastern Kansas, Guidebook 6. Kansas Geological Survey, Lawrence, KS, pp. 213–223. Rothwell, G.W., Warner, S., 1984. Cordaixylon dumusum n. sp. (Cordaitales). I. Vegetative structures. Bot. Gaz. 145, 275–291. Rust, B.R., Gibling, M.R., 1990. Braidplain evolution in the Pennsylvanian South Bar Formation, Sydney Basin, Nova Scotia, Canada. J. Sediment. Petrol. 60, 59–72. Rust, B.R., Gibling, M.R., Legun, A.S., 1984. Coal deposition in an anastomosing-fluvial system: the Pennsylvanian Cumberland Group south of Joggins, Nova Scotia, Canada. Int. Assoc. Sedimentol. Spec. Publ. 7, 105–120. Salati, E., 1987. The forest and the hydrological cycle. In: Dickinson, R.E. ( Ed.), The Geophysiology of Amazonia. Wiley, New York, pp. 273–296. Scheihing, M.H., Pfefferkorn, H.W., 1984. The taphonomy of land plants in the Orinoco Delta: a model for the incorpora-
241
tion of plant parts in clastic sediments of Late Carboniferous age of Euramerica. Rev. Palaeobot. Palynol. 41, 205–240. Scott, A.C., 1978. Sedimentological and ecological control of Westphalian B plant assemblages from West Yorkshire. Proc. Yorks. Geol. Soc. 41, 461–508. Scott, A.C., 1979. The ecology of coal measure floras from northern Britain. Proc. Geol. Assoc. 90, 97–116. Scott, A.C., 1980. The ecology of some Upper Palaeozoic floras. In: The terrestrial environment and origin of land vertebrates. Panchen, A.L. ( Ed.), Syst. Assoc. Spec. vol. 15. Academic Press, London, pp. 87–115. Scott, A.C., 1989. Observations on the nature and origin of fusain. Int. J. Coal Geol. 12, 443–475. Scott, A.C., 1998. The legacy of Charles Lyell: advances in our knowledge of coal and coal-bearing strataLyell: The Past is the Key to the Present, Blundell, D.J., Scott, A.C. (Eds.), Geol. Soc. London Spec. Publ. 143, 243–260. Scott, A.C., Chaloner, W.G., 1983. The earliest fossil conifer from the Westphalian B of Yorkshire. Proc. R. Soc. London Series B 220, 163–182. Scott, A.C., Jones, T.P., 1994. The nature and influence of fire in Carboniferous ecosystems. Palaeogeogr., Palaeoclimatol., Palaeoecol. 106, 96–112. Scott, A.C., Galtier, J., Mapes, R.H., Mapes, G., 1997. Palaeoecological and evolutionary significance of anatomically preserved terrestrial plants in Upper Carboniferous marine goniatite bullions. J. Geol. Soc. London 154, 61–68. Smith, D.G., 1986. Anastomosing river deposits, sedimentation rates and basin subsidence, Magdalena River, northwestern Columbia, South America. Sediment. Geol. 46, 177–196. Stanistreet, I.G., Cairncross, B., McCarthy, T.S., 1993. Low sinuosity and meandering bedload rivers of the Okavango Fan: channel confinement by vegetated levees without fine sediment. Sediment. Geol. 85, 135–156. Swanson, F.J., 1981. Fire and geomorphic processes. In: Fire Regimes and Ecosystem Properties, Mooney, H.A., Bonnicksen, T.H., Christensen, N.L., Lotan, J.E., Reiners, W.A. ( Eds.), General Tech. Rep. WO-26. USDA Forest Service, pp. 401–420. Tandon, S.K., Gibling, M.R., 1994. Calcrete and coal in late Carboniferous cyclothems of Nova Scotia, Canada: climate and sea-level changes linked. Geology 22, 755–758. Trivett, M.L., Rothwell, G.W., 1985. Morphology, systematics, and paleoecology of Palaeozoic fossil plants: Mesoxylon priapi sp. nov. (Cordaitales). System. Bot. 10, 203–223. Turner, N., Spinner, E., Spode, F., Wignall, P.B., 1994. Palynostratigraphy of a Carboniferous transgressive system tract from the earliest Alportian (Namurian) of Britain. Rev. Palaeobot. Palynol. 80, 39–54. Vaughn, A., Nichols, G., 1995. Controls on the deposition of charcoal: implications for sedimentary accumulations of fusain. J. Sediment. Res. A 65, 129–135. Walker, R.G., Cant, D.J., 1984. Sandy fluvial systems. In: Facies Models: Geoscience Canada, Walker, R.G. ( Ed.), Reprint Series 1. Geological Society of Canada, Waterloo, pp. 71–91.
242
H.J. Falcon-Lang, A.C. Scott / Palaeogeography, Palaeoclimatology, Palaeoecology 156 (2000) 225–242
Walter, H., 1973. Vegetation of the Earth in Relation to Climate and Ecophysiological Conditions. 2nd edn., SpringerVerlag, Berlin. translated by J. Weiser. Way, J.H., 1968. Bed thickness analysis of some Carboniferous fluvial sedimentary rocks near Joggins, Nova Scotia. J. Sediment. Petrol. 38, 424–433. Wells, W.G., 1981. Some effects of brushfires on erosion pro-
cesses in coastal Southern California. In: Davies, T.R.H., Pearce, A.J. ( Eds.), Sediment Transport in Pacific Rim Steeplands Symposium. Int. Ass. Hydrol. Sci. R. Society, New Zealand, pp. 305–342. Winston, R.B., 1983. A Late Pennsylvanian upland flora in Kansas: systematics and environmental implications. Rev. Palaeobot. Palynol. 40, 5–31.
