Palaeogeography, Palaeoclimatology, Palaeoecology, 106 ( 1994): 9 l - 112 91

Elsevier Science B.V., Amsterdam

The nature and influence of fire in Carboniferous ecosystems

A n d r e w C. Scott a a n d T i m o t h y P. Jones a'b'l

aDepartment of Geology, Royal Holloway University of London, Egham, Surrey TW20 OEX, UK
bDepartment of Biology, Royal Holloway University of London, Egham, Surrey TW20 OEX, UK
(Received August 6, 1992; revised version accepted March 15, 1993)

ABSTRACT

Scott, A.C. and Jones, T.P., 1994. The nature and influence of fire in Carboniferous ecosystems. Palaeogeogr., Palaeoclimatol.,
Palaeoecol., 106:91-112.

Fusain occurs widely in Carboniferous coals and sediments. It is now recognised to represent charcoal and be the product
of wildfire. The occurrence of fire is partly constrained by atmospheric oxygen levels, availability and nature of fuel and by
aspects of climate (rainfall and seasonality in particular). The majority of fires in the Carboniferous were probably started by
lightning strikes or by volcanic activity. Experiments on the charring of modern plants has shown that the reflectance of
charcoal (and hence fusain) is directly related to temperature of formation. Different fire types may yield fusain assemblages
of differing reflectance spectrums, but it may be significant that many modern charcoal assemblages yield only semifusinites
(as seen by reflectance microscopy). The significance of these findings is assessed in relation to the use of fusinites and
semifusinites as depositional indicators, as interpreted from coal petrology. Fires may have a dramatic effect on ecosystems,
not only causing changes in vegetational succession but also severe erosion can occur following a major fire which can be
traced in depositional systems. In this paper we document three major Carboniferous sedimentary systems affected by fire:
clastic sedimentary systems, using extensive fusain deposits in mid-Lower Carboniferous, near-shore sediments in Donegal,
Ireland; volcanic systems using late Early Carboniferous, volcaniclastic sequences in the Midland Valley of Scotland; and coal
and coal-bearing sequences in the Upper Carboniferous (Westphalian B) of the Pennine Basin, England. In the later settings
the influence of fire in peat formation and succession is assessed. In addition, data on the vegetational composition of charcoal
assemblages is considered. It is concluded that fire plays a major role in many Carboniferous ecosystems.

Introduction fires. The association of fusain and dead animals

in Permian sediments has been recorded (Sander,
Fusain occurs commonly, and in some cases 1987). Here the assemblage is interpreted as a
abundantly, in the Carboniferous coals and sedi- single mass death event where (possibly) a wildfire
ments of Euramerica. Most of the discussion about drove the animals into a lake and they perished
fusain has concerned its nature and origin (Scott, either directly or indirectly as a result of the fire.
1989), but there has been relatively little discussion Later, rising water levels and prevailing winds
concerning the implication that fire, the primary accumulated the carcasses and charcoal together
producer of fusain (fossil charcoal), may have on the lake shore as a drift deposit.
played an important role in Carboniferous eco- Lightning is believed to be the principle source
systems (Arens, 1991). Indeed, the principle evi- of ignition of wildfires (Komarek, 1972; Cope,
dence for palaeowildfires is the fossil record of 1984); there is a fossil record of lightning strikes
fusain (Jones and Chaloner, 1991). There are, on trees which were later petrified (Wuerthner,
however, several other forms of indirect evidence 1988), and on sandy soils producing fulgarites,
that wildfires were prevalent in the past. These which are fused, branched tubes of silica or silicates
particularly relate to the evidence of the effects of (e.g. Harland and Hacker, 1966). After a wildfire
has burnt the vegetation cover, the ground is liable
1Present address: UniversitiitT/Jbingen.lnstitut ffir Geologieand Paleontologie,
Sigwartstrasse10, D-7400,Tfibingen,Germany. to enhanced erosion (Swanson, 1981; Meyer et al.,

0031-0182/94/$07.00 © 1994 - - Elsevier Science B.V. All rights reserved.

SSDI 0031-0 182(93)E0084-7
92 A.C. SCOTT AND T.P. JONES

1992). A relationship between charcoal/fusain and trees (Plate I,B). These fires may spread at different
sedimentation rates has been observed for the rates depending on the plant material. Fires in
Quaternary (Patterson et al., 1987), and for the crowns of conifers would spread faster than those
Lower Carboniferous (Nichols and Jones, 1992). in deciduous angiosperm trees, partly because of
In this paper we consider the effects that the their low moisture and volatile oil content.
fires had upon Carboniferous ecosystems rather (2) S u r f a c e f u e l s - - including loose litter on the
than just discuss the detailed comparison of forest floor, consisting of dead plant material from
modern charcoal and fusain as materials. To do the canopy as well as living herbaceous and
this, however, we have taken three complementary shrubby vegetation. The different decay rates of
approaches linking our charcoalification experi- plants make some litter more susceptible to fire
ments, studies of modern charcoals produced by than others, for example conifer needles decay less
Recent wildfires, and detailed studies of three readily than angiosperm leaves.
different occurrences of fusain in the (3) G r o u n d f u e l s - - comprising all the burnable
Carboniferous of the British Isles: Shalwy, material below the loose litter layer. This material
Donegal, Ireland; East Kirkton, Kingswood and usually supports glowing combustion but not
Pettycur, Scotland; and Thorpe and SwiUington, flames and there is a rapid fall off of temperature
Yorkshire, England. with depth (Rundell, 1981).
Three types of fire may also be recognised
Modern fires and charcoal formation (Davis, 1959; Scott, 1989): ground fire, surface fire
and crown fire as explained below.
Fires are an integral part of modern ecosystems (1) Crown fires burn aerial fuels and may
(Davies, 1959; Moore, 1978, 1982, 1989; Swanson, advance from top to top of the trees in the canopy,
1981; Plate I). They range from being regular and even if the soil surface is damp or wet.
common through to being rare, large and cata- (2) Surface fires burn the loose surface litter on
strophic (Fig. 1). The various effects of these fires the forest floor and may also advance rapidly
are diverse (Swanson, 1981). In areas where there (Plate I,A).
are regular fires the vegetation may have become (3) Ground fires may burn material below the
fire adapted (Rundel, 1981). In some cases where litter and, especially in mires, may burn the under-
fires have been less regular, or even supressed, lying peat layers. Such fires are slower and longer
burns may be more catastrophic, causing a change lasting, often remaining active for months and are
in the vegetational succession (Cypert, 1972) or characterised by smouldering with little flame and
producing a sedimentological effect, such as the may even spread underground.
formation of fire splays (Cohen et al., 1987), lakes Wildfires occur in countries all over the world,
(Cypert, 1972), or increased erosion and the filling and they have been documented and studied in
of lakes (Swanson, 1981). Major fires have been numerous different climates and environments.
recorded in various extensive modern peat-forming Wildfire is now recognised as an integral part of
areas [e.g. Borneo, East Kalimantan (Johnson, many ecosystems, with the ecosystem's evolution
1984), which may be linked to drought induced by closely tied to a history of naturally occurring fires
cyclic ocean current change, the so-called E1 Nifio (Ahlgren, 1974; Rundel, 1981; Walker, 1982;
Southern Oscillation event]. Moore, 1982; Cope and Chaloner, 1985). Some
Fires may burn in a complex manner. After areas, such as North America (Ahlgren, 1974),
ignition a fire may be controlled by a number of southern France and Australia (Gill et al., 1981),
factors including fuel type and weather. There are are prone to wildfires on a seasonal basis; other
three main types of burnable fuels produced by areas, such as Borneo (Johnson, 1984), occasion-
vegetation (Davis, 1959; Scott, 1989, fig. 12). ally suffer from large catastrophic fires. Although
These are: a significant proportion of modern fires are started
(1) A e r i a l f u e l s - - comprising material above (either intentionally or unintentionally) by man, a
1 m from the soil surface such as understory trees, sizable proportion ignite naturally; natural ignition
shrubs and lianas as well as the crowns of forest being principally by lightning, but also from sparks
THE NATURE AND INFLUENCE OF FIRE IN CARBONIFEROUS ECOSYSTEMS 93

PLATE I

Modern fires.
A. Surface fire through coniferous forest on edge of heather heath land, near Thursley, Surrey, 1991.
B. Partiallyburnt mixed conifer/Quercus woodland, north of Montpellier, France, 1989.

generated by rock falls, spontaneous combustion ignited wildfires (Wuerthner, 1988). In addition,
and volcanic activity. One study of wildfire in the heavy storms may follow the fires leading to the
North American Rocky Mountains, found that catastrophic transport and burial of sediment and
between 1940 and 1975 there were 79,131 lightning- charcoal.
94 A.C. SCOTT AND T.P. JONES

Rare Common
catastrophic ~ seasonal
fires fires

Atmosperic
13%-35% ~ 13%-35%
oxygen levels

Climate Unusual droughts, Regular (annual)

dry periods 'fire seasons'

Vegetation Catastrophic effect ~ Fire adaption

No specific plant Specific plants

VegetatiOn communities adapted ~ tending to recolonise
succession for rapid recolonisation burnt terrain
(i.e. fireweed)

Severe, major inputs Moderate, regular

Erosion into sedimentary ~ inputs into
systems sedimentary systems

Ignition Volcanic activity,

lightning ~ Lightning

Regular occurrence of
fusain throughout
Fossil record Single isolated sequence
deposit, (i.e. Donegal) (i.e. fusain lenses
in coal seams)

Modern Yellowstone
Borneo ~ Okefenokee National
analogues 1982-3 Swamp Park

Fig. 1. Wildfires a n d their effects.

A curious mode of formation, known from the swamps, and destroying 98-100% of trees in these
Recent, is sub-surface charring in peat during swamps. The Greater Yellowstone Area of North
unusually dry and warm weather (as described by America is fire-prone, experiencing an annual "fire
Boyd, 1982a,b). Wildfires often play important season", the severity of which depends upon
roles in ecosystems that one would not initially weather conditions and fuel buildup. In 1988, the
expect to be fire-prone, for example, swamps and exceptional conditions of a long hot summer, high
tundra (Racine et al., 1985). Between 1954 and winds, lightning storms and a fire suppression
1955, in exceptional drought conditions, five large policy with resultant fuel buildup, caused fires
fires burnt 318,000 acres of swamp and 140,000 which affected 1.38 million acres (Wuerthner,
acres of adjoining uplands in the Okefenokee 1988).
Swamp (Cypert, 1972). Charcoal found in soils of Studies on the ground temperatures of modern
the upper Rio Negro of Venezuela indicates that fires indicate a significant range of vegetation type
the mature rainforest has been repeatedly disturbed that is susceptible to fire over a great range of
by wildfire for the last six million years (Sandford temperatures (Rundel, 1981). In our studies of
et al., 1985). A huge fire in East Kalimantan, experimentally produced charcoal (Fig. 2a), we
Borneo (Johnson, 1984), affected an area between have speculated that most charcoals from modern
35,000 and 37,000km 2, including peat-forming fires would have reflectances more in the range of
THE NATURE AND INFLUENCE OF FIRE IN C A R B O N I F E R O U S ECOSYSTEMS 95

6- 07 ,
0
E Pinus sylvestris 0
0.6-1
cs. $ Picea abies
a ScotttJones(B. pendula) ¢= 0 a o.54
0 Cope (P. sylvesttis)
0 0.4-1
ID
E @ B SG ,
o
0.3"1
B @ e
[] @
~2- 0.2-1

1-
.,;;:'oO ° 0.1"1

0.0 I i i
I 100 200 300 4~0 5~0 6~) 700
d ,& ,~o 800 IOQO
T e m p e r a t u r e "C
T e m p e r a t u r e "C
(a) (b)

Montpelller Thursley
N o . of n=60
n = 180 readings
min = 0.547 12
rain = 0.646
max = 6.103 max = 2.711
- 30 10
mean = 2.473 mean = 1.331
S.D. =1.369 S.D. =0.599
2o

10

"I' 0% 3.0% 6.0%

0% 3.0% 6.0%
Roil (c) (d)

Yellowstone
q
m n = 360 ~ 100 200
• min = 0 04 l
• .... 5.819 4 80
I • mean= 1.696 /
.... 60
100

20

0% 3.0% 6.0% 0% 1.0% 2.0% 3.0% 4.0% 5.0% 6.0% 7.0%

Rro
(e) (f)
(250 "C) (450'C) (600 ~C) (700"C) (800"0)
Equivalent temperature of f o r m a t i o n "C.
(Grasslands) (Heath/Garrigue) (Spruce/Fir)

Fig. 2. Measurements on natural and artificially produced charcoals. (a) Mean random reflectance of artificially produced charcoals
(after Jones et al. 1991; Scott and Jones, 1991). (b) Specific gravity against temperature of formation of artificially produced
charcoals. (c) Reflectance of modern charcoals after wildfires through woodland near Montpellier, France (1989). (d) Reflectance of
modern charcoals after wildfires through mires with heather, Thursely, Surrey (1976). (e) Reflectance of modern charcoals after
wildfires through forest, Yellowstone National Park, USA (1989). (f) Composite of all data from modern charcoals showing the
equivalent temperature of formation derived from experimental work and ground temperatures recorded from modern fire types
derived from data included in Rundel (1981). For further explanation, see text.

semifusinite rather than fusinite (see also c o m m e n t s nificant transport o f charcoal would be likely to
o f Scott, 1989). H o w e v e r , it is not just reflectance separate charcoals produced at different
that changes with temperature but also specific temperatures.
gravity (Fig. 2b). We can speculate that any sig- Little w o r k has been undertaken on the sedi-
96 A.C. SCOTT AND T.P. JONES

mentation of natural charcoals. Clearly the abun- ciple assemblages of "charcoal" macroclasts: (1)
dant occurrence of particulate charcoal in deep Isolated large clasts, generally in fine-grained sedi-
marine sediments (Smith et al., 1973) may imply a ments. (2) Accumulations of solely "charcoal"
resistance to sinking and sedimentation. Skolnick clasts. (3) Mixed assemblages of "charcoalified"
(1958) briefly experimented on the floatation of and coalified clasts.
different sized particles of Recent charcoal and He suggests that the occurrence of isolated large
Cretaceous fusain from the Newcastle Sandstone, clasts could be related to charcoal's ability to
USA. Cope (1984) undertook an analysis of the remain afloat for long periods, and therefore travel
relationship between charcoal clast size and time long distances from the site of the wildfire. These
the clast would remain afloat. Using natural Pinus large clasts would eventually be deposited in sedi-
charcoal, the clasts were screened into size frac- ments such as off-shore marine or fluviatile over-
tions, 1-2, 2-4, 4-8 and 7-8 mm and 200 clasts bank deposits.
of each size fraction were placed in flasks of The accumulation of "charcoal" clasts usually
distilled water on a mechanical agitator. The consists of fragments <10mm, and represent
number of clasts sinking was recorded over a rapidly waterlogged, winnowed, traction-load
deposits, with a possible input from larger flotation
period of seven months. A relationship was noted
transported clasts. Deposits can grade from those
between clast size and time to become waterlogged
containing fusain-rich laminations to poor fusite-
for particles up to 8 mm and a predictive diagram
coals. The third assemblage of "charcoalified" and
produced.
coalified clasts receives its "charcoal" component
Skolnick (1958) suggests that special conditions
in a similar fashion to the solely "charcoal"
are required to produce fusain-rich sediments, since
assemblages.
it's initial buoyant nature would normally result
At least most, if not all, macroscopic fusain is
in wide dispersal of floating material with corres-
now thought to represent fossil charcoal (Scott,
pondent dilution of fusain in the sediments once
1989). Within coals the organic material can be
it had sunk. He proposed two mechanisms to
studied by examining polished blocks under oil.
explain the occurrence of fusain-rich beds: The inertinite maceral group includes fusinite and
(1) Rapid incorporation of fusain into sediments semifusinite, at least a proportion of which is
almost immediately upon entering a marine envi- believed to have a fire origin (Teichmfiller, 1989).
ronment, and (2) accumulation of fusain in lee In addition to these macerals, highly reflectant
areas protected from currents with subsequent material known as inertodetrinite is also recog-
heavy deposition upon saturation. nised. High temperatures of formation result in
Therefore it is postulated that conditions such the formation of highly friable or delicate charcoal,
as those associated with beach, bar, bay or and this plant material can break up into small
lagoonal environments, provided the depositing highly reflecting particles (Scott and Jones, 1991),
areas is close to the original charring source, are so that material from a fire could range from large
ideal for the entrapment of sufficient quantities to fragments with low to high reflectance, to small
result in deposits like the Newcastle Sandstone. highly reflecting particles. The proportions of each
Support for Skolnicks suggestion that fusain is will primarily depend on the type of vegetation
likely to accumulate in near-shore marine sedi- burnt and the severity of the fire. The use of such
ments is seen in the fusain-rich laminations in inertinite material in interpreting peat-forming
Lower Carboniferous sediments in Donegal, environments, taking into account the possible
Ireland. Here "charcoal" appears to have concen- long distance transportation of material, may be
trated in slack water between tidal bars in an unwise.
estuary (Nichols and Jones, 1992). The volume of In order to test these hypotheses and to provide
"charcoal" raises the interesting questions of the a comparison with fossil deposits, we have studied
source of that amount of charcoal, and the effect naturally produced charcoals from modern wild-
upon the sedimentary systems that such an input fires. We present here data from three fires and
would have had. Cope (1984) proposes three prin- comment on several others which we have also
THE NATURE AND INFLUENCE OF FIRE IN CARBONIFEROUS ECOSYSTEMS 97

investigated. Reflectance data from these fires are giving an annual natural-ignited charcoal pro-
shown on Fig. 2c, d and e. The fires occurred in: duction of 105 tons. We have made additional
(1) extensive woodland dominated by Quercus ilex observations to those of Scott (1989) who noted
and conifers near Montpellier, south France in that following wildfires is the deposition of sedi-
late 1989 (Plate I,B); (2) in peats with Sphagnum ment containing significant proportions of semi-
and heather, Erica and Calluna, at Thursley Bog, charred material. This has implications for the
southwest Surrey in mid-1976, (Plate IA; Plate types of material found in the fossil record from
II,B); and (3) in coniferous dominated woodland palaeowildfires. A wildfire's destruction of vegeta-
at Yellowstone National Park, USA, in mid-1988 tion cover can also lead to increased water run-off
(Wuerthner, 1988). We measured the random by up to 60% (Moore, 1978, 1982), and also can
reflectance under oil of charcoal samples from the enhance erosion, with sediment yields up 30-fold
litter, embedded and polished. In all three examples (Swanson, 1981).
the mean was less than 3% RRandomoil with a large The nature and appearance of the charred debris
number of readings at or below 2% Rro. When all after a wildfire depends on the nature of the pre-
the data are combined the mean is only 1.25% Rro fire environment, and the severity and duration of
which would fall within the semifusinite range of the wildfire (Despain, et al., 1989). This was also
reflectance (I.C.C.P., 1975). We have compared confirmed by our studies of char debris from
our reflectance data with temperatures derived natural fires (Plate II,B). The conifer wood fires
from our experimental results (Fig. la) and with tend to produce more and larger fragments of
the ground temperatures from modern fires charcoal, mostly consisting of wood; whereas the
(Fig. l f). We note the preponderance of readings scrubland fires have debris dominated by charred
less than 3% Rro which corresponds to temper- leaves, grasses and bark (Plate II,B). There are
atures less than 600°C, typical of many fire types. certain physical characteristics of charred material,
Indeed the mean of all readings is 1.2 Rro which when compared with non-charred material, which
corresponds to only 300°C, typical of many sur- are worthy of note. The surface of a piece of
face fires. charred wood, if deeply enough charred, has a
We have noted that in Yellowstone (Fig. 2e) the distinctive morphology of cracking (Plate II,A),
bulk of the material possesses a reflectance of often orientated radially and tangentially (Cope,
3.0% Rro or less with the exception of a small 1981; Scott, 1989). The break-up of this charcoal
proportion which show reflectances up to 6.0%. results in rectangular blocks of charcoal. Harris
This higher-reflecting material probably results (1958) measured a selection of fragments of Pinus
from the distribution of some of the plants in sylvestris charcoal and records a mean
Yellowstone National Park as opposed to the other length : breadth ratio of 3 : 1, however, when gently
locations. In forests where there are stands of shaken with water, this ratio was reduced to 1.6 : 1.
conifers growing close together, crown fires are It was noted that waterlogged twigs from the same
common. These fires race through the crown of location had a mean length : breadth ratio of 33 : 1.
the stands generating intense heat (Wuerthner, Scott (1989) produced charcoal in open fires which
1988). As is seen in Fig. 2a. there is a correlation were either allowed to cool naturally or quenched
between temperature of formation and reflectance, with water. With the naturally cooled fires, Picea
and these higher-reflecting charcoal fragments were charcoal had a mean length : breadth ratio of 1.5 : 1
probably produced in the crowns of stands of and Larix a ratio of 1.2:1; with quenched fire
conifers. Larix gave a mean ratio of 1.1 : 1. Together with
The aftermath of all these Recent wildfires is the reflectance data the size gives a useful compari-
huge quantities of charred material, charcoal and son with the fossil material.
ash. Smith et al. (1973) estimated that 18x 106
acres of forest burn each year producing 3 x 105 Carboniferous case histories
tons of charcoal. Cope (1984) extrapolating from
Smith's figures, suggests that approximately one In order to illustrate the effects of fire on
third of fires, on a global basis, ignite naturally, Carboniferous ecosystems we give three contrast-
98 A.C. SCOTT AND T.P. JONES

PLATE II

Modern charcoal.
A. Partiallycharred log showingcharacteristiccracking(after Scott, 1989) ( × ½).
B. Charred litter from heather heathland/Sphagnum peat fire, Thursley,Surrey, 1976 ( x 1).

ing examples; the "catastrophic" Early Donegal

Carboniferous fires in Donegal, the "volcanically"
ignited fires of the Lower Carboniferous of the A transgressive Lower Carboniferous (Visean)
Midland Valley of Scotland, and the fires associ- sequence of sediments overlies the metamorphic
ated with peat-forming mires of the Upper Dalradian in the Donegal syncline of northwest
Carboniferous of Euramerica. Ireland (George and Oswald, 1957; Fig. 3a, b).
T H E N A T U R E A N D I N F L U E N C E O F F I R E IN C A R B O N I F E R O U S E C O S Y S T E M S 99

. . . . . . . . . . . i, . . . .

rr BEDS

irn~61onesand grey shale=

:usaln
mi
Rinn Polttt Beds
Shak~y Beds / BtucSdessur~l
Mstareorphics
(a)
ALWY BEDS

o~s-bedded sandstones

~.-- CroM -I:,edd,nO

IALWY BESS m I=olaledfu~lin k ~ n e n t =
=~ldmOne= Me~u¢~ sandl~tone
R~e und=lccle
Very ine =aedstone
NOLOMERATES
)mor=tes
Iroc~one
RIAN SCHISTS LImelllorm wilh lusain
(b) Verl~al ~ l e kl metres (C)

LSh~TmY
Pr~lnt Fusain unit

I(Mc

i
30 ' Equivalent area
of wildfire and
'Shalwy River Syste
20

10 500 km
I I
Land
0.0% 1.0% 2.0% 3.0% 4.0% 5.0% 6.0% 7.0%
% Random reflectance under 011 (e) Sea

Fig. 3. The occurrence of fusain horizons in the Lower Carboniferous of Donegal, Ireland. (a) Ireland (inset) showing a geological
map of the Largymore syncline (after Nichols and Jones, 1992). (b) The stratigraphy of the Lower Carboniferous of the Largymore
syncline (after Nichols and Jones 1992). (c) Detail of the Upper Shalwy Beds showing the fusain unit. (d) Field sketch of the fusain
unit. (e) Random reflectance of fusain from the fusain unit. (f) Proposed area of wildfire to produce charcoal found in the fusain
unit (after Nichols and Jones, 1992).
100 A.C. SCOTT AND T.P. JONES

The Upper Shalwy Beds at Shalwy Point, include Muckros Head, which are believed to be part of
a lense-shaped bedform (Fig. 3d; Plate III,A), "the the same laterally extensive formation (Nichols
fusain unit", with fusain-rich laminations (Plate and Jones, 1992). The unit contains abundant
III,B; Plate IV,A; Scott and Collinson, 1978). Two fusain (Fig. 3c; Plate IV), angular quartz and shell
new sites have been located, at Rinn Point and fragments in a muddy limestone matrix, and is less

PLATE III

Fusain unit, Lower Carboniferous, Donegal, Ireland.

A. The Upper Shalwy Beds at ShalwyPoint, Donegal showing the position of fusain unit (arrowed).
B. Banding within the fusain unit.
T H E N A T U R E A N D I N F L U E N C E O F F I R E IN C A R B O N I F E R O U S E C O S Y S T E M S 101

PLATE IV

Fusain unit, Lower Carboniferous, Donegal, Ireland.

A. Layersof fusain weathered in cliff face.
B. Beddingsurface of fusain layer. Small brachiopod shells also occur.

well sorted than the sandstone units above and dominantly to putative gymnospermous (as yet
below it (Nichols and Jones, 1992). The fusain unknown) plants, and as small cylindrical steles
unit r~nges in thickness from 1.10 to 2.45 m and (Scott and Collinson, 1978) from the pseudo-
contains up to 20% fusain (Plate IV,A). The fusain herbaceous lycopod Oxroadia (Plate V).
appears to be concentrated into distinctive layers Reflectance measurements made on fusain
but is believed to have been deposited by a single (Fig. 3e) from within the unit ranged from 1.5 to
event. The fusain occurs as typical wedge or cube- 6.5% Rro. The absence of reflectances below 1.5%
shaped fragments (Plate IV,B), which belong pre- Rro is explained by the degree of maturation that
102 A.C. S C O T T A N D T.P. J O N E S

PLATE V

Fusain from the fusain unit, Lower Carboniferous, Donegal, Ireland.

A. Detail of fusain horizon showing isolated fusainized (charcoalified) steles of the lycopod Oxroadia with cubes of gymnosperm
wood (after Scott, 1989) (x 1).
B. Scanning electron micrograph of Oxroadia stele ( x 40).
C. Detail of Oxroadia stele showing characteristic tracheids ( x 400).
D. Reflectancemicrograph of fusainized Oxroadia stele (after Scott, 1989) ( x 100).

the formation has undergone. Vitrinite reflectance measured were isotopically anomalously heavy
readings taken on compression fossils near the 613C = - 16.4 to - 19.3%, when compared to sim-
formation give a m a x i m u m of 1.8% Rro (Nichols ilar Lower Carboniferous assemblages from other
and Jones, 1992), and from the same formation locations in the British Isles (613C = - 2 5 % ) . It is
Rro = 1.95% (Clayton et al., 1989); this is equivalent hypothesized that the anomalously isotopically
to a thermal alteration index (TAI) of 3. With heavy Donegal material is the end product of an,
increased maturation the reflectance of fusain frag- as yet, undetermined isotopic fractionation process
ments increases so as to approximatly correlate associated with maturation or diagenesis.
with the corresponding vitrinite reflectance (Alpern The shapes and sizes of fusain fragments vary,
and De Sousa, 1970). The inference that is drawn depending on the plant tissues involved. Secondary
from the proportion of high reflecting fusain in gymnosperm wood often occurs as cubes (Plate
the "fusain unit" assemblage is that a part of the IV,B) but the lycopod steles are often long and
original Carboniferous fire must have been of thin, isolated from other tissue (Plate V,A).
su ,fficient intensity (temperature) to produce a pro- Nichols and Jones (1992) interpret the deposi-
portion of high reflecting charcoal. tional environment of the Upper Shalwy Beds as
Jones (1991) measured the carbon stable isotope representing tidally deposited sands and muds.
values of samples from the fusain unit at Shawly Strong currents were active, as indicated by the
Point, and Jones et al. (1993) measured the carbon nature of the cross bedding, which is sand-wave
stable isotopes of a fusain-vitrinite-fusain trans- scale. These currents, however, did not disperse
ition fossil from Muckros Head. All the samples the charcoal (fusain) which was probably kept
T H E N A T U R E A N D I N F L U E N C E O F F I R E 1N C A R B O N I F E R O U S E C O S Y S T E M S I03

within the estuary by reverse tidal flow. The sedi- of gymnospermous wood, through to small
mentation of the charcoal and sand together with (1-5mm) fusainized pteridosperm leaves and
the effects of wind led to the formation of reverse pollen organs (Plate VI). It is clear that, although
graded units, as observed in the field. Fifty-nine this plant material has been charred and trans-
discrete coarse-fine couplets can be observed at ported prior to deposition, this transport may not
Shalwy Point: Nichols and Jones (1992) believe have been far. It is noteworthy, however, that the
that these could have represented a period of only uncharred flora is quite different from the
30 days. Hence the complete fusain unit may be fusainized assemblage and includes a new as yet
the result of a single fire. undescribed gymnosperm and the pseudo-
A conservative estimate of between 5 and 10 km 2 herbaceous lycopod Oxroadia.
has been calculated for the extent of the fusain In contrast, the fusain in the Pettycur sequence
unit. Taking a conservative mean of 1.00 m thick- nearby is quite different. Plants occur within
ness and 10% fusain, it is possible to calculate reworked permineralized peat within the lava, as
that the volume of fusain in the Upper Shalwy well as isolated permineralized specimens within
beds is 1,000,000 m 3. It has been estimated (Cope the associated ashes. For example, one peat type
and Chaloner, 1985) that Recent fires give a char- (reworked within the lava) comprises mainly the
coal production of 4,150 kg/km 2 burnt. Taking an ferns Diplolabis and Metaclepsydropsis (Rex and
average bulk density of 0.4g/cm 3 (Cope, 1984; Scott, 1987). In all cases fusainized plants are of
Jones, 1991), Nichols and Jones (1992) calculated the same taxa as normally permineralized plants.
that the average volume of charcoal generated per The fusain may be scattered, in layers, or some
square kilometer burnt was 10.4 m3/km z. On that specimens may also show partial fusainization
basis, these authors calculated the burned area (Plate VI,A). Some of this charring may have been
that generated the Shalwy charcoal as 95,000 km 2. a direct result of contact with the hot lava or ash,
This result is about three times the size of the although studies by Jones (1991) on charcoals in
Great Fire of Borneo (1982/1983) (Johnson, 1984) Recent pyroclastic flows show only low reflec-
tances. Scott and Galtier (1985) believed that such
and approximately equivalent to the present size
of Ireland (Fig. 3f). fires may have had an effect on the evolution of
ferns, and Scott (1990) and Bateman (1991) com-
It is probable that the effects of the Shalwy
mented on its effect on the evolution of some
wildfire would have not lasted long (maybe only a
gymnosperms.
few years) with vegetation quickly re-establishing
We have studied fusain quantitatively from the
and rejuvenating the ecosystem (Ahlgren, 1974).
important site of East Kirkton (Rolfe et al., 1990),
The increased sediment supply to the estuary
which has not only yielded the oldest known
would also have ceased and quickly returned to its
reptile, four amphibian groups, harvestman spiders
pre-fire levels.
and terrestrial scorpions, but also several new
plants (e.g. Galtier and Scott, 1991). The plants in
East Kirkton, Kingswood and Pettycur the sequence, which has been interpreted as formed
by a volcanic hot spring, are preserved as compres-
Fusain is associated commonly with Lower sions, permineralizations and as fusain (Scott, in
Carboniferous plant-bearing volcanogenic Rolfe et al., 1990; Brown et al., 1993). Fusain
sequences in Scotland (Scott, 1990; Fig. 4a, b). It occurs at several levels and consists of mostly
has already been observed that distinct assem- cube-like fragments of gymnospermous wood
blages of plants may be preferentially preserved as (Plate VII,A,B,C) as well as of rare fusainized
fusain, such as at Kingswood, near Pettycur, Fife lycopod leaves at one level.
(Scott et al., 1986; Rex and Scott 1987; Scott, Rarely does the fusain occur in sufficient abun-
1990). dance to constitute a single layer. Usually large
At Kingswood, fusain occurs in a wide variety isolated fragments (> 2 cm) are found isolated in
of sizes, from wood more than 10 cm long and the ash or limestone (Plate VII). Furthermore, in
4 cm wide to 1-2 cm cube-shaped fragments, also some cases partially charred fragments are encoun-
104 A.C. SCOTT A N D T.P. JONES

PLATE VI

Fusain from the Lower Carboniferous of Scotland (scale = 1 cm).

A. Peal of zygopterid peat from lava sequence, Pettycur, showing layers (large arrows) of fusainized ferns and scattered fusainized
fern fragments (small arrows) in permineralized peat.
B. Fusain from the Kingswood Limestone, Pettycur, (arrowed), showing range of sizes.
C. Thin section of Kingswood Limestone, Pettycur, showing large piece of fusainized wood and pteridosperm leaf fragments
together with unfusainized permineralized lycopod fragments (p).
D. Cut block of ashy Kingswood Limestone with abundant small fusain fragments mainly of pteridosperms.
 THE NATURE AND INFLUENCE OF FIRE IN CARBONIFEROUS ECOSYSTEMS 105

PLATE VII

Fusain from the Lower Carboniferous of East Kirkton, Bathgate, Scotland.

A. Fusainized wood fragments (f) and permineralized wood (p) in ash, Bed 88 (scale 1 cm).
B: Large fragment of fusainized wood is ash, Bed 71 (scale 1 cm).
C. Interlayered ash and laminated limestone with fusain fragment (arrowed), Bed 69 ( x ¼).
D. Section in light microscopy showing opaque cell walls of fusainized gymnosperm wood. Bed 59 ( x 100).
E. Scanning electron micrograph of D (scale 200 Nm).
F. Scanning electron micrograph of pitting in gymnosperm wood, Bed 69 (scale 100 Nm).

tered in p e r m i n e r a l i z a t i o n s o r v i t r a i n - f u s a i n trans- Section (Fig. 4c). T h e cellular spaces o f the fusain

ition fossils (Jones et al., 1993). L a r g e scattered are infilled with chlorite, a m i n e r a l which can be
fusain f r a g m e n t s are a b u n d a n t in the w a t e r - l a i n a s s o c i a t e d with l o w - t e m p e r a t u r e h y d r o t h e r m a l sys-
volcanic ashes a n d f r e s h w a t e r limestones f o u n d in tems ( D e e r et al., 1992). M e a s u r e m e n t s o f the
the lower p a r t o f the p r e s e n t E a s t K i r k t o n Project fusain f r a g m e n t s (Fig. 4d) indicate a m e a n
106 A.C. S C O T T A N D T.P. J O N E S

200 (Nat. Gi'id) 250 300 350 400 Oed Lithology

I If ~I 1o . . _ . ~ / / ~r / ~. I numbers log H i s t o g r a m of fusain distribution
• 700 32
Key: 1

,650
bll
e o-...... i
/ _ / t~///////////'A" I ~. I

42

57
m
BRIGANTLAN NC LOWERGROUP LOWER LIME- 64
LIMESTONE STONE GROUP
VF UPPER,.~EDIMENTAR¥ UPPER OIL v
.Q~#,.~q, ? . GROUP
?L SHALE V
NM ,.' ~ v v "./"~"~.*
ASBIAN • , ~ v v v v v v ' LOWER OIL
,CLYDE PLATEAU
~ow~go,L ~'~
TC
¢ V "rL#')'JA~)V V ' GROUP
HOLNERIAN TS v v v v v v ' i
ARUNDIAN PU ~ V V V V V '

2m
cvv ~v,

CEMENTSTONE 81
COURCEYAN CM GROUP GROUP

(c)
(b)

4O

J Fusain
mean ratio 2:1
30 m

Fusain-vitrinite
E transition fossils
E mean ratio 4:1
.E
20
{,~'-~;
,. a ; % R a n d o m r e f l e c t a n c e u n d e r oil
(e)

jZI~W~ --Bin fossils

[] []

0 []
0 20 40 60 80 100
Length in mm

(d)
Fig. 4. The occurrence of fusain horizons in the Lower Carboniferous of East Kirkton, Scotland. (a) Geological map of Scotland
and its borders showing East Kirkton and other localities mentioned in the text. (b) The stratigraphic position of Scottish Lower
Carboniferous localities mentioned in the text. (c) The distribution of fusain at the East Kirkton site. (d) Length/breadth
measurements of fusain and fusain transition fossils• (e) Random reflectance of fusain fossils.

length:breadth ratio o f 2: 1. In contrast, There is no evidence in the East Kirkton section

fusain-vitrain transition fossils have a m e a n ratio o f continuous fusain bands, as seen at Shalwy.
o f 4 : 1 . Reflectance m e a s u r e m e n t s s h o w a broad H o w e v e r , the regular occurrence o f large fusain
spread o f readings (Fig. 4e), f r o m less than 0.5% fragments in the lower part o f the section m a y be
to 5.0% R,o. o f s o m e significance. M o s t o f the fusain fragments
THE NATURE AND INFLUENCE OF FIRE IN CARBONIFEROUS ECOSYSTEMS 107

appear relatively angular and we do not envisage varied effect (e.g. Cypert, 1972; Komarek, 1972;
extensive transport. The occurrence of an extensive Cohen, 1974; Cohen et al., 1987). In some cases
and articulated animal assemblage within the lime- the burn may be quick and the vegetation may re-
stones, allows us to speculate that fires may have establish. In other cases a more severe burn may
driven the animals towards the hot spring and lead to a change in vegetation type or in some
ultimately to their death. To test this hypothesis, cases affect the sedimentological regime (e.g.
however, would be difficult, if not impossible. Cypert, 1972; Cohen et al., 1987). Charcoal may
then be incorporated into the peat (Cohen et al.,
Swillington and Thorpe 1978, 1987). Occasionally, discrete fire horizons
may be recognized with a significant change in the
Fusain is common both in coals and clastic overlying peat or a change to clastic sedimentation
sediments in the Upper Carbonifeous Coal (Cohen et al., 1987).
Measures of Yorkshire (Scott, 1978, 1979). Fusain Vegetational and sedimentological changes have
occurs as scattered fragments in clastic rocks. been observed with relation to Coal Measure
Many different plant organs have been reported fusain bands in coals from Thorpe Brickworks,
as fusain including the earliest known conifer near Leeds, Yorkshire (Scott, 1978). It has been
leaves (see Scott, 1989, for a review). Material noted that a thick fusain band occurs at the top
from below the Lidget coal at Swillington of the. 14 Yard Coal and this is immediately
Brickworks, near Leeds, Yorkshire (Scott, 1978) followed by a lacustrine sequence (Scott, 1979).
has been studied. Reflectance values of isolated Vegetational change through the Top Haigh Moor
fusain fragments ranged from 0.5 to 2.8 Rfo Coal was investigated palynologically (Scott,
(Fig. 5), the majority being semi-fusinite. Studies 1978). Figure 6a shows lycopod megaspore distri-
on the identity of the fusain indicate a wide bution through the top of the coal. A prominent
spectrum of plants, from pteridosperm pinnules, fusain band occurs near the top of the coal. A
cordiate and conifer leaves, to gymnospermous marked change in plants occurs above the band
wood and even calamite stems. The predominance which may either reflect a changing environment
of this material, however, is from plants living or a fire-initiated vegetational change. Below
outside of peat-forming mires and may represent the band spores belonging to Sigillaria
remnants of vegetation from extrabasinal or ( Tubereulatisporites and Laevigatisporites),
upland fires. Porostrobus (Setosisporites) and Chaloneria
In contrast, fusain bands in coals may represent (Valvisisporites) are most common. Above the
remnants of fires from within the peat-forming band there is a small representation of spores from
system. In modern wetland settings, such as the Paralycopodites and then a dominance of spores
Everglades and the Okefenokee, fires may have a probably belonging to Sporangiostrobus. It is clear
from other studies, however, that vegetational
Swillington changes do not occur above every fusain horizon
(Bartram, 1987), nor is there any theoretical reason
min = 0.58 why this should be so.
n=50
max = 2.59 I 810 A vegetational change associated with abundant
mean = 1.32 fusain has been noted by others in Upper
S.D. = 0.46 6
Carboniferous coals of Euramerica. Using data
4 from coal balls, Phillips and DiMichele (1981)
2 traced the vegetational changes through the Herrin
No. 6 Coal. The dominant plants were lycopods.
0.0% 1.0% 2.0% 3.0% 4.0% 5.0% DiMichele and Phillips (1985) have demonstrated
% Random reflectance under 0il the ecology of these lycopods indicating
Fig. 5. Random reflectance of fusain from Bed 20, Upper Lepidophloios living in the wettest conditions
Carboniferous of Swillington, Yorkshire. though to Lepidodendron s.l. (Diaphodendron) and
I08 A.C. SCOTTAND T.P. JONES

Sigillaria. More detail of this unit from other

sections indicated that there were three coal ball
zones located between the paired clastic bands
with Diaphodendron and Lepidophloios, then abun-
dant Paralycopodites and then mostly Sigillaria.
The interpretation of this data is equivocal. Either
lacustrine
there has been a drying of the mire and
mudstones Paralycopodites and Sigillaria have colonized the
area, in which case fires may have been a regular
part of the ecosystem at this interval or else a dry
coal phase was followed by fires and different plants
colonized the mire. It is possible that both
Fusain band ......................... Paralycopodites and Sigillaria favoured disturbed
sites. Both of these possibilities imply that the
coal cl°mI fusain is from fires within the mire. Alternatively,
I the fusain may have been washed into the mire
1000 (a)
from outside and not be responsible for, or con-
TOP nected with, the vegetation within the zone.
Taxonomic identification of the charcoalified
plants may help resolve this question, although we
do not know the relative charcoal production
capacity of each of the plant groups when burnt.
Some pteridosperms, often thought to be living in
zoneFUsaln-I'ich [Paraly~ ~Sigillaria>~ .~ • iv%
drier sites, are often found predominantly as fusain
shale partings .-~
ro"MT-;~/• V • \/• V •\ .... (Phillips, 1981) again suggesting a regular occur-
rence of fires in the Euramerican coal sequences.

Discussion and conclusions

lnT • L.sclerotlcum • • • •
cml • • • • • • • • We have indicated that fires may range from
being commonly seasonal to rare and catastrophic
BASE ~'~ unidentfedAZk~ ~:~::~.`~.~::~;~i~q~:~:~)~:~::~:~:~;:;~;;;:.~?:~q~q:~
"..'i":.~":",.'7..':.";~".'"::".:.?:"-.:):.';'::.';~'::.i':
'v;~':
(Fig. 1). Climate may also play a significant role
Lycopod Blomass 1oo0/.
(b) in controlling the frequency of fires. It is possible
Fig. 6. (a) The changing megaspore flora across a fusain band
that in addition to large-scale climatic changes,
in the upper part of the Top Haigh Moor Coal, Upper smaller-scale changes in climate may play a signifi-
Carboniferous, Thorpe, Yorkshire (data after Scott, 1978). (b) cant role in the occurrence of fires. It is possible
Lycopod biomass obtained from coal balls in the lower part that in some everwet environments that fires are
of the Herren No 6 Coal, Illinois, showing distinctive flora
associated with fusain-rich zone (after DiMichelle and
unusual. Disturbance in the normal climatic regime
Phillips, 1981). (e.g. El Nifio events) may cause temporary changes
in the rainfall pattern, for example, leading to a
dry period when a catastrophic fire may occur.
Paralycopodites and Sigillaria living in the driest This may be the reason for the Great Fire of
environments. It is probable that Sigillaria also Borneo in 1982-1983. Small fluctuations in rainfall
lived in disturbed sites. The fusain-rich zone in the pattern may allow frequent small fires within a
Herren No. 6 coal in Illinois is bound by shale mire-forming environment. Storms may follow a
partings within the coal (Fig. 6b). The flora within major fire and the increased erosion may eventually
this zone is significantly different from that in the give rise to a sediment/charcoal mix being trans-
rest of the coal, comprising Paralycopodites and ported and deposited in several sites. The occur-
THE NATURE AND INFLUENCE OF FIRE IN CARBONIFEROUS ECOSYSTEMS 109

rence of fusain layers in the fossil record may be coals. Much of this is likely to result from regular
used to distinguish between events of different fires within the mire itself, as seen in the Everglades
time scales. or in the Okefenokee swamps, where vegetational
Within fossil deposits the distinction between change can be demonstrated. It appears that in
regular seasonal fires and more catastrophic fires some ancient cases (as least the few that have been
is only possible by examining the occurrences of so far studied) these fires do not always affect
fusain within a sequence. It may be possible, vegetational succession but in some examples not
however, using a mixture of taxonomic informa- only is there evidence of vegetational change but
tion from the fusain and its reflectance, to be able also of more severe effects upon the sedimentology.
to make additional interpretations. Within a mire Whilst some evidence suggests local fires, data on
regular ground or surface fires should yield char- the taxonomic composition of fusain may suggest
coal of a different taxonomic mix and reflectance more distant fires with greater transport of the
than major crown fires. Material introduced from resulting charcoal.
outside the mire-forming environment may also be Both in the Carboniferous and Recent, fusain
identified. and charcoal are associated with volcanogenic
The use of inertinite macerals within coal seams sediments and lavas. Fusain occurs both sporadi-
as a palaeoenvironmental indicator must be viewed cally as well as in bands in many volcanic ashes
with caution. The idea that charcoal is equivalent and associated sediments. Whilst in some cases
only to high-reflectance fusinite clearly is not cor- local vegetation may have been burnt, evidence of
rect. Charred plant material may yield a variety of taxonomic affinity, as in the case of Kingswood,
coal macerals from fusinite to semifusinite, depend- suggests more distant fires of different vegetation
ing on the nature of the fire. In addition, charcoal type. Such situations have modern analogues in
produced at very high temperatures tends to frag- North Island, New Zealand (see Jones, 1991, for
ment (Jones and Chaloner, 1991; Scott and Jones, details). We note, however, that the charcoal found
1991) and would therefore yield small highly in some pyroclastic flows has a relatively low
reflecting particles of inertodetrinite. Some of this reflectance and may not be completely comparable
material may represent the result of local fires but, to the material from East Kirkton (Fig. 7).
equally likely, other material may be transported The effects of large-scale fires are illustrated by
considerable distances. Inertinites should not, the Lower Carboniferous marginal marine fusain
therefore, be used to interpret dry conditions with- deposits at Shalwy Point, Donegal. Here, there is
out supporting evidence. no doubt that the fire was catastrophic. The wide-
The transportability of charcoal fragments is spread nature of the fire not only yielded significant
supported by the fact that small (1-2 lam) frag- amounts of charcoal but also would have led
ments are common in modern deep oceanic sedi- directly to increased soil erosion which in turn
ments (Smith et al., 1973; Herring, 1985). Data on would have led to a major increase in bed load in
the specific gravity of charcoals produced at the alluvial system. A modern analogue may be
different temperatures (Fig. 2b) suggests that their the great fire of Borneo (Johnson, 1984).
transport and depositional behaviors may differ. While fires may be regular and even seasonal at
We need, therefore, more extensive data on the certain locations, the intervention of man may
nature of charcoal transport. We are able, however, cause additional hazards. It is shown that the
to take the data further and consider several quite extensive fires at Yellowstone in 1988 may have
different deposits and cite modern analogues. We been made worse by the fire control measures
have contrasted the data from our three case employed in the National Park prior to 1988
studies in Fig. 7. We believe that these three exam- (Wuerthner, 1988). Equally, however, climatic
ples are typical of the range encountered in fluctuations, may also lead to changes in fire type
Carboniferous sequences. and severity. Meyer et al. (1992) have demon-
The most commonly encountered fusain occurs strated the effect of fire on alluvial systems in
scattered or in bands in Upper Carboniferous Yellowstone National Park over the past 3500
ll0 A.C. SCOTT AND T.P. JONES

Shalwv Point, East Kirkton, Swillin2ton.

Age. Lower Carboniferous. LowerCarboniferous. UpperCarboniferous.
Lithoiogy. Limestone and poor Water-lainvolcanic Coals and shales.
fusite coals, ash and freshwater
limestones.

Fusain size. Up to 15ram. 5ram to 40ram Up to 10ram.

Deposition Estuary tidal sands Freshwaterlakes Mire and floodplain.
environment, and mudstones, in volcanicprovince.
Fusain Concentrated in Scattered throughout Bandsin coals
distribution, one bedform, bottom of sequence, scatteredin shales.
Fusain Proximal to Proximal. Proximal.
transportation, moderate.

Probablecause Catastrophic Volcanic activity Seasonal wildfires,

ignition / charring, wildfire, related, lightning ignited.
Vegetation. Lycopods and Lyeopods and Lyeopods dominating
putative gynmosperms, putativegymnosperms, in mires, gymnosperms
dominate elsewhere.
- Rt,~. 1.67% - 6.23%. 0.70% - 4.69%. 1.00% - 5.0%.

Modern Great fire of Borneo. NorthIsland, New Everglades.

analogues. Zealand.
Fig. 7. Comparison of the fusain occurrences at Shalwy Point, Donegal, East Kirkton, Scotland and Swillington, Yorkshire with
their interpretation and modern analogues.

years. These authors note that the dominant allu- concentrations may allow fires to occur in wetter
vial activity is strongly modulated by climate, with environments and reduced oxygen concentrations
fire acting as a drought-actuated catalyst for sedi- m a y suppress fires (Watson et al., 1978). It is
ment transport. In their paper Meyer et al. (1992) equally likely that the temperatures reached, and
note the abundance of charcoal debris in some hence also reflectance of the charcoal, may alter sig-
facies. They also note the occurrence of eighteen nificantly with increased oxygen. This variable is yet
fire-related debris-flow events and seventeen prob- one more to consider in assessing the occurrence
able fire-related sedimentation events spanning the and significance of wildfires in the fossil record.
past 3500 years. We believe that this type of study F r o m our studies we believe that more attention
is of major significance to understanding the geolo- needs to be paid to both charcoal and fusain
gical importance of fusain. Detailed sedimentologi- taphonomy. We may argue at length about the
cal studies such as those of Nichols and Jones detailed origin of Carboniferous fusains but it is
(1992) on the Carboniferous and that of Meyer perhaps more important to examine the conse-
et al. (1992) on more Recent deposits must surely quences of our interpretations. If, as we believe,
elucidate the relationships between climate, climate most fusain represents charcoal produced by wild-
change and the evolution of terrestrial landscapes fire then we must begin to look at the effects of
and sedimentary patterns. these fires on Carboniferous ecosystems using
In our discussions we have not considered the diverse data from the fossil record. We tend to
likely effect of changing oxygen levels in the atmo- imagine a tropical Euramerican Carboniferous
sphere. Cope and Chaloner (1980) outline the world as being wet and stable, but the evidence
range of oxygen content which must be available we present here may suggest a much more varied
for fires to occur. However, increased oxygen and dynamic system.
THE NATURE A N D I N F L U E N C E OF FIRE IN C A R B O N I F E R O U S ECOSYSTEMS 111

Acknowledgements Cope, MJ., 1981. Products of natural burning as a component

of the dispersed organic matter of sedimentary rocks. In:
J. Brooks (Editor), Organic Maturation Studies and Fossil
We thank Neil Holloway for preparing polished Fuel Exploration. Academic Press, London, pp. 89-109.
blocks, Kevin D'Souza for photography and Cope, M.J. 1984. Some studies of the origin, nature and
Sandra Muir for drafting the diagrams. We thank occurrence of charcoalified plant fossils. Unp. Ph.D. Thesis,
Univ. London, 407 pp.
W.D. Ian Rolfe and Margaret E. Collinson for
Cope, M.J. and Chaloner, W.G., 1980. Fossil charcoal as
providing material from Yellowstone, M.E. evidence of past atmospheric composition. Nature, 283:
Collinson and W.G. Chaloner for their help in 647-649.
collecting material from Thursely and G. Nichols Cope, M.J. and Chaloner, W.G., 1985. Wildfire, an interaction
of biological and physical processes. In: B.H. Tiffney
for help in Donegal. We thank J. Calder, A. Cohen
(Editor), Geological Factors and the Evolution of Plants.
and T. Phillips for constructive comments on the Yale University Press, Hartford. CT, pp. 257-277.
manuscript. T.J. acknowledges the receipt of a Cypert, E., 1972. Plant succession on burned areas in
NERC studentship and ACS NATO grants which Okefenokee swamp following the fires of 1954 and 1955.
permitted fieldwork in France. Proc. Tall Timbers Fire Ecology Conf., 12: 199-217.
Davis, K.P., 1959. Forest Fire - - Control and Use. McGraw-
Hill, New York, NY, 584 pp.
References Deer, W.A., Howie, R.A. and Zussman, J.. 1992. An
Introduction to the Rock Forming Minerals. Longman
Ahlgren, C.E., 1974. Introduction. In: T.T. Kozolwski and Group, London, 2nd ed.
C.E. Ahlgren (Editors), Fire and Ecosystems. Academic Despain, D., Rodman, A., Schullery, P. and Shovic., 1989.
Press, New York, NY, pp. 1 5. Burned area survey of Yellowstone National Park: The fires
Alpern, B. and De Sousa, L., 1970. Sur le pouvoir reflecteur of 1988. Div. Res. Geogr. Inf. Syst. lab., Yellowstone
de la vitrinite et de la fusinite des houilles. C. R. Acad. Sci. National Park, USA.
Paris, 271: 956-959. DiMichele, W.A. and Phillips, T.L., 1985. Arborescent lycopod
Arens, N.C., 1991. Vegetation replacement following wildfire reproduction and paleocology in a coal-swamp environment
disturbance in Middle Pennsylvanian clastic wetlands: exam- of late Middle Pennsynvanian age (Herrin Coal, Illinois).
ples from Joggins, Nova Scotia, Canada. Geol. Soc. Am., Rev. Palaeobot. Palynol., 44:1 26.
Abstr. Progr., p. A456. Galtier, J. and Scott, A.C., 1991. Stanwoodia, a new genus of
Bartram, K.M., 1987. Lycopod succession in coals: an example
probable early gymnosperms from the Dinantian of East
from the Low Barnsely Seam (Westphalian B), Yorkshire,
Kirkton, Scotland. Trans. R. Soc. Edinburgh, Earth Sci.,
England. In: A.C. Scott (Editor), Coal and Coal-bearing
82: 113-123.
strata: Recent Advances. Geol. Soc. London, Spec. Publ.,
George, T.N. and Oswald, D.H., 1957. The Carboniferous
32: pp. 187-199.
rocks of the Donegal Syncline. Q. J. Geol. Soc. London,
Bateman, R., 1991. Palaeoecology. In: C.J. Cleai (Editor),
113: 137-179.
Plant fossils in Geological Investigation: The Palaeozoic.
Gill, A.M., Groves, R.H. and Noble, I.R. (Editor), 1981. Fire
Ellis Horwood, London, pp. 34-116.
and the Australian Biota. Australian Acad. Sci.
Boyd, W.E., 1982a. Subsurface formation of charcoal and its
Harland, W.B. and Hacker, J.L., 1966. "Fossil" lightning
possible relevance to the interpretation of charcoal remains
in peat. Quat. Newsl., 37: 6-8. strikes 250 million years ago. Adv. Sci., 663-671.
Boy& W.E. 1982b. Subsurface formation of charcoal an unex- Harris, T.M., 1958. Forest fire in the Mesozoic. J. Ecol.,
plained event in peat. Quat. Newsl., 38: 15-16. 46: 447-453.
Brown, R.E., Scott, A.C. and Jones, T.P., 1993. Taphonomy Herring, J.R., 1985. Charcoal fluxes into sediments of the
of plant fossils from the Vis6an of East Kirkton, West North Pacific Ocean: the Cenozoic record of burning. In:
Lothian, Scotland. Trans. Soc. Edinburgh: Earth Sci., 84: The Carbon Cycle and Atmospheric CO2: Natural Variations
267-274. Archean to Present. Geophys. Monogr., 32: 419-442.
Clayton, G., Haughey, N., Sevastopulo, G.D. and Burnett, R., International Committee for Coal Petrology, (I.C.C.P.), 1975.
1989. Thermal maturation levels in the Devonian and International Handbook of Coal Petrology. C.N.R.S. Paris,
Carboniferous rocks of Ireland. Geol. Surv. Ireland, 36 pp. 2nd Suppl. to 2nd ed.
Cohen, A.D., 1974. Evidence of fires in the ancient Everglades Johnson, B., 1984. The great fire of Borneo. Report of a visit
and coastal swamps of Southern Florida. In: Environments to Kalimantan-Timur a year later, May 1984. World Wild
of South Florida, Present and Past. Miami Geol. Soc. Mem., Life Fund, Godalming, UK, 24 pp.
pp. 213 218. Jones, T.P., 1991. The nature, origin and recognition of fusain.
Cohen, A.D., Spackman, W. and Raymond Jr, R., 1987. Unpubl. Ph.D. Thesis, Univ. London, 2 vol.
Interpreting the characteristics of coal seams from chemical, Jones, T.P. and Chaloner, W.G., 1991. Fossil charcoal,
physical and petrographic studies of peat deposits. In: A.C. its recognition and Palaeoatmospheric significance.
Scott (Editor), Coal and Coal-bearing Strata: Recent Palaeogeogr., Palaeoclimatol., Palaeoecol. (Global Plan.
Advances. Geol. Soc. London Spec. Publ., 32: 107-125. Change Sect.), 97: 39-50.
112 A.C. SCOTT AND T.P. JONES

Jones, T.P., Scott, A.C. and Cope, M., 1991. Reflectance Palaeoclimatol., Palaeoecol., 61: 221-236.
measurements and temperature of formation of modern Sanford Jr., R.L., Saldarriage, J., Clark, K.E., Uhl, C. and
charcoals and implications for studies of fusain. Bull. Geol. Herra, R., 1985. Amazon Rainforest fires. Science, 227:
Soc. Fr., 162: 193-200. 53-55.
Jones, T.P., Scott, A.C. and Mattey, D., 1993. Investigations Scott, A.C., 1978. Sedimentological and ecological control of
of "fusain transition fossils" from the Lower Carboniferous: Westphalian B plant assemblages from West Yorkshire. Proc.
comparisons with modern partially charred wood. Int. J. Yorks. Geol. Soc., 41: 461-508.
Coal Geol., 22: 37-59. Scott, A.C., 1979. The ecology of Coal Measure Floras from
Komarek, E.V., 1972. Ancient fires. Proc. Annu. Tall Timbers Northern Britain. Proc. Geol. Assoc., 90: 97-116.
Fire Ecology Conf., 12: 219-240. Scott, A.C., 1989. Observations on the nature and origin of
Meyer, G.A., Wells, S.G., Bailing, R.C. and Jull, A.J.T., 1992. fusain. Int. J. Coal Geol., 12: 443-475.
Reponse of alluvial systems to fire and climate change in Scott, A.C., 1990. Preservation, evolution and extinction of
Yellowstone National Park. Nature, 357:147-150. plants in Lower Carboniferous volcanic sequences in
Moore, P.D., 1978. Forest Fires. Nature, 272: 754. Scotland. In: M.G. Lockley and A. Rice (Editor), Volcanism
Moore, P.D., 1982. Fire: catastrophic or creative? Impact of and Fossil Biotas. Geol. Soc. Am., Spec. Publ., 244: 25-38.
science on society. UNESCO, 32: 5-14. Scott, A.C. and Collinson, M.E., 1978. Organic sedimentary
Moore, P.D., 1989. No smoke without fire. Nature, 342: particles: results from Scanning Electron Microscope studies
226-227. of fragmentary plant material. In: W.B. Whalley (Editor),
Nichols, G. and Jones, T.P., 1992. Fusain in Carboniferous Scanning Electron Microscopy in the Studies of Sediments.
shallow marine sediments, Donegal, Ireland: The sedimento- Geoabstracts, Norwich, pp. 137-167.
logical effects of wildfire. Sedimentology, 39: 487-502. Scott, A.C. and Galtier, J., 1985. The distribution and ecology
Patterson III, W.A., Edwards, K.J. and Maguire, D.J., 1987. of early ferns. Proc. R. Soc. Edinburgh, B, 86: 141-149.
Microscopic charcoals as a fossil indicator of fire. Quat. Sci. Scott, A.C. and Jones, T.P., 1991. Microscopical observations
Rev., 6: 13-23. of Recent and fossil charcoal. Microsc. Anal., 25: 13-15.
Phillips, T.L., 1981. Stratigraphic occurrences and vegetational Scott, A.C., Meyer-Berthaud, B., Galtier, J., Rex, G.M.,
patterns of Pennsylvanian pteridosperms in Euramerican Brindley, S. and Clayton, G., 1986. Studies on a new Lower
coal swamps. Rev. Palaeobot. Palynol., 32: 5-26. Carboniferous flora from Kingswood near Pettycur,
Phillips, T.L. and DiMichele, D.A., 1981. Palaeoecology of Scotland. 1. Preliminary report. Rev. Palaeobot. Palynol.,
Middle Pennsylvanian age coal swamps in southern 48: 161-180.
Illinois/Herrin coal member at Sahara Mine No 6. In: Skolnick, H., 1958. Observations on fusain. Bull. Am. Assoc.
K.J. Niklas, (Editor), Paleobotany, Paleoecology and Pet. Geol., 42: 2223-2236.
Evolution. Prager. New York, NY, Vol. 1, pp. 231-284. Smith, D.M., Griffin, J.J. and Goldberg, E.D., 1973. Elemental
Racine, C.H., Dennis, J.G. and Patterson III, W.A., 1985. Carbon in marine sediments: a baseline for burning. Nature,
Tundra fire regimes in the Noatak River watershed, Alaska: 241: 268-270.
1956-83. Artic, 38: 194-200. Swanson, F.J., 1981. Fire and geomorphic processes. In: H.A.
Rex, G.M. and Scott, A.C., 1987. The sedimentology, palaeo- Mooney, T.H. Bonnicksen, N.L. Christensen, J.E. Lotan
ecology and preservation of the Lower Carboniferous plant and W.A. Reiners (Editor), Fire Regimes and Ecosystem
deposits at Pettycur, Fife, Scotland. Geol. Mag., 124: 43-66. Properties. USDA For. Serv. Gen. Tech. Rep., WO-26:
Rolfe, W.D.I., Durant, G.P., Fallick, A.E., Hall, A.J., Large, 401-420.
D.J., Scott, A.C., Smithson, T.R. and Walkden, G.M., 1990. Teichmiiller, M., 1989. The genesis of coal from the viewpoint
An early terrestrial biota preserved by Visean vulcanicity in of coal petrology. Int. J. Coal Geol., 12: 1-87.
Scotland. In: M.G. Lockley and A. Rice (Editor), Volcanism Walker, D., 1982. The development of resistance in burned
and Fossil Biotas. Geol. Soc. Am., Spec. Publ., 244: 13-24. vegetation. In: E.I. Newman (Editor), The Plant Community
Rundel, P.W., 1981. Fire as an ecological factor. In: O.L. as a Working Mechanism. The British Ecological Society.
Lange, P.S. Nobel, C.B. Osmond and H. Ziegler (Editor), Watson, A., Lovelock, J.E. and Margulis, L., 1978.
Physiological Plant Ecology. I. Response to the Physical Methanogenesis, fires and the regulation of atmospheric
Environment. Springer Verlag, Berlin, pp. 501-538. oxygen. Biosystems, 10: 293-298.
Sander, P.M., 1987. Taphonomy of the Lower Permian Wuerthner, G., 1988. Yellowstone and the Fires of Change.
Geraldine Bonebed in Archer County, Texas. Palaeogeogr., Haggis House Publ., Salt Lake City, UT.