• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Constraints on the Thermal Energy Released From the Chicxulub Impactor: New Evidence From Multi-Method Charcoal Analysis

Posted on: Saturday, 9 July 2005, 03:00 CDT

Abstract:

It has been suggested by various workers that an extraterrestrial impact at the K-T boundary delivered sufficient thermal power to ignite globally extensive wildfires. Numerous models have sought to predict the amount of thermal power released by the impact, but none have considered the distribution of wildfire indicators in K-T rocks. Probably the most distinctive product from combustion of biomass is charcoal. The abundance of charcoal across the K-T boundary at eight non-marine sites in North America, stretching from Colorado in the south to Saskatchewan in the north, is recorded using three separate methods that allow quantitative analyses of microscopic to macroscopic charcoal particles. This study not only provides the first extensive study of charcoals across the K-T boundary but also uses the presence or absence of charred material to predict the extent and severity of the thermal pulse released by the K-T impact across the area predicted to have suffered the most extreme environmental effects. The K-T rocks contain on average between four and eight times (according to the method used) less charcoal than the Cretaceous rock record and non-charred plant remains are abundant in the K-T rocks. The below-background charcoal abundance and the high proportion of noncharred material in the K-T and lowermost Tertiary rocks across the Western Interior of North America suggest that there were no significant wildfires in this area associated with the K-T event. Although soot and polyaromatic hydrocarbons (PAHs) have been reported in the K-T rocks we suggest that the soot morphology and PAH types are more consistent with a source from the vaporization of hydrocarbons rather than biomass. For spontaneous ignition of vegetation temperatures >545 C are necessary, whereas smouldering will begin at 325 C. The below- background levels of charcoal in the K-T rocks allow the ground temperatures following the K-T impact to be constrained to between no more than 545 C at any point and not above 325 C for any significant period. This implies a maximum irradiance of <19 kW m^sup -2^ at the ground surface and that no more than 6 kW m^sup - 2^ of thermal power was delivered to the ground for more than a few hours. Therefore our results show that the fossil record indicates that the impact at Chixculub did not generate sufficient thermal power to ignite extensive wildfires.

Keywords: Chicxulub, charcoal, K-T boundary, fires, temperature, impacts.

The Western Interior of North America contains some of the most exceptional non-marine Cretaceous-Tertiary boundary (K-T; also termed K-P Cretaceous-Palaeogene boundary) sequences in the world, with sites stretching from New Mexico, USA, in the south to Alberta, Canada, in the north. Numerous workers (e.g. Izett & Bohor 1986; Izett 1990; Hildebrand 1993) have described two intimately associated layers forming the boundary interval. The uppermost layer has been traced and studied around the world, leading most to agree that this reflects a global event.

The fossil record reveals a considerable biotic turnover across the K-T boundary, which is arguable evidence for a mass extinction (MacLeod et al. 1997). However, the nature and cause of the K-T event are still much in debate. Alvarez et al. (1980) published evidence that an impact of a large asteroid with the Earth was temporally linked to the biotic turnovers. The subsequent discovery of the Chicxulub crater, linked to the K-T boundary (Hildebrand & Boynton 1990; Hildebrand et al. 1991), suggested that at last the 'smoking gun' had been found. However, the exact environmental consequences, timing and effects of this impact event across the globe remain a topic of hot debate.

One proposed consequence, of considerable environmental significance, is the claim that the impact led to a global fire event (Wolbach et al. 1985, 1988, 1990; Melosh et al. 1990; Ivany & Salawitch 1993). Melosh et al. (1990) proposed that wildfires were ignited by the thermal energy radiated by re-entering ejecta following the asteroid impact. Wolbach et al. (1990) suggested that lightning strikes, produced from charge separation during settling of ejecta through the atmosphere, ignited the vegetation.

Evidence for a global wildfire is supported by several sources. Wolbach et al. (1985, 1988, 1990) reported an enrichment of soot in the K-T rocks that is not only isotopically uniform, suggesting a single source, but also bears an isotopic signature consistent with burning of biomass. Arinbou et al. (1999) investigated the abundance of polyaromatic hydrocarbons (PAHs) in K-T rocks from Caravaca in Spain, observing a 116-fold increase in some PAHs compared with the Cretaceous background. They suggested that the most likely source for the PAHs was the combustion of terrestrial organic matter.

This paper aims to evaluate further the hypothesis of a global wildfire event, by studying the most distinctive product of wildfire, charcoal, across the K-T boundary at eight sites across the Western Interior of North America (the area predicted to have suffered the maximum environmental perturbations). This will allow constraints to be placed on the amount of thermal radiation released by the impact at the K-T boundary. The relative abundance of charcoals has been quantified using three methods. Proportions of inertinites were documented in situ from polished blocks of sediment (Belcher et al. 2003). Macroscopic charcoal particles were quantified from the coarse fraction of sieved demineralized sediment. Microscopic charcoal particles were counted from palynological slides. Although polished blocks are essential to establish the exact location of charcoals they represent only a 'snapshot' in section of charcoal content. Residues from demineralized sediment have lower spatial resolution but reveal the sizes and shapes of particles and the overall charcoal content of each horizon. This quantitative multi-method approach records the abundance and distribution of microfossil to macrofossil charcoal and provides the first extensive study of charcoals across the K-T boundary.

Location and site information

Abundant evidence supports the idea that an extraterrestrial body of significant size struck the Yucatan Penisula 65 Ma ago (Hildebrand et al. 1991). Working with this hypothesis, eight non- marine K-T sites, stretching from Colorado (Raton Basin) in the south (proximal to the crater) to Saskatchewan in the north (distal from the crater) were studied, allowing the effects of the thermal radiation to be compared with distance from the crater. All sites studied contain an iridium anomaly, fern spike, shocked minerals and two boundary clay layers marking the K-T event. Table 1 and Figure 1 provide locations and information about the sites studied.

Materials and methods

Sampling

Intact lithological sequences of the K-T boundary and the adjoining sedimentary rocks were collected in the field (see Fig. 2). The blocks were contained in dental plaster to protect them during transit from the field to the laboratory. The plaster blocks were sawn in half, and one half was impregnated with resin for polishing whereas the other half was left untouched for sediment sampling. A slab of resin-impregnated rock was cut and polished, providing a permanent record of the complete lithological sequence. Smaller polished blocks of sediment were prepared, also at right angles to bedding, and examined in reflected light under oil using a Nikon Microphot Microscope. Using the small polished blocks the microstratigraphy of the sediments was studied, allowing microstratigraphic-scale samples to be taken from the block. The small polished blocks have been deposited in the Natural History Museum, London (samples BMNH V69801-V69824) (see Appendix).

Table 1. Location, distance from Chicxulub, lithologies and appropriate references for the sites studied

Fig. 1. Map showing location of numbered sites in Table 1.

Petrography and in situ study of inertinites

The in situ abundances (%) of inertinites (charred material), vitrinite, liptinites (non-charred plant material) and mineral matter were recorded from the small polished blocks using a point count method (100 per sample level) with the aid of a Whipple grid.

It is generally accepted that most of the inertinite group macerais represent fossil charcoal derived from wildfire (Scott 1989, 2000; Jones el al. 1997). The inertinite group macerais in this study include fusinite (Fig. 3a, (I)), which is identified by having a clearly structured appearance with highly reflecting (bright white) cell walls, semifusinite (Fig. 3a, (2)), which shows some structure but is pale grey in appearance, and inertodetrinite (Fig. 3a, (3)), which is small highly reflecting particles and in this study includes those below 33 m in size.

Two further macerai groups are recognized in this study: vitrinite (Fig. 3a, (4)), which is unstructured and a dull pale grey, and liptinite, which is unstructured and has a low reflectance (almost black). Both these represent non-charred organic matter. The petrographie approach provides a quantitative assessment of in situ charcoal of all size fractions as seenin a polished section of sediment.

Preparation and analysis of sieved macrofossil and microfossil charcoal particles

Samples (5 g) were demineralized to release resistant plant macro- and microfossil particles (including charcoal) for study. Treatment with cold hydrochloric acid (10% to concentrated HCl) for 72 h and cold concentrated hydrofluoric acid for a further 72 h removed carbonates and silicates, respectively. The process was completed with a final treatment in cold concentrated hydrochloric acid overnight to avoid calcium fluoride precipitation and then samples were rinsed with distilled water until a neutral pH was achieved. The remaining organic material was sieved through a 150 m mesh and both fractions were collected.

Fig. 2. Excavated lithological block across the K-T boundary at Madrid East South, USA, prior to encasing in plaster. Scale bar is in 1 cm units.

Fig. 3. Characteristics of organic material, (a) Charred material (1,2 and 3) and non-charred organic material (4) as recognized in a polished block seen under oil using reflected light microscopy: (1) shows fusinite, which is identified by having a clearly structured appearance with highly reflecting (bright white) cell walls; (2) represents semifusinite, which shows some structure but is pale grey in appearance; (3) is inertodetrinite, which is small highly reflecting particles and in this study includes those below 33 m in size; (4) is vitrinite (non-charred material), which is unstructured and a dull grey, (b) A transmitted light micrograph of a palynology slide. The charcoal is distinguished by its structure, sharp fracture and black colour (where partially translucent and seen in colour charcoal is grey-brown). Some charcoal particles (e.g. near centre of the image) are black fibrous grains sometimes exhibiting cellular structure.

The > 150 m residues were examined using a low-power binocular microscope. The samples were dispersed in twice the amount of distilled water necessary to cover the organic debris in the pots. A drop or two of this slurry was dispersed on a large slide and covered with a floating large cover slip (to stop the water evaporating under the heat of the microscope lights but floating so that it did not squash any specimens). The quantity of non-charred particles and charred particles on the temporary slide was recorded. The counted material was then transferred to a new pot and further temporary slides were made until all the material in each sample had been studied. If charcoal was particularly abundant (e.g. tens of thousands of particles) one-fifth of the residue was counted (approximately the amount of charcoal in 1 g of rock as opposed to the 5 g disaggregated). This method was found to be easier than simply picking through the material with a fine brush, as the sample need only be dispersed in a little water, making each particle easy to see (whereas dispersing the sample on a Petri dish, as is more typical, was found to be awkward, as the particles floated or moved around). Charcoals were recognized by their almost graphite-like sheen or their very dark (essentially black) colour, along with a cuboid or needle-like 3D shape, allowing them to be distinguished from other woody or coaly fragments that lack shine (Scott 1989). Charcoalification preserves 3D cellular structure of plant material; however, it does involve an homogenizatton of the layers in the cell wall. These features are visible using an SEM (Scott 2000). Representative pieces of charcoal were selected randomly from the samples and studied uncoated using an S3000N Hitachi variable pressure SEM, which confirmed their identity as charcoal.

The <150 m fractions were dispersed evenly in a known volume of distilled water (between 2.5 and 20 ml depending on the amount of residue remaining). This solution (50 or 100 m) was added to a cover slip using an Eppendorf pipette and mixed with two droplets of polyvinyl alcohol and dried. The cover slips were then mounted onto slides using petropoxy resin 154. These palynological slides will be deposited in the Natural History Museum. London, when other research on them has been completed. The numbers of charred particles present across two transects of a slide were recorded and converted to charcoal particles per gram of sediment (based on knowing the amount of liquid mounted on the slide from a known volume of solution). Charcoals were distinguished by their translucent grey-brown structured nature or occurred as black fibrous grains sometimes exhibiting cellular structure (Fig. 3b). This distinguishes charcoal from coalified material (which is typically dark orange-brown with an irregular shape) and any remaining dark opaque mineral matter.

The study of macro- and microfossil fractions provides quantitative data on the numbers of charcoal particles >150 m and <150 m in size per gram of sediment.

Results

Figure 4a-c illustrates the proportions of charcoal across all the sites using the three methods. One sample Shapiro-Wilkes statistical tests and normal probability plots of the charcoal counts using all methods (not shown) confirm that the charcoal count data do not show a normal distribution. Table 2 gives the mean and median values of charred and non-charred material present in the Cretaceous, K-T and Tertiary rocks rocks using the three methods. Figure 5a-h provides a graphic representation of the data in Table 2 using box plots. In these non-normally distributed data the mean is clearly influenced by extreme outliers and so the median is the more appropriate measure. Table 2 shows that on average (median) the K-T rocks contain between four and eight times less charred remains (depending on the measure used) than the Cretaceous or Tertiary rocks. On average (median) 99% of the total organic material in the K-T rocks is not charred (based on the >150m fraction counts), compared with averages (median) of 94% and 62% non-charred material in the Cretaceous and Tertiary, respectively.

A non-parametric test (Mann-Whitney) was applied to the count data and confirmed that there is a significant difference in charcoal abundance in the K-T rocks compared with the Cretaceous and Tertiary rocks. No significant difference was found between the charcoal proportions in the Cretaceous and Tertiary samples (P > 0.05, Table 3). In contrast, both the Cretaceous and Tertiary charcoal proportions are very different from those of the K-T rocks (P < 0.05, Table 3), indicating that the low abundance of charcoal in the K-T rocks is not due to chance.

Discussion

Could taphonomic factors account for below-background K-T charcoal?

Charcoal pieces <100 m in size can be carried by wind (Patterson el al. 1987), whereas larger fragments are generally transported by water and float (Nichols et al. 2000). Studies of Holocene sediments (Innes & Simmons 2000) reveal that peaks in the abundance of macroscopic and microscopic charcoals correlate well; however, microscopic charcoal often remains when macroscopic particles do not. It has been argued that assemblages comprising both macroscopic and microscopic charcoal represent local fire events occurring at or very close to the site of deposition, whereas assemblages containing only microscopic particles provide a regional pattern of burning (Innes & Simmons 2000). Studies conducted on Recent lake sediments suggest that sieved fractions of charcoal provide a record of fires that occur near lake shores, providing a very local signal (Laird & Campbell 2000). The charcoal particles in the Cretaceous, Tertiary and K-T rocks are dominantly woody in nature; other charred organs such as leaves, cuticle and flowers are largely absent. This suggests that either (1) some sorting or winnowing has occurred, separating woody fragments from other plant parts such as leaves and flowers, or (2) smaller, more delicate organs (e.g. leaves and flowers) were fully burned and have not survived as charcoal. Scott et al. (2000a) showed that winnowing separation can occur over very short distances (a few hundred metres or less). Furthermore, the assemblages do contain both macroscopic and microscopic charred particles of a wide range of sizes with no evidence of transport- induced size sorting or shape modification. Together these data indicate that the charcoal was produced by fires close to the sites of deposition.

All three methods of analysis (in situ proportions of inertinites and sieved macro-fossils and micro-fossils) reveal the same trend of charcoal distribution. Greater amounts of charcoal occur in the Cretaceous sediments with less in the K-T boundary rocks. Charcoal occurrences show that wildfires were evidently a typical part of the Cretaceous ecosystems across the North American continent. The K-T boundary rocks yield abundances of charcoal well below the background levels for the Cretaceous. The charcoal proportions at the eight K-T sites demonstrate that the same phenomenon is present in both siliciclastic and peatforming sequences (see Table 1), providing evidence that the trend is not biased by certain environments or due to preservational conditions.

It is probable that not all fires will leave evidence in the fossil record in the form of charcoal. An absence of charcoal (or particular sizes, shapes or types of charcoals) in the K-T rocks, combined with their presence above and below, could have been argued to indicate a special type of K-T fire that left no charcoal in the sediments. However, charcoal is not totally absent from the K-T rocks and the charcoal signature is very similar throughout the sequences we have studied. There is little or no difference between the size ranges, particle shapes or botanical nature of the K-T charcoal particles and those found in the Cretaceous and Tertiary rocks. Larger particles are more frequent in the Cretaceous and Tertiary rocks but this is likely to be due to greater abundances of charcoal in these layersincreasing the likelihood of encountering larger particles.

At two sites (Wood Mountain Creek and Madrid East South) the K-T rocks are contained within a coal. The organic (macerai) composition of the coals remains similar on either side of the boundary interval. Charred peat surfaces have been reported in the fossil record (Petersen 1998) but are absent in the K-T peat sequences. Wood Mountain Creek and Madrid East South reveal that the K-T rocks simply interrupt an otherwise apparently continous episode of peat formation. A similar situation also existed at Sugarite, New Mexico, nearer the impact crater (Scott et al. 2000b).

The K-T sediments are considered to have been deposited in minutes to hours (Sweet 2001) for the impact-lower claystone layer, to hours to days (Pope 2002) or days to months (Hildebrand 1993; Toon et al. 1997) for the fireball-upper claystone layer. Although this represents rapid sedimentation the actual volume of sediment deposited is not large, now representing a 2.5 cm thickness of rock. Sample levels that are 1 mm thick in the Cretaceous and Tertiary (e.g. see Wood Mountain Creek in Fig. 4) can be considered to represent a short time interval and thus are somewhat comparable with the K-T interval. Tn contrast to the K-T, these thin horizons contain a considerable amount of charcoal, suggesting that the short duration of sedimentation cannot explain the lack of charcoal in the K-T rocks. On the contrary, it would be expected that an all- encompassing raging wildfire would be most likely to leave abundant charcoal, the preservation of which would have been aided by the falling K-T impact debris.

If we hypothesize that the charcoal in the K-T rocks represents a single fire resulting from the K-T event then the amount of charcoal released into the sedimentary record from this single fire can be estimated from the K-T charcoal abundances. The hypothetical K-T fire across the Western Interior of North America produced on average 75 macroscopic particles of charcoal per gram of sediment. This can be compared with modern fire charcoal production, for example, the 1996 wildfire in the Cascade range, USA, where fire destroyed 95% of the biomass in the area of the fire (37.7 km^sup 2^) (Gardner & Whitlock 2001). This single fire produced around 1500 macroscopic particles of charcoal per gram of sediment (this value is from a burn area measuring 0.16 ha) (Gardner & Whitlock 2001), 20 times more charcoal than is present in the K-T rocks. The largest macroscopic charcoal peak that may represent a single fire event (2 mm thick sample level) found at any horizon in this study occurs in the Tertiary sequence of Berwind Canyon. This Tertiary fire event produced just under 61 000 macroscopic particles of charcoal per gram of sediment, over 800 times more charcoal than is present in the K-T rocks.

Fig. 4. (a) Percentage of inertinite (charred material) across the K-T boundary following a transect of sites across North America, measured in situ in polished blocks of sediment, (b) Quantity of > 150 m sized charcoal particles across the K-T boundary (number of charred particles per gram of sediment). Three intervals at Madrid East South and Berwind Canyon have been truncated (bars containing asterisks), as they contain some very large peaks in abundance of charcoal. The large peak at Madrid East South is 19707 particles of charcoal per gram of sediment and the largest peak at Berwind Canyon is 60 988 particles of charcoal per gram of sediment, (c) Quantity of < 150 m sized charcoal particles across the K-T boundary (number of charred particles per gram of sediment).

Table 2. Average (mean and median) amounts of charred and noncharred material found in the K-T rocks compared with the Cretaceous and Tertiary rocks

Charcoal is present in the K-T rocks but considerably below Cretaceous background levels. Taphonomic factors are not responsible for this difference. The charcoal throughout the sequences was produced by local fires and the distribution patterns observed are very similar in three independent analytical approaches. Furthermore, there is nothing unusual about the KT charcoals and no charred peat surface or other modification of organic matter in K-T peat-forming sequences. The K-T rocks do represent a short time interval but thin (millimetre-scale) sedimentary units elswhere in the succession probably also represent short time intervals and they contain very high abundances of charcoal. Charcoal abundances at the K-T are less than produced by single modern fire events and considerably less than probable fire events elsewhere in the sequence studied.

Could charcoal have been destroyed by intense thermal radiation during the K-Tevent?

Taphonomic factors (see above) cannot account for the belowbackground levels of charcoal in the K-T rocks. However, it could be suggested that the K-T impact released such a vast amount of thermal power (e.g. 5000 kW m^sup -2^ calculated from Hildebrand (1993)) that plant material was converted directly to CO2, leaving no or very little charcoal in the fossil record. Robertson et al. (20046) suggested that the high temperatures (predicted by Melosh et al. (1990) to be c. 827 C) during the K-T event would have destroyed any charcoal formed.

To test these suggestions we have studied the effects of pure radiant heat (that is, with no flame for ignition as would be the case for the thermal power released from an asteroid impact) on wood (pine). Small (c. 6 cm 2 cm 2 cm) (not fresh) and large (fresh) (c. 20 cm 5 cm 5 cm) pieces of wood were subjected to various temperatures (as radiant heat) in a Carbolite CWFIlOO furnace for 2 h with air included or excluded. The pieces of pine were labelled, weighed, then soaked in distilled water for 5 days and kept wet prior to burning. Although significant weight loss occurs during the heating process the actual physical amount of material remaining (Fig. 6) is most informative when considering charcoal production and abundance in sediments. After 2 h of heating in air charcoal remains up to 750 C (Fig. 6a). If air is excluded from the heating process (as would be the case at the base of natural wildfires) then the amount of the material remaining on heating up to 1000C is almost the same as the original piece of wood prior to heating (Fig. 6b). When larger pieces of fresh pine were heated for 2 h in the presence of air material remains up to 900 C with ash produced at 1000C (Fig. 6c). Therefore our experiments show that a high proportion of initial biomass can survive as charcoal even at the high temperatures previously predicted for the K-T event by Melosh et al. (1990).

Our results show that the destruction of plant (wood) material on heating is influenced more by oxygen availability than by temperature. Robertson et al. (2004?) suggested that local oxygen deficiencies occurred under the K-T fires, as in firestorms over burning cities during World War II. In our experiment, with air excluded, most of the plant material survived as charcoal, even at temperatures of 1000 C (Fig. 6b). Therefore an oxygen-deficient K-T fire would increase, not decrease, charcoal production. Scott (2000) demonstrated that charcoal reflectance increases with temperature. If the K-T vegetation had been subjected to very high temperatures we would expect to find very highly reflecting charcoals. However, reflectances of charcoal in the K-T rocks are comparable with those of charcoal found in the Cretaceous and Tertiary rocks. Furthermore, the presence of abundant non-charred plant fragments in the K-T rocks (e.g. an average (median) 5281 macroscopic particles per gram, or an average of 22% vitrinite (non-charred material) as seen in polished blocks of sediment) (see also Belcher et al. 2003) is incompatible with any scenario that invokes very high temperatures or total destruction of plant material during the K-T event.

Absence of evidence of geomorphological and sedimentological impact of such widespread and intense fires

One of the potential consequences of wildland fires is post-fire erosion and deposition. Indeed, this is currently a major hazard in many parts of the world, including the western USA (Elsenbeer & Robichaud 2001). Such effects have been documented both in the recent and fossil record (Swanson 1981; Nichols & Jones 1992; SaIa & Rubio 1994; Whitlock & Millspaugh 1996; Falcon-Lang 1998). The effects include increased soil water repellency following fire (DeBano 2000) and increased erosion and deposition in a range of sedimentological settings (Cannon et al. 2001; Elliott & Parker 2001; Meyer et al. 2001; Moody & Martin 2001). If there had been a major worldwide fire, as has been suggested, we would have expected to see evidence from the sedimentological record. We see no evidence of this, which supports our conclusions that there were no large fires as a result of the impact.

Reassessment of the soot record

Wolbach et al. (1990) observed that the coarse K-T carbon contained no particles that were diagnostic of wood; moreover, the soot particles from the K-T rocks were described as fluffy aggregates of soot spherules, with a characteristic 'chained cluster' morphology (Wolbach et al. 1990). To compare the morphology of biomass soot with other soot, samples were collected from the 2002 Hayman wildfire, Colorado, USA (biomass sourced), car exhaust (hydrocarbon source), and the chimney of a coal-burning stove (coal source). Soot was studied uncoated using an S3000N Hitachi variable pressure SEM (Fig. 7). The soot from the Hayman wildfire (Fig. 7a) contains numerous fragments characteristic of plants, such as particle I, which shows cell-wall pits. Komarek et al. (1973) have shown that smoke from biomass burning will contain identifiable parts of vegetation. The soot consists mainly of tiny plant fragments that are generally smaller than 30 m in length. There are some tarry-like spherical dr\oplets in the sample; it is likely that these are unrelated to the soot from the wildfire itself and are more probably sourced from combustion of the diesel fuel of the firefighting equipment. The soot from the Hayman wildfire is dominated by 'large' solid combustion residues. Figure 7b shows soot from the Hayman wildfire following treatment with a 600 h dichromate etch, as used by Wolbach et al. (1990) to extract soot from the K-T rocks. The characteristics of the soot remain the same following this treatment, confirming that the method used by Wolbach et al. (1990) does not alter the morphology of the soot.

Fig. 5. Box plots to demonstrate the spread of the charcoal data and the strong influence of extreme outliers on the mean values, showing that the median is the most appropriate measure of central tendency for these data. Each pair of graphs compares data with and without extreme outliers, (a), (b) Percentage of inertinite (polished block counts) (the median for the K-T is zero); (c), (d) number of > 150 m sized charcoal particles per gram of sediment; (e), (f ) number of < 150 m sized charcoal particles per gram of sediment; (g), (h) number of > 150 m sized non-charred particles per gram of sediment.

Table 3. P value results from the Mann - Whitney test, for the amounts of charcoal found in K-T rocks compared with the Cretaceous and Tertiary, and the Cretaceous compared with the Tertiary

Soot produced from the combustion of hydrocarbons (unleaded petroleum gas) forms fluffy aggregates of soot particles (Fig. 7c). Most of the particles are spherical to subspherical in shape and are clustered together. The petroleum soot is not formed by solid-phase charring of fuel but is produced by recombination and coagulation of aromatic molecules in the gas phase. The soot produced from the burning of coal (Fig. 7d) contains particles less rounded in shape than those produced by petroleum gas and consists of fragmentai material generally angular to subspherical in shape (with some particles exhibiting a rough, pitted surface texture) aggregated and welded into characteristic clusters and chains.

Fig. 6. The amount of material (pine wood) remaining as charcoal or ash following exposure to various amounts of radiant heat, (a), (b) Small pieces; (c) large pieces; (a) 2 h at temperatures between 500 and 800 C in the presence of air; (b) 2 h at temperatures between 700 and 1000 0C in the absence of air; (c) 2 h at temperatures between 600 and 1000 C in the presence of air.

The morphology of the K-T soot as described by Wolbach et al. (1990) appears to be inconsistent with it being sourced from combustion of biomass because it is morphologically similar to that produced by burning petroleum or coal. Chicxulub drill cores reveal that the target rocks contain hydrocarbons (Gilmour et al. 2003), the vaporization of which could produce soot. The isotopic unformity of the K-T soot need not exclude such a source.

Reassessment of the PAH record

Gilmour & Guenther (1988), Wolbach et al. (1988, 1990) and Gilmour et al. (1990) argued that the presence of the PAH retene in the K-T rocks was strong evidence that the K-T soot was sourced from wildfires. However, retene is not exclusively formed by biomass burning but is also derived through diagenesis of pimaric and abietic acids from organic matter (Killops & Killops 1993).

Oros & Simoneit (2000, 2001a, b) and Simoneit (2002) published extensive lists of the organic tracers (including PAHs) from incomplete combustion of biomass and coals. They listed no PAHs produced exclusively by biomass burning; however, there are PAHs that are formed only during the combustion of hydrocarbon material. Both Venkatesan & Dahl (1989) and Arinbou et al. (1999) found coronene to be one of the dominant PAHs in the K-T rocks. Coronene is formed from the combustion of coals and not biomass (Oros & Simoneit 2000, 2001a b; Simoneit 2002), whereas the other PAHs found in the K-T rocks (phenanthrene, 2 methylfluorenc, benzo(b + k) fluoranthene, chrysene/triphenylene, pyrene, fluoranthene, benzopyrene, benzo(g,h,i)perylene (Venkatesan & Dahl 1989), benzo(g,h,i)perylene and benzo(e)pyrene (Arinbou et al. 1999)) can be formed through combustion of both biomass and hydrocarbons (Oros & Simoneit 2000, 2001a, b; Simoneit 2002). The presence of a significant quantity of coronene suggests that the PAHs might be sourced from combustion of hydrocarbons rather than biomass material.

Fig. 7. The morphologies of various kinds of soot as seen using the SEM. (a) Soot from the 2002 Hayman wildfire in Colorado, which represents soot produced by biomass burning, (b) Soot from the Hayman wildfire that has been subjected to a 600 h dichromate etch process as used by Wolbach et al. (1990) to extract soot from the K- T sediments. This confirms that the process used by Wolbach et al. (1990) does not alter the key characteristics of the soot, (c) Soot produced from the combustion of petroleum gas. (d) Soot produced from the combustion of coal.

Venkatesan & Dahl (1989) concluded that the PAH distributions in the K-T rocks were consistent with the suggestion of massive global fires. However, they considered that the PAHs at Woodside Creek and Gubbio could be characteristic of wood or kerosene, whereas those at Stevns Klint were more characteristic of production by the combustion of coal (Venkatesan & Dahl 1989). This would accord with the interpretation that the PAHs were sourced from the combustion of hydrocarbon material rather than biomass burning.

Problems with K- T impact models

K-T ejecta re-entry models (e.g. Melosh et al. 1990; Toon et al. 1997; Kring & Durda 2002) have assumed that the upper K-T rock layer is composed of ballistically distributed ejecta. This may be true for the lower claystone, which decreases in thickness with distance from Chicxulub following a power-law relation with exponent of -3 (Hildebrand & Stansberry 1992), a characteristic of ballistically distributed ejecta. However, this is not the dispersal mechanism for the upper layer, which is uniform in thickness (c. 3 mm) across the world. The presence of this upper globally extensive layer was a puzzle finally explained by remote observations of the SL9 impacts on Jupiter, which showed that impact fireballs collapse hydrodynamically (Hamell et al. 1995). Ballistic dispersal does not produce a layer of uniform thickness across the globe. However, hydrodynamic collapse of an impact fireball would globally disperse impact products but whilst doing so would release an order of magnitude less energy. This implies that previous models of the power released from the K-T impact event (e.g. Melosh et al. 1990; Toon et al. 1997; Kring & Durda 2002) have over-estimated the amount of thermal radiation released across the globe and that indeed there may have been no globally extensive thermal pulse (Belcher et al. 2004).

K-T thermal power and ground temperatures

The models produced by Melosh et al. (1990), Toon et al. (1997) and Kring & Durda (2002) predicted the thermal power released by frictional heating of K-T melt spherules on ballistic re-entry to the atmosphere; therefore these models can be used as an indication of the amount of thermal power released across the area where the ballistically distributed ejecta is found (i.e. North America) but not the globe. Kring & Durda (2002) modelled the power delivered to the atmosphere above specific geographical locations and how this diminished with time after the impact, providing a model that considered the duration of the thermal pulse following the K-T event (unlike Melosh et al. (1990) or Toon et al. (1997)). Using their projected power delivery to the atmosphere above Colorado for the first 35 h following the K-T impact (Kring & Durda 2002, fig. 9d) and the Stefan-Boltzmann Law (Young 1992)

P/A = &963;T^sup 4^

where P is the radiative power (W), A is area (m^sup -2^) (receiving the power), e is the emissitivity of the object (in this case for a cloudless sky, the atmosphere, e = 0.76 (Brooks 2002), s is the Stefan-Boltzmann constant (σ = 5.67 10^sup -8^), and T is the temperature (K) of the atmosphere, we have calculated the probable K-T ground temperatures for the Western Interior of North America, (assuming a cloudless sky) (Fig. 8). By using the data from Kring & Durda (2002) the duration of the ground temperature is also considered (Fig. 8). For the first 15 min following the impact ground temperatures would have been of the order of 639 C. Over the next 2 h 45 min ground temperatures would have averaged 339 C (range 78.6-414C). In the following 7 h, the average ground temperature would have been 201 C (range 23-402 C) and for the next 17 h ground temperature would have averaged 158 C (range, no delivery to 309 C). Kring & Durda (2002) suggested that the thermal power released from the K-T impact was delivered in a series of pulses, so that peaks in thermal power were not delivered for extensive periods. Thus the 2 h exposure to radiant heat in the charring experiments discussed herein is comparable in duration with the pulses of heat described by Kring & Durda (2002). The average ground temperature over the 24 h following the impact (excluding the first 15 min) would have been 266 C.

Fig. 8. Ground temperatures in the 24 h following the K-T impact event. *, Minimum; [black square], maximum; [black triangle up], mean. Kring & Durda (2002) modelled the thermal power delivered to the atmosphere above Colorado. We have used the Stefan-Boltzmann Law to estimate the ground temperatures based on the Kring & Durda (2002) model. Those workers suggested that the thermal radiation was supplied in a series of pulses, one initial maximum pulse seconds after the impact (peak at O h on the graph) followed by subsequent pulses of thermal radiation (represented by the maximum values on the graph). The minimum values on the graph represent the amount of thermalradiation experienced at ground level some time after each thermal pulse. The mean temperatures have been calculated as an overall guide to the temperatures experienced following the K-T event.

Cellulose comprises up to 70% of wood cells, is stable up to 250 C but at temperatures around 325 C begins to break down, generating flammable gases (Pyne et al. 1996). If biomass material is subject to temperatures around 325 C for an extended period smouldering may begin (Pyne et al. 1996). Flames are typically produced between 425 C and 980 C (Pyne et al. 1996); however, for spontaneous ignition to occur a temperature of 545 C is necessary (Simms & Law 1967). The limited amount of charcoal in the K-T rocks constrains ground temperatures to be no more than 545 C at any point (contrary to Kring & Durda's calculation for the first 15 min following the impact) and no more than 325 C for several hours. These temperatures translate to 19kWm^sup -2^ and 6 kW m^sup -2^ of thermal power delivered to the ground surface following the K-T impact event, respectively (Belcher et al. 2003). The temperatures calculated from Kring & Durda (2002) suggest that ground temperatures may have remained just above 325 C for the 3 h following the impact, a duration that might have allowed the vegetation to become dry but would probably not have sufficiently broken down the cellulose to have allowed smouldering. This is further confirmed by the limited amount of charred remains in the K-T rocks, as smouldering wildfires typically produce abundant charcoal (Scott 2000). In the following 7 h ground temperatures were on average 201 C (range 23-402 C, with the peak temperatures being delivered as short pulses); cellulose is still stable at this temperature and smouldering would not be initiated, confirming our view that extensive fires did not occur as part of the K-T events.

Conclusions

Quantitative data from three different measures of charcoal abundance (in situ in polished blocks of rock, and macro- and microscopic charcoal particles released from sieving of demineralized sediment) reveal that the Cretaceous-Tertiary boundary rocks across the Western Interior of North America contain significantly less charcoal than is typical of the Cretaceous background of this area. The Cretaceous sedimentary rocks contain between four and eight times (according to the measure used) more charcoal particles than the K-T rocks. Taphonomic factors cannot explain this difference. Abundant charcoal can be produced experimentally at 1000C so destruction by intense thermal radiation cannot account for the below-background levels in the K-T rocks. Furthermore, non-charred plant remains are also abundant in the K-T rocks. Reassessment of the record of soot and PAHs reported in the K- T rocks suggests that the morphology of the soot and the signature of the PAHs is more consistent with them being sourced from the vaporization of hydrocarbon material rather than biomass burning. We conclude that there was no significant wildfire across North America as part of the K-T events.

The below-background levels of charcoal in the K-T rocks allow the ground temperatures following the K-T impact to be constrained to no more than 545 C at any point and not above 325 C for more than several hours. These values are largely consistent (except for the first 15 min) with those calculated from a model (Kring & Durda 2002) of thermal power released from the K-T impact. These temperatures imply a maximum irradiance of <19kWm^sup -2^ at the ground surface and that no more than 6 kW m^sup -2^ of thermal power was delivered to the ground for more than a few hours. Therefore both the fossil record and a K-T impact model indicate that the impact at Chixculub did not generate sufficient thermal power to ignite extensive wildfires.

Hildebrand (1993) concluded that 'the K-T impact turned the Earth's surface into a living hell, a dark, burning, sulphurous world'. It may have been a dark and sulphurous hell, but the Earth's surface was not burning.

We thank P. Finch, A. Hildebrand, A. Sweet, R. Taylor and W. Wolbach for helpful discussions and exchange of ideas on the K-T event and/or wildfires. We thank J. Hanova, A. Hildebrand, K. Johnson, D. Pearson, A. Sweet, the staff from Grasslands National Park, Canada, and the landowners (M. Andersen and R. Hordenchuck) for help with fieldwork. We thank members of the Colorado fire crews who kindly collected soot samples from the Hayman wildfire on our behalf. J. Wolfe is gratefully acknowledged for providing the plaster jacketed block from Rick's Place. We also thank N. Holloway for expert preparation of polished blocks of sediment, S. Gibbons for laboratory assistance, K. D'Souza for photography and P. Goggins for help with SEM. C.M.B. gives thanks and appreciation to B. Steel and M. Kucera for guidance and discussions regarding data analysis. Constructive reviews by J. Hilton and W. Clemens improved the quality of the manuscript. M.E.C. acknowledges funding from National Geographic to collect at Teapot Dome with Jack Wolfe. A.C.S. acknowledges a Royal Society grant for the purchase of the oven. Funding for C.M.B. through a NERC studentship (NER/S/A / 2001/ 06342) and case support from Royal Botanic Gardens, Kew, is gratefully acknowledged.

References

ALVAREZ, L.W., ALVAREZ, W., ASARO, F. & MICHAL, H.V. 1980. Extraterrestrial cause for the Cretaceous-Tertiary extinction. Science, 208, 1095-1108.

ARINBOU, T., ISHIWATARI, R., KAIHO, K. & LAMOLDA. M.A. 1999. Spike of pyrosynthetic polycyclic aromatic hydrocarbons associated with an abrupt decrease in δ^sup 13^C of a terrestrial biomarker at the Cretaceous-Tertiary boundary at Caravaca, Spain. Geology, 27, 723-726.

BELCHER, C.M., COLLINSON, M.E., SWEET, A.R., HILDEBRAND, A.R. & SCOTT, A.C. 2003. 'Fireball passes and nothing burns'-the role of thermal radiation in the K-T event: evidence from the charcoal record of North America. Geology, 31, 1061-1064.

BELCHER, C.M., COLLINSON, M.E., SWEEI, A.R., HILDEBRAND, A.R. & SCOTT, A.C. 2004. Reply to comment on 'Fireball passes and nothing burns'-the role of thermal radiation in the K-T event: evidence from the charcoal record of North America by Robertson et al. Geology, online forum.

BROOKS, E. 2002. http://snow.ag.uidaho.edu/Classes/Hydrogeology/ Examples/Snow/ snow_eb.html

CANNON, S.H., BIGIO, E.R. & MINE, E. 2001. A process for fire- related debrisflow initiation, Cerro Grande Fire, New Mexico. Hydmlogical Processes, 15, 3011-3024.

DEBANO, L.F. 2000. The role of fire and soil heating on water repellency in wildland environments: a review. Journal of Hydrology, 231, 195-206.

ELLIOTT, J.G. & PARKER, R.S. 2001. Developing a post-fire flood chronology and recurrence probability from alluvial stratigraphy in the Buffalo Creek Watershed, Colorado, U.S.A. Hydrological Processes, 15, 3039-3054.

FALCON-LANG, H. 1998. The impact of wildfire on an early Carboniferous coastal system. Palaeogeography; Palaeoclimatology, Palaeoecology, 139, 121-138.

GARDNER. JJ. & WHITLOCK, C. 2001. Charcoal accumulation following a recent fire in the Cascade Range, northwestern USA, and its relevance for firehistory studies. Holocene, 11, 541-549.

GILMOUR, I. & GUENTHER, F. 1988. The global Cretaceous-Tertiary fire: biomass or fossil carbon? In: Global Catastrophes in Earth History. Lunar and Planetary Institute Contributions, 673, 60-61.

GILMOUR, I., WOLBACH, W.S. & ANDERS, E. 1990. Major wildfires at the Cretaceous-Tertiary boundary. In: CLUBE, S.V.M. (ed.) Catastrophes and Evolution: Astronomical Foundations. Cambridge University Press, Cambridge, 195-213.

GILMOUR, I., SEPHTON, M.A. & MORGAN, J.V. 2003. Organic geochemistry of a hydrocarbon-rich calcarenite from the Chicxulub Scientific Drilling Program. Lunar and Planetary Science, XXXIV. abstract 1771 (CD-ROM).

HAMELL, H.B., BEEBE. R.F., INGERSOLL, A.P. ET AL. 1995. HST imaging of atmospheric phenomena created by the impact of comet Shoemaker-Levy 9. Science, 267, 1288-1296.

HILDEBRAND, A.R. 1993. The Cretaceous/Tertiary boundary impact (or the dinosaurs didn't have a chance). Journal of the Royal Astronomical Society of Canada, 87, 77-118.

HILDEBRAND, A.R. & BOYNTON, W.V. 1990. Proximal Cretaceous- Tertiary boundary impact deposits in the Caribbean. Science, 248, 843-847.

HILDEBRAND, A.R. & STANSBERRY, J.A. 1992. KVT boundary ejecta distribution predicts size and location of Chicxulub crater. Lunar and Planetary Science, XXIII, 537-538.

HILDHBRAND, A.R. PENFIELD, G.T., KRING, D.A., PILKINGTON, M., CAMARGO, A., JACOBSEN, S.B. & BOYNTON. W.W. 1991. Chicxulub crater: a possible Cretaceous-Tertiary boundary impact crater on the Yucatan Peninsula, Mexico. Geology, 19, 867-871.

HOTTON, C. 2002. Palynology of the Cretaceous-Tertiary boundary in central Montana: evidence for extraterrestrial impact as a cause of the terminal Cretaceous extinctions. In: HARTMAN, J.H., JOHNSON, K.R. & NICHOLS, DJ. (eds) The Hell Creek Formation of the Cretaceous- Tertiary Boundary in the Northern Great Plains. Geological Society of America, Special Papers, 361, 473-501.

INNES, J.B. & SIMMONS, I.G. 2000. Mid Holocene charcoal stratigraphy, fire history and palaeoecology at North Gill, North York Moors, UK. Palaeogeography, Palaeoclimatology, Palaeoecology, 164, 151-166.

IVANY, L.C. & SALAWITCH, RJ. 3993. Carbon isotopic evidence for biomass burning at the K-T boundary. Geology, 21, 487^90.

IZETT, G.A. 1990. The Cretaceous/Tertiary Boundary Interval, Raton Basin, Colorado and New Mexico, and its Content of Shock- Metamorphosed Minerals; Evidence Relevant to the KJT Boundary Impact- Extinction Theory. Geological Society of America, Special Papers, 249.

IZETT, G.A. & BOHOR, B.F. 1986. Microstratigraphy of continental sedimentary rocks in the Cretaceous-Tertiary boundary interval in the Western Interior of North America. Geological Society of America, Abstracts with Programs, 18. 644.

JONES, T.P., CHALONER, W.G. \& KUHLBUSCH, T.AJ. 1997. Proposed biogeological and chemical based terminology for fire altered plant matter. In: CLARK, J.S. ET AL. (ed.) Sediment Records of Biomass Burning and Global Change. NATO ASI Series, 151, 9-22.

KILLOPS, S.D. & KILLOPS, VJ. 1993. An Introduction to Organic Geochemistry. Longman Scientific, New York.

KOMAREK, E.V., SR., KOMAREK, B. & CARLYSLE, P. 1973. The Ecology of Smoke Particulates and Charcoal Residues from Forest and Grassland Fires: a Preliminary Atlas. Tall Timbers Research Station, TaHahasse, FL. Publications, 3.

KRING, D.A. & DURDA, D.D. 2002. Trajectories and distribution of material ejected from the Chicxulub impact crater: implications for post impact wildfires. Journal of Geophysical Research, 107, 6-22.

LAIRD, L.D. & CAMPBELL, I.D. 2000. High resolution palaeofire signals from Christina Lake, Alberta: a comparison of the charcoal signals extracted by two different methods. Palaeogeography, Palaeoclimatology, Palaeoecology, 164, 111-123.

MACLEOD. N., RAWSON, P.P., FOREY, P.L. ET AL. 1997. The Cretaceous-Tertiary biotic transition. Journal of the Geological Society, London, 153, 1-28.

MELOSH, HJ., SCHNEIDER, N.M., ZAHNLE, KJ. & LATHAM, D. 1990. Ignition of global wildfires at the Cretaceous/Tertiary boundary. Nature, 343, 251-254.

MEYER, G.A., PIERCE, J.L., Woon, S.H. & JI;LL, A.J.T. 2001. Fire, storms and erosional events in the Idaho Batholith. Hydrological Processes, 15, 3025-3038.

MOODY, J.A. & MARTIN, D.A. 2001. Hydrologie and sedimentologic response of two burned watersheds. US Geological Survey, Water- Resources Investigation Report, 01-4122.

NIC-HOLS, GJ. & JONES, T.P. 1992. Fusain in Carboniferous shallow marine sediments, Donegal, Ireland: the sedimentological effects of wildfire. Sedimentology, 39, 487-502.

NICHOLS, GJ., CRIPPS, J.A., COLLINSON, M.E. & SCOTT, A.C. 2000. Experiments in waterlogging and sedimentology of charcoal: results and implications. Palaeogeography, Palaeoclimatology, Palaeoecology, 164, 43-57.

NICHOLS, DJ. & JOHNSON, K.R. 2002. Palynology and microstratigraphy of Cretaceous-Tertiary boundary sections in southwestern North Dakota. In: HARTMAN, J.H., JOHNSON, K.R. & NICHOLS, DJ. (eds) The Hell Creek Formation of the Cretaceous- Tertiary Boundary in the Northern Great Plains. Geological Society of America, Special Papers, 361, 95-143.

OROS, D.R. & SIMONEIT, B.R.T. 2000. Identification and emission rates of molecular tracers in coal smoke particulate matter. Fuel, 79, 515-536.

OROS. D.R. & SIMONEIT, B.R.T. 2001a. Identification and emission factors of molecular tracers in organic aerosols from biomass burning. Part 1. Temperate climate conifers. Applied Geochemistry, 16, 1513-1544.

OROS, D.R. & SIMONEIT, B.RT. 200Ib. Identification and emission factors of molecular tracers in organic aerosols from biomass burning. Part 2. Decidous trees. Applied Geochemistry, 16, 1545- 1565.

ORTH, CJ., GILMORE, J.S., KNIGHT, J.D., PILLMORE, C.L., TSCHUDY, R.H. & FASSETT, J.E. 1981. An iridium abundance anomaly at the palynological Cretaceous-Tertiary boundary in northern New Mexico. Science. 214, 1341-1342.

PATTERSON, W.A. Ill, EDWAROS, KJ. & MAGUIRE, DJ. 1987. Microscopic charcoal as a fossil indicator of fire. Quaternary Science Reviews, 6, 3-23.

PETERSEN, HJ. 1998. Morphology, formation and palaeoenvironmental implications of naturally formed char particles in coals and carbonaceous mudstones. Fuel, 77, 1177-1183.

PILLMORE, C.L., TSCHUDY, R.H., ORTH, CJ., GILMORE. J.S. & KNIGHT, J.D. 1984. Geologic framework of nonmarine Cretaceous-Tertiary boundary sites, Raton Basin, New Mexico and Colorado. Science, 223, 1180-1183.

POPE, K.O. 2002. Impact dust not the cause of the Cretaceous- Tertiary mass extinction. Geology, 30, 99-102.

PYNE, SJ., ANDREWS, P.L. & LAVEN, R.D. 1996. lnimduciion to Wildlund Fire. Wiley, New York.

ROBERTSON, O.S., MCKENNA, M.C., TOON, O.B., HOPE, S. & LILLEGRAVEN, J.A. 2004a. Survival in the first hours of the Cenozoic. Geological Society of America Bulletin, 116, 760-768.

ROBKRTSON, O.S., MCKENNA, M.C., TOON, O.B., HOPE. S. & LILLEGRAVEN, J.A. 2004b. Comment on 'Fireball passes and nothing burns'-the role of thermal radiation in the K-T event: evidence from the charcoal record of North America. Geology, online forum.

SALA. M. & RUBIO, J.L. (EDS) 1994. Soil Erosion as a Consequence of Forest Fires. Geoforma, Longrono.

SCOTT, A.C. 1989. Observations on the nature and origin of fusain, international Journal of Coal Geology, 12, 443-475.

SCOTT, A.C. 2000. The Pre-Quaternary history of fire. Palaeogeographv, Palaeoclimatology, Palaeoecology, 164, 281-330.

SCOTT, A.C., CRIPPS, J.A., COLLINSON, M.E. & NICHOLS, GJ. 200Oa. The taphonomy of charcoal following a recent heathland fire and some implications for the interpretation of fossil charcoal deposits. Palaeogeography, Palaeoclimatology, Palaeoecology, 164, 1-32.

SCOTT, A.C., LOMAX, B.H., COLLINSON, M.E., UPCHURCH, G.R. & BEERLING, DJ. 200Ob. Fire across the K/T boundary: initial results from the Sugarite Coal, New Mexico, USA. Palaeogeography. Palaeoclimalology, Palaeoecology, 164,381-395.

SIMMS, D.L. & LAW, M. 1967. The ignition of wet and dry wood by radiation. Combustion and Flame, 11, 377-388.

SIMONEIT, B.R.T. 2002. Biomass burning-a review of organic tracers for smoke from incomplete combustion. Applied Geochemistry, 17, 129-162.

SWANSON, FJ. 1981. Fire and geomorphic processes. In: MOONEY, H.A., BONNICKSEN, T.H., CHRISTENSEN, N.L., LOTAN, J.E. & REINERS, W.A. (eds) Fire Regimes and Ecosystem Properties. USDA Forest Service General Technical Report, WO-26, 401^120.

SWEET, A.R. 2001. Plants, a yardstick for measuring the environmental consequences of the Cretaceous-Tertiary boundary event. Geoscience Canada, 28, 127-138.

SWFF.T. A.R., BRAMAN, D.R. & LERBF-KMO, J.F. 1999. Sequential palynological changes across the composite Cretaceous-Tertiary (K- T) boundary claystone and contiguous strata, western Canada and Montana, USA. Canadian Journal of Earth of Science, 36, 743-768.

TOON, O.B., ZAHNLE, K., MORRISON, D.. TURCO, R.P. & COVEY, C. 1997. Environmental perturbations caused by the impacts of asteroids and comets. Reviews of Geophysics, 35, 41-78.

TSHUDY, R.H., PILLMORE, C.L., ORTH, CJ., GILLMORE, J.S. & KNIGHT, J.D. 1984. Disruption of the terrestrial plant ecosystem at the Cretaceous-Tertiary boundary, Western Interior. Science, 225, 1030- 1032.

VENKATESAN. M.I. & DAHL, J. 1989. Further geochemical evidence for global fires at the Cretaceous-Tertiary boundary. Nature, 338, 57-60.

WHITLOCK, C. & MILLSPAUGH, S.H. 1996. Testing the assumptions of fire-history studies. The Holocene, 6, 7-15.

WOLBACH, W.S., LEWIS, R.S. & ANDERS, E. 1985. Cretaceous extinctions: evidence for wildfires and search for meteoritic material. Science, 230, 167-230.

WOLBACH, W.S., GILMOUR, I., ANDERS, E., ORTH, CJ. & BROOKS, R.R. 1988. Global fire at the Cretaceous-Tertiary boundary. Nature, 334, 665-669.

WOLBACH, W.S., GILMOUR, I. & ANDERS, E. 1990. Major wildfires at the Cretaceous/Tertiary boundary. In: SHARPTON, V.L. & WARD, P.O. (eds) Global Catastrophes in Earth History: an interdisciplinary conference on impacts, volcanism and mass mortality. Geological Society of America, Special Papers, 247, 391^tOO.

WOLFE, I.A. 1991. Palaeobotanical evidence for a June 'impact winter' at the Cretaceous/Tertiary boundary. Nature, 352, 420^122.

YOUNG, H.D. 1992. University Physics, 8th edn. Addison Wesley, Reading, MA, 437-438 and 1129-1131.

Received 13 August 2004; revised typescript accepted 3 February 2005.

Scientific editing by Jane Francis

C. M. BELCHER, M. E. COLLINSON & A. C. SCOTT

Department of Geology, Royal Holloway University of London, Egham TW20 OEX, UK (e-mail: c.belcher@gl.rhul.ac.nk)

Appendix: Polished blocks deposited with the Natural History Museum, London

Copyright Geological Society Publishing House Jul 2005

Source: Journal of the Geological Society

More News in this Category