Widely used charcoal analysis method in paleo studies involving NaOCl results in loss of charcoal formed below 400°C

Introduction

The quantification of charcoal in sediments has been used as an indicator of past fires since the early work of Iversen (1941). The production of highly stable polyaromatics during combustion (Bird and Ascough, 2012; Scott, 2010) and the common presence of charcoal in sediments (Novakov et al., 1997) makes it a useful proxy for the identification or quantification of past fires (Clark, 1984; Patterson et al., 1987). Charcoal analysis has progressed from quantification alongside pollen in palynological studies (e.g. Clark, 1982) to its own discipline with an overlapping group of methods. The recognition that ‘macrocharcoal’ (e.g. >150 µm) more accurately reflected local (within ≈10¹–10³ m) fires (Clark, 1988), particularly shifted attention to the quantification of larger charcoal particles (e.g. Clark et al., 1998; Lynch et al., 2004; Ohlson and Tryterud, 2000), and the common use of wet sieving (Long et al., 1998). Soon after (see, e.g. Carcaillet et al., 2001; Gardner and Whitlock, 2001) the use of an oxidant was added to the method to aid in removal of extraneous organic material.

Charcoal is the result of the incomplete combustion of biomass (Braadbaart and Poole, 2008; Clark, 1984; Mooney and Tinner, 2011; Patterson et al., 1987). Pyrolysis and oxidative heating thermally decompose and volatize organic material (Scott, 2010) and this results in structural and chemical changes to cellulose and lignin, eventually producing highly stable polyaromatics, which are often described as extremely resistant to environmental degradation (Ascough et al., 2010). This attributed stability and its common presence in sediments makes charcoal a useful proxy for identification or quantification of past fires (Patterson et al., 1987; Power et al., 2008). The recalcitrance of charcoal is affected by the condition of its organic source material. Wildfire charcoal is composed of material from a continuum of sources, from green plants (both woody and grasses) to long dead and rotting litter on the forest floor. Guo and Bustin (1998) suggested that wood subjected to fungal attack shrinks and cracks at lower temperatures and shorter heating durations than fresh organic material, likely caused by the fungal removal of cellulose or lignin. They argued fungal rotted wood requires less heat application to char than fresh wood. Naisse et al. (2013) also suggested that environmental conditions present during heating and the amount of decomposition the material is subjected to (e.g. mechanical weathering, freeze-thaw cycles, exposure to soil chemicals, microbial attack, etc.) prior to charring can affect its recalcitrance.

Several papers have discussed the use of various oxidants for charcoal analysis (e.g. Schlachter and Horn, 2010; Mooney and Tinner, 2011; Tsakiridou et al., 2021). Schlachter and Horn (2010) found that commercial charcoal was unaffected by oxidants, but stronger treatments of H₂O₂ resulted in lower particle counts in fossil charcoal, which they attributed to chemical digestion of particles that were not fully charred. Similarly, Tsakiridou et al. (2021) found this to be the case with modern charcoal (charred at 400°C and 800°C), but also noted that the treatment likely resulted in fragmentation and expansion of charcoal particles as well as digestion. Preliminary work by Ascough et al. (2011) noted that charcoal formed below 400°C is more reactive to oxidants than that produced at high temperature (Belcher et al., 2018), suggesting a mechanism for any charcoal loss.

The recalcitrance of charcoal to degradation is a function of the arrangement of C-C bonds in its chemical structure. Biomass subjected to higher temperatures, both in controlled experiments (Nguyen et al., 2010) and in wildfires (Alexis et al., 2010) results in charcoal with more stable structures, decreasing the number of reactive sites, thereby increasing its stability. In palaeoenvironmental experiments where bleach or other light oxidants are used to concentrate charcoal, there is a potential for the preferential loss of partially and lightly charred materials, possibly resulting in fire records biased towards higher intensity.

Here, we examined the effects that a frequently used method (4% sodium hypochlorite) for bleaching/deflocculation has on charcoal produced at temperatures commonly achieved in wildfires. We acknowledge that temperature is not a perfect metric to consider a wildfire, as for example, the most closely related aspects of a fire regime are fire intensity, the energy released by combustion of plant matter and fire severity, which describes the loss or decomposition of organic matter (Keeley, 2009). Nonetheless charcoal chemistry, which is at the crux of this work, is definitely affected by temperature (e.g. Belcher et al., 2018).

Methods

Charcoal was produced using a dead, small (diameter < 5 cm) air-dried (35°C fan-forced oven for 24 h) branch of Snow Gum (Eucalyptus pauciflora) wood sampled from Kosciuszko National Park in south-eastern Australia. Samples were charred at 250°C, 300°C, 350°C, 400°C, 450°C, 500°C, 600°C and 800°C under a Nitrogen gas flow (4.0 L/min) using a ramp rate of 30 min and a hold time of 1 h following the protocol described in McBeath et al. (2015) to ensure we achieved the chemical differences we seek to highlight.

The charcoal was treated using an often-standard method for the isolation of charcoal from sediments (Mooney and Tinner, 2011). Twenty replicates of each charring temperature were gently crushed to approximate charcoal size fractions common to charcoal analysis (>125 µm), then immersed for 24 h in 4% bleach (NaOCl). Charcoal area (mm²) of pre- and post-bleach treatments was quantified via image analysis using the software program ImageJ (Rueden et al., 2017). A single factor ANOVA was performed on the percentage change in area (∆A) of the pre- and post-treatment (250°C–800°C) to consider differences between charcoal temperatures.

Results

A summary of the mean and standard deviation of the percentage change in area (∆A) is presented in Table 1. Charcoal formed between 250°C and 400°C was nearly completely bleached in this treatment, and charcoal formed above 450°C (with greater variability amongst replicates), showed much more recalcitrance. ∆A of the samples charred at 450°C and 500°C indicated greater bleaching than those formed at higher temperatures (600°C–800°C). Figure 1 visually depicts the change in area before and after treatment, arranged by progressively higher maximum temperature of pyrolysis. Figure 2 presents a box and whisker plot of ∆A. Results of the single factor ANOVA demonstrated a significant difference between groups (F(7, 152) = 123.697, p = 1.85⁻⁵⁹). The average ∆A of charcoal formed ⩽400°C was 96% (σ 7.9) and the average ∆A for charcoal formed >400°C was 46% (σ 20.3).

OPEN IN VIEWER

Table 1.

Summary (mean and standard deviation) of findings presented in Figure 2, which shows the percent change ∆A of the treatment of charcoal with 4% sodium hypocholrite (bleach). Charcoal formed at temperatures ⩾400°C showed nearly complete bleaching. Charcoal formed at temperatures >400°C was much more recalcitrant to bleaching.

Temp (°C)	Reps	µ ∆Area	σ ∆Area
250	20	100	0
300	20	100	0
350	20	100	0.3
400	20	84	7
450	20	30	16
500	20	40	19
600	20	65	14
800	20	50	13

OPEN IN VIEWER

Figure 1.

Before and after images of charcoal subjected to 4% bleach treatments. Charcoal formed ⩾400°C shows clear evidence of bleaching in the post-treatment.

OPEN IN VIEWER

Figure 2.

Change in charcoal area expressed as percentage change in area of the pre- and post-treatment groups (250°C–800°C). There is a clear step-wise change in bleaching of low temperature (<400°C) and high temperature (⩾400°C) charcoal. The mean, first and third quartiles are displayed in the boxes. Whiskers represent the range of the data and dots represent outliers.

Discussion

This experiment shows a stepwise relationship between the maximum pyrolysis temperature of charcoal and its susceptibility or resistance to bleach treatments commonly used in charcoal analysis. Charcoal produced at temperatures below 400°C was nearly completely bleached. Charcoal produced at temperatures of 400°C and above displayed much greater variation between samples. The variability in bleaching within temperature groups could be the result of heterogeneity in the wood samples or due to variations in temperature across the heating surface in the oven. A recent study (Constantine et al., 2021) using the same production methods as this experiment applied FTIR and chemometrics to model charring intensity (CI) across species and temperatures. That study showed that there is variation in the infrared spectra within and between species charred in identical conditions. This could be the result of inconsistent carbonization, irregularity in the surface texture or result from variations in cellulose and lignin composition within charcoal samples.

McBeath et al. (2015) and Bird et al. (2015) have shown that the abundance of stable polycyclic aromatic carbon (SPAC), a component of charcoal produced in fire, increases relative to the total carbon content as pyrolysis temperature increases. McBeath et al. (2015) found a very steep change in the proportion of SPAC between about 400°C and 600°C, which is not dis-similar to our threshold. The potential loss of recalcitrance when subjected to long-term exposure to environment conditions (e.g. Naisse et al., 2013), also suggests that laboratory produced charcoal in this experiment may be more resistant to oxidation and may underrepresent the amount of charcoal that is bleached in sedimentary charcoal.

Ascough et al. (2010, 2011) also found that oxidation (using K₂Cr₂O₂) occurred more rapidly in material charred at lower temperatures (300°C), suggesting stable aromatic structures were smaller in size, with more irregularities in the C structure and a larger number of edges, increasing the number of O atoms subject to oxidation. They note that the loss of cellulose in material charred above 400°C left the wood more aromatic, further suggesting a threshold where the C in charcoal becomes stable and is relatively unaffected by oxidation (Ascough et al., 2011). The chemical decomposition of charcoal in soil (vs recalcitrance) also appears to decrease as its formation temperature increases (Bruun et al., 2008).

It has been recognized for some time that identifying fire events in sediment-based charcoal records is more reliable when fires are large and of high severity (Hennebelle et al., 2020; Higuera et al., 2007, 2010), suggesting a bias towards the preservation of high intensity fires in sediment catchments. The potential preferential removal of low temperature charcoal during sample preparation is likely to further exacerbate the weighting of the paleorecords towards high severity wildfires.

Given that many of the modern (macro-charcoal) palaeofire records were produced using oxidants in the preparation, it raises the distinct possibility that a number of fire records of the last few decades exclude any low severity fire. For example, the Global Paleofire Database (http://www.paleofire.org) is a repository of over 1200 sites covering all regions of the world between 70°N and 70°S. Of these, more than 300 sites used sieving on at least 935 charcoal records and are likely to have used an oxidant in the preparation. As per the Australian situation, our findings suggest that palaeofire records of the last few decades which have used an oxidant might be biased towards high intensity fire events. Even if the records (Daniau et al., 2012; Power et al., 2008) exclude any low severity fire, they are however relevant to debates concerning the recent apparent trends towards high severity fires. The potential limitations we have addressed concerning use of an oxidant could be of potential use in improving the interpretation of data in the Global Paleofire Database and could be useful in helping researchers choose laboratory methods for charcoal isolation.

Conclusions

Charcoal generally preserves well over long time periods in a variety of anoxic settings and its analysis is commonly used to reconstruct fire history. Our findings demonstrate that often-standard methods for charcoal analysis using oxidants almost completely bleach charcoal formed below 400°C and that charcoal only becomes resistant to oxidation above temperatures of between about 400°C and 450°C: notably this charcoal is also affected, but to a lesser extent. This suggests that charcoal formed in milder fire events might not be captured in the quantification of charcoal, as it is hard to distinguish optically or with image analysis after treatment with bleach.

These results have potentially serious ramifications for palaeofire records where an oxidant has been used, with a potential bias towards higher intensity/severity fires. This also suggests that future charcoal analyses may have to forgo oxidative treatments unless an effective alternative is found that does not preferentially remove charcoal formed under milder conditions.