Rapid Communication Permineralization process promotes preservation of Holocene macrofossil charcoal in soils GUILLAUME DE LAFONTAINE, 1,2 * PIERRE-LUC COUILLARD 1 and SERGE PAYETTE 1 1 NSERC Northern Research Chair, Centre d’e ´tudes nordiques, De ´partement de Biologie, Universite ´ Laval, Que ´bec, QC, Canada 2 INRA, UMR BIOGECO 1202, 69 Route d’Arcachon, 33610, Cestas, France Received 31 March 2011; Revised 16 June 2011; Accepted 19 June 2011 ABSTRACT: The use of macrofossil soil charcoal as a palaeoecological tool to reconstruct past vegetation, climate or fire history has gained much interest in recent years. Yet little is known about taphonomy of charcoal in soils. Here we assessed the putative loss of palaeoecological information due to charcoal fragmentation after burial. We found no significant loss of charcoal mass with time. Instead, we found a significant positive relationship between the mass of charcoal particles and their age. Permineralization of charcoal particles older than ca. 5200 a explained the increased charcoal mass with time in mineral soils. The permineralization process increased the density of charcoal particles (resulting in a twofold particles mass increase) and thus offers a protection against subsequent degradation. Our results suggest high stability of palaeoecological information from charcoal macrofossils buried in mineral soils at least over the Holocene timescale. Copyright # 2011 John Wiley & Sons, Ltd. KEYWORDS: 14 C AMS dating; charcoal analysis; macrofossil; permineralization; taphonomy. Introduction Macrofossil soil charcoal analysis has been used to reconstruct past vegetation, climate or fire for more than half a century (Salisbury and Jane, 1940; Godwin and Tansley, 1941). Early studies focused primarily on wood charcoal remains recovered from archaeological sites (Vernet, 1970, 1980) before sampling charred macrofossils directly from natural ecosystems (Filion, 1984; Payette and Gagnon, 1985). With the recent technical advances in accelerator mass spectrometry (AMS) radiocarbon dating, the use of macrofossil soil charcoal analysis has gained much interest, especially during the last decade (e.g. Gavin et al., 2003; Carnelli et al., 2004; Willis and van Andel, 2004; Asselin and Payette, 2005; Figueiral and Carcaillet, 2005; Talon et al., 2005; Sanborn et al., 2006; Hart et al., 2008; Auger and Payette, 2010; Fesenmyer and Christensen, 2010; Henry et al., 2010; Poschlod and Baumann, 2010; Talon, 2010; Touflan et al., 2010; de Lafontaine and Payette, 2011a, b). Today’s method includes botanical identification and direct radiocarbon AMS dating of macrofossil charcoal pieces extracted from soils (mineral or organic) in order to reconstruct stand-scale (i.e. in situ) past tree communities and fire history over the Holocene timescale. Macrofossil soil charcoal analysis might be essential to infer past local vegetation, climate or wildfire long-term history, at sites where other palaeoecolo- gical proxies are unavailable (Talon et al., 2005; de Lafontaine and Payette, 2011a, b). Despite this, little is known about taphonomy of charcoal in soils (but see Scott, 2010; Scott and Damblon, 2010; The ´ry- Parisot et al., 2010; for recent reviews on the topic). We stress that a sound knowledge of the taphonomical processes should help in better planning sampling and interpreting the results of future palaeoecological studies using soil charcoal analysis. The present study aims to explore some aspects of the taphonomy of charcoal recovered from mineral soils. Specifi- cally, we are interested in examining the putative loss of palaeoecological information due to charcoal degradation after burial. Soil charcoal appears highly resistant to post-depositional alteration (Czimczik et al., 2005; Preston and Schmidt, 2006). Charcoal stability is due to the existence of aromatic chemical structures produced during charcoal formation (Eckmeier et al., 2007). However, some charcoal degradation processes including physical fragmentation, biological and chemical alteration (e.g. Bird et al., 1999; Carcaillet, 2001; Glaser et al., 2002; Hockaday et al., 2006; Braadbaart et al., 2009; Ascough et al., 2011a,b) were identified. We expect more degradation of soil charcoal particles with increasing residence time in soils (as suggested by Preston and Schmidt, 2006; and references therein). The rationale for this hypothesis is that a longer charcoal residence time in soil implies more time for post- depositional alteration processes, causing increased charcoal degradation into smaller and lighter particles. Our objective is to assess the expected decrease of the charcoal particles mass with an increase in the age of the charcoal particles. Methods Our dataset includes a total of 264 AMS radiocarbon-dated charcoal particles (Ø  2 mm) spanning most of the Holocene period (from ca. 9600 cal. a BP to modern). These charcoal particles were recovered from 20 sites (plots: 50 m  10 m or 50 m  20 m) distributed across the closed-crown boreal forest of Que ´bec (see Supporting information, Fig. S1) and have already been used in macrofossil soil charcoal analyses in previous palaeoecological studies (de Lafontaine and Payette, 2011a, b; Couillard, 2011). Sampling sites are dominated either by balsam fir (Abies balsamea (L.) Mill.) or white spruce (Picea glauca (Moench) Voss), with white birch (Betula papyrifera Marsh.) and black spruce (Picea mariana (Mill.) BSP) as companion species in various densities. All sampled mineral soils were well-drained podzols. The sampled charred particles were all wood parts (no other charred plant organs were radiocarbon dated), botanically identified as Picea, Abies, Betula, or Pinus based on wood anatomy features recognized by episcopic optical microscopy. The similarity between extent of vegetation and Holocene charcoal records suggests that there was no significant historical change of the forest composition at the study sites (de Lafontaine and Payette, JOURNAL OF QUATERNARY SCIENCE (2011) 26(6) 571–575 ISSN 0267-8179. DOI: 10.1002/jqs.1529 Copyright ß 2011 John Wiley & Sons, Ltd. *Correspondence: G. De Lafontaine, INRA, UMR BIOGECO 1202, as above. E-mail: gdelafon@pierroton.inra.fr