Characteristics of soil organic matter 14 years after a wildfire: A pyrolysis-gas-chromatography mass spectrometry (Py-GC-MS) study

https://doi.org/10.1016/j.jaap.2020.104922Get rights and content

Highlights

  • After a decade of recovery, PyC markers were not sensitive to post-fire changes in SOM composition.

  • Long-term impacts of wildfire on SOM can be tracked using PyGCMS coupled with factor analysis.

  • SOM aromaticity increased with fire severity and depth in burned soils a decade after wildfire.

Abstract

Severe wildfires combust most above ground vegetation and detritus layers, altering the content and chemical composition of soil organic matter (SOM). To evaluate the lasting effects of wildfire on SOM and the recovery of burned soils, we sampled surface (Oa horizon) and mineral soils (0−5 and 5−15 cm depths) in unburned areas and areas burned at moderate and high severity 14 years after the 2002 Hayman Fire, in Colorado, USA. We characterized SOM using Pyrolysis Gas Chromatography Mass Spectrometry (Py-GC–MS) and identified 106 pyrolysates within eight chemical classes [aromatic hydrocarbon (ArH), carbohydrate (Carb), lignin compound (LgC), nitrogen containing compound (Ntg), polyaromatic hydrocarbon (PAH), phenol compound (PhC), saturated hydrocarbon (SaH), and unsaturated hydrocarbon (UnSaH)]. Burned soils had greater total quantified peak areas (TQPA) for the pyrogenic C indicator (PyC) benzene, compared to unburned soils; however, other common PyC markers were not abundant in burned relative to unburned soils. Factor analysis on the individual pyrolysates suggests that factors 1 and 2 correlated with pyrolysate aromaticity and hydrophobicity, respectively. Sample factor scores potentially suggest that SOM aromaticity increases with fire severity, though difference between moderate and high severity was slight. Factor analysis also indicates that the ratio of [ArH + Ntg] / [PhC + LgC] may serve as index of PyC content in SOM. This study shows that wildfire effects on SOM character may persist for more than a decade of ecosystem recovery and that Py-GC–MS coupled with factor analysis has utility for evaluating how disturbance alters SOM and PyC in complex environments.

Introduction

The number of severe wildfires has increased worldwide during the last two decades [1,2]. Not only do severe wildfires combust most vegetation, they also alter the thickness, extent and composition of soil organic layers [3,4]. A significant portion of combusted biomass is lost to the atmosphere as CO2 and volatile organic carbon, but large amounts of partly burned residues and pyrogenic organic carbon (PyC) are also integrated into the forest floor [5,6], where it may play an important role in regional and global C cycles. Since PyC is not completely inert, abiotic and biotic oxidation processes alter its structure and chemical properties [7,8]. Fine, charred organic particles are common in upper mineral soil layers (0−3 cm). However, oxidized PyC is relatively polar and easily transported as dissolved black carbon (DBC) both within soil horizons and downslope [7,9,10]. Through there has been considerable effort to characterize PyC distribution and composition shortly after wildfires [11,12], few studies have evaluated the long-term fate and vertical movement of PyC (e.g., > 1 decade after wildfire) within soil profiles. In fire-affected mineral soils, PyC is comprised of both labile (i.e., ∼ 0.03 yr mean resident time), and slowly degraded fractions (i.e, ∼ 40 yr mean resident time) [13]. The signal intensities of pyrolyzed materials including aromatic hydrocarbons (i.e., benzene and toluene) and polyaromatic hydrocarbons (i.e., pyrene) increase across a range of charcoal ages and demonstrate changes with time since burning [[14], [15], [16]].

Pyrolysis gas chromatography - mass spectrometry (Py-GC–MS) is a versatile analytical technique with wide utility for natural organic matter characterization. The approach has been used to study foliar litter decomposition [17] and determine the chemical composition of pyrogenic organic matter in biochar and charcoal [14,18,19]. Specifically, total quantified peak areas (TQPA) of the pyrolysates benzene, naphthalene and benzonitrile increase with pyrolysis temperature and have been used as indices of thermal alteration [14]. The abundance of three and four benzene ring polyaromatic hydrocarbons has been suggested as an indicator of the presence of charred or burnt materials [15]. Almendros, et al. [20] reported accumulations of condensed polyaromatic and lignin-derived, phenolic compounds following high- and moderate-intensity fires, respectively. Recently, the van Krevelen (VK) diagram, commonly used to visualize Fourier-transform ion cyclotron resonance mass spectrometer (FT-ICR MS) data [[21], [22], [23]], has been used to classify pyrolysates quantified by Py-GC–MS [16,20]. However, the classification is based on two dimensions of O/C and H/C atom ratios and without appropriate optimization and justification may lead to inconclusive and/or confounding results [22]. Most previous Py-GC–MS studies have been conducted under laboratory conditions or shortly (< 2 years) after wildfire, and few have examined the long-term effects of wildfire on soil PyC. Therefore, the effectiveness of the approach to evaluate changes as ecosystems recover from wildfire remains unknown. A recent statistical approach (i.e., ANOVA and factor analysis) was used to increase analysis efficiency, sensitivity and data throughout for tens of Py-GC–MS peaks [24]. In this study, we will compare different data analysis approaches to identify the best Py-GC–MS method for tracking PyC in natural environments.

The 2002 Hayman Fire was the largest fire (558 km2) in recent Colorado history [25]. A century of fire suppression increased fuel loads in these and many Colorado forests and a severe drought in 2002 created conditions that results in high severity wildfire across large areas of the burn. Post-fire forest vegetation recovery following the Hayman Fire has been extremely slow [26] and stream nutrients and carbon has remained elevated relative to unburned watersheds [27,28]. We hypothesis that the influence of the 2002 Hayman Fire on soil organic composition especially PyC is still evident 14 years after the fire. To study the long-term fate of PyC in 2016, we collected soil across fire severity gradient within and adjacent to the Hayman Fire (Fig S1) from three depths and analyzed them using Py-GC–MS, to: (1) evaluate the effect of fire on soil C and N content 14 years after fire; (2) examine the effectiveness of PyC markers as indicators of fire 14 years after fire; and (3) characterize the fire burned SOM using factor analysis on pyrolysates.

Section snippets

Soil sampling

The Hayman Fire burned ponderosa pine (Pinus ponderosa Dougl. ex Laws) and Douglas-fir [Pseudotsuga menziesii (Mirb.) Franco] forests of the lower montane elevation zone (1980–2750 m) [29]. The majority of the fire was underlain by Pikes Peak batholith, a coarse-grained biotite, and hornblende-biotite granite [30] that weathers to form weakly developed, excessively drained, coarse sandy loams (Typic Ustorthents) [31]. Soils were sampled in 2015 across a severity gradient that includes unburned

Total soil C and N content

The effect of the Hayman Fire remained evident 14 years after the fire with significantly lower C content (p < 0.05) in the O-horizon of area burned at high (13.7 ± 6.2 %) and moderate (18.4 ± 4.9 %) severity compared to unburned soils (28.6 ± 5.0 %). There was also a decreasing trend of soil N (1.04 ± 0.15, 0.95 ± 0.33 to 0.70 ± 0.30 %) in O-horizon for unburned, moderate, and high severity wildfire areas, though differences were not statistically significant. The O-horizon C/N ratio followed

Conclusion

Soil C declined after the Hayman Fire and the changes in SOM quantity and chemical characteristics have persisted for 14 years. However, due to vegetation recovery, litter inputs and decomposition during that period, traditional PyC markers were no longer sensitive to post-fire changes in SOM composition. Some individual pyrolysates (i.e., benzene and p-vinylguaiacol) and chemical classes (i.e., ArH) had utility as wildfire indicators but provided little information about fire severity or SOM

Author statements

Dr. Huan Chen was the first author of the manuscript. He was in charge of the data analysis and manuscript preparation, including peak assignment on the mass spectrometry, statistical analysis on the collected data, as well as the layouts of figures and tables.

Dr. Charles Rhoades was the co-author of the manuscript. He was in charge on the field sampling design and sampling collection. He reviewed and assisted manuscript preparation. He was the principal investigator of the Joint Fire Science

Declaration of Competing Interest

The authors report no declarations of interest.

Acknowledgements

We are grateful for financial support from the Joint Fire Sciences Program (JFSP#14-1-06-11), the US Forest Service, Rocky Mountain Research Station and Tim Fegel and Derek Pierson for assistance with field and laboratory work. This project was also based on work supported by NIFA/USDA under project number SC-1700517.

References (56)

  • J. Campo et al.

    Application of pyrolysis-gas chromatography/mass spectrometry to study changes in the organic matter of macro- and microaggregates of a Mediterranean soil upon heating

    J. Chromatogr. A

    (2011)
  • K. Nathe et al.

    Solid-state 13C NMR characterization of surface fire effects on the composition of organic matter in both soil and soil solution from a coniferous forest

    Geoderma

    (2017)
  • J.A. Gonzalez-Perez et al.

    The effect of fire on soil organic matter - a review

    Environ. Int.

    (2004)
  • S.R. Faria et al.

    Molecular characterization of wildfire impacts on organic matter in eroded sediments and topsoil in Mediterranean eucalypt stands

    Catena

    (2015)
  • J.M. De la Rosa et al.

    Use of pyrolysis/GC-MS combined with thermal analysis to monitor C and N changes in soil organic matter from a Mediterranean fire affected forest

    Catena

    (2008)
  • M.A. Jimenez-Gonzalez et al.

    Post-fire recovery of soil organic matter in a Cambisol from typical Mediterranean forest in Southwestern Spain

    Sci. Total Environ.

    (2016)
  • M. Lopez-Martin et al.

    Distribution of black carbon and black nitrogen in physical soil fractions from soils seven years after an intense forest fire and their role as C sink

    Sci. Total Environ.

    (2018)
  • A. Hilscher et al.

    Carbon and nitrogen degradation on molecular scale of grass-derived pyrogenic organic material during 28 months of incubation in soil

    Soil Biol. Biochem.

    (2011)
  • H. Knicker

    "Black nitrogen" - an important fraction in determining the recalcitrance of charcoal

    Org. Geochem.

    (2010)
  • D. Torres-Rojas et al.

    Nitrogen speciation and transformations in fire-derived organic matter

    Geochim. Cosmochim. Acta

    (2020)
  • A.L. Westerling et al.

    Warming and earlier spring increase western US forest wildfire activity

    Science

    (2006)
  • T. Schoennagel et al.

    Adapt to more wildfire in western North American forests as climate changes

    Proc. Natl. Acad. Sci. U. S. A.

    (2017)
  • M.I. Bird et al.

    The pyrogenic carbon cycle

    Annu. Rev. Earth. Pl. Sc.

    (2015)
  • C. Santin et al.

    Towards a global assessment of pyrogenic carbon from vegetation fires

    Global Change. Biol.

    (2016)
  • C.P. Ward et al.

    Insights into the complete and partial photooxidation of black carbon in surface waters

    Environ. Sci-Proc. Imp.

    (2014)
  • R.B. Abney et al.

    Pyrogenic carbon erosion after the Rim Fire, Yosemite National Park: the role of burn severity and slope

    J. Geophys. Res-Biogeo.

    (2019)
  • T. Dittmar et al.

    Continuous flux of dissolved black carbon from a vanished tropical forest biome

    Nat. Geosci.

    (2012)
  • C. Santin et al.

    Pyrogenic organic matter production from wildfires: a missing sink in the global carbon cycle

    Glob. Change Biol. Bioenergy

    (2015)
  • Cited by (9)

    View all citing articles on Scopus
    View full text