<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE article PUBLIC "-//NLM//DTD Journal Publishing with OASIS Tables v3.0 20080202//EN" "journalpub-oasis3.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:oasis="http://docs.oasis-open.org/ns/oasis-exchange/table" xml:lang="en" dtd-version="3.0">
  <front>
    <journal-meta><journal-id journal-id-type="publisher">BG</journal-id><journal-title-group>
    <journal-title>Biogeosciences</journal-title>
    <abbrev-journal-title abbrev-type="publisher">BG</abbrev-journal-title><abbrev-journal-title abbrev-type="nlm-ta">Biogeosciences</abbrev-journal-title>
  </journal-title-group><issn pub-type="epub">1726-4189</issn><publisher>
    <publisher-name>Copernicus Publications</publisher-name>
    <publisher-loc>Göttingen, Germany</publisher-loc>
  </publisher></journal-meta>
    <article-meta>
      <article-id pub-id-type="doi">10.5194/bg-15-2835-2018</article-id><title-group><article-title>A model based on Rock-Eval thermal analysis to quantify the size of the
centennially persistent organic carbon pool in temperate soils</article-title><alt-title>Quantifying persistent organic carbon in temperate soils</alt-title>
      </title-group><?xmltex \runningtitle{Quantifying persistent organic carbon in temperate soils}?><?xmltex \runningauthor{L. C\'{e}cillon et al.}?>
      <contrib-group>
        <contrib contrib-type="author" corresp="yes" rid="aff1 aff2">
          <name><surname>Cécillon</surname><given-names>Lauric</given-names></name>
          <email>lauric.cecillon@irstea.fr</email>
        <ext-link>https://orcid.org/0000-0001-7795-2382</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff3">
          <name><surname>Baudin</surname><given-names>François</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff4">
          <name><surname>Chenu</surname><given-names>Claire</given-names></name>
          
        <ext-link>https://orcid.org/0000-0001-9054-0489</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff4">
          <name><surname>Houot</surname><given-names>Sabine</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff2">
          <name><surname>Jolivet</surname><given-names>Romain</given-names></name>
          
        <ext-link>https://orcid.org/0000-0002-9896-3651</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff5">
          <name><surname>Kätterer</surname><given-names>Thomas</given-names></name>
          
        <ext-link>https://orcid.org/0000-0002-1751-007X</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff2">
          <name><surname>Lutfalla</surname><given-names>Suzanne</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff6">
          <name><surname>Macdonald</surname><given-names>Andy</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff4">
          <name><surname>van Oort</surname><given-names>Folkert</given-names></name>
          
        <ext-link>https://orcid.org/0000-0002-9948-154X</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff7">
          <name><surname>Plante</surname><given-names>Alain F.</given-names></name>
          
        <ext-link>https://orcid.org/0000-0003-0124-6187</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff3">
          <name><surname>Savignac</surname><given-names>Florence</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff2">
          <name><surname>Soucémarianadin</surname><given-names>Laure N.</given-names></name>
          
        <ext-link>https://orcid.org/0000-0002-9340-3268</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff2">
          <name><surname>Barré</surname><given-names>Pierre</given-names></name>
          
        <ext-link>https://orcid.org/0000-0002-0822-0556</ext-link></contrib>
        <aff id="aff1"><label>1</label><institution>Université Grenoble Alpes, Irstea, UR LESSEM, 38402
St-Martin-d'Hères, France</institution>
        </aff>
        <aff id="aff2"><label>2</label><institution>Laboratoire de Géologie de l'ENS, PSL Research University, CNRS
UMR 8538, 75005, Paris, France</institution>
        </aff>
        <aff id="aff3"><label>3</label><institution>Sorbonne Université-UPMC-Univ Paris 06, Institut des Sciences de
la Terre de Paris, 75005 Paris, France</institution>
        </aff>
        <aff id="aff4"><label>4</label><institution>AgroParisTech – INRA, UMR 1402 ECOSYS, 78850 Thiverval-Grignon,
France</institution>
        </aff>
        <aff id="aff5"><label>5</label><institution>Department of Ecology, Swedish University of Agricultural Sciences,
750 07 Uppsala, Sweden</institution>
        </aff>
        <aff id="aff6"><label>6</label><institution>Department of Sustainable Agriculture Sciences, Rothamsted Research,
Harpenden, Hertfordshire, AL5 2JQ, UK</institution>
        </aff>
        <aff id="aff7"><label>7</label><institution>Earth and Environmental Science, University of Pennsylvania,
Philadelphia, USA</institution>
        </aff>
      </contrib-group>
      <author-notes><corresp id="corr1">Lauric Cécillon (lauric.cecillon@irstea.fr)</corresp></author-notes><pub-date><day>9</day><month>May</month><year>2018</year></pub-date>
      
      <volume>15</volume>
      <issue>9</issue>
      <fpage>2835</fpage><lpage>2849</lpage>
      <history>
        <date date-type="received"><day>8</day><month>January</month><year>2018</year></date>
           <date date-type="rev-request"><day>7</day><month>February</month><year>2018</year></date>
           <date date-type="rev-recd"><day>18</day><month>April</month><year>2018</year></date>
           <date date-type="accepted"><day>25</day><month>April</month><year>2018</year></date>
      </history>
      <permissions>
        
        
      <license license-type="open-access"><license-p>This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this licence, visit <ext-link ext-link-type="uri" xlink:href="https://creativecommons.org/licenses/by/4.0/">https://creativecommons.org/licenses/by/4.0/</ext-link></license-p></license></permissions><self-uri xlink:href="https://bg.copernicus.org/articles/15/2835/2018/bg-15-2835-2018.html">This article is available from https://bg.copernicus.org/articles/15/2835/2018/bg-15-2835-2018.html</self-uri><self-uri xlink:href="https://bg.copernicus.org/articles/15/2835/2018/bg-15-2835-2018.pdf">The full text article is available as a PDF file from https://bg.copernicus.org/articles/15/2835/2018/bg-15-2835-2018.pdf</self-uri>
      <abstract>
    <p id="d1e231">Changes in global soil carbon stocks have considerable potential to influence
the course of future climate change. However, a portion of soil organic
carbon (SOC) has a very long residence time (<inline-formula><mml:math id="M1" display="inline"><mml:mo lspace="0mm">&gt;</mml:mo></mml:math></inline-formula> 100 years) and may not
contribute significantly to terrestrial greenhouse gas emissions during the
next century. The size of this persistent SOC reservoir is presumed to be
large. Consequently, it is a key parameter required for the initialization of
SOC dynamics in ecosystem and Earth system models, but there is considerable
uncertainty in the methods used to quantify it. Thermal analysis methods
provide cost-effective information on SOC thermal stability that has been
shown to be qualitatively related to SOC biogeochemical stability. The
objective of this work was to build the first quantitative model of the size
of the centennially persistent SOC pool based on thermal analysis. We used a
unique set of 118 archived soil samples from four agronomic experiments in
northwestern Europe with long-term bare fallow and non-bare fallow treatments
(e.g., manure amendment, cropland and grassland) as a sample set for which
estimating the size of the centennially persistent SOC pool is relatively
straightforward. At each experimental site, we estimated the average
concentration of centennially persistent SOC and its uncertainty by applying
a Bayesian curve-fitting method to the observed declining SOC concentration
over the duration of the long-term bare fallow treatment. Overall, the
estimated concentrations of centennially persistent SOC ranged from 5 to
11 g C kg<inline-formula><mml:math id="M2" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil (lowest and highest boundaries of four 95 %
confidence intervals). Then, by dividing the site-specific concentrations of
persistent SOC by the total SOC concentration, we could estimate the
proportion of centennially persistent SOC in the 118 archived soil samples
and the associated uncertainty. The proportion of centennially persistent SOC
ranged from 0.14 (standard deviation of 0.01) to 1 (standard deviation of
0.15). Samples were subjected to thermal analysis by Rock-Eval 6 that
generated a series of 30 parameters reflecting their SOC thermal stability
and bulk chemistry. We trained a nonparametric machine-learning algorithm
(random forests multivariate regression model) to predict the proportion of
centennially persistent SOC in new soils using Rock-Eval 6 thermal parameters
as predictors. We evaluated the model predictive performance with two
different strategies. We first used a calibration set (<inline-formula><mml:math id="M3" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M4" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 88) and a
validation set (<inline-formula><mml:math id="M5" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M6" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 30) with soils from all sites. Second, to test the
sensitivity of the model to pedoclimate, we built a calibration set with soil
samples from three out of the four sites (<inline-formula><mml:math id="M7" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M8" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 84). The multivariate
regression model accurately predicted the proportion of centennially
persistent SOC in the validation set composed of soils from all sites
(<inline-formula><mml:math id="M9" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M10" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.92, RMSEP <inline-formula><mml:math id="M11" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.07, <inline-formula><mml:math id="M12" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M13" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 30). The uncertainty of the
model predictions was quantified by a<?pagebreak page2836?> Monte Carlo approach that produced
conservative 95 % prediction intervals across the validation set. The
predictive performance of the model decreased when predicting the proportion
of centennially persistent SOC in soils from one fully independent site with
a different pedoclimate, yet the mean error of prediction only slightly
increased (<inline-formula><mml:math id="M14" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M15" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.53, RMSEP <inline-formula><mml:math id="M16" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.10, <inline-formula><mml:math id="M17" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M18" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 34). This model
based on Rock-Eval 6 thermal analysis can thus be used to predict the
proportion of centennially persistent SOC with known uncertainty in new soil
samples from different pedoclimates, at least for sites that have similar
Rock-Eval 6 thermal characteristics to those included in the calibration set.
Our study reinforces the evidence that there is a link between the thermal
and biogeochemical stability of soil organic matter and demonstrates that
Rock-Eval 6 thermal analysis can be used to quantify the size of the
centennially persistent organic carbon pool in temperate soils.</p>
  </abstract>
    </article-meta>
  </front>
<body>
      

<sec id="Ch1.S1" sec-type="intro">
  <title>Introduction</title>
      <p id="d1e383">Soils exert a key regulation on atmospheric greenhouse gas concentrations
on a decadal timescale through the net carbon source and sink status of their
organic carbon reservoir (Amundson, 2001; Eglin et al., 2010). However, a
portion of the soil organic carbon (SOC) reservoir may not contribute
significantly to the net exchange of <inline-formula><mml:math id="M19" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M20" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CH</mml:mi><mml:mn mathvariant="normal">4</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> between
atmosphere and land during the next century because its residence time
exceeds 100 years and its rate of carbon input is low (Trumbore, 1997; He et
al., 2016). The size of the centennially persistent SOC pool is presumed to
be large (i.e., between one- and two-thirds of total SOC) and dependent on
geochemical parameters such as soil texture and mineralogy (Buyanovsky and
Wagner, 1998a; Trumbore, 2009; Mills et al., 2014; Mathieu et al., 2015).
However, the amount of centennially persistent organic carbon in soils is
highly uncertain as it cannot be estimated accurately by current analytical
methods (Post and Kwon, 2000; von Lützow et al., 2007; Bruun et al.,
2008). Physicochemical procedures attempting to isolate SOC with a high
residence time from bulk SOC have proven unsatisfactory because of
indications that such fractions are a mixture of ancient and recent SOC (von
Lützow et al., 2007; Trumbore, 2009; Lutfalla et al., 2014). Even the
well-established radiocarbon (<inline-formula><mml:math id="M21" display="inline"><mml:mrow class="chem"><mml:msup><mml:mi/><mml:mn mathvariant="normal">14</mml:mn></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula>) analytical technique cannot
precisely determine the size of the centennially persistent SOC pool
(Schrumpf and Kaiser, 2015; Menichetti et al., 2016). The importance of
better information on the size of the centennially persistent SOC pool has
been emphasized recently (Soil Carbon Initiative, 2011; Bispo
et al., 2017; Bailey et al., 2018; Harden et al., 2018), stressing the need
for operational and standardized metrics or proxies to accurately quantify
SOC persistent at the centennial timescale. The general lack of information
on the size and turnover rate of measurable SOC pools hampers the
initialization of SOC pools in dynamic models, questioning their predictions
of the evolution of the global SOC reservoir (Falloon and Smith, 2000; Luo et
al., 2014; Feng et al., 2016; He et al., 2016; Sanderman et al., 2016). Luo
et al. (2016) and He et al. (2016) recently claimed that optimizing
parameter estimations with global datasets on SOC pools and fluxes was the
highest priority to reduce biases among Earth system models.</p>
      <p id="d1e420">In the past decade, the thermal stability of organic carbon has been proposed as
a good surrogate for its biogeochemical stability in litter and soils (e.g.,
Rovira et al., 2008; Plante et al., 2009; Gregorich et al., 2015). Several
studies using thermal analysis techniques, such as thermogravimetry and
differential scanning calorimetry with ramped combustion, have shown that the
fast-cycling SOC pool determined as the amount <inline-formula><mml:math id="M22" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> respired in
laboratory incubation experiments was thermally labile (Plante et al., 2011;
Leifeld and von Lützow, 2014; Campo and Merino, 2016). Recently, studies
using thermal analysis under an oxidative or inert (pyrolysis) reaction
atmosphere coupled with evolved gas analysis have shown a high and positive
correlation between the thermally stable SOC and persistent SOC determined
using <inline-formula><mml:math id="M23" display="inline"><mml:mrow class="chem"><mml:msup><mml:mi/><mml:mn mathvariant="normal">14</mml:mn></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula> measurements (Plante et al., 2013) and between thermally
stable SOC and mineral-associated SOC isolated by a classical SOC physical
fractionation scheme (Saenger et al., 2015). Using long-term bare fallow
(LTBF) soils kept free of vegetation for several decades (i.e., with
negligible carbon inputs), Barré et al. (2016) recently showed that
persistent SOC was low in energy and thermally stable. While there appear to
be strong qualitative links between the thermal and biogeochemical stability of
SOC, there is to date no established quantitative link between the size of
the persistent SOC pool and SOC thermal characteristics.</p>
      <p id="d1e446">The objective of this work was to design a reliable, routine method based on
a thermal analysis technique (Rock-Eval 6; RE6) to quantify centennially
persistent SOC in a range of temperate soil types. First, we compiled a set
of reference soil samples from four long-term agronomic experiments in
northwestern Europe with long-term bare fallow treatments. The SOC
concentration of LTBF treatments can be used to estimate the size of the
persistent SOC pool of a particular site, as proposed by Rühlmann (1999)
and Barré et al. (2010). Here, we refined estimates of the persistent SOC
concentration previously published by Barré et al. (2010) for the four
sites used in this study. We then used these values to estimate the
proportion of centennially persistent SOC in 118 archived soil samples (time
series) from LTBF and non-LTBF treatments of these four sites. The last step
consisted of analyzing these reference samples using RE6 thermal analysis
and building a multivariate regression model to relate RE6 information on SOC
thermal stability and bulk chemistry to the estimated proportion of
centennially persistent SOC. In this work, we aimed to deliver a model
based on thermal analysis with reliable prediction intervals around the
predicted values of the size of the centennially persistent SOC pool. We thus
focused on the uncertainty in the estimated<?pagebreak page2837?> proportion of centennially
persistent SOC and its propagation in the multivariate regression model.</p>
</sec>
<sec id="Ch1.S2">
  <title>Materials and methods</title>
<sec id="Ch1.S2.SS1">
  <title>Reference soil sample set with estimated size of the centennially
persistent SOC pool</title>
      <p id="d1e460">The reference soil sample set was built using samples from four long-term
agronomic experimental sites in northwestern Europe (Versailles and Grignon in France,
Rothamsted in the United Kingdom and Ultuna in Sweden; Supplement,
Table S1). Each of the four sites includes an LTBF
treatment, with bare fallow duration ranging from 48 years at Grignon to
79 years at Versailles. For all experimental sites, we also included non-LTBF
treatments that have increased or maintained their total SOC concentrations
over time or sustained smaller losses than the LTBF treatment. The selected
non-LTBF treatments included manure amendments (Versailles), straw or
composted straw amendments (Grignon), continuous grassland (Rothamsted) and
continuous cropland (Ultuna). Soil samples from each site and treatment have
been regularly collected and archived since the initiation of the
experiments. A total of 118 topsoil samples (0–20 to 0–25 cm of depth;
Table S1) were selected from the archives of LTBF and non-LTBF treatments to
build the reference sample set. Samples were selected from two or three field
replicate plots with a decadal frequency from the initiation of the
experiments up to 2007 (Grignon), 2008 (Versailles, Rothamsted) or 2009
(Ultuna) to obtain a sample set with the widest possible range of proportions
of centennially persistent SOC. The non-LTBF treatments and multiple sites
also added to the diversity of land-use, climate and parent material. For
each sample, total SOC concentration was measured by dry combustion with an
elemental analyzer (SOC<inline-formula><mml:math id="M24" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">EA</mml:mi></mml:msub></mml:math></inline-formula>, g C kg<inline-formula><mml:math id="M25" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil) after
the removal of soil carbonates
when necessary according to NF ISO 10694 (1995).</p>
      <p id="d1e484">Based on the decline in total SOC concentration over the duration of the LTBF
treatment, Barré et al. (2010) estimated the concentration of
centennially persistent SOC at each site using a Bayesian curve-fitting
method applied to each LTBF field replicate plot. Here, we refined those
site-specific estimates by (i) applying a similar Bayesian curve-fitting
method to combined SOC concentration data from all LTBF field replicate plots
of each site (four field replicate plots for Ultuna and Rothamsted, six field
replicate plots for Versailles and Grignon) and (ii) using new SOC
concentration data up to 2014 for Rothamsted and 2015 for Ultuna, increasing
their LTBF duration to 55 years for Rothamsted and 59 years for Ultuna.</p>
      <p id="d1e487">For each site, we assumed that the temporal evolution of LTBF SOC
concentration, <inline-formula><mml:math id="M26" display="inline"><mml:mrow><mml:mi mathvariant="italic">γ</mml:mi><mml:mo>(</mml:mo><mml:mi>t</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>, followed an exponential decay function:
            <disp-formula id="Ch1.E1" content-type="numbered"><mml:math id="M27" display="block"><mml:mrow><mml:mi mathvariant="italic">γ</mml:mi><mml:mo>(</mml:mo><mml:mi>t</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:mi>a</mml:mi><mml:msup><mml:mi>e</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mi>b</mml:mi><mml:mi>t</mml:mi></mml:mrow></mml:msup><mml:mo>+</mml:mo><mml:mi>c</mml:mi><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>
          <?xmltex \hack{\newpage}?><?xmltex \hack{\noindent}?>where <inline-formula><mml:math id="M28" display="inline"><mml:mrow><mml:mi mathvariant="italic">γ</mml:mi><mml:mo>(</mml:mo><mml:mi>t</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> (unit <inline-formula><mml:math id="M29" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> g C kg<inline-formula><mml:math id="M30" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil) is the LTBF SOC
concentration at time <inline-formula><mml:math id="M31" display="inline"><mml:mi>t</mml:mi></mml:math></inline-formula>, <inline-formula><mml:math id="M32" display="inline"><mml:mi>t</mml:mi></mml:math></inline-formula> (unit <inline-formula><mml:math id="M33" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> year) is the time under bare
fallow, and <inline-formula><mml:math id="M34" display="inline"><mml:mi>a</mml:mi></mml:math></inline-formula>, <inline-formula><mml:math id="M35" display="inline"><mml:mi>b</mml:mi></mml:math></inline-formula> and <inline-formula><mml:math id="M36" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula> are fitting parameters. Parameter <inline-formula><mml:math id="M37" display="inline"><mml:mi>a</mml:mi></mml:math></inline-formula>
(unit <inline-formula><mml:math id="M38" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> g C kg<inline-formula><mml:math id="M39" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil) corresponds to the amplitude of the decay
and <inline-formula><mml:math id="M40" display="inline"><mml:mi>b</mml:mi></mml:math></inline-formula> (unit <inline-formula><mml:math id="M41" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> yr<inline-formula><mml:math id="M42" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>) is the characteristic decay rate. The
parameter <inline-formula><mml:math id="M43" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula> (unit <inline-formula><mml:math id="M44" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> g C kg<inline-formula><mml:math id="M45" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil) represents a theoretically
inert portion of SOC. We considered this parameter as a site-specific metric
of the centennially persistent SOC concentration. In our view, the
centennially persistent SOC pool is not biogeochemically inert; it has a mean
age and a mean residence time that are both assumed to be high (e.g.,
centuries) though not precisely defined in this study. As a result, its
decline is minimal at the timescale of this study and we thus considered the
centennially persistent SOC concentration at each experimental site to be
constant. We used a Bayesian inference method to compute site-specific
estimates of centennially persistent SOC concentration and associated
uncertainties (detailed in the Sect. 2.3.1).</p>
      <p id="d1e697">The proportion of centennially persistent SOC (CP<inline-formula><mml:math id="M46" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>) in each
soil sample was then calculated as the ratio of the site-specific
CP<inline-formula><mml:math id="M47" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentration to the total SOC concentration of the sample:

                <disp-formula specific-use="align" content-type="numbered"><mml:math id="M48" display="block"><mml:mtable displaystyle="true"><mml:mtr><mml:mtd><mml:mstyle class="stylechange" displaystyle="true"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:msub><mml:mi mathvariant="normal">CP</mml:mi><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mi mathvariant="normal">proportion</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mo>[</mml:mo><mml:mi mathvariant="normal">sample</mml:mi><mml:mo>]</mml:mo><mml:mo>=</mml:mo></mml:mrow></mml:mtd></mml:mtr><mml:mlabeledtr id="Ch1.E2"><mml:mtd/><mml:mtd><mml:mstyle displaystyle="true" class="stylechange"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mspace width="1em" linebreak="nobreak"/><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:msub><mml:mi mathvariant="normal">CP</mml:mi><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mi mathvariant="normal">concentration</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>[</mml:mo><mml:mi mathvariant="normal">site</mml:mi><mml:mo>]</mml:mo></mml:mrow><mml:mrow><mml:mi mathvariant="normal">SOC</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mi mathvariant="normal">concentration</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>[</mml:mo><mml:mi mathvariant="normal">sample</mml:mi><mml:mo>]</mml:mo></mml:mrow></mml:mfrac></mml:mstyle><mml:mo>,</mml:mo></mml:mrow></mml:mtd></mml:mlabeledtr></mml:mtable></mml:math></disp-formula>

            where the unit of CP<inline-formula><mml:math id="M49" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentration [site] and SOC
concentration [sample] is g C kg<inline-formula><mml:math id="M50" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil. The CP<inline-formula><mml:math id="M51" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportions of five samples that were slightly above 1 were set to 1. In
these calculations, we assumed that at each site, the concentration of
CP<inline-formula><mml:math id="M52" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> was the same in the LTBF and non-LTBF treatments and was
constant with time. The details related to the estimation of the uncertainty
in the CP<inline-formula><mml:math id="M53" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion of each sample are reported in
Sect. 2.3.2.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F1" specific-use="star"><caption><p id="d1e840">Detail of the sequential pyrolysis and oxidation stages of Rock-Eval
6 (RE6) thermal analysis and of the five thermograms used to derive the 30
RE6 parameters reflecting SOC thermal stability and bulk chemistry. The grey
area under each RE6 thermogram represents the portion of the signal
unaffected by soil carbonates that was used to calculate RE6 temperature
parameters (modified after Behar et al., 2001; Saenger et al., 2013).</p></caption>
          <?xmltex \igopts{width=298.753937pt}?><graphic xlink:href="https://bg.copernicus.org/articles/15/2835/2018/bg-15-2835-2018-f01.png"/>

        </fig>

</sec>
<sec id="Ch1.S2.SS2">
  <title>Thermal analysis of soil samples by Rock-Eval 6</title>
      <p id="d1e855">The 118 soil samples from the reference set were analyzed with an RE6 Turbo
device (Vinci Technologies) using the basic setup for the analysis of soil
organic matter (Behar et al., 2001; Disnar et al., 2003). The RE6 technique
provided measurements from the sequential pyrolysis and oxidation of ca.
40 mg of finely ground (<inline-formula><mml:math id="M54" display="inline"><mml:mo lspace="0mm">&lt;</mml:mo></mml:math></inline-formula> 250 <inline-formula><mml:math id="M55" display="inline"><mml:mi mathvariant="normal">µ</mml:mi></mml:math></inline-formula>m) soil per sample (Fig. 1).
Volatile hydrocarbon effluents from pyrolysis were detected and quantified
with flame ionization detection (FID), while the evolution of CO and
<inline-formula><mml:math id="M56" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> gases was quantified by infrared detection during both the pyrolysis
and oxidation stages. Pyrolysis was carried out from 200 to 650 <inline-formula><mml:math id="M57" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C
in an <inline-formula><mml:math id="M58" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">N</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> atmosphere with a heating rate of 30 <inline-formula><mml:math id="M59" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C min<inline-formula><mml:math id="M60" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>,
while the oxidation was carried out from 300 to 850 <inline-formula><mml:math id="M61" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C in a
laboratory air atmosphere (with O<inline-formula><mml:math id="M62" display="inline"><mml:mrow><mml:msub><mml:mi/><mml:mn mathvariant="normal">2</mml:mn></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> with a heating rate of
20 <inline-formula><mml:math id="M63" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C min<inline-formula><mml:math id="M64" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>. The RE6 technique generated five thermograms per
sample (Fig. 1, i.e., volatile hydrocarbon (HC) effluents during pyrolysis,
<inline-formula><mml:math id="M65" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> during pyrolysis, <inline-formula><mml:math id="M66" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> during oxidation, CO during
pyrolysis and CO during oxidation).<?pagebreak page2838?> On average, the organic carbon yield of
the RE6 analysis was greater than 96.5 % of SOC<inline-formula><mml:math id="M67" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">EA</mml:mi></mml:msub></mml:math></inline-formula> for the
soils of the reference sample set
(SOC<inline-formula><mml:math id="M68" display="inline"><mml:msub><mml:mi/><mml:mrow><mml:mi mathvariant="normal">RE</mml:mi><mml:mn mathvariant="normal">6</mml:mn></mml:mrow></mml:msub></mml:math></inline-formula> <inline-formula><mml:math id="M69" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.966 <inline-formula><mml:math id="M70" display="inline"><mml:mo>×</mml:mo></mml:math></inline-formula> SOC<inline-formula><mml:math id="M71" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">EA</mml:mi></mml:msub></mml:math></inline-formula> <inline-formula><mml:math id="M72" display="inline"><mml:mo>+</mml:mo></mml:math></inline-formula> 0.403,
<inline-formula><mml:math id="M73" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M74" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.97, <inline-formula><mml:math id="M75" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M76" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 118).</p>
      <p id="d1e1074">For each RE6 thermogram, we determined the temperatures corresponding to each
incremental proportion of the amount of gases evolved during the pyrolysis
and oxidation stages. Upper temperatures of 850 <inline-formula><mml:math id="M77" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C (CO oxidation
thermogram), 650 <inline-formula><mml:math id="M78" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C (HC pyrolysis thermogram), 611 <inline-formula><mml:math id="M79" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C
(<inline-formula><mml:math id="M80" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> oxidation thermogram) and 560 <inline-formula><mml:math id="M81" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C (CO and <inline-formula><mml:math id="M82" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>
pyrolysis thermograms) were chosen for signal integration (Fig. 1), thereby
excluding any interference of soil carbonates (Behar et al., 2001). The thermal
decomposition of carbonates was indeed observed beyond 560 <inline-formula><mml:math id="M83" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C (CO
and <inline-formula><mml:math id="M84" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> pyrolysis thermograms) and 611 <inline-formula><mml:math id="M85" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C (<inline-formula><mml:math id="M86" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>
oxidation thermogram) for the site of Grignon (data not shown). For each RE6
thermogram, signal integration was performed on the offset-corrected
thermogram using sample-specific offset values estimated by the RE6 Turbo
device. For the three pyrolysis thermograms, signal integration started after
an isotherm step of 200 s at 200 <inline-formula><mml:math id="M87" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C. Finally, we retained five of
these temperature parameters per thermogram: <inline-formula><mml:math id="M88" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mn mathvariant="normal">10</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>, <inline-formula><mml:math id="M89" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mn mathvariant="normal">30</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>, <inline-formula><mml:math id="M90" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mn mathvariant="normal">50</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>,
<inline-formula><mml:math id="M91" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mn mathvariant="normal">70</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M92" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mn mathvariant="normal">90</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>, which respectively represent the temperatures
corresponding to the evolution of 10, 30, 50, 70 and 90 % of the amount
of evolved gases for each sample and for each of the five different
thermograms (HC, <inline-formula><mml:math id="M93" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> pyrolysis, <inline-formula><mml:math id="M94" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> oxidation, CO pyrolysis,
CO oxidation).</p>
      <?pagebreak page2839?><p id="d1e1264">For the HC pyrolysis thermogram we also determined three parameters
reflecting a proportion of thermally resistant or labile hydrocarbons: a
parameter representing the proportion of hydrocarbons evolved between 200 and
450 <inline-formula><mml:math id="M95" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C (thermo-labile hydrocarbons, TLHC index; modified from
Saenger et al., 2015), with the <inline-formula><mml:math id="M96" display="inline"><mml:mi>I</mml:mi></mml:math></inline-formula> index representing the preservation of thermally
labile immature hydrocarbons (after Sebag et al., 2016) and the <inline-formula><mml:math id="M97" display="inline"><mml:mi>R</mml:mi></mml:math></inline-formula> index
representing the proportion of hydrocarbons thermally stable at
400 <inline-formula><mml:math id="M98" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C (after Sebag et al., 2016). Those three RE6 parameters were
calculated as follows:

                <disp-formula specific-use="align" content-type="numbered"><mml:math id="M99" display="block"><mml:mtable displaystyle="true"><mml:mtr><mml:mtd><mml:mstyle class="stylechange" displaystyle="true"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mtext>TLHC index</mml:mtext><mml:mo>=</mml:mo></mml:mrow></mml:mtd></mml:mtr><mml:mlabeledtr id="Ch1.E3"><mml:mtd/><mml:mtd><mml:mstyle displaystyle="true" class="stylechange"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mspace linebreak="nobreak" width="1em"/><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:mi mathvariant="normal">Area</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mi mathvariant="normal">of</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mi mathvariant="normal">HC</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mi mathvariant="normal">pyrolysis</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mi mathvariant="normal">thermogram</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>[</mml:mo><mml:mn mathvariant="normal">200</mml:mn><mml:mtext>–</mml:mtext><mml:mn mathvariant="normal">450</mml:mn><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mtext>C</mml:mtext><mml:mo>]</mml:mo></mml:mrow><mml:mrow><mml:mi mathvariant="normal">Total</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mi mathvariant="normal">area</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mi mathvariant="normal">of</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mi mathvariant="normal">HC</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mi mathvariant="normal">pyrolysis</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mi mathvariant="normal">thermogram</mml:mi></mml:mrow></mml:mfrac></mml:mstyle><mml:mo>,</mml:mo></mml:mrow></mml:mtd></mml:mlabeledtr><mml:mtr><mml:mtd><mml:mstyle class="stylechange" displaystyle="true"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mi>I</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mtext>index</mml:mtext><mml:mo>=</mml:mo></mml:mrow></mml:mtd></mml:mtr><mml:mlabeledtr id="Ch1.E4"><mml:mtd/><mml:mtd><mml:mstyle class="stylechange" displaystyle="true"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mspace linebreak="nobreak" width="1em"/><mml:msub><mml:mi>log⁡</mml:mi><mml:mn mathvariant="normal">10</mml:mn></mml:msub><mml:mfenced close=")" open="("><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:mstyle displaystyle="true"><mml:mtable rowspacing="0.2ex" columnspacing="0.4em" class="substack"><mml:mtr><mml:mtd><mml:mrow><mml:mi mathvariant="normal">proportion</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mi mathvariant="normal">of</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mi mathvariant="normal">HC</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mi mathvariant="normal">pyrolysis</mml:mi></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mrow><mml:mi mathvariant="normal">thermogram</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mo>[</mml:mo><mml:mn mathvariant="normal">200</mml:mn><mml:mtext>–</mml:mtext><mml:mn mathvariant="normal">400</mml:mn><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mtext>C</mml:mtext><mml:mo>]</mml:mo></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:mstyle></mml:mrow><mml:mrow><mml:mstyle displaystyle="true"><mml:mtable columnspacing="0.4em" class="substack" rowspacing="0.2ex"><mml:mtr><mml:mtd><mml:mrow><mml:mi mathvariant="normal">proportion</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mi mathvariant="normal">of</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mi mathvariant="normal">HC</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mi mathvariant="normal">pyrolysis</mml:mi></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mrow><mml:mi mathvariant="normal">thermogram</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mo>[</mml:mo><mml:mn mathvariant="normal">400</mml:mn><mml:mtext>–</mml:mtext><mml:mn mathvariant="normal">460</mml:mn><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mtext>C</mml:mtext><mml:mo>]</mml:mo></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:mstyle></mml:mrow></mml:mfrac></mml:mstyle></mml:mfenced><mml:mo>,</mml:mo></mml:mrow></mml:mtd></mml:mlabeledtr><mml:mtr><mml:mtd><mml:mstyle class="stylechange" displaystyle="true"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mi>R</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mtext>index</mml:mtext><mml:mo>=</mml:mo></mml:mrow></mml:mtd></mml:mtr><mml:mlabeledtr id="Ch1.E5"><mml:mtd/><mml:mtd><mml:mstyle class="stylechange" displaystyle="true"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mspace linebreak="nobreak" width="1em"/><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:mi mathvariant="normal">Area</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mi mathvariant="normal">of</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mi mathvariant="normal">HC</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mi mathvariant="normal">pyrolysis</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mi mathvariant="normal">thermogram</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>[</mml:mo><mml:mn mathvariant="normal">400</mml:mn><mml:mtext>–</mml:mtext><mml:mn mathvariant="normal">650</mml:mn><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup><mml:mtext>C</mml:mtext><mml:mo>]</mml:mo></mml:mrow><mml:mrow><mml:mi mathvariant="normal">Total</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mi mathvariant="normal">area</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mi mathvariant="normal">of</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mi mathvariant="normal">HC</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mi mathvariant="normal">pyrolysis</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mi mathvariant="normal">thermogram</mml:mi></mml:mrow></mml:mfrac></mml:mstyle><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mlabeledtr></mml:mtable></mml:math></disp-formula>

            Using the HC pyrolysis thermogram, we determined a parameter reflecting SOC
bulk chemistry, the hydrogen index (HI, mg HC g<inline-formula><mml:math id="M100" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>C) that corresponds to
the quantity of pyrolyzed hydrocarbons relative to SOC<inline-formula><mml:math id="M101" display="inline"><mml:msub><mml:mi/><mml:mrow><mml:mi mathvariant="normal">RE</mml:mi><mml:mn mathvariant="normal">6</mml:mn></mml:mrow></mml:msub></mml:math></inline-formula>. Using
the CO and <inline-formula><mml:math id="M102" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> pyrolysis thermograms, we determined another parameter
reflecting SOC bulk chemistry, the oxygen index (OI<inline-formula><mml:math id="M103" display="inline"><mml:msub><mml:mi/><mml:mrow><mml:mi mathvariant="normal">RE</mml:mi><mml:mn mathvariant="normal">6</mml:mn></mml:mrow></mml:msub></mml:math></inline-formula>,
mgO<inline-formula><mml:math id="M104" display="inline"><mml:msub><mml:mi/><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:math></inline-formula> g<inline-formula><mml:math id="M105" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>C) corresponding to the oxygen yield as CO and <inline-formula><mml:math id="M106" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>
during the thermal pyrolysis of soil organic matter divided by the total SOC
(SOC<inline-formula><mml:math id="M107" display="inline"><mml:mrow><mml:msub><mml:mi/><mml:mrow><mml:mi mathvariant="normal">RE</mml:mi><mml:mn mathvariant="normal">6</mml:mn></mml:mrow></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> of the sample. The HI correlates with the elemental
H : C atomic ratio of SOC and the OI<inline-formula><mml:math id="M108" display="inline"><mml:msub><mml:mi/><mml:mrow><mml:mi mathvariant="normal">RE</mml:mi><mml:mn mathvariant="normal">6</mml:mn></mml:mrow></mml:msub></mml:math></inline-formula> correlates with the
elemental O : C atomic ratio of SOC (Espitalié et al., 1977).</p>
      <p id="d1e1671">Overall, we thus calculated for each soil sample a series of 30 RE6
parameters reflecting SOC thermal stability and bulk chemistry to be used in
subsequent statistical and modeling analyses.</p>
      <p id="d1e1675">The signal integration of the RE6 thermograms and the calculation of the RE6
temperature parameters were performed with R v.3.4.3 (R Core Team, 2017) and
the hyperSpec (Beleites and Sergo, 2014), pracma (Borchers, 2015) and
stringr (Wickham, 2015) packages.</p>
</sec>
<sec id="Ch1.S2.SS3">
  <title>Statistical analysis</title>
<sec id="Ch1.S2.SS3.SSS1">
  <?xmltex \opttitle{Bayesian inference of site-specific CP${}_{\mathrm{SOC}}$
concentrations and uncertainties}?><title>Bayesian inference of site-specific CP<inline-formula><mml:math id="M109" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
concentrations and uncertainties</title>
      <p id="d1e1699">At each site, the CP<inline-formula><mml:math id="M110" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentration was estimated as the model
parameter <inline-formula><mml:math id="M111" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula> of the exponential decay function described in Eq. (1). To
estimate the value of this parameter and assess its uncertainty, we sampled
the posterior probability density function (PDF) of the model parameters in
Eq. (1), which is given by Bayes' theorem as a function of the prior PDF
(i.e., what we know before collecting data) and the likelihood (i.e., how
likely is it to predict the data given a set of parameters). The posterior
PDF is the combination of our prior knowledge and the information carried
by the data, including measurement uncertainties. For a model vector
<inline-formula><mml:math id="M112" display="inline"><mml:mi mathvariant="bold-italic">m</mml:mi></mml:math></inline-formula> (containing the parameters <inline-formula><mml:math id="M113" display="inline"><mml:mi>a</mml:mi></mml:math></inline-formula>, <inline-formula><mml:math id="M114" display="inline"><mml:mi>b</mml:mi></mml:math></inline-formula> and <inline-formula><mml:math id="M115" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula>) and a data vector
<inline-formula><mml:math id="M116" display="inline"><mml:mi mathvariant="bold-italic">d</mml:mi></mml:math></inline-formula> of all the measurements of SOC concentrations, the posterior PDF,
<inline-formula><mml:math id="M117" display="inline"><mml:mrow><mml:mi>P</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">d</mml:mi><mml:mo>|</mml:mo><mml:mi mathvariant="bold-italic">m</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>, is <inline-formula><mml:math id="M118" display="inline"><mml:mrow><mml:mi>P</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">d</mml:mi><mml:mo>|</mml:mo><mml:mi mathvariant="bold-italic">m</mml:mi><mml:mo>)</mml:mo><mml:mo>∝</mml:mo><mml:mi>P</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">m</mml:mi><mml:mo>)</mml:mo><mml:mi>P</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">m</mml:mi><mml:mo>|</mml:mo><mml:mi mathvariant="bold-italic">d</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>, with <inline-formula><mml:math id="M119" display="inline"><mml:mrow><mml:mi>P</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">m</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> the prior PDF on the model parameters and
<inline-formula><mml:math id="M120" display="inline"><mml:mrow><mml:mi>P</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">m</mml:mi><mml:mo>|</mml:mo><mml:mi mathvariant="bold-italic">d</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> the likelihood.</p>
      <p id="d1e1845">We chose uniform PDFs for the model parameters <inline-formula><mml:math id="M121" display="inline"><mml:mi>a</mml:mi></mml:math></inline-formula>, <inline-formula><mml:math id="M122" display="inline"><mml:mi>b</mml:mi></mml:math></inline-formula> and <inline-formula><mml:math id="M123" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula> to be as
uninformative as possible. We use the Gaussian form of the likelihood
function, such as <inline-formula><mml:math id="M124" display="inline"><mml:mrow><mml:mi>P</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">m</mml:mi><mml:mo>|</mml:mo><mml:mi mathvariant="bold-italic">d</mml:mi><mml:mo>)</mml:mo><mml:mo>∝</mml:mo><mml:msup><mml:mi>e</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mfrac><mml:mn mathvariant="normal">1</mml:mn><mml:mn mathvariant="normal">2</mml:mn></mml:mfrac><mml:msup><mml:mfenced close=")" open="("><mml:mrow><mml:mi mathvariant="bold-italic">d</mml:mi><mml:mo>-</mml:mo><mml:mi mathvariant="italic">γ</mml:mi><mml:mfenced open="(" close=")"><mml:mi mathvariant="bold-italic">t</mml:mi></mml:mfenced></mml:mrow></mml:mfenced><mml:mi mathvariant="normal">T</mml:mi></mml:msup><mml:msubsup><mml:mi mathvariant="bold">C</mml:mi><mml:mi mathvariant="normal">d</mml:mi><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msubsup><mml:mfenced close=")" open="("><mml:mrow><mml:mi mathvariant="bold-italic">d</mml:mi><mml:mo>-</mml:mo><mml:mi mathvariant="italic">γ</mml:mi><mml:mfenced close=")" open="("><mml:mi mathvariant="bold-italic">t</mml:mi></mml:mfenced></mml:mrow></mml:mfenced></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, where <inline-formula><mml:math id="M125" display="inline"><mml:mi mathvariant="bold-italic">t</mml:mi></mml:math></inline-formula> is
the vector of all observation times and <inline-formula><mml:math id="M126" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold">C</mml:mi><mml:mi mathvariant="normal">d</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is the data
covariance matrix describing the uncertainties in the SOC measurements. We
consider a conservative standard deviation for SOC concentration data
(0.75 g C kg<inline-formula><mml:math id="M127" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil) estimated by Barré et al. (2010) for the same
soils. We use a Metropolis algorithm to draw 3 <inline-formula><mml:math id="M128" display="inline"><mml:mo>×</mml:mo></mml:math></inline-formula> 10<inline-formula><mml:math id="M129" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">4</mml:mn></mml:msup></mml:math></inline-formula> samples
from the posterior PDF with a burning phase of 3.7 <inline-formula><mml:math id="M130" display="inline"><mml:mo>×</mml:mo></mml:math></inline-formula> 10<inline-formula><mml:math id="M131" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">5</mml:mn></mml:msup></mml:math></inline-formula> steps.
We can then derive the mean and standard deviation for the parameter <inline-formula><mml:math id="M132" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula> from
the posterior PDF.</p>
</sec>
<sec id="Ch1.S2.SS3.SSS2">
  <?xmltex \opttitle{Estimating the uncertainty of CP${}_{\mathrm{SOC}}$ proportion in
each sample}?><title>Estimating the uncertainty of CP<inline-formula><mml:math id="M133" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion in
each sample</title>
      <p id="d1e2022">Based on our assessment of the uncertainties in SOC concentration data and
site-specific CP<inline-formula><mml:math id="M134" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentrations (see above), we propagated
these errors to estimate the uncertainty in the CP<inline-formula><mml:math id="M135" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion
in each soil sample. This was estimated by calculating the standard deviation
(SD) of the CP<inline-formula><mml:math id="M136" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion for each sample as
follows:

                  <disp-formula specific-use="align" content-type="numbered"><mml:math id="M137" display="block"><mml:mtable displaystyle="true"><mml:mtr><mml:mtd><mml:mstyle displaystyle="true" class="stylechange"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mi mathvariant="normal">SD</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mo>(</mml:mo><mml:msub><mml:mi mathvariant="normal">CP</mml:mi><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mi mathvariant="normal">proportion</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>[</mml:mo><mml:mi mathvariant="normal">sample</mml:mi><mml:mo>]</mml:mo><mml:mo>)</mml:mo><mml:mo>=</mml:mo></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mstyle class="stylechange" displaystyle="true"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mspace width="1em" linebreak="nobreak"/><mml:msub><mml:mi mathvariant="normal">CP</mml:mi><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mi mathvariant="normal">proportion</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mo>[</mml:mo><mml:mi mathvariant="normal">sample</mml:mi><mml:mo>]</mml:mo></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mstyle class="stylechange" displaystyle="true"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mspace linebreak="nobreak" width="1em"/><mml:mo>×</mml:mo><mml:msqrt><mml:mrow><mml:msup><mml:mfenced close=")" open="("><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:mi mathvariant="normal">SD</mml:mi><mml:mfenced close=")" open="("><mml:mrow><mml:msub><mml:mi mathvariant="normal">CP</mml:mi><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mi mathvariant="normal">concentration</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>[</mml:mo><mml:mi mathvariant="normal">site</mml:mi><mml:mo>]</mml:mo></mml:mrow></mml:mfenced></mml:mrow><mml:mrow><mml:msub><mml:mi mathvariant="normal">CP</mml:mi><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mi mathvariant="normal">concentration</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>[</mml:mo><mml:mi mathvariant="normal">site</mml:mi><mml:mo>]</mml:mo></mml:mrow></mml:mfrac></mml:mstyle></mml:mfenced><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:msqrt></mml:mrow></mml:mtd></mml:mtr><mml:mlabeledtr id="Ch1.E6"><mml:mtd/><mml:mtd><mml:mstyle displaystyle="true" class="stylechange"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mspace width="1em" linebreak="nobreak"/><mml:mover accent="true"><mml:mrow><mml:mo>+</mml:mo><mml:msup><mml:mfenced close=")" open="("><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:mi mathvariant="normal">SD</mml:mi><mml:mfenced open="(" close=")"><mml:mrow><mml:mi mathvariant="normal">SOC</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mi mathvariant="normal">concentration</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mo>[</mml:mo><mml:mi mathvariant="normal">sample</mml:mi><mml:mo>]</mml:mo></mml:mrow></mml:mfenced></mml:mrow><mml:mrow><mml:mi mathvariant="normal">SOC</mml:mi><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mi mathvariant="normal">concentration</mml:mi><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mo>[</mml:mo><mml:mi mathvariant="normal">sample</mml:mi><mml:mo>]</mml:mo></mml:mrow></mml:mfrac></mml:mstyle></mml:mfenced><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow><mml:mo mathvariant="normal">‾</mml:mo></mml:mover><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mlabeledtr></mml:mtable></mml:math></disp-formula></p>
</sec>
<sec id="Ch1.S2.SS3.SSS3">
  <?xmltex \opttitle{Statistical relationships between RE6 parameters and
CP${}_{\mathrm{SOC}}$ proportion}?><title>Statistical relationships between RE6 parameters and
CP<inline-formula><mml:math id="M138" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion</title>
      <p id="d1e2231">The reference sample set was randomly split into a calibration set
(<inline-formula><mml:math id="M139" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M140" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 88 samples) and a validation set (<inline-formula><mml:math id="M141" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M142" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 30 samples), each one
containing soils from all sites. The correlations between the 30 RE6
parameters and the CP<inline-formula><mml:math id="M143" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion were assessed with a
nonparametric Spearman's rank correlation test on the calibration set
(<inline-formula><mml:math id="M144" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M145" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 88). A principal component analysis (PCA) of the 30 centered and
scaled RE6 parameters was performed for the calibration set to (i) summarize
the variance of SOC thermal stability and bulk chemistry on a single
factorial map and (ii) illustrate the correlations among RE6 parameters.
Correlations between the CP<inline-formula><mml:math id="M146" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion in calibration soils
and their principal component scores were determined using Spearman's rank
correlation tests, and its relationships with the 30 RE6 parameters were
further illustrated by projecting the CP<inline-formula><mml:math id="M147" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion variable
in the PCA correlation plot. The RE6 data of the soils from the validation
set were projected on the same PCA factorial map to check that the validation
set was representative of the calibration set.</p>
</sec>
<sec id="Ch1.S2.SS3.SSS4">
  <?xmltex \opttitle{Random forests regression model to predict CP${}_{\mathrm{SOC}}$
proportion from RE6 parameters}?><title>Random forests regression model to predict CP<inline-formula><mml:math id="M148" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportion from RE6 parameters</title>
      <p id="d1e2320">A multivariate regression model was built to relate the CP<inline-formula><mml:math id="M149" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportion in the reference samples from the calibration set with soils from
all sites (response vector or dependent variable <inline-formula><mml:math id="M150" display="inline"><mml:mi mathvariant="bold-italic">y</mml:mi></mml:math></inline-formula>, <inline-formula><mml:math id="M151" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M152" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 88) to
their SOC thermal stability and bulk chemistry, summarized by a matrix of
predictor variables (<inline-formula><mml:math id="M153" display="inline"><mml:mi mathvariant="bold">X</mml:mi></mml:math></inline-formula>) made up of the 30 centered and scaled RE6
parameters. The nonparametric and nonlinear machine-learning technique of
random forests (RF, Breiman, 2001; Strobl et al., 2009) was used to build
this model. The random forests regression model was based on a forest of 1000
diverse regression trees made of splits and nodes. A random forests learning
algorithm combines bootstrap resampling and random variable selection. Each
of the 1000 regression trees was thus grown on a bootstrapped subset of the
calibration set (i.e., containing about two-thirds of “in-bag” calibration
samples) by randomly sampling 10 out of the 30 RE6 parameters as candidates
at each split of the tree and using a minimum size of terminal tree nodes of
five soil samples. The random forests<?pagebreak page2840?> regression model was then used to
predict the proportion of CP<inline-formula><mml:math id="M154" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> in the validation set
(<inline-formula><mml:math id="M155" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M156" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 30), a prediction corresponding to the mean of the predicted
values across the 1000 regression trees.</p>
      <p id="d1e2384">The performance of the random forests regression model for predicting
CP<inline-formula><mml:math id="M157" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion was assessed by statistics comparing the
RF-predicted vs. reference (estimated) values of the sample set. The
performance statistics were calculated on (i) the “out-of-bag” soil
samples of the calibration set and (ii) the soil samples of the validation
set. Out-of-bag samples are observations from the calibration set not
included in the learning sample set for a specific tree that can be used as a
“built-in” test set for calculating its prediction accuracy (Strobl et al.,
2009). The performance statistics included the coefficient of determination
(pseudo-<inline-formula><mml:math id="M158" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> for the calibration set or <inline-formula><mml:math id="M159" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> for the validation set) and
the root mean square error of calibration or prediction (RMSEC for the
calibration set or RMSEP for the validation set). The ratio of performance to
deviation (RPD) and the bias of the random forests regression model were
additionally calculated for the validation set. The relative importance
(i.e.,
ranking) of each of the 30 RE6 parameters for the prediction of the
proportion of CP<inline-formula><mml:math id="M160" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> in the RF regression model was computed as
the unscaled permutation accuracy (Strobl et al., 2009).</p>
      <p id="d1e2427">Additionally, the sensitivity of the RF regression model to pedoclimate was
assessed by examining its predictive performance for a calibration set based
on soils from three sites (Versailles, Grignon, Rothamsted, <inline-formula><mml:math id="M161" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M162" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 84) and
a fully independent validation set based on soils from a different
pedoclimate (Ultuna, <inline-formula><mml:math id="M163" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M164" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 34) but with similar RE6 thermal
characteristics to those of the calibration set (see Sect. 3.2). Soils from
Ultuna indeed have a higher clay content (from 11 to 20 % more clays)
and experience a lower mean annual temperature (from 4 to 5.2 <inline-formula><mml:math id="M165" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C
lower temperature) and a lower mean annual precipitation (from 116 to
179 mm lower precipitation) than the soils of the three other sites
(Table S1).</p>
</sec>
<sec id="Ch1.S2.SS3.SSS5">
  <title>Error propagation in the random forests regression model</title>
      <p id="d1e2474">Since our objective was to deliver a model based on thermal analysis with
reliable prediction intervals around the predicted values of the
CP<inline-formula><mml:math id="M166" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion, we estimated the prediction uncertainty of the
random forests model for new soil samples. We used a methodology recently
published by Coulston et al. (2016) to approximate prediction uncertainty for
random forests regression models and adapted it to explicitly take into
account the uncertainty in the reference values of the CP<inline-formula><mml:math id="M167" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion
(Eq. 6) that were used to build the model (Supplement,
Fig. S1).</p>
      <p id="d1e2495">Briefly, we sampled with replacement (i.e., bootstrapped) the calibration set
(<inline-formula><mml:math id="M168" display="inline"><mml:mrow><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mo>,</mml:mo><mml:mi mathvariant="bold">X</mml:mi></mml:mrow></mml:math></inline-formula>) 2000 times to obtain 2000 bootstrap samples
(<inline-formula><mml:math id="M169" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mrow><mml:mo>∗</mml:mo><mml:mi mathvariant="normal">b</mml:mi></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, <inline-formula><mml:math id="M170" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="bold">X</mml:mi><mml:mrow><mml:mo>∗</mml:mo><mml:mi mathvariant="normal">b</mml:mi></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) that were used
to parameterize 2000 random forest models (RF<inline-formula><mml:math id="M171" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>∗</mml:mo><mml:mi mathvariant="normal">b</mml:mi></mml:mrow></mml:msup></mml:math></inline-formula>). To
incorporate the uncertainty in the reference values of the CP<inline-formula><mml:math id="M172" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportion, each of the 2000 bootstrapped vectors (<inline-formula><mml:math id="M173" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mrow><mml:mo>∗</mml:mo><mml:mi mathvariant="normal">b</mml:mi></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) contained values of CP<inline-formula><mml:math id="M174" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion that were
randomly sampled from normal distributions with means and standard deviations
of the CP<inline-formula><mml:math id="M175" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion of the corresponding soil samples from
the calibration set (Eq. 6). For each bootstrap sample of the calibration
set, resampling discarded approximatively 37 % of the data
(<inline-formula><mml:math id="M176" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mrow><mml:mo>∗</mml:mo><mml:mo>-</mml:mo><mml:mi mathvariant="normal">b</mml:mi></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, <inline-formula><mml:math id="M177" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="bold">X</mml:mi><mml:mrow><mml:mo>∗</mml:mo><mml:mo>-</mml:mo><mml:mi mathvariant="normal">b</mml:mi></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>) that were
used for prediction. We obtained an error assessment dataset made up of 2000
vectors of observed (reference) values <inline-formula><mml:math id="M178" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mrow><mml:mo>∗</mml:mo><mml:mo>-</mml:mo><mml:mi mathvariant="normal">b</mml:mi></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>,
predicted values <inline-formula><mml:math id="M179" display="inline"><mml:mrow><mml:msup><mml:mover accent="true"><mml:mover accent="true"><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mo stretchy="false" mathvariant="normal">^</mml:mo></mml:mover><mml:mo mathvariant="normal">¯</mml:mo></mml:mover><mml:mrow><mml:mo>∗</mml:mo><mml:mo>-</mml:mo><mml:mi mathvariant="normal">b</mml:mi></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> (mean of the predictions
across 1000 regression trees for each observation) and
<inline-formula><mml:math id="M180" display="inline"><mml:mrow><mml:mi mathvariant="normal">var</mml:mi><mml:mo>(</mml:mo><mml:mover accent="true"><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mo mathvariant="normal" stretchy="false">^</mml:mo></mml:mover><mml:msup><mml:mo>)</mml:mo><mml:mrow><mml:mo>∗</mml:mo><mml:mo>-</mml:mo><mml:mi mathvariant="normal">b</mml:mi></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> (variance of the predictions across
1000 regression trees for each observation). For each observation of the 2000
bootstrap samples, we calculated a metric <inline-formula><mml:math id="M181" display="inline"><mml:mi mathvariant="italic">τ</mml:mi></mml:math></inline-formula> allowing us to scale between
<inline-formula><mml:math id="M182" display="inline"><mml:mrow><mml:mi mathvariant="normal">var</mml:mi><mml:mo>(</mml:mo><mml:mover accent="true"><mml:mi>y</mml:mi><mml:mo mathvariant="normal" stretchy="false">^</mml:mo></mml:mover><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>, which can be calculated for any soil sample by the
random forests regression model, and the squared prediction error (<inline-formula><mml:math id="M183" display="inline"><mml:mrow><mml:mi>y</mml:mi><mml:mo>-</mml:mo><mml:mover accent="true"><mml:mover accent="true"><mml:mi>y</mml:mi><mml:mo stretchy="false" mathvariant="normal">^</mml:mo></mml:mover><mml:mo mathvariant="normal">¯</mml:mo></mml:mover><mml:msup><mml:mo>)</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula>, which is only available for the reference sample set. The
metric <inline-formula><mml:math id="M184" display="inline"><mml:mi mathvariant="italic">τ</mml:mi></mml:math></inline-formula> was calculated as follows (Coulston et al., 2016):
              <disp-formula id="Ch1.E7" content-type="numbered"><mml:math id="M185" display="block"><mml:mrow><mml:mi mathvariant="italic">τ</mml:mi><mml:mo>=</mml:mo><mml:msqrt><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:mo>(</mml:mo><mml:mi>y</mml:mi><mml:mo>-</mml:mo><mml:mover accent="true"><mml:mover accent="true"><mml:mi>y</mml:mi><mml:mo stretchy="false" mathvariant="normal">^</mml:mo></mml:mover><mml:mo mathvariant="normal">¯</mml:mo></mml:mover><mml:msup><mml:mo>)</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow><mml:mrow><mml:mi mathvariant="normal">var</mml:mi><mml:mo>(</mml:mo><mml:mover accent="true"><mml:mi>y</mml:mi><mml:mo stretchy="false" mathvariant="normal">^</mml:mo></mml:mover><mml:mo>)</mml:mo></mml:mrow></mml:mfrac></mml:mstyle></mml:msqrt><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>
            A Monte Carlo approach was used to estimate <inline-formula><mml:math id="M186" display="inline"><mml:mover accent="true"><mml:mi mathvariant="italic">τ</mml:mi><mml:mo mathvariant="normal" stretchy="false">^</mml:mo></mml:mover></mml:math></inline-formula>, the 95th
percentile of all calculated <inline-formula><mml:math id="M187" display="inline"><mml:mi mathvariant="italic">τ</mml:mi></mml:math></inline-formula> values for all out-of-bag observations
of the 2000 bootstraps (Fig. S1). This <inline-formula><mml:math id="M188" display="inline"><mml:mover accent="true"><mml:mi mathvariant="italic">τ</mml:mi><mml:mo stretchy="false" mathvariant="normal">^</mml:mo></mml:mover></mml:math></inline-formula> value was such that
95 % of the predictions of the CP<inline-formula><mml:math id="M189" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion lie within
<inline-formula><mml:math id="M190" display="inline"><mml:mover accent="true"><mml:mi mathvariant="italic">τ</mml:mi><mml:mo mathvariant="normal" stretchy="false">^</mml:mo></mml:mover></mml:math></inline-formula> <inline-formula><mml:math id="M191" display="inline"><mml:mo>×</mml:mo></mml:math></inline-formula> <inline-formula><mml:math id="M192" display="inline"><mml:mrow><mml:mi mathvariant="normal">SD</mml:mi><mml:mo>(</mml:mo><mml:mover accent="true"><mml:mi>y</mml:mi><mml:mo mathvariant="normal" stretchy="false">^</mml:mo></mml:mover><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> of the true value of the
CP<inline-formula><mml:math id="M193" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion (i.e., 95 % prediction intervals). As
<inline-formula><mml:math id="M194" display="inline"><mml:mrow><mml:mi mathvariant="normal">SD</mml:mi><mml:mo>(</mml:mo><mml:mover accent="true"><mml:mi>y</mml:mi><mml:mo stretchy="false" mathvariant="normal">^</mml:mo></mml:mover><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>, the standard deviation of the predictions of the
CP<inline-formula><mml:math id="M195" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion across 1000 regression trees can be calculated
by the random forests regression model for any soil sample; this approach
allows for the calculation of 95 % prediction intervals on new soil
samples for which only <inline-formula><mml:math id="M196" display="inline"><mml:mi mathvariant="bold">X</mml:mi></mml:math></inline-formula> data (30 RE6 parameters) are available. We
calculated the 95 % prediction intervals
(<inline-formula><mml:math id="M197" display="inline"><mml:mover accent="true"><mml:mover accent="true"><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mo stretchy="false" mathvariant="normal">^</mml:mo></mml:mover><mml:mo mathvariant="normal">¯</mml:mo></mml:mover></mml:math></inline-formula> <inline-formula><mml:math id="M198" display="inline"><mml:mo>±</mml:mo></mml:math></inline-formula> <inline-formula><mml:math id="M199" display="inline"><mml:mover accent="true"><mml:mi mathvariant="italic">τ</mml:mi><mml:mo mathvariant="normal" stretchy="false">^</mml:mo></mml:mover></mml:math></inline-formula> <inline-formula><mml:math id="M200" display="inline"><mml:mo>×</mml:mo></mml:math></inline-formula> <inline-formula><mml:math id="M201" display="inline"><mml:mrow><mml:mi mathvariant="bold">SD</mml:mi><mml:mo>(</mml:mo><mml:mover accent="true"><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mo stretchy="false" mathvariant="normal">^</mml:mo></mml:mover><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>) for the
validation set (<inline-formula><mml:math id="M202" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M203" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 30) to examine whether those intervals included the
true (estimated) values of the CP<inline-formula><mml:math id="M204" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion. More details on the
procedure to approximate prediction uncertainty for random forests regression
models are provided in Coulston et al. (2016). We finally checked how the
error in the CP<inline-formula><mml:math id="M205" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion propagated into the random forests
regression model by (i) comparing the value of <inline-formula><mml:math id="M206" display="inline"><mml:mover accent="true"><mml:mi mathvariant="italic">τ</mml:mi><mml:mo stretchy="false" mathvariant="normal">^</mml:mo></mml:mover></mml:math></inline-formula> with or without
incorporating the uncertainty in the reference values of the CP<inline-formula><mml:math id="M207" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportion in the algorithm and (ii) by comparing the sizes of the 95 %
prediction intervals calculated for the validation soil samples with their
respective 95 % confidence intervals (determined by multiplying their
standard deviation calculated in Eq. 6 by 1.96).</p>

<?xmltex \floatpos{t}?><table-wrap id="Ch1.T1" specific-use="star"><caption><p id="d1e3013">Measured total SOC concentrations, estimated site-specific
CP<inline-formula><mml:math id="M208" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentrations and resulting CP<inline-formula><mml:math id="M209" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportions
in four long-term agronomic experimental sites used to generate calibration
and validation soil sample sets. Abbreviations: LTBF, long-term bare fallow;
min, minimum; max, maximum; SD, standard deviation; CI, confidence interval.</p></caption><oasis:table frame="topbot"><oasis:tgroup cols="5">
     <oasis:colspec colnum="1" colname="col1" align="left"/>
     <oasis:colspec colnum="2" colname="col2" align="left"/>
     <oasis:colspec colnum="3" colname="col3" align="right"/>
     <oasis:colspec colnum="4" colname="col4" align="right"/>
     <oasis:colspec colnum="5" colname="col5" align="right"/>
     <oasis:thead>
       <oasis:row>
         <oasis:entry colname="col1">Site</oasis:entry>
         <oasis:entry colname="col2">Treatments</oasis:entry>
         <oasis:entry colname="col3">SOC concentration</oasis:entry>
         <oasis:entry colname="col4">CP<inline-formula><mml:math id="M210" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentration</oasis:entry>
         <oasis:entry colname="col5">CP<inline-formula><mml:math id="M211" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2">(number of samples)</oasis:entry>
         <oasis:entry colname="col3">(g C kg<inline-formula><mml:math id="M212" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil)</oasis:entry>
         <oasis:entry colname="col4">(g C kg<inline-formula><mml:math id="M213" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil)</oasis:entry>
         <oasis:entry colname="col5">mean (min, max, SD)</oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2"/>
         <oasis:entry colname="col3">mean (min, max, SD)</oasis:entry>
         <oasis:entry colname="col4">mean (95 % CI)</oasis:entry>
         <oasis:entry colname="col5"/>
       </oasis:row>
     </oasis:thead>
     <oasis:tbody>
       <oasis:row>
         <oasis:entry colname="col1">Versailles</oasis:entry>
         <oasis:entry colname="col2">Manure (<inline-formula><mml:math id="M214" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M215" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 20)</oasis:entry>
         <oasis:entry colname="col3">27.9 (17.1, 45.5, 8.2)</oasis:entry>
         <oasis:entry colname="col4">6.22 (5.62–6.82)</oasis:entry>
         <oasis:entry colname="col5">0.24 (0.14, 0.36, 0.07)</oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2">LTBF (<inline-formula><mml:math id="M216" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M217" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 20)</oasis:entry>
         <oasis:entry colname="col3">10.5 (5.4, 19.7, 4.4)</oasis:entry>
         <oasis:entry colname="col4"/>
         <oasis:entry colname="col5">0.67 (0.32, 1.00, 0.24)</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Rothamsted</oasis:entry>
         <oasis:entry colname="col2">Grassland (<inline-formula><mml:math id="M218" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M219" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 8)</oasis:entry>
         <oasis:entry colname="col3">36.8 (31.8, 42.6, 4.8)</oasis:entry>
         <oasis:entry colname="col4">10.46 (9.93–10.99)</oasis:entry>
         <oasis:entry colname="col5">0.29 (0.25, 0.33, 0.04)</oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2">LTBF (<inline-formula><mml:math id="M220" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M221" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 12)</oasis:entry>
         <oasis:entry colname="col3">17.7 (9.7, 30.5, 7.5)</oasis:entry>
         <oasis:entry colname="col4"/>
         <oasis:entry colname="col5">0.68 (0.34, 1.00, 0.25)</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Ultuna</oasis:entry>
         <oasis:entry colname="col2">Cropland (<inline-formula><mml:math id="M222" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M223" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 23)</oasis:entry>
         <oasis:entry colname="col3">15.8 (12.4, 20.3, 2.2)</oasis:entry>
         <oasis:entry colname="col4">6.95 (5.22–8.68)</oasis:entry>
         <oasis:entry colname="col5">0.45 (0.34, 0.56, 0.06)</oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2">LTBF (<inline-formula><mml:math id="M224" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M225" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 11)</oasis:entry>
         <oasis:entry colname="col3">12.0 (9.1, 16.3, 2.4)</oasis:entry>
         <oasis:entry colname="col4"/>
         <oasis:entry colname="col5">0.60 (0.43, 0.76, 0.12)</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Grignon</oasis:entry>
         <oasis:entry colname="col2">Straw or composted straw (<inline-formula><mml:math id="M226" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M227" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 12)</oasis:entry>
         <oasis:entry colname="col3">12.9 (11.7, 14.2, 0.8)</oasis:entry>
         <oasis:entry colname="col4">7.12 (5.16–9.08)</oasis:entry>
         <oasis:entry colname="col5">0.55 (0.50, 0.60, 0.03)</oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2">LTBF (<inline-formula><mml:math id="M228" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M229" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 12)</oasis:entry>
         <oasis:entry colname="col3">11.8 (8.4, 14.7, 1.9)</oasis:entry>
         <oasis:entry colname="col4"/>
         <oasis:entry colname="col5">0.62 (0.48, 0.85, 0.11)</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry namest="col1" nameend="col2">Calibration set with soils from all sites (<inline-formula><mml:math id="M230" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M231" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 88) </oasis:entry>
         <oasis:entry colname="col3">18.0 (5.5, 45.5, 9.5)</oasis:entry>
         <oasis:entry colname="col4"/>
         <oasis:entry colname="col5">0.50 (0.14, 1.00, 0.21)</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry namest="col1" nameend="col2">Validation set with soils from all sites (<inline-formula><mml:math id="M232" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M233" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 30) </oasis:entry>
         <oasis:entry colname="col3">16.1 (5.4, 38.8, 7.9)</oasis:entry>
         <oasis:entry colname="col4"/>
         <oasis:entry colname="col5">0.53 (0.16, 1.00, 0.23)</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry namest="col1" nameend="col2">All samples (<inline-formula><mml:math id="M234" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M235" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 118) </oasis:entry>
         <oasis:entry colname="col3">17.5 (5.4, 45.5, 9.1)</oasis:entry>
         <oasis:entry colname="col4"/>
         <oasis:entry colname="col5">0.51 (0.14, 1.00, 0.21)</oasis:entry>
       </oasis:row>
     </oasis:tbody>
   </oasis:tgroup></oasis:table></table-wrap>

      <p id="d1e3481">The Bayesian inference method was performed with Python 2.7 and the PyMC
library (Patil et al., 2010). All other statistical analyses were performed
with R v.3.4.3 (R Core Team, 2017) and the factoextra package for running PCA
(Kassambara, 2015), the randomForest package for running the random forests
regression models (Liaw and Wiener, 2002) and the boot package for
bootstrapping (Davison and Hinkley, 1997; Canty and Ripley, 2015).</p>
</sec>
</sec>
</sec>
<?pagebreak page2841?><sec id="Ch1.S3">
  <title>Results</title>
<sec id="Ch1.S3.SS1">
  <?xmltex \opttitle{CP${}_{\mathrm{SOC}}$ concentration at each site and CP${}_{\mathrm{SOC}}$
proportion in reference soil samples}?><title>CP<inline-formula><mml:math id="M236" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentration at each site and CP<inline-formula><mml:math id="M237" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportion in reference soil samples</title>
      <p id="d1e3516">The Bayesian inference of the parameter <inline-formula><mml:math id="M238" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula> of the exponential decay function
(Eq. 1) yielded site-specific estimates of the CP<inline-formula><mml:math id="M239" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
concentration with 95 % confidence intervals (Eq. 1, Table 1, Fig. 2).
Estimated CP<inline-formula><mml:math id="M240" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentrations ranged from
6.22 g C kg<inline-formula><mml:math id="M241" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil at Versailles to 10.46 g C kg<inline-formula><mml:math id="M242" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil at
Rothamsted. The uncertainty in CP<inline-formula><mml:math id="M243" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentration was lower at
Rothamsted (standard deviation of 0.27 g C kg<inline-formula><mml:math id="M244" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil) and Versailles
(standard deviation of 0.31 g C kg<inline-formula><mml:math id="M245" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil) than at Ultuna (standard
deviation of 0.88 g C kg<inline-formula><mml:math id="M246" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil) and Grignon (standard deviation of
1.00 g C kg<inline-formula><mml:math id="M247" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil).</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F2"><caption><p id="d1e3628">Evolution of SOC concentration (g C kg<inline-formula><mml:math id="M248" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil) with time for
the bare fallow plots of each experimental site and representation of the
3 <inline-formula><mml:math id="M249" display="inline"><mml:mo>×</mml:mo></mml:math></inline-formula> 10<inline-formula><mml:math id="M250" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">4</mml:mn></mml:msup></mml:math></inline-formula> fitted exponential decay functions (Bayesian curve-fitting method) from which a site-specific CP<inline-formula><mml:math id="M251" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentration
(model parameter <inline-formula><mml:math id="M252" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula>) and its 95 % confidence interval were determined
(histogram in the upper right corner of each scatter plot). At each site, the
95 % confidence interval around the CP<inline-formula><mml:math id="M253" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentration was
determined as <inline-formula><mml:math id="M254" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M255" display="inline"><mml:mo>±</mml:mo></mml:math></inline-formula> 1.96 <inline-formula><mml:math id="M256" display="inline"><mml:mi mathvariant="normal">SD</mml:mi></mml:math></inline-formula>(<inline-formula><mml:math id="M257" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula>), where <inline-formula><mml:math id="M258" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula> is the model
parameter <inline-formula><mml:math id="M259" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula> in Eq. (1) and <inline-formula><mml:math id="M260" display="inline"><mml:mi mathvariant="normal">SD</mml:mi></mml:math></inline-formula>(<inline-formula><mml:math id="M261" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula>) is its standard deviation.</p></caption>
          <?xmltex \igopts{width=210.550394pt}?><graphic xlink:href="https://bg.copernicus.org/articles/15/2835/2018/bg-15-2835-2018-f02.png"/>

        </fig>

      <p id="d1e3748">Overall, the wide range in total SOC concentrations within and across sites
(from 5 to 46 g C kg<inline-formula><mml:math id="M262" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil; Table 1), combined with an assumed
constant CP<inline-formula><mml:math id="M263" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentration within each site, resulted in a
reference sample set with a wide spectrum of CP<inline-formula><mml:math id="M264" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportions
ranging from 0.14 to 1 (Eq. 2, Table 1). The uncertainty (standard deviation)
in the values of CP<inline-formula><mml:math id="M265" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion ranged from 0.01 to 0.15 for
the reference sample set (Eq. 6, Fig. S2). High
uncertainties were found for high values of CP<inline-formula><mml:math id="M266" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion
(i.e., samples with longer time periods under bare fallow treatment), with a
modulation by the site-specific CP<inline-formula><mml:math id="M267" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentration uncertainty
(Grignon <inline-formula><mml:math id="M268" display="inline"><mml:mo>&gt;</mml:mo></mml:math></inline-formula> Ultuna <inline-formula><mml:math id="M269" display="inline"><mml:mo>&gt;</mml:mo></mml:math></inline-formula> Versailles <inline-formula><mml:math id="M270" display="inline"><mml:mo>&gt;</mml:mo></mml:math></inline-formula> Rothamsted; Table 1), as
expected from Eq. (6) (Fig. S2).</p>

<?xmltex \floatpos{t}?><table-wrap id="Ch1.T2" specific-use="star"><caption><p id="d1e3834">Spearman's rank correlation coefficient test between the 30 RE6
parameters and the CP<inline-formula><mml:math id="M271" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion and variable importance
(ranking) of the 30 RE6 parameters to predict CP<inline-formula><mml:math id="M272" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion in
the random forests model based on Rock-Eval 6 thermal analysis (RE6-RF,
calibration sample set with soils from all sites, <inline-formula><mml:math id="M273" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M274" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 88). Symbols for
<inline-formula><mml:math id="M275" display="inline"><mml:mi>p</mml:mi></mml:math></inline-formula> values: <inline-formula><mml:math id="M276" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>*</mml:mo><mml:mo>*</mml:mo><mml:mo>*</mml:mo></mml:mrow></mml:msup></mml:math></inline-formula> <inline-formula><mml:math id="M277" display="inline"><mml:mi>p</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M278" display="inline"><mml:mo>&lt;</mml:mo></mml:math></inline-formula> 0.001; <inline-formula><mml:math id="M279" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>*</mml:mo><mml:mo>*</mml:mo></mml:mrow></mml:msup></mml:math></inline-formula> <inline-formula><mml:math id="M280" display="inline"><mml:mi>p</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M281" display="inline"><mml:mo>&lt;</mml:mo></mml:math></inline-formula> 0.01; <inline-formula><mml:math id="M282" display="inline"><mml:msup><mml:mi/><mml:mo>*</mml:mo></mml:msup></mml:math></inline-formula> <inline-formula><mml:math id="M283" display="inline"><mml:mi>p</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M284" display="inline"><mml:mo>&lt;</mml:mo></mml:math></inline-formula> 0.05; NS <inline-formula><mml:math id="M285" display="inline"><mml:mi>p</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M286" display="inline"><mml:mo>&gt;</mml:mo></mml:math></inline-formula> 0.05 <inline-formula><mml:math id="M287" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> not significant.</p></caption><oasis:table frame="topbot"><oasis:tgroup cols="4">
     <oasis:colspec colnum="1" colname="col1" align="left"/>
     <oasis:colspec colnum="2" colname="col2" align="right"/>
     <oasis:colspec colnum="3" colname="col3" align="center"/>
     <oasis:colspec colnum="4" colname="col4" align="right"/>
     <oasis:thead>
       <oasis:row>
         <oasis:entry colname="col1">RE6 parameter</oasis:entry>
         <oasis:entry colname="col2">Spearman's <inline-formula><mml:math id="M288" display="inline"><mml:mi mathvariant="italic">ρ</mml:mi></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col3"><inline-formula><mml:math id="M289" display="inline"><mml:mi>p</mml:mi></mml:math></inline-formula> value</oasis:entry>
         <oasis:entry colname="col4">Variable importance to predict</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2">with CP<inline-formula><mml:math id="M290" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col3"/>
         <oasis:entry colname="col4">CP<inline-formula><mml:math id="M291" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion in the RE6-RF</oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2">proportion</oasis:entry>
         <oasis:entry colname="col3"/>
         <oasis:entry colname="col4">regression model (rank)</oasis:entry>
       </oasis:row>
     </oasis:thead>
     <oasis:tbody>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M292" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">10</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">HC</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M293" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.36</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">20</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M294" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">30</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">HC</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M295" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.12</oasis:entry>
         <oasis:entry colname="col3">NS</oasis:entry>
         <oasis:entry colname="col4">30</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M296" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">50</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">HC</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.00</oasis:entry>
         <oasis:entry colname="col3">NS</oasis:entry>
         <oasis:entry colname="col4">29</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M297" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">70</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">HC</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.09</oasis:entry>
         <oasis:entry colname="col3">NS</oasis:entry>
         <oasis:entry colname="col4">21</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M298" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">90</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">HC</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.55</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">18</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M299" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">10</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">CO</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M300" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.21</oasis:entry>
         <oasis:entry colname="col3">*</oasis:entry>
         <oasis:entry colname="col4">7</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M301" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">30</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">CO</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M302" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.39</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">12</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M303" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">50</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">CO</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M304" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.33</oasis:entry>
         <oasis:entry colname="col3">**</oasis:entry>
         <oasis:entry colname="col4">14</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M305" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">70</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">CO</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M306" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.23</oasis:entry>
         <oasis:entry colname="col3">*</oasis:entry>
         <oasis:entry colname="col4">23</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M307" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">90</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">CO</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M308" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.09</oasis:entry>
         <oasis:entry colname="col3">NS</oasis:entry>
         <oasis:entry colname="col4">22</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M309" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">10</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.07</oasis:entry>
         <oasis:entry colname="col3">NS</oasis:entry>
         <oasis:entry colname="col4">13</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M310" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">30</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.71</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">4</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M311" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">50</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.80</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">1</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M312" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">70</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.78</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">2</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M313" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">90</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.66</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">17</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M314" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">10</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">CO</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">OX</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.50</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">5</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M315" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">30</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">CO</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">OX</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.71</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">16</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M316" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">50</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">CO</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">OX</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.64</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">28</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M317" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">70</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">CO</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">OX</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.42</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">27</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M318" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">90</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">CO</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">OX</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.14</oasis:entry>
         <oasis:entry colname="col3">NS</oasis:entry>
         <oasis:entry colname="col4">25</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M319" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">10</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">OX</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.72</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">10</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M320" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">30</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">OX</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.83</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">3</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M321" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">50</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">OX</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.82</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">6</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M322" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">70</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">OX</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.80</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">8</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M323" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">90</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">OX</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.54</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">19</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M324" display="inline"><mml:mi>I</mml:mi></mml:math></inline-formula> index</oasis:entry>
         <oasis:entry colname="col2">0.04</oasis:entry>
         <oasis:entry colname="col3">NS</oasis:entry>
         <oasis:entry colname="col4">26</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1"><inline-formula><mml:math id="M325" display="inline"><mml:mi>R</mml:mi></mml:math></inline-formula> index</oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M326" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.01</oasis:entry>
         <oasis:entry colname="col3">NS</oasis:entry>
         <oasis:entry colname="col4">24</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">TLHC index</oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M327" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.02</oasis:entry>
         <oasis:entry colname="col3">NS</oasis:entry>
         <oasis:entry colname="col4">15</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">HI</oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M328" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.78</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">9</oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">OI<inline-formula><mml:math id="M329" display="inline"><mml:msub><mml:mi/><mml:mrow><mml:mi mathvariant="normal">RE</mml:mi><mml:mn mathvariant="normal">6</mml:mn></mml:mrow></mml:msub></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col2">0.42</oasis:entry>
         <oasis:entry colname="col3">***</oasis:entry>
         <oasis:entry colname="col4">11</oasis:entry>
       </oasis:row>
     </oasis:tbody>
   </oasis:tgroup></oasis:table></table-wrap>

      <p id="d1e5094">The random splitting of the reference sample set generated calibration and
validation sample sets with similar mean values, range of values and standard
deviations for both total SOC concentration and CP<inline-formula><mml:math id="M330" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion
(Table 1).</p>
</sec>
<sec id="Ch1.S3.SS2">
  <?xmltex \opttitle{Relationships between RE6 parameters\hack{\break} and CP${}_{\mathrm{SOC}}$
proportion}?><title>Relationships between RE6 parameters<?xmltex \hack{\break}?> and CP<inline-formula><mml:math id="M331" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportion</title>
      <p id="d1e5124">The 30 RE6 parameters showed very different and even contrasting correlations
with the CP<inline-formula><mml:math id="M332" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion in the calibration set with soils from
all sites (Table 2). Most RE6 temperature parameters showed positive
correlations with the CP<inline-formula><mml:math id="M333" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion, with Spearman's <inline-formula><mml:math id="M334" display="inline"><mml:mi mathvariant="italic">ρ</mml:mi></mml:math></inline-formula>
above 0.8 for four of them (the RE6 temperature parameter corresponding to
50 % of <inline-formula><mml:math id="M335" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> gas evolution during the pyrolysis stage,
<inline-formula><mml:math id="M336" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">50</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>, and the RE6 temperature parameters
corresponding to 30, 50 and 70 % of <inline-formula><mml:math id="M337" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> gas evolution during the
oxidation stages <inline-formula><mml:math id="M338" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">30</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">OX</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>,
<inline-formula><mml:math id="M339" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">50</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">OX</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>,
<inline-formula><mml:math id="M340" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">70</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">OX</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>;
Table 2).</p>
      <p id="d1e5271">Conversely, five RE6 temperature parameters showed significant negative
correlations with the CP<inline-formula><mml:math id="M341" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion
(<inline-formula><mml:math id="M342" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">10</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">HC</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>,
<inline-formula><mml:math id="M343" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">10</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">CO</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>,
<inline-formula><mml:math id="M344" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">30</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">CO</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>,
<inline-formula><mml:math id="M345" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">50</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">CO</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>,
<inline-formula><mml:math id="M346" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">70</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">CO</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>; Table 2). The three RE6 parameters
reflecting a proportion of thermally resistant or labile hydrocarbons
(TLHC index, <inline-formula><mml:math id="M347" display="inline"><mml:mi>I</mml:mi></mml:math></inline-formula> index and the <inline-formula><mml:math id="M348" display="inline"><mml:mi>R</mml:mi></mml:math></inline-formula> index) showed no correlations with the
CP<inline-formula><mml:math id="M349" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion (Table 2). The two RE6 parameters reflecting SOC
bulk chemistry showed highly significant correlations with the
CP<inline-formula><mml:math id="M350" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion (Table 2), the HI being negatively correlated
and the OI<inline-formula><mml:math id="M351" display="inline"><mml:msub><mml:mi/><mml:mrow><mml:mi mathvariant="normal">RE</mml:mi><mml:mn mathvariant="normal">6</mml:mn></mml:mrow></mml:msub></mml:math></inline-formula> being positively correlated.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F3" specific-use="star"><caption><p id="d1e5430">Principal component analysis (PCA) of the 30 RE6 parameters of the
calibration sample set with soils from all sites (<inline-formula><mml:math id="M352" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M353" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 88).
<bold>(a)</bold> The scores of the calibration samples on the first two principal
components are represented in the factorial map, as are the projected
principal component scores of the validation samples (<inline-formula><mml:math id="M354" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M355" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 30; with
soils from all sites). A color scale is used to represent the
CP<inline-formula><mml:math id="M356" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion (determined using Eq. 2) in all samples.
<bold>(b)</bold> PCA loadings of the 30 RE6 parameters and projection of the
CP<inline-formula><mml:math id="M357" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion variable in the PCA correlation circle. Note
that despite some of the 30 RE6 parameters being significantly correlated, all
parameters are included in the analysis.</p></caption>
          <?xmltex \igopts{width=361.35pt}?><graphic xlink:href="https://bg.copernicus.org/articles/15/2835/2018/bg-15-2835-2018-f03.png"/>

        </fig>

      <?pagebreak page2842?><p id="d1e5493">The PCA of the centered and scaled RE6 parameters illustrates the
correlations among those 30 variables in the calibration set with soils from
all sites (Fig. 3). A continuum of CP<inline-formula><mml:math id="M358" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion values was
observed in the reference samples along the first two principal components
(Fig. 3a), and projecting the CP<inline-formula><mml:math id="M359" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion in the PCA
correlation circle further highlighted the relationships between this
variable and the 30 RE6 parameters (Fig. 3b). The CP<inline-formula><mml:math id="M360" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportion variable had a strongly negative projected loading score on PC1
(Fig. 3b), as well as negative projected loadings on PC2 (Fig. 3b) and PC3
(data not shown). The scores of the calibration soils on the first three
principal components were indeed significantly and negatively correlated with
the CP<inline-formula><mml:math id="M361" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion (<inline-formula><mml:math id="M362" display="inline"><mml:mi mathvariant="italic">ρ</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M363" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> <inline-formula><mml:math id="M364" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.61, <inline-formula><mml:math id="M365" display="inline"><mml:mi>p</mml:mi></mml:math></inline-formula> value <inline-formula><mml:math id="M366" display="inline"><mml:mo>&lt;</mml:mo></mml:math></inline-formula> 0.001
for PC1, <inline-formula><mml:math id="M367" display="inline"><mml:mi mathvariant="italic">ρ</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M368" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> <inline-formula><mml:math id="M369" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.49, <inline-formula><mml:math id="M370" display="inline"><mml:mi>p</mml:mi></mml:math></inline-formula> value <inline-formula><mml:math id="M371" display="inline"><mml:mo>&lt;</mml:mo></mml:math></inline-formula> 0.001 for PC2,
<inline-formula><mml:math id="M372" display="inline"><mml:mi mathvariant="italic">ρ</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M373" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> <inline-formula><mml:math id="M374" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.25, <inline-formula><mml:math id="M375" display="inline"><mml:mi>p</mml:mi></mml:math></inline-formula> value <inline-formula><mml:math id="M376" display="inline"><mml:mo>&lt;</mml:mo></mml:math></inline-formula> 0.05 for PC3) such that a large part
(82 %) of the variance in the 30 RE6 parameters was linked to the
CP<inline-formula><mml:math id="M377" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion in the calibration set.</p>
      <p id="d1e5649">The random splitting of the reference sample set generated calibration and
validation sample sets with similar RE6 thermal characteristics as
illustrated by their similar distribution on the factorial map of the first
two principal components of the PCA (Fig. 3a). Soils from the site of Grignon
(with carbonates) showed RE6 thermal characteristics different from the other
sites (Fig. 3a). Some soils from the sites of Rothamsted and Versailles with
high CP<inline-formula><mml:math id="M378" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportions also showed specific RE6 thermal
signatures (Fig. 3a). Conversely, all soils from the site of Ultuna showed
similarities regarding their RE6 thermal characteristics with certain soil
samples from other sites (Versailles and Rothamsted; Fig. 3a).</p>
</sec>
<sec id="Ch1.S3.SS3">
  <?xmltex \opttitle{Performance of the regression model using RE6 parameters to predict
CP${}_{\mathrm{SOC}}$ proportion}?><title>Performance of the regression model using RE6 parameters to predict
CP<inline-formula><mml:math id="M379" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion</title>
      <p id="d1e5677">The random forests regression model performed very well in predicting the
CP<inline-formula><mml:math id="M380" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion in the reference sample set using the 30 RE6
parameters as predictors (Fig. 4). Both performance statistics on the
calibration set with soils from all sites (pseudo-<inline-formula><mml:math id="M381" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M382" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.89,
RMSEC <inline-formula><mml:math id="M383" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.07, <inline-formula><mml:math id="M384" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M385" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 88) and on the validation set with soils from all
sites (<inline-formula><mml:math id="M386" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M387" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.92, RMSEP <inline-formula><mml:math id="M388" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.07, <inline-formula><mml:math id="M389" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M390" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 30) demonstrated the
good predictive power of the regression model based on RE6 thermal analysis.
The<?pagebreak page2843?> predictive performance of the random forests model based on RE6 thermal
analysis (RE6-RF) was altered when soil samples from a pedoclimate different
(site of Ultuna, <inline-formula><mml:math id="M391" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M392" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 34) from the calibration set (Versailles,
Rothamsted, Grignon), but with similar RE6 thermal characteristics (see
Sect. 3.2), were used for validation (Fig. 5). The coefficient of
determination of the model decreased (<inline-formula><mml:math id="M393" display="inline"><mml:mrow><mml:msup><mml:mi>R</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M394" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.53), yet its mean error
of prediction did not increase strongly (RMSEP <inline-formula><mml:math id="M395" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.10; Fig. 5).</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F4"><caption><p id="d1e5810">Performance of the random forests regression model based on
Rock-Eval 6 thermal analysis (RE6-RF) for predicting the CP<inline-formula><mml:math id="M396" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportion. The performance statistics on the calibration set with soils from
all sites (<inline-formula><mml:math id="M397" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M398" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 88) and on the validation set with soils from all sites
(<inline-formula><mml:math id="M399" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M400" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 30) of the RE6-RF multivariate regression model are shown.
Horizontal bars represent the estimated uncertainty (95 %
confidence
intervals) in the reference CP<inline-formula><mml:math id="M401" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion values of the
validation set calculated as CP<inline-formula><mml:math id="M402" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion
[sample] <inline-formula><mml:math id="M403" display="inline"><mml:mo>±</mml:mo></mml:math></inline-formula> 1.96 <inline-formula><mml:math id="M404" display="inline"><mml:mo>×</mml:mo></mml:math></inline-formula> <inline-formula><mml:math id="M405" display="inline"><mml:mi mathvariant="normal">SD</mml:mi></mml:math></inline-formula>(<inline-formula><mml:math id="M406" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="normal">CP</mml:mi><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> proportion
[sample]). Vertical bars represent the propagated errors (95 % prediction
intervals) in the RE6-RF-predicted CP<inline-formula><mml:math id="M407" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion values of the
validation sample set calculated as <inline-formula><mml:math id="M408" display="inline"><mml:mover accent="true"><mml:mover accent="true"><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mo stretchy="false" mathvariant="normal">^</mml:mo></mml:mover><mml:mo mathvariant="normal">¯</mml:mo></mml:mover></mml:math></inline-formula> <inline-formula><mml:math id="M409" display="inline"><mml:mo>±</mml:mo></mml:math></inline-formula> <inline-formula><mml:math id="M410" display="inline"><mml:mover accent="true"><mml:mi mathvariant="italic">τ</mml:mi><mml:mo stretchy="false" mathvariant="normal">^</mml:mo></mml:mover></mml:math></inline-formula> <inline-formula><mml:math id="M411" display="inline"><mml:mo>×</mml:mo></mml:math></inline-formula> <bold>SD</bold>(<inline-formula><mml:math id="M412" display="inline"><mml:mover accent="true"><mml:mi mathvariant="bold-italic">y</mml:mi><mml:mo mathvariant="normal" stretchy="false">^</mml:mo></mml:mover></mml:math></inline-formula>) (see Sect. 2.3.5), with a <inline-formula><mml:math id="M413" display="inline"><mml:mover accent="true"><mml:mi mathvariant="italic">τ</mml:mi><mml:mo stretchy="false" mathvariant="normal">^</mml:mo></mml:mover></mml:math></inline-formula> value of
2.12 (Fig. S1). Abbreviations: RMSEC, root mean square error of
calibration; RMSEP, root mean square error of prediction;
RPD, ratio of performance to deviation; SD, standard deviation.</p></caption>
          <?xmltex \igopts{width=236.157874pt}?><graphic xlink:href="https://bg.copernicus.org/articles/15/2835/2018/bg-15-2835-2018-f04.png"/>

        </fig>

      <p id="d1e5978">Propagating the estimated uncertainties in the values of CP<inline-formula><mml:math id="M414" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportion increased the size of the prediction intervals of the RE6-RF
regression model. Indeed, the value of <inline-formula><mml:math id="M415" display="inline"><mml:mover accent="true"><mml:mi mathvariant="italic">τ</mml:mi><mml:mo mathvariant="normal" stretchy="false">^</mml:mo></mml:mover></mml:math></inline-formula> increased from 1.83 to
2.12 when the uncertainty in CP<inline-formula><mml:math id="M416" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion was integrated into
the algorithm described in Sect. 2.3.5. The horizontal and vertical error
bars in Fig. 4 illustrate the global error propagation of the
CP<inline-formula><mml:math id="M417" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion estimates in the RE6-RF regression model for the
validation soil sample set. The values of the total width of the 95 %
confidence interval (reference estimations of CP<inline-formula><mml:math id="M418" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion;
horizontal error bars in Fig. 4) were 0.03 (minimum total width), 0.58
(maximum total width) and 0.24 (mean total width) for the soil samples of the
validation set (<inline-formula><mml:math id="M419" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M420" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 30). For the 95 % prediction intervals (RE6-RF
predictions of CP<inline-formula><mml:math id="M421" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion; vertical error bars in Fig. 4),
the uncertainties increased to 0.11 (minimum total width), 0.66 (maximum
total width) and 0.37 (mean total width). The 30 total 95 % prediction
intervals for RE6-RF predictions of the CP<inline-formula><mml:math id="M422" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion in the
validation set all included their respective reference estimation of the
CP<inline-formula><mml:math id="M423" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion (Fig. 4).</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F5"><caption><p id="d1e6072">Performance of the random forests regression model based on
Rock-Eval 6 thermal analysis (RE6-RF) for predicting the CP<inline-formula><mml:math id="M424" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportion for a calibration set based on soils from three sites
(Versailles, Rothamsted, Grignon; <inline-formula><mml:math id="M425" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M426" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 84) and a validation set based on
soils from a different pedoclimate but with similar RE6 thermal
characteristics than those of the calibration set (Ultuna; <inline-formula><mml:math id="M427" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M428" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 34).</p></caption>
          <?xmltex \igopts{width=236.157874pt}?><graphic xlink:href="https://bg.copernicus.org/articles/15/2835/2018/bg-15-2835-2018-f05.png"/>

        </fig>

      <p id="d1e6118">Out of the 30 RE6 parameters tested by the random forests model as possible
predictor variables of the CP<inline-formula><mml:math id="M429" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion in the calibration
set with soils from all sites, the RE6 temperature parameters corresponding
to 50 and 70 % of <inline-formula><mml:math id="M430" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> gas evolution during the pyrolysis stage
(<inline-formula><mml:math id="M431" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">50</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>,
<inline-formula><mml:math id="M432" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">70</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>) and to 30 % of <inline-formula><mml:math id="M433" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> gas
evolution during the oxidation stage (<inline-formula><mml:math id="M434" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">30</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">OX</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>)
showed the highest importance scores (based on permutation accuracy
importance calculations; Table 2). The eight most important RE6 parameters
for predicting the CP<inline-formula><mml:math id="M435" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion were temperature parameters
calculated on the five different RE6 thermograms (Table 2). The two RE6
parameters reflecting SOC bulk chemistry (OI<inline-formula><mml:math id="M436" display="inline"><mml:msub><mml:mi/><mml:mrow><mml:mi mathvariant="normal">RE</mml:mi><mml:mn mathvariant="normal">6</mml:mn></mml:mrow></mml:msub></mml:math></inline-formula> and HI) were of
medium importance to predict the CP<inline-formula><mml:math id="M437" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion, while the RE6
parameters reflecting a proportion of thermally resistant or labile
hydrocarbons (<inline-formula><mml:math id="M438" display="inline"><mml:mi>I</mml:mi></mml:math></inline-formula> index, <inline-formula><mml:math id="M439" display="inline"><mml:mi>R</mml:mi></mml:math></inline-formula> index and TLHC index) were of weak importance
(Table 2).</p>
</sec>
</sec>
<sec id="Ch1.S4" sec-type="conclusions">
  <title>Discussion</title>
<sec id="Ch1.S4.SS1">
  <?xmltex \opttitle{A unique soil sample set with accurate and contrasting values of
CP${}_{\mathrm{SOC}}$}?><title>A unique soil sample set with accurate and contrasting values of
CP<inline-formula><mml:math id="M440" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula></title>
      <?pagebreak page2844?><p id="d1e6290">Adding new SOC concentration data for Rothamsted (up to 2014) and Ultuna (up
to 2015) and combining SOC concentration data from all LTBF field replicate
plots of each site decreased the uncertainty in the site-specific estimates
of the CP<inline-formula><mml:math id="M441" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentration (Fig. 2) compared with the previous
estimations published by Barré et al. (2010). Indeed, the total width of
the 95 % confidence interval around the estimation of the site-specific
CP<inline-formula><mml:math id="M442" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentration slightly decreased from 1.4 to
1.2 g C kg<inline-formula><mml:math id="M443" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil at Versailles and from 4.96 to
3.92 g C kg<inline-formula><mml:math id="M444" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of
soil at Grignon; it also strongly decreased from 7.24 to 3.46 g C kg<inline-formula><mml:math id="M445" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil
at Ultuna and from 5.98 to 1.06 g C kg<inline-formula><mml:math id="M446" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil at Rothamsted (Table 1,
Fig. 2; Barré et al., 2010). The mean estimated values of the
CP<inline-formula><mml:math id="M447" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentration were marginally changed at Versailles (6.22
vs. 6.12 g C kg<inline-formula><mml:math id="M448" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil in Barré et al., 2010) and Grignon (7.12 vs.
6.80 g C kg<inline-formula><mml:math id="M449" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil in Barré et al., 2010), but strongly modified
(increased) at Ultuna (6.95 vs. 3.90 g C kg<inline-formula><mml:math id="M450" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil in Barré et al.,
2010) and Rothamsted (10.46 vs. 2.72 g C kg<inline-formula><mml:math id="M451" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil in Barré et al.,
2010; Table 1).</p>
      <p id="d1e6417">Our results obtained under four contrasting pedoclimates of northwestern
Europe indicate a minimum value of 5 g C kg<inline-formula><mml:math id="M452" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil (the lowest boundary of
the four 95 % confidence intervals; Table 1) and a maximum value of
11 g C kg<inline-formula><mml:math id="M453" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil (the highest boundary of the four 95 % confidence
intervals; Table 1) for CP<inline-formula><mml:math id="M454" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentration in topsoils (0–20 to
0–25 cm of depth). These estimates are close, yet below the CP<inline-formula><mml:math id="M455" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
concentration value of 12 g C kg<inline-formula><mml:math id="M456" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil estimated by Buyanovsky and
Wagner (1998b) for the topsoil (0–20 cm of depth) of the Sanborn long-term
(100 years) agronomic experiment (Columbia, Missouri, USA). Our estimates of
topsoil CP<inline-formula><mml:math id="M457" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentration are also well below the value of
16 g C kg<inline-formula><mml:math id="M458" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil estimated by Franko and Merbach (2017) in the topsoil
(0–30 cm of depth) of the long-term (28 years) bare fallow experiment of Bad
Lauchstädt (central Germany). The soil type in Bad Lauchstädt (Haplic
Chernozem) and its high concentration of slow-cycling black carbon (estimated
at 2.5 g C kg<inline-formula><mml:math id="M459" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> of soil; Brodowski et al., 2007) may explain this
difference, as well as the relatively short time period under bare fallow
(higher uncertainty in the inferred CP<inline-formula><mml:math id="M460" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentration).</p>
      <?pagebreak page2845?><p id="d1e6517">Among the wide range of CP<inline-formula><mml:math id="M461" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportions (0.14 to 1) of our
reference sample set, high values of CP<inline-formula><mml:math id="M462" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportions (<inline-formula><mml:math id="M463" display="inline"><mml:mo lspace="0mm">&gt;</mml:mo></mml:math></inline-formula> 0.6)
were obtained only for soils that had been under bare fallow for a long
period of time: after several years or decades with negligible C inputs and
sustained SOC decomposition (Table 1). Similarly, the low values of
CP<inline-formula><mml:math id="M464" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportions (<inline-formula><mml:math id="M465" display="inline"><mml:mo lspace="0mm">&lt;</mml:mo></mml:math></inline-formula> 0.25) of our reference sample set were
obtained for soils without vegetation but receiving high amounts of manure
amendments at Versailles (Table 1). It could be argued that the CP<inline-formula><mml:math id="M466" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportion values obtained for bare soils with or without organic matter
amendments may not be representative of the CP<inline-formula><mml:math id="M467" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportions of
soils under conventional management practices. However, it is interesting to
note that the soils of the reference sample set with vegetation and
experiencing classical management practices (grassland at Rothamsted,
cropland at Ultuna) also showed a wide range of CP<inline-formula><mml:math id="M468" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportions, from 0.25 to 0.56 (Table 1). Moreover, other studies have shown
the high variability of CP<inline-formula><mml:math id="M469" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportions in soils. For instance,
Falloon et al. (1998) listed a series of published values of
CP<inline-formula><mml:math id="M470" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportions ranging from 0.13 to 0.59. More recently, Mills
et al. (2014) published a large dataset of CP<inline-formula><mml:math id="M471" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportions in
uncultivated topsoils (ca. 15 cm of depth). They estimated CP<inline-formula><mml:math id="M472" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportions using a global dataset of topsoil radiocarbon (<inline-formula><mml:math id="M473" display="inline"><mml:mrow class="chem"><mml:msup><mml:mi/><mml:mn mathvariant="normal">14</mml:mn></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula>)
data and a steady-state SOC turnover model with a fixed mean residence time
of 1000 years for persistent SOC. Their estimates of CP<inline-formula><mml:math id="M474" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportions varied greatly from 0.03 to 0.98 (mean <inline-formula><mml:math id="M475" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.48, standard
deviation <inline-formula><mml:math id="M476" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 0.22, <inline-formula><mml:math id="M477" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula> <inline-formula><mml:math id="M478" display="inline"><mml:mo>=</mml:mo></mml:math></inline-formula> 232; soils with inconsistent negative modeled
SOC pools values were removed), with significantly higher CP<inline-formula><mml:math id="M479" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportions in non-forest than in forest uncultivated ecosystems (Mills et
al., 2014).</p>
      <p id="d1e6685">Overall, those combined results illustrate the wide range of
CP<inline-formula><mml:math id="M480" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentrations and proportions in topsoils that may depend
upon pedoclimate, land-use and management practices.</p><?xmltex \hack{\newpage}?>
</sec>
<sec id="Ch1.S4.SS2">
  <title>A quantitative link between the long-term biogeochemical stability
of SOC and its thermal stability and bulk chemistry</title>
      <p id="d1e6704">This work reinforces the evidence that there is a link between SOC
persistence in ecosystems and its thermal stability, providing evidence of
the first quantitative link between thermal and in situ long-term
(<inline-formula><mml:math id="M481" display="inline"><mml:mo lspace="0mm">&gt;</mml:mo></mml:math></inline-formula> 100 years) biogeochemical SOC stability (Fig. 4). The regression model
yields accurate RE6-RF predictions of CP<inline-formula><mml:math id="M482" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportions with
95 % prediction intervals that fully propagate the uncertainties
originating from the calibration set that was used to build the model.
Predictions on the validation set illustrate that the error propagation
scheme provides highly conservative 95 % prediction intervals of the
CP<inline-formula><mml:math id="M483" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion in new samples at all intervals including their
respective reference estimate of CP<inline-formula><mml:math id="M484" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion (Fig. 4).
Despite rather large prediction intervals, the RE6-RF regression model
clearly discriminates soils with small CP<inline-formula><mml:math id="M485" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportions from
samples with large CP<inline-formula><mml:math id="M486" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportions (Fig. 4). This model based
on RE6 thermal analysis can thus be used to predict the size of the
CP<inline-formula><mml:math id="M487" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> pool with known uncertainty in new soil samples from
similar pedoclimates and with thermal characteristics (i.e., RE6<?pagebreak page2846?> predictor
variables) similar to those of the reference sample set.</p>
      <p id="d1e6769">Our results also highlight the sensitivity of the RE6-RF regression model to
pedoclimate. Decreased predictive performance of the model (as assessed by
the coefficient of determination) was indeed observed when predicting the
CP<inline-formula><mml:math id="M488" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion in new soils with similar RE6 thermal
characteristics but from a different pedoclimate (Fig. 5). However, the mean
error of prediction of the model only slightly increased when predicting the
CP<inline-formula><mml:math id="M489" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion in soils from the fully independent site of
Ultuna (Fig. 5). Overall, those results illustrate the potential of the model
based on RE6 thermal analysis to predict the proportion of CP<inline-formula><mml:math id="M490" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
in new soil samples from different pedoclimates, at least for sites that have
similar RE6 thermal characteristics to those of the calibration set.</p>
      <p id="d1e6799">Our results also illustrate the complex relationships between thermal-analysis-based parameters of SOC stability and the CP<inline-formula><mml:math id="M491" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportion. The hypothesis behind the use of SOC thermal stability as a proxy
for its biogeochemical stability implies positive correlations between the
size of the CP<inline-formula><mml:math id="M492" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> pool and temperature parameters derived from
thermal analysis such as the 25 RE6 temperature parameters calculated in this
study. Significant positive correlations with the CP<inline-formula><mml:math id="M493" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportions were indeed found for the majority (14 out of 25) of RE6
temperature parameters, with very high and positive Spearman's <inline-formula><mml:math id="M494" display="inline"><mml:mi mathvariant="italic">ρ</mml:mi></mml:math></inline-formula> values
for some of them (Table 2). This was notably the case of the RE6 temperature
parameter corresponding to 50 % of <inline-formula><mml:math id="M495" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> gas evolution during the
oxidation stage, <inline-formula><mml:math id="M496" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">50</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">OX</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>, which had been
previously shown to systematically increase with bare fallow duration in the
same soils by Barré et al. (2016). This study extends the results of
Barré et al. (2016) towards a quantitative link between RE6 temperature
parameters and SOC persistence (direct correlations and predictions of the
size of the CP<inline-formula><mml:math id="M497" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> pool rather than time under bare fallow
treatment). It also extends those results to non-bare fallow soils: bare
soils receiving organic amendments (at Grignon and Versailles), cropland
soils (Ultuna) and grassland soils (Rothamsted). Conversely, 11 RE6
temperature parameters showed no significant correlation or significant
negative correlations with the CP<inline-formula><mml:math id="M498" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion. Weak or negative
correlations occurred principally for temperature parameters calculated on
thermograms of the pyrolysis stage of the RE6 analysis: for parameters of the
HC and CO thermograms (except <inline-formula><mml:math id="M499" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mrow><mml:mn mathvariant="normal">90</mml:mn><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">HC</mml:mi><mml:mi mathvariant="italic">_</mml:mi><mml:mi mathvariant="normal">PYR</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula>) and
the lowest temperature parameter of the <inline-formula><mml:math id="M500" display="inline"><mml:mrow class="chem"><mml:msub><mml:mi mathvariant="normal">CO</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula> thermogram (Table 2).
Negative correlations contradict the abovementioned hypothesis, with the
evolution of a similar proportion of the total amount of gases (HC pyrolysis
effluents or CO) occurring at lower temperatures for samples with high
CP<inline-formula><mml:math id="M501" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportions than for soils with low CP<inline-formula><mml:math id="M502" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportions. A possible explanation for this unexpected observation could be
that the pyrolysis of SOC in samples with a high proportion of
CP<inline-formula><mml:math id="M503" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> may undergo an enhanced pyrolysis catalyst effect by soil
minerals (Auber, 2009), which are relatively more abundant in those samples
generally characterized by low total SOC concentrations.</p>
      <p id="d1e6949">Despite the fact that the TLHC index, the <inline-formula><mml:math id="M504" display="inline"><mml:mi>I</mml:mi></mml:math></inline-formula> index and the <inline-formula><mml:math id="M505" display="inline"><mml:mi>R</mml:mi></mml:math></inline-formula> index had
originally been proposed as useful qualitative metrics of soil carbon
dynamics, reflecting a proportion of thermally resistant or labile
hydrocarbons (Disnar et al., 2003; Sebag et al., 2006, 2016; Saenger et al.,
2013, 2015), those parameters were not correlated with the CP<inline-formula><mml:math id="M506" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportion. Furthermore, they also had a weak importance in the random
forests model predictions of the CP<inline-formula><mml:math id="M507" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion (Table 2). The
poor link between those three RE6 parameters and the CP<inline-formula><mml:math id="M508" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
proportion may be explained by the high residence time of CP<inline-formula><mml:math id="M509" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula>
(<inline-formula><mml:math id="M510" display="inline"><mml:mo lspace="0mm">&gt;</mml:mo></mml:math></inline-formula> 100 years). Indeed, so far those parameters have been related to the
proportion of SOC present in the particulate organic matter fraction (size
<inline-formula><mml:math id="M511" display="inline"><mml:mo>&gt;</mml:mo></mml:math></inline-formula> 50 <inline-formula><mml:math id="M512" display="inline"><mml:mi mathvariant="normal">µ</mml:mi></mml:math></inline-formula>m, density <inline-formula><mml:math id="M513" display="inline"><mml:mo>&lt;</mml:mo></mml:math></inline-formula> 1–1.6), an SOC pool characterized by a
residence time in soils generally below 20 years (Saenger et al., 2015;
Soucémarianadin et al., 2018).</p>
      <p id="d1e7032">The two RE6 parameters reflecting SOC bulk chemistry showed highly
significant correlations with the CP<inline-formula><mml:math id="M514" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> proportion. This
confirms, and extends to vegetated soils, the observed decreasing trend for
HI and increasing trend for OI<inline-formula><mml:math id="M515" display="inline"><mml:msub><mml:mi/><mml:mrow><mml:mi mathvariant="normal">RE</mml:mi><mml:mn mathvariant="normal">6</mml:mn></mml:mrow></mml:msub></mml:math></inline-formula> (except at Versailles where
soils have a high pyrogenic C content) with bare fallow duration observed by
Barré et al. (2016) in the bare fallow treatments of the same
experimental sites. Soils with high proportions of CP<inline-formula><mml:math id="M516" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> are thus
characterized by an oxidized and H-depleted organic matter.</p>
</sec>
<sec id="Ch1.S4.SS3">
  <?xmltex \opttitle{Perspectives to improve and foster RE6 thermal-analysis-based
predictions of the size\hack{\break} of the CP${}_{\mathrm{SOC}}$ pool}?><title>Perspectives to improve and foster RE6 thermal-analysis-based
predictions of the size<?xmltex \hack{\break}?> of the CP<inline-formula><mml:math id="M517" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> pool</title>
      <p id="d1e7083">Future developments of this work must extend the Rock-Eval 6 thermal analysis
regression model to a wider range of pedoclimates and to other biomes. As
sites with LTBF treatments are not widespread, complementing the reference
sample set may be achieved by using soils that have different soil forming
factors (e.g., climate, parent material) and (i) which are sampled from
long-term (<inline-formula><mml:math id="M518" display="inline"><mml:mo lspace="0mm">&gt;</mml:mo></mml:math></inline-formula> 50 to 100 years) experiments with contrasting SOC inputs,
enabling the estimation of their CP<inline-formula><mml:math id="M519" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> concentration (Buyanovsky
and Wagner, 1998a, b), or (ii) for which the mean SOC age is known from
radiocarbon data, enabling the estimation of the size of their persistent SOC
pool (Trumbore, 2009; Mills et al., 2014).</p>
      <p id="d1e7102">Another development of this work will involve elucidating the fundamental
mechanisms linking the biogeochemical stability of SOC with its thermal
stability (e.g., Leifeld and von Lützow, 2014). This was beyond the scope
of this work, yet it constitutes an exciting field of research that should be
addressed in the future, as highlighted by the unexpected observations
discussed in Sect. 4.2 and by other recent works that found no relationships
between the thermal oxidation of SOC between 200 and 400 <inline-formula><mml:math id="M520" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C and the
size of SOC pools with shorter residence times in soils (below or above ca.
18 years; Schiedung et al., 2017).</p>
      <?pagebreak page2847?><p id="d1e7114">Overall, this work demonstrates the value of Rock-Eval 6 as a routine method
for quantifying the size of the centennially persistent SOC pool with known
uncertainty in temperate soils. The relatively low cost of the Rock-Eval 6
technique and the robustness of the thermal analysis regression model make
it possible to apply it to soil monitoring networks across a continuum of
scales as a reliable proxy for SOC persistence. This may be part of the
framework proposed by O'Rourke et al. (2015) to better understand SOC
processes at the biosphere to biome scale and should be added to the soil
carbon cycling proxies recently listed by Bailey et al. (2018). Mapping
persistent SOC at large scales may allow for the identification of regional
hotspots of centennially persistent SOC that may contribute little to climate
change by 2100. It may also provide information on the sustainability of
additional SOC storage from soil carbon sequestration strategies such as
those promoted by the international 4 per 1000 initiative in agriculture and
forestry (<uri>https://www.4p1000.org/</uri>, last access: 6 May 2018; Dignac et al., 2017; Minasny et al.,
2017; Soussana et al., 2018). A global map of centennially persistent SOC
based on this empirical RE6 thermal analysis model could also be useful for
improving the parameterization of SOC dynamics in Earth system models
(Falloon and Smith, 2000; Luo et al., 2014; He et al., 2016). Indeed, this
model based on RE6 thermal analysis appears as a robust and operational
alternative to existing techniques used to initialize the size of the
CP<inline-formula><mml:math id="M521" display="inline"><mml:msub><mml:mi/><mml:mi mathvariant="normal">SOC</mml:mi></mml:msub></mml:math></inline-formula> pool in models of SOC dynamics (such as the methods of
Falloon et al. (1998) or Zimmermann et al. (2007) that estimate the size of
the inert SOC pool in the RothC model). The integration of large-scale
information on the size of SOC kinetic pools may provide an adequate
complement to the global datasets on SOC fluxes that are currently under
development and restructuring (Hashimoto et al., 2015; Luo et al., 2016;
Harden et al., 2018).</p>
</sec>
</sec>

      
      </body>
    <back><notes notes-type="dataavailability">

      <p id="d1e7135">Data can be accessed upon request to Lauric Cécillon and Pierre Barré.</p>
  </notes><app-group>
        <supplementary-material position="anchor"><p id="d1e7138">The supplement related to this article is available online at: <inline-supplementary-material xlink:href="https://doi.org/10.5194/bg-15-2835-2018-supplement" xlink:title="pdf">https://doi.org/10.5194/bg-15-2835-2018-supplement</inline-supplementary-material>.</p></supplementary-material>
        </app-group><notes notes-type="authorcontribution">

      <p id="d1e7147">LC and PB designed the study. PB and CC collected the archived soils. TK,
SH, FvO and AM provided the archived soils and associated metadata. FB and FS
performed the RE6 thermal analyses. RJ wrote the Python code. LC wrote the R
codes and performed all statistical analysis. All authors contributed to the
interpretation of the results. LC prepared the paper with contributions
from all coauthors.</p>
  </notes><notes notes-type="competinginterests">

      <p id="d1e7153">The authors declare that they have no conflict of
interest.</p>
  </notes><ack><title>Acknowledgements</title><p id="d1e7159">This work was funded by ADEME and EC2CO (CARACAS project). Pierre Barré,
Lauric Cécillon, Laure Soucémarianadin and Suzanne Lutfalla thank
the Mairie de Paris (Emergences Programme) for financial support. We thank
Thomas Eglin (ADEME), Nicolas Bouton and Jean Espitalié (Vinci
Technologies), David Sebag (Normandie Univ., University of Lausanne), the
associate editor and two anonymous reviewers for their valuable suggestions
on this work. We thank INRA and AgroParisTech for access to and maintenance
of the Versailles 42 plots and the Grignon 36 plots with their long-term experiments and
sample archives. We thank Rothamsted Research and the Lawes
Agricultural Trust for access to archived samples and the BBSRC for support
under the Institute National Capabilities program grant (BBS/E/C/000J0300).
We gratefully acknowledge the Faculty of Natural Resources and Agricultural
Sciences of the Swedish University of Agricultural Sciences (SLU) for
providing funds for the maintenance of the Ultuna long-term field experiment and
for its sample archive.<?xmltex \hack{\newline}?><?xmltex \hack{\newline}?> Edited by: Marcel van
der Meer<?xmltex \hack{\newline}?> Reviewed by: two anonymous referees</p></ack><ref-list>
    <title>References</title>

      <ref id="bib1.bib1"><label>1</label><mixed-citation>
Amundson, R.: The carbon budget of soils, Annu. Rev. Earth Pl. Sc., 29,
535–562, 2001.</mixed-citation></ref>
      <ref id="bib1.bib2"><label>2</label><mixed-citation>
Auber, M.: Effet catalytique de certains inorganiques sur la
sélectivité des réactions de pyrolyse rapide de biomasses et de
leurs constituants, PhD Thesis, Institut National Polytechnique de Lorraine,
296 pp., 2009.</mixed-citation></ref>
      <ref id="bib1.bib3"><label>3</label><mixed-citation>
Bailey, V. L., Bond-Lamberty, B., DeAngelis, K., Grandy, A. S., Hawkes, C.
V., Heckman, K., Lajtha, K., Phillips, R. P., Sulman, B. N., Todd-Brown, K.
E. O., and Wallenstein, M. D.: Soil carbon cycling proxies: Understanding
their critical role in predicting climate change feedbacks, Glob. Change
Biol., 24, 895–905, 2018.</mixed-citation></ref>
      <ref id="bib1.bib4"><label>4</label><mixed-citation>Barré, P., Eglin, T., Christensen, B. T., Ciais, P., Houot, S.,
Kätterer, T., van Oort, F., Peylin, P., Poulton, P. R., Romanenkov, V.,
and Chenu, C.: Quantifying and isolating stable soil organic carbon using
long-term bare fallow experiments, Biogeosciences, 7, 3839–3850,
<ext-link xlink:href="https://doi.org/10.5194/bg-7-3839-2010" ext-link-type="DOI">10.5194/bg-7-3839-2010</ext-link>, 2010.</mixed-citation></ref>
      <ref id="bib1.bib5"><label>5</label><mixed-citation>
Barré, P., Plante, A. F., Cécillon, L., Lutfalla, S., Baudin, F.,
Bernard, S., Christensen, B. T., Eglin, T., Fernandez, J. M., Houot, S.,
Kätterer, T., Le Guillou, C., Macdonald, A., van Oort, F., and Chenu, C.:
The energetic and chemical signatures of persistent soil organic matter,
Biogeochemistry, 130, 1–12, 2016.</mixed-citation></ref>
      <ref id="bib1.bib6"><label>6</label><mixed-citation>
Behar, F., Beaumont, V., and De B. Penteado, H. L.: Rock-Eval 6 technology:
performances and developments, Oil Gas Sci. Technol., 56, 111–134, 2001.</mixed-citation></ref>
      <ref id="bib1.bib7"><label>7</label><mixed-citation>
Beleites, C. and Sergo, V.: hyperSpec: a package to handle hyperspectral data
sets in R, R package version 0.98-20140523, 2014.</mixed-citation></ref>
      <ref id="bib1.bib8"><label>8</label><mixed-citation>Bispo, A., Andersen, L., Angers, D., Bernoux, M., Brossard, M., Cécillon,
L., Comans, R., Harmsen, J., Jonassen, K., Lamé, F., Lhuillery, C., Maly,
S., Martin, E., Mcelnea, A., Sakai, H., Watabe, Y., and Eglin, T.: Accounting
for carbon stocks in soils and measuring GHGs emission fluxes from soils: do
we have the necessary standards?, Front. Environ. Sci., 5, 41, <ext-link xlink:href="https://doi.org/10.3389/fenvs.2017.00041" ext-link-type="DOI">10.3389/fenvs.2017.00041</ext-link>,
2017.</mixed-citation></ref>
      <ref id="bib1.bib9"><label>9</label><mixed-citation>
Borchers, H. W.: pracma: Practical Numerical Math Functions, R package
version 1.8.8, 2015.</mixed-citation></ref>
      <ref id="bib1.bib10"><label>10</label><mixed-citation>
Breiman, L.: Random forests, Mach. Learn., 45, 5–32, 2001.</mixed-citation></ref>
      <ref id="bib1.bib11"><label>11</label><mixed-citation>
Brodowski, S., Amelung, W., Haumaier, L., and Zech, W.: Black carbon
contribution to stable humus in German arable soils, Geoderma, 139, 220–228,
2007.</mixed-citation></ref>
      <ref id="bib1.bib12"><label>12</label><mixed-citation>Bruun, S., Thomsen, I. K., Christensen, B. T., and Jensen, L. S.: In search
of stable soil organic carbon fractions: a comparison of methods applied to
soils labelled with <inline-formula><mml:math id="M522" display="inline"><mml:mrow class="chem"><mml:msup><mml:mi/><mml:mn mathvariant="normal">14</mml:mn></mml:msup><mml:mi mathvariant="normal">C</mml:mi></mml:mrow></mml:math></inline-formula> for 40 days or 40 years, Eur. J. Soil Sci.,
59, 247–256, 2008.</mixed-citation></ref>
      <ref id="bib1.bib13"><label>13</label><mixed-citation>
Buyanovsky, G. A. and Wagner, G. H.: Carbon cycling in cultivated land and
its global significance, Glob. Change Biol., 4, 131–141, 1998a.</mixed-citation></ref>
      <ref id="bib1.bib14"><label>14</label><mixed-citation>
Buyanovsky, G. A. and Wagner, G. H.: Changing role of cultivated land in the
global carbon cycle, Biol. Fert. Soils, 27, 242–245, 1998b.</mixed-citation></ref>
      <ref id="bib1.bib15"><label>15</label><mixed-citation>
Campo, J. and Merino, A.: Variations in soil carbon sequestration and their
determinants along a precipitation gradient in seasonally dry tropical forest
ecosystems, Glob. Change Biol., 22, 1942–1956, 2016.</mixed-citation></ref>
      <ref id="bib1.bib16"><label>16</label><mixed-citation>
Canty, A. and Ripley, B.: boot: Bootstrap R (S-Plus) Functions, R package
version 1.3-17, 2015.</mixed-citation></ref>
      <ref id="bib1.bib17"><label>17</label><mixed-citation>
Coulston, J. W., Blinn, C. E., Thomas, V. A., and Wynne, R. H.: Approximating
prediction uncertainty for random forest regression models, Photogramm. Eng.
Rem. S., 82, 189–197, 2016.</mixed-citation></ref>
      <ref id="bib1.bib18"><label>18</label><mixed-citation>
Davison, A. C. and Hinkley, D. V.: Bootstrap Methods and Their Applications,
Cambridge University Press, Cambridge, 1997.</mixed-citation></ref>
      <ref id="bib1.bib19"><label>19</label><mixed-citation>Dignac, M. F., Derrien, D., Barré, P., Barot, S., Cécillon, L.,
Chenu, C., Chevallier, T., Freschet, G. T., Garnier, P., Guenet, B., Hedde,
M., Klumpp, K., Lashermes, G., Maron, P. A., Nunan, N., Roumet, C., and
Basile-Doelsch, I.: Increasing soil carbon storage: mechanisms, effects of
agricultural practices and proxies. A review, Agron. Sustain. Dev., 37, 14, <ext-link xlink:href="https://doi.org/10.1007/s13593-017-0421-2" ext-link-type="DOI">10.1007/s13593-017-0421-2</ext-link>,
2017.</mixed-citation></ref>
      <ref id="bib1.bib20"><label>20</label><mixed-citation>
Disnar, J. R., Guillet, B., Keravis, D., Di-Giovanni, C., and Sebag, D.: Soil
organic matter (SOM) characterization by Rock-Eval pyrolysis: scope and
limitations, Org. Geochem., 34, 327–343, 2003.</mixed-citation></ref>
      <ref id="bib1.bib21"><label>21</label><mixed-citation>
Eglin, T., Ciais, P., Piao, S. L., Barré, P., Bellassen, V., Cadule, P.,
Chenu, C., Gasser, T., Koven, C., Reichstein, M., and Smith, P.: Historical
and future perspectives of global soil carbon response to climate and
land-use changes, Tellus B, 62, 700–718, 2010.</mixed-citation></ref>
      <ref id="bib1.bib22"><label>22</label><mixed-citation>Espitalié, J., Madec, M., Tissot, B., Mennig, J. J., and Leplat, P.:
Source rock characterization method for petroleum exploration, 9th annual
offshore technology conference, 2–5 May 1977, Houston, USA, OTC 2935,
<ext-link xlink:href="https://doi.org/10.4043/2935-MS" ext-link-type="DOI">10.4043/2935-MS</ext-link>, 1977.</mixed-citation></ref>
      <ref id="bib1.bib23"><label>23</label><mixed-citation>
Falloon, P. and Smith, P.: Modelling refractory soil organic matter, Biol.
Fert. Soils, 30, 388–398, 2000.</mixed-citation></ref>
      <ref id="bib1.bib24"><label>24</label><mixed-citation>
Falloon, P., Smith, P., Coleman, K., and Marshall, S.: Estimating the size of
the inert organic matter pool from total soil organic carbon content for use
in the Rothamsted carbon model, Soil Biol. Biochem., 30, 1207–1211, 1998.</mixed-citation></ref>
      <ref id="bib1.bib25"><label>25</label><mixed-citation>
Feng, W., Shi, Z., Jiang, J., Xia, J., Liang, J., Zhou, J., and Luo, Y.:
Methodological uncertainty in estimating carbon turnover times of soil
fractions, Soil Biol. Biochem., 100, 118–124, 2016.</mixed-citation></ref>
      <ref id="bib1.bib26"><label>26</label><mixed-citation>
Franko, U. and Merbach, I.: Modelling soil organic matter dynamics on a bare
fallow Chernozem soil in Central Germany, Geoderma, 303, 93–98, 2017.</mixed-citation></ref>
      <ref id="bib1.bib27"><label>27</label><mixed-citation>
Gregorich, E. G., Gillespie, A. W., Beare, M. H., Curtin, D., Sanei, H., and
Yanni, S. F.: Evaluating biodegradability of soil organic matter by its
thermal stability and chemical composition, Soil Biol. Biochem., 91,
182–191, 2015.</mixed-citation></ref>
      <ref id="bib1.bib28"><label>28</label><mixed-citation>
Harden, J. W., Hugelius, G., Ahlström, A., Blankinship, J. C.,
Bond-Lamberty, B., Lawrence, C. R., Loisel, J., Malhotra, A., Jackson, R. B.,
Ogle, S., Phillips, C., Ryals, R., Todd-Brown, K., Vargas, R., Vergara, S.
E., Cotrufo, M. F., Keiluweit, M., Heckman, K. A., Crow, S. E., Silver, W.
L., DeLonge, M., and Nave, L. E.: Networking our science to characterize the
state, vulnerabilities, and management opportunities of soil organic matter,
Glob. Change Biol., 24, e705–e718, 2018.</mixed-citation></ref>
      <ref id="bib1.bib29"><label>29</label><mixed-citation>Hashimoto, S., Carvalhais, N., Ito, A., Migliavacca, M., Nishina, K., and
Reichstein, M.: Global spatiotemporal distribution of soil respiration
modeled using a global database, Biogeosciences, 12, 4121–4132,
<ext-link xlink:href="https://doi.org/10.5194/bg-12-4121-2015" ext-link-type="DOI">10.5194/bg-12-4121-2015</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bib30"><label>30</label><mixed-citation>
He, Y., Trumbore, S. E., Torn, M. S., Harden, J. W., Vaughn, L. J. S.,
Allison, S. D., and Randerson, J. T.: Radiocarbon constraints imply reduced
carbon uptake by soils during the 21st century, Science, 353, 1419–1424,
2016.</mixed-citation></ref>
      <ref id="bib1.bib31"><label>31</label><mixed-citation>
Kassambara, A.: factoextra: extract and visualize the results of multivariate
data analyses, R package version 1.0.3, 2015.</mixed-citation></ref>
      <ref id="bib1.bib32"><label>32</label><mixed-citation>
Leifeld, J. and von Lützow, M.: Chemical and microbial activation
energies of soil organic matter decomposition, Biol. Fert. Soils, 50,
147–153, 2014.</mixed-citation></ref>
      <ref id="bib1.bib33"><label>33</label><mixed-citation>
Liaw, A. and Wiener, M.: Classification and Regression by randomForest, R
News, 2, 18–22, 2002.</mixed-citation></ref>
      <ref id="bib1.bib34"><label>34</label><mixed-citation>Luo, Y., Ahlström, A., Allison, S. D., Batjes, N. H., Brovkin, V.,
Carvalhais, N., Chappell, A., Ciais, P., Davidson, E. A., Finzi, A.,
Georgiou, K., Guenet, B., Hararuk, O., Harden, J. W., He, Y., Hopkins, F.,
Jiang, L., Koven, C., Jackson, R. B., Jones, C. D., Lara, M. J., Liang, J.,
McGuire, A. D., Parton, W., Peng, C., Randerson, J. T., Salazar, A., Sierra,
C. A., Smith, M. J., Tian, H., Todd-Brown, K. E. O., Torn, M., van Groenigen,
K. J., Wang, Y. P., West, T. O., Wei, Y., Wieder, W. R., Xia, J., Xu, X., Xu,
X., and Zhou, T.: Toward more realistic projections of soil carbon dynamics
by Earth system models, Global Biogeochem. Cy., 30, 40–56,
<ext-link xlink:href="https://doi.org/10.1002/2015GB005239" ext-link-type="DOI">10.1002/2015GB005239</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bib35"><label>35</label><mixed-citation>
Luo, Z. K., Wang, E. L., Fillery, I. R. P., Macdonald, L. M., Huth, N., and
Baldock, J.: Modelling soil carbon and nitrogen dynamics using measurable and
conceptual soil organic matter pools in APSIM, Agr. Ecosyst. Environ., 186,
94–104, 2014.</mixed-citation></ref>
      <ref id="bib1.bib36"><label>36</label><mixed-citation>
Lutfalla, S., Chenu, C., and Barré, P.: Are chemical oxidation methods
relevant to isolate a soil pool of centennial carbon?, Biogeochemistry, 118,
135–139, 2014.</mixed-citation></ref>
      <ref id="bib1.bib37"><label>37</label><mixed-citation>
Mathieu, J. A., Hatté, C., Balesdent, J., and Parent, E.: Deep soil
carbon dynamics are driven more by soil type than by climate: a worldwide
meta-analysis of radiocarbon profiles, Glob. Change Biol., 21, 4278–4292,
2015.</mixed-citation></ref>
      <ref id="bib1.bib38"><label>38</label><mixed-citation>Menichetti, L., Kätterer, T., and Leifeld, J.: Parametrization
consequences of constraining soil organic matter models by total carbon and
radiocarbon using long-term field data, Biogeosciences, 13, 3003–3019,
<ext-link xlink:href="https://doi.org/10.5194/bg-13-3003-2016" ext-link-type="DOI">10.5194/bg-13-3003-2016</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bib39"><label>39</label><mixed-citation>
Mills, R. T. E., Tipping, E., Bryant, C. L., and Emmett, B. A.: Long-term
organic carbon turnover rates in natural and semi-natural topsoils,
Biogeochemistry, 118, 257–272, 2014.</mixed-citation></ref>
      <ref id="bib1.bib40"><label>40</label><mixed-citation>
Minasny, B., Malone, B. P., McBratney, A. B., Angers, D. A., Arrouays, D.,
Chambers, A., Chaplot, V., Chen, Z. S., Cheng, K., Das, B. S., Field, D. J.,
Gimona, A., Hedley, C. B., Hong, S.<?pagebreak page2849?> Y., Mandal, B., Marchant, B. P., Martin,
M., McConkey, B. G., Mulder, V. L., O'Rourke, S., Richer-de-Forges, A. C.,
Odeh, I., Padarian, J., Paustian, K., Pan, G., Poggio, L., Savin, I.,
Stolbovoy, V., Stockmann, U., Sulaeman, Y., Tsui, C. C., Vågen, T. G.,
van Wesemael, B., and Winowiecki, L.: Soil carbon 4 per mille, Geoderma, 292,
59–86, 2017.</mixed-citation></ref>
      <ref id="bib1.bib41"><label>41</label><mixed-citation>
NF ISO 10694: Determination of organic and total carbon after dry combustion
(elementary analysis), AFNOR, 1995.</mixed-citation></ref>
      <ref id="bib1.bib42"><label>42</label><mixed-citation>
O'Rourke, S. M., Angers, D. A., Holden, N. M., and McBratney, A. B.: Soil
organic carbon across scales, Glob. Change Biol., 21, 3561–3574, 2015.</mixed-citation></ref>
      <ref id="bib1.bib43"><label>43</label><mixed-citation>
Patil, A., Huard, D., and Fonnesbeck, C. J.: PyMC: Bayesian stochastic
modelling in Python, J. Stat. Softw., 35, 1–81, 2010.</mixed-citation></ref>
      <ref id="bib1.bib44"><label>44</label><mixed-citation>
Plante, A. F., Fernández, J. M., and Leifeld, J.: Application of thermal
analysis techniques in soil science, Geoderma, 153, 1–10, 2009.</mixed-citation></ref>
      <ref id="bib1.bib45"><label>45</label><mixed-citation>
Plante, A. F., Fernández, J. M., Haddix, M. L., Steinweg, J. M., and
Conant, R. T.: Biological, chemical and thermal indices of soil organic
matter stability in four grassland soils, Soil Biol. Biochem., 43,
1051–1058, 2011.</mixed-citation></ref>
      <ref id="bib1.bib46"><label>46</label><mixed-citation>
Plante, A. F., Beaupré, S. R., Roberts, M. L., and Baisden, T.:
Distribution of radiocarbon ages in soil organic matter by thermal
fractionation, Radiocarbon, 55, 1077–1083, 2013.</mixed-citation></ref>
      <ref id="bib1.bib47"><label>47</label><mixed-citation>
Post, W. M. and Kwon, K. C.: Soil carbon sequestration and land-use change:
processes and potential, Glob. Change Biol., 6, 317–327, 2000.</mixed-citation></ref>
      <ref id="bib1.bib48"><label>48</label><mixed-citation>
R Core Team: R: a language and environment for statistical computing, Vienna,
Austria: R Foundation for Statistical Computing, 2017.</mixed-citation></ref>
      <ref id="bib1.bib49"><label>49</label><mixed-citation>
Rovira, P., Kurz-Besson, C., Coûteaux, M. M., and Ramón Vallejo, V.:
Changes in litter properties during decomposition: A study by differential
thermogravimetry and scanning calorimetry, Soil Biol. Biochem., 40, 172–185,
2008.</mixed-citation></ref>
      <ref id="bib1.bib50"><label>50</label><mixed-citation>
Rühlmann, J.: A new approach to estimating the pool of stable organic
matter in soil using data from long-term field experiments, Plant Soil, 213,
149–160, 1999.</mixed-citation></ref>
      <ref id="bib1.bib51"><label>51</label><mixed-citation>
Saenger, A., Cécillon, L., Sebag, D., and Brun, J. J.: Soil organic
carbon quantity, chemistry and thermal stability in a mountainous landscape:
A Rock-Eval pyrolysis survey, Org. Geochem., 54, 101–114, 2013.</mixed-citation></ref>
      <ref id="bib1.bib52"><label>52</label><mixed-citation>
Saenger, A., Cécillon, L., Poulenard, J., Bureau, F., De Danieli, S.,
Gonzalez, J. M., and Brun, J. J.: Surveying the carbon pools of mountain
soils: A comparison of physical fractionation and Rock-Eval pyrolysis,
Geoderma, 241–242, 279–288, 2015.</mixed-citation></ref>
      <ref id="bib1.bib53"><label>53</label><mixed-citation>
Sanderman, J., Baisden, W. T., and Fallon, S.: Redefining the inert organic
carbon pool, Soil Biol. Biochem., 92, 149–152, 2016.</mixed-citation></ref>
      <ref id="bib1.bib54"><label>54</label><mixed-citation>Schiedung, M., Don, A., Wordell-Dietrich, P., Alcántara, V., Kuner, P.,
and Guggenberger, G.: Thermal oxidation does not fractionate soil organic
carbon with differing biological stabilities, J. Plant Nutr. Soil Sci., 180,
18–26, 2017.
 </mixed-citation></ref><?xmltex \hack{\newpage}?>
      <ref id="bib1.bib55"><label>55</label><mixed-citation>
Schrumpf, M. and Kaiser, K.: Large differences in estimates of soil organic
carbon turnover in density fractions by using single and repeated radiocarbon
inventories, Geoderma, 239–240, 168–178, 2015.</mixed-citation></ref>
      <ref id="bib1.bib56"><label>56</label><mixed-citation>
Sebag, D., Disnar, J. R., Guillet, B., Di Giovanni, C., Verrecchia, E. P.,
and Durand, A.: Monitoring organic matter dynamics in soil profiles by
“Rock-Eval pyrolysis”: bulk characterization and quantification of
degradation, Eur. J. Soil Sci., 57, 344–355, 2006.</mixed-citation></ref>
      <ref id="bib1.bib57"><label>57</label><mixed-citation>
Sebag, D., Verrecchia, E. P., Cécillon, L., Adatte, T., Albrecht, R.,
Aubert, M., Bureau, F., Cailleau, G., Copard, Y., Decaens, T., Disnar, J. R.,
Hetényi, M., Nyilas, T., and Trombino, L.: Dynamics of soil organic
matter based on new Rock-Eval indices, Geoderma, 284, 185–203, 2016.</mixed-citation></ref>
      <ref id="bib1.bib58"><label>58</label><mixed-citation>
Soil Carbon Initiative: Outcomes Document Soil Carbon Summit,
Dow-Sustainability-Fund and ALCOA, 11 pp., The United States Study Center, The University of Sydney, Sydney, Australia, 2011.</mixed-citation></ref>
      <ref id="bib1.bib59"><label>59</label><mixed-citation>
Soucémarianadin, L., Cécillon, L., Chenu, C., Baudin, F., Nicolas,
M., Girardin, C., and Barré, P.: Is Rock-Eval 6 thermal analysis a good
indicator of soil organic carbon lability? – A method comparison study in
forest soils, Soil Biol. Biochem., 117, 108–116, 2018.</mixed-citation></ref>
      <ref id="bib1.bib60"><label>60</label><mixed-citation>Soussana, J. F., Lutfalla, S., Ehrhardt, F., Rosenstock, T., Lamanna, C.,
Havlík, P., Richards, M., Wollenberg, E. L., Chotte, J. L., Torquebiau,
E., Ciais, P., Smith, P., and Lal, R.: Matching policy and science: Rationale
for the “4 per 1000 – soils for food security and climate” initiative,
Soil Till. Res., <ext-link xlink:href="https://doi.org/10.1016/j.still.2017.12.002" ext-link-type="DOI">10.1016/j.still.2017.12.002</ext-link>, online first, 2017.</mixed-citation></ref>
      <ref id="bib1.bib61"><label>61</label><mixed-citation>
Strobl, C., Malley, J., and Tutz, G.: An introduction to recursive
partitioning: rationale, application and characteristics of classification
and regression trees, bagging and random forests, Psychol. Methods, 14,
323–348, 2009.</mixed-citation></ref>
      <ref id="bib1.bib62"><label>62</label><mixed-citation>
Trumbore, S. E.: Potential responses of soil organic carbon to global
environmental change, P. Natl. Acad. Sci. USA, 94, 8284–8291, 1997.</mixed-citation></ref>
      <ref id="bib1.bib63"><label>63</label><mixed-citation>
Trumbore, S. E.: Radiocarbon and soil carbon dynamics, Annu. Rev. Earth Pl.
Sc., 37, 47–66, 2009.</mixed-citation></ref>
      <ref id="bib1.bib64"><label>64</label><mixed-citation>
von Lützow, M., Kögel-Knabner, I., Ekschmitt, K., Flessa, H.,
Guggenberger, G., Matzner, E., and Marschner, B.: SOM fractionation methods:
relevance to functional pools and to stabilization mechanisms, Soil Biol.
Biochem., 39, 2183–2207, 2007.</mixed-citation></ref>
      <ref id="bib1.bib65"><label>65</label><mixed-citation>Wickham, H.: stringr: Simple, Consistent Wrappers for Common String
Operations, R package version 1.0.0, available at:
<uri>http://CRAN.R-project.org/package=stringr</uri> (last access: 6 May 2018), 2015.</mixed-citation></ref>
      <ref id="bib1.bib66"><label>66</label><mixed-citation>
Zimmermann, M., Leifeld, J., Schmidt, M. W. I., Smith, P., and Fuhrer, J.:
Measured soil organic matter fractions can be related to pools in the RothC
model, Eur. J. Soil Sc., 58, 658–667, 2007.</mixed-citation></ref>

  </ref-list></back>
    <!--<article-title-html>A model based on Rock-Eval thermal analysis to quantify the size of the centennially persistent organic carbon pool in temperate soils</article-title-html>
<abstract-html><p>Changes in global soil carbon stocks have considerable potential to influence
the course of future climate change. However, a portion of soil organic
carbon (SOC) has a very long residence time ( &gt; &thinsp;100 years) and may not
contribute significantly to terrestrial greenhouse gas emissions during the
next century. The size of this persistent SOC reservoir is presumed to be
large. Consequently, it is a key parameter required for the initialization of
SOC dynamics in ecosystem and Earth system models, but there is considerable
uncertainty in the methods used to quantify it. Thermal analysis methods
provide cost-effective information on SOC thermal stability that has been
shown to be qualitatively related to SOC biogeochemical stability. The
objective of this work was to build the first quantitative model of the size
of the centennially persistent SOC pool based on thermal analysis. We used a
unique set of 118 archived soil samples from four agronomic experiments in
northwestern Europe with long-term bare fallow and non-bare fallow treatments
(e.g., manure amendment, cropland and grassland) as a sample set for which
estimating the size of the centennially persistent SOC pool is relatively
straightforward. At each experimental site, we estimated the average
concentration of centennially persistent SOC and its uncertainty by applying
a Bayesian curve-fitting method to the observed declining SOC concentration
over the duration of the long-term bare fallow treatment. Overall, the
estimated concentrations of centennially persistent SOC ranged from 5 to
11&thinsp;g&thinsp;C&thinsp;kg<sup>−1</sup> of soil (lowest and highest boundaries of four 95&thinsp;%
confidence intervals). Then, by dividing the site-specific concentrations of
persistent SOC by the total SOC concentration, we could estimate the
proportion of centennially persistent SOC in the 118 archived soil samples
and the associated uncertainty. The proportion of centennially persistent SOC
ranged from 0.14 (standard deviation of 0.01) to 1 (standard deviation of
0.15). Samples were subjected to thermal analysis by Rock-Eval 6 that
generated a series of 30 parameters reflecting their SOC thermal stability
and bulk chemistry. We trained a nonparametric machine-learning algorithm
(random forests multivariate regression model) to predict the proportion of
centennially persistent SOC in new soils using Rock-Eval 6 thermal parameters
as predictors. We evaluated the model predictive performance with two
different strategies. We first used a calibration set (<i>n</i>&thinsp; = &thinsp;88) and a
validation set (<i>n</i>&thinsp; = &thinsp;30) with soils from all sites. Second, to test the
sensitivity of the model to pedoclimate, we built a calibration set with soil
samples from three out of the four sites (<i>n</i>&thinsp; = &thinsp;84). The multivariate
regression model accurately predicted the proportion of centennially
persistent SOC in the validation set composed of soils from all sites
(<i>R</i><sup>2</sup>&thinsp; = &thinsp;0.92, RMSEP&thinsp; = &thinsp;0.07, <i>n</i>&thinsp; = &thinsp;30). The uncertainty of the
model predictions was quantified by a Monte Carlo approach that produced
conservative 95&thinsp;% prediction intervals across the validation set. The
predictive performance of the model decreased when predicting the proportion
of centennially persistent SOC in soils from one fully independent site with
a different pedoclimate, yet the mean error of prediction only slightly
increased (<i>R</i><sup>2</sup>&thinsp; = &thinsp;0.53, RMSEP&thinsp; = &thinsp;0.10, <i>n</i>&thinsp; = &thinsp;34). This model
based on Rock-Eval 6 thermal analysis can thus be used to predict the
proportion of centennially persistent SOC with known uncertainty in new soil
samples from different pedoclimates, at least for sites that have similar
Rock-Eval 6 thermal characteristics to those included in the calibration set.
Our study reinforces the evidence that there is a link between the thermal
and biogeochemical stability of soil organic matter and demonstrates that
Rock-Eval 6 thermal analysis can be used to quantify the size of the
centennially persistent organic carbon pool in temperate soils.</p></abstract-html>
<ref-html id="bib1.bib1"><label>1</label><mixed-citation>
Amundson, R.: The carbon budget of soils, Annu. Rev. Earth Pl. Sc., 29,
535–562, 2001.
</mixed-citation></ref-html>
<ref-html id="bib1.bib2"><label>2</label><mixed-citation>
Auber, M.: Effet catalytique de certains inorganiques sur la
sélectivité des réactions de pyrolyse rapide de biomasses et de
leurs constituants, PhD Thesis, Institut National Polytechnique de Lorraine,
296 pp., 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib3"><label>3</label><mixed-citation>
Bailey, V. L., Bond-Lamberty, B., DeAngelis, K., Grandy, A. S., Hawkes, C.
V., Heckman, K., Lajtha, K., Phillips, R. P., Sulman, B. N., Todd-Brown, K.
E. O., and Wallenstein, M. D.: Soil carbon cycling proxies: Understanding
their critical role in predicting climate change feedbacks, Glob. Change
Biol., 24, 895–905, 2018.
</mixed-citation></ref-html>
<ref-html id="bib1.bib4"><label>4</label><mixed-citation>
Barré, P., Eglin, T., Christensen, B. T., Ciais, P., Houot, S.,
Kätterer, T., van Oort, F., Peylin, P., Poulton, P. R., Romanenkov, V.,
and Chenu, C.: Quantifying and isolating stable soil organic carbon using
long-term bare fallow experiments, Biogeosciences, 7, 3839–3850,
<a href="https://doi.org/10.5194/bg-7-3839-2010" target="_blank">https://doi.org/10.5194/bg-7-3839-2010</a>, 2010.
</mixed-citation></ref-html>
<ref-html id="bib1.bib5"><label>5</label><mixed-citation>
Barré, P., Plante, A. F., Cécillon, L., Lutfalla, S., Baudin, F.,
Bernard, S., Christensen, B. T., Eglin, T., Fernandez, J. M., Houot, S.,
Kätterer, T., Le Guillou, C., Macdonald, A., van Oort, F., and Chenu, C.:
The energetic and chemical signatures of persistent soil organic matter,
Biogeochemistry, 130, 1–12, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib6"><label>6</label><mixed-citation>
Behar, F., Beaumont, V., and De B. Penteado, H. L.: Rock-Eval 6 technology:
performances and developments, Oil Gas Sci. Technol., 56, 111–134, 2001.
</mixed-citation></ref-html>
<ref-html id="bib1.bib7"><label>7</label><mixed-citation>
Beleites, C. and Sergo, V.: hyperSpec: a package to handle hyperspectral data
sets in R, R package version 0.98-20140523, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib8"><label>8</label><mixed-citation>
Bispo, A., Andersen, L., Angers, D., Bernoux, M., Brossard, M., Cécillon,
L., Comans, R., Harmsen, J., Jonassen, K., Lamé, F., Lhuillery, C., Maly,
S., Martin, E., Mcelnea, A., Sakai, H., Watabe, Y., and Eglin, T.: Accounting
for carbon stocks in soils and measuring GHGs emission fluxes from soils: do
we have the necessary standards?, Front. Environ. Sci., 5, 41, <a href="https://doi.org/10.3389/fenvs.2017.00041" target="_blank">https://doi.org/10.3389/fenvs.2017.00041</a>,
2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib9"><label>9</label><mixed-citation>
Borchers, H. W.: pracma: Practical Numerical Math Functions, R package
version 1.8.8, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib10"><label>10</label><mixed-citation>
Breiman, L.: Random forests, Mach. Learn., 45, 5–32, 2001.
</mixed-citation></ref-html>
<ref-html id="bib1.bib11"><label>11</label><mixed-citation>
Brodowski, S., Amelung, W., Haumaier, L., and Zech, W.: Black carbon
contribution to stable humus in German arable soils, Geoderma, 139, 220–228,
2007.
</mixed-citation></ref-html>
<ref-html id="bib1.bib12"><label>12</label><mixed-citation>
Bruun, S., Thomsen, I. K., Christensen, B. T., and Jensen, L. S.: In search
of stable soil organic carbon fractions: a comparison of methods applied to
soils labelled with <sup>14</sup>C for 40 days or 40 years, Eur. J. Soil Sci.,
59, 247–256, 2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib13"><label>13</label><mixed-citation>
Buyanovsky, G. A. and Wagner, G. H.: Carbon cycling in cultivated land and
its global significance, Glob. Change Biol., 4, 131–141, 1998a.
</mixed-citation></ref-html>
<ref-html id="bib1.bib14"><label>14</label><mixed-citation>
Buyanovsky, G. A. and Wagner, G. H.: Changing role of cultivated land in the
global carbon cycle, Biol. Fert. Soils, 27, 242–245, 1998b.
</mixed-citation></ref-html>
<ref-html id="bib1.bib15"><label>15</label><mixed-citation>
Campo, J. and Merino, A.: Variations in soil carbon sequestration and their
determinants along a precipitation gradient in seasonally dry tropical forest
ecosystems, Glob. Change Biol., 22, 1942–1956, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib16"><label>16</label><mixed-citation>
Canty, A. and Ripley, B.: boot: Bootstrap R (S-Plus) Functions, R package
version 1.3-17, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib17"><label>17</label><mixed-citation>
Coulston, J. W., Blinn, C. E., Thomas, V. A., and Wynne, R. H.: Approximating
prediction uncertainty for random forest regression models, Photogramm. Eng.
Rem. S., 82, 189–197, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib18"><label>18</label><mixed-citation>
Davison, A. C. and Hinkley, D. V.: Bootstrap Methods and Their Applications,
Cambridge University Press, Cambridge, 1997.
</mixed-citation></ref-html>
<ref-html id="bib1.bib19"><label>19</label><mixed-citation>
Dignac, M. F., Derrien, D., Barré, P., Barot, S., Cécillon, L.,
Chenu, C., Chevallier, T., Freschet, G. T., Garnier, P., Guenet, B., Hedde,
M., Klumpp, K., Lashermes, G., Maron, P. A., Nunan, N., Roumet, C., and
Basile-Doelsch, I.: Increasing soil carbon storage: mechanisms, effects of
agricultural practices and proxies. A review, Agron. Sustain. Dev., 37, 14, <a href="https://doi.org/10.1007/s13593-017-0421-2" target="_blank">https://doi.org/10.1007/s13593-017-0421-2</a>,
2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib20"><label>20</label><mixed-citation>
Disnar, J. R., Guillet, B., Keravis, D., Di-Giovanni, C., and Sebag, D.: Soil
organic matter (SOM) characterization by Rock-Eval pyrolysis: scope and
limitations, Org. Geochem., 34, 327–343, 2003.
</mixed-citation></ref-html>
<ref-html id="bib1.bib21"><label>21</label><mixed-citation>
Eglin, T., Ciais, P., Piao, S. L., Barré, P., Bellassen, V., Cadule, P.,
Chenu, C., Gasser, T., Koven, C., Reichstein, M., and Smith, P.: Historical
and future perspectives of global soil carbon response to climate and
land-use changes, Tellus B, 62, 700–718, 2010.
</mixed-citation></ref-html>
<ref-html id="bib1.bib22"><label>22</label><mixed-citation>
Espitalié, J., Madec, M., Tissot, B., Mennig, J. J., and Leplat, P.:
Source rock characterization method for petroleum exploration, 9th annual
offshore technology conference, 2–5 May 1977, Houston, USA, OTC 2935,
<a href="https://doi.org/10.4043/2935-MS" target="_blank">https://doi.org/10.4043/2935-MS</a>, 1977.
</mixed-citation></ref-html>
<ref-html id="bib1.bib23"><label>23</label><mixed-citation>
Falloon, P. and Smith, P.: Modelling refractory soil organic matter, Biol.
Fert. Soils, 30, 388–398, 2000.
</mixed-citation></ref-html>
<ref-html id="bib1.bib24"><label>24</label><mixed-citation>
Falloon, P., Smith, P., Coleman, K., and Marshall, S.: Estimating the size of
the inert organic matter pool from total soil organic carbon content for use
in the Rothamsted carbon model, Soil Biol. Biochem., 30, 1207–1211, 1998.
</mixed-citation></ref-html>
<ref-html id="bib1.bib25"><label>25</label><mixed-citation>
Feng, W., Shi, Z., Jiang, J., Xia, J., Liang, J., Zhou, J., and Luo, Y.:
Methodological uncertainty in estimating carbon turnover times of soil
fractions, Soil Biol. Biochem., 100, 118–124, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib26"><label>26</label><mixed-citation>
Franko, U. and Merbach, I.: Modelling soil organic matter dynamics on a bare
fallow Chernozem soil in Central Germany, Geoderma, 303, 93–98, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib27"><label>27</label><mixed-citation>
Gregorich, E. G., Gillespie, A. W., Beare, M. H., Curtin, D., Sanei, H., and
Yanni, S. F.: Evaluating biodegradability of soil organic matter by its
thermal stability and chemical composition, Soil Biol. Biochem., 91,
182–191, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib28"><label>28</label><mixed-citation>
Harden, J. W., Hugelius, G., Ahlström, A., Blankinship, J. C.,
Bond-Lamberty, B., Lawrence, C. R., Loisel, J., Malhotra, A., Jackson, R. B.,
Ogle, S., Phillips, C., Ryals, R., Todd-Brown, K., Vargas, R., Vergara, S.
E., Cotrufo, M. F., Keiluweit, M., Heckman, K. A., Crow, S. E., Silver, W.
L., DeLonge, M., and Nave, L. E.: Networking our science to characterize the
state, vulnerabilities, and management opportunities of soil organic matter,
Glob. Change Biol., 24, e705–e718, 2018.
</mixed-citation></ref-html>
<ref-html id="bib1.bib29"><label>29</label><mixed-citation>
Hashimoto, S., Carvalhais, N., Ito, A., Migliavacca, M., Nishina, K., and
Reichstein, M.: Global spatiotemporal distribution of soil respiration
modeled using a global database, Biogeosciences, 12, 4121–4132,
<a href="https://doi.org/10.5194/bg-12-4121-2015" target="_blank">https://doi.org/10.5194/bg-12-4121-2015</a>, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib30"><label>30</label><mixed-citation>
He, Y., Trumbore, S. E., Torn, M. S., Harden, J. W., Vaughn, L. J. S.,
Allison, S. D., and Randerson, J. T.: Radiocarbon constraints imply reduced
carbon uptake by soils during the 21st century, Science, 353, 1419–1424,
2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib31"><label>31</label><mixed-citation>
Kassambara, A.: factoextra: extract and visualize the results of multivariate
data analyses, R package version 1.0.3, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib32"><label>32</label><mixed-citation>
Leifeld, J. and von Lützow, M.: Chemical and microbial activation
energies of soil organic matter decomposition, Biol. Fert. Soils, 50,
147–153, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib33"><label>33</label><mixed-citation>
Liaw, A. and Wiener, M.: Classification and Regression by randomForest, R
News, 2, 18–22, 2002.
</mixed-citation></ref-html>
<ref-html id="bib1.bib34"><label>34</label><mixed-citation>
Luo, Y., Ahlström, A., Allison, S. D., Batjes, N. H., Brovkin, V.,
Carvalhais, N., Chappell, A., Ciais, P., Davidson, E. A., Finzi, A.,
Georgiou, K., Guenet, B., Hararuk, O., Harden, J. W., He, Y., Hopkins, F.,
Jiang, L., Koven, C., Jackson, R. B., Jones, C. D., Lara, M. J., Liang, J.,
McGuire, A. D., Parton, W., Peng, C., Randerson, J. T., Salazar, A., Sierra,
C. A., Smith, M. J., Tian, H., Todd-Brown, K. E. O., Torn, M., van Groenigen,
K. J., Wang, Y. P., West, T. O., Wei, Y., Wieder, W. R., Xia, J., Xu, X., Xu,
X., and Zhou, T.: Toward more realistic projections of soil carbon dynamics
by Earth system models, Global Biogeochem. Cy., 30, 40–56,
<a href="https://doi.org/10.1002/2015GB005239" target="_blank">https://doi.org/10.1002/2015GB005239</a>, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib35"><label>35</label><mixed-citation>
Luo, Z. K., Wang, E. L., Fillery, I. R. P., Macdonald, L. M., Huth, N., and
Baldock, J.: Modelling soil carbon and nitrogen dynamics using measurable and
conceptual soil organic matter pools in APSIM, Agr. Ecosyst. Environ., 186,
94–104, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib36"><label>36</label><mixed-citation>
Lutfalla, S., Chenu, C., and Barré, P.: Are chemical oxidation methods
relevant to isolate a soil pool of centennial carbon?, Biogeochemistry, 118,
135–139, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib37"><label>37</label><mixed-citation>
Mathieu, J. A., Hatté, C., Balesdent, J., and Parent, E.: Deep soil
carbon dynamics are driven more by soil type than by climate: a worldwide
meta-analysis of radiocarbon profiles, Glob. Change Biol., 21, 4278–4292,
2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib38"><label>38</label><mixed-citation>
Menichetti, L., Kätterer, T., and Leifeld, J.: Parametrization
consequences of constraining soil organic matter models by total carbon and
radiocarbon using long-term field data, Biogeosciences, 13, 3003–3019,
<a href="https://doi.org/10.5194/bg-13-3003-2016" target="_blank">https://doi.org/10.5194/bg-13-3003-2016</a>, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib39"><label>39</label><mixed-citation>
Mills, R. T. E., Tipping, E., Bryant, C. L., and Emmett, B. A.: Long-term
organic carbon turnover rates in natural and semi-natural topsoils,
Biogeochemistry, 118, 257–272, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib40"><label>40</label><mixed-citation>
Minasny, B., Malone, B. P., McBratney, A. B., Angers, D. A., Arrouays, D.,
Chambers, A., Chaplot, V., Chen, Z. S., Cheng, K., Das, B. S., Field, D. J.,
Gimona, A., Hedley, C. B., Hong, S. Y., Mandal, B., Marchant, B. P., Martin,
M., McConkey, B. G., Mulder, V. L., O'Rourke, S., Richer-de-Forges, A. C.,
Odeh, I., Padarian, J., Paustian, K., Pan, G., Poggio, L., Savin, I.,
Stolbovoy, V., Stockmann, U., Sulaeman, Y., Tsui, C. C., Vågen, T. G.,
van Wesemael, B., and Winowiecki, L.: Soil carbon 4 per mille, Geoderma, 292,
59–86, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib41"><label>41</label><mixed-citation>
NF ISO 10694: Determination of organic and total carbon after dry combustion
(elementary analysis), AFNOR, 1995.
</mixed-citation></ref-html>
<ref-html id="bib1.bib42"><label>42</label><mixed-citation>
O'Rourke, S. M., Angers, D. A., Holden, N. M., and McBratney, A. B.: Soil
organic carbon across scales, Glob. Change Biol., 21, 3561–3574, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib43"><label>43</label><mixed-citation>
Patil, A., Huard, D., and Fonnesbeck, C. J.: PyMC: Bayesian stochastic
modelling in Python, J. Stat. Softw., 35, 1–81, 2010.
</mixed-citation></ref-html>
<ref-html id="bib1.bib44"><label>44</label><mixed-citation>
Plante, A. F., Fernández, J. M., and Leifeld, J.: Application of thermal
analysis techniques in soil science, Geoderma, 153, 1–10, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib45"><label>45</label><mixed-citation>
Plante, A. F., Fernández, J. M., Haddix, M. L., Steinweg, J. M., and
Conant, R. T.: Biological, chemical and thermal indices of soil organic
matter stability in four grassland soils, Soil Biol. Biochem., 43,
1051–1058, 2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib46"><label>46</label><mixed-citation>
Plante, A. F., Beaupré, S. R., Roberts, M. L., and Baisden, T.:
Distribution of radiocarbon ages in soil organic matter by thermal
fractionation, Radiocarbon, 55, 1077–1083, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib47"><label>47</label><mixed-citation>
Post, W. M. and Kwon, K. C.: Soil carbon sequestration and land-use change:
processes and potential, Glob. Change Biol., 6, 317–327, 2000.
</mixed-citation></ref-html>
<ref-html id="bib1.bib48"><label>48</label><mixed-citation>
R Core Team: R: a language and environment for statistical computing, Vienna,
Austria: R Foundation for Statistical Computing, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib49"><label>49</label><mixed-citation>
Rovira, P., Kurz-Besson, C., Coûteaux, M. M., and Ramón Vallejo, V.:
Changes in litter properties during decomposition: A study by differential
thermogravimetry and scanning calorimetry, Soil Biol. Biochem., 40, 172–185,
2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib50"><label>50</label><mixed-citation>
Rühlmann, J.: A new approach to estimating the pool of stable organic
matter in soil using data from long-term field experiments, Plant Soil, 213,
149–160, 1999.
</mixed-citation></ref-html>
<ref-html id="bib1.bib51"><label>51</label><mixed-citation>
Saenger, A., Cécillon, L., Sebag, D., and Brun, J. J.: Soil organic
carbon quantity, chemistry and thermal stability in a mountainous landscape:
A Rock-Eval pyrolysis survey, Org. Geochem., 54, 101–114, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib52"><label>52</label><mixed-citation>
Saenger, A., Cécillon, L., Poulenard, J., Bureau, F., De Danieli, S.,
Gonzalez, J. M., and Brun, J. J.: Surveying the carbon pools of mountain
soils: A comparison of physical fractionation and Rock-Eval pyrolysis,
Geoderma, 241–242, 279–288, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib53"><label>53</label><mixed-citation>
Sanderman, J., Baisden, W. T., and Fallon, S.: Redefining the inert organic
carbon pool, Soil Biol. Biochem., 92, 149–152, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib54"><label>54</label><mixed-citation>
Schiedung, M., Don, A., Wordell-Dietrich, P., Alcántara, V., Kuner, P.,
and Guggenberger, G.: Thermal oxidation does not fractionate soil organic
carbon with differing biological stabilities, J. Plant Nutr. Soil Sci., 180,
18–26, 2017.

</mixed-citation></ref-html>
<ref-html id="bib1.bib55"><label>55</label><mixed-citation>
Schrumpf, M. and Kaiser, K.: Large differences in estimates of soil organic
carbon turnover in density fractions by using single and repeated radiocarbon
inventories, Geoderma, 239–240, 168–178, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib56"><label>56</label><mixed-citation>
Sebag, D., Disnar, J. R., Guillet, B., Di Giovanni, C., Verrecchia, E. P.,
and Durand, A.: Monitoring organic matter dynamics in soil profiles by
“Rock-Eval pyrolysis”: bulk characterization and quantification of
degradation, Eur. J. Soil Sci., 57, 344–355, 2006.
</mixed-citation></ref-html>
<ref-html id="bib1.bib57"><label>57</label><mixed-citation>
Sebag, D., Verrecchia, E. P., Cécillon, L., Adatte, T., Albrecht, R.,
Aubert, M., Bureau, F., Cailleau, G., Copard, Y., Decaens, T., Disnar, J. R.,
Hetényi, M., Nyilas, T., and Trombino, L.: Dynamics of soil organic
matter based on new Rock-Eval indices, Geoderma, 284, 185–203, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib58"><label>58</label><mixed-citation>
Soil Carbon Initiative: Outcomes Document Soil Carbon Summit,
Dow-Sustainability-Fund and ALCOA, 11 pp., The United States Study Center, The University of Sydney, Sydney, Australia, 2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib59"><label>59</label><mixed-citation>
Soucémarianadin, L., Cécillon, L., Chenu, C., Baudin, F., Nicolas,
M., Girardin, C., and Barré, P.: Is Rock-Eval 6 thermal analysis a good
indicator of soil organic carbon lability? – A method comparison study in
forest soils, Soil Biol. Biochem., 117, 108–116, 2018.
</mixed-citation></ref-html>
<ref-html id="bib1.bib60"><label>60</label><mixed-citation>
Soussana, J. F., Lutfalla, S., Ehrhardt, F., Rosenstock, T., Lamanna, C.,
Havlík, P., Richards, M., Wollenberg, E. L., Chotte, J. L., Torquebiau,
E., Ciais, P., Smith, P., and Lal, R.: Matching policy and science: Rationale
for the “4 per 1000 – soils for food security and climate” initiative,
Soil Till. Res., <a href="https://doi.org/10.1016/j.still.2017.12.002" target="_blank">https://doi.org/10.1016/j.still.2017.12.002</a>, online first, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib61"><label>61</label><mixed-citation>
Strobl, C., Malley, J., and Tutz, G.: An introduction to recursive
partitioning: rationale, application and characteristics of classification
and regression trees, bagging and random forests, Psychol. Methods, 14,
323–348, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib62"><label>62</label><mixed-citation>
Trumbore, S. E.: Potential responses of soil organic carbon to global
environmental change, P. Natl. Acad. Sci. USA, 94, 8284–8291, 1997.
</mixed-citation></ref-html>
<ref-html id="bib1.bib63"><label>63</label><mixed-citation>
Trumbore, S. E.: Radiocarbon and soil carbon dynamics, Annu. Rev. Earth Pl.
Sc., 37, 47–66, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib64"><label>64</label><mixed-citation>
von Lützow, M., Kögel-Knabner, I., Ekschmitt, K., Flessa, H.,
Guggenberger, G., Matzner, E., and Marschner, B.: SOM fractionation methods:
relevance to functional pools and to stabilization mechanisms, Soil Biol.
Biochem., 39, 2183–2207, 2007.
</mixed-citation></ref-html>
<ref-html id="bib1.bib65"><label>65</label><mixed-citation>
Wickham, H.: stringr: Simple, Consistent Wrappers for Common String
Operations, R package version 1.0.0, available at:
<a href="http://CRAN.R-project.org/package=stringr" target="_blank">http://CRAN.R-project.org/package=stringr</a> (last access: 6 May 2018), 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib66"><label>66</label><mixed-citation>
Zimmermann, M., Leifeld, J., Schmidt, M. W. I., Smith, P., and Fuhrer, J.:
Measured soil organic matter fractions can be related to pools in the RothC
model, Eur. J. Soil Sc., 58, 658–667, 2007.
</mixed-citation></ref-html>--></article>
