Soil & Tillage Research 215 (2022) 105203

Contents lists available at ScienceDirect

Soil & Tillage Research

journal homepage: www.elsevier.com/locate/still

Soil organic carbon content and mineralization controlled by the

composition, origin and molecular diversity of organic matter: A study in
tropical alpine grasslands
Songyu Yang a, b, *, Boris Jansen a, Samira Absalah a, Karsten Kalbitz c, Fresia O. Chunga Castro d,
Erik L.H. Cammeraat a
a
Institute for Biodiversity and Ecosystem Dynamics, University of Amsterdam, Amsterdam, Netherlands
b
Biology Centre of the Czech Academy of Sciences, Institute of Soil Biology, Na Sádkách 7, České Budějovice CZ-37005, Czech Republic
c
Soil Resources and Land Use, Institute of Soil Science and Site Ecology, Technische Universität Dresden, Dresden, Germany
d
Consultora, Urbanizacion Jose Galvez - FONAVI II - 10 Edificio 5 Dpto 402, Cajamarca, Peru

A R T I C L E I N F O A B S T R A C T

Keywords: The consensus for mechanisms controlling soil organic matter (SOM) persistence has shifted from traditional
Plant-derived views based on SOM recalcitrance to a new paradigm based on SOM stabilization controlled by soil minerals and
Microbial-derived aggregates. Recent studies indicate that the origin, composition and molecular diversity of SOM are crucial to the
Peruvian Andes
decomposition and stabilization of SOM. However, it is not fully understood how the decomposition and sta­
Stabilization
Ecosystem services
bilization of SOM are controlled at the molecular level. The objectives of this study were to investigate whether
soil organic carbon (SOC) contents and mineralization are controlled by the composition, origin and molecular
diversity of SOM. Soil samples were collected from contrasting bedrocks with different precipitation levels at
tropical alpine grasslands of the Peruvian Andes. We applied a combination of a 76-day soil incubation exper­
iment and pyrolysis-GC/MS assisted by thermochemolysis to investigate SOM decomposition and stabilization at
the molecular level. The results indicated that soil samples with high SOC contents (92.6 ± 7.6 g kg− 1 soil) and
low SOC mineralization had abundant derivates of lignin, polysaccharides and n-alkanes. After the incubation,
we observed neither a selective decomposition of any compound groups nor a decline of molecular diversity. In
contrast, soil samples with low SOC contents (30.7 ± 2.8 g kg− 1 soil) and higher SOC mineralization showed a
depletion of plant-derived compounds, an accumulation of microbial-derived compounds and declined molecular
diversity after the incubation. Furthermore, the SOC mineralization of these samples was positively correlated to
the depletion of unsaturated fatty acids and the decrease in molecular diversity after the incubation. Therefore,
we proposed that SOC contents and mineralization in our soils are (1) controlled by selective preservation of
SOM molecular groups (e.g. plant-derived compounds), and (2) associated with changes in molecular diversity of
SOM during microbial decomposition. Due to the selective preservation of organic compounds under different
environmental conditions, we propose that environmental factors should be considered for the management of
ecosystem services such as SOC sequestration in the studied region.

1. Introduction
Soil organic carbon (SOC) is one of the largest terrestrial carbon
pools and plays an important role in global carbon dynamics. The
storage and stability of SOC are largely dependent on the persistence of
soil organic matter (SOM). In recent decades, there has been a shift from
traditional views regarding SOM persistence based on SOM recalcitrance
to a new paradigm based on SOM stabilization as an ecosystem property
with an important role for interactions of SOM with the soil matrix that
limit its accessibility for microorganisms (Lützow et al., 2006; Schmidt
et al., 2011). Models explaining SOM persistence also have changed
from the “humification” model based on aqueous extractions to a
consolidated continuum model. In the continuum model, SOM persis­
tence is controlled by a progressive SOM decomposition and SOM sta­
bilization regulated by interaction with soil mineral surfaces and
aggregates (Han et al., 2016; Lehmann and Kleber, 2015). Although

* Corresponding author at: Institute for Biodiversity and Ecosystem Dynamics, University of Amsterdam, Amsterdam, Netherlands.
E-mail addresses: songyu.yang@upb.cas.cz, syang_science@163.com (S. Yang).

https://doi.org/10.1016/j.still.2021.105203
Received 6 May 2021; Received in revised form 28 August 2021; Accepted 17 September 2021
Available online 4 October 2021
0167-1987/© 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
S. Yang et al.
molecular recalcitrance is now generally deemed less important as a
SOM stabilization mechanism than previously assumed (Dungait et al.,
2012), the effects of SOM molecular composition on SOM persistence are
not fully understood. For example, a consensus is still absent on the
extent to which the composition and origin of SOM control the stabili­
zation of SOM via governing interaction with minerals and aggregates.
To unravel underlying mechanisms of SOM persistence, it is crucial
to have insight at the molecular level (Kögel-Knabner, 2017, 2002).
Carbohydrates are usually distributed in non-stabilized soil fractions
and act as ready-to-use priming sources for microorganisms (Gunina and
Kuzyakov, 2015). Lignin has been proved not as persistent as previously
believed because it may disappear quickly in favorable environmental
conditions (Schmidt et al., 2011; Thevenot et al., 2010). In contrast,
compounds such as long-chain alkanes and amino sugars are reported to
have a long residence time and can dominate stabilized OM under the
right environmental circumstances (Barré et al., 2018; Jansen and
Wiesenberg, 2017; Joergensen, 2018). Lehmann et al. (2020) proposed a
conceptual framework to show the importance of molecular diversity in
controlling SOM persistence. In general, SOM decomposition is allevi­
ated at high molecular diversity because the microbial utilization of
substrate is restricted by the complexity of compound composition
(Cotrufo et al., 2013). However, it is not sufficiently understood how to
quantify SOM molecular diversity to predict the decomposition and
persistence of SOM (Lehmann et al., 2020).
Microbial transformation of plant-derived materials is considered as
a key process for the formation of stable SOM (Cotrufo et al., 2013; Liang
et al., 2017). Kallenbach et al. (2016) gave the first evidence that soil
microorganisms produce chemically diverse SOM from limited types of
organic compounds. Sokol et al. (2019) and Zhu et al. (2020) reported
that microorganisms utilize labile plant compounds (e.g. simple mole­
cules from the rhizosphere) to produce microbial compounds that
accumulate as stabilized SOM in the form of microbial necromass (e.g.
amino sugars). Recent studies indicated that a large proportion of SOM
is microbial-derived (Liang et al., 2019; Miltner et al., 2012). However,
the relative contributions of plant- versus microbial-derived compounds
are largely dependent on environmental and pedogenic factors (Angst
et al., 2021). It is not well understood how these factors control the
preservation of plant- and microbial-derived compounds, as well as the
relevant microbial processes (Angst et al., 2021; Lehmann et al., 2020).
Therefore, more studies are required to unravel the roles of plant- and
microbial-derived organic matter (OM) in the stabilization of SOM,
especially at the molecular level and under various environmental
conditions.
Tropical alpine ecosystems preserve abundant SOC and act as hot
spots under the stress of global change, although they cover limited
areas of the terrestrial surface (Buytaert et al., 2011). Alpine grassland
soils of the Peruvian Andes are characterized by high SOC stocks and
contribute to important ecosystem services such as food production,
water provision and carbon sequestration (Rolando et al., 2017; Yang
et al., 2018). Therefore, it is important to gain insights into the stabili­
zation of SOC (e.g. at the molecular level) in this region. Ecosystems of
the Peruvian Andes are characterized by heterogeneous environmental
conditions such as contrasting bedrocks and a gradient of precipitation
decreasing from north to south (Geo GPS Perú, 2014; Rolando et al.,
2017). Our previous studies showed that these environmental factors
had clear effects on not only soil aggregation and mineralogy but also
SOC stocks and stabilization mechanisms in the studied regions (Yang
et al., 2020b, 2020c). Therefore, the environmental heterogeneity will
provide sufficient variation in soil properties to investigate the effects of
SOM molecular composition on the persistence of SOC. Meanwhile, the
outcome of this study will contribute to the management of SOC storage
and ecosystem services in the studied region.
The objectives of this study were to investigate whether SOC con­
tents and mineralization are controlled by the composition, origin and
diversity of SOM, using (1) individual compound groups, (2) overall
molecular diversity, and (3) plant- versus microbial-derived OM. To gain
2
Soil & Tillage Research 215 (2022) 105203
insights into SOM decomposition and stabilization at the molecular
level, we applied a combination of soil incubation experiments and
pyrolysis-GC/MS assisted with thermochemolysis using soil samples
collected from the grassland of Peruvian Andes.
2. Materials and methods
2.1. Soil sampling
Soil samples were collected from high-altitude grasslands of the
Peruvian Andes. A detailed description of the sampling sites is provided
by Yang et al. (2020b). Briefly, soil samples were collected from con­
trasting bedrocks (limestone and acid igneous rocks) and two precipi­
tation levels (wet and dry sites), controlled with similar altitude and
average temperature (Table 1). Limestone soils (LSs) had A and B ho­
rizons overlying the parent material and belong to the soil group of
Phaeozems. In contrast, acid igneous rock soils (ASs) only had A hori­
zons directly overlying parent materials and can be classified as Ando­
sols or Umbrisols (WRB, 2014). Soil samples (18 samples) were collected
in three replicates in each combination of precipitation levels and bed­
rocks from A horizons (LSs and ASs) and B horizons (only LSs). Two of
the three replicates (12 samples) were used for analyses after drying at
40 ◦ C.
2.2. Soil analyses and incubation experiment
Soil pH was measured using demi-water (w:v = 1:5). Soil bulk den­
sities were calculated using samples collected using Kopecky rings (100
cm3) by weighing them after freeze-drying. Total carbon (C) and ni­
trogen (N) contents were determined using a VarioEL Elementar
analyzer (Elementar, Germany). For samples with pH values > 5.5,
carbonate contents were measured using a TIC module in the Elementar
analyzer. As carbonate contents were found to be negligible in all
samples, SOC contents were equal to total C contents.
Bulk soil samples were separated by aggregate size using a dry-
sieving method. Between 170 g and 230 g air-dried samples were
placed on two mesh sieves (diameter 2 and 0.25 mm, top and bottom)
and shaken with a horizontal shaker (30 Hz for 20 s). The sieving pro­
duced 3 fractions: large macroaggregates (LM, >2 mm), small macro­
aggregates (SM, 0.25− 2 mm) and microaggregates (Mi, <0.25 mm),
respectively. Total C and N contents of these fractions were also
measured by the Vario EL Elementar Analyzer (Elementar, Germany).
The LM and SM fractions (24 samples) were used for incubation and
molecular analyses because they contained more than 60% of the carbon
of the bulk soils (Table 1).
The incubation experiment was reported in detail by Yang et al.
(2020b). Briefly, samples were rewetted on a pF tray at pF = 2 (− 100
mbar) for 10 days to reach favorable moisture contents and to activate
microorganisms. The moistened samples were incubated at 20 ◦ C in
sealed jars (120 cm3) and the headspaces were sampled at Day 1, 2, 6, 9,
13, 20, 28, 48 and 76. After each sampling, synthetic air (CO2 free) was
injected into jars to maintain the internal pressure and avoid the
depletion of O2. The CO2 concentrations of the headspaces were
measured using a gas chromatography flame ionization detector
(GC-FID with a methanizer, Thermo Scientific, Trace GC Ultra). Based
on the CO2 concentrations for all sampling days, cumulative SOC
mineralization (g CO2-C kg− 1 SOC) during the 76-day incubation was
calculated for each sample and used for further statistical analyses. We
used the unit of g CO2-C kg− 1 SOC rather than g CO2-C kg− 1 soil mass
because the former is generally used to estimate the decomposability or
the stability of the SOC while the latter is applied to evaluate CO2
emission from soils and implies information on SOC stability and con­
tents. The incubation was terminated at Day 76 and the incubated soil
samples were dried at 40 ◦ C for pyrolysis-GC/MS analyses.
S. Yang et al. Soil & Tillage Research 215 (2022) 105203

2.3. Pyrolysis-GC/MS analyses

SOC: soil organic carbon, C/N: C/N ratio, BD: bulk density, LM: weight percentage of large macroaggregates (>2 mm), SM: weight percentage of small macroaggregates (0.25–2 mm). 1 bulk density was measured every 10
SOC distributed in SM (%)
Before pyrolysis-GC/MS analyses, homogenized samples (20 – 50
mg) before and after the incubation (48 samples) were mixed with
20–60 µL tetramethylammonium hydroxide (TMAH) solution (25% by
weight in H2O) for derivatization (He et al., 2020). The purpose of

29.1 ± 2.9
14.8 ± 2.9

28.3 ± 1.4

29.1 ± 2.2
22.3 ± 5.8

45.8 ± 3.7
derivatization (methylation) was to improve the identification of polar
and non-volatile compounds, such as fatty acids, lignin and monomers of
cutin and suberin. These compounds provide information on SOM origin
and microbial processes (Section 2.4). After the TMAH addition, a small
amount of the mixture for each sample was placed on a ferromagnetic
SOC distributed in LM (%)

wire and dried under a halogen lamp for the pyrolysis. A Curie-point
pyrolyzer (Horizon Instruments; 600 ◦ C; 5 s) was applied for pyrolysis
and it was coupled to a ThermoQuest Trace GC gas chromatograph
(Waltham, USA). Helium was used as the carrier gas. The initial tem­
perature of the oven was 40 ◦ C for 1 min and heated to 320 ◦ C with a rate
63.5 ± 3.8
84.2 ± 3.3

40.8 ± 2.4

61.2 ± 3.0
71.6 ± 6.9

19.5 ± 3.6

of 7.0 ◦ C min− 1 and a hold time of 10 min. The separation of analytes

was performed on a ZB1-MS column (Phenomenex: 30 m, 0.25 mm i.d,
0.50 µm df). The column was further connected to a Finnigan Trace MS
mass spectrometer (Waltham, USA; m/z: 47–500, ionization energy: 70
eV, cycle time: 0.45 s).
28.5 ± 3.0
11.3 ± 2.1

37.1 ± 1.4

28.2 ± 2.3
21.9 ± 6.2

49.9 ± 3.1
SM (%)

2.4. Peak identification and calculation

Peak identification and quantification were conducted using the

Xcalibur software (Version 3.1, Thermo Fisher Scientific Inc.). Peak
64.4 ± 3.9
88.1 ± 2.3

39.4 ± 2.2

62.6 ± 3.2
73.1 ± 7.2

19.2 ± 3.4

cm and the averaged value was calculated for each horizon. Original data and detailed information are from Yang et al. (2020b).

identification was based on quantifying ions using the NIST library

LM (%)

(Gaithersburg, USA) and information from other publications (e.g.

Brock et al., 2019; Nierop et al., 2007; Schulten and Schnitzer, 1997). A
peak was deemed valid only when the signal of the peak was at least 3
times larger than that of the background noise and when the spectra
BD1 (g cm¡3)

pattern of the compound could be clearly identified. We identified 104

0.79 ± 0.05
1.09 ± 0.06

0.83 ± 0.09

1.03 ± 0.02
1.16 ± 0.05

0.74 ± 0.02

compounds in total and these compounds were classified into 12

chemical groups. The groups were short-chain saturated fatty acids
(FASS, less than 20 C), short-chain unsaturated fatty acids (FASU, less
than 20 C), long-chain fatty acids (FAL, at least 20 C), α,ω-dioic acids
Dry-AS (9.19 S, 77.60 W, 3490 – 3690 m a.s.l., granodiorite or granodiorite-rich glacier material, grassland)

(DA), ω-hydroxyl alkanoic acids (HA), n-alkanes (Alka), n-alcohols

Wet-AS (7.17◦ S, 78.63◦ W, 3570 – 3590 m a.s.l., granite or ignimbrite, grassland or abandoned cultivation)
5.97 ± 0.18
6.56 ± 0.22

5.09 ± 0.09

5.26 ± 0.12
5.34 ± 0.21

5.29 ± 0.15

(Alch), esters (Ester), lignin and phenols (Lg), polysaccharides (Ps),

nitrogen-containing compounds (N) and other compounds (Other). We
pH

classified fatty acids into three sub-groups because: (1) the carbon chain
Wet-LS (7.16 S, 78.61 W, 3510 – 3720 m a.s.l., limestone, grassland or abandoned cultivation)

Dry-LS (9.25◦ S, 77.59◦ W, 3520 – 3580 m a.s.l., limestone, grassland or abandoned cultivation)

length reveals the origin of the fatty acids (i.e. microorganisms or plants,
Wiesenberg et al., 2010); and (2) previous studies for these samples
indicated that the carbon chain length and the presence of double bonds
12.41 ± 0.29
10.37 ± 0.67

13.01 ± 0.46

9.34 ± 0.52
6.86 ± 1.14

14.25 ± 0.17

are important for the stabilization of fatty acids (Yang et al., 2020a).
Peak quantification was conducted by manual peak integration. The sum
C/N

of peak areas of each sample was set to 100% and the relative amount of
each compound to the total peak area was calculated as relative abun­
dance (RA). The RA of each compound group was the sum of RAs for all
SOC content (g kg¡1)

compounds in this group.

General information of sampling sites and soil samples.

To estimate the molecular diversity of SOM, we calculated the

Shannon index, the Simspon Index and the Pielou Index, based on the RA
28.4 ± 4.5

27.5 ± 3.7
14.8 ± 3.7

48.8 ± 6.5
99.5 ± 12.9

59.5 ± 10.6

of each compound (n = 104). We decide to use the Shannon index to

estimate the molecular diversity of SOM because (1) it is widely used to
estimate diversities in ecology (Spellerberg and Fedor, 2003), and (2) it
had strong correlations with the Simpson Index and the Pielou Index
(Supplementary Table S1). The Shannon index was calculated by the
following equation:
Depth (cm)

49 ± 10
46 ± 5
61 ± 3

46 ± 8
61 ± 6

51 ± 5

∑
i=k
RAi RAi
(3)
◦

ShannonRA = ln
i=1
RAk RAk

In which ShannonRA is the Shannon index, RAi is the RA of compound

◦

i, and RAk is the total relative abundance.

A horizon

A horizon

A horizon

A horizon
B horizon

B horizon

The sums of the relative abundances of DAs and HAs were used to
Table 1

estimate plant-derived compounds because they are widely used as

biomarkers of cutin and suberin in soils (Otto and Simpson, 2006).

3
S. Yang et al.
Short-chain saturated fatty acids have a large contribution from micro­
organisms (Wiesenberg et al., 2010) and were reported to be accumu­
lated during the incubation in our previous study (Yang et al., 2020a).
Therefore, we use short-chain saturated fatty as an indicator for
microbial-derived compounds in this study.
2.5. Statistics
Linear regressions were applied to identify significant factors con­
trolling the molecular composition of SOM. The results showed that the
molecular composition of SOM was controlled by precipitation, lithol­
ogy and horizon but was poorly predicted by aggregate size (Supple­
mentary Table S2). As a result, our analyses only focused on
precipitation, lithology and horizon. As the LM and SM fractions from
the same soil samples were not strictly independent samples, a linear
mixed model was applied to compare SOC properties and SOM molec­
ular composition between different samples. A Fisher’s least significant
difference (LSD) test was used for post hoc analyses. In addition, A
principal component analysis (PCA) based on the correlation matrix was
performed to explore the shift in RAs for different soil samples after the
76-day incubation.
A partial least square regression (PLSR) was used to identify signif­
icant chemical groups controlling SOC contents and SOC mineralization.
A multiple linear regression could not be used because of the collinearity
between chemical groups. The PLSR solved this problem as it is a
combination of PCA and linear regression that allows for: (i) the
decomposition of multiple predictors into several latent variables
despite collinearity, and (ii) the prediction of dependent variation using
Fig. 1. Soil organic carbon (SOC) contents, C/N ratios and SOC mineralization for different soils (n ¼ 4). The linear mixed model is used to identify significant
differences between soil samples, as indicated by different letters (P < 0.05). In graph C, letters on the right indicate differences on Day 76. Wet: the wet site, Dry: the
dry site, LS: limestone soils, AS: acid igneous soils, A: A horizons, B: B horizons.
4
Soil & Tillage Research 215 (2022) 105203
predictors via the transformation into latent variables (Abdi, 2003). The
data was normalized (mean = 0, variance = 1) prior to the PLSR anal­
ysis, whereas the cross-validation was conducted using the “leav­
e-one-out” method. Significant predictors were selected using the
variable importance for projection (VIP) method, in which predictors
characterized by VIP scores > 0.8 were considered significant (Chong
and Jun, 2005). Predictors and dependent variables were visualized in
figures by their correlation coefficients at the levels of latent variable 1
(LV1) and latent variable 2 (LV2).
R software (version 4.0.0) was applied for the PLSR, the PCA and the
calculation of VIP using packages plsdepot (Sanchez, 2012), FactoMineR
(Husson et al., 2020) and plsVarSel (Liland et al., 2020). Other analyses
were performed using SPSS 24.0 (SPSS Inc., USA).
3. Results
3.1. Soil properties
For A horizons, Wet-LS samples had the highest SOC contents
(92.6 ± 7.6 g kg− 1) and the lowest SOC mineralization
(7.77 ± 2.44 g kg− 1 C, Day 76). In contrast, Dry-LS samples had the
lowest SOC contents (30.7 ± 2.8 g kg− 1) and the highest SOC mineral­
ization (22.08 ± 2.23 g kg− 1 C, Day 76). Medium levels of SOC contents
and mineralization were found in both Wet-AS and Dry-AS samples
(Fig. 1-A and 1-C). Soil C/N ratios were the highest in Dry-LS samples,
followed by Wet-AS and Wet-LS samples, and were the lowest in Dry-LS
samples (Fig. 1-B). Only LS samples had B horizons, which had lower
SOC contents and C/N ratios compared to A horizons (Fig. 1-A and 1-B).
S. Yang et al.
In addition, Wet-LS samples had higher pH values than other samples,
whereas LS samples were characterized by larger aggregate sizes than
AS samples (Table 1).
3.2. Molecular composition of soil organic matter before the incubation
For A horizons, most chemical groups were not significantly different
between different samples, except for (1) higher ω-hydroxyl acids in the
Wet-AS compared to Dry-LS and Dry-AS samples, (2) higher poly­
saccharide derivates in Wet-LS and Wet-ASs compared to Dry-AS sam­
ples, and (3) higher lignin derivates in Wet-LS than Dry-AS samples
(Fig. 2). For B horizons, short-chain fatty acids were abundant, whereas
lignin derivates were almost absent (Fig. 2-A and 2-I). In addition, Wet-
LS samples had a larger differentiation between A and B horizons than
Dry-LS samples (Fig. 2-A, 2-C, 2-D and 2-E).
In the graph of partial least square regression (Fig. 3-A and 3-B), LV1
and LV2 explained 34.6% and 18.8% of the total variation for predictors
(i.e. RA of chemical groups), whereas 58.8% of the total variation for
dependent variables (i.e. SOC contents and mineralization) was
explained. Zone (1) showed that Wet-LS-A samples had high SOC con­
tents, which were associated with abundant derivates of lignin, poly­
saccharides and n-alkanes (Fig. 3-A and 3-B). Zone (2) indicated that
Dry-LS samples had high SOC mineralization, which was positively
correlated to the presence of unsaturated fatty acids (Fig. 3-A and 3-B).
3.3. Shift of soil organic molecular composition after the incubation
In Fig. 3-C and 3-D, LV1 and LV2 explain 53.0% and 19.0% of the
total variation for both predictors and the dependent variable. Zone (3)
showed that Dry-LS samples were characterized by high SOC minerali­
zation, which corresponds to the large depletion of unsaturated fatty
acids and n-alcohols (Fig. 3-C and 3-D).
The shift of different chemical groups after the incubation is shown
by the PCA that explained 54.6% of the total variation (Fig. 4). The
Fig. 2. Relative abundances of compound groups for different soils before the incubation. The linear mixed model is used to identify significant differences
between soil samples, as indicated by different letters (P < 0.05). Wet: the wet site, Dry: the dry site, LS: limestone soils, AS: acid igneous soils, A: A horizons, B:
B horizons.
5
Soil & Tillage Research 215 (2022) 105203
principal component 1 (PC1) was positively loaded by derivatives of
short-chain unsaturated fatty acids, long-chain fatty acids, α,ω-dioic
acids, ω-hydroxyl acids, n-alcohols, lignin and polysaccharides, and was
negatively loaded by short-chain saturated fatty acids and N compounds
(Fig. 4-A). The PC2 had positive contributions from N-compounds and n-
alkanes (Fig. 4-A). After the incubation, Dry-LS samples showed a clear
pattern moving towards the negative side of PC1, indicating the accu­
mulation of short-chain saturated fatty acids and N-compounds, as well
as depletions of other compounds (Fig. 4-A and 4-C). In contrast, other
soil samples showed no clear pattern of chemical group changes after the
incubation (Fig. 4-B to 4-D).
The molecular diversity before the incubation was not correlated
with SOC contents or mineralization (P > 0.05, Fig. 5-A). After the in­
cubation, the decline of molecular diversity was positively correlated
with SOC mineralization (P < 0.05, Fig. 5-B). In addition, the Dry-LS
samples showed a trend of consistent decrease of molecular diversity
after the incubation, whereas other samples did not show a clear trend
(Fig. 5-C). For plant- and microbial-derived compounds, the Dry-LS
samples showed consistent depletion of plant-derived compounds (i.e.
cutin and suberin biomarkers) and an accumulation of microbial-
derived compounds (i.e. short-chain saturated fatty acids) after the in­
cubation (Fig. 6-B). In contrast, the other samples had no clear trend
(Fig. 6-A and 6-C).
4. Discussion
4.1. Overview
Pyrolysis-GC/MS assisted with TMAH is a useful analytical method
to get an overview of SOM composition, but quantitative analyses are
difficult due to the lack of analytical standards and the different
response factors per compound (Derenne and Quéné, 2015; Shadkami
and Helleur, 2010). In addition, the TMAH derivatization only improves
the detection of certain compounds (e.g. fatty acids and lignin) in the
S. Yang et al. Soil & Tillage Research 215 (2022) 105203

Fig. 3. Soil organic carbon (SOC) contents and mineralization rates (MR) predicted by molecular composition of soil organic matter using partial least
square regressions (PLSR). A: PLSR model showing correlations of SOC, MR with relative abundances of molecular groups before the incubation, B: scores of
different soil samples corresponding to graph A, C: PLSR model showing correlations between MR and relative abundance changes of molecular groups after the
incubation, D: scores of different soil samples corresponding to graph C. n = 24 in total. The PLSR model for graph A and B explains 53.4% and 58.8% of the total
variation for predictors and respond variables (R2 (X) = 0.534 and R2 (Y) = 0.588), whereas the PLSR for graph C and D explains 72.0% of the total variation for both
predictors and respond variables (R2 (X) = R2 (Y) = 0.720). VIP ≥ 0.8 indicates significant predictors. FASS: short-chain saturated fatty acid, FASU: short-chain
unsaturated fatty acid, FAL: long-chain fatty acid, DA: α,ω-dioic acids, HA: ω-hydroxy acids, Alka: n-alkane, Alch: n-alcohol, Ester: ester, Lg: lignin, Ps: poly­
saccharides, N: N compounds, Other: other compounds, Wet: the wet site, Dry: the dry site, LS: limestone soils, AS: acid igneous soils, A: A horizons, B: B horizons.
Zone (1): Wet-LS-A samples have high SOC contents, which are associated to abundant Lg, Ps and Alka. Zone (2): Dry-LS samples have high SOC mineralization,
which are associated to abundant FASU. Zone (3): High SOC mineralization are correlated to large depletion of FASU and Alch, which is found in Dry-LS samples. In
addition, FASU is a better predictor than Alch for SOC mineralization (Supplementary Table S3).

subsequent GC/MS analysis (He et al., 2020). Therefore, the detected
molecular groups of SOM do not necessarily correspond to the original
composition of SOM in soil samples. To stay on the safe side, we focused
on relative shifts in patterns only, by limiting ourselves to direct com­
parisons between our samples and with the results of other studies that
used the same method.
For A horizons, significant differences in SOM composition between
different soils were only found for three molecular groups (i.e. lignin, n-
alkanes and ω-hydroxyl acids). In contrast, the other nine groups were
not significantly different (Fig. 2). The similarity in the molecular
composition of SOM in the A horizons can be explained by the similar
types of vegetation and land use (Table 1). Similarly, Nierop et al.
(2007) also reported homogeneity in SOM composition in the Ecua­
dorian Andes. For B horizons, which are only present in LS samples, SOM
molecular composition showed more differences compared to the A
horizons. Lignin derivates were almost depleted in subsoils (B horizons)
of the LS samples (Fig. 2). This is similar to the results of many studies as
summarized by Thevenot et al. (2010) and is consistent with the
situation in grassland soils of the Ecuadorian Andes as reported by
Nierop et al. (2007). In addition, the B horizons were characterized by a
large relative amount of short-chain saturated fatty acids (Fig. 2). This
suggests that the B horizons contain more microbial-derived compounds
than the A horizons, which is consistent with general views (Rumpel and
Kögel-Knabner, 2010). In the following discussion, we only compare the
A horizons between different soil plots because (1) B horizons were only
presented in the LS samples, and (2) SOM properties and stabilization
mechanisms can be different between A and B horizons (Rumpel and
Kögel-Knabner, 2010).
4.2. SOC contents and mineralization controlled by SOM molecular
composition
The high SOC contents in the Wet-LS samples, which are associated
with low SOC mineralization, are characterized by more abundant
derivates of lignin, polysaccharides and n-alkanes than other samples
(Figs. 1 and 3). After the incubation, these samples showed no selective

6
S. Yang et al. Soil & Tillage Research 215 (2022) 105203

Fig. 4. Principal component analysis (PCA) indicating changes in molecular composition of soil organic matter after the incubation. PC1 and PC2 explained
54.6% of total variance. Arrows indicate changes in molecular composition after the incubation for individual samples. FASS: short-chain saturated fatty acid, FASU:
short-chain unsaturated fatty acid, FAL: long-chain fatty acid, DA: α,ω-dioic acids, HA: ω-hydroxy acids, Alka: n-alkanes, Alch: n-alcohols, Ester: ester, Lg: lignin, Ps:
polysaccharides, N: n compounds, Other: other compounds, Wet: the wet site, Dry: the dry site, LS: limestone soils, AS: acid igneous soils, A: A horizons, B: B horizons.

Fig. 5. Molecular diversity and its change after the incubation as correlated with soil organic carbon (SOC) contents and mineralization. A: original
molecular diversity correlated with SOC content and SOC mineralization, B: decline of molecular diversity after the incubation as correlated with SOC content and
SOC mineralization, C: decline of molecular diversity after the incubation for individual samples. Significant level: P < 0.05, Wet: the wet site, Dry: the dry site, LS:
limestone soils, AS: acid igneous soils, A: A horizons, B: B horizons.

accumulation or depletion of any compound groups, as indicated by no
clear pattern of changes in compound groups (Figs. 4 and 6). Therefore,
the abundant and stable SOC in the Wet-LS samples can be explained by
the accumulation of lignin, n-alkanes and polysaccharides, as well as the
fact that they were not clearly depleted after the incubation compared to
the Dry-LS samples.
The low SOC contents and high SOC mineralization in the Dry-LS
(Fig. 1) suggest that SOC is not well stabilized. The high SOC
mineralization was further associated with a greater depletion of un­
saturated fatty acids (Figs. 2 and 3). This is in line with previous studies
showing that unsaturated fatty acids are more susceptible to decompo­
sition in soils (Moucawi et al., 1981; Yang et al., 2020a). After the in­
cubation, the selective depletion of compound groups, except for
short-chain saturated fatty acids and N-compounds (Fig. 4), indicates
the low capacity of the Dry-LS samples to stabilize certain OM com­
pounds (e.g. monomers of cutin, suberin and lignin).

7
S. Yang et al.
Fig. 6. Changes in relative abundances of short-chain saturated fatty acids (FASS) and the sum of α,ω-dioic acids and ω-hydroxyl acids (DA þ HA) after
the incubation. FASS is used to estimate microbial-derived compounds, whereas DA + HA are cutin and suberin biomarkers for the estimation of plant-derived
compounds. Arrows indicate the shifts in FASS and DA + HA for individual samples after the incubation. SOC: soil organic carbon, OM: organic matter, Wet: the
wet site, Dry: the dry site, LS: limestone soils, AS: acid igneous soils.
In our previous studies on these soils, we found that SOM stabiliza­
tion was controlled by organo-mineral interactions, whereas SOM
occluded in aggregates played a minor role (Yang et al., 2020b, 2020c).
The results of our present study suggest that SOC contents and miner­
alization are associated with the selective preservation of different OM
molecules. The higher contents of lignin in the Wet-LS samples (high
SOC content) can be explained by the higher pH values compared to
other samples (Table 1), as the decomposition of lignin is closely related
to white rot fungi that prefer acidic soils (Thevenot et al., 2010). This is
consistent with results reported for instance by Brock et al. (2019),
showing that lignin degradation was more pronounced in acidic soils.
The selective preservation of polysaccharides and n-alkanes in the
Wet-LS samples might be related to soil mineralogy, as the soils had high
contents of pedogenic Fe and Al oxides and Ca2+ bridges (Yang et al.,
2020b, 2020c). These oxides and Ca2+ bridges can (1) promote stabili­
zation of SOM compounds (e.g. polysaccharides and lignin) by associ­
ation with mineral surfaces via ligand exchange and cation bridges
(Lützow et al., 2006), and (2) improve soil f to protect compounds (e.g.
n-alkanes) by occlusion in aggregates (Rowley et al., 2018). For the
Dry-LS samples (low SOC content), the high SOC mineralization and the
depletion of most compounds after the incubation are consistent with
the previous findings that these soils had low capacity to stabilize
organic molecules (Yang et al., 2020b). In addition, the abundant un­
saturated fatty acids and other less stabilized SOM might promote mi­
crobial activities, through processes similar to priming effects
(Kuzyakov, 2010), to decompose substrates with relatively high acti­
vation energy (see Section 4.4).
4.3. Stabilization of plant- versus microbial-derived OM
Lignin, n-alkanes, α,ω-dioic acids and ω-hydroxyl acids are widely
used as biomarkers for plant-derived OM (Amelung et al., 2008). In
contrast, short-chain fatty acids, especially branch-chain fatty acids
identified in our samples (Supplementary Table S4), have a major origin
of microorganisms as parts of the cell membrane (Kaneda, 1991; Zelles,
1999). In addition, compounds such as polysaccharides and N-com­
pounds are ubiquitous with both plant and microbial origins (Barré
et al., 2018).
We observed no selective degradation of plant-derived OM or accu­
mulation of microbial-derived OM in the Wet-LS, Wet-AS and Dry-AS
samples (high and medium SOC contents). This is indicated by (1) no
clear trend in depletion of lignin, cutin and suberin biomarkers after the
incubation, and (2) no clear trend in accumulation of short-chain satu­
rated fatty acids after the incubation, compared to the Dry-LS samples
(Figs. 4 and 6). A possible explanation is that plant-derived OM is well
8
Soil & Tillage Research 215 (2022) 105203
protected in these soils and receives less microbial transformation, as
many monomers of plant-derived compounds (i.e. lignin, cutin and su­
berin) can be stabilized by association with mineral surfaces via hy­
droxyl and carboxyl groups (Angst et al., 2021; Lützow et al., 2006). In
these soils, especially the Wet-LS samples, plant-derived OM can be
stabilized by interacting with active Fe- and Al-oxides, as well as Ca2+
bridges (see Section 4.2.). In addition, the higher pH values in the
Wet-LS samples (Table 1) suggest a reduction of fungal activity, which
can restrict the decomposition of plant-derived macromolecules
(Koranda et al., 2014). In contrast, Dry-LS samples (low SOC contents)
showed a strong transformation from plant- to microbial-derived com­
pounds, as indicated by the depletion of most plant-derived OM and an
accumulation of short-chain fatty acids after the incubation (Figs. 4 and
6). The strong microbial transformation might be attributed to a poor
stabilization of the plant-derived compounds in these soils, as mentioned
in Section 4.2. Moreover, it is also possible that these soils provide
favorable pH and moisture to microorganisms for the intensive micro­
bial transformation of plant-derived OM.
Recent studies found that a large proportion of SOM consists of
stabilized compounds with a microbial origin (Liang et al., 2019; Miltner
et al., 2012). However, the relative contributions of plant- versus
microbial-derived SOM are largely dependent on environmental and
pedogenic factors (Angst et al., 2021). Our results are consistent with
this and suggest that the variation in plant- and microbial-derived SOM
can be attributed to the degree of overall preservation of plant-derived
OM. Similarly, Barré et al. (2018) found that persistent SOM is
contributed by a substantial and variable fraction of plant-derived OM.
Moreover, it is also possible that the persistence of microbial-derived
OM is controlled by factors such as soil pH, texture or microbial com­
munities (Hu et al., 2020).
4.4. Molecular diversity
Our results highlighted the relationship between the molecular di­
versity and the decomposition of SOM, as indicated by the fact that the
molecular diversity declined for the Dry-LS samples (low and unstable
SOC) during the incubation but had no consistent trend for the other soil
samples with medium to high SOC contents (Fig. 5). To our knowledge,
this has not been reported before, but relevant studies showed that the
molecular diversity of SOM was controlled by the type of OM input
(Chen et al., 2021), soil mineralogy (Hall et al., 2020) and ecosystem
characteristics (Tfaily et al., 2017). Also, the molecular diversity affects
microbial decomposition and transformation of SOM and, therefore,
controls the stabilization of SOM (Cotrufo et al., 2013; Lehmann et al.,
2020).
S. Yang et al.
We propose two possible explanations for the significantly decreased
molecular diversity in the Dry-LS samples. The first explanation is that
most of the molecular groups were poorly stabilized and depleted under
the detection limit of the instrument for the Dry-LS samples after the
incubation. Therefore, we observed declined molecular diversity for
these samples. Alternatively, the presence of abundant unsaturated fatty
acids, which are easily degradable and act as potential priming sub­
strates, could promote the activity and functional diversity of soil mi­
croorganisms (Kuzyakov, 2010). The activated microbial community
would be able to utilize substrates with relatively higher activation
energy (e.g. recalcitrant or weakly stabilized compounds) and, there­
fore, lead to decreased molecular diversity of SOM after the incubation
(Blagodatskaya and Kuzyakov, 2008; Kuzyakov, 2010). To test the two
explanations, further experiments on the priming effects of unsaturated
fatty acids might be helpful.
4.5. Practical issues for soils in the Peruvian Andes
Tropical alpine grasslands, including Peruvian Andes, provide
important ecosystem services including carbon sequestration and food
provision (Chai et al., 2020; Rolando et al., 2017). To maintain SOC
storage and soil fertility, local farmers perform land management such
as a rotation in the order of cultivation, abandonment and grazing
within a period of 2–5 years (Yang et al., 2018). However, the effec­
tiveness of these measures depends on whether the OM from plant input
can be stabilized in soils. Our results indicate that the decomposition of
plant-derived OM in these soils is dependent on parent material and
precipitation (Fig. 7). This suggests that traditional management to in­
crease SOC storage and fertility (e.g. green manure addition) might have
restricted effects in some soils (e.g. Dry-LS) because a part of
plant-derived OM will be quickly mineralized rather than stabilized. As
plant-derived compounds account for ≥ 50% of the SOM and contribute
to stabilized SOC pools (Angst et al., 2021), we propose that environ­
mental factors (e.g. lithology and precipitation) should be considered to
Fig. 7. Overview of relationships between soil organic carbon (SOC) contents, stability and molecular composition of soil organic matter (OM) in different
soil samples. Only A horizon samples are involved. Wet: the wet site, Dry: the dry site, LS: limestone soils, AS: acid igneous soils.
9
Soil & Tillage Research 215 (2022) 105203
maintain the ecosystem services such as carbon sequestration in the
studied region.
Climate change can have crucial impacts on SOC sequestration in the
Peruvian Andes. The shifting of forest lines to higher altitudes caused by
warming in the Andes (Tovar et al., 2013) will potentially affect the
SOM input and stabilization. In general, the shift from grassland to forest
will increase the input of low-quality OM and the proportion of plant- to
microbial-derived OM in soils (Angst et al., 2021). Because soils in our
study area had different capacities to stabilize plant-derived OM, we
expect contrasting feedbacks of SOC sequestration in these soils (e.g.
Wet-LS vs. Dry-LS) to the shift from grassland to forest caused by climate
change. The change in precipitation, in the short term, can alter the soil
microenvironment and the microbial decomposition and transformation
of SOM. For the long-term effects, precipitation controls the pedogenesis
such as weathering processes and further controls soil mineralogy and
SOM stabilization. As the effects of precipitation are stronger for LS
samples compared to AS samples, the LS samples are more likely to be
affected than the AS samples by climate change regarding the shift of
precipitation.
5. Conclusions
Results of this study were summarized as an overview in Fig. 7. In the
studied region in the Peruvian Andes, soil samples with high SOC con­
tents and low SOC mineralization (Wet-LS) had abundant derivates of
lignin, polysaccharides and n-alkanes. After the incubation, we observed
neither selective decomposition of any compound groups nor decline of
molecular diversity. In contrast, soil samples with low SOC contents and
high SOC mineralization (Dry-LS) showed a strong transformation from
plant-derived OM to microbial-derived OM and declined molecular di­
versity after the incubation. Overall, SOC mineralization was positively
correlated to the decline of unsaturated fatty acids and that of molecular
diversity after the incubation. Our results indicate that variation of SOC
contents and mineralization are associated with selective preservation of
S. Yang et al.
SOM molecules and changes in molecular diversity under microbial
decomposition.
The selective preservation and depletion of plant-derived OM in our
samples help to understand the stabilization and the relative contribu­
tion of plant- versus microbial-derived OM in different environmental
conditions. For SOM stabilization mechanisms, future studies could
focus on specific stabilization mechanisms for different SOM molecules
regarding their interaction with soil minerals and aggregates. Before the
extrapolation of our conclusions to other soils, a crucial check is
required for the applicability of our findings to soils with different
environmental conditions. In addition, long-term experiments could be
helpful to evaluate SOC stability and persistence as our 76-day incuba­
tion experiment could not fully represent the long-term stability of SOC.
Finally, for practical issues, we propose that environmental factors
should be considered for the management of ecosystem services such as
SOC sequestration in the studied region of the Peruvian Andes.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Acknowledgements
We thank for the help from Amedée Wever, Rutger van Hall, Eva de
Rijke and Chiara Cerli for the laboratory work. We would also thank Olaf
Brock, Klaas Nierop, Liz Veerman and Xiang Wang, as they kindly gave
their support for the peak identification, data interpretation, statistics
and incubation experiment. Soil sampling in the field was supported by
the Mountain Institute (TMI) and Dr. Jorge Recharte. We also thank the
three anonymous reniewers for their help to improve this manuscript.
This work is supported by the Institute for Biodiversity and Ecosystem
Dynamics (IBED) of the University of Amsterdam, the China Scholarship
Council (CSC) and the project International Mobilities of Researchers
and Administrative Workers of the Biology Centre (MEMOVA,
CZ.02.2.69/0.0/0.0/18_053/0016982) from the European Structural
and Investing Funds Operational Programme Research, Development
and Education.
Appendix A. Supporting information
Supplementary data associated with this article can be found in the
online version at doi:10.1016/j.still.2021.105203.
References
Abdi, H., 2003. Partial least-square regression. Encycl. Res. Methods Soc. Sci. 6. https://
doi.org/10.1002/9783527809080.cataz12383.
Amelung, W., Brodowski, S., Sandhage-Hofmann, A., Bol, R., 2008. Chapter 6 Combining
biomarker with stable isotope analyses for assessing the transformation and turnover
of soil organic matter. Advances in Agronomy, 1st ed.,. Elsevier Inc.,. https://doi.
org/10.1016/S0065-2113(08)00606-8.
Angst, G., Mueller, K.E., Nierop, K.G.J., Simpson, M.J., 2021. Plant- or microbial-derived
? A review on the molecular composition of stabilized soil organic matter. Soil Biol.
Biochem. 156, 108189 https://doi.org/10.1016/j.soilbio.2021.108189.
Barré, P., Quénéa, K., Vidal, A., Cécillon, L., Christensen, B.T., Kätterer, T.,
Macdonald, A., Petit, L., Plante, A.F., van Oort, F., Chenu, C., 2018. Microbial and
plant-derived compounds both contribute to persistent soil organic carbon in
temperate soils. Biogeochemistry 140, 81–92. https://doi.org/10.1007/s10533-018-
0475-5.
Blagodatskaya, E., Kuzyakov, Y., 2008. Mechanisms of real and apparent priming effects
and their dependence on soil microbial biomass and community structure: critical
review. Biol. Fertil. Soils 45, 115–131. https://doi.org/10.1007/s00374-008-0334-y.
Brock, O., Kooijman, A., Nierop, K.G.J., Muys, B., Vancampenhout, K., Jansen, B., 2019.
Disentangling the effects of parent material and litter input chemistry on molecular
soil organic matter composition in converted forests in Western Europe. Org.
Geochem. 134, 66–76.
Buytaert, W., Cuesta-Camacho, F., Tobón, C., 2011. Potential impacts of climate change
on the environmental services of humid tropical alpine regions. Glob. Ecol. Biogeogr.
20, 19–33. https://doi.org/10.1111/j.1466-8238.2010.00585.x.
10
Soil & Tillage Research 215 (2022) 105203
Chai, L.L., Hernandez-Ramirez, G., Hik, D.S., Barrio, I.C., Frost, C.M., Chinchilla Soto, C.,
Esquivel-Hernández, G., 2020. A methane sink in the Central American high
elevation páramo: topographic, soil moisture and vegetation effects. Geoderma 362,
114092. https://doi.org/10.1016/j.geoderma.2019.114092.
Chen, S., Ding, Y., Xia, X., Feng, X., Liu, X., Zheng, J., Drosos, M., Cheng, K., Bian, R.,
Zhang, X., Li, L., Pan, G., 2021. Amendment of straw biochar increased molecular
diversity and enhanced preservation of plant derived organic matter in extracted
fractions of a rice paddy. J. Environ. Manag. 285, 112104 https://doi.org/10.1016/j.
jenvman.2021.112104.
Chong, I.G., Jun, C.H., 2005. Performance of some variable selection methods when
multicollinearity is present. Chemom. Intell. Lab. Syst. 78, 103–112. https://doi.org/
10.1016/j.chemolab.2004.12.011.
Cotrufo, M.F., Wallenstein, M.D., Boot, C.M., Denef, K., Paul, E., 2013. The Microbial
Efficiency-Matrix Stabilization (MEMS) framework integrates plant litter
decomposition with soil organic matter stabilization: do labile plant inputs form
stable soil organic matter? Glob. Change Biol. 19, 988–995. https://doi.org/
10.1111/gcb.12113.
Derenne, S., Quéné, K., 2015. Analytical pyrolysis as a tool to probe soil organic matter.
J. Anal. Appl. Pyrolysis 111, 108–120. https://doi.org/10.1016/j.jaap.2014.12.001.
Dungait, J.A.J., Hopkins, D.W., Gregory, A.S., Whitmore, A.P., 2012. Soil organic matter
turnover is governed by accessibility not recalcitrance. Glob. Chang. Biol. 18,
1781–1796. https://doi.org/10.1111/j.1365-2486.2012.02665.x.
Geo GPS Perú, 2014. GEO GPS PERÚ: Base de datos Perú - Shapefile - *.shp - MINAM -
IGN - Límites Políticos [WWW Document]. URL 〈https://www.geogpsperu.com/2
013/09/cuadro-de-empalme-de-la-cartografia.html〉.
Gunina, A., Kuzyakov, Y., 2015. Sugars in soil and sweets for microorganisms: review of
origin, content, composition and fate. Soil Biol. Biochem. 90, 87–100. https://doi.
org/10.1016/j.soilbio.2015.07.021.
Hall, S.J., Ye, C., Weintraub, S.R., Hockaday, W.C., 2020. Molecular trade-offs in soil
organic carbon composition at continental scale. Nat. Geosci. 13, 687–692. https://
doi.org/10.1038/s41561-020-0634-x.
Han, L., Sun, K., Jin, J., Xing, B., 2016. Some concepts of soil organic carbon
characteristics and mineral interaction from a review of literature. Soil Biol.
Biochem. 94, 107–121. https://doi.org/10.1016/j.soilbio.2015.11.023.
He, Y., Buch, A., Szopa, C., Williams, A.J., Millan, M., Guzman, M., Freissinet, C.,
Malespin, C., Glavin, D.P., Eigendbrode, J.L., Coscia, D., Teinturier, S., Pin, lu,
Cabane, M., Mahaffy, P.R., 2020. The search for organic compounds with TMAH
thermochemolysis: from Earth analyses to space exploration experiments. TrAC -
Trends Anal. Chem. 127. https://doi.org/10.1016/j.trac.2020.115896.
Hu, Y., Zheng, Q., Noll, L., Zhang, S., Wanek, W., 2020. Direct measurement of the in situ
decomposition of microbial-derived soil organic matter. Soil Biol. Biochem. 141,
107660 https://doi.org/10.1016/j.soilbio.2019.107660.
Husson, F., Josse, J., Le, S., Mazet, J., 2020. FactoMineR: Multivariate Exploratory Data
Analysis and Data Mining [WWW Document]. R Packag. Version 3.5.0. URL 〈https
://cran.r-project.org/package=FactoMineR〉.
Jansen, B., Wiesenberg, G.L.B., 2017. Opportunities and limitations related to the
application of plant-derived lipid molecular proxies in soil science. Soil 3, 211–234.
https://doi.org/10.5194/soil-3-211-2017.
Joergensen, R.G., 2018. Amino sugars as specific indices for fungal and bacterial residues
in soil. Biol. Fertil. Soils 54, 559–568. https://doi.org/10.1007/s00374-018-1288-3.
Kallenbach, C.M., Frey, S.D., Grandy, A.S., 2016. Direct evidence for microbial-derived
soil organic matter formation and its ecophysiological controls. Nat. Commun. 7,
1–10. https://doi.org/10.1038/ncomms13630.
Kaneda, T., 1991. Iso- and anteiso-fatty acids in bacteria: biosynthesis, function, and
taxonomic significancet. Microbiol. Rev. 55, 288–302. https://doi.org/10.14743/
apem2017.1.239.
Koranda, M., Kaiser, C., Fuchslueger, L., Kitzler, B., Sessitsch, A., Zechmeister-
Boltenstern, S., Richter, A., 2014. Fungal and bacterial utilization of organic
substrates depends on substrate complexity and N availability. FEMS Microbiol.
Ecol. 87, 142–152. https://doi.org/10.1111/1574-6941.12214.
Kögel-Knabner, I., 2017. The macromolecular organic composition of plant and
microbial residues as inputs to soil organic matter: fourteen years on. Soil Biol.
Biochem. 105, A3–A8. https://doi.org/10.1016/j.soilbio.2016.08.011.
Kögel-Knabner, I., 2002. The macromolecular organic composition of plant and
microbial residues as inputs to soil organic matter. Soil Biol. Biochem 34, 139–162.
Kuzyakov, Y., 2010. Priming effects: interactions between living and dead organic
matter. Soil Biol. Biochem. 42, 1363–1371. https://doi.org/10.1016/j.
soilbio.2010.04.003.
Lehmann, J., Hansel, C.M., Kaiser, C., Kleber, M., Maher, K., Manzoni, S., Nunan, N.,
Reichstein, M., Schimel, J.P., Torn, M.S., Wieder, W.R., Kögel-Knabner, I., 2020.
Persistence of soil organic carbon caused by functional complexity. Nat. Geosci. 13,
529–534. https://doi.org/10.1038/s41561-020-0612-3.
Lehmann, J., Kleber, M., 2015. The contentious nature of soil organic matter. Nature
528, 60–68. https://doi.org/10.1038/nature16069.
Liang, C., Amelung, W., Lehmann, J., Kästner, M., 2019. Quantitative assessment of
microbial necromass contribution to soil organic matter. Glob. Change Biol. 25,
3578–3590. https://doi.org/10.1111/gcb.14781.
Liang, C., Schimel, J.P., Jastrow, J.D., 2017. The importance of anabolism in microbial
control over soil carbon storage. Nat. Microbiol. 2, 1–6. https://doi.org/10.1038/
nmicrobiol.2017.105.
Liland, K.H., Mehmood, T., Sæbø, S., 2020. plsVarSel: Variable Selection in Partial Least
Squarese [WWW Document]. R Packag. Version 2.1.0. URL 〈https://cran.r-project.or
g/package=plsVarSel〉.
Lützow, M.V., Kögel-Knabner, I., Ekschmitt, K., Matzner, E., Guggenberger, G.,
Marschner, B., Flessa, H., 2006. Stabilization of organic matter in temperate soils:
S. Yang et al.
mechanisms and their relevance under different soil conditions - a review. Eur. J.
Soil Sci. 57, 426–445. https://doi.org/10.1111/j.1365-2389.2006.00809.x.
Miltner, A., Bombach, P., Schmidt-Brücken, B., Kästner, M., 2012. SOM genesis:
Microbial biomass as a significant source. Biogeochemistry 111, 41–55. https://doi.
org/10.1007/s10533-011-9658-z.
Moucawi, J., Fustec, E., Jambu, P., Jacquesy, R., Fuste, E., Jambi, P., 1981.
Decomposition of lipids in soils: free and esterified fatty acids, alcohols and ketones.
Soil Biol. Biochem. 13, 461–468. https://doi.org/10.1016/0038-0717(81)90035-3.
Nierop, K.G.J., Tonneijck, F.H., Jansen, B., Verstraten, J.M., 2007. Organic matter in
volcanic ash soils under forest and páramo along an Ecuadorian altitudinal transect.
Soil Sci. Soc. Am. J. 71, 1119–1127. https://doi.org/10.2136/sssaj2006.0322.
Otto, A., Simpson, M.J., 2006. Sources and composition of hydrolysable aliphatic lipids
and phenols in soils from western Canada. Org. Geochem. 37, 385–407. https://doi.
org/10.1016/j.orggeochem.2005.12.011.
Rolando, J.L., Turin, C., Ramírez, D.A., Mares, V., Monerris, J., Quiroz, R., 2017. Key
ecosystem services and ecological intensification of agriculture in the tropical high-
Andean Puna as affected by land-use and climate changes. Agric. Ecosyst. Environ.
236, 221–233. https://doi.org/10.1016/j.agee.2016.12.010.
Rowley, M.C., Grand, S., Verrecchia, É.P., 2018. Calcium-mediated stabilisation of soil
organic carbon. Biogeochemistry 137, 27–49. https://doi.org/10.1007/s10533-017-
0410-1.
Rumpel, C., Kögel-Knabner, I., 2010. Deep soil organic matter—a key but poorly
understood component of terrestrial C cycle. Plant Soil 338, 143–158. https://doi.
org/10.1007/s11104-010-0391-5.
Sanchez, G., 2012. plsdepot: Partial Least Squares (PLS) Data Analysis Methods [WWW
Document]. R Packag. Version 2.15.1.
Schmidt, M.W.I., Torn, M.S., Abiven, S., Dittmar, T., Guggenberger, G., Janssens, I.A.,
Kleber, M., Kögel-Knabner, I., Lehmann, J., Manning, D.A.C., Nannipieri, P.,
Rasse, D.P., Weiner, S., Trumbore, S.E., 2011. Persistence of soil organic matter as an
ecosystem property. Nature 478, 49–56. https://doi.org/10.1038/nature10386.
Schulten, H.R., Schnitzer, M., 1997. The chemistry of soil organic nitrogen: a review.
Biol. Fertil. Soils 26, 1–15. https://doi.org/10.1007/s003740050335.
Shadkami, F., Helleur, R., 2010. Recent applications in analytical thermochemolysis.
J. Anal. Appl. Pyrolysis 89, 2–16. https://doi.org/10.1016/j.jaap.2010.05.007.
Sokol, N.W., Sanderman, J., Bradford, M.A., 2019. Pathways of mineral-associated soil
organic matter formation: Integrating the role of plant carbon source, chemistry, and
point of entry. Glob. Change Biol. 25, 12–24. https://doi.org/10.1111/gcb.14482.
Spellerberg, I.F., Fedor, P.J., 2003. A tribute to Claude-Shannon (1916-2001) and a plea
for more rigorous use of species richness, species diversity and the “Shannon-
11
Soil & Tillage Research 215 (2022) 105203
Wiener” Index. Glob. Ecol. Biogeogr. 12, 177–179. https://doi.org/10.1046/j.1466-
822X.2003.00015.x.
Tfaily, M.M., Chu, R.K., Toyoda, J., Tolić, N., Robinson, E.W., Paša-Tolić, L., Hess, N.J.,
2017. Sequential extraction protocol for organic matter from soils and sediments
using high resolution mass spectrometry. Anal. Chim. Acta 972, 54–61. https://doi.
org/10.1016/j.aca.2017.03.031.
Thevenot, M., Dignac, M.-F.F., Rumpel, C., 2010. Fate of lignins in soils: a review. Soil
Biol. Biochem. 42, 1200–1211. https://doi.org/10.1016/j.soilbio.2010.03.017.
Tovar, C., Seijmonsbergen, A.C., Duivenvoorden, J.F., 2013. Monitoring land use and
land cover change in mountain regions: an example in the Jalca grasslands of the
Peruvian Andes. Landsc. Urban Plan. 112, 40–49. https://doi.org/10.1016/j.
landurbplan.2012.12.003.
Wiesenberg, G.L.B., Dorodnikov, M., Kuzyakov, Y., 2010. Source determination of lipids
in bulk soil and soil density fractions after four years of wheat cropping. Geoderma
156, 267–277. https://doi.org/10.1016/j.geoderma.2010.02.026.
WRB, I.W.G., 2014. World reference base for soil resources 2014. International Soil
Classification System for Naming Soils and Creating Legends for Soil Maps. FAO,,
Rome.
Yang, S., Cammeraat, E., Jansen, B., den Haan, M., van Loon, E., Recharte, J., 2018. Soil
organic carbon stocks controlled by lithology and soil depth in a Peruvian alpine
grassland of the Andes. Catena 171, 11–21. https://doi.org/10.1016/j.
catena.2018.06.038.
Yang, S., Jansen, B., Absalah, S., Kalbitz, K., Cammeraat, E.L.H., 2020a. Selective
stabilization of soil fatty acids related to their carbon chain length and presence of
double bonds in the Peruvian Andes. Geoderma 373, 114414. https://doi.org/
10.1016/j.geoderma.2020.114414.
Yang, S., Jansen, B., Absalah, S., Van Hall, R.L., Kalbitz, K., Cammeraat, E.H.E., 2020b.
Lithology-and climate-controlled soil aggregate-size distribution and organic carbon
stability in the Peruvian Andes. Soil 6, 1–15. https://doi.org/10.5194/soil-6-1-2020.
Yang, S., Jansen, B., Kalbitz, K., Chunga Castro, F.O., van Hall, R.L., Cammeraat, E.L.H.,
2020c. Lithology controlled soil organic carbon stabilization in an alpine grassland
of the Peruvian Andes. Environ. Earth Sci. 79, 66. https://doi.org/10.1007/s12665-
019-8796-9.
Zelles, L., 1999. Fatty acid patterns of phospholipids and lipopolysaccharides in the
characterisation of microbial communities in soil: a review. Biol. Fertil. Soils 29,
111–129. https://doi.org/10.1007/s003740050533.
Zhu, X., Jackson, R.D., DeLucia, E.H., Tiedje, J.M., Liang, C., 2020. The soil microbial
carbon pump: from conceptual insights to empirical assessments. Glob. Chang. Biol.
26, 6032–6039. https://doi.org/10.1111/gcb.15319.