Introduction
Over the last decade, numerous studies have demonstrated the role of soil
microorganisms in regulating the fate and transformation of organic
compounds. Soil microorganisms produce exoenzymes to carry out the primary
degradation of plant as well as microbial polymers to monomers. Further
transformations of monomers then take place within the microbial cells.
Monomeric substances are taken up by the living microorganisms and are
partly mineralized to CO2, while part is assimilated into cell
polymers and ultimately incorporated into soil organic matter (SOM) after
cell death (Kindler et al., 2006). Understanding the fate of substances
originated from plants and microbial residues into living biomass is
therefore crucial for estimating the recycling of carbon (C) in soil and its
stabilization as SOM.
Living microbial biomass (MB) is a highly active and heterogeneous pool
(Malik et al., 2015), although it accounts for only 2–4 % of the total SOM
(Jenkinson and Ladd, 1981). Heterogeneity is evident at the level of single
cells in the various cellular compartments with different properties,
structures and biochemistry: from the highly heterogeneous cytosol (Malik
et al., 2013) to well-structured cell membranes and cell walls. Due to
their chemical composition and functions, compounds of cell membranes
(phospholipid fatty acids (PLFAs)) and cell walls (amino sugars) have
different turnover times within the cell as well as different stabilities
within SOM.
Organic compounds that are taken up by microorganisms first enter the
cytosol (Gottschalk, 1979), which has a high heterogeneity in composition
(includes components of various chemical structure and molecular weight).
However, due to the heterogeneity of this pool, the calculated C turnover
time is a mean of C turnover times in various components. The calculated
turnover time of intact PLFAs in soil after microbial death is 2.8 days
(Kindler et al., 2009); resulting PLFAs are mainly used to characterize the
living microorganisms (Frostegard et al., 2011; Rethemeyer, 2004). However,
no data concerning turnover time of C in PLFA of living biomass are
currently published. The formation of amino sugars from plant biomass is
relatively rapid at 6.2–9.0 days (Bai et al., 2013), whereas their
turnover times in soil vary between 6.5 and 81.0 yr-1 (Glaser et al.,
2006). Thus, PLFAs and amino sugars can be used to trace the fate of C
within the living microorganisms as well as to estimate the contribution of
microbial residues to SOM (Schmidt et al., 2007).
Some cell compartments, such as the cytosol, are not specific for various
microbial groups, whereas phospholipids are partly specific and consequently
can be used to estimate microbial community structure. Thus, PLFAs of
bacterial (i16:0, a16:0, i15:0, a15:0, 16:1ω7, 18:1ω7) and fungal communities (18:2ω6,9; 18:3ω6,9,12;
16:1ω5) are used to draw conclusions about the qualitative
composition of living microbial communities, their contribution to
utilization of C by various origin (plant or microbial) and to understand
trophic interactions within the soil (Ruess et al., 2005). In contrast, amino
sugars (glucosamine, galactosamine, mannosamine and muramic acid) are usually
used to assess the contributions of bacterial and fungal residues to
SOM (Engelking et al., 2007; Glaser et al., 2004). Muramic acid is of
bacterial origin, whereas glucosamine is derived from both fungal and
bacterial cell walls (Glaser et al., 2004). Galactosamine is more abundant in
fungal than in bacterial cell walls (Engelking et al., 2007; Glaser et al.,
2004).
Bacteria and fungi have different chemical composition, which strongly
contributes to their turnover rates in soil: for bacteria it constitutes 2.3–33
days, whereas for fungi it accounts for 130–150 days (Moore et al., 2005;
Rousk and Baath, 2007; Waring et al., 2013). Despite the fact that the turnover of
microorganisms directly affects the C turnover rates in intercellular
compounds (cell membrane and cell wall biomarkers), this relationship has
rarely been investigated so far. However, the comparison of C turnover for
cell membrane and cell wall components can be used to characterize the
contribution of various microbial groups to medium-term C utilization and to
the stabilization of microbially derived C in SOM.
Combination of PLFAs and amino-sugar biomarker analyses, as well as cytosol
C measurement with isotope tracing techniques (based on 13C natural
abundance or 13C / 14C labelling), has been used in various
studies to characterize organic C utilization by microbial community (Bai et
al., 2013; Brant et al., 2006). However, to date no systematic studies have
compared these contrasting cell compartments in a single soil within a C
turnover experiment. Therefore, this study aimed to examine C allocation to
various cell compartments following 13C labelling with a ubiquitous
monomer, glucose. Glucose has a higher concentrations in the soil solution
compared to other low-molecular-weight organics (Fischer et al., 2007), due
to its diverse origin: from cellulose decomposition, presence in
rhizodeposition (Derrien et al., 2004; Gunina and Kuzyakov, 2015), and
synthesis by microorganisms. It is also used by most of the microbial groups
and is thus the most suitable substance for such a study.
We analysed glucose-derived 13C partitioning into the cytosol, cell
membranes and cell walls in order to evaluate the turnover time of C in each pool
and to assess the contribution of bacterial and fungal biomass to SOM. We
hypothesized that (1) turnover times of C in pools follow the order
cytosol < PLFA < amino sugars, because substances taken
up by cells first are transported by membrane proteins into cytosol, from
where they get distributed to other cellular pools, and (2) recovery of
13C glucose should be faster and higher for bacterial than for fungal
biomarkers, because bacterial biomass has a faster cell turnover than fungal
biomass.
Material and methods
Field site and experimental design
The 13C labelling field experiment was established at an agricultural
field trial in Hohenpölz, Germany (49∘54′ N,
11∘08′ E; at 500 m a.s.l.). Triticale, wheat and barley were
cultivated by a rotation at the chosen site. The soil type was a loamy Haplic
Luvisol (IUSS Working group WRB, 2014) and had the following
chemical properties in the uppermost 10 cm: total organic C content
1.5 %, C / N 10.7, pH 6.6, clay content 22 %, CEC 13 cmolC kg-1. The annual precipitation is 870 mm and mean annual temperature is
+7 ∘C.
In summer 2010, following harvest of the triticale, columns (diameter 10 cm
and height 13 cm) were installed to a depth of 10 cm. Each column contained
1.5 kg of soil and bulk density was 1.36 g cm-3. The 50 mL of
uniformly labelled 13C glucose (99 atom % 13C) was injected
into the columns via a syringe at five points inside the column to spread the
tracer homogeneously. The syringe was equipped with a special pipe (13 cm long) with perforations along the whole length, while the end of the pipe was
sealed to prevent glucose injection below the column. Each column received
93.4 µmol 13C of tracer (0.06 µmol
13C g-1soil), and similar amounts of non-labelled glucose were
applied to the control columns to make the experimental conditions equal.
The concentration was chosen to trace the natural pool of glucose in soil
solution (Fischer et al., 2007), rather than stimulate the activity or growth
of microorganisms.
The experiment was done in four field replicates, which were organized in a
randomized block design. Labelled and control columns were present within
each block. For the first 10 days of the experiment rainfall was excluded
by a protective shelter, which was removed thereafter, with the experiment running for
50 days in total. The rainfall was excluded to prevent the added glucose from
being leached out from the soil profile, due to processes of microbial uptake
progressing slower in the field conditions than in the controlled laboratory. After 3,
10 and 50 days, separate soil columns (four columns where 13C was
applied and four control columns) were destructively sampled. The columns had
no vegetation by the collecting time, nor was there any when the 13C glucose
was applied.
The soil was removed from the column and weighed, and the water content was
determined in a subsample. Soil moisture was determined by drying samples for
24 h at 105 ∘C and was essentially constant during the experiment,
ranging between 21 and 25 % (25.7 ± 1.2 (3 days), 23.3 ± 1.3
(10 days), and 21.4 ± 0.7 (50 days)). Each soil sample was sieved to
< 2 mm and divided into three parts. One part was stored frozen
(-20 ∘C) for PLFA analysis, another was cooled (+5 ∘C)
(over 1 week) before the microbial biomass analysis, and the rest was
freeze-dried and used for amino-sugar analysis and for measurement of the
total amount of glucose-derived 13C remaining in the soil.
Bulk soil δ13C analysis
The soil for the δ13C analysis was milled and δ13C values
of bulk SOM were determined using a Euro EA elemental analyser (Eurovector,
Milan, Italy) unit coupled via a ConFlo III interface (Thermo Fisher,
Bremen, Germany) to a Delta V Advantage isotope
ratio mass spectrometer (IRMS; Thermo Fisher, Bremen,
Germany). The amount of glucose-derived 13C remaining in the soil was
calculated based on a mixing model (Eqs. 1 and 2), where the amount of C in
the background sample in Eq. (1) was substituted according to Eq. (2).
[C]soil⋅atom%soil=[C]BG⋅atom%BG+[C]glc⋅atom%glc[C]soil=[C]BG+[C]glc,
where [C]soil/BG/glc is the C amount of enriched soil sample/background
soil sample/glucose-derived C in soil (mol × gsoil-1)
and atom %soil/BG/glc is 13C in enriched soil
sample/background soil sample/applied glucose (atom %).
Cytosol C pool
The cytosol pool was determined by the fumigation–extraction technique from
fresh soil shortly after sampling, according to Wu et al. (1990) with slight
changes. Briefly, 15 g of fresh soil was placed into glass vials, which were
exposed to chloroform over 5 days. After removing the rest of chloroform
from the soil, the cytosol C was extracted with 45 mL of 0.05 M
K2SO4. As the fumigation–extraction technique produces not only
soluble components but also cell organelles and cell particles, we referred
to the pool of C in fumigated extracts as “cytosol” only for simplification of
terminology. Organic C was measured with a high-temperature combustion
TOC analyser (multi N/C 2100 analyser, Analytik Jena, Germany). The cytosol
pool was calculated as the difference between organic C in fumigated and
unfumigated samples without correcting for extraction efficiency. After
organic C concentrations were measured, the K2SO4 extracts were
freeze-dried and the δ13C values of a 30–35 µg subsample
were determined using EA-IRMS (instrumentation identical to soil
δ13C determination). The recovery of glucose-derived 13C in
fumigated and unfumigated samples was calculated according to the
above-mentioned mixing model (Eqs. 1 and 2). The 13C in the microbial
cytosol was calculated from the difference in these recoveries.
Phospholipid fatty acid analysis
The PLFA analysis was performed using the liquid–liquid extraction method of
Frostegard et al. (1991) with some modifications (Gunina et al., 2014).
Briefly, 6 g of soil was extracted with a 25 mL one-phase mixture of
chloroform, methanol and 0.15 M aqueous citric acid (1:2:0.8 v/v/v) with
two extraction steps. The 19:0-phospholipid
(dinonadecanoylglycerol-phosphatidylcholine, Larodan Lipids, Malmö,
Sweden) was used as internal standard one (IS1) and was added directly to
soil before extraction (25 µL with
1 µg µL-1). Additional chloroform and citric acid
were added to the extract to achieve a separation of two liquid phases, in
which the lipid fraction was separated from other organics. Phospholipids
were separated from neutral lipids and glycolipids by solid-phase extraction using
a silica column. Alkaline saponification of the purified phospholipids was
performed with 0.5 mL of 0.5 M NaOH dissolved in dried MeOH, followed by
methylation with 0.75 mL of BF3 dissolved in methanol. The resulting fatty
acid methyl esters (FAMEs) were purified by liquid–liquid extraction with
hexane (three times). Before the final quality and quantity measurements,
internal standard two (IS2) (13:0 FAME) (15 µL with
1 µg µL-1) was added to the samples (Knapp, 1979).
All PLFA samples were analysed by gas chromatograph (GC) (Hewlett Packard
5890 GC coupled to a mass-selective detector 5971A) (Gunina et al., 2014). A
25 m HP-1 methylpolysiloxane column (internal diameter 0.25 mm, film
thickness 0.25 µm) was used (Gunina et al., 2014). Peaks were
integrated and the ratio to IS2 was calculated for each peak per
chromatogram. Substances were quantified using a calibration curve, which was
constructed using 29 single standard substances (13:0, 14:0, i14:0,
a14:0, 14:1ω5, 15:0, i15:0, a15:0, 16:0, a16:0, i16:0,
16:1ω5, 16:1ω7, 10Me16:0, 17:0, a17:0, i17:0,
cy17:0, 18:0, 10Me18:0, 18:1ω7, 18:1ω9, 18:2ω6,9, 18:3ω6,9,12, cy19:0, 19:0, 20:0, 20:1ω9,
20:4ω6) at six concentrations. The recovery of extracted PLFA was
calculated using IS1, and the PLFA contents of samples were individually
corrected for recovery. Based on the measured PLFA contents, the PLFA C was
calculated for the each single compound.
The 13C / 12C isotope ratios of the single fatty acids were
determined by an IRMS Delta PlusTM coupled to a gas chromatograph (GC; Trace
GC 2000) via a GC-II/III-combustion interface (all units from Thermo Fisher,
Bremen, Germany) (Gunina et al., 2014). A 15 m HP-1 methylpolysiloxane
column coupled with a 30 m HP-5 (5 % phenyl)-methylpolysiloxane column
(both with an internal diameter of 0.25 mm and a film thickness of
0.25 µm) was used. The measured δ13C values of the fatty
acids were corrected for the effect of derivative C by analogy to Glaser and
Amelung (2002) and were referenced to Pee Dee Belemnite by external
standards. The enrichment of 13C in single fatty acids was calculated by
analogy to bulk soil and cytosol according to Eqs. (1) and (2), following a
two-pool dilution model (Gearing et al., 1991).
Amino-sugar analysis
Acid hydrolysis was performed to obtain amino sugars from soil and further
ion removal was performed according to the method of Zhang and Amelung (1996)
with optimization for δ13C determination (Glaser and Gross, 2005).
Methylglucamine (100 µL, 5 mg mL-1) was used as IS1 and was
added to the samples after hydrolysis. Following iron and salt removal,
non-cationic compounds such as monosaccharides and carboxylic acids were
removed from the extracts using a cation exchange column (AG 50W-X8 resin,
H+ form, mesh size 100–200, Biorad, Munich, Germany) (Indorf et al.,
2012). For final measurement, IS2 – fructose (50 µL,
1 mg mL-1) – was added to each sample. The amino-sugar contents and
13C enrichments were determined by LC-O-IRMS (ICS-5000 SP ion
chromatography system coupled by an LC IsoLink to a Delta V Advantage IRMS (Thermo-Fischer, Bremen, Germany)) (Dippold et al.,
2014). Amino sugars were quantified using a calibration curve, which was
constructed using four single standard substances (glucosamine,
galactosamine, mannosamine and muramic acid) as external standards at four
different concentrations (Dippold et al., 2014).
Calculations and statistical analysis
Factor analysis with the principal component extraction method of mass %
of individual PLFAs was done. The final assignment of fatty acids to distinct
microbial groups was made by combining the results of the factor loadings table
with databases of the presence of particular fatty acids in microbial
groups (Zelles, 1997). Fatty acids which were loaded into the same factor
with the same sign (+ or -) and belonged to one group (base of the table
provided in Zelles, 1997) were related to one specific microbial group and
their PLFA contents were summed. This method enables quality separation of
microbial groups within the soils (Apostel et al., 2013; Gunina et al.,
2014). The results of the factor analysis are presented in Supplement
Table S1.
Recovery of glucose-derived 13C (13Crec) (means 13C
recovery represented as % of total applied 13C) and enrichment
(13Cenrichm) (means 13C recovery represented as % of total
C pool) of the cytosol, PLFAs and amino sugars was calculated according to
Eqs. (3) and (4), respectively. The C turnover times in the cell pools were
calculated as 1/k; the k values were obtained from Eq. (5).
13Crec=CGlc13CApplied×100%,13Cenrichm=CGlcTotalCPool×100%,
where CGlc
is the amount of glucose-derived C incorporated into a distinct cell
compartment calculated by Eqs. (1) and (2) (µmol 13C per column), 13CApplied is the amount of applied glucose
13C (µmol 13C per column), and
TotalCPool is the amount of pool C (µmol C
per column).
Cenrichm(t)=Cenrichm(0)⋅exp-kt,
where
Cenrichm(t) is the 13C enrichment of the compartment, obtained
from Eq. (4) at time t (%), Cenrichm(0) is the 13C
enrichment of the compartment at time 0 (%), k is the decomposition rate
constant (% d-1), and t stands for time (days).
One-way ANOVA was used to estimate the significance of differences in total
13C recovery and enrichment of non-specified SOM pool, cytosol, PLFAs
and amino sugars. The data always represent the mean of four
replications ± SE. The 13C in the non-specified SOM was calculated by subtracting 13C incorporated into cytosol, PLFAs and amino sugars from total 13C measured in the soil. To describe decomposition
rate of 13C, a single first-order kinetic equation was applied to the
enrichment of 13C in the pool of cytosol, PLFAs and amino sugars
(Eq. 5) (Kuzyakov, 2011; Parton et al., 1987).
Results
Glucose utilization and its partitioning within microbial
biomass pools
The amino-sugar C pool was the largest, due to accumulation of these substances
in SOM, whereas pools that mainly characterize living MB showed smaller C
contents (Table 1). The cytosol pool (C content 210 ± 7.10 for day 3;
195 ± 14.8 for day 10; 198 ± 19.9 mg C kg-1 soil for
day 50) as well as nearly all PLFA groups (Table S2) remained
constant during the experiment.
Amount of microbial biomass compartments, their C content, PLFA
content of microbial groups and composition of microbial residues in
investigated soil. G-1 and G-2 are Gram-negative group one and two,
respectively; G+1 and G+2 are Gram-positive group one and two,
respectively; Ac – actinomycetes; 16:1ω5 – saprotrophic fungi. Data present
mean of three time points (with four replications for each time point)
±SE.
Compartment
mg component C kg-1 soil
mg kg-1 soil
Ratio
Cytosol
201.0 ± 7.1
–
Phospholipid fatty acids
39.4 ± 4.7
51.9 ± 6.2
Specific phospholipid fatty acids
G-1
8.9 ± 3.6
11.6 ± 4.6
G-2
5.6 ± 0.8
7.4 ± 1.1
G+1
5.9 ± 1.2
7.9 ± 1.6
G+2
0.7 ± 0.3
1.0 ± 0.4
Ac
2.3 ± 0.7
3.0 ± 1.0
16:1ω5
1.7 ± 0.3
2.2 ± 0.3
Fungi
1.0 ± 0.2
1.3 ± 0.2
Bacteria/fungi
6–8.5
Amino sugars
560.7 ± 68.2
1393.8 ± 170.0
Glucosamine
460.7 ± 79.3
1146.5 ± 197.3
Galactosamine
90.9 ± 11.3
226.3 ± 28.2
Muramic acid
9.1 ± 1.8
21.1 ± 4.1
Glucosamine / muramic acid
17–55
Glucosamine / muramic acid (literature data for pure cultures*)
Bacteria
5.3
Fungi
271
Galactosamine / muramic acid
12–19
Galactosamine / muramic acid (literature data for pure cultures*)
Bacteria
2.8
Fungi
59
* Data are taken from Glaser et al. (2004).
The highest recovery of 13C was found for cytosol pool (15–25 % of
applied 13C), whereas the lowest was reported for amino sugars
(0.8–1.6 % of applied 13C) (Fig. 1). The recovery of glucose-derived 13C in the cytosol pool decreased over time, with the largest
decline from day 3 to day 10, and then remained constant for the following
month (Fig. 1). The 13C recovery into PLFA was generally very low and
was in the same range as recovery into amino sugars (Fig. 1). The 13C
recovery in PLFA showed no clear trend between the sampling points (high
standard error) (Fig. 1). In contrast, 13C recovery in amino sugars
increased 2-fold on the 50th day of the experiment (p<0.05).
Partitioning of glucose-derived 13C in SOM presented as the
13C recovery (% of initially applied 13C) between the following
pools: non-specified SOM (calculated as total 13C recovery subtract
13C recovery in cytosol, PLFAs and amino sugars), cytosol, PLFAs and
amino sugars. Brown line indicates the total remaining glucose-derived 13C
glucose in the soil and is a sum of 13C in non-specified SOM, cytosol,
PLFAs and amino sugars. Small letters reflect differences between the
sampling points for the distinct pool. Data present mean (n=4) and bars
present standard errors (SEs). The SEs for the amino sugars are not fully
shown.
Turnover time of C in microbial biomass pools
To evaluate C turnover in the cytosol, PLFAs and amino sugars, we calculated
the enrichment (% of incorporated 13C relatively to pool C) of each
pool by glucose-derived 13C. The pool enrichment was the highest for
PLFAs and the lowest for amino sugars (Fig. 2).
Based on the decrease in 13C enrichments over time (Fig. 2), the C
turnover times in the cytosol and PLFAs were calculated as 151 and 47 days,
respectively. The C turnover time in the amino-sugar pool could not be
calculated by this approach because the maximum enrichment had not yet been
reached and, consequently, a decomposition function could not be fitted.
Phospholipid fatty acids
Fatty acids of bacterial origin dominated over those of fungal origin within
the living microbial community characterized by PLFA composition (Table 1).
The PLFA content of most groups did not change significantly during the
experiment, reflecting steady-state conditions for the microbial community
(see Table S2).
Higher 13C recovery was found in bacterial than in fungal PLFAs (Fig. 3, top). Remarkably, the 13C enrichment decreased over time for all
bacterial PLFAs, whereas it increased or remained constant for 16:1ω5, fungi
and actinomycetes (Fig. 3, bottom), indicating differences in C turnover in
single-celled organisms compared to filamentous organisms.
Amino sugars
The content of amino sugars followed the order muramic acid <galactosamine < glucosamine (Table 1). The galactosamine / muramic
acid ratio ranged between 12 and 19 (Table 1), showing that bacterial
residues were dominant in the composition of microbial residues in SOM.
The recovery of glucose-derived 13C into amino sugars increased in the
order muramic acid = galactosamine < glucosamine (Fig. 4, top),
partly reflecting their pool sizes. The 13C recovery showed no increase
from day 3 to day 50 for any amino sugars. The ratio of galactosamine / muramic
acid, calculated for the incorporated 13C, was about six. This is much
lower than the ratio observed for the pools of amino sugars. The 13C
enrichment did not increase from day 3 to day 50 for any of the amino sugars.
The highest enrichment was observed for muramic acid and the lowest for
galactosamine (Fig. 4, bottom). The 13C enrichment in amino sugars was
10–20 times lower than for PLFA.
Discussion
Glucose decomposition
The amount of glucose-derived 13C remaining in soil after 50 days was in
the range 80 % which was higher than reported by other studies. Glanville
et al. (2012) observed that 50 % of glucose C remained in SOM after 20
days; Wu et al. (1993) reported that 55 % of glucose-derived 14C
remained after 50 days; Perelo and Munch (2005) reported the mineralization
of 50 % of 13C glucose within 98 days. The amounts of applied
C (Bremer and Kuikman, 1994; Schneckenberger et al., 2008), as well as
differences in microbial activity (Bremer and Kuikman, 1994; Schimel and
Weintraub, 2003) in the investigated soils, explain the variations between
studies in the portion of remaining glucose C.
13C enrichment in the cytosol, PLFA and amino-sugar cell pools
as well as functions to calculate the C turnover times in these microbial
cell pools. The left y axis represents the PLFA pool, the first right
y axis the cytosol, and the second y axis the amino-sugar pool. Data
present mean (n=4) and bars present standard errors.
The highest mineralization of glucose-derived 13C (20 %) was found
within the first 3 days after tracer application (Fig. 1), whereas at day
50 mineralization was much slower. Glucose is decomposed in soil in two
stages (Gunina and Kuzyakov, 2015): during the first one, part of glucose C
is immediately mineralized to CO2 and part is incorporated into the
microbial compartments; and second one, when C incorporated into MB is
further transformed and is used for microbial biosynthesis, and
mineralization of glucose-C to CO2 occurs much slower (Bremer and
Kuikman, 1994). This first stage takes place in the first day after substrate
addition and is 30 times faster than the second stage (Gregorich et al.,
1991; Fischer et al., 2010). Due to the first sampling point in our
experiment was 3 days after glucose addition, the obtained data on glucose
mineralization can be mainly related to the second stage.
Recovery of glucose-derived 13C (top) and 13C enrichment
(bottom) of the microbial PLFAs. Note that the values for 16:1ω5 and fungi
are scaled up 10 times (secondary y axis) compared to those of other groups
(y axis on the left). Data present the mean (n=4) and bars present standard
errors. Small letters reflect differences between the microbial groups for
13C recovery and 13C enrichment from glucose; letters (a–d) are for
day 3, (l–o) are for day 10, and (x–z) are for day 50.
Recovery of glucose-derived 13C (top) and 13C enrichment
(bottom) of amino sugars and muramic acid. Letters reflect significant
differences in the recovery and 13C enrichment from glucose 13C
into amino sugars on a particular day; letters (a–b) are for day 3, (l–m)
are for day 10, and (x–y) are for day 50. No significant differences were observed
between the three sampling days. Data present mean (n=4) and bars present
standard errors.
A significant portion of glucose-derived C was stored in the non-specific
pool in SOM (Fig. 1), e.g., as microbial storage compounds and other
cellular building blocks, which can contribute to C accumulation in
microbial residues (Wagner GH, 1968; Zelles et al., 1997; Lutzow et al.,
2006). This part cannot be extracted by the methods applied in this study.
The amino-sugar method detects only the peptidoglycan and chitin proportions
of the cell walls, whereas other constituents cannot be determined (Glaser
et al., 2004). Chloroform fumigation only partially extracts the cytosol
cell compounds, and high-molecular-weight components, which interact with
the soil matrix, cannot be extracted with low-molarity salt solution.
Partitioning of 13C-derived glucose between cell
compounds
To estimate the residual amount of C derived from applied 13C-labelled
low-molecular-weight organic substances (LMWOS), the 13C in SOM or in
the total MB pool is frequently determined. This approach, however, does not
allow the portions of 13C incorporated into stable and non-stable C
pools to be estimated, because the 13C in SOM includes the sum of
13C in living biomass and 13C in microbial residues. Furthermore,
the living MB contains cell compartments with a broad spectrum of C turnover
times. The approach applied in the present study allows the partitioning of
glucose-derived C in living MB to be estimated, as well as the contribution
of LMWOS-C to SOM composition.
Cytosol
We calculated the 13C enrichment of the cytosol C pool, extracted after
chloroform fumigation. The estimated turnover time of C in this pool was
about 151 days. This value lies close to the previously reported range of
87–113 days, for the same pool for soils incubated for 98 days with 13C
glucose (Perelo and Munch, 2005), but was lower than MB C turnover time
calculated using a conversion factor (2.22) for soils incubated
for 60 days with 14C glucose (Kouno et al., 2002). The long C turnover
time in cytosol is related to the high heterogeneity of this pool, which
includes compounds with various molecular masses (Malik et al., 2013) and
functions, with different turnover times. Thus, C turnover time in cytosol
presents the mean value of turnover times of these compounds.
Phospholipid fatty acids
Phospholipid fatty acid content and turnover
Phospholipid fatty acid C comprised 0.27 % of the soil organic carbon
(SOC). The 13C recovery into PLFAs, in the case of constant PLFAs content
during the experiment, reflects microbial activity under steady-state
conditions (growth and death of microorganisms occur with the same rates)
and processes of the exchange and replacement of existing PLFA C within
living cells.
Few studies have estimated the C turnover time in PLFAs or the turnover time
of PLFAs themselves in soil, as very few options exist to estimate these
parameters under steady-state conditions. The turnover time of
13C-labelled PLFAs contained in dead microbial cells was 2.7 days
(Kindler et al., 2009). The PLFAs turnover times estimated in the field
conditions using a C3–C4 vegetation change (Amelung et al.,
2008; Glaser, 2005) or 14C dating (Rethemeyer et al., 2005) were between
1 and 80 years. However, these approaches estimate the turnover time of C
bound in PLFA, which can be much older than the PLFA molecules due to
repeated C recycling before incorporation. In contrast, 13C pulse
labelling is an approach that enables direct estimation of the turnover of
freshly added C by the initial recovery peak. The approach used in the
present study showed that the C turnover time in PLFA is about 47 days
(Fig. 2). Accordingly, if the decomposition after cell death is about 3
days, the PLFA turnover time in living cells is about 44 days. This short
turnover time of PLFAs is significantly lower than the C turnover time in the
cytosol (Figs. 2, 5). This is because the membrane is an interacting surface
between the cell and the environment, and thus frequent and rapid adaptations
of its structure are crucial for active microorganisms (Bossio et al., 1998;
Kieft et al.,1997). In contrast, the extracted cytosol pool includes C from
both active and dormant microorganisms (Blagodatskaya and Kuzyakov, 2013),
and the latter can dilute the 13C signal incorporated into the active
pool with non-labelled C, yielding a lower turnover of this pool.
Dynamic relationships between microbial glucose utilization and C
turnover times in cytosol, cell membrane and cell wall components.
Contribution of microbial groups to glucose-derived C
utilization
More glucose-derived 13C was incorporated into bacterial PLFAs (Fig. 3,
top), than into filamentous microorganisms. This can be a consequence of low
C loading rates (less than 4 mg C g-1 soil; see Reischke et al.,
2014), under which conditions the added C is utilized primarily by bacterial
communities, whereas at higher concentrations of applied substrate the
dominance of fungi in substrate utilization is observed (Reischke et al.,
2014).
The 13C recovery into Gram-negative fatty acids was higher (taking both
G- groups together) compared to G+ bacterial PLFAs (Fig. 3, top), which
might be due to (i) the abundance of their fatty acids, which was higher
(Table 1), or (ii) glucose uptake activity, which was higher for G- than
G+ groups. In contrast, the 13C enrichment (13C recovery related
to total C in particular biomarkers) for G- bacterial PLFAs was not higher
than that for G+ (Fig. 3, bottom). Thus, the high 13C recovery into
G- bacterial biomarkers can mainly correspond to their high content in the
soil, not to higher activity of microbial groups. However, enrichment of
PLFA C by glucose-derived 13C is only a proxy of microbial activity and
can only partly estimate the real activity of microbial groups. This clearly
suggests that the analysis of isotope data after labelling in general requires
the calculation and combined interpretation of both the total tracer C
recovery and the 13C enrichment in the investigated pool.
In contrast to our results, a higher recovery of glucose-derived 13C
into G+ than G- PLFAs was observed in other studies (Dungait et al.,
2011; Ziegler et al., 2005). However, in these studies, much higher amounts
of C were applied to the soil (15 µg C g-1 soil), which
stimulated the growth of G+ bacteria. In contrast, under steady-state
conditions with low glucose concentrations in soil, G- bacteria were the
most competitive group for glucose uptake (Fig. 3).
The 13C enrichment of bacterial PLFAs decreased from day 3 to day 50,
whereas 13C in fungal PLFAs increased (in the case of 16:1ω5) or
stayed constant (Fig. 3, bottom). The decrease in 13C enrichment in
bacterial fatty acids indicates a partial turnover of bacterial lipid
membranes, which is much faster than turnover in fungal membranes. This
result is consistent with the turnover time of bacterial biomass in
soil (Baath, 1998), which is about 10 days, whereas fungal biomass turnover
times range between 130 and 150 days (Rousk and Baath, 2007). Consequently, the
increase in 13C enrichment in fungal PLFAs at late sampling points
indicates that fungi consume the exudation products of bacteria or even dead
bacterial biomass (Zhang et al., 2013; Ziegler et al., 2005).
Amino sugars
Amino-sugar content and amino-sugar C turnover in total and
living microbial cell walls
Amino sugars represented the largest microbial pool investigated in this
study (Table 1) and comprised 3.7 % of SOC. Chitin and peptidoglycan, the
direct sources of the amino sugars, comprise no more than 5 % of cell
biomass (Park and Uehara, 2008; Wallander et al., 2013). Therefore, the high
amount of amino sugars, relative to PLFAs, can only be explained by their
high proportion in microbial residues/necromass (Glaser et al., 2004; Liang
et al., 2008). Irrespective of the large pool size of the amino sugars, their
recovery and pool enrichment with glucose-derived 13C was the lowest
compared to other compartments in living cells and increased during the
experiment. Consequently, amino sugars can have the slowest turnover in
soils, presumably even within living cells, for three reasons: (1) cell walls
are polymers that require a rather complex biosynthesis of the amino-sugar
fibres, (2) cell wall polymerization occurs extracellularly (Lengeler et al.,
1999) and (3) microorganisms do not need to synthesize peptidoglycan unless
they multiply. To calculate C turnover time in this pool, it is necessary to conduct
long-term experiments.
The majority of amino sugars, extracted after acid hydrolysis, represent
microbial necromass, which does not incorporate any glucose-derived 13C
but strongly dilutes the 13C incorporated into the walls of living
cells. To estimate the 13C enrichment into amino sugars of living cells,
we first calculated the amount of amino sugars in the living MB pool, which
constituted 0.87 µmol g-1 soil, and was about 11 % of the
total amino-sugar pool (please see Supplement calculations for further
details). This estimate agrees with that of Amelung et al. (2001a) and Glaser
et al. (2004), who reported that the amount of amino sugars in living biomass
is one to two orders of magnitude lower than in the total amino-sugar pool.
We calculated the 13C enrichment in amino sugars for the first sampling
point, assuming that all replaced C is still contained within living MB after
3 days of glucose C utilization, and it constituted 0.57 % of the C
pool. Comparison of these data with the 13C enrichment into PLFAs and
the cytosol allowed us to conclude that the enrichment of amino-sugar C with
glucose-derived 13C in living biomass is 2-fold lower than the
enrichment in PLFAs, and higher than in the cytosol pool. This reflects the fact that
microbial C turnover is a phenomenon that is not restricted to the death or
growth of new cells but that, even within living cells, highly polymeric cell
compounds, including cell walls, are constantly replaced and renewed (Park
and Uehara, 2008).
Contribution of bacterial and fungal cell walls to SOC
Glucosamine was the dominant amino sugar in the soil, whereas muramic acid
was the least abundant (Table 1), which agrees with the most literature data
(Engelking et al., 2007; Glaser et al., 2004). To conclude about the
proportions of bacterial and fungal residues in the SOM, the ratio of
galactosamine / muramic acid (Glaser et al., 2004) was calculated (Table 1), showing bacteria to be the dominant within the soil microbial community.
The bacterial origin of microbial residues in the soil is supported by (1) the dominance of bacterial PLFA biomarkers and (2) the environmental
conditions of the site, namely long-term agricultural use, which promotes
the development of bacterial communities.
Three-fold more glucose-derived 13C was recovered in glucosamine than
in galactosamine and muramic acid (Fig. 4, top). This correlates with the
pool size and indicates that glucosamine is the most dominant amino sugar
not only in total amino sugars but also within the walls of living cells.
The galactosamine / muramic acid ratio of the incorporated 13C was 6
and, consequently, was significantly lower than the ratio calculated for the
amount of amino sugars (Table 1). This indicates that bacteria are more
active in glucose-derived 13C utilization than fungi, a conclusion also
supported by the 13C-PLFA data (Fig. 3, top). Thus, even if the
composition of amino sugars does not allow a clear conclusion concerning
living microbial communities in soil, amino-sugar analysis combined with
13C labelling reveals the activity of living microbial groups in terms
of substrate utilization.
The calculated 13C enrichment was the highest in muramic acid (Fig. 4,
bottom). This is in agreement with the high 13C enrichment of bacterial
PLFAs compared to 16:1ω5 and fungi (Fig. 3). Due to differences in
cell wall architecture, G+ bacteria contain more muramic acid
(approximately 4 times) than G- bacteria (Lengeler et al., 1999) and
thus make a higher contribution to the 13C enrichment of muramic acid.
The 13C enrichment of glucosamine was 2-fold lower than muramic acid
(Fig. 4, bottom). This confirms the hypothesis that glucosamine originates
from bacterial as well as fungal cell walls and, consequently, has a mixed
enrichment between the fungal galactosamine and bacterial muramic acid.