ARTICLE IN PRESS

Soil Biology & Biochemistry xxx (2009) 1e8

Contents lists available at ScienceDirect

Soil Biology & Biochemistry

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

Amounts of carbon mineralised and leached as DOC during decomposition

of Norway spruce needles and ne roots
Karna Hansson a, *, Dan Berggren Kleja b, Karsten Kalbitz c, Hanna Larsson b
a
Department of Ecology, Box 7044, Swedish University of Agricultural Sciences, S-75007 Uppsala, Sweden
b
Department of Soil and Environment, Box 7001, Swedish University of Agricultural Sciences, S-75007 Uppsala, Sweden
c
Earth Surface Science, Institute for Biodiversity and Ecosystem Dynamics, University of Amsterdam, Nieuwe Achtergracht 166, NL-1018 WV Amsterdam, The Netherlands

a r t i c l e i n f o a b s t r a c t

Article history: Changes in climate or forest management practices leading to increased litter production will most likely
Received 24 June 2009 cause increased leaching rates of dissolved organic carbon (DOC) from the O horizon. The rhizosphere is
Received in revised form often assumed to have a large carbon ux associated with root turnover and exudation. However, little
1 October 2009
has been done to quantify the amount of DOC originating from root litter. We studied decomposition of
Accepted 14 October 2009
Available online xxx
ne root and needle litter of Norway spruce (Picea abies) through a combined incubation and leaching
experiment in the laboratory using ve different litter types: fresh needle litter, aged needles from the
litter layer, fresh and dead roots from mineral soil samples, and seven-year-old roots from a previous
Keywords:
Dissolved organic carbon litterbag study. After respiration measurements, the samples were percolated with articial throughfall
Mineralisation water and DOC and UV absorbance were measured in the leachate. Mineralisation of dissolved organic
Fine roots matter in the leachate and sorption of DOC to ferrihydrite were determined as a measure of DOC ability
Needles to be stabilised by iron (hydr)oxide surfaces.
Litter decomposition The mineralisation rate and DOC production rate of root samples were always lower than that of needle
Norway spruce samples. However, root and needle derived dissolved organic matter (DOM) were similar in terms of
Carbon dioxide aromaticity, as indicated by their specic UV absorbance, and ability to be sorbed by ferrihydrite. For seven-
year-old roots, a signicantly higher fraction of carbon was lost as DOC (30%) than for younger roots (20%).
Furthermore, DOM from old roots bound more strongly to ferrihydrite and is mineralised at a lower rate
than DOC from younger roots, suggesting that roots at late stages of decomposition, although a small
fraction of total litter, signicantly contribute to carbon build-up in mineral soils. The slower decomposition
rate of roots compared with needles must be taken into account when modelling litter decomposition.
2009 Elsevier Ltd. All rights reserved.

1. Introduction (Callesen et al., 2003). The major carbon inputs to mineral soil
layers are from ne root litter and dissolved organic carbon (DOC)
Temperate forest ecosystems account for about 25% of the leached from the forest oor. Thus, the carbon pool and its
carbon stock in global terrestrial ecosystems, of which about half is dynamics in the mineral soil are determined by the input rates of
stored in soil organic matter (King et al., 1997). Thus, a small change root litter and retained DOC, as well as their decomposition rates.
in the carbon balance of soils in these ecosystems might affect In a recent study, uxes of carbon into the mineral soil in the
atmospheric CO2 concentration. Compared to the processes form of DOC and ne root litter were measured in three Norway
controlling soil organic matter (SOM) turnover in the forest oor, spruce ecosystems situated along a climate gradient in Sweden
our understanding of the chemical and microbial processes (Kleja et al., 2008). The annual inputs of carbon as ne root
controlling turnover of SOM in mineral soil layers is poor. This is (<1 mm) litter to the mineral soil (0e50 cm) ranged between 73
a problem when assessing the potential of forest soils to act as and 78 g m
2 yr
1, whereas the corresponding range for DOC was
carbon sinks or sources since a major fraction of carbon is normally 9e26 g m
2 yr
1. Thus, root litter clearly dominates the carbon
found in the mineral soil. According to an inventory of soil organic input. However, the net contribution of root litter to the steady-
carbon (SOC) pools in boreal forest soils in Scandinavia, 70e80% of state carbon pool is less clear, because detailed information on the
the organic carbon in the upper 100 cm is found in the mineral soil decomposition rates of root litter and DOC in the mineral soil e and
their determining factors e is not yet available.
Michalzik et al. (2003) used the Dynamic DOC model (DyDOC) to
* Corresponding author. Tel.: 46 18 672412; fax: 46 18 672890. estimate the contribution of DOC input and root litter to the steady-
E-mail address: karna.hansson@ekol.slu.se (K. Hansson). state carbon pool. According to their simulations, 73e89% of the

0038-0717/$ e see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.soilbio.2009.10.013

Please cite this article in press as: Hansson, K., et al., Amounts of carbon mineralised and leached as DOC during decomposition of Norway
spruce..., Soil Biology & Biochemistry (2009), doi:10.1016/j.soilbio.2009.10.013
 ARTICLE IN PRESS

2 K. Hansson et al. / Soil Biology & Biochemistry xxx (2009) 1e8

mineral soil carbon originated from DOC. However, in producing located 190e200 m above sea level in the boreonemoral vegetation
this estimate they assumed that root litter behaved as needle litter zone. Mean annual air temperature is 5.5  C and mean annual
in terms of carbon mineralisation and DOC production rates. precipitation 688 mm. The duration of the growing season (temper-
Furthermore, they assumed that the quality of dissolved organic ature >5  C) is 190 days. Field samples were collected in LUSTRA plots
matter (DOM) produced from the two substrates was identical. with a mesic moisture regime. Stand age was 44e47 years in 2007.
These assumptions are critical and might not be valid. For example, Site productivity ranges from 10.1 to 11.3 m3 ha
1 yr
1 and the eld
a litterbag study by Majdi (2004) found that the mass loss of ne and ground vegetation is grass or no vegetation.
root litter of Norway spruce was half that of needle litter after one According to FAO. (1990) the soil is classied as a Haplic Podzol,
year of decomposition. Palviainen et al. (2004) reported similar developed on a glacial till. The texture is a stony sandy loam with
values, with 34% mass loss of ne root litter compared with 59% a medium boulder frequency. Site productivity ranges from 10.1 to
mass loss of needle litter after three years of decomposition in 11.3 m3 ha
1 yr
1 and the eld and ground vegetation is grass or no
a Norway spruce stand. However, litterbag studies provide no vegetation (Berggren et al., 2004).
information on the relative losses as CO2 and DOC. Needle litter is
known to produce substantial amounts of DOC during the 2.2. Root and needle litter samples
decomposition process (Fro berg et al., 2005). The extent to which
this occurs for ne root litter is less well known. In a recent study, Five different litter types were sampled: fresh needle litter, aged
Uselman et al. (2007) incubated 14C-labelled ne root material and needles from litter layer, fresh roots from mineral soil, dead roots
leaf litter in 50-cm soil microcosm columns and measured the from mineral soil and seven-year-old roots from a previous litterbag
production of CO2 and DOC. Their experiment showed that roots study. Each litter type was a mix of several subsamples. Needle litter
decomposed more slowly than leaf litter and that DOC made was collected in December 2006 and stored in the freezer. Fresh
a signicant contribution (w60%) to total carbon losses during the needle litter samples were obtained by shaking trees and collecting
47-day experimental period. The experimental setup did not allow the falling needles. Green needles were excluded. Aged, slightly
for any qualitative characterisation of the DOM produced by the decomposed needles were taken from the litter (Oi) layer. The
two substrates. The quality of DOM formed during decomposition turnover time of this layer is about 5 years (Fro berg et al., 2005).
of root litter is probably crucial for its contribution to the build-up Mineral soil samples (0e10 cm soil depth) were collected in October
of soil carbon stocks in mineral soil layers because the sorption of and November 2007. Roots were carefully removed from the soil,
DOM to mineral surfaces such as ferrihydrate is inuenced by its placed in deionised water and gently stirred to remove soil particles.
chemical composition. Constituents with higher molecular weight They were carefully cleaned and sorted using forceps under 10
have been shown to adsorb preferentially, and fractions rich in magnication into living and dead roots, based on visual criteria
aromatic structures such as lignin-derived hydrophobic described by Vogt and Persson (1991). Grass roots were excluded.
compounds, fulvic and humic acids show stronger sorption than All roots used in the incubation experiment had a diameter of
compounds rich in carbohydrates (Chorover and Amistadi, 2001; <2 mm. Strongly decomposed roots were obtained from a previous
Kaiser, 2003). As shown by Mikutta et al. (2007), the binding mode litterbag study. Fresh roots with a diameter < 2 mm were cut into
of DOM to mineral surfaces is decisive for its bioavailability. 1e4 cm-long pieces and placed in litterbags in 1999. These litter-
The decomposition rate and DOC production of a substrate bags were buried in the mineral soil at 10 cm depth and recovered in
changes with time, due to changes in substrate quality during December 2006 and stored in the freezer. All roots used in the
decomposition (Moore and Dalva, 2001; Don and Kalbitz, 2005). In experiment were cut into pieces of approximately needle length,
a study of the ne root dynamics of black spruce (Picea mariana L.), 1e2 cm. Water content in the material was determined by weighing
decomposition was found to be rapid soon after the roots were litter samples, drying them at 105  C for 24 h and calculating the
identied as being dead, but decreased with time (Ruess et al., weight loss. Total carbon and nitrogen (N) content in the dried
2003). Berg (2000) suggests that the decomposition rate of plant samples were analysed by dry combustion (CN2000, LECO Corpo-
litter at late decomposition stages is very slow and approaches zero. ration). Samples used in incubation were not dried.
Different stages of decomposition should therefore be considered
when estimating DOC originating from litter. To our knowledge, 2.3. Column incubation and measurements
there is no previous study on DOC production from roots in
different stages of decomposition. In the present study we focused Litter samples were incubated in glass columns (35 cm long
on determining leached DOC and respired carbon for Norway with an inner diameter of 2.4 cm) using a method adapted from
spruce ne roots and needles at different stages of decomposition. Sjo berg et al. (2003). Each column had a bottom plug made of
Our specic objectives were (i) to investigate the extent to which silicone, containing a glass drain pipe connected to a silicone tube
the fraction of DOC lost during decomposition depended on the closed with a clip. A glass bre lter (1.0 mm pore size, Whatman
stage of decomposition of the substrate; (ii) to make a brief quali- GF/B) was placed in the bottom of the column to avoid leaching of
tative comparison of DOM leached from root and needle litter at particles. The columns were lled with litter (equivalent to 1 g dry
different stages of decomposition; (iii) to determine the ability of weight) mixed with 25 g quartz sand (washed with acid and heated
DOM originating from root and needle litter to be sorbed by ferri- to 600  C to remove carbon), and a second glass bre lter (0.7 mm
hydrite; and (iv) to determine the mineralisation of DOM derived pore size, Pall Corporation) was placed on top. During incubation,
from roots and needles. plastic lms were placed on the opening of the column to allow gas
exchange but prevent evaporation of water. Four replicates of each
2. Materials and methods litter type (in total 20 columns) were incubated. Prior to each
percolation, column outlets were connected by silicone tubes to
2.1. Site description vacuum chambers in which borosilicate glass bottles were placed
to collect the leachate. A suction of approximately
0.2 bar was set
All litter samples were taken from Asa Experimental Forest to create unsaturated ow conditions. The chemical composition of
(57 080 N, 14 450 E), in southern Sweden. The site is one of three the leaching solution resembled throughfall water at the site. The
Norway spruce (Picea abies (L.) Karst.) stands used within the LUSTRA solution consisted of deionised water with addition of ions to give
research programme (Berggren et al., 2004; Kleja et al., 2008). Asa is a concentration of Na: 0.066 mM, K: 0.054 mM, Ca2: 0.014 mM,

Please cite this article in press as: Hansson, K., et al., Amounts of carbon mineralised and leached as DOC during decomposition of Norway
spruce..., Soil Biology & Biochemistry (2009), doi:10.1016/j.soilbio.2009.10.013
 ARTICLE IN PRESS

K. Hansson et al. / Soil Biology & Biochemistry xxx (2009) 1e8 3

Mg2: 0.01 mM, NH
2

4 : 0.014 mM, NO3 : 0.014, SO4 : 0.027 mM and For inoculation, a uniform microbial community was chosen for
Cl
: 0.114 mM. all samples so that any variation measured would only be the result
Prior to the start of the experiment, litter samples were inocu- of variation in DOM properties. Before extraction of the inoculum,
lated with a litter extract from the site. Needle litter (14.55 g) was the air-dried Oh horizon of a Norway spruce soil was rewetted to
ground and mixed with 1 L deionised water. After 30 min of sedi- a water capacity of 60% and incubated for two weeks at 20  C to
mentation, 5 mL of the solution, with large particles removed, were reactivate the microorganisms. The soil was then shaken for 30 min
added to each column. Columns were incubated in a dark room at with a 5 mM CaCl2 solution (soil:solution ratio 1:2) and ltered
a constant temperature of 15  C. through a 5 mm lter (SMWP 4700, Millipore, Bedford, MA, USA).
Production of carbon dioxide (CO2) was measured after 1, 2, 3, 6, This inoculum was added to the sample in a sample/inoculum
9, 12, 15, 19 and 28 weeks of incubation. The columns were left volume ratio of 100:1.
uncovered for 30 min and to assist with circulation, air was blown Air in the headspace of the asks was sampled using a syringe
into the columns using a rubber air pump. Columns were then and analysed for CO2 using a gas chromatograph with ame ion-
closed with silicone plugs and samples extracted after 10 min (t0) isation detector/methanizer (SRI 8610C, SRI Instruments, Scham-
using a syringe. Total air volume in the columns was 0.09 L. The beck, Bad Honnef, Germany). CO2 was measured 7 times during
columns were incubated for 3e7 h, dependent on mineralisation the incubation period, at short intervals at the beginning of the
rate of substrate, after which a second set of samples was extracted experiment and at long intervals at the end. Before starting the
(t1). The samples were analysed on a gas chromatograph (Hewlett incubation, air was applied to each ask to maintain a pressure of
Packard 5890A with Thermal Conductivity Detector, helium in2 m about 30 kPa for proper sampling of the headspace. The CO2 in the
Porapak T-column and a carrier ow of 25 mL min
1). The CO2 asks was calculated using the general gas equation for the gas
production rate was calculated as the difference between the total phase concentration, and from solubility constants and the pH
amount of inorganic carbon in gas and pore water phases at t1 and measured at the end of incubation for the liquid phase
t0, divided by the length of the incubation period (t1
t0). The concentration.
amount of dissolved inorganic carbon (DIC) in the pore water was The CO2 evolution by mineralisation of the inoculum carbon
calculated using thermodynamic equilibrium calculations. Before was determined in control samples (ultrapure water, inoculum
each respiration measurement, the columns were weighed to and nutrients) and subtracted from the values obtained for the
calculate water content. other samples. A glucose solution was used as a second control to
On the day after CO2 measurements, the columns were perco- test the functioning of the microbial community. After 8 weeks,
lated with 50 mL throughfall water solution at a rate of 0.55 mL/min. more than 80% of the glucose-C had been mineralised.
A preliminary study with spruce needles showed that >90% of
accumulated DOC was leached with the rst 50 mL of solution 2.5. Sorption of DOM to iron hydroxide
(results not shown). Leachate was ltered using a 0.2 mm Acrodisc
PF-lter to remove microorganisms and then analysed for DOC Sorption of dissolved organic matter (DOM) in the mineral
concentration (Shimadzu TOC-5000A analyser), pH (Radiometer soil horizons is probably the main process by which DOM is
Copenhagen PHM93 reference pH meter) and UV absorbance at retained in forest soils (Kalbitz et al., 2005). In Sweden, 60% of all
285 nm (Jasco V-530 spectrophotometer). Specic ultraviolet forest soils are podsolic, with most soil organic matter concen-
absorbance at 285 nm (SUVA285) was calculated as absorbance trated in the iron-rich B-horizon (Olsson et al., 2009). The ability
divided by DOC concentration. SUVA has been shown to be strongly of DOM to be sorbed by ferrihydrite was investigated on leachate
positively correlated to the aromaticity of the compound tested obtained after 6, 9 and 19 weeks. Due to the limited sample
(Traina et al., 1990; Chin et al., 1994; Aitkenhead-Peterson and Kal- volume obtained, it was not possible to conduct DOC biodegra-
bitz, 2005). Kalbitz et al. (2003) used UV absorbance at l 280 nm as dation and sorption experiments using leachate from the same
a parameter to investigate the aromaticity of DOM, whereas percolation.
Weishaar et al. (2003) used UV absorbance at l 254 nm. In the near Ferrihydrite was synthesised using a method adapted from
UV spectrum (l 200e380 nm), conjugated systems such as those Swedlund and Webster (1999) and Schwertmann and Cornell
present in aromatic molecules have the highest absorptivities, (2000). NaOH (4 M) was added drop-wise under stirring to
whereas other electronic structures do not absorb in this spectrum a solution containing 36 mM Fe(NO3)3 and 12 mM NaNO3 until
(Weishaar et al., 2003). Samples from percolation after 1, 2, 6, 12 and pH 8 was reached. The resulting suspension was aged for 18 h at
19 weeks were analysed for ammonium (NH4) and nitrate (NO3) 20  C and then back-titrated to pH 4.69 with 0.1 M HNO3. To
using a colorimetric method (FIASTAR 5000, FOSS AS). release CO2 the suspension was stirred for 1 h and then 1.40 mL of
ferrihydrite suspension was added to 15 mL of sample, corre-
2.4. DOC biodegradation sponding to 0.33 g Fe(OH)3 per litre. Samples for each litter type
and time were pooled. Sorption was investigated in the pH range
To determine carbon mineralisation of leachate obtained after 4.0e6.5, adjusted with NaOH or HNO3. The samples were shaken
12 weeks of the column experiment, samples were frozen directly end-over-end for 5 h in darkness at 20  C, and then centrifuged
after percolation and kept frozen until further usage in the exper- for 20 min at 2000 rpm and 20  C (Beckman Coulter J6-MI
iment. All replicate samples (20) were incubated in 60-mL sealed Centrifuge). The supernatant was extracted and pH was
asks at 20  C in the dark for 8 weeks, with 3 replications. 20 mL of measured. The residual supernatant was ltered using a 0.2 mm
sample was added to each ask. Aqueous samples with more than Acrodisc PF lter and UV absorbance and DOC concentration were
10 mg C L
1 were diluted before incubation to avoid overgrowth of analysed. To conrm that there were no ferrihydrite particles in
microorganisms (Hongve et al., 2000) and to minimise concentra- the ltered supernatant, original solutions and ltered superna-
tion effects on DOM biodegradation (Zsolnay, 2003). Nutrients tant were analysed for iron (using ICP Optima 3000 DV) on the
(equal weights of NH4NO3 and K2HPO4) were added in order to rst occasion the sorption test was performed. The iron concen-
adjust the C:N ratio to about 10:1 and facilitate DOM biodegrada- tration in ltered solutions was found to be low (<0.5 mg/L),
tion (McDowell et al., 2006). All incubation solutions were adjusted suggesting that all DOC in the ltrate was fully dissolved. Original
to pH 5.5 by adding HCl or NaOH. The asks were gently shaken by solutions were analysed for iron, sodium, potassium, calcium,
hand every day. magnesium, and aluminium.

Please cite this article in press as: Hansson, K., et al., Amounts of carbon mineralised and leached as DOC during decomposition of Norway
spruce..., Soil Biology & Biochemistry (2009), doi:10.1016/j.soilbio.2009.10.013
 ARTICLE IN PRESS

4 K. Hansson et al. / Soil Biology & Biochemistry xxx (2009) 1e8

2.6. Statistical analysis

a 1.6 Fresh needle litter
One-way analysis of variance (ANOVA) with pair-wise compar- 1.4 Aged needles, litter layer

CO2 (mg C g-1 day-1)

isons was carried out on the accumulated values of DOC and 1.2
respired C and on C mineralisation values, all using SAS software. Fresh roots, mineral soil
Analysis on Fe, Na, K, Ca, Mg, and Al content was carried out using 1
Dead roots, mineral soil
the same analysis, treating values from different weeks as repli- 0.8
cates. The correlations between SUVA285 and mineralisation rate Seven-year-old roots
and between SUVA285 and sorption to ferrihydrite were deter- 0.6
mined using regression analysis in Minitab 15 software. 0.4
0.2
3. Results
0
3.1. Total carbon and nitrogen content in litter 0 50 100 150 200
Time (days)
Carbon content was similar in different litter types and ranged
between 44 and 48%. However, nitrogen content differed between b
litter types, ranging from 0.6% N in fresh needle litter to 1.2% N in 0.6
seven-year-old roots (Table 1). Consequently, the C:N ratio was

DOC (mg g-1 day -1)

highest (82) for fresh needle litter and lowest (40) for seven-year-
old roots. For both needles and roots, the C:N ratio decreased with
increasing degree of decomposition. 0.4

3.2. Respiration and DOC production

Respiration decreased with time for all litter types except the 0.2
seven-year-old roots from the litterbag study, for which respiration
was very low and almost constant over time (Fig. 1a). The minerali-
sation rate of needle samples was always higher than that of root
0.0
samples. Accordingly, the accumulated mass of respired carbon was 0 50 100 150
signicantly higher for both needle types than for all root types Time (days)
(Fig. 2). There were also signicant differences in accumulated mass of
respired carbon in fresh and older material, both between fresh and Fig. 1. Losses of carbon (mg C g
1 substrate day
1) in the form of a) respired C and b)
dissolved organic C (DOC) leached from columns (n 4, mean  SE).
old needle litter and between fresh and seven-year-old root litter.
Leaching rate of DOC tended to decrease with time for all litter
types and was most pronounced for fresh roots from mineral soil 3.3. Water chemistry
and aged needles from the litter layer, which had the highest DOC
leaching on the rst measurement occasion (Fig. 1b). After the rst For fresh and dead roots, pH in leachate decreased with time, from
two months, changes over time became small. As for mineralisation 6.8 to 5.9 and 6.5 to 5.8 respectively, whereas the pH for other litter
rates, DOC production rates for needles were higher than those for types remained fairly constant over time (Fig. 5). Leachate from
roots. The accumulated mass of leached DOC per gram (dry weight) seven-year-old roots had the lowest pH, varying between 5.0 and 5.6.
of litter was signicantly higher (P < 0.005) for needles than for Ammonium and nitrate concentration in leachate remained low
roots (Fig. 2). for all litter types except seven-year-old roots (Fig. 5). Ammonium
Fresh litter types showed increasing SUVA285, i.e. an increasing concentration decreased and nitrate concentration increased with
aromaticity of DOM, through the rst four measurements, after time in leachate from the seven-year-old roots, indicating onset of
which values stabilised, whereas SUVA285 values for other litter nitrication. The relatively high concentration of inorganic nitrogen
types remained almost constant over time (Fig. 3). Seven-year-old in leachate from the seven-year-old roots is probably due to
roots had the highest SUVA285 throughout the study, indicating
DOM with a high aromaticity. 80
Accumulat ed CO 2 and DO C

The fraction of carbon leached as DOC decreased with time for a

all litter types except fresh needle litter (Fig. 4). After the rst three
60 CO2-C
percolations, all litter types except seven-year-old roots stabilised b
( mg C g-1)

at around 20% (Fig. 4). The seven-year-old roots had the highest DOC
proportion of carbon lost as DOC, ranging from 60% initially to 30% 40
during the later phase of the experiment.
c
cdd
A A
Table 1
20 d
B BC
Total carbon (Tot-C) and nitrogen (Tot-N) content (%) and C:N ratio in the ve C
different litter types.
0
Substrate Tot-C (%) Tot-N (%) C:N Fresh needle Aged Fresh roots, Dead roots, Seven-year-
litter needles, mineral soil mineral soil old roots
Fresh needle litter 48.0 0.59 82
Aged needles, litter layer 47.9 1.15 42
litter layer
Fresh roots, mineral soil 44.8 0.88 51
Fig. 2. Accumulated losses of respired C (CO2eC) and leached dissolved organic C
Dead roots, mineral soil 43.6 0.75 58
(DOC) during 19 weeks of incubation (n 4, mean  SE). Bars with different letters
Seven-year-old roots 46.4 1.17 40
differ signicantly (p < 0.05).

Please cite this article in press as: Hansson, K., et al., Amounts of carbon mineralised and leached as DOC during decomposition of Norway
spruce..., Soil Biology & Biochemistry (2009), doi:10.1016/j.soilbio.2009.10.013
 ARTICLE IN PRESS

K. Hansson et al. / Soil Biology & Biochemistry xxx (2009) 1e8 5

0.04 Fresh needle litter

Aged needles, litter layer
7.0 Fresh roots, mineral soil
Seven -year-old roots
0.03 Dead roots, mineral soil
SUVA285 (L cm g-1)

6.5

6.0
0.02

pH
Fresh needle litter
5.5
Aged needles, litter layer

0.01 Fresh roots, mineral soil

Dead roots, mineral soil 5.0
Seven-year-old roots
4.5
0.00 1.0
0 5 10 15 20
Time (weeks) 0.8

NH4 ( mg l-1)
Fig. 3. SUVA285 (ultraviolet absorbance at 285 nm divided by DOC concentration)
measured in leachate (n 4, mean  SE). Increasing SUVA indicates increasing 0.6
aromaticity of DOC.

0.4
a combination of low litter quality and low C:N ratio compared with
the other litter types.
0.2
3.4. DOC biodegradation
0.0
Most mineralisation of DOC in leachate took place in the rst 1.0
3 days, when 7e45% of the DOC was mineralised (Table 2). DOM
originating from fresh and dead roots had signicantly higher 0.8
mineralisation than that originating from needles (P < 0.005),
NO 3 (mg l -1)

whereas DOM from seven-year-old roots did not signicantly differ

0.6
from DOM from needles. For the root litters, DOM produced during
decomposition became more recalcitrant with increasing degree of
decomposition, i.e. the fraction mineralised decreased in the 0.4
following order: fresh roots > dead roots > seven-year-old roots.
No correlation was found between the fraction mineralised and 0.2
SUVA285 (P 0.715).
0.0
3.5. Sorption to ferrihydrite 0 5 10 15 20
Time (weeks)
Initially, DOM from more decomposed litter types sorbed more
strongly to ferrihydrite than DOM from fresh litter types (Fig. 6a). Fig. 5. Leachate chemistry. Differences in pH measured in leachate (n 4, mean  SE).
Nitrogen as NH4eN and NO3eN measured in leachate (n 4, mean  SE).
Later, the fraction of sorbed DOM did not differ markedly between
litter types and ranged between 80 and 95%. As indicated by the
small standard error bars in Fig. 6, the pH dependency of DOM
80 Fresh needle litter sorption was low for all litter types in the pH range 4.5e6.5.
Aged needles, litter layer We found no signicant correlation between percentage DOC
70
sorbed and SUVA285 (P 0.078), even though such a relationship
Fresh roots, mineral soil
60 between aromaticity and ability to be sorbed by the ferrihydrate may
Dead roots, mineral soil
still exist. The ratio between SUVA285 before and after contact with
50 Seven-year-old roots the ferrihydrite was highest at the rst test for all litter types, indi-
% DOC

40 cating initial preferential sorption of aromatic compounds (Fig. 6b).

30
Table 2
20 Carbon mineralisation (% of initial C) after 3 and 57 days from DOM obtained after 12
weeks of column incubation (n 4, mean, SE in brackets). Means with different
10 letters differ signicantly (p < 0.05).

Substrate Day 3 Day 57

0
0 5 10 15 20 Fresh needle litter 6.8 (5.5) a 14 (3.4) a
Aged needles 21 (11) ab 24 (9.5) ab
Time (weeks) Fresh roots from mineral soil 45 (5.3) c 50 (5.4) c
Dead roots from mineral soil 31 (9.7) bc 36 (8.3) bc
Fig. 4. Percentage of dissolved organic carbon (DOC/(DOC CO2)  100) leached from
Seven-year-old dead roots 11 (5.2) ab 21 (1.2) ab
columns (n 4, mean  SE).

Please cite this article in press as: Hansson, K., et al., Amounts of carbon mineralised and leached as DOC during decomposition of Norway
spruce..., Soil Biology & Biochemistry (2009), doi:10.1016/j.soilbio.2009.10.013
 ARTICLE IN PRESS

6 K. Hansson et al. / Soil Biology & Biochemistry xxx (2009) 1e8

Fresh roots had the highest ratio (2.9), whereas dead roots from decomposition rates in forest soils (Silver and Miya, 2001). Bird et al.
mineral soil had a ratio close to one and consequently no preferential (2008) attribute the lower decomposition rate of roots (compared
sorption. In the experiment carried out on leachate obtained after 19 with needles) to lower litter quality, with less labile constituents.
weeks of incubation, all litter types had a ratio between 1.0 and 1.4, Taylor et al. (1991) reported differences in initial chemical quality of
indicating only weak preferential sorption of aromatic compounds. litter for different litter types, with 35e37% lignin and 17e29% labile
Leachate from both needle litter types had signicantly higher compounds for coniferous roots, compared with 15% lignin and 49%
Mg and Ca concentrations than leachate from all root litter types labile compounds for spruce needles. They concluded that lignin
(P < 0.05), (Table 3). For Fe, Na, K, and Al there were no clear content is the most reliable indicator of decomposition rate, fol-
differences between litter types. lowed by nutrient content. In a litterbag study, Palviainen et al.
(2004) suggest that lower initial N and P concentrations in roots
4. Discussion compared with needles can explain the lower mass loss in roots, in
combination with higher lignin content and a smaller fraction of
Respiration rates were highest for fresh litter types (needles and soluble compounds. In our study, however, initial N concentration
roots), indicating that they contained a relatively large amount of was higher in roots than in needles, with a lower C:N ratio for roots
easily degradable substances (Fig. 1a). After a few weeks of incuba- (Table 1). The difference in mass loss is therefore more probably
tion the easily degradable substances in the fresh litter had explained by differences in the structure of carbon compounds.
decomposed and fresh and aged litter no longer differed. However, For root litter, leached DOC followed a similar pattern to
the signicant difference between needles and roots persisted, respired C, with initially high DOC leaching for fresh roots, in
suggesting that roots decompose more slowly than needles. This is agreement with ndings by Uselman et al. (2007). The fresh needle
consistent with other studies on root and foliage litter decomposi- litter differed from this pattern, with lower initial DOC leaching
tion (Taylor et al., 1991; Heim and Frey, 2004; Palviainen et al., 2004; (Figs. 1b and 4). A possible explanation for the lower initial DOC
Bird et al., 2008). Litter quality is important for predicting litter leaching in fresh needles compared with roots is that the needles
contained more easily degradable substances (carbohydrates,
sugars, amino acids), resulting in losses as CO2 rather than DOC.
a 100 Another explanation is that the fresh needles still had a protective
wax layer that prevented leaching of DOC.
90
The SUVA285 was initially lower for DOM from fresh litter types,
80 indicating a larger proportion of organic compounds with low
aromaticity. This is in agreement with results reported in a study on
70
% C adsorbed

DOM leaching from forest litter, where UV absorbance of initially

60 leached DOM was very low (Hagedorn and Machwitz, 2007). In the
present study, seven-year-old roots had the highest SUVA285 and
50
the lowest pH and respiration. In this litter type, most of the easily
40 Fresh needle litter degradable compounds were already decomposed, resulting in
Aged needles, litter layer DOM with a high degree of aromaticity. The SUVA285 for seven-
30
Fresh roots, mineral soil year-old roots ranged between 0.030 and 0.034 L mg
1 cm
1, which
20 Dead roots, mineral soil could be compared with the average SUVA285 value for O horizon
Seven-year-old roots leachates at the site of 0.023 L mg
1 cm
1 (Fro berg et al., 2005). The
10
latter value represents a mixture of DOM originating from a range
0 of different substrates and decomposition stages; in our study
5 10 15 20
Time (weeks)
Table 3
b 3.5 Fe, Na, K, Ca, Mg and Al concentration (mg L
1) in pooled samples of leachate
obtained after 6, 9 and 19 weeks of column incubation.
Relative change in SUVA 285

3 Week Fe Na K Ca Mg Al
Fresh needle litter
2.5 6 0.02 6.62 5.80 2.14 1.04 0.52
9 0.42 4.58 4.16 3.58 1.21 1.05
2 19 0.02 5.80 5.36 4.74 1.54 1.04

Aged needles, litter layer

1.5 6 0.05 5.56 9.82 9.04 2.36 0.76
9 0.09 2.82 1.78 4.14 0.83 0.36
1 19 0.02 5.06 4.90 4.34 1.00 0.39

Dead roots, mineral soil

0.5 6 0.18 10.04 5.26 0.43 0.24 0.49
9 0.19 4.84 4.46 0.48 0.37 0.38
0 19 0.18 5.90 6.88 0.64 0.80 0.80
5 10 15 20 Fresh roots, mineral soil
Time (weeks) 6 0.24 13.00 9.28 0.58 0.49 0.59
9 0.58 4.34 3.16 0.29 0.15 0.29
Fig. 6. Sorption to ferryhydrite shown as a) percentage of C in leachate adsorbed to 19 0.20 5.70 7.50 0.33 0.51 0.55
ferrihydrite and b) ratio between SUVA285 before and after sorption. A ratio of 1 means
Seven-year-old roots
no effect of quality on sorption, while a ratio larger than 1 indicates preferential
6 0.09 6.54 5.92 0.32 0.14 0.25
sorption of hydrophobic compounds and a ratio of less than 1 preferential sorption of
9 0.14 7.20 5.58 0.47 0.08 0.39
hydrophilic compounds. Pooled samples of leachate after 6 (n 6), 9 (n 5) and 19
19 0.11 7.28 6.94 0.47 0.13 0.57
(n 5) weeks of incubation (mean  SE).

Please cite this article in press as: Hansson, K., et al., Amounts of carbon mineralised and leached as DOC during decomposition of Norway
spruce..., Soil Biology & Biochemistry (2009), doi:10.1016/j.soilbio.2009.10.013
 ARTICLE IN PRESS

K. Hansson et al. / Soil Biology & Biochemistry xxx (2009) 1e8 7

varying from 0.010 L mg
1 cm
1 (fresh needles) to the ferrihydrite showed preferential sorption of hydrophobic
0.034 L mg
1 cm
1 (seven-year-old roots) (Fig. 3). compounds (Fig. 6b), but at the third adsorption test such prefer-
In general, the fraction of carbon lost from the columns as DOC ential sorption was very weak. This was expected, since most easily
decreased over time (Fig. 4). After the rst three weeks of degraded, hydrophilic substances were decomposed by then and
decreasing percentage DOC, all litter types seemed to stabilise, differences in SUVA285 between litter types were small (Fig. 3).
although seven-year-old roots stabilised at a higher fraction of DOC When estimating the contribution of DOC and root litter input to
than the other litter types. To our knowledge, there are no previous the steady-state carbon pool, using DyDOC, Michalzik et al. (2003)
studies of DOC production from such old ne roots. In a study on assumed that root litter had similar DOC production rates to needle
the release of DOC by plant tissues (Moore and Dalva, 2001), fresh litter. They also assumed the DOC quality of roots and needles to be
maple leaves were found to release 58% of lost carbon as DOC, similar. Our results suggest that the rst assumption probably not
whereas old (over-wintered) maple leaves only released 28% as holds, whereas the second does. Berg et al. (2000) suggest that
DOC. Those authors concluded that chemical composition and substrate quality can be the main controlling factor for litter
degree of decomposition of the substrate are important in decomposition rates. However, they ignore the contribution of root
controlling DOC production, with less DOC released from more litter in the forest oor when trying to explain the present carbon
decomposed materials. However, Kalbitz et al. (2006) reported that stocks and dynamics in the forest oor of Swedish spruce forests
DOC production from decaying needles decreases in the rst phase and they only include above-ground litter, mainly needles. Our
of litter decomposition, whereas an increase takes place with results show that it is important to take litter origin (i.e. above- vs.
further ongoing litter decomposition because of degradation of below-ground litter) into account when estimating DOC production
lignin. Therefore, lignin-derived compounds can comprise a large and its contribution to soil carbon sequestration. At our study site,
proportion of total DOC. The large proportion of DOC from seven- Asa in southern Sweden, above-ground litter production is esti-
year-old roots in our study indicates that roots follow the same mated to be 118 g C m
2 yr
1 and ne root litter production in the O
pattern. Seven-year-old roots had the lowest C:N ratio, the highest horizon 27 g C m
2 yr
1(Kleja et al., 2008). If we assume the mass
total nitrogen content in litter and the highest aromaticity loss rate, i.e. carbon losses as CO2 and DOC, to be twice as large for
(SUVA285) of DOC. Nitrate and ammonium content in leachate were needles as for roots throughout the decomposition, then the rela-
also highest for seven-year-old roots. These are all signs of strong tive contribution from ne roots to carbon sequestration in the O
decomposition, with all carbon bound in complex structures, horizon is 31% (27/(27(118/2))). Regarding the role of root litter-
leading to carbon deciency. derived DOC compared with DOC originating from forest oor
DOM biodegradation and the ability of DOM to bind to iron leachate to carbon build-up in mineral soil layers, our results
(hydr)oxide surfaces in the soil are important factors affecting the suggest that root litter-derived DOC makes a signicant contribu-
contribution of ne roots to carbon sequestration in mineral soils. tion. During root decomposition about 20% will be lost as DOC. In
Biodegradation of DOM and sorption capacity are both closely the mineral soil at Asa, ne root (<1 mm) litter input is
related to chemical properties of DOM (Kalbitz et al., 2003, 2005). 74 g C m
2 yr
1, resulting in an approximate input of DOC to the
In our study, DOM from roots had a higher mineralisation rate than mineral soil of 15 g C m
2 yr
1. Compared to 28 g C m
2 yr
1, which
DOM from needles (Table 2), which was unexpected since needles is the DOC input originating from the O horizon, this means that
had higher respiration and DOC production. However, as decom- 35% of the DOC in the mineral soil comes from root litter decom-
position of roots proceeds, the recalcitrance of root-derived DOM position. Even though this gure is a rough estimate, it clearly
seems to become similar to that of needle litter, as indicated by the suggests that root litter-derived DOC will make a signicant
low fraction of DOM mineralised in the seven-year-old root contribution to the carbon build-up in mineral soil layers.
leachate. There were no signicant differences in mineralisation
rates between fresh and older litter types except for seven-year-old 5. Conclusions
root leachate, which differed signicantly from fresh root leachate
but not from needle leachate. However, the mineralisation study Root and needle litter differ in both CO2 and DOC production.
was carried out on water samples from week 12, when differences Our results suggest that root-derived DOC can signicantly
in SUVA285 and consequently in DOC quality were small (Fig. 3). In contribute to carbon sequestration in mineral soil layers. Roots had
such cases, a close relationship between SUVA and degradability a lower total mass loss than needles, increasing their relative
cannot be expected. Most mineralisation of DOC in leachate took contribution to carbon build-up in the soil. Old roots had a signi-
place in the rst 3 days of incubation (Table 2). This is in agreement cantly higher fraction of carbon lost as DOC than fresher litter types.
with previous studies. Kalbitz et al. (2003) reports a half-life of the Furthermore, DOM from roots in later stages of decomposition
labile DOC pool of between 2.6 and 5.0 days for DOC originating bound more strongly to ferrihydrite than DOC from fresh litter
from the O horizon in spruce forest, while in a study by Don and types and was mineralised at a lower rate, suggesting that roots at
Kalbitz (2005) a half-life of 0.9 and 4.0 days is reported for DOC late stages of decomposition, although a small fraction of total litter,
leached from decomposed and fresh spruce litter. signicantly contribute to carbon build-up in mineral soils.
In a recent study, Mikutta et al. (2007) showed that binding of Even though roots and needles seem to follow the same
organic matter to mineral surfaces generally decreased its biode- decomposition pathways, roots have a slower decomposition rate.
gradability. Consequently, compounds capable of sorbing strongly This needs to be taken into account when modelling litter
to mineral surfaces are more likely to be preserved for a long time decomposition.
in the soil than other compounds. In our study, DOM from older
litter types was initially strongly sorbed to ferrihydrite (Fig. 6a), Acknowledgements
whereas DOM from fresh litter types, with easily degradable
substances with a low aromaticity, was not as strongly sorbed. At This work is dedicated to Professor Hooshang Majdi, who passed
the second and third adsorption test, after 9 and 19 weeks of away in the initial phase of this study. He contributed to initiate this
incubation, most easily degradable, hydrophilic compounds were study and with old roots from a litterbag study. Thanks to Go ran
degraded, even in the fresh litter types, as indicated by the strong 
Agren and Tryggve Persson for valuable comments. Thanks to
sorption to ferrihydrite for all litter types. At the rst adsorption Tomas Gro nqvist for laboratory assistance, and to personnel at Asa
test, the high ratio between SUVA285 before and after contact with Research Station for helping collect litter. This research was

Please cite this article in press as: Hansson, K., et al., Amounts of carbon mineralised and leached as DOC during decomposition of Norway
spruce..., Soil Biology & Biochemistry (2009), doi:10.1016/j.soilbio.2009.10.013
 ARTICLE IN PRESS

8 K. Hansson et al. / Soil Biology & Biochemistry xxx (2009) 1e8

supported by the Swedish Research Council for Environment, King, A.W., Post, W.M., Wullschleger, S.D., 1997. The potential response of terrestrial
carbon storage to changes in climate and atmospheric CO2. Climatic Change 35,
Agricultural Sciences and Spatial Planning.
199e227.
Kleja, D.B., Svensson, M., Majdi, H., Langvall, O., Jansson, P.E., Lindroth, A.,
Weslien, P., Bergkvist, B., Johansson, M.B., 2008. Pools and uxes of carbon in
three Norway spruce ecosystems along a climatic gradient in Sweden.
References Biogeochemistry 89, 7e25.
Majdi, H., 2004. Root and needle litter decomposition responses to enhanced
Aitkenhead-Peterson, J.A., Kalbitz, K., 2005. Short-term response on the quantity supplies of N and S in a Norway spruce forest in southwest Sweden. Plant
and quality of rhizo-deposited carbon from Norway spruce exposed to low and Biosystems 138, 225e230.
high N inputs. Journal of Plant Nutrition and Soil Science 168, 687e693. McDowell, W.H., Zsolnay, A., Aitkenhead-Peterson, J.A., Gregorich, E.G., Jones, D.L.,
Berg, B., 2000. Litter decomposition and organic matter turnover in northern forest Jodemann, D., Kalbitz, K., Marschner, B., Schwesig, D., 2006. A comparison of
soils. Forest Ecology and Management 133, 13e22. methods to determine the biodegradable dissolved organic carbon from
Berg, B., Johansson, M.B., Meentemeyer, V., 2000. Litter decomposition in a transect different terrestrial sources. Soil Biology and Biochemistry 38, 1933e1942.
of Norway spruce forests: substrate quality and climate control. Canadian Michalzik, B., Tipping, E., Mulder, J., Lancho, J.F.G., Matzner, E., Bryant, C.L., Clarke, N.,
Journal of Forest Research 30, 1136e1147. Lofts, S., Esteban, M.A.V., 2003. Modelling the production and transport of
Berggren, D., Bergkvist, B., Johansson, M.-B., Langvall, O., Majdi, H., Melkerud, P.-A., dissolved organic carbon in forest soils. Biogeochemistry 66, 241e264.
Nilsson,  A., Weslien, P., Olsson, M., 2004. A Description of LUSTRA's Common Mikutta, R., Mikutta, C., Kalbitz, K., Scheel, T., Kaiser, K., Jahn, R., 2007. Biodegra-
Field Sites. Report 87. Department of Forest Soils, SLU, Uppsala. dation of forest oor organic matter bound to minerals via different binding
Bird, J.A., Kleber, M., Torn, M.S., 2008. 13C and 15N stabilization dynamics in soil mechanisms. Geochimica et Cosmochimica Acta 71, 2569e2590.
organic matter fractions during needle and ne root decomposition. Organic Moore, T.R., Dalva, M., 2001. Some controls on the release of dissolved organic
Geochemistry 39, 465e477. carbon by plant tissues and soils. Soil Science 166, 38e47.
Callesen, I., Liski, J., Raulund-Rasmussen, K., Olsson, M.T., Tau-Strand, L., Vesterdal, L., Olsson, M.T., Erlandsson, M., Lundin, L., Nilsson, T., Nilsson,  A., Stendahl, J., 2009.
Westman, C.J., 2003. Soil carbon stores in Nordic well-drained forest soils - Organic carbon stocks in Swedish podzol soils in relation to soil hydrology and
relationships with climate and texture class. Global Change Biology 9, 358e370. other site characteristics. Silva Fennica 43, 209e222.
Chin, Y.-P., Aiken, G., O'Loughlin, E., 1994. Molecular weight, polydispersity, and Palviainen, M., Finer, L., Kurka, A.M., Mannerkoski, H., Piirainen, S., Starr, M., 2004.
spectroscopic properties of aquatic humic substances. Environmental Science & Decomposition and nutrient release from logging residues after clear-cutting of
Technology 28, 1853e1858. mixed boreal forest. Plant and Soil 263, 53e67.
Chorover, J., Amistadi, M.K., 2001. Reaction of forest oor organic matter at goethite, Ruess, R.W., Hendrick, R.L., Burton, A.J., Pregitzer, K.S., Sveinbjornsson, B., Allen, M.E.,
birnessite and smectite surfaces. Geochimica et Cosmochimica Acta 65, 95e109. Maurer, G.E., 2003. Coupling ne root dynamics with ecosystem carbon cycling in
Don, A., Kalbitz, K., 2005. Amounts and degradability of dissolved organic carbon black spruce forests of interior Alaska. Ecological Monographs 73, 643e662.
from foliar litter at different decomposition stages. Soil Biology and Biochem- Schwertmann, U., Cornell, R.M., 2000. Iron Oxides in the Laboratory. Preparation
istry 37, 2171e2179. and Characterization. Wiley-VCH, Weinheim.
FAO., 1990. FAO-Unesco Soil Map of the World. Revised Legend. In: Soils Bulletin, Silver, W., Miya, R., 2001. Global patterns in root decomposition: comparisons of
vol.60. climate and litter quality effects. Oecologia 129, 407e419.
berg, M., Kleja, D.B., Bergkvist, B., Tipping, E., Mulder, J., 2005. Dissolved organic
Fro berg, G., Bergkvist, B., Berggren, D., Nilsson, S.I., 2003. Long-term N addition
Sjo
carbon leaching from a coniferous forest oor e a eld manipulation experi- effects on the C mineralization and DOC production in mor humus under
ment. Biogeochemistry 75, 271e287. spruce. Soil Biology and Biochemistry 35, 1305e1315.
Hagedorn, F., Machwitz, M., 2007. Controls on dissolved organic matter leaching Swedlund, P.J., Webster, J.G., 1999. Adsorption and polymerisation of silicic acid on
from forest litter grown under elevated atmospheric CO2. Soil Biology & ferrihydrite, and its effect on arsenic adsorption. Water Research 33,
Biochemistry 39, 1759e1769. 3413e3422.
Heim, A., Frey, B., 2004. Early stage litter decomposition rates for Swiss forests. Taylor, B.R., Prescott, C.E., Parsons, W.J.F., Parkinson, D., 1991. Substrate control of
Biogeochemistry 70, 299e313. litter decomposition in four Rocky Mountain coniferous forests. Canadian
Hongve, D., van Hees, P.A.W., Lundstrom, U.S., 2000. Dissolved components in Journal of Botany 69, 2242e2250.
precipitation water percolated through forest litter. European Journal of Soil Traina, S.J., Novak, J., Smeck, N.E., 1990. An ultraviolet absorbance method of
Science 51, 667e677. estimating the percent aromatic carbon content of humic acids. Journal of
Kaiser, K., 2003. Sorption of natural organic matter fractions to goethite (a-FeOOH): Environmental Quality 19, 151e153.
effect of chemical composition as revealed by liquid-state 13C NMR and wet- Uselman, S.M., Qualls, R.G., Lilienfein, J., 2007. Contribution of root vs. leaitter to
chemical analysis. Organic Geochemistry 34, 1569e1579. dissolved organic carbon leaching through soil. Soil Science Society of America
Kalbitz, K., Kaiser, K., Bargholz, J., Dardenne, P., 2006. Lignin degradation controls Journal 71, 1555e1563.
the production of dissolved organic matter in decomposing foliar litter. Euro- Vogt, K.A., Persson, H., 1991. Measuring Growth and Development of Roots. Tech-
pean Journal of Soil Science 57, 504e516. niques and Approaches in Forest Tree Ecophysiology. CRC Press, pp. 477e501.
Kalbitz, K., Schmerwitz, J., Schwesig, D., Matzner, E., 2003. Biodegradation of soil- Weishaar, J.L., Aiken, G.R., Bergamaschi, B.A., Fram, M.S., Fujii, R., Mopper, K., 2003.
derived dissolved organic matter as related to its properties. Geoderma 113, Evaluation of specic ultraviolet absorbance as an indicator of the chemical
273e291. composition and reactivity of dissolved organic carbon. Environmental Science
Kalbitz, K., Schwesig, D., Rethemeyer, J., Matzner, E., 2005. Stabilization of dissolved & Technology 37, 4702e4708.
organic matter by sorption to the mineral soil. Soil Biology and Biochemistry 37, Zsolnay, A, 2003. Dissolved organic matter: artefacts, denitions, and functions.
1319e1331. Geoderma 113, 187e209.

Please cite this article in press as: Hansson, K., et al., Amounts of carbon mineralised and leached as DOC during decomposition of Norway
spruce..., Soil Biology & Biochemistry (2009), doi:10.1016/j.soilbio.2009.10.013