Response of soil C and N transformations to tannin fractions originating

from Scots pine and Norway spruce needles

Sanna Kanerva
a,
*
, Veikko Kitunen
a
, Oili Kiikkila
a
, Jyrki Loponen
b
, Aino Smolander
a
a
Finnish Forest Research Institute, Vantaa Research Centre, P.O. Box 18, FI-01301 Vantaa, Finland
b
Department of Chemistry, University of Turku, Vatselankatu 2, FI-20014 Turku, Finland
Received 17 May 2005; received in revised form 4 October 2005; accepted 13 October 2005
Available online 4 January 2006
Abstract
Tannins are polyphenolic compounds that may inuence litter decomposition, humus formation, nutrient (especially N) cycling and ultimately,
plant nutrition and growth. The aim of this study was to determine the response of C and N transformations in soil to tannins of different molecular
weight from Norway spruce (Picea abies (L.) Karst) and Scots pine (Pinus sylvestris L.) needles, tannic acid and cellulose. Arginine was added to
test whether the soil microbial community was limited by the amount of N, and arginineCtannin treatments were used to test whether the effects
of tannins could be counteracted by adding N. Soil and needle samples were taken from adjacent 70-year-old Scots pine and Norway spruce stands
located in Kivalo, northern Finland. Tannins were extracted from needles and fractioned based on molecular weight; the fractions were then
characterized by LCMS and GCMS. Light fractions contained tannin monomers and dimers as well as many other compounds, whereas heavy
fractions consisted predominantly of polymerized condensed tannins. Spruce needles contained more procyanidin than prodelphinidin units, while
in pine needles prodelphinidin units seemed to be dominant. The fractions were added to soil samples, pine fractions to pine soil and spruce
fractions to spruce soil, and incubated at 14 8C for 6 weeks. CO
2
evolution was followed throughout the experiment, and the rates of net
mineralization of N and net nitrication, concentration of dissolved organic N (DON) and amounts of microbial biomass C and N were measured
at the end of the experiment. The main effects of the fractions were similar in both soils. Light fractions strongly enhanced respiration and
decreased net N mineralization, indicating higher immobilization of N in the microbial biomass. On the contrary, heavy fractions reduced
respiration and slightly increased net N mineralization, suggesting toxic or protein-precipitating effects. The effects of tannic acid and cellulose
resembled those of light fractions. DON concentrations generally decreased during incubation and were lower with heavy fractions than with light
fractions. No clear differences were detected between the effects of light and heavy fractions on microbial biomass C and N. Treatments that
included addition of arginine generally showed trends similar to treatments without it, although some differences between light and heavy
fractions became more obvious with arginine than without it. Overall, light fractions seemed to act as a labile source of C for microbes, while
heavy fractions were inhibitors.
q 2005 Elsevier Ltd. All rights reserved.
Keywords: Forest soil; Microbial activities; Mineralization; Nitrogen cycling; Norway spruce; Polyphenols; Scots pine; Tannins
1. Introduction
Tannins are polyphenolic compounds with the ability to
form stable complexes with proteins and other compounds.
They can be divided in to two main classes: condensed tannins,
which are also called proanthocyanidins, and hydrolyzable
tannins. Condensed tannins can be further divided into
subclasses such as procyanidins and prodelphinidins. While
gymnosperms and monocots produce only condensed tannins,
dicots can produce either condensed or hydrolyzable tannins or
a mixture of the two (reviewed by Kraus et al., 2003). In woody
species, foliar concentrations of tannins commonly range from
15 to 25% dry weight (reviewed by Kraus et al., 2004). Tannins
may inuence litter decomposition rates, humus formation,
nutrient (especially N) cycling and ultimately, plant nutrition
and growth (e.g. Schimel et al., 1996; Bradley et al., 2000;
Fierer et al., 2001). Tannins from various plant species have
been shown to affect N mineralization, induce toxicity in
microbes and affect enzyme activities in soil (Schimel et al.,
1996; Bradley et al., 2000; Fierer et al., 2001; Kraus et al., 2003
and references therein). Hence, there is very strong evidence
that tannins play an important role in interspecic competition,
and many studies have suggested that individual plants may be
important in nutrient cycling on the ecosystem level (Schimel
Soil Biology & Biochemistry 38 (2006) 13641374
www.elsevier.com/locate/soilbio
0038-0717/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.soilbio.2005.10.013
* Corresponding author. Tel.: C358 10 211 2595; fax: C358 10 211 2206.
E-mail address: sanna.kanerva@metla. (S. Kanerva).
et al., 1996; Chen and Stark, 2000; Fierer et al., 2001; Castells
et al., 2003; Kraus et al., 2004).
Concentration and composition of phenolic compounds in
the soil seem to vary depending on the plant species growing in
it (e.g. Kuiters and Denneman, 1987; Smolander et al., 2005).
Tannin reactivity in soil has been suggested to be based on such
characteristics as condensed versus hydrolyzable tannins and
procyanidin versus prodelphinidin content of the tannins
(Kraus et al., 2004). In addition, molecular weight or degree
of polymerization of tannins or phenolic compounds seems to
be an important factor when their inuence on soil nutrient
cycling is considered. Schimel et al. (1996) and Fierer et al.
(2001) demonstrated that high molecular weight phenolics
from balsam poplar acted as a general microbial inhibitor,
while the effects of lower molecular weight phenolics were less
predictable and depended on prior exposure of the soil
microbial community to related molecules; microbial commu-
nities previously exposed to smaller chain tannins were more
likely to use them as a C substrate, while in the communities
that had limited exposure to tannins they were more likely to
prove toxic.
Norway spruce (Picea abies (L.) Karst) and Scots pine
(Pinus sylvestris L.) are the dominant tree species in Finland.
Both similarities and differences have been reported in the C
and N transformations of soil under these species (Priha and
Smolander, 1997, 1999; Smolander and Kitunen, 2002). The
aim of this study was to nd out the response of soil C and N
transformations to tannins of different molecular weight from
Norway spruce and Scots pine needles. Fractions prepared
from spruce and pine needles were added to spruce and pine
soils, respectively, to examine their effects on microbial
activities by measuring CO
2
evolution, net mineralization of
N and net nitrication rates, concentrations of dissolved
organic nitrogen (DON) and amounts of C and N in the
microbial biomass. The availability or inhibition of these
fractions to bacteria and fungi was also assessed.
2. Materials and methods
2.1. Study site, soil and needle sampling and chemical analysis
The stands used in this study were adjacent 70-year-old
stands in Kivalo, northern Finland (66820
0
N/26840
0
E), which
were dominated by Scots pine, Norway spruce or silver birch
(Betula pendula Roth) growing in soil that originally was
similar in all three stands. The soil type was podzolic and
humus type mor. Three study plots (25!25 m) were placed in
each stand. The coniferous stands also contained species other
than the dominant one (the spruce stand contained 76% spruce
and 24% other species, and the pine stand contained 88% pine
and 12% tree species other than pine). For a more detailed
description of the study site and the tree stands, see Smolander
and Kitunen (2002).
In August 2001 soil samples (2030 cores, core diameter
58 mm) were taken systematically from the humus layer of
spruce and pine plots. The samples were combined to give one
composite sample per plot, and the composite samples from
each plot were combined to give one sample that represented
one stand. After the green plant material was removed, the
samples were sieved through a 4.0 mm mesh and stored in
plastic bags at 4 8C until used. Content of soil organic matter
(o.m.) was measured as loss-on-ignition at 550 8C. The soil
characteristics have been described earlier (Smolander and
Kitunen, 2002); the pH (H
2
O) of both conifer soils was 4.0, and
the C-to-N ratio in spruce soil was 37 and in pine soil 32.
Undamaged bulk green needles were collected from the pine
and spruce plots in spring 2001. After collection, the needles
were freeze dried and nely ground.
2.2. Extraction, fractionation and analysis of tannin fractions
Tannins were extracted and fractioned as described in Fierer
et al. (2001) but with some modications. The ground plant
material (1300 g spruce/1070 g pine needles) was placed in a
steel container and 5 l of hexane was added. The material was
soaked in hexane and stirred occasionally with a power drill; at
the end the solvent was decanted from the plant material. The
plant material was extracted again with 2.5 l hexane, stirred and
decanted. This procedure was repeated two more times. After
that, the remaining plant material was extracted with acetone
water (70:30) overnight. The next day the suspension was
stirred and ltered through a lter paper (S & S 589
3
). The
ltrate was collected in a glass bottle. Needles were extracted
again with acetonewater (70:30), stirred 30 min and ltered.
The procedure was repeated once overnight. The three
acetonewater extracts were combined and concentrated by
roto-evaporation. This concentrated extract was then extracted
with 100% ethyl acetate for 30 min at 200 rev min
K1
. Ethyl
acetate and water were separated with a siphon, and the
procedure was repeated three more times. The ethyl acetate
fractions were combined and roto-evaporated to dryness. This
fraction was labelled F1.
The acetonewater fraction was loaded onto a Sephadex
LH-20 column that had previously been equilibrated with
methanolwater (50:50). The column was eluted with
methanolwater (50:50) followed by acetonewater (70:30)
until the eluate was colourless. The acetonewater fraction was
concentrated by roto-evaporation.
The concentrated acetonewater fraction was loaded into a
clean Sephadex LH-20 column that had previously been
equilibrated with methanolwater (50:50). The acetonewater
fraction was eluted with 100% ethanol, and the eluate was
concentrated by roto-evaporation and labelled F2.
The extract in the LH-20 column was then eluted with 100%
methanol. The eluate was concentrated and labelled F3.
Finally, the extract in the column was extracted with
acetonewater (70:30), and the eluate was concentrated and
labelled F4.
All four fractions were analyzed using thin layer chroma-
tography (TLC) to determine their composition. The solvent
system for TLC analyses was toluene/acetone/formic acid
(3/6/1). The components in each fraction were detected by UV-
light. In addition, the compositions of the fractions, in terms of
number and types of monomer units, molecular weights of
S. Kanerva et al. / Soil Biology & Biochemistry 38 (2006) 13641374 1365
proanthocyanidins (up to heptamers) and identication of
anthocyanidin monomers (procyanidin/prodelphinidin), as
well as a tannic acid product (Merck, Tannic acid powder
pure, DAB7, FU, pH Helv, USP) were conrmed by reversed-
phase and normal-phase high-performance liquid chromatog-
raphy (RP- and NP-HPLC) coupled with an ultraviolet (UV)
detector and electrospray-ionization mass spectrometer (ESI
MS) (for methods see Loponen et al., 2001; Karonen et al.,
2004). Soluble condensed tannins were quantied by the
modied acid-butanol assay (proanthocyanidin assay) (Porter
et al., 1986; Terrill et al., 1992; Ossipova et al., 2001). The
standard curves for the calculations were determined by using
previously puried proanthocyanidin from leaves of mountain
birch (Betula pubescens ssp. czerepanovii) (Bae et al., 1993;
Ossipova et al., 2001). The content of low molecular weight
substances other than condensed tannins in the fractions were
analysed with GCMS. A known amount of each fraction was
weighed for GCMS analysis. Samples were analysed before
and after BSTFA derivatization and 4-chlorobenzoic acid,
erytritol and heptadecanoic acid were used as internal
references.
The fractions and the tannic acid product were mixed with
silica gel in a ratio of 1:2 (w:w), slurried in acetone or acetone
water and dried with rotary evaporation. The dry mixture was
ground in a mortar. This procedure binds tannin components to
the silica gel, which aids handling and application to soils.
2.3. Incubation experiment
Soil (corresponding to a volume of 20 ml) from both
coniferous stands was weighed into 120 ml glass bottles. The
treatments consisted of the four needle fractions (see above),
tannic acid and cellulose, 45 mg g
K1
soil o.m. each, and
control. All treatments were done in three replicates. Tannic
acid was added to test how the soils respond to hydrolyzable
tannins, since birch, which is often mixed with conifers in
natural forests, contains both hydrolyzable and condensed
tannins (Ossipova et al., 2001). Cellulose was used as a
compound that would supply C to the soil microbia without
having any specic physiological effects (such as toxicity). All
the same treatments were also done with arginine addition
(1 mg arg-N g
K1
o.m.). Arginine was added to test whether the
microbial community in the soil was limited by the amount of
N, and adding arginineCneedle fractions tested whether the
effects of the fractions could be counteracted by adding N.
Fraction-silica mixture and cellulose were added to soil dry,
while arginine was added in solution. To determine how
addition of silica gel affected soil processes, soil with and
without addition of silica was also incubated. Since the changes
in the soil properties measured were affected not much by silica
gel alone, we conclude that adding silica gel with tannin
fractions did not distort the results of this experiment, and
therefore silica gel alone is not discussed later in this article.
The four fractions isolated from spruce needles were added
to the spruce soil and those from pine needles were added to the
pine soil. After the additions, the soils were adjusted to 60%
water saturation.
The samples were incubated at 14 8C for 42 or 43 days, and
their moisture content was adjusted weekly. CO
2
evolution
bottles were sealed with rubber septa, and the other bottles
were capped with aluminium foil. Some samples of pure soil
and controls had been frozen immediately before incubation
and were extracted concurrently with the incubated samples.
The reason for this was to determine the initial concentrations
of NH
C
4
N, NO
K
2
CNO
K
3
N and total N in the soils.
2.4. Chemical and microbiological analyses
During the 6-week incubation, CO
2
evolution was measured
10 times by sampling the headspace and analyzing the amount
of CO
2
on a gas chromatograph (Priha and Smolander, 1997).
At the beginning of incubation, CO
2
was analyzed more
frequently than in the end of the run. The bottles were aerated
24 h before each sampling. To determine net N mineralization
rates over the 6-week incubation period, samples were
extracted at time zero and after the incubation with 40 ml
1 M KCl (2 h 200 rev min
K1
) and ltered through S & S 589
3
lter papers. Concentrations of NH
C
4
N, NO
K
2
CNO
K
3
N and
total N were measured using a ow injection analyzer (FIAstar
5012 analyzer C5042 detector, Tecator). To calculate net
ammonication and nitrication, initial concentrations of NH
C
4
N and NO
K
2
CNO
K
3
N from non-incubated samples were
subtracted from the nal (post-incubation) NH
C
4
N and NO
K
2
CNO
K
3
N concentrations. DON was calculated as the
difference between total N and inorganic N.
At the end of the incubation period, the amounts of C (C
mic
)
and N (N
mic
) in the microbial biomass were measured using the
chloroform fumigation-extraction method (Priha and Smo-
lander, 1997). Carbon and nitrogen contents in the microbial
biomass were calculated by subtracting the total C and N in
unfumigated samples from that in fumigated samples. Because
the results represent a ush of C and N instead of the total
microbial biomass, ushes were converted to microbial
biomass with the formulas of Martikainen and Palojarvi
(1990), which are for C
mic
in humus samples (1.9 C
f
C428)
mg g
K1
dry mass (d.m.) and for N
mic
(1.86N
f
C74.82) mg g
K1
d.m (C
f
and N
f
represent amounts of C and N measured in
fumigation-extraction).
In addition to the incubation experiment, water extracts of
the needle fractions (F1F4) and tannic acid were used to
assess the availability or inhibition of these amendments to soil
bacteria and fungi. Fractions (45350 mg) were shaken
(250 rev min
K1
) for 1 h with 20 ml of water and ltered
through a 0.45 mm PES membrane. The concentrations of C
were measured, and two concentrations of C were chosen for
testing the conditions where the extracts of the fractions could
act as substrates (30 mg C l
K1
) for or inhibitors (700 mg C l
K1
)
to bacterial and fungal growth. A preliminary experiment was
done with F3 and F4, which were expected to inhibit microbes
(Schimel et al., 1996; Fierer et al., 2001). Based on the results
of that experiment, 700 mg C l
K1
was chosen as an inhibitory
concentration, being near the 50% inhibition of bacterial
growth compared to growth in water solution. As an inoculum
of bacteria and fungi we used a mixture of humus layer
S. Kanerva et al. / Soil Biology & Biochemistry 38 (2006) 13641374 1366
collected from the same study site and also including a birch
stand (Kiikkila et al., 2005).
The rate of bacterial growth was measured using the
3
H-
thymidine incorporation technique on bacteria extracted from
soil (Baath, 1992), with the modications introduced by Baath
et al. (2001) and Kiikkila et al. (2005). Inoculum was prepared
by shaking a soil sample (3 g fresh weight, fw) with 100 ml of
water, followed by low-speed centrifugation and ltration
through quarz wool. Two replicates of the mixture of bacterial
suspension (8.6 ml) and the extracts of the fractions (11.4 ml)
were prepared, and after pre-incubation of the suspensions
(shaken in 80 rev min
K1
) at 20 8C for 1, 2, 3, 7 and 8 days, the
rate of thymidine incorporation was measured. Then 1.4 ml of
the suspension was transferred to Eppendorf tubes, 3.5 ml of
[methyl-
3
H]thymidine (925 GBq mmol
K1
) was added, and the
samples were incubated for 2 h. Washing of excess tracer and
measurement of radioactivity are described in detail by Baath
et al. (2001).
The rate of fungal growth was measured using the technique
of
14
C-acetate incorporation into ergosterol devised by Newell
and Fallon (1991) and modied by Baath (2001) for use in the
soil habitat. Extracts of fractions (30 mg C l
K1
, 30 ml) were
transferred into Erlenmeyer asks, and 0.03 g fw of soil was
added. To diminish the bacterial growth, streptomycin and
ampicillin (50 mg l
K1
) were added. The bottles were shaken at
80 rev min
K1
in the dark at 20 8C for 1, 2, 7 and 8 days. After
this pre-incubation, the suspension was ltered through a
Whatman GF/D glass ber lter. The soil and the lter were
transferred to a test tube with 1.5 ml of the extract. Samples of
700 mg C l
K1
were used only after one-day pre-incubation in
test tubes (0.03 g fw soil and 1.5 ml extract). After pre-
incubation,
14
C-acetate solution (0.05 ml, 1,2,-[
14
C]acetic acid,
sodium salt, 2.07 GBq mmol
K1
) and 1 mM non-radioactive
acetate (0.45 ml) were added to the test tubes. After the mixture
was incubated for 20 h at 20 8C, 1 ml of 5% formalin was
added, the test tubes were centrifuged and the supernatant was
discarded. The ergosterol was then extracted (Baath, 2001) and
measured with HPLC (Hitachi, Merck), and
14
C-ergosterol was
determined with the HPLC radioactivity monitor (Berthold, LB
506 C-1). The proportion of radioactivity of the total ergosterol
was calculated.
The mean of three samplings (30 mg C l
K1
) was calculated
(bacterial growth: after 1, 2, and 3 days pre-incubation; fungal
growth: after 1, 7 and 8 days pre-incubation). For 700 mg C l
K1
the sampling after one-day pre-incubation was used so that
the inhibition effect could be measured before the microbes
adapted. Relative availability (30 mg C l
K1
) or inhibition
(700 mg C l
K1
) of fractions to bacteria (TdR 30, TdR 700) or
fungi (Ac-erg 30, Ac-erg 700) was calculated by dividing the
incorporation of the sample by the incorporation in water
solution (valueZ1). Thus the higher the TdR 30 or Ac-erg 30
value the better the fraction is as substrate. Low TdR 700 or
Ac-erg 700 values mean that the fraction inhibited growth of
bacteria.
2.5. Statistical analysis
Differences in the measured characteristics between
treatments were compared with one-way analysis of variance.
ANOVA was performed separately for the two soil-treatment
combinations. When needed, transformations were made to
fulll the assumptions of the analysis of variance. Signicant
differences of the means by treatments were determined by
Tukeys test using a signicance level of P!0.05.
3. Results
Chromatographic and mass spectrometric results from
TLC, RP- and NP-HPLC-UV/ESI-MS analyses revealed that
both Norway spruce and Scots pine needle fractions F3 and
F4 contained polymers of condensed tannins that were
longer than those in the F1 and F2 fractions. Acid-butanol
assay showed that fractions F3 and F4 consisted mainly of
condensed tannins (5587%) while F1 and F2 contained
only 1.75.5% condensed tannins (Table 1). Minor amounts
of several compounds other than condensed tannins were
found in light fractions, especially in F1 (Table 2), the rest
containing other needle constituents.
Table 1
Qualitative and quantitative data for fractions of condensed tannins from spruce and pine needles
Fraction Polymeric composition of fractions Molecular weight
of fractions
Estimation of the
most abundant
CT unit
CT concentration
in fractions
(g kg
K1
)
Spruce F1 Mono-dimers 290610 PC 39
F2 Mono-trimers 290914 PC 55
F3 Tetra-heptamers and higher tannin polymers, minor
amounts of mono-trimers
11542130 and
higher
PC 622
F4 Tetra-hexamers and higher tannin polymers, trace
amounts of mono-trimers
11541826 and
higher
PC 853
Pine F1 Mono-dimers 290610 PD 20
F2 Mono-dimers 290610 PD 17
F3 Tri-tetramers and higher tannin polymers, minor
amounts of mono-dimers
8661218 and
higher
PD 551
F4 Tetramers, pentamers and higher tannin polymers,
trace amounts of mono-trimers
11541522 and
higher
PD 870
CT, condensed tannins; PC, procyanidin; PD, prodelphinidin.
S. Kanerva et al. / Soil Biology & Biochemistry 38 (2006) 13641374 1367
Spruce needles contained more procyanidin than prodel-
phinidin units while in pine needles prodelphinidin units were
dominant. In addition, HPLC-ESI-MS analysis conrmed that
the tannic acid product contained a mixture of galloylglucoses
of different molecular sizes (see also Hagerman and Butler,
1989). Acid-butanol assay with tannic acid gave no indication
of condensed tannins.
In both soils, and with and without arginine, fractions F1
and F2 and tannic acid caused a sharp increase in CO
2
production during the rst days after addition; but after that C
mineralization settled to the same level as the control (Fig. 1).
In contrast, throughout the incubation experiment both spruce
and pine F3 and F4 fractions decreased C mineralization rates
relative to the control. Cellulose showed slightly higher CO
2
production in the spruce soil during last weeks of the
incubation than the other treatments did.
Net N mineralization rates measured over the 6-week
incubation were mostly negative in the absence of arginine
Table 2
Compounds other than condensed tannins in spruce and pine needle fractions analysed after BSTFA derivatization
Compound (mg kg
K1
) Spruce Pine
F1 F2 F3 F4 F1 F2 F3 F4
Low molecular weight
phenols
518 4.5
Sesqui-and diterpenes 170 10
Resin acids 108 148 2.7
Glucose 34 14 7.9 7.6 6.7
Phenolic glucosides 45 21 1.1
Stilbene glucosides 41 292
Flavonoid monomers 228 117 6.3 4.3 233 80 38 4.0
Sterols 6.4 5.6
Cyclitols 44 23
Fatty acids 23 4.6
Fig. 1. Respiration rates during 6-week incubation for spruce and pine control soils and soils treated with spruce or pine needle fractions F1F4, respectively, tannic
acid or cellulose and the corresponding treatments with arginine addition. Mean of three replicates.
S. Kanerva et al. / Soil Biology & Biochemistry 38 (2006) 13641374 1368
(initial concentrations of NH
4
N were in the spruce soil
without and with arginine 442 and 450 mg kg
K1
o.m.,
respectively, and in the pine soil without and with arginine
229 and 264 mg kg
K1
o.m., respectively). In the spruce soil, F2
and especially F1 showed signicantly lower rate of net N
mineralization whereas F3 and F4 showed signicantly higher
rates of net N mineralization than the control did (Fig. 2).
Tannic acid and cellulose also showed signicantly lower net N
mineralization than the control did. In the pine soil the trends
were similar, but the differences were smaller and F3 did not
differ from the control (Fig. 2). With arginine addition, the
trends observed in both soils were similar and net N
mineralization rates were positive, with the exception of
cellulose added to spruce soil (Fig. 2). In the spruce soil the rate
of net nitrication was very low while in the pine soil there was
no clearly measurable NO
K
3
production (results not shown).
In both of the soils in the absence of arginine, all needle
fraction and tannic acid treatments showed lower DON
concentrations than the control did (pine F1 and F2 not signif.)
(Fig. 3). Cellulose had no effect on concentration of DON.
Similar trends were observed with arginine addition (Fig. 3),
but in both soils only the fractions F3 and F4 clearly decreased
DON and in the pine soil the F1 fraction increased DON.
In neither of the soils in the absence of arginine did the
treatments signicantly affect the amount of microbial biomass
C, except for pine F3, which decreased it signicantly (Fig. 4).
With arginine addition, F1 in spruce soil and F2 in pine soil
increased C
mic
signicantly during the incubation while the
other treatments had no effect (Fig. 4).
In the spruce soil in the absence of arginine, addition of
tannic acid resulted in signicantly higher N
mic
values than in
the control, while the other treatments had no effect (Fig. 5). In
the pine soil, F1 and cellulose increased N
mic
slightly but
signicantly and F4 decreased it signicantly, while the other
treatments had no effect (Fig. 5). In the spruce soil, with
arginine addition all treatments (except F3) showed signi-
cantly lower N
mic
values than the control did, while in the pine
soil none of the treatments differed from the control.
The relative availability of the fractions to bacteria and
fungi was studied by
3
H-thymidine and
14
C-acetate incorpor-
ation using concentrations of 30 and 700 mg C l
K1
. The
highest relative availability to bacteria at the concentration of
30 mg C l
K1
(TdR 30) was in pine F1 and F2, while all other
fractions and tannic acid were less available to bacteria. At the
concentration of 700 mg C l
K1
(TdR 700) all fractions
inhibited bacterial growth; F1 and tannic acid showed the
highest inhibition (Fig. 6). When assessed by Ac-erg 700
(valuesO1), none of the treatments, except for tannic acid,
inhibited fungi (Fig. 6). However, if growth at the higher
concentration (Ac-erg 700) is lower than at the lower
concentration (Ac-erg 30), inhibition is also possible. In spruce
and pine F3, Ac-erg 700 was clearly lower than Ac-erg 30, thus
Fig. 2. Rate of net N mineralization during 6-week incubation for spruce and pine control soils and soils treated with spruce or pine needle fractions F1F4,
respectively, tannic acid or cellulose and the corresponding treatments with arginine addition. ANOVA was performed separately for the two soil-treatment
combinations. Signicant differences (P!0.05) between the means of one soil-treatment combination are marked with different letters. Mean and SE for three
replicates.
S. Kanerva et al. / Soil Biology & Biochemistry 38 (2006) 13641374 1369
indicating inhibition. In pine F4 and in spruce F1, however, Ac-
erg 700 was very high and Ac-erg 30 was very low; but the
variation was high. In spruce F2 and pine F1, Ac-erg 700 was
nearly the same as Ac-erg 30.
4. Discussion
The procedure used here to extract and fractionate tannins
was similar to that of Fierer et al. (2001) with balsam poplar.
Light fractions F1 and F2 from spruce and pine needles also
contained compounds other than tannins. In addition to the
analyzed compounds (Table 2), both spruce and pine F1
probably contained chemically neutral and also higher
molecular weight compounds like waxes, chlorophyll and
terpenoids, which are soluble in organic solvents. F2 probably
contained more polar compounds, such as phenols and
oligomeric phenols. Therefore, we cannot specify all the
effects of fractions F1 and F2 treatments as tannin effects, and
therefore we nd it more justied to discuss only effects of
light fractions.
Our results indicated mainly parallel effects of spruce- and
pine-needle fractions in spruce and pine soils, respectively; but
some differences, mostly in magnitude, were also detected. The
effects caused by the lighter fractions F1 and F2 were mainly
Fig. 3. Concentrations of DON after 6-week incubation for spruce and pine control soils and soils treated with spruce or pine needle fractions F1F4, respectively,
tannic acid or cellulose and the corresponding treatments with arginine addition. ANOVA was performed separately for the two soil-treatment combinations.
Signicant differences (P!0.05) between the means of one soil-treatment combination are marked with different letters. Mean and SE for three replicates.
Fig. 4. C in microbial biomass after 6-week incubation for spruce and pine control soils and soils treated with spruce or pine needle fractions F1F4, respectively,
tannic acid or cellulose and the corresponding treatments with arginine addition. ANOVA was performed separately for the two soil-treatment combinations.
Signicant differences (P!0.05) between the means of one soil-treatment combination are marked with different letters. Mean and SE for three replicates.
S. Kanerva et al. / Soil Biology & Biochemistry 38 (2006) 13641374 1370
contrary to those caused by the heavier fractions F3 and F4, as
was also observed Fierer et al. (2001) with Populus
balsamifera leaf fractions.
Based on the results for soil respiration, the lighter fractions
seemed to act as a C source for microbes, while the heavier
fractions were inhibitors. Results for the net N mineralization
also indicated the difference between light and heavy fractions;
in both soils net N mineralization was clearly lower with F1
and F2 than with F3 and F4. Effects of tannic acid and cellulose
resembled the effects of light fractions more than those of
heavy fractions.
One reason for inhibition of respiration by heavy fractions
may be the same as Schimel et al. (1996) suggested for balsam
poplar tannins: inhibition of exoenzyme activity or complexa-
tion of proteinaceous substrates. Bradley et al. (2000) also
expected that decreases in mineral N cycling would be a result
of the ability of Kalmia angustifolia and balsam r tannins to
bind and sequester organic sources of N. Moreover, Kumar and
Horigome (1986) showed that in black locust (Robinia pseudo
acacia) the protein-precipitating capacity and the percentage of
protein-precipitable phenolics increased with increasing degree
of polymerization of the tannin fractions, and Kraus et al.
(2003) reported that high molecular weight tannins precipitate
more protein than low molecular weight tannins do. In our
study, fractions F3 and F4 contained longer tannin polymers
than fractions F1 and F2 did (Table 1). Therefore it is likely,
that protein precipitation could have played an important role
in inhibiting respiration in soils treated with heavy fractions.
Most fractions seemed to slightly reduce DON, but heavy
fraction treatments slightly more than light fraction treatments.
This points to protein precipitation by the heavy fractions,
since probably most of protein-tannin complexes do not appear
in DON, due to their weak extractability. According to Northup
et al. (1995), the soil DON:mineral-N ratio correlates with soil
tannin concentration; but our soils did not generally show
increased DON:mineral-N ratios (results not shown) due to
addition of the needle fractions. However, it is possible that
added tannins do not affect soil in the same way as natural soil
tannins do.
Inhibition of soil respiration by heavy fractions may also
have been due to toxic effects. Since heavy-fraction treatments
with addition of arginine showed similar results for soil
respiration as treatments without added arginine, it is possible
that N was not a limiting factor in our soils, but more probably,
arginine addition was not large enough to overcome the
Fig. 5. N in microbial biomass after 6-week incubation for spruce and pine control soils and soils treated with spruce or pine needle fractions F1F4, respectively,
tannic acid or cellulose and the corresponding treatments with arginine addition. ANOVA was performed separately for the two soil-treatment combinations.
Signicant differences (P!0.05) between the means of one soil-treatment combination are marked with different letters. Mean and SE for three replicates.
Fig. 6. Relative availability (30 mg C l
K1
) and inhibition (700 mg C l
K1
) of
spruce and pine needle fractions F1F4 and tannic acid (TA) to bacteria (TdR)
and fungi (Ac-erg). Bars represent SE for three replicates.
S. Kanerva et al. / Soil Biology & Biochemistry 38 (2006) 13641374 1371
negative effects of the precipitation of organic N compounds or
that the inhibition mechanism of heavy fractions was other than
substrate complexation, e.g. toxicity. This conclusion is also
supported by C in microbial biomass, since in some cases C
mic
was slightly decreased by heavy fractions, although not
signicantly, which could indicate direct toxic effects of
tannins on the microbial community or decreased enzyme
activities (Kraus et al., 2004). Therefore slight increases in net
N mineralization by heavy fractions relative to the control were
likely a consequence of reduced microbial activity and N
uptake rather than the result of gross mineralization of N
becoming more effective. The rate of N mineralization may not
have decreased as much as the rate of N immobilization, which
would result in accumulation of mineral N in the soil.
Decreased net mineralization of N by light fractions does
not necessarily indicate a reduction in gross mineralization of
N. More probably, mineralized N was immobilizated by soil
microbes since soil respiration increased shortly after addition
of the fractions. This indicates that the compounds in those
fractions were easily metabolized, as has also been seen in
other studies with different plant species (Basaraba, 1964;
Schimel et al., 1996; Fierer et al., 2001; Castells et al., 2003;
Kraus et al., 2004). However, microbial biomass C and N were
not signicantly increased with additions of light fractions,
except only N
mic
with pine F1. Kraus et al. (2004) also found
no increase in the amounts of C and N in the microbial biomass
with additions of puried tannins from different plant species.
Fierer et al. (2001) observed an increased C-to-N ratio in the
microbial biomass due to addition of tannins from balsam
poplar, and especially additions of the lighter fractions. This
was also observed in our study, but only in spruce soil with F1,
and when arginine was added, in spruce soil with both light
fractions and in pine soil with F2 (results not shown). In spruce
soil with arginine the addition of tannic acid and cellulose
clearly increased the C-to-N ratio in the microbial biomass, but
in pine soil with added arginine increased it only slightly. In
general, while the trends with arginine were mainly similar to
those without it, in pine soil, arginine addition made the
differences in net N mineralization and DON concentrations
between light and heavy fractions more obvious, which
indicates that for determining the maximal effects of the
fractions, nitrogen addition was essential. However, better
understanding of the N ux in both soils could be obtained
using
15
N labelling.
In some studies, nitrication has been found to be inhibited
by phenolic compounds (Basaraba, 1964; Thibault et al., 1982;
Baldwin et al., 1983), but it is unclear, is the mechanism
specic or only dependent on the change in net ammonica-
tion. In this study, however, nitrate concentrations were always
very low, close to the limit of determination; therefore, it is not
relevant to make conclusions about net nitrication based on
them.
Some differences were found between spruce soil with
spruce-needle fractions and pine soil with pine-needle
fractions, but in both soils the main trends were similar.
Spruce needles contained more procyanidin units, while in
pine needles prodelphinidin units were dominant (Table 1).
This is in agreement with the results of Maie et al. (2003)
where the procyanidin/prodelphinidin ratio was 70:30 in
Norway spruce needles, and those of Kraus et al. (2004)
with bishop pine needles in which the percentage of
procyanidin monomer units versus prodelphinidin monomer
units was 22%. Procyanidins and prodelphinidins differ from
each other in terms of the hydroxylation pattern in the
B-ring, which in terms of reactivity is a critical factor for
condensed tannins (Kraus et al., 2004). Hernes et al. (2001)
reported that, compared to procyanidin units, prodelphinidin
units may be structurally less stable and thus more prone to
chemical transformation by abiotic processes. In addition,
different protein-binding capacity for the procyanidin-type
and the prodelphinidin-type condensed tannins have been
suggested to affect the total amount of extractable free
condensed tannins in forest soils (Hernes et al., 2001; Maie
et al., 2003).
When the control, tannic acid and cellulose treatments
were compared in order to ascertain the differences between
spruce and pine soils, spruce soil seemed to be somewhat
more active than the pine soil, because it showed higher
respiration rates than pine soil did. This difference was to be
expected because spruce soil from this particular stand has
already previously been shown to have higher rates of C
mineralization and net N mineralization than the soil under
pine (Smolander and Kitunen, 2002). No toxic effects of
tannins in light fractions were observed in either of the soils,
which was probably due to the addition of fractions only in
their natural soils where the microbes were adapted to use the
compounds contained as C substrates. On the other hand,
concentrations of condensed tannins found in the light
fractions were so low that it may not be possible to detect
their potential toxic effects.
At the concentration of 700 mg C l
K1
, none of the spruce or
pine needle fractions was clearly inhibitory to fungi. It is
possible, however, that heavy fractions (except pine F4) may
have been somewhat inhibitory to fungi because fungal growth
rate was lower at the concentration of 700 mg C l
K1
than at the
concentration of 30 mg C l
K1
. In contrast, at the concentration
of 700 mg C l
K1
, all fractions seemed to be inhibitory to
bacteria and at the concentration of 30 mg C l
K1
, available to
bacteria. These results did not clearly support the results of the
soil incubation experiment. The soil incubation experiment,
where fungi may out-compete the bacteria, may indicate more
fungal than bacterial activity. In addition, bacteria and fungi
may play different roles in mineralization of soil organic
matter; fungi and bacteria have been shown to be specialized to
grow on different types of phenolic compounds (Blum and
Shafer, 1988; Ley and Schmidt, 2002). It is also possible that
the concentration of 700 mg C l
K1
was not enough to inhibit
fungi, which are suggested to degrade the most refractory
compounds (Mller et al., 1999). These results must, however,
be interpreted and compared to those from the soil incubation
experiment with caution, because only the water-extractable
part of tannin fractions was studied and the application of the
used methods was novel.
S. Kanerva et al. / Soil Biology & Biochemistry 38 (2006) 13641374 1372
4.1. Conclusions
Spruce and pine needle fractions seem to play an important
role in controlling transformations of C and N in spruce and
pine soils. Light fractions, consisting of tannin mono- and
dimers and also many other compounds, were used as C
substrates, while heavy fractions, consisting mainly of the
polymerised condensed tannins, inhibited growth. However,
these results cannot be applied directly to natural forest soil,
since in nature, light and heavy fractions are introduced to soil
concurrently through litter and leachates, addition is more or
less continuous and it is not known whether in natural
conditions the effects of light fractions overcome the effects
of heavy fractions or vice versa. In the future, it would be
reasonable to nd out the differences between the effects of
spruce and pine needle fractions, and to explore whether
differences between spruce and pine soils can be explained by
those compounds.
Acknowledgements
We are grateful to Anneli Rautiainen and Pauli Karppinen
for excellent laboratory work, to Anne Siika for making the
gures and to Dr Joann von Weissenberg for checking the
English language of this paper. Many thanks to Prof Erland
Baath for discussions and comments on the results of
14
C-
acetate-in-ergosterol method. This study was supported by the
Academy of Finland.
References
Baath, E., 1992. Thymidine incorporation into macromolecules of bacteria
extracted from soil by homogenizationcentrifugation. Soil Biology &
Biochemistry 24, 11571165.
Baath, E., 2001. Estimation of fungal growth rates in soil using
14
C-acetate
incorporation into ergosterol. Soil Biology & Biochemistry 33, 20112018.
Baath, E., Pettersson, M., Soderberg, K.H., 2001. Adaptation of a rapid and
economical microcentrifugation method to measure thymidine and leucine
incorporation by soil bacteria. Soil Biology & Biochemistry 33,
15711574.
Bae, H.-D., McAllister, T.A., Muir, A.D., Yanke, L.J., Bassendowski, K.A.,
Cheng, K.-J., 1993. Selection of a method for condenced tannin analysis for
studies with rumen bacteria. Journal of Agricultural and Food Chemistry
41, 12561260.
Baldwin, I.T., Olson, R.K., Reiners, W.A., 1983. Protein binding phenolics and
the inhibition of nitrication in subalpine balsam r soils. Soil Biology &
Biochemistry 15, 419423.
Basaraba, J., 1964. Inuence of vegetable tannins on nitrication in soil. Plant
and Soil 21, 816.
Blum, U., Shafer, S.R., 1988. Microbial populations and phenolic acids in soil.
Soil Biology & Biochemistry 20, 793800.
Bradley, R.L., Titus, B.D., Preston, C.P., 2000. Changes to mineral N cycling
and microbial communities in black spruce humus after additions of
(NH
4
)
2
SO
4
and condensed tannins extracted from Kalmia angustifolia and
balsam r. Soil Biology & Biochemistry 32, 12271240.
Castells, E., Penuelas, J., Valentine, D.W., 2003. Inuence of the phenolic
compound bearing species Ledum palustre on soil N cycling in boreal
hardwood forest. Plant and Soil 251, 155166.
Chen, J., Stark, J.M., 2000. Plant species effects and carbon and nitrogen
cycling in a sagebrush-crested wheatgrass soil. Soil Biology &
Biochemistry 32, 4757.
Fierer, N., Schimel, J.P., Cates, R.G., Zou, J., 2001. Inuence of balsam poplar
tannin fractions on carbon and nitrogen dynamics in Alaskan taiga
oodplain soils. Soil Biology & Biochemistry 33, 18271839.
Hagerman, A.E., Butler, L.G., 1989. Choosing appropriate methods and
standards for assaying tannin. Journal of Chemical Ecology 15, 17951810.
Hernes, P.J., Benner, R., Cowie, G.L., Goni, M.A., Bergamaschi, B.A.,
Hedges, J.I., 2001. Tannin diagenesis in mangrove leaves from a tropical
estuary: a novel molecular approach. Geochimica et Cosmochimica Acta
65, 31093122.
Karonen, M., Loponen, J., Ossipov, V., Pihlaja, K., 2004. Analysis of
procyanidins in pine bark with reversed-phase and normal-phase high-
performance liquid chromatography-electrospray ionisation-mass spec-
trometry. Analytica Chimica Acta 522, 105112.
Kiikkila, O., Kitunen, V., Smolander A., 2005. Degradability of dissolved
organic carbon and nitrogen in relation to tree species. FEMS Microbiology
Ecology 53, 3340.
Kraus, T.E.C., Dahlgren, R.A., Zasoski, R.J., 2003. Tannins in nutrient
dynamics of forest ecosystemsa review. Plant and Soil 256, 4166.
Kraus, T.E.C., Zasoski, R.J., Dahlgren, R.A., Horwath, W.T., Preston, C.M.,
2004. Carbon and nitrogen dynamics in a forest soil amended with puried
tannins from different plant species. Soil Biology & Biochemistry 36,
309321.
Kuiters, A.T., Denneman, C.A.J., 1987. Water-soluble phenolic substances in
soils under several coniferous and deciduous tree species. Soil Biology &
Biochemistry 19, 765769.
Kumar, R., Horigome, T., 1986. Fractionation, characterization, and protein-
precipitating capacity of the condensed tannins fromRobinia pseudo acacia
L. leaves. Journal of Agricultural and Food Chemistry 34, 487489.
Ley, R.E., Schmidt, S.K., 2002. Fungal and bacterial responses to phenolic
compounds and amino acids in high altitude barren soils. Soil Biology &
Biochemistry 34, 989995.
Loponen, J., Lempa, K., Ossipov, V., Kozlov, M.V., Girs, A., Hangasmaa, K.,
Haukioja, E., Pihlaja, K., 2001. Patterns in content of phenolic compounds
in leaves of mountain birches along a strong pollution gradient. Chemo-
sphere 45, 291301.
Maie, N., Behrens, A., Knicker, H., Kogel-Knabner, I., 2003. Changes in the
structure and protein binding ability of condensed tannins during
decomposition of fresh needles and leaves. Soil Biology & Biochemistry
35, 577589.
Martikainen, P.J., Palojarvi, A., 1990. Evaluation of the fumigationextraction
method for the determination of microbial C and N in a range of forest soils.
Soil Biology & Biochemistry 22, 797802.
Mller, J., Miller, M., Kjller, A., 1999. Fungal-bacterial interaction on beech
leaves: inuence on decomposition and dissolved organic carbon quality.
Soil Biology & Biochemistry 31, 367374.
Newell, S.Y., Fallon, R.D., 1991. Towards a method for measuring
instantaneous fungal growth rates in eld samples. Ecology 72,
15471559.
Northup, R.R., Yu, Z., Dahlgren, R.A., Vogt, K.A., 1995. Polyphenol control of
nitrogen release from pine litter. Nature 377, 227229.
Ossipova, S., Ossipov, V., Haukioja, E., Loponen, J., Pihlaja, K., 2001.
Proanthocyanidins of mountain birch leaves: quantication and properties.
Phytochemical Analysis 12, 128133.
Porter, L.J., Hrstich, L.N., Chan, B.C., 1986. The conversion of procyanidins
and prodelphinidins to cyanidin and delphinidin. Phytochemistry 25,
223230.
Priha, O., Smolander, A., 1997. Microbial biomass and activity in soil and
litter under Pinus sylvestris, Picea abies and Betula pendula at
originally similar eld afforestation sites. Biology and Fertility of
Soils 24, 4551.
Priha, O., Smolander, A., 1999. Nitrogen transformations in soil under Pinus
sylvestris, Picea abies and Betula pendula at two forest sites. Soil Biology
& Biochemistry 31, 965977.
Schimel, J.P., Van Cleve, K., Cates, R.G., Clausen, T.P., Reichardt, P.B., 1996.
Effects of balsam poplar (Populus balsamifera) tannins and low molecular
weight phenolics on microbial activity in taiga oodplain soil: implications
for changes in N cycling during succession. Canadian Journal of Botany 74,
8490.
S. Kanerva et al. / Soil Biology & Biochemistry 38 (2006) 13641374 1373
Smolander, A., Kitunen, V., 2002. Soil microbial activities and characteristics
of dissolved organic C and N in relation to tree species. Soil Biology &
Biochemistry 34, 651660.
Smolander, A., Loponen, J., Suominen, K., Kitunen, V., 2005. Organic matter
characteristics and C and N transformations in the humus layer under two
tree species, Betula pendula and Picea abies. Soil Biology & Biochemistry
37, 13091318.
Terrill, T.H., Rowan, A.M., Douglas, G.B., Barry, T.N., 1992. Determination of
extractable and bound condensed tannin concentrations in forage plants,
protein concentrate meals and cereal grains. Journal of the Science of Food
and Agriculture 58, 321329.
Thibault, J.-R., Fortin, J.-A., Smirnoff, W.A., 1982. In vitro allelopathic
inhibition of nitrication by balsam poplar and balsam r. American
Journal of Botany 69, 676679.
S. Kanerva et al. / Soil Biology & Biochemistry 38 (2006) 13641374 1374