Page 1
Biogeosciences, 2, 159–174, 2005
www.biogeosciences.net/bg/2/159/
SRef-ID: 1726-4189/bg/2005-2-159
European Geosciences Union
Biogeosciences
Aggregate and soil organic carbon dynamics in
South Chilean Andisols
D. Huygens
1
, P. Boeckx
1
, O. Van Cleemput
1
, C. Oyarz ´un
2
, and R. Godoy
3
1
Laboratory of Applied Physical Chemistry – ISOFYS, Ghent University, Coupure 653, B-9000 Gent, Belgium
2
Institute of Geosciences, Universidad Austral de Chile, Casilla 567, Valdivia, Chile
3
Institute of Botany, Universidad Austral de Chile, Casilla 567, Valdivia, Chile
Received: 20 December 2004 – Published in Biogeosciences Discussions: 9 February 2005
Revised: 11 May 2005 – Accepted: 6 June 2005 – Published: 24 June 2005
Abstract. Extreme sensitivity of soil organic carbon (SOC)
to climate and land use change warrants further research
in different terrestrial ecosystems. The aim of this study
was to investigate the link between aggregate and SOC dy-
namics in a chronosequence of three different land uses
of a south Chilean Andisol: a second growth Nothofagus
obliqua forest (SGFOR), a grassland (GRASS) and a Pi-
nus radiata plantation (PINUS). Total carbon content of the
0–10 cm soil layer was higher for GRASS (6.7 kg C m
−
2
)
than for PINUS (4.3 kg C m
−
2
)
, while TC content of SG-
FOR (5.8 kg C m
−
2
)
was not significantly different from ei-
ther one. High extractable oxalate and pyrophosphate Al
concentrations (varying from 20.3–24.4 g kg
−
1
, and 3.9–
11.1 g kg
−
1
, respectively) were found in all sites. In this
study, SOC and aggregate dynamics were studied using size
and density fractionation experiments of the SOC,
δ
13
C and
total carbon analysis of the different SOC fractions, and C
mineralization experiments. The results showed that elec-
trostatic sorption between and among amorphous Al com-
ponents and clay minerals is mainly responsible for the for-
mation of metal-humus-clay complexes and the stabiliza-
tion of soil aggregates.
The process of ligand exchange
between SOC and Al would be of minor importance re-
sulting in the absence of aggregate hierarchy in this soil
type. Whole soil C mineralization rate constants were high-
est for SGFOR and PINUS, followed by GRASS (respec-
tively 0.495, 0.266 and 0.196 g CO
2
-C m
−
2
d
−
1
for the top
soil layer). In contrast, incubation experiments of isolated
macro organic matter fractions gave opposite results, show-
ing that the recalcitrance of the SOC decreased in another
order: PINUS
>
SGFOR
>
GRASS. We deduced that electro-
static sorption processes and physical protection of SOC in
soil aggregates were the main processes determining SOC
stabilization. As a result, high aggregate carbon concentra-
Correspondence to: D. Huygens
(dries.huygens@ugent.be)
tions, varying from 148 till 48 g kg
−
1
, were encountered for
all land use sites. Al availability and electrostatic charges are
dependent on pH, resulting in an important influence of soil
pH on aggregate stability. Recalcitrance of the SOC did not
appear to largely affect SOC stabilization. Statistical correla-
tions between extractable amorphous Al contents, aggregate
stability and C mineralization rate constants were encoun-
tered, supporting this hypothesis. Land use changes affected
SOC dynamics and aggregate stability by modifying soil pH
(and thus electrostatic charges and available Al content), root
SOC input and management practices (such as ploughing and
accompanying drying of the soil).
1
Introduction
Andisols develop from volcanic ash and are characterized
by a high soil organic matter (SOM) content, good physical
properties, and a high potential anion- or cation exchange ca-
pacity (AEC and CEC) (Galindo and Bingham, 1977). Their
clay fraction consists mainly of poorly crystalline allophane,
imogolite, and oxides and oxyhydroxides of Fe and Al (Be-
soain, 1958). Allophane is a group of clay minerals made
up of silica, aluminum, and water in a chemical combina-
tion (Parfitt, 1990). In southern Chile (35
◦
–49
◦
S), Andisols
cover a large part of the surface area of the Central Valley
and Andes mountain ranges (FAO-UNESCO, 1971).
Temperate native forests covered most of southern Chile
for the last 3,000 years, remaining almost intact until the
arrival of the Europeans, 450 years ago (Lara et al., 1999).
Significant land use changes have taken place in this region
(Lara et al., 1999). Today, only 56% (or 10.3 million ha) of
this original forest cover remains. Most affected are the Scle-
rophyllus and the Nothofagus forest types, with 3 and 30%
of the original forest cover remaining, respectively (Lara
et al., 1999).
On the other hand, grasslands and shrub-
lands have doubled in area, from 2.5 to 5.5 million ha (Lara
© 2005 Author(s). This work is licensed under a Creative Commons License.
Page 2
160
D. Huygens et al.: Aggregate and soil organic carbon dynamics in South Chilean Andisols
et al., 1999).
New land use types, such as urban areas
(55 000 ha), arable lands (2.8 million ha) and forestry planta-
tions (2.1 million ha), now occupy a significant area as well.
General overviews of changes in soil organic carbon
(SOC) stocks after land use conversions are given by Guo
and Gifford (2002) and Paul et al. (2002). Guo and Gifford
(2002) revised changes in SOC stocks as a result of land use
conversion, using data from 74 publications. They indicated
that SOC stocks increase after land use change from native
forest to pasture (+8%, based on 170 observations), but de-
clined after pasture conversion to plantation (
−
10%, based
on 83 observations). However, rainfall has a strong influence
on changes in SOC stocks. It was concluded that in areas re-
ceiving 1500–2000 mm of rainfall per year a conversion from
forest to pasture did not result in changes in SOC stock. A
land use change from grassland to plantation resulted in a sig-
nificant decrease in SOC stock with
−
23%. Paul et al. (2002)
concluded that afforestation decreased soil stocks by 3.5%
per year during the first five years of afforestation. Later
on, a decreasing rate of decline or eventually a recovery of
SOC would occur. De Koning et al. (2003) concluded more
specific that in Al-rich soils, changes in SOC stocks follow
a pattern according to land use history and accompanying
changes in soil characteristics.
Soil organic matter has a very complex and heterogeneous
composition and is generally mixed or associated with the
mineral soil constituents to form soil aggregates (Del Galdo
et al., 2003). SOC dynamics are mainly influenced by its re-
calcitrance and accessibility, and interactions between SOC
and soil components (Sollins et al., 1996). Recalcitrance is
the inherent molecular level resistance of a molecule to mi-
crobial and enzymatic breakdown. Accessibility is the extent
to which the location of substrates controls access by mi-
crobes, fungi or enzymes. Interactions of organic substrates
with organic or inorganic molecules may alter the degrada-
tion rate of the substrates (Swanston et al., 2002). Several
studies have shown that in Al rich soils, such as Andisols,
chemical stabilization is the determining process for carbon
accrual (Zunino et al., 1982; Percival et al., 2000). However,
the exact role of SOC and Al in the process of chemical sta-
bilization, and which kind of interactions occurs between the
different soil components is not yet fully understood.
The objective of this study was to evaluate the role of Al
on SOC dynamics in Chilean Andisols and to investigate
SOC stabilization by electrostatic Al-C binding, and occlu-
sion within soil aggregates. Via a case study, we tried to
assess the effect of changes in certain physical and chem-
ical soil characteristics – which resulted from the land use
conversions – on SOC, physical protection of SOC in aggre-
gates, quality of SOC, and organo-mineral associations. A
chronosequence of three different land uses was studied: a
second growth Nothofagus obliqua forest (SGFOR), a grass-
land (GRASS), and a plantation of Pinus radiata (PINUS).
2
Materials and methods
2.1
Site description and soil sampling
Soil samples were taken in August 2002 near Paillaco, in
the Central Valley of southern Chile (40
◦
07 S, 72
◦
51 W). The
mean annual precipitation varies from 1600–1800 mm, while
mean annual temperature is 10.8
◦
C. All soils were classified
as trumao’s or Andisols, Typic Dysandrepts (Tosso, 1985).
They have typical variable electrical charge characteristics
(i.e. pH dependent) as a result of their particular clay mineral
and humus content (Radcliffe and Gillman, 1985; Nanzyo
et al., 1993). A mixture of 1:1 and 2:1 clay minerals (e.g.
halloysite, vermiculite, etc.) has been documented in Chilean
Andisols as well as the generally observed poorly crystalline
minerals (Besoain, 1958, 1969; Galindo and Bingham, 1977;
Besoain et al., 2000).
The three sites are all located in close proximity (100 m
apart) and were originally all under the same forest vegeta-
tion dominated by Nothofagus obliqua. Nowadays, only the
SGFOR site is still under this original vegetation. The land
use of the GRASS and PINUS sites changed 50–100 years
ago from second growth N. obliqua forest to grassland for
agricultural livestock activity. Finally, 16 years ago, a Pinus
radiata plantation was established on part of the GRASS site.
The second growth N. obliqua forest (SGFOR) has a mean
tree height of 35 m, a stand density of 757 ha
−
1
and a canopy
cover of 70% during the growing season. The average tree
age is 120 years, while mean diameter at breast height equals
42.5 cm. In this forest, the overstorey is dominated by N.
obliqua (32 m) with a cover of 40%. The understorey (25 m)
consists of Aextoxicon punctatum, Persea linque, Nothofa-
gus obliqua, Eucriphia cordifolia, Laurelia sempervirens,
Nothofagus dombeyi and Gevuina avellana with a cover of
70%. The lower understorey (10 m) is dominated by Aextox-
icon punctatum. In the shrub stratum (
<
4 m) Chusquea quila
and Raphithamnus spinosus are dominant.
Since the establishment of the pasture, agricultural activ-
ities (cattle breeding) have taken place on the GRASS and
PINUS sites with an animal density of 1.5–2 cows ha
−
1
. The
grassland has a species composition consisting of Lolium
perenne and Trifolium repens, and is ploughed and resown
every two years. A mixture of various forms of acidifying
fertilizers has been applied yearly on the GRASS site at high
levels, though no data were available on the exact fertilizer
type or application rate.
On the PINUS site, fertilization was stopped with the es-
tablishment of the plantation, 16 years ago. This P. radiata
monoculture has a mean tree height of 15–18 m, while stand
density is around 1500 ha
−
1
.
Sampling was done after selection of representative plots
in the three ecosystem sites (land use treatments). In each
land use treatment, three sampling plots were selected based
on pedological, forestry and ecological experience, and were
located approximately 20 m apart from each other. Samples
Biogeosciences, 2, 159–174, 2005
www.biogeosciences.net/bg/2/159/
Page 3
D. Huygens et al.: Aggregate and soil organic carbon dynamics in South Chilean Andisols
161
Soil
Wet sieving
250
µ
m
150
µ
m
50
µ
m
Decantation
in water
Mineral fraction (150 µm-2000 µm)
Macro-organic matter
Size fraction <50
µ
m
Size fraction 50
µ
m -150
µ
m
Density separation
in Ludox
*
LF
IF
HF
*
Ludox : silica suspensions with different density
Fig. 1. Scheme of the density fractionation of SOC; LF=light density fraction (density
<
1.13 Mg m
−
3
), IF=intermediate density fraction
(1.13 Mg m
−
3
<
density
<
1.37 Mg m
−
3
), HF=heavy density fraction (density
>
1.37 Mg m
−
3
) (adapted from Accoe et al., 2002).
originating from a different sampling plot in the same land
use treatment were considered as replicates, and were kept
separate for later analysis. In each sampling plot, one soil
sample has been taken at 0–10, 10–20 and 20–30 cm depth
from a soil pit. Thus, in total 27 soil samples were taken.
After transport to the laboratory, the recognizable plant ma-
terial was removed. All soil samples were dried, homoge-
nized and sieved (2 mm) for later analyses. Soil samples were
dried in open air to a gravimetric water content of approxi-
mately 20%. Volcanic-ash soils can granulate irreversibly
when completely dried (Kubota, 1972), and this can affect
the subsequent fractionation of SOC. We will refer to these
partly dried samples using the term “air dried samples”.
Bulk densities were determined by pushing steel cylinders
with a known volume into the undisturbed soil (Klute, 1986).
After transport to the laboratory, the cylinders were dried at
105
◦
C during 24 h, after which its weight was determined.
Three replicates per land use treatment were performed. Soil
texture was determined in the field according to a method
described by Milford (1997).
2.2
Size and density fractionation of SOC
The size fractionation procedure was performed by sepa-
rately wet sieving a soil sample of 100 g over three sieves
(250
µ
m, 150
µ
m and 50
µ
m) with tap water. The suspen-
sion passing the bottom size was collected and left about
48 h at room temperature to settle. After settling, the clear
solution was removed and the size fraction
<
50
µ
m was
collected. Four size fractions (2000–250
µ
m, 250–150
µ
m,
150–50
µ
m and
<
50
µ
m) were obtained, dried at 40
◦
C for
48 h and homogenized using a planetary ball mill (PM400,
Retsch, Germany) for subsequent analysis.
To evaluate the soil aggregate stability or susceptibility of
the aggregates against disruption processes, we used a mod-
ification of the procedure described by De Leenheer and De
Boodt (1959). The size fractionation was performed at two
different times, using the same procedure: (1) starting from
“air dried” soil (slaking) and (2) after an 80 day laboratory
incubation experiment (wet sieving of field moist soil) (see
below). In the incubation experiment, the soil was exposed to
several supplementary disruption processes such as rewetting
and mixing of the soils. The distribution of the soil weight
among the different size fractions was expressed using the
weighted average diameter (WAD) and could thus be quan-
tified as one single parameter for further statistical analyses.
The difference in WAD between the two fractionation exper-
iments can be related to the stability of the soil aggregates
(De Leenheer and De Boodt, 1959). A smaller difference in
WAD indicates a higher stability of the soil aggregates.
Density fractionation of the SOC was performed using the
method of Meijboom et al. (1995) (Fig. 1). Air dried soil
samples (400 g) were wet sieved over two sieves with tap
water (top sieve mesh size 250
µ
m, bottom sieve mesh size
150
µ
m), until the passing water became clear in order to de-
stroy aggregates
>
150
µ
m. Only a water jet was applied as
dispersive energy in order to destroy soil aggregates. Next,
www.biogeosciences.net/bg/2/159/
Biogeosciences, 2, 159–174, 2005
Page 4
162
D. Huygens et al.: Aggregate and soil organic carbon dynamics in South Chilean Andisols
the two size fractions retained on the sieves were washed
into a bucket and swirled with a jet of water to separate
the organic matter (OM) from the mineral fraction by de-
cantation. Swirling and decanting were repeated until no
more floating organic matter appeared. The obtained macro
organic matter (MOM) from both size fractions (250–2000
and 150–250
µ
m) was recombined to be further separated
into a light (LF, density
<
1.13 Mg m
−
3
)
, intermediate (IF,
1.13 Mg m
−
3
<
density
<
1.37 Mg m
−
3
)
and heavy (HF, den-
sity
>
1.37 Mg m
−
3
)
fraction by subsequent submersion in
two colloidal silica suspensions (Ludox, Dupont, USA) with
a respective density of 1.13 Mg m
−
3
and 1.37 Mg m
−
3
. The
obtained density fractions were washed with distilled water
and dried at 40
◦
C for 48 h.
The C concentrations increase with decreasing density of
the fractions (LF
>
IF
>
HF). The IF and HF show a higher
association of the SOC of these heavier fractions with soil
minerals (Hassink, 1995). The LF consists mainly of recog-
nizable plant material in an early stage of decomposition and
serves as readily decomposable substrate for the microbial
biomass (Six et al., 1999; Solomon et al., 2000).
2.3
C mineralization experiments
The C mineralization experiments consisted out of two dif-
ferent experiments. In a first experiment, whole soil C min-
eralization was measured. In a second experiment, isolated
macro organic matter fractions (LF and HF) were incubated
to determine the specific C mineralization of these fractions.
Before starting the whole soil C mineralization experi-
ment, the air dried soil was rewetted and pre-incubated dur-
ing six days at a temperature of 15
◦
C. Air dried bulk sam-
ples were brought to a gravimetric water content of 46% for
SGFOR and PINUS, respectively, and 32% for GRASS, cor-
responding to a water-filled pore space of 25% at the bulk
density in the field (using a bulk density for all the layers
of 0.45 Mg m
−
3
for SGFOR and PINUS, and 0.60 Mg m
−
3
for GRASS (see Table 1) and a soil particle density of
2.65 Mg m
−
3
)
. As such, a realistic microbial soil activity
could be restored. An exact quantity of soil was put in PVC
tubes (4.6 cm diameter, 13 cm high) and compressed manu-
ally to the corresponding bulk density. The tubes were cov-
ered with parafilm, which permits sufficient gas exchange
and aeration, but reduced moisture loss.
During the en-
tire experiment, the moisture content and temperature were
kept constant. Carbon dioxide emission was measured on
eight different time points (day 1, 3, 7, 11, 17, 35, 43 and
62). This was done by putting the PVC-tubes into glass jars,
which were closed airtight. The headspace of the glass jars
was sampled after 0, 27, 54 and 81 minutes through a rub-
ber septum. Before sampling, the CO
2
concentration in the
headspace was homogenized by repeatedly sucking and re-
injecting the headspace. A gas sample of 1 mL was injected
into a gas chromatograph (GC-14B, Shimadzu, Japan) (Pora-
pack Q, mesh size 80/100 column specification, electron cap-
ture detector (ECD)). The CO
2
concentration was calculated
via a linear calibration using a certified CO
2
concentration
(1% CO
2
in dry air, Air Liquide, France). Linear regression
between the measuring points at 0, 27, 54 and 81 min gave
the CO
2
mineralization flux.
The incubation of the isolated HF and LF was only per-
formed for SOC originating from the 0–10 cm soil layer of
each land use treatment, following the method described by
Swanston et al. (2002). For LF, 2.00 g substrate was mixed
with 18.00 g inert quartz sand (acid washed, 580–650
µ
m di-
ameter) in a glass flask. For HF, 5.00 g substrate was mixed
with 15.00 g inert quartz sand. A microbial inoculum was
added at the beginning of the incubation. A water extract
from the mineral soil originating from the same soil layer
was used as inoculum source. The inoculum was made by
adding distilled water to a wet soil in a 1/1.5 ratio, followed
by placing the mixture on a rotary shaker for 30 min. Af-
ter settling down, 4 mL of the supernatant was added to each
flask and the substrate was pre-incubated for two days. Dur-
ing the entire experiment, the inoculated flasks were stored in
an incubation room at a constant temperature of 15
◦
C. The
CO
2
concentration was measured through a septum in the cap
of the flasks after 0, 27, 54 and 81 min and on six different
time points (day 1, 3, 7, 11, 17 and 35) with a gas chromato-
graph (GC-14B, Shimadzu, Japan), using the same method
as described above.
2.4
Physico-chemical analyses
Measurement of
13
C and total C (TC) content in the different
soils and SOC fractions was performed using an elemental
analyser (ANCA-SL, PDZ-Europa, UK) coupled to an Iso-
tope Ratio Mass Spectrometer (20-20, PDZ Europa, UK).
The measured
13
C/
12
C ratios are expressed as
δ
13
C values
(‰) vs. the international Vienna Pee Dee Belemnite (VPDB)
standard:
δ
13
C=
R
sample
−
R
standard
R
standard
×
1000
(
‰
)
(1)
R
sample
and R
standard
refer to the
13
C/
12
C ratio of the mea-
sured sample and the standard, respectively. The working
standard was flour with a
δ
13
C value of
−
27.01
±
0.04‰ vs.
VPDB and TC of 39.95
±
0.40‰ (certified by Iso-Analytical,
UK).
A modified procedure of the method described by Schw-
ertmann (1964), using a mixture of ammonium oxalate
(0.2 M) and oxalic acid as extractant, was applied for the
determination of the extractable aluminum oxalate (Al
ox
),
and silicon oxalate (Si
ox
) content. In the absence of light,
a soil:solution (1:10) mixture was placed on a rotary shaker
for 2 hours. A pyrophosphate reagent (0.1 M) was used in
soil:solution ratio of 1:10 (shaking time 16 h) for the deter-
mination of the pyrophosphate extractable Al content (Al
py
).
All supernatants were centrifuged at 3000 rpm for 10 min,
filtrated (folded filters, Schleicher & Sch¨uell, Germany) and
Biogeosciences, 2, 159–174, 2005
www.biogeosciences.net/bg/2/159/
Page 5
D. Huygens et al.: Aggregate and soil organic carbon dynamics in South Chilean Andisols
163
Table 1. Chemical and physical properties of the second growth forest (SGFOR), grassland (GRASS) and plantation (PINUS) for the 0–10,
10–20 and 20–30 cm soil layer (Standard deviations between brackets, significant differences (
p<
0.05) between land use treatments within
one soil depth layer are indicated by different letters).
Soil layer
Land use class
Texture
pH
H
2
O
ρ
b
Al
ox
Al
py
Si
ox
Allophane
TC
TN
C/N
cm
Mg m
−
3
—
g kg
−
1
—
%
kg C m
−
2
g kg
−
1
kg N m
−
2
g kg
−
1
0–10
SGFOR
sandy loam
6.1
a
0.42
20.3
b
6.4
b
9.6
6.00
a
5.8
a,b
139
a
0.31
b
7.4
b
18.9
(0.1)
(0.06)
(0.5)
(0.9)
(0.4)
(0.33)
(0.5)
(7)
(0.05)
(0.2)
(1.2)
GRASS
sandy loam
5.7
b
0.60
23.6
a
11.1
a
8.1
5.22
b
6.7
a
111
b
0.53
a
8.9
a
12.5
(0.1)
(0.09)
(0.5)
(0.9)
(0.6)
(0.28)
(1.1)
(1)
(0.08)
(0.1)
(4.5)
PINUS
sandy loam
6.2
a
0.45
22.9
a
6.3
b
9.1
6.46
a
4.3
b
94
c
0.31
b
6.8
b
13.8
(0.0)
(0.07)
(0.4)
(0.8)
(0.2)
(0.21)
(0.9)
(6)
(0.06)
(0.6)
(5.0)
10–20
SGFOR
sandy loam
6.3
a
0.44
21.2
b
5.4
b
11.4
6.99
a
4.0
b
93
b
0.23
b
5.5
b
16.9
(0.0)
(0.12)
(0.6)
(0.7)
(0.6)
(0.12)
(0.9)
(6)
(0.02)
(1.2)
(0.6)
GRASS
sandy loam
5.4
b
0.46
24.1
a
10.5
a
8.5
5.57
b
4.8
b
105
a
0.37
a
8.1
a
13.0
(0.1)
(0.10)
(0.2)
(1.0)
(0.3)
(0.27)
(1.0)
(5)
(0.08)
(0.1)
(0.8)
PINUS
sandy loam
6.4
a
0.46
23.4
a
6.1
b
9.1
6.63
a
3.3
a
71
c
0.26
b
5.6
b
12.7
(0.1)
(0.04)
(0.6)
(0.6)
(0.6)
(0.47)
(0.3)
(3)
(0.03)
(0.2)
(1.5)
20–30
SGFOR
sandy loam
6.3
a
0.52
22.3
b
4.3
b
11.8
7.59
a
3.4
b
64
b
0.21
b
4.0
b
16.1
(0.1)
(0.05)
(0.5)
(0.7)
(0.8)
(0.28)
(0.3)
(9)
(0.05)
(1.0)
(2.2)
GRASS
sandy loam
5.8
b
0.60
24.4
a
10.7
a
9.0
5.75
b
5.8
a
98
a
0.44
a
7.3
a
13.4
(0.3)
(0.08)
(0.3)
(0.6)
(0.1)
(0.05)
(1.0)
(5)
(0.10)
(0.8)
(1.1)
PINUS
sandy loam
6.7
a
0.47
22.2
b
3.9
b
10.2
7.19
a
2.4
b
51
b
0.18
b
3.9
b
13.1
(0.1)
(0.05)
(0.4)
(1.0)
(0.3)
(0.38)
(0.4)
(7)
(0.03)
(0.6)
(1.0)
ρ
b
=Bulk density
Al
ox
=Ammonium oxalate extractable Al content
Al
py
=Pyrophosphate extractable Al content
Si
ox
=Ammonium oxalate extractable Si content
Allophane=Estimated allophane content
TC=Total carbon content
TN=Total nitrogen content
the Al and Si content of the solutions were determined with
respectively ICP-OES (Inductive Coupled Plasma – Optical
Emission Spectrometry) (VISTA-MPX CCD Simultaneous
ICP-OES, Varian, USA), and AAS (SpectrAA 600, Varian,
USA). Allophane contents were calculated using the formula
proposed by Parfitt and Wilson (1985), and later simplified
by Mizota and van Reeuwijk (1989):
Allophane=100Si
ox
×
23
.
4
−
5
.
1
Al
ox
−
Al
py
Si
ox
(
%
)
(2)
2.5
Statistical analyses
Statistical analyses were performed using SYSTAT version
10.0 (SPSS, USA). Statistical differences of means (95% sig-
nificance level) were distinguished using ANOVA (3 repli-
cates) followed by a Duncan’s multiple range post-hoc test.
Paired samples T-tests were used to determine differences
in C mineralization rates between LF and HF fractions.
Log transformation was used to improve normality and ho-
moscedastisticy in the data when necessary. In case no ho-
mogeneity of variance was reached after log transformation,
Tamhane’s T2 test was used as post-hoc test. Correlations
were determined by the Pearson’s correlation factor.
3
Results
3.1
Physico- chemical soil properties
The chemical and physical properties of the soils are listed in
Table 1. Averaged over the three soil layers, land use changes
significantly affected SOC stocks with +31 and
−
42% for the
conversion from SGFOR to GRASS and GRASS to PINUS,
respectively. The N content was highest in GRASS, followed
by SGFOR and PINUS. Nitrogen stocks varied from 0.53 to
0.18 kg m
−
2
. Fertilizer additions on the GRASS site are re-
sponsible for the higher N contents in this site. The bulk
www.biogeosciences.net/bg/2/159/
Biogeosciences, 2, 159–174, 2005
Page 6
164
D. Huygens et al.: Aggregate and soil organic carbon dynamics in South Chilean Andisols
Fig. 2. Distribution of the soil weight over the different size classes (2000–250
µ
m, 250–150
µ
m, 150–50
µ
m, and
<
50
µ
m) after the first
fractionation procedure (starting from “air-dried” soil) for the second growth forest (SGFOR), grassland (GRASS) and plantation (PINUS)
and for the three soil layers (0–10 cm, 10–20 cm, and 20–30 cm).
Fig. 3. Distribution of the soil weight over the different size classes (2000–250
µ
m, 250–150
µ
m, 150–50
µ
m, and
<
50
µ
m) after the second
fractionation procedure (starting from field moist soil) for the second growth forest (SGFOR), grassland (GRASS) and plantation (PINUS)
and for the three soil layers (0—10 cm, 10–20 cm, and 20–30 cm).
density was higher in GRASS than in SGFOR and PINUS
and is as well a result of agricultural livestock activities.
Soil texture was categorized as sandy loam (USDA texture
class) for all land use treatments. The soils in SGFOR and
PINUS showed an average pH
H
2
O
between 6.1 and 6.7. In
contrast, GRASS showed a lower pH
H
2
O
(around 5.5). The
extractable Al
ox
and Al
py
concentrations in the GRASS site
were higher than for SGFOR and PINUS, though differences
between GRASS and PINUS were not always significant for
Al
ox
. Soil pH was negatively correlated with Al
ox
(
r
=
−
0.45;
p
=0.02), and Al
py
(
r
=
−
0.83;
p
=0.00). Allophane contents
were, for all three soil layers, significantly lower for GRASS
compared to SGFOR and PINUS.
Biogeosciences, 2, 159–174, 2005
www.biogeosciences.net/bg/2/159/
Page 7
D. Huygens et al.: Aggregate and soil organic carbon dynamics in South Chilean Andisols
165
Table 2. Distribution of the soil weight according to the size fraction in Fig. 2 and Fig. 3, expressed as weighted average diameter (WAD), of
the first and second fractionation procedure, and difference in WAD between the two fractionation procedures for the three soil layers of the
second growth forest (SGFOR), grassland (GRASS), and plantation (PINUS) (Standard deviations between brackets, significant differences
(
p<
0.05) between land use treatments are indicated by different letters).
0–10 cm
10–20 cm
20–30 cm
SGFOR
GRASS
PINUS
SGFOR
GRASS
PINUS
SGFOR
GRASS
PINUS
WAD first fractionation procedure (
µ
m)
786
A
788
A
790
A
794
A
760
A
758
A
739
A
785
A
647
B
(40)
(26)
(13)
(15)
(27)
(50)
(19)
(14)
(34)
WAD second fractionation procedure (
µ
m)
511
A
662
A
572
A
522
A
590
A
585
A
565
B
720
A
526
B
(118)
(59)
(124)
(84)
(73)
(54)
(33)
(53)
(5)
Difference in WAD (
µ
m)
275
A
126
A
218
A
273
A
170
A
173
A
174
A
65
B
120
A,B
(84)
(84)
(125)
(90)
(97)
(103)
(25)
(43)
(34)
3.2
Size and density fractionation of SOC
The highest weight percentage after the first size fractiona-
tion was found in the size fraction 250–2000
µ
m (Fig. 2).
The other size fractions contain smaller amounts of soil
weight, varying from 23.1 to 7.0% for all land use treatments.
The WAD was not significantly different between the differ-
ent land use treatments, except for PINUS 20–30 cm where
a significant lower value was found (Table 2). TC concentra-
tion (%) was not significantly different among the different
size fractions in all the sites, with exception of the 0–10 cm
layer of SGFOR (Table 3).
The results of the weight distribution of the second size
fractionation experiment (after an 80 day incubation period,
starting from field moist soil) are shown in Fig. 3. A con-
siderable shift in soil weight towards the finer size fractions
could be noticed, resulting in a smaller WAD compared to
the first fractionation procedure (Table 2). Between land use
treatments, GRASS showed the coarsest aggregate distribu-
tion but due to relative large standard deviations (on average
50
µ
m), differences in WAD were only significant for the 20–
30 cm layer. As in the first fractionation procedure, the TC
concentration did not differ significantly among the different
size classes for all land uses, though a few exceptions were
observed here (data not shown).
Averaged
over
the
three
soil
layers,
the
aggre-
gate
stability
decreased
in
the
following
order:
GRASS
>
PINUS
>
SGFOR.
However,
only
significant
differences were observed for the 20–30 cm layer between
SGFOR and GRASS (Table 2). The difference in WAD was
negatively correlated with the extractable Al
ox
content of
the soil (
r
=
−
0.50;
p
=0.01), but not with the TC content
(
r
=
−
0.18;
p
=0.38), extractable Al
py
content (
r
=
−
0.23;
p
=0.25) or allophane content (
r
=0.11;
p
=0.60) of the soil.
The relation between extractable Al
ox
content and difference
in WAD for the different land use treatments and soil depths
is given in Fig. 4.
y = -0.0152x + 25.401
R
2
= 0.25
19
20
21
22
23
24
25
0
50
100
150
200
250
300
350
400
difference in weighted average diameter (µm)
A
l
o
x
c
o
n
c
e
n
tr
a
ti
o
n
(g
k
g
-1
)
Fig. 4. The Al
ox
concentration in function of the difference in
weighted average diameter for the different land use treatments and
soil depths (error bars indicate plus/minus one standard deviation).
The amounts of MOM are expressed relative to the TC
content (i.e. as % of total C) (Table 4). The differences be-
tween land use treatments diminished with depth and were
only significant between SGFOR, GRASS and PINUS for
the 0–10 cm and 10–20 cm layer.
Table 5 shows the complete mass and SOC balance for
the 0–10 cm. It was found that the C content of the MOM
is only a small fraction of the TC content (2.4%–8.3%) in
the soil. Though the concept of the fractionation procedure
assumes that aggregates larger than 150
µ
m are broken and
the enclosed organic matter is released, still 56–74% of the
total SOC content of the soil was found in the remaining min-
eral fraction (i.e. the fraction 150–2000
µ
m, from which the
MOM is isolated).
3.3
δ
13
C signature in size and density fractions
The
δ
13
C values of the whole SOC and the different size
fractions in the three land use treatments and soil layers are
presented in Table 6. Within the different size fractions of a
www.biogeosciences.net/bg/2/159/
Biogeosciences, 2, 159–174, 2005
Page 8
166
D. Huygens et al.: Aggregate and soil organic carbon dynamics in South Chilean Andisols
Table 3. Carbon concentrations (g kg
−
1
)
for the soil layers of the second growth forest (SGFOR), grassland (GRASS), and plantation
(PINUS) (Standard deviations between brackets, significant differences (
p<
0.05) between aggregate size classes for a certain land use
treatment and soil depth are indicated by different letters).
Carbon concentration (g kg
−
1
)
0–10 cm
10–20 cm
20–30 cm
SGFOR
GRASS
PINUS
SGFOR
GRASS
PINUS
SGFOR
GRASS
PINUS
250
µ
m–2000
µ
m
141
a,b
111
a
93
a
90
a
105
a
71
a
66
a
98
a
52
a
(16)
(13)
(13)
(6)
(5)
(15)
(6)
(9)
(8)
150
µ
m–250
µ
m
146
a
111
a
98
a
88
a
106
a
69
a
63
a
99
a
50
a
(7)
(17)
(12)
(6)
(1)
(13)
(11)
(7)
(7)
50
µ
m–150
µ
m
148
a
107
a
99
a
93
a
105
a
72
a
67
a
97
a
52
a
(19)
(19)
(14)
(3)
(4)
(10)
(10)
(6)
(7)
<
50
µ
m
122
b
110
a
84
a
81
a
106
a
68
a
61
a
96
a
48
a
(5)
(15)
(12)
(2)
(2)
(6)
(8)
(4)
(6)
Table 4. Percentage of total C of the macro organic matter fractions (MOM) for the second growth forest (SGFOR), grassland (GRASS)
and plantation (PINUS) (Standard deviations between brackets, significant differences between land use treatments in the same soil layer are
indicated by different letters).
Soil layer (cm)
% of total C
SGFOR
GRASS
PINUS
0–10
Macro organic matter
8.3
A
(1.9)
2.4
B
(0.5)
5.0
C
(0.8)
HF
0.8
A
(0.1)
0.4
B
(0.0)
1.1
A
(0.3)
IF
1.0
A
(0.1)
0.3
B
(0.0)
0.8
C
(0.1)
LF
6.5
A
(1.9)
1.7
B
(0.5)
3.1
B
(0.7)
10–20
Macro organic matter
2.8
A
(0.6)
1.6
B
(0.1)
3.6
C
(0.2)
HF
0.3
A
(0.1)
0.3
A
(0.0)
0.8
B
(0.1)
IF
0.4
A
(0.2)
0.2
A
(0.1)
0.7
B
(0.2)
LF
2.1
A
(0.3)
1.1
B
(0.1)
2.1
A
(0.3)
20–30
Macro organic matter
3.8
A
(1.4)
1.1
B
(0.1)
3.4
A
(0.5)
HF
0.4
A
(0.2)
0.3
A
(0.0)
1.0
B
(0.2)
IF
0.7
A
(0.3)
0.1
B
(0.1)
0.8
A
(0.2)
LF
2.7
A
(0.9)
0.6
B
(0.1)
1.6
A,B
(0.4)
HF=Heavy density fraction (density
>
1.37 Mg m
−
3
)
IF=Intermediate density fraction (1.13 Mg m
−
3
<
density
<
1.37 Mg m
−
3
)
LF=Light density fraction (density
<
1.13 Mg m
−
3
)
certain land use, no significant differences in
δ
13
C value were
found. However, a significant difference of 1–2‰ was found
between the
δ
13
C value of the size classes and MOM density
classes for the different land use treatments, with exception
for PINUS 0–10 cm.
3.4
C mineralization experiments
3.4.1
Whole soil C mineralization
The cumulative C mineralization curves were calculated for
all land use treatments as shown in Fig. 5 for SGFOR. The
linear part of each curve (from day 17 to 62) was used to
estimate the potential C mineralization rate via linear regres-
sion (Table 7). The differences in potential C mineralization
rates between the different land use treatments were signif-
icant in the 0–10 cm soil layer: SGFOR
>
PINUS
>
GRASS.
However, differences between the three land use treatments
disappeared gradually with depth. In the 10–20 cm and 20–
30 cm layer SGFOR still showed the highest potential miner-
alization rate, but differences were no longer significant. For
all land use treatments, potential C mineralization rates de-
creased with soil depth. The C mineralization rate constant,
Biogeosciences, 2, 159–174, 2005
www.biogeosciences.net/bg/2/159/
Page 9
D. Huygens et al.: Aggregate and soil organic carbon dynamics in South Chilean Andisols
167
Table 5. Weight, C concentration, and % of total C of the different soil organic carbon fractions for the second growth forest (SGFOR), grass-
land (GRASS) and plantation (PINUS) for the 0–10 cm soil layer (Standard deviations between brackets, significant differences (
p<
0.05)
between land use treatments are indicated by different letters).
Weight
C-concentration
% of total C
(g kg
−
1
soil)
(g C kg
−
1
fraction)
(%)
NO
GR
PR
NO
GR
PR
NO
GR
PR
Whole soil
1000
1000
1000
139
A
111
B
94
C
(7)
(1)
(6)
SOC fractions
MOM
31.5
A
8.6
B
14.4
B
366
A
309
B
325
B
8.3
A
2.4
B
5.0
C
(5.8)
(1.8)
(2.8)
(5)
(12)
(32)
(1.9)
(0.5)
(0.8)
HF
4.6
A
1.7
B
4.5
A
239
A
251
A
230
A
0.8
A
0.4
B
1.1
A
(0.4)
(0.1)
(1.6)
(28)
(8)
(34)
(0.1)
(0.0)
(0.3)
IF
4.5
A
1.1
C
2.4
B
328
A
284
A
305
A
1.0
A
0.3
C
0.8
B
(0.1)
(0.2)
(0.6)
(20)
(18)
(47)
(0.1)
(0.0)
(0.1)
LF
22.5
A
5.8
C
7.6
B
399
A
332
B
385
A
6.5
A
1.7
B
3.1
B
(5.5)
(1.7)
(1.5)
(5)
(16)
(14)
(1.9)
(0.5)
(0.7)
Mineral fraction
738
A
756
A
749
A
108
A
109
A
90
B
56.4
A
74.4
B
71.5
B
150
µ
m–2000
µ
m
(28)
(16)
(8)
(11)
(9)
(3)
(3.8)
(3.2)
(4.3)
50–150
µ
m
152
A
130
A
143
A
149
A
107
B
99
B
16.7
A
12.5
A
15.0
A
(40)
(23)
(9)
(19)
(4)
(10)
(6.3)
(1.8)
(0.9)
<
50
µ
m
91
A
104
A
93
A
122
A
110
A
84
B
7.9
A
10.3
B
8.3
A,B
(15)
(6)
(14)
(5)
(2)
(8)
(0.7)
(0.7)
(1.7)
Sum of the fractions
89.5
A
99.6
A
99.7
A
(8.2)
(3.6)
(5.5)
SOC=Soil organic carbon
MOM=Macro organic matter (
>
150
µ
m)
HF=Heavy density fraction (density
>
1.37 Mg m
−
3
)
IF=Intermediate density fraction (1.13 Mg m
−
3
<
density
<
1.37 Mg m
−
3
)
LF=Light density fraction (density
<
1.13 Mg m
−
3
)
which can be used as an indicator of the degradability of the
SOC, was calculated by dividing the observed potential C
mineralization rate (expressed per gram soil) by the corre-
sponding TC content in each soil layer (Table 7). For the
0–10 cm soil layer, these values varied from 0.029 for SG-
FOR to 0.023 for PINUS and 0.011 yr
−
1
for GRASS. In
all soil layers, the decomposability of the SOC was signif-
icantly lower for GRASS than for PINUS and SGFOR. The
SOC from GRASS showed the highest stability against mi-
crobial breakdown. Statistical correlation tests showed that
the C mineralization rate constants were significantly corre-
lated with the difference in WAD (
r
=0.39;
p
=0.04) and the
extractable Al
ox
content (
r
=
−
0.68;
p
=0.00).
3.4.2
Incubation of the isolated MOM density fractions HF
and LF
In this procedure, the MOM was isolated by the density
separation procedure, through which the physical protec-
tion of SOC in soil aggregates could be removed (Fig. 1).
Fig. 5. Evolution of the cumulative C mineralization for the second
growth forest (SGFOR) for the 0–10 cm, 10–20 cm and 20–30 cm
soil layer (day 1 to day 62) (error bars indicate plus/minus one stan-
dard deviation), and linear regression of the curves (from day 17 to
day 62).
www.biogeosciences.net/bg/2/159/
Biogeosciences, 2, 159–174, 2005
Page 10
168
D. Huygens et al.: Aggregate and soil organic carbon dynamics in South Chilean Andisols
Table 6.
δ
13
C values (‰) of the soil organic carbon fractions for the second growth forest (SGFOR), grassland (GRASS) and plantation
(PINUS) (Standard deviations between brackets, significant differences (
p<
0.05) between soil organic carbon fractions are indicated by
different letters).
δ
13
C (‰)
Fraction
0–10 cm
10–20 cm
20–30 cm
SGFOR
GRASS
PINUS
SGFOR
GRASS
PINUS
SGFOR
GRASS
PINUS
Whole soil
−
27.2
−
26.7
−
27.8
−
26.4
−
26.8
−
26.7
−
26.2
−
26.7
−
26.4
(0.1)
(0.3)
(0.2)
(0.4)
(0.1)
(0.2)
(0.2)
(0.3)
(0.1)
Macro organic matter
−
28.3
b
−
28.7
b
−
29.2
b
−
27.7
b
−
27.8
b
−
28.2
b
−
27.7
b
−
27.5
b
−
28.2
b
(0.4)
(0.6)
(0.6)
(0.1)
(0.1)
(0.3)
(0.1)
(0.3)
(0.2)
Size fractions
250
µ
m–2000
µ
m
−
27.2
a
−
26.7
a
−
27.9
a
−
26.4
a
−
26.8
a
−
26.7
a
−
26.3
a
−
26.7
a
−
26.4
a
(0.1)
(0.6)
(0.6)
(0.4)
(0.1)
(0.2)
(0.2)
(0.4)
(0.1)
150
µ
m–250
µ
m
−
27.2
a
−
26.7
a
−
27.9
a
−
26.4
a
−
26.8
a
−
26.7
a
−
26.2
a
−
26.7
a
−
26.3
a
(0.2)
(0.5)
(0.6)
(0.4)
(0.1)
(0.1)
(0.3)
(0.3)
(0.2)
50
µ
m–150
µ
m
−
27.1
a
−
26.6
a
−
27.7
a
−
26.4
a
−
26.8
a
−
26.7
a
−
26.2
a
−
26.7
a
−
26.3
a
(0.2)
(0.7)
(0.7)
(0.4)
(0.1)
(0.1)
(0.2)
(0.3)
(0.2)
<
50
µ
m
−
27.2
a
−
26.7
a
−
27.7
a
−
26.4
a
−
26.8
a
−
26.7
a
−
26.1
a
−
26.6
a
−
26.2
a
(0.2)
(0.6)
(0.7)
(0.3)
(0.1)
(0.1)
(0.3)
(0.1)
(0.1)
Table 7. Potential C mineralization rates and decomposition rate constants for the second growth forest (SGFOR), grassland (GRASS) and
plantation (PINUS) for 0–10 cm, 10–20 cm and 20–30 cm soil layer (Standard deviation between brackets, significant differences between
land use treatments (
p<
0.05) are indicated by different letters).
Land use class
Potential mineralization rate
Mineralization rate constant
0–10 cm
10–20 cm
20–30 cm
0–10 cm
10–20 cm
20–30 cm
g CO
2
-C m
−
2
d
−
1
—
yr
−
1
—
SGFOR
0.495
a
0.192
a
0.142
a
0.029
a
0.017
a
0.018
a
(0.108)
(0.061)
(0.041)
(0.005)
(0.005)
(0.003)
GRASS
0.196
c
0.092
a
0.087
a
0.011
b
0.005
b
0.005
b
(0.021)
(0.021)
(0.034)
(0.001)
(0.001)
(0.002)
PINUS
0.266
b
0.135
a
0.087
a
0.023
a
0.015
a
0.014
a
(0.041)
(0.034)
(0.044)
(0.001)
(0.008)
(0.006)
Especially, the accessibility of the HF has been increased, as
this fraction is more associated with soil minerals and com-
ponents compared to the LF. The LF is mainly present in the
soil as free MOM. Only a small fraction is bounded in soil
aggregates (Hassink, 1995; Swanston et al., 2002). Potential
C mineralization rates and C mineralization rate constants of
the isolated density fractions were calculated using the same
method as the whole soil C mineralization experiment (Ta-
ble 8). In this experiment, the mineralization rate constants
are an indicator of the recalcitrance of the MOM fractions.
Statistical analyses showed that the potential C mineraliza-
tion rates of the LF and HF were significantly different for
all land use treatments. The C mineralization rate constants
of the LF and HF are significantly different for GRASS, but
not for SGFOR and PINUS.
The C mineralization rate constants showed that the
MOM from GRASS decomposes fastest.
The recal-
citrance of the LF decreased in the following order:
PINUS
>
SGFOR
>
GRASS. For the HF, PINUS showed a
higher recalcitrance than GRASS, whereas the value of SG-
FOR was not significantly different from PINUS and GRASS
(Table 8).
4
Discussion
In a first section of the discussion, we will focus on the in-
teraction of Al, SOC and soil minerals. The potential role of
Al-humus complexes and electrostatic-C binding on aggre-
gate structure will be assessed in this Andisol. In a second
Biogeosciences, 2, 159–174, 2005
www.biogeosciences.net/bg/2/159/
Page 11
D. Huygens et al.: Aggregate and soil organic carbon dynamics in South Chilean Andisols
169
Table 8. Potential mineralization rates and mineralization rate constants of the light (LF) and heavy (HF) density fraction for the second
growth forest (SGFOR), grassland (GRASS) and plantation (PINUS) (Standard deviations between brackets, significant differences (
p<
0.05)
are indicated by different letters (upper case letters between LF and HF, lower case letters between land use treatments)).
Land use class
Potential mineralization rate
Mineralization rate constant
LF
HF
LF
HF
mg CO
2
-C kg
−
1
fraction d
−
1
yr
−
1
SGFOR
131.2
A
a
59.7
B
a,b
0.120
A
a
0.090
A
a,b
(11.4)
(14.6)
(0.010)
(0.022)
GRASS
137.4
A
a
73.0
B
a
0.151
A
b
0.106
B
a
(6.2)
(7.6)
(0.007)
(0.011)
PINUS
74.7
A
b
37.6
B
b
0.071
A
c
0.060
A
b
(14.2)
(8.4)
(0.013)
(0.013)
part, we will evaluate the results of C mineralization experi-
ments. In a third section, a link between aggregate structure,
C mineralization and SOC stabilization will be made. In a
last section, the potential role of pH on aggregate structure
and SOC stabilization will be discussed.
4.1
Aggregate structure in south Chilean Andisols
In literature, two main sorption mechanisms are documented
in Al rich soils. As sorption tends to promote aggregation
(Sollins et al., 1996), these mechanisms lead to aggregate
build up and might be responsible for the occurrence of high
amounts of water-stable aggregates in this soil type (e.g.
Hoyos and Comerford, 2005; see Sect. 3.2). Firstly, metal-
humus complexes would act as binding agents between ag-
gregates or soil particles. In metal-humus complexes, SOC
can be sorbed by allophane through ligand exchange between
functional groups of SOC and Al (Parfitt et al., 1977; Skjem-
stad, 1992; Parfitt et al., 1999). In this mechanism, an or-
ganic ligand replaces a hydroxyl group, which is coordinated
to the metal ion on the surface of the oxide to form a surface
complex. Since surface hydroxyls are characteristic for vari-
able charge surfaces, sorption of negatively charged organics
by ligand exchange should be most important in oxide-rich
and allophanic soils, such as Andisols (Sollins et al., 1996).
Other bounding types involved in sorption between SOC and
metals include cation bridging (Varadachari et al., 1995), wa-
ter bridging, hydrogen bounding, and van der Waals forces
(Burchill et al., 1981). However, distinction between poten-
tial types is impossible bearing in mind the methodology of
this study. In this first case, the mechanism of aggregate hi-
erarchy (Tisdall and Oades, 1982) would be adequate to de-
scribe C dynamics as SOC (bonded to Al) acts as principal
binding agent between soil components.
Secondly, electrostatic attractions could occur between
and among amorphous Al components and clay minerals
due to the simultaneous existence of positive and negative
charges at field pH (Schofield and Samson, 1954; El-Swaify,
1980; Six et al., 2000b; Six et al., 2000c; Denef et al., 2002).
Thus, here aggregate formation is partly induced by electro-
static interactions between soil minerals and Al components
and aggregate hierarchy should be less pronounced (Oades
and Waters, 1991).
We conceptually tested the occurrence of aggregate hi-
erarchy for the different land use treatments. Three con-
sequences of this theory are: (1) a gradual breakdown of
macroaggregates into microaggregates before they dissolve
into primary particles, as increasing dispersive energy is ap-
plied to the soil (Oades and Waters, 1991), (2) an increase
in percentage C concentration with increasing aggregate-size
class because large aggregate-size classes are composed of
small aggregate-size classes plus organic binding agents (El-
liott, 1986), and (3) younger and more labile organic matter
is contained in macroaggregates rather than in microaggre-
gates (Elliott, 1986; Puget et al., 1995; Jastrow et al., 1996).
Firstly, the fractionation of the soils at two different periods
allowed us to test the aggregate stability. A gradual break-
down of macroaggregates into microaggregates could be ob-
served (Figs. 2, 3). It was shown that this breakdown could
not be explained by the TC content, allophane content or ex-
tractable Al
py
content of the soil, but by the extractable Al
ox
content of the soil (Fig. 4). In the latter Al-extraction proce-
dure, only amorphous and poorly crystalline Al-oxides and
Al bound to organic complexes could be isolated (Bertsch
and Bloom, 1996). Secondly, the TC concentration did not
differ among the different size classes for a certain land use
treatment (Table 3). Absence of differences in C concen-
tration between aggregate-size classes in Andisols was also
documented by Hoyos and Comerford (2005). Thirdly, there
were no significant differences in
δ
13
C values among the dif-
ferent size classes for a certain soil (Table 6). In contrast to
e.g. Accoe et al. (2002), no gradual shift in
δ
13
C value to-
wards finer aggregate classes could be observed as a result of
increasing decomposition (O’Brien and Stout, 1978). In this
study, we could only distinguish two different SOC pools:
the MOM and the mineral soil pool. It was indicated that the
www.biogeosciences.net/bg/2/159/
Biogeosciences, 2, 159–174, 2005
Page 12
170
D. Huygens et al.: Aggregate and soil organic carbon dynamics in South Chilean Andisols
SOC in the different size fractions is at the same degree of
decomposition.
The results of these experiments indicated that metal-
humus complexes are probably not the main binding agents
between soil aggregates in these Andisols. Aggregate build-
up would be mainly a result of electrostatic attractions be-
tween and among Al and Al-oxides and clay minerals, as
documented as well by Schofield and Samson (1954), Vi-
olante and Huang (1992) and Denef et al. (2002).
SOC
would only play a minor role in these strong bindings be-
tween larger aggregates. Absence of aggregate hierarchy
in this soil type has earlier been documented by Hoyos and
Comerford (2005).
On the other hand, ploughing (Beare et al., 1994; Six et
al., 1999) and drying of the soil might influence the ag-
gregate structure. Every two year, the GRASS site is dis-
turbed by ploughing. Drying of Andisols might cause dra-
matic changes of the physical properties of aggregates, and
lead to a coarser aggregate size distribution (Shoji et al.,
1993). Even air-drying can affect the particle size and micro-
aggregate distribution. Tillage on the GRASS site might have
exposed the soil unprotected to the sun and resulted in drying
of the topsoil layer, leading to an increase in stable aggre-
gates. Continued ploughing would lead to a mixture of the
aggregates in the upper 30 cm of the soil.
4.2
C mineralization experiments
Land use changes affect SOC dynamics by differences in
SOC input, and by direct effects of accompanying soil distur-
bances (Guo and Gifford, 2002). Pastures maintain a contin-
uous vegetation cover on the soil and have high turnover rates
and productivity, particularly belowground. As a result more
intense humus formation would take place (Brown and Lugo,
1990; Yakimenko, 1998). Jobbagy and Jackson (2000) indi-
cated that the annual turnover of SOC from dying tree roots
is smaller than from grass roots. In forested ecosystems, the
SOC inputs originate partly from aboveground and accumu-
late more in and on the surface soil due to non-mixing of the
litter (Paul et al., 2002). SOC tends to decompose partly be-
fore it is incorporated into the forest soil, involving a smaller
formation of SOC. Thus, under grassland a thicker A horizon
is formed compared with forests (e.g. Birkland, 1984). Ac-
cording to Guo and Gifford (2002), the lower SOC content in
plantations may partly be attributed to soil disturbances and
consequent loss in physical protection of the SOC involved
in the establishment of plantations. The transformation from
a multi-storey forest to a single storey plantation lacking un-
derstorey vegetation results in less SOC input (Guo and Gif-
ford, 2002). Also, during the first years after the establish-
ment of a plantation, reduced SOC sequestration takes place
as a result of low SOC input (due to a small forest biomass
and low litter fall rate) (Wilde, 1964) and a continued de-
composition of the SOC present in the soil profile. Carbon
storage as well is affected by forest type (species, deciduous,
evergreen) (Lal et al., 1995).
In
our
study,
the
whole
soil
C
mineralization
rate
constant
decreased
in
the
following
order:
SGFOR
≈
PINUS
>
GRASS. Our values are in accor-
dance to those found in Columbian Andisols by Boudot et
al. (1986), who measured a mineralization rate constant of
approximately 0.01 yr
−
1
in the 5–20 cm layer. Based on
the C mineralization rate constants of the incubated MOM
fractions, we were able to distinguish differences in the
recalcitrance of the SOC of the different land use treatments
(see below).
4.3
Evaluation of SOC stabilization mechanisms
As stated by Sollins et al. (1996), SOC can be stabilized
by different mechanisms. We will consider each factor or
process that controls the rate of stabilization, and discuss its
potential effects on SOC stabilization in our soils. Sollins
(1996) brought forward following sets of characteristics
affecting the stability of SOC: recalcitrance, interactions
and accessibility. Though we mainly focus on the aspect
of accessibility, one has to bear in mind that several mech-
anisms and processes interact to determine the stability or
biodegradability of SOC in soils (Marschner and Kalbitz,
2003). Keeping in mind the used methodology of this study,
we cannot deduce conclusions on all aspects. In order to
assess the importance of certain mechanisms, we will refer
to the state-of-the-art in literature.
Recalcitrance
We tried to assess the importance of this parameter for
the studied land use treatments by performing mineralization
experiments of the isolated MOM fractions. We encountered
that the degree of the whole soil C mineralization rate
constants was opposite to the degree of the C mineralization
rate constants of the isolated MOM fractions. Therefore,
recalcitrance of the SOC did not appear to control SOC
stabilization. Thus, other stabilization mechanisms should
overrule the influence of litter quality on whole soil C
mineralization rates.
Interactions between soil components
We showed that in our soils, Al is associated to clay
minerals by electrostatic sorption.
These clay minerals
are associated to SOC, leading to the formation of metal-
humus-clay complexes.
Direct adsorption of SOC to Al
by ligand exchange is considered of minor importance in
our soils. Many studies (e.g. Percival et al., 2000) assume
ligand exchange to be very important in Andisols, based
on data of pyrophosphate extractable Al. In the literature
(Bertsch and Bloom, 1996), the extractable Al
py
content is
often related to the directly organically bound Al, while the
Biogeosciences, 2, 159–174, 2005
www.biogeosciences.net/bg/2/159/
Page 13
D. Huygens et al.: Aggregate and soil organic carbon dynamics in South Chilean Andisols
171
quantity of poorly ordered amorphous or noncrystalline Al
phases plus the organically bound Al would be proportional
to Al
ox
. However, we believe that in a very close association
between SOC, Al and soil minerals, as occurring in our
soils, extractants cannot be completely selective for one
certain phase. This as well was brought forward by Bertsch
and Bloom (1996). Kaiser and Zech (1996) indicated that
pyrophosphate also removed Al hydroxides and stated that it
should not be used to estimate Al in humus complexes.
The formation of metal-humus-clay complexes by sorp-
tion onto clay minerals is usually assumed to stabilize
SOC. It is known that such metal-humus-clay complexes
are resistant to microbial decomposition (Juste et al., 1975;
Martin et al., 1982) in Andisols. These adsorption reactions
would provide a mechanism of stabilizing SOC against
biological attack (Baldock and Skjemstad, 2000). However,
the relevance of SOC stabilization remains poorly under-
stood (Sollins et al., 1996). Sollins et al. (1996) indicated
the difficulty to separate physico-chemical processes and
physical protection (see below) as factors influencing SOC
stabilization. Physico-chemical associations namely tend to
promote aggregation. Boudot et al. (1989) observed that the
binding of metals (such as Al and Fe) per se in metal-humus-
clay complexes could not be regarded as responsible for
any significant protective effect against biodegradation. The
formation of an insoluble metallic hydroxide trapping and
wrapping SOC molecules, and therefore either preserving
it from access by soluble soil enzymes or preventing their
movement to immobile enzymatic constituents associated
with microbial cells (=inaccessibility or physical protection),
would better account for the observed protective effect than
the chemical binding per se.
Physical protection
Physical protection and aggregation can influence ac-
cessibility of substrate to microbes and fauna. Theoretical
calculations suggest that aggregation should limit access to
SOC (e.g. Van Veen and Kuikman, 1990). Six et al. (2000a)
underscored the importance of soil aggregation and more
specifically the interactions of SOC and aggregate dynamics
in controlling SOC sequestration in soils. Gijsman and Sanz
(1998) investigated the physical protection of macroaggre-
gates in Andisols. After crushing of macroaggregates, these
authors observed a significant increase in CO
2
mineraliza-
tion compared to intact soil samples. Plante and McGill
(2002) indicated that macroaggregate turnover results in
the exposure of labile organic matter.
Jastrow (1996)
found evidence for a relationship between macroaggregate
formation and SOC accrual.
Aggregate stability affects
the C mineralization process in Andisols by a physically
protecting SOC from microbial breakdown. In our soils, the
latter was confirmed by the significant correlation between
the whole soil C mineralization rate constants and the
difference in WAD (
r
=0.39,
p
=0.04).
We believe that this phenomenon is the most important
SOC stabilization parameter occurring in our soils. How-
ever, it is probably an indirect effect of sorption mechanisms
due to the presence of amorphous Al minerals as described
above. Further statistical evidence for this conclusion was
found in a significant correlation between the C mineraliza-
tion rate constant and Al
ox
content (
r
=
−
0.68,
p
=0.00), and
the difference in WAD (Fig. 4). High Al
ox
contents lead
to the formation of stable soil aggregates which offer phys-
ical protection to the SOC. The importance of amorphous
clay materials for SOC stabilization and aggregation in al-
lophanic Andisols has often been documented (e.g. Boudot
et al., 1989; Nanzyo et al.,1993; Torn et al., 1997), and is
assumed to be the main stabilization mechanism in this soil
type.
4.4
Effect of soil pH on SOC stabilization
Under lower pH conditions, Al is more soluble and available
for complexation with soil components (Wesselink et al.,
1996). The pH differences between GRASS and both SG-
FOR and PINUS were significant for all soil layers, and most
probably a result of the acidifying fertilizer applications. As
a result of a lower pH, the extractable Al
ox
and Al
py
concen-
trations in the GRASS site were higher than for SGFOR and
PINUS. We found that allophane contents were significantly
lower for GRASS than for SGFOR and PINUS. It was in-
dicated that as a result of a lower pH, Al was released from
allophanic compounds and consequently resulted in a higher
Al content associated with the Al-humus complexes. Statisti-
cal evidence for this conclusion was found by the strong sig-
nificant correlation between soil pH and allophane (
r
=0.74,
p
=0.00). Increased Al release from allophane at decreasing
pH was also found by Farmer and Lumsdon (2002). Taka-
hashi et al. (2005) found that as a result of liming treatments,
the amounts of organically complexed Al decreased signifi-
cantly. Dahlgren and Saigusa (1994) found in Andisols ev-
idence for the precipitation of released Al from allophanic
components on soil surfaces, while Walker et al. (1988)
found more specifically that soluble Al was adsorbed ex-
tremely rapid on common soil minerals such as vermiculite.
Violante and Huang (1992) found in OM rich soils that sorp-
tion and aggregation between soil particles was promoted
by increasing Al precipitation. Tama and El-Swaify (1978)
stated that soil pH affects the ability of soil particles to ag-
gregate differently in variable charge soils. The electrostatic
contribution of oxides to aggregate stability is pH depen-
dent and affects the extent of aggregate breakdown at dif-
ferent pH values. In lowering soil pH, the amount of pos-
itive charges of the Al-oxides and allophane increases (Es-
pinoza et al., 1975). As such, more electrostatic forces are
involved in aggregating soil particles. However, in our sta-
tistical analyses of the aggregate stability test no significant
correlation was found between soil pH and difference in
WAD between the two fractionation procedures. Possibly,
www.biogeosciences.net/bg/2/159/
Biogeosciences, 2, 159–174, 2005
Page 14
172
D. Huygens et al.: Aggregate and soil organic carbon dynamics in South Chilean Andisols
the sigmoid retention curves, relating the effect of soil pH
and positive charges, documented in south Chilean Andisols
(Espinoza et al., 1975) might be responsible for the absence
of a linear correlation.
5
Conclusions
It can be concluded that by electrostatic sorption between
and among amorphous Al compounds and clay minerals,
physico-chemical stabilization of the SOC occurs. As clay
minerals are closely associated with organic matter com-
pounds, stable metal-humus-clay complexes can be formed.
The principle of aggregate hierarchy was not valid in this soil
type as metal-humus complexes formed as a result of ligand
exchange did not appear to be important binding agents be-
tween soil aggregates. Soil pH is an important parameter
in the studied Andisol as it influences aggregate stability. It
was shown that SOC stabilization occurred via the formation
of metal-humus-clay complexes and physical protection of
SOC in soil aggregates. The recalcitrance of the SOC did not
appear to influence SOC stabilization.
The whole soil C mineralization constants of the differ-
ent land use treatments decreased in the following order:
SGFOR
≈
PINUS
>
GRASS. In this study, land use conver-
sion from a second growth forest (SGFOR) to a grassland
(GRASS) showed, indirectly, a positive effect on C stabiliza-
tion in the soil. As a result of management practices soil
pH decreased in GRASS and more metal-humus-clay com-
plexes and stable soil aggregates could be formed. A higher
physical protection of the SOC was reached on this site, re-
sulting in lower C mineralization rates. Other influences due
to management practices (ploughing and direct exposition to
sunlight due to the absence of a vegetation cover), and differ-
ences in root C input might influence as well SOC stabiliza-
tion in the different sites.
Acknowledgements. This research was supported by the Fund for
Scientific Research – Flanders (Belgium) (FWO), the Fondecyt
Project No. 1020989 Millenium Nucleus Forecos P01-057-F, and
the Bilateral Scientific and Technological Cooperation between
Flanders and Chile. We thank the Flemish Interuniversity Council
(VLIR) for the assignment of a scholarship for the research work in
Chile.
Edited by: E.-D. Schulze
References
Accoe, F., Boeckx, P., Van Cleemput, O., Hofman, G., Hui, X.,
Bin, H., and Guanxiong, C.: Characterization of soil organic
matter fractions from grassland and cultivated soils via C con-
tent and
δ
13
C signature, Rapid Commun. Mass Spectrom., 16,
2157–2164, 2002.
Baldock, J. A. and Skjemstad, J. O.: Role of the soil matrix and
minerals in protecting natural organic materials against biologi-
cal attack, Org. Geochem., 31, 697–710, 2000.
Beare, M. H., Hendrix, P. F., and Coleman, D. C.: Water-stable
aggregates and organic matter fractions in conventional and no-
tillage soils, Soil Sci. Soc. Am. J., 58, 777–786, 1994.
Bertsch, P. M. and Bloom, P. R.: Aluminum, in: Methods of soil
analysis Part 3- Chemical methods, edited by: Sparks, D. L.,
SSSA Book Series, SSSA, Madison, 1996.
Besoain, E. M.: Clay mineralogy of volcanic ash soils, Panel of Vol-
canic ash soils in Latin America, Soil B.1.1–B.1.16., Turrialba,
Costa Rica, 1969.
Besoain, E. M.: Mineralogia de las arcillas de algunos suelos
volc´anicos de Chile, Agr. T´ec., 18, 110–165, 1958.
Besoain, E. M., Peralta, M. P., and Massaro, S. M.: Mineralog´ıa y
genesis de algunas suelos de cenizas volc´anicas de Chilo´e conti-
nental, Chile, Agr. T´ec., 60, 127–153, 2000.
Birkland, P.: Soils and geomorphology, Oxford University Press,
New York, 1984.
Boudot, J.-P., Bel Hgadj, B. A., and Chone, T.: Carbon mineraliza-
tion in Andosols and aluminium-rich highland soils, Soil Biol.
Biochem., 18, 457–461, 1986.
Boudot, J.-P., Bel Hadj, B. A., Steiman, R., and Seigle-Murandi,
F.: Biodegradation of synthtic organo-metallic complexes of iron
and aluminium with selected metal to carbon ratios, Soil Biol.
Biochem., 21, 961–966, 1989.
Brown, S. and Lugo, E. A.: Effects of forest clearing and succession
on the carbon and nitrogen content of soils in Puerto Rico and US
Virgin Islands, Plant Soil, 124, 53–64, 1990.
Burchill, S., Greenland, D. J., and Hayes, M. H. B.: Adsorption of
organic molecules, in: The chemistry of soil processes, edited
by: Greenland, D. J. and Hayes, M. H. B., Wiley, Chichester,
221–400, 1981.
Dahlgren, R. A. and Saigusa, M.: Aluminum release rates from
allophanic and nonallophanic andosols, Soil Sci. Plant Nutr., 40,
125–136, 1994.
de Koning, G. H. J., Veldkamp, E., and L´opez-Ulloa, M.: Quantifi-
cation of carbon sequestration in soils following pasture to forest
conversion in northwestern Ecuador, Global Biogeochem. Cy-
cles, 17 (4), Art. No. 9, doi:101029/2003GB002099, 2003.
De Leenheer, L. and De Boodt, M.: Determination of aggregate
stability by the change in Mean Weight Diameter, Proceedings
of the international symposium on soil structure, Gent, Belgium,
Rijkslandbouwhogeschool, Gent, Belgium, 290–300, 1959.
Del Galdo, I., Six, J., Peressotti, A., and Cotrufo, M. F.: Assessing
the impact of land-use change on soil C sequestration in agricul-
tural soils by means of organic matter fractionation and stable C
isotopes, Global Change Biol., 9, 1204–1213, 2003.
Denef, K., Six, J., Merckx, R., and Paustian, K.: Short-term effects
of biological and physical forces on aggregate formation in soils
with different clay mineralogy, Plant Soil, 246, 185–200, 2002.
Elliott, E. T.: Aggregate structure and carbon, nitrogen and phos-
phorus in native and cultivated soils, Soil Sci. Soc. Am. J., 50,
627–633, 1986.
El-Swaify, S. A.: Physical and mechanical properties of Oxisols,
in: Soils with variable charge, edited by: Theng, B.K.G., Offset
Publ., Palmerston North, 303–324, 1980.
Espinoza, W., Gast, R. G., and Adams, R. S. J.: Charge charac-
teristics and nitrate retention by two Andepts from south-central
Chile, Soil Sci. Soc. Am. J., 39, 842–846, 1975.
Biogeosciences, 2, 159–174, 2005
www.biogeosciences.net/bg/2/159/
Page 15
D. Huygens et al.: Aggregate and soil organic carbon dynamics in South Chilean Andisols
173
FAO-UNESCO: Soil map of the world, Volume IV, South America,
UNESCO, Paris, 1971.
Farmer, V. C. and Lumsdon, D. G.: A re-interpretation of ’Alu-
minium solubility mechanisms in moderately acid Bs horizons
of podzolized soils’ by Gufstafsson et al., Eur. J. Soil Sci., 53,
671–673, 2002.
Galindo, G. G., and Bingham, T.: Homovalent and heterovalent
cation exchange equilibria in soils with variable surface charge,
Soil Sci. Soc. Am. J., 41, 833–866, 1977.
Gijsman, A. J. and Sanz, J. I.: Soil organic matter pools in a
volcanic-ash soil under fallow or cultivation with applied chicken
manure, Eur. J. Soil Sci., 49, 427–436, 1998.
Guo, L. B. and Gifford, R. M.: Soil carbon stocks and land use
change: a meta analysis, Global Change Biol., 8, 345–360, 2002.
Hassink, J.: Density fractions of soil macroorganic matter and mi-
crobial biomass as predictors of C and N mineralization, Soil
Biol. Biochem., 27, 1099–1108, 1995.
Hoyos, N. and Comerford, N. B.: Land use and landscape effects on
aggregate stability and total carbon of Andisols from the Colom-
bian Andes, Geoderma, accepted, 2005.
Jastrow, J. D.: Soil aggregate formation and the accrual of particu-
late and mineral-associated organic matter, Soil Biol. Biochem.,
28, 665–676, 1996.
Jobbagy, E. G. and Jackson, R. B.: The vertical distribution of soil
organic carbon and its relation to climate and vegetation, Ecol.
Appl., 10, 423–436, 2000.
Juste, C., Delas, J., and Langon, M.: Comparison de la stabilit
biologique de diff´erents humates m´etalliques, C. R. Acad. Sci.,
281, 1685–1688, 1975.
Kaiser, K. and Zech, W.: Defects in estimation of aluminum in hu-
mus complexes of podzolic soils by pyrophosphate extraction,
Soil Sci., 161, 452–458, 1996.
Klute, A.: Methods of Soil Analysis (2nd Ed.), Agronomy Mono-
graph Vol. 9, ASA-SSSA, Madison, 1986.
Kubota, T.: Aggregate formation of allophonic soils: effect of dry-
ing on the dispersion of the soils, Soil Sci. Plant Nutr., 18, 79–87,
1972.
Lal, R., Kimble, J. M., Levine, E., and Stewart, B. A.: World soils
and greenhouse effect: an overview, in: Soils and global change,
edited by: Lal, R., Kimble, J. M., Levine, E., and Stewart, B. A.,
CRC Lewis Publishers, Boca Raton, 1–7, 1995.
Lara, A., Solari, M. E., Rutherford, P., Thiers, O., Tracaman,
R., Molina, R., Prieto, R., and Montory, C.: Cobertura de
la vegetaci´on original de la Ecoregi´on de los bosques valdi-
vianos en Chile hacia 1550, Informe T´ecnico, Proyecto FB 49-
WWF/Universidad Austral de Chile, Valdivia, Chile, 1999.
Marschner, B. and Kalbitz, K.: Controls of bioavailability and
biodegradability of dissolved organic matter in soil, Geoderma,
113, 211–235, 2003.
Martin, J. P., Zunino, H., Peirano, P., Caiozzi, M., and Haider,
K.: Decomposition of 14C-labilled lignins, model humic acid
polymers and fungal melanins in allophanic soils, Soil Biol.
Biochem., 14, 289–293, 1982.
Meijboom, F. W., Hassink, J., and Van Noordwijk, M.: Density
fractionation of soil macroorganic matter using silica suspen-
sions, Soil Biol. Biochem., 27, 1109–1111, 1995.
Milford, M. H.: Soils and Soils Science: Laboratory excercises, 4th
Ed., Texas A&M University, Kendall/Hunt Publishing Company,
Dubuqua, Iowa, 1997.
Mizota, C. and van Reeuwijk, L. P.: Clay mineralogy and chemistry
of soils formed in volcanic material in diverse climatic regions,
Soil Monogr. 2, Int. Soil Reference and Information Centre, Wa-
geningen, The Netherlands, 1989.
Nanzyo, M., Dahlgren, R. A., and Shoji, S.: Chemical charac-
teristics of volcanic ash soils, in: Volcanic ash soils: Genesis,
properties and utilization, edited by: Shoji, S., Nanzyo, M. and
Dahlgren, R. A., Developments in Soil Science 21, Elsevier, Am-
sterdam, the Netherlands, 145–187, 1993.
Oades, J. M. and Waters, A. G.: Aggregate hierarchy in soils, Aust.
J. Soil Res., 29, 815–828, 1991.
O’Brien, B. J. and Stout, J. D.: Movement and turnover of soil or-
ganic matter as indicated by carbon isotope measurements, Soil
Biol. Biochem., 10, 309–317, 1978.
Parfitt, R. L.: Allphane in New Zealand – a review, Aust. J. Soil
Res., 28, 343–360, 1990.
Parfitt, R. L., Fraser, A. R., and Farmer, V. C.: Adsorption on hy-
drous oxides. III. Fulvic acid and humic acid on goethite, gibbsite
and imogolite, J. Soil Sci., 28, 289–296, 1977.
Parfitt, R. L. and Wilson, A. D.: Estimation of allophane and
halloysite in three sequences of volcanic soils, New Zealand,
in: Volcanic soil, edited by: Caldas, E. F. and Yaalon, D. H.,
CATENA suppl. 7, Catena, Braunschweig, Germany, 1–8, 1985.
Parfitt, R. L., Yuan, G., and Theng, B. K. G.: A 13C-NMR study of
the interactions of soil organic matter with aluminium and allo-
phane in podzols, Eur. J. Soil Sci., 50, 695–700, 1999.
Paul, K. I., Polglase, P. J., Nyakuengama, J. G., and Khanna, P.
K.: Change in soil carbon following afforestation, Forest Ecol.
Manag., 168, 241–257, 2002.
Percival, H. J., Parfitt, R. L., and Scott, N.: Factors controlling soil
carbon levels in New Zealand grasslands: Is clay content impor-
tant?, Soil Sci. Soc. Am. J., 64, 1623–1630, 2000.
Plante, A. F. and McGill, W. B.: Intraseasonal soil macroaggre-
gate dynamics in two contrasting field soils using labeled tracer
spheres, Soil Sci. Soc. Am. J., 66, 1285–1295, 2002.
Puget, P., Chenu, C., and Balasdent, J.: Total and young organic
matter distributions in aggregates of silty cultivated soils, Eur. J.
Soil Sci., 46, 449–459, 1995.
Radcliffe, D. J. and Gillman, G. P.: Surface charge characteristics
of volcanic ash soils from the southern highlands of Papua New
Guinea, in: Volcanic soil, edited by: Caldas, E. F. and Yaalon, D.
H., CATENA suppl. 7., Catena, Braunschweig, Germany, 35–46,
1985.
Schofield, R. K. and Samson, H. R.: Flocculation of kaolinite due to
the attraction of oppositely charged crystal faces, Faraday Dis.,
18, 135–145, 1954.
Schwertmann, U.: Differenzierung der Eisenoxide des Bodens
durch photochemische Extraction mit saurer ammoniumoxalate
L¨osung, Z. Pflanzenernaehr. Bodenkd., 105, 194–202, 1964.
Shoji, S., Nanzyo, M., and Dahlgren, R. A.: Volcanic ash soils-
Genesis, properties and utilization, Developments in Soil Sci-
ence, 21, Elsevier, Amsterdam, 1993.
Six, J., Elliott, E. T., and Paustian, K.: Aggregate and soil organic
matter dynamics under conventional no-tillage systems, Soil Sci.
Soc. Am. J., 63, 1350–1358, 1999.
Six, J., Elliott, E. T., and Paustian, K.: Soil macroaggregate turnover
and microaggregate formation: a mechanism for C sequestration
under no-tillage agriculture, Soil Biol. Biochem., 32, 2099–2103,
2000a.
www.biogeosciences.net/bg/2/159/
Biogeosciences, 2, 159–174, 2005
Page 16
174
D. Huygens et al.: Aggregate and soil organic carbon dynamics in South Chilean Andisols
Six, J., Elliott, E. T., and Paustian, K.: Soil structure and soil or-
ganic matter: II. A normalized stability index and the effect of
mineralogy, Soil Sci. Soc. Am. J., 64, 1042–1049, 2000b.
Six, J., Paustian, K., Elliott, E. T., and Combrink, C.: Soil structure
and soil organic matter: I. Distribution of aggregate size classes
and aggregate associated carbon, Soil Sci. Soc. Am. J., 64, 681–
689, 2000c.
Skjemstad, J. O.: Genesis of Podzols on coastal dunes in southern
Queensland: III. The role of aluminium-organic complexes in
profile development, Aust. J. Soil Res., 30, 645–665, 1992.
Sollins, P., Homann, P., and Caldwell, B. A.: Stabilization and
destabilization of soil organic matter: mechanisms and controls,
Geoderma, 74, 65–105, 1996.
Solomon, D., Lehmann, J., and Zech, W.: Land use effects on soil
organic matter properties of chromic Luvisols in semi-arid north-
ern Tanzania: carbon, nitrogen, lignin and carbohydrates, Agr.
Ecosyst. Environ., 78, 203–213, 2000.
Swanston, C. W., Caldwell, B. A., Homann, P. S., Ganio, L., and
Sollins, P.: Carbon dynamics during a long-term incubation
of separate and recombined density fractions from seven forest
soils, Soil Biol. Biochem., 34, 1121–1130, 2002.
Takahashi, T., Ikeda, Y., Fujita, K., and Nanzyo, M.: Effect of lim-
ing on organically complexed aluminum of nonallophanic An-
dosols from northeastern Japan, Geoderma, accepted , 2005.
Tama, K. and El-Swaify, S. A.: Charge, colloidal, and structural sta-
bility relationships in oxidic soils, in: Modification of soil struc-
ture, edited by: Emerson, W. W., Bond, R. D., and Dexter, A. R.,
John Wiley & Sons, Chicester, 41–52, 1978.
Tisdall, J. M. and Oades, J. M.: Organic matter and water-stable
aggregates in soils, J. Soil Sci., 33, 141–163, 1982.
Torn, M. S., Trumbore, S. E., Chadwick, O. A., Vitousek, P. M.,
and Hendricks, D. M.: Mineral control of soil organic storage
and turnover, Nature, 389, 170–170, 1997.
Tosso, J.: Suelos volc´anicos de Chile, Instituto de Investigaciones
Agropecuarias, Santiago, Chile, 1985.
Van Veen, J. A. and Kuikman, P. J.: Soil structural aspects of de-
composition of organic matter by microorganisms, Biogeochem-
istry, 11, 213-233, 1990.
Varadachari, C., Mondal, A. H., and Ghosh, K.: The influence of
crystal edges on clay-humus complexation, Soil Sci., 159, 185–
190, 1995.
Violante, A. and Huang, P. M.: Effect of tartaric acid and pH on
the nature and physicochemical properties of short-range ordered
aluminum precipitation products, Clays Clay Miner., 40, 462–
469, 1992.
Walker, W. J., Cronan, C. S., and Patterson, H. H.: A kinetic-
study of aluminum adsorption by aluminosilicate clay-minerals,
Geochim. Cosmochim. Acta, 52, 55–62, 1988.
Wesselink, L. G., van Breemen, N., Mulder, J., and Janssen, P. H.:
A simple model of soil organic matter complexation to predict
the solubility of aluminium in acid forest soils, Eur. J. Soil Sci.,
47, 373–384, 1996.
Wilde, S. A.: Changes in soil productivity induced by pine planta-
tions, Soil Sci., 97, 276–278, 1964.
Yakimenko, E. Y.: Soil comparative evolution under grasslands and
woodlands in the forest zone of Russia, in: Management of car-
bon sequestration in soil, edited by: Lal, R., Kimble, J. M., Fol-
let, R. F., and Stewart, B. A., CRC Press, Boca Raton, 391–404,
1998.
Zunino, H., Borie, F., Aguilera, S., Martin, J. P., and Haider, K.: De-
composition of 14C labeled glucose, plant and microbial prod-
ucts and phenols in volcanic ash-derived soils of Chile, Soil Biol.
Biochem., 14, 37–43, 1982.
Biogeosciences, 2, 159–174, 2005
www.biogeosciences.net/bg/2/159/