Content uploaded by Mari Könönen
Author content
All content in this area was uploaded by Mari Könönen on Jul 07, 2016
Content may be subject to copyright.
ORIGINAL PAPER
Land use increases the recalcitrance of tropical peat
M. Ko
¨no
¨nen .J. Jauhiainen .R. Laiho .P. Spetz .
K. Kusin .S. Limin .H. Vasander
Received: 20 November 2015 / Accepted: 23 June 2016
ÓSpringer Science+Business Media Dordrecht 2016
Abstract Tropical peat carbon compound composi-
tion (CCC) is a highly understudied subject. Advanced
understanding of peat CCC and carbon dynamics in
differing conditions is desperately needed due to large-
scale utilization of these peatlands. We studied the
CCC—i.e. the hemicellulosic carbohydrate and uronic
acid composition and concentrations of extractives,
cellulose, acid-soluble lignin and acid-insoluble lig-
nin—in association with peat profile depth and phys-
ical structure of peat, under representative, common
land uses. Samples were gathered from an undrained
forest and three sites altered 20–30 years prior to the
study, which in aggregate form a continuum of
increasing land-use intensity (drainage-affected forest;
drained and deforested degraded open site; drained and
deforested site under cultivation) in Central Kaliman-
tan, Indonesia. Peat samples were taken from depths
between 10 and 115 cm that covered mostly oxic,
frequently waterlogged and permanently waterlogged,
anoxic conditions. Our results demonstrated greater
modification of peat properties when both vegetation
and hydrological conditions were altered. The differ-
ences between sites were mainly present in the topmost
peat and decreased with depth. Peat located at the
surface contained more labile compounds (hemicellu-
loses, extractives, uronic acids, cellulose) on forest
sites than at the most intensively altered open sites,
where peat was enriched with recalcitrant acid insol-
uble lignin. The effect of drainage was evident in the
drained forest site, where at the approximate median
water table depth peat more closely resembled open
sites in terms of the peat properties. The increased
recalcitrance of peat in reclaimed areas has been a
result of enhanced decomposition, reduced litter input
rates and, at open sites also by repeated fires.
Keywords Tropical peat Land-use
Decomposability Carbohydrates Polymers
Introduction
The carbon compound composition (CCC) of organic
matter determines the amount and quality of energy
Electronic supplementary material The online version of
this article (doi:10.1007/s11273-016-9498-7) contains supple-
mentary material, which is available to authorized users.
M. Ko
¨no
¨nen (&)J. Jauhiainen H. Vasander
Department of Forest Sciences, University of Helsinki,
Latokartanonkaari 7, P.O. Box27, 00014 Helsinki, Finland
e-mail: mari.kononen@helsinki.fi
R. Laiho
Natural Resources Institute Finland, Kaironiementie 15,
39700 Parkano, Finland
P. Spetz
Natural Resources Institute Finland, Jokiniemenkuja 1,
01370 Vantaa, Finland
K. Kusin S. Limin
Center for International Cooperation in Sustainable
Management of Tropical Peatland (CIMTROP),
University of Palangkaraya, Palangkaraya, Indonesia
123
Wetlands Ecol Manage
DOI 10.1007/s11273-016-9498-7
available for decomposers (i.e. decomposability), and
is thus an important factor regulating decomposition
(Berg 2000; Bragazza et al. 2007; Zhang et al. 2008;
Sanaullah et al. 2010; Strakova
´et al. 2011). In
particular, the amount of relatively labile C com-
pounds, such as hemicelluloses and cellulose, and
their ratio to more decomposition-resistant (recalci-
trant) polymers, such as lignin, together with nitrogen
(N) and phosphorus (P) concentrations influence the
rate at which decomposition progresses (Berg 2000;
Rejmankova 2001; Strakova
´et al. 2011; Talbot and
Treseder 2012). Organic matter CCC generally firstly
depends on the characteristics of the parent material,
e.g., litter type (woody, roots, leaf, branches, etc.), and
later on the decomposition stage (Berg 2000). Peat is
organic matter formed mainly of partly decomposed
plant litter. Woody, carbon (C)-rich peat is formed in
tropical ombrotrophic forested swamps in Southeast
Asia from incompletely decomposed above- and
below-ground vegetation mainly originated from trees
(Page et al. 1999;Wu
¨st et al. 2008; Hoyos-Santillan
et al. 2015). These thick peat deposits are significant C
stores containing 11–14 % of all C stored in peat soils
globally (Page et al. 2011). However, large-scale land
reclamation including drainage and deforestation has
taken place during the past few decades (Miettinen
et al. 2016). Land reclamation has drastically influ-
enced peat decomposition processes and the quantity
and quality of organic matter inputs leading to
enhanced peat loss through microbial decomposition
and wildfires (Hoscilo et al. 2011; Hooijer et al. 2012;
IPCC 2014), and thus land reclamation is the largest
modifier of peat CCC. An understanding of the impact
of various anthropogenic activities (e.g. drainage,
deforestation, fire, agriculture) on CCC is needed to
understand and predict soil organic matter stability in
tropical peat soils.
Peat is generally formed when the rate of decom-
position does not exceed the litter input rate. Decom-
position in peat soils is regulated by abiotic
environmental factors such as oxygen, moisture,
nutrient availability and pH (Rieley and Page 2005;
Laiho 2006). In an undisturbed peat swamp forest,
there is a typically high soil water-table level (WL).
This creates anoxic conditions in the peat below the
water level where decomposition is slower than the
aerobic decomposition, which takes place in oxic
conditions above the water table. Thus peat accumu-
lates over time (Clymo 1984). These abiotic factors
determine to what extent the decomposability of the
litter inputs or the accumulated peat is realized as
actual decomposition through microbial enzymatic
activity (Freeman et al. 2001; Sjo
¨gersten et al. 2011),
and they may be greatly altered by land use.
Land use reclamation in tropical peatlands usually
includes drainage and deforestation. Drainage lowers
the WL, allowing an increasing potential for aerobic
decomposition deeper in the peat profile. The physical
structure of peat changes due to advancing decompo-
sition, which leads to increased peat dry bulk density
(BD) and decreased porosity (Hooijer et al. 2012;
Ko
¨no
¨nen et al. 2015). Deforestation lowers litter input
rates and alters litter quality, since reclaimed land
uses, such as plantations, cultivated and fallow lands,
typically have lower vegetation cover and biomass
production and less variation in species (crops and
monocultures) than the original peat swamp forests
(Sulistiyanto 2004; Hoscilo et al. 2011; Mehta et al.
2013; Blackham et al. 2014; Yule et al. 2016).
Reclaimed, unmanaged peatlands are vulnerable to
wildfires, and up to several decimetres of peat can be
lost in recurring wildfires (Hoscilo et al. 2011;
Konecny et al. 2016), which results in exposure of
aged peat from the deeper layers and keeps the
vegetation cover low. All these changes in turn likely
contribute to changes in peat CCC.
Despite the increasing utilization of tropical peat-
lands (Miettinen and Liew 2016; Murdiyarso et al.
2010), there is almost no information available
concerning tropical peat CCC from any land-use type.
As far as we know, tropical peat CCC has previously
been reported only by Andriesse (1988). Yet, the
connection between the land-use type and peat CCC
should be understood to minimize both the loss of
substrate and the release of greenhouse gases (GHG).
Here we conduct a detailed analysis on tropical peat
CCC in different land-use types in association with
depth in peat profile and peat physical structure. We
established study sites at both (i) an undrained peat
swamp forest, and at three sites altered by human
activities. The altered sites included in order of
increasing land-use intensity: (ii) a drained forest,
and two sites at clear-cut open peatlands, i.e. (iii) a
drained site burned several times and (iv) a cultivated
site under controlled drainage. To determine the peat
CCC we analysed the concentrations of C, N, neutral
and acidic sugars in non-cellulosic polysaccharides
(hemicellulose), cellulose, extractives, acid-soluble
Wetlands Ecol Manage
123
lignin (ASL) and acid-insoluble lignin (AIL). AIL
forms the most recalcitrant portion of organic matter
and is often referred to as Klason’s lignin. The CCC
data were combined with the peat physical structure
data, defined both by peat BD and by the proportion of
different sized particles forming the peat, as an
indication of the physical decomposition stage of the
peat.
We hypothesized that (i) the impact of land use on
CCC is greatest when both WL and vegetation are
altered, (ii) the differences between land-use types are
greatest in the topmost peat, (iii) the differences
between land-use types can be seen as an increase in
the amount of resistant biopolymers in the peat and
(iv) the peat physical properties implying decompo-
sition stage will correlate with the peat CCC/recalci-
trance level of the peat.
Materials and methods
Study area and sites
The study area was an ombrotrophic lowland peatland
in Central Kalimantan, Indonesia (2°200S, 113°550E).
The reported mean annual temperature in the area was
26.2 ±0.3 °C and the mean annual rainfall was
2540 ±596 mm year
-1
between 2002 and 2010
(Sundari et al. 2012). Climatically, there are two rainy
seasons per year divided by one dry season typically
occurring from June through to the end of August.
We established four study sites located within a
10-km radius in the same watershed of the Sebangau
River. All sites can be assumed to originate from
similar ombrotrophic peat swamp forests, where the
peat depth exceeded 3 m (Fig. 1). The original
Fig. 1 Study area and
sampling sites in Central
Kalimantan, Indonesia
(2°200S, 113°550E)
Wetlands Ecol Manage
123
hydrology at three of the sites was altered due to
human activities, while the vegetation was also altered
at two sites. Two of the sites had original, close to
steady state forest cover (comprising species of a
number of plant families, including Dipterocarpaceae
sp.), although selective logging had occurred in the
past in both areas (Page et al. 1999; Jauhiainen et al.
2005,2008). These forest sites are called undrained
(UF) and drained forest (DF). The hydrology of the UF
was close to natural, but a large drainage canal has
impacted the WL of the drained forest since 1997. Soil
surface microtopography of both forests was formed
of a mosaic of low peat surfaces (forming pools when
high WL) and hummocks around tree bases (Lampela
et al. 2014), although the microtopography was less
apparent in the DF. The other two sites were open due
to deforestation, relatively flat, and both were drained.
The first open site, the degraded site (DO), was
deforested and had been drained by the same large
canal as the drained forest in 1997. The surface at the
degraded site had burned several times (1997, 1999,
2002, and 2006) prior to sampling. The recurring fires
had removed more than 0.5 m of the topmost peat
(Page et al. 1999; Hoscilo et al. 2011), and the
topography was therefore relatively flat excluding
some deeper fire scars. The dominant vegetation at the
degraded site was formed of ferns, and scattered
shrubs and small trees. The second open site, called the
agricultural site (AO), had been under shallow,
controlled drainage and cultivated by smallholder
farmers since the 1980s, but it had been unutilised
fallow land prior to sampling. The agricultural site had
been fertilized during cultivation. Due to repeated soil
surface tillage and management fires the soil
topography was flat. Details of the average annual
long-term WL at the sites are shown in Table 1.
Soil sampling and physical fractioning
Peat samples were collected when annual WL was
deepest at the end of the dry season in September
2009. Samples in the forested sites were taken from
the low peat surfaces between trees and in the open
sites from the flat vegetation-free surfaces. Sampling
depths were 10–15, 40–45, 80–85 and 110–115 cm
below the soil surface. The topmost depth (10–15 cm)
represented surface peat that is mostly above WL and
thus oxic, while deeper peat (40–45 cm) was mostly
waterlogged and thus anoxic at all sites. The long-term
median WL of the UF was relatively high (Table 1)
and thus the conditions under permanent waterlogging
and consequent anoxia were reached at the depth of
80–85, which formed the deepest sampling depth at
the site. The WL of the uncontrollably drained sites
has been deeper more frequently, and the 80–85 cm
depth provided mostly waterlogged and 110–115 cm
depth provided permanently waterlogged conditions.
Where present, recently deposited litter on the soil
surface (e.g. loose cover of leaves or branches) was
removed before sampling. The exposed soil surface
was marked as the 0 cm depth. Samples at all sites
were taken from one wall of a dug pit hole immedi-
ately after reaching each sampling depth. At the end of
sampling the pit holes were approximately
100 9100 cm wide and 150 cm deep. We cut six
samples from each depth using a sharp knife and a
volume-exact frame (10 910 cm wide and 5 cm
high). All samples (total n =90) were sealed in plastic
Table 1 Water table level characteristics at the study sites
UF DF DO AO
Period 1997–2006 2005–2007 2004–2008 2001–2003
n 2733 1231 1421 355
Average -18.3 -50.0 -26.6 -25.7
Median -10.3 -38.6 -16.8 -22.0
Maximum 33.4 1.7 20.2 -5.2
Minimum -60.8 -177.6 -160.0 -71.9
First quartile -30.7 -58.7 -32.3 -30.2
Third quartile 3.1 -27.2 -8.3 -16.6
Negative values indicate depth in centimetres below the soil surface and positive values above the soil surface. Period indicates the
observation period in years, and n indicates the number of observations based on diurnal means
Wetlands Ecol Manage
123
bags and stored in a refrigerator (4 °C) for maximum
of 1 week prior to analyses.
Living roots were removed from all samples. The
samples from each site and depth were divided into
two subsets (n =3 per subset). One sample subset was
dried at 70 °C to determine the dry bulk density
(g cm
-3
, BD). This sample type is called ‘bulk soil’.
The other sample subset was fractioned gently with
running water and two sieves (mesh [1.5 mm and
0.15 mm) to describe the physical decomposition
stage. The physically least decomposed matter cap-
tured by the larger mesh is called a ‘woody’ sample,
and the more decomposed matter, captured by the
smaller mesh, is called a ‘fibric’ sample. Both sieved
fractions were dried at 70 °C and their proportion of
the mass of the corresponding bulk soil samples was
calculated. A summary of the physical structure
characteristics is presented in Table 2, while the
detailed data is reported in Ko
¨no
¨nen et al. (2015).
The proportions of wood and fibre were calculated
from the average BD of the samples from the same site
and depth. The finer the material forming the peat and
the higher the BD, the higher the physical decompo-
sition stage was considered to be.
Chemical analyses
First all samples of the same type (bulk soil, fibric,
woody) sampled from the same depth and site were
pooled and milled to fine powder. Loss in mass on
ignition (LOI, 550 °C for 4 h) was measured. Con-
centrations of total C and N were determined (LECO
CHN-1000). All these analyses were performed using
three laboratory replicates.
We measured several organic matter quality
parameters to examine the CCC of all the fractions.
The parameters were hemicellulose (including pectin),
cellulose, extractives, ASL and AIL. All analyses were
performed using two laboratory replicates.
The carbohydrate composition of hemicellulose and
cellulose concentrations were analysed from
10 ±2 mg of dry sample. The total hemicellulose
concentration was calculated by adding together the
non-cellulosic polysaccharide concentrations formed of
neutral sugars (arabinose, rhamnose, xylose, mannose,
galactose, glucose) and uronic acids (glucuronic,
galacturonic and 4-O-Me-glucuronic acid), which were
determined by acid methanolysis followed by gas
chromatography (GC). The cellulosic glucose concen-
tration was determined by subtracting the concentration
of non-cellulosic glucose determined by acid methanol-
ysis from the total glucose concentration. The total
glucose concentration was determined by acid hydrol-
ysis and silylation followed by GC. The methods for
methanolysis and hydrolysis are described in Sundberg
et al. (1996,2003), respectively.
Concentrations of extractives and lignin-derived
compounds were determined using gravimetric
Table 2 Summary of peat physical structure based on data published in Ko
¨no
¨nen et al. (2015)
Site Depth (cm) BD (g cm
-3
) Wood (% of the sample) Fibre (% of the sample)
UF 10–15 0.13 ±0.01 22.8 ±4.0 48.3 ±8.4
40–45 0.14 ±0.01 5.8 ±1.2 46.1 ±2.8
80–85 0.15 ±0.03 5.8 ±1.7 15.8 ±2.5
DF 10–15 0.17 ±0.03 9.2 ±1.5 65.1 ±9.2
40–45 0.22 ±0.03 2.0 ±1.0 49.0 ±9.1
80–85 0.15 ±0.02 7.2 ±5.1 26.7 ±1.8
110–115 0.12 ±0.01 19.1 ±7.9 36.5 ±5.9
AO 10–15 0.18 ±0.01 4.3 ±2.2 37.8 ±1.1
40–45 0.17 ±0.01 9.9 ±1.0 37.2 ±4.3
80–85 0.13 ±0.01 6.8 ±2.4 32.0 ±50
110–115 0.13 ±0.00 17.3 ±12.0 32.1 ±9.7
DO 10–15 0.20 ±003 2.7 ±1.8 14.1 ±0.5
40–45 0.12 ±0.03 12.1 ±1.3 34.1 ±2.1
80–85 0.14 ±0.02 18.2 ±14.4 30.7 ±6.3
110–115 0.15 ±0.02 15.0 ±6.4 30.2 ±3.6
Wetlands Ecol Manage
123
methods. The concentration of extractives was deter-
mined by mixing 1.2 g of the dry sample with 20 ml of
an acetone:water solution (v:v, 9:1). The mixture was
then sonicated (45 min), filtered (through a no. 4
Pyrex) and the residual was dried (24 h, 105 °C). The
mass loss resulted from dissolving indicated the
amount of extractives in the sample. AIL concentra-
tion was determined from 0.3 g of the extractive-free
sample with a standard procedure (TAPPI Test
Method T 222-om-88, 2000). ASL concentration
was determined from the solution produced as a by-
product when determining AIL using UV spec-
troscopy (Shimadzu UV-2401 PC UV–VIS Recording
Spectrophotometer) at an absorbance of 203 nm
(Brunow et al. 1999). These methods are originally
developed for wood fibres, but have also been
successfully applied to organic soils (Merila
¨et al.
2010; Strakova
´et al. 2010). The chemical character-
ization using these methods is discussed further in
Merila
¨et al. (2010). All concentrations were
expressed as mg g
-1
sample dry mass.
Data analyses
Unconstrained principal component analysis (PCA)
using the determined C compounds as response
variables, and the environmental variables (site,
sampling depth, sample type, and physical structure)
as passive explanatory variables was first conducted to
summarize the variation in peat CCC. Next, con-
strained redundancy analysis (RDA) with variation
partitioning was conducted to test the effect of site,
depth and physical structure and to analyse how much
of the total variation (inertia) was explained by each
environmental variable. This RDA was conducted
with a reduced data set, which included only the
topmost three depths and bulk soil as the only sample
type, to reduce the influence of the lack of the fourth
depth from the undrained site, and to examine the
unique effects of the environmental variables on bulk
soil. Next, to test the sample type effect on peat CCC,
and to further test hypothesis iv, a partial RDA
analysis was conducted with the full data set using
sample type as the explanatory variable, and depth and
site as the covariates. To further test the effect of
sample type a ttest between all samples of the same
type was conducted separately for each measured
variable. Finally, RDA was conducted with interac-
tively chosen explanatory variables, to pinpoint the
best explaining single environmental factors. All
RDAs were followed by Monte Carlo permutation
tests with 999 permutations. All ordination tests were
conducted with Canoco5 for Windows and the t-tests
were conducted with RStudio version 0.98.1102. The
p-value significance level used was 0.05. Otherwise,
due to limitations arising from a relatively small data
pool for any advanced statistical analyses, the main
focus was on comparisons of the calculated means of
the measured variables from different sites and depths.
Results
General patterns in peat carbon compound
composition
In PCA, the strongest gradient in the CCC data, PC1,
separated samples with high AIL (i.e. acid insoluble
lignin) from samples with high hemicellulosic carbo-
hydrates and uronic acids, and explained 42.8 % of the
total variation in CCC. PC2 separated samples rich in
extractives and ash from samples rich in cellulose, and
explained 16.5 % of the total variation in CCC
(Fig. 2). Concentrations of hemicelluloses, uronic
acids and ASL increased towards the UF and two
topmost depths of DF, which were separated from the
deeper depths of DF and the open sites by PC1. Ash
and extractives concentrations increased towards the
80–85 cm depth of DF and two topmost depths of the
degraded site, which were separated from their deepest
depths and all depths of the agricultural site and AIL
concentration by PC2. Of the physical characteristics,
the proportion of wood fraction correlated positively
with cellulose, and BD correlated positively with
extractives and ash. The proportion of fibre fraction
correlated with hemicellulosic carbohydrates and
uronic acids. Of the sample types, the bulk soil and
fibric sample type were separated from the woody
sample type by PC1. Extractives, ASL and AIL
pointed towards, i.e., were higher in, the bulk soil and
fibric sample, and cellulose, hemicelluloses and uronic
acids pointed towards the woody sample type. RDA
with variation partitioning showed that the site
accounted for 27.3 % (p =0.056), depth 9.7 %
(p =0.124) and physical structure (BD and wood
and fibre proportion) 23.6 % (p =0.066) of all the
variation in peat CCC. The best single environmental
factors explaining variation in peat CCC were the
Wetlands Ecol Manage
123
woody sample, UF, fibre proportion and BD, which
explained 23.1, 19.1, 14.1 and 4.4 %, respectively.
The carbon compound composition of bulk soil
in relation to site, depth and physical structure
The average concentrations of C, N, extractives, total
hemicellulose, and ASL were generally higher (2.2,
50.7, 37.8, 646.6 and 37.8 %, respectively) in the bulk
soil of forest sites while concentrations of, ash,
cellulose, and AIL were lower (25.6, 3.6 and
11.3 %, respectively) than at the open sites (Figs. 3,
4; Table 3). The CCC of both undrained and drained
forests was similar in the surface peat (10–15 cm
depth), but at the depth of 40–45 cm and deeper the
peat properties in drained forest and open sites were
increasingly similar, reflecting a higher physical
decomposition stage than in the UF (Tables 2,3;
Fig. 3).
All sites showed a decreasing trend in N concentration
and increasing trends in C concentration and CN-ratio
with increasing peat depth. The hemicellulosic carbohy-
drate concentration of bulk soil was higher in the forests
than in the open sites, but always contained compounds
in the following order: glucose [xylane [man-
nose [galactan [galacturonic acid [arabinose [
rhamnose [glucuronic acid [4-O-Me-glucuronic acid
(Fig. 3;Table4). The total hemicellulose concentration
in the bulk soil of the forest sites usually showed a clear
decreasing trend with depth. At the open sites the total
hemicellulose concentrations in bulk soil was on average
87 % smaller than at the forest sites, and the differences
in concentrations between sites decreased with depth.
The concentration of extractives in bulk soil had a
decreasing trend with depth at the degraded site, and an
increasing trend in the forests and agricultural site
(Fig. 3;Table3). Furthermore, the concentration of
extractives was highest in the drained forest at the depth
of 40–45 cm. At all sites, the cellulose concentration and
ASL decreased and AIL increased with depth.
Generally, fibres comprised 14.1–65.1 % and
woody matter 2.0–22.8 % of the dry mass of bulk soil
(Table 2). Of the physical structure, BD correlated
negatively with the concentrations of cellulose, hemi-
celluloses and ASL, while proportions of wood and
fibres correlated positively with them (Fig. 5). BD
correlated positively with the concentrations of AIL,
ash and extractives.
The carbon compound composition in relation
to sample type
The CCC of fibric samples resembled bulk soil and
responded to depth and site in the same way, but in all
measured variables the woody sample was signifi-
cantly(p \0.05) different (Figs. 3,6). This was in line
with the physical decomposition stage, which was
generally higher in the bulk soils and the fibric sample
type than in the woody sample type at all sites and
depths (Table 2). On average the bulk soil and fibric
samples contained 33.8 % more extractives, 12.7 %
more AIL and 68.4 % more ASL than the woody
samples (Table 5). The woody samples contained on
average over 4.5 times more cellulose and nearly 3
times more total hemicellulose than the other sample
types. The concentrations of hemicellulosic
Fig. 2 Results of the unconstrained PCA. The first principal
component, PC1 (x-axis) explained 42.8 %, and PC2 (y-axis)
explained 28.4 % of the total CCC variation. The response
variables indicated by black arrows were: ASL acid-soluble
lignin, hemicelluloses, uronic acids, extractives, cellulose and
AIL. Site, sampling depth, sample type, and physical structure
(BD and proportion of wood and fibres) were used as passive
explanatory variables. Site: UF undrained forest, DF drained
forest, DO degraded open site, AO agricultural open site,
number following dash indicates sampling depth (e.g.
UF*10 =undrained forest at a depth of 10 cm from the soil
surface). Sample type: bulk soil, fibric sample (between 0.15 and
1.5 mm; fibric) and woody sample ([1.5 mm; woody). Physical
structure: dry bulk density (g cm
-3
)BD,wood % volumetric
proportion of wood, fibre % volumetric proportion of fibres
Wetlands Ecol Manage
123
Wetlands Ecol Manage
123
carbohydrates, especially the concentrations of man-
nose, glucose, xylane, arabinose and galacturonic acid
were significantly higher in the woody samples
throughout the whole sampling profile than in the
fibric samples or bulk soil (Fig. 6). On average, C
concentration was 57 % and N less than 1 % in all
sample types. N concentration was the lowest in the
woody samples, resulting in an average 30 % higher
CN-ratio than other sample types. Generally the CN-
ratio increased with depth at all sample types, but the
increase was remarkable in woody samples, in which
the CN-ratio was as high as 111–167 at the depth of
110 cm (Fig. 4).
Discussion and conclusions
This study provides strong evidence that long-term
land use involving drainage (with or without defor-
estation) leads to the enrichment of recalcitrant C
compounds in peat. This observation is in line with
our hypothesis (iii). Although peat in the UF was
also rich in recalcitrant AIL, it is notable that the
concentrations of labile hemicelluloses and uronic
acids declined especially when both vegetation and
WL had been altered. Our results thus support the
hypothesis (i) that land-use intensity greatly impacts
peat CCC. The differences between land-use types
were greatest at the surface peat down to a depth of
40–45 cm, which supports our hypothesis (ii). The
concentrations of total hemicellulose (including
pectin) and cellulose were higher in the physically
less decomposed bulk soil of surface peat
(10–15 cm) of the forest sites than in the open sites
and in the least physically decomposed woody
sample type at all sites. AIL concentrations were
highest in the physically more decomposed bulk soil
of the surface peat in intensively altered open sites
and in the fibric samples compared to the woody
samples. This supports our fourth hypothesis (iv).
Both vegetation, providing litter supply with a
higher concentration of labile carbohydrates and
decomposition, of the litter and of peat, contribute to
peat CCC structure. Originally the sites were woody
peat-accumulating forest-covered swamps, but peat
formation and decomposition capacity at the
reclaimed sites has been altered in the past. The AIL
content was generally very high in the bulk soil (on
average 77 % of the dry mass). AIL in this study is
comparable to concentrations of lignin measured by
Andriesse (1988), who reported relatively high lignin
concentrations (64–76 %) for tropical peat in com-
parison to boreal Sphagnum peat (18 %) and woody
bFig. 3 CCC of the different sample types in tropical peat. The
right-hand paragraph cumulatively presents the concentrations
of AIL, ASL, extractives, hemicelluloses and uronic acids and
cellulose. Cellulose concentration of wood sample from a depth
of 110 cm is lacking from DF. The left-hand paragraph
cumulatively presents the hemicellulosic carbohydrates (Ara
arabinose, Glc glucose, Gal galactan, GalA galaturonic acid,
GlcA glucuronic acid, Man =mannose, Rha rhamnose, Xyl
xylane, 4-O-me =4-O-methyl glucuronic acid). Ash concen-
tration varied between 0.22 and 1.1 mg g
-1
. Laboratory
analyses were performed with two replicates (n =2). Site: UF
undrained forest, DF drained forest, DO degraded open site, AO
agricultural open site
Fig. 4 The C and N concentrations and CN-ratio for all sample
types, depths and sites. At each depth the observations of woody
samples are presented on the top, the non-fractioned samples in
the middle and the fibric sample at the bottom. Site: UF
undrained forest, DF drained forest, DO degraded open site, AO
agricultural open site
Wetlands Ecol Manage
123
sedge peat (38 %). The relatively high AIL content in
tropical peat is mostly due to the woody origin of the
peat matter (Yule and Gomez 2009; Mehta et al.
2013). In this study, it is shown that the peat CCC
correlates also with the physical decomposition stage,
expressed as BD and proportions of different sized
particles in the bulk soil. Yet, the CCC of bulk soil
cannot be estimated directly from the proportions of
different sized fractions, because the CCC of the
fractions differed between land-uses. However, in all
land-use types the recalcitrance of bulk soil and fibric
samples was higher than in the woody sample type.
This is because during the decomposition process the
physical structure of the peat becomes finer, and
microbes deplete the most labile energy sources
relatively faster from the peat substrate leading to
the enrichment of more resistant compounds such as
AIL (Berg 2000; Hooijer et al. 2012;Ko
¨no
¨nen et al.
2015). Subsequently, peat CCC determines the
decomposability of the residual organic matter, and
thus the amount of GHG emissions released from peat.
Reduced peat decomposability may have contributed
to the lower CO
2
emissions released in oxidation at
degraded peatlands as compared to forests (Jauhiainen
et al. 2008; Hirano et al. 2014; IPCC 2014).
The observed differences in peat CCC (higher AIL
and lower hemicellulose concentrations in the
reclaimed sites) were greatest between forests and
open sites, but drainage also independently affected
peat CCC. In both forest-covered peatlands the litter
decomposition rate depends on both the quality and
position of the litter in the peat profile (Hoyos-
Santillan et al. 2015). Root litter is slower to decom-
pose in comparison to leaves and stems, and their
location deeper in peat profile (often in waterlogged
conditions) makes them a major contributor to peat
accumulation (Page et al. 1999;Wu
¨st et al. 2008;
Hoyos-Santillan et al. 2015). Forest vegetation is also
an important component in the nutrient cycle, because
nutrients in ombrotrophic peatlands mainly bind to
plant biomass and are released during the early phase
of decomposition (Aerts et al. 1999; Rieley and Page
2005). In the topmost peat of the forests the conditions
for decomposition are thus more favourable due to the
higher availability of nutrients and easily decompos-
able substrates, implied in our study as high total
hemicellulose and N concentrations, a CN-ratio close
to 30 and mostly oxic conditions. The easily decom-
posable organic C compounds therefore become
depleted mostly in the topmost peat leading to the
Table 3 The ash content and C compound composition of non-fractioned samples from every site and depth
Site Depth
(cm)
Ash
a
(mg g
-1
)
Extractive
(mg g
-1
)
Total hemicelluloses
(mg g
-1
)
Cellulose
(mg g
-1
)
ASL
(mg g
-1
)
AIL
(mg g
-1
)
UF 10–15 0.60 ±0.04 131.3 ±4.0 56.7 ±5.8 39.3 ±11.8 17.7 ±0.9 683.6 ±14.7
40–45 0.36 ±0.03 135.0 ±3.6 37.2 ±3.9 16.8 ±1.1 19.1 ±4.0 708.6 ±10.8
80–85 0.63 ±0.03 290.9 ±55.9 5.4 ±1.0 14.8 ±0.9 4.8 ±5.2 717.5 ±35.1
DF 10–15 0.61 ±0.08 111.6 ±2.3 59.5 ±7.3 25.7 ±1.8 20.0 ±0.0 674.1 ±20.3
40–45 0.39 ±0.02 380.4 ±11.8 9.3 ±0.3 5.7 ±1.4 4.6 ±1.8 671.0 ±18.2
80–85 0.43 ±0.04 220.9 ±9.0 1.0 ±0.0 12.0 ±0.2 5.1 ±3.1 751.0 ±29.4
110–115 0.32 ±0.06 130.8 ±2.1 2.7 ±0.6 28.8 ±11.3 1.1 ±0.5 811.0 ±14.8
DO 10–15 1.10 ±0.05 207.8 ±17.0 1.0 ±0.6 6.7 ±0.5 12.9 ±6.0 764.5 ±29.2
40–45 0.75 ±0.04 167.2 ±17.5 1.9 ±0.8 17.4 ±0.9 14.3 ±0.5 774.1 ±3.9
80–85 0.62 ±0.36 132.1 ±3.0 1.0 ±0.1 28.7 ±12.7 6.2 ±3.3 808.9 ±1.2
110–115 0.60 ±0.02 129.9 ±12.8 1.2 ±0.7 22.9 ±2.2 5.0 ±1.7 816.2 ±4.5
AO 10–15 0.88 ±0.01 118.5 ±3.5 2.6 ±0.0 19.7 ±0.3 2.9 ±0.0 863.1 ±10.2
40–45 0.73 ±0.03 103.5 ±5.0 8.4 ±9.2 18.7 ±0.9 1.8 ±0.2 820.4 ±5.6
80–85 0.22 ±0.01 173.7 ±8.8 5.0 ±0.5 24.3 ±0.7 3.8 ±0.1 815.8 ±1.0
110–115 0.23 ±0.01 129.6 ±7.6 5.2 ±1.6 31.2 ±1.4 6.1 ±0.8 797.6 ±6.0
Values are mean ±SD of two laboratory replicates
a
Proportion of ash published previously in Ko
¨no
¨nen et al. (2015)
Wetlands Ecol Manage
123
Table 4 The hemicellulosic carbohydrate composition of non-fractioned samples from every site and depth
Site Depth
(cm)
Mannose
(mg g
-1
)
Glucose
(mg g
-1
)
Galactan
(mg g
-1
)
Xylane
(mg g
-1
)
Arabinose
(mg g
-1
)
Rhamnose
(mg g
-1
)
Glucuronic acid
(mg g
-1
)
Galacturonic acid
(mg g
-1
)
4-O-methyl glucuronic
acid (mg g
-1
)
UF 10–15 6.0 ±0.4 21.4 ±2.5 6.5 ±0.4 9.7 ±0.6 2.6 ±0.2 3.2 ±0.2 1.2 ±0.6 5.3 ±0.6 0.7 ±0.2
40–45 5.3 ±0.5 19.8 ±2.6 3.7 ±0.4 3.4 ±0.3 1.7 ±0.4 1.3 ±0.1 1.0 ±0.4 1.1 ±0.0 0.0 ±0.0
80–85 0.9 ±0.1 3.0 ±0.5 0.1 ±0.2 1.3 ±0.3 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0
DF 10–15 8.2 ±0.4 3.4 ±2.5 6.9 ±0.9 6.1 ±1.2 1.7 ±1.3 1.8 ±0.1 0.0 ±0.0 1.5 ±0.9 0.0 ±0.0
40–45 2.1 ±0.0 6.2 ±0.1 0.7 ±0.0 0.3 ±0.5 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0
80–85 0.0 ±0.0 1.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0
110–115 0.0 ±0.0 2.7 ±0.6 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0
DO 10–15 0.3 ±0.5 0.7 ±0.2 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0
40–45 0.3 ±0.5 1.5 ±0.3 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0
80–85 0.0 ±0.0 1.0 ±0.1 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0
110–115 0.0 ±0.0 1.2 ±0.7 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0
AO 10–15 0.0 ±0.0 1.6 ±0.0 0.0 ±0.0 0.0 ±0.0 1.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0
40–45 0.5 ±0.8 1.6 ±0.6 0.0 ±0.0 5.8 ±8.2 0.0 ±0.0 0.1 ±0.1 0.0 ±0.0 0.5 ±0.7 0.0 ±0.0
80–85 0.9 ±0.1 4.1 ±0.4 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0±0.0 0.0 ±0.0 0.0 ±0.0
110–115 0.6 ±0.0 3.3 ±0.3 0.0 ±0.0 0.0 ±0.0 1.3 ±1.9 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0 0.0 ±0.0
Values are mean ±SD of two laboratory replicates
Wetlands Ecol Manage
123
enrichment of more recalcitrant compounds deeper in
the peat over time, where decomposition is more
restricted also due to more frequent waterlogging and
anaerobic conditions. In drained forest the surface peat
CCC is maintained by continuous litter deposition in a
similar manner to the UF, but in the drained forest the
effects of enhanced drainage were seen especially at
the depth of 40–45 cm, where the peat CCC began to
resemble corresponding depths at the open sites. At the
40–45 cm depth, with declining influence of litter
input, the peat properties in drained forest are greatly
impacted by improved conditions for aerobic decom-
position and compaction of peat after WL drawdown
(Laiho 2006; Hooijer et al. 2012; IPCC 2014). It has
also been suggested that vertical and horizontal water
level movements can relocate finer particles (Lampela
et al. 2014), which could have contributed to the
enrichment of finer particles and the high concentra-
tions of extractives detected near the WL median (at
approximately -40 cm depth) in the drained forest.
At open sites, deforestation, drainage and fires have
modified peat CCC properties, and resulted in a low N
content and a high CN-ratio and AIL content seen
especially in the topmost peat. Deforestation had
permanently reduced the vegetation biomass and litter
deposition rate (Sulistiyanto 2004; Mehta et al. 2013;
Blackham et al. 2014), and due to this, recent litter
deposition now barely contributes to the formation of
new peat, and decomposition mainly happens in the
peat formed prior to deforestation and drainage. The
peat at the open sites has also been repeatedly fire-
affected, and especially at the degraded site peat
combustion has exposed older peat from a depth of
several decimetres (Hoscilo et al. 2011; Konecny et al.
2016). The combustion process in smouldering fires
also modifies organic matter into a more recalcitrant
form i.e. ‘black carbon’ (Knicker 2007 and references
within). The consequent decomposition together with
the formation of black carbon has likely contributed to
the AIL concentrations in the surface peat in the open
sites, where it was found to be 35 % higher than in
non-burnt peat at the forest sites. The slightly higher
concentrations of recalcitrant compounds at the agri-
cultural site than at the degraded site may be due to
increased microbial decomposition, which is an out-
come of both longer lasted drainage and active soil
management including fertilizer applications and
tilling, which aerates the topmost peat.
Fig. 5 Results of partial redundancy analysis (RDA) testing the
effect of peat physical structure on CCC. The concentration of
extractives, ash and AIL increased with dry BD, while the
concentrations of cellulose and hemicelluloses and uronic acids
were higher in bulk soils with high proportions of wood and
fibres. Ara arabinose, Extractives acetone: water extractives,
Cellulos cellulose, Glc glucose (hemicellulosic), Gal galactan,
GalA galaturonic acid, GlcA glucuronic acid, AIL acid-insoluble
lignin, Man mannose, Rha rhamnose, ASL acid soluble lignin,
Xyl xylane, 4-O-me 4-O-methyl glucuronic acid Fig. 6 Results of the partial RDA testing of the effect of sample
type to CCC. Variation of the measured chemical properties is
summarized by classifying it according to sample type. The
woody sample type differs from the bulk soil and fibric sample
type. Site: UF undrained forest, DF drained forest, DO degraded
open site, AO agricultural open site
Wetlands Ecol Manage
123
Peat CCC determines the decomposability of peat
and thus will influence to the rate of release of gaseous
C emissions. Several studies have highlighted the
importance of substrate quality in controlling decom-
position processes (Berg 2000; Bragazza et al. 2007;
Merila
¨et al. 2010; Strakova
´et al. 2011). Thus, the
results of our study partly explain why the manage-
ment intensity and time from land-use change have
been observed to reflect the amount of GHG emissions
released from the decomposition. Lower gaseous C
emissions have been reported in the most intensively
altered land-use types, which have been both drained,
deforested and burned, and where litter feed from the
vegetation is low, than from land-use types with
substantial vegetation (i.e. oil palm, acacia) (Couwen-
berg et al. 2010; Carlson et al. 2013; Hirano et al.
2014; IPCC 2014). However, regardless of land-use
intensity, the average C losses are high a few years
after land reclamation and then decrease with time
(Hooijer et al. 2012 and references within). In
reclaimed lands this reduction in C loss with time is
likely due to the progressive enrichment of recalcitrant
compounds in the surface peat under conditions where
litter supply is limited, which together with the altered
environmental conditions (i.e. drought, nutrient lim-
itations, high temperature), have led to reduced rates
of decomposition and gaseous C emissions from the
most intensively managed sites. In undisturbed peat
swamp forest, where litter input provides large
continuous supply of labile carbohydrates and nor-
mally no droughts occur, CO
2
emissions can be higher
than at intensively altered sites (Sundari et al. 2012;
Hirano et al. 2014). However, despite the higher level
of CO
2
emissions, the undisturbed forest supports litter
production and thus preserves existing peat deposits
and indicates the potential for the accumulation of new
peat.
Tropical soils are naturally poor in N and P, and our
study showed that with intensive management they
tend to become rich in recalcitrant compounds, which
limit decomposition and plant growth. In drained and
deforested peatlands, peat enriched with recalcitrant
substrates will continue to decompose especially if the
conditions for microbial activities are improved (e.g.
quality and amount of substrate, and nutrient moisture
and oxic conditions) (Dungait et al. 2012). Productive
cultivation in these nutrient limited soils requires
frequent fertilization (Andriesse 1988; Rieley and
Page 2005; Murdiyarso et al. 2010), which increases
labile N and P availability in peat. This may increase
the microbial activity and thus decomposition of
recalcitrant substrates (Dungait et al. 2012; Jauhiainen
et al. 2014; Comeau et al. 2016).
Drainage and deforestation largely alter peat decom-
posability and, despite the increased recalcitrance, in
intensively managed drained and deforested sites lead to
continuous net C loss from peat deposits. For the
existing large degraded areas the best option to reach
original C dynamics and in the end,hopefully, also peat
accumulation could be rewetting and reforestation.
Some drained areas support the economy by remaining
in production, subject to fertilizing and tilling, which
maintains continued C losses, but this is a trade-off
between productivity and carbon storage. Therefore, the
best sustainable and carbon–neutral management con-
dition for these fascinating ecosystems is to keep them
as undisturbed peat swamp forests.
Acknowledgments We thank Dr. Hidenori Takahashi and Dr.
Takashi Inoue for providing the data used in calculating average
groundwater depths of the study sites prior to 2004. We
acknowledge the Centre for International Cooperation in
Sustainable Management of Tropical Peatland (CIMTROP)
for providing facilities and assistance during field sampling. We
thank Satu Repo from the Natural Resources Institute Finland
Table 5 The average C compound composition of different sample types
Sample
type
Extractives
(mg g
-1
dry
mass)
Total
hemicelluloses and
Uronic acids (mg
g
-1
dry mass)
Cellulose
(mg g
-1
dry
mass)
Acid-soluble
lignin (mg
g
-1
dry mass)
Acid-
insoluble
lignin (mg
g
-1
dry mass)
C (%) N (%) CN (%)
Bulk soil 171.0 ±12.9
a
13.2 ±3.0
a
20.8 ±4.5
a
8.3 ±1.8
a
765.2 ±10.9
a
57.6 ±3.2
a
0.97 ±0.3
a
64.5 ±17.0
a
Fibres 162.0 ±25.9
a
14.7 ±5.4
a
24.5 ±2.3
a
11.0 ±2.7
a
758.7 ±7.6
a
58.0 ±2.3
a
0.95 ±0.4
a
69.5 ±21.6
a
Woody 124.0 ±8.0
b
53.1 ±3.2
b
128.6 ±5.1
b
5.7 ±1.0
b
676.3 ±10.2
b
55.9 ±1.4
b
0.70 ±0.3
b
95.9 ±39.0
b
Woody sample: particle size [1.5 mm; 0.15 mm\fibric sample \1.5 mm. Upper indexes denote statistical differences (p\0.05)
between the sample types. Values are mean ±SD of 30 observations for each sample type
Wetlands Ecol Manage
123
(Luke) and Marjut Wallner from the University of Helsinki for
their patience and expertise in the laboratory analyses. This
research was supported by the Academy of Finland -funded
‘Restoration Impact on Tropical Peat Carbon and Nitrogen
Dynamics’ -project (RETROPEAT), University of Helsinki
prize money for Peatlanders and the Jenny and Antti Wihuri
foundation.
References
Aerts R, Verhoeven J, Whigham D (1999) Plant-mediated
controls on nutrient cycling in temperate fens and bogs.
Ecology 80(7):2170–2181
Andriesse JP (1988) Nature and management of tropical peat
soils. FAO SOIL BULL 59. Food and agriculture organi-
zation of the United Nations, Rome. pp 179
Berg B (2000) Litter decomposition and organic matter turnover
in northern forest soils. Ecol Manag 133:13–22
Blackham G, Webb E, Corlett R (2014) Natural regeneration in
a degraded tropical peatland, Central Kalimantan,
Indonesia: implications for forest restoration. Ecol Manag
324:8–15
Bragazza L, Siffi C, Iacumin P, Gerdol R (2007) Mass loss and
nutrient release during litter decay in peatland: the role of
microbial adaptability to litter chemistry. Soil Biol Bio-
chem 39:257–267
Brunow G, Lundquist K, Gellerstedt G (1999) Lignin. In:
Sjo
¨stro
¨m E, Ale
´n R (eds) Analytical methods in wood
chemistry, pulping, and papermaking. Springer, Berlin,
pp 77–125
Carlson K, Curran L, Asner G, Pittman A, Trigg S, Marion
Adeney J (2013) Carbon emissions from forest conversion
by Kalimantan oil palm plantations. Nat Clim Chang
3:283–287
Clymo R (1984) The limits to peat bog growth. Philos Trans R
Soc Lond B 303:605–654
Comeau L-P, Hergoualc’h K, Hartill J, Smith J, Verchot L, Peak
D, Salim A (2016) How do the heterotrophic and the total
soil respiration of an oil palm plantation on peat respond to
nitrogen fertilizer application? Geoderma 268:41–51
Couwenberg J, Dommain R, Joosten H (2010) Greenhouse gas
fluxes from tropical peatlands in south-east Asia. Glob
Chang Biol 16:1715–1732
Dungait J, Hopkins D, Gregory A, Whitmore A (2012) Soil
organic matter turnover is governed by accessibility not
recalcitrance. Glob Change Biol 18:1781–1796
Freeman C, Ostle N, Kang H (2001) An enzymic ‘latch’ on a
global carbon store. Nature 409:149
Hirano T, Kusin K, Limin S, Osaki M (2014) Carbon dioxide
emissions through oxidative peat decomposition on a burnt
tropical peatland. Glob Chang Biol 20:555–565
Hooijer A, Page S, Jauhiainen J, Lee W, Lu X, Idris A, Anshari
G (2012) Subsidence and carbon loss in drained tropical
peatlands. Biogeosciences 9:1053–1071
Hoscilo A, Page S, Tansey K, Rieley J (2011) Effect of repeated
fires on land-cover change on peatland in southern Central
Kalimantan, Indonesia, from 1973 to 2005. Int J Wildland
Fire 20:578–588
Hoyos-Santillan J, Lomax B, Large D, Turner B, Boom A,
Lopez O (2015) Getting to the root of the problem: litter
decomposition and peat formation in lowland Neotropical
peatlands. Biogeochemistry 126:115–129
IPCC (2014) 2013 Supplement to the 2006 IPCC guidelines for
national greenhouse gas inventories: wetlands. In: Hiraishi
T, Krug T, Tanabe K, Srivastava N, Baasansuren J, Fukuda
M, Troxler TG (eds). IPCC, Geneva, p 353
Jauhiainen J, Takahashi H, Heikkinen J, Martikainen P,
Vasander H (2005) Carbon fluxes from a tropical peat
swamp forest floor. Glob Chang Biol 11:1788–1797
Jauhiainen J, Limin S, Silvennoinen H, Vasander H (2008)
Carbon dioxide and methane fluxes in drained tropical peat
before and after hydrological restoration. Ecology
89(12):3503–3514
Jauhiainen J, Kerojoki O, Silvennoinen H, Limin S, Vasander H
(2014) Heterotrophic respiration in drained tropical peat is
greatly affected by temperature—a passive ecosystem
cooling experiment. Environ Res Lett 9:18
Knicker H (2007) How does fire affect the nature and stability of
soil organic nitrogen and carbon? A review. Biogeo-
chemistry 85:91–118
Konecny K, Ballhorn U, Navratil P, Jubanski J, Page SE, Tansey
KJ, Hooijer A, Vernimmen R, Siegert F (2016) Variable
carbon losses for recurrent fires in drained tropical peat-
lands. Glob Change Biol 22(4):1469–1480
Ko
¨no
¨nen M, Jauhiainen J, Laiho R, Kusin K, Vasander H (2015)
Physical and chemical properties of tropical peat under
stabilised land uses. Mires Peat 16(8):1–13
Laiho R (2006) Decomposition in peatlands: reconciling
seemingly contrasting results on the impacts of lowered
water levels. Soil Biol Biochem 38:2011–2024
Lampela M, Jauhiainen J, Vasander H (2014) Surface peat
structure and chemistry in a tropical peat swamp forest.
Plant Soil 382:329–347
Mehta N, Dinakaran J, Patel S, Laskar A, Yadava M, Ramesh Y,
Krishnayya N (2013) Changes in litter decomposition and
soil organic carbon in a reforested tropical deciduous cover
(India). Ecol Res 28:239–248
Merila
¨P, Malmivaara-La
¨msa
¨M, Spetz P, Stark S, Vierikko K
(2010) Soil organic matter quality as a link between
microbial community structure and vegetation composi-
tion along a successional gradient in a boreal forest. Appl
Soil Ecol 46:259–267
Miettinen J, Shi C, Liew S (2016) Land cover distribution in the
peatlands of Peninsular Malaysia, Sumatra and Borneo in
2015 with changes since 1990. Global Ecol Conserv
6:67–78
Murdiyarso D, Hergoualc’h K, Verchot V (2010) Opportunities
for reducing greenhouse gas emissions in tropical peat-
lands. Proc Natl Acad Sci USA 107(46):19655–19660
Page S, Rieley JO, Shotyk Ø, Weiss D (1999) Interdependence
of peat and vegetation in a tropical peat swamp forest.
Philos Trans R Soc B 354(1391):1885–1897
Page SE, Rieley JO, Banks C (2011) Global and regional
importance of the tropical peatland carbon pool. Glob
Chang Biol 17:798–818
Rejmankova E (2001) Effect of experimental phosphorus
enrichment on oligotrophic tropical marshes in Belize,
Central America. Plant Soil 236:33–53
Wetlands Ecol Manage
123
Rieley J, Page S (eds) 2005. Wise use of tropical peatlands:
focus on South East Asia. ALTERRA - Wageningen
University and Research Centre and the EU INCO -
STRAPEAT and RESTORPEAT. 237 p. http://www.
restorpeat.alterra.wur.nl/download/WUG.pdf. Accessed
13 June 2015
Sanaullah M, Chabbi A, Lemaire G, Charrier X, Rumpel C
(2010) How does plant leaf senescence of grassland species
influence decomposition kinetics and litter compounds
dynamics? Nutr Cycl Agroecosys 88:159–171
Sjo
¨gersten S, Cheesman A, Lopez O, Turner B (2011) Bio-
geochemical processes along a nutrient gradient in a trop-
ical ombrotrophic peatland. Biogeochemistry
104:147–163
Strakova
´P, Anttila J, Spetz P, Kitunen V, Tapanila T, Laiho R
(2010) Litter quality and its response to water level draw-
down in boreal peatlands at plant species and community
level. Plant Soil 335:501–520
Strakova
´P, Niemi R, Freeman C, Peltoniemi K, Toberman H,
Heiskanen I, Fritze H, Laiho R (2011) Litter type affects
the activity of aerobic decomposers in a boreal peatland
more than site nutrient and water table regimes. Biogeo-
sciences 8:2741–2755
Sulistiyanto Y (2004) Nutrient dynamics in different sub-types
of peat swamp forest in central Kalimantan, Indonesia.
PhD thesis, University of Nottingham, UK. p 388
Sundari S, Hirano T, Yamada H, Kusin K, Limin S (2012) Effect
of groundwater level on soil respiration in tropical peat
swamp forests. J Agr Meteor 68:121–134
Sundberg A, Sundberg K, Lillandt C, Holmbom B (1996)
Determination of hemicelluloses and pectins in wood and
pulp fibres by acid methanolysis and gas chromatography.
Nord Pulp Pap Res J 11:216–219
Sundberg A, Pranovich AV, Holmbom B (2003) Chemical
characterization of various types of mechanical pulp fines.
J Pulp Pap Sci 29:173–178
Talbot J, Treseder K (2012) Interactions among lignin, cellu-
lose, and nitrogen drive litter chemistry–decay relation-
ships. Ecology 93(2):345–354
Wu
¨st R, Jacobsen G, van der Gaast H, Smith A (2008) Com-
parison of radiocarbon ages from different organic frac-
tions in tropical peat cores: insights from Kalimantan,
Indonesia. Radiocarbon 50(3):359–372
Yule C, Gomez L (2009) Leaf litter decomposition in a tropical
peat swamp forest in Peninsular Malaysia. Wetl Ecol
manag 17:231–241
Yule C, Lim YY, Lim TY (2016) Degradation of Malaysian
peatlands decreases levels of phenolics in soil and in leaves
of Macaranga pruinosa. Front Earth Sci 4(45):1–9
Zhang D, Hui D, Luo Y, Zhou G (2008) Rates of litter decom-
position in terrestrial ecosystems: global patterns and
controlling factors. J Plant Ecol 1(2):85–93
Wetlands Ecol Manage
123
A preview of this full-text is provided by Springer Nature.
Content available from Wetlands Ecology and Management
This content is subject to copyright. Terms and conditions apply.