Вы находитесь на странице: 1из 114

Universit degli Studi di Milano Bicocca

SCUOLA DI SCIENZE
Dipartimento di Scienze dellAmbiente e del Territorio e di Scienze della Terra
Corso di Laurea Magistrale in Scienze e Tecnologie per lAmbiente e il Territorio

The relationships between water


table and redox potential in
peatlands

Relatore:
Prof. Roberto Comolli

Tesi di Laurea di:


Iacopo Federico Ferrario

Correlatore:
Prof. Ruurd Van Diggelen

Matricola:

770166

______________________________________________________________________________
_________
Anno Accademico 2013 /2014

TABLE OF CONTENTS
1

INTRODUCTION ................................................................................................................... 4
1.1

Peatlands............................................................................................................................... 4

1.2

Ecohydrology ....................................................................................................................... 6

1.2.1

Height above water table .............................................................................................. 9

1.2.2

Micro-topography....................................................................................................... 10

1.3

Carbon cycle ....................................................................................................................... 10

1.3.1

Decomposition ........................................................................................................... 11

1.3.2

Spatial process of decomposition ............................................................................... 12

1.4

Microbial community ......................................................................................................... 15

1.5

Vegetation season: drying and rewetting ........................................................................... 15

1.6

Redox potential .................................................................................................................. 17

1.7

Redox potential is a complex indicator .............................................................................. 22

1.7.1
1.8

Aim and research questions................................................................................................ 35

1.8.1
2

Hypotheses ................................................................................................................. 35

MATERIALS AND METHODS .......................................................................................... 36


2.1

Field work .......................................................................................................................... 36

2.1.1

Site description ........................................................................................................... 36

2.1.2

Experimental design ................................................................................................... 37

2.1.3

Redox measurements.................................................................................................. 41

2.1.4

Water table measurements ......................................................................................... 42

2.1.5

Pore water sampling ................................................................................................... 43

2.1.6

Decomposition ........................................................................................................... 44

2.1.7

Peatsoil sampling........................................................................................................ 45

2.2

Assembling the picture ............................................................................................... 33

Laboratory work ................................................................................................................. 45

2.2.1

Water chemistry analysis ........................................................................................... 45

2.2.2

Peat analysis ............................................................................................................... 46

2.2.3

Data calculation .......................................................................................................... 53

RESULTS AND DISCUSSION ........................................................................................... 54


3.1

Introduction ........................................................................................................................ 54
2

3.2

Redox potential and water table ......................................................................................... 56

3.3

pH, Dissolved Inorganic Carbon and Dissolved Organic Carbon ..................................... 68

3.4

Cellulose decomposition and TBI ...................................................................................... 78

3.5

Peat quality and degradability ............................................................................................ 83

CONCLUSIONS ................................................................................................................... 93

Acknowledges ....................................................................................................................... 98

Appendices ............................................................................................................................ 99

References ........................................................................................................................... 102

1 INTRODUCTION
1.1 Peatlands
Peatland ecosystems are terrestrial environments where, over the long term, net primary
production exceeds organic matter decomposition, leading to the accumulation of a deposit of
poorly decomposed organic matter, named peat, thicker than 30 cm (Wieder and Vitt, 2006).
Peat formation is the result of complex interaction among anoxic conditions, low
decomposability of the plant material and hydrology. Peatlands are distinguished by the presence
of water close to the surface, they often have unique soil conditions that differ from the adjacent
uplands, they support vegetation adapted to wet conditions and they are characterised by an
absence of flooding-intolerant vegetation. Climate and geomorphology control at the larger scale
the degree to which peatland can exist, but hydrology underpins its ultimate development,
affecting the physicochemical environment and influencing the biota (Gosselink and Mitsch,
2000). The persistence of peatlands depends from the constant supply of water, while source of
water influences its structure and function. Origin of water has been long employed for
classifying peatlands. When water comes from surrounding or underlying mineral soil, peatlands
are termed minerogenous (or geogenous). If peat has grown thick enough to progressively
become isolated from the mineral soil, precipitation becomes the only source of water and
peatlands are termed ombrogenus. Minerogenous water carries cations, anions and nutrients and
the resulting chemical composition has great effects on influencing flora, vegetation and
ecosystem functions. Minerotrophy is the term describing this condition and forms the ecological
basis for the peatland type named fen. On the contrary, ombrogenous water has very low
dissolved minerals and provides the ombrotrophic condition for the development of the peatland
type named bog.
Two main environmental gradients are responsible for the differentiation of peatland
habitat types. One gradient follows the variation in moisture and aeration as a function of water
table position, which changes in time and space, and of the pore structure of the peat. The
distance between the soil surface and the water table is the height above water table (HWT) and
represents roughly the depth of the aerated zone, which has high biological and ecological
relevance. The other is a complex gradient that combines the variation of pH, base richness and
nutrient availability. The gradient of nutrient availability and increasing productivity is described
using three terms borrowed from limnology, which are respectively oligotrophic, mesotrophic
and eutrophic. This gradient changes independently from the mineral gradient composed of pH,
4

base cations and associated anions (Wieder and Vitt, 2006), indeed nitrogen, potassium and
phosphorus have their own chemistry and variation and all assumptions of strong correlation
with pH and base richness are not obvious (Rydin and Jeglum, 2013). The factors governing the
differentiation into the main habitats are the same as those causing vegetation differentiation, so
that a useful system to describe habitat variation in peatlands relies on vegetation pattern.
Therefore, the floristic composition mirrors change in pH and base richness, creating the bogpoor fen-rich series. Bogs are always oligotrophic while rich fens can be either eutrophic or
oligotrophic. As the peatland becomes more isolated from the groundwater, the pH usually
decreases because Sphagnum mosses have the ability to acidify water, which cannot be
contrasted by the low buffering capacity of rainwater. The pH of bogs is comprised between 3.5
- 4.2, while it increases along the poor-fen rich-fen gradient from 4 to 8 (Rydin and Jeglum,
2013). The complete absence of bicarbonate alkalinity below pH 5.5 is a fundamental dividing
point in the habitat limits of many peatland species. Below 5.5 pH fens are dominated by
oligotrophic species of Sphagnum, while above 5.5 pH Sphagnum abundance decreases and true
mosses mostly dominate (Wieder and Vitt, 2006). Mire ecologists observed that the measure of
the water table from the surface (HWT) was a very strong predictor of vegetational gradients in
peatlands. The water table gradient allows the micro-topography of peatlands to be classified in
the following elements, which were first described in Sjrs (1948; cf. in Rydin and Jeglum,
2013):

Hummocks are mounds of 20-50 cm raised above the lowest surface level. The thick
aerobic peat supports dwarf shrubs and vascular plants.

Lawns are intermediate microforms, 5-20 cm above the water table, where graminoids
are most common. Mosses reach the highest species diversity in lawns. Firm surfaces that
you can walk on.

Carpets are from 5 cm below to 5 cm above the water table. They are characterised by a
soft, green layer of mosses and a sparse cover of cyperaceous plants. Softer surface into
which your feet sinks.

Mud-bottoms are most of the time submerged; they lack vascular plants and are
dominated by mosses or they can also expose the bare peat, usually covered by algae.

Pools are water-filled depressions.

Within the peatland, a vegetational gradient develops largely in parallel to HWT, which is deeper
at the uplands and shallower towards the open mire, passing through the mire margin and the
mire expanse. The mire margin has usually thinner peat and vascular plants can reach the
5

mineral layer, so that small, creeping trees and shrubs can grow. pH is lower in the centre and
higher in margin and calcium ion shows the same pattern. In bogs and especially in raised bogs,
where the peat meets the upland soil, usually a narrow fen called lagg develops. The lagg is a
feature of the margin that surrounds the bog and receives water both from the bog and from the
surrounding mineral soils.
The peculiar shape of a peatland, its surface morphology and patterning is the result of
interactions between substrate, climate, hydrology and vegetation. Raised bogs are ombrotrophic
mires raised above the level of surrounding uplands, which are usually at least 500 m in diameter
and with a convex cupola that can be several metres higher than the edges and surrounding
uplands (Rydin and Jeglum, 2013). Water slowly flows from the centre to the edge. Where raised
bogs develop over round-shaped lakes, they grow circular and hummocks and hollows form
concentric patterns, perpendicular to the lateral through-flow.

1.2 Ecohydrology of raised bogs


Raised bogs are ombrotrophic mires that depend only on precipitation for their supply of water,
so that they can be found only in climatic regions where the input of water exceed the loss.
Raised bogs are characterised by vertical oscillation of the surface during drying and rewetting
(the so called mire breath), small temporal fluctuation of the water table, reduction of
evapotranspiration occurring at shallow water level and large storage coefficients (Schaaf, 1999).
In saturated condition, the content of water in undisturbed bog peatsoil can range between 88-97
% (Ivanov, 1981).
Peatlands such as raised bogs are not made of a uniform body, but they are characterised
by a double layer, and mire researchers introduced the terms acrotelm and catotelm to define this
particular feature (figure 1.1).

Figure 1.1 Bogs are formed by a double layer. The acrotelm, thin and biologically active and the catotelm,
thick and inactive.

The acrotelm is the superficial layer where water table fluctuation occurs and where
biogeochemical cycles and biological processes are more active, while under a hydrological
point of view, the high hydraulic conductivity and low degree of humification identifies the
acrotelm as an aquifer (Rydin and Jeglum, 2013; Schaaf, 1999). The acrotelm thickness can vary
from 7-8 cm in herb rich fen to 60-70 cm in moss-rich raised mire (Ivanov, 1981). The catotelm
is the main body of a bog and can be about several metres depth. It is the less active layer where
biological processes are slow, peat is generally well decomposed and hydraulic conductivity is
so low that, hydrologically, it is described as an aquitard. The classic Ingrams theory describes
raised bogs as raised water mounds, with the catotelm as a body that loses water by lateral
outflow and that is compensated for by infiltration by the overlying acrotelm (figure 1.2). The
mound shape is a function of the hydraulic conductivity of the catotelm (assumed low and
spatially constant) and the infiltration rate from the acrotelm (assumed constant).

Figure 1.2 Precipitation is the only source of water in raised bogs. Rainwater flows outwards from the
centre via runoff, lateral through flow and groundwater flow.

The theory describes raised bogs as fed only by precipitation and isolated from the mineral
surrounding groundwater. However, recent studies have demonstrated that the isolation from
groundwater may not be always a rule in bogs. Relatively high hydraulic conductivity, ranging
from 10-2 to 10-1 m d-1, was measured even in the catotelm (Chason and Siegel, 1986) so that, if
there is no net hydrological distinction between acrotelm and catotelm, there might be significant
downward and upward exchanges of water (Sirin et al., 1996 cited in Rydin and Jeglum, 2013)
(figure 1.3). Internal hydrological mechanisms, the change in precipitation, evapotranspiration
and the degree by which water table fluctuation affects the hydraulic head in the catotelm govern
the process and can cause the reversal of hydraulic gradient (Fraser et al., 2001; Devito et al.,
1997). Consequently, during droughts, the water table drops and mineral groundwater can move
upward reaching 1-2 m depth from the surface thus affecting pore water chemistry (Gosselink
and Mitsch, 2000).
7

Figure 1.3 For raised bog with particular climatic and peat conditions upwelling of mineral water can
occur.

Hydraulic conductivity (K) is the property that governs the flow of water and transport of
solutes in a porous medium. In a bog, hydraulic conductivity varies with peat type, degree of
humification and bulk density. Peat type is characterised by different structures influencing
hydraulic conductivity, for instance K is lowest for Sphagnum peat and higher for lignoid peat
(Rydin and Jeglum, 2013). Generally, K correlates negatively to degree of humification. The
degree of humification can be measured with the von Posts method, which assigns an increasing
number to the change of some diagnostic features that correlate with degree of decomposition.
Humification can vary either vertically on a scale of decimetres as horizontally on a scale of few
metres (Schaaf, 1999). Humification, however, does not increase necessarily with depth because
it also depends on different factors: vegetation (Silamikele et al., 2007), peat forming conditions,
different decaying rate in the catotelm, physical conditions, micro-topography and climatic
conditions (Schaaf, 1999). Hydraulic conductivity correlates negatively to bulk density. Bulk
density tends to increase with depth since compaction caused by overlaying peat layers reduces
the volume of pores. All the factors described above vary spatially so that also K is spatially
dependent, despite the Ingrams theory assumed constant hydraulic conductivity throughout the
catotelm. Indeed, more recent works showed that in the catotelm K tends to have a vertical
decreasing trend with depth while it tend to decrease from the centre to the bog margin (Beer et
al., 2008; Schaaf, 1999), invalidating the assumption of the Ingrams theory. The way K varies in
a bog controls the vertical and horizontal flow of water. In bogs, water has a preferential
horizontal flow because vertical hydraulic conductivity can be as three orders of magnitude
lower than horizontal K (Beer et al., 2008). In the acrotelm, compaction and decomposition
cause the hydraulic conductivity to decrease strongly with depth from about 10 5 m d-1 at the
surface to 110 m d-1 at some decimetres below it (Schaaf, 2004; 1999). Fraser et al. (2001)
measured a great decrease of K that reached 10-3 m d-1 at 45 cm below the surface. Yet, hydraulic
8

conductivity can be dynamic. Drying can decrease K compressing the pores while precipitation
can cause peat to swell, increasing its hydraulic conductivity. Given the vertical trend of K and
the peat properties change during drying-rewetting cycles, the rate of lateral subsurface flow in
the acrotelm depends on water table position, so that flow decreases as water table decreases.
This mechanism provides an important feedback that augments the water retention power in bogs
during droughts. In the acrotelm, hydraulic conductivity is usually 4-5 times greater horizontally
than vertically (Beer et al., 2008; Schaaf, 1999). Runoff can occurs if the rate of precipitation
fills the storage capacity of hollows and exceeds the rate of infiltration. In ecosystems like raised
bogs, characterised by a small slope and a well developed hummock-hollow topography, runoff
occurs as sheet flow and channel flow in micro channels. The micro-topography adds complexity
to the flow pattern because depressions can be connected during wetting and can be disconnected
during drying and the flow pattern can reverse due to the different filling and emptying pattern of
depressions while some depression can also not participate to surface flow (van der Ploeg et al.,
2011). The transition between subsurface flow and runoff is controlled by complex threshold
mechanisms, and Frei et al. (2010) found that surface flow is mainly a function of depression
storage and that runoff occurred only during intense rainfall. Subsurface flow is then the
dominant flow during dry and wet periods while surface flow occurs only during intense rain
events because micro-topography acts as inhibitors of surface flow (Frei et al. 2010).

1.2.1 Height above water table


Height above water table (HWT, i.e. the distance from peat surface to the water table) is
an important indicator used to predict a number of important eco-hydrological variables in
peatland hydrology, ecology and biogeochemistry including run-off, saturation, redox potential,
biodiversity, soil structure, methane emission, peat quality and organic matter decomposition
(Waddington et al., 2014). Water table governs the water availability to bog plants and
delimitates the zones of aerobic and anaerobic respiration in peat, so that it is considered a main
factor controlling vegetation distribution (Rydin and Jeglum, 2013), microbial diversity and
niche differentiation (Andersen et al., 2013). Water table fluctuation governs the oxygen
penetration in peat and affects redox potential (Niedermeier and Robinson, 2007; Mansfeldt,
2003; Seybold et al., 2002), thus determining the spatial distribution of aerobic conditions. That
condition enhances decomposition not only because aerobic respiration is more efficient than
anaerobic respiration, but also because it allows the degradation of anti-microbial Sphagnum
phenols in shallow peat (Abbott et al., 2013; Fenner and Freeman, 2011). Water table level is
9

also the most important control on the relative importance of methane production in peat and of
the amount of CH4 fluxes out of peatlands (Bridgham et al., 2013). Yet, from a hydrological
point of view, water table level can provoke flow reversal in peatlands (Fraser et al., 2001) and it
is an indicator of water discharge and solute transport out of the peatland into the watershed.

1.2.2 Micro-topography
The surface of raised bogs is characterised by distinct hummocks and hollows micro-topographic
elements. Micro-topography features can alter hydrological, physicochemical and biological
attributes (Courtwright et al., 2011; Baldwin et al., 2007; Haraguchi, 1992), adding spatial and
temporal heterogeneity to the turnover of redox-sensitive solutes in peatlands. Therefore, they
are not just depending on peat quality, peat properties and temperature. The most important
difference between hummocks and hollows is water-table depth (Bragazza et al., 1998).
Hydrologically, in continental bogs, above the water table, hummocks have a dominance of
macropores associated with vascular plant roots, so that they have higher hydraulic conductivity
than hollows and the water flow is predominantly vertical. The difference in peat physical
properties and hydraulic conductivity accounted by microforms become less important with
depth to water table. Branham and Strack (2014) showed that hydraulic conductivity was not
influenced any more by microforms at a depth of 20 cm below the water table. Frei et al. (2012)
showed that micro-topography could induce water flow patterns that eventually determined
biogeochemical hot spots, defined as sites characterised by higher biogeochemical activity than
the surrounding areas. For example, when water table increases, hummocks can be the only place
where oxidation can occur. Indeed, a higher concentration of nitrate is usually found in
hummocks (Frei et al. 2012; Wolf et al., 2011; Bragazza et al., 2005). Through the infiltration
zone inside the hummocks, oxidised nutrients can be transported by precipitation to deeper and
more redox-reducing levels where they can be reduced. In this occasion and at the water table
level, hummocks can become hot spots for reducing reactions (Frei et al. 2012).

1.3 Carbon cycle


The study of biogeochemical cycles is of primary importance to unravel the ecological role of
peatlands at different spatial and temporal scales. Regarding the carbon cycle (figure 1.4),
Northern peatlands store half (approximately 1672 Gt) of the global soil carbon pool though
cover an estimated 3.6x106 km2 (Tarnocai et al., 2009), which is equal to about 3 % of the total
10

lands. Peatlands can be source, sink or transformers of carbon compounds, and many efforts
were directed to measure its fluxes and to understand their functioning.

Figure 1.4 Carbon cycle in bogs. Circle boxes represent gas phase; broken lines represent microbialmediated processes.

1.3.1 Decomposition
Measurement of decomposition rate (k) is important for the biogeochemistry of nutrients and
carbon fluxes (Prescott, 2010). The availability of nutrients is due in large part to the decay
dynamics of the organic matter. The process also supports diversity in the microbial population
by supplying a rich set of intermediate degradation products (Berg and McClaugherty, 2014).
Bogs receive very low amount of nutrients from wet and dry deposition that made up their entire
external sources. For that reason, decomposition and recycling of nutrients become an essential
internal source of nutrients in bogs (Bragazza et al., 2013).
Zhang et al. (2008) reviewed the factors that controlled decomposition rate. These were:
(i) climatic factors (mean annual temperature, MAT; mean annual precipitation, MAP; annual
actual evapotranspiration, AET); (ii) litter quality (nitrogen content; carbon:nitrogen ratio, C:N;
lignin content, LIGN; lignin:N ratio, LIGN:N); (iii) vegetation and litter types; (iv) geographical
variables (latitude, LAT and altitude, ALT). The authors found that C:N ratio and the total
nutrient content of the litter (both litter quality factors) were the two most important factors
influencing decaying rate on a global scale. A threshold-based mechanism dominates the relation
between decomposition rate and its factors. There is not a single factor that is, in every

11

conditions, more important than the other ones, so that decomposition can slow down despite all
the factors but one is adequate (Prescott, 2010).
In bogs, the litter quality of Sphagnum overwhelms the importance of other factors.
Sphagnum species forming hummocks have relatively higher content of recalcitrant structural
carbohydrates (lignin-like polymeric phenolics) than hollows species. Thus, despite hummocks
are well-aerated and aerobic decomposition rate is higher than anaerobic decomposition, the
decomposition of Sphagnum litter is lower in hummocks (Bragazza et al., 2013). After litter
quality, moisture (i.e. HWT) is the most important factor for decomposition that can show a
threshold effect. In fact, a lower water table caused by droughts can desiccate the litter
hampering decomposition. Laiho (2006), in a review on decomposition constraint in northern
peatlands, included pH as important factor because it can limit the activity of phenol oxidase.
Finally, several studies have shown that temperature increased decomposition only under nonlimiting moisture condition (Bragazza et al., 2013; Mkiranta et al., 2009).
Hydrologic dynamics, interactions between surface ombrogenous water and deeper
groundwater, influence the distribution and transformation of nutrients in wetlands (Frei et al.
2010) and the decomposition rate in the catotelm (Beer and Blodau, 2007). If in the catotelm
diffusion dominates over advection there will be a lack of solute transport and an accumulation
of product of microbial metabolism, such as CH4, dissolved inorganic carbon (DIC) and
recalcitrant dissolved organic carbon (DOC) (Beer et al., 2008). Beer and Blodau (2007) showed
that these phenomena can lead to a thermodynamic constraint for microorganism metabolism
that slows or even cuts off the decay of peat in anoxic layers. It follows that the groundwater
residence time becomes a fundamental factor in assessing the spatial decomposition rate in bogs
(Morris et al., 2011). Advective flux like upwelling of mineral water or downwelling of
rainwater can break the constraint. The influence of rainwater and upwelling depends on the
profile of hydraulic conductivity. It is reported that in bogs at most the first 100 cm can be
affected by rainwater (Dobrovolskaya, 2013; Morris et al., 2011).

1.3.2 Spatial process of decomposition


Main inputs of carbon in peatland are through photosynthetic production of mosses and then
vascular plants. In peatlands, plant biomass can be divided into above ground, rhizome and
coarse roots, and fine roots (diameter < 0.5 mm). Carbon input below ground takes place through
root exudations and decaying vascular plant materials. Most of the organic matter decomposition
occurs in the acrotelm, where rate of decomposition has been estimated to be one hundred times
12

higher than in the catotelm. There is an important biogeochemical connection between the upper
layers and the deep peat, so that about 10 % of litter mass produced in the acrotelm reaches the
catotelm (Frolking et al. 2001 cited in Beer et al. 2008), albeit the occurrence of this
translocation is actually strongly dependent on hydraulic conductivity (Beer et al. 2008).
Products of decompositions are CO2 and CH4 gases and DOC. The relative importance of
aerobic and anaerobic respiration depends on the HWT, which controls the oxygen supply into
the peatsoil. The principal product of aerobic respiration is CO2, while anaerobic respiration
produces DOC, CH4 and CO2 (Fenner and Freeman, 2011). Oxic conditions favour CO2
production and inhibit methanogenesis, while anaerobic respiration favours lower CO2
production but greater CH4 production rate (Estop-Aragons and Blodau, 2012). Respiration rate
is higher at water table level and then decreases (Shoemaker et al., 2012) so that most of the
organic matter is decomposed in the acrotelm (Beer and Blodau, 2007). In anoxic layers,
respiratory pathway depends on Eh conditions, peat quality and relative concentration of electron
acceptors. Microorganisms compete for electron acceptors in presence of nitrate, ferric ion,
sulphate and finally CO2. Since bogs are fed only by precipitations, the common electron
acceptors and nutrients have very low concentrations compared to other minerogenic wetlands
and they even decrease with depth (Beer et al., 2008; Proctor, 2003; Steinmann and Shotyk,
1996; Lundin and Bergquist, 1990). Keller and Bridgham (2007) measured that Fe(III) and
nitrate reduction accounted for less than 1% of anaerobic carbon mineralisation, while sulphate
reduction was responsible for 6 26 %. These results are consistent with the frequent
observation that reactive inorganic sulphur (RIS) pool can be very dynamic in bogs (Wieder and
Lang, 1988). Many studies have stressed that respiration driven by common electron acceptors
accounts for only a minor fraction of the CO2 produced in bogs. The sources for the unexplained
CO2 can be the regeneration of electron acceptors during water table drawdown and/or other
respiratory pathways like bacterial respiration with humic substances (Knorr et al., 2009; Deppe
et al., 2009; Heitmann et al., 2007), with organic sulphur species (Kertesz, 2000), and
fermentative processes in absence of electron acceptors. Fermentation can be very important in
bogs and can lead to high production of DOC (Vile et al., 2003). Fermenting microorganisms are
important in degrading complex polymers yielding simpler products used in methanogenesis.
Fermentation, together with DOC reduction, is addressed as responsible of the high fraction of
unexplained carbon mineralization in anoxic layers (Keller and Bridgham, 2007). There is an
increasing need to assess the use of organic electron acceptors (DOC) by microorganisms in
anoxic environment. Comparing sites with different hydrology (bog and fen) Shoemaker et al.
(2012) found a high production of CO2 in anaerobic layer at the ombrotrophic site and found a
13

local correlation with concentration of DOC, suggesting that DOC may have driven anaerobic
respiration. A further complication to the role of DOC is that the chemical composition rather
than its amount distinguishes its role as electron donor or electron acceptor (Keller et al., 2009).
The vertical profile of DIC generally increases with depth (Deppe et al., 2009). Several
authors observed that DIC increased at all depths during summer months, along with
temperature. Despite the temperature drop during autumn, DIC is reported to be still increasing
and Shoemaker et al. (2012) reported that the profile of DIC decreased only after frequent rain
events occurred, since rain dilutes bog water and favours CO2 degassing. The DIC profile
depends also by emission to atmosphere, which can occur by diffusive flux or by non-diffusive
flux via roots of vascular plant or via ebullition (Glaser et al., 2004; Tokida et al., 2007).
Substrate quality decreases with depth and the main substrate for decomposition in
deeper layer becomes DOC. DOC concentration is higher in surface layer and decreases with
depth (Deppe et al., 2009; Beer et al., 2008; Fraser et al., 2001). Pore water concentration of
DOC usually increases in summer due to evapotranspiration and enhanced decomposition
(Waddington and Roulet, 1997).
Supply of high quality substrate and water table depth (Rydin and Jeglum, 2013) are the
main factors controlling the production of methane in peatlands. In general, acetoclastic
methanogenesis is favoured in the upper peat layers where abundant labile organic carbon from
decaying vegetation are found, and H2 and CO2-dependent methanogenesis predominates in the
more recalcitrant deeper peat layers (Beer and Blodau, 2007). The zone of most active CH 4
production was measured about 10 cm below the average water table (Sundh et al., 1994), where
there is input of fresh organic matter. This zone is closely related to the source of products from
anaerobic respiration, in fact Shoemaker et al. (2012) found a peak of methane production just
below the respiration rate peak near the water table surface (often between 0-5 cm below the
water table). Regarding the role of micro-topography, the hummocks are reported to have lower
methanogenesis than hollows because there is less quality input of substrate (Bubier et al., 1993).
It is also arguable that, being oxygenated, the hummocks provide more energetically favourable
microbial pathways.
Peak of oxidation of methane is found at the water table level and above it and it is higher
in hummocks than in hollows (Frenzel and Karofeld, 2000). Occasionally methanotrophy can be
found in saturated layer. In this case it is driven by oxygen released by roots and by anaerobic
oxidation of methane.

14

1.4 Microbial community


Mires are active systems characterized by organic matter turnover (Biester et al. 2012).
Microorganisms play a key role in wetland soils. They are involved in fundamental
biogeochemical cycles and transformation of elements. The rate and extent of their role in these
processes depend on the variability and change of environmental variables on which their
metabolism and growth depend (Andersen et al., 2012). The activity of microorganism depends
on temperature, pH, hydrological regime, oxygen, peat quality, nutrients availability and redox
potential. These variables are spatially dependent and, in peatlands, different habitats can show
changes in the distribution of microbial groups and in their total biomass. Distribution of
microorganism is also depth dependent. Microorganisms face increasing energetic constraints
with depth, caused by a combination of factors such as the availability of oxygen and
thermodynamic disadvantage that are characteristic of a stratified peat with redox and peat
quality change (Robroek et al., 2013). Fungi, as general rule, are dominant in oxic and acid
habitats, so they are abundant in the aerated surface layer. Microhabitats as hollows and
hummocks have different vegetation, hydroperiod and redox dynamic. In Robroek et al. (2013)
an analysis of microbial community between microhabitats has revealed a change in the
fungal/bacterial ratio toward higher values in hummocks than hollows. Generally fungi decreases
in biomass with depth (Andersen et al., 2012), but recently Jassey et al. (2011) have found larger
number in lower peat layer, a finding described also in Dobrovilskaya (2014). Bacteria instead
dominate in anoxic and neutral environment. The bacterial abundance generally decreases with
depth, but some studies have revealed increasing biomass with depth (Golovchenko et al., 2007),
or peaks at some depth (Grodnitskaya and Trusova (2009) cited in Rydin and Jeglum, 2013).
In a peatland, microorganisms respond to alteration of hydrological regime following the
time and space changes and the intensity of the event. Persistent drought could trigger
vegetational shifts and indirectly affect microorganisms. During drought methanogenic bacteria
can survive in anoxic pores still present in aerobic layers (Kotiaho et al., 2013) and aerobic
bacteria can increase in number following the lowering of water table (Karsisto, 1979).
Methanotrophic bacteria have been found throughout the profile but in greater concentration at
certain depth so that under particular condition the population can become active.
Methanotrophic bacteria, which can tolerate extended periods of anoxia, can resume methane
oxidation within few hours of re-exposure to oxygen.

1.5 Vegetation season: drying and rewetting


15

Drought is a period of scarcity of water that is statistically derived from climatic data. Regarding
the effect on biogeochemical processes in peatlands, droughts can be defined according to their
degree of intensity. Severe droughts can affect microorganisms either directly or indirectly, on a
long-term range, after driving a shift in vegetation. Short droughts, although having minor effect
on vegetation, can also affect directly the microbial community. Droughts influence microbes by
lowering the moisture content of peat (i.e. water table) and enabling the penetration of oxygen in
deeper peat layers so that aerobic respiration can establish. If drought is severe, shortage of water
can lead to microbial mortality (Mettrop et al., 2014). Regarding climate change, the effects of
droughts on the carbon cycle is under high concern since an increase of their severity or
frequency could change the role of peatlands from being sinks of carbon to being sources.
Drought intensity and following rewetting are the most important factors that can
enhance decomposition of peat soil. Water table drawdown allows oxygen to reach previous
anaerobic layers of peat. Usually, although the pattern is very peat type-dependent, CO2
increases during drying until an optimum moisture level is reached and afterward it decreases
(Estop-Aragons and Blondau, 2012), because moisture becomes limiting. The higher
decomposition rate occurring during droughts, allows the release of nutrients in the peat, a
process called eutrophication. During the rewetting phase, these nutrients can be transported
downward and can serve to enhance microbial decomposition in low nutrient layers.
Fenner and Freeman (2011) have proposed a theory, called the enzymatic latch theory,
which aims to explain the mechanism underlying the bogs response to droughts. As mentioned
above, the persistence of bogs is due mainly to waterlogged anoxic condition and recalcitrance of
Sphagnum. Sphagnum spp. have high amount of phenolic compounds, with inhibiting effects on
microbes, that can be degraded only in presence of oxygen by phenol oxidases enzymes. When
water table is high, oxygen cannot penetrate in peat and that can prevent the peat deposit to be
released as CO2. When drought introduces oxygen, phenol oxidase can remove phenolic
inhibitors, enabling hydrolases to resume normal mineralization of organic matter and increase
decomposition.
Consequently, droughts influence microbial activity and lead to a hydrochemical shift in
surface water. As mentioned above, eutrophication releases nutrients, while another important
effect of aeration is the re-oxidation and regeneration of electron acceptors (Deppe et al. 2009).
At the same time, the oxidation of reduced redox couples by aerobic respiration releases
hydrogen ions bringing acidification (Brouns et al., 2014; Mettrop et al., 2014; Clark et al.,
2009), which in turn further limits microbial activity. In Fenner and Freeman (2011), sulphate
and nitrate were released after 23 days of drought while potassium and P after 48 days. Sulphate
16

was released within weeks in Brouns et al. (2014) and Proctor (1994), followed by acidification.
The time scale of response of oxidising and reducing processes depends on process, depth and
peat heterogeneity (Knorr et al., 2009). According to the enzymatic latch theory, if a drought is
not intense enough, it is possible that phenolics will not be degraded and then eutrophication and
increasing of CO2 will not occur (Mettrop et al., 2014). Fenner and Freeman (2011) have
observed a strong increase of CO2 after a severe drought. Despite the increasing production of
CO2, in the zone of aeration, DIC can be very low because of degassing (Deppe et al., 2009). The
production of DOC seems related to drought intensity. During aeration, DOC becomes the
substrate of aerobic respiration (Hughes et al. 1997) and it decreases (Mettrop et al., 2014;
Fenner and Freeman, 2011). However, in a mesocosm experiment, Mettrop et al. (2014)
observed that after strong desiccation DOC greatly increased, which the authors proposed that
this is a consequences of microbial mortality or a change in microbial community composition.
Rewetting in bogs occurs by means of precipitations. The infiltration of rainwater to
lower layer is important after droughts (Deppe et al. 2009) because causes the pH to increase
(Fiedler et al. 2007). Fenner and Freeman (2011) measured increasing flux of CO2 after
rewetting, which the authors explained as an effect of eutrophication and removal of pH
constraints. Deppe et al (2009) measured a rapid increase of DIC and CH4 after flooding,
whereas drying and following rewetting had not strong effects on DOC. Clark et al. (2009)
measured that the drying-rewetting cycles influenced DOC down to 55 cm depth. Proctor (2003)
observed a sharp peak of SO4 after rewetting (Proctor, 2003).

1.6 Redox potential


Redox reactions are chemical reactions that involve the transfer of electrons between two
species, molecules, atoms or ions, so that their oxidation state changes. Oxidation is the loss of
electrons or an increase of oxidation state and reduction is the gain of electrons or a decrease in
oxidation state. Reduction and oxidation must occur simultaneously, since, in a redox reaction,
the reducing agent transfers the electrons to an oxidised agent and free electrons cannot exist in
solution. Conceptually, redox reactions are described as two half-reactions, one releasing
electrons and the other gaining electrons, combined to form a whole reaction. The two related
species that exchange electrons and change their oxidation state in a redox reaction are called
redox pair or redox couple. In analogy with pH, the hypothetical electron activity of a solution is
given by p = - log{e-}. The quantity p measures the relative tendency of a solution (with one or
more redox couples) to accept or transfer electrons and it is a measure of the Gibbs free energy
of the redox reaction. Also, the redox potential (Eh) is a measure of the electron availability in a
17

solution, though it is made with an electrochemical cell, where the potential is expressed in
relation to the standard hydrogen electrode. In making an electrochemical measurement of the
redox intensity, an electro motive force (Volts) is measured. Eh is related to P by

F
Eh
2.3RT

Eq. 1.1

Where Eh is the redox potential [V] of the solution (in relation to reference electrode); F is the
faraday constant (= 96940 Cmol-1 e-); R is the gas constant (= 8.314 J mol-1K-1); T is
temperature in K.
The thermodynamic definition of the redox potential of a solution is given by the
Nernsts equation, which describes the relationship between Eh and the activities of oxidised and
reduced species

Eh Eh 2.3

Ox
RT
log
Re d
nF

Eq. 1.2

Where Eh is the redox potential [V] of the solution (in relation to reference electrode); Eh is the
redox potential [V] under standard conditions (all activities = 1, pH2= 1atm, [H+] = 1M) and is
related to the free energy change for the cell reaction (G); F = 1 is the faraday constant (=
96940 Cmol-1); n is the number of exchanged electrons; R is the gas constant (= 8.314 J mol1

K-1); T is the temperature in K; 2.3RT/F = 0.059 V (at 25 C); {Ox}/{Red}is the activity of

oxidised/reduced couples.
From equation 1.2, it follows that Eh depends on the ratio of oxidized and reduced forms
(i.e. the relative activities) and not on their absolute quantities. For that reason, Eh is considered
an intensity factor for the reduction/oxidation (i.e. the greater the ratio the more oxidised the
system). Contrarily, redox capacity, a function of the amount of oxidised and reducing
compounds, measures the buffer effect of the solution. The intensity factor determines the
relative ease of the reduction/oxidation whereas the capacity factor denotes the extent to which
the shift will take place. In aqueous natural environments, the water solvent exerts a levelling
influence on the system and restricts the range of accessible Eh between the intensity of water
reduction (E= - 0.83 V) and the intensity of water oxidation (E= + 1.23 V).
Redox reactions are common in nature and they are primary driver of biogeochemical
processes. The metabolism of living organisms relies on redox reactions for energy and for
provision of building blocks, such as for example respiration. Microbes are an important control
18

of redox reactions in soils. They catalyse the kinetic of redox reactions, transferring electrons
from reduced inorganic or organic matter to inorganic or organic terminal electron acceptors
(TEAs), to produce energy for their metabolism. The kind of redox couple, electron donor and
electron acceptor, determines the amount of energy (i.e. Gibbs energy) gained by the
microorganisms out of the process. In a closed aquatic system with organic matter as energy
source and microbes, it is possible to calculate the sequence of oxidation reaction coupled to
TEAs expected on the base of their thermodynamic possibility. Oxygen is the most abundant and
the strongest oxidised agent that is readily available in the atmosphere so that, in aerobic
condition, microbial respiration uses oxygen as final electron acceptor to oxidise organic matter
and produce energy. As water table rises, water saturates the soil and slows down the rate of
oxygen diffusion. Consequently, microorganisms consume oxygen more rapidly than it is
supplied, and, after oxygen is completely depleted, the oxidation of organic matter is coupled to
other TEAs. The temporal sequence predicted on the basis of decreasing Gibbs free energy is the
following: NO3-, Fe(III), Mn(IV), SO42- and finally CO2. To put it differently, nitrate has a better
affinity for electrons than sulphate so that the microbes require a higher electron pressure (lower
Eh) to transfer electrons from organic matter to sulphate.
Each of these reactions occurs within defined values of redox potential, or thresholds,
which can be used to predict the occurrence of the relative redox reactions in soils, though there
are currently arguments about where to draw these boundaries and the matter needs to be verified
more extensively. According to Gosselink and Mitsch (2007), local factors like pH and
temperature, influence the values of these thresholds. Nevertheless, if caution is taken in
interpreting the results they can still be useful in characterising redox condition in soils (Reddy
and DeLaune, 2008). In addition, it has to be borne in mind that such boundaries are rather
smooth stripes than straight lines, therefore they should not be taken as exact borders between
soil redox reactions. According to Gosselink and Mitsch (2007), the microbial oxidation of
organic substrate uses oxygen as terminal electron acceptor at a redox potential of between +400
and +600 mV (Reddy and DeLaune (2008) suggested that below 300 mV oxygen is completely
absent). Below +400 mV nitrate starts to being used as electron acceptor; at about +225 mV
manganese is reduced; between +100 and -100 mV iron is transformed from ferric to ferrous
form; from -100 to -200 mV sulphate is reduced to sulphide; eventually, below -200 mV carbon
dioxide is reduced to methane. There are different thresholds and classifications of redox zones
in literature. Figure 1.5 shows a comparison among other classifications.

19

Figure 1.5. Redox zones defined by different authors, based on Fiedler (2007).

As mentioned before, an electrochemical cell measures the difference in potential


(electron motive force, emf) between an inert indicator electrode in contact with a redox couple
in solution and a reference electrode. The indicator electrode is usually an inert metal like
platinum and the reference electrode can be H+/H2 or Ag/AgCl 1mol/l KCl. The reaction of the
cell where the reference electrode is hydrogen is
Pt, H2 (pH2=1)| H+ (a=1) | Mn+ | M

The overall cell reaction is


Mn+ + nH2 M(s) + 2nH+

If the electrode potential is positive, the above reaction is the spontaneous reaction in the
direction left to right. If the electrode potential is negative, the spontaneous reaction is in
the opposite direction. The voltage of the cell, when the activities of all ions in the cell are
20

unity, when gases are at 1 atm pressure and solids are in their most stable form at 25C, is
calculated as
Ecell = Eh(ox-red) Eref

Eq. 1.3

Eh(ox-red) = Ecell + Eref

Eq. 1.4

Rearranging eq. 1.3 gives

That is the redox potential for the solution measured.


The measurement of redox potential with platinum electrodes has wide applications. The
assessment of anaerobic condition in soils and sediments is important for many disciplines such
as ecology, ecotoxicology, soil science and agronomy. Redox potential influences the
availability of redox sensitive nutrients, their removal and/or translocation in soil profiles and the
efflux of solutes in percolating water (Chadwick and Chorover, 2001). For example, redox
condition influences the nitrogen cycle in soils, determining the nitrogen emission as N 2O, N2 or
NH3, or the plant uptake as NH+4 + or NO3-, with obvious consequence for agriculture and water
quality. Regarding the carbon cycle and global warming, methane, a strong green house gas, is
produced in wetlands depending on redox condition. Most agricultural systems rely on aerated
soil so that roots respiration can occur. Reducing conditions can produce toxic compounds for
plants (reduced forms of Fe and Mn, cyanogenic compounds, ethanol, lactic acid, acetaldehyde
and aliphatic acids such as formic, acetic, butyric acids, and H2S) impairing their growth and the
crops yield. Again, the fate of toxic compounds in the environment depends on redox potential
since the mobility of heavy metals, and that of non-metal like arsenic as well, is redox-sensitive.
Moreover, the degradation of organic pollutants and pesticides in groundwater are affected by
oxidising/reducing condition.
The U.S. National Academy of Sciences Definition states that the minimum essential
characteristics of a wetland are the recurrent, sustained inundation or saturation at or near the
surface and the presence of physical, chemical, and biological features reflective of recurrent,
sustained inundation or saturation. (Gosselink and Mitsch, 2007). In peatlands, where water
saturation and anaerobic conditions are essential characteristics, it is possible to use an electrode
to measure oxygen penetration, in order to characterise the redox conditions and to relate them to
the dominant redox processes (Pezeshki, 2001). When Eh time series is measured, and the
21

relative frequency of redox potential values taken at certain depths is calculated, is possible to
determinate to which redox zones the peat layer corresponds. For instance, Fiedler and Sommer
(2004) related the frequency of Eh to diagnostic features in hydric soil, finding that the
thresholds for Mn reduction was Eh < 450 mV, for Fe(III) reduction was Eh < 170 mV and for
CH4 oxidation was Eh > 75 mV. Measured Eh must be seen as an integrated operational
parameter, which is influenced, as stated, by the activity of living microorganisms but also by
external factors such as variation of water table, precipitation, source of water and chemistry,
pH, temperature and organic matter (Fielder et al., 2007). So far, a number of problems have
hindered the broader application of Eh direct measurement. First, the employment of platinum
electrode probe in in-situ condition requires rugged electrode device to tackle harsh
environmental conditions like storms, prolonged submersion, extreme temperatures and animal
disturbance. Second, irreversibility of coating reactions at the electrode, slow reaction kinetics
and mixed potentials can complicate the interpretation of redox measurements (Peiffer et al.,
1992; Stumm and Morgan, 1981). Last but not least, redox potential shows high spatial and
temporal variability. To conclude, on one hand unpredictable incidents or technical problems
could prevent the collection of reliable data and, on the other hand, the high number of
interplaying variables might prevent simple interpretation of redox potential (Husson, 2013). To
date these issues have slowed the measurement of redox potential in wetlands, leading to a
scarcity of studies (De Mars and Wassen, 1999). Frequently, when a study on redox potential
was performed, it was designed without permanent continuous measurements or without vertical
profiles measurements (Shoemaker et al., 2013; Fiedler et al., 2007). The common use of single
time measurement with intervals in the order of days or weeks cannot account for the variability
expressed by wetland systems. Indeed, in wetland soil redox potential may have periodical or
occasional fluctuations within time of hours and at different depths (Vorenhout et al., 2004).
Moreover, most of the available studies have focused on minerogenous systems with an even
greater lack of publications on redox measurements in ombrogenous bogs.
In the next section, the factors that control redox potential in peatlands are examined, in
order to clarify, where possible, the nature of the relation, the relative importance and the
potential feedbacks.

1.7 Redox potential is a complex indicator


Redox potential in peatland depends on several factors, namely: peat property and quality,
temperature, pH, microbial community, type of vegetation, hydrology and water chemistry
22

(figure 2.2). Since redox potential is an indicator of several processes, the aim here is to
understand how each factor can influence it. The method is based on a review of previous
researches in order to trace emerging trends, which will be discussed in the light of the theory
discussed above and the peculiar characteristic of bog peatlands. In this section, each factor will
be analysed separately for the sake of clarity, though it must be stressed that, in nature, they are
not isolated but they closely interact.

Hydrology
Precipitation
Water table

Temperature
Groundwater

pH

Eh

Water chemistry

Microbial community

Vegetation
Peat properties and
peat quality

Figure 1.6 Controls of redox potential.

Redox potential has a highly dynamic nature and is spatially dependent. The discussion
starts with an evaluation of spatial and temporal variability in the measurement of redox
potential.

Spatial and temporal variability of Eh


In typical wetland soils, Eh values vary from 300 to 700 mV (Reddy and DeLaune, 2008).
According to the work of Fiedler et al. (2007), the range of temporal variation of redox potential
in wetlands can be identified in: short-term, diurnal, single event, seasonal and annual variations.
Periodic diurnal fluctuations can be explained in term of Nernsts equation and the vant Hoffs
law. In this case, the fluctuations are driven by temperature with a temperature maximum
followed by Eh minimum. Plant living roots have daily metabolic cycles. In saturated anaerobic
soils, if aerenchymatous roots are enough dense, redox potential may fluctuate following the
23

release of oxygen that has its peak during the highest photosynthetic rate (i.e. light intensity).
Periodical redox variation can also be driven by periodic physical phenomenon. For example,
Catallo et al. (1999) studied the effect of a 12-hours tide fluctuation in a salt marsh and found
that it corresponded to redox potential fluctuation of about 40-300 mV. In Fiedler et al. (2007),
drying and rewetting are considered single event changes that may bring redox potential to vary
up to 900 mV. Seasonal changes can be related to gradients in soil temperature that affects
microbial activity. Short-term changes are the most difficult to characterise. They can be caused
by soil chemistry dynamics that may result from the production of carbon dioxide, the release of
hydroxyls in ferric iron reduction, the chemistry of precipitation and from inputs of redox
sensitive species. Measurement of redox potential often showed short-term peaks or daily
variation, though so far, there have not been studies that addressed and explained the underlying
causes.
Redox potential can vary with respect to centimetres or even millimetres in soils. This
variability is induced by partial water saturation of soil structure, absorption on soil minerals and
organic matters, plant roots and microbial activity (Yang et al., 2006). Each of these factors
varies spatially and the soil heterogeneity may cause the electrodes positioned at the same site
and depth to show simultaneously a wide range of Eh values (Urquhart et al, 1972). However,
De Mars and Wassen (1999) investigated the role of heterogeneity in different peatlands and,
placing from 4 to 12 electrode replicates within an area of 10 m2 and at 15 cm below surface,
showed that the variability of Eh among the replicates was low, even at low water level. Yang et
al. (2006) measured that spatial redox variability is higher vertically than horizontally, which
seems reasonable given that peat properties that are reported to vary more vertically than
horizontally (Schaaf, 1999). In conclusion, the degree and scale of heterogeneity will affect the
variability of redox potential, though in permanently saturated soils or in areas with minimal
hydrological fluctuations the variability could be minimal (Reddy and DeLaune, 2008).
In wetlands, redox potential generally follows a depth gradient. The vertical profile is
created by local hydrology and addition of electron acceptors and donors and typically decreases
with depth due to oxygen intrusion from the surface (Reddy and DeLaune, 2008). Nonetheless,
other studies showed quite different redox profile patterns. For instance, Vorenhout et al. (2011)
observed that redox potential at deeper soil layers occasionally increased and exceeded the redox
potential at upper layers and that some sensors located at greater depth showed steadily higher
redox potential than overlying layers. Urquhart et al. (1972) observed in a bog a similar pattern,
where some of the redox potential profiles (0 - 30 cm) showed deeper layers having higher Eh
than overlying layers. It follows that redox profile in peatlands can be dynamic and that a simple
24

decreasing gradient is not the general pattern. Hydrology and water circulation are important
factors that can influence the vertical profile and can disrupt the vertical redox stratification that
might be predicted by oxygen intrusion alone (Thompson et al., 2009). However, as mentioned
before, there have not yet been direct field investigations of these eventual relations in wetlands.
Finally, a bogs surface is characterised by micro-topography that adds complexity and affects
the variability of redox potential (Haraguchi, 1992).

Water table
Water table level controls the diffusion of oxygen in peatlands and several studies have found a
good correlation between oxygen concentration and redox potential at low (6-15%) O2 contents
(Callebaut et al., 1982). Almost all the literature examined in the present work involves research
carried out in wetlands other than bogs, highlighting a general lack of studies on bogs. Generally,
redox potential decreases when water table increases so that they are negatively correlated
(Niedermeier and Robinson, 2007; De Mars and Wassen, 1999) and the change of redox
potential trails shortly the drop of water table (Seybold et al., 2002). In a mineral Calcaric
Gleysol soil, Mansfeldt et al. (2003) showed that the principal variable explaining temporal and
spatial variation of redox potential through a vertical profile was water table (r = - 0.97). De
Mars and Wassen (1999) studied different peatlands and showed that the temporal variation
explained by water table level was 2/3 of the total variance. The authors suggested that soil
conditions like physical properties (capillarity), pH and nutritional status (quality of organic
matter) might have explained the residual variance. Rezanezhad et al. (2014) set up two columns
filled with homogenised riparian soil: in one column they imposed a fluctuating water table
regime whereas in the other the water table was kept stable. Redox potential was measured at
10 cm and - 30 cm every 60 seconds for 75 days and they clearly showed that the imposed
regime controlled the spatial and temporal distribution of the soil redox potential. The authors
observed also short-term spikes during high water table extending for hundreds of mV and
probably caused by gas transport and heterogeneity of water composition. Interestingly, these
spikes did not show up in the column with stable water table, so that soil water circulation could
have been responsible of their generation.
While the degree to which water table level influences redox potential is depth
dependent, the major part of studies on redox potential in wetland soils did not carry out
measurements with depth resolution (Shoemaker et al., 2013; Fiedler et al., 2007). The oxygen
concentration in soil depends on diffusion rate and consumption rate by microbial activity. The
process of diffusion as described by Ficks law is driven by the concentration of oxygen at
25

interfaces, diffusion path length and diffusion coefficient. Water table influences indirectly redox
potential lowering the rate of diffusion of O2 by a factor of 10000 in respect to the gaseous
phase (Gosselink and Mitsch, 2007). Soil properties, like porosity, are also important to develop
the degree of soil aeration; in fact, a strong capillarity can support saturation above the water
table level (Knorr et al., 2009; Thompson et al., 2009; Thompson et al., 2007). Previous studies
have measured that oxygen penetration length in peat can be very short; for example Armstrong
and Boatman (1967) found that oxygen, even in extreme cases, would not diffuse further than 6
centimetres. Benstead and Lloyd (1995) measured oxygen in different hollows in a Sphagnum
bog, when water table was above and below 2 cm from the surface, and showed that, at 2 cm
depth, almost all hollows lacked O2; only in one site, the authors detected oxygen at depth of 5
centimetres. These studies suggest that water table efficiently prevents oxygen to penetrate in
peat, though local peat properties and microbial activity can still be important to determine the
length of the vertical extinction curve of oxygen. It is worth adding that during summer the water
table drop might not entail oxygen penetration to deeper layers if oxygen in the upper layer has
been consumed because of high microbiological activity (Barber et al., 2004).
As mentioned before, many studies have measured decreasing redox potential with
increasing depth. Fiedler et al. (2004) found decreasing redox potential in a wetland mineral soil,
while Thomas et al. (2009) found a decreasing gradient in Florida Everglades wetlands. The
relative importance of sources of water shapes the vertical redox profile: for instance, lateral
water flowing from uplands may be responsible of changes of Eh (Wheeler and Shaw, 1995
cited in Thompson et al., 2009) and upwelling of groundwater (when occurs) was suggested to
be responsible of local increasing of Eh in deeper layers. Considering redox profile again, when
two or more redox probes were employed, the measurements have shown that spatial and
temporal variability of Eh decreased with depth (Thomas et al., 2009; Mansfeldt et al., 2003;
Fiedler et al., 2004). However, Mansfeldt et al. (2003) found that the highest variation of redox
potential was at 60 cm below the surface, where the soil most frequently changed between
saturated and unsaturated condition. Therefore, standard deviation is associated to the extent of
water table fluctuation, because it affects the relative proportion of air and water in the pores. In
conclusion, the effects of water table on redox potential should be limited to the zone of water
table fluctuation, or at most, should extend very shortly above or below the water level.
Contrarily, below that zone, other factors should play a major role.

Precipitation

26

Rainfall is a source of dissolved oxygen and oxidised nutrients, like nitrate and sulphate, which
may temporally increase redox potential in surface peat layers. Rainfall also raises water table
and if rainfall rate exceeds storage capacity, runoff water is produced. The effect of rainfall on
oxidation-reduction cycles may be important in peatland and can be greater during dry periods,
because rain can reach deeper anoxic layers (Deppe et al. 2009). Niedermeier and Robinson
(2006) measured redox potential in a fen during two summer rain events, when water table was
below 20 cm. The authors observed a sharp increase of Eh at 10 cm depth during the first rain
(45 mm d-1) and a broad peak in Eh at 10 cm during the second rain (70 mm d-1). The peaks were
about 400-500 mV. There were no effects at 30 cm depth, below the water table. Precipitation
also raised water table lowering the rate of oxygen diffusion and counteracting the former
oxidising effect of rainwater (Niedermeier and Robinson, 2006). The influence of rain is not
always outstanding, in fact Mitchell et al. (2005) has observed just little Eh increment during
rainfall events in the surface of a peatland, and no influence at layers deeper than 15 centimetres.
Contrarily, in the first 50 cm of peat, Haavisto (1974) has measured an averaged decrease of 47
mV, although the procedure used may have produced unreliable values. Fiedler (1999) observed
that at depths > 30 cm, precipitation influenced the potentials only indirectly by raising the water
table.

Water chemistry
Peatland bog soils are electron acceptors limited, but have plenty of electron donors (i.e. organic
matter). Addition of easy degradable organic matter can enhance the electron pressure and can
reduce redox potential while the input of oxidised species can increase it, as explained by the
Nernsts equation (Reddy and DeLaune, 2008). The redox potential measured at the electrode is
characterised by the dominant redox couple, and it depends on standard rate constant for the
redox couple, concentration of oxidised and reduced species, number of electrons transferred per
molecule and electrode surface (Peiffer et al, 1992). Some redox couples can be present in very
low concentration in oligotrophic bog water. For example, Fe(III)/Fe(II) has a high standard rate
constant, so that is often dominant in mineral soils, but total dissolved Fe is very low in bogs
(Wieder and Vitt, 2006) and the major part is bound to and stabilised by DOC (Steinmann and
Shotyk, 1996), so that its reaction may be insignificant (Keller and Bridgham, 2007). The same
can be said for manganese. Reduced inorganic sulphur (RIS) is much less abundant than
organically bound S in bogs, but it is reported to have a very dynamic turnover (Wieder and
Lang, 1988). The concentration of RIS is higher than iron and it can be more important in a bog,
explaining up to 30 % of anaerobic respiration (Keller and Bridgham, 2007). DOC is not only an
27

electron donor but it can also act as electron acceptor for anaerobic respiration. There is an
increasing need of assessing the use of organic electron acceptors by microorganisms in anoxic
environment. Many studies hypothesised their involvement to account for the great fraction of
unexplained carbon mineralization in low nutrient bogs (Knorr et al., 2009; Deppe et al., 2009).
However, the great importance of DOC has been evaluated only in laboratory and, so far, it has
not been studied directly in the field (Blodau et al., 2009). To complicate the issue, the DOC
chemical composition rather than its amount distinguishes its reducing/oxidising role in
peatlands (Keller et al., 2009).

Microbial community
Microorganisms control oxidation-reduction reactions in soils, modifying their environment by
consuming TEAs and lowering redox potential, which, at the same time, determines the
functional microbial type emergent in the bacterial community (Husson, 2013). Despite the low
nutrient content and pH and the recalcitrance of Sphagnum (Dobrovolskaya, 2014), microbial
community in bogs is frequently reported to be as active as other richer peatlands fens
(Mettrop et al., 2014; Fisk et al., 2003). In bogs, microbial activity is greater in surface layers (015 cm below surface) than subsurface ones (15-30 cm below surface) (Fisk et al. 2003). In
general, the physical-chemical and environmental factors that influence microbial activity will
cause the redox potential to change. Studies that took into account the role of microorganisms in
driving Eh changes in bogs are lacking, albeit an important role for them, as observed in other
ecosystems, might be hypothesized. In a tidal marsh, Catallo et al. (1999) showed that microbial
activity alone produced quick Eh variation to the extent of more than 370 mV in less than 48
hours. Oligotrophy of bog ecosystems represents a constraint to microbial activity and input of
nutrients from deposition or other sources can alter this constraint. Bragazza et al. (2012) have
shown that N deposition could alter microbial communities and favour bacterial growth.
Temperature can influence microbial community composition, growth rate, enzyme synthesis
and response. As a rule of thumb, the decomposition rate of organic matter doubles every 10C
increment, following the Arrhenius equation. However, other environmental constraints affect
the rate of decomposition. These physical and chemical constraints are themselves affected by
temperature and climatic factor like flooding, droughts and freezing, which have the effect to
making the relation complex (Davidson and Janssens, 2006) and giving different Q10 for
different ecosystems (Peng et al., 2009). Nevertheless, regardless of the extent of their effect,
warmer conditions increase microbial activity both in oxic and anoxic environments (Estoparagons and Blondau, 2012). Acidity affects the microbial community. Acidic pH means a
28

greater amount of energy spent by bacteria pumping hydrogen ions out of their cells to survive.
Enwall et al. (2007), showed a negative correlation between soil pH and the microbial metabolic
quotient, which indicate a decreased efficiency to convert organic carbon in microbial biomass in
acidic soils. Notwithstanding these constraints, N-mineralisation and P-mineralisation can be
higher in bogs than fens (Mettrop et al. 2014; Verhoeven et al. 1990 cited in Wieder and Vitt,
2006). It is possible that bacterial groups in bogs are not constrained by pH because they evolved
for that specific acid condition (Dobrovolskaya, 2014). Moreover, it has to be kept in mind that
fungi are less affected by low pH than bacteria.

Vegetation
Plants have the potential to change redox stratification in soil through at least three distinct
processes: (1) release of O2 through roots into the rhizosphere, (2) primary production,
increasing the quantity of labile organic carbon and releasing roots exudates and (3) direct
nutrients uptake into roots, rhizomes, stems and leaves. The plant species can account for
differences in electron acceptor renewal in anaerobic soils. It is reported that in bogs this process
is usually less important than in fens because of lack of sedges (Deppe et al., 2009). During the
day, photosynthetic activity and active transport of oxygen create an oxic environment in the
rhizosphere. Nikolaustz et al. (2008) observed that, in an artificial Juncus effusus wetland,
reducing condition during night (-320 mV) and oxic condition during day (+300 mV) were a
function of light intensity and dissolved oxygen in the rhizosphere. Instead, microbial
consumption of electron acceptors could explain the reducing condition measured at night. The
authors suggested that this pattern might be applicable to all aerenchymatous plants in wetlands.
Respiration of roots produces a dynamic trend of carbon dioxide in soil. Benstead and Lloyd
(1996) incubated a solid peat core extracted from a Sphagnum-Eriophorum bog and observed a
diurnal fluctuation of CO2 at the depth of 15 cm and 20 cm. They found a minimum at 18:00 and
a maximum at 7:00. The authors placed the same peat in dark and fluctuations disappeared,
suggesting a key role of vegetation in the process. Shoemaker et al. (2012) build a mesocosmos
with Myriophyllum acquaticum and Leersia oryzoides and found that an increase of temperature
occurred almost simultaneously with a rise of redox potential leading to daily fluctuation of
about 100 mV in the top 10 cm. The authors did not find the same pattern in a non-vegetated
mesocosm, which suggested that the observed fluctuations were produced by vegetation roots
and not by temperature.

29

Peat structure
Bulk density is measured as the dry weight of peat per unit volume (g cm-3). Bulk density
increases with degree of humification and increases with depth. Porosity is the fraction of soil
volume that is not filled by the solid phase. In Sphagnum peat near the surface porosity is very
high, ranging between 88 and 97 % (Ivanov, 1981). Porosity decreases with increasing
humification, bulk density and degree of compaction, so it decreases with depth. Porosity
correlates positively to hydraulic conductivity (Schaaf, 1999). Porosity can decrease during
water table drawdown (Schlotzhauer and Price, 1999). Peat of different origins can vary in
porosity and in bulk density influencing the capillarity and the air-filled porosity of the peatsoil.
Sphagnum peat may have low capillarity: indeed, Deppe et al. (2009) showed that air filled
porosity sharply increases at 1 cm above the water table in a bog peatsoil. Fen peat with
dominant sedges instead is denser than Sphagnum peat and has higher capillarity, which can
bring to a disconnection between oxic/anoxic boundary and water table dynamic (Knorr et al.,
2009).

Peat chemistry
Peat is composed of an enormous mixture of organic compounds, including carbohydrates
(cellulose, hemicellulose), nitrogenous compounds (proteins, amino acids), phenolics (including
lignin), lipids (waxes, resins, steroids, terpenes) and humic substances (Rydin and Jeglum,
2013). The amount of organic matter, C quality and nutrient availability influence redox
potential through microbial activity. Low input of C, lower nutrient availability and less labile
organic matter can slow the redox pathway catalysed by microbes and the release of electrons,
diminishing the electron pressure and resulting in higher Eh (De Mars and Wassen, 1999). The
acid soluble components (e.g. cellulose) tend to be metabolised by microorganisms more readily
than acid insoluble components (e.g. lignin), so that the relative fraction can be used as index of
degradability. The C:N ratio is an important indices of degradability or recalcitrance, so that
where it is high the decomposition is low (Biester et al., 2014). The relative C:N and C fraction
in litter depends on vegetation type. In Sphagnum bogs, where vascular plants are presents there
is a higher input of lignin (Biester et al. 2014). C:N ratio also varies with plant species, for
example in Sphagnum spp. it is generally around 50-60 but it can reach value of 300 in S.fuscum
litter (Rydin and Jeglum, 2013). Sphagnum tissue is recalcitrant not only because it has high C:N
ratio, but also because it has high content of sphagnum acid in the cell walls (Rudolph and
Sampland, 1985) and other phenolic compounds that act as anti-septic (Dobrovolskaya, 2014).
Phenolic compounds are degraded by phenolic oxidase only in presence of oxygen. Therefore, in
30

anoxic conditions phenolic compounds can inhibit the hydrolases, the main enzyme involved in
peat decomposition (Freeman et al., 2001). New fresh organic matter is added into the acrotelm
from the layer of living Sphagnum spp. and from vascular plants. Different vegetation grows on
hummocks and hollows and they have also different HWT, so that litter chemical precursors and
importance of aerobic respiration differ between the two micro-structures, influencing the quality
and quantity of organic matter that is added to the acrotelm. In the hollows, organic matter is
added directly to the zone of water table fluctuation, while in the hummock it has to pass through
the aerobic body where a great part of the labile carbon is decomposed to CO2. Therefore, by the
time that litter has reached the water fluctuation zone, the amount of labile organic matter added
would be lower below hummocks (Nilsson and quist, 2009). Moreover, the hummocks
S.fuscum and vascular plants (e.g. Calluna vulgaris) should produce litter with higher lignin
fraction and higher C:N ratio than hollows, increasing its recalcitrance. Since the effect of water
table fluctuation on decomposition depends on the intrinsic liability of that zone to sustain
decomposition in terms of nutrients, C quality and microbial community (Artz, 2009), also redox
potential will be affected by these factors by showing different sensitivity to water table
fluctuation. Ombrotrophic plants evolved mechanisms for dealing with a nutrient-deficient
environment. Sphagnum spp. and Eriophorum spp. show strong resorption and recycling of
nutrients during tissue senescence (Wang et al., 2014; Rydin and Jeglum, 2013; Bragazza et al.,
2003), resulting in a nutrient-depleted litter. Microbial activity and decomposition rate will be
controlled by the limiting nutrient, however it is also true that a greater amount of nutrients
sustains a higher microbial activity and decomposition. Nutrients exert an indirect effect on Eh
through microbial activity. Thomas et al. (2009) found that the phosphorous gradient in the
Everglades influenced how sensitive the redox potential was to changing water table only in the
surface layer. Thomas et al. (2009) found that the higher the P content was the lower was the Eh.
So far, to the writers knowledge, there have not been studies that addressed in detail the
difference in nutrients between micro-topographic structures. Regarding the pore-water
chemistry, there is some interesting data in the work of Bragazza et al. (2005), who studied
Ryggmossen bog, and found higher K+ and slightly higher orthophosphate, nitrate and ammonia
in hummocks. However, the authors sampled water using wells rather than rhizons. Other studies
found higher nitrate in hummocks (Frei et al., 2012; Wolf et al., 2011). Bragazza et al. (2005)
also analysed peat samples and found that total P was two times higher in hummocks, total
nitrogen was lower and total K was higher. Along the burial process of peat within Sphagnum
bogs, the variation of litter fraction and C:N depends more on decomposition pattern (climate,
hydrology) than plant species (Biester et al., 2014). In the acrotelm the C:N ratio usually
31

decreases with depth (Wang et al., 2014), while in more decomposed and deeper layer it
increases (Silamikele et al., 2007). As decomposition proceeds the organic matter becomes
enriched in lignin and N, with lignin suppressing decomposition rate and raised N-level
suppressing lignin degradation (Berg and Meentemeyer, 2002). As mentioned before, the
absence of signicant external P input and relatively elevated atmospheric N deposition in
ombrotrophic systems requires efcient internal cycling of P so that the usual increase of C:P
and N:P ratio with depth might be due to preference of P uptake (Wang et al., 2014). Another
factor of recalcitrance varies with depth, Beer et al. (2008) observed in a bog that aromatic and
phenolic functional groups of organic matter increased with depth, suggesting that degradability
decreased accordingly. The humic fraction of peat is important in affecting redox potential and
electron transfer to humic substances in anoxic systems is considered to competitively suppress
reduction of other terminal electron acceptors (TEAs), including CO2 under methanogenic
conditions (Klpfel et al., 2014). Gondar et al. (2005) measured that the extractable fraction of
humic acid (HA) and fulvic acid (FA) was 10 % of the TOM in the fibric layer and 2 % in the
sapric. Peat has high cations exchange capacity due to humified organic matters (Rydin and
Jeglum, 2013). Most cations can be bound to organic matter diminishing the availability to
microorganisms. For example, Fe(III) can be stabilised by complexation to DOC and can still be
found in anoxic and reducing environment (Steinmann and Shotyk, 1996).
Temperature
Nernsts equation describes how temperature directly affects redox potential, but that the direct
effect of temperature may not be relevant (Shoemaker et al., 2012). Temperature exerts a more
important indirect effect triggering microbial activity (Fiedler et al., 2007). A more subtle
indirect effect occurs through the influence on ionic activity, though in this case a shift from 3 to
22 C produces only a shift of around 25-30 mV (de Mars and Wissen, 1999). It is possible that
temperature effects are more important in explaining variance of Eh in anaerobic layer than in
surface layer because the lack of oxygen in the former does not influence redox potential.
Urquhart et al. (1972) studying four different bogs found significant negative relation between
Eh and temperature at the depth of 30 cm. The authors argued that it was a consequence of
microbial activity. The higher sensibility to temperature of decomposition rate of deep
recalcitrant peat than surface peat (Hilasvuori et al., 2013) could corroborate the importance of
temperature effect on redox potential in anaerobic peat. The heat wave varies in different soil
ecosystem, and, for example, Barber et al (2004) did not find seasonal variation (i.e.
temperature) of redox at depth of 40 cm in a wet grassland.
32

pH
Bogs are acid environments and the main source of H+ stems from humic substances and
Sphagnum phenolics. In surface layers, humic substances mainly supply hydrogen ions while at
deeper layers carbonic acid becomes the main source of hydrogen ions (Steinmann and Shotyk,
1996). Humic acids have a more acidic dissociation constant than carbonic acid and for this
reason pH is reported to increase with depth in bogs (Deppe et al., 2009; Steinmann and Shotyk,
1996; Lundin and Berquist, 1990). The Eh responds inversely to change in pH according to the
Nernsts equation: as pH increases, Eh decreases. The most common value in the literature is a
decrement of 59 mV per unit increment of pH at 25C, although experimental values have been
found to range from -6 to -256 mV, with the latter taken in acid sulphate-rich soil solution
(Charoenchamratcheep et al. (1987) cited in Fiedler et al (2007). Given their low concentration
of iron and sulphate, it is preferable to use 59 mV in bogs (Reddy and DeLaune, 2008). The
concentration of hydrogen ions both influences microbial activity and is influenced by microbes,
so that feedback mechanisms are at stake. For example, microbial activity can decrease pH that
in turn slows their metabolism, when stimulated by labile organic carbon, electron acceptors
availability or temperature (Mettrop et al, 2014; Fenner and Freeman, 2011). Redox conditions
that promote the oxidation of reduced species (i.e. organic matters, Fe2+, H2S, NH4+) can
decrease pH (DAngelo and Reddy, 2003). In bogs, the constraint of acidity developed during
droughts on microbial activity is suppressed by precipitation because it increases pH and
enhances microbial activity and decomposition (Fenner and Freeman, 2011).

1.7.1 Assembling the picture


Water table affects redox potential only close to the water level because it abruptly decreases
oxygen diffusion. In the zone of water table fluctuation, Sphagnum peat has low capillarity so it
is expected that water table acts as a defined boundary between oxic and anoxic conditions. Peat
porosity decreases with depth so that, when water table drops at depth, oxygen will penetrate less
easily. The other relevant process that can affect redox potential in surface layer is wet
deposition. Rainwater unfolds a two-tiers and double-fast effect on redox potential; the first
effect is chemical, carrying oxidised species that increase redox potential, and the second effect
is physical, increasing water table and decreasing oxygen diffusion. The effect should be
important only in the first centimetres and should be more marked after intense droughts and
intense precipitation. Below the zone of water table fluctuation, other factors than oxygen
33

penetration and precipitation are important in determining redox potential. First, chemically, peat
and dissolved organic matter (DOM) are very different from mineral soils. In the latter,
microorganisms use inorganic alternative electron acceptors in a sequential thermodynamic
cascade, but in the former the low concentration of inorganic redox couples has never been able
to explain the high CO2 and low CH4 production that was found in bogs. Consequently,
researchers focused on alternative anaerobic pathways and shed light on a class of electron
acceptors that is fundamental in peatland and in bogs: the humic substances, which can occur in
solid or dissolved phase (dissolved organic matter, DOM). Researches showed that in anoxic
conditions, humic substance was able to oxidise acetate (Lovley et al, 1999), sulphide (Heitmann
and Blodau, 2006) and CH4 (Smemo and Yavitt, 2011) and that they could be thermodynamic
favourable for microbes at intermediate values between Fe(III) and sulphate reduction (Klpfel
et al., 2014; Cervantes et al., 2000). Moreover, electron acceptor capacity of DOM could inhibit
methanogenesis (Keller et al., 2009; Heitmann et al., 2007).
The presence of vascular plant roots is likely to affect redox potential. It is not always
clear how roots would drive redox potential. In fact, an important root process is the release of
organic exudates that increase microbial activity and decrease redox potential. These two
mechanisms are likely to act simultaneously in anaerobic soil and Eh could show contrasting and
complex spatial trends. It is possible to hypothesise a different importance of the two processes
in different microforms, given the different vegetation. A factor that can be important both
within and below the zone of water table fluctuation is temperature: increasing temperature
should increase microbial activity and should decrease redox potential. A further
geohydrological factor can influence the redox profile. Depending on the geohydrological
features of the catotelm, during the summer drying upwelling of mineral water can change the
vertical stratification of redox potential. Despite that, the dominance of diffusive forms of
transport in catotelm bogs decreases greatly the occurrence of upwelling. In hummocks the
vertical hydraulic conductivity is higher than hollow and it is dominant over horizontal hydraulic
conductivity so that rainwater will infiltrate, leaching solutes (more labile DOC and oxidised
nutrients) to lower layer. Hummocks though should not be affected by runoff flow as much as
hollows are, so that leached solutes may be entrapped inside hummocks for longer. Redox
potential will be affected by these temporal and spatial changes of chemistry composition,
accordingly to the circulation of oxidised species and DOC. Microorganisms activity, nutrients
availability, TEAs (i.e. terminal electron acceptors) and peat quality decrease with depth. The
quality of DOC, the main source of labile carbon in the catotelm, also decreases with depth. For
this reasons it is sensible to expect a higher spatial and temporal variability of redox potential
34

approaching the surface. This should apply either to the zone of the water table fluctuation or the
catotelm.

1.8 Aim and research questions


The thesis focuses on the temporal variation in redox potential in vertical peat profiles of a
boreal bog and in different micro-topographic structures during the vegetation season. The aim
of the present work was to assess the relation between redox potential and water table dynamic
on one hand, and peat quality (i.e. peat degradability) and decomposition processes.
The research questions were

Is the water table an important control of redox potential?

Are redox potential measurements able to describe spatial decomposition processes?

Is peat quality influencing the value of redox potential?

1.8.1 Hypotheses
The hypotheses were

(i)

Redox potential will decrease with depth, and particularly drop at the water table.

(ii)

Given that DOC is mainly a product of anaerobic respiration and fermentation while
it is consumed by aerobic respiration, the accumulation rate of DOC will be
negatively related to redox potential.

(iii)

In anoxic peat, the production of CO2 is balanced by consumption via autotrophic


methanogenesis, so that more CO2 would be consumed and accumulation rate of DIC
will decrease in correspondence to more reducing redox potential favourable for
methanogenesis.

(iv)

Since microbial activity is facilitated by nutrient availability and labile carbon


fraction, peat quality will be related to redox potential.

(v)

Hummocks and hollows have different peat quality and hence different depth
variation in redox potential.

35

2 MATERIALS AND METHODS


2.1 Field work
2.1.1 Site description
The project aimed to measure redox potential through the vegetation period, between the end of
May and the end of September. The logging period started on 31 May 2014 and ended on 22
September 2014. Unfortunately, the HYPNOS III data logger broke on 17 July and the
technician was not able to repair it immediately. The data gap is showed in figure 3.1 and it
ranged from 21 July to 1 September. Given the gap in the data, after having replaced the
HYPNOS III, it was decided to continue the measurements until the 24 October. Still, the
incident probably caused by a summer thunderstorm, spoiled some sensors so that it was
necessary to leave them out from analysis. The period when redox data lacked corresponded to
the lowest water table level (figure 3.1). Therefore, it was not possible to measure redox
potential when water levels fell below 18 cm from mean hollow surface.
The research project took place at Ryggmossen (60 2 40.88 N, 17 19 51.05 E), a
boreo-nemoral bog situated in central Sweden, devoid of anthropogenic disturbance and that
from 1997 has been declared a protected natural reserve. The bog originated from the filling up
of a lake-basin with bedrock consisting of igneous rock, and the depth of peat increases from the
margin to the centre where it reaches a thickness of ca. 4.5 m (Bragazza et al., 2003). The bog is
located 25 km north-west of Uppsala; its area is ca. 60 ha with a nearly circular shape
surrounded by forest dominated by Pinus sylvestris and Picea abies.

36

Figure 2.1 The picture shows the lands north to Uppsala. The thick white arrow indicates Ryggmossen bog
(circular shape, light green). Images: 2015 TerraMetrics; map data: 2015 Google

Ryggmossen can be classified as a raised bog, which means that it has a convex surface that is
higher than the surrounding upland. The difference in height between the mire margin and the
inner portion is ca 1.5 m. The bog is characterised by a distinct micro topography of hollowlawn-hummock, which are typical shapes of northern ombrotrophic peatland. Proceeding from
the forest toward the bog centre, the first feature encountered is the lagg, where bog and
minerogenous groundwater meets. The lagg is habitat for graminoids (Carex lasiocarpa,
Phragmites australis), peat mosses (Sphagnum cuspidatum, S. fallax) and other plants. Further to
the centre, there is the bog margin, which is well drained and sparsely wooded with low P.
sylvestris and shrubs like Calluna vulgaris and with S. fuscum hummocks and S. angustifolium
hollows. Finally, the bog expanse is characterised by hummocks in concentric strings and
hollows. On the hummocks the dominant plants species are S. fuscum and C. vulgaris , while S.
balticum, S. cuspidatum and S. tenellum grows together with Eriophorum vaginatum in the
hollows.

2.1.2 Experimental design

37

The research site is located in the northeast bog expanse, shown in figure 2.2, where a
water level station has been previously installed, providing a long history of water level time
series. Peat depth in this location is about two to two and half metres.

Figure 2.2 Ortophoto of Ryggmossen bog. The research site is indicated with the red point. Images: 2015
Ones/Spot Image, Digital Globe, Lantmteriet/Metria; map data: 2015 Google.

The research site was characterised by a definite hummock and hollow micro-topography. In the
present work, the term hummock was used according to Sjrss classification, while the term
hollow was used as general term to indicate carpets. Two micro-topographic pairs were
investigated each one composed of two close microforms or hummock-hollow elements. The
microforms were chosen within a range of approximately 25 m from the water level station. The
northern pair was named A and the southern one was named B. Figure 2.3 shows a schematic
map of the site.

38

Figure 2.3 Schematic map of the research site. Northern microforms belongs to station A and southern
belong to station B.

Each microform was equipped with in-situ redox probes and water samplers in order to take
vertical profile of redox potential and peat pore water. Each redox probe had six platinum
electrodes positioned at intervals of 12 cm and sealed into a fibreglass rod. The water samplers
consisted in PVC tube of different lengths with RHYZONS connected to vacuum syringes
(detailed information below).
The water level station measured the absolute level of water table in metres above sea
level. Therefore, the position of the bog surface could have been measured in relation to water
level in order to obtain the HWT. Assuming that the water table within the research area was flat
and taking the water table level as reference, the redox probes were inserted in each microform
in order to have redox sensors positioned in relation to water table level. On 27 May, the site was
equipped and water level and HWT was measured at each microform and at the water level
station. HWT measured 6 cm at the water level station, 6 cm at A and B hollows, 44 cm at A
hummock and 32 cm at B hummock. Six water samplers were inserted in each microform, close
to and at the same depth of the redox sensors. They were stuck into the peat with an angle of 45,
in order to prevent potential leakage of rainwater from the surface. One further redox sensor and
one water sampler were inserted at 150 cm depth, close to B hollow, in order to have data form
deeper groundwater. Details of the installation are given in table 2.1 while figure 2.4 shows the
experimental design for one station.

39

Table 2.1 Coding and information of the objects of the experimental setting.

Code

Station

Structure Level

Depth;
HWT (cm)

Altitude (m a.s.l)

Ahum1

Hummock

60,204

Ahum2

Hummock

20

60,084

Ahum3

Hummock

32

59,964

Ahum4

Hummock

44

59,844

Ahum5

Hummock

56

59,724

Ahum6

Hummock

68

59,604

Ahol1

Hollow

59,844

Ahol2

Hollow

18

59,724

Ahol3

Hollow

30

59,604

Ahol4

Hollow

42

59,484

Ahol5

Hollow

54

59,364

Ahol6

Hollow

66

59,244

Bhum1

Hummock

60,084

Bhum2

Hummock

20

59,964

Bhum3

Hummock

32

59,844

Bhum4

Hummock

44

59,724

Bhum5

Hummock

56

59,604

Bhum6

Hummock

68

59,484

Bhol1

Hollow

59,844

Bhol2

Hollow

18

59,724

Bhol3

Hollow

30

59,604

Bhol4

Hollow

42

59,484

Bhol5

Hollow

54

59,364

Bhol6

Hollow

66

59,244

Deep

Deep

Deep

150

58,344

40

Figure 2.4 Sketch of the vertical profile of station B hollow, hummock and deep redox sensors and water
samplers. The only difference from station A is that in A there is no Deep sampler. The horizontal dotted
lines represent max and min water table fluctuation in Ryggmossen from historical data.

2.1.3 Redox measurements


The redox potential was measured by a Hypnos III data logger (MVH Consult, Leiden, the
Netherlands; Vorenhout et al. 2011) that can run autonomously with lithium batteries. Hypnos III
was composed of a waterproof box enclosing the hardware and LCD screen (figure 2.5), where
one could check the status of the logger and can command the downloading of data to a SD
memory card.

41

Figure 2.5 Hypnos III redox data logger after installation. Photo M. Vorenhout.

In this study, a 3M Ag/AgCl (QIS, the Netherlands) reference electrode was used, whereas
platinum wires were used as working electrode. Hypnos III was set to collect redox measurement
every 15 minutes. Standardised redox potential (Eh) was calculated adding the measured redox
potential (Em) to the potential of the reference probe (Eref)

Eh Em Eref

Eq. 2.1

Afterward, Eh was corrected for pH using formula 2.2

Ehcorrected Eh 59 pH 7

Eq. 2.2

All redox potential measurements are reported in mV.

2.1.4 Water table measurements


A Trutrack water height logger recorded the position of the water level every hour. It also
recorded air temperature and peat temperature at 75 cm below the surface.

42

2.1.5 Pore water sampling


In the present work, Rhizon water sampler (pore size 0.16 m, length 10 cm, diameter 3 mm,
bre glass support, Eijkelkamp, Giesbeek, The Netherlands) have been employed to collect pore
water (figure 2.6).

Figure 2.6- Detail of the ceramic cup tip of the Rhizon sampler.

Rhizon is composed of a pipe that connects a


ceramic cup filter to a syringe. The syringe is
connected to the tube with a valve that can
regulate the flux. Ceramic cup filters are made
up of pores of 0.16 m diameters. The
ceramic cup is introduced into the soil and the
syringes piston is pulled to produce negative
air pressure, which sucks up filtered water
from the soil. The syringes piston can be
secured at required volume with a nail. The
following procedure was applied: when water
reached the required level, the needle was
removed and the piston was slowly returned
to normal position. Before opening the valve,
sampled water was left five minutes in the
syringes in order to equilibrate with gases.
After five minutes, the water was poured into
the sample bottle. 20 ml of water was
collected for cations analysis in PE bottle and
acidified in the field with 140 l of a solution

Figure 2.7 Rhizons, location: A hummock. The tube is


inserted in peat at a determinate depth and connect the
ceramic cup to the syringe for collecting pore water.

43

of 65% HNO3. Approximately 15 ml was extracted to measure pH and EC directly in the field
with a portable analyser. 60 ml of water was collected to analyse anions and carried inside a cold
bag to the freezing room in the lab. The samples for anions were stored at -20 C, while the
samples for cations at 4C. Finally, 30 ml of water was collected and left inside the syringe
without opening the valves. The syringes were stored in a cold bag and carried to the lab. The
samples were stored in a cold room at 4 C and analysed for DIC and DOC the day after. pH and
electric conductivity (EC, S.cm-1) were measured directly in the field with a portable HI 98129
HANNA Instruments, that was calibrated the same day in laboratory with standard solutions.
The instrument automatically corrected both pH and EC to 25 C. EC was corrected for pH
according to the following equation (Rydin and Jeglum, 2013):

Ecorr EC (3.25105 10 pH )

Eq. 2.3

2.1.6 Decomposition
Weight loss of a determinate degradable material is an index of decomposition in soil. Pure
cellulose was used in order to assess the decomposition of labile organic carbon. Each cellulose
filter was enclosed in a nylon envelope sieve with a mesh size of 16 m, so that only bacteria and
fungi could reach the cellulose. The cellulose filters were attached at different depths to a PVC
rod in order to evaluate decomposition in three functional zones: the aerobic zone, the partially
aerobic/anaerobic zone and the anaerobic zone. The filters were positioned on 29 May and
retrieved on 22 September, after 117 days. In the laboratory, the filters were carefully washed
and dried in a stove at 70C for 48 hours. Afterwards, they were shipped to the University of
Antwerp where they were put at 105C for 24 hours and then weighed. Each filter was put in a
ceramic cup and then it was burned in a muffle oven at 550C for 4 hours to measure the mineral
content. Finally, weight loss was calculated subtracting final weight to initial weight, where the
initial weight for each filter, given the low variation, was taken equal to the mean of three filters.
The Tea Bag Index is a simplified method for measuring decomposition rate and it is
based on the use of standardised commercial tea bags, with known chemical composition, as
representative dead plant material. The Tea Bag Index integrates two indicators representative of
the decomposition process. The stabilisation factor (S) accounts for the transformation of labile
organic fraction into recalcitrant fraction as decomposition proceeds and, since it depends on
temperature and moisture factors, it is an indicator of environmental constraints. The
decomposition rate (k) is a constant rate calculated applying an exponential asymptotic model of
44

decomposition to the measured loss of labile organic material. The labile organic fraction is
composed by the sum of non-polar extractive (e.g. fats ad waxes), water-soluble (e.g. simple
sugars and phenolics) and acid-insoluble (e.g. cellulose). The tea bags (Lipton, Unilever) were
placed in two hummocks and two hollows close to the ones supplied with redox probes and
cellulose filters, because there was not enough space and in order not to disturb the redox
measurements. The bags were weighed, enclosed in a nylon mesh, tied with plastic rope and
inserted with a wood stick into the peat on 7 July, down to every 12 cm, starting from 6 cm
below the surface in the hollows and 8 cm below the surface in the hummocks. The plastic ropes
that emerged from the peat were secured to a plastic stick in order to mark the location. They
were retrieved on 22 September, after 77 days. After collection, the tea bags were cleaned from
peat remains and they were put in a stove at 70C for 48 hours. They were subsequently weighed
and the calculation of the Tea Bag Index was done according to the protocol and the equations
presented in Keuskamp et al. (2013).

2.1.7 Peatsoil sampling


The peat sampling was carried out with a Russian peat core. The peat was taken on 22
September, one metre away from the redox probes in order not to disturb the measurements. The
peat was sampled from just below the living Sphagnum down to 100 cm. At location A, the peat
was sampled at the slope of the hummock, because the mound could not provide enough space.
Near the Deep redox probe, a peat core was extracted from 100-200 cm. The peat cores were put
in PVC containers and carried to the laboratory where they were sealed and stored at 4C for
further analysis.

2.2 Laboratory work


2.2.1 Water chemistry analysis
Analysis of DIC and DOC was carried out with a Shimadzu TOC-L with sample changer
ASI-L (Shimadzu Corporation, 1. Nishiokyo-Kuwabaracho, Nakagyo-ku, Kyoto 604, JAPAN).
Injected volume was 50 l for TC and 50 l or 325 l (at conc. <5 mg/l) for IC. The average of
three replicate injections (out of max. six) was used for calculations. For total carbon (TC)
analysis injected samples are combusted (catalytic) at 720 C in a stream of carbon-free air. After
cooling, drying and halogen removal the formed CO2 is detected in a NDIR infrared detector that
gives peaks that are proportional to the carbon content. Peak areas are compared to areas for
45

calibration samples (potassium hydrogen ftalat) with known content of carbon. For total
inorganic carbon (TIC) injected samples are acidified with phosphoric acid and sparged with
carbon-free air. All IC is converted to CO2 which is then detected the same way as for TC. The
calibration samples are made up from Na2CO3 and NaHCO3. Total organic carbon (TOC) is then
calculated as the difference between TC and TIC.
In this report, TIC is named dissolved inorganic carbon (DIC) and TOC is named
dissolved organic carbon (DOC).

2.2.2 Peat analysis


The peat cores, when still wet, were analysed for determining the von Post degree of
humification. The von Post degree was identified using the structure of the plant remains (Table
2).

Table 2 - The von Post scale, H1H10 (von Post 1924; from Rydin and Jeglum, 2013).
Degree of
humification
H1: Completely
undecomposed
H2 Almost completely
undecomposed
H3 Very slightly decomposed

H4 Slightly decomposed
H5 Moderately decomposed
H6 Moderately to strongly
decomposed

H7 Strongly decomposed

H8 Very strongly
decomposed

Plant structures
Unaltered

Distinct

Distinct
Mostly distinct, but a slight
amount unidentifiable
Mostly distinct, but some
indistinct
Mostly indistinct, some
structures identifiable
Very indistinct, some
structures identifiable
Very indistinct, a few
resistant structures like roots,
wood

H9 Almost completely
decomposed

Hardly any recognizable


structures

H10 Completely decomposed

No discernible structures

The top layers containing the remains of living Sphagnum were removed from the peat cores.
The remaining peat cores were cut in sections of 12 centimetres, in order to have the subsamples
46

centred on the level of redox sensors. The length of the first section was less than 12 cm because
living Sphagnum was removed from the top. Each subsample was weighed and then oven dried
at 40C for two weeks. To calculate the bulk density (g.cm-3), the volume of the subsample was
calculated measuring the diameter of the peat corer and calculating its cross sectional area (11.03
cm2) and the peat subsamples length. To calculate the fibres content, the method illustrated in
Sneddon et al. (1970) was implemented. A longitudinal slice of approximately 1.5 g was cut
from the peat subsample and dried at 105 C for 4 hours. Afterwards, the samples were weighed
to obtain the total dried weight. The samples were rewetted for 24 hours, then 25 ml of a
dispersing solution of sodium esametaphosphate + sodium carbonate (25 ml = 0.02 g) was added
and the samples were incubated overnight. The day after, the sample were poured in a sieve with
mesh size of 0.150 mm and gently rubbed with the fingers under a light water jet, in order to
dissolve the lumps of peat. The procedure continued until the water discharged was clean. The
fibrous material that remained on the sieve was dried in oven at 105C for 48 hours and
weighted.
Peatsoil was classified according to the Rfrentiel pdologique (AFES, 2008). Peat soils
belong to the class of Histosols and three pedologic horizons are described, mainly determined
by the fibres content and von Post scale. A short description of the horizons follows
Fibric horizon (Hf): the content of fibres is between 40 to 100 % in dry weight. The
decomposition is from null to very low. The vegetation structures are identifiable and peat is
devoid of amorphous organic matter. In the von Post scale ranges between 1 and 5. The peat is
not heavy and the liquid is amber coloured. The bulk density is < 0.1 gcm3
Mesic horizon (Hm): the content of fibres is between 10-40 % in dry weight. The decomposition
is advanced, the vegetation structures (woods, grasses and mosses) are almost indistinct and the
proportion of amorphous organic matter is elevated. In the von Post scale, it ranges between 5
and 8. It is more solid that the sapric Horizon and often the colour of the liquid is brown. The
bulk density is between 0.1 to 0.2 g.cm-3.
Sapric horizon (Hs): the content of fibres is less than 10 % in dry weight. The decomposition is
strong and plant structures are not recognizable. In the von Post scale, it ranges between 8 and
10. The peat is a heavy paste, and the fluid is dark. Hard fragments of ligneous non-decomposed
plants can be still present. The bulk density is close to 0.2 g.cm-3.
The remaining analysis required that the peat samples were grinded. Each sample was
crushed with a ball mill with Agatha balls for two minutes and 450 rpm. If, after the treatment,
the sample still carried coarse remains it was ground again.

47

MINERAL CONTENT

For the determination of the mineral content, the following procedure was used for each sample.

Procedure

The ceramic cups was weighed

1 g of the peat was put in the ceramic cup

The ceramic cup was placed in a oven at 550 C for two hours

The ceramic cup containing the ash was weighed at room temperature

Calculation
Mineral content (M.C., g.g-1) is calculated
M.C. =

ba
100
s

Eq. 2.4

b = ceramic cup + initial sample weight before the oven treatment


a = ceramic cup + sample weight after the oven treatment
s = initial sample weight

ORGANIC AND INORGANIC PHOSPHOROUS

For the determination of organic and inorganic P, the procedure as described in Arduino &
Barberis (2000) was used.

Method

The grinded-peat sample was split in two parts, one was burned at 550 C in an oven and the
other was not burned. Then, the two samples underwent the same extraction procedure. The
treated sample gave the total P, the untreated sample gave the inorganic P and the organic P
results from the difference.

Procedure
48

The ceramic cup is weighed

1 g of the peat sample is put in the ceramic cup

The ceramic cup + sample is put in the oven at 550 C for two hours

The ceramic cup is cooled and the ash content is transferred in a Falcon tube

Another 1 g of the grinded-peat sample is put directly in a Falcon tube without


any pre-treatment

25 ml of H2SO4 0.5 M is added to both samples

The Falcons are mixed for 16 hours

After mixing, the untreated sample is centrifuged in a centrifuge Joan C312

2 ml of the treated sample is poured in a volumetric flask

10 ml of the untreated sample is poured in a volumetric flask

30 ml of reagent sulph-molybdic is added to each volumetric flask

The solution is left for 1.5 hours to react

After that time deionised water is added to 50 ml

The absorbance value is measured with a spectrophotometer

Calculation

The calibration curve in figure 2.8 is used to calculate the amount of P from absorbance.

95
0

85
0

75
0

65
0

55
0

45
0

35
0

25
0

15
0

90

60

40

20

4
3.5
3
2.5
2
1.5
1
0.5
0
-0.5
1

P (mg/kg)

Calibration line

Absorbance

Figure 2.8- Calibration curve of the spectrophotometer obtained with reference sample set.

The dilution factor is calculated as


49

D.F. =

e 50

p a

Eq. 2.5

e = H2SO4 (ml)
p = sample weight (g)
a = volume of treated/untreated solution added in the volumetric flask (ml)
50 = volume of solution in the flask after reaction (ml)
Phosphorous content (mg.kg-1) is calculated using the curve in figure 2.8 as
0.005260 (0.001839 ABS) (0.0000011612 ABS 2 )
P = D.F .

352 ABS 3 ) (0.0000000000


027155 ABS 4 )
(0.0000000022

Eq. 2.6

ORGANIC CARBON AND TOTAL NITROGEN

Carbon and nitrogen content were determined with a Thermo Fisher Flash 1112 EA NC soil
analyser.

Procedure

Weigh about 5 mg of grinded-peat sample and put in a tin capsule and seal it

Introduce the capsule into the elemental analyzer

Calculation
No calculation is needed, the organic carbon and total nitrogen are expressed in g kg-1

CELLULOSE AND LIGNIN DETERMINATION

There are various methods to determine the relative fraction of structural plant constituents. In
this report the so-called forage fibre method or proximate analysis (Gessner, 2005) is used,
which consists of determining the residual weight of samples following successive removal of
50

various tissue constituents. The method allows the separation of three categories: (a) extractives
(non-polar compounds such as fatty acids and lipids and polar compounds such as phenolics), (b)
cellulose plus hemicellulose, (c) acid-insoluble aromatic compounds (lignin) (figure 2.10). Each
of these compounds decomposes at different rate and may influence nutrient dynamic.

sample

Ashed
ASH

Extract in H2SO4
and CTAB solution

EXCTRACTIVES

Insoluble residue =
Acid-detergent fiber
Exctract in 72 % H2SO4

CELLULOSE

Acid-detergent lignin + ash

Ashed
Ash free (acid-detergent) lignin

Figure 2.9 Scheme of the proximate analysis of carbon fraction, redrawn from Ryan et al. (1989).

Reagents

Acid detergent solution:


o Sulphuric acid 0.5 M 1000 ml
o Hexadecyltrimethylammonium bromide = Cetyltrimethylammonium bromide
(CTAB), 20 g/l

Acetone

Sulphuric acid, 72% by weight

Instruments

Beakers volume 600 ml

Analytical balance, precision 0.1 mg

Hot plate for boiling H2O


51

6 crucibles, Gooch type, porosity no. 2, volume 50 ml

Heat digester holding 6 glass tubes

Pump for creating vacuum in filtration systems

Desiccator

Drying oven set at 105 C

Muffle furnace set at 500 C

Procedure

Weigh clean and oven-dry crucibles to the nearest 1 mg

Weigh 250 mg of the sample with precision of 1 mg

Add 20 ml of acid-detergent solution to tubes

Put tubes in heat digester (put the tubes when the plates are around 170C and let the
temperature drop to 130C, afterward keep 130C constant)

Boil for 60 min after onset of boiling

Filter tube content on a tared Gooch crucible set on a filter manifold. Wash with hot
water (90-100C) until discharge water is clean

Repeat wash with acetone until it removes no more colour. Break up all lumps so that the
solvent comes into contact with all particles of fibre

Dry overnight at 105 C, then put in desiccator and weigh

Acid-detergent fibre (ADF) determination


ADF (% ) = (tared Gooch + initial sample weight) ( Gooch + sample after treatment) x 100
initial sample weight

Eq. 2.8

Put the Gooch filter upon a beaker in order to recover acid. Cover the contents of the
crucible with cooled 72% H2SO4 and stir with a spatula smooth paste breaking all lumps

Refill with 72% H2SO4 and stir at hourly intervals as acid drains away. Crucibles
do not need to be kept full at all times, but samples must be covered continuously

Filter off after 3 hours as much acid as possible with vacuum. Wash contents abundantly
with hot water until free from acid

Dry crucible overnight at 105 C

Weigh crucibles with residual sample

52

ADC (%) = (tared Gooch + ADF sample) ( Gooch + sample after treatment)

x 100

Eq. 2.9

initial sample weight

Ignite crucible in a muffle furnace at 500 C for 2 h and then weigh

Acid-detergent lignin determination

ADL (%) =

(tared Gooch + ADC sample) ( Gooch + sample after ignition)


initial sample weight

x 100

Eq. 2.10

2.2.3 Data analysis


Data was processed with the statistical software R (R Core Team (2014); R Foundation for
Statistical Computing, Vienna, Austria. URL http://www.R-project.org/.)
The principal source of DOC is degradation of OM by exoenzyme and fermentation
processes (Moore, 2009) so it was used as a proxy for anaerobic decomposition. To investigate
the relation between DOC and redox potential it was assumed that the balance between DOC
production and consumption would have been reflected by the difference between DOC at the
end and DOC at the beginning of the research period, which is in other word a rate of
accumulation. The data from hummocks above the mean hollow surface was excluded because
greatly influenced by evapotranspiration and concentration effects. The rate of

DOC

accumulation (rDOC) was calculated as


rDOC = DOC22 September DOC7 July

Eq. 2.11

Since the redox potential distribution was skewed, the median was chosen over the mean to
describe central tendency of redox potential. The median was calculated for each sensor. Finally,
rDOC was plot against median redox potential. The correlation was evaluated against a linear
regression.
The relation between DIC and Eh was based on the same assumption as DOC vs. Eh. A
further restriction was that the relation was investigated only for the depths were degassing was
53

assumed negligible (i.e. below lowest water table level), so that the only consumption of DIC
relied on methanogenesis. The median redox potential was used to characterise the different
redox depths.
rDIC = DIC22 September DIC7 July

Eq. 2.12

Finally, rDIC was plotted against median redox potential. The correlation was evaluated against
a linear regression.
To assess the difference between aerobic and anaerobic respiration on cellulose
decomposition using redox potential, it was decided to use 400 mV as threshold for aerated soil
(Husson, 2013; Gosselink and Mitsch, 2007). According to Reddy and DeLaune (2008), 300 mV
is the limit under which oxygen is completely depleted, though for the sake of this analysis it
was decided that the threshold of 400 mV would represent better the effect of a developed
aerobic microbial community. Since oxic redox potential was measured only at 6 cm, the
analysis was performed using the sensors and the cellulose filters at that depth. The A hummock
was excluded from the analysis because of lack of redox data. For each sensor, the time in which
redox potential was above 400 mV was divided by the total time and the results expressed as
percent of time the sensor measured aerobic conditions. Percentage of cellulose weight loss
served as measure of decomposition. Finally, for each site, percentage of weight loss was
confronted against percentage of aerobic redox potential.
Total P and N, acid-detergent lignin (ADL), C:N and ADL:N were chosen as proxy of
peat quality and plotted against median redox potential. The regression analysis was carried out
only on the samples where redox potential sensors were not affected by water fluctuation, since
oxygen would have masked the difference in redox potential caused by peat quality factors
alone.

3 RESULTS AND DISCUSSION


3.1 Introduction
The time series of water table level measured in relation to mean hollow surface (zero
topographic level in this work) are shown in figure 3.1. The time series showed that water table
quickly rose after rain events, while it required longer time to drop. It is well known that bogs
are adapted to retain water during drier periods lowering the level of evapotranspiration (Schaaf,
1999). There were two drying periods, one lasted from beginning of July to beginning of August
54

and the other spanned from 25 August to 20 September. A period of frequent precipitations,
where water table slowly approached the surface, followed the first drying. Instead, the second
drying ended with a great precipitation event that, in one time, restored the water table to the
surface. In Rydin and Jeglum (2013) water table fluctuation measured in Ryggmossen for
significant years during the vegetation season (June-September) is reported. During the wettest
year (2012), the water table was stable at the hollow surface. During the last two driest years
(1983 and 2010), the water table dropped down respectively to 35 cm and 30 cm below mean
hollow surface. Compared to these years, it is possible to state that the present year represented
an average between wettest and driest historic conditions.

Figure 3.1 - Water table time series measured relatively to mean hollow surface (the zero level on the y-axis).
The two vertical dotted lines enclose the period where redox data lacked. The arrows on the top of the graph
represent water sampling date. Horizontal dashed lines represent the position of redox sensors.

In the present research, July was the month with the highest air temperature (figure 3.2). Water
temperature slowly increased from June to approximately half August and then decreased. From
approximately the second half of September air temperature fell below water temperature.

55

Figure 3.2 - Water table and temperature measured at the water table station during the research period.

3.2 Redox potential and water table

Results
The current work investigated redox potential in vertical profiles, among micro-topographic
structures and among different sites. The following figures show redox potential time series at
different depths combined with water table dynamic. After restoring the HYPNOS III, the sensor
Ahol6 and all the sensors of A hummock did not respond reliably, and were deleted from
analysis. During the vegetation season, water table never fell below 20 cm from the hollow
surface. Instead, considering the time in which the HYPNOS III was working properly, water
table never fell below -15 cm (figure 3.1). Only at 6 cm the redox sensor laid frequently in
saturated/unsaturated condition.

56

Figure 3.3 - Redox potential measured at different depths at A hollow. Blue line shows the water table as
measured nearby A. Data not presented for Ahol6 during autumn period.

Figure 3.3 shows the redox potential time series at the hollow structure of the station A. Redox
potential measured at 6 cm had broad fluctuations (ca. 900 mV) that seemed influenced by
water table dynamic. For example, on 15 June redox potential shifted from almost 300 mV to
400 mV in three hours. Afterward, in the following four days, it slowly approached 660 mV and
again dropped below -200 mV in about 24 hours. Apart from Ahol1 (-6 cm) the other sensors did
not show any particular fluctuation. During the vegetation season, there was no vertical
stratification of redox potential: the upper layers were more reduced than underlying layers.
Ahol1 measured the most reduced environment of the redox profile (- 300 mV) during the
summer. After the beginning of September, the situation changed. Water table increased and
redox potential measured at 6 cm decreased but not as much as during the summer.

57

Figure 3.4 - Redox potential measured at different depths at A hummock. Blue line shows the water table as
measured nearby A. Data not presented for autumn period.

Figure 3.4 shows the redox potential time series of the hummock structure of the station A. The
hummock body was characterised by stable oxidised condition during the period with available
data (Eh > 600 mV). The redox channel at - 6 cm had broad redox potential fluctuations,
seemingly correlated to water table dynamic. The Eh reached -200 mV for a short period at the
beginning of June and thereafter it stood in oxidised condition. The redox potential at other
sensors did not show any particular fluctuation, except from Ahum6.

Figure 3.5 - Redox potential measured at different depths at B hollow. Blue line shows the water table as
measured nearby A

58

Figure 3.5 shows the redox potential time series of the hollow structure of the station B and of
the Deep sensor. The data showed strong variation at 6 cm, however, redox potential seemed
not closely correlated to water table dynamic, unless when water table fell below 6 cm on the
second half of July, when it shifted to oxidised condition. The sensor at - 18 cm depth showed a
clear trend toward positive redox potential as the water table decreased during July. Similarly to
station A the shallow sensor was the most reduced of the whole redox profile during the summer.
In September-October, the redox potential measured at 6 cm decreased again but were not
below 200 mV. Instead, the redox potential at -18 cm decreased and reached highly reduced
conditions (- 300 mV). The redox potential at the other sensors converged and approached 200
mV.

Figure 3.6 - Redox potential measured at different depths at B hummock. Blue line shows the water table as
measured nearby A

Figure 3.6 shows the redox potential time series of the hummock structure at station B. Redox
potentials above the water table were oxidised (Eh > 600 mV). The Eh at - 6 cm had a first swift
drop from 600 mV to 200 mV but it quickly returned to over 600 mV where it stood on for the
entire period of measurement.

To have a deeper insight into the trends of vertical stratification and dynamic of redox
profile, the mean and standard error for every sensor were calculated and afterward examined.
The descriptive parameters were calculated and sorted over three different periods, represented
in the graph by different colours (figure 3.7). These periods were selected because they
59

represented different system conditions, defined by water table position, and to evaluate the
temporal progression of the system looking at trends in mean and standard error. The black
colour refers to the wet period from the beginning of measurement to the first time the water
table dropped below 6 cm (31 May-16 June). The red colour represents the dry period and it
refers to the following period excluded the period of missing data (17 June-21 September). The
blue colour represents the rewetting period and it refers to the period after the water table level
has risen above 6 cm (22 September-24 October).

Figure 3.7 - Mean and standard error calculated for three periods. Black colour refers to first wet period; red
colour refers to drying period; blue colour refers to rewetting (data missing for Ahum). The dotted horizontal
lines represent the max and min water table level during the given period.

Figure 3.7 shows that the sensors located in the zone of water table fluctuation had the highest
variability, excluding Bhum3. During the initial wet period redox potential slightly increased
with depth in A and B hollows (including Deep). The variability of redox potential below the
60

water table fluctuation did not vary much with depth. During the drying period, the A
microforms showed that in the zone of water table fluctuation Eh increased but decreased in the
rest of the underlying layers. In B structures the situation was different: in the zone of water table
fluctuation Eh increased but, in the catotelm it did not show any significant difference from the
previous period except in the hollow at 18 cm, where it decreased. The period of rewetting,
showed that redox potential in the zone of water table fluctuation decreased and it decreased in
every sensor in the catotelm and in all the structures. In B hollow it decreased particularly at 18
cm. The sensor located at - 150 cm, kept decreasing during the season passing from around 0 mV
to around 200 mV.

The following figures investigate the relation between redox potential and water table.
Each figure sorts the sensor by station and absolute level, so that, for example, figure 3.8 shows
the sensors Ahum4, Ahol1, Bhum3 and Bhol1 that are at the same absolute level (i.e. - 6 cm
below mean hollow surface). The HWT in the x-axis is calculated in relation to redox sensor
level, so that when it is negative the sensor is below water table. The figures with no significant
relation were not displayed. The relation was examined over two periods, the first period referred
to the time before the redox logger had broke and the second period referred to the time after it
was restored. That was chose for clarity and to test if there might have been differences between
early summer and late summer.

Figure 3.8 - First period, relation between redox potential measured at 6 cm and HWT relatively to the
same sensor. Left panel refers to A station and right panel refers to B station.

61

Figure 3.8 shows that in A station and at 6 cm the relation between redox potential and HWT
was positive (i.e. relation between redox potential and water table level was negative). When
both sensors were below water table, the hollow reached more reduced value than hummock. In
station B, the relation was still positive but much less clear than A, and seemed that oxidised
condition persisted while the soil was saturated and vice versa. Yet, the B hollow redox potential
showed sensibility to water table fluctuation but B hummock seemed not sensible to water table
at all.

Figure 3.9 - Second period, relation between redox potential measured at 6 cm and HWT relatively to the
same sensor. Left panel refers to A station and right panel refers to B station.

Figure 3.9 shows the relation between redox potential and height above water table at -6 cm for
the second period. The A hollow showed that the relation between HWT and redox potential was
not as strong as the first period but that was still positive. In B hollow the sensor at 6 cm
showed positive relation but less marked than in the A station. B hummock did not show any
relation.

62

Figure 3.10 - First period, relation between redox potential measured at 18 cm and HWT relatively to the
same sensor. Left panel refers to A station and right panel refers to B station.

Figure 3.10 shows the redox potential at 18 cm for the first period of measurement. The A
microforms did not show any relation. In B hollow, the sensor at -18 cm expressed a slightly
positive relation with HWT. In B hummock redox potential was not affected.

Figure 3.11 - Second period, relation between redox potential measured at 18 cm and HWT relatively to the
same sensor. Left panel refers to A station and right panel refers to B station.

Figure 3.11 shows the redox potential at 18 cm for the second period. The A hollow did not
show any relation with water table. The great jump between the values on the left of the panel to
the right is caused by the quick rise of water table occurred in the second half of September. For,
the B station the relation was slightly positive.
63

Duration lines are frequency curves of discharge used in hydrology to investigate the
regime of rivers. In analogy, redox duration line can be used to illustrate the fraction of time that
redox potential was below a certain level. When redox zones are used to track these levels, it is
possible to investigate the dominant redox processes mediated by microbes occurring in the soil.
In the following figures, redox zones are identified by dotted horizontal lines. These zones were
taken from Husson (2013): dissolved oxygen > 400 mV; anaerobic conditions < 400 mV; nitrate
reduction < 250 mV; manganese reduction < 225 mV; ferric ion reduction between +100 and 100 mV; sulphate reduction between -100 and -200 mV; carbon dioxide reduction < -200 mV.

Figure 3.12 Redox potential duration lines at A hollow. Colours refer to different depths. The horizontal
dotted lines represent microbial reduction thresholds. The dominant redox process is written between the
lines.

Figure 3.12 shows the duration line for A


hollow (Ahol6 is drawn in figure 3.13 because it
is referred only to the period before the data
logger broke). CH4 production mainly occurred
at -6 cm, though just for approximately 20 % of
the time. CH4 might have been produced also at
-18 cm and - 30 cm because redox potential was
close to the 200 mV threshold but, for most of
the time, SO4- reduction characterised these
depths. SO4- and Fe(III) reduction equally
characterised the layers at - 42 cm and - 54 cm

Figure 3.13 - Redox potential duration lines at


66 cm at A hollow. The horizontal dotted
lines represent microbial reduction thresholds.
The dominant redox process is written
between the lines.

64

depth. Figure 3.13 shows that redox potential at - 66 cm, was most of the time in Fe(III)
reduction and the rest in SO4- reduction. Methanogenesis was fully favoured at -6 cm and likely
decreased at - 18 cm and - 30 cm. The underlying peat was principally in SO4- reduction. The
dominance of Fe(III) reduction at - 66 cm showed that the vertical redox profile was increasing
with depth.

Figure 3.13 - Redox potential duration lines at A hummock. Colours refer to different depths. The horizontal
dotted lines represent microbial reduction thresholds. The dominant redox process is written between the
lines.

In figure 3.13, the A hummock body was clearly different from the layers below mean hollow
surface, in facts the sensors at 30 cm, 18 cm and 6 cm always laid in highly oxidised conditions.
At 6 cm, redox potential was almost never below 200 mV. Into the underlying layers, at - 18
cm and - 30 cm, SO4- reduction dominated with a marginal time of Fe(III) reduction.

65

Figure 3.14 - Redox potential duration lines at B hollow. Colours refer to different depths. The horizontal
dotted lines represent microbial reduction thresholds. The dominant redox process is written between the
lines.

Figure 3.14 shows that in B hollow CH4 production was favoured at 6 cm, for approximately
20 to 30 % of the time, and at - 18 cm, for almost 50 % of the time. Interestingly, the other
underlying sensors, excluded the one at 42 cm, approached - 200 mV threshold for about 30 %
of the time. The rest of the time, SO4- reduction was dominant at 54 cm and -66 cm, while
Fe(III) reduction was dominant at 150 cm. The sensor at - 42 cm showed an odd behaviour
since it was most of the time in Fe(III) reduction. Methanogenesis was clearly possible at 6 cm
and 18 cm, while it was less favoured in underlying layers. The redox vertical profile was
increasing with depth, shifting from SO4- reduction dominance at 54 cm and -66 cm, to Fe(III)
reduction dominance at 150 cm

Figure 3.15 - Redox potential duration lines at B hummock. Colours refer to different depths. The horizontal
dotted lines represent microbial reduction thresholds. The dominant redox process is written between the
lines.

Figure 3.15 shows that in B hummock at 18 cm, 6 cm and at 6 cm the layers were aerobic.
Below mean hollow surface, CH4 production was possible only at 42 cm. The layer at 18 cm
was in SO4- and Fe(III) reduction and the layer at - 30 was in Fe(III) reduction. Also, in this case
redox potential was increasing with depth.

Discussion
The zone of water table fluctuation showed the highest redox potential variability, with changes
approximately about 900 mV. During the summer, redox potential at 6 cm frequently switched
from oxic to highly reducing condition. The very low redox potential could have been a
66

combination of high microbial activity and low dissolution of O2 due to warm temperature. The
gap in the data corresponded to a period when the sensors at 18 cm might have become
unsaturated. Thus, it was not possible to show the effect of aeration there. In September, air
temperature became lower than water temperature and, in October, air temperature frequently
fell below zero at night (Figure 3.2). At the beginning of fall, redox potential at 6 cm was not
reducing as during the summer, which could have been determined by the lower temperature of
surface layers that hampered microbial activity and increased oxygen dissolution, and in addition
by the oxidising effect of electron acceptor renewal associated to the previous water table
drawdown (Knorr et al., 2009). Considering the sensors at -18 cm, the inversion of temperature
profile caused a relatively higher temperature at that depth than at overlying layers. This might
explain the low redox potential values measured there (< - 200 mV), although it cannot explain
the respectively gentle to fast Eh decrease in that period. It is arguable, albeit there are no
supporting data, that during the temperature inversion, colder and heavier water may have sunk,
replacing underlying warmer water due to buoyancy effect, and carrying labile DOC and TEAs
down to 18 cm, enhancing microbial activity and reducing Eh. Indeed, on 22 September, the
only place where DOC was still increasing corresponded to these two highly reduced locations
(data presented in next section), suggesting that anaerobic respiration and production of DOC
was occurring. It is reported that buoyancy transport can be an important mean of transport of
solutes and can occur in the top 10-15 cm of the acrotelm in Sphagnum bogs, driven by nigh-day
temperature cycles and controlled by hydraulic conductivity (Patberg, 2011; Adema et al., 2005;
Rappoldt et al., 2003). The hummocks above the water table were always in oxic conditions,
reflecting completely different redox conditions from hollows. The redox potential in the
catotelm was quite stable, showing no abrupt variation during the vegetation season and ranging
from moderate reducing to reducing condition. In the zone of water table fluctuation, station A
and B greatly differed. Station A showed a clear influence of water table on redox potential at
both microforms. Redox potential in station B had a somewhat disconnected and more complex
relation to water table dynamic. Beside, B hummock did not show the common broad redox
fluctuation at 6 cm. Looking at figures 3.8-11 it seemed that micro-topography influenced
redox potential at the zone of water table fluctuation in station A. When the sensors at 6 cm are
compared, the hollow responded more sensibly to change in water table level (i.e.
saturated/unsaturated conditions) and the redox potential reached always more reduced values
than hummock. In station B, micro-topography brought a more obscured influence on redox
potential. Figures 3.8-11 showed that the effects of water table on redox potential were evident at
6 cm and from less to no evident at 18 cm. The data showed that water table was the most
67

important driving factor for redox potential only in the zone of water table fluctuation. The
saturated/unsaturated boundary in figures 3.8-3.11 clearly represented a threshold for redox
potential. Notwithstanding that, water table level was not the precise boundary between anoxic
and oxic environment that was expected. Peat properties and capillarity were possibly involved
in complicating that relation, causing a disconnection between Eh and water table. Figures 3.3-7
and figures 3.12-16 showed that the redox potential profile was not decreasing with depth as
expected, but it reached the lowest value within the first 30 cm of peat below mean hollow
surface, more frequently at the shallower sensors. Contrarily to the hypothesis (i), the redox
potential showed a tendency to increase at deeper peat layers. The duration lines showed that in
the acrotelm redox potential differed between hollows and hummocks (at -6, -18 and - 30 cm)
not only regarding the relation between water table dynamic and redox potential, but also in
relation to which process was dominant. In the hollows, redox potentials at 6 cm (A and B), at
-18 cm (B) and to a less extent at 30 cm (A and B) were so low that methane production was
possible. This means that electric donors, local peat and DOC, were relatively easily oxidised by
microorganisms (or the other way around the organic matter was quite decomposable) and that
the TEAs were quickly reduced. In the hummocks, the layers at - 6 cm, - 18 cm and - 30 cm
were in sulphate and ferric reduction and only in B hummock at - 42 cm (no sensor at that depth
in A) redox potential reached 200 mV. In conclusion, below the hummocks, differently than
hollows, the condition for methane production was not favourable. This finding is in accord to
the facts that below hummocks there is a smaller input of fresh organic matter and that it is of
lower quality than hollows, since it stems from vascular plant lignin and Sphagnum fuscum
(Nilsson and Oquist, 2009). It is interesting to add that redox potential in the catotelm was
largely in the reduction zone of humic substances. Finally, two sensors Bhol4 and Bhum3
showed that their measured redox potentials were questionable. In facts, Bhol4 was significantly
higher than nearby sensors and was not changing during the research period. Besides, Bhum3
after an initial drop stood unchanged in oxic condition even when it fell under water level. These
considerations will support their exclusion to further statistical analysis.

3.3 pH, Dissolved Inorganic Carbon and Dissolved Organic Carbon

Results
Figure 3.16 shows pH measurements taken on different time and at increasing depth at both
hollows and hummocks. On 17 June and 7 July, the water table was high and the system
reflected wet condition. Considering the hollows, at that time, pH was around 3.9 and it just
68

slightly increased at the bottom of the profile. The A hollow was slightly more acidic than B
hollow at almost all depths. The top of the hummocks had higher pH than the underlying layers.
The A hummock was slightly more acidic than B hummock in this period. Between 7 July and 4
August, the position of water table dropped to - 20 cm, temperature raised and the system shifted
to dry condition. At both hollows, pH did not change very much compared to previous sampling.
The pH of the oxic zone of the hummocks decreased. From 4 August to 22 September, the bog
got through a partial rewetting and afterward a second drying. The pH measured on 22
September occurred just after an important rain event, which restored the water table to surface.
At that time, pH decreased at both hollows in particular down to - 20 cm deep. Both hummocks
were also more acid than before, although rainwater strongly influenced the top layers where pH
was above 4. The oxic hummocks showed a strong acidification, reaching almost pH 3.4 in A.

Figure 3.16. pH values in relation to peat surface. Different colours are used to identify the date of sampling.
Dotted horizontal lines show the position of the water table at the beginning of the research period.

Figure 3.17 shows that from June to September mean pH calculated over the vertical profile of
each microform, decreased at all stations. pH was significantly higher in B hollow on 7 July
69

(mean = 3.93) than in B hollow on 22 September (mean = 3.83; t = 3.666, p = 0.015, paired ttest). pH was significantly higher in B hummock on 7 July (mean = 4.01) than in B hummock on
22 September (mean = 3.80; t = 3.808, p = 0.013, paired t-test). pH did not differ significantly
between A hummock on 7 July (mean = 3.95) and A hummock on 22 September (mean = 3.80; t
= 1.677, p = 0.154, paired t-test). pH was significantly higher in A hollow on 7 July (mean =
3.90) than in A hollow on 22 September (mean = 3.80; t = 2.796, p = 0.038, paired t-test). The
groundwater collected at - 150 cm was always above pH 4.2. The difference between station A
and B were small, within a decimal of pH, with A hollow being more acid than B hollow.

Figure 3.17 On the left: mean value of pH calculated for each microform. The value for the groundwater
pH (Deep) is in blue. On the right: box plot of pH. Black boxes refer to pH taken on 7 July and red boxes
refer to pH taken on 22 September.

Figure 3.18 shows the vertical profile of DIC at the relative structures in relation to peat
surface. In general, the A pair showed higher concentration than B during all the summer. DIC
was approximately similar at both hollows on 7 July, when the system was still wet. Thereafter,
on 12 August and on 22 September, A hollow exceeded B hollow by 5 mg/l to 15 mg/l.
Considering the hummocks, A had greater DIC than B, but the magnitude of the difference was
less pronounced (max around 5 mg/l). In oxic hummocks, DIC was almost absent.

70

Figure 3.18 - DIC values measured at depth to surface. Each graph represents a microform. The sampling
dates are coded with different colours.

Below the mean hollow surface, when absolute levels are compared, differences between A and
B pairs were greater than difference between microforms. In the zone of water table fluctuation,
where DIC concentration increased with a great leap, hollows showed slightly higher DIC
concentration than hummocks. The vertical profile concentration was somewhat increasing with
depth in all the structures on 7 July. During the summer, the steepness of the vertical profile
accentuated toward higher concentration at bottom, reaching on 22 September a maximum of 25
mg/l at 66 cm . Significant amount of DIC was measured at 150 cm depth that showed an
almost four-fold increase during the summer, passing from 5 mg/l to ca. 21 mg/l. Figure 3.18
showed that, down to - 40 cm, DIC did not increase between 12 August and 22 September except
that in the B hummock. Figure 3.19 shows the relation between redox potential and the rate of
accumulation of DIC (rDIC). To calculate the linear regression Bhol4 was excluded, for it

71

showed unreliable redox potential behaviour. The regression gave a strong relationship between
the two variables (r2 = 0.86; p < 0.001; std. intercept = 1.643; std. slope = 0.015).

Figure 3.19 Rate of accumulation of DIC (rDIC) vs. redox potential. Labels of redox sensors are added near
the points. Black straight line is the regression line.

Figure 3.20 shows the vertical profile of DOC concentration at the relative structures in
relation to peat surface. The temporal and spatial variation of DOC was more complex than that
of DIC. In the zone of water table fluctuation, B hollow showed higher DOC concentration than
A hollow, though below that zone the situation reversed. B hummock showed lower DOC than A
hummock along the entire profile except at 18 cm above mean hollow surface. Regarding the
micro-topography, comparing the zone of water table fluctuation, A hollows had higher DOC
than A hummocks at 6 cm while the situation reversed at - 18 cm except on 22 September. Yet,
comparing the zone of water table fluctuation, B hollows had higher DOC than B hummocks at
6 cm and at - 18 cm while DOC was similar at - 32 cm. Considering the temporal variation, from
7 July to 12 August, the concentration increased along the entire profile, in all the structures and
particularly within the first 20 cm of peat. On 22 September, DOC was decreasing almost
everywhere. In the zone of water table fluctuation, in A and B hollows, DOC decreased at 6
cm, increased at 18 cm and decreased in the lower profile. In both hummocks on 22
September, DOC decreased greatly at the top, whereas it increased down to 6 cm and
decreased below 6 cm. Inside the oxic hummocks, DOC strikingly increased during the
vegetation season (A: 140 mg/l and B: 70 mg/l).

72

Figure 3.20 - DOC values measured at depth from surface. Each graph represents a microform. The
sampling dates are coded with different colours.

Figure 3.21 shows the relation between redox potential and the rate of accumulation of
DOC (rDOC). The negative values of rDOC represented net consumption and the positive values
net production. To calculate the linear regression Bhum3 was excluded, for it showed unreliable
redox potential behaviour. The regression gave a weak relationship between the two variables (r2
= 0.33; p < 0.05; std. intercept = 0.679; std. slope = 0.004).

73

Figure 3.21 Rate of accumulation of DOC (rDOC) vs. redox potential. The horizontal dotted line represents
the net consumption/production boundary. Black straight line is the regression line.

Figure 3.22 shows the relation between pH and DOC with samples taken on 7 July at the
beginning of the research season during the wet period and on 22 September at the end of the dry
season. The correlation test for the two periods showed that on 7 July r = - 0.47 and p < 0.05,
while on 22 September r = - 0.70 and p < 0.001.

Figure 3.22 pH and DOC. The colours are black for 7 July and red for 22 September.

74

Discussion
By figure 3.16, it is clear that in the zone of water table fluctuation, where oxic and
anoxic condition alternated (ca. 20 cm), pH became more acid than in deeper anoxic layers. The
enhanced decomposition of moss and plant remains, stimulated by higher redox potential, drove
the local acidification. Moreover, data showed that during the summer the mean pH decreased.
This trend, consistent with observation from other works, results from the increasing aerobic
microbial activity that tends to acidify the environment (Brouns et al., 2014; Mettrop et al., 2014;
Clark et al., 2009; Proctor, 1994). The acidification process was even greater within the oxic
body of the hummocks, although it was not a consequence of aerobic respiration alone, but of
concentration of dissolved organic carbon (figure 3.22). Precipitation had the effect to diluting
the acidity: oxic hummocks showed a clear influence of rainwater during the rewetting, while the
hollow did not show any effect, unless very slightly on 4 August and just in A. The two stations
did not differ very much, but station A was slightly more acid. The hummocks were more acid
than hollows only above the water table. Dissolved inorganic carbon in peat water relates to
microbial decomposition and it is used as indices of decomposition of organic matter. Given the
low pH (< 4), DIC was essentially present as carbon dioxide. The data showed that DIC
increased during the summer in relation to increasing temperature, accordingly to other
observations found in literature (Shoemaker et al., 2012; Thompson et al., 2009; Waddington and
Roulet, 1997). The vertical profile of DIC is the result of different processes. The greatest
respiration rate and production of carbon dioxide found just below the water table (Shoemaker et
al., 2012) is balanced by the loss of carbon dioxide caused by high porosity, high
saturated/unsaturated hydraulic conductivity,

Sphagnum assimilation and dilution by

precipitations (Waddington and Roulet, 1997). Into deeper anaerobic layers, respiration is much
lower but the dominance of diffusion over advection can bring CO2 to accumulate. The vertical
profile of temperature is also responsible of differential rate of decomposition and CO2
production, in combination with substrate temperature sensibility to decomposition. Generally
during summer, the temperature is higher in surface layers where it follows day-night cycles,
while during winter the profile reverses and deeper layers show higher temperature. In the
present project, temperature was not measured along the peat profiles, though the water table
logger measured air and water temperature at 75 cm depth (figure 3.2), and it revealed the point
when profile inversion occurred. On the other hand, it is possible to argue that at 150 cm depth
the temperature difference between summer and winter was hardly important (Shoemaker et al.,
2012; Clymo and Bryant, 2008; Van der Molen and Wijmstra, 1994). Figure 3.16 suggests that
emissions from peat and assimilation of DIC by Sphagnum were dominant over production near
75

the peat surface. The production gradually became more important than consumption with depth
so that carbon dioxide concentration increased greatly from the mean hollow surface to the
bottom of the water table fluctuation zone. In the catotelm, CO2 increasingly accumulated so that
production exceeded consumption. Yet, figure 3.18 shows the great importance of temperature
on decomposition and on the production of CO2. Indeed, on 22 September, when air temperature
was lower than water temperature the respiration rate in the surface layers slowed down and DIC
decreased, while it kept increasing below -40 cm. CO2 is consumed in anoxic peat via
methanogenesis, the only possible sink in the catotelm. Acetoclastic methanogenesis is important
just below the zone of high respiration rate, where it produces methane and CO2, and where it is
usually found the higher production of methane in bogs (Rydin and Jeglum, 2013; Shoemaker et
al., 2012). In the catotelm, CO2 and H2-dependent (autotrophic) methanogenesis became
negligible, so that CO2 is normally the principal gas dissolved and rate of consumption of carbon
dioxide decreases greatly. Methanogenesis is favoured below 200 mV, so that lower Eh values
correspond to higher methane emissions (Updegraff et al., 1995). Assuming that below the zone
of water table fluctuation degassing did not occur, diffusion dominated and that the only sink for
CO2 was CH4, figure 3.19 illustrates that rDIC (i.e. rate of accumulation of DIC) was positively
related to median Eh (r2= 0.86; p < 0.001; s.d. intercept= 1.643; s.d. slope= 0.015). In other
words, if the DIC production along the profile was constant, the consumption of CO2 by
methanogenesis increased when median Eh approached - 200 mV so that rate of DIC
accumulation decreased. If production of DIC was not constant, but decreased with depth
because of temperature or peat quality changes, the consumption by methanogenesis, indicated
by low median Eh, might have been even greater in shallow layers. In conclusion, figure 3.19
shows that redox potential increased with depth, so that methanogenesis became increasingly
important approaching the surface, in accord to other studies (Sundh et al., 1994). In figure 3.18,
it was shown that B hollow had lower DIC concentration than A hollow. Under the light of the
positive relation illustrated in figure 3.19, that difference might have been a result of greater
consumption by methanogenesis, as indicated by a lower redox potential in B hollow. The
samples taken at 150 cm depth showed that DIC increased from 5 to 21 mg/l. This suggested that
a local production of carbon dioxide by either fermentation or anaerobic respiration prevailed
over the consumption by methanogenesis. Dissolved organic carbon (DOC) is a complex class of
biogenic organic compounds found in water, with complex and often undetermined
characteristics. Within the carbon cycle, DOC can be an intermediate product of organic matter
decomposition and can also be a substrate for fermentation, anaerobic and aerobic respiration.
The principal source of DOC is degradation of OM by exoenzyme and fermentation processes
76

and the sinks are CO2 and CH4 (Moore, 2009). During the summer, from 7 July to 12 August,
DOC increased along the profile so that production exceeded consumption. On 7 July, the profile
of DOC was decreasing with depth, and the trend accentuated on 12 August. Within each vertical
profile, DOC was higher in the zone of water table fluctuation, whereas hollows had higher DOC
than hummocks. It follows that, in that zone, production of DOC has been greater than in
underlying layers in part also because hollows receives greater input of fresh organic matter
(Nilsson and quist, 2009). It is interesting to notice that on 22 September, the trend of dominant
DOC production over consumption shifted, and consumption started to exceed production. The
only layers that from 12 August to 22 September still showed increasing DOC concentration
were found in both hollows at 18 cm, in both hollows at 66 cm and in both hummocks at 6
cm and at 150 cm. Data showed that there was still activity at 150 cm depth. At this depth,
some degradation of organic matter still occurred because DOC increased from 41 mg/l in July
to 44.7 mg/l in September and DIC increased from 5 to 21 mg/l. Since DOC is mostly produced
by anaerobic respiration while it is mostly consumed by aerobic respiration, it was expected that,
along with increasing frequency of oxic condition, consumption would have increasingly
prevailed over production. Figure 3.21 shows that rDOC (i.e. rate of accumulation of DOC) and
redox potential had a negative relation (r2 = 0.33; p < 0.05; std. intercept = 0.679; std. slope =
0.004), though not very strong. For example, when median Eh was near 600 mV, reflecting oxic
conditions, rDOC was below zero so that consumption by aerobic respiration prevailed over
production during the vegetation season. The outlier with high Eh and high rDOC comes from
Bhum3. While it was already showed its Eh odd behaviour, the high rDOC measured in Bhum3
might have been a result of the rain event on 22 September that flushed DOC from the top
hummock and thus percolated to lower layer increasing the local pool with allochthonous
DOC, masking local consumption and resulting in a high rDOC.

The present study showed that both DOC and DIC increased while pH decreased during
the vegetation season, triggered by temperature effects on microbial activity. On 22 September,
the trend was already reversing, particularly in surface layers, again witnessing the importance of
temperature in peat biological processes. According to the measured variables, the zone of water
table fluctuation, extending to approximately 20 cm below mean hollow surface, was identifiable
as separate microenvironment against the catotelm and the aerobic hummock. It had, in absolute
value, the highest amount of DOC (excluding aerobic hummocks that were influenced by
concentration effects) and it was also the zone where DOC was rapidly consumed on 22
September, or in other words, where DOC had the most rapid turn-over. This zone received
77

direct input of fresh organic matter from Sphagnum and vascular plants whereas the
intermittently oxic/anoxic condition provided its quick degradation and the onset of more acidic
pH. Micro-topography created a microenvironment by mean of HWT difference. In fact, the
aerobic hummocks showed the greatest acidification brought by evapotranspiration and
concentration of DOC during the summer drying (it was hardly possible to extract water from the
top hummocks), as showed in figure 3.22, where DOC and pH was negatively correlated. It is
worth noticing the potential negative effects of acidification and moisture deficiency on
decomposition rate (Mkiranta et al., 2009). Regarding the decomposition process, redox
potential was positively correlated to rate of accumulation of DIC, while redox potential was
negative correlated to rate of DOC accumulation. If the relation between Eh and net DIC
mirrored the role of methanogenesis as sink of carbon dioxide, methanogenesis was more
important in surface layers and it decreased with depth. This relation could also evaluate the
spatial-dependent variability of methanogenesis, showing that it was more important in B hollow
than A hollow (i.e. B had lower Eh values and lower net DIC). Finally, it is worth noticing that
decomposition in the catotelm still occurred, though likely at a low rate. In the catotelm, DOC
and DIC accumulated, on one hand because anaerobic respiration and redox potential favoured
production of DOC over consumption and on the other hand because methanogenesis, or
consumption of carbon dioxide, decreased with depth.

3.4 Cellulose decomposition and TBI

Results
Figure 3.23 represents the weight loss of cellulose filters from 29 May to 22 September, which is
117 days. In both hollows, the weight loss was higher at - 6 cm, in the zone of water level
fluctuation. At that depth, the cellulose in A hollow was almost completely consumed, while
cellulose in B lost about 30 % of its initial weight. In general, A hollow had a greater loss of
weight than B hollow. Considering the hummocks, the weight loss was higher at the zone of
water table fluctuation. The A hummock showed higher weight loss than B hummock. Below the
zone of water table fluctuation, weight loss fell between 2 % and 4 % of the initial weight, and
approached 1 % at 150 cm depth.

78

Figure 3.23 Cellulose weight loss in relation to mean hollow surface. The two horizontal dashed blue lines
represent max and min water table level during the research period.

When frequency of oxic condition at 6 cm and weight loss was compared (data not showed),
cellulose in Bhol1 lost 28 % of initial weight and was in aerobic condition 38 % of the time,
while cellulose in Ahol1 lost 87 % of initial weight and was in aerobic condition 56% of the
time. Data in Bhum3 was not compared because redox potential showed unreliable behaviour.
Also, Ahum4 was not included in the analysis because the redox sensor could not be repaired
after the incident.
Regarding the Tea Bag Index (TBI), TBI was not measured in the same structures where
cellulose filters was displaced, but it was measured in nearby micro-topographic structures (the
letter C designed the station nearby A and letter D the one nearby B). Therefore, cellulose and
Tea Bag Index should be compared critically. Unluckily, it was impossible to retrieve all the tea
bags inserted into the peat because the plastic rope that tied some of them broke when they were
pulled out. Table 3 shows the decomposition rate measured from the 7 July to the 22 September,
which is 77 days.
Table 3 Tea bags decomposition rates from the Tea Bag Index.
Site
Chol1
Chol2
Chol3
Chol4
Chol5
Chol6
Chum1
Chum2

Depth
(cm)
-6
-18
-30
-42
-54
-66
30
18

K (d-1)

Site

K (d-1)

-0.008
-0.010
-0.010

0.217
0.330
0.287

-0.009

0.215

No data

No data

No data

No data

No data

No data

No data

No data

No data

No data

No data

No data

No data

No data

No data

No data

-0.009
-0.009

0.077
0.129

Dhol1
Dhol2
Dhol3
Dhol4
Dhol5
Dhol6
Dhum1
Dhum2

-0.009
-0.009

0.150
0.106

79

Chum3
Chum4
Chum5
Chum6

6
-6
-18
-30

-0.008
-0.009
-0.013

0.087
0.083
0.397

No data

No data

Dhum3
Dhum4
Dhum5
Dhum6

-0.008
-0.009
-0.010

0.122
0.247
0.341

No data

No data

Figure 3.24 shows that the TBI of oxic hummocks differed from that of the water fluctuation
zone.

Figure 3.24 Tea Bag Index. The numbers associated to the points in the graphs correspond to height above
mean hollow surface.

Figure 3.25 shows that a drop in decomposition rate occurred above water table.

Figure 3.25 Decomposition rate of tea bags at different depths.

80

Discussion
The decomposition rate depends on the method used to measure it (Berg and McClaugherty,
2014) and the time scale of the research determines which factors can be considered to show
their effects. Substrate quality is defined by C fractions (soluble, cellulose and hemicellulose,
lignin) and nutrients content. Cellulose is made up of only one labile fraction and no nutrients. If
the substrate has no nutrients, the microorganisms may translocate them from the surrounding
(Lindalh et al., 2001 cited in Laiho, 2006). Hence, the decomposition of cellulose filters is
representative for the local environmental condition and local microbial activity. Previous
studies have pointed that the determinants of weight loss in bogs micro topography were
primarily plant litter composition and secondarily microhabitat (i.e. HWT) (Johnson and
Damman, 1991). The use of cellulose stripes excluded the effects of plant litter and reflected
only the effect of environmental condition and microbial activity (i.e. microhabitat). Given that,
the results showed that the zone of water table fluctuation had the most suitable ecological
condition for decomposition. One might expect that, since hummocks were always in aerobic
condition and aerobic respiration is more efficient than anaerobic respiration, the greater weight
loss would have been measured above the water table in the hummocks. However, figure 3.23
shows that this was not the case. Above the water table, other factors seemed involved in
determining the decomposition of cellulose. It is possible that, during the summer, moisture or
pH decreased in the top of the hummocks to the point that they hampered decomposition. Figure
3.23 shows that at the zone of water table fluctuation, cellulose was more degraded in the
hollows than hummocks. These results suggested a differentiation of the oxic/anoxic zone in two
further microhabitats, influenced by micro-topographic structures. On the contrary, other studies
have found that decomposition of Sphagnum litter in the zone of water table fluctuation did not
differ between hummocks and hollows (Johnson and Damman, 1991). Two conclusions can be
drawn from this data. First, since the experimental design did not use replicates, it may have not
account for environmental variability so that hummocks and hollows actually did not differed at
the water fluctuation zone. Second, the microforms truly split the water fluctuation zone in two
sub-microhabitats and the difference in decomposition between hummocks and hollows resulted
from stronger microbial inhibition exerted by Sphagnum species growing on hummocks or lower
nutrient content of hummock peat (Wieder and Vitt, 2006; Johnson and Damman, 1991), which
was detected by cellulose degradation. The feature of the sub-microhabitat overwhelmed the role
of carbon quality alone (i.e. pure cellulose), so that inhibiting mechanisms and low amount of
nutrients released by slow decomposition of hummocks Sphagnum determined the
decomposition rate of cellulose. In conclusion, above the water table in the hummocks, a
81

combination of low moisture, increasing acidity (circumscribed to summer) and Sphagnum


specie-specific inhibition of microbial activity possibly constrained decomposition and created a
definite microhabitat. Provided water was not limiting, inhibition and nutrient deficiency below
hummocks might have influenced the decomposition rate in the water table fluctuation zone,
creating a different sub-microhabitat. The relation between cellulose weight loss and redox
potential measured at 6 cm below mean hollow surface showed that in hollows weight loss
increased along with frequency of oxic conditions, although the amount of data cannot be
adequate to test definitively that correlation.
The tea bags are composed of plant material so, according to Johnson and Damman
(1991), litter composition should have overwhelmed the microhabitat control on decomposition
rate. The Tea Bag Index is a simplified methodology for measuring decomposition rate (k) and
for giving information on environmental constraints (S). The method was tested and showed
sensibility and ability to discriminate among different ecosystems (Keuskamp et al., 2013).
Keuskamp et al. (2013) studied the correlation between the TBI parameters and data on mean
annual temperature, precipitation and soil carbon sequestration potential (function of
waterlogged conditions in bogs). They found that k increased and S decreased with mean annual
temperature and precipitation, while S increased with soil carbon sequestration potential. In this
project, the Tea Bag Index was used to evaluate k and S in different microhabitats. TBI should be
influenced by temperature, moisture and waterlogged condition. Figure 3.24 shows that the TBI
formed two distinct groups. The one with lower k and S corresponded to the hummocks
microhabitat above the water table. The other with higher k and S corresponded to the zone of
water table fluctuation of both hummocks and hollows. The results suggested that there were two
different microhabitats: the water table fluctuation zone and the oxic zone of hummocks. The
TBI could not rule out any difference between hummocks and hollows in the intermittent
oxic/anoxic zone, probably because of the small amount of data, though this is in accordance to
other studies (Johnson and Damman, 1991). Keuskamp et al. (2013) measured and interpreted S
and k in different ecosystems but not within the same ecosystem, indeed the differences between
S and k found in the present work were not as great as the ones reported in Keuskamp et al.
(2013). Looking at the set of data seemed that k increased as S increased, which is
counterintuitive. However, looking within a single microhabitats, k and S showed a negative
relation. In figure 3.25, the TBI showed that k increased in the zone of water fluctuation zone
and identified clearly the zone of water fluctuation from the oxic hummocks but it did not sort
out any difference between the two structures. TBI is calculated from the labile C fraction of tea
(soluble and cellulose) and, differently from pure cellulose, it contains nutrients. It is possible
82

that, in this case, C quality together with nutrient supplied by tealeaves shaded the constraint
exerted by inhibiting mechanisms (i.e. plant composition was more important than microhabitat).

Cellulose weight loss showed that the intermittently oxic/anoxic zone had the greatest
decomposition. At this location, decomposition between hummocks and hollows differed greatly,
with hollows having a greater decay. This suggested that micro-topography was influencing the
environment. Secondly, in the zone of water table fluctuation, also spatial variability affected
decomposition, with A having higher decay than B, whereas it was partially explained by
different frequency of oxic/anoxic conditions. Above the water table, in the aerobic hummocks,
decomposition decreased. Interestingly that was the zone where during the summer acidity and
DOC greatly increased while moisture decreased. Below the zone of water table fluctuation, in
the catotelm, degradation abruptly decreased. In the catotelm, the lack of oxygen forces
microorganisms to less efficient anaerobic respiration and fermentation and inhibits the activity
of phenol oxidase (Freeman et al., 2001), causing accumulation of phenolic compounds that
inhibit the activity of hydrolase enzymes responsible for decomposition (Davidson and Janssens,
2006). The TBI, similarly to cellulose, identified the oxic/anoxic zone as a hotspot for
decomposition and it measured a lower decomposition rate in oxic hummocks. Interestingly it
was able to distinguish, in terms of k and S, between these two microhabitats. Contrarily to
cellulose, in the zone of water table fluctuation the TBI was not able to distinguish between
hummocks and hollows. It can be that the dataset was too small to discern differences or that
plant litter was more important than microhabitat in affecting decomposition rate (Davidson and
Janssens, 2006). Unluckily, the TBI was not calculated in the catotelm. However, the still high
decomposition rate at 30 cm could have been due to the addition of nutrients plant litter in a
depleted environment.

3.5 Peat quality and degradability

Results
Figure 3.26 both hollow profiles showed a less decomposed superficial layer overlying a more
decomposed strata, below which the decomposition degree decreased although not as much as in
the shallow strata. The hollow B seemed more decomposed than hollow A. The hummocks
superficial layers were more decomposed than hollows. The hummock A had very thin well
preserved superficial peat, which became more decomposed in the middle strata and then the
83

humification degree decreased again. The hummock B had a superficial layer slightly more
decomposed than the hollows, below which the humification increased. The Deep samples
showed that below 140 cm depth the von Post increased to H5, reaching H7 at 200 cm. The
common characteristic among the peat cores was that humification reached a maximum at some
depth, between 12 and 40 cm below the surface, and decreased again at the bottom. Yet, the peat
was more decomposed at the surface strata than at the bottom. There was no evident accordance
between microforms. Station B seemed somewhat more decomposed than A. Bulk density was
very low, < 0.1 g.cm-3, probably also caused by underestimation by the method, and it did not
show any vertical trend. The only high values (> 0.1 g.cm-3) were in A hollow at 24-36 cm and in
Deep at 190-200 cm. The fibres content was very high, and it increased slightly with depth.
However, the variability was not so great since the lowest fibres content value was 80 % and the
highest was 100 %. The A hollow had a minimum at 24-36 cm depth while A hummock had a
maximum at 36-48 cm depth. The Deep samples showed that the fibres content decreased again
at 145-155 cm reaching a minimum at 190-200 cm.

84

Figure 3.26 von Post degree of humification (H1-H10) of 100 cm long peat cores and the three subsamples
took at 100-110 cm, 145-155 cm and 190-200. Solid lines identifying the subsamples of 12 cm length divide
each core horizontally. By each subsamples, the relative bulk density (BD, g.cm-3) and fibres content (F, %)
are written beside the relative 12 cm length subsample.

85

Figure 3.27- peat cores. From top left: A hummock; A hollow; B hummock; B hollow; Deep; Deep zoom at 150 cm.

86

The great part of phosphorus was present in organic form (appendix table). In general, total P
was decreasing with depth down to -100 cm (figure 3.28). In station A, the Sphagnum and
proximate layer of peat had low P, which then increased producing a peak near the surface in A
hummock, while A hollow showed a peak about 20 cm deeper. In station B, phosphorous
decreased from the top layers without showing any peak. Below -50 cm, the concentration of P
was changing frequently. In the zone of water table fluctuation, total P was lower below
hummocks than hollows.

Figure 3.28 the total P profile in relation to absolute position of microforms. Blue horizontal dotted lines
represent max and min water table during the research period.

Figure 3.29 illustrates the relation between total P and Eh. The linear regression (r2 < 0.1; p >
0.5) showed that there was no relation between the two variables.

Figure 3.29 total P and median redox potential. Numbers are depth below mean hollow surface.

All over the sampling points the N:P ratio was between 10 to 60 (figure 3.30). Higher N:P ratio
was found in hummocks, in particular in the water fluctuation zone or nearby.

Figure 3.30 N:P ratio profile measured in each microform. Blue horizontal dotted lines represent max and
min water table during the research period.

Figure 3.31 shows the C:N profile in the four micro-topographies. The ratio decreased between
the living Sphagnum layer and the peat about 20 cm below the surface. In the underlying layers,
the ratio increased and approached the value of almost 150 at -100 cm. In B hummock, the ratio

88

was changing back and forth. From -100 to -200 cm the C:N ratio was again decreasing. In the
zone of water table fluctuation, the C:N ratio was higher below hummocks than hollows.

Figure 3.31 - The C:N profile measured in each microform. Blue horizontal dotted lines represent max and
min water table during the research period.

Figure 3.32 illustrates the relation between redox potential and the C:N ratio. Though the figure
seems in some way suggesting a positive relation, the linear regression showed that no positive
correlation between Eh and C:N was found (r2 < 0.1; p > 0.5).

Figure 3.32 C:N and median redox potential. Numbers are depth below mean hollow surface.

89

Figure 3.33 shows the profile of the acid-detergent lignin (ADL) in the four micro-topographies.
If B hummock is excluded, ADL was decreasing with depth. In station B the lignin content was
lower than A station. Regarding the entire profile, the hummocks did not have explicit higher
lignin than hollows. Contrarily, in the zone of water table fluctuation lignin was lower in the
hummocks.

Figure 3.33 - ADL (%) profile in relation to absolute position of microforms. Blue horizontal dotted lines
represent max and min water table during the research period.

Figure 3.33 shows the relation between median redox potential and ADL. The linear regression
determined no relation between the two variables. (r2 < 0.1; p > 0.5).

Figura 3.34 - Eh vs. ADL.

90

Figure 3.35 shows the profile of the acid-detergent cellulose (ADC) in the four microtopographies. The cellulose did not show a clear trend with depth. A hollow had lower level of
ADC than B hollow. A hummock had higher level of ADC than B hummock.

Figure 3.35 - ADC (%) profile in relation to absolute position of microforms. Blue horizontal dotted lines
represent max and min water table during the research period.

Figure 3.36 shows the vertical profile of ADL:N ratio in the four micro-topographies. The ratio
did not show particular trend with depth. Except from A hummock there were not great
difference between the sites. In the zone of water table fluctuation ADL:N was higher in A
hollow than B hollow and A hummock had the greatest value.

Figure 3.36 ADL:N profile in relation to absolute position of microforms. Blue horizontal dotted lines
represent max and min water table during the research period.

91

Figure 3.37 shows the relation between median redox potential and ADL:N ratio. There was no
relation between the two variables (r2 < 0.1; p > 0.5).

Figure 3.37 Eh vs. ADL:N.

Discussion
The peat soil belonged to the taxonomic class of Histosol (AFES, 2008). Content of fibres, von
Post scale and the bulk density showed that all the four peat cores represented fibric horizons
(figure 3.31). Instead, the deepest Deep sample, from -190 to -200 cm represented mesic horizon.
Regarding the bulk density, it has to be remind that the Russian peat core is not an ideal tools for
its determination, because it tends to cut roughly the surface peat and to underestimate the bulk
density. However, the bulk density was very low also at depth of 70-100 cm were the cut was
more precise. The content of fibres increased down to -100 cm, which, in accord to von Post,
means that surface peat was more decomposed than bottom peat. It has been suggested that the
N:P ratio can be used to evaluate if plant growth (Wang et al., 2014) and microbial community
(Gusewell and Gessner, 2009) are N or P limited. Figure 3.30 shows that in the zone of water
table fluctuation the hollows were N-limited as a probable consequence of the nutrient recycling
strategy of Sphagnum and Eriophorum vaginatum, while the hummocks were P-limited.
Regarding the difference between microforms, nutrients content in the zone of water table
fluctuation of hummocks showed that total P was lower, total N did not differ significantly
(appendix) and mineral content was three times lower (appendix). Consequently, it is possible to
affirm that in general hummocks were nutrient depleted in respect to hollows. The analysis of the
carbon fraction did not reveal the expected higher lignin content in hummocks supposedly
92

caused by the inhabiting lignified vascular plants. Indeed, ADL was even lower in A hummock.
Among the suggested indices of recalcitrance that were investigated, C:N ratio showed much
higher values below hummocks, while ADL:N was significantly higher only below A hummock.
These results, also in the light of the lower cellulose decomposition found in hummocks,
might highlight a possible nutrient deficiency in the water table fluctuation zone of hummocks.
Yet, the fact that in that zone the TBI index failed to discriminate between micro-topographies,
can lead to think that the tea leaves nutrient supply might have masked the lower peat nutrient
content of hummocks. It is interesting to notice that in that zone median redox potential,
measured in the 20 cm of shallow peat, was always higher below hummocks. Ahol: 344 mV (-6
cm), -168 mV (-18 cm); Ahum: 597 mV (-6 cm), -129 mV (-18 cm); Bhol: 318 mV (-6 cm), 185 mV (-18 cm); Bhum: 685 mV (-6 cm), -78 mV (-18 cm). Moreover, the redox potential
showed less strong response to water table dynamic. Despite the possible relation between redox
potential and factors of peat quality in the water table fluctuation zone, when this relation was
explicitly investigated along the anoxic vertical profile, it did not show any significant
relationships. In a previous study was found that P-rich fen had lower redox potential than
moderately rich fens, though the authors did not found any correlation between redox potential
and depth gradient of phosphorous (Thomas et al., 2009). In the present work, the relation
between total P and redox potential along the anoxic profile was not significant. In addition nor
organic or inorganic phosphorous did correlate (data not showed).

4 CONCLUSIONS
The present work showed that water table was driving redox potential in the acrotelm.
The zone above the water table fluctuation (above water fluctuation zone, AWFZ hereafter) was
always in aerobic condition. In the zone of water table fluctuation (water fluctuation zone, WFZ
hereafter), redox potential was greatly affected by water table dynamic, and followed shortly its
movement shifting about 900 mV in few hours. However, other factors played a role in the
relationship. Peat properties and capillarity probably confounded the relation, while microforms,
influencing peat quality, obfuscated the strength of the relation. Indeed, water table was not the
good boundary for oxic/anoxic condition that was expected given the supposed low capillarity of
fibrous Sphagnum peat. Quite often, when water table fluctuated, reducing conditions persisted
above the water table and oxidised condition persisted below it. In conclusion, the present work
showed that water table was a strong driving factor of redox potential, though its poor resolution

93

suggested that it should be coupled to redox potential measurement for a better characterisation
of oxic/anoxic boundary in soils.
The redox potential profile below the mean hollow surface was not decreasing with depth
as supposed in the hypothesis, indeed highly reducing conditions were measured in the first 30
cm of peat, while less reducing conditions were measured in the underlying peat and at 150 cm
depth. The inverted redox profile was consistent with the higher degree of humification found in
superficial peat, which reflected an active process of degradation of organic matter. The negative
relation between redox potential and rate of accumulation of DOC explained that increasing
reducing conditions favoured the relative accumulation of DOC, which combined with higher
concentration of DOC in the WFZ witnessed the greater activity of that layer.
The redox time series of the vertical profile showed that the suitable conditions for CH4
production occurred just within the first 30 cm of surface peat and only in hollows (figure 4.2).
Below the WFZ (below water fluctuation zone, BWFZ hereafter), in the catotelm,
methanogenesis was not thermodynamically favoured, and redox potential increased
progressively with depth. There, the significant positive relationship between redox potential and
rate of accumulation of DIC reflected a gradual decreasing importance of consumption by
methanogenesis with depth.
Contrarily to what stated in the hypothesis, the analysis on peat did not reveal a
difference in carbon quality between hummocks and hollows. The difference between the two
microforms was found only in the WFZ on nutrients (total N and P) and mineral content, and on
indices of degradability such as C:N and ADL:N. Since an important control of methanogenesis
is quality of organic matter, the lower P and N and higher C:N and ADL:N ratio of WFZ of
hummocks might have hindered methanogenesis and caused redox potential to be less reducing.
The cellulose weight loss showed that in the WFZ, oxic condition boosted the decomposition
rate of organic matter, though the extent of the process seemed also related to difference in
nutrient content of peat, or litter if the TBI index is considered. If the occurrence of low peat
quality factors with high redox potential in the WFZ may tempt to bring to a generalisation,
when the relation was investigated including the data of BWFZ, it did not revealed any
significant pattern. This might mean that other factors were more important in controlling redox
potential in the catotelm.

94

Figure 4.1 Vertical profile of the favourable conditions of methane production and of decomposition rate.
The broken horizontal lines represent max and min level of water table. The three microhabitats are named
AWFZ (above water fluctuation zone), WFZ (water fluctuation zone) and BWFZ (below water fluctuation
zone). The dominant specie on the hummock was Sphagnum fuscum (thick orange line), where also grew the
bush Calluna vulgaris. The dominant species on the hollows were S. balticum, S.cuspidatum and S.tenellum
(thick dark green line) where was also frequent the herbaceous Eriophorum vaginatum.

The summer season represents a pulse in the activity of the bog ecosystem brought by
energetic input from the sun, which increases temperature, decomposition rate and plant
production. The biogeochemistry of the bog results accelerated. The picture that came out from
the present work was composed of redox potential profile, decomposition rate and products of
degradation, which were representative measurements of short-term processes occurring during a
relatively not-dry summer. The measured weight loss and decomposition rate, pH, redox
potential and DOC identified at least three zones along the vertical profile of the bog, with
peculiar process rate and microenvironment, controlled by HWT and water table fluctuation
(figure 4.2).

95

Figure 4.1 Three main microhabitats are identified in the vertical profile with roman numbers. A further
distinction (II1 and II2) might follow the effect of micro-topography in the WFZ. Horizontal dotted lines
represent max and min water table position. (AWFZ, above water fluctuation zone; WFZ, water fluctuation
zone; BWFZ, below water fluctuation zone).

The AWFZ was characterised by a peculiar dynamic of decomposition rate, redox potential,

pH and DOC. It had low decomposition rate, measured with both cellulose and tea bags
methods, intermediate between values of WFZ and catotelm. Redox potential was always stable
and oxic. Interestingly, during the summer drying, when evapotranspiration and HWT increased,
DOC and acidity increased greatly due to concentration effect. This suggested a possible
feedback mechanism that counteracted the effect of increasing temperature on decomposition
rate.

The WFZ had the highest decomposition rate of the profile, for both cellulose and tea bags

methods. The redox potential was most variable than in any other places shifting from -300 mV
to 600 mV. There, Eh reached the lowest values of the entire profile. DOC was more abundant
and had higher turnover than underlying peat. Intermittent aerobic respiration caused
acidification largely than in lower profile. The influence of micro-topography on biogeochemical
processes seemed extending into the WFZ. The hypothesis of lower peat quality below
hummocks has been partially verified showing lower P, higher C:N and ADL:N than below
hollows, but no difference was found in lignin content. The lower peat quality caused higher Eh,
less sensible Eh-water table relation and unfavourable conditions for methanogenesis below
hummocks. Moreover, confronting hollow and hummock WFZ, the decomposition of cellulose
was lower below hummocks. Therefore it is argued that the water fluctuation zone could be
subdivided in two further zone influenced by micro-topography (figure 4.1) and with important
consequences for their persistence.

96

In the BWFZ, or catotelm, the decomposition rate fell abruptly to negligible values. Redox

potential, also considering the sensor at -150 cm, was not decreasing with depth but increasing.
Yet, the characteristic redox zone of the catotelm was between iron and sulphate reduction,
which is also the zone of anaerobic reduction of organic electron acceptors. DOC was slightly
less abundant than in WFZ, and in comparison, it had lower production and consumption (i.e.
low turnover). DIC and DOC were produced and were accumulated even at -150 cm depth,
whereas cellulose lost 1 % of its initial weight. The only factor that at that depth was arguably
responsible for enhanced decomposition was temperature. In the catotelm methanogenesis
decreased, removing a sink for CO2 that accumulated. The causes of this pattern are different and
still in need of clarification: lack of nutrients, thermodynamic constraints, accumulation of antiseptic phenols, alternative organic electron acceptors and recalcitrance of peat are among the
possible causes.

97

5 ACKNOWLEDGMENTS

98

6 APPENDICES

Figure 6.1 - pH values in relation to water table level. The two sub-graphs represent vertical profile of pH
measured at both hollows and both hummocks. Different colours are used to identify the date of sampling.
The position of the water table at the moment of sampling is showed with reversed coloured triangles located
just above the water table curve.

Figure 6.2- DIC values at different micro-topographies. Different colours refer to DIC concentration.

99

Figure 6.3 - DOC values at different micro-topographies. Different colours refer to DOC concentration.
Hummock and hollow differ in range of DOC concentration.

Table 6 Overview of peat chemistry

Samples

Ahol0
Ahol1
Ahol2
Ahol3
Ahol4
Ahol5
Ahol6
Ahol7
Ahol8
Ahum0
Ahum1
Ahum2
Ahum3
Ahum4
Ahum5

0-2
2-12
12-24
24-36
36-48
48-60
60-72
72-84
84-96
0-2
2-12
12-24
24-36
36-48

Min
eral
Cont
ent
(%)
1.48
0.60
1.54
0.87
0.78
0.60
0.59
0.36
0.49
1.34
1.37
0.75
0.37
0.55

48-60

0.50

Ahum6

60-72
72-84
84-96
0-2

0.56
0.50
0.58
1.21

2-12

0.99

12-24
24-36
36-48
48-60
60-72
72-84
84-96
0-2
2-12
12-24
24-36
36-48
48-60

0.90
0.50
0.60
0.42
0.58
0.48
0.50
0.99
1.39
1.19
0.69
0.50
0.59

Ahum7
Ahum8
Bhol0
Bhol1
Bhol2
Bhol3
Bhol4
Bhol5
Bhol6
Bhol7
Bhol8
Bhum0
Bhum1
Bhum2
Bhum3
Bhum4
Bhum5

Depth
(cm)

Carbo
n
(g/kg)

Nitroge
n (g/kg)

C:N

Total P
(mg/kg)

293.8
390.2
519.8
653.0
320.8
200.5
263.0
243.9
116.6
352.9
583.7
354.4
301.4
322.5

Inorganic
P(mg/kg)

Organi
cP
(mg/kg)

8.3
7.6
2.8
3.1
4.5
1.0
6.8
11.4

381.9
512.3
650.1
317.7
196.1
262.0
237.1
105.2

5.9
5.2
2.6
2.6

577.8
349.2
298.8
319.9

19.4
16.1
16.9
14.7
18.1
21.9
14.1
14.4
34.3
13.3
16.3
16.6
15.6
13

455.5
472.5
469.8
500.6
456.7
453.9
457
461
472.4
451
489.7
473.9
466.5
456.5

5.7
6.3
8.8
9.6
5.8
4.4
3.7
3.5
4
4.7
9.5
5.9
4.7
4.2

80.03
74.76
53.66
52.01
79.08
102.97
122.05
132.52
117.72
96.18
51.40
80.89
99.39
107.44

464.6

4.4

104.92

94.8

12.9

81.9

464.5
457.4
453
446.7

4.4
4.1
3.4
5.6

106.72
111.02
134.20
79.53

282.1
141.0
195.9
376.2

1.9
10.9
1.2

280.2
130.0
194.7

458.4
472.4
462.8
472.6
465.5
466.5
461.8
464.2
447.5
456.3
476.5
453.8

6.4
7.1
4.5
5.1
4.5
4.6
4.4
3.7
5.3
6.3
7.7
4.4

71.72

246.3

12.9

233.4
245.6
150.2
157.7
146.2
602.4
229.1
135.4

3.8
8.4

253.7
154.4
163.1
149.7
606.0
232.9
145.9
352.9
212.8
133.6
121.4
92.4
283.8

8.0
4.2
5.4
3.5
3.5
3.8
10.4

456.2
578.6

66.38
101.74
91.90
102.82
100.45
104.21
125.44
83.89
72.76
61.53
102.31
120.28
68.57

10.0
14.1
13.9
11.7
8.0

202.9
119.5
107.6
80.7
275.7

N:P

Lignin
(%)

Ligni
n:N

25.5
23.5
27.6
25.3
27.8
24.3

27.1
21.6
31.2
15.1
9.4
12.5

43.1
24.6
32.5
26.1
21.3
33.9

34.9
34.1
36.9
35

34.2
48.3
47
42.6

30.9
17.6
16.1
22.4

32.6
29.8
34.3
53.4

46.4

39.2

39.3

21.5

48.8

15.6
29.1
17.4
14.9

55.9

34.7

9.4

21.3

49.8

31.3

18.9

29.5

47.6
60.1
45.6
48.6
48.3

34.4
27.7
37.9
38
37.8

18
12.2
16.5
13.4
13.9

25.4
27.1
32.3
29.8
30.3

52.3
47.7
63.7
55.6
52.7

32.8
30.7
25
29.7
22.1

14.9
21.6
11.3
14.7
25.2

23.7
28.1
24.5
30.3
30

26
28
29.1
31.3
30.1
7.6
18.9
25.4
15
29.6
57.6
36.2
41.1
29.6

Extractives
(%)

Cellulose
(%)

47.3
54.9
41.2
59.6
62.8
63.2

100

Bhum6
Bhum7

60-72

0.48

389.8

7.1

54.78

150.8

13.6

137.2

47.1

72-84

0.49

453.6

3.2

143.06

215.3

9.7

205.6

14.9

Bhum8

84-96

0.59

460.3

3.6

126.89

172.8

2.1

170.7

20.8

Deep1

100-110

0.40

4.2

109.39

245.5

4.9

240.6

17.1

Deep2

145-155

0.70

518.5

10.6

49.02

173.0

2.4

170.6

61.3

Deep3

190-200

1.19

465.4

6.5

72.11

237.9

8.8

229.2

27.3

456.6

47.1

34.9

18

25.3

50.5

29.1

20.3

19.2

101

7 References
Abbott G.D., Swain E.Y., Aminu, B.M., Allton, K., Belyea, L.R., Laing, C.G., Cowie, G.L.
(2013). Effect of water-table uctuations on the degradation of Sphagnum phenols in surcial
peats. Geochimica et Cosmochimica Acta 106: 177-191
Adema, E.B., Baaijen, G.J., Belle J., Rappoldt, C., Grootjans, A.P., Smolders, A.J.P. (2005).
Field evidence for buoyancy-driven water ow in a Sphagnum dominated peat bog. Journal
of Hydrology 327: 226 234
Andersen, R., Chapman, S.J., Artz, R.R.E. (2013). Microbial communities in natural and
disturbed peatlands: A review. Soil Biology & Biochemistry 57: 979-994
Arduino, E., Barberis, E. (2000). Determinazione del fosforo organico. In Metodi di analisi
chimica del suolo. Coord. Pietro Violante, anno 2000.
Armstrong, W. and Boatman, D.J. (1967). Some field observations relating the growth of bog
plants to conditions of soil aeration. Journal of Ecology 55: 101-110
Artz, R.R.E. (2009). Microbial community structure and carbon substrate use in northern
peatlands. Carbon Cycling in Northern Peatlands in Geophysical Monograph Series 184
Baize, D. and Girard, M-C. (2008). Association franaise pour ltude du sol (Afes). ditions
Qu 405 pp
Barber, K.R., Leeds-Harrison, P.B., Lawson, C.S., Gowing, D.J.G. (2004). Soil aeration status in
a lowland wet grassland. Hydrological Processes 18: 329341
Beer, J. and Blodau, C. (2007). Transport and thermodynamics constrain belowground carbon
turnover in a northern peatland. Geochimica et Cosmochimica Acta 71: 29893002
Beer, J., Lee, K., Whiticar, M., Blodau, C. (2008). Geochemical controls on anaerobic organic
matter decomposition in a northern peatland. Limnology Oceanography 53: 13931407
Benstead, J. and Lloyd, D. (1996). Spatial and temporal variations of dissolved gases (CH4, CO2,
and O2) in peat cores. Microbial Ecology 31: 57-66

102

Berg, B. and McClaugherty, C. (2014). Plant litter: decomposition, humus formation, carbon
sequestration (third edition). Springer
Berg, B. and Meentemeyer, V. (2002). Litter quality in a north European transect versus carbon
storage potential. Plant and Soil 242: 83 92
Biester, H., Hermanns, Y-M., Cortizas, A.M. (2012). The inuence of organic matter decay on
the distribution of major and trace elements in ombrotrophic mires a case study from the
Harz Mountains. Geochimica et Cosmochimica Acta 84: 126136
Biester, H., Knorr, K-H., Schellekens, J., Basler, A., Hermanns, Y-M. (2014). Comparison of
different methods to determine the degree of peat decomposition in peat bogs. Biogeosciences
11: 26912707
Blodau, C., Bauer, M., Regenspurg, S., Macalady D. (2009). Electron accepting capacity of
dissolved organic matter as determined by reaction with metallic zinc. Chemical Geology
260: 186195
Bragazza, L., Alber, R., Gerdol, R. (1998). Seasonal chemistry of pore water in hummocks
and hollows in a poor mire in the southern alps (Italy). Wetlands 18: 320-328
Bragazza, L., Buttler, A., Habermacher,, J., Brancaleoni, L., Gerdol, R, Fritze, H., Hanajk, P.,
Laiho, R., Johnson, D. (2012). High nitrogen deposition alters the decomposition of bog plant
litter and reduces carbon accumulation. Global Change Biology 18: 11631172
Bragazza, L., Buttler, A., Siegenthaler, A., Mitchell, E.A.D. (2013). Plant litter decomposition
and nutrient release in peatlands. Carbon Cycling in Northern Peatlands Geophysical
Monograph Series 184: 99-110
Bragazza, L., Gerdol, R., Rydin, H. (2003). Effects of mineral and nutrient input on mire biogeochemistry in two geographical regions. Journal of Ecology 91: 417-426
Bragazza, L., Rydin, H., Gerdol, R. (2005). Multiple gradients in mire vegetation: a comparison
of a Swedish and an Italian bog. Plant Ecology 177: 223-236
Branham, J.,E. and Strack, M. (2014). Saturated hydraulic conductivity in Sphagnum-dominated
peatlands: do microforms matter? Hydrological Processes 28: 4352-4362

103

Bridgham, S., Cadillo-Quiroz, H., Keller, J., Zhuang, Q. (2013). Methane emissions from
wetlands: biogeochemical, microbial, and modeling perspectives from local to global scales.
Global Change Biology doi: 10.1111/gcb.12131
Brouns, K.,Verhoeven, J.T.A., Hefting, M.M. (2014). Short period of oxygenation releases latch
on peat decomposition. Science of the Total Environment 481: 6168
Bubier, J., Costello, J., Moore, T.R., Roulet, N.T., Savage, K. (1993). Microtopography and
methane flux in boreal peatlands, northern Ontario, Canada. Canadian Journal Botany 71:
1056-1063
Callebaut, F., Gabriels, D., Minjauw, W., de Boodt, M. (1982). Redox potential, oxygen
diffusion rate, and soil gas composition in relation to water table levels in two soils. Soil
Science 132: 149-156
Catallo, W.J. (1999). Hourly and daily variation of sediment redox potential in tidal wetland
sediments. U.S. Geological Survey, Biological Resources Division Biological Science Report
USGS/BRD/BSR-1999-0001, 10 pp.
Cervantes, F.J., de Bok F.A.M., Duong-Dac, T., Stams, A.J.M., Lettinga, G., Field, J.A. (2002).
Reduction of humic substances by halorespiring, sulphate-reducing and methanogenic
microorganisms. Environmental Microbiology 4: 5157
Chadwick, A.O. and Chorover, J. (2001). The chemistry of pedogenic thresholds. Geoderma
100: 321353
Chason, D.B. and Siegel, D.L. (1986). Hydraulic conductivity and related physical properties of
peat, Lost River Peatlands, Northern Minnesota. Soil Science 142: 91-99
Clark, J.M., Ashley, D., Wagner, M., Chapman, P.J., Lanes, N., Evans, C.D., Heatwait, A.L.
(2009). Increased temperature sensitivity of net DOC production from ombrotrophic peat due
to water table draw-down. Global Change Biology 15: 794807.
Clymo, R.S., Bryant, C.L. (2008). Diusion and mass ow of dissolved carbon dioxide,
methane, and dissolved organic carbon in a 7-m deep raised peat bog. Geochimica et
Cosmochimica Acta 72: 20482066

104

Courtwright, J. and Findlay, S.E.G. (2011). Effects of microtopography on hydrology,


physicochemistry, and vegetation in a tidal swamp of the Hudson river. Wetlands 31: 239249
DAngelo, E.M. and Reddy, K.R. (2003). Effect of aerobic and anaerobic conditions on
chlorophenol sorption in wetland soils. Soil Science Society America 67: 787794
Davidson, E.A. and Janssens, I.A. (2006). Temperature sensitivity of soil carbon decomposition
and feedbacks to climate change. Nature 440: 165-173
de Mars, H. and Wassen, M.J. (1999). Redox potentials in relation to water levels in different
mire types in the Netherlands and Poland. Plant Ecology 140: 4151
Deppe, M., McKnight, D.M., Blodau, C. (2009). Effects of Short-term drying and irrigation on
electron flow in mesocosms of a Northern bog and an Alpine fen. Environmental Science
Technology 44: 8086
Devito, K.J., Waddington, J.M., Branfireun, B.A. (1997). Flow reversal in peatlands influenced
by local groundwater systems. Hydrological Processes 11: 103-110
Dobrovolskaya, T.G., Golovchenko, A.V., and Zvyagintsev, D.G. (2014). Analysis of
ecological factors limiting the destruction of High-Moor Peat. Eurasian Soil Science 47: 182
193
Enwall, K., Nyberg, K., Bertilsson, S., Cederlund, H., Stenstrm, J., Hallin, S. (2007). Long-term
impact of fertilization on activity and composition of bacterial communities and metabolic
guilds in agricultural soil. Soil Biology Biochemistry 39: 106-115.
Estop-Aragons, C. and Blodau, C. (2012). Effects of experimental drying intensity and duration
on respiration and methane production recovery in fen peat incubations. Soil Biology &
Biochemistry 47: 1-9
Estop-Aragons, C., Knorr, K-H., Blodau, C. (2012). Controls on in situ oxygen and dissolved
inorganic carbon dynamics in peats of a temperate fen. Journal of Geophysical Research 117:
1-14
Fenner, N. and Freeman, C. (2011). Drought-induced carbon loss in peatlands. Nature
Geoscience 4: 895-900

105

Fiedler, S. (1999). In situ long-term-measurement of redox potential in redoximorphic soils.


Chapter 7 in Schring, J., Schulz, H.D, Fischer, W.R., Bttcher, J., Duijnisveld, W.H.M.
(Eds.). Redox: fundamentals, processes and applications. Springer-Verlag New York
Fiedler, S. and Sommer, M. (2004). Water and redox conditions in wetland soils-their influence
on pedogenic oxides and morphology. Soil Science Society American Journal 68: 326335
Fiedler, S., Vepraskas, M.J., Richardson, J.L. (2007). Soil redox potential: importance, field
measurements, and observations. Advances in Agronomy 94, 54 pp
Fisk, M.C., Ruether, K.R., Yavitt, J.B. (2003). Microbial activity and functional composition
among northern peatland ecosystems. Soil Biology & Biochemistry 35: 591602
Fraser, C.J.D, Roulet, N.T., Moore , T.R. (2001). Hydrology and dissolved organic carbon
biogeochemistry in an ombrotrophic bog. Hydrological Processes 15: 31513166
Fraser, C.J.D., Roulet, N.T., Lafleur, M. (2001). Groundwater flow pattern in a large peatland.
Journal of Hydrology 246: 142-154.
Freeman, C., Ostle, N., Kang, H. (2001). An enzymic latch on a global carbon store. Nature
409 : 149
Frei, S., Knorr, K.H., Peiffer, S., Fleckenstein, J.H. (2012). Surface micro-topography causes hot
spots in wetland systems: a virtual modelling experiment. Journal Of Geophysical Research
117: 1-18
Frei, S., Lischeid, G., Fleckenstein, J.H. (2010). Effects of micro-topography on surfacesubsurface exchange and runoff generation in a virtual riparian wetland A modelling study.
Advances in Water Resources 33: 13881401
Frenzel, P., and Karofeld, E. (2000). CH4 emission from a hollow-ridge complex in a raised bog:
the role of CH4 production and oxidation. Biogeochemistry 51: 91112
Gessner, M.O. (2005). Proximate lignin and cellulose. In: Graa M.A.S. et al. (eds.) Methods to
study litter decomposition. A practical guide. Springer, Dordrecht, The Nederlands
Glaser, P.H., Chanton, J.P., Morin, P., Rosenberry, D.O., Siegel, D.I., Ruud, O., Chasar, L.I. and
Reeve, A.S. (2004). Surface deformations as indicators of deep ebullition fluxes in a large
northern peatland. Global Biogeochemical Cycles 18: 1-15
106

Golovchenko, A.V., Tikhonova, E.Y., Zvyagintsev, D.G. (2007). Abundance, biomass, structure,
and activity of the microbial complexes of minerotrophic and ombrotrophic peatlands.
Microbiology 76: 630637
Gondar, D., Lopez, R., Fiol, S., Antelo, J.M., Arce, F. (2005). Characterization and acidbase
properties of fulvic and humic acids isolated from two horizons of an ombrotrophic peat bog.
Geoderma 126: 367374
Gorham, E. (1991). Northern peatlands: role in the carbon cycle and probable responses to
climatic warming. Ecological Applications 1: 182-195
Gosselink, J.G. and Mitsch, W.J. (2000). Wetlands, third edition. John Wiley and Sons. New
york, USA.
Gosselink, J.G. and Mitsch, W.J. (2007). Wetlands, fourth edition. John Wiley and Sons.
Hoboken, New Jersey, USA.
Gsewell, S., Gessner, M.O. (2009). N : P ratios inuence litter decomposition and colonization
by fungi and bacteria in microcosms. Functional Ecology 23: 211219
Haavisto, V.F. (1974). Effects of a heavy rainfall on redox potential and acidity of a
waterlogged peat. Canadian Journal of Soil Science 54: 133-135.
Haraguchi, A. (1992). Seasonal change in the redox property of peat and its relation to
vegetation in a system of floating mat and pond. Ecological Research 7: 205-212
Heitmann, T. and Blodau, C. (2006). Oxidation and incorporation of hydrogen sulfide by
dissolved organic matter. Chemical Geology 235: 12 20
Heitmann, T., Goldhammer, T., Beer, J., Blodau, C. (2007). Electron transfer of dissolved
organic matter and its potential signicance for anaerobic respiration in a northern bog.
Global Change Biology 13: 17711785
Hilasvuori, E., Akujrvi, A., Fritze, H., Karhu, K., Laiho, R., Mkiranta, P., Oinonen, N.,
Palonen, V., Vanhala, P., Liski, J. (2013).Temperature sensitivity of decomposition in a peat
prole. Soil Biology & Biochemistry 67: 47-54

107

Husson, O. (2013). Redox potential (Eh) and pH as drivers of soil/plant/microorganism systems:


a transdisciplinary overview pointing to integrative opportunities for agronomy. Plant Soil
362: 389 417
Ivanon, K.E. (1981). Water movement in mirelands. Academic press, 1981
Jassey, V.J., Vechiapusio, G., Gilbert, D., Buttler, A., Toussaint, M-L., Binet, P. (2011).
Experimental climate effect on seasonal variability of polyphenol/phenoloxidase interplay
along a narrow fenbog ecological gradient in Sphagnum fallax. Global Change Biology 17:
29452957
Johnson, L.C. and Damman, A.W.H. (1991). Species-controlled Sphagnum decay on a south
Swedish raised bog. Oikos 61: 234-242
Karsisto, M. (1979). Effect of forest improvement measures on activity of organic matter
decomposing microorganisms in forested peatland. Suo 30: 49-58
Keller, J.K. and Bridgham, S.D. (2007). Pathways of anaerobic carbon cycling across an
ombrotrophic-minerotrophic peatland gradient. Limnology Oceanography 52: 96107
Keller, J.K., Weisenhorn, P.B., Megonigal, J.B. (2009). Humic acids as electron acceptors in
wetland decomposition. Soil Biology & Biochemistry 41: 15181522
Keuskamp, J.A., Dingemans, J.J.B., Lehtinen, T., Sarneel, J.M., Hefting, M.M. (2013). Tea Bag
Index: a novel approach to collect uniform decomposition data across ecosystems. Methods in
Ecology and Evolution 4: 1070-1075
Klpfel, L., Piepenbrock, A., Kappler, A., Sander, M. (2014). Humic substances as fully
regenerable electron acceptors in recurrently anoxic environments. Nature Geoscience 7: 195200
Knorr, K-H. and Blodau, C. (2009). Impact of experimental drought and rewetting on redox
transformations and methanogenesis in mesocosms of a northern fen soil. Soil biology &
Biochemistry 41: 11871198
Knorr, K-H., Lischeid, G., Blodau, C. (2009). Dynamics of redox processes in a minerotrophic
fen exposed to a water table manipulation. Geoderma 153: 379392

108

Kotiaho, M., Fritze, H., Meril, P., Tuomivirta, T.,Vliranta, M., Korhola, A., Karofeld, E.,
Tuittila, E-S. (2013). Actinobacteria community structure in the peat profile of boreal bogs
follows a variation in the microtopographical gradient similar to vegetation. Plant Soil 369:
103 114
Laiho, R. (2006). Decomposition in peatlands: reconciling seemingly contrasting results on the
impacts of lowered water levels. Soil biology & Biogeochemistry 38: 2011 2024.
Lovley, D.R., Fraga, J.L., Coates, J.D., Blunt-Harris, E.L. (1999). Humics as an electron donor
for anaerobic respiration. Environmental microbiology 1: 89-98
Lundin, L. and Bergquist, B. (1990). Effects on water chemistry after drainage of a bog for
forestry. Hydrobiologia 196: 167-181
Makiranta, P., Laiho, R., Fritze, H., Hytonen, J., Laine, J., Minkkinen, K. (2009). Indirect
regulation of heterotrophic peat soil respiration by water level via microbial community
structure and temperature sensitivity. Soil Biology & Biochemistry 41: 695703
Mansfeldt, T. (2003). Redox potential of bulk soil and solution concentration of nitrate,
manganese, iron, sulphate in two Gleysols. Journal of Plant Nutrient Soil Science 167: 7-16
Mettrop, I.S., Cusell, C., Kooijman, A.M., Lamers, L.P.M. (2014). Nutrient and carbon
dynamics in peat from rich fens and Sphagnum-fens during different gradations of drought.
Soil Biology & Biochemistry 68: 317-328
Mitchell C.P.J. and Branfireun, B.A. (2005). Hydrogeomorphic controls on reductionoxidation
conditions across boreal uplandpeatland. Interfaces Ecosystems 8: 731747
Moore, T.R. (2009). Dissolved organic carbon production and transport in Canadian peatlands.
In Carbon Cycling in Northern Peatlands Geophysical Monograph Series 184: 229-236
Morris, P.J. and Waddington, J.M. (2011). Groundwater residence time distributions in
peatlands: implications for peat decomposition and accumulation. Water Resources Research,
47: 1-12
Niedermeier, A., Robinson, J.S. (2007). Hydrological controls on soil redox dynamics in a peatbased, restored wetland. Geoderma 137: 318 326

109

Nikolaustz, M., Kappelmeyer, U., Szkely, A., Rusznyk, A., Mrialigeti,K., Kstner, M.
(2008). Diurnal redox uctuation and microbial activity in the rhizosphere of wetland plants.
European Journal of Soil Biology 44: 324 333
Nilsson, M. and quist, M. (2009). Partitioning litter mass loss into carbon dioxide and methane
in peatland ecosystems. Carbon Cycling in Northern Peatlands in Geophysical Monograph
Series 184
Patberg, W. (2011). Solute transport in Sphagnum dominated bogs. The ecophysiological effects
of mixing by convective flow. PhD thesis, University of Groningen
Peiffer, S., Klemm, O., Pecher, K., Hollerung, R. (1992). Redox measurements in aqueous
solutions - A theoretical approach to data interpretation, based on electrode kinetics.
Journal of Contaminant Hydrology 10: 1-18
Peng S., Piao, S., Wang, T., Sun, J., Shen, Z. (2009). Temperature sensitivity of soil respiration
in different ecosystems in China. Soil biology and Biochemistry 41: 1008-1014
Pezeshk, S.R. (2001). Wetland plant responses to soil ooding. Environmental and Experimental
Botany 46: 299312
Prescott, C.E. (2010). Litter decomposition: what controls it and how can we alter it to sequester
more carbon in forest soils? Biogeochemistry 101: 133 149
Proctor, M.C.F. (1994). Seasonal and shorter-term changes in surface-water chemistry on four
English ombrogenous bogs. Journal of Ecology 82: 597-610
Proctor, M.C.F. (2003). Malham tarn moss: the surface-water chemistry of an ombrotrophic bog.
Field Studies 10: 553 578
Rappoldt, C., Pieters, G-J.J.M., Adema, E.B., Baaijens, G.J., Grootjans, A.P., Duijn, C.J. (2003).
Buoyancy-driven flow in a peat moss layer as a mechanism for solute transport. PNAS 100:
1493714942
Reddy, K.R. and DeLaune, R.D. (2008). The biogeochemistry of wetlands: science and
applications. CRC Press Taylor & Francis Group

110

Rezanezhad, F., Couture, R.-M., Kovac, R., OConnell, D., Van Cappellen, P. (2014). Water
table uctuations and soil biogeochemistry: an experimental approach using an automated soil
column system. Journal of Hydrology 509: 245256
Robroek, B.J.M., Heijboer, A., Jassey, V.E.J., Hefting, M.M., Rouwenhorst, T.G., Buttler, A.,
Bragazza, L. (2013). Snow cover manipulation effects on microbial community structure and
soil chemistry in a mountain bog. Plant Soil 369: 151 164
Ryan, G.R., Melillo, J.R., Ricca, A. (1989). A comparison of methods for determining proximate
carbon fractions of forest litter. Canadian Journal of Forest Research 20: 166 171
Rydin, H. and Jeglum, J.R. (2013). Biology of peatland (Biology of habitats). 2nd ed. Oxford
University Press
Schaaf, S. van der (1999). Analysis of the hydrology of raised bogs in the Irish Midlands - A
case of study of Raheenmore Bog and Clara Bog. Doctoral thesis, Wageningen Agricultural
University. 375 pp
Scholotzhauer, M.S., Price, J.S. (1999). Soil water flow dynamics in a managed cutover peat
field, Quebec: field and laboratory investigations. Water Resources Research 35: 3675-3683
Seybold, C.A., Mersie, W., Huang, J., McNamee, C. (2002). Soil redox, pH, temperature, and
water-table patterns of a freshwater tidal wetland. The Society of Wetland Scientists 22: 149158
Shoemaker, C., Kroger, R., Reese, B., Pierce, S.C. (2013). Continuous, short-interval redox data
loggers: verication and setup considerations. Environmental Science: Processes Impacts 15
Shoemaker, C.M., Kroger, R., Pierce, S. (2012). Assessing a novel method for verifying
automated oxidation-reduction potential data loggers: laboratory and field tests. Society of
America Journal 73: 668-674.
Shoemaker, J.K., Varner R.K., Schrag D.P. (2012). Characterisation of subsurface methane
production and release over 3 years at a New Hampshire wetlands. Geochimica et
Cosmochimica Acta 91: 120-139
Silamikele, I., Nikodemus, O., Kalnina, L., Purmalis, O., Sire, J., Klavins, M. (2007). Properties
of peat in ombrotrophic bogs depending on the humification process.

111

Sjrs, H. (1948). Myrvegetation i Bergslagen. Acta Phytogeographyca Suecica 21: 1-299


Smemo, K.A. and Yavitt, J.B. (2011). Anaerobic oxidation of methane: an underappreciated
aspect of methane cycling in peatland ecosystems? Biogeosciences 8: 779793
Sneddon, J.I., Farstad, L., Lavkulich, L.M. (1971). Fiber content determination and the
expression of results in organic soils. Canadian Journal of Soil Science 51: 138-141
Steinmann, P. and Shotyk, W. (1996). Chemical composition, pH, and redox state of sulfur
and iron in complete vertical porewater profiles from two Sphagnum peat bogs, Jura
Mountains, Switzerland. Geochimica et Cosmochimica Acta 61: 1143-l 163
Stumm, W. and Morgan, J.J. (1981). Oxidation and reduction, in Stumm, W., Morgan, J.J.
(Eds.), Aquatic chemistry an introduction emphasizing chemical equilibria in natural waters.
Sundh, I., Nilsson, M., Granberg, G., Svensson, B.H. (1994). Depth distribution of microbial
production and oxidation of methane in northern boreal peatlands. Microbial Ecology 27:
253-265
Tarnocai, C., Canadell, J.G., Schuur, E.A.G., Khury, P., Mazhitova, G., Zimov, S. (2009). Soil
organic carbon pools in the northern circumpolar permafrost region. Global Biogeochemical
Cycles 23 doi: 10.1029/2008GB003327
Thomas, C.R., Miao, S., Erik Sindhj, E. (2009). Environmental factors affecting temporal and
spatial patterns of soil redox potential in Florida everglades wetlands. Wetlands 29: 1133
1145
Thompson, Y., Sandefur, C.B., Karathanasis, A.D., DAngelo, E. (2009). Redox potential and
seasonal porewater biogeochemistry of three mountain wetlands in Southeastern Kentucky,
USA. Aquatic Geochemistry 15: 349370
Thompson, Y., Sandefur, C.B., Miller, J.O., Karathanasis, A.D. (2007). Hydrologic and edaphic
characteristics of three mountain wetlands in southeastern Kentucky, USA. Wetlands 27: 174188
Tokida, P.T., Miyazaki, T., Mizoguchi, M., Nagata, O., Takakai, F., Kagemoto, A., Hatano, R.
(2007). Falling atmospheric pressure as a trigger for methane ebullition from peatland. Global
Biogeochemical Cycles 21: 1-8

112

Updegraff, K., John Pastor, J., Bridgham, S.D., Johnston, C.A. (1995). Environmental and
substrate controls over carbon and nitrogen mineralization in northern wetlands. Ecological
Applications 5: 151-163
Urquhart, C. and Gore, A.J.P. (1973). The redox characteristics of four peat profiles. Soil
Biology Biochemistry 5: 659-672.
Van der Molen, P.C. and Wijmstra, T.A. (1994). The thermal regime of hummock-hollow
complexes on Clara Bog, Co. Offaly. Biology and Environment: Proceedings of the Royal
Irish Academy 94B: 209-221
Van der Ploeg, M.J., Appels, W.M., Cirkel, D.G., Oosterwoud, M.R., Witte, J.-P.M., van der
Zee, S.E.A.T.M. (2011). microtopography as a driving mechanism for ecohydrological
processes in shallow groundwater systems. Vadose Zone Journal
Vile, M.A., Bridgham, S.D., Wieder, K.R. (2003). Response of anaerobic carbon mineralization
rates to sulfate amendments in a boreal peatland. Ecological Applications 13: 720734
Vorenhout, M., van der Geest, H.G., Hunting, E.R. (2011). An improved datalogger and novel
probes for continuous redox measurements in wetlands. International Journal of
Environmental Analytical Chemistry 91: 801 - 810
Waddington, J. and M., Roulet, N.T. (1997). Groundwater flow and dissolved carbon
movement in a boreal peatland. Journal of Hydrology 191: 122-138
Waddington, J.M. and Roulet, N.T. (1997). Groundwater flow and dissolved carbon movement
in a boreal peatland. Journal of hydrology 191: 122-138
Waddington, J.M., Morris, P.J., Kettridge, N.,Granath, G.,Thompson, D.K., Moore, P.A. (2014).
Hydrological feedbacks in northern peatlands. Ecohydrology DOI: 10.1002/eco.1493
Wang, M., Moore, T.R., Talbot, J., Richard, P.J.H. (2014). The cascade of C:N:P stoichiometry
in an ombrotrophic peatland: from plants to peat. Environmental Research Letters 9: 1-7
Wieder, K.R. and Vitt, H.D. (Eds) (2006). Boreal peatland ecosystems. Springer
Wieder, R.K. and Lang, G.E. (1988). Cycling of inorganic and organic sulphur in peat from Big
Run Bog, West Virginia. Biogeochemistry 5: 221-242

113

Wolf, K.L., Ahn, C., Noe, G.B. (2011). Microtopography enhances nitrogen cycling and removal
in created mitigation wetlands. Ecological Engineering 37: 1398-1406
Yang, J., Hu, Y., Bu, R. (2006). Microscale spatial variability of redox potential in surface soil.
Soil Science 171: 747-753
Zhang, D., Hui, D., Luo, Y., Zhou, G. (2008). Rates of litter decomposition in terrestrial
ecosystems: global patterns and controlling factors. Journal of Plant Ecology 1: 85-93

114

Вам также может понравиться