Interactive comment on “ The effect of a reciprocal peat transplant between two contrasting Central European sites on C cycling ”

The discussion phase has ended and we have two reviews available to this paper. Both reviewers recommend major revisions will be necessary before the paper becomes acceptable for publication in BG in particular the reviewers found that the scientific quality of the paper was only fair. However, both reviewers expressed that they believe that provided appropriate changes are made that the paper will become acceptable for publication. I agree with the reviwers and think that they made a number of very useful suggestion for improving the paper which the authors should consider when revising their paper. Any revised submission should be line-numbered and include a point-by-point reply to the reviewer and my (see below) comments.


Introduction
It is estimated that one third of global soil carbon is stored in wetlands, mainly at northern latitudes (Woodwell, and Mackenzie, 1995).According to most atmospheric circulation models, it is these northern latitudes where global warming will be most pronounced during the 21st century (IPCC, 2007).Microbial activities and solubilities of dissolved gases are strongly temperature-dependent (Moore et al., 1998).Therefore, recent studies have focused on whether thinning of peat deposits and higher emissions of greenhouse gases may contribute to stronger climate forcing, and accelerate temperature increases (Updegraff et al., 2001;Yavitt et al., 2005).Using inverse modelling, Mikaloff Fletcher et al. (2004) showed greater importance of the wetland global source of CH 4 than previously thought.Some carbon (C) sequestration studies have suggested that local hydrological factors are more important than climate (Yu et al., 2001).For Holocene peat profiles, concave plots of total mass vs. age have indicated continuous decomposition throughout the peat column (Wieder and Vitt, 2006).Over time, the efficiency of peatlands as atmospheric C sinks diminishes (Belyea, and Clymo, 2001).
Part of C photosynthesized in peatlands is returned to the atmosphere as CO 2 .Autotrophic and heterotrophic respiration comprise ca.one third of CO 2 uptake via photosynthesis (Heikkinen et al., 2002).Methane is formed either from acetate dissimilation (acetate pathway) or bicarbonate reduction (hydrogen pathway).In northern peatlands, acetate pathway dominates in shallow anaerobic peat layers, benefiting from warm summers and a relatively large supply of fresh labile organic matter from the acrotelm.The importance of the hydrogen pathway increases at greater peat depths with lower availability of labile organic carbon molecules (Hornibrook et al., 1997).
Isotopes can be used to constrain the global budgets of greenhouse gases, including CO 2 and methane (Bouwman, 1999).When isotope abundances in an atmospheric gas change, a shift in sources or sinks of the gas is indicated.Thus, isotopes may be instrumental in identifying a previously unknown source or sink of greenhouse gases.Anaerobic decay produces CH 4 depleted in 13 C, and CO 2 enriched in 13 C, relative to the source substrate (Gu and Schelske, 2004).Several attempts have been made to isotopically link two complementary reservoirs of carbon, solid peat substrate, and atmospheric gases (Lansdown et al., 1992;Jedrysek et al., 2005;Clymo and Bryant, 2008;Knorr et al., 2008).The δ 13 C values can potentially be a useful tracer of past methane emissions because biogenic methane contains isotopically extremely light C (Woodwell, and Mackenzie, 1995).Can removal of this isotopically light C from peat be detected in isotopically heavier residual organic C remaining in situ?An analysis of various aspects of this question was presented by Clymo and Bryant (2008).Before any isotope mass balance considerations are made, the degree of withinsite and between-site homogeneity in δ 13 C values of vertical peat profiles must be assessed.Upscaling the results of research plot studies is possible only if a variety of sites exhibit similar isotope features.Our first objective was to isotopically analyze bulk C in a number of replicate peat cores within two contrasting Sphagnum wetlands, and compare the vertical δ 13 C profiles.We discuss possible explanations of the found vertical δ 13 C profiles in solid peat.
Our second objective was related to the fact that the current rate of increase in methane concentrations in the atmosphere is lower than several decades ago (Dise and Verry, 2001).Suggestions have been made that elevated sulfur depositions in industrial countries have led to suppression of methane emissions from wetlands and rice paddies (Wieder and Vitt, 2006).Thermodynamically, anaerobic bacterial sulfate reduction outcompetes methanogenesis (Lamers et al., 2002;Blodau, and Moore, 2003).We used a sharp north-south pollution gradient in Central Europe to investigate whether higher S inputs lead to lower rates of methanogenesis.Our third objective was to evaluate whether CO 2 and CH 4 emission rates in Sphagnum peat bogs are governed mainly by organic matter quality in the substrate, or by environmental conditions.Again, we used the north-south pollution gradient in Central Europe, and conducted a reciprocal peat transplant experiment between a highly polluted and a relatively unpolluted Sphagnum bog.Reciprocal transplants of organic soil among contrasting sites were previously used to study the effect of site-specific controls on biogeochemical (Mulligan and Gignac, 2001;Keller and Bridgham, 2007;Westbrook et al., 2006) and biological (Sohlenius and Bostrom, 1999) variables.Reciprocal Sphagnum transplant experiments were performed by Wieder and Yavitt (1994), Yavitt and Wieder (2000), and Yavitt et al. (2005).We hypothesized that if organic matter quality in the substrate, i.e., the ratio between labile and recalcitrant C forms, is the primary control of terminal C mineralization, the emissions of CO 2 and CH 4 from the host site will remain similar to those at the home site.If environmental parameters are the primary control, the emissions of CO 2 and CH 4 will become similar to the host site.

Study sites
The two peat bogs selected for the reciprocal transplant experiment (Fig. 1, Table 1) represented extremes in atmospheric S deposition.The northern site, Velke jerabi jezero (VJJ), is located in the "Black Triangle" region, one of the most polluted parts of the world.The Black Triangle comprises the northern Czech Republic, southeast Germany   (Novak et al., 2005).After 1987, industrial emission rates decreased rapidly, and are now 90% lower.The southern site, Cervene blato (CB, Fig. 1, Table 1), is located in the relatively unpolluted Trebon Basin, a natural preserve near the Czech border with Austria.The bog is surrounded by healthy mature spruce stands.The distance between CB and VJJ is 220 km.CB is currently four times less polluted with atmospheric S and twice less polluted with reactive forms of atmospheric N, compared to VJJ.

Design of peat transplant experiment
A total of 30 peat cores, 40-cm long, 10-cm in diameter, were collected in April 2002 using polyvinylchloride (PVC) cylinders with a sharpened bottom edge.Fifteen peat cores were collected at VJJ and 15 peat cores at CB.All cores were taken from peat lawns whose surface was 100% Sphagnumcovered, with no vascular plants.Predominating Sphagnum species are given in Table 2. Five cores from each site were transported to the laboratory and analyzed.At each site, additional five cores were transplanted back to their original positions.Before replacing the five cores into the soil at their home sites, the cores were removed from the PVC cylinders, wrapped individually with plastic mesh (openings 2.5 mm×2.5 mm), leaving the surface free.The mesh was fastened around each peat core and secured with cable ties.At each site, the remaining five cores were capped and placed vertically into a cooler for immediate transport to the host site.Five replicate peat cores were transplanted from VJJ to CB, and five replicate cores from CB to VJJ.Upon arrival at the host site, the five peat cores were removed from PVC cylinders, wrapped with mesh and inserted into the positions of five local peat cores, which had been previously collected and transported to their host site.
The transplanted cores remained in the soil for 18 months.In September 2003, we retrieved 10 cores from VJJ (five control cores originating from VJJ and five transplanted cores originating from CB) and 10 cores from CB (five control cores originating from CB and five transplanted cores originating from VJJ).The cores were transported to the laboratory.
Cylindrical incisions in peat may impact the field experiments (see Yavitt et al., 2005 and references therein).For example, exudate dynamics of severed roots are affected for several hours.The length of our experiment (18 months), however, permits equibrium to be reached soon after its beginning.In a transplanted peat core, pore water of the host site replaces pore water of the home site within several days after insertion into its new position.The reason is high hydraulic conductivity in topmost 30 to 40 cm of the bog.
In month 18, we were able to recover peat cores that were somewhat shorter than peat cores analyzed at time 0 (30 vs. 40 cm).With no roots of vacular plants, peat at the bottom of the cores was nearly fluid and some of it may have been lost by advection of bog water when being transplanted.
One 50-cm long, 10-cm in-diameter, peat core was collected from each VJ and CB in April 2002 for 210 Pb dating of peat accretion history.

Analysis
Five 40-cm long peat cores per site taken into the laboratory at time t = 0 (April 2002) were frozen, cut into 2-cm thick segments, dried at 60 • C, weighed, and homogenized.Ash content was determined on a 0.5 g aliquot at 550 • C. Carbon concentrations were measured on a Carlo Erba Elemental Analyzer and corrected for ash content, C isotope compositions were determined on a Finnigan MAT 251 mass spectrometer following on-line combustion of the peat.Carbon isotope data are given in the usual notation as δ 13 C values relative to the V-PDB standard.Reproducibility of δ 13 C determinations was better than ± 0.3‰.
The two 50-cm long peat cores were processed as above.Three grams of pulverized peat from each 2-cm section were digested with HCl and HNO 3 , spiked with 209 Po as a chemical yield tracer, and 210 Po, along with 209 Po were plated on Ag discs (Vile et al., 2000).Polonium activities were measured on an EG&G Ortec 576 A dual α spectrometer, and 210 Pb dates calculated.
A 27-hour laboratory incubation study was performed to assess CO 2 and methane emission potentials from transplanted and control peat cores extruded in September 2003.The duration of the incubation was chosen in accord with previous studies that reported linearity of both CO 2 and methane emission rates over time (Vile et al, 2003;Turetsky, and Ripley, 2005).Three peat cores were randomly selected from each of the following four treatments: CB cores extruded from CB, CB cores extruded from VJJ, VJJ cores extruded from VJJ, and VJJ cores extruded from CB.The headspace was purged with N 2 for 10 min.The 12 capped peat cores were anaerobically incubated at 20 • C and gas samples taken 2, 4, 8, 12 and 27 h after the beginning of the incubation.At each time, 60 mL of headspace gas were removed using a syringe and injected into an evacuated Hungate tube.An additional 60 mL of N 2 were added to each headspace for vacuum prevention, and headspace dilution corrections were made.Concentrations of CO 2 and CH 4 were determined on a Shimadzu 14A gas chromatograph.Rates of CO 2 and CH 4 emissions were estimated from changes in headspace concentrations over time.Hungate tube content was transferred by He carrier into a line for trace gas measurements.CO 2 was collected in a condensation trap, CH 4 was combusted over Pd catalyst.The δ 13 C values were determined with reproducibility better than ± 0.5‰.
Following the 27-hour laboratory incubation, the 10 peat cores previously transplanted to their host site (five from CB to VJJ, and five from VJJ to CB) were frozen, cut into 2-cm thick segments, dried at 60 • C, weighed, and homogenized.Bulk C concentrations and δ 13 C values were determined for each 2-cm thick peat segment.
Statistical analysis was performed using linear mixed models (Laird and Ware, 1982) and the R 2.6.2 software (R Development Core Team, 2008).For CO 2 and CH 4 emissions, a linear mean evolution was used with random in- tercept and slope for each sample to account for only conditional independence of repeated observations.Using the method of restricted maximum likelihood, estimates of the mean intercepts, mean slopes, covariance matrix and residual variance were computed.The estimated mean evolution of CO 2 and CH 4 emissions is shown in Figs.A1 and A2 in the Electronic Annex (http://www.biogeosciences.net/7/921/2010/bg-7-921-2010-supplement.pdf).

Peat accretion history
Cumulative dry mass increased smoothly with an increasing depth and age at both sites, and was slightly higher at CB than at VJJ (Fig. 2a).Lead-210 dates revealed faster peat accumulation at the northern, colder site VJJ, compared to the southern, warmer site CB (Fig. 2b).At a depth of 34 cm, peat at CB was 140 years old, whereas peat at VJJ was just 55 years old.

Downcore C concentration trends
The vertical C concentration patterns differed between the two sites (Fig. 3).Whereas C concentrations at VJJ steadily increased with an increasing peat depth, those at CB showed no vertical gradient.

Downcore δ 13 C trends
Also the vertical trends in δ 13 C values of bulk peat differed between the two sites (Fig. 3).With an increasing depth, peat at CB became progressively enriched in the heavier isotope 13 C (higher δ 13 C values).In contrast, peat from VJJ exhibited a clear-cut δ 13 C maximum at a depth of 20 cm below surface.Between 0 and 20 cm, peat at VJJ became progressively enriched in the heavier isotope 13 C, similar to CB.However, at depths greater than 20 cm, peat at VJJ became progressively enriched in the lighter isotope 12 C, and the δ 13 C values decreased with increasing depth.Importantly, at both sites, when a specific vertical δ 13 C trend was seen in one peat core,  the same trend was also seen in all the remaining peat cores.There was a high degree of within-site homogeneity in vertical δ 13 C trends.

The effect of peat transplant on substrate C
Figures 4 and 5 show mean C concentrations and δ 13 C values for five replicate peat cores per treatment, along with standard errors.The four treatments are marked as follows: "CB" are peat cores taken from CB to the laboratory at the begin- ning of the experiment."VJJ" are peat cores taken from VJJ to the laboratory at the beginning of the experiment."CB to VJJ" are peat cores originating from CB which were transplanted to VJJ for 18 months, and then analyzed."VJJ to CB" are peat cores originating from VJJ, which were transplanted to CB for 18 months, and then analyzed.
As seen in Fig. 4, transplanting peat cores from VJJ to CB had no effect on the vertical C concentration gradient: both sets of peat cores originating from VJJ had higher C concentrations at greater depth than at peat surface (cf., Fig. 4a  and 4b).CB peat cores had similar C concentrations at all depths.When peat cores were transplanted from CB to VJJ, higher C concentrations were seen at greater depths.The CB to VJJ carbon concentration profiles were similar to VJJ concentrations at the beginning of the experiment (Fig. 4b).A complementary representation of the C concetration data is in Fig. A5 (Electronic Annex (http://www.biogeosciences.net/7/921/2010/bg-7-921-2010-supplement.pdf); an overlay plot for CB and CB to VJJ).
As seen in Fig. 5, transplanting peat cores between sites had little effect on their vertical δ 13 C trends.CB to VJJ peat cores exhibited higher δ 13 C values downcore, identical to CB peat cores (Fig. 5a).VJJ to CB peat cores exhibited a peak in δ 13 C values at 20-cm depth, similar to VJJ peat cores (Fig. 5b).

Emissions of CO 2 and CH 4 from incubated peat
The 27-hour incubation was performed on peat cores taken into the laboratory at time t = 18 months.The four preceding treatments (transplant types

Comparison of CO 2 and CH 4 emissions among treatments
Figure 6 compares CO 2 and CH 4 emission rates and δ 13 C values among the four transplant types, obtained from the 27-hour laboratory incubation.All data in Fig. 6 were measured at room temperature (20 • C).For both CO 2 and CH 4 , the same relative magnitude of emission rates was observed: CB to CB had the lowest gas emission rates, VJJ to VJJ had the highest gas emission rates.CO 2 emissions from the peat cores kept at their home site VJJ were almost four times higher than CO 2 emissions from the peat cores kept at their home site CB.Methane emissions from the peat cores kept at their home site VJJ were nine times higher than those in any other treatment.Interestingly, all four measured parameters in every graph in Fig. 6 (concentrations and isotope compositions of CO 2 and CH 4 ) had converged to parameters of their host site by the end of the transplant experiment.Transplanting peat cores from CB to VJJ resulted in a 3.5-fold increase in CO 2 emission, approaching that of the host site.
Transplanting peat cores from VJJ to CB resulted in a 50% decrease in CO 2 emissions, approaching those of the host site.A similar pattern was seen for methane emissions, e.g., transplanting peat cores from VJJ to CB resulted in 15 times lower methane emissions compared to the home site VJJ.

Comparison of δ 13 C-CO 2 and δ 13 C-CH 4
Also the δ 13 C values of both gases emitted from transplanted peat cores converged to those of the host site (Fig. 6).The δ 13 C values of both CO 2 and methane were lower for VJJ to VJJ peat cores than for CB to CB peat cores.The peat cores producing more CO 2 and methane at their home location (i.e., VJJ) always had isotopically lighter carbon.In general, transplanting peat cores to a host site resulted in either a decrease or an increase in δ 13 C so that the C isotope composition became more similar to the host site (Fig. 6).

C sequestration in peat
Over millennia, peat deposits developing in colder climate tend to accumulate more carbon and are thicker compared to peat deposits developing in warmer climate (Yu et al., 2001;Beleya, and Baird, 2006).Such a relationship was found for VJJ and CB for the most recent 150 years.The mean annual temperature at the northern site VJJ is 3.8 • C lower compared to the southern site CB.The vertical growth rate of VJJ was twice that of CB (Fig. 2b).Substrate accumulates whenever the rates of net primary production are greater than the rates of decomposition.While higher temperature leads to higher Sphagnum growth, also the decomposition rate is greater.Lower net primary production at lower temperature is usually associated with slower decomposition.A number of records of recent peat show similar litter additions and decay rates despite climatic differences (Belyea and Clymo, 2001;Wieder and Vitt, 2006;Novak et al., 2008).The positive relationship between C accumulation and colder climate valid for entire Holocene deposits is not always valid for the topmost peat layers.Net primary production appears to be more sensitive to environmental parameters than decay (Moore et al., 1998).The main difference between VJJ and CB is in fertilization levels by anthropogenic nutrients (Table 3).Higher availability of nutrients at VJJ may have played a role in peat accumulation (see Sect. 4.8).Temperature may not be the main control of the greater peat accumulation at VJJ.Two more aspects of peat accumulation at VJJ and CB bear mention.First, the peat accumulation rates at our sites are unusually high compared to other sites in the temperate climate zone.This is illustrated in Fig. A6, in which depth vs. age profiles of VJJ and CB are compared to 10 different sites in Europe and the US.More peat (and thus C) accumulation at the Czech sites is seen especially in layers older than 50 years.
Second, previous studies have documented that Sphagnum growth rates differ between species (Vitt et al., 2003;Asada, and Warner, 2005), and that following extended periods of exposure to nutrients, Sphagnum species distribution may change (Hobbie et al., 2000).Also the decay rate is affected by species, e.g., S. fuscum decays more slowly than S. cuspidatum (Johnson, and Dammon, 1991).At our sites, a number of various Sphagnum species are present (Table 2), and sofar we have no indication of a relationship between the Sphagnum species community and peat accumulation rate.
Several authors have questioned the validity of 210 Pb dating of peat accretion history (Urban et al., 1990).However, Pb immobility in peat was verified experimentally (Vile et al., 1999).Turetsky et al. (2004) published a critical review of the potential of residual 210 Pb activities as a dating tool, and of techinques used to corroborate 210 Pb dates in peat.

Controls on δ 13 C of ageing peat
The δ 13 C values of vertical peat profiles are a useful tool for studying maturation and diagenesis of peat if four conditions are met: (i) δ 13 C gradients exist in peat substrate with age, (ii) δ 13 C gradients are widespread, (iii) C maturation in peat is isotopically selective, and (iv) other isotope-selective processes can be distinguished from peat maturation.As seen from Figs. 3 and 5, clear-cut vertical δ 13 C signals are found in Sphagnum peat.From peat surface to a depth of 15-20 cm, these signals (higher δ 13 C downcore) seem to be common (Knorr et al., 2008).In contrast, we do not know of any other peatland where a systematic decrease in δ 13 C downcore (such as the one starting at a depth of 20 cm at VJJ) was reported.
A small kinetic C isotope fractionation during peat degradation was reported by Wynn et al. (2006).In an open system, Rayleigh distillation may result in isotopically heavier residual C in situ while low-δ 13 C is removed.Indeed, biogenic methane can have δ 13 C values as low as −110‰.There is an analogy between degradation of organic C in wellaerated upland soils, and wetland soils.For upland soils, Nadelhoffer and Fry (1988) concluded that mineralizationrelated preferential release of the light isotope 12 C may contribute to higher δ 13 C in deeper soil horizons.Isotopically heavier residuum in deeper soils has been reported also for N (Nadelhoffer, andFry, 1988 andreferences therein, Novak et al., 2003), and for S (Novak et al., 1996(Novak et al., , 2005)).Degradation of organic S-containing molecules in peat also preferentially releases the light isotope 32 S (Novak et al., 1994(Novak et al., , 1999)).
Two processes linked to assimilation of atmospheric CO 2 by plants can mask a possible C isotope effect of peat degradation.Burning of fossil fuels over the past 200 years has injected into the atmosphere isotopically light CO 2 -C (Friedli et al., 1986), and this lower δ 13 C (decrease of 2‰ since the early 1800s) may be seen in the uppermost layers of organic upland soils.While this effect is well documented, we propose that it is not the dominant control of δ 13 C values in young peat layers in our study.As seen in Figs. 3 and 5, gradual admixture of isotopically light fossil-fuel C from the atmosphere is not consistent with the VVJ δ 13 C profile below the depth of 20 cm.When moving upward, higher, instead of lower, δ 13 C were observed.In these peat layers, dating from the early and mid-20th century (Fig. 2), admixture of the industrial low-δ 13 C carbon is not seen.Instead, the lower δ 13 C at greater depth at VJJ could be explained by drier conditions before 1987 (the year corresponding to the 20-cm peat depth).Lower water content leads to lower external CO 2 diffusion resistance, and greater C-isotope discrimination upon assimilation.If so, δ 13 C records in peat could be used as a geochemical archive of site-specific moisture conditions.
Under different climatic conditions, different Sphagnum species predominate.If the climate changes at a single site, this may result in a third proces that may mask a diagenesisrelated downcore shift in δ 13 C of peat substrate.An assumption would be that different Sphagnum species carry different δ 13 C signatures.Rice (2000), however, showed that changes of species composition had no effect on δ 13 C of Sphagnum.Skrzypek et al. (2007) found that higher temperature leads to slightly lower δ 13 C of Sphagnum cellulose within the same species.Loader et al. (2007) found small differences in δ 13 C of stems, pendants and horizontal branches of S. capillifolium.

The C isotope mass balance approach
Clymo, and Bryant ( 2008) studied a 7-meter deep peat deposit in the UK.Their results illustrate that closing a historical C isotope budget may be difficult, due to several poorly understood parameters.At the UK site, δ 13 C values www.biogeosciences.net/7/921/2010/Biogeosciences, 7, 921-932, 2010 of bulk peat substrate and DOC did not change with depth and were close to −27‰.In contrast, porewater methane C was isotopically much lighter (−70‰).The δ 13 C value of porewater CO 2 increased with increasing peat depth from ca. −10 to + 10‰.DOC and dissolved gases at each depth were hundreds to thousands years younger than solid peat.The amount of CO 2 in pore water was 5-10 times greater than the amount of CH 4 .In case of methane production via acetate fermentation, we would expect the same magnitude of two opposite C isotope shifts relative to δ 13 C of the source substrate (more negative δ 13 C-CH 4 , and more positive δ 13 C-CO 2 ), but also the same concentrations/fluxes of both gases.Neither was observed (Clymo and Bryant, 2008).The unknown parameters include δ 13 C of historical emissions of CO 2 and CH 4 , and differences in diffusion coefficients of 12 C and 13 C of dissolved gases.Hornibrook et al. (2000) reported a confounding factor for C isotope systematics in peat profiles: at greater depths, autotrophically formed acetate may be present, leading to somewhat lower fractionation than otherwise would be expected in the CO 2 -reduction dominated part of deep catotelm (typical fractionation factors α < 1.032 for acetoclastic methanogenesis and 1.045-1.082for hydrogenotrophic methanogenesis; Whiticar, 1999).Hornibrook et al. (1997) found increasing δ 13 C-CO 2 and decreasing δ 13 C-CH 4 values from peat surface to a depth of 1 m (from −12 to −6‰, and from −55 to −65‰, respectively), attributing them mainly to a switch from the acetate to the hydrogen pathway of methanogenesis.Steinmann et al. (2009) found, similar to Clymo and Bryant (2008), also 5 times higher dissolved CO 2 concentrations compared to CH 4 in deep peat, and interpreted the difference as a result of more intense ebullition of methane, which is the less soluble of the two gases.Knorr et al. (2008) manipulated water table level in peat monoliths in a 300-day experiment, and applied the isotope mass balance approach to individual depth segments.The δ 13 C value of solid substrate was −27.1 to −27.7‰.The δ 13 C value of emitted CO 2 was between −20 and −23‰, i.e., higher than δ 13 C of the substrate (heavier C).Our data do not permit construction of an isotope mass balance similar to that of Knorr et al. (2008), due to a different sampling protocol.However, we note that a major problem associated with these calculations is uncertainty related to the soil mass.A very small error in dry mass (or presence of silicate grains) affects the uncertainty of the gas partitioning, for example, of the flux of anaerobically produced CO 2 .The reason is much lower mass of the gas compared to the mass of the soil.We encountered this difficulty in our previous work, specifically when calculating δ 34 S of emanating H 2 S in laboratory soil incubations (Novak et al., 2001).

Decoupling of C concentrations and δ 13 C in peat profiles
In intact peat profiles, increasing C concentrations at greater depth are often viewed as a manifestation of substrate degradation, with faster removal of functional -OH groups than removal of carbon (Wieder and Yavitt, 1994).Surprisingly, we found that, in some sample types, the two signs of substrate degradation were decoupled.For example, at CB, δ 13 C values increased downcore, possibly reflecting substrate degradation (emission of low-δ 13 C methane), while C concentration did not increase.Further study is needed to ascertain how common the decoupling of C concentration and isotope trends in maturating peat is.

Linking sulfate reduction and methane emission
Atmospheric S inputs at VJJ were four times higher than at CB at the time of the transplant experiment, and 10 times higher 15 years earlier (Novak et al., 2005).Table 3 illustrates that sulfate concentration in pore water at VJJ is still 10 times higher compared to CB, despite easing atmospheric pollution.Because microbial sulfate reducers outcompete methanogens (Dise, and Verry, 2001), we expected higher methane emissions at CB than at VJJ.In fact, the opposite was true, and peat cores originating from the polluted site VJJ had nine times higher CH 4 emission rate than other peat cores.Methane emissions at CB were close to zero.This is an unexpected result, because the most obvious explanation, the presence of vascular plants, is ruled out.There were no vascular plants on the research plots.Vascular plants contain conduits transporting methane directly to the atmosphere, protecting it from oxidation in the acrotelm.The measured emission rates of methane at our sites were affected by methane oxidation, which mainly takes place at the lower edge of the green Sphagnum plants.This is documented by δ 13 C-CH 4 values in the range of −20 to −45‰, a result of preferential oxidation of low-δ 13 C methane (Chanton, 2005).Fresh biogenic methane has δ 13 C values between −50 and −110‰, depending on the methanogenic pathway (Wieder and Vitt, 2006).

Comparison of δ 13 C-CO 2 and δ 13 C-CH 4 in incubated peat and in the field
The ranges of δ 13 C values of CO 2 and methane in our incubation experiment were quite narrow (Fig. 6), and different from those reported by previous studies.Landsdown et al. (1992) to −40.8‰), compared to the laboratory incubation (−31.3 to −52.7‰).On average, carbon of both CO 2 and CH 4 was isotopically heavier in the field than in the laboratory incubation.This was true for both sites.The laboratory conditions (identical for all treatments) resulted in a somewhat lower degree of methane oxidation than under field conditions.We offer a similar explanation also for the observed lower δ 13 C values of both CO 2 and methane at VJJ compared to CB in the laboratory (Fig. 6).VJJ exhibited higher CO 2 and methane emission rates than CB.In peat cores from both sites with the same water level during incubation, methane oxidation was more similar than methane production.Thus a lower percentage of the produced methane was oxidized at VJJ, leaving behind isotopically lighter methane C compared to CB (freshly produced CH 4 has a very low δ 13 C value on average close to −70‰).In local peat cores from both sites, C of CO 2 was isotopically somewhat heavier than C of methane, providing additional evidence of methane oxidation (Fig. 6).

The effect of peat transplant on CO 2 and CH 4 emissions
The main objective of the reciprocal peat transplant experiment was to evaluate whether organic carbon quality in the substrate, or environmental parameters are the main control of CO 2 and CH 4 emission rates.At the studied depths, peat from VJJ was younger and the substrate could potentially provide more labile organic C than the substrate from CB.
If organic C quality in the substrate was the main control of greenhouse gas emissions, we would expect high CH 4 emissions from VJJ peat transplanted to CB.That was not observed.Over the 18 months of the experiment, metabolism and trace gas emissions from all transplanted peat cores converged to those of the host sites (Fig. 6).This allows us to conclude that environmental parameters were by far the dominant control of terminal C mineralization potentials at our study sites.Availability of labile C forms in solid peat played a much less significant role, as illustrated by a dramatic suppression of CH 4 emissions when VJJ peat was transplanted to CB (Fig. 6).The role of DOC will be discussed separately in Sect.4.9.

Effect of elevated inputs of nutrients on C cycling in shallow peat
Aerts et al. (2003) found in a 12-year field manipulation study, during which N and P were added to a riparian peatland and grassland, that the initial plant community was the main determinant of C losses through decomposition.Nutrient additions enhanced litter production, but had no effect on litter decomposition (i.e., on terminal C mineralization, and, by implication, also on greenhouse gas emissions).In contrast, Limpens and Berendse (2003) concluded from a litterbag study that Sphagnum decomposition was stimulated by nitrogen deposition.Bragazza et al. (2006) showed along a European pollution transect that increased N availability resulted in higher CO 2 emissions and increased DOC production.Increased N availability favoured microbial decomposition by removing N contraints on microbial metabolism, by higher litter quality, and by stimulated microbial enzymatic activity.As seen in Table 3, VJJ bog waters had higher concentrations of both NH + 4 and NO − 3 , compared to CB.Based on the existing literature, we suggest that the higher industrial pollution of the northern site VJJ may have contributed to higher CO 2 emission rates.In theory, also methane production may be ehanced by a slight nutrient enrichment through higher production of free aminoacids and sugars.

The role of DOC
Microbial populations in peat core transplanted to a host site start using the local DOC for their metabolism within hours to days, due to high permeability of the substrate to the lateral flow of local pore waters.In this respect, DOC resembles www.biogeosciences.net/7/921/2010/Biogeosciences, 7, 921-932, 2010 environmental parameters of the host site, (temperature, atmospheric deposition of nutrients/pollutants) which also start affecting the imported substrate almost immediately.
In general, DOC production decreases with increasing peat depth.The contribution of recent Sphagnum photosynthate C to DOC may be as much as 4% (Fenner et al., 2004).As mentioned previously, Clymo and Bryant (2008) found younger 14 C ages in individual depth levels of a peat bog compared to in-situ peat substrate.This agrees with the finding of Chasar et al. (2000) that recently photosyntesized DOC is transported downcore.Fresh DOC laterally entering transplanted peat cores plus freshly formed plant exudates may affect C cycling by supplying labile organic C forms (Crow and Wieder, 2005).At our sites, we did not study individual chemical C forms in DOC, however, DOC concentrations were known.During the growing season, with the water table only 2 cm below Sphagnum capitula, Hruska (1994) reported significantly higher DOC concentrations at CB (95-101 mg L −1 ), compared to any other peat bogs in the Czech Republic.sites in the south near CB had 7 to 35 mg DOC L −1 .VJJ and other sites in the north of the country had 5 to 42 mg DOC L −1 .We conclude that DOC concentrations themselves do not explain higher rates of terminal C mineralization at VJJ compared to CB.

Conclusions
Sensitivity of carbon cycling in peatlands to various environmental parameters, such as mean annual temperature and sulfur and nitrogen deposition, was studied at two Sphagnum-dominated peat bogs in Central Europe.Both sites were waterlogged throughout the year, with negligible differences in hydrology.It was found that the colder northern site VJJ accumulated more organic carbon over the past 150 years than the southern warmer site CB.The emission rates of both products of terminal C mineralization, CO 2 and CH 4 , were also higher at the colder site VJJ.One possible explanation is that peat at the northern site VJJ had been continuously fertilized since the beginning of the Industrial Revolution.While net primary productivity increased due to higher nutritional status of the northern site, decomposition rates, which are believed to be less sensitive to environmental parameters, were also higher at the northern than at the southern site.The difference between organic mass production and decomposition was still higher at the northern site compared to the southern site.
Since microbial sulfate reducers outcompete methanogens, we hypothesized that the sulfur-polluted site VJJ should have lower methane emissions than CB.However, the opposite was true, with higher methane emissions from VJJ.This finding cannot be explained by presence of vascular plants at VJJ, known to facilitate methane transport to the atmosphere, because such plants were not found at the sampling sites.
A replicated peat transplant experiment revealed that imported peat adjusted its carbon cycling to that of the host site.It follows that the organic carbon quality of the substrate did not control C transformation rates.Rather, environmental parameters, such as nutritional status, were the main controls of C storage in the wetland.Fluxes of CO 2 and CH 4 from the peat to the atmosphere were controlled mainly by local plus imported microbial communities, metabolizing DOC of local porewaters under local environmental and nutritional conditions.
Extensive replication of peat monoliths, needed for a peat transplant experiment, was used for a complementary C isotope study of both the solid substrate and emitted gases, CO 2 and CH 4 .Vertical δ 13 C trends in solid peat exhibited a high degree of within-site homogeneity, but were site specific.In all 20 studied peat cores, the δ 13 C value increased downcore in the topmost ca.20 cm.This trend could have been caused by both preferential removal of the light isotope 12 C from older peat due to decomposition, or by growing fossil fuel burning which spikes 12 C-rich CO 2 into the atmosphere.In peat deeper than 20 cm at VJJ, however, δ 13 C increased upcore.VJJ peat at the 20 cm depth was relatively young (year 1987).Between 1880 and 1987, fossil fuel burning was already adding isotopically light CO 2 -C to the atmosphere, and could not have dominated the increasing δ 13 C trend upcore at VJJ.Consequently, a different C isotope selective mechanism must have dominated the C isotope systematics between 1880 and 1987 at VJJ, overprinting both peat degradation and fossil-fuel derived C input.This mechanism might be related to historical moisture conditions (most severe 150 years ago), because droughts lead to greater C-isotope discrimination during photosynthesis.In general, both assimilation of atmospheric C by Sphagnum, and decay of Sphagnum, are associated with C isotope fractionation, which may potentially serve as an archive of past changes in the peatland C sink.Dry conditions lead to lower δ 13 C values of Sphagnum, substrate maturation (itself depending on climate) leads to higher δ 13 C values of the residual organic matter.The vertical δ 13 C profile in peat is a result of an interplay of these isotope selective processes.The steepness of the downcore δ 13 C shift may contain information about cummulative removal of isotopically light C via methane production.
) included: CB to CB, CB to VJJ, VJJ to VJJ and VJJ to CB. Detailed time series of concentrations of CO 2 and CH 4 and their δ 13 C values during the 27-hour incubation are depicted in Figs.A1-A4 in the Electronic Annex (http://www.biogeosciences.net/7/921/2010/bg-7-921-2010-supplement.pdf).CO 2 concentrations increased linearly in all four treatments.Methane concentrations increased in three treatments during the first 12 h.

Fig. 6 .
Fig. 6.The concentration and C isotope composition of CO 2 and methane emanating from peat at the end of the transplant experiment (mean ± SE, n = 3).Different letters denote statistically significant differences (p <0.05).

Table 1 .
Characteristics of the studied peat bogs.Atmospheric S and N inputs into semi-open peat bogs were calculated as arithmetic mean of open-area deposition and spruce canopy throughfall. a

Table 2 .
Sphagnum species present at the two study sites.
and southern Poland.VJJ is situated in the Krusne hory Mts.near a cluster of coal-fired power plants, which were built in the 1950s, but had desulphurization units installed only in the mid-1990s.Spruce (Picea abies) stands died back on 1000 km 2 in the Black Triangle between 1975 and 1995, partly as a result of high atmospheric SO 2 concentrations.In the vicinity of VJJ (25 km north-west of the power plants), symptoms of crown thinning and needle yellowing were observed.The industrial air pollution in

Table 3 .
Concentration of nutrients in surface peat bog waters.Years of sampling: 2002-2003, number of samples n = 12 per site.Means ± standard errors (mg L −1 ) are given.

Table 4 .
, for example, reported δ 13 C-CH 4 values of −73‰ (isotopically lighter C compared to CB and VJJ) and δ 13 C-CO 2 values of around −2‰ (isotopically heavier C compared to CB and VJJ).In Table4we summarized unpublished C isotope data for CO 2 and methane, captured in a passive chamber at CB and VJJ under field conditions.The field data exhibited greater variability in δ 13 C-CH 4 values (−7.5 Concentrations and δ 13 C values of CO 2 and methane emitted from VJJ and CB Sphagnum-covered bog surfaces.The gas samples were taken on 9 July 2002 (VJJ), and 11 July 2002 (CB) under ambient air temperature of 20 ± 2 • C. SE=standard error.