Recent paludification rates and effects on total ecosystem carbon storage in two boreal peatlands of Northeast European Russia

Abstract

Forest and peatland ecosystems constitute the two major carbon pools in the boreal region. We assess the evolution in total storage and partitioning of ecosystem carbon following recent paludification of forest into peatland at two sites in Northeast European Russia. Based on radiocarbon dating of basal peat and quantification of total ecosystem carbon storage, our results show that paludification rates and its consequences for carbon storage vary significantly between sites. A peatland expanding on ground with steeper slopes has experienced a slow lateral advance in recent times, about 2.6 m on average per century, whereas a peatland in flatter terrain has expanded much more rapidly, about 35 m on average per century. The total ecosystem carbon storage (sum of phytomass, top soil organics or peat, and 30 cm of underlying mineral soil) showed a long-term trend toward increased ecosystem C storage following the replacement of forest (mean value = 20.8 kg C/m², range = 13.0–43.4 kg C/m²) by peatland (>100 kg C/m² in the deepest peat deposits). However, the transitional stage in which the forest is replaced by the margin of the peatland results in a short-term decrease of carbon stored in the ecosystem with a mean loss of 7.5 kg C/m². After the initiation of a peatland through paludification, a period of decades to centuries of peat accumulation is needed to compensate for the initial loss of carbon. In the short term, an intensification of the paludification process could lead to a loss of carbon stored in the boreal region.

Keywords

boreal forest carbon storage ecotone expansion rate late Holocene paludification peatland peatland margins

Introduction

Forests and peatlands represent two main ecosystem types in the boreal (taiga) region. Both of these ecosystems are important global carbon (C) pools. Peatlands cover an estimated 24% of the boreal region, and boreal peatlands represent 80% of the world’s peatlands (Wieder et al., 2006). Gorham (1991) estimates that northern peat deposits contain c. 450 Pg C (10¹⁵ g) and represent one-third of the world’s organic soil C pool. This C is stored in peat which has accumulated over long periods of time as rates of net primary production exceeded organic matter decomposition. Changes in the area and C balance of peatland ecosystems could have important effects on the global C cycle.

The main process for the formation of boreal peatlands is paludification (Charman, 2002), by which an upland forested habitat changes into a peatland habitat through lateral expansion of the peatland area. This change could be a result of continued local peat accumulation raising local and nearby water tables, but often external environmental factors trigger (or accelerate) the paludification process. On large scales, periods of climate change toward cooler and wetter conditions have been shown to correlate with periods of peatland development through paludification (Crawford et al., 2003; Kuhry and Turunen, 2006; Vitt, 2006). The lateral expansion rate is also significantly affected by local topography (Bauer et al., 2003; Charman, 2002; Korhola, 1994; Kuhry and Turunen, 2006; Lavoie et al., 2005). Forest fires, which reduce plant evapotranspiration, have been reported to favor paludification (Kuhry and Turunen, 2006). The development of peatland plant communities and the associated changes in local hydrology lead to the accumulation of peat deposits, mainly consisting of partly decomposed plant remains, which accumulate above the original humus-rich forest soils.

Carbon dynamics and storage in boreal peatlands have been widely investigated, and extensive databases exist for both fen and bog ecosystems (Blodau, 2002; Gorham, 1991; Gorham et al., 2012; Novak et al., 2008; Vitt et al., 2000; Weishampel et al., 2009). The C storage in taiga forest is also well documented (Bhatti et al., 2002; Kobak et al., 1998; Shvidenko et al., 2000; Stolbovoi, 2006). However, the transitional stage from forest to peatland ecosystem and its effect on C storage have not been studied in detail. Some studies regarding the effect of paludification on C dynamics have been carried out from a forestry perspective, mainly in Canada (Lavoie et al., 2005), focusing on the decrease of tree productivity more than on the total ecosystem C balance. The more recent dynamics of peatland development is especially poorly investigated as many paleoenvironmental studies of peatland ecosystems have been focused on the deepest peat profiles with old basal ages (Kuhry and Turunen, 2006). Hartshorn et al. (2003) showed that ecosystem properties were not continuous during the transition from forest to peatland, but rather that some of them such as vegetation growth were mostly related to the forest ecosystem, while other properties like soil respiration rate and water table level were more related to peatland features. Two studies in forested margins of Canadian peatlands suggest a continuous increase in total ecosystem C storage from upland margin to peatland (Bauer et al., 2009; Bhatti et al., 2006). The relationship between lateral peatland expansion and ecosystem C storage remains largely unexplored.

This study describes the rates of late-Holocene paludification and the effects of shifting forest to peatland on total ecosystem C storage, including both aboveground phytomass and belowground soil/peat C pools, at two sites in the taiga of Northeast European Russia. The investigated sites include the present transitions from forest to peatland ecosystems as well as the peatlands proper.

Study area

The two studied peatland complexes, Sludka (61°55′55″N, 50°13′19″E, 80 m a.p.s.l. (above present sea level)) and Laly (62°12′35″N, 50°35′06″E, 120 m a.p.s.l.), are situated in the middle taiga region of Northeast European Russia (Kozubov et al., 1999). They are located in the Komi Republic, 30 km from each other and from the city of Syktyvkar (Figure 1). These sites were chosen for their undisturbed character, accessibility, and the ongoing paludification process indicated by field observations of dead trees along the peatland margins.

Figure 1.

Location of the two investigated peatlands in Northeast European Russia, and positions of transects and sampling points within the study areas.

The local climate is characteristic for the boreal region. Based on the weather station situated in Syktyvkar (61°42′35″N, 50°49′48″E, 119 m a.p.s.l.), the mean annual temperature for the period 1971–2000 is +0.9°C, the mean July temperature is +17.0°C, and the mean January temperature is −15.2°C. The average total annual precipitation for the same period is 565 mm, of which slightly less than 50% falls as snow (http://www.tutiempo.net; Tutiempo Network, n.d.). The Sludka peatland is located in a former macro-meander of a river terrace landscape (Sidorchuk et al., 2001), whereas the Laly peatland site developed in an area with little topographic relief and no apparent paleolandforms.

The studied peatland area in Sludka is c. 220 ha, located in the southwestern part of a larger peatland complex (c. 1320 ha). The central and deeper parts of the studied peatland area are weakly minerotrophic to ombrotrophic and largely devoid of trees. In the marginal areas of the Sludka peatland, the open tree layer is mainly characterized by stunted spruce (Picea abies), birch (Betula pubescens Ehrh.), and pine (Pinus sylvestris L.). Vaccinium spp., Carex spp., Equisetum spp., and Sphagnum spp., among others, dominate the undergrowth. The peatland area in Laly is smaller (about 17 ha), and the peat depth is shallower than in Sludka. The Laly peatland is mostly an oligotrophic fen with scattered and stunted pine trees; dead tree trunks are found in the marginal part of the peatland. An area of swamp forest with tall pine trees is found to the north of the investigated Laly peatland. Surrounding forests in both Sludka and Laly are spruce or pine dominated.

Materials and methods

Field transects and profile sampling

Field sampling of forest soils and peat deposits took place during the summer of 2008. It was based on three transects from forest to peatland, perpendicular to the peatland edge, in each study area (Figure 1). Peat deposits were cored along the transects every 100 m in the peatland proper and every 50 m near the peatland/forest transition zone. By this approach, we increased the resolution in the assessment of the most recent paludification. The reported peatland edges were based on field observations identified by the transition from mixed spruce or pine forest with tall trees and closed canopies to more open vegetation characterized by the presence of stunted trees, tall dead tree trunks, and open water pools. For the purpose of this study, the transitional zone between forest and peatland is defined as a buffer area of 25 m on both sides of the peatland edge. The marginal area of the peatland is defined by the presence of sparse, living, stunted trees. Forest sites were subdivided into moist and dry habitats, based on floristic criteria. Species such as spruce, Equisetum spp., and mosses were considered indicators of a moist forest habitat. In contrast, the dominance of pine and lichens, as main undergrowth species, were characteristic of a dry forest habitat.

Peat samples were recovered using a Russian peat corer. In most of the peatland sampling points, only the basal peat at the peat–mineral contact was collected. The peat–mineral contact was often a sharp visual transition from organic to mineral soil horizon, and field observations were later confirmed in the laboratory by loss-on-ignition (LOI) measurements. Complete cores were collected in at least one site per transect for determination of total peat C storage (Figure 1). These complete peat cores were subsampled with a vertical resolution of 5 or 10 cm. Mineral soil samples beneath the peat–mineral contact were also collected in some points along the transects.

Forest soils were sampled to a reference depth of 1 m. Soil samples, with a vertical resolution of 5–10 cm (5-cm samples in organic layers), were collected with small fixed volume corers inserted horizontally into the wall of open soil pits. In some cases, deep mineral soil horizons were sampled using a steel pipe with known volume hammered into the subsoil from the bottom of the pits. Three replicate samples of the O-horizon were randomly collected within 10 m radius to capture some of the micro-topographic variability in the thickness of surface organic horizons in the forest soils (analytical results are later averaged and presented as a single value for each forest site).

Phytomass sampling

The phytomass sampling program in forest and peatland ecosystems covered only transects T1 and T2 in the Sludka area (Figure 1). The aboveground phytomass consisted of both understory, including the living upper 2 cm of lichen/moss layer and tree phytomass. Root biomass is included in the soil C storage. Trees that were smaller than 1 cm of diameter at 1.3 m height (dbh) were sampled as part of the understory. To account for local variability, the mean value from three sub-plots in both forest and peatland was used for each phytomass sampling point. Tree measurements were done using Bitterlich (1984) sampling and circular sample plots. On one of the three sub-plots, both circular and Bitterlich plots were measured, whereas on the two side plots, only Bitterlich sampling was conducted. On each circular sample plot, dbh was measured from every tree (tally trees, dbh > 1 cm) within a radius of 5.64 m, and tree height (h) and diameter at 6 m height (d₆) were measured from median diameter trees of each species. On the Bitterlich sample plots, species-specific basal area (m²/ha) as well as dbh and height of the median diameter tree of each species were measured. The understory phytomass was measured by removing all living vegetation from one 30 × 30 cm² on each of the sub-plots. Calculated phytomass averages for moist and dry forests and the marginal and central parts of the peatland from transect T1 and T2 at Sludka were applied to Sludka transect T3 and all three transects in the Laly study area. Based on field observations and measurements not reported here, we conclude that the forest and peatland vegetation types in the Sludka and Laly areas are very similar with regard to floristic composition and apparent tree sizes. Moreover, the relatively low within-class variability of phytomass storage in Sludka indicates a rather homogenous vegetation composition in different landscape and vegetation types, and we assume that phytomass storage in the two areas is similar.

Chemical analyses and total ecosystem C storage calculations

For the phytomass C assessment, the removed understory vegetation was stored in paper bags separated by functional group, and the samples were dried in 60°C for at least 24 h and weighed. The phytomass C is estimated as 50% of the vegetation dry weight. Tree aboveground phytomass was estimated by the following procedure: stem volume was first calculated for each sample tree on the circular sample plots using species-specific equations of Laasasenaho (1982) and those reviewed by Zianis et al. (2005). For trees smaller than 3 m of height, the stem volume was estimated with the so-called Smalian II formula (0.5 × dbh × h). Mixed models (Lappi et al., 2006) were then constructed to predict the volumes of the sample trees based on dbh and site fertility. These models were based on measurements from eight transects (thus including data from six additional transects not included in this study). The models were used to estimate the volumes of the tally trees on the circular sample plots. Once the volume of all trees on the circular sample plots was known, the species-specific growing stock per hectare (m³/ha) was computed. The growing stock on the Bitterlich sample plots could then be predicted from the dependency of growing stock on basal area and mean height on the circular sample plots. Tree aboveground phytomass was calculated from the growing stock with four different methods: two based on Russian (Alexeyev and Birdsey, 1998; Shvidenko et al., 1998) and two on Nordic (Kauppi et al., 1995; Lehtonen, 2005) studies. The phytomass values produced using these four different methods were averaged to yield the aboveground tree phytomass on each plot.

The peat C storage values were based on seven complete peat cores, including two cores in the open ombrotrophic part and two in the minerotrophic part (whereof one in the marginal area) of the Sludka peatland, and three in the oligotrophic fen of the Laly peatland (Figure 1). The mineral soil underlying the peat was described based on a total of 31 samples, which include buried organic-rich horizons (O- and A-horizons) in Sludka and buried sylvic peat layers in Laly. Samples of known field volume were analyzed for dry bulk density (BD, g/cm³) after drying at 100°C for 24 h. LOI at 550°C (for 2 h) was used to determine organic content (Dean, 1974). Elemental organic C and N content was determined in 13 full profiles (56 samples from four peat profiles and 63 samples from nine forest soil profiles; elemental analyzer CE-instruments, EA 1110). Based on these samples, two second-order polynomial regression models (Figure 2a) were used to predict the percentage of organic C for the peat/soil samples for which only LOI measurements are available (12 profiles representing 184 samples, including mineral subsoil beneath peat). Total peat C storage in sites where only peat depth was known (35 sites) was estimated using regression models based on known relationships between the depth of peat and peat C storage. As the two peatlands showed different trends due to their different age and landscape setting, a separate regression model was developed for each peatland (Figure 2b). The regressions represent the average cumulative C storages at 10 cm peat depth increments for Laly (24 samples from 3 profiles) and Sludka (42 samples from 4 profiles). By using cumulative peat C storage with increasing peat depth, we capture the more loose peat and relatively low C storage near the top of the peat sequences, which is characteristic for the shallow peat profiles in the marginal areas of the Sludka and Laly peatlands. C storages in the Sludka peatland profiles and in the Laly peatland profiles were estimated by second-order polynomial regressions (Figure 2b).

Figure 2.

Regression analyses used to calculate (a) percentage of carbon (C) from percentage of loss-on-ignition (LOI) and (b) total peat C storage from peat depth using cumulative values.

The soils at all forest sites were analyzed down to a standard depth of 1 m. Mineral soil underlying peat deposits has been shown to contain significant amounts of C (Moore and Turunen, 2004). The C storage in the top 30 cm beneath the surface organic horizons in the forest soils and beneath the peat–mineral contact in the peatlands was taken into account when comparing and reporting the total ecosystem C storage. Our samples provide a good representation of the 0–10 cm interval of mineral subsoil beneath peat (19 analyzed samples from 18 points, 15 from central areas, and 4 from marginal parts of the peatlands). Only two analyzed samples are available from the interval 10–30 cm, which were complemented with published data (three samples) from the deeper mineral subsoil in a northern taiga peatland in the Komi Republic (Oksanen et al., 2004). Because of limited data, we apply the same carbon density to the 10- to 30-cm depth interval of mineral subsoil across the peatland areas. In the Sludka peatland, the subsoil at some sites (n = 7) included a distinct paleosol surface organic horizon buried under the recent peat deposits. The samples of the seven concerned profiles were analyzed, and the C storage of this layer was included in the 30 cm of top mineral subsoil beneath the peat deposits. In the Laly peatland, a layer of sylvic peat occurred in many points under the recent peat deposits, recognizable by its darker color and more compact structure indicative of a higher degree of humification. It is considered as a buried organic peat layer, not related to the recent paludification process. Thus, the depth (and corresponding C storage) of this layer of sylvic peat is treated separately from the total 30 cm of mineral subsoil.

Radiocarbon dating and derived calculations

The initiation of peat accumulation at selected sampling points was determined by radiocarbon dating of 16 basal bulk peat samples (Table 1). Focusing on recent paludification, nine selected samples were from basal peat within 100 m from the edge of the peatlands. Another seven samples were from deep peat cores taken in the central part of the Sludka peatland. An additional three samples were from (buried) sylvic peat in Laly. The radiocarbon analyses were carried out at the Poznan dating laboratory (Poland). Obtained radiocarbon dates were calibrated to calendar years Before Present (cal. yr BP) using OxCal v3.8 for ‘older’ dates (Bronk Ramsey, 2001) and OxCal v4.1.7 for the three ‘modern’ dates (Bronk Ramsey, 2010), adding 58 years since 1950 (summer 2008 collection date). In many dated samples, the radiocarbon dating calibration gave several possible ranges within the 68% confidence interval. The selected bold dates in Table 1 provide the most plausible history of lateral peatland expansion and net peat C accumulation rates considering a logical trend in the transects from deeper and older points in the more central parts of the peatland areas toward shallower and more recent points at the peatland edges.

Table 1.

Radiocarbon dates, total peat carbon (C) storage, and peat C accumulation rates sorted according to dated intervals. From top to bottom: dates older than 2500 cal. yr BP without shading, late-Holocene dates between 250 and 2500 cal. yr BP in light grey shading, and dates younger than 250 cal. yr BP in dark grey shading. The last three dates correspond to buried sylvic peat in Laly peatland related to previous paludification events. Bold values are dates used in further analyses (see text).

Lab no.	Sample	Depth interval (cm)	Radiocarbon date	Uncertainty	Calendar years before 2008 1st^a	Calendar years before 2008 2nd	Calendar years before 2008 3rd	Peat C (kg C/m²)	Peat C accumulation (g C/m²/yr)
Poz-28395	SLU T1-6	204–206	8570	50	9583	9653		114.8	12.0
Poz-28394	SLU T1-10	183–185	7060	40	7973	7913	7858	99.7	12.5
Poz-28393	SLU T1-16	185–187	5110	40	5848	5958		102.1	17.5
Poz-28400	SLU T2-6	195–197	4170	35	4753	4873		105.9	22.3
Poz-28402	SLU T3-4	90–93	3380	35	3668	3733		23.4	6.4
Poz-28396	SLU T1-4	87–90	3325	35	3618	3553	3683	40.9	11.3
Poz-28399	SLU T2-4	89–91	1680	30	1628	1738		36.7	22.5
Poz-28397	SLU T1-3	60–62	1585	30	1498	1558		21.2	14.2
Poz-34617	LAL T4-5	102–104	980	35	983	898	868	12.5	12.7
Poz-28408	LAL T4-3	45–47	690	30	715	630		3.2	5.1
Poz-28407	LAL T5-5	83.5–85	540	30	598	670		9.1	15.2
Poz-28411	LAL T6-5	75–77	190	30	233	338	68	9.3	27.6
Poz-28413	LAL T6-2	46–48	225	30	348	223	63	3.3	52.7
Poz-28398	SLU T2-3	26–28	102.9%	0.32	53			7.2	136.6
Poz-28401	SLU T3-3	29–31	104.0%	0.32	52			8.3	159
Poz-28405	LAL T5-2	32–34	112.27%	0.39	14	51	<10	1.9	137.0
Poz-28412	LAL T6-5	154–156	7890	50	8743	8898	8983
Poz-35123	LAL T6-8	40–50	1485	30	1428
Poz-35124	LAL T6-8	90–92	8250	40	7648	7583

Median of the highest 68% probability interval in the calibration.

Estimates of the recent lateral expansion rate of the peatlands were based on basal peat ages and distances from the edge of the peatland. In Laly, we considered basal peat to be the base of the upper unhumified and loose, recently deposited peat, which was underlain at several places by older sylvic peat. In the cases where young basal dates were not available (SLU T1-2 and LAL T4-2), we assume that the peatland edge corresponds to the present (0 years). Net peat C accumulation rates were calculated for 16 profiles using basal peat ages and total peat C storage, where ‘zero’ time represents ‘zero’ accumulation. Slope was calculated as the peat depth difference (from local water table) between adjacent dated profiles divided by their distance. Profiles with modern basal peat ages were not considered because of large uncertainties in these dating results. Instead, we used the adjacent profiles with older basal peat depths and their distance to the peatland edge (considered 0 years).

Results

Peat age–depth assessment and lateral peatland expansion rates

The radiocarbon dating results show different time periods of initial paludification in the two peatlands (Table 1). Basal dates >2500 cal. yr BP are confined to the more central parts of the Sludka peatland. The oldest basal peat date in Sludka is around 9600 cal. yr BP (SLU T1-6), but we cannot assume that this corresponds to the earliest peatland development in the area. Basal dates between 2500 and 250 cal. yr BP are found in the marginal areas of the Sludka peatland and in central and marginal parts of the Laly peatland. The youngest dates <250 cal. yr BP are found in the peatland–forest transition zones of both peatlands.

The margins of the three Sludka transects have steep slopes with a rapid increase of peat depth to a maximum observed depth of 205 cm (SLU T1-6) within a few hundred meters from the peatland edge (Figure 3). A linear regression model, including the recent eight data points from the three transects, shows a consistent slow replacement of forest by peatland at c. 2.6 m/century (Figure 4a). A large part of the central area of the Sludka peatland has peat depths between 160 and 200 cm. Basal dates for this older part of the peatland overlying flatter terrain give a much higher mean lateral expansion rate of c. 30 m/century (not shown).

Figure 3.

Depth of organic soil horizons (separated into organic layer for aerated upland soils, peat, and sylvic peat) and basal radiocarbon dates along the six studied transects.

Figure 4.

Panels (a) and (b) present the regressed expansion rates in the marginal areas of Sludka peatland and Laly peatland, respectively. For Sludka peatland, only dates close to the present peatland edge (see Figure 3) were used in the regression giving the recent expansion rate. Panel (c) shows the regression between slope and expansion rate for the Sludka and Laly peatlands combined.

The peat depth in the Laly peatland is significantly lower than in Sludka, with 103 and 84 cm at the two deepest measured points (LAL T4-5 and T5-5, respectively) The Laly peatland also has a much more recent development. The deepest recent peat has the oldest age, at about 1000 cal. yr BP (Figure 3). Slopes are gentler than in the Sludka transects. The average lateral peatland expansion rate at Laly based on a linear regression model using all seven data points in the three transects was more rapid, c. 35 m/century (Figure 4b). Direct interpolations between adjacent dated profiles give a wide range of expansion rates, from 27 to 73 m/century, depending on the location within the peatland. However, the particularly large uncertainties associated with the young Laly dates (see Table 1) should be kept in mind, and results should not be over-interpreted.

An exponential regression model based on 12 data points shows a strong correlation between slope and expansion rate (Figure 4c). The flatter terrain in the Laly peatland and the older sections of the Sludka peatland promoted fast lateral expansion rates, whereas the steeper slopes at the present-day margins of the Sludka peatland show much lower rates.

Total ecosystem C storage

The vertical stratigraphy of soil profiles illustrates the main differences in BD, LOI, and C/N among five complete profiles considered representative of the different ecosystem types investigated in this study (Figure 5). The upland sites LAL T4-1 (moist forest) and SLU T2-1 (dry forest) show variable surface organic horizon thickness (31 and 3 cm, respectively), with a rapid decline in LOI values in the mineral horizons. The peatland sites (LAL T6-5, SLU T1-4, and SLU T1-16) have much deeper peat deposits (down to 200 cm). C/N ratios generally decline from the top downwards into the profile, suggesting increased decomposition of the organic material (Kuhry and Vitt, 2006). At Laly, the recent peat layer is often underlain by sylvic peat. This layer has a highly fluctuating thickness, particularly in the more central parts of the basin, and was absent at some places (Figure 3, Laly transects). The shift from recent, unhumified peat to older, buried sylvic peat below 77 cm depth in LAL T6-5 (see Figure 5) is reflected in a much higher BD and lower C/N ratio indicative of a higher degree of compaction and decomposition. An analog to this layer was found in a 90-cm-thick sylvic peat sequence in the pine swamp forest site at LAL T6-8 with a similar old basal date (Table 1).

Figure 5.

Physico-chemical variables analyzed in soil/peat profiles representing different ecosystem types: LAL T4-1 (moist forest site), SLU T2-1 (dry forest site), LAL T6-5 (oligotrophic fen site with buried sylvic peat), SLU T1-4 (marginal, sparsely forested peatland site), and SLU T1-16 (central, open ombrotrophic peatland site). Age–depth models for the peatland profiles are derived from the available radiocarbon dates (Table 1).

The C allocation (storage in phytomass and in different soil layers) at the different sites along sampled transects illustrates changing total C storage and its partitioning as the ecosystem shifts from upland forest into peatland (Figure 6). In the upland sites, the forest aboveground phytomass represents about 25% of the total ecosystem C with a mean value of 4.2 ± 1.2 kg C/m² (n = 9), with 4.6 ± 0.04 kg C/m² (n = 4) in dry forest types and 3.9 ± 1.7 kg C/m² (n = 5) for moist forest types (Table 2). The forest soil C storage for the reference 1 m depth ranges from 14.4 to 40.3 kg C/m² (supplementary Table S1, available online). The moist forest sites (n = 5) store more C, with 27.4 ± 8.8 kg C/m², than the drier sites (n = 4), with 16.4 ± 1.9 kg C/m² (Table 2). The surface organic horizon is usually deeper and the mineral horizon richer in C in moist than in dry forest sites (Table 2 and supplementary Table S1, available online). In fact, the depth of the surface organic horizon shows a significant correlation to total soil C storage (Pearson product–moment correlation, R² = 0.82, p < 0.01).

Figure 6.

Total ecosystem carbon storage distribution along the six studied transects. The ‘0’ distance represents the edge of the peatland.

Table 2.

Summary table illustrating total ecosystem carbon (C) storage of the different forest and peatland types in the investigated study sites.

Ecosystem type (number of sites)	Phytomass C storage (kg C/m²)	OL/peat C storage (kg C/m²)	Mineral (30 cm) C storage (kg C/m²)^a	Total ecosystem C storage (kg C/m²)^a	Forest soil (1 m) C storage (kg C/m²)
Moist forest (n = 5)	3.9 ± 1.7^b	9.8 ± 7.7	9.5 ± 5.6	23.2 ± 13.4	27.4 ± 8.8
Dry forest (n = 4)	4.6 ± 0.04	2.6 ± 2.1	5.4 ± 1.4	12.5 ± 1.4	16.4 ± 1.9
All forest (n = 9)	4.2 ± 1.2	6.6 ± 6.7	7.6 ± 4.6	18.5 ± 11.1	22.5 ± 8.6
Sludka margin (n = 6)	0.8 ± 1.0	24.8 ± 21.1	6.8 ± 2.1	32.5 ± 20.8
Laly margin (n = 8)	0.8 ± 0.5	4.4 ± 2.3	6.2^c	11.3 ± 2.2
All margin (n = 14)	0.9 ± 0.5	15.5 ± 17.0	6.6 ± 1.3	22.1 ± 17.3
Sludka open peatland (n = 21)	0.2 ± 0.1	97.4 ± 25.0	7.6 ± 1.1	105.3 ± 24.9
Laly open peatland (n = 7)	0.3 ± 0.3^d	8.3 ± 2.5	7.1^e	15.7 ± 2.3
All open peatland (n = 28)	0.2 ± 0.2	81.6 ± 41.0	7.5 ± 1.1	82.9 ± 48.9

OL: organic layer.

Including the 30 cm of mineral soil below the top organic soil or peat layer, but excluding the buried ‘older’ sylvic peat at some Laly sites.

Typical moist forest phytomass is 4.7 kg C/m2, but moist forest type includes one site at a peatland edge with typical moist forest soil characteristics but already reduced phytomass.

Same mean value applied to all Laly marginal peatland sites.

Typical open peatland phytomass is 0.2 kg C/m2, but Laly open peatland type includes one site with increased phytomass.

Same mean value applied to all Laly open peatland sites.

The total C storage in peatland sites varies from 13.5 kg C/m² in the marginal parts to >100 kg C/m² in the central parts of the Sludka peatland, and from 8.3 to 19.8 kg C/m² (80.9 kg C/m² including the buried sylvic peat) in the Laly peatland (supplementary Table S1, available online). The phytomass C storage is on average 0.9 ± 0.5 kg C/m² (n = 14) in the marginal parts, with sparse and stunted trees, and 0.2 ± 0.2 kg C/m² (n = 28) in the central, more open parts (Table 2). Phytomass C storage in these different parts of the peatland represents 10% and 0.1% of the total ecosystem C storage, respectively. The peat layer is the major reservoir of C in the Sludka peatland. The buried organic soil horizon accounts for only 1.6 to 6.3 kg C/m² (supplementary Table S1, available online). However, in the Laly peatland, the buried sylvic peat layer and/or the mineral horizons are the main reservoirs of C. The sylvic peat has a thickness of up to 80 cm and contains up to 64.3 kg C/m² (supplementary Table S1, available online). Moreover, the recent and shallow peat in the marginal parts of the Sludka peatland and at all points of the Laly peatland has a very low BD with a high water content (profiles SLU T1-4 and LAL T6-5 in Figure 5), which leads to a low C mass. The C storage in the 30 cm of mineral subsoil below the peat–mineral contact was estimated at 6.2 kg C/m² in the marginal parts and 7.1 kg C/m² in the central parts of the peatlands (up to 11.3 kg C/m² where buried surface organic horizons are present; supplementary Table S1, available online).

Overall, the peatland ecosystem stores more C than the forest ecosystem. The decrease of phytomass C storage is more than compensated by peat C accumulation. However, the increase of the total ecosystem C storage from forest to peatland is not linear. All six transects have a decrease of total ecosystem C storage in the transition zone from forest to peatland (mean loss = −7.5 ± 7.4 kg C/m²; n = 6; Figure 6). The losses are caused both by reductions in the aboveground phytomass and belowground carbon pools, even though in four cases, the belowground loss was <1 kg C/m². The estimated mean loss in Sludka (n = 3) was −5.0 ± 4.9 kg C/m²; the mean loss in Laly (n = 3) was −9.9 ± 9.9 kg C/m². The buried sylvic peat layers under the recent Laly peat are not taken into account as they are not related to the recent paludification. Only one forest-to-peatland transition zone in Laly (at LAL T5-2) was underlain by a buried thin sylvic peat layer (Figure 3), which, if included, would still have shown a (somewhat reduced) C loss following recent paludification. Estimated C losses were much greater in the transition from moist forest to peatland margin (mean = −11.9 ± 6.9 kg C/m²; n = 3) than from dry forest to peatland margin (mean = −3.0 ± 4.5 kg C/m²; n = 3).

Peat C accumulation rates and C loss recovery times after initial paludification

Basal dates are available for a total of 16 profiles, 6 of which have C storage estimates derived from LOI/C measurements, and the 10 others are obtained from the regressions against depth depicted in Figure 2b. These 16 profiles can be subdivided into sites with recent basal dates (<250 cal. yr BP; n = 4), late-Holocene basal dates (>250 to <2500 cal. yr BP; n = 6), and early–middle Holocene basal dates (>2500 cal. yr BP; n = 6). Table 1 shows peat C storage and net peat C accumulation rates for each of these profiles grouped according to age interval. The mean net peat accumulation rate for all profiles within each of these age intervals (from young to old) is 120.0, 15.7, and 13.7 g C/m²/yr, respectively. Net peat accumulation rates are highest for the youngest interval because the original high net primary productivity is reduced only to a limited extent by peat decomposition. On longer time scales, the cumulative aerobic (in the acrotelm) and anaerobic (in the catotelm) decay significantly reduces net C accumulation rates in the peat deposits (Clymo et al., 1998).

The recovery times required to compensate for initial ecosystem C losses following paludification are estimations based on the measured C loss (−7.5 kg C/m², range = −0.4 to −21.2 kg C/m²) and C accumulation rates in the investigated peat profiles (Table 1) or empirical observations from each transect. We obtain a period of c. 60 years to compensate for the initial losses based on the average net peat accumulation rate calculated for the youngest interval (<250 cal. yr BP) and c. 440 years when considering the average rate for the interval 250–2500 cal. yr BP. When recovery times are reported separately for the different transects, two transitions (SLU T2 and SLU T3) recovered the estimated C losses within 55 years, one transition (LAL T5) between 14 and 598 years, one transition (LAL T6) between 63 and 338 years, one transition (LAL T4) after 983 years, and one transition (SLU T1) between 1498 and 3618 years. The three latter transects correspond to transitions from moist forest with highest initial C storage. There was a limited temporal resolution in Sludka because of the slow lateral peatland expansion rate in recent times. Results from the Laly peatland suggest that most of the C losses might occur in the earliest stages of paludification (Figure 6). Overall, C loss recovery times range from less than 55 years to more than 1500 years. We ascribe this to natural spatial variability mostly related to initial forest C ecosystem storage, but results should be carefully interpreted because of methodological uncertainties associated with C storage estimates and radiocarbon dating.

Discussion

Historical development of the studied peatlands

This study gives new insight into the Holocene development, paludification rates, and C dynamics of two investigated peatlands in Northeast European Russia. Initial peat formation in Sludka began in the flat part of the basin, more than 9600 years ago. The peatland expanded quickly (c. 30 m/century) until it reached the steeper slopes of the river terrace paleolandscape where the lateral expansion rate slowed down considerably (to c. 3 m/century).

The recent paludification in the Laly peatland took place over relatively flat terrain, and had a rapid lateral expansion rate since its inception about 1000 cal. yr BP (c. 35 m/century). The basal peat depths suggest that there were two separate points of peatland initiation which linked up around point LAL T4-7 (Figure 3). The estimated expansion rate of the Laly peatland was approximately eight times higher than the recent expansion of the Sludka peatland. The topography appears to be a main driving factor behind changes in lateral expansion rates corroborating the results of Korhola (1994) who showed the same relation between slope and paludification rate in Finnish raised bogs.

The sylvic peat layer underlying much of the more recent peat at Laly shows that the central part of the present peatland area had accumulated thick organic layers prior to the recent paludification. It could originate from a historic cycle of peatland paludification–forest recolonization, as described by Crawford et al. (2003). The presence of the buried peat likely leads to an increased local water holding capacity favoring (re)paludification. We can note that this sylvic peat layer underlying the central expanses of the Laly peatland would not affect our assessment of recent (<1000 years) C storage dynamics following paludification if it would have been taken into account.

Carbon storage in the boreal environment

This study focuses on the two main components of total ecosystem C storage in forest and peatland ecosystems: the aboveground phytomass and the belowground soil/peat (including roots) C pools. Our phytomass C storage estimates range from about 4.6 kg C/m² in forest sites to less than 1 kg C/m² in open peatlands. Our forest phytomass values are comparable with the reported values of 3.6 kg C/m² for the whole boreal forest region (Alexeyev and Birdsey, 1998), between 4 and 8 kg C/m² for the middle taiga (Bobkova and Galenko, 2001) and 4.4 kg C/m² in the forested lands of European Northern Russia (Shvidenko et al., 1998). Some previously reported values for wetlands include 1.6 kg C/m² for Russian peatlands (Shvidenko et al., 2000) and 0.2 kg C/m² in a peatland site of the southern boreal forest in North America (Weishampel et al., 2009).

Reported mean C storage in peat deposits ranges from 63.6 to 241.6 kg C/m² in the northern taiga of East European Russia (Hugelius and Kuhry, 2009) and from 74.1 to 178.2 kg C/m² in Canadian boreal peatlands (Bhatti et al., 2002). The peatland C storage values in this study are somewhat lower (12.6–132.3 kg C/m²), but this can partly be explained by the sampling methodology which focused on the shallower, marginal areas of the peatlands. In this study, a 30-cm deep mineral subsoil layer beneath the peat–mineral contact was included in the total C storage of peatland sites. Our mean value of 7.1 kg C/m² for 30 cm depth is likely somewhat higher than 7.8 kg C/m² for 70 cm depth reported by Turunen and Moore (2003) and Moore and Turunen (2004), but soils with a finer texture such as those underlying the Sludka and Laly peatlands (silty and clayey loam) typically store more C compared with, for instance, sandy soils (Jenny, 1980). The estimated forest soil C storage (down to 1 m depth) in this study (between 14.4 and 40.3 kg C/m²) is higher in comparison with other reported values such as 10.2–13.2 kg C/m² in Canadian boreal forests (Bhatti et al., 2002), 11.4 kg C/m² on average in Russian forest soils (Stolbovoi, 2006), and between 11.1 and 19.0 kg C/m² in the global boreal forest biome (Post et al., 1982). However, Hugelius and Kuhry (2009) report higher values for moist (herbaceous) spruce forests in northern Komi (on average 20.3 kg C/m²). The higher values (on average 27.4 kg C/m²) for our moist forest sites could result from the proximity to the peatlands indicating early stages of paludification.

Carbon storage in the transitional zone of paludification

Results from the Sludka and Laly peatlands indicate that there was an initial break in the general trend of increased C storage following paludification from forest to peatland. Based on observational (n = 1) and regression-based (n = 5) data, all six studied transitional zones are characterized by C losses (Figure 4) with an average loss of −7.5 ± 7.4 kg C/m². Due to limited observational data, the analyses of these marginal transition zones are vulnerable to systematic, nonlinear prediction errors (such as underestimation of transitional zone C stocks in the regression models of Figure 2b). However, the measured data points available from transitional zones (n = 8) show similar C stocks in relation to peat depth than what is predicted by the regression model, indicating that systematic under-prediction of transitional zone C stocks is unlikely (Figure 2b).

The highest losses are predicted in the transition from moist forest to peatland, which can be ascribed to the fact that moist forest ecosystems stored initially more C (mean = 23.2 ± 13.4 kg C/m²; n = 5) compared with dry ones (mean = 12.5 ± 1.4 kg C/m²; n = 4). Here, we assume that the total ecosystem C storage of the recently paludified sites was similar to that of the currently adjacent forest sites. Recovery times for the estimated total ecosystem C losses after initial paludification are estimated to be in the order of decades to centuries, or more (<55 to >1500 years), with longer times required if the original forest ecosystem C pool was higher.

Our estimates show that in this transition zone phytomass values decreased because of rising water tables resulting in the drowning of the forest. In the marginal zone of peatlands, flowing water may transport organic material away as dissolved organic C and added nutrients could accelerate decomposition. Earlier studies have identified many factors influencing the mineralization rates in peatlands. The fluctuations of the water table level, the species composition of the plant litter, and/or the concentrations of nutrients (such as nitrogen and phosphorus) are all factors that affect the decomposition rate (Blodau, 2002). For example, fluctuations of the water table level can increase C mineralization rate by a factor of 1.5–3 (Aerts and Ludwig, 1997). Leaching can also transfer part of the surface organic/peat matter to the mineral part of the soil profile (Moore and Turunen, 2004; Turunen and Moore, 2003). This latter process does not affect the total C storage, whereas lateral flow leads to a loss of C from the ecosystem. The marginal zones of the investigated Sludka and Laly peatlands had on average slightly less C in the upper 10 cm of mineral subsoil (beneath the peat deposit) than central parts, which suggests that C was not leached more extensively into the mineral subsoil and cannot account for the observed losses in total ecosystem storage in the transitional zones.

Few other studies have focused on the transitional stage between boreal forest and peatland ecosystem. Research in the margins of two peatlands in Saskatchewan (Canada) indicates a continuous increase in ecosystem C storage from upland margin to peatland (Bauer et al., 2009; Bhatti et al., 2006). However, the upland margin forest sites in their studies would be considered as marginal peatland in our definition considering the deep organic layer (>45 cm) and shallow water table. Their transects did not extend to the edges of the investigated peatlands and the adjacent upland spruce and jack pine forest sites. The observed increase in total ecosystem C storage from the peatland margins to more central parts of the peatlands in these two Canadian sites is, therefore, consistent with the results from this study. However, differences in historical development of peatlands as well as local variability should not be neglected. As a first-order difference, the study sites in Saskatchewan are continental ‘treed’ peatlands compared with the ‘open’ peatlands typical of more oceanic climates such as prevailing in our study sites in Northeast European Russia. More research in the marginal parts of peatlands is needed to obtain a wider perspective on the effects of initial paludification on total ecosystem C storage.

Paludification and global change

Wetter and colder climates have been shown to correlate with more rapid peatland paludification over the Holocene period (Crawford et al., 2003; Kuhry and Turunen, 2006; Vitt, 2006). Climate can thus be seen as a large-scale driver of peatland dynamics in the boreal region. Climate affects key factors in peatland formation such as hydrologic conditions linked to precipitation and/or temperature changes (Charman, 2002). For the coming century, increases in annual temperature and precipitation in the boreal region are projected, both during summer and winter seasons (Christensen et al., 2007). There is potential for a larger extent of boreal peatlands because of increased excess soil water under a wetter climate leading to accelerated rates of paludification. However, while higher precipitation may contribute to more waterlogged soils, warmer temperatures tend to increase evapotranspiration, reducing soil moisture. Periods of summer drought are also expected to increase resulting in more frequent boreal forest fires (Stocks et al., 1998), which in turn reduce vegetation cover and subsequent evapotranspiration. Because of difficulties in predicting the relative magnitude of these possible changes and because of regional variability, there are large uncertainties regarding the response of boreal peatland ecosystems to future climatic changes. Topography has previously been shown to be a key factor for local expansion rate (Korhola, 1994), and this is confirmed by our results. The Laly peatland, established on flatter terrain, shows an expansion rate eight times higher than that of the Sludka peatland underlain by steeper slopes. This difference emphasizes the local nature of the paludification process, which contributes to the difficulty of upscaling local studies to biome scales. There are, thus, large uncertainties about the future rates of paludification in the boreal biome. Our study suggests that if there would be a rapid pulse of paludification, the short-term total ecosystem C loss would lead to a positive feedback to global warming before this initial loss is compensated by long-term peat accumulation.

Conclusion

This study provides estimates for the rates and effects on total ecosystem C storage of late-Holocene paludification in two peatlands of the middle taiga in Northeast European Russia. The edge of the Sludka peatland situated in steeper topography expanded more slowly, at about 3 m/century, than the Laly peatland in flatter topography, which expanded at about 35 m/century. The transitional stage from forest to peatland leads to a loss of C from the ecosystem, before renewed peat accumulation compensates for the initial phytomass and soil C losses. The estimated period for this recovery is estimated at decades to centuries (or more), which is a relevant time span in the context of anthropogenically induced climate change and the C balance of the boreal region. If future climate change (increased precipitation and/or fire frequencies) would result in a pulse of paludification, this could contribute on the short term to a diminished total C storage in the boreal region because of transitory C losses during initial peatland expansion.

Footnotes

Acknowledgements

We would like to thank Suzanne Baer (University of Greifswald, Germany), Sanna Susiluoto, Tarmo Virtanen, and Malin Ek (University of Helsinki) for field assistance and an anonymous reviewer for helpful comments.

Funding

This study was funded through the EU 6th Framework CARBO-North project (contract 036993) and a grant of the Swedish Research Council to P. Kuhry.

References

Aerts

Ludwig

(1997) Water-table changes and nutritional status affect trace gas emissions from laboratory columns of peatland soils. Soil Biology & Biochemistry 29: 1691–1698.

Alexeyev

Birdsey

(1998) Carbon Storage in Forests and Peatlands of Russia. General technical report NE-244, April. Delaware, OH: USDA Forest Service, p. 137.

Bauer

Gignac

Vitt

(2003) Development of a peatland complex in boreal western Canada: Lateral site expansion and local variability in vegetation succession and long-term peat accumulation. Canadian Journal of Botany 81: 833–847.

Bauer

Bhatti

Swanston

. (2009) Organic matter accumulation and community change at the peatland-upland interface: Influences from ¹⁴C and ¹²⁰Pb dated profiles. Ecosystems 12: 636–653.

Bhatti

Apps

Tarnocai

(2002) Estimates of soil organic carbon stocks in central Canada using three different approaches. Canadian Journal of Forest Research 32: 805–812.

Bhatti

Errington

Bauer

. (2006) Carbon stock trends along forested peatland margins in central Saskatchewan. Canadian Journal of Soil Science 86: 321–333.

Bitterlich

(1984) The Relascope Idea: Relative Measurements in Forestry. Slough: Commonwealth Agricultural Bureau, p. 242.

Blodau

(2002) Carbon cycling in peatlands – A review of processes and controls. Environmental Reviews 10: 111–134.

Bobkova

Galenko

(2001) Process of bioproduction in forest ecosystems of the north. Ural Division, Komi Research Center, Institute of Biology, Russian Academy of Sciences, Nauka, St Petersburg, pp. 40–54 (in Russian).

10.

Bronk Ramsey

(2001) Development of the radiocarbon program OxCal. Radiocarbon 43(2A): 355–363.

11.

Bronk Ramsey

(2010) OxCal v4.1.7. Radiocarbon Calibration Software. Oxford: Research Lab for Archaeology. Available at: http://c14.arch.ox.ac.uk.

12.

Charman

(2002) Peatlands and Environmental Change. Chichester: John Wiley & Sons Ltd.

13.

Christensen

Hewitson

Busuioc

. (2007) Regional climate projections. In: Solomon

Qin

Manning

. (eds) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge; New York: Cambridge University Press, pp. 848–940.

14.

Clymo

Turunen

Tolonen

(1998) Carbon accumulation in peatland. Oikos 81: 368–388.

15.

Crawford

RMM

Jeffree

Rees

(2003) Paludification and forest retreat in Northern oceanic environment. Annals of Botany 91: 213–226.

16.

Dean

(1974) Determination of carbonate and organic matter in calcareous sediments and sedimentary rocks by loss on ignition: Comparison with other methods. Journal of Sedimentary Petrology 44: 242–248.

17.

Gorham

(1991) Northern peatlands: Role in the carbon cycle and probable responses to climatic warming. Ecological Applications 1: 182–195.

18.

Gorham

Lehman

Dyke

. (2012) Long-term carbon sequestration in North American peatlands. Quaternary Science Reviews 58: 77–82.

19.

Hartshorn

Southard

Bledsoe

(2003) Structure and function of peatland-forest ecotones in Southeastern Alaska. Soil Science Society of America Journal 67: 1572–1581.

20.

Hugelius

Kuhry

(2009) Landscape partitioning and environmental gradient analyses of soil organic carbon in a permafrost environment. Global Biogeochemical Cycles 23: GB3006. DOI: 10.1029/2008GB003419.

21.

Kauppi

Tomppo

Ferm

(1995) C and N storage in living trees within Finland since 1950s. Plant and Soil 168–169: 633–638.

22.

Kobak

Kondrasheva

Turchinovich

(1998) Changes in carbon pools of peatland and forests in northwestern Russia during the Holocene. Global and Planetary Change 16–17: 75–84.

23.

Korhola

(1994) Radiocarbon evidence for rates of lateral expansion in raised mires in southern Finland. Quaternary Research 42: 299–307.

24.

Kozubov

Taskaev

Degteva

. (1999) Lesa Respubliki Komi. Moscow: Dizain, 332 pp.

25.

Kuhry

Turunen

(2006) The postglacial development of boreal and subarctic peatlands. In: Wieder

Vitt

(eds) Boreal Peatland Ecosystems: Ecological Studies 188. Berlin, Heidelberg: Springer-Verlag, pp. 25–45.

26.

Jenny

(1980) The Soil Resource: Origin and Behaviour (Ecological Studies). vol. 37. New York: Springer-Verlag, 377 pp.

27.

Laasasenaho

(1982) Taper curve and volume functions for pine, spruce and birch. Communicationes Instituti Forestalis Fenniae 108: 1–74.

28.

Lappi

Mehtätalo

Korhonen

(2006) Generalizing sample tree information. In: Kangas

Maltamo

(ed.) Forest Inventory Methodology and Applications. Managing Forest Ecosystems. Vol 10. Dordrecht: Springer, pp. 85–106.

29.

Lavoie

Paré

Fenton

. (2005) Paludification and management of forested peatlands in Canada: A literature review. Environmental Reviews 13: 21–50.

30.

Lehtonen

(2005) Carbon stock and flows if forest ecosystems based on forest inventory data. PhD Thesis, University of Helsinki, Helsinki.

31.

Moore

Turunen

(2004) Carbon accumulation and storage in mineral subsoil beneath peat. Soil Science Society of America Journal 68: 690–696.

32.

Novak

Brizova

Adamova

. (2008) Accumulation of organic carbon over the past 150 years in five freshwater peatlands in western and central Europe. Science of the Total Environment 390: 425–436.

33.

Oksanen

Kuhry

Alekseeva

(2004) Holocene development and permafrost history of the Usinsk Mire, Northeast European Russian. Geographie Physique et Quaternaire 57: 169–187.

34.

Post

Emanuel

Zinke

. (1982) Soil carbon pools and world life zones. Nature (London) 208: 156–159.

35.

Shvidenko

Nilsson

Stolbovoi

. (1998) Background Information for Carbon Analysis of the Russian Forest Sector. Laxenburg: International Institute for Applied Systems Analysis.

36.

Shvidenko

Nilsson

Stolbovoi

. (2000) Aggregated estimation of the basic parameters of biological production and the carbon budget of Russian terrestrial ecosystems: 1. Stocks of plants organic mass. Russian Journal of Ecology 31: 371–378.

37.

Sidorchuk

Borisova

Panin

(2001) Fluvial response to the late Valdai/Holocene environmental change on the East European Plain. Global and Planetary Change 28: 303–318.

38.

Stocks

Fosberg

Lynham

. (1998) Climate change and forest fire potential in Russian and Canadian boreal forests. Climatic Change 38: 1–13.

39.

Stolbovoi

(2006) Soil carbon in the forests of Russia. Mitigation and Adaptation Strategies for Global Change 11: 203–222.

40.

Turunen

Moore

(2003) Controls on carbon accumulation and storage in the mineral subsoil beneath peat in Lakkasuo mire, central Finland. European Journal of Soil Science 54: 279–286.

41.

Tutiempo Network (n.d.) S.L. Climate data. Available at: http://www.tutiempo.net (accessed 2009).

42.

Vitt

(2006) Functional characteristics and indicators of boreal peatlands. In: Wieder

Vitt

(eds) Boreal Peatland Ecosystems: Ecological Studies 188. Berlin, Heidelberg: Springer-Verlag, pp. 9–24.

43.

Vitt

Halsey

Bauer

. (2000) Spatial and temporal trends in carbon storage of peatlands of continental western Canada through the Holocene. Canadian Journal of Earth Sciences 37: 683–693.

44.

Weishampel

Kolka

King

(2009) Carbon pools and productivity in a 1-km² heterogeneous forest and peatland mosaic in Minnesota, USA. Forest Ecology and Management 257: 747–754.

45.

Wieder

Vitt

Benscoter

(2006) Peatlands and the boreal forest. In: Wieder

Vitt

(eds) Boreal Peatland Ecosystems: Ecological Studies 188. Berlin, Heidelberg: Springer-Verlag, pp. 1–8.

46.

Zianis

Muukkonen

Mäkipää

. (2005) Biomass and stem volume equations for tree species in Europe. Silva Fennica Monographs 4: 63 pp.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB