Stable isotopes in Sphagnum fuscum peat as late-Holocene climate proxies in northeastern European Russia

Abstract

The environment of the northern taiga to tundra transition is highly sensitive to climate fluctuations. In this study from northeastern European Russia, stable carbon and oxygen isotope ratios (δ¹³C, δ¹⁸O) in α-cellulose of Sphagnum fuscum stems subsampled from hummocks and peat plateau profiles have been used as climate proxies. The entire isotope time series, dated by lead (²¹⁰Pb), caesium (¹³⁷Cs) and AMS-radiocarbon (¹⁴C) dating, spans the past 2500 years. Plant macrofossil analyses were used as an aid in single species selection, but are also helpful in identifying past surface moisture conditions. The most significant relationships were found between the recent δ¹³C record and summer (July–August) temperatures (R² = 0.58, p < 0.01), and the recent δ¹⁸O record and winter (October–May) precipitation anomalies in the tundra region (R² = 0.36, p < 0.01). The study demonstrates that stable isotopes preserved in northern peat deposits are useful indicators for summer temperature and winter precipitation at decadal to millennial timescales.

Keywords

climate northern taiga stable isotopes tundra

Introduction

The boreal and subarctic regions of the Northern Hemisphere are characterized by large areas of peatlands, with deposits that provide records for natural climate variability since Postglacial times (Kuhry and Turunen, 2006). Since the mechanisms behind the climatic variability act at different timescales and magnitudes, and the responses are spatially not synchronous (Gray et al., 2010; Wanner et al., 2008), temporal and spatial coverages of reliable reconstructions still need to be improved. The Arctic is a key area, where the past century warming has been amplified (e.g. Miller et al., 2010). In northeastern European Russia, Holocene climate variability has mainly been studied by semi-quantitative methods, such as pollen and plant macrofossil analysis (Kaakinen and Eronen, 2000; Kultti et al., 2004; MacDonald et al., 2000; Oksanen et al., 2001, 2003; Väliranta et al., 2003). Quantitative reconstructions from the region were made by Andreev and Klimanov (2000) and Salonen et al. (2011) using transfer functions of pollen records, and from the eastern side of northern Urals by Briffa (1995) and Hantemirov and Shiyatov (2002) using tree-ring width as climate proxies. Because these studies cover different time intervals during the Holocene and differ in methodology and time resolution, more studies are needed from the region.

Stable carbon and oxygen isotope ratios (δ¹³C, δ¹⁸O) of moss remains preserved in peat deposits have become valuable tools for studies of fluctuations in regional air temperature and precipitation (Daley et al., 2010; Hong et al., 2010; Kaislahti Tillman et al., 2010a; Loisel et al., 2010). Discrimination (Δ) of ¹³C in relation to ¹²C by plants can be calculated by the equation:

Δ = a + (b - a) (p_{i} / p_{a})

(1)

where a denotes diffusion and b carboxylation; p_i and p_a express concentrations of intracellular CO₂ (in leaf) and CO₂ in ambient air, respectively (Farquhar et al., 1989). Variation in isotope discrimination due to diffusion fractionation (parameter a in previous equation) is limited in mosses compared with vascular plants because of their lack of stomata (Farquhar et al., 1989). The grade of filling of hyaline cells (water reservoir cells) is regarded as the main mechanism controlling diffusion of CO₂ into photosynthetic chlorophyll cells (Rice, 2000; Williams and Flanagan, 1996), a fact that may be relevant for species-specific sensibilities to climate parameters (Loisel et al., 2009). As Sphagnum species belonging to the section Acutifolia (e.g. Sphagnum fuscum) have their chlorophyll cells situated close to the leaf surface, they are plausibly not controlled by hyaline cells in the same degree as species with embedded chlorophyll cells. Furthermore, S. fuscum occupy mainly hummocks and are less sensitive to water-table fluctuations than species growing in wet hollows (Loisel et al., 2010; Rydin, 1985). Temperature variation has a direct impact in kinetic reactions, such as photosynthesis (Clark and Fritz, 1997). As increased rate of photosynthesis leads to depletion of intracellular concentration of CO₂ and thereby decreased carboxylation discrimination of ¹³C, higher air temperatures can indirectly lead to increasing δ¹³C values in trees (McCarroll and Loader, 2004). However, different C3 plants have shown large variations in δ¹³C responses to temperature in laboratory and field studies (Ménot and Burns, 2001). A positive relationship was found in an isotope study of S. fuscum peat from subarctic west central Canada, where temperature may be the main limiting control on growth for that species (Kaislahti Tillman et al., 2010b). Negative correlation of δ¹³C and temperature without sensitivity to relative humidity have been reported from altitudinal transects (e.g. Skrzypek et al., 2007), whereas other studies based on the δ¹³C variation in different peat-forming species have shown positive correlation to water-table depths (e.g. Lamentowicz et al., 2008; Loisel et al., 2009, 2010). Stable oxygen isotopes display the pathway of meteoric water from the source to the plant tissues; the source values of δ¹⁸O are depleted during the air-mass transport of vapor by latitudinal, altitudinal and continental effects on air temperature (Rozanski et al., 1993). ¹⁸O isotopes are enriched in moss plants during photosynthesis and especially in hummock species by evapotranspiration (Brenninkmeijer et al., 1982; Zanassi and Mora, 2005). Indirectly, temperature variation affects precipitation patterns, relative humidity and evaporative enrichment of ¹⁸O that can be limited by high temperatures and/or low precipitation amounts, as evapotranspiration is inhibited when Sphagnum plants dry out (Kim and Verma, 1996). In mosses, relationships between δ¹⁸O values and both temperature and precipitation anomalies have been found (e.g. Daley et al., 2010; Kaislahti Tillman et al., 2010b, Ménot-Combes et al., 2002).

Because of differences found between species and even plant components of the same species in stable isotope records (Kaislahti Tillman et al., 2010b; Loader et al., 2007; Ménot and Burns, 2001; Ménot-Combes et al., 2002; Moschen et al., 2009), we have constrained our study to investigate relationships of climate parameters and stable carbon and oxygen isotope values in relatively stable α-cellulose fractions extracted from Sphagnum fuscum stems. It is a favorable species in isotope studies for many reasons: it has an ubiquitous distribution in boreal and subarctic peatlands, high decay resistance and consequently high accumulation rates, high capacity to out-compete other species, and it is drought tolerant and mainly fed by meteoric water (Rydin and Jeglum, 2006). The aim of the study is to reconstruct climate shifts during the late Holocene in northeastern European Russia, where, as far as we know, no other isotope studies from moss plants have been performed. We compare our results with some of the Holocene paleoclimate reconstructions that have been conducted in the region using other proxies (macrofossils, tree rings, pollen) with varying time resolution and span.

Study area

Two peat plateau areas in the Bolshezemelskaya Tundra, about 50–100 km west of the Ural Mountains in northeastern European Russia, were investigated in summer 2008. These areas lie in the lowlands of the Usa River Basin, one in the vicinity of Seida village in the Komi Republic, the other one close to the Rogovaya River in the Nenets Autonomous District, about 50 km away from each other (see Figure 1 and Table 1 for location of sites). The Rogovaya sites are called ROG3-1 and ROG3-2 in order to avoid confusion with the sampling site of Oksanen et al. (2001), here called ROG3-O, and the study areas ROG 1 and ROG 2 of Hugelius et al. (2011). The subarctic tundra landscape in the sporadic/discontinuous permafrost area is a patchwork of upland tundra heath, flat peat plateaus, fens and thermokarst ponds. The present forest-line in the area follows approximately the 13.9°C July isotherm (Virtanen et al., 2004). The regional climate west of the Ural Mountains is mainly influenced by cold and dry air from the Arctic High or by (warmer) air originating from the North Atlantic low pressure systems bringing more moisture. Most of the precipitation falls in summer months; the eastern parts closer to the Ural Mountains experience more precipitation because of an orographic effect. In ad 1967–1986, mean annual air temperature (MAAT) varied from −6.1°C in Vorkuta to −3.3°C in Ust-Usa; July temperatures from 12.3°C in Vorkuta to 14.1°C in Ust-Usa; and mean annual precipitation (MAP) from 456 mm in Khoseda to 531 mm in Vorkuta (for locations of climate stations, see Figure 1). The active layer depth in the investigated peat plateaus varies between c. 30 and 60 cm; the fens, thermokarst lakes, and upland tundra and forest patches are mostly permafrost free or underlain by unfrozen supra-permafrost taliks.

Table 1.

Study sites and weather stations in northern taiga and tundra: location, elevation and annual averages of temperature (°C) and total amounts of precipitation (mm) in ad 1967–1986.

	Latitude	Longitude	Elevation (m a.s.l.)	Mean annual/July temperature (°C)	Mean annual precipitation (mm)
Tundra sites
Rogovaya ROG3-O	67°15′N	62°05′E	74
Rogovaya ROG3-1	67°09′N	61°51′E	69
Rogovaya ROG3-2	67°09′N	61°51′E	69
Seida SE2	67°03′N	62°56′E	98
*Khoseda	67°05′N	59°23′E	84	−5.3/+12.7	456
*Petrun	66°28′N	60°45′E	58	−4.6/+13.3	471
*Salekhard	66°32′N	66°32′E	16	−6.8/+13.5	456
*Vorkuta	67°29′N	64°06′E	172	−6.1/+12.3	531
Northern taiga site
Usinsk USI1	65°45′N	57°30′E	52
*Ust-Usa	65°58′N	56°55′E	78	−3.3/+14.1	498

Note: *weather stations.

Figure 1.

Location of climate stations and study sites: Usinsk (USI1) in the northern taiga, Rogovaya (ROG3-1,2 and ROG3-O) and Seida (SE2) in the tundra. Main land cover types are divided into forest, open tundra and peatland (colour figure available online).

The landscape in Rogovaya is characterized by a mosaic of peat plateaus, fens, and shrub tundra (all occupying 25–30% of the landscape), while the surroundings of the Seida sampling site consists of c. 50% shrub tundra, 20–25% tundra heath and 15–20% peat plateaus (Hugelius et al., 2011). About 15–20% of the Rogovaya area is covered by sparse forest dominated by spruce (Picea abies ssp. obovata) and mountain birch (Betula pubescens ssp. czerepanovii), which occur only in some isolated patches around the Seida site. The dwarf shrub vegetation at both sites is mainly dominated by dwarf birch (Betula nana) and heaths (e.g. Andromeda polifolia, Empetrum nigrum ssp. hermaphroditum, Ledum palustre, Vaccinium spp.). Peat plateau surfaces are partly bare because of wind erosion or cryoturbation and partly vegetated by Sphagnum mosses (mainly section Acutifolia, e.g. S. fuscum), other mosses (e.g. Polytrichum spp., Dicranum spp.), herbs (e.g. Drosera rotundifolia, Rubus chamaemorus) and lichens (e.g. Cladonia spp.). The fens, often surrounded by willows (Salix spp.), are covered by sedges (Carex spp., Eriophorum vaginatum, E. russeolum) and Sphagnum mosses (mainly section Cuspidata).

At the Seida site, the surface of hummock SE2 covered by Sphagnum fuscum, Rubus chamaemorus and Ledum palustre rose some tens of centimetres above a minor depression in the peat plateau (Supplementary Figure 2a, available online). The sampling site ROG3-1 was selected from a group of low Sphagnum hummocks growing on a peat plateau close to a fen with Carex spp. (Supplementary Figure 2b, available online). The sampling site ROG3-2 was on the edge of a peat plateau at a thermokarst lake where the dry plateau surface was mainly covered by mosses and lichen, Vaccinium spp., Empetrum nigrum and Rubus chamaemorus (Supplementary Figure 2c, available online), about 250 m southeast from ROG3-1. For the descriptions of peat plateau profiles and environmental settings of the Rogovaya site ROG3-O and Usinsk site USI1 in the northern taiga included in this study (Figure 1 and Supplementary Figure 2d, available online), the reader is referred to Oksanen et al. (2001, 2003).

Material and methods

Sampling

Peat samples were collected from short exposures in more or less circular Sphagnum fuscum hummocks (with a diameter of 1–2 m and a height of 25–35 cm) on the peat plateaus, and from profiles cut out from plateau edges eroding into adjacent thermokarst lakes and collapse scars. The hummocks were pre-sawed close to the highest point in order to avoid bending layers; 1 cm thick samples were subsequently cut by scissors from a c. 4 × 5 cm² area. At thermokarst shores, collapsing peat layers were first removed by spade and then 1 cm thick slices were cut from the cleaned peat exposures down to the permafrost table. For sampling methods of previously collected (freeze stored) peat profiles from Rogovaya (ROG3-O) and the Usinsk Mire (USI1), see Oksanen et al. (2001, 2003).

Dating methods

The upper (most recent) bulk peat samples from Seida and Rogovaya (SE2, ROG3-1 and ROG3-2) were analyzed for the activity of ²¹⁰Pb, ²²⁶Ra and ¹³⁷Cs via gamma spectrometry at the Gamma Dating Center, Department of Geography and Geology, University of Copenhagen. CRS-modeling was applied assuming a non-linear peat accumulation rate and that the hummocks were not severely influenced by post-depositional mobility (Appleby and Oldfield, 1983). Modern AMS radiocarbon dates of selected Sphagnum fuscum stems from SE2, ROG 3-1 and ROG 3-2 were calibrated by OxCal F14C and the calibration curve Bomb 04 NH1 for the Northern Hemisphere (Hua and Barbetti, 2004). Other AMS radiocarbon dates from ROG3-2, ROG3-O and USI1, and previous results from ROG3-O and USI1 profiles (Oksanen et al., 2001, 2003) were (re-) calibrated by OxCal 4.1.6 (Bronk Ramsey, 2010) and IntCal09 calibration curve (Reimer et al., 2009) and expressed in calendar years before present ±1σ (cal. BP; 0 = ad 1950). The modeled ages were based on median values obtained from the OxCal P-sequence model (based on Poisson processes and approximate proportionality to the depth of deposition) where k (increments per unit) was set to a value of 10 (Bronk Ramsey, 2008). Stratigraphic changes were considered in the age model of ROG3-2; modern ages were extrapolated to the top of a rootlet layer, from where ages were interpolated to the top of the P-sequence modelled part below the rootlet layer.

Macrofossil analysis

Plant macrofossils were identified with a stereo binocular (25–40× magnification) and reference literature (Hallingbäck and Holmåsen, 1982; Mossberg and Stenberg, 2003). Sphagnum species were identified (stained by methyl violet) by leaf morphology under a light microscope (100–400× magnification) and reference literature (Laine et al., 2009; Lange, 1982). The fractions are presented as volume percentages of the total sample, rare elements as counts recalculated to a standard sample volume of 5 cm³ (< 1 finding/5 cm³ given as value 1).

Meteorological data

For the tundra study sites Seida and Rogovaya, monthly temperature averages and precipitation amounts were based on observations from meteorological stations in Petrun, Khoseda, Vorkuta and Salekhard (Figure 1 and Table 1) (KNMI climate explorer). The 20 yr period ad 1967–1986 was chosen as a reference period, because both temperature and precipitation data were available from all four stations during the whole year. During the reference period, June to September temperatures had high pairwise correlations between observations from all three meteorological stations west of the Ural Mountains (R² = 0.97–0.98, p < 0.01, Bonferroni correction applied). The meteorological station in Salekhard on the east side of the Ural Mountains was included in the composite tundra temperature record, because summer temperatures in Salekhard correlated rather well with those from the tundra stations west of the mountain chain (R² = 0.79–0.85, p < 0.01), and it provided the longest continuous record (ad 1883–2008). When observations were only obtained from Salekhard (ad 1883–1925), values were decreased by 0.6°C, i.e. the difference between temperatures in Salekhard and composite tundra July–August averages during the reference period. Salekhard was excluded from tundra precipitation averages because of an assumed effect of the mountain chain on the precipitation amount, and Petrun because of a relatively short period of observed data (ad 1967–1993). In the northern taiga, temperature data (ad 1929–1994) and precipitation data (ad 1936–1992) from a weather station in Ust-Usa were used for the study site at Usinsk. Precipitation anomalies were calculated as deviations (%) from mean values during the reference period. Additionally, temperature observations recorded during a carbon flux field campaign in Seida in 2007–2008 (Marushchak et al., 2011) provided a control for how well the composite tundra records display climate conditions at the Seida site. Winter/annual precipitation amounts were not recorded in Seida.

Stable isotope analyses

Sphagnum fuscum stems were manually selected and cleaned from rootlets and detritus under a stereo binocular. A minimum number of pieces of stems was set at 30 in order to diminish the influence of differences between individual plants and to purify a sufficient amount of α-cellulose by the method developed for wood (Green, 1963) and modified to suit moss samples (Kaislahti Tillman et al., 2010b; Loader et al., 2007). Stable carbon and oxygen isotopes (δ¹³C, δ¹⁸O) were analyzed at the Stable Isotope Laboratory (SIL), Department of Geological Sciences, Stockholm University.

For stable carbon isotope analysis, 0.5±0.1 mg of α-cellulose was combusted in tin capsules with a Carlo Erba NC2500 elemental analyzer connected via ConfloIV open split interface to reduce the gas volume, and analyzed by a Finnigan MAT Delta V mass spectrometer. The laboratory reports that the reproducibility is < 0.15‰ and the relative error is < 1%. The results were expressed in δ¹³C_cellulose = [(R_cellulose − R_ref)/R_ref] × 1000 (‰); R = ¹³CO₂/¹²CO₂; reference VPDB. The ¹³C-level in the atmosphere was considered relatively stable in the middle and late Holocene (Elsig et al., 2009), but modern δ¹³C values were corrected for the depletion of ¹³C during the industrial time, i.e. after ad 1850, by adding correction factors adapted from table 2 in McCarroll and Loader (2004).

For oxygen isotope analysis, 1.0±0.1 mg of α-cellulose was weighted in silver capsules and analyzed with a Finnigan DeltaV mass spectrometer connected to a Temperature Conversion Elemental Analyzer (TC/EA) through a ConfloIV open split interface. Standard deviation for IAEA-standard was better than ±0.4‰. The results were expressed in δ¹⁸O_cellulose = [(R_cellulose − R_ref)/R_ref] × 1000 (‰); R = ¹⁸O/¹⁶O; reference VSMOW. ROG3-1 samples were analyzed separately together with a re-analysis of some samples from other sites, which showed +0.99‰ higher values in average compared with previous results (n=7, SD=0.4‰); ROG3-1 values were subsequently corrected for this difference.

Climate correlation

An attempt was made to correlate isotope data with local climate observations separately from tundra and northern taiga. Relationships between isotope data and temperature were studied for different combinations of months when photosynthesis occurs (MJJAS), between isotope data and precipitation for annual data (October previous year until September growth year) and periods with and without snow cover (October–May and June–September). In the topmost part of Seida hummock SE2 (peat accumulation > 1 cm/yr), δ¹³C and δ¹⁸O values were averaged from two to three samples of 1 cm thickness by using the ²¹⁰Pb age model adjusted within the dating error (±2 yr, Figure 2a). Annual climate data were similarly averaged from the lower part of the hummock SE2 in order to match the age of isotope signals (peat accumulation < 1 cm/yr below 14 cm depth). In adjusting the limits of averaging, peaks and troughs were considered in both data sets within the dating error. The approach is justified because of varying annual Sphagnum fuscum growth rate depending on climate conditions (e.g. Dorrepaal et al., 2003; Rydin and Jeglum, 2006) and increasing compaction and decomposition of peat with depth (e.g. Clymo, 1984). Furthermore, moss increments originating from one growing season may be spread into two adjacent samples during the subsampling. The relationships between stable isotope ratios and climate parameters were regarded significant if the p-value was < 0.05, evaluated by the statistical programme PAST version 2.07 (Hammer et al., 2001).

Figure 2.

Age models of (a) the Seida hummock SE2, based on ²¹⁰Pb dating and one radiocarbon date (green line with error bars), and on averaged depths used in climate reconstruction (red line); (b) the Rogovaya hummock ROG3-1, based on a radiocarbon date and linear extrapolation to the surface sample collected ad 2008; (c–e) peat profiles ROG3-2, ROG3-O and USI-1, respectively, based on radiocarbon dates and the OxCal P-sequence model (Bronk Ramsey, 2008). Calibrated ‘bomb peak’ radiocarbon ages by OxCal F14C and the calibration curve Bomb 04 NH1 (Hua and Barbetti, 2004) are marked by circles (higher probability) and triangles (lower probability), see Table 2 for probabilities. Samples from ROG3-O and USI1 previously published by Oksanen et al. (2001, 2003) are recalibrated (colour figure available online).

Results

Chronology

The ²¹⁰Pb age model of the hummock SE2 was modified by setting a reference-date (Appleby, 2001) at ad 1963 (bomb test maximum) at the lowermost ¹³⁷Cs peak found at a depth of 33.5 cm, coherent with an independent post-bomb AMS ¹⁴C date ad 1972–1973 at 31.5 cm depth (Figure 2a, Table 2).The surface contents of unsupported ²¹⁰Pb were about 100 Bq/kg. Contrary to what is generally expected, the vertical variation of unsupported ²¹⁰Pb was irregular and contents were generally increasing with depth. Assuming constant flux of unsupported ²¹⁰Pb from the atmosphere this pattern indicates increasing dilution of the ²¹⁰Pb caused by a significant increase in accumulation rate towards the present (e.g. Appleby and Oldfield, 1983). However, some leaching of ²¹⁰Pb as well as ¹³⁷Cs cannot be excluded. The age models of the hummock ROG3-1 and of the upper part of profile ROG3-2 rely on AMS ¹⁴C-dating (Figures 2b,c, Table 2), because ²¹⁰Pb dating of these sections were similarly hampered by fluctuations of unsupported ²¹⁰Pb and distinct ¹³⁷Cs peaks were not found in these cores. The results of new radiocarbon dates and recalibrated previously published radiocarbon dates for profiles ROG3-O and USI1 (Figure 2d,e) are compiled in Table 2.

Table 2.

AMS-radiocarbon dates from Seida (SE2), Rogovaya (ROG3-1, ROG3-2, ROG3-O) and Usinsk (USI-1) dated at the Poznan Radiocarbon Laboratory (Poz), Angstrom laboratory of Uppsala University (Ua) and van de Graaff Laboratory, Utrecht University (UtC), and conventional dates at Helsinki University (Hel). For the modern dates, the minima and maxima from post-bomb calibration curve are given in ages ad. Other calibrated ages are given in calendar years ±1σ before present (ad 1950 = 0 cal. BP). Modeled ages are median values BP ± 1σ, calculated by OxCal P-sequence model (Bronk Ramsey, 2008).

Site depth (cm)	Lab code	Radiocarbon age (BP)	Median (cal. BP)	Min (cal. BP)	Max (cal. BP)	Model age	Min age	Max age
SE2 (31–32)	Poz-27008	145.46 ± 0.45 pMC	Modern ad	1972–1973 (88.9%)	1963 (6.5%)
ROG3-1 (21–22)	Poz-45134	169.61 ± 0.51 pMC	Modern ad	1965–1966 (95.4%)
ROG3-2 (5–6)	Poz-37727	114.83 ± 0.34 pMC	Modern ad	1990–1992 (88.7%)	1957–1958 (6.7%)
ROG3-2 (10–11)	Poz-35974	120.53 ± 0.36 pMC	Modern ad	1983–1986 (61.5%)	1959–1961 (34.0%)
ROG3-2 (20–21)	Poz-35941	95±30	110	33	255	92	59	121
ROG3-2 (30–31)	Poz-35940	65±30	98	35	251	127	115	139
ROG3-2 (40–41)	Poz-35939	215±30	179	−2	301	159	145	186
ROG3-2 (50–51)	Poz-27004	165±30	177	1	282	192	169	261
ROG3-2 (55–56)	Poz-37728	200±30	178	−2	290	209	181	295
*ROG3-O (5–10)	UtC-10062	1556±36	1453	1397	1516	1467	1407	1518
*ROG3-O (20–25)	Hel-3798	1920±100	1865	1735	1986	1946	1915	1985
ROG3-O (31–34)	Ua-41695	2164±34	2181	2117	2302	2278	2260	2300
*ROG3-O (45–50)	Hel-3799	2860±90	3001	2862	3141	2780	2756	2802
*ROG3-O (65–75)	UtC-5724	3100±45	3324	3265	3376	3369	3335	3440
*USI1 (40–50)	UtC-7402	315±45	388	307	437	342	300	386
USI1 (83–87)	Ua-41696	1530±33	1415	1370	1510	1503	1487	1522
*USI1 (120–130)	UtC-8675	2775±40	2871	2796	2925	2797	2774	2816
*USI1 (160–170)	UtC-7403	3580±50	3883	3831	3972	3949	3885	4061

Note: *Previously published radiocarbon dates by Oksanen et al. (2001, 2003).

Stratigraphy and macrofossil analysis

In Seida hummock SE2, small ice crystals at 36–35 cm depth marked a transition to frozen soil. Between 36 cm and 32 cm, rootlet peat was dark and humified. Peat at 32–31 cm depth consisted of 83% Sphagnum section Acutifolia (uncertain identification at species level), 5% wood, 8% roots and Betula nana leaf fragments. The upper part of the hummock consisted of Sphagnum fuscum peat with some rootlets; the 31–24 cm interval with 72–91% S. fuscum was slightly darker and more compacted than the top 24 cm with 93–99% S. fuscum. For a detailed plant macrofossil analysis diagram, see Supplementary Figure 1a, available online.

In Rogovaya hummock ROG3-1, the permafrost table was situated at 31 cm depth below the surface. Rootlet peat at 31–25 cm was dark and rather humified. The upper part of the hummock consisted of Sphagnum fuscum peat, which was somewhat darker and more decomposed at 25–24–(20) cm depth compared with the top 20 cm. At 23–21 cm, S. fuscum made up 49–55% of the total assemblage while the content above 21 cm depth was 82–98%. In the Rogovaya profile ROG3-2, the lowermost layer at 56–51 cm depth was characterized by dark humified peat. The peat layer between 51 cm and 35 cm, mainly formed by S. fuscum and Dicranum spp., and a short period by Drepanocladus spp., was more compacted and darker than above. From 35 cm up to 20 cm depth, the proportion of S. fuscum varied between 55% and 84%, while a rootlet peat layer from 20 cm to 17.5 cm contained only 20–45% S. fuscum. The top 17 cm was low-humified peat with 68–94% S. fuscum. Plant macrofossil diagrams of ROG3-1 and ROG3-2 are provided in Supplementary Figure 1b and c, available online.

Based on previously published macrofossil analyses, portions of the peat sections of ROG3-O (0–35 cm) and USI1 (30–110 cm) with >40% Sphagnum fuscum and <20% other Sphagnum species were resampled for isotope analyses. For complete stratigraphies and macrofossil analyses, see Oksanen et al. (2001, 2003).

Stable carbon and oxygen isotope records

Stable carbon and oxygen isotope ratios from Rogovaya hummock ROG3-1 and Seida hummock SE2 with the highest temporal resolution of peat sequences are shown against sample depths in Figure 3a–d. Relationships between instrumental climate records (tundra composite records and Ust-Usa record from the northern taiga separated) and isotope data from the Seida hummock SE2 were tested in order to understand which season had the greatest impact on carbon and oxygen ratios. There was a positive correlation between climate from ad 1978 to 2008 and Sphagnum isotopes from the same period. If we consider the dating uncertainty of ±2 years of the ²¹⁰Pb age–depth model, then a shift of the SE2 chronology within dating uncertainty reveals even closer correspondence between them. The most significant (positive) linear correlations were found between δ¹³C and average July–August tundra and northern taiga temperatures (R² = 0.58, p < 0.01 and R² = 0.44, p < 0.01, respectively, see Table 3 and Figure 3e). The lower grade of significance in taiga correlations compared with tundra values is explained by the lower number of data points because of a shorter instrumental record. There was some correlation between δ¹³C and winter precipitation anomalies (October–May, mainly falling as snow) in the tundra (R² = 0.30, p ≤ 0.01). δ¹⁸O ratios showed stronger positive relationships with variations in amounts of winter precipitation in the tundra and the northern taiga (R² = 0.36, p < 0.01 and R² = 0.41, p = 0.05, respectively, Table 3 and Figure 3f), and a weaker relationship with annual precipitation in the tundra (R² = 0.23, p = 0.02, Table 3). Notable is that there was no correlation between carbon and oxygen isotope ratios (R² = 0.06, p = 0.25) despite the common relationship to winter precipitation.

Figure 3.

(a) δ¹³C (VPDB, ‰) and (b) δ¹⁸O (VSMOW, ‰) records of Rogovaya hummock ROG3-1 (black line, open quadrants); (c) δ¹³C (VPDB, ‰) and (d) δ¹⁸O (VSMOW, ‰) records of Seida hummock SE2 (grey line, closed circles). δ¹³C values corrected for ¹³C_atm by factors adapted from McCarroll and Loader (2004) are shown in broken lines, error bars ± SD for mean δ¹⁸O values. (e) Linear regression models between δ¹³C values from the Seida hummock SE2 and recorded mean July–August temperatures ad 1978–2008 in the tundra (red circles, solid trend line) and the northern taiga (black crosses, broken trend line). (f) Linear regression models between δ¹⁸O values from SE2 and winter (October–May) precipitation anomalies ad 1978–2008 in the tundra (blue diamonds, solid trend line) and the northern taiga (slanted crosses, broken trend line), calculated as deviations (%) from mean value during the reference period ad 1967–1986 (colour figure available online).

Table 3.

Determinants of coefficients (R²) for relationships between tundra/northern taiga climate data (temperature T and precipitation P) and stable isotope ratios (δ¹³C, δ¹⁸O), and statistical significances (p < 0.05 indicates significant relationship). In the left upper corner, n indicates the number of data points in correlation models.

Tundra: n=23 (T and P)Taiga: n=23 (T), n=10 (P)	δ¹³C R² linear/polynomial	δ¹³C p-valuelinear/polynomial	δ¹⁸O R² linear/polynomial	δ¹⁸O p-valuelinear/polynomial
May–Aug T	0.31	<0.01	0.11	0.13
June–Aug T	0.50	<0.01	0.11	0.12
June–Sept T	0.39	<0.01	0.08	0.20
July T	0.37	<0.01	0.03	0.43
July–Aug T tundra July–Aug T taiga	0.58/0.580.44/0.44	<0.01/< 0.01<0.01/< 0.01	0.08/0.090.12/0.13	0.18/0.410.10/0.24
June–Sept P tundra June–Sept P taiga	0.010.01	0.640.81	0.010.08	0.650.41
Annual P tundra Annual P taiga	0.110.06	0.130.50	0.230.01	0.020.82
Oct–May P tundra Oct–May P taiga	0.30/0.340.30/0.38	<0.01/0.010.10/0.19	0.36/0.370.41/0.45	<0.01/0.010.05/0.12

Stable isotope ratios of northern taiga site USI1 and tundra sites ROG3-O, ROG3-1, ROG3-2, SE2 and climate data are shown in Figures 4 and 5, where ages of peat sections are adapted from age models shown in Figure 2. Trend lines show that during the period ad 1979–2008 when winter precipitation increased on average by 16% and July–August temperatures by 1.8°C in the tundra area, δ¹⁸O increased on average by 0.3‰ and δ¹³C by 0.8‰ in Seida SE2 moss samples. Longer trend lines show that winter precipitation has increased by 34% during the period ad 1937–2008, while summer temperatures increased by 0.3°C in ad 1910–2008 (Figure 4). Stable isotope ratios had the largest variations in the ROG3-2 peat profile: δ¹⁸O from 16.2‰ to 20.3‰ and δ¹³C from −28.2‰ to −23.4‰ (Figures 4 and 5). Low resolution series showed less variation: δ¹⁸O varied from 16.4‰ to 19.5‰ in the ROG3-O tundra record between c. 2500 and 1300 cal. BP and from 18.6‰ to 20.6 ‰ in the USI1 record from northern taiga during the past 2200 years, and δ¹³C varied from −28.0‰ to −25.0‰ in ROG3-O and from −28.7‰ to −25.8‰ in USI1.

Figure 4.

(a) Composite tundra October–May precipitation anomalies (in % compared with the reference period ad 1968–1987, black solid line with diamonds); (b) δ¹⁸O records of SE2 (grey line, closed circles) and ROG3-2 (black line, crosses) with error bars ± SD; (c) composite tundra July–August temperature record (black line), northern taiga record (grey line), Seida flux site July–August temperatures ad 2007–2008 (open diamonds, from Marushchak et al., 2011); (d) δ¹³C records of SE2 and ROG3-2 (corrected for ¹³C_atm). The ROG3-2 chronology is tentative, ROG3-1 values are not shown in the time scale because of poor chronology. Trend lines of whole periods are shown by grey lines and trend lines of ad 1978–2008 by red lines (colour figure available online).

Figure 5.

(a) δ¹³C records of tundra sites ROG3-O (black line, open circles) and ROG3-2 (black line, crosses); (b) δ¹³C record of the northern taiga site USI-1 (grey line, slanted crosses); (c) δ¹⁸O records of tundra sites ROG3-O and ROG3-2; (d) δ¹⁸O record of the northern taiga site USI1 with error bars ± SD. Interpretations of isotope data on the right.

Discussion

Stable carbon and oxygen isotope ratios showed no long-term trends in the studied peat sequences covering several thousands of years of peat accumulation and decay (ROG3-O and USI1 in Figure 5). This suggests that they were not affected by the increasing age or increasing decomposition rate of peat because of anaerobic decay in the catotelm in addition to aerobic decay in the acrotelm. Furthermore, the study was based on α-cellulose fractions of a single Sphagnum species partly because of selective decay among species (Clymo, 1984) and partly because the α-cellulose fraction is regarded as a relatively stable plant component (Loader et al., 2007). However, isotope variability is plausibly more smoothed in older parts of ROG3-O and USI1 because of lower net peat accumulation rate and higher compaction, compared with the younger records of ROG3-2, ROG3-1 and SE2. The hummock SE2 had highest time resolution of peat sections and the reliability of its age model based on ²¹⁰Pb dating was considerably improved by independent ¹³⁷Cs- and ¹⁴C-analyses. In the following text, we interpret our stable isotope records based on found relationships between corrected δ¹³C values of SE2 and July–August average temperatures and between δ¹⁸O values of SE2 and winter precipitation anomalies, shown in Figure 3e, f.

If we accept the regressions and consider δ¹³C as a proxy for summer temperature, a general cooling period is observed from 2500 cal. BP to 1850 cal. BP in the tundra, and from 2200 cal. BP to 1850 cal. BP in the northern taiga (Figure 5). In comparison, the macrofossil study from the Rogovaya peat plateau by Oksanen et al. (2001) showed permafrost initiation at c. 3100 cal. BP and, following collapse, a renewed period of permafrost aggradation c. 2200–1900 cal. BP, which corroborates the cooling trend seen in our study. At the Usinsk peat plateau site USI1 in the northern taiga, first tentative permafrost aggradation is suggested at c. 2900 cal. BP when Sphagnum fuscum was established, but is more certain after c. 2300 cal. BP, when there is no more S. magellanicum in the record (Oksanen et al., 2003). A pollen and plant macrofossil study by Kultti et al. (2004) concludes that the southernmost position of the treeline in the late Holocene occurred between 2700 and 2100 BP in this region, and a pollen-based temperature reconstruction by Andreev and Klimanov (2000) from Khaipudurskaya Guba section north of Vorkuta also showed the lowest July temperatures between 2 and 3 (non-calibrated) ka BP (thousand years before present, ad 1950), i.e. some centuries earlier than in our study. The latter authors also suggest negative precipitation anomalies around 2 ka BP in their investigated tundra region. In our record, δ¹⁸O curves deviate considerably between 2500 and 1650 cal. BP implying that winter precipitation was low at the beginning of this period but reached amounts comparable with our reference period at c. 1700 cal. BP in the tundra, while the overall precipitation amounts were generally high but slightly declining in the northern taiga (Figure 5).

During the first millennium ad, our record shows fluctuating summer temperatures in the tundra, similar to the pollen based reconstruction north of Vorkuta (Andreev and Klimanov, 2000). Unfortunately, the S. fuscum record in ROG3-O reached only until c. ad 640 (~1300 cal. BP; Figure 5). In the northern taiga, a warming trend occurred till around ad 1030 (~1000 cal. BP), coinciding approximately with the so-called ‘Medieval Climate Anomaly’ that has been recorded in the Northern Hemisphere as a period with close to modern temperatures (Moberg et al., 2005). At that time, winter precipitation increased slightly in the northern taiga, but does not show much additional variability during the last millennia (Figure 5).

Only four samples analysed from the northern taiga were dated to the past millennium, of which the lowest values of δ¹³C and δ¹⁸O were around ad 1320 (~630 cal. BP; Figure 5). Because of the low resolution of our northern taiga record and a gap in the tundra record until the latter part of the 18th century, no conclusions can be drawn about possible magnitude and timing of the so-called ‘Little Ice Age’ in the region. A pollen-based reconstruction by Andreev and Klimanov (2000) had also very low resolution, and showed two warmer periods interrupted by a cold stage in the middle of the millennium. Oksanen et al. (2001) suggested a new permafrost aggradation phase in Rogovaya ad 1350–1850 (600–100 cal. BP), connected to the ‘Little Ice Age’. δ¹⁸O values in tundra profiles suggest that the amount of winter precipitation was lowest in the latter part of the 18th century (Figure 5), followed by a generally increasing trend until the middle of the 19th century, similar to higher δ¹³C values. We interpret the latter as increasing summer temperatures. There are short periods of cooling around ad 1815 and ad 1835 that possibly coincide with the Dalton solar minimum (Bard et al., 2000; Magny et al., 2010) and cold spells recorded in a tree-ring study near Salekhard (Briffa et al., 1995), keeping in mind the uncertainty of the age model. The oxygen isotope curve also shows fluctuations, interpreted as decreased winter precipitation in time periods when influence from dry and cold arctic air masses dominated. A gap in our stable isotope record between c. ad 1855 and 1915 was caused by a rootlet layer that indicates dry local conditions. After the gap, δ¹³C values of the Rogovaya peat profile ROG3-2 start low but show a generally increasing trend during the 20th century that is not seen in the long instrumental temperature record from the tundra area (Figure 4). The large uncertainty of ROG3-2 chronology is a possible reason for this discrepancy. δ¹⁸O values show a corresponding decline as low winter precipitation observations in the middle of the 20th century. From the late 1970s, there is an increasing trend in stable carbon and oxygen isotope data and in contemporary instrumentally observed summer temperatures and winter precipitation anomalies in the tundra. No isotope data were obtained from the northern taiga peat profile from the last century.

Various climatic conditions, such as annual variation in seasonality and growing season length, may impact the rate of photosynthesis of Sphagnum species in different ways. For example, as S. fuscum is a drought-tolerant species (Rydin, 1985), its growth is more inhibited in periods with low temperatures (Gunnarsson, 2005). The found relationship between δ¹³C values and summer temperatures was plausibly caused by the impact of variations of photosynthetic rate on carboxylation discrimination of ¹³C. In this study from northeastern European Russia, we found correlation between δ¹³C values and July–August temperature and winter precipitation, whereas a study from central Canada showed correlation only to July temperatures (Kaislahti Tillman et al., 2010a, 2010b). Additional correlations may depend on slightly different optimal conditions for moss productivity, since climate has a more maritime influence with prolonged warm period and more variation in rain patterns in the Russian sites compared with continental Canada. Higher average and minimum temperatures at the beginning of the growing season and increased protection of the photosynthetic tissues in the capitula were indicated as main reasons for a positive effect of increased winter snow cover on the biomass productivity of Sphagnum fuscum in a climate manipulation experiment in the Swedish subarctic (Dorrepaal et al., 2003). Therefore, the positive correlation of δ¹³C with precipitation (snow) amounts in the preceding winter in our S. fuscum record may be explained by varying timing of the initiation of the photosynthetic period. Carbon fractionation resulting from moisture conditions or pCO₂ variation during the growing season may be some of the determinants for the isotope variation that was not explained by direct and/or indirect temperature control in this study, as it is difficult to disentangle these effects (Ménot and Burns, 2001). However, our results are not comparable with studies showing strong correlation to the surface wetness because they rely on different species growing in various wetness conditions and habitats, or with studies showing δ¹³C dependence on ¹³C depleted methane uptake (Raghoebarsing et al., 2005) because our study is solely based on one species S. fuscum growing as insulating mats on dry hummocks, where the active layer depth (peat thawed during summer) is shallow and methane formation is inhibited. A possible link between wetness and temperature is that besides releasing more methane, the wetter parts of fen peatland have higher evapotranspiration rates and dampen surface temperature increase that can be considerably high in non-shaded dry surfaces during sunny summer days (Rydin and Jeglum, 2006). Hence, the proximity to wet pools may partly explain local differences. Isotope signals deviated between tundra sites ROG3-1/SE2 surrounded by a wet fen/depression (lower δ¹³C values) and ROG3-2 situated on the dry tundra plateau edge (higher δ¹³C values). On the contrary, δ¹³C values were similar between the northern taiga site USI1 surrounded by wet pools and the tundra site ROG3-O on peat plateau with less open pools, despite the ~1°C higher modern air temperatures in the northern taiga. Consequently, variation of δ¹³C should be seen as a signal for local absolute temperature, because the sensitivity of δ¹³C to temperature changes may differ from site to site for example because of different latent heat fluxes.

In a study of δ¹⁸O variation in plants from a temperate peat bog, it was shown that precipitation is the main source of oxygen found in α-cellulose fractions of mosses and used in photosynthesis by plants (Ménot-Combes et al., 2002). In tundra peat plateaus, the local hydrology is also affected by frost heave and the permafrost table that binds more of the percolating moisture in expanding ice during colder times. Although hummocky species mainly obtain water from precipitation during summers, the initial moisture in the beginning of growth seasons originates from the snowmelt, as discharge is inhibited by permafrost. The positive relationship found in our study between δ¹⁸O and winter precipitation may therefore be a consequence of different source areas of precipitation in winter time: since arctic air masses are cold and dry, snow accumulated in winters with low precipitation have lower ¹⁸O content because of latitudinal depletion, compared with snow-rich winters with moisture origin from more westerly air masses. Calculated by the modeled equation for spatial distribution of δ¹⁸O in precipitation (δ¹⁸O_ppt) of Bowen and Wilkinson (2002), δ¹⁸O_ppt varies from −16.3‰ in tundra sites Seida and Rogovaya to −15.1‰ in Usinsk in the northern taiga. Between 2500 and 1700 cal. BP, the difference in precipitation δ¹⁸O values between the Usinsk and Rogovaya sites was much larger (~4‰). The discrepancy between sites may depend on a greater contribution of westerly air masses to the northern taiga, and arctic air to the tundra region, local hydrology (permafrost history), or varying evapotranspiration (different ground surface temperatures). Thereafter, the air masses with higher δ¹⁸O_ppt may have reached further to the north, as reflected by similarly increasing δ¹⁸O values at tundra and taiga sites in the overlapping records. As several processes obviously impact the oxygen isotope record, our interpretation of winter precipitation anomalies should be regarded as tentative.

Conclusions

This study provides stable carbon and oxygen isotope data with palaeoclimatic interpretations, derived from a single moss species from peat archives in subarctic Russia. We conclude that increased decomposition of peat with age does not affect the long-term trends seen in stable isotope time series. Temperature controlled isotope fractionation was the most important factor explaining stable carbon isotope variability in α-cellulose of Sphagnum fuscum stems. A new application of moss stable isotopes in climate research is suggested by the link between winter precipitation and oxygen isotope ratios. The varying dominance of some of the factors in different time periods is probably related to changes in air-mass trajectories, local hydrology and permafrost history. Therefore, in addition to plant macrofossil analyses, surface wetness reconstructions of peat by for example testate amoebae and comparisons with other isotope records from lake sediments and tree rings could be helpful for interpretations. Although our results confirm the potential of stable carbon and oxygen isotope ratios derived from single moss plants for climate reconstruction, a wider distribution of calibration sets over a range of temperature and precipitation regimes and landscape settings is needed to further evaluate the impact of local and microtopographical differences.

Footnotes

Acknowledgements

We acknowledge Maija Marushchak and Pertti Martikainen (Department of Environmental Science, University of Eastern Finland) and Isabell Kiepe and Thomas Friborg (Department of Geography and Geology, University of Copenhagen) for providing climate data from Seida. We also thank the editor and two anonymous reviewers for comments which improved the manuscript.

Funding

The Ahlmann foundation, the Bolin Centre for Climate Research, Kungstenen foundation, The Swedish Society for Anthropology and Geography, and the EU-funded 6FP CARBO-North project (036993) are acknowledged for financial support.

References

Andreev

Klimanov

(2000) Quantitative Holocene climatic reconstruction from Arctic Russia. Journal of Paleolimnology 24: 81–91.

Appleby

(2001) Chronostratigraphic techniques in recent sediments. In: Last

Smol

(eds) Tracking Environmental Change Using Lake Sediments. Volume 1: Basin Analysis, Coring and Chronological Techniques. Dordrecht, Boston, London: Kluwer Academic Publishers, pp. 171–203.

Appleby

Oldfield

(1983) The assessment of Pb-210 data from sites with varying sediment accumulation rates. Hydrobiologia 103: 29–35.

Bard

Raisbeck

Yiou

. (2000) Solar irradiance during the last 1200 years based on cosmogenic nuclides. Tellus 52B: 985–992.

Bowen

Wilkinson

(2002) Spatial distribution of δ¹⁸O in meteoric precipitation. Geology 30: 315–318.

Brenninkmeijer

CAM

van Geel

Mook

(1982) Variations in the D/H and ¹⁸O/¹⁶O ratios in cellulose extracted from a peat bog core. Earth and Planetary Science Letters 61: 283–290.

Briffa

Jones

Schweingruber

. (1995) Unusual twentieth-century summer warmth in a 1,000-year temperature record from Siberia. Nature 376: 156–159.

Bronk Ramsey

(2008) Deposition models for chronological records. Quaternary Science Reviews 27: 42–60.

Bronk Ramsey

(2010) OxCal v4.1.6. Radiocarbon Calibration Software. Research Lab for Archaeology, Oxford, U.K. http://c14.arch.ox.ac.uk.

10.

Clark

Fritz

(1997) Environmental Isotopes in Hydrogeology. Florida: CRC Press LLC.

11.

Clymo

(1984) The limits to peat bog growth. Philosophical Transactions of the Royal Society B 303: 605–654.

12.

Daley

Barber

Street-Perrott

. (2010) Holocene climate variability revealed by oxygen isotope analysis of Sphagnum cellulose from Walton Moss, northern England. Quaternary Science Reviews 29: 1590–1601.

13.

Dorrepaal

Aerts

Cornelissen

JHC

. (2003) Summer warming and increased winter snow cover affect Sphagnum fuscum growth, structure and production in a sub-arctic bog. Global Change Biology 10: 93–104.

14.

Elsig

Schmitt

Leuenberger

. (2009) Stable isotope constraints on Holocene carbon cycle changes from an Antarctic ice core. Nature 461: 507–510.

15.

Farquhar

Ehleringer

Hubick

(1989) Carbon isotope and photosynthesis. Annual Review of Plant Physiology & Plant Molecular Biology 40: 503–537.

16.

Gray

Beer

Geller

. (2010). Solar influences on climate. Reviews of Geophysics 48: 1–53.

17.

Green

(1963) Wood cellulose. In: Green

Bemiller

(eds) Methods in Carbohydrate Chemistry. New York, London: Academic Press, pp. 9–20.

18.

Gunnarsson

(2005) Global patterns of Sphagnum productivity. Journal of Bryology 27: 269–279.

19.

Hallingbäck

Holmåsen

(1982) Mossor – En fälthandbok. Stockholm: Interpublishing.

20.

Hammer

Harper

DAT

Ryan

(2001) Paleontological statistics software package for education and data analysis. Palaeontologia Electronica 4(1): 9 pp.

21.

Hantemirov

Shiyatov

(2002) A continuous multimillennial ring-width chronology in Yamal, northwestern Siberia. The Holocene 12: 717–726.

22.

Hong

Lin

. (2010) Anti-phase oscillation of Asian monsoons during the Younger Dryas period: Evidence from peat cellulose δ¹³C of Hani, Northeast China. Palaeogeography, Palaeoclimatology, Palaeoecology 297: 214–222.

23.

Hua

Barbetti

(2004) Review of tropospheric bomb ¹⁴C data for carbon cycle modeling and age calibration purposes. Radiocarbon 46: 1273–1298.

24.

Hugelius

Virtanen

Kaverin

. (2011) High-resolution mapping of ecosystem carbon storage and potential effects of permafrost thaw in periglacial terrain, European Russian Arctic. Journal of Geophysical Research 116: 1–14.

25.

Kaakinen

Eronen

(2000) Holocene pollen stratigraphy indicating climatic and treeline changes derived from a peat section at Ortino, in the Pechora lowland, northern Russia. The Holocene 10: 611–620.

26.

Kaislahti Tillman

Holzkämper

Kuhry

. (2010a) Long-term climate variability in continental subarctic Canada: A 6200-year record derived from stable isotopes in peat. Palaeogeography, Palaeoclimatology, Palaeoecology 298: 235–246.

27.

Kaislahti Tillman

Holzkämper

Kuhry

. (2010b) Stable carbon and oxygen isotopes in Sphagnum fuscum peat from subarctic Canada: Implications for palaeoclimate studies. Chemical Geology 270: 216–226.

28.

Kim

Verma

(1996) Surface exhance of water vapour between an open Sphagnum fen and the atmosphere. Boundary-Layer Meteorology 79: 243–264.

29.

Kuhry

Turunen

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

Vitt

(eds) Boreal Peatland Ecosystems. Springer-Verlag, pp. 25–46.

30.

Kultti

Oksanen

Väliranta

(2004) Holocene tree line, permafrost, and climate dynamics in the Nenets Region, East European Arctic. Canadian Journal of Earth Sciences 41: 1141–1158.

31.

Laine

Harju

Timonen

. (2009) The intricate beauty of Sphagnum mosses – A Finnish guide to identification. University of Helsinki Department of Forest Ecology Publications 39: 1–190.

32.

Lamentowicz

Cedro

Gałka

. (2008) Last millennium palaeoenvironmental changes from a Baltic bog (Poland) inferred from stable isotopes, pollen, plant macrofossils and testate amoebae. Palaeogeography, Palaeoclimatology, Palaeoecology 265: 93–106.

33.

Lange

(1982) Key to the northern boreal and arctic species of Sphagnum, based on characteristics of the stem leaves. Lindbergia 8: 1–29.

34.

Loader

McCarroll

van der Knaap

. (2007) Characterizing carbon isotope variability in Sphagnum. The Holocene 17: 403–410.

35.

Loisel

Garneau

Hélie

J-F

(2009) Modern Sphagnum δ¹³C signatures follow a surface moisture gradient in two boreal peat bogs, James Bay lowlands, Québec. Journal of Quaternary Science 24: 209–214.

36.

Loisel

Garneau

Hélie

J-F

(2010) Sphagnum ¹³C values as indicators of palaeohydrological changes in a peat bog. The Holocene 20: 285–291.

37.

McCarroll

Loader

(2004) Stable isotopes in tree rings. Quaternary Science Reviews 23: 771–801.

38.

MacDonald

Velichko

Kremenetski

. (2000). Holocene tree-line history and climate change across northern Eurasia. Quaternary Research 53: 302–311.

39.

Magny

Arnaud

Holzhauser

. (2010) Solar and proxy-sensitive imprints on paleohydrological records for the last millenium in west-central Europe. Quaternary Research 73: 173–179.

40.

Marushchak

Pitkämäki

Koponen

. (2011) Hot spots of nitrous oxide emissions found in different types of permafrost peatlands. Global Change Biology 17: 2601–2614.

41.

Ménot

Burns

(2001) Carbon isotopes in ombrogenic peat bog plants as climatic indicators: Calibration from an altitudinal transect in Switzerland. Organic Geochemistry 32: 233–245.

42.

Ménot-Combes

Burns

Leuenberger

(2002) Variations of ¹⁸O/¹⁶O in plants from temperate peat bogs (Switzerland): Implications for paleoclimatic studies. Earth and Planetary Science Letters 202: 419–434.

43.

Miller

Alley

Brigham-Grette

. (2010) Arctic amplification: Can the past constrain the future? Quaternary Science Reviews 29: 1779–1790.

44.

Moberg

Sonechkin

Holmgren

. (2005) Highly variable Northern Hemisphere temperatures reconstructed from low- and high-resolution proxy data. Nature 433: 613–617.

45.

Moschen

Kühl

Rehberger

. (2009) Stable carbon and oxygen isotopes in sub-fossil Sphagnum: Assessment of their applicability for palaeoclimatology. Chemical Geology 259: 262–272.

46.

Mossberg

Stenberg

(2003) Den nya nordiska floran. Tangen: Wahlström & Widstrand.

47.

Oksanen

Kuhry

Alekseeva

(2001) Holocene development of the Rogovaya River peat plateau, European Russian Arctic. The Holocene 11: 25–40.

48.

Oksanen

Kuhry

Alekseeva

(2003) Holocene development and permafrost history of the Usinsk mire, northeast European Russia. Géographie physique et Quaternaire 57: 169–187.

49.

Raghoebarsing

Smolders

AJP

Schmid

. (2005) Methanotrophic symbionts provide carbon for photosynthesis in peat bogs. Nature 436: 1153–1156.

50.

Reimer

Baillie

MGL

Bard

. (2009) IntCal09 and Marine09 radiocarbon age calibration curves, 0–50,000 years cal BP. Radiocarbon 51: 1111–1150.

51.

Rice

(2000) Variation in carbon isotope discrimination within and among Sphagnum species in a temperate wetland. Oecologia 123: 1–8.

52.

Rozanski

Araguas-Araguas

Gonfiantini

(1993) Isotopic patterns in modern global precipitation. In: Swart

Lohmann

McKenzie

. (eds) Climate Change in Continental Isotopic Records. Geophysical Monographs Series 78: 1–36.

53.

Rydin

(1985) Effects of water level on desiccation of Sphagnum in relation to surrounding Sphagna. Oikos 45: 374–379.

54.

Rydin

Jeglum

(2006) The Biology of Peatlands. New York: Oxford University Press.

55.

Salonen

Seppä

Jones

. (2011) The Holocene thermal maximum and late-Holocene cooling in the tundra of NE European Russia. Quaternary Research 75: 501–511.

56.

Skrzypek

Kałużny

Wojtuń

. (2007) The carbon stable isotopic composition of mosses: A record of temperature variation. Organic Geochemistry 38: 1770–1781.

57.

Väliranta

Kaakinen

Kuhry

(2003) Holocene climate and landscape evolution East of the Pechora Delta, East-European Russian Arctic. Quaternary Research 59: 335–344.

58.

Virtanen

Mikkola

Nikula

. (2004) Modeling the location of the forest line in northeast European Russia with remotely sensed vegetation and GIS-based climate and terrain data. Arctic, Antarctic and Alpine Research 36: 314–322.

59.

Wanner

Beer

Bütikofer

. (2008) Mid- to later Holocene climate change: An overview. Quaternary Science Reviews 27: 1791–1828.

60.

Williams

Flanagan

(1996) Effect of changes in water content on photosynthesis, transpiration and discrimination against ¹³CO₂ and C¹⁸O¹⁶O in Pleurozium and Sphagnum. Oecologia 108: 38–46.

61.

Zanassi

Mora

(2005) Paleoclimatic implications of the relationship between oxygen isotope ratios of moss cellulose and source water in wetlands of Lake Superior. Chemical Geology 222: 281–291.

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

3.60 MB

0.71 MB