(This is a sample cover image for this issue. The actual cover is not yet available at this time.) This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright  Author's personal copy Geoderma 175–176 (2012) 9–20 Contents lists available at SciVerse ScienceDirect Geoderma journal homepage: www.elsevier.com/locate/geoderma Initial soil development and carbon accumulation on moraines of the rapidly retreating Werenskiold Glacier, SW Spitsbergen, Svalbard archipelago C. Kabala ⁎, J. Zapart Institute of Soil Science and Environmental Protection, University of Environmental and Life Sciences, ul. Grunwaldzka 53, 50-357 Wroclaw, Poland a r t i c l e i n f o a b s t r a c t Article history: The rapid retreat of the Werenskiold Glacier (the Svalbard archipelago, High Arctic) is leading to an extensive Received 21 April 2011 broadening of the proglacial zone covered with recent moraine till on older glacigenic deposits or directly on Received in revised form 8 January 2012 bedrock schists. To study the type and intensity of initial soil development under a harsh periglacial climate, a Accepted 15 January 2012 chronosequence of six soils was established between the glacier front and its terminal moraine on 1 to about Available online xxxx 80 year-old moraines. Although the surface layer over the entire area of study is frost-active, the patterned Keywords: features are not well developed. The succession of vegetation, mainly Saxifraga sp. and lichens, starts Deglaciation 5–6 years after deglaciation, successively covering up to 30% of the soil surface, then stagnating . Present- High Arctic day soil-forming processes within the uppermost soil layer comprise initial weathering of primary minerals Cryosols (chlorites and amphiboles), carbonate dissolution and base cation leaching associated with pH lowering, Soil formation accumulation of organic matter and nitrogen, and an increase in pedogenically-derived Fe. Soil development Organic carbon measures are time-related and, in general, fit a logarithmic model. The intensity of transformation, including Soil chronosequence organic carbon and nitrogen accumulation, started at high rates comparable to those reported in Low Arctic and Alpine environments; however, in the fourth/fifth decade after deglaciation it reached a quasi steady- state. Low annual precipitation is probably a crucial factor that controls plant succession and leaching of carbonates, thus limiting mineral weathering, organic matter accumulation and soil development on the Werenskiold moraines. © 2012 Elsevier B.V. All rights reserved. 1. Introduction Tang (2008). Regular recession of subarctic glaciers on Spitsbergen (the largest island of the Svalbard Archipelago, the High Arctic) Global climate warming significantly influences the functioning of began at the end of 19th century or in first decades of 20th century. glacier systems on all continents. The most often reported symptoms Unequivocal assessment of the starting point of glacial retreat in of the ongoing changes are: the shortening of mountain glacier this area is highly problematic as the tongues of many glaciers reach tongues (Barry, 2006), reduction of the mass of arctic and subarctic the tidewater and deglaciation rates were initially low. Significant glaciers (Hagen et al., 2006), as well as gradual deepening of the shortening of glacier snouts in the High Arctic was confirmed in the active layer in areas of permafrost occurrence (Humlum et al., 2003; second and third decades of 20th century (Ziaja, 2001). The rate of Rachlewicz and Szczucinski, 2008). Contrary to widespread popular deglaciation in the Svalbard archipelago has accelerated rapidly in the opinion, glacier regression did not start in recent decades, but has last two decades, up to 700% in some cases. Based on new original occurred for a minimum 100–200 years and followed the last period calculations, Rachlewicz et al. (2007) distinguished four types of of relatively colder climate — the Little Ice Age (LIA). Glaciers of glaciers differing in deglaciation rate: very dynamic, surging tidewater southern Alaska reached their maximum extent at the end of 18th glaciers with post-LIA retreat rates of between 100 and 200 m a − 1, century, documented by the oldest moraines on the forefield of the other tidewater glaciers receding at a rate of 15 to 70 m a − 1, land Mendenhall Glaciers, assessed to be no younger than 240 years old terminating valley polythermal glaciers with an average retreat of 10 (Alexander and Burt, 1996). The recession of the Hailogou Glacier in to 20 m a − 1, and small, usually cold glaciers with retreat rates below the Gongga Mountains (Sichuan province, south China) and forma- 10 m a− 1. tion of neoglacial moraines started in 1820, as reported by He and Uncovering new land is an environmentally important result of glacier retreat. According to Rachlewicz et al. (2007), the total surface of glaciers in the central part of Spitsbergen has decreased by between 5 and 53.5% during the 20th century, and the area of recently ⁎ Corresponding author. Tel.: + 48 71 3201943; fax: +48 71 3205631. uncovered ground and rock surfaces has increased by several E-mail address: cezary.kabala@up.wroc.pl (C. Kabala). thousand hectares. Cannone et al. (2008) expect that most small 0016-7061/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.geoderma.2012.01.025  Author's personal copy 10 C. Kabala, J. Zapart / Geoderma 175–176 (2012) 9–20 glaciers in the Alps (80% of total glacial coverage and an important moraine, formed probably during a large and rapid glacial surge at contribution to water resources) will disappear in the next few the beginning of the 20th century. Both within the frontal and lateral decades. Therefore, recently deglaciated areas have increasingly moraines there are well-preserved glacial cores, making the moraine attracted the interest of the scientific community, because soils on surface thermoactive in many places and causing frequent landslides glacier forefields offer unique possibilities to study biogenic and on their slopes (Karczewski et al., 1984). abiogenic processes and their interactions induced by accelerated The Werenskiold Glacier basin is located in the contact zone of three climate changes, yet still under extreme conditions. The rate of glacier tectonic blocks of the Caledonian Hecla Hoek formation (Czerny et al., transgression and regression (thawing) plays a key role in postglacial 1993). The southern neighborhood of the glacier is built of metamorphic landscape formation (Karczewski, 1982; Lonne and Lysa, 2005). It Proterozoic amphibolites, quartzites, chlorites and amphibolite-quartzite generally decides on the type and quantity of non-cemented deposits schists. The eastern surroundings of the accumulation field consist of in the proglacial zone. Moreover, the rate of deglaciation influences phyllites, laminated schists and quartzite conglomerates with dolomite the local microclimate, water drainage patterns, the intensity of precipitations. The Jens Erikfjellet formation, which surrounds the glacier erosion, occurrence of cryogenic phenomena in the ground, basin from the north-west, is built predominantly of greenschists and succession of plant communities, as well as soil evolution mica-calcite-quartzite schists. The rock bed at the glacier terminus, in (Haugland, 2004; Ziaja, 2004). Ecological changes in the forefields of the area exposed from ice cover, consists of Precambrian chlorite-mica- retreating glaciers are mostly characterized in the context of the quartzite schists, marbles and marble–quartzite conglomerates (Czerny succession of plant species or communities (Burga et al., 2010; et al., 1993). Due to the glacial mixing of these materials, the soil parent Cannone et al., 2008; Frenot et al., 1998) and microorganisms material is considered to be polymineral and relatively homogenous in (Kastovska et al., 2005). Reports focusing on soil development or particular sedimentary zones of the proglacial area. Petrographical transformations of soil properties in deglaciated areas are currently analyses of pebbles in the southern part of the forefield documented 5– rare and concerned mainly with seasonal or long-term changes in 20% of quartz and quartz–mica aggregates, 20–30% of quartzites, 12– soil respiration and its influence on regional or global carbon turnover 21% of chlorite schists, 5–12% of mica schists, 1–9% of amphibolites, and as well as on the climate (e.g. Bekku et al., 2004). A limited number of 23–38% of marbles and marble–quartzite conglomerates (Kowalska and research papers have focused on soil development in the context of Sroka, 2008). Quartzites are not abundant in the source areas, thus — as weathering processes and the physical transformation of glacial the most resistant to weathering — are probably over-represented in deposits (e.g. Bernasconi and BigLink Consortium, 2008; Burt and the pebble fraction. Due to the regular occurrence of marbles in bedrock, Alexander, 1996; Egli et al., 2001; Haugland, 2004; Mavris et al., the “fresh” glacial sediments always contain calcium carbonate and have 2010), and only a few papers involve the problems of the morphological an alkaline reaction (Bukowska-Jania, 2007). development of the soil profile and discuss the classification of soils Long-term climate data for the Hornsund area have been collected currently developing in specific, near-glacial environments (Alexander by the permanent polar observatory (Polish Polar Station) at Isbjorn- and Burt, 1996; He and Tang, 2008; Jacobson and Birks, 1980). hamna, located ca. 12 km south-east of the Werenskiold Glacier. Extremely rare are reports that present soil chronosequences in the Mean annual precipitation (water equivalent) in the period High Arctic (Mann et al., 1986), probably because of the relatively 1979–2006 was about 430 mm, relatively high when compared to young age and superimposed weak transformation of uncovered glacial other records from Spitsbergen. Mean monthly precipitation ranges deposits. from 20 mm in May to 63 mm in September (Fig. 1). The mean annual The objectives of the present study in the forefield of the Werenskiold air temperature was − 4.4 °C, with mean monthly air temperatures Glacier were therefore (i) to examine how the morphology and ranging from −11.3 °C in January to 4.4 °C in July. A distinct upward properties of initial soils develop during the first decades after trend of mean annual air temperature as high as +0.095 °C/year has deglaciation; and (ii) to quantify changes in soil properties, including been detected during the last three decades of measurements (Marsz organic carbon accumulation in the upper layer of the soil profile. and Styszynska, 2007). Soils have a gelic temperature regime, with a mean annual temperature − 4.7 °C at a depth of 50 cm below the 2. Study site and methods soil surface. Mean summer (June–August) and winter (December– February) soil temperatures are 1.8 °C and −10.3 °C, respectively 2.1. Site characteristics (Leszkiewicz and Caputa, 2004). Climate conditions in the foreland of the Werenskiold Glacier are considered to be slightly milder than The study area is in the proglacial foreland of the Werenskiold those in the direct proximity of the polar station at Isbjornhamna Glacier (77°05′ N, 15°15′ E), located in Wedel-Jarlsberg Land in (Migała et al., 2008). south-west Spitsbergen, the largest island of the Svalbard Archipelago in the High Arctic. The Werenskiold Glacier is a typical land-based, polythermal glacier of approximately 27.4 km 2 in area and 9.5 km in length (Hagen et al., 1993). Previous studies focused mostly on mass balance and geometry of the glacier (e.g. Baranowski, 1977), and hydrological (e.g. Krawczyk, 1992), climatic (e.g. Migała et al., 2008), geologic (e.g. Czerny et al., 1993), geomorphologic (e.g. Karczewski et al., 1984), and geochemical (e.g. Bukowska-Jania, 2007) aspects, while the study of soil development on the glacier's foreland and adjacent areas still remains at an initial stage (Piroznikow and Gorniak, 1992; Szerszen, 1968). The southern part of its proglacial zone is covered with a fluted moraine, changing towards the west into a flat moraine. Moraine tills rest immediately on bedrock schists or glacio-fluvial sands and gravels. The thickness of the neoglacial till does not in general exceed 50 cm. The northern part of the foreland is predominantly covered with glaciofluvial sediments forming a mosaic of plains and fans. Fig. 1. Mean monthly air temperature and precipitation at the Polish Polar Station The proglacial zone of the glacier is delimited in the north and (Isbjornhamna-Hornsund, south-west Spitsbergen) in the period 1979–2006. south by lateral moraines, and in the west by a terminal (frontal) Based on: Marsz and Styszynska, 2007.  Author's personal copy C. Kabala, J. Zapart / Geoderma 175–176 (2012) 9–20 11 2.2. Soil sampling the soils was made based on the FAO World Reference Base (IUSS, 2007). Sampling depth was different in particular profiles depending A preliminary investigation, preceding the final selection of on the individual pedon features; however, the first, uppermost sampling sites, revealed heterogeneous site conditions which are sample was always taken at the depth of 0–3 cm. Before sampling, typical on glacier forelands and demonstrated by several soil parame- the soil surface was carefully cleaned of lichens and higher plants to ters: proportion of rock fragments, soil texture and wetness, in particu- avoid an overestimation of organic matter content. All samples were lar. It was decided to include only the sites on typical moraine till into air-dried and crushed to sieve >2 mm and ≤2 mm size fractions. the present study, while soils on outwash sands and gravels, as well as glaciofluvial and limnic sediments were rejected. The soil pits were 2.3. Soil analysis placed on relatively well-drained sites, never in local depressions, channels, or on steep slopes. Finally, six sites with increasing distance Soil samples were analyzed for texture and basic chemical properties from the glacier were selected, considering that each sampling site using laboratory methods applied to soil classification purposes (Van represents a different period and stage of soil development. Soil Reeuwijk, 2002). Particle size distribution of the ≤2 mm fraction, after formation on the moraines began at a time of stabilization of the removing the organic matter and sample dispersion with heksameta- deposits and has been continued to the present. The age of parent phosphate–bicarbonate solution, was conducted using sand separation material is therefore considered to approximate the period of soil on sieves and the hydrometer method to fine earth fractions development. The estimated ages of moraine materials in the proglacial (b0.1 mm). Soil pH was potentiometrically measured in a 1:2.5 (soil: zone were assessed based on numerous previous delineations of the distilled water) suspension. Calcium carbonate was measured by the glacier snout position (e.g. Baranowski, 1977; Bukowska-Jania, 2007; gas volumetric method. Total content of organic carbon was determined Karczewski et al., 1984; Piroznikow and Gorniak, 1992). by dry combustion using an automatic analyzer (Ströhlein CS-mat 5500) The soil chronosequence was located on the southern foreland of and total nitrogen — by the standard Kjeldahl technique. Exchangeable the Werenskiold Glacier (Fig. 2). Four sampling sites were placed on ions (Ca2 +, Mg2 +, K+, Na+) were extracted initially with 1 M a typical fluted moraine (Table 1), one site (WER10) on a flat ammonium acetate at standard pH =7. However, due to unrealistically moraine, and the last one (WER17) in the western (marginal) part high calcium and magnesium concentrations (resulting from carbonate of the end (terminal) moraine. All the studied sites were located dissolution), extraction with 1 M NH4Cl at pH =8.2 was additionally within a direct distance of 1650 m. The difference in altitude between applied. Iron in non-silicate forms (“free” iron, Fed) was extracted with sites did not exceed 50 m, in the range 25–75 m asl. a bicarbonate–dithionite–citrate buffer (BDC) according to the standard Basic geomorphic information and vegetation cover were Mehra–Jackson method, and iron in amorphous oxides and hydroxides recorded at each site. All the soil pedons were described according (“active” iron, Feo) was extracted with acid ammonium oxalate to FAO guidelines (FAO, 2006). Soil texture, structure, consistence, according to the standard Tamm method (Van Reeuwijk, 2002). The color, redoximorphic features and root abundance in all the derived concentration of metallic elements in all extracts was measured by layers of the soil profile were noted (Table 2). An attempt to classify atomic absorption spectroscopy (Phillips-Unicam AAS). Fig. 2. Location of soil profiles in the forefield of the Werenskiold Glacier. 1 — Mountains without ice cover, 2 — terminal and lateral moraines, 3 — soil pits.  Author's personal copy 12 C. Kabala, J. Zapart / Geoderma 175–176 (2012) 9–20 Table 1 Localization of pedons; landforms, ages, kind of parent materials, and vegetative cover of the soils. Site Distance to Surface age Landform Parent material Bare Structure of vegetation glacier (meters) (years) ground (%) WER4 20 1 Fluted Thin layer of till directly on bedrock schists 100 No vegetation moraine WER6 150 6 Fluted Thin layer of till on bedrock schists 95 Saxifraga oppositifolia (50%a), Saxifraga caespitosa (50%), moraine single individuals of Poa arctica WER7 280 12 Fluted Thin layer of till on bedrock schists 90 S. oppositifolia (40%), S. caespitosa (10%), lichens (50%) moraine WER10 450 30 Flat Moraine till on older stratified outwash 75 S. oppositifolia (50%), S. caespitosa (5%), lichens (45%) moraine gravels; permafrost at 90 cm WER16 750 45 Fluted Till on older stratified till and outwash 70 S. oppositifolia (60%), S. caespitosa (20%), lichens, single moraine sands; permafrost at 95 cm individuals of P. arctica and Cerastium alpinum WER17 1650 70–80 Terminal Deep moraine till; permafrost at 125 cm 70 S. oppositifolia (70%), S. caespitosa (10%), lichens, single moraine individuals of P. arctica and C. alpinum a Percentage of surface covered with vegetation. Soil organic carbon and nitrogen stocks in the upper 15 cm of soil initial humic A horizon having a thickness of up to 3 cm in soils were calculated according to the following equation (Mavris et al., 12–45 years old and 5–6 cm in soils older than 70 years. The initial 2010): subsurface B horizon may be distinguished only in better drained soils on the frontal moraine, on materials older than 70 years. Initial Cstock ðNstock Þ ¼ ∑ci hi di ð1−RMÞ; B horizons are slightly more brown colored than the subsoil (have a 10Y Munsell color hue instead of 5GY, and chrome 2 instead of 1), where Cstock (Nstock) denotes organic C (total N) abundance have a weak to moderate subangular blocky structure, and are usually (kg m − 2), ci — concentration of C or N (kg t − 1), hi — thickness of i interlaced with plant roots throughout. These B layers, however, do layer (m), di — bulk soil density (t m − 3), and RM the mass proportion not fulfill the requirements of diagnostic horizon cambic (IUSS, of rock fragments. 2007), due to insufficient thickness and too little differentiation of Mineralogical composition of the clay fraction (b0.002 mm) was color when compared to the subsoil. determined by x-ray diffractometry (XRD). The clay fraction was Soils derived from very shallow till on hard rocks in the forefield of separated by centrifugation, followed by ultrasonic sample dispersion. the glacier were classified as Leptosols (Table 2, pedon WER 4). More To remove organic matter, all clay samples were treated with a 10% problematic is classification of soils developed from till resting on H2O2 solution. Initially, air dried specimens were x-rayed. The Mg2 +- hard rock at the depth of 45–50 cm below ground level. The entire saturated slides were treated with glycerol and rescanned. The K +- volume of till thaws during the polar summer, thus diagnostic horizon treated slides were subsequently heated to 300° C and rescanned. The cryic does not occur (Bockheim et al., 2006). However, features of slides were finally heated again to 550°C and rescanned. Diffraction cryoturbation are visible, particularly as separation of coarse particles patterns were obtained using CuKα radiation and step-scanning within the soil and on its surface. Soils are seasonally saturated with between 3 and 40° 2θ, using 0.05° 2θ increments with a 3-s counting melting water and are “gleyed”, however, they do not meet the time. The differential thermal gravimetry (DTA-DTG) method was requirements of “gleyic color pattern” (IUSS, 2007) as the dominant additionally used to allow for the differentiation of primary versus hue of soil color is 10GY. Finally, soils represented in pedons WER6 pedogenic minerals (detailed results were not shown, only the final and WER7 (Table 2) were classified as Leptic Regosols (Calcaric, Turbic, determination). Oxyaquic, Skeletic). All other soils, developed of deep till or till on sand The Statistica 8.0 software system (StatSoft Inc., Tulsa, OK) was both having a permafrost layer, were classified as Turbic Cryosols used to test the statistical significance of trends in changes of soil (Table 2: pedons WER10, WER16, and WER17). properties. Various linear and nonlinear models were tested, and In the pedon WER16, under the thin cover of young moraine, a the model yielding the highest correlation coefficients (at least 0.05 unique, complete profile of buried soil was found. It consists of well confidence level) was chosen as the best chronofunctions. developed A and B horizons whose cumulative thickness exceeds 25 cm. The subsurface Bw layer meets the requirements of diagnostic 3. Results horizon cambic (IUSS, 2007), and the buried soil represents typical “arctic brown earths” (Tedrow and Hill, 1955) or “rusty tundra 3.1. Morphology features and classification of soils soils” (Kowalkowski, 1998), which corresponds to Haplic Cambisols (IUSS, 2007). The present soil (complete pedon WER16) must be, The youngest till in the forefield of the Werenskiold glacier is, in however, classified as Cryosols due to the occurrence of permafrost general, only 20–40 cm thick and rests directly on bedrock schists in subsoil and cryoturbation features at the soil surface (Bockheim or on older glaciofluvial gravels and sands. The permafrost layer et al., 2006; IUSS, 2007). Due to unfavorable climate conditions, started at a depth of 90–100 cm below ground level, provided no Cambisols are not expected to occur in the High Arctic (Goryachkin hard rock occurred. The thickness of till was greater on the lateral et al., 2004). Thus, if identified on older Holocene surfaces, are and frontal moraines, reaching 100–130 cm to contact with the probably relic soils (Kabala and Zapart, 2009). permafrost layer. Recent till in the proximity of the glacier snout is greenish gray (10GY 5-6/1), massive (structureless in a pedological 3.2. Particle-size distribution sense) and gleyed throughout due to prolonged saturation with melting water (Table 2, Pedon WER4). The first morphological The share of individual particle-size fractions within the profiles of features of soil development are changes in the soil color to dark soils developed from till varies only insignificantly in terms of space greenish gray (10Y 4/1) due to an initial accumulation of organic and depth. The skeleton fraction (fine and medium gravel) usually matter, and formation of a weak subangular blocky (pedogenic) forms 40–60% of the overall soil mass. The content of the silt fraction structure. These transformations lead to the distinguishing of the (0.002–0.05 mm) is in the range of 40–45% of the fine-earth fractions  Author's personal copy C. Kabala, J. Zapart / Geoderma 175–176 (2012) 9–20 13 and the clay fraction (b0.002 mm) is usually in the range of 9–15% 3.3. Carbonates, pH and exchangeable cations (Table 3). The common texture class of soils developed on till is there- fore loam (according to the USDA classification). The only exception is The dissolution and leaching of carbonates is reported to be the pedon WER17 located on the end moraine, where sandy loam was found first feature of the chemical transformation of glacial sediments in the upper soil layers due to the lower content of clay (6–7%) and silt (Bukowska-Jania, 2007). All soils developed on the recent glacial (25–36%) fractions. Decrease in the soil skeleton due to the duration of sediments contain variable, but considerable amounts of CaCO3 (up physical weathering was weak, if any. The mean content of rock frag- to 5%). Significant diversification of CaCO3 content was observed in ments within the layer 0–15 cm changed from 42% in WER4 soils to sediments which are 10–12 years old (pedon WER7) but only in the 38% in WER10, however, it exceeded 50% in older soils. Despite nearly upper 0–3 cm layer, while in the 45–70 year old formations uniform texture, a decrease in the clay fraction in the uppermost soil (WER16) the leaching of CaCO3 reaches the depth 12–14 cm from layer (0–3 cm) is distinctly visible (Fig. 3a) when compared to the chron- the soil surface (Fig. 4). However, a temporal trend in CaCO3 leaching osequence from the glacier snout (15% of clay in pedon WER4) to the ter- is not evident, and excluding the uppermost layer of pedon WER16, minal moraine (5% of clay in pedon WER17). the CaCO3 content was at least as high as 2.3%. Only the buried Table 2 Soil morphology and profile descriptions. Horizon Depth (cm) Color (moist) Texture, rock fragments Structure Consistence (moist) Carbonates Roots Boundary Pedon WER4: Haplic Leptosol (Calcaric, Oxyaquic, Skeletic) age of glacial till — 1 year Cg1 0–3 10GY 6/1 L, MC — A MA FR MO N G Cg2 3–18 10GY 5/1 L, MC — M MA FR MO N A R 18 + Hard, fractured bedrock schists Pedon WER6: Leptic Regosol (Calcaric, Turbic, Oxyaquic, Skeletic) age of glacial till — 6 years Cg1 0–3 10GY 5/1 L, MC — A PM FR MO VF G Cg2 3–12 10GY 5/1 L, MC — M PM/PL, WE FR MO VF G Cg3 12–30 10GY 5/1 L, ME — M PL,WE FR MO N C Cg4 30–45 10GY 5/1 L, MC — A MA FR MO N C R 45 + Hard, fractured bedrock schists Pedon WER7: Leptic Regosol (Calcaric, Turbic, Oxyaquic, Skeletic) age of glacial till — 12 years A 0–3 10Y 4/1 L, MC — A SB, WE FR MO F G, W AC 3–6 5GY 5/1 L, MC — M SB, WE FR MO F G Cg1 6–15 10GY 5/1 L, MC — M AB, WE FR MO VF G Cg2 15–30 10GY 5/1 L, MC — M AB/PL, MO FR MO N G Cg3 30–45 10GY 5/1 L, MC — A AB, MO FR MO N C R 45 + Hard, fractured bedrock schists Pedon WER10: Turbic Cryosol (Calcaric, Oxyaquic, Skeletic) age of glacial till — 30 years (glacial till on older stratified outwash gravels) A 0–3 5 G 4/1 L, MC — M SB, WE FR MO F G, W AC 3–8 5 G 5/1 L, MC — M SB/PL, WE FR MO VF G Cg 8–20 5 G 5/1 L, MC — M PL, WE FR MO N A 2C 20–90 2.5Y 3/2–3 S, MC — A/D SG LO MO N C, W Stone lines at the depths 35–44 cm and 60–70 cm I 90 + Permafrost layer Pedon WER16: Thaptocambic Turbic Cryosol (Calcaric, Oxyaquic, Skeletic) age of glacial till — 45 years (glacial till on older stratified glacial till and outwash gravels) A 0–3 10Y 4/1 L, MC — A SB, WE FR SL C G AC 3–10 10GY 5/1 L, MC — A PL, WE FR MO F G Cg 10–14 2.5Y 5/3 SiL, FI — V MA FI MO VF A, S No pavement at the boundary of horizons 2Ab 14–20 2.5Y 4/3 SL, MC — M SB, ST (fine) FR SL C G 2ABb 20–26 2.5Y 3/3 SL, MC — M AB, MO FR SL C G 2Bwb 26–40 2.5Y 5/4 SL, MC — M AB/PL, MO FR SL F C Thin iron pan at the depth ca 35 cm Stone pavement at the boundary of horizons 3C 40–70 2.5Y 6/6 SL/S, MC — D MA/SG FI/LO SL VF C 4Cx 70–90 2.5Y 5/5 SL, MC — A AB/PL, ST FI SL N C,W I 90 + Permafrost layer Pedon WER17: Turbic Cryosol (Calcaric, Oxyaquic, Skeletic) age of glacial till — more than 70 years A 0–6 10Y 4/1 SL, MC — A SB, MO FR MO C G ABg 6–12 10Y 5/2 SL, MC — A SB, MO FR MO C G BCg 12–22 5GY 5/2 L, MC — A SB, MO FR MO F G Cg1 22–40 5GY 5/1 L, MC — A AB, MO FR MO VF G Cg2 40–125 5GY 4/1 L, MC — A AB/PL, MO FR MO N A, W I 125+ Permafrost — “ice nucleus” of the moraine Textural classes: S — sand, SL — sandy loam, L — loam, SiL — silt loam; dominant size of rock fragments: FI — fine (2–6 mm), ME — medium (6–20 mm), MC — medium and coarse gravel (6–60 mm); abundance of rock fragments: V — very few (0–2%), F — few (2–5%), C — common (5–15%), M — many (15–40%), A — abundant (40–80%), D — dominant (> 80%); types of soil structure: SG — single grain, MA — massive, PM — porous massive, AB — angular blocky, AS — angular and subangular blocky, SB — subangular blocky, PL platy; grade of development: WE — weak, MO — moderate, ST — strong; consistence (moist): LO — loose, FR — friable, FI — firm; carbonates content: SL — slightly calcareous (b 2%), MO — moderately calcareous (2–10%), abundance of roots: N — none, V — very few, F — few, C — common; horizon boundary (distinctness, topography): A — abrupt, C — clear, G — gradual, S — smooth, W — wavy.  Author's personal copy 14 C. Kabala, J. Zapart / Geoderma 175–176 (2012) 9–20 Table 3 Particle-size distribution and chemical characteristics of soils in the foreland of the Werenskiøld Glacier. Horizon Depth Particle size distribution, % pHH2O CaCO3 Organic carbon Total N Feo Fed Feo/ (cm) (particle diameters in mm) (%) Fed > 2a 2.0–0.05b 0.05–0.002b b0.002b g kg− 1 Site WER4 Cg1 0–3 50 40 45 15 8.35 3.0 4.5 0.2 4.7 10.4 0.46 Cg2 3–10 42 40 44 16 8.36 3.3 4.6 0.2 4.6 10.3 0.45 Cg2 10–18 38 40 45 15 8.38 3.6 4.2 n.d. 4.6 10.3 0.45 Site WER6 Cg1 0–3 45 42 45 13 8.12 2.7 6.4 0.3 5.1 10.5 0.49 Cg2 3–12 38 42 43 15 8.35 3.5 5.2 0.2 4.9 10.2 0.48 Cg3 12–30 35 41 43 16 8.32 3.8 4.5 n.d. 4.8 10.1 0.48 Cg4 30–45 49 49 39 12 8.33 3.8 3.9 n.d. 4.4 9.5 0.46 Site WER7 A 0–3 47 54 35 11 7.95 2.9 10.4 0.8 5.6 10.8 0.52 AC 3–6 38 49 39 12 8.13 4.6 5.0 0.3 5.3 10.7 0.50 Cg1 6–15 39 43 45 12 8.29 4.5 4.3 n.d. 5.0 9.5 0.52 Cg2 15–30 33 42 45 12 8.34 4.6 5.1 n.d. 4.2 8.9 0.47 Cg3 30–45 47 41 45 14 8.29 4.9 4.8 n.d. 4.6 9.7 0.47 Site WER10 A 0–3 42 45 45 10 7.82 3.2 11.0 1.1 6.9 12.4 0.56 AC 3–8 38 41 45 14 8.20 4.1 5.6 0.4 6.7 12.7 0.53 Cg1 8–20 37 44 42 14 8.41 4.6 5.5 n.d. 5.9 11.9 0.50 2C 20–33 72 91 7 2 8.36 1.6 1.6 n.d. 1.7 13.5 0.13 2C 50–60 82 89 8 3 8.33 1.5 1.4 n.d. 1.5 8.7 0.17 2C 70–80 60 90 8 2 8.35 1.8 1.2 n.d. 0.8 11.2 0.07 Site WER16 A 0–3 55 44 47 9 7.73 0.9 13.3 1.1 7.0 13.1 0.53 AC 3–10 50 40 48 12 8.20 2.8 5.7 0.5 6.8 13.2 0.52 Cg 10–14 6 33 62 5 8.25 3.6 4.5 0.4 6.3 13.5 0.47 2Ab 14–20 37 69 27 4 7.31 1.7 14.6 1.7 8.7 14.8 0.59 2ABb 20–26 37 62 24 4 7.22 1.6 13.1 1.2 7.3 14.3 0.51 2Bwb 26–40 36 68 29 3 7.08 0.8 4.3 n.d. 3.9 15.5 0.25 3C 40–70 80 67 31 2 7.09 0.6 3.3 n.d. 2.5 14.0 0.18 4Cx 70–90 81 68 30 2 7.21 1.8 1.6 n.d. 2.3 11.0 0.21 Site WER17 A 0–6 55 58 35 7 7.74 2.3 19.2 1.7 7.2 12.3 0.59 ABg 6–12 58 58 36 6 7.92 2.6 6.7 0.6 6.3 11.1 0.57 BCg 12–22 60 68 25 7 8.16 3.3 4.4 0.4 4.7 9.2 0.51 Cg1 22–40 55 51 40 9 8.23 3.3 4.0 n.d. 4.7 10.2 0.46 Cg2 40–70 60 49 41 10 8.23 3.2 4.4 n.d. 4.3 9.5 0.45 Cg2 70–90 60 48 41 11 8.24 4.1 4.9 n.d. 4.4 10.9 0.40 a Percentage of bulk soil sample. b Percentage of fine earth fractions (b 2 mm). older outwash gravels, found in the lower part of pedons WER10 and Ammonium oxalate at pH = 7 extracted in such soils ca. 5-fold higher WER16, contained less CaCO3, namely 0.6–1.8% (Table 3). amounts of calcium, twofold more magnesium, 10-fold more Despite the relatively short time of transformation, the change in potassium and twofold more sodium when compared to extraction soil pH was evident (p b 0.05). Initial pH of the recent moraine till at pH = 8.2 (Table 4). The calcium is therefore “extractable” or was about 8.3–8.4 throughout the till layer. The value of pH in the “soluble” rather than “exchangeable” in these soils. Decrease in the uppermost 0–3 cm soil layer falls below 8.0 already in 12-year old total sum of base cations observed in the uppermost layers of soil in sediments and below 7.8 in 30-year-old sediments. The pH value the chronosequence is correlated with the dissolution of carbonates decreased logarithmically with time (Fig. 3b); however, it tended to and leaching of calcium ions. Change in the vertical distribution of stabilize at pH = 7.7 in sediments older than 40 years due to the “exchangeable” Ca in particular soil profiles was more evident than presence of carbonates. The acidification of a deeper layer (3–6/ differences in carbonate content or pH. While in the youngest till 8 cm) was very weak: although the pH value decreased below 8.2 in (pedon WER4) Ca content increased towards the soil surface, in 12-year-old sediments, the level of 8.0 was reached only in soils sediments 6-year old and older the calcium content significantly older than 70 years (Table 3). No significant acidification was decreased towards the surface (Table 4, extraction with ammonium observed in moraine till below the depth of 10 cm within the chloride). The loss of “exchangeable” Ca in the layer 0–3 cm reached observed time span. 23% of the initial Ca level after 6 years of leaching, 45% after The overall sum of exchangeable base cations (Ca 2 +, Mg 2 +, K +, 12 years and 57% after 70 years of cation leaching, whereas the loss + Na ) in the soils under investigation is influenced primarily by the of CaCO3 reached maximally 30% (in the same horizons). This time- large amounts of calcium released easily from CaCO3, which explains dependent trend was however not significant statistically, as some the differences between its amount when extracted at pH = 7 (with medium-age soils (pedons WER 10 and 16) revealed a smaller loss ammonium oxalate) and at pH = 8.2 (with ammonium chloride). of Ca than expected on the basis of their age.  Author's personal copy C. Kabala, J. Zapart / Geoderma 175–176 (2012) 9–20 15 Table 4 Comparison of exchangeable ions extracted with NH4OAc pH 7.0 and NH4Cl pH 8.2. Horizon Depth Exchangeable ions Sum Exchangeable ions Sum (cm) extracted with NH4OAc extracted with NH4Cl pH 7.0 (cmol[+] kg− 1) pH 8.2 (cmol[+] kg− 1) Ca Mg K Na Ca Mg K Na Site WER4 Cg1 0–3 30.0 3.83 0.70 0.20 34.7 6.64 1.39 0.16 0.16 8.4 Cg2 3–10 34.4 4.05 0.81 0.31 39.6 6.42 1.45 0.18 0.12 8.2 Cg2 10–18 30.4 4.02 0.73 0.30 35.5 6.00 1.60 0.21 0.12 7.9 Site WER6 Cg1 0–3 24.0 0.68 3.02 0.27 28.0 4.74 0.99 0.13 0.14 6.0 Cg2 3–12 26.4 0.73 2.54 0.28 29.9 6.16 1.19 0.15 0.14 7.6 Cg3 12–30 28.4 0.67 2.72 0.34 32.1 6.20 1.38 0.19 0.16 7.9 Cg4 30–45 26.8 0.63 2.36 0.24 30.0 5.98 1.02 0.11 0.10 7.2 Site WER7 A 0–3 12.9 5.62 0.46 0.10 19.1 3.30 0.95 0.06 0.28 4.6 AC 3–6 18.7 2.78 0.55 0.13 22.2 4.48 1.36 0.08 0.12 6.0 Cg1 6–15 26.8 3.96 0.62 0.18 31.6 5.26 1.15 0.11 0.16 6.7 Cg2 15–30 25.2 3.83 0.57 0.17 29.8 5.50 1.13 0.12 0.16 6.9 Cg3 30–45 25.6 3.26 0.82 0.26 29.9 5.94 1.19 0.13 0.14 7.4 Site WER10 A 0–3 27.2 0.91 1.64 0.06 29.8 5.38 1.84 0.12 0.26 7.6 AC 3–8 34.8 0.75 1.53 0.05 37.1 6.32 1.57 0.14 0.19 8.2 Cg1 8–20 37.6 0.62 1.40 0.05 39.7 6.62 1.66 0.16 0.23 8.7 2C 20–33 35.6 0.76 1.35 0.05 37.8 5.56 1.41 0.03 0.17 7.2 2C 50–60 32.8 0.83 1.86 0.03 35.5 4.80 1.08 0.02 0.16 6.1 2C 70–80 26.8 0.65 1.35 0.05 28.8 5.78 1.64 0.05 0.33 7.8 Site WER16 A 0–3 18.7 1.58 0.46 0.24 21.0 4.94 1.35 0.14 0.14 6.5 AC 3–10 21.2 2.68 0.68 0.21 24.8 5.72 1.53 0.15 0.10 7.5 Cg 10–14 29.2 5.10 0.67 0.16 35.1 5.00 0.91 0.12 0.07 6.1 2Ab 14–20 36.8 6.97 0.82 0.04 44.6 1.82 0.89 0.04 0.02 2.8 Fig. 3. Change in clay content (a) and soil pH (b) in the uppermost soil layer (0–3 cm) 2ABb 20–26 32.4 5.10 0.75 0.09 38.3 0.88 0.65 0.06 0.05 1.6 with time of soil development. 2Bwb 26–40 37.6 6.05 1.02 0.13 44.8 0.72 0.48 0.07 0.04 1.3 3C 40–70 39.6 4.71 0.92 0.11 45.3 0.38 0.34 0.05 0.03 0.8 4Cx 70–90 37.6 10.10 0.98 0.23 48.9 1.32 1.02 0.03 0.02 2.4 Enhanced Ca leaching narrowed the proportion of “exchangeable” Ca to Mg, which in recent till had an initial value 4.5 or more, and Site WER17 A 0–6 27.2 2.55 0.52 0.26 30.5 2.62 1.03 0.17 0.24 4.1 decreased with age to 2.5. A similar change was found in the ABg 6–12 31.2 1.75 0.59 0.16 33.7 3.92 0.80 0.13 0.09 4.9 relationship of “exchangeable” Ca to K, which decreased from the BCg 12–22 33.2 1.82 0.44 0.18 35.6 4.54 0.79 0.13 0.10 5.6 level of 38–40 to 15–20 (in the soil layer 0–3 cm). The most spectacular Cg1 22–40 34.3 2.56 0.38 0.17 37.4 4.60 1.26 0.16 0.14 6.2 change in the proportion of exchangeable cations was however found Cg2 40–70 38.6 2.50 0.27 0.18 41.5 6.68 1.07 0.16 0.12 8.0 Cg2 70–90 37.8 3.42 0.15 0.16 41.5 5.88 1.16 0.16 0.10 7.3 between Ca and Na, which decreased from 42–45 in recent till to about 10 in 70-year old sediments. It resulted both from Ca loss and from significant increase in Na content in the layer 0–3 cm. 3.4. Soil organic carbon and soil nitrogen Accumulation of organic carbon in the uppermost 3 cm of soil showed a polynomial increase with time over the observed period of time (Fig. 5a). Its concentration in fine earth fractions ranged from 4.5 g kg − 1 in the newly deglaciated till (throughout till layer), to 19.2 g kg − 1 in the oldest soil. Increase in organic C concentration was observed mainly within the uppermost 3 cm of the soil, and a weak increase was initiated at greater depths (3–8/10 cm) starting from the fourth decade of soil development (Table 3). Organic C stocks, calculated for the upper 15 cm of soils, increased gradually with soil development, but rates of accumulation declined logarith- mically with soil age (Fig. 5b). In the 80-year-old soil, organic C stock increased to 0.76 kg m − 2 with an average accumulation rate of 4.4 g m − 2 a − 1, however, the annual rate decreased from ca. 20 g m − 2 a − 1 in the first decade after deglaciation to 1 g m − 2 a − 1 in the fifth and following decades. It is noteworthy that the content of organic C in the buried soil (pedon WER16) was relatively high, up to 14.6 g kg − 1 throughout 12 cm deep A and AB layers. Carbon stock calculated for the upper 15 cm of buried soil was 1.52 kg m − 2, Fig. 4. Calcium carbonate distribution in the upper soil layers. that is at least twice as much as in 80-year-old modern soil.  Author's personal copy 16 C. Kabala, J. Zapart / Geoderma 175–176 (2012) 9–20 Fig. 6. Increase of total nitrogen content in a soil layer 0–3 cm (a) and stocks of nitro- Fig. 5. Increase of organic carbon content in a soil layer 0–3 cm (a) and stocks of organ- gen in the upper 15 cm of soil (b). ic carbon in the upper 15 cm of soil (b). logarithm model gives the best fit to Feo accumulation in the 0–3 cm Distribution of total soil N within and among profiles was similar to layer (Fig. 7a). As early as the second decade after deglaciation a weak that for organic C but differed in magnitude (Table 3). Accumulation of Feo increase was observed at the 6–8 cm depth and more, but no deeper soil N in the uppermost 0–3 cm layer showed a rather polynomial than 12 cm below the soil surface even in the oldest soil (pedon function of time and its concentration ranged from nearly zero in the WER17). Older glacio-fluvial materials had a significantly lower Feo newly deglaciated till to 1.7 g kg− 1 in the 80-year-old soil (Fig. 6a). content (0.8–2.5 g kg − 1 in parent material), thus the pedogenically- Within 80 years, soil N in the upper 15 cm of soil had accumulated to induced increase of Feo in buried soil was evident in the AB horizon, about 85 g m− 2, at an average rate of 1.1 g m − 2 a − 1. In contrast to where it was nearly 4-fold higher that in its parent material (Table 3). the logarithm model of organic carbon stock, the polynomial model The ratio of Feo:Fed has been called an “activity ratio” and has been provided a better fit to the increasing N accumulation across the used as a relative measure of the crystallinity of free iron oxides chronosequence (Fig. 6b). The annual rate of N accumulation decreased, (Loeppert and Inskeep, 1996). Similar to Feo, the ratio Feo:Fed but not so radically when compared to carbon, e.g. from ca. increased logarithmically in the 0–3 cm soil layer across the 1.1 g m− 2 a− 1 in the first four decades after deglaciation to chronosequence. Initially, the ratio was below 0.46, while in soils 0.6 g m− 2 a− 1 in the following decades. 30–50 years old it exceeded a level of 0.53–0.55 and maximally rose to 0.59 indicating a greater proportion of poorly crystalline Fe 3.5. Pedogenically-derived Fe (relative to total pedogenic Fe) in older soils (Fig. 7b). Both pedogenically-derived Fe forms and the “activity ratio” increased “Free” iron in these soils, as indicated by dithionite citrate– towards the soil surface and no features of iron translocation were extractable Fe (Fed) was relatively low, in the range 8.7– detected within pedons across the chronosequence (Table 3). 13.5 g kg − 1, where up to 10.4 g kg − 1 of Fed appears to be inherited from the parent glacial material, and was weakly differentiated with- 3.6. Clay fraction mineralogy in and among pedons across the chronosequence (Table 3). More evident was a rise of acid oxalate–extractable Fe (Feo) that is a Numerous authors have reported the low intensity of mineral measure of the most reactive Fe forms, e.g. water-soluble, exchange- transformation and a lack of detectable differences in mineralogy able, poorly crystalline or short-range ordered, and a fraction of within the profiles of recent arctic soils on Spitsbergen (Klimowicz et organically-bound Fe (Loeppert and Inskeep, 1996). Starting from al., 2009; Skiba et al., 2002; Szerszen and Chodak, 1983). However, an initial 4.6 g kg − 1 (inherited from the parent glacial material) these conclusions are based on soil profiles situated on various parent throughout the till in pedon WER4, Feo increased up to 7.2 g kg − 1 materials, but never on those combined in chronosequences. It was in the uppermost soil layer at site WER17. The annual increase of therefore decided to compare the clay fraction (b0.002 mm) Feo was significantly higher during the first three decades of soil mineralogy of the uppermost layer of soils across the chronosequence development when compared to the following decades, thus the in the forefield of the Werenskiold Glacier.  Author's personal copy C. Kabala, J. Zapart / Geoderma 175–176 (2012) 9–20 17 inherited from mica-chlorite schists-trioctahedral chlorites and micas (presumably sericite) — dominated the clay fraction of all samples (Fig. 8). In a recent till (site WER4) they were accompanied by amphiboles, mainly Fe–hornblende, feldspars and dolomite. Quartz was absent in the clay fraction while in the coarser fractions it was regularly present. Both XRD and DTA-DTG methods did not indicate clearly the presence of kaolinite, and peaks 0.716 and 0.353–0.355 were often stronger after heating to 550 °C than at room temperature. However, kaolinite identification in the presence of chlorites in a clay fraction requires application of an infrared adsorption technique (Kodama and Oinuma, 1963) that was not conducted and needs to be completed in the course of further studies. Some distinct changes in clay fraction mineralogy were found among soils across the chronosequence: disappearance of dolomite, decrease in amphibole content, and inversion of the relative ratio of chlorite to mica. The last ratio was calculated based on intensities of the strongest peaks: 1.021 nm for mica and 0.716 nm for chlorite. In a recent till the chlorite:mica ratio was nearly 3.5, but in the first three decades it decreased to ca. 1.0 and finally, in 80-year-old soil, to 0.74, indicating relatively rapid transformation of the clay-size primary chlorites. The changes in the chlorite:mica ratio fitted the logarithmic model (Fig. 9). 4. Discussion The results of the present study indicate that soil development on moraines of the Werenskiold glacier is, in general, a function of time. However, the accumulation of substances and other changes in soil chemical properties rarely fit a simple linear model over the observed time span of 80 years. Even in such a short time the logarithm model provided a better fit for the studied phenomena. Similar trends were also observed on other glacial chronosequences, usually covering a Fig. 7. Changes in (a) Feo concentration, and (b) Feo:Fed ratio in soil layer 0–3 cm as a function of time. much longer period of soil development (e.g. Burt and Alexander, 1996; Egli et al., 2001; He and Tang, 2008; Jacobson and Birks, 1980; Mavris et al., 2010). The clay fraction appears to be relatively uniform among pedons Weathering processes and soil development are controlled by and in close relation to the mineral composition of rocks surrounding climate and succession of vegetation (Velde and Meunier, 2008). and underlying the glacier. XRD scanning of all clay samples after Rapid mineral weathering accompanied by high rates of organic standard preparation (Mg 2 + and K + saturation, glycol solvation, matter accumulation in soil are common in cool, humid climates and heating at 350, and 550 °C), supported by DTA-DTG analyses, (Burt and Alexander, 1996). Thus, physical weathering (freeze-thaw confirmed the occurrence of primary minerals and the lack or only weathering), in particular, is supposed to proceed quickly under an traces of secondary (pedogenic) clay minerals. Primary minerals, arctic climate (Skiba et al., 2002; Szerszen and Chodak, 1983). As Fig. 8. The x-ray diffractograms of clay fraction (b 0.002 mm), air-dried specimens. Diagnostic peaks of mica (1.021 nm) and chlorite (0.716 nm) for calculation of chlorite:mica ratio are indicated. A — amphibole, C — chlorite, D — dolomite, F — feldspars, M — mica.  Author's personal copy 18 C. Kabala, J. Zapart / Geoderma 175–176 (2012) 9–20 Vermont, Canada. Other studies demonstrated the apparent instability of micas in glacial sediments, in particular that of biotite, and the relatively rapid transformation to hydrobiotite and illite, then to vermiculite, and — in a last step — to smectite (Egli et al., 2001; He and Tang, 2008; Mavris et al., 2010). Despite the successive disappearance of hornblenda and chlorite, no evidence of secondary illite, hydrobiotite, vermiculite or smectite were detected in the clay fraction of soils across the chronosequence. Similarly, Szerszen and Chodak (1983) reported primary mineral occurrence and a lack of secondary phyllosilicates in the clay fraction of moraine soils at the Werenskiold glacier. A possible explanation for mica stability in the soils under study is the predominance of sericite inherited from bedrock. Dioctahedral muscovites and sericites are more resistant to chemical weathering than trioctahedral biotites, particularly under alkaline conditions and an excess of Ca2 + and Mg2 + ions (Fanning et Fig. 9. Changes of chlorite:mica ratio in clay fraction of soils as a function of time. al., 1989). The beginnings of mica chemical weathering and the transformation of primary chlorite to vermiculite are assumed for indicated by the particle size distribution pattern, all of the studied these soils as evidenced at least by an increasing content of “pedogenic” soils have a relatively uniform texture. The stable content of the soil iron (Table 3). Initial secondary products of these processes are skeleton and decrease in the clay fraction in the uppermost soil however still “masked” by intense signals from prevailing primary layer contradict the observations made in other chronosequences on micas, or leached from the uppermost soil layer (as evidenced by the glacial moraines (He and Tang, 2008; Mavris et al., 2010). Righi et decrease in clay content). Moreover, both the decrease of chlorite: al. (1999) found that the increase in clay particle-size in soils derived mica (Fig. 9) and Feo:Fed (Fig. 7b) ratios fit the logarithm model, on post-glacier moraines was time-dependent and relatively fast indicating the weakening of weathering intensity with time. The other during the early stages of soil development. The loss of the clay explanation of the relative increase of primary mica in clay-size fraction from the uppermost layer of soils across the chronosequence fractions is the successive fragmentation of sand- and silt-sized micas on the Werenskiold moraines may be therefore apparent and result due to physical weathering (Szerszen and Chodak, 1983). Any time- from simple spatial (zonal) variability of glacial sediments. Moreover, dependent trend in these fractions was however not evident across the washing out of fine earth fractions from the soil surface is possible the chronosequence. Further study is therefore necessary to understand every spring, when melting water creates numerous seasonal micro- the relationships between time and clay-sized particles, and secondary streams on the moraine. Washing out, most intense close to the mineral formation in these soils. glacier, may explain the larger clay loss on recent moraines and the Low weathering intensity, including weak carbonate leaching, is relative stabilization on older ones (Fig. 3a). However, clay content assumed to be at least in part related to the slow succession of increases with depth in some soil profiles. The possibility of clay vegetation. It is well documented that plants and microorganisms translocation in arctic soils due to cryoturbation or formation of can significantly increase the rate of weathering of minerals, and internal aggregate soil structure has been mentioned by Alexander through the exudation of protons, low molecular weight organic and Burt (1996) and Jacobson and Birks (1980). acids, or siderophores, leading to ligand-promoted mineral The 80-year-long period is too short to expect a spectacular trans- dissolution (Bernasconi and BigLink Consortium, 2008). The intensity formation in chemical properties or formation of secondary clay of soil development is relatively more intense under continuous minerals (Velde and Meunier, 2008). However, the results of vegetation cover, and significantly accelerates as the forest becomes chemical and mineralogical analyses provide some evidence of established (Burt and Alexander, 1996; He and Tang;, 2008; Mavris chemical weathering in this environment. The primary feature of et al., 2010). Unfortunately, the harsh arctic climate on the Svalbard moraine till at the Werenskiold glacier is the relatively high content Islands eliminates forest vegetation and controls the succession of of carbonates, inherited from bedrock or newly precipitated from vascular plants in general (Piroznikow and Gorniak, 1992). We stagnating glacial water (Bukowska-Jania, 2007). Progressive found that the first flowering plants (scattered individual examples carbonate dissolution is viewed as being controlled by a combination of Saxifraga oppositifolia and Saxifraga caespitosa) invaded young of the influences of water supply and percolation within the soil, CO2 deglaciated surfaces after about 5–6 years, and the density of plant production, temperature, surface properties of carbonates (Anderson cover (including a few species of flowering plants and lichens) et al., 2000; Mavris et al., 2010) and production of organic acids (Burt increased during the first three decades after deglaciation. In the and Alexander, 1996). Many studies have shown that carbonate fourth decade, however, both the low number of plant species and minerals disappear from parent materials in humid regions within a the percentage of surface covered with vegetation stabilized at few hundred years of soil development (e.g. Alexander and Burt, relatively low levels (Table 1). A similar limited plant succession 1996; Bain et al., 1993; He and Tang, 2008), and the leaching of the was documented earlier by Piroznikow and Gorniak (1992) and soil material may be initially rapid (Crocker and Dickson, 1957). Our more recently by Wojtun et al. (2007). Piroznikow and Gorniak results evidence significant but not total dissolution of carbonates (1992) concluded that the plant succession is controlled by low even in the uppermost soil layer, where at least 0.8–2% of CaCO3 temperatures and unstable soil surface, but above all by climate was still found 80 years after deglaciation (Fig. 4). The front of dryness and the “low amount” of soil water available for plants on carbonates dissolution descended maximally to 10–12 cm below the moraines. Plant succession in the foreland of the Werenskiold glacier, soil surface. as evidenced by the percentage of the soil surface covered with The only obvious primary mineral change in the clay fraction vegetation, fits therefore to a logarithm model, where the “flattening” (excluding carbonates) within 80 years of soil development was of the curve started at the same time as in the case of weathering amphibole and chlorite depletion (Fig. 8). Similar successive evidence (Feo concentration, Feo:Fed ratio, chlorite:mica ratio etc.). disappearance of hornblenda was reported by Mavris et al. (2010) In addition to vascular plants, lichens and (in wet sites) mosses, and assumed its transformation into smectite. Munroe et al. (2007) and also cyanobacteria colonise the soil surface on recent moraines documented a total transformation of chlorites inherited from bed- at the Werenskiold glacier, often being the first oxyphototrophic rocks to hydrobiotite and vermiculite under the alpine conditions of colonisers (Kastovska et al., 2005). Cyanobacteria and eukaryotic  Author's personal copy C. Kabala, J. Zapart / Geoderma 175–176 (2012) 9–20 19 microalgae play an indispensable role in soil stabilization against wind 1990), but He and Tang (2008) suggested that the logistic model and water erosion, and accumulation of organic matter; however, often provides a better fit to observed phenomena. Based on such a their impact on mineral weathering and soil development in arctic model they estimated the time of soil organic carbon accumulation environments is still not definitively established (Bernasconi and to a steady-state (under present climate conditions) to about BigLink Consortium, 2008; Nakatsubo et al., 2005). 800–1000 years. Lichter (1998) calculated that at least 400 years As plant succession proceeded, an increase of soil organic carbon may be required to attain steady-state on sand dunes. We may con- and nitrogen was detected. It is worth noting that here the carbon clude, therefore, that despite the presently limited plant succession accumulation process started not from zero content, but nearly and reduced C stock rate the accumulation of organic carbon in soils 4–5 g kg − 1. Although the metamorphic rocks surrounding the at the Werenskiold glacier has not yet reached a steady-state. Werenskiold glacier do not contain organic matter, it may be Available reference levels for the potential content of organic carbon inherited from regoliths, marine sediments and other sediments in soils on recent moraines are offered by the buried soils on the enriched in the preglacial warmer period and then incorporated glacier foreland or relic soils on uplifted marine terraces (Kabala into the glacier body. Moreover, Kastovska et al. (2005) found that and Zapart, 2009). microbial communities can occur both in subglacial environments As plant cover extends on moraines, the organic acids from and on the glacier surface (in cryoconite sediments) and produce decaying plant residues cause a progressive decrease in the pH organic matter which is then added to glacial waters or sediments. value of the soils. The rate of decrease, however, has slowed with Relic organic matter inherited from older deposits is common in time and finally stabilized. Similar trends for pH were also found in profiles of soils developed from glacial materials in the Russian arctic the Hailougou (He and Tang, 2008) and the Morteratsch (Mavris et and subarctic regions (Gubin and Veremeeva, 2010). al., 2010) chronosequences, but the steady-state was never reached Already after 80 years of soil formation, the soil organic carbon as early as after four to seven decades. The pattern of limited decline and soil nitrogen stocks reached 0.76 kg m − 2 and 0.085 kg m − 2, in pH was closely related to limited accumulation of organic carbon, respectively, rather low when compared to other chronosequences carbonate leaching and pedogenic Fe oxide formation indicating in subarctic or alpine environments. He and Tang (2008) reported interdependences between climatically-controlled succession of carbon accumulation reaching 3.5 kg m − 2 after 180 years on vegetation, weathering intensity and intensity of soil development. moraines of the Hailogou Glacier (Sichuan, China), and Mavris et al. Similarly, in the opinion of Jacobson and Birks (1980), low tempera- (2010) reported up to 5.5 kg C m − 2 after 150 years in the tures, a short growing season, and low mean annual precipitation Morteratsch proglacial area (Swiss Alps). Our findings are consistent limit plant growth and account for the delayed soil development on only with other reports from the High Arctic, such as from the the Klutlan moraines in the Yukon Territory. deglaciated area near Ny-Alesund on Spitsbergen (Nakatsubo et al., 2005). Increases in soil organic C and N concentration in the Werens- 5. Conclusions kiold forefield showed a non-linear (polynomial) pattern over time in the uppermost layer 0–3 cm. Organic C and N stocks within the 15 cm Soils in the proglacial area of the Werenskiold glacier have a very soil layer fit similar polynomial or logarithmic models with the young age. However, within only 80 years of soil development maximal rates of accumulation being in the first decades of soil transformation trends could be estimated. The development of soil development. It is consistent with a general opinion based on properties shows mostly logarithmic or polynomial patterns with chronosequences of several thousands of years and 14C dating that soil age over the observed time span. Immediately after deglaciation, indicates the greatest rates of net carbon accumulation during the transformation of chemical soil properties started at high rates; initial phase of soil formation (Schlesinger, 1990). The mean annual however, most of the observed indicators of weathering and soil accumulation rates of organic C and N in the upper soil layer development, including stocks of organic carbon and nitrogen, (0–15 cm) over the observed time span were 4.4 g m − 2 a − 1 and carbonate dissolution, pH value, chlorite:mica ratio in clay fraction, 1.1 g m − 2 a − 1, respectively (Figs. 5b and 6b). These findings are pedogenically-derived Fe etc. decreased with time. These results lower than results from other short-term chronosequences. Mean highlight the importance of the local climate and the intensity of annual rates of C and N accumulation at the Hailogou Glacier were plant succession on weathering processes and soil development. In 28 g m − 2 a − 1 and 3.5 g m − 2 a − 1, respectively (He and Tang, 2008), contrast to the Low Arctic and alpine environments of Eurasia, the and up to 36 g C m − 2 a − 1 at Morteratsch (Mavris et al., 2010). plant succession on proglacial areas in the High Arctic is inhibited However, the rates of C and N accumulation in soils at the by low precipitation and limited water availability for plants. The Werenskiold glacier are similar to those reported by Egli et al. same factor limits the carbonate dissolution and leaching that (2001) from a soil chronosequence in the Swiss Alps that ranged stabilizes soil pH and inhibits primary mineral weathering. Based on from 6.7 to 9 g C m − 2 a − 1 and from 0.33 to 0.50 g N m − 2 a − 1. This observations of the initial stage of soil development in the foreland chronosequence covers a period of 400 years and involves the stage of the rapidly retreating Werenskiold glacier we would therefore of soil “maturity” with a steady-state of organic carbon. Similarly, a conclude that climate warming, not accompanied by an increase in quasi steady-state is observed in a chronosequence at the climate humidity, does not accelerate plant succession and soil Werenskiold glacier, where the rate of carbon accumulation (in a development. Thus, the globally discussed potential for carbon and layer 0–15 cm) decreased from 20 g m − 2 a − 1 (in the first decade nitrogen fixation in arctic soils currently developing in the proglacial after deglaciation) to only 1 g m − 2 a − 1 in the fifth and following areas of retreating glaciers should not be overestimated in further decades (Fig. 5b). The varying length of a steady-state period may models of carbon sequestration. significantly influence the final result of calculations. It seems, therefore, that the accumulation ratios for the period of “increase” (better for comparison), and for the whole observed period of soil Acknowledgments development (informative only) should be calculated separately. The increase over time in organic carbon stocks in soils of the This paper is dedicated to Prof. Leszek Szerszen who in 1957 Werenskiold chronosequence can be fitted to a logarithmic model, initiated Polish studies on the soil cover of south-west Spitsbergen. where the carbon accumulation is an ever-increasing function of We are grateful to Dr Jan Klementowski, head of our expedition, for time, however, where a quasi steady-state is approached after a his assistance in the field. The study was financially supported by certain period of time. Similar patterns are reported in other the University of Environmental and Life Sciences in Wroclaw, chronosequences (e.g. Egli et al., 2011; Lichter, 1998; Schlesinger, Poland.  Author's personal copy 20 C. Kabala, J. Zapart / Geoderma 175–176 (2012) 9–20 References Kastovska, K., Elster, J., Stibal, M., Santruckova, H., 2005. Microbial assemblages in soil microbial succession after glacial retreat in Svalbard (High Arctic). Microbial Ecology Alexander, E.B., Burt, R., 1996. Soil development on moraines of Mendenhall Glacier, 50, 396–407. southeast Alaska. 1. The moraines and soil morphology. Geoderma 72, 1–17. Klimowicz, Z., Melke, J., Uziak, S., Chodorowski, J., 2009. Specifity of arctic tundra soils Anderson, S.P., Drever, J.I., Frost, C.D., Holden, P., 2000. Chemical weathering in the of Spitsbergen. Polish Journal of Soil Science 42, 97–109. foreland of a retreating glacier. Geochimica et Cosmochimica Acta 64, 1173–1189. Kodama, H., Oinuma, K., 1963. Identification of kaolinite minerals in the presence of Bain, D.C., Mellor, A., Robertson-Rintoul, M.S.E., Buckland, S.T., 1993. Variations in chlorite by x-ray diffraction and infrared absorption spectra. Clays and Clay Minerals weathering processes and rates with time in a chronosequence of soils from Glen 11, 236–249. Feshie, Scotland. Geoderma 57, 275–293. Kowalkowski, A., 1998. Holocene rusty and rusty podzolic soils in the tundra and taiga Baranowski, S., 1977. The subpolar glaciers of Spitsbergen seen against the climate of of Middle Sweden. Roczn. Glebozn. (Soil Science Annual) 49, 29–44. this region. Acta Universitatis Wratislaviensis 387, 31–37. Kowalska, A., Sroka, W., 2008. Sedimentary environment of the Nottinghambukta Barry, R.G., 2006. The status of research on glaciers and global glacier recession: a delta, SW Spitsbergen. Polish Polar Research 29, 245–259. review. Progress in Physical Geography 30, 285–306. Krawczyk, W.E., 1992. Chemical Characteristics of Water Circulating in Werenskiold Bekku, Y.S., Kume, A., Masuzawa, T., Kanda, H., Nakatsubo, T., Koizumi, H., 2004. Soil Glacier (SW Spitsbergen). Proc. 2nd Int. Symp. of Glacier Caves and Kares, pp. respiration in a high arctic glacier foreland in Ny-Ålesund, Svalbard. Polar Bioscience 65–80. Sosnowiec, Poland. 17, 36–46. Leszkiewicz, J., Caputa, Z., 2004. The termal condition of the active layer in the Bernasconi, S.M., BigLink Consortium, 2008. Weathering, soil formation and initial permafrost at Hornsund, Spitsbergen. Polish Polar Research 25, 223–229. ecosystem evolution on a glacier forefield: a case study from the Damma Glacier, Lichter, J., 1998. Rates of weathering and chemical depletion in soils across of chrono- Switzerland. Mineralogical Magazine 72, 19–22. sequence of Lake Michigan sand dunes. Geoderma 85, 255–282. Bockheim, J.G., Mazhitova, G., Kimble, J.M., Tarnocai, C., 2006. Controversies on the Loeppert, R.H., Inskeep, W.P., 1996. Iron. In: Bartels, J.M. (Ed.), Methods of Soil Analysis: genesis and classification of permafrost-affected soils. Geoderma 137, 33–39. Part 3. Chemical Methods. SSSA Book Series, 5, pp. 639–664. Madison, Wisconsin. Bukowska-Jania, E., 2007. The role of glacier system in migration of calcium carbonate Lonne, I., Lysa, A., 2005. Deglaciation dynamics following the Little Ice Age on Svalbard: on Svalbard. Polish Polar Research 28, 137–155. implication for shaping of landscapes at high latitudes. Geomorphology 72, Burga, C.A., Krüsi, B., Egli, M., Wernli, M., Elsener, S., Ziefle, M., Mavris, C., 2010. Plant 300–319. succession and soil development on the foreland of the Morteratsch glacier Mann, D.H., Sletten, R.S., Ugolini, F.C., 1986. Soil development at Kongsfjorden, (Pontresina, Switzerland): straight forward or chaotic? Flora 205, 561–576. Spitsbergen. Polar Research 4, 1–16. Burt, R., Alexander, E.B., 1996. Soil development on moraines of Mendenhall Glacier, Marsz, A.A., Styszynska, A., 2007. The climate of Polish Polar Station at Hornsund southeast Alaska. 1. Chemical transformations and soil micromorphology. (Klimat rejonu Polskiej Stacji Polarnej w Hornsundzie). Akademia Morska, Gdynia, Geoderma 72, 19–36. Poland. Cannone, N., Diolaluti, G., Guglielmin, M., Smiraglia, C., 2008. Accelerating climate Mavris, C., Egli, M., Ploetze, M., Blum, J.D., Mirabella, A., Giaccai, D., Haeberli, W., 2010. change impacts on alpine glacier forefield ecosystems in the European Alps. Initial stages of weathering and soil formation in the Morteratsch proglacial area Ecological Applications 18, 637–648. (Upper Engadine, Switzerland). Geoderma 155, 359–371. Crocker, R.L., Dickson, B.A., 1957. Soil development on the recessional moraines of Herbert Migała, K., Nasiółkowski, T., Pereyma, J., 2008. Topoclimatic conditions in the Hornsund and Mendenhall Glaciers, South-eastern Alaska. Journal of Ecology 45, 169–185. area (SW Spitsbergen ) during the ablation season 2005. Polish Polar Research 29, Czerny, J., Kieres, A., Manecki, M., Rajchel, J., 1993. Geological Map of the SW Part of 73–91. Wedel-Jarlsberg Land, Spitsbergen. Institute of Geology and Mineral Deposits, Munroe, J.S., Farrugia, G., Ryan, P.C., 2007. Parent material and chemical weathering in University of Mining and Metallurgy, Krakow, Poland. alpine soils on Mt. Mansfield, Vermont, USA. Catena 70, 39–48. Egli, M., Mirabella, A., Fitze, P., 2001. Clay mineral formation in soils of two different Nakatsubo, T., Bekku, Y.S., Uchida, M., Muraoka, H., Kume, A., Ohtsuka, T., Masuzawa, T., chronosequences in the Swiss Alps. Geoderma 117, 81–98. Kanda, H., Koizumi, H., 2005. Ecosystem development and carbon cycle on a glacier Egli, M., Wernli, M., Burga, C., Kneisel, C., Mavris, C., Valboa, G., Mirabella, A., Plötze, M., foreland in the high Arctic, Ny-Ålesund, Svalbard. Journal of Plant Research 118, Haeberli, W., 2011. Fast but spatially scattered smectite-formation in the proglacial 173–179. area Morteratsch: an evaluation using GIS. Geoderma 164, 11–21. Piroznikow, E., Gorniak, A., 1992. Changes in the characteristics of the soil and vegetation Fanning, D.S., Keramidas, V.Z., El-Desoky, M.A., 1989. Micas, In: Dixon, J.B., Weed, S.B. during the primary succession in the marginal zone of the Werenskiold glacier, (Eds.), Minerals in Soil Environments, 2nd Edition. Soil Science Society of America, Spitsbergen. Polish Polar Research 13, 19–30. Madison, Wisconsin, pp. 551–634. Rachlewicz, G., Szczucinski, W., 2008. Changes in thermal structure of permafrost FAO, 2006. Guidelines for Soil Description, 4th Edition. Rome. active layer in a dry polar climate, Petuniabukta, Svalbard. Polish Polar Research Frenot, Y., Gloaguen, J.C., Cannavacciuolo, M., Bellido, A., 1998. Primary succession on 29, 261–278. glacier forelands in the subantarctic Kerguelen Islands. Journal of Vegetation Science Rachlewicz, G., Szczucinski, W., Ewertowski, M., 2007. Post-“Little Ice Age” retreat of 9, 75–84. glaciers around Billefjorden in central Spitsbergen, Svalbard. Polish Polar Research Goryachkin, S.V., Blume, H.P., Beyer, L., Campbell, I., Claridge, G., Bockheim, J.G., Karavaeva, 28, 159–186. N.A., Targulian, V., Tarnocai, C., 2004. Similarities and differences in Arctic and Antarctic Righi, D., Huber, K., Keller, C., 1999. Clay formation and Podzol development from post- soil zones. In: Kimball, J.M. (Ed.), Cryosols. Permafrost-affected Soils. Springer, Berlin, pp. glacier moraines in Switzerland. Clay Minerals 34, 319–332. 49–70. Schlesinger, W.H., 1990. Evidence from chronosequence studies for a low carbon- Gubin, S.V., Veremeeva, A.A., 2010. Parent materials enriched in organic matter in the storage potential of soils. Nature 348, 232–234. northeast of Russia. Eurasian Soil Science 43, 1238–1243. Skiba, S., Drewnik, M., Kacprzak, A., 2002. Soils of the western coast of Sorkappland. In: Hagen, J.O., Liestroy, O., Roland, E., Jorgensen, T., 1993. Glacier atlas of Svalbard and Jan Ziaja, W., Skiba, S. (Eds.), Sorkappland Landscape Structure and Functioning Mayen. Meddelser, Oslo. (Spitsbergen, Svalbard). Uniw. Jagiellonski, Krakow, Poland, pp. 51–86. Hagen, J.O., Kohler, J., Melvold, K., Winther, J.G., 2006. Glaciers in Svalbard: mass balance, Szerszen, L., 1968. Preliminary investigations of soil cover in the region of Hornsund, runoff and freshwater flux. Polar Research 22, 145–159. Vestspitsbergen. In: Birkenmajer, K. (Ed.), Polish Spitsbergen Expeditions. Polish Haugland, J.E., 2004. Formation of patterned ground and fine scale soil development Academy of Sciences, Warszawa, pp. 217–227. within two late Holocene glacial chronosequences: Jotunheimen, Norway. Szerszen, L., Chodak, T., 1983. Clay minerals in soils of the arctic climatic zone exemplified by Geomorphology 61, 287–301. south-west Spitsbergen. Proc. 5th Meeting European Clay Group. Charles University, He, L., Tang, Y., 2008. Soil development along primary succession sequences on moraines Prague, pp. 445–451. of Hailuogou Glacier, Gongga Mountain, Sichuan, China. Catena 72, 259–269. Tedrow, J.C.F., Hill, D.E., 1955. Arctic brown soil. Soil Science 80, 265–275. Humlum, O., Instannes, A., Sollid, J.L., 2003. Permafrost in Svalbard: a review of research. Van Reeuwijk, L.P., 2002. Procedures for soil analysis. 6th ed. ISRIC — World Soil Information History, climatic background and engineering challenges. Polar Research 22, 191–215. Centre. Technical Reports 9. Wageningen, Netherlands. IUSS Working Group WRB, 2007. World Reference Base for Soil Resources 2006, first Velde, B., Meunier, A., 2008. The origin of clay minerals in soils and weathered rocks. update 2007. World Soil Resources Reports No. 103. FAO, Rome. Springer, Berlin-Heidelberg. Jacobson, G.L., Birks, H.J.B., 1980. Soil development on recent end moraines of the Klutlan Wojtun, B., Matula, J., Pereyma, J., 2007. Plant colonization of the Werenskiold glacier Glacier, Yukon Territory, Canada. Quaternary Research 14, 87–100. forefield (W Spitsbergen). History, Evolution and Future of Arctic and Alpine Kabala, C., Zapart, J., 2009. Recent, relic and buried soils in the forefield of Werenskiold Flora. Botanical Society of Scotland. : Symposium abstracts. St Andrews University, Glacier, SW Spitsbergen. Polish Polar Research 30, 161–178. Scotland, UK, pp. 1–17. Karczewski, A., 1982. The deglaciation zonality of some glaciers in the Hornsund region Ziaja, W., 2001. Glacial recession in Sorkappland and Central Nordenskioldland, (South–West Spitsbergen). Acta Universitatis Wratislaviensis 525, 115–121. Spitsbergen, Svalbard, during the 20th century. Arctic, Antarctic, and Alpine Karczewski, A., Andrzejewski, L., Chmal, H., Jania, J., Kostrzewski, A., Lindner, L., Marks, Research 33, 36–41. L., Pulina, M., Pekala, K., 1984. Hornsund, Spitsbergen, Geomorphology — Map. Ziaja, W., 2004. Spitsbergen landscape under 20th century climate change: Sorkapp University of Silesia, Katowice, Poland. Land. Ambio 33, 295–299. View publication stats