This study was carried out as part of the project No N N 306 279 035 financed by the Ministry of Science and Higher Education. We thank Prof. Anna Pazdur for the radiocarbon dating of fen deposits from the two sites under analysis. Laboratory work was performed at the Adam Mickiewicz University Geoecological Station at Storkowo and in the laboratory of the Institute of Geoecology and Geoinformation of Adam Mickiewicz University in Poznań. We are also grateful to two anonymous reviewers for their helpful comments on the manuscript. The manuscript is translated by Maria Kawińska.
1Spring-fed fens belong to the rare group of geoecosystems the formation and functioning of which depend on groundwater nourishment. Groundwater controls peatland hydrology, ecology and morphology, and the geochemical properties of its sediments. Apart from a characteristic botanical composition (spring hygrophytic species), alkaline soligenous peatlands are also distinguished by their pattern of lithological deposits featuring interbedded peat and calcareous tufa (Dobrowolski, 2011). Considering the great usefulness of peat-tufa series from such objects for palaeoclimatic and palaeohydrological reconstructions, so emphasised in the recent years, the determination of the geological and geochemical conditions of their operation seems to be of key importance for Late-Glacial /Holocene palaeogeographical reconstructions, including changes brought about by human activity (Dobrowolski et al., 2005, 2010b; Pazdur et al., 2002a).
2Spring mires are found in the temperate zones of North America, Europe, and Asia (Atmendinger and Leete, 1998, Grootjans et al., 2006) in a variety of landscape types differing in geological, morphological and climate conditions. In Europe spring-fed fens occur mainly in the mountains and their piedmont zones (Boyer and Wheeler 1989, Grootjans et al., 2005; Hájek et al., 2002; Hájková et al., 2012; Koczur and Nicia, 2013; Kovanda, 1971) as well as on carbonate uplands (Dobrowolski et al., 2002, 2005; Pazdur et al., 2002a, b; Alexandrowicz, 2004). They can also be found, though much less frequently, in the belt of the Central European lowlands, in the glacial landscape connected with the Saale Glaciation (Alexandrowicz and Żurek, 1996; Dobrowolski et al. 2012). Spring mires (today rarely accompanied by the formation of tufa) are widespread in postglacial lakeland plateaux within the range of the Weichselian Glaciation, e.g. in Germany, Poland, Sweden and Estonia (Dobrowolski et al., 2010a, Kloss 1965; Kukla 1965; Osadowski et al., 2009; Succow and Joosten 2001; Tyler 1981,Wołejko 2000). In spite of the growing knowledge of the distribution and formation of the deposits building this type of peatland, those data still find only limited use in palaeoenvironmental analyses. But the morphological similarity of the sites, the uniform mechanism of their formation together with a great diversity of their locations offer an opportunity for a comparison of depositional records.
3On the basis of multi-proxy studies it is possible to interpret, e.g., the intensity and character of alimentation of a peatland, sources supplying the components of its deposits (nutrients and other inorganic elements), or its trophic state (Hájková et al., 2012). Those parameters provide a basis for a reconstruction of Late-Glacial /Holocene climate changes (Swindles et al., 2009), extreme events (Booth et al., 2005; Franzén and Malmgren, 2011; Swindles et al., 2010; Gałka et al., 2013), and various aspects of human impact (Lamentowicz et al., 2013).
4Biological proxies, such as pollen (e.g., Pidek et al., 2012), testate amoeba (e.g., Mitchell et al., 2008; Lamentowicz et al., 2010) or peat-forming macrofossils (e.g., Rybníčková et al., 2005; Hájek et al., 2011; Hájková et al., 2012; Drzymulska et al., 2012), have long been a focus of climatic studies or those reconstructing the trophic state of wetlands. In the research on environmental changes, mostly in ombrotrophic bogs, increasingly popular has been the geochemical record. A survey of recent studies in this field can be found, e.g., in R. Bindler (2006), while F.M. Chambers and D.J. Carman (2004) state that geochemical proxies combined with improvements in chronological techniques have resulted in renewed interest in peatland records of environmental change.
5To a limited extent, geochemical analyses have also been made of deposits of spring-fed fens (West et al., 1997; Dobrowolski et al., 2005, 2012). In this type of peatland the geochemical properties not only of peat, but also of calcareous tufas have been considered an excellent record of palaeoenvironmental information (Pentecost, 2005; Andrews, 2006; Pedley, 2009; Domínguez-Villar et al., 2011). In comparison with ombrotrophic peatlands (bogs), entirely dependent on the atmospheric supply of their water and chemical elements, geochemical features of deposits in a groundwater-fed fen (including a spring-fed one) depend mostly on minerotrophic water, but atmospheric supply also plays a part, hence in this case it can be hard to establish the source of mineral matter input unequivocally (see Shotyk, 1996).
6The geochemistry of spring-fed fen deposits is a resultant of (West, 1997): (1) features of the natural environment within the groundwater catchment feeding the peatland (allogenic drivers), (2) determinants and processes occurring within it, e.g. its biological production, redox and thermal conditions (autogenic drivers), and (3) post-depositional diagenesis. Combined with the determination of their absolute age, the geochemical features of fen deposits can be helpful in the reconstruction of conditions of and variations in sedimentary processes (= stratigraphic markers), especially hydrological and climatic determinants, and provide a basis for an assessment of the effect of human activity on the state of the environment.
7In the postglacial landscape within the Weichselian Glaciation, there are many rich calcareous fens, but high-resolution multi-proxy studies of groundwater-fed fens, especially those fed by springs, are scarce (Lamentowicz et al., 2013). Consequently, in this research it was assumed that the proxy in the form of the geochemistry of deposits of a spring-fed fen can be employed in an analysis of:
-
the chemistry of groundwater dependent on substratum lithology, the intensity of chemical weathering, and hydrodynamic conditions determining groundwater flow routes,
-
the hydrological regime resulting from climate changes and/or human activity and,
-
the hydrogeochemical environment.
8The first objective of the research was to characterise the geology and geochemistry of deposits from two spring-fed fens at Bobolice and Ogartowo (Drawa Lakeland, West Pomerania) in the postglacial landscape of Northern Poland. The obtained results of standardised geochemical analyses and radiocarbon dating provided a basis for: (1) establishing the source of water and mineral matter input, (2) reconstructing main stages in the development of the fens (the last 12 ka), and (3) identifying environmental conditions in the period preceding biogenic accumulation (18-12 ka).
9The studied sites are located in the Drawa Lakeland, a part of the macroregion of the West Pomeranian Lake District in Northern Poland (Kondracki, 2002; fig. 1).
Fig. 1 – Study area.
Fig. 1 – Secteur d'étude.
A: Locations of study sites in Western Pomerania. B: Locations of study sites in Parsęta drainage basin. C: Geological sketch map of the Parsęta River catchment. Holocene. 1: alluvia; peats and organic silts; 2: lake sands, silts, clays and gytjas; Weichselian glaciations. 3: eolian sands; 4: fluvial sands, gravels and silts; 5: lake sands and silts; 6: ice-dam clays, silts and gravels; 7: outwash sands and gravels; 8: kame sands and silts; 9: end moraine gravels, sands, boulders and tills; 10: tills, weatherd tills, glacial sands and gravels; Paleogene. 11: Sands locally containing amber, silts, clays and ignite; 12: river network, lakes; 13: the Pomeranian Phase of Weichselian Glaciation limit; 14: study site.
A : Localisation des sites d'étude dans l'ouest de la Poméranie. B : Localisation des sites d'étude dans le bassin versant de la rivière Parsęta. C : Carte schématique de la géologie du bassin versant de la Parsęta. Holocène. 1 : alluvions, tourbes et limons organiques ; 2 : sables lacustres, limons, argiles et gyttjas, glaciation weichsélienne ; 3 : sables éoliens ; 4 : sables, graviers et limons fluviatiles ; 5 : sables et limons lacustres ; 6 : argiles, limons et graviers lacustres bloqués par la glace ; 7 : sables et graviers des sandurs ; 8 : sables et limons des kames ; 9 : graviers, sables, blocs et tills des moraines terminales ; 10 : tills, tills météorisés, sables et graviers glaciaires ; 11 : sables avec contenu local en ambre, limons, argiles et lignite ; 12 : réseau hydrographique ; 13 : limite de la phase poméranienne de la glaciation weichsélienne ; 14 : site d'étude.
After Geological Map of Poland, 1:500 000, PGI, Warsaw 2006. Marks L., Ber A., Gogołek W., Piotrowska K. (Eds.), modified (http://www.pgi.gov.pl/mapy/mgp500/MGP500_main.html).
D'après la Carte géologique de la Pologne au 1:500000, PGI, Warsaw 2006. Marks L., Ber A., Gogołek W., Piotrowska K. (Eds.), modifié. (http://www.pgi.gov.pl/mapy/mgp500/MGP500_main.html).
10The area lies within the morphogenetically diversified marginal zone of the Pomeranian Stage of the Weichselian Glaciation. It stands out for its great hypsometric diversity, the altitudes connected with groups of postglacial, slope and fluvial landforms varying from 37 to 221 m a.s.l. In the complex geological structure of its substratum glaciofluvial sands and glacial tills predominate. Spring-fed fens in the postglacial area tend to concentrate in zones of contact of geomorphological units differing in their geological structure, and in areas of considerable hypsometric gradients.
11The study area is drained by the Parsęta river system. Important in its water cycle, especially in the formation of water-bearing horizons, is the lithostratigraphy of its postglacial deposits: well-permeable sands and gravels alternating with very poorly permeable series of glacial till. The Quaternary aquifer is a multi-level structure embracing several water-bearing units: near-surface, inter-morainic (upper and lower), and sub-morainic. The diversified hydrogeological conditions of the Quaternary level in the study area is presented in a geological cross-section of the Chociel river valley located north of Bobolice (Marszałek and Szymański, 2005; fig. 2). Hydrogeological conditions determine groundwater outflows in postglacial areas and give spring-fed fens their specific lithological and morphological characteristics.
Fig. 2 – Geological cross-sections in the vicinity of Bobolice.
Fig. 2 – Coupe géologique à proximité de Bobolice).
1: alluvial sands of valley floors and floodplains; 2: colluvial sands and clays; 3: alluvial sands of river terraces; 4: fluvioglacial sands and gravels; 5: kame silty sands, sands and gravels; 6: glacial silty sands with gravels; 7: glacial tills; 8: alluvial sands.
1 : sables fluviatiles des fonds des vallées et des plaines d'inondation ; 2 : sables et argiles colluviaux ; 3 : sables des terrasses fluviatiles ; 4 : sables et graviers fluvioglaciaires ; 5 : sables limoneux des kames, sables et graviers ; 6 : sables limoneux avec graviers glaciaires ; 7 : tills glaciaires ; 8 : sables fluviatiles.
The cross-section line is shown in figure 1, after Marszałek, Szymański 2005, modified.
Tracé précisé sur la figure 1 d'après Marszałek, Szymański 2005, modifié.
12The distribution of the identified spring-fed fens in the Parsęta basin is associated with the pattern of its valley network (fig. 1). Among privileged areas in terms of the number and discharge of outflows are valleys and fluvioglacial plains with a river terrace system, sections of the Pomeranian marginal spillways, and subglacial troughs and ravines. Deep incisions of the valleys and high hydraulic gradients create favourable conditions for groundwater drainage of a local and regional range from recharge areas, i.e. the surrounding morainic plateaux and fluvioglacial plains (Mazurek, 2008, 2010). River valleys are nourished by water from the near-surface and inter-morainic aquifers. Those groundwater levels are also drained by an abundance of springs and seepage sites, as in the upper part of the Chociel valley near Bobolice, or the Dębnica valley near Ogartowo (fig. 1). Unconfined groundwater supplies layer-type outflows around which hanging spring-fed fens have formed on the valley slopes. In the footslope zone where inter-morainic confined water nourishes efficient ascending springs, dome-shaped spring-fed fens have developed. Fen complexes comprising different types of soligenous fens can develop in the drainage zones of a few aquifer levels, as stressed, e.g., by Wołejko (2000).
13Spring-fed fens in the Bobolice site ( 55°56'33" N; 16°37'07" E; 113-117 m a.s.l.) are located within a vast complex of valley fens (ca. 1,300 ha, vide Osadowski and Wołejko, 1997; Wołejko, 2000). They embrace 5 such objects situated at the bottom of the upper Chociel river valley (the Parsęta drainage basin, fig. 3) and on its slopes (Pidek et al., 2012). The Chociel flows in a subglacial channel 30-40 m deep, clearly marked in the relief and cutting through an undulating morainic plateau. The plateau, composed of glacial sands locally overlying tills, is the supply area for Quaternary aquifers drained in the Chociel valley.
Fig. 3 – Geological cross-sections.
Fig. 3 – Coupes stratigraphiques.
A: Geological cross-section through the Bobolice spring-fed fen. B: Geological cross-section through the Ogartowo spring-fed fen. 1: mineralized peat; 2: sedge peat; 3: reed peat; 4: reed-sedge peat; 5: sedge-wood peat; 6: moss peat; 7: calcareous tufa; 8: organic clay; 9: silty sand; 10: varigrained sand; 11: location of borings.
A : Coupe stratigraphique à travers la tourbière de source de Bobolice. B : Coupe stratigraphique à travers la tourbière de source de Ogartowo. 1 : tourbe minéralisée ; 2 : tourbe à carex ; 3 : tourbe à roseaux ; 4 : tourbe de canne et de laîches ; 5 : tourbe à carex et débris ligneux ; 6 : tourbe à mousses ; 7 : tuf calcaire ; 8 : argile organique ; 9 : sable limoneux ; 10 : sable hétérométrique ; 11 : localisation des forages.
14The deep incision of the Chociel valley, and in effect its high hydraulic gradients, create favourable conditions for the drainage of groundwater from the surrounding morainic plateaux that are its alimentation areas. Groundwater outflows, occurring at various heights of the valley slopes, are mostly located at the contact of sand-gravel fluvioglacial or fluvial deposits with semi-permeable morainic tills that underlie them (fig. 2). The outflows in the bottom of the valley and at the foot of its slopes are nourished primarily by confined groundwater from the inter-morainic level. The extensive bottom of the Chociel valley (over 180 ha) is filled with various hydro-ecological types of spring-fed fens (Succow, 1988). Many fen cupolas have formed on the valley slopes and at their foot. Today peatlands are drained by the river and its tributaries as well as an intricate network of drainage ditches.
15The outflowing water has a neutral reaction (7.1-7.8 pH) and average mineralisation (specific conductivity 297-370 μS/cm), and belongs to the group of bicarbonate-calcium and bicarbonate-sulphate-calcium waters (concentrations of Ca2+: 35-60 mgdm-3; Mg2+: 4.2-4.5 mgdm-3, Na+: 3.9-4.8 mgdm-3, K+: 0.8-0.9 mgdm-3, HCO3-: 98-169 mgdm-3, Osadowski, 2010).
16The fen cupolas, intensively fed by groundwater, are occupied by rushes of the sedge association Caricetum acutiformis Sauer 1937 and eutrophic common reed communities Urtico-Phragmitetum Succ. 1970. The fens, used as hay-growing places, are often part of moor-grass meadows Junco-Molinietum Prsg. 1951 and globe-flower meadows Polygono bistortae-Trollietum europaei (Handt 1964) Bal.-Tul. 1981. In many places, after farming use has been abandoned, the dominant herbal associations often growing on fen cupolas are Valeriano-Filipenduletum Siss. (Westh. et al., 1946) and patches of the sedge association Caricetum cespitosae (Steffen, 1931) Klika et Šmarda 1940 (Osadowski, 2000; Pidek, et al., 2012).
17The spring-fed fen in the Ogartowo site (53°45'57" N; 16°07'56" E; 97-98 m a.s.l.) is located in the marginal part of an extensive peatland complex some 5.5 ha in area that occupies the headwater zone of an anonymous tributary of the Dębnica river (fig. 1). The fen in the Dębnica valley is characterised by the occurrence of two distinct spring cupolas (protruding 1-2 m above the surface of the peat plain). The cupolas have formed around springs located at the foot of the morainic plateau scarp, in the contact zone with the Dębnica valley. The deep dissection of the undulating plateau has created conditions for the drainage of inter-morainic confined waters. Several efficient foothill springs are supplied from a series of lower glaciofluvial sand-gravel deposits. Since the fen is surrounded by arable land, the cupolas are being intensely drained by farming-related drainage ditches and consist of highly decomposed peat. In the past, this place used to be exposed to a much greater human impact.
18In modern times, the concentrated flow of water from the fen cupola causes intensive erosion of biogenic-mineral deposits in the lower part of the fen plain. The outflowing water has a neutral reaction (7.6-7.8 pH), average mineralisation (specific conductivity 420-440 μS/cm), and belongs to the bicarbonate-calcium group (concentrations of Ca2+: 73-81 mgdm-3; Mg2+: 6.3-6.8 mgdm-3, Na+: 5.6-6.2 mgdm-3, K+: 1.3-1.5 mgdm-3, HCO3-: 204-235 mgdm-3; Mazurek, 2010). This type of water is characteristic of the entire Drawa Lakeland (Mazurek, 2008). At present the deposition of carbonates is not observed in this area.
19The examined cupola of the Ogartowo fen is situated within a eutrophic reed-rush association Urtico-Phragmitetum Succ. 1970. Physiognomically resembling aquatic reed rushes, this association includes nitrophilous and heliophilous species, attesting to the drainage and mineralisation of the fen surface layer (Osadowski, 2000; Pidek et al., 2012).
20A geological identification of spring-fed fens was made for 6 such objects in West Pomerania (Osadowski, 2010; Dobrowolski et al., 2010a; Pidek et al., 2012). Two were chosen for a detailed geochemical study (fig. 1): at Bobolice (coded BOB) and Ogartowo (coded OGA). In each, drillings were made using an Instorf sampler (length 0.50 m; diameter 0.05 m), at an interval of ca. 10-50 m, along orthogonally orientated transects (fig. 1). A total of 21 geological drillings were performed. From the central part of cupolas of each peatland core deposits with an undisturbed structure were sampled. Each core was analysed in sedimentological terms using Troels-Smith's method for the description of the deposits (Troels-Smith, 1955), with Dobrowolski's (2011) modifications for the peat-tufa sequences (tab. 1 and tab. 2). The sampled deposits were used for detailed laboratory analyses and radiocarbon dating.
Tab. 1 – Lithology of deposits building the spring-fed dome-shaped fen in the Chociel valley near Bobolice (core BOB-8).
Tab. 1 – Lithologie des sédiments qui forment le dôme de la tourbière de source dans la vallée de Chociel, près de Bobolice (forage BOB-8).
Depth (m)
|
Unit
|
Lithology
|
T-S formuła
(after Troels-Smith, 1955
and Dobrowolski, 2011)
|
Remarks
|
0.00-0.10
|
6
|
Sphagnum peat, strongly decomposed, mineralized, dark brown to black, with single sand grains
|
Th44 (Sph), Ag++, sicc.3, nig.4, elas 0, str.0
|
|
0.10-1.15
|
5
|
Sedge peat, medium decomposed, in places with an admixture of brown and peat mosses, dark brown to black, with accessory wood
|
Th33(Car), Tb31, Dh++, sicc.2, nig.3, elas 2, str.0, trunki, lim.1
|
C14
BOB-8/57
|
1.15-1.50
|
Sedge peat, medium-decomposed, dark brown, with accessory reed and wood
|
Th33(Car), Th31(Phra), Dh++, Ld+, sicc.2, nig.3, elas2, str.0, trunki, lim.1
|
C14
BOB-8/142
|
1.50-2.30
|
Sedge peat, medium decomposed, in places well decomposed, dark brown, with accessory wood
|
Th3-43, Tl3-41, Ld+, sicc.2, nig.3, str.0, trunki, lim.1
|
|
2.30-3.26
|
Sedge peat, medium and well decomposed, dark brown to black, with accessory Cladium mariscus
|
Th34(Car), sicc.2, nig.4, elas2, str.0, lim.0
|
C14
BOB-8/326
|
3.26-6.64
|
4
|
Peat-tufa rhythmite (calcareous tufa with varying grain size, with interlayers of well decomposed sedge peat)
|
Cm(min.)2, Cm(maj.)1, Th41, sicc.2, nig.0-1, str.2, lim.1
|
C14
BOB-8/347
BOB-8/500
BOB-8/562
|
6.64-6.75
|
3
|
Sedge-moss peat, well decomposed, contaminated by amorphous calcium carbonate
|
Tb44, Cp(min.)+, sicc.2, nig.4, str.1, lim.1
|
C14
BOB-8/666
BOB-8/675
|
6.75-7.40
|
2
|
Peat-mineral rhythmite (silty sands and sandy silts)
|
Ag2, As1, Tb21, sicc.2, nig.3, str.2, lim.1
|
|
7.40-7.50
|
1
|
Humopeat (sedge peat, strongly decomposed), with a small admixture of sand, with accessory wood
|
Th24, Ag+, sicc.2, nig.4, str.0, trunki, lim.0
|
C14
BOB-8/750
|
7.50-7.70
|
Bedrock
|
Coarse sand with single gravels of Scandinavian rocks
|
Gmin2, Gma2, sicc.2, nig.0, str.0, lim.2
|
|
Tab. 2 – Lithology of deposits building the spring-fed dome-shaped fen in the Ogartowo site (core OGA-7).
Tab. 2 – Lithologie des sédiments qui forment le dôme de la tourbière de source à Ogartowo (forage OGA-7).
Depth
(m)
|
Unit
|
Lithology
|
T-S formuła
(after Troels-Smith, 1955
and Dobrowolski, 2011)
|
Remarks
|
0.0-0.10
|
3
|
Sedge peat, strongly mineralized, with single sand grains, with not decomposed roots of herbaceous plants
|
Th43, Gmin1, nig.4, strf.0, sicc.3
|
|
0.10-0.45
|
Sedge peat, well decomposed, with single sand grains, contaminated by amorphous calcium carbonate
|
Th43, Gmin1, nig.4, strf.0, sicc.2, lim.0
|
C14
OGA -7/34
|
0.45-0.65
|
2
|
Calcareous tufa, fine-grained, with inserts of plant (sedge) detritus
|
Cm(min.)2, Cp(maj.)1, Dh+, nig.1, strf.1, sicc.2, lim.3
|
|
0.65-0.70
|
Calcareous tufa, fine-grained, with inserts of medium decomposed sedge peat, with visible malacofauna
|
Cm(min.)3, Th11, nig.2, strf.2, sicc.2, lim.3, teste moll
|
|
0.70-0.75
|
Sedge peat, well decomposed, contaminated by amorphous calcium carbonate, with visible malacofauna
|
Th33, Cp(maj.)1, nig.3, strf.1, sicc.2, lim.3, teste moll
|
|
0.75-0.78
|
Calcareous tufa, fine-grained, with dispersed plant (sedge) detritus
|
Cp(maj.)4, Dh+, nig.2, strf.1, sicc.2, lim.3
|
|
0.78-0.80
|
Sedge peat, well decomposed, contaminated by amorphous calcium carbonate
|
Th33, Cp(maj.)1, nig.3, strf.1, sicc.2, lim.3
|
|
0.80-0.95
|
Calcareous tufa, fine-grained, with abundant malacofauna
|
Cp(maj.)4, nig.1, strf.1, sicc.2, lim.3
|
|
0.95-1.00
|
Calcareous tufa, fine-grained, strongly contaminated by amorphous organic matter
|
Cp(maj.)3, Sh1, nig.2, strf.0, sicc.2, lim.3
|
|
1.00-1.10
|
Calcareous tufa, medium-grained
|
Cm(min.)4, Sh++, nig.2, strf.0, sicc.2, lim.3
|
|
1.10-1.30
|
Sedge peat, well decomposed, contaminated by amorphous calcium carbonate, with visible traces of streaks
|
Th33, Cp(maj.)1, Dh+, nig.3, strf.2, sicc.2, lim.2
|
C14
OGA-7/110
OGA-7/123
|
1.30-1.70
|
Calcareous tufa, medium and fine-grained, with inserts of sedge-moss peat and amorphous humus material
|
Cp(maj.)2, Cm(min.)2, Dh+, Sh+, nig.2, strf.1, sicc.2, lim.2
|
|
1.70-2.15
|
Sedge peat, well decomposed, with abundant malacofauna
|
Th34, nig.4, strf.0, sicc.2, lim.2
|
C14
OGA-7/172
|
2.15-2.18
|
1
|
Sedge peat, well decomposed, strongly contaminated by amorphous calcium carbonate
|
Th32, Cp(maj.)2, nig.3, strf.1, sicc.2, lim.3
|
C14
OGA-7/215
|
2.18-2.32
|
Sedge peat, strongly decomposed, with visible malacofauna
|
Th34, nig.4, strf.0, sicc.2, lim.3, teste moll, lim.0.
|
C14
OGA-7/230
|
2.32-2.60
|
Bedrock
|
Sandy silt with single gravels of Scandinavian rocks
|
Ag4, As+, Dh+, Gmin+, nig.3, strf.0, sicc.2, lim. sup.0
|
|
After Pidek et al. (2012), modified.
D'après Pidek et al. (2012), modifié.
21To analyse the chemical composition of deposits of spring-fed fens, samples were taken from two profiles (at an interval of 5 cm): at Bobolice (BOB-8, 71 samples) and Ogartowo (OGA-7, 25 samples). After drying at 105°C and homogenisation in an agate mortar, the samples were roasted at 550°C to constant weight, which allowed the determination of their per cent content of ash and organic matter OM. After roasting, the ash was extracted in aqua regia at room temperature for 16 h, then boiled on a water bath for 2 h. Total contents of macroelements (Na, K, Ca, Mg) and microelements (Fe, Mn, Cu, Zn) were determined. Their concentrations were analysed by the AAS (Atomic Absorption Spectrophotometry) method using a SpectraAA 20 Plus Varian apparatus. The obtained results were calculated as percentages of dry mass; they are presented in figure 4 and figure 5.
22The amount of carbonates reacting with HCl, expressed in CaCO3, was determined with the help of the volumetric method, using a Scheibler calcimeter (Bednarek et al., 2004). Determinations were also made of actual soil acidity (pHH2O) in the 1:5 soil: water suspension by glass electrode (Bednarek et al., 2004).
23Radiocarbon dating was performed for 15 samples of peat and calcareous tufa layers from the BOB-8 (9 samples) and OGA-7 cores (6 samples). Bulk sediments were dated (each sample was 2 cm thick). Relatively small volume of samples is acceptable in conventional technique of radiocarbon dating and the results are comparable with those obtained by AMS technique. The samples were treated with 2% HCl to remove carbonates and converted to benzene for liquid scintillation counting with the use of a Quantulus 1220 spectrometer (GdS samples, see tab. 3; Pazdur et al., 2003) or ICELS (GdC samples, see tab. 3; Tudyka and Pazdur, 2012).
24The mineral substratum of the Bobolice fen consists of Pleni-Weichselian fluvioglacial coarse sands and/or Late Weichselian fluvial sands overlying them at places (fig. 3A). Those deposits, drilled below a depth of 7.50 m, are the top part of fillings of a subglacial trough used today by the headwater section of the Chociel river valley. The biogenic series (tab. 1, Pidek et al., 2012), directly overlying mineral deposits, starts with highly decomposed peats with features of humopeat (unit 1) and with fine wood detritus (7.40-7.50 m). They pass upward into a mineral-organic series (unit 2) composed of silty sands and sandy silts (6.75-7.40 m). This is overlain by a thin layer (6.64-6.75 m) of highly decomposed sedge-moss peats contaminated by amorphous calcium carbonate (unit 3). They pass gradationally into a peat-tufa rhythmite (3.26-6.64 m), which is the fundamental part of the headwater sediment series of the fen (unit 4). The rhythmite consists of calcareous tufa, highly diversified in fractional terms (from coarse-grained to silt-sized), and of highly decomposed sedge peats. The size of individual sequences varies from a few millimetres to some dozen centimetres. Overlying the peat-tufa series are highly and medium decomposed sedge peats (0.10-3.26 m) with numerous intercalations of poorly decomposed alder wood (fragments of branches, trunks, and wood detritus), building unit 5. The top segment of the biogenic series (unit 6) is made up of highly decomposed Sphagnum peat (0.00-0.10 m) (fig. 3A).
25Three basic lithological units can be distinguished in the Ogartowo vertical profile (tab. 2; Dobrowolski et al., 2010a; Pidek et al., 2012) embracing: (unit 1) a thin (up to 0.25 m), highly decomposed layer of sedge peats (humopeat), and a usually continuous layer of bottom sedge-reed and/or sedge peats; (unit 2) a peat-tufa rhythmite, which is the basic part of the headwater series of the deposit, and which consists of thin (0.01-0.20 m) intercalations of calcareous tufa (heterogeneous in granulometric terms) and highly decomposed sedge and sedge-moss peats; and (unit 3) a discontinuous layer (up to 0.45 m) of highly mineralised sedge peats, at places contaminated by amorphous calcium carbonate (fig. 3B).
26The results of the radiocarbon dating of deposits of the fens under study are presented in table 3. The dates obtained do not give a clear picture of the chronology of events in the evolution of the spring-fed fens at the scale of the region, and require reference to the published results of other proxies (see Pidek et al., 2012). While there is full agreement between the palaeobotanical record and the radiocarbon dates in the case of the Bobolice site, it is hard to give their definitive interpretation in the entire OGA-7 profile (the Ogartowo site) unless palynological data have been accommodated too.
Tab. 3 – List of radiocarbon dates from the spring-fed fens in Bobolice (BOB-8) and in Ogartowo (OGA-7).
Tab. 3 – Liste des datations radiocarbone réalisées dans les tourbières de source de Bobolice (BOB-8) et de Ogartowo (OGA-7).
No.
|
Sample name
|
Lab. No.
|
δ13C PDB
(‰)
|
T
(14C BP)
|
Calibrated Age
(range 95%)
|
1
|
BOB-8/57
(0.55-0.57 m)
|
GdS-1022
|
-25.0
|
2400 ± 120
|
800 (86.6%) 345calBC
325 (8.8%) 205calBC
|
2
|
BOB-8/142
(1.40-1.42 m)
|
GdS-1025
|
-25.0
|
3920 ± 110
|
2860 (2.3%) 2810calBC
2750 (1.0%) 2720calBC
2700 (90.4%) 2125calBC
2090 (1.7%) 2045calBC
|
3
|
BOB-8/326
(3.24-3.26 m)
|
GdC-398
|
-25.0
|
6975 ± 60
|
5985 (95.4%) 5735calBC
|
4
|
BOB-8/347
(3.45-3.47 m)
|
GdS-1068
|
-25.0
|
7530 ± 110
|
6610 (93.4%) 6205calBC
6170 (0.3%) 6160calBC
6145 (1.7%) 6105calBC
|
5
|
BOB-8/500
(4.98-5.00 m)
|
GdS-1061
|
-25.0
|
9050 ± 150
|
8630 (95.4%) 7750calBC
|
6
|
BOB-8/562
(5.60-5.62 m)
|
GdS-1071
|
-25.0
|
9780 ± 190
|
10010 (1.7%) 9930calBC
9880 (93.3%) 8700calBC
8675 (0.4%) 8650calBC
|
7
|
BOB-8/666
(6.64-6.66 m)
|
GdC-402
|
-25.0
|
8920 ± 75
|
8280 (95.4%) 7820calBC
|
8
|
BOB-8/675
(6.73-6.75 m)
|
GdS-1043
|
-25.0
|
10510 ± 160
|
10440 (6.0%) 10295calBC
10285 (89.4%) 9310calBC
|
9
|
BOB-8/750
(7.48-7.50 m)
|
GdS-1053
|
-25.0
|
10630 ± 150
|
10910 (95.4%) 10135calBC
|
|
10
|
OGA -7/34
|
GdS-1026
|
-25.0
|
3830 ± 120
|
2620 (95.4%) 1935calBC
|
11
|
OGA-7/110
|
GdS-1070
|
-25.0
|
7930 ± 130
|
7175 (95.4%) 6495calBC
|
12
|
OGA-7/123
(1.20-1.23cm)
|
GdS-987
|
-25.0
|
4390 ± 120
|
3495 (0.8%) 3465calBC
3375 (90.2%) 2850calBC
2815 (3.1%) 2740calBC
2730 (1.4%) 2675calBC
|
13
|
OGA-7/172
(1.70-1.72 m)
|
GdS-982
|
-25.0
|
8390 ± 90
|
7590 (91.0%) 7245calBC
7235 (4.4%) 7185calBC
|
14
|
OGA-7/215
(2.13-2.15 m)
|
GdS-977
|
-25.0
|
8110 ± 170
|
7505 (95.4%) 6645calBC
|
15
|
OGA-7/230
(2.28-2.30 m)
|
GdS-981
|
-25.0
|
7750 ± 75
|
6770 (95.4%) 6440calBC
|
27The examined deposits of spring-fed fens show different levels of organic and mineral matter and calcium carbonate in the profile. Against this background an analysis was made of variations in the levels of major and trace elements as well as selected geochemical environmental indices. The geochemical diagrams obtained (fig. 4 and fig. 5) document differences in sedimentary environments resulting from conditions of groundwater circulation and its physico-chemical properties as well as changes in the conditions of sedimentation. A characteristic feature is a clear connection between the concentration of the individual geochemical components (OM, pH, micro- and macro-elements) and the change in the lithology of deposits building the fens.
28Four basic geochemical zones were distinguished in the Bobolice core BOB-8, correlated with lithological divisions and differing fundamentally in terms of the chemical composition of their deposits (fig. 4).
Fig. 4 – Geochemistry of deposits in the BOB-8 core.
Fig. 4 – Géochimie des dépôts du forage BOB-8.
Geochemical zones are based on stratigraphically constrained cluster analysis CONISS. 1: sedge peat (medium decomposed); 2: sedge peat (well decomposed); 3: calcareous tufa (fine-grained); 4: calcareous tufa (coarse-grained); 5: sedge-moss peat (well decomposed); 6: Sphagnum peat (well decomposed); 7: silty sand; 8: varigrained sand; 9: coarse sand with gravel.
Les zones géochimiques sont définies par le programme CONISS. 1 : tourbe à carex (moyennement décomposée) ; 2 : tourbe à carex (bien décomposée) ; 3 : tuf calcaire (à grain fin) ; 4 : tuf calcaire (grossier) ; 5 : tourbe à carex et à mousses (bien décomposée) ; 6 : tourbe à sphaignes (bien décomposée) ; 7 : sable limoneux ; 8 : sable hétérométrique ; 9 : sable grossier avec gravier.
29Geochemical zone I (6.64-7.50 m, lithological units: 1-3) occurs on a substratum of coarse sands. Its characteristic is a stepwise increase in the content of calcium carbonate to 39%. The content of organic matter grows in two peat intercalations at depths of 7.40-7.50 m (IA = unit 1) and 6.64-6.75 m (IC = unit 3). A characteristic feature of this zone is the highest share of lithophilous elements: magnesium and potassium, as well as zinc, in the entire profile. There are also high levels of sodium and iron in the deposits (fig. 4).
30Geochemical zone II (3.26-6.64 m; lithological unit 4) has the highest mean content of calcium carbonate -66% (the maximum for the entire profile being 99%). The calcareous tufa found here in intercalations has the highest levels of calcium and manganese, while its mean magnesium content drops to 1 g/kg DM. The upper limit of the zone is marked by a decline in calcium carbonate content to 1% and an increase in that of organic matter to 82%, which is connected with a change in the lithological character of the deposits (a transition from calcareous tufa to peat sedge). Within the entire zone, organic matter accounts for 2.4 to 39.4% of deposits, and they show the lowest concentration of potassium in the profile. Instead, they maintain a high concentration of iron, with a maximum of 131 g/kg DM in a peat intercalation. This intercalation also displays the highest Fe:Mn rate in the profile that can be indicative of highly reducing conditions during accumulation (fig. 4). In the deposits of zone II a high correlation was only found in relation to the pairs of elements Ca and Fe (r = 0.67, p < 0.001), and Na and K (r = 0.62, p < 0.001).
31Geochemical zone III (0.10-3.26 m), embracing mainly sedge peats (lithological unit 5), displays a high content of organic matter - on average, 81.6%, while that of calcium carbonate amounts to about 1% in the entire profile. There is an increase in the mean content of lithophilous elements: magnesium and sodium, but their concentrations show great fluctuations. The share of potassium grows too; it attains the highest concentration at a depth of ca. 1.70-2.40 m. Also noted at this depth is a drop in organic matter to 57%. Towards the top of the profile, there is a decline in the content of iron to ca. 1.7 g/kg DM, and in the values of the Fe:Mn index. In the entire zone the content of manganese is very low. A high correlation was found for the pairs of elements Ca and Mn (r = 0.79, p < 0.001) and between the contents of organic matter OM and potassium (r = -0.79, p < 0.001). Correlations at a lower significance level were recorded for pairs of elements: K and Zn (r = 0.67, p < 0.001), Fe and Mn (r = 0.68, p < 0.001), and Zn and Cu (r = 0.55, p < 0.02), as well as between organic matter OM and potassium (r = -0.58, p < 0.03).
32Geochemical zone IV (.00-0.10 m; lithological unit 6) embraces the top layer of deposits with a 50% content of organic matter, a minimum share of carbonates < 1%, and an increase in the Zn and Cu levels.
33On the basis of the varying levels of the main components of deposits in the Ogartowo OGA-7 core as well as three geochemical indices, three main geochemical zones were distinguished here (fig. 5).
Fig. 5 – Geochemistry of deposits in the OGA-7 core.
Fig. 5 – Géochimie des dépôts du forage OGA-7.
Geochemical zones are based on stratigraphically constrained cluster analysis CONISS. Explanation of lithological units as in Figure 4.
Les zones géochimiques sont définies par le progamme CONISS. Description des unités lithologiques : voir figure 4.
34Geochemical zone I (1.70-2.32 m), corresponding to the sedge peats of lithological unit 1 and the bottom part of the peat-tufa rhythmite (bottom part of lithological unit 2), lies on sandy silts which were found to be high in lithophilous elements: potassium and magnesium, as well as zinc. This zone has a high average content of organic matter, at 56%, although the figures vary widely, from a maximum of 84% to a minimum of 25%. In the top part OM falls to 22%. With a low mean content of CaCO3 (ca. 17%), there is a short rise in calcium carbonate (to 59%) that accompanies a drop in organic matter. Changes in calcium carbonate are accompanied by an increase in the concentrations of calcium, magnesium and manganese. The content of sodium shows the highest values in both, the bottom and the top part of this zone. Iron shows here the highest average concentration (108.6 mg/kg DM) in the entire core. With a low mean content of Mn, the Fe:Mn index attains the highest values.
35Geochemical zone II (0.45-1.70 m), connected with the peat-tufa rhythmite of lithological unit 2, displays wide differences in the content of organic matter and calcium carbonate, which can indicate variable conditions of sedimentation. There are three growth stages in the calcium carbonate levels corresponding to the occurrence of calcareous tufa. The pattern is similar in the case of the CaCO3:OM index and the proportions of Ca, Mg and Mn. In turn, the iron concentration is lower when compared with those in the underlying deposits of zone I. Iron attains the highest level with a higher content of organic matter, which can mean a worsening of oxygen conditions.
36Geochemical zone III (0.00-0.45 m) embraces the highly mineralised sedge peats of the top part of lithological unit 3 in which organic matter averages 25%. Carbonates remain high, above 34% on average, which shows peat sedentation to have occurred in the conditions of their abundant supply and accumulation, especially in the bottom part. The distribution of calcium and magnesium is similar. In the bottom part the highest shares go to sodium and manganese, and in the top part, to potassium and zinc. Iron concentrations are low here, and the average Fe:Mn index equals 27.
37To identify sources supplying the chemical components found in the deposits of the Ogartowo and Bobolice spring-fed fens, factor analysis was conducted with the help of the principal components method (Currás et al., 2012, Kylander et al., 2013). The input variables were standardised values of OM, CaCO3 and the eight macro- and microcomponents determined. Using Kaiser's criterion, those factors (principal components) were distinguished that had a value greater than 1.
38When assessing 10 chemical components of the Bobolice deposits, four complementary principal components were distinguished that jointly accounted for 74.1% of the geochemical variance of the deposits. The first component PC1 accounted for 31.7% of the total variance, and the second PC2 for 20.3% (tab. 4). PC1 is inversely proportional to the content of organic matter, and positively correlated with calcium carbonates and cations. The second component PC2 reflects primarily the variability of such macrocomponents as magnesium and potassium. Less important are principal component PC3 (tab. 4), directly proportional to iron and inversely proportional to copper, and principal component PC4 identifying manganese.
39In the case of the Ogartowo site, the first principal component (PC1) accounts for 54.5% of the total variance, and the second, for 19.4%. The first component is positively correlated with zinc, organic matter and iron, and inversely proportional to the content of calcium carbonate and calcium, magnesium and manganese cations (tab. 4), while the second component mainly characterises the variability of potassium content.
Tab. 4 – Factor loadings obtained from the principal component method on the basis of the geochemical properties of sediment from the spring-fed fens in Bobolice (BOB-8) and in Ogartowo (OGA-7).
Tab. 4 – Coefficients de saturation obtenus par l'analyse en composantes principales portant sur les propriétés géochimiques des dépôts des tourbières de source de Bobolice (BOB-8) et d’Ogartowo (OGA-7).
Parameter
|
Bobolice
(BOB-8)
|
Ogartowo
(OGA-7)
|
Factor 1
|
Factor 2
|
Factor 3
|
Factor 4
|
Factor 1
|
Factor 2
|
OM
|
-0.799
|
-0.494
|
-0.067
|
0.085
|
0.762
|
-0.570
|
CaCO3
|
0.906
|
0.205
|
-0.004
|
-0.108
|
-0.882
|
-0.132
|
Ca
|
0.859
|
0.143
|
-0.209
|
-0.131
|
-0.918
|
-0.194
|
Mg
|
-0.444
|
0.686
|
-0.321
|
-0.158
|
-0.888
|
0.243
|
Na
|
-0.350
|
0.135
|
-0.061
|
-0.291
|
-0.417
|
-0.474
|
K
|
-0.534
|
0.746
|
0.043
|
0.054
|
0.286
|
0.894
|
Fe
|
0.210
|
0.342
|
0.685
|
-0.453
|
0.743
|
-0.560
|
Mn
|
0.508
|
0.149
|
0.008
|
0.676
|
-0.828
|
-0.305
|
Zn
|
-0.245
|
0.575
|
0.421
|
0.452
|
-0.362
|
-0.278
|
Cu
|
0.090
|
0.455
|
-0.623
|
-0.034
|
0.915
|
-0.039
|
Eigenvalues
|
3.17
|
2.03
|
1.19
|
1.02
|
5.45
|
1.95
|
Explained variance %
|
31.73
|
20.28
|
11.93
|
10.16
|
54.52
|
19.49
|
Factor loads statistically significant at α ≤ 0.05 are given in bold.
En caractères noirs, les coefficients de saturation statistiquement représentatifs avec α ≤ 0.05.
40The groups distinguished with the help of the principal components method correspond well with the geochemical zones obtained via stratigraphically constrained cluster analysis CONISS (fig. 6).
Fig. 6 – Principal components analysis (PCA) based map and a geochemical classification of sediment samples by the first and second component (A) for the Bobolice spring-fed fen (B) for the Ogartowo spring-fed fen.
Fig. 6 – Représentation graphique de l'analyse en composantes principales (ACP) et classification géochimique des échantillons sédimentaires par le premier et le second facteur (A) pour la tourbière de source de Bobolice (B) pour la tourbière de source d'Ogartowo.
41The documentary material collected, based on the results of sedimentological and geochemical analyses and supplemented with radiocarbon dates and the results of palaeobotanical analyses published earlier (Pidek et al., 2012), shows close correspondence between the records of palaeoenvironmental events in the chief evolutionary stages of the two spring-fed fens under study.
42The results of factor analysis reveal two groups of determinants: (1) allogenic drivers and (2) autogenic drivers, controlling, according to West (1997), the geochemistry of spring-fed fen deposits. Those determinants affect the chemical composition of deposits to varying degrees, as shown by the number of factors distinguished and their correlations with individual components.
43In both fens the first principal component PC1 can be interpreted to some extent as one expressing the share of chief allogenic components coming from the leaching of postglacial deposits and transported to the sedimentary basin by water. The mineral-petrographic composition of the loose Quaternary deposits and their area of contact with infiltrating water are favourable to the weathering of aluminosilicates, solution of calcium carbonate, and ion exchange. Those processes lead to the release of such ions as Ca, Mg and K, as well as iron, manganese and copper compounds that determine the chemical composition of groundwater feeding the fens. The presence of those compounds in the deposits depends not only on their concentration in groundwater, but also on the geochemical conditions obtaining in a sedimentary basin that affect their precipitation and bio-accumulation in peat, and then reactivation, e.g. by acid rainwater.
44The first principal component PC1 is also one that accounts for the variability of the content of the autogenic part of the deposits, i.e. organic matter. The high values of the first component PC1 reflect the proportions of products of chemical denudation (fig. 6), while low values express the share of the biogenic component of the deposits. Additionally, organic matter is positively correlated with such metals as copper and iron (the Ogartowo site, tab. 4), which shows peat to have sorptive properties with respect to those elements (and to zinc, Shotyk, 1996). Organic matter also influences redox conditions and the reaction of the geochemical environment, thus affecting the pattern of geochemical processes and changes in the levels of some elements. In a sedimentary environment with low pH values and microbiological oxidation of organic substance occurring via a reduction of sulphates, hydrogen sulphide is produced and Fe precipitates in the form of sparingly soluble iron sulphides (Macioszczyk and Dobrzyński, 2002), as noted, e.g., in the organic deposits at Ogartowo (high OM-Fe correlation, r=0.82, p<0.001). The same conditions are extremely unfavourable to the precipitation of manganese compounds. This last element has a major weight in the third principal component PC3 (tab. 4). Those compounds have a variable ability to migrate depending on the current redox conditions, hence the partial geochemical separation of those elements (Kabata-Pendias and Pendias, 1999) and their different positions in the system of the first two principal components (fig. 6). The highest concentrations of manganese, and lower ones of iron, were noted in tufa deposits (Ogartowo, Bobolice) that have formed in oxidation conditions at a high pH. Manganese compounds show a high affinity for many metals (among them Cu). Like manganese ions, copper ions migrate easily, especially in an oxidising and slightly acidic environment, and when bound to organic matter, turn into hardly mobile forms. The high average content of copper in the deposits under study (Bobolice 5.4 mgkg-1 DM, Ogartowo 6.2 mg/kg DM) against its level in bogs (2 mg/kg DM; Borówka, 1992), as well as its position on the diagram of the first two principal components (fig. 6), justifies the conclusion that an important source of this element is chemical denudation in the recharge area of the fens.
45The second principal component PC2 is identified with magnesium (in the Ogartowo site, tab. 4) and potassium (in both sites), less active migrants, bound much more strongly than calcium in the sorption complex of the deposits, hence their content in them is interpreted by Borówka (1992) and Wojciechowski (2000) as an indicator of mechanical denudation and of the supply of allogenic clastic material to the fen (a proportion of zinc is also supplied passively). However, what should also be taken into consideration is the supply of those ions from the processes of chemical denudation, especially in the case of magnesium, as demonstrated by the position of this element in the first two principal components for the Ogartowo site (fig. 6B). It is the second-ranking cation, in terms of quantity, in the migration series of elements in the water nourishing the fen, even if its concentrations can be as low as one-tenth of those of calcium (Osadowski et al., 2009).
46The elevated levels of heavy metals, especially Zn and Cu in the top part of the profile, can indicate a supply of matter from anthropogenic processes. However, the principal components analysis can hardly differentiate between wet and dry atmospheric deposition, which also embraces elements coming from anthropogenic sources. The examined fens lie in an area exposed to a relatively weak human impact, with a possibility of a transborder supply of pollutants. The dominant cation in rainwater here is sodium (Mazurek, 2010), but Na has a low affinity for organic matter, and thus may be easily lost from peatland (Shotyk, 1988). Shotyk (1996) emphasises that in the case of a peatland nourished by various types of water (groundwater, surface water, rainwater), it is impossible to distinguish between atmospheric and hydrospheric metal inputs.
47The mineral substratum of both fens consists of fractionally diversified glaciofluvial and fluvial sands stratigraphically connected with the Pleni-Weichselian (fig. 2 and fig. 7).
Fig. 7 – Model of spring-fed fen evolution in West Pomerania (detailed description in the text).
Fig. 7 – Modèle du développement d'une tourbière de source dans l'ouest de la Poméranie (description détaillée dans le texte).
1: varigrained sand; 2: silty sand; 3: sedge peat; 4: moss-sedge peat; 5: calcareous tufa; 6: Sphagnum peat; 7: directions of overland, subsurface and/or groundwater flow.
1 : sable de granulométrie variée ; 2 : sable limoneux ; 3 : tourbe à carex ; 4 : tourbe à carex et à mousses ; 5 : tuf calcaire ; 6 : tourbe à sphaignes ; 7 : direction du ruissellement, de l'écoulement de subsurface et/ou de l'écoulement souterrain.
48Another source of mineral deposits in the substratum of the fens was also Late-Glacial mechanical denudation occurring in the conditions of a scant vegetation cover. In those deposits the content of organic matter and carbonates amounts to a few per cent, while the levels of lithophilous elements: magnesium and potassium, as well as zinc, are high (fig. 4 and fig. 5). The predominance of potassium over sodium in the deposits of the substratum suggests the domination of mechanical denudation and a passive supply of its products via overland flow to the valley bottom (Borówka, 1992; West, 1997). This type of alimentation of the study areas in that period was due to the presence of permafrost in the substratum which hindered a vertical migration of water.
49This stage is only documented in the Bobolice site. Its lithological record consists of bottom humopeats and highly decomposed sedge peats of lithological unit 1 (fig. 7). Radiocarbon dating indicates the Younger Dryas as the beginning of the biogenic succession here. This is corroborated by the published palynological data (Pidek et al., 2012) showing the presence of juniper and the maximum values of herbs. Predominant among herb pollen are Poaceae, Cyperaceae, Artemisia and Filipendula. Those are taxa of heliophilous plants, and the described spectra are evidence of the occurrence of loose communities of park tundra with juniper shrubs, typical of the Younger Dryas. Thus, in this stage peat accumulated in the conditions of permafrost present in the substratum and no contribution from confined groundwater. Substantial admixtures of mineral components (silty sand and sandy silt) were found in the deposits, probably transported to the fen surface by river waters, or they may be products of valley slope denudation by overland flow.
50In the period of permafrost persisting in the substratum, the development of the fens was still controlled by local conditions. The water nourishing them came then from a shallow circulation (fig. 7). What favoured this type of alimentation of the fen in the Chociel valley was permafrost in the substratum. Its peat-mineral rhythmite has a high level of magnesium and potassium resulting from a passive supply of its products by water via overland flow and throughflow. Low Na:K rates (fig. 4) corroborate the predominance of mechanical over chemical denudation in that period. Also the high content of iron in bottom deposits can be partly due to the passive supply of this element to the sedimentary basin with products of mechanical denudation.
51In periglacial conditions, there was already redeposition of carbonates present in the active layer of permafrost (Nowaczyk and Tobolski, 1980), while the acidic calcium carbonate contained in water precipitated as a result of biochemical and physical processes. Given the alkaline reaction (pH 7.2-7.5, fig. 4) caused by an increase in the content of carbonates in the deposits of this stage (8-39%), it was also possible that iron ions got oxidised and precipitated from water in the form of colloid suspensions or complexes with organic matter. The oxidation conditions obtaining during sedimentation were also favourable to the precipitation of manganese. The relatively high levels of Fe, Mn, Cu and Zn in the deposits under analysis can be seen as connected with the availability of those compounds in the surface layers of postglacial deposits and their susceptibility to leaching. The supply and transport of those compounds occurred probably as a result of the drainage of the active layer of permafrost in the summer season. The increase in the Zn content in the organic deposits of stage II was probably partly due to its bio-accumulation.
52At Ogartowo, the bottom organic deposits starting this stage (Pidek et al., 2012), with a very low pH of ca. 2.6 and high Fe:Mn rates (fig. 5), are indicative of highly reducing and acidic conditions of phytogenic deposition in the initial stage of development of this fen. In a reducing environment, Fe and Zn precipitated in the form of sparingly soluble sulphides (Macioszczyk and Dobrzyński, 2002). Those conditions were extremely unfavourable to the precipitation of manganese compounds, hence the low concentrations of this element in the bottom deposits (fig. 5).
53The functioning of the spring-fed fens in the two sites under study is connected with the unblocking of the vertical circulation of groundwater as a result of the degradation of permafrost (= start of the ascending type of alimentation) in the middle of the Early Holocene (Preboreal/ Boreal transition). This was the period of formation of the regional system of groundwater flow in West Pomerania (Lewandowski and Nita 2008) and the start of groundwater outflows supplied from inter-morainic levels (fig. 7). In lithological terms, its evidence is the tufa-peat rhythmite with its high concentrations of Ca, Fe and Mn, the highest content of calcium carbonate, and rising values of the CaCO3:OM rate (fig. 4 and fig. 5). In the deposits, the pH values generally exceed 7. In the groundwater present in postglacial areas today, calcium is the chief cation (Mazurek, 2008, Osadowski et al., 2009) occupying the first place in the migration series of elements; since the migration series in deposits is the same, this corroborates their alimentation by groundwater. This very high content of calcium carbonate in sediments indicates that the spring areas were supplied by groundwater rich in products leached from glacial and fluvioglacial sediments, while the concentrations of lithophilous elements (Mg and especially K) decreased. However, the high level of calcium carbonate in deposits is not only a result of the availability of calcium compounds in the substratum and a record of their intensive leaching by groundwater, but also follows from conditions favourable to the biogeochemical precipitation of those compounds in zones of groundwater outflows to the surface. At this stage of fen development the intensity of accumulation of calcareous tufa increased, helped by spring-related moss communities from the class Montio-Cardaminetea or the vegetation of eutrophic waterlogged sites of the alliance Caricion davallianae from the class Scheuchzerio-Caricetea nigrae (Osadowski et al., 2009; Dobrowolski et al., 2010a), in which bio-deposition of calcium carbonate is a physiological ability. In carbonate deposits the proportion of organic matter often drops below 10%.
54According to Borówka (1992), another indication of the predominance of leaching processes in a drainage basin can be a higher content of sodium than potassium in its biogenic-mineral deposits. The low content of K can also result from its limited sorption in the deposits of this stage, and in effect, from transporting this cation away from sedimentary basins. In the outflow zone of alkaline groundwater there were favourable conditions for oxidation and precipitation of iron and, especially, manganese ions, as indicated by high concentrations of these elements and low values (<50) of the Fe:Mn index (Kabata-Pendias and Pendias, 1999).
55The basic period in the development of the two fens is the Atlantic stage of the Holocene (fig. 7). In lithological terms, its record is the peat-tufa rhythmite (sensu Dobrowolski, 2011) documenting moisture and thermal changes in the environment, with warm and moist periods corresponding to the deposition of calcareous tufa. In both study sites, as in most other localities in southern (Pazdur et al., 1988) and eastern Poland (Dobrowolski et al., 2002, 2005), the stage of maximum deposition of calcareous tufa is correlated with the climatic optimum of the Atlantic. This is also the time to which Lamentowicz (2005) dates the development of spring-fed fens in East Pomerania. The considerable discharge of the confined springs nourishing the fens persisted until the late Atlantic, while the migration of carbonates attained a wider range then and has prevailed with different intensity until today (Petelski and Sadurski, 1987).
56The oxidising conditions and the alkaline reaction of water still favoured an accumulation of chemical denudation products resulting in high concentrations of Ca, Na, Mn and Fe in the deposits of this stage. A periodic increase in the supply of those elements indicates temperature and moisture conditions favouring intensive denudational processes (Davison, 1993; Wojciechowski, 2000; Borówka and Tomkowiak, 2010). The high levels of Ca and Na in relation to Mg and K, respectively, are indicative of persistent fen alimentation by groundwater enriched with products of the chemical denudation of glacial deposits. A decrease in groundwater supply that could result in a slowdown of denudational processes was connected with a gradual accumulation of organic matter and a change in redox conditions in the depositional basins developing around the springs. With an elevated content of organic matter and a drop in the pH value, iron attained its highest concentrations then, which followed from the reducing conditions of the environment. At the same time the levels of Ca, Mn and Cu as products of chemical denudation decreased. The sequence of carbonate-organic deposits in the two fens (especially at Ogartowo, fig. 5) can provide a basis for establishing stratigraphic differences in the intensity of denudational processes.
57In the deposits of the Bobolice fen, an abrupt fall in calcium carbonate accompanied by an increase in organic matter (fig. 5) is evidence of environmental changes, probably a decreasing intensity of the ascending groundwater supply. The more than 3-metre-thick series of the peat-tufa rhythmite could limit the alimentation of this sedimentary basin by groundwater as a result of the substratum becoming sealed and a decline in vertical hydraulic conductivity and the volume of groundwater outflows.
58In the late part of this stage at Bobolice, the peat-tufa rhythmite transformed into medium- and highly decomposed sedge peat with low concentrations of Ca, Fe and Mn and a slight increase in those of lithophilous elements: Mg and K (fig. 4). The low concentration of Ca connected with a trace presence of calcium carbonate in deposits (< 2%) indicates that the supply and/or precipitation of chemical denudation products to the spring-fed fen decreased radically. Organic matter was gradually deposited in an environment with limited access to oxygen, under Mn-reducing conditions, which could have resulted in its lower concentration in the deposits, but greater mobility and transport by water leaving the fen. A change in the reaction of deposits to acidic (pH < 5) probably reduced the accumulation of copper and allowed its greater leaching from the deposits (Kabata-Pendias and Pendias, 1999). A slight increase in the levels of Mg and K can indicate a growing importance of mechanical denudation and the supply of its products to the fen. The fen passed from a krenic environment (sensu Dobrowolski, 2011) to a helocrenic one. It probably gradually switched to rainwater alimentation, which changed the volume and quality of solutes brought to deposits, while acid rains could contribute to the leaching of the deposited components and their transport to surface waters. Recorded in the deposits of the spring-fed fen is also the history of atmospheric supply. Wehrli et al. (2010) suggest that the end of tufa formation coincided with the transition from fen to raised bog in the Lorze Valley (central Switzerland). It also coincided with the end of the mid-Holocene temperature optimum, when climate became cooler and wetter.
59A marked worsening of the thermal and moisture conditions at the start of the Subboreal brought about a substantial decline in the activity of confined springs, a drop in the rate of organic accumulation, and a radical limitation of the deposition of calcareous tufa (fig. 7). The dry conditions with several minor wet shifts are also observed in the ombrotrophic bog situated in Swiss Jura mountains (Roos-Barraclough et al., 2004) and in the raised bog in East Pomerania (Gałka et al., 2013). At the close of the Atlantic and during the Subboreal, the natural development of the fens could have been weak influenced by the appearance of the first Neolithic farming population.
60In lithological terms, this stage is represented by sedge and Sphagnum peat low in ash and with calcium carbonate content falling below 5%. In these deposits the concentrations of Ca, Fe and Mn are below the mean values for the whole core, while those of Cu and K are average or greater (fig. 4 and fig. 5). The increased levels of Zn and Cu in the near-surface deposits result from present-day anthropogenic activities, i.e. industrial air pollution (Zgłobicki et al., 2011). Changes in the chemical composition of the near-surface deposits can now also be due to the mineralisation of organic deposits.
61The conducted analyses of the peat-tufa series supplied geochemical data for spring-fed fens, which are examined even more rarely than other cation-rich fens. The mean content of elements in sediments from the two spring-fed fens in Northern Poland are higher than in those of fens and bogs (see for example Shotyk, 1988, 1996, Borówka, 1992, Bragazza et al., 2003). The chemical composition of sediments, especially the levels of macrocomponents Ca, Mg, Na and K, reflect the character and intensity of the natural processes of chemical denudation that occur in the recharge area of the analysed sites. Changes in the levels of macrocomponents Mg and K in the geochemical record can also result from the supply of clastic material, e.g. from extreme slope or fluvial processes. The relationships observed in the stratigraphic pattern of some geochemical elements, like iron, manganese and zinc, also reveal local conditions obtaining in the zone of sediment accumulation that can be favourable to precipitation, chemical sorption, or oxidation and reduction of some chemical components. Hence, the tufa-peat rhythmite can provide a basis for an assessment of relative changes in the structure and magnitude of denudation processes in the catchments of sedimentary basins. Changes in those relations can reflect changes in the water cycle, climatic conditions and land cover in the postglacial area of Northern Poland. The geochemical record can also provide a basis for an analysis of sources and pathways of the supply of chemical components to a fen. The research performed shows that there were geogenic, biogenic, atmospheric and anthropogenic sources contributing them in various proportions over time.
62Compared with the results of radiocarbon dating, the record of sedimentary succession indicates a great variability, both temporal and spatial, of sedentation/deposition processes in Ogartowo and Bobolice sites, even though they lie at a distance of some 50 km from each other. The geochemical features reflect the regional and local conditions of the alimentation of the two spring-fed fens: the different starting time of groundwater alimentation, changes in the volume of groundwater outflows, and an increase in the role of rainwater nourishment.
63The peat-tufa series are a more complex accumulation environment than the usually analysed deposits of bogs with their single dominating source of supply in the form of rainwater and the local accretion of organic matter. The complex origins of chemical components of deposits of spring-fed fens in postglacial areas makes it hard in many cases to identify their source unequivocally; one should also remember that their content can be controlled by several factors, among which are local conditions of the geochemical environment, like the reaction, redox potential, the CO2 content in water, the presence of organic matter, and microbial activity. Neither can one exclude chemical changes occurring after the deposits have formed, caused by such processes as rainwater leaching their near-surface part, decomposition of organic matter, or the uptake of some elements, especially Ca, Mg and K ions, by plants (West, 1997). The best interpretation of sediment geochemistry in terms of environmental changes will be possible when we have learned how to combine biological and geochemical proxies.