Climate and human induced hydrological change since AD 800 in an ombrotrophic mire in Pomerania (N Poland) tracked by testate amoebae, macro-fossils, pollen and tree rings of pine

Climate and human induced hydrological change since AD 800 in anombrotrophic mire in Pomerania (N Poland) tracked by testate amoebae,macro-fossils, pollen and tree rings of pine

MARIUSZ LAMENTOWICZ, KRYSTYNA MILECKA, MARIUSZ GAyKA, ANNA CEDRO, JACEK PAWLYTA,NATALIA PIOTROWSKA, yUKASZ LAMENTOWICZ ANDWILLEM O. VAN DER KNAAP

BOREAS Lamentowicz, M., Milecka, K., Gazka, M., Cedro, A., Pawlyta, J., Piotrowska, N., Lamentowicz, y. & van derKnaap, W. O.: Climate and human induced hydrological change since AD 800 in an ombrotrophic mire inPomerania (N Poland) tracked by testate amoebae, macro-fossils, pollen and tree rings of pine. Boreas, Vol. 38,pp. 214–229. 10.1111/j.1502-3885.2007.00004.x. ISSN 0300-9843.

This high-resolution, multiproxy, palaeoenvironmental study of the Szowinskie Bzota raised bog in N Poland,10 km from the Baltic Sea, covering the last 1200 years reveals different aspects of environmental change in a rangeof spatial scales from local to regional. Testate amoebae allowed quantitative reconstruction of the local watertable using a transfer function based on a training set fromN andW Poland. Special attention is paid to the testateamoeba Arcella discoides, which responds to rapid water-table fluctuations more than to average surface wetness.Macrofossils supported by local pollen tracked the local vegetation dynamics caused by local human impact anddisturbance, including nutrients. Regional pollen showed human-induced landscape change outside the bog. Treerings of Pinus sylvestris reflected the history of tree establishment and desiccation of the bog. Strong correlationsbetween DCA axes 1 of regional pollen, of macrofossils and of testate amoebae, and a testate-amoebae-basedwater-table reconstruction that excludes A. discoides, indicate that changes on all spatial scales are linked, which isexplained by a strong hydrologic connection between bog and surroundings. The combination of proxies showsthat groundwater levels were modified by both human impact and climate change.

Mariusz Lamentowicz (e-mail: [email protected]), Krystyna Milecka and Mariusz Gazka, Department of Bio-geography and Palaeoecology, Adam Mickiewicz University (Faculty of Geographical and Geological Science), Dzie-gielowa 27, PL-61-680 Poznan, Poland; Anna Cedro, University of Szczecin Laboratory of Climatology and MarineMeteorology, Institute of Marine Science Faculty of Natural Science, University of Szczecin, Poland; Jacek Pawlyta andNatalia Piotrowska, Department of Radioisotopes, Institute of Physics, Silesian University of Technology, Krzywoustego2, PL-44-100 Gliwice, Poland; yukasz Lamentowicz, Department of Hydrobiology, Faculty of Biology, Adam Mick-iewicz University, Umultowska 89, PL-61-614 Poznan, Poland; Willem O. van der Knaap, Institute of Plant Sciences,University of Bern, Altenbergrain 21, CH-3013 Bern, Switzerland; received 29th April 2008, accepted 20th June 2008.

The sensitivity of ombrotrophic mires to climatechange has often been brought up in the literature(Mauquoy & Barber 1999a, b; Hughes et al. 2007). Inmost recent publications, climate has been consideredthe most important driver of change in raised bogs(Mauquoy et al. 2004b; Schoning et al. 2005; Charmanet al. 2006; Tarnocai 2006; Yeloff & Mauquoy 2006).However, human disturbance is also considered sig-nificant, especially in the past 1000 years (Chambers &Charman 2004; van der Linden & van Geel 2006;Rasanen et al. 2007; Yeloff et al. 2007a). The use ofseveral proxies in palaeoecological studies (Sillasooet al. 2007; Blundell et al. 2008) may help to disentangleclimate from factors such as autogenous vegetationchange, peat-cutting, drainage or deforestation. Pollenanalysis, in particular, gives the important backgroundof human-induced land-use changes that can then becompared with local habitat and hydrological change.

Peat archives are not extensively documented in Po-land, especially those of Baltic raised bogs (Wodziczko& Thomaschewski 1932; Otzuszewski 1948; Otzuszewski& Borowko 1954; Szafranski 1961; Herbichowa 1998;Latazowa & Pedziszewska 2003). Kluczynski (1939,

1940, 1949) described the process of origin and devel-opment of Baltic and continental raised bogs on thebasis of observations in the field, but he carried out in-vestigations only on continental raised bogs in the east-ern part of former Poland. Most studies of mires haveconcentrated on modern vegetation, with the palaeoe-cological aspect underused so far. Features of this pre-sent study, filling a gap in the knowledge, are that itcombines testate-amoebae analysis, pollen analysis,macrofossil analysis and dendroecology; it covers thelast millennium in high (sub-decadal) resolution, andpalaeoecological studies in this part of Poland close tothe Baltic Sea are anyway rare. The site studied here,Szowinskie Bzota, was investigated by Herbichowa(1998), who has provided useful background informa-tion on the developmental history of the site and itssurroundings.

Combining four proxies in one study means thatinferences can be drawn on the dynamics of hydrology,trophic state, local vegetation and regional vegetation.The challenge is in disentangling anthropogenic,climatic and other causes of change. This approachaddresses a range of spatial scales and functions, as

DOI 10.1111/j.1502-3885.2008.00047.x r 2008 The Authors, Journal compilation r 2008 The Boreas Collegium

follows. Pollen analysis is most suitable for re-constructing land-use change in the surroundings of themire (Birks & Birks 1980; Berglund & Ralska-Jasie-wiczowa 1986); dendroecological analysis of treesgrowing on the mire surface may help in determiningthe precise moment of anthropogenic disturbance(Frelechoux et al. 2003); pollen analysis should informus how this affected the mire vegetation (Koff et al.1998); and testate-amoebae analysis should show howpeat formation was affected (Buttler et al. 1996).

One main aim with this study was in determining theclimate during the past millennium and thus contribut-ing towards answering the key question posed by theEU project Millennium (http://geography.swan.ac.uk/millennium/Millennium2a.htm): ‘Does the magnitudeand rate of 20th century climate change exceed thenatural variability of European climate over the pastmillennium?’ Peat deposits from various places mayhelp us understand climatic change, even though mostpeatlands have a human-impact signal that is difficultto separate from the climatic signal, especially for thepast 200 years.

In this study, we aim: (1) to provide a multiproxyhigh-quality palaeoenvironmental record for the past1200 years in a little-studied region, (2) to investigatehow different aspects of environmental change are re-flected by various biotic proxies and how they interact,with emphasis on human impact and climate change,and (3) to extract any climatic signal.

Study site

Szowinskie Bzota, studied in this article, is located innorthern Poland (54121052.5300N, 16129024.7200E, 30ma.s.l.) (Fig. 1), 10 km from the Baltic shoreline. It is aBaltic bog (Osvald 1925; Kulczynski 1939, 1940, 1949)lying in the Baltic bog region (Baltisches Re-genmoorbezirk) in Succow & Joosten’s (2001) classifi-cation of mire regions of Europe. The mire is 120 ha insize and is situated in a shallow melt-out depression ona flat morainic plateau on the watershed of the riversWieprza and Grabowa; it is believed to be exclusivelyombrotrophic (Herbichowa et al. 2007). The dome ofthe raised bog is elevated 1.2m above the surroundings,thus not visibly elevated due to its large size (Herbi-chowa 1998). The first effort to drain Szowinskie Bzotain 1880 was by ditching around the edges; the next in1970 with two ditches extending into the central part ofthe dome, which were renewed in 1985. SzowinskieBzota was not covered by forest until the end of the 18thcentury. The present-day pine forests on the mire mar-gins are the result of spontaneous Pinus sylvestris ex-pansion (mainly in the central part of the bog) c. 120years ago, as well as forest management practices in themarginal part. Betula is admixed in this forest. Themodern vegetation of the open, central part of the bog

is Sphagnum lawn (mostly S. fallax and S. magellani-cum) with Eriophorum vaginatum, Trichophorum cespi-tosum and scattered dwarf P. sylvestris (Herbichowa1998).

Herbichowa’s (1998) palaeoecological study of Szo-winskie Bzota was based on plant macrofossils, geo-chemistry and radiocarbon dating. Her geologicalsurvey based on 59 cores showed an average maximumpeat depth of about 2.8m and a constant peatstratigraphy over most of the bog. She found that themire started to develop 3965–3693 yr BC probably asa result of climatic change. The maximum peat-accumulation rate found was 1mm/year. Herbaceousand woody peat was deposited up to 1887–1530 yr BC,when the peatland was burned. Oligotrophic mire(raised-bog vegetation) with a dominance of Sphagnumand E. vaginatum began to develop AD 50–356 (now at2m peat depth). Sphagnum fuscum dominated peat

0 250 500 m

Location ofpeatland

Active ditches

Coring site

Central open part of the mirewith scattered dwarf pines

Pine forest on the bog

Deciduousforest

Crop fields

Fig. 1. Location of the study site.

BOREAS Hydrological change in an ombrotrophic mire in Pomerania, N Poland 215

started to develop between AD 470 and 1400 (approx-imate date due to low resolution).

The region of Szowinskie Bzota has the temperatetransitional climate of eastern Europe, influenced by bothhumid maritime and dry continental air masses. How-ever, this coastal part of northern Poland has the stron-gest maritime impact of the country. Westerlies move inrapidly from the Atlantic without any barriers, makingthe climate similar (though drier) to NW Europe. Ac-cording to the climatic regionalization byWos (1999), thestudy area is located in climatic division II (central coastalarea). Average annual temperature is 17.81C. January isthe coldest month, with average �0.41C, and August thewarmest, with 116.71C. The vegetation season has c. 220days. Average annual precipitation at Ustka (34km eastof the study site) is 701mm, varying from 451mm in dryyears (1959) to 1036mm in moist years (1970). February,March and April are the driest months (about 40mm);July is the wettest (81mm). Average annual snow cover is50 days per year, but this is highly variable.

Material and methods

Sampling and identification

In August 2005, two 1m long peat monoliths were col-lected with a Wardenaar sampler at 20m distance from

the central part of the mire (Wardenaar 1987). Mono-lith SL4 was analysed for Pb isotopes and pollutantsand the results are reported by De Vleeschouwer(2007). Monolith SL2 studied here for biotic proxieswas divided into 1 cm slices that were then subdividedfor the different analyses.

Subsamples of 4ml at 1 cm intervals for testate-amoebae analysis were prepared with sieving andback-sieving (Hendon & Charman 1997). The testateamoebae were identified and counted at 200–400�magnification to a total of 150 individuals per sub-sample whenever possible. The identification was basedon the available literature (Grospietsch 1958; Charmanet al. 2000; Clarke 2003; Mitchell 2003).

Subsamples (2ml at 1 cm intervals) for pollen analysiswere treated with 10% KOH and acetolysed for 3min(Berglund & Ralska-Jasiewiczowa 1986; Faegri & Iversen1989). A minimum of 500 pollen grains of trees andshrubs (AP) was counted in each subsample wheneverpossible, but in subsamples with low pollen frequencies allpollen grains were identified in two slides of 22�22mm.Charcoal was not analysed. Percentage calculation wasbased on the sum of pollen grains of trees, shrubs andherbs (AP1NAP=100%). The results are presented aspercentages in the pollen diagram (Fig. 3).

Subsamples of 4ml at 1 cm intervals for plant mac-rofossils analysis were rinsed over 0.25 and 0.5mm

800 1000 1200 1400 1600 1800 2000Calendar date (AD)

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100SL-I-a

SL-II

SL-III

SL-IV-b

SL-V-b

AD 1830

AD 1100

AD 860

AD 840

SL-I-b

SL-IV-a

SL-V-a

Common zonesand time

Dep

th (

cm)

Fig. 2. Age–depth model for peat monolith SL2 of Szowinskie Bzota bog. Dashed vertical lines and horizontal grey bands delimit commonzones.

216 Mariusz Lamentowicz et al. BOREAS

2000

1900

1800

1700

1600

1500

1400

1300

1200

1100

1000 900

800

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

AP

NA

P

Poaceae

Ana

lyst

: Kry

styn

a M

ileck

a

SL-I-a

SL-II

SL-III

SL-IV-b

SL-V-b

AD

183

0

AD

110

0

AD

860

AD

840

SL-I-b

SL-IV-a

SL-V-a

Com

mon

zon

esan

d tim

e A

D

Fig.3.Percentagepollen

diagram

forpeatmonolith

SL2ofSzowinskieBzota

bogshowingregionalvegetationchanges.


mesh size sieves. The residues were identified under astereoscopic microscope at 10–100� magnification.The percentage of each taxon was estimated on thebasis of one slide from each subsample examined underthe microscope at 200–400� magnification. Macro-fossils were identified in accordance with the availableliterature (Katz et al. 1965, 1977; Grosse-Brauckmann1972, 1974; Tobolski 2000) and compared with areference collection. Special attention was paid toSphagnum sections, because precise identification ofparticular species is difficult.

Dating

The 10 samples for radiocarbon dating were extractedfrom 1 cm thick peat slices and consist of Sphagnumstems carefully cleaned of rootlets and other con-tamination (Table 1). All samples underwent the stan-dard AAA procedure. Sample pretreatment andgraphite target preparation were done at the GliwiceRadiocarbon Laboratory, Gliwice, Poland. 14C con-centration was measured in the AMS facility ofthe Poznan Radiocarbon Laboratory. The resultingconventional radiocarbon dates were calibrated withOxCal 4 software (Bronk Ramsey 2001). The IntCal04calibration curve (Reimer et al. 2004) was usedfor samples older than AD 1950 and the Kueppers04post-bomb calibration curve for younger samples(Kueppers et al. 2004). The Kueppers04 curve isdesigned especially for the biosphere and takes intoconsideration 14C concentration changes in the north-ern hemisphere atmospheric CO2 only during the vege-tation period. The age–depth model (Fig. 2) wasconstructed using the P_Sequence deposition model ofOxCal 4 (Bronk Ramsey 2008), with the functionparameter k set to 0.08. The modelled calibrated ages ofthe 14C-dated depths were interpolated using the Aki-ma spline algorithm (Akima 1970).

Data processing

Depth to the water table (DWT) was estimated fromtestate amoebae using a training set of 123 recent sam-ples taken mainly from natural, not drained, Sphagnummires in the central part of Pomerania (N Poland) andW Poland (Lamentowicz & Mitchell 2005b; Lamento-wicz et al. 2008b). The performance of four modelswas tested: partial least squares (PLS), weightedaveraging (WA), tolerance down-weighted averaging(WA-tol) and weighted-averaging partial least squares(WA-PLS), using the C2 software (Juggins 2003). Theroot mean square error of prediction (RMSEP) wascalculated using jack-knifed cross-validation (Crowley1992). For DWT, the lowest RMSEP was provided byWA-tol, with a maximum prediction bias of 8.9 cm anda jack-knifed RMSEP (RMSEP(jack)) of 4.3 cm. Somesubsamples had very low testate-amoebae concentra-tions, resulting in a total count as low as 10–20. DWTexperimentally estimated for such subsamples appearedto make sense compared with macrofossil results,so these are presented. DWT was estimated in twodifferent ways: DWT(1) using the complete list of taxaand DWT(2) with exclusion of Arcella discoides. Incontrast to other abundant taxa, hydrologicalinstability (rapidly fluctuating water levels) stronglyfavours A. discoides (Lamentowicz & Mitchell 2005a;Lamentowicz et al. 2008a). DWT(1) estimations(A. discoides included) are therefore ambiguous insections in which such conditions occur. We presentboth versions of DWT because it is not commonpractice to exclude A. discoides from DWT estimates,and it is not required in many studies because hydro-logical instability is the exception rather than the rule.

Statistically significant zones were determined in-dependently for testate amoebae, pollen and plant mac-rofossils in accordance with the recommendations ofBennett (1996), i.e. using optimal sum-of-squares parti-tioning (Birks & Gordon 1985; Birks 1986) and testing

Table 1. Results of radiocarbon dating of SL2 monolith of Szowinskie bog.

Depth Lab. no. Age 14C yr BP or pMC Dated material Calibrated range 68% Calibrated range 95%

10–11 cm Gda-1101 111.15�0.32 pMC Sphagnum stems AD 1995–1998 (68.2%) AD 1994–2000 (95.4%)15–16 cm Gda-1102 125.03�0.37 pMC Sphagnum stems AD 1958–1962 (68.2%) AD 1958–1962 (80.6%)

AD 1981–1982 (14.8%)30–31 cm Gda-1103 355�60 Bark, wood AD 1482–1601 (68.2%) AD 1452–1636 (95.4%)40–41 cm Gda-1104 750�30 Sphagnum stems AD 1255–1284 (68.2%) AD 1225–1289 (95.4%)50–51 cm Gda-1105 1053�35 Sphagnum stems AD 1000–1031 (58.2%) AD 986–1046 (69%)

AD 1106–1117 (10%) AD 1106–1117 (10%)AD 1137–1150 (3.4%)

60–61 cm Gda-1106 1275�30 Sphagnum stems AD 930–944 (68.2%) AD 925–953 (95.4%)70–71 cm Gda-1107 1109�30 Sphagnum stems AD 887–908 (68.2%) AD 871–920 (95.4%)80–81 cm Gda-1108 1158�30 Sphagnum stems AD 825–856 (68.2%) AD 807–877 (95.4%)90–91 cm Gda-1109 1174�30 Sphagnum stems AD 774–795 (68.2%) AD 728–744 (4.7%)

AD 766–820 (90.7%)99–100 cm Gda-1110 1186�30 Sphagnum stems AD 690–704 (25%)

AD 710–735 (43.2%) AD 682–743 (95.4%)


statistical significance with the broken-stick method(MacArthur 1957). The zonation for testate amoebaewas adopted for all proxies as a common zonation, asthis shows the main phases of hydrological change. Allzones are prefixed SL; testate-amoebae zones SL-T, pol-

len zones SL-P, macrofossil zones SL-M and commonzones SL followed by a roman numeral (SL-I to SL-V).

Detrended correspondence analysis (DCA) (terBraak & Prentice 1988) was used to determine themain gradients of change in pollen, macrofossils and

Table 2. Correlation among summarized proxies shown in Fig. 7. Shown is positive or negative correlation (1/�) and R value.

Simple linear regression Regional-pollen DCAaxis 1

Macrofossils DCAaxis 1

Testate amoebae DCAaxis 1

Testate amoebaeDWT(1)


Peat-accumulation rate 1, 0.001 1, 0.369�� , 0.220� �, 0.187 �, 0.200Testate amoebae

DWT(2)�, 0.662�� , 0.708�� , 0.653�� 1, 0.471��


�, 0.161 �, 0.262� 1, 0.018

Testate amoebae DCAaxis 1

1, 0.756�� 1, 0.515��

Macrofossils DCA axis 11, 0.736��

�Statistically significant at P � 0.05; ��Statistically significant at Po0.01.

Sphag

num sec

. Acu

tifolia (%

)

Sphag

num sec

. Cus

pida

ta (%

)

Sphag

num sec

. Sph

agnu

m (%

)

Ericac

eae (%

)

Herbs

- un

iden

tified (%

)

Trich

opho

rum cae

spito

sum -

seed

s

Trich

opho

rum cae

spito

sum -

epider

ms

Calluna

vulga

ris -

seed

s

Calluna

vulga

ris -

flower

s

Calluna

vulga

ris -

stem

s with

leav

es

Andro

med

a po

lifolia

- se

eds

20 40 60 80 20 40 60 20 20 20 20 40

Erioph

orum

vag

inatum

- epide

rms (%

)

20 20 40 20 40 20

Oxy

ccoc

us palus

tris - lea

ves

Analyst: Mariusz Gałka

SL-M1

SL-M2

SL-M3

SL-M4

SL-M5

Sphag

num fu

scum

(%)

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

2000

1900

1800

1700

1600

1500

1400

1300

1200

1100

1000

900

800

Depth (c

m)

SL-I-a

SL-II

SL-III

SL-IV-b

SL-V-b

AD 1830

AD 1100

AD 860

AD 840

SL-I-b

SL-IV-a

SL-V-a

ZONESCommon zones

and time AD

Fig. 4. Plant–macrofossil diagram for peat monolith SL2 of Szowinskie Bzota bog showing local vegetation changes.


0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

Ana

lyst

: Ł.L

amen

tow

icz

& M

. Lam

ento

wic

z

SL-

T5

SL-

T4

SL-

T3

SL-

T2

SL-

T1

SL-I-a

SL-II

SL-III

SL-IV-b

SL-V

-b

AD

183

0

AD

110

0

AD

860

AD

840

SL-I-b

SL-IV-a

SL-V

-a

C

omm

on z

ones

and

time

AD

2000

1900

1800

1700

1600

1500

1400

1300

1200

1100

1000 90

0

800

Fig.5.Testate

amoebaepercentagediagram

forpeatmonolith

SL2ofSzowinskieBzota

bogwithreconstructed

depth

towatertable.


testate amoebae. The first DCA axis is consideredrepresentative.

The diagrams of plant macrofossils, pollen and tes-tate amoebae were constructed with the PC programsTilia and TGView (Grimm 1992) and C2 (Juggins2003). Testate amoebae, pollen and macrofossils areshown as percentages, except countable macrofossils(e.g. seeds), which are presented as counts. The peat-accumulation rate (cm/yr) derives directly from theslope of the age–depth model (Fig. 2).

Tree rings

Thirty-five P. sylvestris trees were sampled in 2006 witha Pressler increment borer, 1.3m above ground level.Widths of annual rings were measured with 0.01mmaccuracy. Two groups of trees were sampled: seriesSLM of 16 trees growing just outside the mire on mi-neral-peaty ground, covering AD 1880–2005 and with amean ring-width of 2.00mm, and series SLK of 19 pinetrees growing on the central dome of the bog close tothe site of peat collection, covering AD 1880–2005 andwith a mean ring-width of 1.05mm. Two main factorsdetermining the ring widths are climate and hydrology.Average SLK/SLM, which eliminates the effect of cli-mate and thus highlights the differences in hydrologybetween the two tree-ring series, is shown in Fig. 6.

Results

The radiocarbon dates are given in Table 1 and thepercentage diagrams of pollen, plant macrofossils andtestate amoebae are presented in Figs 3, 4 and 5. Table2 gives the general features reflected in all proxies. Therecord for Szowinskie Bzota covers the last 1200 years.In addition to a statistically significant zonation of eachindividual proxy shown on the respective diagrams, acommon zonation is provided in a graphic summary(Fig. 7). The common zonation follows the statistically

significant zonation of testate amoebae data, because,by its nature, this proxy has the closest connection withhydrology, which is the main focus of this study, andthis zonation tracks the changes in the other proxiessufficiently closely. In the lower half of the studied sec-tion, the statistically determined zones of the individualproxies are in close agreement with each other, butthere are differences in the upper half.

Spatial scales of change

The main gradients of change summarized in DCA axes1 of regional pollen, plant macrofossils, testate amoe-bae, water-table fluctuations DWT(1) and DWT(2)reconstructed from testate amoebae and peat accumu-lation rates are shown in Fig. 7.

The relationships between proxies are complex notjust because they reflect different spatial scales but alsobecause they respond in different ways to environ-mental change. Nevertheless, there are similaritiesbetween the curves. Due to differences in the biology ofthe groups and in the relevant spatial scales, surfacewetness is recorded without any time-lag in the DWTestimated from testate amoebae (macrofossils have thepotential for some delay), and the connection of DWTwith regional pollen (derived from outside the mire)depends on the hydrological connection of the mirewith its surroundings. Indeed, the high variability ofDWT(1) in subzone SL-IV-a matches a simultaneouschange in macrofossils, whereas regional pollen seemsto respond later. The change in macrofossils representsa shift from S. fuscum to E. vaginatum. Today, thedominance of E. vaginatum plants in the vegetationmostly coincides with fluctuating water levels anda drier bog surface. Testate amoebae support thefluctuating water levels in subzone SL-IV-a, but suggesta later start of drier bog surface. Also, the pollen ofpeatland plants (Fig. 3, right) indicates that the drierbog surface started about 50 years (3 samples) later

0.2

0.4

0.6

0.8

1

1.2

1.4

1875

1880

1885

1890

1895

1900

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

Tre

e rin

g di

ffere

nce

(mm

)

SLK/SLM

Fig. 6. Growth differences of Pinus sylvestris tree rings between trees growing on the central dome of Szowinskie Bzota bog (SLK, 19 trees) andtrees on the margin of the bog (SLM, 16 trees). The average SLK/SLM year-ring ratio is shown. The grey area indicates the period of decreasedwater table caused by drainage.


0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

SL-

P1

SL-

P2

SL-

P3

SL-

P4

SL-

P5

SL-

M1

SL-

M2

SL-

M3SL-

M4

SL-

M5

SL-

T1

SL-

T2

SL-

T3

SL-

T4

SL-

T5

DWT (c

m) - co

mplete training se

t

Macrofoss

ils D

CA sample sc

ores

A

xis 1 - 4

3% of v

ariance

explained

SL-I-a

SL-II

SL-III

SL-IV-b

SL-V

-b

Pollen D

CA sample sc

ores

Axis 1 - 1

8% of v

ariance

explained

Testa

te amoebae DCA sa

mple scores

Axis 1 - 2

2% of varia

nce exp

lained

DWT (c

m) - A

rcella

discoides e

xcluded

AD

183

0

AD

110

0

AD

860

AD

840

AD

790

AD

200

6

Depth (cm)

Peat a

ccumulatio

n rate

(mm/yr

)

SL-I-b

SL-IV-a

SL-V

-a

Inte

rpre

tatio

n

Common zo

nes

and time

AD

2000

1900

1800

1700

1600

1500

1400

1300

1200

1100

1000 90

0

800

Fig.7.

Composite

diagram

presentingDCA

samplescoresobtained

from

macrofossils,pollen

andtestate

amoebaedata

comparedto

testate

amoebaeinferred

depth

tothewater

tablewithan

dwithoutArcelladiscoides

(errorbarsrepresentbootstrap-estim

atedsample-specificerrors)an

dpeataccumulationrates.


(Calluna expansion around AD 1250). The DCA axes 1of pollen, macrofossils and testate amoebae all corre-late strongly with each other, and also with DWT(2).This supports our interpretation that all spatial scalesand all proxies are linked by common driving factors,i.e. climate or human impact.

DWT(1) correlates only weakly with DCA axis 1 ofpollen, macrofossils and testate amoebae, whereasDWT(2) shows strong correlations. The estimate ofDWT(2) differs from that of DWT(1) in exclusion ofA. discoides. This species is thought to respond stronglyto rapid fluctuations of the water table. The correla-tions therefore indicate that these rapid fluctuations onthe local scale are not paralleled with change on aregional scale.

With the exception of peat-accumulation rates, thecurves in Fig. 7 are relatively stable during zones SL-Ito SL-III. Only during zone SL-II (AD 840–860) is awet period reflected in DWT(1) and macrofossils. Ma-jor changes first show up in zones SL-IV and SL-V,dated c. AD 1100–1300, 1830 and 1975, both in themire itself and in regional vegetation. The shape of theregional-pollen DCA axis resembles that of the NAPcurve in Fig. 3, so we can consider both curves as ameasure of landscape openness and forest decline. Thehighest landscape openness is recorded at AD1830–1900, followed by forest regeneration. The re-generation of Sphagnum around c. AD 1830 is con-nected with increased habitat moisture, as shown byDWT(1). This rapid wet shift was large enough to in-itiate the Sphagnum regeneration process resulting inincreased peat-accumulation rate.

Vegetation history and palaeohydrology

During zone SL-I (AD 790–840), Szowinskie Bzota wasa rapidly peat-accumulating Baltic bog with dominantS. fuscum and surrounded by broad-leaved forest.Carpinus and Quercus dominant in subzone SL-I-awere replaced by Fagus in subzone SL-I-b (around AD805). Some agriculture is shown by, for example, Secalepollen, and anthropogenic grasslands by Rumex andPlantago lanceolata pollen. Calluna was present on themire surface. The high peat-accumulation rate (4mm/year) and the presence of wetness indicators amongtestate amoebae (e.g.Archerella flavum,Nebela militarisand Assulina spp.) suggest moderately wet hydrologicalconditions in the mire (average DWT(1)=17.0 cm).The main peat-forming plant of the local vegetationwas S. fuscum. This, together with abundant Callunamacrofossils, suggests a hummock at the exact sam-pling location. The low and even pollen percentages ofPinus and Betula and the absence of tree macrofossilsindicate that no trees were growing on the bog.

Zone SL-II (AD 840–860) shows an abrupt change inmoisture of the mire and a decline in Fagus pollen, but

with unchanged peat-accumulation rate. The wet shift(average DWT(1)=5.1 cm) is shown by A. discoides,Amphitrema stenostoma and Amphitrema wrightianum.An increased water table is also indicated by the tem-porary occurrence of Sphagnum sec. Cuspidata c. AD850 and the decline ofCallunamacrofossils. Outside thebog, pollen indicates a moderate decline of Fagus andsome expansion of Quercus, whereas Carpinus re-mained stable. Human activity is indicated by pollen ofgrassland plants such as Rumex and P. lanceolata, andthe low and discontinuous Secale curve.

Zone SL-III (AD 860–1100) shows constant taxo-nomic composition in each proxy and initially adecreasing water table. In the second part of the zone,peat accumulation declined considerably (0.78mm/year)and the water table again increased, reaching the highestlevel at the end of the zone (average DWT(1)=3.1 cm).This increased wetness, however, appears to be fluctua-tions not favourable for peat accumulation in the toppart of the zone. S. fuscum was the most important com-ponent of the vegetation until AD 1200.Calluna occurredin very small quantities and Alnus became more sig-nificant, probably expanding along the edges of the mirecomplex. This suggests a raised groundwater level alsooutside the mire. Fagus had an initial maximum and thengradually declined in the surrounding forest, whereasCarpinus and Quercus remained stable.

Zone SL-IV (AD 1100–1830) is in the most decom-posed part of the peat. Most testate amoebae shells aredegraded, and the reconstruction indicates a stronglyfluctuating water table. At its beginning, it has the wet-test conditions of the entire sequence, as shown by tes-tate amoebae (average DWT(1)=–2.35 cm). This wassoon followed by a rapid wet–dry shift. After somefluctuations, the driest conditions of the entire sequencewere reached (average DWT=24 cm), suggesting thatthe mire experienced several rapid water-table fluctua-tions during this phase. The testate amoebae in thislayer are much corroded (as the result of peat humifi-cation caused by the drop in the water table) and theestimated DWT has a high uncertainty due to the lowcount. However, other proxies provide additionalindications for wet–dry fluctuations. Dominance ofE. vaginatum and scarcity of Sphagnum (indicated bymacrofossils) suggest hydrological instability.

The vegetation changed around AD 1300 (subzoneSL-IV-b) after the period of unstable hydrology(subzone SL-IV-a). Increased Calluna pollen and, later,Trichophorum macrofossils suggest local desiccation atthe sampling location, and declined Alnus and Betulapollen indicates a drop in the water table at the edge ofthe bog. Increasing agricultural activity is apparentfrom increasing Secale and Humulus and/or Cannabispollen and a continuous curve of Centaurea cyanuspollen, and landscape openness increased (pollen ofArtemisia, Rumex, etc.). Fagus almost disappearedfrom the surroundings mainly due to human impact.


In subzone SL-IV-b starting AD 1500, Quercuspollen has a maximum ending around 1750, Pinusincreases and Fagus, Betula and Populus decline. Twoalternative explanations are (1) drier and warmerclimate, and (2) forest grazing or some form offorest management. Quercus may have been favoureddeliberately as a source of food for swine, in part re-placing Fagus and Carpinus. Around AD 1750, Quer-cus, Corylus, Carpinus and Alnus declined, andP. sylvestris became more abundant as the result offorest management favouring pine as the most produc-tive of the tree species.

Zone SL-V (AD 1830–2006) marks increased wetnessand a gradually increasing peat-accumulation rate. Itstarted with a slow regeneration of the mire after a dryphase, initiated by an abrupt but short-lived water-table rise, as shown by DWT(1). DWT(2) fails to showthis and other short-lived phases of high water tabledue to the exclusion of A. discoides, but instead thegeneral drop in the water table during the intermittentphases of low water table. S. fuscum had no part in thepeat regeneration but was replaced by other Sphagnumtaxa of the sections Acutifolia and Cuspidata. The peat-accumulation rate increased to 1.38mm/year; however,this section consists of non-compacted acrotelm.Around AD 1880 the water table decreased sub-stantially, and simultaneously P. sylvestris started toexpand on the mire surface and the margins of thepeatland. This dry shift resulted from ditches being dugaround the mire edges aimed at afforestation of themire. Around AD 1950 the mire regenerated and thewater table increased, and in AD 1975 it decreasedagain abruptly (average DWT(1) =10.9 cm) due to thedigging of two ditches in the central part of the bog.

Dendroecology

Preliminary results of the relations between the tree-ring chronologies and climate are as follows (Fig. 6).The year ring chronology SLM of 16 P. sylvestris treesgrowing at the edge of the mire shows climate relation-ships typical for northern Poland, i.e. ring widths cor-relating positively with mild winters and early springs,as well as high rainfall in summer. The ring-width re-cord SLK of the 19 P. sylvestris trees growing on thecentral bog dome appears unrelated to weather condi-tions, but responds to changes in the hydrology, which,however, was disturbed by human action.

Discussion

Spatial scales of interpretation

It is important to be aware of differences in spatialscales that the proxies reflect. Testate amoebae reflectstrictly local conditions (moisture, but also nutrients in

the substrate in which they live), and macrofossils inmires are derived from a few metres distance at themost (local scale). For pollen, the scale varies amonggroups, depending on both pollen productivity anddispersal properties and on ecology (e.g. plants growingin bogs, fens or on dry ground). A regional scale up toseveral kilometres applies to pollen of plants that can-not grow on mires, among which are the indicators ofhuman impact on the landscape, whereas a local or ex-tra-local scale up to tens of metres applies to the pollenof mire plants (e.g. Cyperaceae, Calluna). P. sylvestris(pine) is ambiguous, as it can grow on both dry nu-trient-poor soils and raised bogs. Betula and Salix cangrow in relatively dry or relatively nutrient-rich mire,and also on moist soils outside the mires. Alnus gluti-nosa grows on wet, more nutrient-rich, soils, often atmire edges. An increase in Alnus pollen may thereforeindicate paludification of formerly drier soils. Spatialscales are often linked by environmental forcing fac-tors. For example, a change in precipitation and/or inlandscape hydrology may affect the landscape and thebog centre alike. Interpretations in terms of environ-mental change are therefore mostly valid for a largerspatial scale than the individual proxies would suggest.

Disentangling climatic and anthropogenic signals

We are not alone in experiencing difficulties in disen-tangling climate from human disturbance during thelate Holocene; see, for example, Chambers (1993) andvan der Linden & van Geel (2006). Direct disturbanceof mires, such as peat exploitation or mire drainage, isrelatively simple to detect, but indirect disturbance isnot. For example, human activity outside the mire(deforestation, agriculture) may alter the hydrologysubstantially and thus impact on the mire vegetation.We can see this in Jelenia Wyspa mire (Tucholapinewoods, N Poland) (Lamentowicz et al. 2007b). AsSzowinskie Bzota has certainly been influenced by hu-man activity during the past 1200 years, we studiedseveral different proxies reflecting different aspects ofthe effects of human impact. In the end result, bothhuman and climate factors appear to have steered therecorded change during the entire period, with a shiftfrom climate to humans as the dominant factor.Szowinskie Bzota appears to be sensitive to environ-mental change in the catchment, especially hydrology.The hydrological link between bog and catchment ap-pears to be stronger than we had expected, which in-dicates a certain permeability of the till below the peatdeposits. In addition, external change in hydrology hasthe greatest impact on the peat surface when the totalpeat deposit is still relatively thin (2m in SzowinskieBzota), such that the capacity of ‘Mooratmung’ (up anddown movement of the peat surface depending onmoisture availability) is not yet well developed. For


example, Joosten (1995) and van Walsum & Joosten(1994) described and modelled a case study in TheNetherlands (Groote Peel mire area) where anthro-pogenic hydrologic disturbance at a distance of manykilometres impacted on mire hydrology and peat for-mation.

Compared to Staz’ki mire (Lamentowicz et al. 2008a),another Baltic bog located in the same region (Pomer-ania) 100 km to the east, peat-growth conditions inSzowinskie Bzota have been relatively dry during the past800 years. Szowinskie Bzota situated on flat groundmoraine (30m a.s.l.) has a surprisingly shallow peat de-posit of c. 2m, compared to 7m of Staz’ki mire locatedon end moraine (212m a.s.l.). The shallow total peatthickness of Szowinskie Bzota and the relatively late startof peat formation (after the shift to a wetter climate 4000BC; Herbichowa (1998)) suggest that the bog is locatedin less favourable conditions than Staz’ki mire (start ofpeat formation c. 5350 BC on top of lake sediments;Miotk–Szpiganowicz, unpublished data). SzowinskieBzota is therefore more sensitive to external impacts, andthus may respond differently to the same impact. Anexample is the Little Ice Age, during which Staz’ki mireexperienced the expected wet shift, whereas SzowinskieBzota became drier.

The Holocene history of Szowinskie Bzota haspreviously been investigated by Herbichowa (1998)with the use of plant macrofossils, and the generalstratigraphic pattern of change found in both our studyand hers is similar. The main difference is that herhighest peat-accumulation rate is 1mm/year and in ourstudy this is 5mm/year (during AD 840–860), whichcan in part be explained by the higher temporal resolu-tion in our study. This exceptional but short period ofincreased moisture and rapid peat growth is reflected inall proxies. It is interesting that Fagus pollen had aminimum in the same period, for which human impactis an unlikely explanation as there are no extra humanpollen indicators. Possible explanations in terms ofclimate are that increased precipitation and/or lowertemperatures suppressed pollen production. Fagusreached a maximum after this event c. AD 870(69–67 cm) and then gradually declined, where-as A. glutinosa increased, replacing Fagus. The lattersuggests paludification of soils in the surroundings.

S. fuscum (adapted to a relatively low but stablewater table) is reported as the most important peat-forming species in Polish raised bogs during the pastc. 2500 years (Jasnowski 1962, 1972; Pacowski 1967). Itstarted to decline in Szowinskie Bzota around AD 1100and about AD 1350 finally disappeared from the coringlocation in connection with a substantial decrease in thewater table at the beginning of the Little Ice Agefollowed by intensified land-use. The large water-table fluctuations probably exceeded the tolerance ofS. fuscum and enabled E. vaginatum to become domi-nant. The disappearance of S. fuscum parallels that of

S. austinii from Atlantic peatlands of Western Europe(Mauquoy & Barber 1999b; Hughes et al. 2007). Thedifferences in timing are probably the result of differ-ences in human impact and the impact of climaticchange in different parts of Europe and of the differentresilience of mires to hydrological stress.

An unambiguous climatic signal is difficult to extractfrom the analysed proxies. We have to be cautiousbecause land-use change might have had a substantialimpact on peatland hydrology, especially when the peatis shallow (2–3m thickness in Szowinskie Bzota) and thepeatland thus very sensitive to hydrological changes inthe surroundings. The pollen data show transforma-tions of regional vegetation, especially forest clearingfor agriculture. The pollen curves of the cereal Secalecereale and its accompanying weed C. cyanus becamestable c. AD 1200–1300. Intensified cultivation in NEPoland during early Medieval times (period defined forPoland by Kaczanowski & Kozzowski 1998) has beenrecorded previously at various sites (Latazowa 1982,1989; Latazowa & Tobolski 1987, 1989; Ralska-Jasie-wiczowa & Latazowa 1996).

The hydrological relationships of bogs close to theBaltic Sea, with their catchment, are still poorly under-stood. In Szowinskie Bzota, the existence of a hydro-logical link of mire vegetation with nutrient-richgroundwater is shown by the type of peat (miner-otrophic) from the time of peatland initiation shortlyafter 6000 cal. yr BP up to c. 1800 cal. yr BP, when thedome started to form (Herbichowa 1998). The dome isnot high (1.2m) and the total peat deposit is not thick(average maximum 2.8m), so it is possible that thegroundwater pressure from the catchment influencedvertical water movements in the peat at all times. It isalso possible that more strongly elevated raised bogslocated in oceanic western Europe (UK or Ireland)were less susceptible to groundwater fluctuationscaused by increased landscape openness, even thoughlarge-scale deforestation started earlier in oceanic wes-tern Europe than in Pomerania (Odgaard 1994;O’Brien et al. 1995).

Around AD 1100, increasing human impact andclimate change together may have caused the endof relatively stable peat-accumulating vegetation inSzowinskie Bzota. S. fuscum declined, after whichE. vaginatum appeared and became dominant, and thewater table decreased markedly and started to fluctuatestrongly. This caused peat compaction, including theunderlying peat (down to peat dated around AD 900),and even erosion. The pollen signal does not indicate ashift in human impact around AD 1100, so climatechange is the more likely explanation. For example, aone-year case study in the region indicated that a de-crease in precipitation can lead to high hydrologicalstress during summer (Hazas et al. 2008). Allowing asmall dating uncertainty, it may be related to theMedieval Warm Period AD 800–1300, reflected in


other peat archives (Barber 1981) and proxies (Jedryseket al. 2003; Guiot et al. 2005; Goosse et al. 2006). InStaz’ki mire, 100 km to the east (Lamentowicz et al.2008a), a decreased peat-accumulation rate betweenAD 1100 and 1500 also indicates a dry phase, whichsuggests a common cause.

Special attention should be given to the ecology ofA. discoides. This species is generally abundant intestate-amoebae training sets and responds to a differ-ent aspect of hydrology than, for example, A. wright-ianum or A. flavum, which are associated with relativelystable wet conditions. A. discoides is considered to befavoured by hydrological disturbance in the form ofabrupt water-table fluctuations. This is confirmed inour data in which its peaks alternate with dryness in-dicators (e.g. N. militaris). Another example comesfrom Tore Hill (Scotland), where in zone THM-c thisspecies occurs together with the apparent drynessindicator Hyalosphenia subflava (Blundell & Barber2005) and is located in the dry part of the gradient in theDCA scatterplot. We found a similar situation inStaz’ki mire 100 km to the east of Szowinskie Bzota(Lamentowicz et al. 2008a), where A. discoides is abun-dant in well-decomposed E. vaginatum peat, suggestinghydrological disturbance. The Polish testate-amoebaetraining set shows that A. discoides occurs in laggs andditches of Sphagnum mires in Pomerania where thewater table is unstable (Lamentowicz & Mitchell2005b). Samples with a high abundance of both A. dis-coides and taxa indicative of dry conditions shouldtherefore be interpreted with caution as they indicatehighly variable water tables, which may have causedtemporary flooding of dry peat surfaces resulting in themixing of testate amoebae with contrasting wetness re-quirements. The weak correlation of DCA axis 1 oftestate-amoebae with DWT(1) (with A. discoides) andthe strong correlation with DWT(2) (without A. dis-coides) indicated in Table 2 show that A. discoides picksup extreme hydrological events that are not recorded inthe general trend of the remaining testate amoebae.

We noted a strong correlation between the maintrends of the three proxies from peat (DCA axes 1 inTable 2). Independently of the dominant driving factor(climate change or human impact), it is likely that thehydrology of the area links together the spatial scales:changes in groundwater level have their impact fromthe strictly local to the regional scale.

The dry conditions reconstructed for SzowinskieBzota continuing after AD 1500 contrast with thecooler and wetter climate c. AD 1560–1850 generallyinferred for the Little Ice Age. Either the climate con-tinued to be relatively dry in this part of Europe, or themire continued to be dry for non-climatic reasons. Vander Linden & van Geel (2006) report a similar case froma high-resolution study in southern Sweden. The con-flicting interpretations based on testate amoebae, pol-len and the general conditions of the Little Ice Age rule

out climate as the main driving factor. Human impactremains as the more likely, preventing the bog to in-crease in wetness and causing the forest to shift thebalance between the tree taxa.

Macrofossils indicate that at the end of Little Ice Age(AD 1850) Szowinskie Bzota started a succession ofSphagnum sec. Acutifolia and sec. Cuspidata, which aremore tolerant to water-table fluctuations than the pre-viously abundant S. fuscum. The water table graduallydropped. Pinus further increased as the result of in-troduction in forest as well as on the mire. The tree-ringanalysis shows that P. sylvestris established c. AD 1850on the edges of the bog. We cannot expect any climaticsignal in the local hydrology after 1850 because of thisand because of the intensive land-use, the first artificialdrainage around the bog edges around 1880 and cen-trally in the bog in the early 1960s, renewal of marginaldrainage in the 1980s and the peat digging (Herbichowa1998). At 17.5 cm, possibly related to the First WorldWar, some shift in agricultural practice is indicated bydeclining grassland taxa (Rumex div. types, P. lanceo-lata) and Pteridium. The most recent decrease inmoisture, marked by increasing Sphagnum sec. Acuti-folia and Calluna vulgaris macrofossils, is the result ofditches dug c. 1979. From c. 1990 the mire started torecover, as marked by the testate amoebaHyalospheniaelegans. The most recent regeneration of the mire afterAD 2002, resulting from the overgrowing of ditchesand restoration measurements, is not yet recorded bytestate amoebae.

Research on Szowinskie Bzota bog alone is of courseinsufficient to answer the key question posed by the EUproject Millennium, i.e. whether or not the recent cli-matic change exceeds natural climate variability duringthe last millennium, but it is a modest contribution tothe research on a part of Europe where high-resolutionmultiproxy studies are still in short supply. Togetherwith other high-resolution records from Europe (vander Linden & van Geel 2006; Charman & Blundell2007; Lamentowicz et al. 2007a; Sillasoo et al. 2007;Valiranta et al. 2007; Yeloff et al. 2007a, b) and fromother continents (Booth et al. 2004, 2005, 2006;Mitchell et al. 2008) the Szowinskie Bzota and Staz’kimires lying 100 km more east (Lamentowicz et al.2008a) contribute to the regional palaeoclimatic pic-ture. The results highlight two major strengths of re-search on peatlands: one providing a signal of climatechange in the low-frequency domain that is mostlyweak or absent in the widely appreciated archive oftree rings; the other, a multi-proxy approach, enablingclimate and human effects to be separated effectively.Various ongoing scientific projects (e.g. Millennium,ACCROTELM) focus on synthesis of differentarchives, including peatlands, to obtain a wide viewof past global changes. Charman (2007) has arguedthat the natural variability of the climate system and itsimpact on peatland biota may be properly described


only on the basis of multiple peatland sites and replicatecoring per site; and modern palaeoecologicaltools should be used to give the necessary backgroundfor the reconstructions (Belyea 2007). For example,Blundell et al. (2008) search for a regional water-tablecurve valid for NW Europe. Many more peatlandsites will be needed if the regional palaeohydrologicalcharacteristics for (northern) Poland are to be defined.Szowinskie Bzota is only the second peatland innorthern Poland that has been subjected to multiproxyhigh-resolution palaeoenvironmental study. It showsthe difficulties with palaeoclimatic reconstructioncaused by an increasing human impact on landscape,although some interesting results have been achieved.Our future research aims at extending the spatial andtemporal scale of palaeoenvironmental peatlandinvestigations.

Szowinskie Bzota bog is presently managed for natureconservation, and we consider the formerly present,peat-forming, S. fuscum as a good target speciesbecause it was typical of Baltic bogs in the past.

Conclusions

The record of Szowinskie Bzota starts in AD 790.Undisturbed broad-leaved forest surrounded the bog upto AD 860. After AD 860 the peat-accumulation rate de-creased from 4mm/year (the maximum in this bog) to0.78mm, and after AD 1100 to 0.35mm. S. fuscum, themost important peat-forming species, declined after AD1100, disappearing in AD 1350. Other Sphagnum taxaand E. vaginatum took over and, together with a testate-amoebae-based water-table reconstruction that includesA. discoides, this indicates strongly and rapidly fluctuatingwater tables resulting in a drier bog surface. This changewas most likely driven by a warmer/drier climate (Medie-val Warm Period), as pollen does not suggest humanimpact. A reconstructed shift to lower water levels anddrier forest types after AD 1500 leads to interpretationsbeing not in agreement with the supposed cooler/moisterclimate of the Little Ice Age, which indicates that humanaction is the forcing agent. A different peat-formingvegetation type developed after AD 1830.

This study shows once more that the multiproxyapproach is the key to understanding the complexrelationships of the various mechanisms forcingecosystem change. The use of multiple proxies is helpfulbecause the forcing factors (climate, human impact) canshift in relative importance and act on a range of spatialscales (local–regional) in a complex landscape (bog, for-est, cultural areas), and the proxies differ in their sensi-tivity to forcing factors, in the relevant spatial scale andin the parts of the landscape that they reflect. The mainstrength of testate amoebae was found to lie in local hy-drology, of macrofossils in local human impact and dis-turbance including nutrients, of tree rings in either mire

hydrology (for trees growing in the bog centre) or tem-perature (for trees growing at the bog edges), and of pol-len in land-use of the surrounding landscape. Thecombination of proxies enabled climate and human im-pact to be separated in the hydrological signal.

Acknowledgements. – We are indebted to Mazgorzata Suchorska andMilena Obremska for assistance in sample preparation. This work ispart of a research project funded by the Polish Ministry of ScientificResearch and Information Technology: Climatic Changes in Pomer-ania (N Poland) in the Last Millennium Based on the MultiproxyHigh-Resolution Studies (no. 2P04G03228) (Principal In-vestigator–M. Lamentowicz). We thank Hans Joosten for useful sug-gestions concerning the hydrology of mires. The work was carried outin collaboration with the EU-funded Project Millennium (EuropeanClimate 685 of the last millennium; EU 6FP Integrated Project No.017008-GOCE). We also thank the two anonymous referees and JanA. Piotrowski for their comments leading to improvement of themanuscript.

References

Akima, H. 1970: A new method of interpolation and smooth curvefitting based on local procedures. Journal of the Association forComputing Machinery 17, 589–602.

Barber, K. E. 1981: Peat Stratigraphy and Climatic Change.A Palaeoecological Test of the Theory of Cyclic Bog Regeneration.219 pp. A. A. Balkema, Rotterdam.

Belyea, L. R. 2007: Revealing the Emperor’s new clothes: Niche-based palaeoenvironmental reconstruction in the light of recentecological theory. The Holocene 17, 683–688.

Bennett, K. D. 1996: Determination of the number of zones in abiostratigraphical sequence. New Phytologist 132, 155–170.

Berglund, B. E. & Ralska-Jasiewiczowa, M. 1986: Pollen analysis andpollen diagrams. In Berglund, B. E. (ed.): Handbook of HolocenePaleoecology and Paleohydrology, 455–484. Wiley, Chichester andToronto.

Birks, H. J. B. 1986: Numerical zonation, comparison and correlationof Quaternary pollen-stratigraphical data. In Berglund, B. E. B.(ed.): Handbook of Holocene Palaeoecology and Palaeohydrology,743–774. Wiley, Chichester.

Birks, H. J. B. & Birks, H. H. 1980: Quaternary Palaeoecology. 289pp. Arnold, London.

Birks, H. J. B. & Gordon, A. D. 1985: Numerical Methods in Qua-ternary Pollen Analysis. 317 pp. Academic Press, London.

Blundell, A. & Barber, K. 2005: A 2800-year palaeoclimatic recordfrom Tore Hill Moss, Strathspey, Scotland: The need for a multi-proxy approach to peat-based climate reconstructions. QuaternaryScience Reviews 24, 1261–1277.

Blundell, A., Charman, D. J. & Barber, K. 2008: Multiproxy lateHolocene peat records from Ireland: Towards a regional palaeocli-mate curve. Journal of Quaternary Science 23, 59–71.

Booth, R. K., Jackson, S. T., Forman, S. L., Kutzbach, J. E., Bettis,E. A., Kreig, J. & Wright, D. K. 2005: A severe centennial-scaledrought in midcontinental North America 4200 years ago and ap-parent global linkages. The Holocene 15, 321–328.

Booth, R. K., Jackson, S. T. & Gray, C. E. D. 2004: Paleoecology andhigh-resolution paleohydrology of a kettle peatland in upper Mi-chigan. Quaternary Research 61, 1–13.

Booth, R. K., Notaro, M., Jackson, S. T. & Kutzbach, J. E. 2006:Widespread drought episodes in the western Great Lakes regionduring the past 2000 years: Geographic extent and potential me-chanisms. Earth and Planetary Science Letters 242, 415–427.

ter Braak, C. J. F. & Prentice, I. C. 1988: A theory of gradientanalysis. Advances in Ecological Research 18, 235–282.

Bronk Ramsey, C. 2001: Development of the radiocarbon programOxCal. Radiocarbon 43, 355–363.

Bronk Ramsey, C. 2008: Deposition models for chronological re-cords. Quaternary Science Reviews 17, 2115–2237.


Buttler, A., Warner, B. G., Grosvernier, P. & Yvan, M. 1996: Verticalpatterns of testate amoebae (Protozoa: Rhizopoda) andpeat-forming vegetation on cutover bogs in the Jura, Switzerland.New Phytologist 134, 371–382.

Chambers, F. M. 1993: Climate change and human impact on thelandscape: Studies in palaeoecology and environmental archaeology.328 pp. Chapman & Hall, London.

Chambers, F. M. & Charman, D. J. 2004: Holocene environmentalchange: Contributions from the peatland archive.TheHolocene 14, 1–6.

Charman, D. J. 2007: Summer water deficit variability controls onpeatland water-table changes: implications for Holocene palaeocli-mate reconstructions. The Holocene 17, 217–227.

Charman, D. & Blundell, A. 2007: A new European testate amoebaetransfer function for palaeohydrological reconstruction on ombro-trophic peatlands. Journal of Quaternary Science 22, 209–221.

Charman, D. J., Blundell, A., Chiverrell, R. C., Hendon, D. & Lang-don, P. G. 2006: Compilation of non-annually resolved Holoceneproxy climate records: Stacked Holocene peatland palaeo-water tablereconstructions from northern Britain. Quaternary Science Reviews25, 336–350.

Charman, D. J., Hendon, D. & Woodland, W. A. 2000: The Identifi-cation of Testate Amoebae (Protozoa: Rhizopoda) in Peats. Tech-nical Guide, 9. 147 pp. Quaternary Research Association, London.

Clarke, K. J. 2003: Guide to Identification of Soil Protozoa – TestateAmoebae. 40 pp. Freshwater Biological Association, Ambleside.

Crowley, P. H. 1992: Resampling methods for data analysis in com-putation-intensive ecology and evolution. Annual Review ofEcology and Systematics 23, 405–447.

De Vleeschouwer, F. 2007: Heavy Metal Records in Peat. Archives ofNatural vs. Anthropogenic Influences. Ph.D. dissertation, Universityof Liege, 160 pp.

Fægri, K. & Iversen, J. 1989: Texbook of Pollen Analysis. 382pp. John Wiley, Chichester and Toronto.

Frelechoux, F., Buttler, A., Gillet, F., Gobat, J. M. & Schweingruber,F. H. 2003: Succession from bog pine (Pinus uncinata var. rotunda-ta) to Norway spruce (Picea abies) stands in relation to anthropicfactors in Les Saignolis bog, Jura Mountains, Switzerland. Annalsof Forest Science 60, 347–356.

Goosse, H., Arzel, O., Luterbacher, J., Mann, M. E., Renssen, H.,Riedwyl, N., Timmermann, A., Xoplaki, E. & Wanner, H. 2006:The origin of the European ‘MedievalWarm Period’.Climate of thePast Discussions 2, 285–314.

Grimm, E. C. 1992: TILIA/TILIA graph. Version 1.2. Illinois StateMuseum, Springfield, Illinois.

Grospietsch, T. 1958:Wechseltierchen (Rhizopoden), 86 pp. Kosmos,Stuttgart.

Grosse-Brauckmann, G. 1972: Uber pflanzliche Makrofossilien mit-teleuropaischer Torfe. I. Gewebereste krautiger Pflanzen und ihreMerkmale. Telma 2, 19–55.

Grosse-Brauckmann, G. 1974: Uber pflanzliche Makrofossilien mitte-leuropaischer Torfe. II. Weitere Reste (Fruchte und Samen, Moose u.a.) und ihre Bestimmungsmoglichkeiten. Telma 4, 51–117.

Guiot, J., Nicault, A., Rathgeber, C., Edouard, J. L., Guibal, F.,Pichard, G. & Till, C. 2005: Last-millennium summer-temperaturevariations in western Europe based on proxy data. The Holocene15, 489–500.

Hazas, S., Szowinski, M. & Lamentowicz, M. 2008: Relationships be-tween meteorological factors and hydrology of a kettle-hole mire innorthern Poland. Studia Limnologica et Telmatologica 2, 15–26.

Hendon, D. & Charman, D. J. 1997: The preparation of testateamoebae (Protozoa: Rhizopoda) samples from peat. The Holocene7, 199–205.

Herbichowa, M. 1998: Ekologiczne studium rozwoju torfowisk wyso-kich wzasciwych na przykzadzie wybranych obiektow z srodkowejczesci Pobrzez’a Baztyckiego. 119 pp. Wydawnictwo UniwersytetuGdanskiego, Gdansk.

Herbichowa, M., Pawlaczyk, P. & Stanko, R. 2007: Conservation ofBaltic Raised Bogs in Pomerania, Poland. 147 pp. WydawnictwoKlubu Przyrodnikow, �S wiebodzin.

Hughes, P. D. M., Lomas-Clarke, S. H., Schulz, J. & Jones, P. 2007:The declining quality of late-Holocene ombrotrophic communitiesand the loss of Sphagnum austinii (Sull. ex Aust.) on raised bogs inWales. The Holocene 17, 613–625.

Jasnowski, M. 1962: Budowa i roslinnosc torfowisk Pomorza Szcze-cinskiego. Szczecinskie Towarzystwo Naukowe, Wydziaz NaukPrzyrodniczo – Rolniczych 10, 1–339 (in Polish).

Jasnowski, M. 1972: Rozmiary i kierunki przeksztazcen szaty roslin-nej torfowisk. Phytocenosis 1, 193–209 (in Polish).

Jedrysek, M. O., Krapiec, M., Skrzypek, G. & Kazuz’ny, A. 2003: Air-pollution effect and palaeotemperature scale versus d13C records intree rings and in peat core (Southern Poland). Water, Air, and SoilPollution 145, 359–375.

Joosten, J. H. J. 1995: Time to regenerate: Long-term perspectives ofraised bog regeneration with special emphasis on palaeoecologicalstudies. In Wheeler, B. D., Shaw, S. C., Fojt, W. J. & Robertson,R. A. (eds.): Restoration of Temperate Wetlands, 379–404. Wiley,Chichester.

Juggins, S. 2003: C2 User Guide. Software for Ecological and Pa-laeoecological Data Analysis and Visualisation. 69 pp. University ofNewcastle, Newcastle upon Tyne.

Kaczanowski, P. & Kozzowski, J. K. 1998: Wielka historia Polski.Tom 1. Najdawniejsze dzieje ziem polskich (do VII w). 382 pp. Fo-gra, Krakow (in Polish).

Katz, N. J., Katz, S. V. & Kipiani, M. G. 1965: Atlas opredelitel’ plo-dov i semyan vstretchayushchikhsya v chetvertinnykh otuocheniyakhSSSR. 378 pp. Nauka, Moskva (in Russian).

Katz, N. J., Katz, S. V. & Skobiejeva, E. I. 1977: Atlas rastitielnychostatkov v torfach. 370 pp. Nedra, Moskva (in Russian).

Koff, T., Punning, J.-M. & Yli-Hallab, M. 1998: Human impact on apaludified landscape in northern Estonia. Landscape and UrbanPlanning 41, 263–272.

Kueppers, L., Southon, J., Baer, P. & Harte, J. 2004: Dead woodbiomass and turnover time, measured by radiocarbon, along asubalpine elevation gradient. Oecologia 141, 641–651.

Kulczynski, S. 1939: Torfowiska Polesia t. I, 1–394. Uniwersytet Ja-giellonski, Krakow (in Polish).

Kulczynski, S. 1940: Torfowiska Polesia t. II, 395–777. UniwersytetJagiellonski, Krakow (in Polish).

Kulczynski, S. 1949: Peatbogs of Polesie. Memoires de l’AcademiePolonaise des Sciences et des Lettres. 15, 1–356 (in Polish).

Lamentowicz, M. & Mitchell, E. A. D. 2005a: The ecology of testateamoebae (Protists) in Sphagnum in north-western Poland in rela-tion to peatland ecology.Microbial Ecology 50, 48–63.

Lamentowicz, M. & Mitchell, E. A. D. 2005b: Testate amoebae(Protists) as palaeoenvironmental indicators in peatlands. PolishGeological Institute Special Papers 16, 58–64.

Lamentowicz, M., Cedro, A., Miotk-Szpiganowicz, G., Mitchell, E.A. D., Pawlyta, J. & Goslar, T. 2008a: Last millennium pa-laeoenvironmental changes from a Baltic bog (Poland) inferredfrom stable isotopes, pollen, plant macrofossils and testate amoe-bae. Palaeogeography, Palaeoclimatology, Palaeoecology 265,93–106.

Lamentowicz, M., Mitchell, E. A. D. & Knaap, W. O. v. d.2007a: Reconstruction of climate during the last 1000 yearsfrom a high-resolution testate amoebae sequence in Switzerland. InCatto, N. R. (ed.): XVII INQUA Congress. The Tropics: HeatEngine of the Quaternary. Quaternary International, p. 429.

Lamentowicz, M., Obremska, M. & Mitchell, E. A. D. 2008b: Auto-genic succession, land-use change, and climatic influences on theHolocene development of a kettle hole mire in Northern Poland(Northern Poland). Review of Palaeobotany and Palynology 151,21–40.

Lamentowicz, M., Tobolski, K. & Mitchell, E. A. D. 2007b: Pa-laeoecological evidence for anthropogenic acidification of a kettle-hole peatland in northern Poland. The Holocene 17, 1185–1196.

Latazowa, M. 1982: Major aspects of the vegetational history in theeastern Baltic coastal zone of Poland. Acta Palaeobotanica 22,47–63.

Latazowa, M. 1989: Type region P-t: Baltic Coastal Zone. Acta Pa-laeobotanica 29, 103–108.

Latazowa,M. & Pedziszewska, A. 2003: Zbiorowiska lesne z udziazemgrabu (Carpinus betulus) i buka (Fagus sylvatica) na Wysoczy$nieGdanskiej w po$nym Holocenie. Wstepne wyniki badan. In Goze-biewski, R. (ed.): Ewolucja pojezierzy i pobrzez’y pozudniowo-baztyckich, 95–100. Fundacja Rozwoju Uniwesytetu Gdanskiego,Gdansk (in Polish).


Latazowa, M. & Tobolski, K. 1987: A first attempt at a regionalsynthesis for the Baltic Coast and Baltic Coastal Zone of Poland.Lundqua Report 26, 101–104.

Latazowa, M. & Tobolski, K. 1989: Type region P-u: Baltic shore.Acta Paleobotanica 29, 109–114.

MacArthur, R. H. 1957: On the relative abundance of bird species.Proceedings of the National Academy of Sciences USA 43, 293–295.

Mauquoy, D. & Barber, K. E. 1999a: Evidence for climatic dete-riorations associated with the decline of Sphagnum imbricatumHornsch. ex Russ. in six ombrotrophic mires from northern Eng-land and the Scottish Borders. The Holocene 9, 423–437.

Mauquoy, D. & Barber, K. E. 1999b: A replicated 3000 yr proxy-climaterecord from Coom Rigg Moss and Felecia Moss, the Border Mires,northern England. Journal of Quaternary Science 14, 263–275.

Mauquoy, D., van Geel, B., Blaauw, M., Speranza, A. & Plicht, J. v.d. 2004b: Changes in solar activity and Holocene climatic shiftsderived from 14C wiggle-match dated peat deposits. The Holocene14, 45–52.

Mitchell, E., Charman, D. & Warner, B. In press: Testate amoebaeanalysis in ecological and paleoecological studies of wetlands: Past,present and future. Biodiversity and Conservation, doi: 10.1007/s10531-007-9221-3.

Mitchell, E. A. D. 2003: Identification keys for testate amoebae.Available at: http://wslar.epfl.ch/mitchell/edward/Identification_keys/Keys.htm.

O’Brien, S. R., Mayewski, P. A., Meeker, L. D., Meese, D. A., Twick-ler, M. S. & Whitlow, S. I. 1995: Complexity of Holocene climate asreconstructed from a Greenland ice core. Science 270, 1962–1964.

Odgaard, B. V. 1994: The Holocene vegetation history of northernWest Jutland, Denmark. Opera Botanica 123, 1–171.

Oztuszewski, W. 1948: Badania pyzkowe nad torfowiskami dolnejyeby. Badania Fizjograficzne nad Polska Zachodnia 1, 97–128(in Polish).

Oztuszewski, W. & Borowko, Z. 1954: Analiza pyzkowa torfowiska ‘Bie-lawskie Bzoto’. In Czubinski, Z. (ed.): Bielawskie Bzoto – ginace torfo-wisko atlantyckie Pomorza. Ochrona Przyrody, 140–152 (in Polish).

Osvald, H. 1925: Die Hochmore Europas. Veroffentlichungen desGeobotanischen Instituts der Technischen Hochschule Zurich 3,707–723.

Pacowski, R. 1967: Biologia i stratygrafia torfowiska wysokiegoWieliszewo na Pomorzu Zachodnim. Zeszyty ProblemowePostepow Nauk Rolniczych 76, 174–190 (in Polish).

Ralska-Jasiewiczowa, M. & Latazowa, M. 1996: Poland. In Berglund,B. E., Birks, H. J. B., Ralska-Jasiewiczowa, M. & Wright, H. E. Jr.(eds.): Palaeoecological Events During the Last 15000 Years:Regional Syntheses of Palaeoecological Studies of Lakes and Miresin Europe, 403–472. Wiley, Chichester.

Rasanen, S., Froyd, C. & Goslar, T. 2007: The impact of tourism andreindeer herding on forest vegetation at Saariselka, Finnish Lap-land: A pollen analytical study of a high resolution peat profile. TheHolocene 17, 447–456.

Reimer, P. J., Baillie, M. G. L., Bard, E., Bayliss, A., Beck, J. W.,Bertrand, C. J. H., Blackwell, P. G., Buck, C. E., Burr, G. S.,

Cutler, K. B., Damon, P. E., Edwards, R. L., Fairbanks, R. G.,Friedrich, M., Guilderson, T. P., Hogg, A. G., Hughen, K. A.,Kromer, B., McCormac, G., Manning, S., Ramsey, C. B., Reimer,R. W., Remmele, S., Southon, J. R., Stuiver, M., Talamo, S., Tay-lor, F., van der Plicht, J. & Weyhenmeyer, C. 2004: IntCal04terrestrial radiocarbon age calibration, 0–26 cal kyr BP. Radio-carbon 46, 1029–1058.

Schoning, K., Charman, D. J. & Wastegard, S. 2005: Reconstructedwater tables from two ombrotrophic mires in eastern centralSweden compared with instrumental meteorological data. TheHolocene 15, 111–118.

Sillasoo, U., Mauquoy, D., Blundell, A., Charman, D., Blaauw, M.,Daniell, J. R. G., Toms, P., Newberry, J., Chambers, F. M. &Karofeld, E. 2007: Peat multi-proxy data fromMannikjarve bog asindicators of late Holocene climate changes in Estonia. Boreas 36,20–37.

Succow, M. & Joosten, H. 2001: Landschaftsokologische Moorkunde.622 pp. Sweizerbart, Stuttgart.

Szafranski, F. 1961: Polodowcowa historia lasow obszaru na poznocod Wysoczyzny Staniszewskiej. Badania Fizjograficzne nad PolskaZachodnia 8, 91–131 (in Polish).

Tarnocai, C. 2006: The effect of climate change on carbon inCanadian peatlands. Global and Planetary Change 53, 222–232.

Tobolski, K. 2000: Przewodnik do oznaczania torfow i osadowjeziornych. 508 pp. PWN, Warszawa.

Valiranta,M., Korhola, A., Seppa, H., Tuittila, E.-S., Sarmaja-Korjonen,K., Laine, J. & Alm, J. 2007: High resolution reconstruction of wetnessdynamics in a southern boreal raised bog, Finland, during the late Ho-locene: A quantitative approach. The Holocene 17, 1093–1107.

van der Linden, M. & van Geel, B. 2006: Late Holocene climatechange and human impact recorded in a south Swedish ombro-trophic peat bog. Palaeogeography, Palaeoclimatology, Palaeoecol-ogy 240, 649–667.

van Walsum, P. E. V. & Joosten, J. H. J. 1994: Quantification of localecological effects in regional hydrologic modelling of bog reservesand surrounding agricultural lands. Agricultural Water Manage-ment 25, 45–55.

Wardenaar, E. C. P. 1987: A new hand tool for cutting peat profiles.Canadian Journal of Botany 65, 1772–1773.

Wodziczko, A. & Thomaschewski, M. 1932: Staniszewskie Bzoto naKaszubczy$nie. Acta Societatis Botanicorum Poloniae 9, 1–10.

Wos, A. 1999: Climate of Poland. 302 pp. PWN,Warszawa (in Polish).Yeloff, D., Broekens, P., Innes, J. & van Geel, B. 2007a: Late Holo-cene vegetation and land-use history in Denmark: A multi-decadally resolved record from Lille Vildmose, northeast Jutland.Review of Palaeobotany and Palynology 146, 182–192.

Yeloff, D., Charman, D., Geel, B. v. & Mauquoy, D. 2007b: Re-construction of hydrology, vegetation and past climate change bogsusing fungal microfossils. Review of Palaeobotany and Palynology146, 102–145.

Yeloff, D. & Mauquoy, D. 2006: The influence of vegetationcomposition on peat humification: Implications for palaeoclimaticstudies. Boreas 35, 662–673.


Climate and human induced hydrological change since AD 800 in an ombrotrophic mire in Pomerania (N Poland) tracked by testate amoebae, macro-fossils, pollen and tree rings of pine

Documents